Tectonic Evolution

of Southeast Asia
Geological Society Special Publications
Series Editor A. J. FLEET
GEOLOGICAL SOCIETY SPECIAL PUBLICATION NO. 106

Tectonic Evolution
of Southeast Asia

EDITED BY

ROBERT HALL
SE Asia Research Group
Department of Geology
Royal Holloway London University
Surrey, UK

AND

DEREK BLUNDELL
SE Asia Research Group
Department of Geology
Royal Holloway London University
Surrey, UK

1996

Published by
The Geological Society
London
 THE GEOLOGICAL SOCIETY

The Society was founded in 1807 as the Geological Society of London and is the oldest geological society in the world.
It received its Royal Charter in 1825 for the purpose of 'investigating the mineral structure of the Earth'. The Society
is Britain's national society for geology with a membership of 8000. It has countrywide coverage and approximately
1000 members reside overseas. The Society is responsible for all aspects of the geological sciences including
professional matters. The Society has its own publishing house, which produces the Society's international journals,
books and maps, and which acts as the European distributor for publications of the American Association of Petroleum
Geologists, SEPM and the Geological Society of America.
Fellowship is open to those holding a recognized honours degree in geology or cognate subject and who have at least
two years' relevant postgraduate experience, or who have not less than six years' experience in geology or a cognate
subject. A Fellow who has not less than five years' relevant postgraduate experience in the practice of geology may
apply for validation and, subject to approval, may be able to use the designatory letters C Geol (Chartered Geologist).
Further information about the Society is available from the Membership Manager, The Geological Society,
Burlington House, Piccadilly, London W1V 0JU, UK. The Society is a Registered Charity, No. 210161

Published by the Geological Society from: Distributors

The Geological Society Publishing House USA
Unit 7 AAPG Bookstore
Brassmill Enterprise Centre PO Box 979
Brassmill Lane Tulsa
Bath BA1 3JN OK 74101-0979
UK USA
(Orders: Tel 01225 445046 (Orders: Tel (918) 584-2555
Fax 01225 442836) Fax (918) 548-0469)

First published 1996 Australia

Reprinted 1997 Australian Mineral Foundation
63 Conyngham Street
The publisher makes no representation, express or Glenside
implied, with regard to the accuracy of the information South Australia 5075
contained in this book and cannot accept any legal Australia
responsibility for any errors or omissions that may be (Orders: Tel (08) 379-0444
made. Fax (08) 379-4634)

© The Geological Society 1996. All rights reserved. No India

reproduction, copy or transmission of this publication Affiliated East-West Press PVT Ltd
may be made without written permission. No paragraph of G-l/16 Ansari Road
this publication may be reproduced, copied or transmitted New Delhi 110 002
save with the provisions of the Copyright Licensing India
Agency, 90 Tottenham Court Road, London W1P 9HE. (Orders: Tel (11) 327-9113
Users registered with the Copyright Clearance Center, Fax (11) 326-0538)
27 Congress Street, Salem, MA 01970, USA: the item-fee
code for this publication is 0305-8719/95/$07.00. Japan
Kanda Book Trading Co.
British Library Cataloguing in Publication Data Tanikawa Building
3-2 Kanda Surugadai
A catalogue record for this book is available from the Chiyoda-Ku
British Library Tokyo 101
Japan
ISBN 1-897799-52-7 (Orders: Tel (03) 3255-3497
Fax (03) 3255-3495)
Typeset by EJS Chemical Composition,
Midsomer Norton, Bath BA3 4BQ, UK

Printed by The Alden Press, Osney Mead, Oxford, UK

 Contents

HALL, R. & BLUNDELL,D. J. Tectonic evolution of SE Asia: introduction vii

Part 1: Present-day tectonics

MCCAFFREY, R. Slip partitioning at convergent plate boundaries of SE Asia 3
MALOD, J. A. & KEMAL, B. M. The Sumatra margin: oblique subduction and lateral
displacement of the accretionary prism 19
RANGIN, C., DAHRIN, D., QUEBRAL,R. & THE MODEC SCIENTIFICPARTY. Collision and
strike-slip faulting in the northern Molucca Sea (Philippines and Indonesia): preliminary
results of a morphotectonic study 29
RICHARDSON,A. N. & BLUNDELL,D. J. Continental collision in the Banda arc 47
SNYDER,D. B., MILSOM,J. & PRASETYO,H. Geophysical evidence for local indentor tectonics
in the Banda arc east of Timor 61
HUGHES, B. D., BAXTER, K., CLARK,R. A. & SNYDER,D. B. Detailed processing of seismic
reflection data from the frontal part of the Timor trough accretionary wedge, eastern Indonesia 75
MILSOM, J., KAYE,S. & SARDJONO.Extension, collision and curvature in the eastern Banda arc 85

Part 2: Tectonic development of Southeast Asia

METCALFE, I. Pre-Cretaceous evolution of SE Asian terranes 97
PACrd-IAM, G. Cenozoic SE Asia: reconstructing its aggregation and reorganization 123
HALL, R. Reconstructing Cenozoic SE Asia 153
SIMANDJUNTAK,T. O. & BARBER,A. J. Contrasting tectonic styles in the Neogene orogenic
belts of Indonesia 185
RICHTER, B. & FULLER, M. Palaeomagnetism of the Sibumasu and Indochina blocks: implications
for the extrusion tectonic model 203
STOKES, R. B., LOVATTSMITH,P. E & SOUMPHONPHAKDY,K. Timing of the Shan-Thai-Indochina
collision: new evidence from the Pak Lay Foldbelt of the Lao PDR 225
LOVATr SMITH,P. F., STOKES,R. B., BRISTOW,C. & CARTER,A. Mid-Cretaceous inversion in the
Northern Khorat Plateau of Lao PDR and Thailand 233
HtrrCHISON, C. S. The 'Rajang accretionary prism' and 'Lupar Line' problem of Borneo 247
OMANG, S. A. K. • BARBER,m. J. Origin and tectonic significance of the metamorphic rocks
associated with the Darvel Bay Ophiolite, Sabah, Malaysia 263
NGAH, K., MADON,M. & TJIA, H. D. Role of pre-Tertiary fracture in formation and development
of the Malay and Penyu basins 281
TJIA, H. D. & LIEW, K. K. Changes in tectonic stress field in northern Sunda Shelf basins 291
CLENYELL, B. Far-field and gravity tectonics in Miocene basins of Sabah, Malaysia 307
McCOURT, W. J., CROW, M. J., COBBING,E. J. & AMIN, T. C. Mesozoic and Cenozoic plutonic
evolution of SE Asia: evidence from Sumatra, Indonesia 321
SAMUEL, M. A. & HARBURV,N. A. The Mentawai fault zone and deformation of the Sumatran
Forearc in the Nias area 337
WAKITA,K., SOPAHELUWAKAN,J., MIYAZAKI,K., ZULKARNAIN,I., & MUNASRI.Tectonic
evolution of the Bantimala Complex, South Sulawesi, Indonesia 353
vi CONTENTS

WILSON, M. E. J. & BOSENCE, D. W. J. The Tertiary evolution of South Sulawesi: a record in

redeposited carbonates of the Tonasa Limestone Formation 365
BERGMAN, S. C., COFFIELD,D. Q., TALBOT, J. P. • GARRARD,R. A. Tertiary tectonic and
magmatic evolution of western Sulawesi and the Makassar Strait, Indonesia: evidence for
a Miocene continent-continent collision 391
ALI, J. R., MILSOM, J., FINCH, E. M. & MUBROTO,B. SE Sundaland accretion:
palaeomagnetic evidence of large Plio-Pleistocene thin-skin rotations in Buton 431
VROON, P. Z., VAN BERGEN, M. J. & FORDE, E. J. Pb and Nd isotope constraints on the
provenance of tectonically dispersed continental fragments in east Indonesia 445
LINTHOUT, K., HELMERS, H., WIJBRANS,J. R. & VAN WEES, J. D. A. M. 4°Ar/a9Ar constraints
on obduction of the Seram ultramafic complex: consequences for the evolution of the southern
Banda Sea 455
CHARLTON, T. R. Correlation of the Salawati and Tomori Basins, eastern Indonesia: a constraint
on left-lateral displacements of the Sorong fault zone 465
MALAIHOLLO,J. E A. & HALL, R. The geology and tectonic evolution of the Bacan
region, east Indonesia 483
BAKER, S. & MALAIHOLLO,J. Dating of Neogene igneous rocks in the Halmahera region:
arc initiation and development 499
PUBELLIER, M., QUEBRAL,R., AURELIO, M. & RANGIN, C. Docking and post-docking escape
tectonics in the southern Philippines 511
CROWHURST, P. V., HILL, K. C., FOSTER, D. A. & BENNETT, A. R Thermochronological and
geochemical constraints on the tectonic evolution of northern Papua New Guinea 525
WOPFNER, H. Gondwana origin of the Baoshan and Tengchong terranes of west Yunnan 539
ZI-IOU, Z., LAO, Q., CHEN, H., DING, S. & LIAO, Z. Early Mesozoic orogeny in Fujian,
southeast China 549
Index 557
 Tectonic evolution of SE Asia: introduction

R O B E R T H A L L 1 & D. J. B L U N D E L L
SE Asia Research Group, Department of Geology
Royal Holloway, University of London, Egham, Surrey TW20 OEX, UK
1 Fax no. (44) 01784 471780 Email robert.hall@gl.rhbnc.ac.uk

SE Asia is probably the finest natural geological
laboratory in the world yet is still not geologically
well known. It is a spectacular region in which the
manifestations and processes of plate collision can
be observed at present and in which their history
is recorded. It is a region that must be understood
if we are to understand mountain belts, arc
development, marginal basin evolution and, more
generally, the behaviour of the lithosphere in
collision settings. Furthermore, the region is
developing rapidly on the economic front, and a
major part of this rapid development is built on
natural resources. The geological reasons for the
distribution of these resources are therefore of
major importance for the inhabitants of the region
and for attempts to discover and exploit them.
These were some of the thoughts which stimulated
the collection of papers (Fig. 1) in this volume.
In order to understand the development of this
complex region an essential first step is to identify
the key features of the active tectonics and deter-
mine how plates and sub-plate lithospheric frag-
ments are moving. How successfully can rigid plate
tectonics be applied in describing present tectonics?
Where are the boundaries between plates? What are
the rates at which different parts of the region are
moving? The first part of this volume includes a
number of papers which deal with these questions,
based upon the application of GPS (Global
Positioning System) measurements to determining
the nature and rates of plate movements and plate
boundary zone deformation, results from the
BIRPS deep seismic reflection experiment in the
Banda arc, the first to cross a modern active margin,
and other geophysical and geological studies.
MeCaffrey provides a regional overview of recent
GPS results and earthquake data bearing on the
present plate tectonics. It is ironic that in such an
active region the identification of several plates and
determination of some important relative plate
motions, critical to a full kinematic description, are
still very uncertain. McCaffrey discusses the way in
which motion is partitioned in obliquely convergent
settings, almost the rule in SE Asia. Oblique con-
vergence is commonly inferred in the past for SE
Asia (as shown in many of the later papers in the
volume) and is often used as an explanatory tool in
orogenic belts elsewhere in the world. He draws
attention to the deformation of the upper plate in
these convergent settings and emphasizes the
importance of a three-dimensional understanding
of the process. In this rapidly evolving region,
information from slip vectors and geodetic
measurements will allow a fourth dimension of
time to be included as data accumulate, which will
raise the important problem of whether and for how
long the present motions can be assumed to extend
back into the past.
Two well-known areas of oblique convergence,
the Sumatra and Philippine Sea plate margins,
illustrate the realities of present tectonics, first
in identifying small plates or smaller tectonic
elements, and second in providing a kinematic
description. The increase in obliquity of con-
vergence between Java and north Sumatra has long
been considered to result in thrusting normal to the
subduction trench and arc-parallel movement on
the Sumatran fault. However, as McCaffrey points
out, this simple model does not predict some
important features, such as subduction west of the
Andaman Sea and differences in amounts of exten-
sion between north and south Sumatra. Malod &
Kemal discuss evidence from marine surveys off
Sumatra and propose that this area can be under-
stood as a number of plate slivers between the
trench and the Sumatran fault, with variations in
the partitioning of movement in different parts of
the Sunda forearc. This is explained as the result of
differences in coupling between the subducting and
overriding plates, possibly reflecting the presence
of major structures on the subducting slab, notably
an extinct spreading centre on the Indian plate.
Further east, the Philippine fault and trench are also
considered as the joint expression of partitioning of
oblique convergence, but once again applying this
simple model presents problems: how does this
paired trench-fault system link southwards into the
arc-arc collision of the Molucca Sea? Rangin et al.
tackle the first important problem of what is
actually present in the zone of transition by
mapping structures from the south Philippines into
Indonesia and determining which features belong

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, vii

Geological Society Special Publication No. 106, pp. vii-xiii.
viii R. HALL t~z D. J. BLUNDELL

.30*
t/° ,, q
lo~ 1110° SOUTHcHINA /1;~0°' 0 130° 140° 30~-

.-. !
," ,,",'," -" 12
".-,,,.; ,...
200..

" ~ A China
',.)~.. ',CAMBODIA ~ / 1 8 ~ Sea PHILIPPINE Pacific
Ocean
-10"

E
N -~ ~MALAYSIA~
MALAYSIA,, 30 MOLUCCAS

.o- N ~ SULAWES, ~ .~
32

- '),SUMATRA~ ~ ~ 22 ~ ,
: GUINEA
~-' : JAVA 24 - - BANDI.~o ~
Indian ~ B ~ U ~ oooj~----~ J
.lo* Ocean i~ .(-fl 4, 5, 61
AUSTRALIA
1O0° 110* 120* 1

Fig. 1. Areas discussed in the papers in this volume. 1: McCaffrey; 2: Malod & Kemal; 3: Rangin et al.; 4: Richardson
& Blundell; 5: Snyder et al.; 6: Hughes; 7: Milsom et al.; 8: Metcalfe; 9: Packham; 10: Hall; 11: Simandjuntak &
Barber; 12: Richter & Fuller; 13: Stokes et al.; 14: Lovatt Smith et al.; 15: Hutchison; 16: Omang & Barber; 17: Ngah
et al.; 18: Tjia & Liew; 19: Clennell; 20: Mccourt et al.; 21: Samuel & Harbury; 22: Wakita et al.; 23: Wilson &
Bosence; 24: Bergman et al.; 25: Ali et al.; 26: Vroon et al.; 27: Linthout et al.; 28: Charlton; 29: Malaihollo & Hall;
30: Baker & Malaihollo; 31: Pubellier et al.; 32: Crowhurst et al.; 33: Wopfner; 34: Zhou et al.

to which plate, a demanding task. Their use of the set, in the upper plate, is antithetic, and more
technique of multibeam mapping, combined with recent. They argue that these structures changed
geophysical evidence, is producing highly detailed orientation and migrated north because buoyant
maps, in many cases of much higher quality than continental crust blocked the subduction zone,
those available for land areas nearby, and revealing although Australia continued to move north. Using
unsuspected structures. the BIRPS images to estimate the volume of
The arguments surrounding the tectonics of the material in the collision zone they infer that an
Banda arc, and some of the complexities of the earlier Neogene collision event, involving a
collision zone, for example its curvature, age of separate microcontinental fragment or outer margin
the Banda Sea, and significance of deep basins such high, must have preceded the present phase of
as the Weber trough, also illustrate the difficulties collision which began at about 2.4 Ma. Snyder
of linking simple kinematic models to reality. Very e t al. draw attention to unusual features of the
recently two BIRPS deep seismic reflection profiles BIRPS 'Timor' sector of the Banda orogenic zone,
have crossed the Banda arc and imaged the deep including a Bouguer anomaly and thin sediment
structure of the central part of the Australia- cover implying a thicker continental crust than
SE Asia collision zone. Richardson & Blundell normal beneath the accretionary complex, a very
discuss how deep reflectors can be linked to recent narrow forearc, and presence of the volcano
seismicity, and their connection to structures on Gunung Api in the backarc. Like Richardson &
land. They identify two sets of divergent structures Blundell, they infer the presence of additional
in the collision zone: a southern set dipping in the Australian margin crust but suggest it formed either
same direction as subduction is related to the a promontory of the Australian margin or a pre-
subducted Australian margin, whereas a northern breakup extensional basin within the former pas-
 INTRODUCTION
sive margin, now inverted and thickened. Also like
Richardson and Blundell they deduce that the
deformation has advanced across the collision
zone, causing crustal fracturing in the backarc
region and allowing magmatic uprise and under-
plating beneath Gunung Api. Hughes et al. have re-
processed part of the Timor section to bring out the
details of the Timor trough and the accretionary
wedge developed on its northern flank. Australian
shelf sediments are clearly imaged, relatively
undisturbed, and beneath a low-angle decollement
above which are highly deformed sediments
stacked in a complex set of thrust sheets indicative
of the northward underthrusting of the Australian
continent. However, a covering drape of sediment
indicates that deformation in this area has been
inactive for about 1 Ma.
One feature that has bedevilled discussions of the
Banda arc is the significance of the deep troughs
which parallel the arc, which in a superficial way
have characteristics of trenches, but when studied
more carefully seem to be in the wrong position.
Milsom et al. present evidence that the deformation
front between Australia and the Banda arc extends
between the Kai islands, in the region of maximum
curvature of the arc, rather than following the
apparent subduction-related feature of the Aru
trough. They conclude that the arc must have
been a continuous feature before collision with a
curvature acquired recently.
The present setting of the region is both the start
and the end point in attempts to comprehend its
development. The start, because an understanding
of what is present provides the first clues of
what has happened in the past few millions or tens
of millions of years, leaving traces in the slabs
beneath the arcs, the arcs themselves, and in the
stratigraphic record of the arc regions as well as the
Sunda continental shelf further north. The end,
because the present is the result of what has
happened in the past, and many of the present
features of the region may be explicable only in
terms of what has now disappeared and can only be
inferred. Thus, our understanding of the region and
its development is an iterative synthesis, each new
step forward requiring a reconsideration of present
and past. The second section of the volume there-
fore begins with three overview papers intended to
synthesize earlier observations which may provoke
new research improving and extending our present
regional understanding.
Metealfe reviews the pre-Cretaceous develop-
ment of SE Asia and east Asia and shows that the
region has grown by addition of allochthonous
terranes which separated from different parts of
Gondwana. Their northwards movement was
accompanied by the opening and closing of
three successive oceans: the Palaeo-, Meso- and
ix
Ceno-Tethys. Assembly began with the formation
of Cathaysia in the Late Devonian-Early
Carboniferous and its growth continued within the
Palaeo-Tethys in the later Palaeozoic. As the
Meso-Tethys opened by Late Carboniferous-Early
Permian rifting of the northern margin of
Gondwanaland so the Palaeo-Tethys began to
close, and so subsequently did the opening of the
Ceno-Tethys result in elimination of the Meso-
Tethys. The final stage of the reassembly of
Gondwana in Asia is not yet complete but the
present complexity of the region, as well as the
comprehensive grasp of so many disciplines in
earth science required to attempt the task of
describing its development, give an indication of
Metcalfe's achievement in producing a compre-
hensible synthesis.
A description is an essential step in identifying
the driving processes and modelling the develop-
ment of the region. Reconstructing SE Asia in the
Cenozoic requires a description of India-Eurasia
collision, the motion of the Philippine Sea plate and
the present collision of Australia with eastern
Indonesia. Eurasian extrusion models have proved
popular with many workers in SE Asia, partly
because of the striking similarities of plasticine
models to tectonic maps, and partly because they
offer a means to explain and link different events.
As with many major advances, the extrusion model
has also been effective in provoking other new
ideas and the search for new evidence. Paekham
considers the relationships between the observed
geology and different models of the region, review-
ing the timing of different events, the evidence
from crustal volumes for extrusion, and the
relationship between predicted and observed
palaeomagnetic measurements. In east Asia and
western SE Asia the estimates of rates and amounts
of movements predicted by the extrusion model,
and those that can be determined, are becoming
much closer. The inadequacies of the fit cannot just
be explained as weaknesses of the model but also
emphasize the variable quality of data and their
uncertainties. Packham concludes that a regional
understanding of SE Asia requires a better under-
standing of the timing of deformation, especially
uplift, in the Himalayas and central Asia, a clearer
picture of whether rotations detected using palaeo-
magnetic data are regional or local, and much new
stratigraphic information, particularly in the eastern
part of the region.
The key role of palaeomagnetic data in develop-
ing regional models is a theme discussed by Hall in
an attempt to produce a kinematic model for the
whole of Cenozoic SE Asia. At the centre of the
region is the island of Borneo from which rotations
recorded appear to be in conflict with those pre-
dicted by an indentation model. Further east, new
X R. HALL • D. J. BLUNDELL
data from the Philippine Sea plate confirm long-
term clockwise rotations of the whole plate
suggested by many earlier studies. The attempt to
synthesise these results suggests that, even if
Cenozoic extrusion of continental fragments from
east Asia is accepted, this has not been the most
important driving force in the development of the
marginal basins of eastern SE Asia. According to
this model, development of marginal basins was
linked physically and temporally, and opening
appears to be mainly subduction-related rather than
indentor-driven. This new model does suggest
some possible configurations for the region which
are different from those previously accepted and
may provoke reconsideration of some evidence,
especially the habit of looking for local explana-
tions of tectonic phenomena. The animation which
accompanies the paper reminds us that regional
events may have causes outside the immediate area,
and the manifestation of plate movement changes
may propagate gradually across the whole region,
as plate boundary changes in one area cause
changes in others. Several aspects of the model,
such as the proposed ages of some basins, for
example the Banda Sea, remain to be tested by
ocean drilling, palaeomagnetic work, and strati-
graphic and structural studies.
The remaining papers in the volume are arranged
broadly in geographical order. Simandjuntak &
Barber illustrate the variation in orogenic styles
from different parts of Indonesia. The present
diversity and complexity of tectonic processes in
SE Asia may provide keys to the interpretation of
other orogenic belts. The differences in the history
of deformation within the region may leave traces
other then geological structures and Richter &
Fuller discuss the still thorny question of the
implications of palaeomagnetic data from
Sundaland and Indochina. They conclude that
different parts of Sundaland and Indochina have
deformed in different ways; some parts as small
blocks with rotations indicating local deformation,
some extrusion-driven rotation, principally in
Indochina, and some parts recording the results of
deformation dominated by the oblique subduction
of the Indian plate. Distinguishing between such
areas for the whole SE Asia region is an important
task for geologists and palaeomagnetists for the
future.
Even the most carefully constructed regional
kinematic models are totally dependent on basic
data revealing the timing of tectonic events and the
evidence for them. The papers concerned with
areas around the South China Sea suggest the need
for revisions of models as well as new interpreta-
tions of relationships between effects and supposed
causes. In Laos, Stokes et al. argue that suturing of
the Shan-Thai and Indochina blocks occurred in
the Late Jurassic and that the Indosinian Orogeny,
currently assumed to be of Permian or Triassic age
is significantly younger than commonly assumed.
Based largely on seismic data, Lovatt Smith et al.
suggest regional tilting, compressive folding,
reverse faulting and basin inversion in Thailand
record important phases of structural development
which pre-date the currently assumed Tertiary age
of structuring. If correct, these interpretations have
implications for hydrocarbon exploration and
potential. Hutchison draws attention to the contrast
between tectonic models of the South China Sea
region and the geology of Borneo and proposes that
the term Rajang Group should be more carefully
applied. An older turbidite sequence, assigned to
the Rajang Group proper, represents an accre-
tionary prism compressed and uplifted between the
Schwaner Mountains volcano-plutonic arc and a
South China Sea microcontinent during an Eocene
orogeny. Similar but younger turbiditic rocks
deformed by a Miocene orogeny are interpreted not
as deposits of a forearc, but as derived from the
eroding and uplifted Rajang Group, and should be
separated from it. A further record of the late
Mesozoic or early Tertiary subduction setting of
the NE Borneo margin is to be found in the large
Darvel Bay Ophiolite Complex of Sabah.
Mineralogical and geochemical studies by Omang
& B a r b e r suggest its formation in a supra-
subduction zone environment, but with com-
plexities due to high T-low P deformation along a
transform fault. High P-T garnet pyroxenites and
amphibolites found as clasts in Miocene rocks were
derived from a metamorphic sole underlying the
complex, formed during subduction and emplace-
ment of the ophiolite.
Within the Sunda shelf are sedimentary basins of
the South China Sea and adjacent areas which
record a link between east Asian tectonics and the
plates beyond the subduction zones bounding SE
Asia. The importance of pre-existing structures in
controlling tectonic development is often forgotten
and Ngah et al. suggest that the Malay, Penyu and
West Natuna basins originated in the Late
Cretaceous as three rift arms that developed during
doming of continental crust above a mantle plume.
The hydrocarbon potential of these basins was
subsequently influenced by changes in stress
patterns which Tjia & Liew argue resulted from
the interplay of Eurasian extrusion driven by Indian
indentation, and changes in directions and rates of
motion of the plates in the Pacific and Indian
Oceans. Borneo is situated in the middle of this
region at the south side of the South China Sea and
ought to record the effects of these changes. The
geology of Sabah is therefore of considerable
regional interest since, if Borneo has rotated, it is
an area where the consequences should be most
 INTRODUCTION
obvious. Clennell discusses the interplay between
large-scale regional plate motions (his 'far-field
tectonics') and pre-existing structures and local
tectonic influences for the development of the
unusual circular basins of Sabah. These papers on
sedimentary basins show that we are still some way
from clearly linking local and regional tectonics.
Tjia and colleagues show that there were reversals
in the sense of movement on important faults, that
the effects of fault movements differ from area to
area, and that there is still uncertainty in the timing
of fault movements. Clennell infers that basins in
Sabah appear not to record some tectonic events
because they were decoupled due to the thick-
nesses of underlying muds and m61anges.
Sumatra is an area where a long history of sub-
duction should be recognizable since the island is
usually considered to have been situated above the
northward-subducting Indian plate from at least the
Mesozoic. Despite this there appear to be distinct
periods characterized by igneous activity, separated
by intervals with little or none. McCourt et al. use
isotopic dating and geochemistry to identify
plutonic episodes and their character, which they
link to plutonism elsewhere in Sunda margin.
Understanding the tectonic significance of the
igneous episodes needs to be the next step forward.
McCourt et al. speculate that variations in the
obliquity of convergence, and collision of
allochthonous terranes are implicated, although
Neogene and younger strike-slip faulting compli-
cates the picture. It would be useful to consider the
Cenozoic history of this margin in the light of
known plate motions, although the major un-
certainty here is not the motion of the Indian plate
but the orientation and position of Sumatra;
different regional models show very different
configurations for the early Tertiary. The Cenozoic
history of this margin is clearly complex as
indicated by other papers in the volume and
Samuel & H a r b u r y show that this complexity is
still far from understood. In their paper, based on
detailed studies on land in the Sumatran forearc
islands, principally Nias, they interpret the
Mentawai fault system not as a strike-slip fault, but
as an extensional structure with late contractional
reactivation. If correct, at least one of the plate
slivers of the forearc proposed by Malod & Kemal
either does not exist or is of very recent origin.
Samuel & Harbury's work also illustrates the
importance of field-based studies in providing a
firm stratigrapbic basis for the interpretation of
other evidence, such as seismic and marine geo-
physical evidence, and they infer a long exten-
sional history for the Sumatran forearc, with major
extensional structures later reactivated, again with
possible links to changes in plate motions such as
the angle of convergence.
xi
A good stratigraphic base is fundamental to
attempts to interpret the tectonic evolution of the
region and Wakita et aL provide an example of
how this is achieved by detailed radiolarian studies.
The Bantimala Complex and Balangbaru
Formation of south Sulawesi record critical events
in the accretionary growth of the SE Sunda
margin and dating based on micropalaeontological
evidence is difficult to obtain and interpret since the
turbidite sequences yield few fossils, and these may
be reworked. Radiolaria, which can often be
assigned to narrow zones, provide an additional
means of comparing the ages of different litho-
logical units and suggest the Bantimala Complex
and Balangbaru Formation are contemporaneous,
requiring a modification of previous tectonic
interpretations. From the same region Wilson &
Bosenee show how redeposited limestones of the
Tonasa Limestone Formation can be used as indi-
cators of tectonic activity. Detailed measured
sections, well dated by fossils, illustrate how a
very clear palaeogeographic picture can be deduced
and linked to larger-scale tectonics. Their results
provide a basis for regional interpretation and will
be of considerable interest to those exploring for
hydrocarbon in the frontier regions of east
Indonesia. Sulawesi is currently much less well-
known than it deserves to be, especially con-
sidering its large size and critical position at the
Eurasia-Australia-Pacific junction, and recent
results have shown that simple tectonic models for
its development need reconsideration. Bergman et
al. present data from west Sulawesi which will need
to be incorporated in new models and speculate
on possible solutions. Of considerable interest is
the evidence, based on isotopic studies, for a
magmatic contribution from old Australian-type
continental crust to the Tertiary plutonic rocks of
west Sulawesi. Bergman et al. also focus attention
on the structures around the Makassar Strait. This
has previously been widely accepted to be an exten-
sional basin but they interpret it as a foreland basin
bounded by converging Neogene thrust belts, with
the late Miocene western Sulawesi magmatic
arc recording continent-continent collision. The
collapse of the orogenic belt is seen as the cause of
young extension in the region. The rapid changes
predicted by the model are certainly consistent
with the variety and speed of tectonic processes
currently observed in SE Asia.
The Neogene collisions of continental fragments
in Sulawesi are a principal cause of its geological
complexity and Ali et al, provide some insights into
how tectonic models can be tested using palaeo-
magnetic data. In Buton, large rotations are
recorded, but are apparently very local, and were
very rapid. This work reinforces the value of, and
need for, many more palaeomagnetic studies in SE
xii R. HALL • D. J. BLUNDELL
Asia in order to separate local from regional
motions. Buton is one of several continental
fragments which are now being reassembled in SE
Asia. These include Australian and Sundaland
material but their origin can often only be inferred
from indirect arguments, commonly controversial.
Vroon et al. suggest that isotopic evidence can
contribute to solving this problem and show how
different types of continental crust can be charac-
terized by analysis of igneous and sedimentary
rocks. They suggest that different parts of east
Indonesia have provenances in southern New
Guinea, north Australia, Pacific New Guinea and
Sundaland, leaving the tectonicians with an
additional tool but, in this area, some additional
problems to solve.
In the midst of the continental fragments of east
Indonesia are the deep basins of the Banda Sea, as
yet unsampled by the ocean drilling programme,
and of uncertain age. Linthout et al. report new
isotopic ages from Seram implying Neogene
spreading in the southern Banda Sea before
ophiolite obduction on Seram in the Late Miocene.
These ages are broadly consistent with ages of
rocks recovered during recent dredging in the
Banda Sea and with the tectonic reconstructions of
Hall, although the great depths and low heatflow
measurements remain apparent inconsistencies. On
the north side of the Banda Sea the Sorong fault
system separates Australia from the Philippine
Sea and Molucca Sea plates and terminates in the
continental fragments of east Sulawesi. The timing
of movement on the strike-slip faults has never
been clear, although recent work suggests this plate
boundary zone became a strike-slip system in the
early Miocene. The timing of movements and the
distribution of continental crust in this region is of
major interest, not least in the search for hydro-
carbons, since this is an area of established pro-
duction, recent discoveries, as well as active
exploration, all linked to Australian crust. Based on
stratigraphic arguments, Charlton argues that two
of the basins, the Salawati basin of western New
Guinea and the Tomori basin of eastern Sulawesi,
were originally a single sedimentary basin, now
separated by latest Miocene to Quaternary move-
ments on the fault, implying a left-lateral displace-
ment of about 900 km. Movement on the fault
system is one of the latest complications in the
development of east Indonesia; on the south side of
the fault system is Australian crust while on the
north side are the arc-arc collision of the Molucca
Sea and the clockwise-rotating Philippine Sea
plate. In the fault zone, which includes several
major splays, are fragments of both Philippine Sea
and Australian origin and Bacan is one of the
islands which includes rocks of both provenances.
Bacan therefore offers the possibility of elucidating
some of the history of the plate boundary zone, and
establishing the timing of tectonic events.
Malaihollo & Hall report new stratigraphic data
from Bacan which record the early arc history
of the Philippine Sea plate and the arrival of
continental crust, providing a basis for distinguish-
ing different tectonic models. Baker & Malaihollo
discuss the timing of volcanism in the islands of
Halmahera immediately north of Bacan, which
records the initiation of subduction of the Molucca
Sea plate beneath Halmahera and the development
of the present-day arc-arc collision. Volcanism
began in the middle to late Miocene and migrated
northwards implying that the double subduction
system was established between about 15-12 Ma.
This evidence still remains to be incorporated in
models linking east Indonesia to the Philippines.
Pubellier et al. provide further evidence from the
southern Philippines critical to linking the two
areas and understanding the development of the
present tectonic setting described earlier in the
volume by Rangin et al. Once again, the theme of
partitioning of oblique convergence is emphasized.
In addition, there are complications reflecting
Neogene changes in plate motions and the com-
plexities of intra-arc deformation. Strain has been
partitioned between several orientations of faults,
reactivated at different stages as thrusts and wrench
faults, as well as subduction zones. Of particular
interest is the way in which this development has
resulted in intra-arc extension and fragmentation
within the Philippines.
The interplay of subduction and strike-slip
faulting is a theme which appears in many of the
papers in the volume. Ancient strike-slip motion is
often difficult to demonstrate and quantify and is
consequently often neglected. However, there is
evidence, traditionally linked to the convergent
component of collision, which may be differently
interpreted. Crowhurst et al. show that fission
track data suggest that the Papuan metamorphic
rocks may be interpreted as representing early
Neogene extension after arc collision, rather than
contraction-related metamorphism. These argu-
ments may be applicable in other parts of SE Asia
where fission track and isotopic ages are revealing
unsuspected events, very young ages, and short
time periods for the very complex tectonic
evolution of many parts of the region.
The volume concludes with two papers from
south China accompanied by interpretations of
the timing and significance of events. Wopfner
suggests that the Baoshan and Tengchong Blocks
in western Yunnan have a Gondwana origin,
supported by the presence of Upper Palaeozoic
glaciomarine deposits, cold-water faunas and
Glossopteris. The terranes separated from
Gondwana in the Early Permian and docked with

INTRODUCTION
Cathaysia in the Late Triassic, although Tertiary
strike-slip faulting on the Nujiang Line has juxta-
posed the two terranes. Z h o u et al. reinterpret part
of the history of SE China, diverging in particular
from the traditional practice of relating major
unconformities to separate orogenies, and
suggesting instead that they record different stages
in the evolution of a single orogeny following early
Mesozoic collision between the South China and
the South China Sea blocks.
SE Asia is one of the most exciting regions of the
globe for any earth scientist. The size, difficulties
and practicalities of the region demand a long-term
investment of effort but the rewards are illustrated
by the papers in the volume. These give an insight
into how the history of the region will be un-
covered, as well as providing an overview of
present regional tectonics and its development.
Because of the rapid rates of movements in many
parts of the region new geodetic tools and
increasingly refined methods of examining earth-
quake data mean that realistic and rapid tests of
regional and global plate models are possible. The
challenge for the future is to examine how far
interpretations of these data can be pushed back
into the past, to improve kinematic descriptions and
models, and to identify the processes which have
led to the complexity of the region and which can
be applied to older orogenic belts. In these tasks,
there remains an important role for the field
geologist as well as those developing and applying
new technologies. An improvement of our under-
standing will have benefits for knowledge as well
. ° .
Xlll
as for the many inhabitants of the region, in helping
to develop its resources and mitigate its hazards.
This Special Publication arose from a conference on the
Tectonic Evolution of SE Asia held at the Geological
Society in London in December 1994. In addition to the
reasons outlined above for the conference, the London
University SE Asia Research Group wished to mark the
retirement of Dr A. J. Barber in 1994. Tony Barber
initiated several of the studies which are presented as
publications in this volume and was instrumental in
developing the programme of the SE Asia Research
Group over many years, particularly by fieldwork
throughout the region. We wish him a long and happy
retirement.
We thank all of the following who provided reviews of
manuscripts: J. R. Ali, M. Allen, M. G. Audley-Charles,
A. J. Barber, H. Bellon, S. C. Bergman, J. C. Briden, C. S.
Bristow, T. R. Charlton, J. Charvet, B. Clennell, D. Q.
Coffield, M. C. Daly, J. E. Dixon, C. Elders, R. Ellam,
A. Fortuin, M. Fuller, R. J. Garrard, N. S. Haile, N. A.
Harbury, K. C. Hill, A. J. Hurford, C. S. Hutchison, S. J.
Kelly, J. Malod, S. J. Matthews, R. McCaffrey,
M. Menzies, I. Metcalfe, J. S. Milsom, A. H. G. Mitchell,
G. Moore, S. J. Moss, R. J. Murphy, G. J. Nichols,
G. Packham, C. D. Parkinson, S. Polachan, M. Pubellier,
A. J. Racey, C. Rangin, J-P. Rehault, M. A. Samuel,
D. Snyder, P. Styles, R. E. Swarbrick, M. E Thirlwall,
E Tongkul, J. J. Veevers, R. von Huene, G. K. Westbrook,
H. J. Wensink. We are especially grateful to Diane
Cameron who carried out many of the major tasks related
to organizing the conference, in addition to her assistance
in reviewing and editing the manuscripts. Simon Baker,
Steve Moss and Moyra Wilson also provided considerable
assistance during the preparation of the volume, for which
we are extremely grateful.
 Slip partitioning at convergent plate boundaries of SE Asia
ROBERT McCAFFREY
Department of Earth & Environmental Sciences, Rensselaer Polytechnic Institute,
Troy, N e w York 12180, USA

Abstract: The active tectonics of SE Asia can be characterized by the interactions of large, rigid
plates separated by broad zones of deformation. The relative motion of these plates across their
boundaries is often partitioned in the sense that the normal and shear components occur on
different structures. Earthquake slip vectors and geological and geodetic measurements are used
to infer the degree to which oblique convergence, which is ubiquitous in SE Asia, is partitioned.
The active tectonics of Sumatra, the Himalayan thrust, the Philippines, the New Guinea fold-and-
thrust belt, the Huon-Finisterre collision, and the San Cristobal trench can be understood in
terms of upper plate deformation associated with oblique convergence. Western Java may
also exhibit partitioning of oblique subduction. Structures accommodating normal and shear
components of the motion are often very close. Arc-parallel strain rates are estimated for forearcs
of the region. The arc-parallel deformation of forearcs of the SE Asia region demonstrates that
plate convergence, whether normal to structure or not, is a three-dimensionalprocess.

When convergence between two plates is not
perpendicular to their boundary, shear stress
parallel to the plate boundary results in margin-
parallel shear strain within both plates. In a
collision belt or subduction zone, if one of the
plates is weak in shear, then the total slip may be
partitioned into shear and thrust components
(defined by the plate boundary orientation). Often
these shear and normal components of slip across
the boundary are accommodated on different geo-
logical structures. Oblique convergence is globally
much more common than trench-normal conver-
gence and often the obliquity varies along the
margin. The deformation in many convergence
zones may be understood in terms of such slip
partitioning. A widely cited example of these
parallel structures is along the San Andreas system,
which comprises a strike-slip fault and a parallel
fold-and-thrust belt (Mount & Suppe 1987) even
though the relative motion between the Pacific and
North American plates is only a few degrees from
the trend of the San Andreas fault.
Throughout SE Asia and the SW Pacific, we see
several examples of oblique convergence (Fig. 1),
both in subduction and collision settings. Studies of
Sumatra led Fitch (1972) to first propose that slip
was sometimes accommodated by parallel thrust
and strike-slip faults. Since that study, it has been
found that partitioning of oblique subduction is
globally quite common although it is rarely com-
pletely partitioned (Jarrard 1986; McCaffrey 1994).
Similar geometries, in which a major strike-slip
fault accommodates the boundary-parallel shear
strain, exist in the Philippines (Barrier et al. 1991)
and possibly in Irian Jaya (Abers & McCaffrey
1988). Other convergence zones that have clear slip
partitioning are the Himalayan thrust (Molnar &
Lyon-Caen 1989), the Aegean (Gilbert et al. 1994),
New Zealand (Anderson et al. 1993) and the
Aleutians (Ekstr(Sm & Engdahl 1989; McCaffrey
1992). McKenzie & Jackson (1983) argued that,
if plate motions are driven by ductile shear from
below, oblique convergence in continental settings
should be completely partitioned, that is, slip will
occur only on strike-slip and pure thrust faults. This
theory provides a test of crustal dynamics in
oblique collision zones but requires detailed
knowledge of the deformation. SE Asia is a good
region to test such ideas because oblique conver-
gence zones are plentiful and fast.
In the past few years much has been done to
improve our knowledge of the pattems of crustal
deformation in SE Asia. In particular, systematic
studies of large earthquakes, geological studies of
active faults, and geodetic measurements using the
Global Positioning System (GPS), that commenced
in the SW Pacific region in 1988, are revealing
important results, and surprises. The goal of this
paper is to review our knowledge of the crustal
deformation in SE Asia with attention to the
partitioning of slip between thrust and shear faults
in the broad deforming boundaries. First, con-
straints on the motions of the major plates are dis-
cussed in order to put bounds on the convergence
geometries and then the deformation within these
zones is described.

From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia,
Geological Society Special Publication No. 106, pp. 3-18.
 4 R. McCAFFREY

Himalaya } //

30°N
llEUt
<,
o _

India PSP
200N
Marianas

J~INDIAN 10°N
<.;3 •
SEA ~
:/ !
°
).
-/7, I
10 cm_ a -1 ,,,-
50 SV _ "~ New Guine~ 10°S
l SV

800E 100°E 1200E 140°E

Ca)

Io t ',t 0*
olomo° I.--- -~ [~ ~SV ,

?~)/'~~ t , s--z o ~ 10*S

I
-20*S
o

.Y
o
30°S
Pacific
0

o.

Oo

150°E 170°E i70°W

(b)
 SLIP PARTITIONING AT CONVERGENT PLATE BOUNDARIES 5

Plate motions The Philippine Sea plate

The broad kinematic framework of SE Asia,
determined by the relative motions of the
Pacific, Australian, and Eurasian plates, is
described to the first order by NUVEL-1A, the
global plate motion solution of DeMets et al.
(1994a). In addition, the motions of the Philippine
Sea, the SE Asian, and Caroline plates control the
large-scale tectonics of the region, and these are
discussed below.
SE Asia as a separate plate
One of the important regional problems is whether
or not SE Asia (SEA) moves independently of
Eurasia. This question is linked to collision in the
Himalayan region and its answer will provide con-
straints on the mechanism of continental collision.
Recently, the first direct measurements of the
relative motion between Australia and Indonesia
across the Java trench were made. Five GPS
measurements over 4 years on a baseline between
Christmas Island (Fig. la) and western Java reveal
a relative velocity (67 _+7 m m a-t rate at 011 _+4 °
azimuth) that matches the NUVEL-1A vector for
Australia relative to Eurasia (71 _+ 2 m m a -1 rate at
020 _+3 ° azimuth) within 1 cm a-1 (Tregoning et al.
1994) (Fig. l a). This single vector can be inter-
preted in several ways: that SEA moves slowly if
at all relative to Eurasia (within the uncertainties of
GPS); that SEA moves relative to Eurasia but the
pole of rotation is near western Java; or that elastic
strains contaminate the geodetic signal due to the
proximity of the GPS sites to the plate boundary.
Until more results become available, the author
favours the first view - that SEA moves very
slowly, 1 cm a -1 or less, northward relative to
Eurasia. Because the motion of SEA relative to
Eurasia is slow and the pole of rotation is poorly
constrained, in the following discussions SEA will
be treated as part of Eurasia.
The motion of the Philippine Sea plate (PSP) is still
poorly constrained due to lack of reliable kine-
matic data; only subduction zone earthquake slip
vectors are available (Ranken et al. 1984; S e n t
et al. 1993) and these are probably deflected by
local deformation of the leading edges of the upper
plates (McCaffrey 1994). Moreover, earthquake
slip vectors do not contain information about rates
of motion.
Despite these handicaps, S e n t et al. (1993)
revised estimates of PSP motion by constraining it
to comply with the NUVEL-1A global solution.
However, this does not improve the tie of the PSP
to its neighbouring plates; for example the un-
certainty in the latitude of the Pacific-PSP pole is
nearly 50 ° (Sent et al. 1993). Ongoing programmes
by several groups to measure the motion and
deformations of the PSP with GPS will provide the
required data to constrain the PSP motion relative
to the rest of the world. Recent GPS measurements
across Taiwan (Yu & C h e n 1994) show that the PSP
and Eurasia converge at azimuth 3070_+ 1 and
a rate of 86 _+2 m m a -1, which is in a similar
direction but 20% faster than the S e n t et al. (1993)
PSP-Eurasia pole predicts. A very long baseline
interferometry vector (Matsuzaka et al. 1991)
between sites on Japan (North American plate) and
the Izu-Bonin arc (PSP) also agrees with the
azimuth predicted by the combined PSP-North
America Euler pole (Sent et al. 1993; DeMets et al.
1994a) but is 15% faster. Both results suggest that
Sent et al. may have underestimated the angular
velocities for the PSP relative to Eurasia and North
America. In the following, the S e n t et al. (1993)
poles are used for the PSP but other poles are tried
in the kinematic analyses of forearcs around the
PSP.
The Caroline plate
Weissel & Anderson (1978) argued for the
existence of a separate Caroline plate (CAR)

Fig. 1. Tectonic maps. (a) SE Asia. Thrust faults have barbs on the hanging wall. Heavy black arrows show relative
plate motions at boundaries (vectors show upper plate moving relative to lower) using poles given in Table 1. Grey
arrows with error ellipses show GPS vectors of Cocos Island (CO) and Christmas Island (CI) relative to Java
(Tregoning et aL 1994) (black arrows show expected Australia-Eurasia vector). Small arrows pointing landward of
trenches show average slip vector (SV) azimuths of interplate thrust earthquakes along 400 km long segments of the
forearcs (lengths of arrows are scaled to number of SV in average). Thick grey lines show outline of broad deforming
region between the Indian and Australian plates (DeMets et aI. 1994b). SEA, SE Asian plate; PSP, Philippine Sea
plate; ODR, Oki Daito ridge; CAR, Caroline plate; JT, Java trench; BS, Banda Sea; MS, Molucca Sea; PF, Philippine
fault; PT, Philippine trench; AS, Andaman Sea; SS, Sunda Strait; NGT, New Guinea trench. Bathymetry contours at
2000 m intervals. (b) Southwest Pacific. Format same as (a). GPS vectors are relative to Pacific plate (Bevis et al.
1995). The large dot north of New Zealand is the pole of rotation of the N. Tonga forearc relative to the Pacific
determined from GPS vectors. NH, New Hebrides arc; LB, Lau Basin; LR, Louisville Ridge; SV, slip vector.
6 R. McCAFFREY

o ~_~ ,/ ~° ~) o IO*N

~ .j~v@o o YapTrench~--~ <'-~-~ ~

: y.. j ,.3 - -t5"N

,_ t!;~,~/(~lh~' Caroline Plate // ,." .~ ~ ' ) ~ , -J

" <'/.-',, ~ I0 mma -I

~, ,~~1 Slipvector I ~ ~k ~J
~ . b r ~ I I I I I • . 10i
132°E 137*E 1420E 147*E 152*E
Fig. 2. Plot of bathymetry, plate boundaries, relative plate motion vectors (thick arrows), and earthquake mechanisms
and slip vectors (thin black arrows) around the Caroline plate. Grey arrows show the expected motion of the Caroline
plate relative to the Pacific and Philippine Sea (PSP) plates and the thicker black arrows show the motion of the
Pacific relative to the PSP, all based on the Seno et al. (1993) poles. Earthquake mechanisms are from the Harvard
CMT catalogue. PSP, Philippine Sea plate.

between the Pacific, the PSP, and northern New
Guinea (Fig. la). This suggestion was based
largely on the morphologies of the Mussau trench
and Sorol trough (Fig. 2); the former thought to be
a subduction zone and the latter predominantly a
left-lateral shear zone with some extension. Rates
of motion of CAR relative to surrounding plates are
based on the observed increase in sediment thick-
ness away from the Ayu Ridge (Weissel &
Anderson 1978) and on inferences about the
amount of shortening at the Mussau trench
(Hegarty et al. 1983). Such motions are estimated
at less than 1 cm a-l.
Ranken et al. (1984) and Seno et al. (1993)
estimated rotation poles for CAR relative to the
Pacific and PSP by requiring convergence along the
Mussau trench and a combination of extension and
strike-slip at the Sorol trough. Recent earthquake
mechanisms along the Mussau trench and Sorol
trough (Fig. 2) are not consistent with a simple
plate boundary between CAR and the Pacific. The
slip vector for one event near the intersection of
the Sorol trough and Mussau trench is nearly
orthogonal to the expected slip based on the Seno
et al. Pacific-CAR pole, Events along the Mussau
trench itself indicate some thrusting but mostly
right-lateral strike-slip on very steep planes that are
consistent with the strike and polarity of the trench.
Thrust earthquakes along the Yap trench have slip
vectors that are consistent with either Pacific-PSP
or CAR-PSP motion. The alleged Caroline-Pacific
plate boundary is not a very compelling feature and
may simply be a region where stress concentrates
due to locally steep, inherited topography. In this
paper it is presumed that the 'Caroline plate' is
kinematically part of the Pacific plate. Even if the
Caroline plate is a separate block, ignoring it in the
kinematic analysis will not be critical because it
probably moves slowly relative to surrounding
plates.
Deforming boundary zones and
slip partitioning
By far the most elusive aspect of SE Asian
tectonics is the description of the deforming zones
between the rigid portions of the major plates
described above. One way to view the deformation
of the plate boundary zones is in terms of slip
partitioning. Before discussing slip partitioning at
convergent margins of SE Asia, some basic geo-
metrical concepts are reviewed. The mechanics of
slip partitioning are discussed by McCaffrey (1992)
and Platt (1993) but this paper concentrates on the
kinematic consequences. Most of the inferences
 SLIP PARTITIONING AT CONVERGENT PLATE BOUNDARIES
drawn here are based on the deflections of slip
vectors of interplate earthquakes away from the
expected convergence direction between the two
plates (Fig. 1 shows orientations of slip vectors and
plate vectors along plate boundaries). Earthquake
slip vector deflections tell us mostly about the arc-
parallel component of deformation within the upper
plate, because most convergence is at a high angle
to the trench normal, arc-normal deformation of the
forearc does little to deflect the slip vectors
(McCaffrey 1991). If the plate convergence vector
and the orientation of the trench are known, then
the arc-parallel rate of motion v s of the forearc
relative to the upper plate can be inferred from the
slip vector based on a simple kinematic relationship
(Fig. 3). Moreover, if the obliquity varies along the
margin then v s may vary as well, resulting in an
arc-parallel strain rate. Note that the arc-parallel
strain rate arises from the gradient in obliquity, not
from the obliquity itself, and we will see that
arc-parallel strain can occur where the obliquity is
zero. Hence, even though subduction may be
locally perpendicular to the trench and apparently
two-dimensional, rapid arc-parallel deformation of
the upper plate can occur, producing a three-
dimensional strain field.
Earthquakes used are those that reveal slip
between the subducting plate and the overriding
forearc. McCaffrey (1994) discusses the selection
of data (i.e. trench orientations, plate convergence
vectors and slip vectors) needed for this type of
analysis. Slip vectors are averaged in 5 0 - 1 5 0 k m
long segments of the trenches. This paper uses the
Harvard centroid-moment tensor solutions from
1977 through August 1994 and other published
solutions for earthquakes prior to 1977. It is
assumed that the downgoing plate does not deform
rapidly; slip vector deflections require motions of a
large percentage of the plate motion rate and it is
unlikely that subducting plates are sheared at such
high rates. The goal is to infer arc-parallel slip rates
v s (Fig. 3) and arc-parallel strain rates (arc-parallel
gradients in Vs) in forearcs in the SE Asia region
(Table 1). Where available the slip vector derived
slip rates are compared to geological or geodetic
measurements of slip rates.
Sumatra and Java
Fitch (1972) correctly inferred that oblique con-
vergence at Sumatra was partitioned into a compo-
nent of thrusting at the Java trench that was normal
to the trench and a component of arc-parallel shear
on the Sumatra fault. In the past few years, detailed
investigations utilizing more and better data have
refined the kinematics of the Sumatra region.
Attempts to treat the Sumatra forearc, between the
trench and the Sumatra fault, as a rigid plate (e.g.
7
Upper Plate
/~ A A A
VS 1 S l i p vector
~<_....z_.~,
/
~"~ - - ~ / ~ Forearc
z ' ~ IVn
Trench
~" Vp r SubductingPlate
v ; predicted plate convergence vector
T = trench-normal azimuth
@ - plate motion azimuth
13 - slip vector azimuth
g = trench normal - slip vector = T - 13
Obliquity: ~, = T - @
Slip vector residual: A[3 = [3 -
Forearc slip rate relative to Upper Plate:
v s = v ( sin y - cos ~/ tan ~ )
Fig. 3. Geometry of slip partitioning as inferred from
earthquake slip vectors. The important items are ~/, the
plate convergence obliquity, ~, the slip vector obliquity,
and v s, the slip rate of the forearc in the arc-parallel
direction relative to the upper plate.
Curray 1989) generally failed to explain subduction
west of the Andaman Sea, where the trench takes a
turn to the NNE (Fig. l a) or the difference in the
amount of extension in the Sunda Strait (100 km)
(Huchon & Le Pichon 1984; Harjono et al. 1991)
and Andaman Sea (400 km) (Curray et al. 1979).
These features can be explained if the forearc
stretches in addition to being translated northwest-
ward relative to SEA (Diament et al. 1990;
McCaffrey 1991).
Using a convergence direction at the Java trench
of N3°E and a rate of 75 m m a -t, an arc-parallel
extensional strain rate of 3-4 × 10-8 a-1 is required
of the Sumatra forearc to match the deflections of
the slip vectors (McCaffrey 1991). Tregoning et al.
(1994) show that the N U V E L - 1 A Australia-
Eurasia vector is approximately correct for the Java
trench. The NUVEL-1A Australia-Eurasia con-
vergence direction results in a lower inferred arc-
parallel strain rate (1.8_+0.5 × 10-8 a-I; Fig. 4a,
Table 1) than the N3°E direction does because it
matches better the slip vectors west of Sumatra.
Such arc-parallel stretching of the forearc
implies that the slip rate on the Sumatra fault
8 R. McCAFFREY

Table 1. Arc-parallel strain rates for forearcs

Trench name Pole of Number of Number of Arc-parallel Correlation RMS

rotation* trench slip strain rate coefficient (mm a-z)
segments vectors (x 10-8 a -1)

Himalaya N IND-EUR 13 19 2.2 _+0.5 0.79 12

Izu S PAC-PSP 7 67 0.7 _-!-0.7 0.52 6
Java N AUS-EUR 15 40 -0.1 _+0.4 -0.05 12
Mariana S PAC-PSP 10 61 1.7 + 0.5 0.81 6
New Hebrides N PAC-AUS 11 124 12.3 + 2.5 0.96 25
N. Ryukyu S PSP-EUR 5 27 1.3 _+0.3 0.92 2
N. Tonga N PAC-AUS 4 99 16.3 _+5.4 0.93 14
Philippine S PSP-EUR 23 140 2.6 _+ 1.0 0.60 18
S. Ryukyu S PSP-EUR 4 14 0.6 _ 1.6 0.13 7
S. Tonga N PAC-AUS 15 481 -0.1 _+0.4 0.15 11
Sumatra N AUS-EUR 18 102 1.8 _+0.5 0.84 11

* N, Demets et al. (1994a) NUVEL-1A; S, Seno et al. (1993); IND, India; EUR, Eurasia; PAC, Pacific; PSR Philippine
Sea plate; AUS, Australia.
Trench segments are between 50 and 150 km long.
Positive values of strain rates are arc-parallel extrensional, and negative values are compressional.

increases to the N W (McCaffrey 1991), w h i c h is
supported by the few direct m e a s u r e m e n t s of slip
rate on the fault (Fig. 4a). At its southern end, near
5°S, the slip rate is less than 10 m m a -z (Bellier
et al. 1991) but increases to about 10 m m a -1 near
the equator and to about 28 m m a -1 near 2.2°N
(Sieh et al. 1994) (Fig. 4a). The average gradient
in the slip rate on the fault is close to that pre-
dicted f r o m slip vector deflections; an increase in
28 m m a -1 over 1200 k m f r o m the S u n d a Strait to
2.2°N gives a strain rate of 2.3 x 10 -8 a -1.
As is seen in Fig. 4a, slip vectors and the
N U V E L - 1 A Australia-Eurasia pole give approxi-
mately the correct slip rate for the Sumatra fault but
also predict left-lateral, arc-parallel shear at a rate
of 15-20 m m a -1 at the longitude of W Java. A pole
of rotation that fits the earthquake slip vectors south
of Java, and hence does not require upper plate
d e f o r m a t i o n near Java, predicts h i g h e r than
o b s e r v e d slip rates for the S u m a t r a fault
(McCaffrey 1991). Therefore, either the earthquake
slip vectors are p o o r indicators o f upper plate
deformation, there are faults other than the Sumatra
fault that a c c o m m o d a t e forearc deformation off
Sumatra (Karig et al. 1980; D i a m e n t et al. 1992), or
there is left-lateral shear through Java (Dardji et al.
1991; Malod et al. 1996). The constant v s value at
Java (Fig. 4a) indicates a near-zero strain rate. T h e
Java forearc m a y be m o v i n g relative to the Sunda
shelf but, unlike the Sumatra side, not u n d e r g o i n g
internal strain.
The m e c h a n i s m for the arc-parallel stretching of
the Sumatra forearc is unclear but involves some
extension on cross-forearc faults, such as in the
strait b e t w e e n Java and S u m a t r a ( H u c h o n &
Le Pichon 1984; Harjono et al. 1991; M a l o d et al.
1996), and s o m e strike-slip faults crossing the fore-
arc (Karig et al. 1980). T h e largest shallow, strike-
slip earthquake to occur in Sumatra in the past 30
years was beneath the forearc and had nodal planes

Fig. 4. Plot of plate convergence obliquity (shaded curves), slip vector obliquity (open circles), arc-parallel forearc
slip rates (v s, closed circles), and arc-parallel slip rate of subducting plate (v., thin solid curve) along convergence
zones of SE Asia. The horizontal axis is distance in kilometres along the de(ormation front or trench (also labelled
with approximate latitude or longitude). Positive distance is the direction faced if the thrust fault dips to the viewer's
right (for example, positive distance is west if the fault dips N). The vertical axis is angles in degrees and slip rates
in mm a -z. Plate obliquity is inferred from trench-normals, plate convergence directions (poles used are listed in
Table 1), and their uncertainties. Thin dashed curves, when present, show obliquity from poles of rotation based on
fitting slip vectors at the trench. Slip vector obliquity is the angle between the slip vector and the direction normal to
the trench (Fig. 3). Positive values of v s indicate right-lateral shear and negative values are left-lateral. Heavy dark
straight lines show the best-fit to the values of v s, the slope of which (dVs/dX) is the arc-parallel strain rate (Table 1;
positive slopes show arc-parallel extension, negative slopes show arc-parallel compression). Triangles show geodetic
and geological estimates of arc-parallel slip rates, as discussed in text. NZ, New Zealand.
 SLIP PARTITIONING AT CONVERGENT PLATE BOUNDARIES 9
that were rotated by 45 ° relative to the trend of the Himalaya
Sumatra fault (Zwick & McCaffrey 1991). It is
likely that more detailed study of the Sumatra fore- Due to the extensive deformation within Asia north
arc structure will find complex deformation, such of the Himalaya the convergence rate and direction
as in the Aleutian (Geist et al. 1988) and Cascadia at the Himalayan front is poorly constrained. The
(Goldfinger et al. 1992) forearcs. convergence rate is estimated to be 15 m m a -1

/.-., -- 90
i i i i a
"7 E W
Sumatra ~~" 60
Java ~'~' ~'~'~''
¢~ ,
!
Timor 30

, ..... o
C
-30

,0.7E ' k'C---i122.3E 111. 1104"sE I 1"5S 15"4N -60

© 0
(a) 1000 2000 3000 4000 5000

~E ~ f Himalaya_
, ,... ~ + ~.~~J":~Ji
1~
w 59~ o

. 30 -75

~5 4E 90.9E 5.8E 80.9E 76.6E 50 00

0 500 1000 1500 2000 0 250 500 750 1000 1250

-~"
(b) , , , - 25
(e) ~ 60

30
t ... o
-25
-2.'
.::...:.::..~
-50
g" /// ' New -5,
•~ .1~11
~¢~/'~' Hebrides -75
-30

•-Q 75~/~~ ..........."........ 1171.7E 121.1S 17.8SI - 100 - -60

O 0

._.
(c) 400 800 1200
60
0
(t) 500 1000 1500 2000
60

}0

~D

3 4 -30 -30
if-,
-60

1f'4.6S 2.4S 0.9S 9.5S -90 -90

O 0 1000 2000 3000 0 1000 2000 3tRIO
(d) Distance, (km) (g) Distance, (kin)
10 R. McCAFFREY

(Molnar & Lyon-Caen 1989), or about one-third of
the total India-Eurasia motion. Due to the large
curvature of the Himalayan front, the obliquity
varies by more than 90 ° along it (Fig. 4b). Despite
this large variation in obliquity, earthquake slip
vectors tend to remain perpendicular to the thrust
front instead of following the convergence
direction (Fig. 4b). Kinematically this requires
deformation of the overriding plate. The uniform
arc-parallel strain rate estimated from slip vectors

~ i ~ /~~-,,,~, " m

-k. ~i ,

~ "~'--~\ ~
-

SundaShelf
/\ -~ "

j~ \ ~ - ~ B a n d a S e a _ ~ - - - - - ~
°

l~.~*o)~)~?h
"

. . . . - ...... ~ ,o , . ~ . ~ ! ~ j f ~ ' . ~ < ~ 4 ~'~ ¥,~

9.s

....... -

--~ -" ~Z .4 11°S

-,, , I, , , , ,
116°E 118°E 120°E 122°E 124"E 126°E 128°E 130°E 132°E
(a)
I°N

t°S

t°S

i*S

#°S

1330E 135"E 137"E 139"E 141°E 143"E 145"E 147°E 149"E

(b)
Fig. 5. Fault plane solutions (1962-1984; M s > 5.5), slip vectors, and plate vectors for (a) eastern Indonesia and
(b) New Guinea. Size of earthquake symbol is scaled to logarithm of seismic moment. Asterisks show active
volcanoes. Small arrows show slip vectors of interplate thrust earthquakes and large arrows show expected plate
convergence directions based on (a) Australia-Eurasia and (b) Australia-Pacific (DeMets et al. 1994a). MT,
Manokwari trough; BH, Bird's Head; SF, Sorong fault; CB, Cenderwasih Bay; TAF, Tarera-Aiduna fault; MTB,
Mamberambo thrust belt; HFTB, Highlands fold-thrust belt; NGT, New Guinea trench; MV, Markham Valley;
HF, Huon-Finisterre Range; BP, Bismarck plate; NBT, New Britain trench; PNG, Papua New Guinea.
 SLIP PARTITIONINGAT CONVERGENT PLATE BOUNDARIES
is 2.2 + 0.5 × 10-8 a-l (Fig. 4a) if the India-Eurasia
convergence rate is used and one-third of that
(i.e., 0.7 x 10-8 a-l) if two-thirds of India-Eurasia
convergence is taken up in Asia. Molnar &
Lyon-Caen (1989) estimate 1.0 _+0.5 x 10-8 a -1
(18 _ 9 mm a-t over a distance of 1800 km) for
the rate of E-W extension of Tibet.
Extension of the Himalayan frontal arc is thought
to be driven by gravitational collapse of the over-
thickened Tibetan Plateau, instead of shearing due
to oblique plate convergence. The evidence for this
is the correlation of faulting type with elevation.
Normal faulting is confined to elevations above
4000 m, strike-slip occurs at all elevations, and
thrust faulting is confined to elevations below
4 5 0 0 m (e.g. Molnar et al. 1993). However,
because the earthquakes showing rotated slip
vectors occur within 200 km of the thrust front, the
Himalayan forearc must also be extending along
strike at about 10-8 a-1. Therefore, while normal
faulting may be confined to high elevations, E-W
extension of the upper plate is not. As is the case
for the Sumatra forearc, such extension could be
accommodated by strike-slip faulting at low
elevations instead of normal faulting.
Note that the arc-parallel strain rate applies even
to the region near 86°E ( - k m 1000) where sub-
duction is perpendicular to the mountain front. (At
several other subduction zones in Fig. 4, similarly,
the strain rate is not zero even thought the plate
obliquity is near zero.) Even though v s ---0, the
margin-parallel gradient in v s is non-zero. In the
case of normal convergence, strain will occur in the
vertical plane containing the convergence vector
and the normal to the margin. Since v s represents
motion perpendicular to this plane, a gradient in v s
results in strain outside of this plane, producing a
three-dimensional strain tensor. The force that
stretches the forearc along its length arises from
the gradient in the arc-parallel shear stress due to
the increase in obliquity on both sides of the point
where margin-perpendicular convergence occurs
(McCaffrey 1992). Hence, subduction at a curved
margin is a three-dimensional process even where
it may be locally perpendicular to the trench.
Banda arc and Timor
Earthquakes suggest that the present stage of the
collision of the Australian continental margin with
the Banda arc probably involves more rapid con-
vergence at the backarc thrusts than at the Timor
trough (Fig. 5a). The Wetar and Flores backarc
thrust zones are considerably more energetic
seismically than the subduction thrust fault north
of the eastern Java trench and Timor trough
(McCaffrey 1988); all known large thrust earth-
quakes in this decade have occurred at the Flores
||
and Wetar thrusts. The M w = 7.9 December 1992
Flores earthquake is the latest and largest of many
to occur along the backarc thrusts.
Preliminary GPS results show that the slip rate
associated with backarc thrusting increases from
Bali eastward to about 50 mm a-1 at the eastern
end of the Flores thrust and reaches the full
expected rate of Australia-Eurasia relative motion
(70 mm a-1) at the Wetar thrust north of Timor
(Genrich et al. i994). GPS appears to be con-
firming the inference based on earthquakes
(McCaffrey 1988) and crustal structure (Silver
et al. 1983) that the eastern Sunda arc is rotating
anticlockwise about a pole in eastern Java.
Timor appears to be moving northward relative
to the Sunda Shelf (SEA) at about the same rate as
Australia, suggesting that the Timor trough no
longer accommodates subduction (Johnston &
Bowin 1981; McCaffrey & Abers 1991; Genrich
et al. 1994). The northward motion of the extinct
volcanic islands north of Timor is revealed with
GPS to be similar to that of southern Timor
(Genrich et al. 1994); hence, little of the northward
motion of Australia is now accommodated by
internal deformation of the Banda arc structure.
The island arc structure, bounded by the Timor
trough and the Wetar backarc thrust, is rigidly
thrusting over the backarc basin. The present
geometry is probably quite young; the Wetar thrust
exhibits less than 10 km of shortening (McCaffrey
& Nabelek 1986), which would take only 0.15 Ma
at the present rate of 60 mm a-1. Probably between
3 Ma, when collision began, and 0.15 Ma conver-
gence was in part accommodated by shortening and
thickening of the forearc (Snyder et al. 1993). The
Sunda arc now shows a gradual transition from
subduction of oceanic lithosphere south of Java to
a completed accretion of an island arc terrane to a
continental margin at Timor. In the process, con-
vergence has jumped from the forearc trench to the
backarc thrust.
South of Timor, the Timor trough trends about
N70°E while the plate convergence vector is
probably close to N10°E (Tregoning et al. 1994),
resulting in an obliquity of about 30 ° (Fig. 4a). The
geometry predicts that the Timor forearc should be
under ENE-trending, left-lateral shear and possibly
arc-parallel tension. Interplate earthquake slip
vectors (Fig. 5a) are too few to confirm this; values
of v s inferred from slip vectors range from zero to
the full arc-parallel motion rate (Fig. 4a). The few
available slip vectors show motions both parallel to
plate convergence and perpendicular to the trench
(Fig. 5a), suggesting a complex spatial variation in
slip partitioning (below we see that this bimodal
distribution of slip vectors is common at forearcs
with large curvature). Recent earthquakes in the
overriding plate suggest that both shearing and
12 R. M c C A F F R E Y

stretching are occurring. In particular, two shallow

earthquakes beneath Timor show left-lateral, strike-
H i | l a n d g hs u i~i:i~i:i i ~ ~i:!~i!:
slip on NE-trending, vertical planes and shallow, .....~: •!iii~i - 240
N-striking, normal faulting events are found north ~B iii~!i!i
of central Timor (Fig. 5a) (McCaffrey 1988). GPS
results also suggest that slow E - W stretching of the 180
arc may occur (Genrich e t al. 1994).
120
New Guinea
~!Thrust .............(3 ..............oi!ii!!!ii ' ' ' ~ ......................................!!iiiiiiii..........
The island of New Guinea accommodates oblique 60
convergence between the Pacific and Australian .
o
• ii!ii!i,o
o -

plates at a rate of about 12 cm a-1 (Fig. 5b) and

azimuth of N70°E. The young, high mountain belt,
called the Highlands fold-and-thrust belt, trends
at N100°E, so is about 60 ° away from being o iiiiiiiii PVN
n n I ~i~i~i~i I I ili~i i l -60
perpendicular to plate convergence. Large earth- 0 30 60 90 120 150 180
quakes suggest that shortening and shear are (a) Strike
distributed over much of the island (Abers &
McCaffrey 1988; McCaffrey & Abers 1991; Fig. Mamberambo
5b). Earthquakes within the Highlands are pre- I I •l :::i I I :!:i |
dominantly thrust and strike-slip with strikes 0 d 240
generally parallel to structure rather than perpen- FTB • :~i:J
dicular to the plate convergence direction (Fig. 6a).
Oblique convergence within the Highlands belt of
Irian Jaya is therefore largely partitioned. The
Mamberambo thrust belt (Fig. 5b) shows many
more earthquakes with planes that strike perpen-
dicular to the plate vector, suggesting less
partitioning there (Fig. 6b). 60
Recent GPS measurements (Puntodewo et al.
1994) reveal less than 2 cm a-1 of shortening and
less than 3 cm a-1 of left-lateral shear between sites 0
at the north and south coasts of E Irian Jaya (at
about 140°E) even though nearly 6 cm a -1 of • iii!i!i P~N
! ! | iiiiiiill | i i!i!ii~!i!- -60
shortening and 10 cm a-1 of left-lateral shear are
0 30 60 90 120 150 180
predicted by Pacific-Australia relative motion. At
(b) Strike
the longitude of central New Guinea, most of the
motion between Australia and the Pacific occurs
Fig. 6. Plot of rake angles vs. strike angles for earth-
offshore, probably at the New Guinea trench,
quakes in (a) the New Guinea fold-and-thrust belt (FTB)
contrary to inferences made from geology and between 135° and 144°E and (b) the Mamberambo
seismology. thrust belt. For each event, both nodal planes are
In westernmost New Guinea, plate motion shown (using open and closed circles). The vertical line
appears to be accommodated in a different manner. labelled FTB shows the trend of the fold-thrust belt,
GPS sites on the Bird's Head appear to move and PVN shows the trend perpendicular to the plate
relative to Australia at nearly the same rate as the convergence direction. Rake angles near 90 ° indicate
Pacific (Puntodewo et al. 1994; Stevens e t al. thrusting and those near 0 ° and 180° degrees are strike-
1995). In the vicinity of the Bird's Head, the slip (SS). Events with a strike near 100° and rake near
zero reveal left-lateral slip on a plane parallel to the
majority of Pacific-Australia motion occurs by
FTB.
large-scale W S W motion of the Bird's Head
relative to Australia. The fault (or faults) on which
this motion occurs is not clear although in the SW
the main boundary is probably the Tarera-Aiduna
fault (Hamilton 1979; Abers & McCaffrey 1988). shear zones in the world, being at least 7 cm a-l
Manuel Pubellier (pers. comm. 1994) suggests that (Stevens e t al. 1995). The Sorong fault, at least
a large, NE-trending shear zone exists landward of where it passes through the Bird's Head, cannot
the east coast of Cenderwasih Bay (Fig. 5b). If so, now form the main boundary between the Pacific
it could be among the fastest slipping continental and Australia because GPS sites on opposite sides
 SLIP PARTITIONINGAT CONVERGENT PLATE BOUNDARIES
of the fault show little relative motion (Puntodewo
et al. 1994).
In Papua New Guinea (PNG), the Huon-
Finisterre (HF) arc terrane is being emplaced onto
the Australian continent along a north-dipping
thrust fault that crops out in or near the Markham
Valley (MV) (Fig. 5b). Slip vectors of thrust earth-
quakes beneath the HF range north of the valley
remain perpendicular to the valley as it swings to a
NW trend (Fig. 5b). This rotation of the slip vectors
cannot be explained by a single pole of rotation
between rigid plates north and south of the valley.
Schouten & Benes (1996) argue that the rotation
of the slip vectors is caused by rapid deformation
south of the New Britain trench and MV. On the
bases of geological observations (Abbott et al.
1994), microearthquakes (Kulig et al. 1993), and
large earthquakes (Abers & McCaffrey 1994), it is
interpreted that the HF range is cut by cross-faults
that accommodate relative slip and rotations of
blocks of the HF range as it moves southward. In
the case of the HF block, slip partitioning is
accommodated by faulting at high angles to the
trench as in Cascadia and the Aleutians, rather than
arc-parallel faults, as in Sumatra.
Solomons and N e w Hebrides
The Solomon and New Hebrides trenches form the
boundary between the Pacific and Australian plates
and, due to their great curvature, are the sites of
large variations in convergence obliquity. The
deviations of slip vectors from the plate vectors
suggest that both forearcs deform significantly
and partition the slip between the plates. At the
southern end of the New Hebrides trench, west of
173.5°E, slip is partitioned completely so that thrust
earthquakes at the trench have their slip vectors
perpendicular to the trench and nearly perpen-
dicular to the plate convergence vector (Fig. 7a).
Going west around the bend in the trench, the slip
vectors remain perpendicular to the trench to 18°S
where the trench becomes nearly normal to the
plate vector. East of about 173.5°E, the slip vectors
rotate so that they form an angle of about 45 ° to the
trench and plate vector. The extensional strain rate
estimated for the southern New Hebrides forearc
from the slip vectors is 12.3 __+2.5 × 10-8 a -1
(Fig. 4c).
The Solomon trench becomes nearly parallel to
the plate vector at the San Cristobal section (Fig.
7b). Slip vectors for earthquakes between 162 ° and
165°E show two orientations: one nearly perpen-
dicular to the local trend of the trench and the other
nearly parallel to the plate vector. If partitioning
occurred by a detached forearc sliver moving along
the arc, then the events with slip vectors perpen-
dicular to the trench would be closer to the trench
13
(i.e. below the translating sliver) than those
showing oblique slip. The uncertainties in the
epicentral locations of these events preclude dis-
cerning such a pattern. Nevertheless, structures
accommodating shearing and thrusting must be
within a few tens of kilometres (the uncertainties
in epicentres) of each other. Some fault plane
solutions (Fig. 7b) show left-lateral, strike-slip
faulting on E - W orientated, nearly vertical faults;
these faults are consistent with the type needed to
partition the slip by a small sliver of the forearc
being translated eastward at a high rate relative to
the arc.
Tonga trench
According to the NUVEL-1A Pacific-Australia
pole of rotation, the Pacific plate subducts beneath
the Tonga arc in a direction nearly perpendicular to
the trend of the trench throughout most of its length
(Fig. lb), although the obliquity gets large at the
northern and southern ends (Fig. 4d). Earthquake
slip vectors at the Tonga trench from 15°S to
35°S, deviate by 20 ° or less from the NUVEL-1A
Pacific-Australia plate vector (Fig. 8a). If this pole
applies to the Tonga trench then slip vectors
suggest that the southern Tonga forearc translates
very slowly (10 mm a-1 on average), left-laterally
relative to Australia and does so as a rigid block
(Fig. 4d; Table 1).
GPS measurements by Bevis et al. (1995)
suggest that the northern Tonga forearc rotates
rapidly away from Australia as a rigid block (Fig.
lb). Slip vectors also suggest that the forearc is
rigid as they can be matched throughout most of
the length of the island arc with a single pole of
rotation at 54°S, 173°E (Fig. 8a). However, the
poles of rotation estimated from GPS and from
slip vectors are quite distant. The 3 GPS vectors
(relative to fixed Pacific plate) from Tonga forearc
sites are all matched within 2 mm a-1 in rate and 2 °
in azimuth by a pole at 32.6°S, 178.9°W (Fig. lb),
and an angular velocity of-7.1 o Ma (the distance
from the pole to the observation points is
constrained by the gradient in GPS velocities). This
pole falls near the central Tonga trench (Fig. lb)
and therefore cannot describe the motion of the
entire forearc from north to south because it
predicts spreading between the forearc and the
Pacific south of 33°S. There is probably a structure
within the forearc south of the GPS sites that
separates the N and S sections of the forearc. Most
probably this boundary is near 24°S where the
Louisville Ridge intersects the trench and spreading
in the Lau Basin appears to terminate (Fig. lb).
The northern Tonga arc goes through a large
bend until the trench is nearly parallel to plate
14 R. McCAFFREY

ll*S

15*S

190S

~,3"S

159°E 163*E 167°E 171*E 175*E

(a)
' ; ' a-~ Y 10°S

I ~ ,,'~~ Pacific /_ / ~ [ y

t
I*S

I I I
163*E 164*E 1650E 166°E
(b)

Fig 7. (a) Map of earthquake slip vectors (small arrows) and plate convergence vectors (large arrows) based on
Pacific-Australia pole (DeMets et al. 1994a) at the Solomon and New Hebrides arcs. SCT, San Cristobal trench.
Dashed outline box shows area of b. (b) Map of earthquake mechanisms (Harvard CMT 1977-1994; M s > 5,0) and
slip vectors at the San Cristobal trench. The large arrow shows the convergence between Australia and the Pacific.

convergence (Fig. 8b), in the same fashion as
the southern New Hebrides, San Cristobal, and
southern Marianas arcs. Like the San Cristobal
trench, slip vectors at the northern Tonga forearc
appear to describe two sets of directions; parallel to
the plate vector and perpendicular to the trench
(Fig. 8b). The averages of these slip vectors imply
a rapid arc-parallel extensional strain rate (Table 1;
Fig. 4d) for the northern Tonga forearc. However,
because the trench-normal slip vectors are arcward
(SW) of the vectors that parallel plate motion (Fig.
8b), they probably occur either within the sub-
ducted slab or in the upper plate and are not
indicative of the interplate slip direction. (If this

! ! ! !
'l
• NTF-PAC pole
(GPS) ~ s
Trench normal ~ •
i -- • •
• ...... r
(~v)......... ..................".....; ~ a.
. • .
-PAC-AUS **
dl l P
-45 -40 -35 -30 -25
(a) Latitude
174°Wt 173°W
(b)
Fig. 8. (a) Plot of slip vector azimuths (small dots),
GPS vector azimuths (large dots), the azimuth normal
to the Tonga trench, the Pacific-Australia plate
convergence direction (PAC-AUS), and the directions
predicted by a pole of rotation between the north Tonga
forearc and the Pacific based on GPS vectors (labelled
GPS) (Bevis et al. 1995) and by a pole based on slip
vectors (labelled SV). (b) Map of earthquake slip
vectors (small, black arrows) beneath the northern
Tonga forearc, the Pacific-Australia predicted direction
(large, black arrow = 84 mm a-]) and the GPS vector
(grey arrow --- 240 mm a-l).
rotation of the slip vectors resulted from a detached
sliver, these events should be trenchward, NE, of
the vectors that parallel plate motion.) Excluding
these southerly trending slip vectors results in a
near zero arc-parallel strain rate for the northern
Tonga forearc, in agreement with GPS results of
Bevis et al. (1995).
Slip vectors east of northern New Zealand reveal
a large degree of slip partitioning and v s values
suggest approximately 30 m m a -l of right-lateral
shear across the North Island, in agreement with
geodetic measurements (Fig. 4d) (31 +__9 m m a-1
from Walcott 1984). In the South Island the plate
vector becomes nearly parallel to the island and
most of the motion occurs by oblique slip near
the Alpine fault, although there is still a small
SLIP PARTITIONING AT CONVERGENT PLATE BOUNDARIES 15
i i 150 convergent component taken up by thrusting on
• •
faults parallel to the Alpine fault (Anderson et al.
1993).
130
~'*~ eee*%e •
' " Philippines
110
The Philippines are a second region that Fitch
(1972) regarded as revealing slip partitioning -
.!'""~!3.~" 90 between the Philippine trench and the left-lateral
" " " ~;
Philippine fault. Using the Seno et al. (1993)
I _ I i
PSP-Eurasia pole, the Philippine fault and trench
-20 -15 system appears to be strongly partitioned (Barrier
et al. 1991) (Fig. 4e). A difficulty in the kinematic
14°S analysis of the Philippines is the lack of a reliable
plate convergence vector because the presence of
several faults in the Philippines makes it unlikely
that the Philippine islands are rigidly attached to
150S Eurasia. A pole of rotation can be found between
the forearc and the PSP that fits the Philippine
trench slip vectors and removes a large part of the
16*S obliquity (Fig. 4e). Hence, we cannot be sure, on
the basis of slip vectors alone, that the Philippines
are partitioned.
From combined trilateration and GPS,
172*W 171°W Duquesnoy et al. (1994) estimate a left-lateral
slip rate on the Philippine Fault near l l°N of
26 + 10 m m a -1, which is consistent with the
estimates of v s (Fig. 4e) based on the Philippine
Sea-Eurasia pole of rotation (Seno et al. 1993).
Hence, this first reliable slip rate for the Philippine
fault supports the case for the Philippine system
being largely partitioned. The arc-parallel strain
rate for the Philippine forearc is poorly determined
as the v s values do not vary smoothly with distance
along the trench (Table 1; Fig. 4e).
R y u k y u trench
Slip vectors along the Ryukyu trench generally
follow the Seno et al. (1993) PSP-Eurasia pole
(Fig. 4f) and very little slip partitioning is evident
even where the obliquity is high. If there is any
arc-parallel slip of the forearc, it is of the order of
10 m m a-t or less. Near the centre of the Ryukyu
trench (near 28°N), slip vectors are about 30 ° away
from the plate vector even thought the obliquity is
near zero (Fig. 4f). In this region the Oki Daito
Ridge hits the trench (Fig. 1a) and probably causes
the local deflection of slip vectors (these two points
were excluded from estimates of strain rates).
I z u - M a r i a n a trench
Although the convergence direction is poorly
known, the Marianas forearc shows strong
partitioning in that slip Vectors remain nearly
perpendicular to the trench despite large changes in
 16 R. McCAFFREY

the obliquity (Figs 4g and 9). As observed at other

trenches, at the south end of the Marianas trench
there appears to be two groups of slip vector
directions; one nearly perpendicular to the trench
and another, comprising only 2 events, that is much
more oblique (Fig. 9). The values of v s at the 27 °N
Mariana forearc are generally equal to v , (Fig. 4g),
indicating that the upper plate absorbs" all of the
arc-parallel component of plate motion by shear
resulting in rapid arc-parallel extension (Table 1;
Fig. 4g). At the Izu arc, the slip vectors are similar
to the plate vector and no arc-parallel motion is
required. 23 °N

Conclusions
Earthquake slip vectors at convergent margins of
SE Asia are used to infer the degree of partitioning
of slip on margin-parallel and margin-normal 19°N
structures. W h e r e independent geological or
geodetic estimates are available, slip vector-derived
estimates of arc-parallel forearc slip rates appear
accurate. Regions where oblique convergence is
strongly partitioned are the Himalaya, Marianas,
New Hebrides, Philippines, Sumatra, Markham 15°N
Valley, and New Zealand. At some strongly curved
margins, earthquake slip vectors are bimodally
distributed between the convergence direction and
the trench-normal direction, suggesting that slip
partitioning occurs at small spatial scales. Several
forearcs reveal arc-parallel deformation even where ll°N
subduction is normal to the trench, showing that
subduction is a three-dimensional process.
139°E 143°E 147°E
I thank Tony Barber for many years of good work and
cheer, the organizers of the symposium for allowing me Fig. 9. Map of earthquake slip vectors (small, black
to attend, and J. Milsom and D. Snyder for helpful arrows) beneath the Izu-Mariana forearc, and the
reviews. Supported by NSF grants EAR-9105050 and PSP-Pacific (Seno et al. 1993) predicted directions
EAR-9406917. (large, black arrows).

References
ABBOTT, L. D., SILVER, E. A. & GALEWSKY, J. 1994.
Structural evolution of a modern arc-continent
collision in Papua New Guinea. Tectonics, 13,
1007-1034.
ABERS, G. & MCCAFFREY,R. 1988. Active deformation
in the New Guinea Fold-and-Thrust Belt: seismo-
logical evidence for strike-slip faulting and
basement-involved thrusting. Journal of
Geophysical Research, 93, 13 332-13 354.
& 1994. Active arc-continent collision:
Earthquakes, gravity anomalies and fault kinematics
in the Huon-Finisterre collision zone, Papua New
Guinea. Tectonics, 13, 227-245.
ANDERSON, H., WEBB, T. & JACKSON, J. 1993. Focal
mechanisms of large earthquakes in the South Island
of New Zealand: implications for the accommo-
dation of Pacific-Australia motion. Geophysical
Journal of the Royal Astronomical Society, 115,
1032-1054.
BARRIER,E., HUCHON,P. & AURELIO,M. 1991. Philippine
Fault: A key for Philippine kinematics. Geology,
19, 32-35.
BELLtER, O., SEBRIER,M. & PRAMUMIJOYO,S. 1991. La
grande faille de Sumatra: geometrie, cinematique
et quantite de deplacement raises en evidence par
l'imagerie satellitaire. Comptes Rendus de
l'Acad~mie des Sciences de Paris, 312, 1219-1226.
BEVlS, M., TAYLOR, E W., SCHUTZ, B. E., RECY, J.,
ISACKS,B. L. ETAL. 1995. Geodetic observations of
very rapid convergence and back-arc extension at
the Tonga island arc. Nature, 374, 249-251.
CURRAY,J. R. 1989. The Sunda Arc: A model for oblique
plate convergence. In: VAN HINTE, J. E., VAN
WEERING, TJ. C. E. & FORTUXN, A. R. (eds)
Proceedings. Symposium Snellius-H Expedition,
Jakarta, Nov. 23-28, 1989, Vol. 1, Geology and
Geophysics of the Banda Arc and Adjacent Areas.
Netherlands Journal of Sea Research, 24, 131-140.
 SLIP PARTITIONING AT CONVERGENT PLATE BOUNDARIES
--, MOORE, D. G., LAWVER,L. A., EMMEL,E J., RAITT,
R.W., HENRY,M. & KIECKHEFER,R. 1979. Tectonics
of the Andaman Sea and Burma. In: WATKINS,
J. S., MONTADERT,L. & DICKERSON, P. W. (eds)
Geological and Geophysical Investigations of
Continental Margins. AAPG Memoir, 29, 189-198.
DARDJI, N., VILLEMIN, T. & RAMPNOUX, J.-E 1991.
Cenozoic fault systems and paleostress along the
Cimandiri fault zone, west Java Indonesia. In:
PRASETYO, H. & SANTOSO(eds) The Silver Jubilee:
Symposium on the dynamics of subduction and
its products. Proceedings LIPI Yogjakarta,
233-253.
DEMETS, C., GORDON, R. G., ARGUS, D. E & STEIN, S.
1994a. Effects of recent revisions to the geo-
magnetic reversal time scale on estimates of current
plate motions. Geophysical Research Letters, 21,
2191-2194.
& VOGT, E 1994b. Location of the Africa-
Australia-India triple junction and motion between
the Australian and Indian plates: results from an
aeromagnetic investigation of the Central Indian
and Carlsberg ridges. Geophysical Journal
International, 119, 893-930.
DIAMENT, M., DEPLUS, C., HARJONO, H., LARUE, M.,
LASSAL, O., DUBOtS, J., & RENARD, V. 1990.
Extension in the Sunda Strait (Indonesia): a review
of the Krakatau programme. Oceanologica Acta, 10,
31-42.
--, HARJONO, H., KARTA, K., DEPLUS, C., DAHRIN, D.
ErAL. 1992. Mentawai fault zone off Sumatra: a new
key to the geodynamics of western Indonesia.
Geology, 20, 259-262.
DUQUESNOY,TH., BARRIER,E., KASSER,M., AURELIO,M.,
GAULON, R. ET AL. 1994. Detection of creep along
the Philippine Fault: First results of geodetic
measurements on Leyte island, central Philippine.
Geophysical Research Letters, 21, 975-978.
EKSTROM, G. & ENGDAHL,E. R. 1989. Earthquake source
parameters and stress distribution in the Adak
Island region of the central Aleutian Islands,
Alaska. Journal of Geophysical Research, 94,
15 499-15 519.
FITCH, T. J. 1972. Plate convergence, transcurrent faults
and internal deformation adjacent to southeast Asia
and the western Pacific. Journal of Geophysical
Research, 77, 4432-4460.
GEIST, E. L., CHILDS, J. R. & SCHOLL, D. W. 1988. The
origin of summit basins of the Aleutian Ridge:
Implications for block rotation of an arc massif.
Tectonics, 7, 327-341.
GENRICH, J., BOCK, Y., MCCAFFREY, R., CALAIS, E.,
STEVENS, C. E T A L . 1994. Kinematics of the
eastern Indonesian island arc estimated by Global
positioning system measurements. EOS, 75, 162.
GILBERT, L., KASTENS, K., HURST, K., PARADISSIS, D.,
VEIS, G. ET AL. 1994. Strain results and tectonics
from the Aegean GPS experiment. EOS, 75, 116.
GOLDFINGER, C., KULM, L. D., YEATS, R. S., APPLEGATE,
B., MACKAY, M. E. & MOORE, G. E 1992.
Transverse structural trends along the Oregon
convergence margin: Implications for Cascadia
earthquake potential and crustal rotations. Geology,
20, 141-144.
17
HAMILTON, W. 1979. Tectonics of the Indonesian region.
USGS Professional Paper, 1078.
HARJONO, H., DIAMENT, M., DUBOIS, J. & LARUE, M.
1991. Seismicity of the Sunda Strait: Evidence for
crustal extension and volcanological implications.
Tectonics, 10, 17-30•
HEGARTY, K. A., WEISSEL, J. K. & HAYES, D. E. 1983.
Convergence at the Caroline-Pacific plate
boundary: Collision and subduction. In: HAYES,
D. E. (ed.) The Tectonic and Geologic
Evolution of Southeast Asian Seas and Islands,
Part 2. AGU Geophysical Monograph, 27, 349-
359.
HUCHON, P. & LE P~CHON, X. 1984. Sunda Strait and
central Sumatra fault. Geology, 12, 668-672.
JARRARD, R. D. 1986. Relations among subduction
parameters. Reviews of Geophysics and Space
Physics, 24, 217-284.
JOHNSTON, C. R. & BOWrN, C. O. 1981. Crustal reactions
resulting from the mid Pliocene to Recent
continent-island arc collision in the Timor region.
BMR Journal of Australian Geology & Geophysics,
6, 223-243.
KARIG, D. E., LAWRENCE,M. B., MOORE, G. E & CURRAY,
J. R. 1980. Structural framework of the fore-arc
basin, NW Sumatra. Journal of the Geological
Society, London, 137, 77-91.
KULIG, C., MCCAFFREY, R., ABERS, G. A. & LETZ, H.
1993. Shallow seismicity of arc-continent collision
near Lae, Papua New Guinea. Tectonophysics, 227,
81-93.
MALOD, J. A., KARTA, K, BESLIER, M. O. & ZEN, M. T.
1996. From normal to oblique subduction: Tectonic
relationships between Java and Sumatra. Journal
of SE Asian Earth Science, 12, 85-93.
MATSUZAKA,S., TOBITA,n . , NAKAHORI,Y., AMAGAI,J. &
SUGIMOTO, Y. 1991. Detection of Philippine Sea
plate motion by very long baseline interferometry.
Geophysical Research Letters, 18, 1417-1419.
MCCAFFREY, R. 1988. Active tectonics of the eastern
Sunda and Banda arcs. Journal of Geophysical
Research, 93, 15 163-15 182.
- -1991. Slip vectors and stretching of the Sumatran
forearc. Geology, 19, 881-884.
- 1992. Oblique plate convergence, slip vectors, and
-
forearc deformation. Journal of Geophysical
Research, 97, 8905-8915.
- -1994. Global variability in subduction thrust zone-
forearc systems. Pure and Applied Geophysics, 141,
173-224.
• & ABERS, G. A. 1991. Orogeny in arc-continent
collision: the Banda arc and western New Guinea.
Geology, 19, 563-566
--. & NABELEK, J. 1986. Seismological evidence for
shallow thrusting north of the Timor trough.
Geophysical Journal of the Royal Astronomical
Society, 85, 365-381.
MCKENZIE, D. & JACKSON, J. 1983. The relationship
between strain rates, crustal thickening, paleD-
magnetism, finite strain, and fault movements
within a deforming zone. Earth and Planetary
Science Letters, 65, 182-202.
MOLNAR,P. & LYON-CAEN,H. 1989. Fault plane solutions
of earthquakes and active tectonics of the Tibetan
18 R. McCAFFREY
Plateau and its margins. Geophysical Journal
International, 99, 123-153.
--, ENGLAND, P. & MARTINOD, J. 1993. Mantle
dynamics, uplift of the Tibetan Plateau, and the
Indian monsoon. Review of Geophysics, 31,
357-396.
MOUNT, V. S. & SUPPE, J. 1987. State of stress near
the San Andreas fault: Implications for wrench
tectonics. Geology, 15, 1143-1146.
PLATT, J. P. 1993. Mechanics of oblique convergence.
Journal of Geophysical Research, 98, 16239-
16256.
PUNTODEWO, S. S. O., MCCAFFREY, R., CALAIS, E.,
BOCK, Y., RAIS, J. ET AL. 1994. GPS measurements
of crustal deformation within the Pacific-Australia
plate boundary zone in Irian Jaya, Indonesia.
Tectonophysics, 237, 141-153.
RANKZN, B., CAm3WELL, R. K. & KARIG, D. E. 1984.
Kinematics of the Philippine Sea Plate. Tectonics, 3,
555-575.
SCHOUTEN, H., & BENES, V. 1996. Post-Miocene collision
of Australia and the Bismarck arc, western
equatorial Pacific. Earth and Planetary Science
Letters, in press.
SENO, T., STEIN, S. & GRIPe, A. 1993. A model for the
motion of the Philippine Sea plate consistent
with NUVEL-1 and geological data. Journal of
Geophysical Research, 98, 17 941-17 948.
SIEH, K., ZACHARIASEN, J., BOCK, Y., EDWARDS, L.,
TAYLOR, E & GANS, P. 1994. Active tectonics of
Sumatra. GSA Abstracts with Programs, 26, A-382.
SILVER, E. A., REED, D., MCCAEFREY, R., &
JOWOD1WIRYO, Y. 1983. Back-arc thrusting in the
eastern Sunda Arc, Indonesia: a consequence of
arc-continent collision. Journal of Geophysical
Research, 88, 7429-7448.
SNYDEr, D. B., PRASETYO, H., TJOKROSAPOETRO, S.,
BLUNDELL,D. J., BARBER,A. J., RICHARDSON,A. N.
&; MILSOM, J. 1993. A deep seismic reflection
transect across the youngest mountain range in the
world, the active convergent margin between the
Australian craton and the Banda volcanic arc near
Timor, Indonesia. EOS, 74, 444.
STEVENS, C., PUNTODEWO, S. S. O., MCCAFFREY, R.,
BOCK, Y. t~ CALAIS, E. 1995. Out of the pot
and into the frying pan: A Bird's Head's view
of tectonic escape. International Union of
Geodesy and Geophysics, XXI General Assembly,
p. A38.
TREGONING,P., BRUNNER,E K., BOCK, Y,. PUNTODEWO,S.
S. O., McCAFFREY, R. ET AL. 1994. First geodetic
measurement of convergence across the Java
Trench. Geophysical Research Letters, 21, 2135-
2138.
WALCOTT, R. I. 1984. The kinematics of the plate
boundary zone through New Zealand: a comparison
of short- and long-term deformations. Geophysical
Journal of the Royal Astronomical Society, 79,
613-633.
WEISSEL, J. K. • ANDERSON, R. N. 1978. Is there a
Caroline plate? Earth and Planetary Science
Letters, 41, 143-158.
Yu, S.-B. & CHEN, H.-Y. 1994. Global positioning
system measurements of crustal deformation
in the Taiwan arc-continent collision zone.
Terrestrial, Atmospheric and Oceanic Sciences 5(4),
477-498.
ZWICK, P. & MCCAFFREY, R. 1991. Seismic slip rate
and direction of the Great Sumatra Fault based
on earthquake fault plane solutions. EOS, 72,
201.
 The Sumatra margin: oblique subduction and lateral
displacement of the accretionary prism
JACQUES ANDRI~ MALOD 1 & BADRUL MUSTAFA KEMAL 2
1 GEMCO, URA 718 du CNRS, UPMC, Case 129, 4 pl. Jussieu, 75252 Paris, France
2 Padang University, 25117 Padang, Indonesia

Abstract: The Sumatra active margin extends more than 1600 km from northwest to southeast.
The almost north-south direction of subduction is normal in front of Java and oblique in front of
Sumatra. The transition between the two regimes of subduction occurs south of the Sunda Strait.
A structural and stratigraphical study of the forearc domain, based on data collected during a
French-Indonesian programme of cooperation in oceanography, confirms that the convergent
motion is partitioned into a convergent motion, more or less perpendicular to the trench, and a
strike-slip motion parallel to the trench. This latter motion is taken up along two major faults, the
Sumatra and the Mentawai faults. At its northern end, the Mentawai fault is attenuated and seems
to terminate within the accretionary prism. It is relayed and connected to the Sumatra fault by the
Batee fault. This pattern can be explained by a simple model with two sliver plates: the Mentawai
and Aceh sliver plates, on the top of which the forearc basin has developed. The accretionary
prism itself is moving northwestwards along the Mentawai fault. Because there is no major
evidence of extension within the Mentawai plate, we conclude that the motion along the Sumatra
fault may occur at a uniform rate south of latitude 3°N. Strike-slip motion along the Mentawai
fault may be explained by relatively better coupling between the subducting slab and the upper
plate beneath the accretionary prism compared to that beneath the forearc, perhaps because of the
large oceanic structures entering the subduction zone.

Oblique subduction induces the partitioning of the

convergent motion into two components, parallel
and perpendicular to the subduction zone (Fitch Upper
1972; Jarrard 1986; McCaffrey 1991, 1992). This
partitioning results from the coupling effect
~. plate

between the dipping oceanic slab and the upper

plate lithosphere. This coupling effect can be
identified by study of slip vectors at the subduction ~. plate k
zone which indicate the true direction of relative
movement of the oceanic slab with respect to the
forearc (Jarrard 1986; McCaffrey 1992): In the case Oceanic
plate
where the margin is curved, partitioning increases
with obliquity (Fig. 1). If the partitioning is not
complete the slip vector will be situated between Fig. 1. Partitioning of movements during oblique
the convergence vector and the direction perpen- subduction increases with margin curvature. Full
dicular to the margin. The lateral movement results partitioning is shown for sake of clarity.
in the formation of a sliver microplate while the
movement perpendicular to the margin, mainly
taken up by subduction, is also expressed by com-
pressional deformation within the accretionary Structures related to the subduction
prism or the forearc.
off Sumatra
Marine geophysics cruises, in cooperation with
Indonesia, allowed a detailed study of the Sumatra The subducting oceanic lithosphere of the Wharton
margin to be undertaken (Diament et al. 1992; basin has a variable age and structure (Fig. 2). Its
Malod et al. 1993; Mustafa Kemal 1993; Zen Jr. age ranges from more than 110 Ma in the eastern
1993). The object of this paper is to propose a part of the area, along the Java trench, to approxi-
model for the distribution of the movements in the mately 50 Ma south of central Sumatra where the
whole forearc area. Wharton extinct spreading ridge is being subducted

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, 19

Geological Society Special Publication No. 106, pp. 19-28.
20 J . A . M A L O D & B . M. KEMAL

1 0 ~ * E 1 llO*E
break

,amai:se; : i i i
I • I:E~ensi°n460 kin-
;umatra
i . . . . . . . . . . . . . . . .

;.-C)
Sunda
Oblique I
Subduction

:i'i!i:
~:~~~......... Subduction slab
IO°S :~";i;iii~:;;i: : 400-450 km
...; .

I Subduction I

/iiiiiiiiiiiiiill
i:i\::iiiiiiii:iiii

mccumu'a'n 1 a-'ill
~liiiiiiiiiiiiiii
--o o 31;iiiiii;ii;i:ii
. . . . . . . . . . . . . . . .
.......... 14 . . . . . . . . . . . . . . . . .

Wharton
extinct
_ spreading
centre
. . . . . . . . ~ 7 " "
. . . . . . . . . . . . .

' " " iE~'.eO~d.nl . . . . . . . .

I ans'e, I . . 4Fi
....
) ...........
. . . . .
il .......
-- 5os i iiliSundaStrait
~ Extension
constant ~,~ 5 0 - 7 0 km
obliquity

-
i + Ablation
I Zoneo
increasing -
~5°E 100"E obliquity ~?5"E
I I [ [ I I _ t l I [ t I I

Fig. 2. The geodynamic framework of Sumatra subduction. Note the Wharton extinct spreading axis and the
Investigator fracture zone entering the subduction. The Sumatra and Mentawai faults are the tectonic response to the
obliquity of subduction. A schematic distribution of the movements in the forearc is shown. Thick arrows indicate
motions with respect to Sumatra mainland. Thin arrows give the relative motion along strike-slip faults. As pointed
out in the text, the numerical values are realistic but only indicative, based on published studies. The inset shows the
slab geometry indicated by earthquake depths. The stippled areas are high velocity zones reflecting a subducted slab
at depths of 400--450 km according to the seismic tomography of Puspito et al. (1993).
 THE SUMATRA MARGIN
with a velocity of approximately 60-70 mm a-1
(DeMets et al. 1990). The geometry of the
dipping oceanic slab can be deduced from the
distribution of earthquakes linked to subduction
(Fig. 2). The slab curvature follows the island arc of
Java and Sumatra and shows some minor rupture
zones to the south of Sumatra (Zen Jr. 1993;
R. Louat, pers. comm. 1994). Further north in the
region of Lake Toba, a major rupture occurs, which
is also indicated by seismic tomography (Puspito
et al. 1993). This feature corresponds to a marked
change in orientation of the continental margin and
the subduction trench which follow the shape of the
slab and turn to a westward orientation before
resuming a NNW direction.
The direction of convergence indicated by
global kinematic models is 024 ° to 025 ° (DeMets
et al. 1990). In fact, the study of slip vectors shows
that the subduction is normal beneath Java, where
the average of 15 vectors indicates a 003 ° direction
(McCaffrey 1991). The Sunda arc therefore allows
observation of the change from frontal subduction
to oblique subduction (Fig. 2). Subduction becomes
increasingly oblique to the margin as the trench
turns from E-W to NW-SE west of Java. One can
therefore expect that partitioning of the movement
will result in an increasing lateral component of
motion. The obliquity of subduction is constant
along the main linear portion of the Sumatran
margin. The lateral component of the movement is
taken up here by the great dextral fault of Sumatra,
which extends from the south to the north of
Sumatra (Katili 1970).
McCaffrey (1992) showed that, correlated with
the curvature of the margin, slip vectors indicate
that the lateral movement increases gradually along
the Sumatra fault. Consequently, he predicted a
longitudinal extension of the forearc sliver plate
between the trench and the Sumatra fault. This
model may partly explain the discrepancy between
extension at the two ends of the Sumatra fault, from
more than 460 km in the Andaman Sea (Curray
et al. 1978) compared to only 50-70 km in the
Sunda Strait (Huchon & Le Pichon 1984; Lassal
et al. 1989; Diament et al. 1992).
Boundaries of forearc microplates
The first result of the study of single channel
seismic profiles was the discovery that the faulted
zone situated between the accretionary prism and
the forearc, previously interpreted as a flexure or
backthrust (Karig et al. 1978) is linear, has a
highly variable tectonic aspect and looks like a
flower structure in some places. Therefore, this
structure was re-interpreted as the trace of a strike-
slip fault zone, the Mentawai fault, which has also
21
accommodated some compression (Diament et al.
1992; Zen Jr. 1993). The Mentawai fault seems
to have been active during the Neogene, but it
is not possible to identify the relative importance
of strike-slip and compressional movements.
Furthermore, the amount of lateral motion cannot
be measured from seismic profiles.
The Mentawai microplate
The Mentawai fault (Fig. 3) extends from the south
of Sumatra up to the south of Nias. Therefore the
Sumatra fault and the Mentawai fault bound an
elongated sliver microplate, the Mentawai plate,
which corresponds to the forearc domain. Evidence
for extensional deformation within this plate is
weak since no major normal faults are observed to
cross the Mentawai plate. There is no significant
indication of the extension predicted in previous
models (Jarrard 1986; McCaffrey 1992) or that
necessary to explain the difference in the amount of
extension between the north and south of Sumatra.
There remains the possibility of distributed exten-
sion by means of small extensional faults.
However, recent faults with very small throws are
found only in a 50 km long zone southwest of
Sumatra where recent seismicity is concentrated
(Zen Jr. 1993; Fig. 4).
To the north of Nias island, the NNW-SSE
striking Batee fault connects the Mentawai fault
and the Sumatra fault (Beaudry & Moore 1985;
Matson & Moore 1992). On land where it seems
not to have been very active in recent time, its
dextral strike-slip movement is documented from
Spot images (C. Detourbet, pers. comm. 1994). At
sea, it appears as a vertical fault with a large normal
throw explaining the Nias basin formation (Fig. 5).
The transfer of the major part of the movement on
the Sumatra fault may in fact occur by means of
several faults in the Nias basin (Matson & Moore
1992) or on its southern ~rim (archipelago of Pini
island; Fig. 3). Further north, movement would
therefore be concentrated on the Sumatra fault, as
suggested by the opening of the Andaman Sea
north of Sumatra (Fig. 2). Accordingly, to the north
of Nias island, the forearc area moves significantly
with respect to the main part of Sumatra.
The Sumatra fault forms the northern boundary
of a pull-apart in the Sunda Strait (Harjono et al.
1988, 1991, 1993; Lassal et al. 1989; Diament et al.
1990). The dextral strike-slip movement is relayed
to the south by the Ujung Kulon fracture zone
which extends into the accretionary prism (Fig. 6).
To the south of the Sunda Strait the Mentawai fault
becomes difficult to follow in the accretionary
prism. A progressive initiation of the Mentawai
fault could explain why it is no longer strongly
22 J . A . MALOD • B. M. KEMAL

Sumatra / I I
•

J~
i
I

Et
I

Indo-Australial
plate

Aceh basin :" :: \

\,,,

ION
SUMATRA
i ¸¸ ! "~ : ~!i!~i!:
Simeulu~

~ Fault
~iiiiiiiiiii,iiiiiiiiiiiiiii~iiiii!iiiiiiiiiiiiiii
lON Thrusts

I 0ON ~7
~1 Strikeslip

Subductiontrench

",,,,,,,./ Structures of the

oceanic crust
Wharton
fossil I
oceanic |,,.,jr
!

94°E 95OE 96OE ridge | 97°E \ 98°E

Fig. 3. Structural sketch of the northern margin of Sumatra, with the delineation of the main sliver plates.

evident in the structures to the south of the Sunda
Strait.
The Mentawai microplate is greater than
1000 km in length (Fig. 2). Its existence shows that
the accretionary prism has moved relative to the
forearc domain. The southern limit of the Mentawai
plate is coincident with the zone where subduction
changes from oblique along Sumatra to orthogonal
along Java. The Mentawai plate ends to the north of
Nias, where the subducting slab is interrupted by
a large discontinuity.
The Aceh microplate
North of Nias, the Sumatra fault, the northern
extension of the Mentawai fault, and the southern
extension of the West Andaman fault (Curray et al.
1978; Izart et al. 1994), bound another sliver plate,
the Aceh microplate (Fig. 3). The Mentawai fault
may have played a role in the past, but its actual
prolongation does not currently show evidence of
active strike-slip movement. Backthrusting of the
accretionary prism over the forearc basin, along the
 THE SUMATRA MARGIN
NW
3.0
3.5
4.0
4.5
s.d.t.t
t 5 km
Fig. 4. Recent vertical normal faults associated with active seismicity southeast of Sumatra (location on Fig. 2).
southern continuation of the West Andaman fault,
is the probable effect of flexural bending of the
forearc basement under the accretionary prism,
which would explain the deepening of the forearc
basin toward the west (Malod et al. 1993) (Fig. 7).
In the longitudinal direction, the forearc is made up
of large blocks tilted southwards. Some part of the
movement is absorbed by internal deformation of
the Aceh microplate as indicated by the Banyak
Island thrust zone, active in latest Pliocene (Fig. 7)
or the presently active Tuba anticline ridge.
Transfers in the accretionary prism
along Sumatra
The definition of Mentawai and Aceh plates leads
to a consideration of the accretionary prism
between the trench and the Mentawai fault as a kind
W
-o.s~~atee fault
I 5 krn I
Fig. 5. The Batee fault. Note the large normal throw along this vertical fault (location on Fig. 4).
23
SE
of sliver plate. However, deformation within the
prism is difficult to study. Northwest of Nias
island, several NW-SE fault zones, which are
ramifications of the Mentawai fault, cross the
accretionary prism (Fig. 7). The tectonic pattern of
these faults suggests strike-slip motion (although
mud diapirism might be possible). To the north of
this zone, the accretionary prism may be a part of
the Andaman Island and Burma plate. Where the
Mentawai plate is present, south of Nias Island, we
recognize three different zones from south to north
(Fig. 2), as follows.
1. Extension and material ablation in the south
The displacement towards the northwest of the
Mentawai microplate as a single block is related
to the extension in the Sunda Strait and perhaps
further south within the forearc itself. At the same
time, oceanic sediments accreted to the deeper part
E
24 J.A. MALOD • B. M. KEMAL
Fig. 6. Simplified structural sketch of the southern
prolongation of the Mentawai and Sumatra faults. The
Ujung Kulon fault zone relays the Sumatra fault system
to the south of the Sunda Strait pull-apart. The
Mentawai fault zone prolongation indicated by a dashed
line is not observed in this region.
of the margin must have been displaced towards the
northwest. These two mechanisms, of extension in
the basement and removal of sediments from the
accretionary prism, contribute to the concave shape
of the margin south of the Sunda Strait.
2. Transfer in the central part In the linear part
of the margin west of Sumatra, the Mentawai plate
and the accretionary prism are displaced towards
the northwest. There is relative movement between
the plate and the prism along the Mentawai fault.
The amount of relative motion and its timing
during the Neogene cannot be established.
3. Accumulation and relay zone in the north In
the northern part of the region, according to our
interpretation, movement was concentrated mainly
on the Sumatra fault, while materials of the prism
accumulated in several slices west of Simeulue
island. This process occurred as if the subduction
was locally frontal, in agreement with the local
E-W orientation of the oceanic slab, and is con-
sistent with the horsetail shape of the Mentawai
fault in this region and its reduced role further
north.
This description of the Sumatra oblique subduction
and transfer process leaves some unanswered
questions, such as possible deformation within the
accretionary prism or the Sumatra mainland to the
northeast of the Sumatra fault, and mechanisms of
movement partitioning in the prism and the forearc.
However, the general tectonic disposition of the
Sumatra margin can be integrated into a simple
model, which actually involves the transfer of
material at the edge of Sumatra.
Model
The boundaries, the distribution and motion of the
plates and microplates are illustrated on a simpli-
fied structural diagram (Fig. 2).
Sumatra fault and Mentawai fault
These two faults have different natures (Fig. 8).
The Sumatra fault is an intra-continental strike-slip
fault and its geometry can be followed on land and
at sea (Fig. 6). It coincides with the volcanic arc
which suggests a connection between volcanism
and tectonics (Bellier & Srbrier 1994), the volcanic
zone being a zone of weakness. The Mentawai fault
appears more like a boundary between the prism
constituted of accreted oceanic sediments and the
forearc basin established on a crust of intermediate
or continental type as shown by refraction
(Kieckhefer et al. 1980) or expanding seismic
profiles (Liu & Curray 1988; Fig. 8). This fault is
not necessarily vertical and the compressional
component may explain its complicated surface
pattern. This fault separates the relatively rigid
area of the Mentawai microplate, as shown by
the lack of major deformation, from the complex
accretionary prism, possibly more deformable
because of the existence of numerous thrust slices.
The geometry o f the margin
Down to a depth o f 20 or 25 km, the subduction
contact occurs between the oceanic slab and the
accretionary prism (Fig. 8). Coupling along this
contact, assuming an active Mentawai strike-slip
fault, induces partitioning of the movement that
drags the prism towards the northwest. At greater
depths beneath the margin, the coupling occurs
between the oceanic slab and the lithosphere of the
forearc. Consequently, the partitioning of the move-
ment induces two decoupling zones: the Mentawai
fault and the Sumatra fault.
In a horizontal plane, the relative movements
depend on the importance of the partitioning of
the movement in each of these two zones. The
Mentawai fault will have dextral strike-slip move-
ment if coupling is more efficient under the
accretionary prism than under the forearc basin
(Fig. 9); this seems to happen along the rectilinear
part of the fault and the Mentawai plate. If the
 THE SUMATRA MARGIN 25

3.or =W-_~_AAccretionary
~ ~~j~rism Aceh basin E
3.5 I

4.0 I

4.5 I

5"sOd.t.t.

N Banyak
thrust ~ i
i
'_2._ . . . . . . . . . . . . , ,i
2.0

. . . . . . ~ ~ N
~Late Miocene "~''--.'-' "- ~ ~ . ~'.-
~ - - 2.5

:.7"-.~." ..:~/- - 7 : " , - , ;.:- , :;..:.;:2,~. -'-.~-~ ..'.;ZI-... - ., ..;.~ J

SW Middle-Late M i o c e n e " : ...... " " " ' " ': " =" " s.d.t.t. _ 3.0

2.5 ~ ;.~

3.0

3.5 --
s.d.t.t. • 5 krn I

Fig. 7. Seismic reflection profiles showing deformations around the Aceh microplate (location on Fig. 4):
(a) backthrust of the accretionary prism over the Aceh basin; (h) Late Pliocene northward thrust of Banyak Island;
(c) possible strike-slip faults in the accretionary prism forming part of a horsetail termination of the Mentawai fault.

partitioning of movement is the same in both poor expression or even the apparent absence of the
regions there will be no relative movement between Mentawai fault in this region.
prism and forearc (Fig. 9). This may be the case to The occurrence of the Mentawai fault and its
the south of the Sunda Strait and could explain the continuation into the Batee fault indicate that the
26 J.A. MALOD Be; B. M. KEMAL

Mentawai fault Sumatra fault

0
Accretionaryprism
~ ~
l F°rearc coasl
lOaSl(I
i,~ VolcancJarc
~...... ......................
Sumatra

50 ......... ......... :

Fig. 8. Schematic cross-section, without vertical exaggeration, of the Sumatra subduction system.

first case (Fig. 9) is most likely over most pans of
the Sumatra forearc. The coupling between the sub-
ducting slab and the overriding forearc seems to be
more efficient in the prism area than in the forearc
basin. This may be the result of several processes
including:
(1) the accretionary prism may load the down-
going plate and increase the coupling effect;
(2) the deeper part of the accretionary prism may
have become better consolidated when sub-
ruction was more frontal in the Eocene-
Oligocene time;
(3) the large structures of the oceanic sea floor
entering the subduction zone, like the
Investigator Ridge and the Wharton extinct
spreading ridge, may induce a strong coupling
between the slab and the accretionary prism.
In any case, since the accretionary prism rests
directly on the oceanic slab, its lateral transfer is
likely to occur provided there is decoupling along
the Mentawai fault.
Movement distribution
A quantification of motions is possible only for the
most recent period of time during which measure-

Sumatra Sumatra
fault~
Mentaf : i l t ~
fault~ ~L ix

T.rench ~ , , % ~ ~ T~nc. \ X xv \
"X.\ , Fore c

~ Prism ~ Prism

Obliquedirection Obliquedirection
of subduction of subducUon

Q Dextralstrike-slip Q No strike-slip
Fig. 9. Sketch showing two possibilities of motion distribution by decoupling: (a) the situation in central and north
Sumatra with the forearc and prism moving at different rates; (b) the situation south of the Sunda Strait, where it is
proposed that the forearc and prism are moving at the same rate.
 THE SUMATRA MARGIN
ments have been made. Measurements of the recent
displacement on the Sumatra fault have been made
essentially by means of satellite photographs
(Bellier et al. 1991, 1993; Bellier & S6brier 1994).
These measurements indicate a velocity of dis-
placement that grows from 6 _ 4 m m a-t in the
south close to the Sunda Strait (Bellier et al. 1991 ),
to 19 _+ 2 m m a -1 in the Singkarak region at 1°S
(Bellier et al. 1993), and to about 23 mm a -i in the
Lake Toba region at 2°30'N (Detourbet et al. 1993).
The value of 6 _ 4 m m a -1 in the region close to the
Sunda Strait is low and might be explained by
recent extensional deformation occurring in the
forearc (Fig. 4). Along the Batee fault a value of
11_ 2 mm a-l is proposed (C. Detourbet, pets.
comm. 1994).
Figure 9 shows a possible distribution of the
movements that would explain the observed
structures and rates of Fig. 2. In order to give some
numerical values 20 mm a -1 is chosen for the main
Sumatra fault segment. Accordingly movement on
the northern part of the Sumatra fault is fixed to the
sum of this value plus the strike-slip component
on the Batee fault, which is 11 m m a -t, yielding
31 m m a-1. These numerical values serve mainly to
illustrate the model, and should not be regarded as
absolute values. In central and north Sumatra
movement is distributed between the Mentawai and
Sumatra faults with the forearc and prism moving
at different rates (Fig. 9a). In the region south of the
Sunda Strait we propose that the forearc and prism
are moving at the same rate (Fig. 9b). At the the
northern end of the Sunda forearc the movement is
absorbed in three ways: opening of the Andaman
References
BEAUDRY,D. & MOORE, G. E 1985. Seismic stratigraphy
and Cenozoic evolution of west~Sumatra forearc
basin. AAPG Bulletin, 69, 742-759.
BELLmR, O. & SE;BmER,M. 1994. Relationship between
tectonism and volcanism along the Great Sumatran
Fault Zone deduced by SPOT image analyses.
Tectonophysics, 233, 215-231.
, - - , D~TOURBET, C. & PRAMUMIJOYO,S. 1993.
Long term dextral slip rate along the Sumatran
fault system. Terra Abstracts', 1, 253, EUG VII,
4-8 April Strasbourg.
, & PRAMUMnOYO,S. 1991. La grande faille
de Sumatra: gEomEtrie, cinEmatique et quantitE
de dEplacement mises en Evidence par l'imagerie
satellitaire. Comptes Rendus de l'Acadgrnie des
Sciences de Paris, 312, Sgrie II, 1219-1226.
CURRAY, J. R., MOORE, D. G., LAWVER, L. A., EMMEL,
E J., RAITT, R. W., HENRY,M. & KIECKHEFFER,R.
1978. Tectonics of the Andaman Sea and Burma.
In: WATKINS,J. S., MONTADERt,L. & DICKENSON,
P. W. (eds) Geological and Geophysical
Investigations of Continental Margins. AAPG
Memoir, 29, 189-198.
27
Sea, westward migration of the Andaman arc that
overlaps the Indian ocean oceanic crust, and finally
in backthrusting of the accretionary prism over the
forearc basin belonging to the Aceh microplate.
Conclusions
The authors observations suggest that the degree of
partitioning of movement may be different beneath
the accretionary prism and the forearc domain. In
the most representative portion of the Sumatra
margin with respect to oblique subduction,
partitioning is more efficient in the accretionary
prism than in the forearc region. This may result
from the importance of the oceanic structures
entering the subduction zone beneath Sumatra. The
Mentawai microplate movements are now well
understood and partly quantified for recent periods,
but those of the accretionary prism necessitate
further studies, particularly to look at possible
internal deformation. Quantification of the move-
ments also needs further investigation.
We thank all the people who participated in the data
collection, specially the Commandant Soepangat and his
crew on the Indonesian R.V. Baruna Jaya IlL This
research was conducted during an Indonesian-French
cooperative programme in oceanography. We acknowl-
edge the help of Mr. M. Pain, attach6 scientifique at the
French Embassy in Jakarta. This study has been funded by
IFREMER, the French Embassy in Jakarta, CNRS-INSU
(grant ATP 933909) and UniversitE Pierre et Marie Curie
in France and BPPT in Indonesia. S. J. Matthews and
M. A. Samuel reviewed the paper and made many
valuable comments.
DEMETS, C., GORDON, R. G., ARGUS, D. E & STEk~, C.
1990. Current plate motions. Geophysical Journal
International, 10L 425-478.
DETOUm3ET, C., BELL~r, O. & SEBmER, M. 1993. La
caldeira volcanic de Toba et le systbme de faille
de Sumatra (IndonEsie) vues par SPOT. Comptes
Rendus de l'Acad(mie des Sciences de Paris, 316,
1439-1445.
DIAMENT, M., DEPLUS, C., HARJONO, H., LARUE, M.,
LASSAL, O., DUBOIS, J. & RENARD, V. 1990.
Extension in the Sunda Strait (Indonesia): Review
of the Krakatau programme. Oceanologica Acta, 10,
31-41.
~, HARJONO,H., KARTA,K., DEPLUS,C., DAHRIN,D.,
ZEN, M. T. J. & MALOD,J. A. 1992. Mentawai fault
zone off Sumatra: A new key to the geodynamics of
western Indonesia. Geology, 20, 259-262.
FITCH, T. J. 1972. Plate convergence, transcurrent faults,
and internal deformation adjacent to Southeast Asia
and the Western Pacific. Journal of Geophysical
Research, 77, 4432.4460.
HARJONO, H., DIAMENT, M., DUBOIS, J. & LARUE, M.
1988. SEisrnicitE du detroit de la Sonde (IndonEsie):
28 J . A . MALOD t~Z B. M. KEMAL
presentation des rEsultats d'un rdseau local.
Comptes Rendus de l'Acaddmie des Sciences de
Paris, 307, SErie II, 565-571.
--, --, , -- & ZEN, M. T. 1991. Seismicity of
the Sunda Strait: evidence for crustal extension and
volcanological implications. Tectonics, 10, 17-30.
, SI~BRIER, M. & DIAMENT, M. 1993. Correction
and addition to "Seismicity of the Sunda Strait:
evidence for crustal extension and volcanological
implications" by Hery Harjono, Michel Diament,
Jacques Dubois, Michel Larue and Mudaham
Taufick Zen. Tectonics, 12, 787-790.
HUCHOn, E & LE PICHON, X. 1984. Sunda Strait and
Central Sumatra Fault. Geology, 12, 668-672.
IZART, A., MUSTAFA KEMAL, B. & MALOD, J. A. 1994.
Seismic stratigraphy and subsidence evolution of
the northwest Sumatra forearc basin. Marine
Geology, 122, 109-124.
JARRARD, R. D. 1986. Terrane motion by strike-slip
faulting of forearc slivers. Geology, 14, 780-783.
KARIG, D. E., SUPARKA,S., MOORE, G. E & HEHANUSSA,
E E. 1978. Structure and Cenozoic evolution of the
Sunda arc in the central Sumatra region In: Watkins,
J. S., MONTADERT, L. & DICKENSON, P. W. (eds)
Geological and Geophysical Investigations of
Continental Margins. AAPG Memoir, 29, 223-237.
KATILI, J. A. 1970. Large transcurrent faults in southeast
Asia with special reference to Indonesia.
Geologische Rundschau, 59, 581-600.
KIECKHEFER, R. M., SHOR JR., G. G., CURRAY, J. R.,
SUGIARTA, W. & HEHUWAT, F. 1980. Seismic
refraction studies of the Sunda trench and forearc
basin. Journal of Geophysical Research, 85,
863-889.
LASSAL, O., HUCHON,P. & HARJONO,H. 1989. Extension
crustale dans le detroit de la Sonde (IndonEsie), don-
nEes de la sismique rEflexion (Campagne
KRAKATAU). Comptes Rendus de l'Acaddmie des
Sciences de Paris, 309, SErie II, 205-212.
LIU, C. S. & CURRAV, J. R. 1988. Structure and nature
of the forearc basin of west central Sumatra. EOS,
69, 1443.
MALOD, J. A., MUSTAFA KEMAL, B., BESLIER, M. O.,
DEPLUS, C., DIAMENT,M. ETAL. 1993. DEformations
du bassin d'avant-arc au nord-ouest de Sumatra.
Une rEponse ~ la subduction oblique. Comptes
Rendus de l'Acaddmie des Sciences de Paris, 316,
SErie II, 791-797.
MATSON, R. G. & MOORE, G. F. 1992. Structural
influences on Neogene subsidence in the central
Sumatra forearc basin. In: WATKINS,J. S., ZHIQIANG,
F. ET AL. (eds) Geology and Geophysics of
Continental Margins. AAPG Memoir, 53, 157-181.
MCCAFFREY, R. 1991. Slip vectors and stretching of the
Sumatran forearc. Geology, 19, 881-884.
1992. Oblique plate convergence, slip vectors, and
forearc deformation. Journal of Geophysical
Research, 97, 8905-8915.
MUSTAFA KEMAL, B. 1993. La marge active au Nord-
Ouest de Sumatra. Mdcanismes gdodynamiques de
tran~fert lids glla subduction oblique. Th~se de
l'UniversitE E et M. Curie. Paris.
PUSPITO, N. T., YAMANAKA,Y., MIYATAKE,Z., SHIMAZAKI,
K. & HIRAHARA, K. 1993. Three-dimensional
P-wave velocity structure beneath the Indonesian
region. Tectonophysics, 220, 175-192.
ZEN JR., M. T. 1993. DEformation de l'avant-arc en
rdponse ?tune subduction ?1 convergence oblique.
Exemple de Sumatra. Th~se de l'UniversitE P. et
M. Curie, Paris.
 Collision and strike-slip faulting in the northern Molucca Sea
(Philippines and Indonesia): preliminary results
of a morphotectonic study
CLAUDE R A N G I N 1, D A H A R T A D A H R I N 2, R A Y Q U E B R A L 3
& THE MODEC SCIENTIFIC PARTY

ANNE GAELLE BADER, JEAN PAUL CADET, GLEN CAGLARCAN, BENOIT DEFFONTAINES,
CHRISTINE DEPLUS, ERNESTO G. CORPUS, ROBERT HALL, YANN HELLO, JACQUES MALOD,
SERGE LALLEMAND, DOMINGO B. LAYUGAN, RI~MY LOUAT, REYNALDO MORALES,
KEITH PANKOW, MANUEL PUBELLIER, MICHEL POPOFF, REYNALDO T. RODELAS
& TRAMANADI YUDHO
1 URA 1759 CNRS, Ddpartement de Gdotectonique, T 26-0 El, Universitd Pierre et
Marie Curie, 4 Place Jussieu, 75252 Paris, France
e ITB Bandung, Indonesia
3 Mines and Geosciences Bureau, Manila, Philippines

Abstract: A swath mapping, gravity and single channel seismic survey was carried out in the
northern Molucca Sea with R.V. L'Atalante. Preliminary interpretation of these data reveals the
presence of an almost complete Sangihe arc and forearc. The Miangas-Pujada-Talaud ridge in
the central part of the Molucca Sea appears to be a backstop within the Sangihe forearc. East
of the ridge very contrasting terranes are separated by a major NW-SE crustal discontinuity
interpreted as a left-lateral strike-slip fault. North of 6°N the Philippine Trench inner wall is
dissected by NW-SE trending left-lateral strike-slip faults, resulting from the dominantly oblique
convergence between colliding arcs. South of 6°N the westward subduction of the buoyant
Snellius volcanic plateau, a fragment of the Halmahera arc terrane, has induced the formation of
a new plate boundary, the Philippine Trench along what is interpreted as a former strike-slip fault
zone. East of the Miangas-Pujada-Talaud ridge there is a wide sedimentary wedge separated
from the Snellius Ridge to the south by the major NW-SE crustal discontinuity. The thickness of
this wedge cannot be explained by subduction along the very young Philippine Trench. It could
be either an accretionary wedge developed at the deformed leading edge of the Sangihe forearc
or be part of a former intra-arc basin which was part of the colliding Halmahera arc terrane. There
is no clear evidence for accreted oceanic crust belonging to the recently subducted Molucca Sea.
The Miangas-Pujada-Talaud ridge is part of the Sangihe forearc, and the ophiolites could
represent its basement, uplifted along the outer arc ridge. The dog-leg-shaped Philippine Trench
is propagating southward across the fragmented Halmahera arc terrane and its southern segment
could reactivate a former strike-slip fault zone.

The collision of the Hahnahera arc (a fragment
of the Philippine Sea plate) with the Sangihe arc
(Fig. 1) has been described by Silver & Moore
(1978), Hamilton (1979), and Moore & Silver
(1983). The ophiolites and melanges exposed
in Talaud, Mayu and Tifore Islands have been
interpreted as fragments of the almost completely
subducted Molucca Sea floor formerly separating
the two colliding island arcs (Silver & Moore
1978; McCaffrey et al. 1980). The creation of
a new plate boundary, the Philippine Trench,
east of the collision zone therefore marks a step in
the incorporation of the intra-oceanic Halmahera
island arc into the Eurasian continental margin,
if the Sangihe arc is considered to be built on
the thinned margin o f Eurasia (Rangin et al.
1990a).
In this paper the geometry of this collision zone
is documented by a study of the g e o m o r p h o l o g y of
the northern Molucca Sea, based on large coverage
m u l t i b e a m mapping, with gravity and single
channel seismic data. This dataset was recorded
during a cruise of R.V. L'Atalante in April and
May 1994 (MODEC cruise), in co-operation with
the I n d o n e s i a n Institute of D e v e l o p m e n t and
Technological Research, Jakarta (BPPT) and the
Philippines Bureau of Mines and Geosciences,
Manila (BMG).

From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, 29
Geological Society Special Publication No. 106, pp. 29-46.
 -p+~.qutu.ts qzutu -smeiu 3!a+qusoql!l U.teLU +ql S~aoqs I+SU! +q& "~9S eoonlOlN oq+ j o ~laO~a+tueaj 3!tUeUApO+D "I "~!~I
+,wvu ii ~,l un elml. eros P ' / / / / A
ouoz UOllonPlln S V '"~/ ="''' oAn;~Y 'u!dd.i|!qd V / / / / / ~
" o~!i'IOU "q,,,,
:)!]ouOow~
uo#low Oluld - - eleld
~ImmW~
)UOJl uo#s#llOO ~ o;nlosqv Ue!lUJlSni ue.lsLln 3
d!Is-Oq!JtS maonlow
~ ~
•

~ ' "
•
+:..,,---,
Nisvb , ° .
• ~'.

~
.~'o~
:oTvJ.N"
":,

"
•

.
. •

•
•
", . ' . . ' . s
•

•
• •
•

.
:
.
"

.
•

• ~ . +/:
•

.... ~+..~,+.
....,+.~-.+.
• " A
~. " . . . . . r a : m t q Ills
i~O L
• o .
~.i-...' .;..
.... ,.'......'.'........:~.:,..:::...
I~401
" " ~ • qr •
.": ~. . i . . . .: ~. . ' "~.;~."..: ". )
" •
• . ~ . - ~:'~,~,8"" "..'..
....-.
,<. ~ ' '
m."X\+"'',,~-~"
• i ......... :.......~..-...:~',~..
,g~(~)m..
..
• •
" "
I % 1 0 ~"
]]....
• . " • • •
• '~.+~/.+,,+,....
o" ",,~ ~(+)si'
~ • . .
"
.
"
. N.O
I~ Qd~, O;eld oU!Oed
='-~2
• e 0 • •
o 5(.,I, oL;~L o ~ I.I.
 COLLISION AND FAULTING IN THE N MOLUCCA SEA 31

The preliminary interpretations of these data
and attempts to answer some basic questions are
presented here. How does the collision zone
between the Molucca Sea basin margins change
from frontal to oblique collision, after oceanic crust
of the Molucca Sea has been entirely subducted
between the two colliding arcs? What is the
morphotectonic signature of a newly created plate
boundary, the Philippine Trench?
Gravity and seismological data from the
Molucca Sea were published by Silver & Moore
(1978) and McCaffrey et al. (1980) but the only
detailed bathymetric map available before the
cruise was that of Krause (1966). Seabed mapping
should allow a better understanding of the
geometry of this diachronous collision and the
rapid structural variations expected in the transition
zone between collision and oblique subduction.
Here, the focus is on the interpretation of newly
acquired seabed surface data (bathymetry and
reflectivity, 3.5 kHz echo sounder and single
channel seismic reflection with a maximum pene-
tration of 2 s twt). Gravity, magnetic, seismic
reflection and swath mapping surveys were con-
ducted to map geological structures and the thick-
ness of sediments between the Halmahera and
Sangihe arcs and the Philippine and Cotabato
Trenches. The position, speed and direction of the
ship were given by the GPS (Global Positioning
System) navigation system. The preliminary
interpretation of the structures in the survey area
is based on the morphological and geophysical data
gathered on board in real time. A total of 3725
nautical miles were surveyed during the cruise
providing a detailed bathymetric map (Fig. 2).
L'Atalante is equipped with a SIMRAD EM-12
dual multibeam echo sounder which provides
bathymetry and reflectivity imagery of the sea
floor. Only the bathymetric data are discussed here.
The MODEC cruise also involved continuous
gravity and magnetic measurements and six
channel seismic recording. The few seismic data
presented here are single channel records acquired
on board and are not yet reprocessed.

123 ° E ~23"3o 124°E 124°3(~ 125°E msO3OE 126"E ms*3OE 127°E 127o30E 128°E 128°3<~ 129"E
r r i I i l +

7oh 7*N

+°" - 6ON

+'+'_- _- 5"30N

5°N

,++,,,., -
]-+/ t
[
1 +.-', +
\: - 4o30N

4*N

3°3ON

+'• " Morotai

i 3°N

~3ON 2=30N
-- ~ Tagulandang _ ~ --
_ Halmahera~'i~

2°N 2ON
123°E 123°30E 124°E 124°~E 125°E 12S°30E 126°E 1~°3OE 1 27°E +27*3oE 128°E t2B*~E 1290E

Fig. 2. MODEC cruise tracks.

32 C. RANGIN ET AL.
Geodynamic framework
The Molucca Sea is located at the junction of three
convergent major lithospheric plates (Fig. 1). The
Eurasian plate is bounded by active volcanic arcs
(west Mindanao, Sangihe, and north Sulawesi) and
includes the marginal basins of the Celebes Sea and
South China Sea which opened in Palaeogene
times. The northern margin of the Australian plate
extends from New Guinea in the east to central
Sulawesi in the west. The Philippine Sea plate has
a western part formed largely of Palaeogene and
older oceanic crust (the West Philippine Basin)
which probably formed in a backarc environment.
The associated arc currently comprises a major part
of the Philippine archipelago, now orientated
approximately north-south between Luzon and
Mindanao, and extending southwards into the
Halmahera arc.
Kinematically, the Australian plate is moving
northward with respect to Eurasia at a rate of 7-
8 cm a-1. The Philippine Sea plate is moving west-
wards with respect to Eurasia about a pole of
rotation located at approximately 48°N and 157°E
(Seno et al. 1993), at a rate which varies from 8-
10 cm a-1 from north to south along the Philippine
Trench. The West Philippine Basin therefore
subducts along the Philippine Trench more
obliquely in the north than in the south. The
Philippine fault, parallel to the trench, accommo-
dates part of the lateral component of this move-
ment, assuming a simple model of shear
partitioning. As a result the East Philippine crustal
fragment is displaced to the north.
Th~ recent development (3 to 5 Ma) of sub-
duction at the Philippine Trench is suggested by
(1) the presence of a slab extending to less than
200 km depth (Cardwell et al. 1980; McCaffrey
1982); (2) recent arc volcanism in eastern
Mindanao (Quebral et al. 1995); and (3) young arc
volcanism in Leyte, Samar and Bicol (Aurelio et al.
1991). The formation of this new active margin is
interpreted to be the result of a jump in the position
of the subduction zone from west to east after the
Philippine arc collided with the Eurasian margin.
Further south in the Molucca Sea, the Halmahera
arc, carried on the Philippine Sea plate, is in frontal
collision with the Sangihe arc (McCaffrey et al.
1980). The Philippine Trench terminates at this
latitude (Nichols et al. 1990). The collision zone is
marked by a thickening of the crust below the
basin, mainly deduced from gravity data and
interpreted as oceanic crust imbrication in the
central part of the basin (Miangas-Talaud-Mayu-
Tifore ridge). The oblique movement of the
Philippine plate with respect to Eurasia has there-
fore been interpreted to have resulted in the
diachronous closure of the Molucca Sea basin. The
presence of a slab more than 700 km below the
Celebes Sea (Cardwell et al. 1980) provides an
indication of the extent of the subducted oceanic
basin.
The Philippine archipelago is the site of the
active collision between the Philippine Sea plate
carrying an extinct volcanic arc and continental
fragments rifted away from the Eurasian margin
(Holloway 1982; Rangin et al. 1990a). The present
tectonic setting of the Philippines shows an elon-
gated Philippine Mobile Belt, fringed by two recent
subduction zones (Lewis & Hayes 1983), and
traversed axially by the 2000 km long left-lateral
Philippine fault (Willis 1937; Allen 1962; Aurelio
et al. 1991). The volcanic arc of Oligocene age,
which forms the basement to the eastern part of the
mobile belt, is in tectonic contact with the Eurasian
margin represented by continental fragments
exposed in the western part of this belt, and rifted
away from the Eurasian mainland in south China
(Holloway 1982; Rangin et al. 1990a).
Southeast of Mindanao, the collision between the
Halmahera arc and the along-strike equivalent of
the western part of Mindanao occurred during the
Late Pliocene and Pleistocene (Moore & Silver
1983; Hawkins et al. 1985; Mitchell et al. 1986;
Pubellier et al. 1991; Quebral et al. 1995).
According to many authors (Roeder 1977;
Cardwell et al. 1980; McCaffrey et al. 1980; Moore
& Silver 1983; Hawkins et al. 1985), the arc-arc
collision of the Molucca Sea should be traceable
onshore into Mindanao, and the Philippine fault has
been interpreted as the inferred suture. However,
geological observations (Mitchell et al. 1986;
Pubellier et al. 1991) indicate that the Philippine
fault is not a former suture although a possible
suture candidate may be traced through the
Cotabato basin.
Transect across the Northern Molucca Sea
On the basis of the newly acquired bathymetric data
(Fig. 2), this paper presents first a summary of the
key features and preliminary interpretations of a
NE-SW transect across the Molucca Sea south of
Mindanao island. This northern transect extends
from the Philippine Sea plate to the Celebes Sea
immediately south of Mindanao. We then discuss
a shorter but complementary transect carried out
further south, around the Talaud Islands.
1. Philippine Trench to the
Cotabato Trench
Various morphostructural provinces are distin-
guished from east to west along this transect
(Fig. 3).
 COLLISION AND FAULTING IN THE N MOLUCCA SEA 33

7N

Fig. 3. Shaded image of the sea bottom in the northern most part of the Molucca Sea, south of Mindanao. The main
morphological units of the area are shown. DB: Sarangani-Davao Depression; SB: Sangihe forearc Basin;
PR: Pujada ridge; CD: Central Depression.

Outer slope of the Philippine Trench
Close to 6°N, there is a dog-leg bend of the
Philippine Trench. The trench trends north-south in
the north and suddenly changes to a mean direction
of N N W - S S E further south. North of 6°N, the
outer slope shows regularly spaced west-facing
fault scarps interpreted as parallel normal faults,
down-thrown towards the trench, with vertical
offsets which are up to 900 m (Fig. 4). The outer
slope is one of the steepest outer slopes observed at
a trench anywhere in the world with gradients of up
to 11 o to the west. South of 6°N, the outer slope is
less steep with an average gradient of 6 ° . Linear
sub-parallel scarps trend at 160 ° to 170 ° and are
slightly oblique to the north-south trend of the
trench. The normal faults dissecting the outer slope
vary in orientation from 160 ° to 140 ° and are
parallel to the trench axis. In some places the faults
are slightly oblique to the trench axis and trend
125 ° (Fig. 4). The very deep (9600+_100 m)
V-shaped trench is characteristic of a non-
accretionary convergent margin. Its north-south
trend north of 6°N is slightly en echelon. South of
6°N the trench consists of a series of en echelon
basins offset along left-lateral faults resembling
Riedel fractures, and its depth shallows from north
to south by almost 1000 m.
The inner slope of the Philippine Trench
The inner slope of the Philippine Trench, which
descends eastwards towards the trench, can be
divided into three morphological areas in the
northern part of the area surveyed: the lower,
middle and upper slope regions (Fig. 3). South of
6°N the lower and upper slope areas lose their
distinct character and merge with the middle slope.
The middle slope is separated from the trench by
the Snellius Ridge (Figs 3 and 4).
Lower slope. The lower slope of the upper plate
(Figs 3 and 4) is geomorphologically very different
34 C.
5N--
127 E
Fig. 4, Bathymetric map of the Philippine Trench Inner wall. Main structures are also indicated. South of 6°N,
the complex morphostructural zone between the Snellius Ridge and the Miangas ridge is the middle slope unit.
in the areas north and south of 6°N. North of 6°N it
consists of an 8000 m deep terrace. In this area, the
scarps at the base of the inner wall that define the
front of the margin could be interpreted either as
normal or thrust faults. It is very difficult to trace
the true plate boundary since no decollement is
observed on seismic profiles. South of 6°N a
triangular-shaped terrace lies between the
Philippine Trench and the middle slope. The major
feature observed in this area is a 50 km linear scarp
RANGIN ET AL.
~ PHILIPPINE
I~i TRENCH
, \
Outer
~-jSlope
facing the trench with a 140 ° trend. A vertical
offset of more than 2 km is observable along this
scarp that could be interpreted as a major normal
fault. West of the scarp the lower slope dips gently
towards the NE and pinches out along the trench
close to 5°30'N. This surface is part of the flank of
the Snellius Ridge which broadens southwards.
Middle slope. The middle slope has an average
gradient of 10 ° to the east across the whole of the
 COLLISION AND FAULTINGIN THE N MOLUCCA SEA
area surveyed. In the northern part of the area, north
of 6°N, it is a complex morphostructural zone of
interfering structures. The major characteristic of
this area is a series of NW-SE trending features,
interpreted as a strike-slip faults. One of these
fractures appears to be the offshore extension of
the Philippine fault (Quebral et al. 1995) exposed
in Mindanao (Fig. 5). Discrete 045 ° trending
structures are also observable and may represent
a fold axial trend. South of 6°N, the orientation of
the middle slope break changes from 140 ° to north-
south, and 010 ° close to 5°40'N, and then turns
again to 140 ° south of 5°N. This morphology is
interpreted as a combination of NW-SE trending
left-lateral strike-slip faults and 010 ° trending folds
and thrusts. The middle slope break is probably
seismically active in the southernmost part of
the area since thrusts were observed on seismic
profiles.
To the west, the middle slope unit is in tectonic
contact with the Miangas ridge. At the base of the
steep east-facing scarp of this ridge there is a
narrow area of broken topography which descends
over about 200-300 m to the middle slope (Figs 3
and 4). This large area (Fig. 3) has an average depth
of about 3800 m between 5°30'N and 5°50'N
and rises to the south and north. A number of
lineaments can be seen on the bathymetric and
reflectivity maps which mark irregular ridges with
principal orientations of 015 °, 140 °, 170 ° and 180 °.
The intersection of these ridges outline a number of
small rhomb-shaped, flat-bottomed basins.
Upper slope. The upper slope is steeper than the
middle slope north of 6°N and absent further south.
In its southern part east-verging thrusts were
imaged on seismic profiles. Further to the north,
lobate bodies with edges which are convex to the
east, are interpreted as the manifestation of flat
east-verging thrusts crossed by NW-SE trending
strike-slip faults (Figs 3 & 4).
Miangas-Pujada ridges
This north-south trending bathymetric high is a
complex broad feature formed by two ridges which
rise from depths of 3000 m on each side (Fig. 5).
To the north the western ridge continues into the
Pujada peninsula and to the south the eastern ridge
widens considerably to form a bank connecting
southward with the Talaud Islands (Fig. 2). For
most of its length this high is divided by a
depression (the Central Depression of Fig. 5). In the
north, the Central Depression is little more than an
irregular deeper part of a single bathymetrically
high area with steep east- and west-facing slopes
(Figs 3 and 5). Further south the Central Depression
becomes more pronounced and deeper.
35
The Miangas ridge. This eastern ridge remains
high along all its length and is emergent on the
island of Miangas at 5°36'N. In most of the area
surveyed its depth is typically less than 1200 m.
The ridge trends broadly towards 160 ° but in detail
the eastern face can be seen to be composed of
en echelon north- and 140 ° trending segments. It is
asymmetrical in cross-section, and the highest
points are on its eastern side, with a very steep east
slope (average gradient 20% and a much more
gently sloping west face (average gradient 9°). This
ridge has a consistently high reflectivity, and
seismic lines indicate an absence of sediment on
most of the upper parts of the ridge and on the very
steep east-facing slope.
The Miangas ridge is interpreted to be thrust
eastwards over the middle slope of the Philippine
Trench. This interpretation is supported by pre-
liminary seismicity data obtained in an OBS
(Ocean Bottom Seismometer) network installed
around Miangas Island during the MODEC
Cruise. This temporary network revealed the
possible presence of a west-dipping thrust plane
below the ridge. This thrust is not observed on
seismic profiles due to the steepness of the slope
and the absence of sediments. The inferred thrust
trace at the western base of the ridge is offset by
140 ° trending lineaments at 5°30'N and 5°40'N,
south and north of Miangas Island respectively.
These features are interpreted as left-lateral strike-
slip faults offsetting the ridge (Fig. 5).
The Pujada ridge. The summit of the western
ridge deepens southwards, narrows and disappears
at 5°05'N (Fig. 3 & 5). South of 5°30'N, the ridge
turns east and has an overall orientation parallel to
the base of the Sangihe arc (c. 155°). Its western
slope becomes somewhat less steep and more
irregular although there remains a steep section
about half-way up. Its eastern face has a broad,
gently east-dipping upper terrace between 2000 m
and 2250 m which then descends more steeply, but
irregularly, to the well-marked Central Depression
with a flat surface at 3700 m. The ridge has a steep
west-facing slope. The reflectivity map and the
seismic lines show little sediment cover and only a
few discontinuous reflectors (Fig. 6); where there
are reflections they mainly dip west but no structure
can be identified with confidence. It is suggested
that these areas are underlain by basement rocks
such as ophiolites or arc volcanics. Two principal
lineament orientations, 150 ° and 050 ° are parallel
to the steeper slopes on the upper part of the ridge.
There are no clear offsets on these lineaments and
no clear indications that they are faults on seismic
lines although the 150 ° direction is close to the
orientation of the western margin of the Sangihe
trough and inferred axial directions of folds within
 =!
i=
oE
i=l., o
~8
/
J .... , Z .--=j"
Ii
,x- .... ;:'~ <:%d~ - ~.
J
J
,G. 18
I:
,G. li
,-)
p - , , ~ e ~ . . . . /" • _/J ,--~ >
Q.
; .:' ; : f . i' f
,•!., .~-:;~ .. -:'.- u.,
, ,
.2.-~
. <: ," , (-
~k<>~'- ~ ;
oooP ~,!,
.17v~%,, ~ - : ( > ~ -* -
,>':.737~;75;=., . . , ,,j))
\
'~ ~~c'FT, Q': 7:7 ::7~7"{::;~ " <
7
°/
i'l
I:1:1
I-
o f
> !
ill .,,-i
i'-'i
Z~
P~
 125"E +~.~ 126"E ~a~*aOE
........t................. 1 , i ...... I..

,.. Ii+i+~++ii+++++++
:+++i+++++!
++iiii+i!iiiWi~
w + ,'+,+,ar+ +

Sarangani ridge 6ON- ++i! ~,+ +

M l'ta r ;a-= ~ {+i++;~ Z ~g{~:>+V.~:>-';?Lv.+_.:-.::i~:~-:

~+~-i~::~ ~ ~ ll~ ~ ~+,a _ ~ ~-~+2-!*+'.+.:--+-?: :-.:':.:.+:.!=+

<:::!'-"+:~,i;+++.+.''..:, :-5,:~

]4i'7"ie:-:':)'i:::!!:'ili!{i:+'{!i:-.'
+~ ~!":-'~; ~g~{i~{i+{i~!!-<~ " Line 19

Pujada ridge . ~ i
....... Sarangani ridge ~ !

:?:-;~:'.+.k~~
g:~:~:~.~`~:`...:~:~4?.:::~.`~?~:`!i:~5~:~¢?>:::..?.#~i!~`yy?:~.~!~:s#~ }::.<:I:~4P{G#--_:~--~:~?!~.!+;::;:?:.#-~
.

6 -:: :.?.:.~:::57:i..:,7.?:7:-.!:.; .:&~>:-~:::.i .... : ,:. :: +. +i!<e.:?~?~

' ~!{~:~{~>~e~:¢~.{g>;~}~-~`..{r.~...?f~:~-+~.7c~<:-~:~.{:#5~::~{~i~g ~>+gi~+??:.>~£÷:#:.:_;g_.~_~?4---!g-:3--..:~
_~.-; ]-? :::]].::::7'7"r.::':i?i#:i
?Si~::'" :' :i?i ': :;: : " ""':(:': e:+!!).i`~{>!e{~!)5#~{>{~#{~{;i~i{~{~{{!;~.!~z~.~:~i:;f¢!~:¢:~`:~}~:gi{{~#
~~~a~%##i~v~+~!i>{i::~':::::':~-+~i> ,:'{;v<:7'v:~:?5~2:;~:; L in e 2 1

Fig. 6. Single channel seismic profiles and interpreted line drawings across the Sangihe forearc basin. Note increasing deformation in the basin from south to north.
t.aa
-.dl
38 C. RANGIN E T A L .

the Sangihe trough (see below). These lineaments
may represent a conjugate fault/joint set within the
basement. The eastern side of the Pujada Ridge is
locally steeper than the western side, but includes a
number of irregular terraces which become wider
southwards. Medium to low reflectivity and
seismic profiles indicate that there is a sediment
cover although the thickness is not great (< 1 s twt).
Over both the Pujada and Miangas ridges, the
free-air anomaly map is well correlated with the
bathymetry (Fig. 7). The slope descending from
the Sangihe arc is characterized by a regular
gradient of the free-air anomaly with values
decreasing towards the Sangihe forearc basin. The
minimum (-80 mGal) is situated at the northern
end of this basin at 5°50'N and 125°50'E, above an
area of thrusting, where the sedimentary thickness
is probably larger. The Pujada and Miangas ridges
are underlain by north-south aligned positive
anomalies separated by a low corresponding to the
Central Depression. The Miangas ridge, which is
topographically higher than the Pujada ridge, is

126 E 127 E 128 E

I i i

7N + , -~\9~IIT! I | ..,. +

6N

5N

%
4N + ,~

3N

o?
Fig. 7. Free-air gravity anomaly map of the surveyed area.
 COLLISION AND FAULTING IN THE N MOLUCCA SEA
associated with a smaller anomaly. Northwards, the
two anomalies merge into a single one, as the ridges
do. On the eastern edge of the Miangas ridge, a
steep gradient (more than 15 mGal/km) can be
correlated with its steep eastern flank.
The Sangihe forearc
The Sangihe forearc basin (Figs 3 and 5) is located
east of the Sangihe volcanic arc and its northern
extension in Mindanao, the Sarangani peninsula,
and west of prominent ridges (Pujada and Miangas)
formed by ultramafic and mafic material (Fig. 3).
This basin pinches out to the north in the Davao
Gulf (Fig. 5), between the Sarangani and Pujada
peninsulas, where we term it the Sarangani-Davao
depression (Fig. 3). Between 5°37'N and 5°45'N
there is a marked change in orientation of the
eastern slope of the Sangihe volcanic arc, from
155 ° in the south to the 015 ° in the north. At this
latitude the basin becomes wider and structurally
simpler, and has a different orientation.
The Sarangani-Davao depression. With a central
plain at depths of about 3200 m, the basin has an
arcuate shape parallel to the Sarangani peninsula.
Field work in Mindanao (Pubellier et al. 1991) has
shown this arcuate peninsula is a large recumbent
fold verging towards the east. The Sarangani-
Davao depression corresponds to the northern
extension of the Sangihe Basin and sediments in the
basin are folded and thrust in the Davao Gulf. This
is related to the active compressive tectonics
affecting the island of Mindanao (Pubellier et al.
1991; Quebral et aL 1995). At the western edge
of the basin, the slope of the eastern edge of the
Sarangani peninsula is very steep (Fig. 5) and is
interpreted as the manifestation of a buried thrust.
In the central part of the basin, three broadly north-
south trending, arcuate, and west-verging thrust
anticlines are well marked in the bathymetry (Figs
3 & 5). Reverse faults or thrusts are observed along
their eastern flanks. The free-air anomalies deduced
from six gravity profiles across the basin (Fig. 7)
show an asymmetric form with a clear minimum
(-80 reGal) close to Sarangani shoreline on the
west side of Davao Gulf. This anomaly is corre-
lated with the thickness of sediment, suggesting a
flexure of the basin due to loading, by overthrusting
from the west, of the Sarangani peninsula.
The Sangihe basin. Seismic profiles across this
basin show at least 2 s twt of well-stratified sedi-
ments (Fig. 6). Buried folds with a clear west-
vergence are observed on seismic profiles, with an
axial trend of 155 ° . This direction is parallel to
bathymetric features at the basin margins. South of
5°25'N the seabed surface in the basin changes
from a flat smooth surface to an irregular surface
39
cut by numerous channels. Silver & Moore (1978)
extended their East Sangihe Thrust through this
part of the Sangihe trough. Data presented here do
not support the presence of any west-verging thrust
along the western edge of the forearc basin.
Sangihe backarc and Cotabato Trench
East of the Sangihe arc, the survey was extended
into the Celebes Sea and along the southeast
termination of the Cotabato Trench. At about 5°N
the Cotabato Trench bends from north-south to an
approximately ENE-WSW orientation parallel to
the southern coast of Mindanao. In the Celebes Sea
basin the ship track crossed ODP sites 767 and 770
(Fig. 8).
South of 4°50'N the feature corresponding to the
continuation of Cotabato Trench is a broad flat-
bottomed area, sloping gently northwards. A large
meandering canyon running south to north down
to 5°N probably originates from the large island
of the Sangihe group indicating the trench dies
out morphologically further south. To the east
volcanoes are aligned north-south on the western
flank of the Sangihe arc. At 4°40'N a conical
seamount, probably belonging to the nearby
Sangihe volcanic arc, rises 3500 m above the sea
floor. Between 4°50'N and 5°20'N the Cotabato
Trench bends from an orientation of 000 ° in the
south to 120 ° in the north. The trench floor deepens
from 5100-5900 m as it approaches the Sangihe
arc. At the northern end of the north-south trending
section (north of 4°50'N), which is marked by a
100 reGal negative gravity anomaly, a narrow
NNE-trending accretionary wedge is observed at
the foot of the steep western flank of the Sangihe
arC.
At the trench bend there is an intermediate zone
which trends 145 ° and is separated from the
Sangihe arc by a non-linear ridge and a plateau
which could be the physiographic equivalent of an
outer arc high. A NNW-SSE trough 5100 m deep,
is parallel to these on their landward side. Here, the
trench-fill sediment sequence thickens significantly
up to 2 s twt. A re-entrant in the wedge associated
with bending of an anticline axis is interpreted as
the effect of a subducted asperity (ridge or
seamount) on the sea bottom. This is supported by
the presence of a 1600 m high curved scar observed
in the bathymetry of the wedge at 5°20'N,
124°50'E, which could be attributed to the collapse
of the margin following the subduction of a
seamount.
Along the Cotabato active margin, which trends
at 120 °, the lower slope consists of a series of
anticlines, 20-30 km apm't, with curved axial traces
which are concave upslope, forming a wide fold-
and-thrust belt typical of an accretionary wedge.
40 C. RANGIN ET AL.
The trend of the anticlines is roughly parallel to the
trench axis with an en echelon pattern in several
places. An outer arc high, 400-500 m in elevation,
separates the accretionary complex from what
is suspected to be a forearc basin with an upper
surface dipping slightly towards the southeast. A
large gravity low is located over the accretionary
wedge probably reflecting a thick sedimentary
prism.
The 4600-5500 m deep abyssal plain of the
Celebes Sea basin shows several ridges and
seamounts sitting on the oceanic crust. On seismic
profiles the basin floor has a very rough undulating
surface (amplitude of a hundred metres) which
possibly reflects the original grain of the oceanic
crust overlain by no more than 0.5 s twt thick
sediment cover. The first sediments deposited onto
the basement were dated at 43 Ma during ODP Leg
124 (Rangin et al. 1990b). In the vicinity of the
trench no clear normal faults were observed on
seismic lines. The bathymetric map (Fig. 8) shows
lineaments, ridges and troughs trending 110 ° in the
northwestern part of the area surveyed. An
elongated positive feature, 1.5 km above the sea
floor and without a magnetic signature, interpreted
as a group of seamounts, trends WSW, parallel to
the magnetic lineations proposed by Weissel (1980)
in the basin. Another major feature, trending north-
south, is more than 100 km long and l0 km wide
and rises 1 km above the sea floor.
2. Talaud Islands to Morotai Basin
A second transect, orientated NW-SE, south of the
Talaud Islands, was surveyed in the deep waters
between the Morotai Basin and the Sangihe trough.
It was chosen to be perpendicular to the regional
morphology. South of 5°N, the Miangas ridge
merges into a very shallow area covered by reefs
around the Nanusa Islands (Fig. 9). To the east,
the Snellius Ridge which flanks the Philippine
Trench between 4°50'N and 5°30'N, becomes very
shallow south of 4°50'N. Consequently, most of
the area at the latitude of the Talaud Islands was
not surveyed because it is too shallow. Various
morphological provinces were identified from west
to east.
Sangihe trough
Only a small part of the undeformed Sangihe fore-
arc basin (Sangihe trough on Fig. 9) was surveyed
in this area (Fig. 9). The flat part of the trough is
much narrower here than in the north. The area
deeper than 3500 m is less than 25 km wide. The
trough is slightly deeper in its central parts and
reaches depths of 4000 m. The West Talaud Bank
rises from the Sangihe trough towards the Talaud
Ridge and its lower slopes form two terraces
(Fig. 9) which occupy a triangular area which is
narrow in the north and widens southward. The
eastern margin of the West Talaud Bank is bounded
by west-verging thrusts following the base of the
cliffs of the Talaud ridge where a second OBS
temporary network has been deployed (Fig. 9). The
West Talaud Bank is thus thrusted towards the west
above the undeformed Sangihe trough.
Talaud ridge
The Talaud ridge forms a high area between the
2000 m bathymetric contours. Little of the ridge
appears on the map (Fig. 9) because of the very
shallow depths of the ridge summit. In the northern
part of the area surveyed the Taland and Nanusa
Islands are part of the same wide ridge, but further
to the south the Talaud ridge narrows and swings
southeastward. These trends are reflected by the
shape of the islands' coastlines. On seismic profiles
the ridge lacks any continuous reflectors and a very
thin sediment cover was observed above basement
rocks, possibly ophiolites or arc volcanic rocks, as
indicated by the geology of the Talaud Islands on
land (Sukamto et al. 1980; Moore et al. 1981). Both
east and west flanks are fault controlled. These are
most likely to be thrust faults because the fault
contacts are sub-parallel to the curved isobaths.
East Talaud Bank
The eastern margin of the Talaud ridge is a wide
area, named the East Talaud Bank, which connects
northwards to the middle slope of the Philippine
Trench inner slope(Figs 3, 9, 10). South of 3°55'N,
the East Talaud Bank is very wide (about 35 km)
and has an irregular surface which is convex-
upwards. North of 3°55'N, it becomes considerably
narrower (about 18 km wide), and forms a north-
south trending feature between depths of 1500
and 3000 m. South of 4°N (Fig. 9), the bank is
separated from the Talaud ridge by NW-dipping
thrusts. Seismic profiles show that the East Talaud
Bank resembles the West Talaud Bank which is
interpreted as the deformed margin of the Sangihe
Basin to the west.
Along the east flank of the bank, a major low
angle thrust can be traced southwards from 4°40'N
to 4°00'N where it has a north-south trace; it then
turns, first towards 140 ° , and then further south
towards 030 °. There are at least 3 s twt of sedi-
ments in the slice above the thrust plane, which
carries the East Talaud Bank eastwards onto the
Snellius Ridge, an observation reported previously
by Silver & Moore (1978).
The Talaud ridge, and the East and West Talaud
Banks on each side, are cut by 140 ° scarps that
 COLLISION AND FAULTING IN THE N MOLUCCA SEA 41

*d

0
E
O
if}
O.
:1:0 --, o

@
0
i,f l

¢,i
o

c ~
©
l
(1)
k_ o/

/ /
¢)

s II.~l

¢i

, ~~ -i ~ °~
"a

• ©
42 ¢. RANGIN ET AL.

E126 40 E127 20 E128

~J ,,,

N a n u s a Island ( Q k , ~ . . ~-~ ~S'~'-""¢,"tJ,ie '

~- Philippine i
~ Trench

N420
_ ~, ...... i)i:~ ~

!_/2. ?,

/
"¢ .,~(
• . : ~, ,
4,

• N3 20

...... I?:!L..:i.

1
_ ! I San

Fig. 9. Bathymetric map of the area around the Talaud Islands showing the main interpreted structural elements.

could be strike-slip fault zones. This is supported
by the evidence of earthquakes in this area with
strike-slip focal mechanisms (McCaffrey 1991).
The progressive offset of morphological units
towards the northwest suggests left-lateral motion
on these faults.
Snellius Ridge
East of the main thrust separating it from the East
Talaud Bank, the Snellius Ridge is a large sub-
marine plateau sloping to the west with a gradient
of 3-6 °. South of 3°40'N (Fig. 9) the plateau is
dissected by SW-facing fault scarps trending 140 °.
These features also dissect the northern margin of
the plateau (Fig. 9) and are interpreted as the trace
of major strike-slip fault zones (see above). Well
developed reflectors at the top of the Snellius Ridge
could be platform carbonates capping the plateau.
The positive gravity anomaly over the Snellius
Ridge suggests a dense crust compatible with an
ophiolitic or arc origin for the basement.
In the southern and deepest end of the Snellius
Ridge is the Morotai Basin. This basin has an
elongate shape with its major axis orientated 070 ° ,
and has a maximum depth of about 3500 m. Along
its northwestern margin, it is thrust westwards onto
the Snellius Ridge, and folds trending 075 ° defornl
the basin sediments. These structures terminate
abruptly to the west at the curved thrust zone where
the Morotai Basin margin is thrusted on top of
the Snellius Ridge. Further south the basin is filled
by more than 1.5 s twt of almost undeformed
sediments. There is no evidence for significant
 COLLISION AND FAULTINGIN THE N MOLUCCA SEA
deformation at the southern side of the basin close
to Morotai Island within the area surveyed,
although a little further south the sea bottom rises
rapidly from depths of more than 3000 m to emerge
at the steep fault-controlled coasts of the island.
The southwest end of the Snellius Ridge is
situated between two convergent thrust fault zones,
and is beneath the East Tataud Bank and the
Morotai Basin (to the NW and SE respectively).
The superficial convergent thrusts in this area are
thus compatible with interpretations at depth that
show two convergent subduction zones and an
inverted U-shape of the subducted Molucca Sea
oceanic slab. More precise correlations need to be
done between the superficial data collected during
the cruise and structures at depth.
Discussion
During the MODEC cruise in the northern Molucca
Sea it was possible to trace the Sangihe forearc
E 1 2 3 20. E124 E125
CELEBES ¶.
SEA
E123 20 El 4 E125
Fig. 10. Overall interpretation of the northern Molucca Sea. The shaded area is the Miangas-Pujada-Talaud unit
probably formed by ophiolite slivers and melanges. Immediately east of this is the middle slope unit.
43
continuously over 500 km from north to south.
Between the eastern edge of the forearc and the
Philippine Trench, three major crustal units or
terranes were also mapped almost continuously
from north to south (Fig. 10), and are interpreted to
be separated by major lithosphere-scale boundaries.
Miangas-Pujada-Talaud unit
The Miangas-Pujada-Talaud unit is probably
formed by ophiolite slivers and melanges; it can be
traced from the Pujada peninsula to the Talaud
Islands, with a major offset between Miangas and
Talaud north of the Nanusa Islands. The Pujada-
Miangas ridge is onlapped westward by the
sediments of the Sangihe forearc basin; further
south at the southern end of the Talaud ridge there
are west-vergent thrusts moving the ridge onto the
edge of the Sangihe forearc basin. This ridge could
be interpreted as the backstop of the Sangihe
subduction zone, implying it is underlain by
E127 E128 E1281 4 0
r . . . . . r
PHILIPPINE
I BASIN
E126 E;27 EI~8 ~I~ 40
44 C. RANGIN ET AL.
Sangihe forearc basement rather than oceanic crust
of the subducted Molucca Sea. The Miangas-
Pujada-Talaud unit is separated from the Philippine
Trench middle slope terrace by a major discontinu-
ity that could be a lithospheric-scale thrust. Seismic
focal mechanisms associated with this thrust are
clearly compressive with east-west P axes. This
thrust can be traced south at least to 3°N at the foot
of the Talaud ridge.
Middle slope unit
The middle slope unit, lying immediately to the
east of the Pujada-Miangas ridge, is formed by low
density material as indicated by the free-air gravity
data. The middle slope unit can be traced southward
into the East Talaud Bank which is thrusted east-
wards onto the Snellius Ridge, almost certainly a
fragment of the Halmahera arc terrane. North of
5°30'N, this thrust fault approaches the Philippine
Trench. A broad gravity low trending north-south,
about 100 km long and with minimum values of
-260 mGal is centred on the middle slope unit
(Fig. 7) and is clearly shifted to the west with
respect to the Philippine Trench (-220 mGal). This
anomaly cannot therefore be attributed to the
classical gravity low associated with deep trenches.
The juxtaposition of the low density material of the
middle slope unit in the north and the high density
material of the Snellius Ridge in the south, could
be explained by the presence of a major crustal
discontinuity trending 140 ° . The orientation of
isogals on the free-air gravity map above the
middle slope unit changes from north-south to
NW-SE approaching the northern edge of the
Nanusa Islands and the Snellius Ridge (Fig. 7). The
discontinuity is also marked by a pronounced
NW-SE steep gravity gradient. It is also sub-
parallel to the orientation of the Philippine Trench
south of 6°N (Fig. 10). Such a gradient and dis-
continuity is also clear in recent satellite FAA data.
This supports the presence of a major crustal-scale
fault zone separating two contrasting terranes.
The middle slope unit could be composed of
accreted sediments, and be part of the Sangihe
accretionary wedge, with the Snellius Ridge
representing part of the Halmahera arc. In this
interpretation the boundary between the colliding
arcs would be the thrust at the base of the middle
slope and East Talaud Bank. An alternative inter-
pretation is to consider this unit as a part of the
Halmahera arc like the Snellius Ridge to the south.
The middle slope unit would then represent an
intra-arc basin. Such contrasting features are
known within the Halmahera arc system to the
south (Hall et al. 1988a, b; Nichols & Hall 1991).
In this case the major west-dipping thrust located
east of the Miangas-Talaud ridge would be the
main tectonic boundary between the Sangihe and
Halmahera arcs. The major crustal discontinuity
trending I40 ° which offsets the Pujada-Miangas
ridge relative to the Talaud ridge is interpreted as
a left-lateral strike-slip fault zone.
Philippine Trench
The dog-leg shaped Philippine Trench has two
distinct segments. The north-south trending
segment north of 6°N is the trace of the recently
established west-dipping subduction zone as
indicated by the short subducted slab. The trend of
this trench segment is completely independent of
the dominant oblique NW-trending structures
observed in the inner slope. In contrast, south of
6°N, the N W - S E trending segment of the
Philippine Trench is parallel to the interpreted
left-lateral strike-slip fault zones observed in the
whole forearc area. The southern segment of the
Philippine Trench is tentatively interpreted as the
trace of a major left-lateral strike-slip fault zone
dissecting the Halmahera terrane. The small faults
in Riedel position observed along this part of the
trench support this hypothesis. However, thrust
fault focal mechanisms of seismic events reported
by Ranken et al. (1984) along this trench segment
reveal this strike-slip fault zone is now turning into
a thrust (Nichols et al. 1990). This suggests a
migration of the Philippine Trench towards the
south along pre-existing strike-slip fault zones.
It is also possible that, although focal mechanisms
support a thrust interpretation of this fault zone, the
main long-term motion is strike-slip, possibly
aseismic.
Conclusion
The area surveyed in the northern Molucca Sea
reveals the presence of two distinct geodynamic
settings east of the Miangas-Pujada-Talaud ridge.
A major NW-SE trending tectonic boundary
separates these two areas.
South of 6°N, the Snellius Ridge, a volcanic
plateau capped by carbonates, is subducting below
the Sangihe forearc terrane. The buoyancy of this
plateau, a fragment of the Halmahera arc terrane,
has induced deformation in the upper and lower
plates (Talaud ridge and Morotai Basin). Its buoy-
ancy also induces incipient subduction along a new
plate boundary, the Philippine Trench. This net-
tectonic feature is developed along a former strike-
slip fault zone cross-cutting the Halmahera terrane.
North of the major NW-SE crustal discontinuity
crossing the area surveyed, a wide accretionary
prism is present within the inner slope of the
present Philippine Trench. The thickness of this
wedge cannot be explained by subduction along
 COLLISION AND FAULTING IN THE N MOLUCCA SEA
this trench because it is very recent (Cardwell et al.
1980; Quebral et al. 1995). This wedge could
be interpreted either as an accretionary wedge
developed at the deformed leading edge of the
Sangihe forearc or be part of a former intra-arc
basin which was part of the colliding Halmahera
arc terrane. There is no evidence from gravity data
of dense material comparable to the Snellius Ridge
below this wedge. If the ridge extended originally
north of 6°N, it could have been already subducted
to some greater depth, or be considerably displaced
left-laterally, as suggested by the presence of a
similar buoyant terrane in central Mindanao
(Pubellier et al. 1991).
This survey also reveals no clear evidence for
References
ALLEN, C. R. 1962. Circum-Pacific faulting in the
Philippine-Taiwan region. Journal of Geophysical
Research, 67, 4795-4812.
AURELIO,M. A., BARRIER,E., RANGIN,C. & MULLER,C.
1991. The Philippine Fault in the Late Cenozoic
evolution of the Bondoc-Masbate-Leyte area,
central Philippines. Journal of SE Asian Earth
Sciences, 6, 221-238.
CARDWELL,R. K., ISACKS,B. L. & KARIG,D. E. 1980. The
spatial distribution of earthquakes, focal mechanism
solutions and subducted lithosphere in the
Philippine and northeastern Indonesian islands. In:
HAYES, D. E. (ed.) The Tectonic and Geologic
Evolution of South-east Asian Seas and Islands.
American Geophysical Union Monograph, 23,
1-35.
HALL, R., AUDLEY-CHARLES, M. G., BANNER, E T.,
HIDAYAT,S. 8,~ TOBING, S. L. 1988a. The basement
rocks of the Halmahera region, east Indonesia:
a Late Cretaceous-Early Tertiary forearc. Journal of
the Geological Society, London, 145, 65-84.
& --. 1988b. Late
Paleogene-Quaternary geology of Halmahera,
eastern Indonesia: initiation of a volcanic island arc.
Journal of the Geological Society, London, 145,
577-590.
HAMILTON, W. 1979. Tectonics of the Indonesian region.
US Geological Survey Professional Paper, 1078.
HAWKINS,J. W, MOORE,G. E, VILLAMOR,R., EVANS,C. &
WRIGHT, E. 1985. Geology of the composite terrane
of east and central Mindanao. In: HOWELL, D. G.
(ed.) Tectonostratigraphic Terranes of the Circum-
Pacific Region. Circum-Pacific Council for Energy
and Mineral Resources, Earth Sciences Series, 1,
437-463.
HOLLOWAY, N. H. 1982. The stratigraphic and tectonic
evolution of Reed Bank, North Palawan and
Mindoro to the Asian mainland and its significance
in the evolution of the South China Sea. AAPG
Bulletin, 66, 1357-1383.
KRAUSE, D. C. 1966. Tectonics, marine geology and
bathymetry of the Celebes-Sulu Sea region.
Geological Society of America Bulletin, 77,
813-831.
45
accreted oceanic crust belonging to the recently
subducted Molucca Sea. The M i a n g a s - P u j a d a -
Talaud Ridge is clearly in a back-stop position for
the Sangihe arc, and the ophiolites of this ridge
could therefore have very different origin from
either the Molucca Sea or the Halmahera arc.
This work was possible due to the co-operation programs
between France, the Philippines and Indonesia. We want
to thank particularly E. Domingo in the Bureau of Mines
and Geosciences in Manila and Dr Zen of BPPT in
Jakarta. The active co-operation of the L'Atalante crew
and Genavir team was particularly appreciated. We thank
particularly Captain G. Tredunit for his enthusiasm for the
project and his active participation in contributing to the
success of this cruise.
LEWIS, S. D. & HAYES, D. E. 1983. A geophysical study
of the Manila Trench, Luzon, Philippines: forearc
basin structural and stratigraphic evolution. Journal
of Geophysical Research, 89, 9196-9214.
MCCAFFREY, R. 1982. Lithospheric deformation within
the Molucca Sea arc-arc collision: evidence from
shallow and intermediate earthquake activity.
Journal of Geophysical Research, 87, 3663-3678.
- - 1991. Earthquake and ophiolite emplacement in the
Molucca Sea collision zone, Indonesia. Tectonics,
10, 433-453.
--, SILVER, E. A. & RAITT, R. W. 1980. Crustal
structure of the Molucca Sea collision zone,
Indonesia. In: HAYES,D. E. (ed.) The Tectonic and
Geologic Evolution of South-east Asian Seas and
Islands. American Geophysical Union Monograph,
23, 161-177.
MITCHELL, A. H. G., HERNANDEZ, E t~ DE LA CRUZ,
A. E 1986. Cenozoic evolution of the Philippine
archipelago. Journal of SE Asian Earth Sciences,
1, 3-22.
MOORE, G. E & SILVER,E. A. 1983. Collision processes
in the northern Molucca Sea In: HAYES,D. E. (ed.)
The Tectonic and Geologic Evolution of South-east
Asian Seas and Islands. American Geophysical
Union Monograph, 27, 360-372.
- - - , KADARISMAN,D., EVANS,C. A. & HAWKINS,J. W.
1981. Geology of the Talaud Islands, Molucca Sea
collision zone, northeast Indonesia. Journal of
Structural Geology, 3, 467-475.
NICHOLS, G. J. • HALL, R. 1991. Basin formation and
Neogene sedimentation in a backarc setting,
Halmahera, eastern Indonesia. Marine and
Petroleum Geology, 8, 50-61.
, --, MILSOM, J., MASSON,D., PARSON, L. ET AL.
1990. The southern termination of the Philippine
Trench. Tectonophysics, 183, 289-303.
PUBELLIER,M., QUEBRAL,R., RANGIN,C., DEFFONTAINES,
B., MULLER, C., BUTTERLIN,J. MANZANO,J. 1991.
The Mindanao, collision zone, a soft collision event
within a continuous strike-slip setting, Journal of
Southeast Asian Earth Sciences, 6, 239-248.
QUEBRAL,R., PUBELLIER,M. & RANGIN,C. 1995. Eastern
Mindanao, Philippines: transition zone from a
46 C. RANGIN ET AL.
collision to strike-slip environment. Tectonics,
(in press).
RANGIN, C., JOLWET, L. & PUBELHER, M. 1990. A
simple model for the tectonic evolution of southeast
Asia and Indonesia region for the past 43 m.y.
Bulletin de la Soci(t( g~ologique de France, 8 VI,
889-905.
, SILVER, E. A., VON BREYMANN,M. T. ET AL. 1990.
Proceedings of the Ocean Drilling Program Initial
Reports, 124.
RANKEN, B., CARDWELL, R. K. & KARIG, D. E. 1984.
Kinematics of the Philippine Sea Plate. Tectonics,
3, 555-575.
ROEDER, D. 1977. Philippine arc system - collision or
flipped subduction zone? Geology, 5, 203-206.
SENO, T., STEIN, S. A. & GRIPP, A. E. 1993. A model for
the motion of the Philippine Sea plate consistent
with NUVEL-1 and geological data. Journal of
Geophysical Research, 98, 17 941-17 948.
SILVER, E. A. & MOORE, J. C. 1978. The Molucca Sea
collision zone, Indonesia. Journal of Geophysical
Research, 83, 1681-1691.
SUKAMTO,R., SUWAR~O,N., YUSUP, J. & MONOARFA,M.
1980. Geologic map of the Talaud Islands,
1:250 000. Geological Research and Development
Centre, Bandung, Indonesia.
WEISSEL,J. K. 1980. Evidence for Eocene oceanic crust in
the Celebes basin. In: HAYES, D. E. (ed.) The
Tectonic and Geologic Evolution of South-east
Asian Seas and Islands. American Geophysical
Union Monograph, 23, 37--47.
WILLIS, B. 1937. Geologic observations in the Philippine
islands. Natural Resources Council of the
Philippines Bulletin, 13.
 Continental collision in the Banda arc

A. N. R I C H A R D S O N 1 & D. J. B L U N D E L L
Department of Geology, Royal Holloway, University of London,
Egham, Surrey TW20 OEX, UK
1 Present address." Shell Internationale Petroleum Maatschappij B. V., Postbus 162,
2501 A N Den Haag, The Netherlands

Abstract: Two deep seismic reflection profiles across the zone of convergence between
Australia and the Banda arc east of Timor reveal the structures formed during lithosphere
deformation in response to continental collision. Australian continental crust is bent down to the
north to form the lower lithospheric plate. The immediately overlying upper plate is made up
of a former outlier of the Australian continental shelf, now squeezed between the Australian
continent to the south and the Banda arc to the north. The ages of metamorphism and of a
regional unconformity in Timor indicate that the outlier began to accrete to the upper plate in the
late Miocene. Near the Timor trough, at the junction between the two plates, seismic reflection
data show the upper plate to be cut by north-dipping structures interpreted as thrusts in an
accretionary prism formed since the arrival of the Australian continental shelf at the collision
zone at about 2.5 Ma. The northern part of the collision zone is dominated by structures dipping
southwards, antithetic to subduction, which penetrate the lithosphere to depths of at least 50 kin.
Earthquake data indicate that these are active thrust faults which divide the upper plate into
imbricate slices. The seismic reflection sections, supported by gravity data and seismicity
patterns, have been used to develop a model of plate convergence and continental collision which
now consists of a 200 km wide zone in which both oceanic and continental material is being
shortened, thickened and uplifted.

The Banda arc lies at the intersection of the SE
Asian and Indo-Australian plates (Fig. 1). Over the
past few million years, Australian continental
material has arrived at and collided with the
oceanic plate of the Banda Sea in eastern Indonesia.
The collision has progressed to such an extent
that continental lithosphere has subducted under
oceanic lithosphere to a depth of over 100 km. The
collision of continental material with the sub-
duction zone has given rise to a 'collision com-
plex'. This forms a ridge, in places rising above
sea-level (e.g. Timor), sub-parallel to the volcanic
arc. Over the years, three main structural models,
resulting mainly from near-surface observations,
have been proposed for Timor:
(1) The Imbricate Model (Fig. 2a; Fitch &
Hamilton 1974; Hamilton 1 9 7 9 ) i s based
largely on marine geological and geophysical
data (e.g. von der Borch 1979; Silver et al.
1983; Karig et al. 1987). In this model, Timor
is interpreted as an accumulation of chaotic
material imbricated against the hanging wall
of a subduction trench, the Timor trough, and
essentially forms a large accretionary prism.
(2) The Overthrust Model (Fig. 2b) is probably the
oldest model in which Timor was interpreted
in terms of Alpine-style thrust sheets (e.g.
Wanner 1913). This model was dominantly
based on surface geology where overthrust
sheets of the Timor allochthon are well
exposed. Subsequent workers (Carter et al.
1976; Barber et al. 1977; Barber 1979; Haile
et al. 1979; Brown & Earle 1983; Audley-
Charles 1981, 1986a, b; Price & Audley-
Charles 1983, 1987; Harris 1989; Audley-
Charles & Harris 1990) have made a clear
distinction between allochthonous units of
non-Australian origin and parautochthonous
units derived from the Australian continent.
(3) The Rebound Model (Fig. 2c; Chamalaun &
Grady 1978) suggests that the Australian
continental margin entered a subduction zone
in the vicinity of the Wetar Strait. Subse-
quently, the oceanic lithosphere detached from
the continental part, resulting in the uplift of
Timor by isostatic rebound on steep faults.
Combinations of these models have also been
proposed (e.g. Charlton 1989; Charlton et al. 1991;
Harris 1992) in which the parautochthon is divided
into two parts, reflecting the step-wise nature of
the collision. The main part, the 'underplated'
parautochthon, is thought to have accreted in the
early stages of collision by the sequential addition
by underplating of imbricate thrust slices of the

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, 47

Geological Society Special Publication No. 106, pp. 47-60.
48 A. N. RICHARDSON • D. J. BLUNDELL

South

10 °
125 ° 126 ° 127 ° East 128 ° 129 ° 130 °

Fig. 1. Location map of the BANDA deep seismic reflection survey lines DAMAR and TIMOR, with shot point
numbers shown in italic. Earthquake epicentres shown as stars; numbers relate to McCaffrey (1988), table l, letters
a, b and c to events of 30 May 1988, 8 August 1988 and 21 May 1991 (Harvard centroid-moment tensor solutions).
Barbed line represents the plate boundary with barbs in the direction of subduction. Arrow indicates relative plate
motion. Bathymetry in metres.

outermost Australian continental margin to the base
of the forearc complex. A second phase of collision
is thought to have commenced in the Early Pliocene
with the addition to southern Timor of younger
parts of the Australian continental margin, the
'frontally accreted' parautochthon.
Data acquisition and processing
The British Institutions Reflection Profiling
Syndicate (BIRPS) joined with the Indonesian
Marine Geological Institute to record two long,
multichannel normal-incidence reflection profiles
across the Banda arc west of Timor. The two sub-
parallel traverses of the arc were designed to image
the deep structure of the central part of the collision
zone (Fig. 1). The proximity of the two lines to
each other and to Timor and the geologically
contiguous Minor Islands to the east increased
confidence that observed reflections represent
regional structural features consistent with surface
geology. The survey ship used was M/V GECO-
Kappa which towed a 92-channel hydrophone
streamer 4.6 km long. The airgun array of 7324 in 3
(120 litres) capacity was 70 m wide, designed to
produce 97 bar m peak-to-peak pressure at
frequencies of 3-62.5 Hz.
Along the western line, 'TIMOR', shots were
fired at 50 m intervals with 50 m receiver groups
producing records up to 23 s two-way time (TWT)
to show reflections at relatively high spatial
resolution down to about 45 k m depth. The eastern
line, ' D A M A R ' , was located where deep earth-
quakes are frequent and was therefore recorded to
37 s with 100 m shot intervals. This was designed
to compare reflectors, resolved slightly less well
than in the T I M O R profile, with earthquake
hypocentres and mechanisms down to around
140 km depth.
The seabed topography along both profiles is
rugged and ranges from < 100 m to >4000 m below
sea-level. This produced several sets of strong
 CONTINENTAL COLLISION IN THE BANDA ARC
NORTH SOUTH
(a) Imbricate Model
Volcanic Arc -13mor Australian
Continental Shelf
(b) Overthrust Model
Volcanic Arc 13mor Australian
(c) Rebound Model
Volcanic Arc -timer Australian
~ ~ ~ Continental Shelf
Fig. 2. Models proposed for the structure of Timor
(after Barber 1981).
seabed multiples, which partly obscure primary
reflections from deeper levels, and a high level
of random noise. The multiples were so easily
recognized on the brute stack that they could be
distinguished from primary reflections passing
through them. Attempts to remove these multiples
for the final stack sections were largely successful,
but at the same time probably degraded the
primary reflections. The brute stacks displayed the
clearest images of primary reflections, multiples
and diffraction patterns, and so were used more
than the final stacks in preparing line drawing
interpretations.
Following the interpretation method of many
other deep seismic experiments, discontinuous
linear segments of reflections identified as primary
on the seismic sections were traced on to a trans-
parent film overlay. These line drawings more
clearly delineate the overall structure and are
treated as an interpretation of the geometries of the
main reflections (Fig. 3). The principal reflections
interpreted on the line drawings were depth-
corrected and migrated in the plane of the section
to give their depth geometry (Fig. 4).
The only direct velocity inforhaation available
was provided from the moveout analysis of
reflected data from the 4.6 km long streamer. This
offset was insufficient to discriminate between
49
normal moveout curves for reflections at two-way
travel times longer than around 10 s, equivalent
to depths of around 25 km. For greater depths,
indirect evidence from refraction experiments in
the Timor trough and Banda Sea (Bowin et al.
1980), and from a general knowledge of p-wave
velocities in the upper mantle had to be used for the
depth conversion.
Interpretation
Australian Continental Shelf
(TIMOR SP 6400-8000,
DAMAR SP 1-750, Figs 3 & 4)
The upper 2.1 s TWT beneath the seabed of the
continental shelf portions of both sections consist
of fairly parallel reflectors interpreted as sedi-
mentary bedding. This is consistent with the tie
of the TIMOR line with the Troubadour No. 1 oil
exploration well (Fig. 1) which penetrated Recent
to Triassic marine sediments lying unconformably
on granite encountered at a depth of 3315m,
equivalent to 2.16 s TWT. Extensional faults are
evident in the upper 3 s TWT of both sections and
appear to penetrate the crystalline basement as well
as cut the entire sedimentary cover. These faults
were probably initiated in the Jurassic when the
Australian Northwest Shelf was last stretched
(Bradshaw et al. 1988; Powell 1982). In the region
of Sumba Island, where subduction is continuing at
present, many faults that were originally related to
rifting and spreading are still in active extension
and are giving rise to earthquakes with extensional
mechanisms. Some strong, north-dipping reflectors
were also identified in the otherwise transparent
upper crust (Fig. 3). These reflectors appear to sole
out of the Jurassic faults to cut through the entire
upper crust and continue into a reflective zone
in the lower crust. The upper continental crust
consists of crystalline rocks so it is most likely
that these reflectors represent faults, following the
interpretation of other deep seismic reflection
surveys elsewhere (e.g. Matthews et al. 1990).
Between 8 and 12 s TWT at the southern end
of the TIMOR section and 7 and 14 s TWT at the
southern end of the DAMAR section lies a zone
of intense, sub-horizontal, laterally extensive,
layered reflections. The base of this zone is con-
sistent with an increase in interval velocity from
6.3-7.3 to over 8 km s-1 at this level on both
profiles (Fig. 5) and corresponds well with the
normal jump in p-wave velocity across the Moho to
8.1 km s-1 in the upper mantle (Fowler 1990). The
velocity structure and character of reflection
patterns in this zone is therefore much like that of
the lower continental crust imaged by a number of
 N S

Shotpoint
1 2000 4000 6000 8000
0 ~ n n n I I I I

Z
T W T (s)
o
(a)

Shotpoint (/3
©
3000 2000 1000 Z
0 I I I I I I
u

10

20
TWT
(s) 30 . . . . . . . . ~ n km I an

(b)
Fig. 3. Line drawing interpretations of the BANDA seismic sections: (a) Timor; (b) Damar. Major reflectors and zones of high reflectivity are highlighted with dotted lines.
 CONTINENTAL COLLISION IN THE BANDA ARC 51

Shot Point
1000 2000 3000 4000 5000 6000 7000 8000
I ' ' ' I . . . . I , , , , I , , , , I , , , , I , . , . I , , . , I , , , , 0

50

km
q 50

km

100 I I I J 1 O0
NORTH 7o 8o 9* SOUTH
(a)
iI1~, Shot Point
3000 4b, j 2500 2000 1500 1000 500 0
I I I
--
I I I ~I I I i i
~ . - ~ - - - - - - ~ . _ . _
I I
- -
| I I I I l i i i , I , , , , I , , , ,
................
I

i , 6 ~ ~ ,~~ a ~ ~ 50

100 ~ 100

km " / km

150 i i i 150
NORTH 7° 8° 9° SOUTH
(b)
Fig. 4. (a) Interpretation of the major structures recognized on the TIMOR profile (Fig. 3a) after depth correction and
migration, and their association with earthquake data. Earthquake hypocentral locations, projected along the regional
strike (073°E), are shown as black dots. Fault plane solutions are shown as hemisphere projections in the vertical
plane of the profile. Upper crust: light grey; lower crust dark grey. (b) Interpretation of the major structures recog-
nized on the DAMAR profile (Fig. 3b) after depth correction and migration, and their association with
earthquake data.

other deep seismic reflection surveys from other The other observed deformation of the crust is a
areas (e.g. Mooney & Brocher 1987). The base gentle flexure as it is bowed down beneath the
of this zone of high reflectivity has been identi- Timor trough, bringing the Moho down to a depth
fied as the Moho (originally by Fuchs 1969) and there of 45-50 km.
it would be reasonable to apply the same
interpretation here. This reflectivity within the
lower continental crust has been associated with Subduction Trench
crustal extension, involving basaltic sill emplace- (TIMOR SP 6400-6450,
ment (Serpa & Voogd 1987; Warner 1990) and/or D A M A R SP 700-750)
ductile shear zones (Reston 1987). Since this part of
the Australian continent became a rifted margin in Undisturbed, horizontally bedded sediments of
the Jurassic (Bradshaw et al. 1988; Powell 1982), it 544 m thickness (0.48 s TWT) are present in the
would be reasonable to assign the sub-horizontal Timor Trough at the base of the trench on the
reflectors in the lower crust to this age. On this TIMOR profile. Using a sedimentation rate of
basis the crustal thickness at the southern end of 480 m/Ma in the trough axis obtained from analysis
both sections is 35-40 km, consistent with a crustal of DSDP-262 (Johnston & Bowin 1981), situated
thickness of 30-40 km determined in this area SW of Timor, these sediments accumulated in a
using seismic refraction data (Bowin et al. 1980). period of c. 1.1 Ma during which little convergence
 L~h
t~

Shotpoint
1000 2000 3000 4000 5000 6000 7000 8000
T ! i I i I i

10
TWT
(s) 15

20
(a)

Shotpoint
3000 2500 2000 1500 1000 500 1
0 , ! , i i =

T~q" I0
(s)

(b)
Fig. 5. Contour plots of interval velocity obtained from velocity analyses along (a) TIMOR and (b) DAMAR. Velocities in ms-1; contour interval 600 ms -1.
 CONTINENTAL COLLISION IN THE BANDA ARC
has taken place. However, immediately south of
the horizontal sediments, a small, north-dipping
reflector observed on TIMOR at SP 6500 is
interpreted as a thrust which has created a hanging
wall anticline and delaminated the upper 0.6 s TWT
of sediments, sliding them 200-400 m to the south.
Other thrusts have formed beneath. A similar
situation is observed on the DAMAR profile. Thus
a stick-slip system of convergence appears to be
operating SE of Timor in much the same way as has
been observed in the Timor trough by Karig et al.
(1987) SW of Timor.
Interval velocities in excess of 5 km s-1 are
encountered beneath the trench at 3 s and 2 s TWT
below the seabed on the TIMOR and DAMAR
profiles respectively (Fig. 5). This is in
approximate agreement with refraction data (Bowin
et al. 1980). Seismic velocities appropriate for
granitic basement are thus encountered at about the
same depth below the seabed both at the trench and
on the continental shelf at the southern end of the
two profiles.
The crust, measured from the base of the highly
reflective region, appears somewhat thicker at the
trench than at the southern end of both sections
(Fig. 4). Thickening could imply that the crust
has shortened on, or just prior, to its arrival at
the subduction trench. However, upper crustal
shortening through inversion of extensional faults
offsetting sedimentary layers is not evident, except
within the trench. The zone of high reflectivity,
interpreted as representing the lower crust, appears
to terminate abruptly just north of the subduction
trench in the DAMAR section. If the high reflec-
tivity of the lower crust is interpreted as a struc-
tural fabric of Jurassic age, this would be destroyed
if it were folded and sheared by compression. In
this case, reflectivity would be drastically reduced,
which could explain its abrupt termination.
Accretionary Prism
(TIMOR SP 5800--6400,
D A M A R SP 750-1100)
A few kilometres north of the trench, reflectors
within 4 s TWT below the seabed are less coherent.
This is probably because of the disrupted nature
of the sedimentary package. Deeper reflectors tend
to be swamped by strong seabed multiples on
the brute stack and noise on the final stack.
Nevertheless, a series of southward-verging duplex
structures is visible on both sections within what
appears to be an accretionary prism (Fig. 3). This
interpretation is consistent with the results of
seismic reflection surveys to the west and south-
west of Timor and south of Sumba (Karig et al.
1987; Van tier Werff et al. 1994) and southeast of
53
the Tanimbar Islands (Jongsma et al. 1989;
Richardson 1994). Thus there is evidence that a
proportion of the sedimentary sequence is accreted
to the underside of the developing accretionary
prism within a few kilometres of the trench.
In this portion of both TIMOR and DAMAR
profiles there is a general increase in seismic
velocity from south to noah within 4 s TWT below
the seabed (Fig. 5). This is probably the result of
compaction and de-watering of the sedimentary
package as it is broken up and accreted to the upper
plate.
North-dipping reflectors co-linear with sedi-
mentary reflectors dipping beneath the trench
appear to extend only c. 30 km north of the sub-
duction trench on both profiles. North of this,
regions of reflectors parallel to sedimentary
reflectors dipping beneath the trench appear to
occur in discrete 'packages' offset by prominent
south-dipping reflectors (Fig. 4). Shallower
packages appear to be offset to the north of deeper
packages. The upper surface of these packages can
be followed downwards in a relatively straight line
to the base of the profiles and extrapolated to
c. 140 km depth to join the northern (top) surface of
the plate defined by earthquake hypocentres which
has been defined in this area by McCaffrey (1989).
There are three alternative interpretations for the
actual physical nature of the reflectors within the
packages. The north-dipping reflectors may
represent: (1) remnant sedimentary material
attached to the lower plate that has escaped
accretion to the collision complex and has been
carried to great depth; (2) faults or shear zones
representing the zone of decoupling between the
upper and lower plates; (3) accreted (underplated)
sedimentary rocks.
This last interpretation was adopted by Clowes
et al. (1987) for similar deep reflectors dipping
parallel to the subduction zone under Vancouver
Island and by Moore et al. (1991) for the eastern
Aleutian Islands, but the other two possibilities
appear equally plausible. Whichever is preferred,
the packages of reflectors may be interpreted as
broadly representing the upper surface of the
Australian plate beneath the collision complex.
Therefore, it does not appear that the slab has
detached above about 140 km depth.
Central Collision Complex
(TIMOR SP 3200-5800,
D A M A R SP 1100-2350)
North of the accretionary prism, where the
reflectors dip mainly to the north, both profiles pass
into a region of predominantly south-dipping
reflections. This geometry is similar to that seen on
54 A. N. RICHARDSON & D. J. BLUNDELL
high quality oil industry seismic reflection profiles
in the Tanimbar Islands region made available by
Union Texas (SE Asia) Inc. (Richardson 1994) and
in the forearc region around Sumba, along strike
from westem Timor (van der Werff et al. 1994).
The transition zone between north-dipping and
south-dipping reflectors appears complex in most
cases. The south-dipping reflectors represent either
sedimentary layers or low-angle faults because they
do not have the characteristic shape of diffraction
tails. They are unlikely to represent sedimentary
layering because they appear fairly continuous to
considerable depth whereas there is no sign of well-
bedded sediments just below the seabed. On the
TIMOR line (SP 32004400, 4700-5100), the
patterns of the south-dipping reflectors are
reminiscent of large-scale thrust sheets (Fig. 4).
Similar features are observed on the seismic
reflection profiles around the Tanimbar Islands
where south-dipping reflectors terminate upwards
in antiform folds. On the DAMAR profile, major
south-dipping reflectors appear to offset north-
dipping packages of reflectors to the north (Fig. 3),
consistent with north-vergent thrust faulting
involving the subducted plate. In both the BANDA
and Union Texas profiles, interval velocity
inversions coincide with these south-dipping
reflectors (Fig. 5). If a region with a normal
velocity gradient were deformed into a stack of
thrust sheets, velocity inversions could be expected
to develop at the base of some thrust sheets where
high velocity material overrides low velocity
material.
Shallow earthquakes occur in this area and
McCaffrey (1988) has calculated hypocentres and
focal mechanisms for a number of them (Figs 1
& 4). Because they have been projected up to
60 km onto the profile it is impossible to correlate
particular earthquakes with individual reflectors.
However, it is clear that there is a reasonable
association and the dominantly thrust mechanisms
of the earthquakes support the other evidence that
the reflectors represent thrusts. The reflectors'
appearance, their coincidence with interval
velocity inversions and association with thrust
mechanism earthquakes all suggest that they
represent north-vergent thrusts which appear to
reach depths of tens of kilometres, possibly
penetrating into the mantle. Similar reflectors
observed elsewhere have been interpreted as shear
zones and ancient subduction zone traces
(Matthews et al. 1990).
The crest of the upper thrust sheet at SP 3900 on
the TIMOR profile (Fig. 3) is a topographic feature
of the seabed and forms the eastward, submarine
extension of the island of Kisar. Kisar is offset from
the trend of the main archipelago and consists of
tightly folded felsic schists and mylonites which
can be correlated with the Aileu Formation of
northern Timor (Richardson 1994). The presence
of the thrust on the TIMOR profile in alignment
supports the idea that Kisar may have been uplifted
on an out-of-sequence thrust. However, the
structural grain in Kisar runs approximately N-S
(Richardson 1994), at right angles to the structural
trend of the other islands. It is possible that Kisar
has rotated by 90 ° during thrusting, or alterna-
tively that it forms a lateral ramp.
The seabed high at SP 3400 on the TIMOR
section (Fig. 3) is mid-way between Kisar and the
volcanic island of Romang. South-dipping
reflectors with velocity inversions (Figs 3a & 5a)
visible at SP 3200-3400 are interpreted as thrust
faults. Interval velocities below the topographic
high at TIMOR SP 3400 are similar to those below
SP 3900 but are generally lower than the next topo-
graphic high to the north at SP 2900. The crustal
thickness appears to be 18-23 km. The topo-
graphic high at SP 3400 on the TIMOR section is
therefore interpreted to be similar in composition to
Kisar and so may be the northernmost occurrence
of continental rocks at the surface along this profile.
Dating features identified on deep seismic
reflection profiles is one of the most intractable
problems within the field of deep seismics
(Klemperer et al. 1990) unless a reflection can be
traced to outcrop or there is other independent
evidence. Since none of the south-dipping struc-
tures on TIMOR and DAMAR have been identified
before, let alone recognised at the surface, dating
them must rely on more speculative geometrical
observations. Consequently, the cross-cutting
geometric relationship of the south-dipping
reflectors with the packages of north-dipping
reflectors is used as evidence for their age. The
packages of north-dipping reflectors are interpreted
as representing the upper surface of the Australian
plate beneath the collision complex. The amount
of subduction is thought to have been extremely
limited south of western Timor for the last 0.6 Ma
(Karig et al. 1987). Horizontal sediments at the
trench axis on the TIMOR profile imply that sub-
duction has been inactive there for the last 1 Ma
apart from a few hundred metres of south-vergent
thrusting. The shallowest cross-cutting, south-
dipping structure on the TIMOR and DAMAR
profiles cuts the boundary of north-dipping
reflectors about 30 km north of the trench axis.
Assuming the rate of convergence to be decreasing
(Johnston & Bowin 1981), it would have taken this
part of the lower plate c. 1.4 Ma to have reached its
present location. Hence, this thrust cannot be older
than c. 2.4 Ma which is approximately the age of
onset of the current collision derived from evidence
in the Timor trough south of western Timor
(Johnston & Bowin 1981).
 CONTINENTAL COLLISION IN THE BANDA ARC
The top of the north-dipping lower plate and the
deepest of the south-dipping thrusts define the
collision orogen as a triangular-shaped wedge
(Fig. 4). Sedimentary material is being added to
the southern end of the wedge by underplating on
north-dipping thrusts. However, the northern
portion of the collision orogen is deformed on
south-dipping, north-vergent thrust sheets.
The exact depth of the Moho is difficult to
constrain for this portion of the profiles because
the poor resolution of the velocity picks leads to
significant variations between adjacent values of
interval velocity. However, assigning the base
of the crust to the boundary between interval
velocities less than and greater than 8000 ms -l, the
crust appears to be thickest directly under the
topographic highs: SP 4600 (> 46 km), SP 4000
(c. 30 km) and SP 3400 (> 36 km) on the TIMOR
section and SP 1600 (c. 50km) and SP 2150
(c. 20 km) on the DAMAR section (Fig. 5).
A 2D density model was produced to match a
gravity profile along the TIMOR line recorded in
a previous survey by Woodside et al. (1989).
Although poorly constrained, the model serves to
delimit the approximate thickness of the crust in
this region. Two end-member versions of the model
were produced. In one (Fig. 6a), which defines a
minimum value for crustal thickness, the crust of
the central collision complex was made 30 km
thick, the same as the continental crust of the
Australian Northwest Shelf. A reasonable fit
Gravityo
(mgal) /
"1°°I
-200
~ ~ i i
(km) / //
150 ? /.s
2o0 3200
/ S
2~o 2~6 1~o ~o .=b
Distance from Trench(km)
Fig. 6. Gravity models to constrain estimates of crustal thickness in the central part of the Banda orogen. Densities of
polygons are in kg m-3.
55
between calculated and observed gravity was
achieved. In the second (Fig. 6b), crust with
2 6 7 0 k g m -3 density representing the central
collision complex was enlarged at the expense of an
upper mantle region with 3 1 0 0 k g m -3 density,
whilst maintaining the fit of calculated with
observed gravity. The second model gives an upper
limit for which a reasonable match can be made
with the observed gravity data. In this, crust of the
central collision complex attains a maximum
thickness of nearly 60 km.
Volcanic Arc (TIMOR SP 1700-3200,
DAMAR SP 2350-2900)
At the northern end of both seismic sections, sub-
marine volcanoes associated with the subduction-
related volcanic islands of the Banda arc form
prominent topographic highs. The volcanoes are
said to be built on oceanic crust on the basis of
seismic refraction data (Bowin et al. 1980). Long
sub-horizontal reflectors just below the seabed are
interpreted as delineating lava flows (D. Snyder,
pers. comm., 1995). The TIMOR profile crosses the
main volcanic ridge at SP 2950. Volcanoes at SP
2000 and SP 2350 on the TIMOR profile lie to the
north of the main volcanic ridge and have a steep-
sided, narrow trough between them at SP 2130
(Fig. 3a). This sharp trough lies along strike from
the surface expression of the Wetar Thrust visible
• t:
1::o°o
°
2670
o
2800
2800 50
3300 \,........
lOO
3300
150
3
2OO
3200
250
~ -~o -lOO loo ~o ~ -~o
Distance fromTrench(km)
56 A. N. RICHARDSON 8Z; D. J. BLUNDELL
on sonar imagery (Masson et al. 1991). A south-
dipping reflector reaches the surface at the base of
this trough and can be traced 50 km southwards to
a depth of c. 20 km. Two earthquakes with thrust
mechanisms project on to the TIMOR section
within the horizontal range of this reflector (Fig.
4a) although with significantly greater hypocentral
depths. It may be significant, however, that the
south-dipping nodal planes of both these events
dip at an angle close to the migrated dip of the
reflector. This reflector is therefore interpreted as
evidence for the along-strike extension of the Wetar
Thrust, associated with current movement.
A number of major, south-dipping reflectors are
also present in this portion of the DAMAR section
which extend close to the base of the seismic record
(Fig. 3b). These structures therefore appear to be
the dominant fabric of the upper plate on both
profiles. Some shorter north-dipping reflectors are
present on both profiles but none are as long or as
continuous as the south-dipping reflectors (Fig. 4).
The volcanic arc portion of the DAMAR profile
includes the greater proportion of earthquakes less
than 100 km deep along the transect (Fig. 4b). Most
of these earthquakes have thrust mechanisms, with
one nodal plane which has a south-dipping compo-
nent, and plot close to the major reflectors. It is
not possible to correlate specific earthquakes with
individual reflectors because of the large envelope
of uncertainty in hypocentral locations and the
distance that they have been projected on to the
section. However, the fact that earthquakes with
thrust mechanisms are occurring in this region
does suggest that the south-dipping reflectors may
represent active thrust faults which penetrate the
crust and upper mantle to depths of up to 100 km.
On the TIMOR profile, the crustal thickness
(defined where Win t > 8000 m s-1) varies from
c. 12 km to over 20 km beneath the volcanoes.
This is abnormally thick for oceanic crust, even
including the volcanic edifices. Interval velocities
> 8000 m s-1 are attained at depths of 8-19 km
below sea-level on the DAMAR profile (Fig. 5b)
which implies that the crust is thinner than on the
TIMOR profile but still thicker than normal
oceanic crust.
The backarc region (TIMOR SP 1-1700,
D A M A R SP 2900-3200)
A number of marine geophysical surveys have been
conducted in the backarc region. From Lombok
to Maopora, the backarc region is dominated by
south-dipping reflectors beneath the volcanoes,
sometimes overlain by an accretionary prism
(Silver et al. 1983). Likewise, the south-dipping
reflectors in this part of the BANDA profiles are
interpreted as thrusts because of their appearance
on seismic profiles.
The two most northerly volcanoes on the
TIMOR section, at SP 400 and SP 1200, are
isolated seamounts built on the oceanic crust of the
Banda Sea. The northernmost edifice rises above
sea-level to form the active volcano, Gunung Api.
It is not clear why this volcano is active so far to
the north of the subduction zone. The volcanoes
traversed by the TIMOR profile appear to line up
with Gunung Api and the seamount in between
them (Fig. 1). The NNW-SSE lineation of sea-
mounts suggests that they may have formed over
a 'leaky' transform fault (D. Snyder, pers. comm.
1992). The DAMAR profile passes close to one of
the seamounts at SP 2600 and directly over another
at SP 2800, a few kilometres to the north.
The basin areas in between the volcanoes show
clear sedimentary layering near the surface and
long, gently dipping reflectors down to the depth
of the first multiple (Fig. 3). The velocity pick at
TIMOR SP 1599 in the centre of one of these
basins is interpreted as representing 1840 m of
sediments (cf. oceanic lithosphere Layer 1, Vint =
1900-2400m s-1) and 5860 m of volcanics and
intrusives (oceanic lithosphere Layers 2 and 3,
Vin t - - 4 2 1 7 m s-1) overlying upper mantle material
(oceanic lithosphere Layer 4, Vin t -- 8058 m s-1)
at 7.7 km below the sea-bed. Most velocity picks at
the extreme northern end of the DAMAR profile
have velocities typical of the upper mantle
(Vin t > 8000 m s-1) occurring 3 . 7 4 km below the
seabed. All these velocity structures broadly agree
with seismic refraction data (Bowin et al. 1980) and
are therefore interpreted as representing normal
oceanic crust. The northern part of the DAMAR
line appears to traverse rather rough sea floor.
However, one of the major south-dipping thrusts
reaches the surface in this region and the rough
topography may represent a small accretionary
prism.
Summary
The southern end of both profiles, which traverse
the Australian Continental Shelf, is interpreted in a
similar manner to deep seismic reflection profiles
from other passive continental shelf areas. Well-
stratified sediments overlie a largely reflection-
free upper crustal crystalline basement and a
highly reflective lower crust. This arrangement is
common in continental regions that have undergone
extension.
Structures at shallow depths under the northern
slope of the trench (or the southern part of the
collision zone) dip exclusively northwards and are
interpreted to represent thrusts and tilted sediments
in an accretionary prism. Deeper, north-dipping
 CONTINENTAL COLLISION IN THE BANDA ARC
structures are interpreted to be related to the inter-
face between the lower and upper plates. This is
supported by the extrapolation of this surface to
greater depths where it corresponds well with the
upper surface of the Australian plate as defined
by earthquakes.
The northern part of the collision zone is
dominated by predominantly south-dipping
structures which extend to sub-crustal depth. These
are interpreted as thrusts along which the lateral
shortening and crustal thickening have taken
place. The thrusts are estimated to be no older than
2.4 Ma, the age of initiation of the collision of the
Australian margin with the arc, and seem to be
active at the present day at the northern end of the
collision zone.
A maximum crustal thickness of c. 60 km is
calculated under the topographically highest part
of the collision zone on the basis of calculated
seismic interval velocities. If this crust had been
constructed from rifted and thinned Australian
margin material, the crust may now be more
than double its original thickness which would
imply over 50% (100-150 km) shortening in the
continental portion of the collision zone.
The oceanic crust also appears to be anom-
alously thick beneath the volcanoes on the basis
of calculated seismic interval velocities. It appears
likely that the oceanic crust has, like the con-
tinental crust, been shortened and thickened on
predominantly south-dipping thrusts. The fact that
these major thrusts are visible beneath and north of
the volcanic arc suggests that the collision zone
extends north of the non-volcanic arc and is in total
about 200 km wide, comprising both continental
and oceanic lithosphere, and may have accommo-
dated over 200 km of shortening.
The overall, large-scale character of the collision
zone is dominated by two sets of divergent struc-
N Australian Continental S
Volcanic Arc Timor Micro-Continent AccretionaryPrism ~li,helf
~1~ ~ _ Kisaj~ ~,
~" /~5. / ~i~;~
~ ; ~
Fig. 7. Cartoon structural cross-section of the Banda orogen based on our interpretation of the BANDA deep
seismic profiles. Australian continental shelf sediments are shaded light grey. Highly reflective lower crust is shaded
dark grey.
57
tures. The southern set is related to the subducting
(lower) plate and dips in the same direction as
subduction. The northern set, in the upper plate, is
antithetic to this. Figure 7 is a cartoon to highlight
these primary features.
The deep-seated nature of the orogen
In considering the evolution of the collision
complex as an orogen, three dates are significant:
(1) The Banda allochthon in Timor was apparently
uplifted at c. 37 Ma according to K-Ar dating
(Sopaheluwakan 1990).
(2) Accurate radiometric dating on the Aileu
Formation on the north coast of Timor gives an
age of metamorphism associated with gradual
cooling of 8 Ma (Berry & Grady 1981) and a
Mid-Miocene unconformity has been inferred
across Timor (Audley-Charles 1968).
(3) Pliocene deformation is consistently recorded
by sedimentary sequences elsewhere in Timor
(Audley-Charles 1968) and the start of the
most recent collision has been dated at 2.4 Ma
from evidence in the Timor trough (Johnston
& Bowin 1981).
The extent of crustal material in the central
collision complex is defined by seismic velocity
and gravity modelling. The crustal material has
the form of an inverted triangle about 50 km
(37-60) deep and 125 km (135-160) across on the
TIMOR profile, which has a cross-sectional area of
3750 km 2. This thickness of crustal material could,
conceivably, have been constructed by a number of
means. However, it cannot have been constructed
entirely from sediments accreted from the
Australian plate since 2.4 Ma. Using the plate
velocity vectors in DeMets et al. 1990 (Fig. 1),
iSubd
,.)., ~lt~
ctionTrench , , ~
Moho ~60 km
0 km 1O0
I I I ]
58 A.N. RICHARDSON t~ D. J. BLUNDELL
156 km (65 km/Ma for 2.4 Ma) of convergence
normal to the arc could have occurred. This is an
upper estimate as it does not take account of
decreasing convergence velocity after collision. An
upper estimate of the thickness of the sediments
between the Australian coast and Timor from the
Banda profiles is about 3 km. Assuming that this
entire thickness of sediments is added to the
growing orogen and that strike-slip movement has
not removed material from the normal plane, a
cross-sectional area of 468 km 2 would result
(3 × 156 km). This is far less than the estimated
crustal area of 3750 km 2. A collision event, adding
crustal material to the orogen, must therefore have
preceded the Pliocene collision event.
Nor is it possible that the Banda orogen was
constructed entirely from sediments accreted from
the Australian Plate since 8 Ma. Using the same
argument as above, a sediment volume accumu-
lated over 8 Ma of convergence would amount to
only 1560 km 2 in cross-sectional area, which is still
substantially less than the estimated 3750 km 2 of
crust. There must therefore be an additional piece
of crystalline continental crustal material within the
collision complex making up a major portion of its
bulk. It is thus proposed that a micro-continental
fragment, or perhaps an outer margin 'high', lay
a few hundred kilometres to the north of the
Australian Northwest Shelf and this was incor-
porated into the collision complex before the arrival
of the Australian continental margin. The material
separating the micro-continent from the Australian
craton was probably oceanic in origin but could
have been extremely stretched and attenuated
continental lithosphere. This micro-continental
fragment collided with the subduction zone at
c. 8 Ma in the Miocene. This caused metamorphism
R e f e r e n c e s
AUDLEY-CHARLES, M. G. 1968. The Geology of
Portuguese llmor. Geological Society, London,
Memoir, 4.
1981. Geometrical problems and implications of
large-scale overthrusting in the Banda Arc-
Australian Margin collision zone. In: MCCLAY,
K. R. & PRICE, N. J. (eds) Thrust and Nappe
Tectonics. Geological Society, London, Special
Publications, 9, 407-416.
- - 1986a. Timor-Tanimbar Trough: the foreland basin
to the evolving Banda Orogen. In: ALLEN, P. A. 8£
HOMEWOOD, P. (eds) Foreland Basins. International
Association of Sedimentology Special Publication,
8, 91-102.
- - 1986b. Rates of Neogene and Quaternary tectonic
movements in the southern Banda Arc based on
micropalaeontology. Journal of the Geological
Society, London, 143, 161-175.
- - & HARRIS,R. A. 1990. Allochthonous terranes of the
of the Aileu Formation at the leading edge of the
micro-continental fragment ( ' m e t a m o r p h o s e d
parautochthon' on Timor and Kisar) and folding in
the unmetamorphosed rocks. The micro-continent
was thrust beneath the forearc region, thus uplifting
the Banda Allochthon, and may have induced Late
Miocene inversion structures reported from the
Timor Sea (MacDaniel 1988; Pattillo & Nicholls
1990).
The Australian continental margin arrived at the
collision zone 2.4 Ma ago and started downwarping
into the subduction zone. The continental crust
could not be subducted very far. Subsequently, the
Australian continental margin has acted as a
bulldozer, shortening and uplifting continental and
oceanic parts of the collision zone on deep seated
thrusts dipping antithetic to subduction. The oldest
thrusts are presumably the more southerly ones.
Since then the locus of activity has propagated
northwards in a normal fashion until it has reached
the backarc region where thrusts are active at the
present day. Uplift is currently continuing as
evidenced by several hundred metres of elevation
of Pliocene coral reef terraces on Alor, Atauro and
Wetar (Nishimura & Suparka 1986) and Sumba
(Piranzolli et al. 1991).
The authors wish to thank Dr D. Snyder for his major
efforts in conducting the BANDA survey and supervising
the data processing, and to him and their other BIRPS
colleagues for ready access to the deep seismic reflection
data. BIRPS is supported by the Natural Environment
Research Council (NERC) and all BIRPS data are
available, at the cost of reproduction, from the Marine
Geophysics Programme Manager, British Geological
Survey, Murchison House, West Mains Road, Edinburgh
EH9 3LA. ANR is grateful to NERC for the award of
a studentship during the period of this research.
South West Pacific and Indonesia. Philosophical
Transactions of the Royal Society of London, A331,
571-587.
BARBER, A. J. 1979. Structural interpretations of the
island of Timor. South East Asian Petroleum
Exploration Association Proceedings, 4, 9-12.
1981. Structural interpretations of the island of
Timor, eastern Indonesia. In: BARBER, A. J.
WIRYOSUJONO,S. (eds) The Geology and Tectonics
of Eastern Indonesia. GRDC Special Publication,
2, 183-198.
- - - , AUDLEY-CHARLES,M. G. & CARTER, D. J. 1977.
Thrust tectonics in Timor. Journal of the Geological
Society of Australia, 24, 51-62.
BERRY, R. E & GRADY, A. E. 1981. The age of the
major orogenesis in Timor. In: BARBER, A. J. &
WIRYOSUJONO,S. (eds) The Geology and Tectonics
of Eastern Indonesia. GRDC Special Publication,
2, 171-181.
 CONTINENTAL COLLISION IN THE BANDA ARC
BOWlN, C., PURDY, G. M., JOHNSTON, C., SHOR, G.,
LAWVER,L., HARTONO,H. M. S. & JEZEK,P. 1980.
Arc-continent collision in the Banda Sea region.
AAPG Bulletin, 64, 868-918.
BRADSHAW, M. T., YEATES, A. N, BEYNON, R. M.,
BRAKEL,A. T., LANGFORD,R. P., TOTTERDELL,J. M.
& YEUNG, M. 1988. Paleogeographic evolution of
the North West Shelf region. In: PURCELL,P. G. &
PURCELL, R. R. (eds) The North West Shelf
Australia. Proceedings of the NW Shelf Symposium
1988, 29-54.
BROWN, M. & EARLE, M. M. 1983. Cordierite-bearing
schists and gneisses from Timor, Eastern Indonesia:
P-T implications of metamorphism and tectonic
implications. Journal of Metamorphic Geology, 1,
183-203.
CARTER, D. J., AUDLEY-CHARLES,M. G. & BARBER,A. J.
1976. Stratigraphical analysis of island arc-
continental margin collision in eastern Indonesia.
Journal of the Geological Society, London, 132,
179-198.
CHAMALAUN, E H. & GRADY, A. 1978. The tectonic
development of Timor: a new model and its
implications for petroleum exploration. Australian
Petroleum Exploration Association Journal, 18,
102-108.
CHARLTON, T. R. 1989. Stratigraphic correlation across
an arc--continent collision zone: Timor and the
Australian Northwest Shelf. Australian Journal of
Earth Science, 36, 264-274.
~, BARBER, A. J. & BARKHAM, S. T. 1991. The
structural evolution of the Timor collision complex,
eastern Indonesia. Journal of Structural Geology,
13, 489-500.
CLOWES, R. M., BRANDON,M. T., GREEN,A. G., YORATH,
C. J., SUTHERLANDBROWN, A., KANASEWICH,E. R.
& Spencer, C. 1987. LITHOPROBE - Southern
Vancouver Island: Cenozoic subduction complex
imaged by deep seismic reflections. Canadian
Journal of Earth Science, 24, 31-51.
DE METS, C., GORDON,R. G., ARGUS,D. E & STEIN, S.
1990. Current plate motions. Geophysical Journal
International, 101, 425-478.
FITCH, T. J. & HAMILTON,W. 1974. Reply to Audley-
Charles & Milsom (1974). Journal of Geophysical
Research, 79, 4982.
FOWLER, C. M. R. 1990. The Solid Earth. Cambridge
University Press.
Fucns, K. 1969. On the properties of deep seismic
reflectors. Zeitschriftfiir Geophysik, 35, 133-149.
HAILE, N. S., BARBER, A. J. & CARTER, D. J. 1979.
Mesozoic cherts on crystalline schists in Sulawesi
and Timor. Journal of the Geological Society,
London, 136, 65-70.
HAMILTON,W. 1979. Tectonics of the Indonesian region.
US Geological Survey Professional Paper, 1078.
HARRIS, R. A. 1989. Processes of aUochthon emplace-
ment, with special reference to the Brooks Range
Ophiolite, Alaska and Iimor, Indonesia. PhD thesis,
University of London.
1992. Temporal distribution of strain in the active
Banda Orogen: a reconciliation of rival hypotheses.
Journal of SE Asian Earth Sciences, 6, 373-386.
JOmqSTON, C. R. & BOWIN, C. 1981. Crustal reaction
59
resulting from the mid-Pliocene to Recent
continent-island arc collision in the Timor region.
Australian Bureau of Mineral Resources Journal
of Geology and Geophysics, 6, 223-243.
JONGSMA,D., WOODSIDE,J. M., HUSON,W., SUPARKA,S.
& KADARISMAN,D. 1989. Geophysics and tentative
Late Cenozoic seismic stratigraphy of the Banda
Arc-Australian continent collision zone along three
transects. Netherlands Journal of Sea Research, 24,
205-229.
KARIG, D. E., BARBER, A. J., CHARLTON, T. R.,
KLEMPERER,S. & HUSSON~,D. M. 1987. Nature and
distribution of deformation across the Banda
Arc-Australian collision zone at Timor. Geological
Society of America Bulletin, 98, 18-32.
KLEMPERER, S. L., HOBBS, R. W. & FREEMAN,B. 1990.
Dating the source of lower crustal reflectivity using
BIRPS deep seismic profiles across the Iapetus
Suture. Tectonophysics, 173, 445-454.
MACDAmEL, R. P. 1988. The geological evolution
and hydrocarbon potential of the western Timor
Sea region. Australian Petroleum Exploration
Association Journal, 28, 270-284.
MASSON, D. G., MILSOM, J., BARBER, A. J., SIKUMBANG,
N. & DWlYANTO,B. 1991. Recent tectonics around
the island of Timor, eastern Indonesia. Marine &
Petroleum Geology, 8, 35-49.
MATTHEWS, D. & THE BIRPS GROUP. 1990. Progress in
BIRPS deep seismic reflection profiling around the
British Isles. Tectonophysics, 173, 387-396.
MCCAFFREY, R. 1988. Active tectonics of the eastern
Sunda and Banda Arcs. Journal of Geophysical
Research, 93, 15 163-15 182,
1989. Seismological constraints and speculations
on Banda Arc tectonics. Netherlands Journal of
Sea Research, 24, 141-152.
MOONEY, W. D. & BROCHER, T. M. 1987. Coincident
seismic reflection/refraction studies of the
continental lithosphere: a global review.
Geophysical Journal of the Royal Astronomical
Society, 89, 1-6.
MOORE, J. C., DIEBOLD, J., FISHER, M. A., SAMPLe, J.,
BROCHER, T. ET AL. 1991. EDGE deep seismic
reflection transect of the eastern Aleutian arc-trench
layered lower crust reveals underplating and
continental growth. Geology, 19, 420-424.
NISHIMURA, S. & SUPARKA,S. 1986. Tectonic develop-
ment of East Indonesia. Journal of SE Asian Earth
Science, 1, 45-57.
PATT~LLO, J. & NlCnOLLS, P. J. 1990. A tectonostrati-
graphic framework for the Vulcan graben, Timor
Sea region. Australian Petroleum Exploration
Association Journal, 30, 27-51.
PIRANZOLLI, P. A., RADTKE, U., HANTORO, W. S.,
JOUANNIC, C., HOANG, C. T., CAUSSE,C. & BOREL
BEST, M. 1991. Quaternary raised coral reef terraces
on Sumba Island, Indonesia. Science, 252, 1834-
1837.
POWELL, D. E. 1982. The North West Australian
continental margin. Philosophical Transactions of
the Royal Society of London, A305, 45--62.
PRICE, N. J. & Atn)LEY-CHARLES, M. G. 1983. Plate
rupture by hydraulic fracture resulting in over-
thrusting. Nature, 306, 572-575.
 60 A. N. RICHARDSON •
- - & -- 1987. Tectonic collision processes after
plate rupture. Tectonophysics, 140, 121-129.
RESTON, T. J. 1987. Spatial interference, reflection
character and the structure of the lower crust under
extension. Results from 2D seismic modelling.
Annales Geophysicae, 5 (ser. B), 339-48.
RICHARDSON, A. N. 1994. Lithospheric structure and
dynamics of the Banda Arc, Eastern Indonesia. PhD
thesis, University of London.
SERPA, L. 8z VOOGD,B. DE 1987. Deep seismic reflection
evidence for the role of extension in the evolution of
continental crust. Geophysical Journal of the Royal
Astronomical Society, 89, 55-60.
SILVER, E. A., REED, D., MCCAFFREY,R. & JOYODIWIRYO,
Y. 1983. Back arc thrusting in the eastern Sunda
Arc, Indonesia: a consequence of arc-continent
collision. Journal of Geophysical Research, 88,
7429-7448.
SOPAHELUWAKAN, J. 1990. Ophiolite obduction in the
Mutis Complex, Timor, eastern Indonesia: an
D. J. BLUNDELL
example of inverted, isobaric, medium-high
pressure metamorphism. PhD thesis, Free
University, Amsterdam.
VAN DER WERFF, W., KUSNIDA, D., PRASETYO, H. &
WEERING, TJ. C. E. VAN 1994. Origin of the Sumba
forearc basin. Marine and Petroleum Geology, 11,
363-374.
YON DER BORCH, C. C. 1979. Continent-island arc
collision in the Banda Arc. Tectonophysics, 54,
169-193.
WANNER,J. 1913. Geologie von West Timor. Geologische
Rundschau, 4, 136-150.
WARNER, M. 1990. Basalts, water, or shear zones in the
lower continental crust? Tectonophysics, 173,
163-174.
WOODSIDE, J. M., JONGSMA,D., THOMMERET,M., STRANG
VAN HEES, G. & PUNTODEWO. 1989. Gravity and
magnetic field measurements in the eastern Banda
Sea. Netherlands Journal of Sea Research, 24,
185-203.
 Geophysical evidence for local indentor tectonics in the
Banda arc east of Timor

D A V I D B. S N Y D E R 1, J O H N M I L S O M 2, & H A R D I PRASETYO 3
1BIRPS, Bullard Laboratories, University of Cambridge, Madingley Road,
Cambridge CB3 0EZ, UK
2 Department of Geological Sciences, University College London, Gower Street,
London WC1E 6BT, UK
3 Marine Geological Institute, J1. Dr. Junjunan 236, Bandung 40174, Indonesia

Abstract: Two deep seismic reflection profiles and gravity measurements collected across the
Banda arc east of Timor help to characterize crustal and uppermost mantle structures in the region
where arc-continent collision is thought to be furthest advanced. Reflectors beneath the Sahul
Platform of the Australian shelf indicate geometries consistent with older extensional rift
structures overprinted by more recent horizontal shortening. To the north, the negative Bouguer
gravity feature associated with the southern parts of the accretionary complex is unusually broad
and deep. Post-collisional sediments are thin where this feature is crossed by the seismic lines,
implying that older sediments or the crust are anomalously thick. Still further north, the forearc
basin is notably narrow near eastern Timor and contains few sediments, most undeformed. The
backarc region to the north is remarkable for the presence of a N-S trending line of seamounts
culminating in an active volcano, Gunung Api, which is situated 400 km above the Wadati-
Benioff zone. Reflection profiles across and along the seamount chain indicate underplating
at Moho depths and a fault which appears to have acted as a conduit for basalt eruptions at the
surface. Collectively, these observations imply anomalously thick and bouyant crust beneath
the Banda arc east of Timor, and suggest two possible causes. Either a local promontory in the
irregular boundary of the Australian craton was underthrust beneath the volcanic arc and forearc
to 50-70 km depths or a Palaeozoic basin similar to the nearby Bonaparte Basin was underthrust
and its former crustal structure inverted and thickened to form the buoyant crust. The new
seismic reflection data help to locate the anomalously thick crust implied by gravity anomalies
by defining the leaky fracture zone containing Gunung Api. It is inferred that the fracture
propagates "ahead of the indenting wedge of underthrust, thickened Australian crust.

The Banda arc between Indonesia and Australia
possesses some of the best examples of features
associated with the collision of an arc and a
continental margin (e.g. Hamilton 1979, 1988;
McCaffrey 1988), and also displays distinctive
segmentation along strike (Silver et al. 1983; Karig
et al. 1987; Jongsma et al. 1989; Harris 1991). A
pair of recent deep seismic reflection profiles were
sited across one of these distinctive segments,
immediately to the east of the island of Timor, in
order to better characterize structural geometries
throughout the lithosphere (Fig. 1). Focal mech-
anisms of large shallow earthquakes (McCaffrey
1988) and shallow seismic reflection profiles
(Silver et al. 1983) and recent GPS measurements
(Genrich et al. 1994) indicate that the Wetar thrust,
a northward verging thrust on the northern edge of
the inactive volcanic arc (Fig. 1), today absorbs
most of the 75 mm a-1 (DeMets et al. 1990) relative
convergence between Australia and the Banda Sea.
In contrast, shallow seismic reflection profiles
within the Timor Sea show reflector geometries in
Mesozoic and Cenozoic sediments that are typical
of active subduction zone trenches worldwide
(Karig et al. 1987). The Timor trough cannot
simultaneously absorb plate convergence and
transfer the motion of Australia to the volcanic
arc. Loci of horizontal shortening have probably
migrated northward across the 2 0 0 k m wide
convergence zone during the past 2 Ma, with
specific faults, active for short periods, acting as
mechanisms of strain partitioning (e.g. McCaffrey
1992, 1996).
A series of balanced cross-sections of the
accretionary complex showed that m a x i m u m
horizontal shortening, and hence the most evolved
arc-normal convergence, has occurred in eastern
Timor (Johnston & Bowin 1981; Harris 1991).
Eastern Timor lies midway along the inferred
convergence margin of the Australian continent
with SE Asia (Fig. 1). Low 3He-4He ratios in
volcanic rocks east of Flores were interpreted as
due to the subduction and melting of continental
crust at c. 150 km depths whereas to the west the

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, 61

Geological Society Special Publication No. 106, pp. 61-73.
62 D.B. SNYDERETAL

• ~ .. :,
~ ~ ~ " , ,IRIAN JAYA

5o r,L- KAIIS. /
. ,, --)

xt, ( J / SEA

I
~ AUS
3 ~
ff---- 200km
10"

F f,,,I P
/
f /

/
,"it" ; L
15"S
t
/ //
,
q
iiiii!i!!!!iii
......,.....o,..,.,..... •

,J C ii~iiiiii!!::!i~ii::~i!iii~i~
/ AUSTRAUA
/
I ,r I '~ I I
120* 130*E

Fig. 1. Regional location map showing the Archaean blocks on the Australian mainland (shading: ST = Sturt,
K = Kimberley), principal landmasses, the 200 m bathymetric contour (short dash) that approximately defines the
limit of the continental shelf, the 3 km contour (long dash) that approximates the limits of oceanic crust in the Indian
ocean, the new seismic reflection lines (heavy solid lines), older refraction profiles (medium solid lines) with crustal
thickness indicated in kilometres, and the 100, 300, and 500 km contours of the Wadati-Benioff zone (Cardwell &
Isacks 1978). Large solid dots indicate the transition from subducted continental crust to subducted oceanic crust
(to the west) within the subduction zone as determined from He isotope analysis of erupted lavas (Hilton et al. 1993).
The vector labelled AUS---~SEA represents the local relative convergence vector between Australia and Eurasia
(DeMets et al. 1990). C = Canning Offshore Basin, L = Londonderry High, A = Ashmore Platform, V = Vulcan
sub-basin, P = Petrel sub-basin of the Bonaparte Basin, S = Sahul platform, M = Malita Graben, MS = Money Shoal
Graben, B = Browse Basin, and WT = Wetar Thrust.

He source was inferred to be subducted oceanic to the east of eastern Timor, the Wadati-Benioff
(MORB) rocks (Hilton et al. 1992). The change in zone (Fig. 1) curves sharply back onto itself to the
isotope ratios coincides with strike-slip faults at the N W (Cardwell & Isacks 1978). Here the gross
surface (Breen et al. 1989) and crustal-scale fault Australia-SE Asia convergence vector (DeMets
planes defined by micro-earthquakes (McCaffrey et et al. 1990) is almost parallel to the volcanic arc,
al. 1985) which are co-linear with the 3 k m bathy- i m p l y i n g largely strike-slip d i s p l a c e m e n t s on
metric contour defining the N W margin o f the structures parallel to the arc (see also McCaffrey
Australian continental shelf (Fig. 1). Thus defined, 1996).
the western edge of the subducted continental If eastem Timor represents the most evolved
margin lies c. 300 k m west of the study area. A segment of the Banda arc, local anomalously
similar distance along the strike of the volcanic arc thickened crust or thick crust inherited from a
 LOCAL INDENTOR TECTONICS IN THE BANDA ARC
former promontory on the Australian shelf margin
may be the cause. The current complex tectonic
geometries associated with Irian Jaya prevent
precise reconstruction of the former continental
margin east of present-day Tanimbar, but New
Guinea is thought to represent a large indentor
wedge of Australian continental crust plowing into
and bending the margin of SE Asia (Pigram et al.
1989). The east Timor region may contain a
smaller promontory on the western margin of this
larger indentor, but one that produced significant
localized shortening across the Banda arc.
Evidence for a promontory which has already
subducted or underthrust the Banda arc must be
sought using reflection and refraction profiles,
gravity field anomalies and bathymetric observa-
tions. This paper describes and discusses this
evidence. The variability in crustal structure
observed in the unsubducted part of the Australian
shelf south of Timor serves as the starting point,
and it suggests similar complexity in subducted
parts. Inversion of subducted sedimentary basin
structures and thickening of sub-basin crust
previously thinned during Devonian-Permian
rifting provides an additional mechanism that can
produce anomalously thick and buoyant crust east
of Timor.
Structures of the northern australian
shelf
The Australian continental shelf south of Timor
is covered by a relatively dense network of deep
seismic reflection profiles plus gravity, magnetic
and bathymetric measurements (e.g. O'Brien et al.
1993; AGSO North West Shelf Study Group 1994).
Sedimentary basin structures are well understood,
but crustal thickness is poorly constrained. The
nearby Australian mainland contains two large
Archaean cratons, the Kimberley and Sturt blocks
(Fig. 1). Immediately offshore between these two
blocks and continuing northwards to the Malita
Graben lies the Bonaparte basin with its Petrel and
Vulcan sub-basins. It, and the Canning Basin to
the southwest, form composite Late Devonian to
Early Carboniferous intra-cratonic rift basins with
axes perpendicular to the Australian margin and
associated with NE-SW extension (AGSO North
West Shelf Study Group 1994).
Mid-Carboniferous to Early Permian rifting that
led to break-up of Gondwanaland and separation of
'Sibumasu' (Sengor 1987) from NW Australia and
the opening of 'Neo-Tethys' (Veevers 1988), also
initiated a series of generally NE-trending depo-
centres which constitute the Westralian Superbasin
of Yeates et al. (1987). The superbasin's internal
platforms and basins are orientated parallel to the
continental margin and thus angularly super-
63
imposed on the older radial rift basins. In the study
area these later features include the Sahul and
Ashmore Platforms, which show little evidence of
Mesozoic extension, and the Malita, Vulcan and
Browse Basins (Fig. 1). The start of the break-up
phase and the subsequent development of the
continental margin was recorded as a major
Lower-Middle Carboniferous unconformity. Mid-
Carboniferous to Lower Permian basin fill may
actually be a sag-phase sequence deposited
following break-up and the 100-400% extension
thought to have characterized this rift episode
(Etheridge & O'Brien 1994).
The Sahul Platform occupies c. 30 000 km 2 of
the Australian shelf immediately south of eastern
Timor. The Malita Graben separates it from off-
shore parts of the Sturt block of the Australian
mainland (Fig. 1). Seismic reflection profiles and
drill holes indicate that c. 6000 m of Carbon-
iferous(?) and as much as 10 000 m of later sedi-
ments were deposited in the Malita Graben,
confirmed by a core of 3300 m of post-Permian
sediments on the Sahul Platform (AGSO North
West Shelf Study Group 1994). The uniformity
of sediment cover on the Sahul Platform suggests
thermal subsidence rather than subsidence syn-
chronous with rifting and implies little thinning
of this crustal block. To the north, very young
structures associated with convergence tectonics
overprint older extensional structures (Figs 3 & 4).
Although no evidence exists on seismic reflection
sections to indicate thicknesses of post-Permian
sediments as great as those observed in the Malita
Graben, reflectors beneath the Jurassic and Permian
levels indicate a possible older basin north of the
Sahul Platform and currently underlying the Timor
trough (Figs 3 & 4).
Crustal thickness estimates
In contrast to the wealth of information available
about the upper crustal and basin structure of the
Australian continental shelf, its crustal thickness
and that of the Banda arc itself are less well docu-
mented. Three refraction profiles provide general
estimates for the thickness of Australian con-
tinental crust in this area, but these show vari-
ability and uncertainty commensurate with the
basin structure described above (Jacobson et al.
1979). The three profiles were orientated parallel to
the shelf margin, and two lie in the Timor trough
(Fig. 1). Crustal thickness estimates range from
31-40 km within an area of 500 000 km 2, and
profiles 150kin long show crustal thickness
variations of 4-8 km (Bowin et al. 1980). A fourth
profile within the Banda Sea, east of Wetar (Figs
1 & 2), indicated a typical oceanic crustal thickness
of 11 km (Bowin et al. 1980).
64 D. B. SNYDER ET AL.

GUNUNG
4 API
B A N D A SEA

I
I
I 5O IO4
i

i©
!
t-
,~.~.%
.~-""~
"1
i I

//

-/
WETAR ....~ " /
BANDA "". l /
I
!

i
\ KISAR
SER~ATA

:I
I
I
.500o I ARC
I
BANOA I
I
I
I , lOOO
l
+
127* l
+9"s
! 29" E

0~~ ~i"
searnount

island D
~, l q # '

Troubadour

Fig. 2. Location of major known submarine features and the new seismic lines TIMOR, API, and DAMAR. Dashed
lines indicate older scientific expeditions in this area. The Troubadour well (solid dot) on the Australian shelf margin
(Australian Geological Survey Organisation, unpublished data) is located at SP 7995 of the TIMOR line. Onland
areas shaded in black contain exposures of ophiolites. M 12 is a refraction profile (Bowin et al. 1980).

By themselves, these thickness estimates are too
sparse to characterize the crust, but crustal thick-
nesses can be extrapolated or inferred over the
region by using reflection profiles. Penetrative
reflectivity within extended continental margins
generally disappears below the Moho (e.g.
Matthews 1986). The intersecting reflection and
refraction profiles in the study area (Fig. 1) demon-
strate that this relationship applies here, at least
locally. By assuming that it applies for the entire
northern Australian shelf, a gradual eastward
thickening of the crust from about 30-40 km is
inferred between eastern Timor and Tanimbar.
This pattern of crustal thickening is not locally
correlated with the known basin structure unless the
Sahul Platform continues to the NE as argued by
Charlton et al. (1991) using stratigraphic criteria.
Unrecognized basins, similar to the Malita Graben
 LOCAL INDENTOR TECTONICS IN THE BANDA ARC 65

N S

o-- _ .
-- ,. -" ~ _ -_-_~.-. r . . . . . . . . . . --_- ~:..~-~. : . ~- • . -~
"-~.-~÷-=~.-.-:-~----~"-~_--~2=_-~-.~.'-~~ ,, ,, ~ ~ . -~-~-~--~: -~.~.-~_-_~.~.._~

~~~--~,--: ~ ~ ~ ~

SP 6 0 5 0 6500 7000 7500 7975

Fig. 3. Migrated seismic section from the TIMOR profile in the vicinity of the Timor trough. Post-Permian shelf
sediments that can be tied to the Troubadour well (top right) are clearly seen beneath the sea floor on the inner trough
wall, more recent undeformed sediments lie in the bottom of the trough. One arcuate reflector is indicated (arrow at
SP 6500) in the uppermost basement beneath the shelf sediments; it may represent an unrecognized Devonian-
Carboniferous basin, or side-swipe from a fault plane parallel to the seismic profile and bounding such a basin. Moho
lies at about 15 s beneath the trough. Numerous lower crustal reflectors occur between 8-12 s on the right half of the
figure; these appear offset with a thrust sense (half arrows) along south-dipping thrust shear zones near SP 7250.

0 N S

s - ~ _ . ~ ~ -
~ ~ ~ ~ ~ :

"~ 10 ~ ~ ~ ' ~ ' ~ ; ~ ~ ' = ~ - ~ ~ - " ~ - ~ " ~ - -

E
-- ~..-~ax.~..~.-~-..~3.,¢.,~~ - ~.~,,...::-.~_~--~:---z~g.r~,.~..-r,L.~,~,.~7.~;~---.: ~-~-.-,-~.-~.~.~-~.~=,~.~-.~.-.~_-~

sP 755 500 250 120

Fig. 4. Migrated seismic section from the DAMAR profile in the vicinity of the Timor trough. Post-Permian
shelf sediments are clearly seen near the sea floor on the inner trough wall, folded beneath the trough floor. Two
northward-dipping reflectors indicated (arrows) in the uppermost basement beneath the shelf sediments are similar
to the one described in the previous figure. Moho lies at 13-15 s.
66 D. B. SNYDER ET AL.

or Petrel sub-basin, lying north of the Sahul
Platform and beneath the Timor trough may explain
these relationships, but cannot readily account
for the observed regional along-trench crustal
thickness variations. The AGSO North West Shelf
Study Group (1994) note that one important conse-
quence of the concentration of Carboniferous-
Permian extension in the lower crust is that exten-
sional structures appear only as subtle reactivation
features within the upper crust and basin fill, often
displaced or offset from areas of lower crustal
thinning. Parts of the Australian shelf overthrust by
the Timor accretionary prism may therefore contain
additional NE-trending sag basins like the Malita
Graben, or more probably, northern parts of the
north-trending rifts associated with the Petrel
sub-basin.
Gravity field observations and modelling
The Australian continental shelf south of Timor is
now covered by a relatively dense grid of marine
gravity measurements. To the north, coverage is
less complete but recent surveys onshore have
mapped the gravity field on nearly all the outer arc
islands between Timor and Tanimbar (Richardson
1993). The gravity field of the Banda Sea region
is dominated by a steep gradient which forms a
continuous arcuate feature running from northern
Timor past Tanimbar and the Kai Islands to
southern Seram (Bowin et al. 1980). Typically,
values of Bouguer gravity increase from the
slightly negative, as in the south coast region of
Timor, to greater than +150 mGal as at points on
the north coast of Timor (Fig. 5).
Attempts to model this gradient used profiles
crossing the arc at various locations. Of these, the
model of McBride & Karig (1987) incorporates
measurements from an anomalous region of
western Timor where contour trends are locally
transverse to, rather than parallel to, the strike
of the arc, and the model of Milsom & Audley-
Charles (1985) is unsatisfactory because of the lack
of seismic control. The model of Jongsma et al.
(1989; see also Richardson 1993) (Fig. 6a) was

Fig. 5. Bouguer gravity map of the study area. Reduction density is 2.2 Mg m-3. Contour interval is 10 mGal. Solid
lines indicate the (roughly N-S) deep reflection profiles and (roughly E-W) modelled transects. Note the particularly
negative values associated with the accretionary complex between the seismic profiles south of the island of Moa and
the north-south gravity 'notch' over the shelf immediately to the south. Stippled borders are located near inflection
points in east-west transects of the gravity field and define the possible areal extent of an anomalous low-density
body in the crust.
 LOCAL INDENTOR TECTONICS IN THE BANDA ARC 67

based on gravity, magnetic and seismic data along

~o0~,,_
L ~
I¢~. II~I: ouguerAnomaly C
A'J
q
a profile which nearly coincides with the TIMOR
line (Fig. 2). The model shows thick continental
crust in the downgoing plate almost as far north
as the north coast of Timor, with low Bouguer
anomalies in the south produced by a thick, low-
2°°1 ~ I
Wetar TIMOR Australianshelf density accretionary complex and the high values
0 in the north due to the presence of what was
originally sub-oceanic crust and mantle beneath the
forearc basin and the volcanic arc. Subduction,
• ee • Moho rather than limited overthrusting, is hypothesized
at the Wetar thrust and a mantle block is therefore
100 ~ • %1 • shown as virtually isolated by the action of the
opposing subduction zones (Fig. 6a).
"f": .. Of the more localized features of the gravity
field, one of the most remarkable is the negative
200 free-air and Bouguer anomaly feature centred
directly south of the outer arc island of Moa (Figs
• : y" o ~o ,oo~,, 2 & 5). This feature was first indicated by measure-
../ • ~ oontinentacrust
ments on a few widely spaced lines of marine
survey, but was later substantiated by free-air
300 gravity anomalies derived from satellite altimetry
/ ~ oceanic oust
x and measurements on Moa and the adjacent islands.
~ aomet=onarycomplex
The lowest Bouguer values occur 40 km to the
volcanic arc
south of the islands and the free-air minimum lies
(modified lrom ~ mantle further south in deeper water. Bouguer anomalies,
Woodside et al 1989) • earthquakes used in preference to free-air anomalies in Fig. 5,
(a) are less influenced by bathymetric variations and
therefore better at resolving deep structures. On
this map the Moa feature appears nearly elliptical,
.,oL
> C , C'
-I 40 km across from north to south (measured
I ¢''~ ~ observed,ons h e l f ~ ~"=" / between half amplitude points) and 120 km from
~°I-~ ~ ~ _.- --:~-I east to west, with an amplitude of c. 70 mGal. It
=< I~ ~ ' ~ ~ ~ ~ - ~ / is superimposed on a more elongated, lower
r calcula~ / observed~,onprisrn 1 amplitude feature that just reaches into the southern
part of eastern Timor. Directly south of this 'Moa
°-'t-
WEST
.... " - - / "
EAST
-I anomaly', the Bouguer anomalies are 30-50 mGal
o . . . . . . . . . . . . . . . . . . .
less than on nearby parts of the Australian shelf
margin, and NNW of the feature Bouguer anom-
alies are 30-50 mGal less than in other parts of
~2o
41-
~ 3.2
t~ the forearc basin (Fig. 5). Collectively, these form
a NNW trending band of lower gravity.
60~- MANTLE The two deep seismic profiles cross the long axis
of the 'Moa anomaly' on either side of its centre, at
; ,o 8'o ," ,6'° 20'0 210 2'° 3'° about the location of the two half-amplitude points.
(b) D,s,ANcE. K,
The profiles also lie near the base of local east-west
gravity gradients on the Australian shelf (Fig. 5).
Fig. 6. (a) Cross-strike (N-S) gravity profile and cross-
section (Woodsideet al. 1989). Location coincides with Other than the discordantly dipping reflections
the western deep seismic profile shown on Fig. 5. indicated by arrows in Figs 3 & 4, neither seismic
(b) East-west profiles across the 'Moa anomaly' and section appears to show basins or shelf features
along the shelf margin. The simplistic gravity model significantly different from those imaged on
shows that density contrasts at two crustal levels, one numerous unpublished seismic profile sections
associated with a sedimentary basin and one with further east, where the regional gravity minimum
thickened lower crust, can produce the observed gravity has lower amplitude. These discordant reflectors
anomalies. Profiles are located on Fig. 5, on which have geometries consistent with out-of-plane
stippling indicates the possible extent of the anomalous
reflections from dipping reflector surfaces (faults?)
crustal structure modelled here. Vertical bars on the
observed curves indicate where the gravity profiles striking nearly parallel to the seismic profiles and
intersect. offset by 10-15 km.
68 D . B . SNYDER E T A L
Two gravity profiles orientated along structural
strike (Figs 5 & 6b) illustrate the similarity in the
along-strike gravity trends along the prism and
shelf by eliminating the more dominant north-
south gravity gradients associated with the sub-
duction zone. One simple model indicates that a
combination of anomalous, low densities at the
base of the crust and at the top of basement can
successfully match observed Bouguer values
(Fig. 6b).
Three sources of low density material are
considered here. The first possibility, of a deep
post-orogenic slope basin of the type known to
occur south of Timor (e.g. Karig et al. 1987), is
not supported by the seismic stratigraphy inferred
from the seismic profiles (Fig. 3). Similarly, no
direct evidence exists for a significantly thicker
accretionary complex east of Timor, nor for anom-
alously lower density material within it, although
Woodside et al. (1989) showed it as a 50 km thick
tapering wedge. Thickening of the basement of
the downgoing crustal layer by thrusts does seem
plausible, based on duplex thrusts suggested by
lower crustal reflector geometries (Figs 3 & 4) and
on the variations in crustal thickness reported from
the NW shelf of Australia (Bowin et al. 1980;
O'Brien et al. 1993). Because thickening would
effectively add low density material at Moho depth,
it would have to be very local to produce a gravity
field anomaly as localized as that near Moa.
If thickening of the downgoing crustal layer is
a valid explanation, syn-orogenic thrust duplexing
seems more probable than isostatically uncom-
pensated variations in crustal thickness that existed
prior to collision. Elsewhere on the Australian shelf
such variations are compensated. Thickening due
to thrust duplexing could be accomplished as an
independent tectonic mechanism or by restoration
of crust previously thinned during the development
of Palaeozoic sag or rift basins to its normal thick-
ness whilst preserving or thickening the overlying
basin strata.
The resulting bi-level thickening is illustrated
by the density model shown in Fig. 6b. Inherited
crustal structure localizing and intensifying the
thickening effect of syn-orogenic thrusting seems
the most probable mechanism to produce the
anomalously low Bouguer anomalies associated
with the Australian shelf margin in this region.
Forearc crustal structures
The forearc within the study area is unusual in the
narrowness of the forearc basin and in the presence
of the island of Kisar. Between the two deep
seismic profiles the forearc basin is typically only
15-20 km wide and contains only tens of metres
of sediment (Fig. 7). The basin appears to not have
acted as a depocentre, suggesting that it has been
structurally higher than other parts of the forearc,
such as the Savu Sea, enabling along-arc ocean
currents to keep the sea floor relatively free of
sediments during the past few million years.
Recent mapping on the island of Kisar has
revealed, from south to north, low-grade meta-
morphic rocks such as greenstones, amphibolites,
and bands of thick mylonitized quartzites inter-
layered with greenstones and high-grade pelites
(Dropkin et al. 1993; Richardson 1993). Some of
these rocks bear a strong similarity to the Aileu
(Grady & Berry 1977) or Lolotoi/Mutis Complex
Formations on Timor and Moa (interpretations of
Richardson 1993 and Dropkin et al. 1993 respec-
tively), and are interpreted as dismembered ophio-
lites thrust onto thinned Australian shelf or oceanic
crust or the volcanic arc.
Kisar thus is related to the accretionary complex
rather than to the volcanic arc, which emphasizes
the unusual narrowness of the forearc basin in this
segment of the Banda arc. Southward dipping
reflections observed on the deep reflection sections
(Fig. 7) may represent backthrusts that have trans-
lated the forearc (accretionary prism) northward
over the forearc basin and thus narrowed it. The
entire forearc block from Kisar to Sermata may
represent a large rectangular klippe (Fig. 2).
Backarc features: a seamount chain
including Gunung Api
Anomalous features also occur north of this
segment of the volcanic arc. Although the
magnetic patterns are very poorly constrained,
Lapouille et al. ( 1 9 8 5 ) suggested transform offsets
of c. 50 km of anomalies identified as M7-M10
along a NW-SE orientated line passing through
Gunung Api. More recent work (e.g. Rehault, pers.
comm. 1995) suggests that the anomalies have been
incorrectly identified, but the evidence for offset
remains valid. The new seismic reflection profiling
has revealed a number of previously undocumented
sea-floor volcanoes (seamounts) along the TIMOR
line and coincident with this inferred transform
fault. The northern part of one of these seamounts
is shown in Fig. 8. The API cross-line (Fig. 2)
showed that between the seamounts lies a low ridge
of extruded basalt separating two levels of the
Banda Sea ocean floor (Fig. 9). Sediment thick-
nesses appear to be similar on both sides of the
ridge, which therefore coincides with a down-to-
the-east offset of 350-400 m in the sea-floor base-
ment. Arcuate reflectors within the ridge suggest
either folds and thrusts or primary extrusive struc-
tures among the basalt flows. Gunung Api, the
seamounts and the inferred ridge basalt flows
 LOCAL INDENTOR TECTONICS IN THE BANDA ARC 69

-0

u0

o5
E
b-

10
SP 2135 2000 1800 1600

Fig. 7. Part of the DAMAR seismic reflection section within the forearc basin. This section was migrated but has not
been depth converted. Approximately 200 m of relatively undeformed sediment lie within the 10 km wide forearc
basin at the centre of the section. Greater sediment accumulations can be observed (arrows) in slope basins both north
and south of the forearc basin. Curved arrows indicate possible backthrust.

N S
M12 2 -" ~--
-~ ~ G . . . . GAPI 5 K API LINE

~.sJ ~!. . . . . . . . . . . . . . . . . . . . . . . . . . . m 1.48

--I
" I ~ ~ , .z , "' ~ ~ - ~ ~ , - . ~~ . ~' . ~% " -""-
~ .... ~ ......... ~

~.~ .. ~.,~:;,.~~.~. ,, .

/ .~.~'~-
/ ~-~-~
I-I 0 .~e~- 3.6

~..-:-~.
r~ ~,~,~ '//YtiI~'/Z" ~,,,,'IZ~/., ".~, %,"..¢:'~3"~,,\~,~\3~&~,',,,*\'~;~,~..~,'~ ~ ,.:~,~:~,,,, ~k-~"-'.:
SP 500 1 1200

Fig. 8. Part of the TIMOR seismic reflection section just south of Gunung Api. This section was migrated and then
depth converted using the velocities shown at the right in km s-1. In this depth section the Moho appears as a bright
reflection (arrow) at a depth of 10 km where this line crosses the API line. If refraction velocity function M12, of
Bowin et al. (1980), shown for comparison at the left, was used instead for the depth conversions, the bright
reflection occurs at 14-15 km depths, and no obvious feature is associated with the Moho. Numerous convex-upward
reflections (M) are migration noise due to the incomplete removal of the sea-floor multiple before migration and
should be ignored.
70 D . B . SNYDER E T AL.

TIMOR LINE , [E

w .,. • ~..~.,:;
!
5~ , 5KM ~ I ~ ! ~
. . . .:~:~"~
' ! ~ ~~ N ~

>. :~-'~.=_-:-.:,~--=-7-.'~'~:Y~,-.?]-=':":'-~:.:.'":L':.'I)Z~.:"T::. ' ......7-:~...:::"17 Z ....... ,7.;. ::7":d.~'::-:.-::-'-2-~---".~,~-"÷i'.i-':!~-"-~==:?~.;'~-~Q~~ ~

ZZ'_,;.::'~L~";:-~" - ~:-"-~-:,~ - ~"~:->.--.:,:'-':.--,~.;Z :- :- -, " , _ - -:: .7: ": -:"-::::~"£ ~ ~-~_~7"'_':~-=.--'S7 ~ ' 7~~: :c'~L:--:L~-~:Cg=~
".7-,~'-"-"7.'%-:-
.:.."
"=r_.'--..::.-'.." ~ ..... . _ ~ . , . . , ; . . . 7........
, - ' , . ~ ' . . ; . . . - ",~',,.-- .'...~.,,.' ,~ . . . . ., " ~ . , ,- ....
: . . . . . . . a . - .~:T,~;:..:._~:...,~_.~..~.:~:.~:.:::,~_:~7.,.~_~,_7.
*.. ...--_- :. - . - ~ : , * . . . - - ' r . . , : ' , -
~?.~:"
. . ' - : : . ; --...=~.-,,:.-'.-,~.,.-...'2".,-a,: ~. .L'-
i'- 7='-.'7-'..,: .'~:,,.-'::/7:". ,-" - Z "--' .:,,7"--':" ~. " ..7. . . . . "." "..:: 5 "-- -::.'----..-' "~:,,:; . . . . .' ',.7::.;~,~,, . q ' ~ . ~ " - . - - : ' 2 . ~ - : " .".':,~:"_-'-Z ~iee~'-L'.~,~..~Tr.~£"~'=.:~:~"~.:'~

7~--=--""- ' :!:-~'"5::";"".7.."-:~::-.~',,.~-.~::':.--=-::?:.:: -:~=::..:.:i::,c-;~-".:,.:'£7:~%:). ;:",~'~'.,'.~:~:;-:L:-~::~;~-.":.',.-.7~'Z-~i~-t~:'-.::~

•, ,', "--Y'.:"~-":.." : 2 . i ::~:5-.~:S.'%~!!.<======================:
:'::..-:.:-::q::-'-7".':.. : .....:;>~d,a.?7_'. " . J " /7:5"- ":--~~:>::% a7":.7"~-.:,-:~-:::i
/ L I = "i ........ '......... " " ' " "......") ---"-' T................................... "' ~"' ........{"'/~" . . . . . . . . . . . . . . . . . . . . . . ) - °"
SP 3 5 0 I 250 1 S0 50

Fig. 9. Part of the API seismic reflection section just south of Gunung Api where it crosses the TIMOR line. This
section was migrated using the same velocity function as used for the TIMOR line shown in Fig. 8. The reflection
observed at c. 9.3 s (10-15 km depth) on the TIMOR line (see Fig. 8 caption) is not as bright here. A sea-floor ridge
separates oceanic crust, overlain by < 200 m of sediments, at two distinct depths. Reflectors beneath the ridge include
arcuate and short horizontal segments on the flanks that probably represent submarine basalt flows, possibly later
folded (inset).

indicate that the fracture has acted as a m a g m a fault, orientated nearly normal to the regional plate
conduit. Together these features imply a leaky m a r g i n w o u l d s u g g e s t that the d e f o r m a t i o n
transform fault zone nearly coincident with the associated with plate c o n v e r g e n c e m a y have here
T I M O R seismic profile. propagated further into the overlying plate than
A bright horizontal reflector is o b s e r v e d on the elsewhere along the plate boundary. The location
T I M O R line where it intersects the API line (9.3 s of this fracture m a y be due to an older zone of
travel time on Fig. 9; 10.5 k m depth on Fig. 8), weakness in the B a n d a Sea lithosphere or due
although only a w e a k w e s t w a r d dipping reflector is to unusually thick or buoyant structures on the
observed on the API line (Fig. 9). Nearby refraction subducted plate (e.g. Cross & Pilger 1982).
studies indicated that the M o h o occurs at 10.3 k m
depth (M 12 profile of B o w i n e t al. 1980). If the
refraction velocities from the M 1 2 profile are used Discussion: appropriateness of
to migrate and depth convert the reflection section,
an indentor model
the bright reflector occurs within the mantle at
1 4 - 1 5 k m d e p t h s and no o b v i o u s r e f l e c t i o n s T h e part o f the B a n d a arc i m m e d i a t e l y east of
coincide with the M o h o at 10.3 k m depth. Either T i m o r is characterized by several features not
the refraction velocities derived along the nearby generally observed in other parts of the arc. A n
profile M 1 2 are inappropriate here or else the active volcano within the backarc basin 400 k m
mantle is m o r e reflective than the oceanic crust above the W a d a t i - B e n i o f f zone is rare, and here
along this profile. occurs along a line o f seamounts orientated nearly
T h e bright reflections m a y m a r k the location o f normal to the c o n v e r g e n t margin. Uplift rates
m a g m a rather than a change in rock lithologies or are high on both the volcanic arc and forearc
mineralogy. M o r e probably the bright reflector accretionary c o m p l e x islands, bringing parts o f
represents underplated m a g m a w h i c h produces a eastern T i m o r nearly 3 k m above sea-level. T h e
thicker crust i m m e d i a t e l y beneath the sea-floor forearc basin is only 10 k m wide where it was
ridge. A regional fracture zone, or leaky transform crossed by the D A M A R seismic line. A n o m a l o u s l y
 LOCAL INDENTOR TECTONICS IN THE BANDA ARC 71

low Bouguer gravity anomalies are observed in the
corridor defined by the two deep seismic profiles,
anomalies more negative than anywhere else along
the Banda arc. Taken together, these features
suggest that unusual lithospheric strains are associ-
ated with the convergence that occurs within this
part of the arc. Explanations proposed involve three
major lithospheric plates (e.g. Charlton 1986), but
the authors prefer to discuss the local features in
terms of a single two-plate boundary.
Unusually thick parts of the Australian shelf
entering the subduction zone are one possible
explanation. Crust not greatly thinned during
Jurassic rifting, crust thickened as it entered the
modern convergence zone, or unusually thick sedi-
mentary basin fill could all contribute relatively
buoyant, low density continental crustal material to
this segment of the arc. The greater amount of low
density material satisfies the observed gravity
anomalies. Its associated buoyancy, compared to
oceanic crust, would enhance uplift. A thickened
wedge of subducting material resists further sub-
duction by deforming internally to form a core of
continental material. This core may partly under-
thrust and dome the arc, and partly impinge upon
and strain the oceanic crust of the backarc (Fig. 10).
This form of localized strain is here proposed to
have caused the fracture associated with the line
of seamounts including Gunung Api.
Unsubducted parts of the Australian shelf pro-
vide some evidence that thick sedimentary basins
may have underthrust the arc in the area of the deep
seismic profiles. About 500 km to the south, more
than 6000 m of Palaeozoic and younger sediments
occur in the Petrel sub-basin (Schltiter & Fritsch
1985; AGSO North West Shelf Study Group 1994)
between the offshore parts of two Archaean blocks
(Fig. 1). The northern continuation of the Petrel
sub-basin is poorly defined, so that it or a similar
Carboniferous-Jurassic basin may continue
beneath the Sahul platform or align with the band
of lower Bouguer anomalies, including the 'Moa
anomaly', that trend nearly orthogonal to the
convergence margin (Fig. 5). Crustal thinning
associated with the crustal extension that formed
the Petrel subbasin and other similar basins to the
SW has estimated stretching factors of 100-400%
(O'Brien et al. 1993; Etheridge & O'Brien 1994).
These estimates are consistent with the lower
crustal thicknesses of 31 km modelled at the
western ends of the two refraction lines, within
the Moa gravity feature, when compared with the
35-39 km thicknesses (Fig. 1) reported at the
eastern ends (Bowin et al. 1980).
Thinner crust beneath the relatively undeformed
shelf in this region is not obviously consistent
with the proposed model of anomalously thickened
continental crust doming and indenting the arc
and backarc, respectively. Thrust duplexing within
the lower crust could partly restore the pre-rift
crystalline crust's thickness while retaining the full
overlying basin's thickness, and result in local mass

GUNUNG API ~ DAMAR

BAN

....

Fig. 10. Block diagram showing the proposed local promontory of thickened Australian shelf crust and sediments.
This promontory becomes a potential indentor wedge during subsequent convergence by lifting and offsetting the
overlying crust across faults inferred to segment the accretionary complex that includes Timor and to propagate
northward into the backarc Banda Sea. The model contains elements intended to explain the various observed
anomalous features in the bathymetry, gravity field and crustal thickness in the vicinity of the new deep seismic
reflection profiles. The cross pattern indicates crystalline basement within the Australian crust.
72 D.B. SNYDER ET AL.
deficits such as that m o d e l l e d for the ' M o a
anomaly' (Figs 6b & 10). Shear thrusts within the
relatively weak lower crust need not directly
underlie the rift basins.
O v e r l y i n g these deeper, N N E trending
Devonian-Carboniferous basin structures are the
NE trending platforms and graben structures, such
as the Sahul Platform and Malita Graben, that are
associated with the Carboniferous-Permian rifting
of Gondwanaland. A small outboard platform,
similar to the A s h m o r e and Sahul Platforms
(Fig. 1), and effectively forming a promontory on
the pre-collisional Australian margin, may have
already been subducted and contributed to the low
mass a n o m a l y of thicker continental material
beneath the Moa gravity anomaly.
Insufficent details of crustal thicknesses and
structures are available to resolve more clearly the
various superposed pre-collision extension-related
structures from the recent collisional ones. The
coincidence of a north-south orientated boundary
between Archaean crustal blocks and the mid-
Palaeozoic Petrel (Bonaparte) Basin with a N E - S W
orientated late-Palaeozoic rifted m a r g i n and
Neogene convergence zone permits many possi-
bilities for crustal structures. Some of the older
structures may nucleate subsequent strain.
The presently preferred explanation for high
uplift rates on Timor and other anomalous features
within the Banda arc immediately east of Timor is
References
AGSO NORTH WEST SHELF STUDY GROUP. 1994. Deep
reflections on the North West shelf: changing
perceptions of basin formation. Proceedings
Western Australian Basins Symposium, Perth,
63-76.
BOWlN, C., PURDY, G. M., JOHNSON, C., SHOR, G.,
LAWYER, L., HARTONO,H. M. S. & JEZEK, P. 1980.
Arc-continent collision in the Banda Sea region.
AAPG Bulletin, 64, 868-915.
BREEN, N. A., StaYER, E. A. & ROOF, S. 1989. The Wetar
backthrust belt, eastern Indonesia: the effects of
accretion against an irregularly shaped arc.
Tectonics, 8, 85-98.
CARDWELL,R. K. & ISACKS,B. L. 1978. Geometry of the
subducted lithosphere beneath the Banda Sea in
eastern Indonesia from seismicity and fault plane
solutions. Journal of Geophysical Research, 87,
2825-2838.
CHARLTON, T. R. 1986. A plate tectonic model of the
eastern Indonesian collision zone. Nature, 319,
394-396.
, DE SMET, M. E. M., SAMODRA,H. 8z KAYE, S. J.
1991. The stratigraphic and structural evolution of
the Tanimbar Islands, Eastern Indonesia. Journal of
SE Asian Earth Sciences, 6, 343-358.
CROSS, T. A. & PILGER, R. H. 1982. Controls of sub-
duction geometry, location of magmatic arcs, and
partial restoration of crustal thicknesses beneath a
northward continuation of the Petrel sub-basin of
the Bonaparte rift by thrust duplexing within the
lower crust during recent collision. Dynamically
depressed crust is inferred from the gravity field
east of Timor. In eastern Timor the thickened crust
has rebounded isostatically to produce rapid uplift
and neutral gravity anomalies. In both areas, over-
compensated crust within the subducted slab has
acted as a buoyant mass that jams subduction and
translates convergence to the north where it loads
and fractures the oceanic lithosphere of the Banda
Sea.
We would like to especially thank K. Bartram,
A. Macfarlane, B. Situmorang, S. Suparka and M. Takim,
for their interest and support in organizing the seismic
project in Indonesia. The BANDA seismic reflection
profiles were acquired by GECO Singapore Pte. and
processed by Prakla-Seismos, and are available at the cost
of reproduction from the British Geological Survey
(Marine Geophysics Programme Manager), West Mains
Road, Edinburgh EH9 3LA, UK. This research was done
under the auspices of the Indonesian Institute of Sciences
(LIPI) and the Natural Environment Research Council
(NERC) of the UK. BANDA was funded by the British
Natural Environment Research Council and BIRPS'
Industrial Associates Program (Amerada Hess Ltd,
Amoco, BP Exploration Operating Company Ltd,
Chevron (UK) Ltd, Conoco (UK) Ltd, Mobil North Sea
Ltd and Shell (UK) Ltd). This paper is Cambridge Earth
Sciences contribution number 4176.
tectonics of arc and back-arc regions. Geological
Society of America Bulletin, 93, 545-562.
DEMETS, C., GORDON, R. G., ARGUS, D. E & STEIN, S.
1990. Current plate motions. Geophysical Journal
International, 101, 425--478.
DROPKIN, M. J., HARRIS, R. A. & ZEITLER, P. K.
1993. An Oligocene forearc crustal flake exposed
in a contemporary arc-continent collision:
Timor, Indonesia. Geological Society of America
Annual Meeting, Abstracts with Programs, 25, 6,
A--482.
ETHERIDGE, M. • O'BRIEN, G. W. 1994. Structural and
tectonic evolution of the Western Australian margin
basin system. Australian Petroleum Exploration
Association Journal, 34, 906-908.
GENRICH, J., BOCK, Y., MCCAFFREY, R., CALAIS, E.,
STEVENS, C. E T A L . 1994. Kinematics of the
eastern Indonesian island arc estimated by global
positioning system measurements. EOS abstracts,
75, 162.
GRADY, A. E. & BERRY, R. E 1977. Some Paleozoic-
Mesozoic stratigraphic-structural relationships in
east Timor and their significance in the tectonics of
Timor. Journal of the Geological Society Australia,
24, 203-214.
HAMILTON, W. 1979. Tectonics of the Indonesian region.
US Geological Survey Professional Paper, 1078.
 LOCAL INDENTOR TECTONICS IN THE BANDA ARC
1988. Plate tectonics and island arcs. Geological
Society of America Bulletin, 100, 1503-1527.
HARRIS, R. A. 1991. Temporal distribution of strain in
the active Banda orogen: a reconciliation of rival
hypotheses. Journal of SE Asian Earth Sciences, 6,
373-386.
HILTON, D. R., HOOGEWERFF,J. A., VAN BERGEN, M. J.
& HAMMERSCHMIDT, K. 1992. Mapping magma
sources in the west Sunda-Banda arcs, Indonesia:
Constraints from helium isotopes. Geochimica et
Cosmochimica Acta, 56, 851-859.
JACOBSON,R. S., SHOR, G., KIECKHEFER,R. M. & PURDY,
G. M. 1979. Seismic refraction and reflection
studies in the Timor-Aru Trough system and
Australian continental shelf. In: WAT~NS, J. S.,
MONTADERT, L. & DICKERSON, P. W. (eds)
Geological and Geophysical Investigations
of Continental Margins. AAPG Memoir, 29,
209-222.
JOHNSTON, C. R. & BOWlN, C. O. 1981. Crustal reactions
resulting from the mid-Pliocene to recent continent-
island arc collision in the Timor region. Bureau of
Mineral Resources Journal of Australian Geology
& Geophysics, 6, 223-243.
JONGSMA, D., WOODSIDE, J. M., HUSON, W., SUPARKA,S.
& KADARISMAN,D. 1989. Geophysics and tentative
late Cenozoic seismic stratigraphy of the Banda
Arc-Australian continent collision zone along three
transects. Netherlands Journal of Sea Research, 24,
205-229.
KARIG, D. E., BARBER, A. J., CHARLTON, Z. R.,
KLEMPERER, S. & HUSSONG, D. M. 1987. Nature
and distribution of deformation across the Banda
Arc-Australian collision zone. Geological Society
of America Bulletin, 98, 18-32.
LAPOU1LLE,A., HARYONO,H., LARUE, M., PRAMUMIJOYO,
S. & LARDY, M. 1985. Age and origin of the
seafloor of the Banda Sea (eastern Indonesia).
Oceanologica Acta, 8, 379-389.
MATrHEWS, D. H. 1986. Seismic reflections from the
lower crust around Britain. In: DAWSON, J. B.,
CARSWELL, D. A., HALL, J. & WEDEPOHL,K. H.
(eds) The Nature of the Lower Continental Crust.
Geological Society, London, Special Publications,
24, 11-21.
MCBRIDE, J. H. & KARIG, D. E. 1987. Crustal structure of
the outer Banda arc: new free-air gravity evidence.
Tectonophysics, 140, 265-273.
MCCAFFREY, R. 1988. Active tectonics of the eastern
Sunda and Banda Arcs. Journal of Geophysical
Research, 93, 15163-15182.
1992. Oblique plate convergence, slip vectors,
and forearc deformation. Journal of Geophysical
Research, 97, 8905-8915.
1996. Slip partitioning at convergent plate
boundaries of SE Asia. This volume.
73
--, MOLNAR, P., ROECKER, S. W. & JOYOD1WIRYO,
Y. S. 1985. Microearthquake seismicity and fault
plane solutions related to arc-continent collision
in the eastern Sunda Arc, Indonesia. Journal of
Geophysical Research, 90, 4511--4528.
MILSOM, J. & AUDLEY-CHARLES, M. G. 1985. Post-
isostatic readjustment in the southern Banda Arc.
In: COWARD, M. P. & RIES, A. C. (eds) Collision
tectonics. Geological Society, London, Special
Publications, 19, 353-364.
O'BRIEN, G. W., ETHERIDGE, M. A., WILCOX, J. B.,
MORSE, M., SYMMONDS, P., NORMAN, C. &
NEEDHAM, D. J. 1993. The structural architecture
of the Timor Sea, northwestern Australia: implica-
tions for basin development and hydrocarbon
exploration. Australian Petroleum Exploration
Association Journal, 33, 258-278.
PIGRAM, C. J., DAVIES, P. J., FEARY, D. A. & SYMONDS,
P. A. 1989. Tectonic controls on carbonate
platform evolution in southern Papua New Guinea:
passive margin to foreland basin. Geology, 17,
199-202.
RICHARDSON, A. N. 1993. Lithospheric structure and
dynamics of the Banda Arc, eastern Indonesia. PhD
thesis, University of London.
SCHLt2rrER, H. U. & FRITSCH, J. 1985. Geology and
tectonics of the Banda Arc between Tanimbar Island
and Aru Island (Indonesia) - results of R/V Sonne
Cruise SO-16. Geologisches Jahrbuch Reihe E, 30,
3-41.
SENGOR, A. M. C. 1987. Tectonics of the Tethysides:
orogenic collage development in a collisional
setting. Annual Reviews of Earth & Planetary
Science, 15, 213-244.
SILVER, E. A., REED, D., MCCAFFREY, R. &
JOYODIRWmYO, Y. 1983. Back arc thrusting in the
eastern Sunda arc, Indonesia: a consequence of
arc-continent collision. Journal of Geophysical
Research, 88, 7429-7448.
VEEVERS, J. J. 1988. Morphotectonics of Australia's
northwestern margin - a review. In: PURCELL,P. G.
& PURCELL, R. R. (eds) The North West Shelf,
Australia. Proceedings of Petroleum Exploration
Society Australia Symposium, Perth, 1988,
19-27.
WOODSIDE, J. M., JONGSMA,D., THOMMERET,M., STRANG
VAN HEES, G. & PtYNTODEWO 1989. Gravity and
magnetic field measurements in the eastern Banda
Sea. Netherlands Journal of Sea Research, 24,
185-203.
YEATES, A. N., BRADSHAW,M. T., DICKINS,J. M., BRAKEL,
A. T., EXON, N. E Er AL. 1987. The Westralian
Superbasin: an Australian link with Tethys. In:
MCKENZIE, K. G. (ed.) Shallow Tethys 2.
Proceedings of Intenational Symposium on Shallow
Tethys 2, Wagga Wagga, 199-213.
 Detailed processing of seismic reflection data from the frontal part
of the Timor trough accretionary wedge, eastern Indonesia
B. D. H U G H E S 1'3, K. B A X T E R TM, R. A. C L A R K 1 & D. B. S N Y D E R 2
1 Department of Earth Sciences, The University of Leeds, Leeds LS2 9JT, UK
2 BIRPS, Bullard Laboratories, Madingley Rise, Cambridge, CB30EZ, UK
3 Present address: ABG EXPLORATION, 12 Howard Way, Cromwell Business Park,
Newport Pagnell, Bucks MK16 9QR, UK
4 Present address: CSIRO (Division of Exploration & Mining), Floreat Park Laboratories,
Wembley, Perth, WA, Australia 6014

Abstract: The DAMAR deep seismic line across the Banda arc in Indonesia provides a valuable
insight into the Late Cenozoic collision between the northward-moving Australian plate and the
Southeast Asian plate. Previous geophysical investigations across the accretionary wedge have
only imaged relatively shallow structures due to the complex structure of the wedge which
remains poorly resolved following conventional processing techniques. The design of a detailed
processing sequence and its application to a 35 km section of the DAMAR line to include the
Australian Northwest shelf, the Timor trough and the frontal part of the accretionary wedge, has
allowed an improvement in resolution of structures both within and below the frontal part of the
wedge. The Australian margin to the south of the trough contains a number of normal faults
which show inversion features. The reprocessing of the accretionary wedge has allowed an
improved interpretation of the internal structure in the frontal part of the collision zone and this
consists of a number of high angle thrust structures. This has shown that shortening in the frontal
part of the accretionary wedge has developed coherent thrust slices imbricated from the
subducting Australian margin, rather than an incoherent melange.

The Timor region in eastern Indonesia is often
quoted as a present-day fold-and-thrust belt in
its earliest stages, and is one of the youngest
collisional belts presently active. The study of the
Timor region not only provides an understanding of
the active collision between the Australian plate
and Southeast Asia, but allows a view of the early
stages of mountain belt evolution.
The development of the region may be divided
into two main phases:
• break-up of Gondwanaland and the development
of the Australian passive margin;
• development of subduction north of Australia
with eventual collision of the Australian margin
and development of the collision zone presently
exposed on Timor.
In 1992 the British Institutions Reflection Profiling
Syndicate (BIRPS) and the Indonesian Marine
Geological Institute commissioned two deep
seismic lines across the Timor-Tanimbar trough to
the east of Timor, the BANDA survey (Fig. 1). This
paper discusses the development of a processing
sequence to improve the resolution obtained by
conventional processing on part of the DAMAR
line (shotpoints 600-950), and the interpretation
of these data to assess the deformation styles, both
within the accretionary wedge and on the
Australian shelf immediately adjacent to the
trough. Accepted models that assume that deforma-
tion has recently ceased at the Timor trough are
also evaluated in the light of this work. Similar
processing techniques to those described may be
applied to similar datasets from other areas to
increase the resolution of structures both within and
below areas of complex geological structure.
Evolution of the Timor trough region
One of the most extensive features in Indonesia is
the linear, northerly dipping subduction zone
delineated by the Java Trench-Timor trough which
marks the collision of the northerly moving Indo-
Australian plate with the Southeast Asian plate.
During the break-up of Gondwanaland the
Australian margin progressively rifted southwards
with continental break-up on the Northwest Shelf
by the Oxfordian (160 Ma) (Pigram & Pangga-
bean 1984). Following rifting, the drift phase of
the margin during the Cretaceous and Tertiary
was characterized by the deposition of dominantly
carbonate sediments. During the Neogene the

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution o.fSoutheastAsia, 75

76 B.D. HUGHES ET AL.
0
0
I
I
km 10,01 ' 000/ 4
I~ o ~ TIMOR Banda Sea /L) ° /
~'~ GunungApi
DAMAIq~
0

o
Kisar
I
South

5 m r n a -I
T r o u b a d o u r Well N-°1%
Australian Northwest Shelf
10 °
125 ° 126 ° 127 ° East 128 ° 129 ° 130 °

Fig. 1. Part of the Banda arc in eastern Indonesia showing the location of the BIRPS TIMOR and DAMAR deep
seismic lines. The barbed line indicates the trace of the Timor trough, volcanoes are shown as circles (black where
active) and values on bathymetric contours are in metres. The section detailed in this paper is shown by the thickened
line, located in the immediate vicinity of the trough axis on the DAMAR line, shotpoints 600 to 950.

northerly drift of the Indo-Australian plate pro-
duced the development of oceanic subduction to the
north of Australia and ultimately the collision of the
leading edge of the Australian plate with the Banda
arc subduction complex. This resulted in the
development of a fold-and-thrust belt exposed on a
number of islands, the largest of which is Timor.
Subduction is presently active along the Java
Trench with decreased activity on the Timor trough.
Recent geodetic measurements by McCaffrey
(1996) suggest that the Timor collision zone has
recently locked on to the northward-moving
Australian plate with cessation of subduction at the
Timor trough.
Marine geophysical studies in the
Timor trough
Numerous previous workers have studied the
Australian margin and the Timor trough region
using marine geophysical techniques (e.g. Bowin
et al. 1980; Karig et al. 1987; Charlton, 1988;
Masson et al. 1991). Cross-sections across the
Timor trough have previously been imaged using
seismic reflection data, and a number of common
features have been observed (Fig. 2):
• an outer slope increasing in gradient towards the
trough axis, underlain by Australian shelf and
slope strata (A & B, Fig. 2). Normal faults are
present, mainly dipping northwards;
• a trough axis with a flat-topped wedge of trough
fill sediments (F, Fig. 2) representing sediments
trapped between the outer slope and the deforma-
tion front;
• a deformation front marking the southerly extent
of the Timor fold-and-thrust belt;
• an accretionary wedge containing individual
thrusts and folds representing deformation of the
subducted Australian margin.
The complex seismic response of accretionary
wedges causes the geologically important struc-
tures within and beneath them to remain poorly
imaged after conventional 2D processing, as seen
in many previously published sections. An ideal
processing sequence, always data-specific, should
 SEISMIC REFLECTION DATA FROM THE TIMOR TROUGH 77

(a)

TW'T
(s)

(b)
Inner slope Trough fill Outer slope
< >< >< ,,- >
Deformation front

4 --"~ 4

6 6
TWT - -/
(s) ' ' '

Fig. 2. Seismic reflection profile (Gulf line AU-38F) across the Timor trough and outer slope south of Timor
showing the major morphological features. Adapted from Karig et al. (1987).

deliver optimum signal-to-noise ratio (SNR) while
representing subsurface geometries faithfully and
with optimum vertical and lateral resolution. These
objectives are sometimes compromised against
practical limits such as computer resources, but this
study has developed a more detailed processing
sequence which (compared with conventional
processing) has resulted in a significant improve-
ment in SNR, resolution, and hence geological
interpretation.
Data processing
Seismic reflection data processing techniques are
well described in many texts (e.g. Yilmaz 1987).
However, 'deep' reflection surveys (i.e. deep
relative to hydrocarbon exploration, for which
reflection methods are most developed) encounter
particular difficulties: data from the Timor trough
are quite typical in this respect. They show strong
multiple activity, all reflectors are relatively deep,
and numerous very strong diffractions are visible.
Migration is an essential process in such heavily
faulted sections; with marked variations in reflector
dip across the section, choice of suitable migration
velocities was critical, but, as is typical of deep
reflection surveys, velocities were difficult to
measure accurately because of the very limited
amount of moveout on the reflections.
The data were provided by the BIRPS group
already preprocessed and sorted into CMP gathers.
The considerable water depth (2-4 s twt) meant
that water-bottom multiples in fact only appeared
deeper in the section than the zones of interest here:
their removal was straightforward and multiple
suppression was carried out as part of standard
preprocessing as described by Snyder et. al. (1996).
Conventional processing was then completed, to
ascertain the overall geometry of the accretionary
wedge. Velocity analyses were made at 100 CMP
(2.5 km) intervals across the section. Given the
simple structures south of the Timor trough, this
was felt acceptable for that part of the line, but
further north, velocities were difficult to determine
78 B.D. HUGHES ET AL.
reliably. The stacked section so produced was
migrated using a finite-difference time migration,
to allow lateral variations in velocity to be
accommodated. Even so, internal structure of the
accretionary wedge and the subducting Australian
shelf beneath it were poorly resolved and a more
detailed processing strategy was imperative for this
part of the line. Shotpoints 600-950, spanning the
toe of the wedge, are shown conventionally
processed in Fig. 3. Key elements of the improved
processing sequence (Fig. 4), which yielded the
section in Fig. 5 (which shows the same shotpoint
range a~Fig. 3), are now discussed briefly.
Detailed processing
South of shotpoint 748, no revisions to velocity
analyses were thought necessary. As the section
becomes structurally more complex northwards,
within the accretionary wedge, selection of
velocities becomes more important. Instead of
regular spacing, velocity analyses were located at
the apices of all prominent diffractions, resulting
in identification of more accurate velocities. The
average separation between velocity analyses thus
reduced to 1.6 km; still larger than would have been
preferred, but the effect of the strong diffractions
was to make additional velocity analyses elsewhere
of little value.
To refine velocity estimates further, bias from
dip effects (inherent in any moveout-based
velocities for dipping reflectors) was reduced by
an iterative sequence of velocity analysis, DMO
(dip moveout), and repeat velocity analysis.
Common-offset FK-based DMO (Hale 1984) was
used; because of limitations on computer storage,
600 700
I I
Fig. 3. Example of part of the DAMAR line (shotpoints 600-950) processed using a conventional processing
sequence. Note the poor lateral and vertical resolution which is attributed to the low signal-to-noise ratio resulting
from conventional processing. Compare this with the same section in Fig. 5, reprocessed to give better definition,
which has allowed an improvement in the geological interpretation.
DMO was applied separately to near, intermediate,
and far offsets and the full dataset then recon-
stituted. Revised velocities were some 5-10%
different from the original in and below the
accretionary wedge but effectively unchanged for
the shallow-dipping Australian shelf to the south.
Successful migration of the resultant stacked
section was critically dependent of the use of
accurate, true, medium velocity throughout. Below
the seabed, optimum migration velocities were
determined empirically as percentages of stacking
velocities, but that of the water column is well
known and was not allowed to vary. In the complex
parts of the line, test panels from faulted areas at
either end of the shelf slope and from within the
accretionary wedge were used. Optimum percent-
ages (typically 80-90%) were chosen and applied
to a finite-difference time migration of the full line
(shown in Fig. 5). For structurally simple parts of
the line, optimum percentages could be seen to
correlate clearly with dip (see Hughes 1994 for a
full discussion). A depth conversion was done post-
migration, using geologically-constrained interval
velocities.
Interpretation of the DAMAR line
An interpretation of the DAMAR seismic l i n e
involved an iteration between geological observa-
tions and the processing sequence. Structures
identified were examined and additional processing
was applied if required to redefine the velocity field
allowing an improvement in resolution and
interpretation. An interpretation of the reprocessed
section is shown in Fig. 5b.
800 900 SP
| I I I
0 10 km
|,,, I
TWT
(s)
-5
 SEISMIC REFLECTION DATA FROM THE TIMOR TROUGH 79

PREPROCESSING
Carried out by BIRPSgroup independently. Incorporated
multiple suppressionand sortingto CMP gathers

I
" VELOCITY ANALYSIS
Incorporates analysis of diffractions

......

re-sort to CMP~athers [

NMO CORRECTION, CMP STACK

acceptable velocities,I gathers & stack?

no / \yes
/
I sort pre-NMO CMP gathers
DMO to constant-offset gathers:
apply Hale's FK DMO

POST-STACK FILTERING
AGC, bandpass filters, etc.

1
MIGRATION VELOCITY TESTING
On several separate test panels along line

1
l TIME MIGRATION
Constant velocity in water column; variable percentage
of stacking velocities sub-seafloor.

I
DEPTH CONVERSION [

Fig. 4. Outline of the processing sequence used to produce the section in Fig. 5. Emphasis is placed on determination
of accurate and appropriate velocities. Key elements only are shown: background information on processes may be
found in standard texts such as Yilrnaz (1987), and data-specific details are given in Hughes (1994).

The Australian margin slope
The Australian margin slope (shotpoints 600-780)
dips gradually north into the Timor trough at an
angle of approximately 3.5 ° . Reflectors on the
adjacent TIMOR line have been tied to the
Troubador No 1 Well (see Hughes 1994), located
at its southern end, and have been compared and
verified with seismic data from the Timor gap to the
west of the BANDA lines (see Baxter 1993).
Reflectors on DAMAR have been directly corre-
lated with those on the TIMOR section, thus
making it possible to estimate that they relate
stratigraphically to Mid-Late Tertiary (A),
Mid-Base Tertiary (B), Mid-Late Jurassic break-
up unconformity (C) and Top Permian? (D) ages.
The upper sediments represent the ongoing post-rift
subsidence of the Australian passive margin and are
Late Cretaceous to Miocene pelagic mudstones and
shelf carbonates (Karig et al. 1987). These are
laterally consistent across the section with no
prominent angular erosional surfaces.
Normal faulting on the DAMAR line is apparent
in two main areas (shotpoints 608-635 and shot-
points 740-778) and these faults appear to bound a
small graben structure. Some of these faults dip to
the south, rather than north towards the trough as
commonly described by previous workers. The
faults show growth (indicated by thickening of
sediments in the hanging wall) during Cenozoic
 •OU!l pzqsep ~o~oIql e £q po3lxetu st o~po~ Otll ol luotuzlIOOOp Ieseq oq,L "l[not.D!p sT o*pOA~ Lmuo!loazoe oql uTtIl!ax saolooIJoz jo uo!lelO~O o "sou!i poqsnp Xq po:~a~m ore sllne~I
"((I) Lue.[tu~od do& pue (D) Xl.ttu~ojuooun dn-]t-eozq otsseanf ole'-l-P.[IA[ '(~/) ~d~!laO.L os~fl-PtIAI '(V) ~e!lao.L ~le~I-Pt.IA[ ol ole[o~ ut~amtu ue!lezlsn V oql uo szoloolJOH "uoTl~lo.Idaolu!
posodm.uodns e (q) pue uo!loos possooo.tdo.t oql (~) ~u.~taoqs (uo!leoo[ .tol [ "~!~I oos) .tom!,L Jo lseo 'q~noal aotuLL oql sso~oe ou![ ~IVIA[V(I oql jo uo!)oos possoooad oqJ., "~ "B!,tI
... . . . . . . . . . . . . . . . . . . . . . . . . . ,._ . =.~.=. . . . . . . . . . ...... _ _ _ _

(q)
/ ~-.-,--,=-.--r---.---~ ~ ~ ~ -..-.~ - -- ~ -_:--~-~_.----_- _-_ _.~,--'-~'~--___ " - z - z - - - ~ - ~ - . . - . - - - ~ ~ ~ ~ ~ - ~ . . , . . = , , , , . ~ _ -,,~_'.."-- -~.,--'~-~'_-,,..~_'_---,.-.~
/
~
~- ~=-~Z.~~
-"'.--~----~
_ ~ . ~ _~'...-_-"-~---,C--
~ : . .- ~
-'-
- ~ . .:-~-:,-_.~-.:.-~ ~_' = - = ~~._
l u o m o l l O O O ¢ I ~ - - - ~ - ] - ~
- - - ~',~- ~, .-. ,
= ' ~ - ' , ~
~
~
.
- ~-~.~.
~ . .
~' ~
a . . . . .
..... 11:1
" - -~=
,_
- ~ - .~_ -~- - . ,
~- ,.
.
.>.=:
~
.,-.=..
- ; ~
~ ~
- .-:,.~-~.
. .~. . . . .
-
._
- ~
-
~-~==~-~
- . - _ - - =--3,T_
"~.-~__ ~
.... ~.--~:~- -~.~,'Z"-~'--"
~_~
~:.--7~-~?.~ --~--'z~ = :'~ " - . -.-- ." , ~ '= "'---'7 ~ "-: -' --~'L~:-L-~:,~-~;.7.~_ - - . - -==~-':~.~_7~
-~=--: .
' ~ - ~ ' ='-~.~--' :~-"" . . . . " ~ ~ ~ - - - ' ~ _ : "- "-.L~', ':" - ~,~. . . . ~ - " - ~ - ~ - ~ - ~ -- " "
/~- ~-, i~W-~.-'-~_~-~.~~~
.LA~,L II ~- ~ ~ ~_,.-.~.~
- . . . " "-"
, : -~"~'-..-..-~'=.=c.-.'~r,'=.=,'
. 7 _Z , z : .. ,:.=..~ " -
luo~J uoql~tuaoJo(I ~
:1 ~-----------------~"
"~
--- .---.-~
,4-,Z:
~01 . ttl)l~ ~dolsiol.I tu.o~po.,~A6
l, /~UO'I'IOIooVI I I odol
st Jl°qs aolno uN.mmttm!Ie'°'snv
dS 006 0~:8 008 0~L 00L 0~9 009
(~)
•
I ~ ~ ~
~ •~..~,~- - - . . . ~ - ~ - -. -~-~"--~--~
~_..~..~ =-,. . -=
£A~.I.
N S
.... I i | I | I 3r
oo
dS 006 0~8 008 O~L OOL 0~9 009
 SEISMIC REFLECTION DATA FROM THE TIMOR TROUGH
sediment deposition and in places reach the sea
floor. Inversion structures are apparent on the
southernmost, north-dipping, fault (shotpoint 610)
with a compressional anticline developed on
reflector B. Testing this structure using different
migration velocities has shown that this is not
an undermigration effect (Hughes 1994). Some
thickening of the sediments just below this reflector
on the fault can be seen with steep up-thrown
sediments immediately overlying low-angle
dipping strata, possibly indicating a small thrust,
although this is not well imaged.
Some complicated zones are apparent on the
section and are thought to correlate to normal faults
(e.g. in the vicinity of shotpoint 747). The excessive
displacement gradient generated by interpreting
single faults suggests that this area may contain
additional fault terraces, increasing the appearance
of drag on the fault zones. This faulting also creates
the topographic bulge in the sea floor at shotpoint
760. At shotpoint 635, a south-dipping fault with no
apparent inversion similarly shows thickening of
sediments in the hanging wall and steepening of
reflectors immediately adjacent to the fault. This is
also difficult to interpret as a single structure and
probably involves terraces. It should be noted that
the presence of inversion across this region and
potential strike-slip activity may also give out-of-
plane-of-section complications which are not
considered here.
The trough axis
The axis of the trough, marking the surface tran-
sition between the shelf slope and the accretionary
wedge, is characterized by a change in the regional
topographic dip from north to south. However, a
significant thickness of trough sediment, as seen on
the adjacent TIMOR line and in sections published
by previous workers (e.g. Fig. 2) is absent on the
DAMAR line. Karig et al. (1987) also noted a
change in the geometry of the trough fill along axis
with ranges of 0-10 km in width and 0 to >1 km
deep. They proposed that this is due to the wedge
thrust front being laterally discontinuous and that in
some areas the trough fill may have been accreted
into the toe of the wedge. Both the section in Fig. 2
and the adjacent TIMOR line show similar features
of overthrusting of the trough sediments by the
wedge front.
From the reprocessing beneath the frontal part
of the accretionary wedge, a pre-existing thickness
of sediment in the trough does not appear to have
been overridden by the wedge (as seen on the
TIMOR line), unless it has been fully accreted to
the wedge by its recent advancement. The top of
the downgoing Australian plate lies at a depth of
81
3.8 km on the TIMOR line beneath a trough sedi-
ment fill of c. 630 m, and at a depth of c. 3.2 km
at seabed on the DAMAR line at shotpoint 780.
Therefore, the greater seabed depth of the TIMOR
line may have produced a preferential site for
trough sedimentation in comparison with the lack
of sediment fill present on the DAMAR section. It
may be possible that the inversion of the normal
fault beneath the trough axis on the DAMAR line
at shotpoint 760 has also contributed to the
reduced bathymetry in the trough, although this is
speculative.
The accretionary wedge
The accretionary wedge is characterized by a
change in the surface topography as the Timor
collision zone overrides the Australian margin. The
sea floor above the wedge shows a relatively
smooth profile, with a lack of penetration of thrusts
at the wedge front and only minor penetration
further back in the wedge. This suggests that the
frontal part of the wedge is inactive, with minor
out-of-sequence thicknening further back pro-
ducing internal thickening. This may be due either
to the cessation of subduction at the Timor trough
in this region, or that the frontal part of the wedge
is inactive as internal thickening attempts to return
the wedge to a critical state and allow propagation
of the wedge front.
The internal structure of the wedge is seismic-
ally complex. Picking structural and lithological
reflectors is obviously difficult in such a complex
zone which is likely to contain a high amount
of diffracted energy which may not have been
migrated back to a source reflector. The use of
migration velocity analysis during this study has,
however, led to the conclusion that a number of
dipping events within the wedge are real reflectors
and not diffraction effects. These reflectors are
shallow in dip at the toe of the wedge, increasing in
dip southwards and are most likely to be thrust
faults developed during forward propagation of the
thrust front, developing a back-steepening of the
thrusts. These show typical thrust geometries with
rollover anticlines in the hanging wall above a low-
angled detachment.
Correlation of reflectors from the Australian
margin is difficult in such a highly deformed zone.
A strong seismic reflector is defined by a high
acoustic impedance contrast at the contact between
two lithological units. The lithological units
occupying the intervals between the strong
reflectors A-D are considered to be homogeneous
throughout as indicated by their relatively trans-
parent internal seismic structure, although some
internal layering is apparent. These units are
82 B.D. HUGHES ET AL.
thought to be similar to those observed from the
Troubadour 1 well located nearby, and consist of."
a Recent-Maastrichtian sequence of calcarenites
and marls from seabed to 1.55 km depth; a 550 m
thick claystone unit (Turonian-Cenomanian);
Jurassic sandstones and claystones (2.1-2.8km
depth); and Triassic limestones and claystone-
siltstone-sandstone sequence (2.8-3.3 km depth).
Therefore, during thrusting, imbrication which
structurally juxtaposed two of the main lithologies
could be expected to produce a reflector with
acoustic impedance and phase similar to a strati-
graphic contact. The strong convex-upward
nature of many reflectors is thought to correlate
to hanging wall geometries within thrust
sheets rather than being from faults but it is
unclear whether the lithological contacts picked
correlate with the original stratigraphic reflectors
on the Australian margin or whether these are
due to lithological juxtapositions created during
faulting.
Imaging of structures beneath the
accretionary wedge
Previous studies across the Timor trough have not
generally imaged the structure of the Australian
shelf beneath the accretionary wedge. Improved
processing during this study has allowed some
structures to be imaged beneath the frontal part
of the wedge. The decollement itself is not
strongly imaged, although the truncation of stratal
reflections within the uniform shelf sediments
allows an interpretation of its location.
Beneath the wedge in the area of shotpoint 820
there appears to be an upwarping of reflectors C
and D. This could be either a processing anomaly or
a geological feature. Similar anomalous features
are well known during processing and are associ-
ated with inaccurate velocity determination of the
overlying structure, particularly if this is of a high
velocity, resulting in a velocity pull-up. However,
the feature at shotpoint 820 appears to be of high
amplitude and short wavelength, which is incon-
sistent with any possible velocity anomaly within
the overlying wedge. Therefore, this feature is
tentatively interpreted as geological. The upper part
appears to be truncated by the wedge decollement
with no apparent continuation into the wedge.
Thus, this structure must pre-date wedge
imbrication of this part of the shelf. The shape of
the feature on reflector D is similar to that beneath
the zones of structural complexity seen elsewhere
on the section (shotpoint 635 and shotpoint 747)
and may be associated with similar fault structures,
although resolved less clearly due to the overlying
wedge.
Conclusions
Following the design of a reprocessing sequence,
an improvement of SNR and resolution of the
DAMAR line has been achieved, with the improve-
ment in data quality allowing an increased con-
fidence in the interpretation. Application of similar,
data-specific, processing should significantly
improve other datasets from areas of complex
geological structure.
Recent geodetic measurements reported by
McCaffrey (1995) have suggested that subduction
at the Timor trough has almost stopped. However,
observations by Charlton (1988) that the wedge
thrust front in the Timor trough to the south of
Timor has not advanced significantly in the last
450 000 years have been interpreted as indicative
of episodic propagation of the thrust front during
continuing subduction by Masson et al. (1991).
From a detailed examination of the frontal part of
the accretionary wedge on the DAMAR line, it is
difficult to resolve this debate. The lack of recent
sediment deposition to the south of the thrust front
does not allow any estimation of timing of the most
recent advance of the wedge and from Charlton's
observations both scenarios are possible.
A further indication of possible recent thrust
activity within the wedge is the identification of
active faults either at the thrust front or within the
wedge, affecting the more recent sediments which
overlie the wedge. Such faults will produce sedi-
ment thickening and sharp topographic variations
on the wedge top. The reprocessing of the DAMAR
line has allowed an improved interpretation of the
internal structure of the accretionary wedge which
should allow similar structures to be imaged.
However, although some faults appear to outcrop at
the seabed with the development of minor breaks in
topography (e.g. at shotpoints 785, 810, and 922)
and may indicate minor internal thickening, the
thrust front and much of the internal part of the
wedge appear to be recently inactive with con-
tinuous, undeformed sediment drape across the
outer slope and the wedge front. The lack of a
recent significant sediment thickness in the trough
axis is attributed to a lack of deposition, rather than
due to recent wedge movement imbricating any
pre-existing trough fill.
From the improved processing of the DAMAR
line it is possible to identify the large-scale internal
deformation mechanism within the wedge. This is
dominated by high-angle thrusts which have
imbricated sediments from the upper part of the
subducting Australian passive margin. This indi-
cates that the deformation mechanism within the
frontal part of the accretionary wedge consists of
the imbrication of coherent blocks rather than by
a shearing of the subducting Australian margin
sediments into an incoherent melange.
 SEISMIC REFLECTION DATA FROM THE TIMOR TROUGH
References
BAXTER, K. 1993. Quantitative modelling of continent
collision: Application to the Timor region, Eastern
Indonesia. PhD thesis, University of Liverpool.
BowrN, C., PURDY, G. M., JOHNSTON, C., SHOR, G.,
LAWVER, L., HARTONO,H. M. S. & JEZEK, P. 1980.
Arc-continent collision in the Banda Sea region.
AAPG Bulletin, 64, 868-915.
CHARLTON, T. R. 1988. Tectonic accretion and erosion
in steady-state trenches. Tectonophysics, 149,
233-243.
HALE, D. 1984. Dip moveout by Fourier transform.
Geophysics, 49, 741-757.
HUGHES, B. D. 1994. Reprocessing, modelling, and
interpretation of complex seismic reflection data
from accretionary wedges: Application to the Timor
Trough, Indonesia. MSc thesis, University of
Leeds.
KARIG, D. E., BARBER, A. J., CHARLTON, T. R.,
KLEMPERER, S. & HUSSONG,D. M. 1987. Nature and
distribution of deformation across the Banda Arc-
83
Australian collision zone at Timor. Bulletin of the
Geological Society of America, 98, 18-32.
MASSON, D. G., MILSOM, J., BARBER, A. J., SIKUMBANG,
N. & DWIYAYrO, B. 1991. Recent tectonics around
the island of Timor, eastern Indonesia. Marine and
Petroleum Geology, 8, 35-49.
MCCAFFREY,R. 1996. Slip partitioning at convergent plate
boundaries of SE Asia. This volume.
PIGRAM, C. J. & PANGGABEAN,H. 1984. Rifting of the
northern margin of the Australian continent and the
origin of some of the microcontinents in eastern
Indonesia. Tectonophysics, 107, 331-353.
SNYDER, D., PRASETYO, H., BLUNDELL, D. J., PIGRAM,
C. J., BARBER, A. J., RICHARDSON, A. & TJOKO-
SAPROETRO, S. 1996. Style of crustal deformation
across the Banda Arc continent-arc collision zone
as observed on deep seismic reflection profiles.
Tectonics, in press.
YtLMAZ, O. 1987. Seismic data processing. Investigations
in Geophysics, No. 2. Society of Exploration
Geophysicists.
 Extension, collision and curvature in the eastern Banda arc
JOHN MILSOM, STEVE KAYE & SARDJONO
Research School of Geological and Geophysical Sciences, Birkbeck College and
University College London, Gower Street, London WC1E 6BT, UK

A b s t r a c t : The Kai Islands occupy the region of maximum curvature in the east of the Banda arc,
where the Aru trough has been regarded as the surface trace of past subduction and present arc-
continent collision. Eocene to Pleistocene sediments on Kai Besar, the easternmost island, have
not been deeply buried or imbricated but have experienced large-scale extensional faulting. The
associated Bouguer gravity high of more than +200 mGal requires upfaulting of the accretionary
complex, the attenuated Australian continental crust on which it rests and the underlying mantle
at the western side of the Aru trough. Seismic reflection surveys show the deformation front
within the Aru trough SE of the Kai Islands but entirely to its west further north. Instead of
continuing NNE to an offset near the coast of New Guinea, the collision trace passes through
the narrow and shallow strait between Kai Besar and the other islands, and thus mimics the
relatively smooth curve of the Bouguer gravity contours, rather than the discontinuities of
the bathymetric troughs. The continuity in deep and shallow structures is strong evidence for
the existence of the outer arc as a single geological unit prior to the present phase of arc-
continent collision.

The B a n d a arc, in eastern Indonesia, is the site o f extensively studied in Timor, where the processes
the collision b e t w e e n the Australian continental involved are relatively simple and where transfer o f
margin, which is m o v i n g north with the Indo- imbricated material from the Australian margin to
Australian plate, and the B a n d a Sea which lies on SE Asia has been well d o c u m e n t e d (cf. B a r b e r
the SE Asia plate (Fig. 1). The collision has been 1979; Charlton et al. 1991a). Similar processes

]~2*e ::: t~3*~ ~.'..'"]]. -. ¢ ":-

/ Fadol : ~

........................ Ke,:n ..... ~ - - , = ! .............

. .~w i':,-.~J ~ . J ..... -

TI sal ya anndds u I
.......... 6°s I " :: "
....... :!................ *J~:................ ~ " Weber ~ ; "'/e
o : I ........................ Basin i.." ,i#<--~ °-: - ,I: ~ 6.'.'~!.
"S

Banda Sea "~iml>,,_,- o-" ~ii:"

" • ..-.ii[" a,-u
. .........,i?. ...........~ o.>'~..

Timor ' ~6"~"°~

2 0 0 Km
............. t ~ ............. i ............. lO°S

-: ;.ll(d~ 124°F., 128°E ~ 1,52<>E

Fig. 1. The eastern Banda Sea. Inset shows details of the Kai Islands.

From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, 85
Geological Society Special Publication No. 106, pp. 85-94.
86 J. MILSOM ET AL.

operate further to the east, where impressive record collision between Southeast Asia and the
images of continental crust underthrusting the Bird's Head peninsula of New Guinea. The Banda
arc have been provided by multichannel seismic arc thus appears to achieve the impossible by being
surveys near the islands of Kisar and Damar involved in collision simultaneously, and with
(Richardson & Blundell 1996) and Tanimbar the same plate, to the north and to the south. As
(Schltiter & Fritsch 1985). The similarity of the noted by McCaffrey & Abets (1991), either the
seismic sections from these two areas is surprising Australia-New Guinea plate or the SE Asia plate,
since the collision trace swings markedly to the and quite possibly both, must be undergoing
north in the Tanimbar region to an orientation considerable internal deformation.
almost parallel to the present-day SE Asia- The nature of the tectonic regime in the eastern
Australia convergence vector. The problem is part of the Banda arc has been controversial from
further compounded on the northern side of the arc, almost the earliest days of plate tectonics. The
where the island of Seram has been described as Timor, Tanimbar, Am and Seram troughs which
geologically a mirror image of Timor (Audley- separate the islands of the outer Banda arc from the
Charles et al. 1979). The very similar seismic Australian continental margin are generally less
reflection sections from the Seram Trough than 50 km across and only 1.0-2.5 km deep. The
(Hamilton 1979: Jongsma et al. 1989a) presumably Am trough, east of the Kai Islands, is exceptional,

~. .-.-.~.. ~" ,? "7..~7".

/. / / ~ . ~ _ _ _ \ .

~b L_z-" ~',," ' // / 5 °S-

/

Weber '\
i B a s i n /'
~,~
)(:e-xi:N:
/

/
7ooo~/~
.....-"

• IJT-B • ~. "~II-,,-a
-.~. .
' o - " i \J / '/S" / - 7os -
../ / / F- ,
Arafura
. / / / Sea
//" i/ / /. .

/;//
?G /,
I'i/ Ooo 0 100 Km

ARU TROUGH
Bathymetry
Contours at 500m intervals
]OE/~ l~oF] and at 200m,

Fig. 2. The Kai-Tanimbar region, showing locations of seismic sections interpreted or discussed in the text.
Bathymetric contours at 500 m intervals, and at 200 m. Bathymetry reproduced from GEBCO sheet 5.10 compiled by
D. Monahan, R. K. H. Falconer & M. Tharp and made available through the GEBCO Digital Atlas published by the
British Oceanographic Data Centre on behalf of the IOC and IHO, 1994.
 EXTENSION, COLLISION ~ CURVATURE, E BANDA ARC
being about 100 km across at its widest point and
reaching a maximum depth of more than 3.5 km
(Fig. 2). Fitch & Hamilton (1974) identified this
trough as the subduction trace a view contested
by Audley-Charles & Milsom (1974), who cited
Bowin et al. (1980) as their authority for an
extensional origin for the Aru trough and suggested
the 7 km deep Weber Basin west of the Kai Islands
as an alternative collision site. Although the reason
for the great depth of the Weber Basin is still un-
certain (but see McCaffrey 1988), later authors
have generally followed Hamilton (1979) in
placing a collision trace in the Am trough, often
transferring motion from its northern end to the
Scram trough by strike-slip motion along the
Tarera-Aiduna fault in the extreme southwest of
the Bird's Head (Fig. 3a). Geophysical data from
the Kai Islands favour a third possibility which was
hinted at by SchRiter & Fritsch (1985) in one of
their illustrations (Fig. 3b) but was not discussed
in,their text.
Geology of the Kai Islands
The Kai archipelago (Fig. 1 inset) is one of the
major island groups of the outer Banda arc,
occupying the region of maximum arc curvature.
The onshore geology has been described by
Achdan & Turkandi (1982) and by Charlton et al.
(1991b). The latter authors divided the group into
Eastern, Central and Western Provinces, of which
the Eastern Province consisted of the high, rugged
(a)
Fig. 3. The Kai-Bird's Head triple junction, showing plate boundary solutions suggested by (a) Hamilton (1979) and
(b) Schltiter & Fritsch (1985). Bathymetry as Fig. 2 but with values omitted.
87
and strongly N-S elongated island of Kai Besar.
The Central Province included Kai Kecil and the
Tayandu Islands and the Western Province com-
prised a number of small islands, including Kur and
Fadol, close to the edge of the Weber Basin.
Kai Besar, which contains the major exposures
of non-coralline rocks in the archipelago, consists
of heavily block-faulted sediments of the
Australian margin belonging to the Eocene, deep-
water, Elat Formation, the shallow water Oligo-
Miocene Tamangil and Weduar Formations and
the Pliocene Weryahan Formation which consists
of shallow-water limestone and marl. Achdan &
Turkandi (1982) showed the Pliocene rocks in
depositional contact on all older formations but
Charlton et al. (1991b) failed to find any evidence
of this unconformity. All geological discussions
of Kai Besar comment on the strong NNE-SSW
faulting, which Charlton et al. (1991b) considered
to be extensional, both on the basis of the regional
geological setting at the edge of the Am trough and
because of the consistent pattern of large east-
facing escarpments (reaching to a maximum of
800 m above sea-level) and gently west-tilted
sediments. The faults are offset or terminated at
WNW-ESE lineaments which were interpreted as
transfer faults.
Kai Kecil, which is separated from Kai Besar by
Selat Nerong (Nerong Strait), consists mainly of
slightly raised coral reef. The occasional outcrops
of the Weryahan Formation may occupy the cores
of gentle anticlinal upwarps. The other islands of
the Central Province are generally similar and are
....... % ~ ............
(b)
88 J. MILSOM ET AL.
notable for the presence of the mud volcanoes
which are a feature of accretionary complexes
around the entire outer Banda arc (cf. Barber et al.
1986). There is no evidence of imbrication and the
carbonate platform is therefore assumed to have
developed on top of the accretionary complex after
convergence had ceased. High grade metamorphic
rocks (silicic schists, gneiss and migmatite) which
are exposed on the small islands of the Western
Province have been correlated by Charlton et al.
(1991b) with the Kobipoto Complex of Seram.
Seismic reflection data
Examples of multichannel seismic reflection
sections obtained by the research vessel R/v S o n n e
along lines which crossed the southern end of the
Aru trough were published by Schltiter & Fritsch
(1985), and sections obtained by the Dutch Snellius
II project across the trough NE of the Kai Islands
using a 4-channel system have been published by
Jongsma et al. (1989b).
BGR-93, the northernmost of the Sonne lines,
crossed the Am trough only a little south of Kai
Besar but did not extend as far west as the offshore
prolongation of the island (Fig. 2). The portion of
the section reproduced by Schltiter & Fritsch
(1985) shows Australian continental margin sedi-
ments dipping beneath unstratified material,
identified as melange, at a deformation front in the
deepest part of the trough, but ends in the west in a
region where water depths still exceed 2500 m and
so does. not show any of the structures at the
trough's western margin. A significant difference
between the section obtained on this line and those
?rom lines further south is the upfaulting of the
sediments of the continental margin by about 1 s
twt beneath the melange wedge about 15 km west
of the deformation front. Schliiter & Fritsch (1985)
interpreted the fault as a thrust but the evidence for
overstepping of the upper block on the lower is not
strong; the fault seems to be high angle but could be
either normal or reverse.
The southernmost of the Snellius lines (Line A2)
ran from east to west across the Aru trough about
30 km north of the northern tip of Kai Besar. In the
west it fell just short of the prolongation of the axis
of the island but it did reach the East Kai Ridge, a
bathymetric high between Kai Besar and the Am
trough which rises to within 500 m of the sea
surface. Along Line A2, the ridge is in water about
2000 m deep and the floor of the trough is at
a depth of about 3300 m (Jongsma et al. 1989b).
Underthrusting Australian continental margin was
not imaged; this may have been because the seismic
system used lacked the penetration of the system
mounted on the Sonne or because the trough fill is
considerably thicker (approaching 1 km), but the
sediments are virtually horizontal and the section
can be interpreted in purely extensional terms.
Arcward-dipping and arcward:thickening sedi-
ments do occupy the shallower basin between the
East Kai Ridge and the Kai Plateau but there is no
evidence for convergence there and the boundary
between the two appears to be a large normal fault.
The Kai region has also been covered by a high
quality seismic survey which included data
recorded to 8 s twt along lines which passed close
to the northern and southern ends of Kai Besar
(respectively UT-A and UT-B of Fig. 2). Figure 4
shows line-drawing interpretations of these
sections. The interpretation of line UT-B which
passed just south of the Kai Islands may be
compared in part with Line BGR-93. In the south-
east, beneath the shallow waters of the Arafura Sea,
sediments of the Australian margin have been
subject to normal faulting with slight rotation.
Faulting increases into the Am trough and the
eastern flank of the trough is underlain by fault
blocks, some of which are noticeably tilted. The
deepest part of the trough appears to coincide with
the outcrop on the sea floor of the deformation front
and further west the continental margin sequence
can be traced, deepening towards the west, as far
as the steep NW wall of the trough. There is a very
thin layer of flat-lying sediment above the sea-
floor trace of the deformation front but little or no
sediment cover on the accretionary complex within
the trough.
The abrupt change of sea-floor slope at the foot
of the NW wall of the trough strongly suggests
normal faulting, and this interpretation is reinforced
by the reappearance of the west-deepening con-
tinental margin reflectors within the upthrown
block. The throw on the fault appears to amount to
about 5 s twt. The seismic character of the material
which overlies the dipping reflectors and which
forms the southern extension of the Kai Plateau is
similar to that of the accretionary wedge in the
Aru trough, although a little more high-frequency
content has been preserved. The implications are
clear; the Aru trough is an extensional feature
bounded on both sides by normal faults, with the
main basin-controlling fault in the west. The Kai
Plateau, where crossed by the line, is part of the
accretionary prism and therefore equivalent to
southern Timor and perhaps to a part of the inner
slope down to the Timor trough, but has been
displaced upwards along major normal faults which
post-date the main convergent movements. The
time available for extensional faulting has been
very short; there are still volcanic eruptions in the
eastern Banda arc and a considerable level of
seismic activity although, as demonstrated by
McCaffrey & Abers (1991), fault plane solutions
suggest extension in the trough and strike-slip
 EXTENSION, COLLISION & CURVATURE,E BANDA ARC 89

o 1 O0 Km
I I I I ~lr I I I I

- ~ ~ East Kai
Kai Kecil ~,\~Y:~:~'-- ~ Ridge (N)
f..,.._ -- - ~ -~. ,~ . .,, ,/_
Line UT-A

.. "- .- --~ " ^'7"f'-.~-_ x~ -~F~--~'~

; ." t--~ Besar . r " -r~%" ~

6000 - : ~ , . " ' -.~ ~ "~-'=:
W S W -" ENE

(S) Line UT-B

0 50 1 O0 150 Krn
I __ I I I I I I I I I I I i I i

,- Z //°?, / , .
: ..~. A r u Trough ~ ~ ' ~ - ; ~ . . . . -~'L
Z000 ~ / / ',' .~.~_r2~;~\ - . . . . . . . . . . . . . . . ~ - ~ F " . . . . . .
~ ~ , " , ' 7 7 ~ --- ~ ' ~ = ~ ' - ~ _ ~ Australian :
I / 1 " " ,~/'/ " \ "~)g/'~, ~ ~ "--':"~ . . . . Shelf

6000 Plateau
---~ - ' , . \' ~ " \ ' / , , ' f " , / ~ "~"-
\ //
NW SE

Fig. 4. Line drawing interpretation of seismic lines north and south of the Kai Islands. Vertical scale in milliseconds
two-way travel time. Static corrections have been applied, replacing seawater with a layer of velocity 2000 m s-l,
so that vertical scales also give bathymetric depths in metres. Original sections provided by Union Texas (Kai) Inc.,
with the permission of Pertamina. See Fig. 2 for line locations.

rather than thrust motion in the Banda Sea. Gravity data

Although the present Am trough is a young feature,
the graben framework may well have been estab- The very strong gravity gradients associated with
lished in the Mesozoic, when the passive margin of the Banda arcs have been known ever since the
northwest Australia was formed by rifting. pioneering marine surveys of Vening Meinesz
The line drawing in Fig. 4 of the seismic section (1932, 1954). The gradient has since also been
obtained just to the north of Kai Besar, shows a defined on land, notably in eastern Timor where
very different picture. The East Kai Ridge, at about Bouguer gravity increases over a distance of 50 km
2000 m, is underlain by sediments which dip gently from close to zero along the axis of the island to
westward and are similar in seismic character to more than +100 reGal at points on the north coast
the Australian continental margin sediments on (Chamalaun et al. 1976; Kaye 1990). Similar
the southern line. There are no indications of an gradients between similar limiting values have been
accretionary complex beneath the ridge, which is reported from Seram (Milsom 1977) and from
consistent with their absence from the Snellius Tanimbar (Kaye 1990).
seismic sections of Jongsma et al. (1989b), which Apart from its size and steepness, the Banda
were obtained across the whole width of the trough. gravity gradient is also notable for its continuity
However, an apparent accretionary complex can be and smooth curvature in plan view, controlled
seen beneath the Kai Plateau, overlying west- largely by deep crustal structure and therefore seen
dipping reflectors which emerge at the sea floor in most clearly on regional maps of Bouguer gravity
the northern extension of Selat Nerong, the narrow (Fig. 5). The pattern appears to be little affected by
and shallow seaway between Kai Besar and Kai the presence of island groups, confirming the
Kecil. The region between this easternmost part source of the gradient as deep-seated. If the position
of the Kai Plateau and the East Kai Ridge is of the steepest slope can be taken as a guide, then
dominated by diffraction patterns which were in the Tanimbar group the main island is part of
probably generated by closely spaced normal the same geological zone as southern Timor and
faults. Kai Besar is therefore on the Australian side equivalents of northern Timor are to be found in the
of the deformation front and underlain by small islands closer to the Weber Basin (Kaye
Australian continental crust. The deformation front 1990).
is east of Kai Besar in the south but west of Land gravity measurements have been made
Kai Besar in the north. throughout the Kai Islands, with the majority of
90 J. MILSOM E T AL.
Fig. 5. The Bouguer gravity field of the Banda Sea region. Contours, at 25 mGal intervals, after Bowin et al. (1980)
with additional information from Woodside et al. (1989) and unpublished data. Stipple indicates regions of negative
Bouguer gravity.
stations situated within a few metres of the shore-
line which provided a ready, if approximate, height
reference (Kaye 1990). The Bouguer gravity map
(Fig. 6) has been prepared by combining the
onshore data with data from lines UT-A and UT-B.
The Bouguer gravity low departs from the bathy-
metric axis, which trends NNE into the Aru trough
and instead curves smoothly into the Seram trough,
passing to the west of the main islands of the group.
The contour pattern is dominated by a high which
peaks at more than +200 mGal at the east coast of
northern Kai Besar and by a broader low, with a
minimum value of less than -30 mGal, over the
western islands of the Central Province (Fig. 6).
From the low there is a moderately strong gradient
eastwards across Kai Kecil and a steeper one west-
wards across Kur and Fadol towards the Weber
Basin. This pattern suggests that the Tayandu
Islands belong to the same geological zone as
southern Timor and the basinal areas of northern
Seram. The main island of the Central Province,
Kai Kecil, is then gravitationally analogous to the
Timor trough or even to its outer (Australian) flank,
although there is no deep marine basin between
it and the other islands of the group, either to the
east or the west. The strong gravity high which
is associated with Kai Besar (Eastern Province)
has no parallel elsewhere in the outer Banda
arc or on the Australian shelf, and is exceptional,
in global terms, for any area underlain by con-
tinental crust. Marine gravity work has shown
that the high extends, with diminishing intensity,
almost to the coast of New Guinea (Woodside
et al. 1989).
The gravity profile interpreted in Fig. 7 was
extracted from the data shown in Fig. 6 along a line
oriented approximately ESE which crosses the
centre of the Bouguer maximum on Kai Besar. This
is the area in which gravity interpretation has most
to offer. A profile based on seismic line UT-B,
which passed south of Kai Besar, could have been
based on actual gravity observations for a greater
distance but would have been less interesting since
the interpretation would have closely resembled
that already published by Schltiter & Fritsch
(1985). On the other hand, complications would
have arisen in modelling a profile based on line
UT-A because it would run parallel to, rather than
across, the trend of the Bouguer contours immedi-
ately north of Kai Besar.
The profile selected has its origin at the boundary
between the West and Central provinces (i.e.
between the Tayandu group and Kur and Fadol),
close to the start of the Bouguer gravity rise
towards the Weber Basin. About 50 km east of the
origin, the line passes close to the western end of
 EXTENSION, COLLISION ~ CURVATURE,E BANDA ARC
13~2°E
1j~tk
Contour interval
10 reGal
I
Fig. 6. Bouguer gravity of the Kai Islands. Onshore data from Kaye (1990). Offshore data from profiles provided by
Union Texas (Kai) Inc., with the permission of Pertamina.
UT-A, but diverges thereafter since the seismic
line, necessarily, passes to the north of Kai Besar.
Despite this, the seismic section provides an
important constraint on the shallow structure used
in gravity modelling. Deep structures are less well
controlled but a satisfactory fit with observation is
difficult to obtain without supposing that only the
relatively thin outer margin of the Australian
continental crust has been subducted and that
further crustal thinning occurred during the
formation of the Ayu Trough.
The gravity low around the Tayandu Islands and
the high over Kai Besar require lateral changes in
density both at shallow depths, to account for the
steepness of the gradients, and in deep structure, to
explain the large differences in absolute levels. The
peak value of +200 mGal on Kai Besar is typical of
intra-oceanic arcs and oceanic islands but can be
achieved in an area underlain by continental crust
by raising the Moho to an extent which implies
both that the crust is thin and that the Aru trough
western boundary faults cut down to and displace
the Moho. This interpretation is consistent with the
seismic reflection data.
The gravity data also provide direct support for
the existence of an accretionary complex beneath
the Central Province. The gravity low requires both
crustal thickening, which is provided by the
doubling up of crust implied b y subduction, and
low density material, either young stratified sedi-
91
180
IN
6°S
0 50 Km
I I
ments or melange, near the surface. There are no
indications of significant thicknesses of young
undeformed sediments on seismic sections across
the Central Province but Charlton et al. (1989b)
noted the presence of mud volcanoes, which are
characteristic of melange terranes. The model
therefore incorporates an accretionary wedge; its
density and thickness can be varied widely and still
satisfy the gravitational constraints but the essential
features of the model are preserved through these
changes. The models are most easily brought into
agreement with the observed field if a density of
2.2 Mg m-3 or less is used for the main part of the
wedge. It seems fair to use the term melange for
what is obviously a very disordered assemblage,
but geological mapping in the possibly equivalent
terrane of southern Timor has shown that large
blocks of sediment have retained their original
internal structure (cf. Barber 1979).
The model includes a number of layers of sedi-
ment on the downgoing plate, the youngest of
which are incorporated into the melange at the
deformation front while the older are preserved to
considerable depths beneath the wedge. These
features are introduced to maintain compatibility
with the seismic sections and are not dictated by the
gravitational constraints. Gravity modelling can say
little about the degree to which sediments are
subducted or are incorporated with little deforma-
tion within the melange, and the densities assigned
92 J. MILSOMET AL.

200 f' Bouguer gravity (reGal)

I (Reductiondensity 2.2 Mg,m~) ~ ~

J°°t ~ Calculated ~ - .
ot, W N W . ~ ()bscrvcd • •
ESE

_ : ..."...,.,.,.o.:...e_
5~
s~lsm°~"ni[, Australian continental crust

0 50 100 Kilomctrcs 150 200 250

Fig. 7. Gravity profile and model cross-section across northern Kai Besar. The upper part of the accretionary
complex (above the dotted line) has been assigned a density contrast of-0.5 Mg m-3. with standard crust, corre-
sponding to a real density of 2.17 Mg m-3, the lower part a density contrast of -0.3 Mg m-3. Tertiary sediment layers
have density contrasts of -0.47, -0.42 and -0.3 Mg m-3 with standard crust. Arc crust, continental crust and Mesozoic
sediments have been assigned standard density.

to the sediments are not critical to the success or
failure of the model.
Discussion
There are three possible routes by which the
deformation front defined by seismic reflection
surveys south of Timor might link with its mirror
image defined by Jongsma et al. (1989a) in the
Seram trough. The seismic sections of Schltiter &
Fritsch (1985) which imaged the deformation front
in the eastern parts of the southern Am trough
eliminated the western route through the Weber
Basin as a possibility and could also be considered
as support for the eastern route along the Am
trough to a transfer at the Tarera-Aiduna fault. The
absence of the front from the seismic sections
presented by Jongsma et al. (1989b) is not strong
evidence against this latter possibility since the
weak source used lacked deep penetration in the
thick sediments. However, the UT-A seismic
section (Fig. 4) is interpreted as showing a west-
dipping thrust which separates stratified sediments
on Kai Besar from an accretionary wedge in the Kai
Central Province. This interpretation is supported
by the fact that the Eocene to Pliocene sediments on
Kai Besar have been tilted but not imbricated and
show no signs of deep burial. The faulting they
have suffered has been interpreted as extensional
by Charlton et al. (1991b) and this seems to be
in accord with the seismic data. It is almost
impossible to reconcile such an interpretation with
the existence of a true deformation front beneath
the sediments in the Aru trough since fault dis-
placement of the order of l0 km followed by con-
siderable erosion would be needed to expose this
front on the upthrown block, and these processes
would have had to have taken place in the short
period since the subduction zone locked.
The trace of past subduction must therefore pass
through Selat Nerong and then curve into the Seram
trough. The fault locations in Fig. 8, which illus-
trate this solution, are inevitably schematic, being
constrained by seismic data along the seismic lines
of Fig. 4, but elsewhere being drawn on the basis of
bathymetric features which are themselves not well
controlled. Almost certainly the true pattern is more
complex, but such complexity would not affect
the conclusion that the deformation front broadly
parallels the curvature of the Bouguer gravity
contours and does not follow the bathymetric trends
along the Am trough
The cause o f the extreme curvature of the Banda
arc is one of the major unsolved problems of SE
 EXTENSION, COLLISION • CURVATURE, E BANDA ARC 93

5°S -

J k.~ /

0 100 K m

c,.ar
...

1 °E~" 133°E

Fig. 8. The Kai-Tanimbar region, showing suggested interaction of extensional faulting with earlier thrust faulting.
Strong black lines show thrust faults of the deformation front, narrower lines mark possible locations of faults of the
Aru trough extensional system, as suggested by bathymetry. Because of the absence of major geological differences
between northern and southern Kai Besar (cf. Charlton et al, 1991b), the narrow isthmus which links the two is
attributed to transfer faulting associated with the extensional system, rather than with the passage of the thrust from
the Aru trough to Selat Nerong.

Asian geology (cf. Linthout e t al. 1991). Evaluation
of the processes involved has been hampered by
the lack of a universally accepted model of arc
evolution. There are three broad possibilities, one
of which is that the curvature has been a feature
of the arc since its inception. This has found
little favour (i) because of the problems involved
in incorporating a tightly curved arc into recon-
structions of the area during the Neogene,
(ii) because of the difficulty of transporting some
Australian continental fragments to their present
positions within the arc, (iii) because of the
evidence for some 74 ° counter-clockwise rotation
of Seram (Haile 1981) and (iv) because of the
evidence for continuing N - S compression and
E - W extension in the Banda Sea (McCaffrey &
Abers 1991).
An alternative possibility is that the arc was
assembled relatively recently from two or more
elements with quite different previous histories.
Such a solution is attractive since it poses few
mechanical difficulties and places few constraints
on Tertiary reconstructions, but it has always
encountered an obstacle in the evidence for
continuity in deep structures provided by regional
gravity maps. If it could be shown that the sub-
duction trace coincided with the Aru trough and
was offset along the Tarera-Aiduna fault (Fig. 3a),
94 J. MILSOM ETAL.
this objection might be discounte&since the offset
w o u l d imply a major discontinuity in the upper, as
well as the lower, plate. The seismic reflection
evidence that curvature o f the deformation front
m a t c h e s the curvature o f the B o u g u e r gravity
contours, and the detailed gravity interpretation
w h i c h c o n f i r m s this c o n c l u s i o n , nullify this
argument. Whatever the geometrical difficulties
involved in bending not merely a crustal sliver but
References
ACHDAN, A. & TURKANDI,T. 1982. Preliminary geologic
map of the Kai (Tayandu and Tual) quadrangles,
.Maluku, Indonesia (Scale 1:250,000). Geological
Research and Development Centre, Bandung.
AUDLEY-CHARLES, M. G. & MILSOM, J. 1974. Comment
on 'Plate convergence, transcurrent faults and
internal deformation adjacent to south east Asia' by
T. J. Fitch. Journal of Geophysical Research, 79,
4980-4981.
, CARTER, D. J., BARBER, A. J., NORWCK, M. S. &
TJOKROSAPOETRO, S. 1979. Reinterpretation of the
geology of Seram: implications for the Banda Arcs
and northern Australia. Journal of the Geological
Society, London, 136, 547-568.
BARBER, A. J. 1979. Structural interpretation of the island
of Timor. SEAPEX Proceedings, 4, 9-21.
, TJOKROSAPOETRO,S. & CHARLTON,T. R. 1986. Mud
volcanoes, shale diapirs, wrench faults and
melanges in accretionary complexes, eastern
Indonesia. AAPG Bulletin, 70, 1729-1741.
BOWIN, C., PURDY, G. M., JOHNSTON, C., SHOR, G.,
LAWVER, L. HARTONO, H. M. S. & JEZEK, E 1980.
Arc-continent collision in the Banda Sea region.
AAPG Bulletin, 64, 868-915.
CHAMALAUN, E H., LOCKWOOD, K. & WHITE, A. 1976.
The Bouguer gravity field and crustal structure of
eastern Timor. Tectonophysics, 30, 241-259.
CHARLTON,T. R., BARBER, A. J. & BARKHAM,S. T. 1991a.
The structural evolution of the Timor collision
complex, eastern Indonesia. Journal of Structural
Geology, 13, 489-500.
, KAYE, S. J., SAMODRA,H. & SARDJONO. 1991b. The
geology of the Kai Islands: implications for the
evolution of the Aru Trough and the Weber Basin.
Marine and Petroleum Geology, 8, 62-69.
FITCH, T. J. & HAMILTON,W. 1974. Reply to comments on
Audley-Charles, M. G. and Milsom, J. on 'Plate
convergence, transcurrent faults and internal defor-
mation adjacent to south east Asia' by T. J. Fitch.
Journal of Geophysical Research, 79, 4982-4985.
HALLE, N. S. 1981. Palaeomagnetic evidence and the
geotectonic history and palaeogeography of Eastern
Indonesia. In: BARBER, A. J. & WIRVUSUJONO, S.
(eds) The Geology and Tectonics of Eastern
Indonesia. Geological Research and Development
Centre, Bandung, Special Publication 2, 81-87.
an associated subducted slab, it seems that this is
what has happened in the Banda arc.
We thank Union Texas (Kai) Inc. for providing access to
their unpublished seismic sections and marine gravity
data and permitting reproduction of our line drawing
interpretations of the seismic data. Sardjono publishes
with the permission of the Director, Geological Research
and Development Centre, Bandung.
HAMILTON,W. 1979. Tectonics of the Indonesian region.
US Geological Survey Professional Paper, 1078.
JONGSMA, D., WOODSIDE,J. M., HUSON, W., SUPARKA,S.
KADARISMAN, D. 1989a. Geophysics and tenta-
tive Late Cenozoic seismic stratigraphy of the
Banda Arc-Australian continent collision along
three traverses. Netherlands Journal of Sea
Research, 24, 205-229.
--, HUSON, W., WOODSIDE, J. M., SUPARKA, S.,
SUMANTRI, T. & BARBER, A. J. 1989b. Bathymetry
and geophysics of the Snellius II triple junction and
tentative seismic stratigraphy and neotectonics of
the northern Aru Trough. Netherlands Journal of
Sea Research, 24, 231-250.
KAYE, S. K. 1990. The structure of eastern b~donesia:
an approach via gravity and other geophysical
methods. PhD thesis, University of London.
LINTHOUT, K., HELMERS, H. & ANDRIESSEN,E A. M.
1991. Dextral strike-slip in central Seram and 3-
4.5 Ma Rb/Sr ages in pre-Triassic metamorphics
related to early Pliocene counterclockwise rotation
of the Buru-Seram microplate. Journal of SE Asian
Earth Sciences, 6, 335-342.
M¢CAFFREY, R. 1988. Active tectonics of the eastern
Sunda and Banda arcs. Journal of Geophysical
Research, 93, 15163-15182.
-- & ABERS, G. A. 1991. Orogeny in arc-continent
collision: the Banda Arc and western New Guinea.
Geology, 19, 563-566.
MILSOM, J. 1977. Preliminary gravity map of Seram,
eastern Indonesia. Geology, 5, 641-643.
RICHARDSON, A. N. & BLUNDELL,D. J. 1996. Continental
collision in the Banda arc. This volume.
SCHLt3TER, H. U. & FR~TSCH, J. 1985. Geology and
tectonics of the Banda arc between Tanimbar Island
and Aru Island (Indonesia). Geologisches Jahrbuch,
E30, 3-41.
VENING MEINESZ, F. A. 1932. Gravity expeditions at sea,
Vol I. Netherlands Geological Commission, Delft.
-- 1954. Indonesian archipelago-a geophysical study.
Geological Society of America Bulletin, 65,
143-164.
WOODS1DE,J. M., JONGSMA,D., THOMMERET,M., STRANG
VAN HEES, G. L. & PUNTEDEWO. 1989. Gravity and
magnetic measurements in the eastern Banda Sea.
Netherlands Journal of Sea Research, 24, 185-203.
 Pre-Cretaceous evolution of SE Asian terranes

I. M E T C A L F E
Department of Geology & Geophysics, University of New England,
Armidale NSW 2351, Australia

Abstract: During the last decade, a wide range of geological and geophysical data has led to the
recognition of various continental terranes in SE Asia which, on tectonostratigraphic, palaeo-
biogeographic and palaeomagnetic grounds are 'suspect' or allochthonous in nature. Some of the
recognized terranes may be composite and there is some disagreement with regard to the number
of terranes and their boundaries. The continental terranes are bounded by sutures (representing
former oceans), by narrow mobile belts or major fault zones. Comparative studies of the
stratigraphy, palaeontology, and palaeomagnetism of the various pre-Cretaceous continental
terranes of East and SE Asia suggest that they were all derived directly or indirectly from
Gondwanaland. Asian continental terranes that are placed on the India-Australian margin of
Gondwanaland in the Early Palaeozoic include Tarim (here regarded to include the Kunlun and
Ala Shan terranes), Qaidam, Indochina/East Malaya (which includes the Qamdo-Simao block of
western China), North & South China, Sibumasu, Qiangtang, Lhasa, Kurosegawa, NW and SE
Hainan, West Burma and the Woyla terranes. The evolution of Gondwanaland and Tethys during
the Palaeozoic and Mesozoic involved the rifting of continental slivers/fragments from northern
Gondwanaland, and the northwards drift and amalgamation/accretion of these to form proto-East
and SE Asia. Three continental slivers were rifted from the northern margin of Gondwanaland in
the Early to Late Devonian (North China, South China, Indochina/East Malaya/Qamdo-Simao,
Qaidam and Tarim terranes); Early-Middle Permian (the Cimmerian continent including the
Sibumasu and Qiangtang terranes and possibly NW and SE Hainan); and Late Triassic to Late
Jurassic (Lhasa, West Burma and Woyla terranes). The northwards drift of these terranes was
accompanied by the opening and closing of three successive oceans, the Palaeo-Tethys,
Meso-Tethys and Ceno-Tethys.
Assembly of Gondwanaland-derived Asian terranes began with the amalgamation of South
China and Indochina/East Malaya along the Song Ma/Song Da zone during the Late
Devonian/Early Carboniferous to form 'Cathaysialand'. Palaeomagnetic, climatic and bit-
geographic data indicate that Cathaysialand and North China were located within the
Palaeo-Tethys at low northern/equatorial latitudes during the Late Carboniferous and Permian.
Palaeomagnetically determined palaeolatitudes are consistent with the development of tropical
Cathaysian floras on these terranes. The Tarim, Kunlun, Qaidam and Ala Shan terranes accreted
to Kazakstan/Siberia in the Permian.
A major episode of rifting occurred on the northern margin of Gondwanaland in the Late
Carboniferous-Early Permian and the Cimmerian continent separated in the late Early Permian
resulting in the opening of the Meso-Tethys. Suturing of Sibumasu and Qiangtang to
Cathaysialand occurred in the Late Permian-Triassic, closing a major branch of the Palaeo-
Tethys. South and North China amalgamated and then accreted to Laurasia by Late
Triassic-Early Jurassic times. The highly disrupted Kurosegawa terrane of Japan, possibly
derived from Australian Gondwanaland, accreted to Japanese Eurasia, also in the Late Jurassic.
The Lhasa, West Burma and Woyla terranes rifted from NW Australian Gondwanaland in the
Late Triassic to Late Jurassic and drifted northwards during the Jurassic and Early Cretaceous as
the Ceno-Tethys opened and the Meso-Tethys was destroyed by subduction beneath Eurasia.
Accretion of these terranes to proto-SE Asia occurred in the Cretaceous. The SW Borneo and
Semitau terranes were derived from the South China/Indochina margin by the opening of a
marginal basin in the Cretaceous which was subsequently destroyed by southwards subduction
during the rifting of the Reed Bank-Dangerous Grounds terrane from South China when the
South China Sea opened. The NW and SE Hainan terranes, which formed part of Early
Palaeozoic and possibly Late Palaeozoic Gondwanaland, reached their current positions, relative
to South China, sometime in the Jurassic-Cretaceous.

The pre-Cretaceous evolution o f SE Asia has
received m u c h attention during the last decade and
has been the principal focus of two recent I G C P
projects, IGCP 224 (Pre-Jurassic Evolution o f East
Asia) and I G C P 321 ( G o n d w a n a Dispersion and
Asian Accretion). The evolution of East and SE
Asian terranes, in the framework of the evolution of
Gondwanaland, Laurasia, Pangaea and Tethys, has
important scientific implications for both regional
and global earth sciences. Detailed information on

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, 97

Geological Society Special Publication No. 106, pp. 97-122.
98 I. METCALFE
the movements (particularly cross-latitudinal) of
continental blocks and the changing spatial
distribution of seas and oceans provides a basis for
investigating globally important problems. Past
ocean currents and climatic patterns have been
very important in determining biogeographical
distributions of both marine and land organisms
on the Earth and a knowledge of changing
continent/ocean configurations helps us model
these past climates and current patterns and leads to
a greater understanding of the observed changes
in both the distributions of organisms, and sedi-
mentation patterns in the evolving seas and oceans.
This paper forms a contribution to the lUGS Global
Sedimentary Geology Program's Project Pangaea
which is primarily concerned with modelling
global climate during the existence of Pangaea. Of
importance to this modelling is the continent/ocean
configuration, and more specific to East and SE
Asian terranes, was Tethys an empty open gap into
Panthalassa or did it contain continental fragments,
and if so what positions did these occupy relative
to the rest of Pangaea? Biologists, botanists and
zoologists are becoming increasingly interested in
SE Asian terrane movements, particularly in the
Mesozoic and Cenozoic, because the movements of
continental fragments may have provided land
'rafts' or chains of islands that influenced land
plant and animal evolution and dispersal in the
region. Another rationale for detailed studies of
terrane evolution in the Asian region is the
increasing economic importance of these regional
studies in understanding metallogenesis and the
development and evolution of hydrocarbon and
coal-bearing sedimentary basins in the region.
It is now well established that continental SE
Asia (together with large parts of eastern Asia)
comprises a complex assembly of continental
terranes (Figs 1 and 2) which are entirely allochtho-
nous to central and northern Asia (Siberia and
Kazakhstan). All the SE Asian terranes are inter-
preted to have been derived directly or indirectly
from Gondwanaland and their Palaeozoic and
Mesozoic history involved the rifting of terranes
from the northern margin of Gondwanaland, their
northwards drift and amalgamation/accretion to
form proto-SE Asia (Seng0r 1979, 1984; Ridd
1980; Mitchell 1981; Stauffer 1983; Audley-
Charles 1983, 1984, 1988, 1991; Metcalfe 1986,
1988, 1990, 1993, 1994a, c ; Klimetz 1987; Seng/3r
et al. 1988; Burrett et al. 1990; Nie et al. 1990). The
northwards translation of terranes involved the
opening and closing of successive Tethys oceans.
Previous workers have recognized a two Tethys
ocean system comprising Palaeo-Tethys ('Old
Tethys') and Neo-Tethys or 'New Tethys' (e.g.
Seng6r 1979, 1984, 1989). The rift-drift-
suturing (amalgamation/accretion) history of the
SE Asian terranes is here considered within a
framework of a three Tethys ocean system
following Metcalfe (1994a, c). It is proposed that
three continental slivers rifted from the northern
margin of Gondwanaland during the Palaeozoic to
Mesozoic, opening and closing three successive
Tethys oceans, the Palaeo-Tethys, Meso-Tethys
and Ceno-Tethys. The Palaeo-Tethys is broadly
equivalent to the Palaeo-Tethys of previous
workers, although timings of the closure of
the different branches of Palaeo-Tethys differ in
detail. The Meso-Tethys and Ceno-Tethys are
approximately equivalent in age to what has
previously been referred to as 'Neo-Tethys' but are
here regarded as successive oceans developed by
the Triassic-Jurassic rifting of a third continental
sliver from northeast Gondwanaland.
SE Asian pre-Cretaceous terranes and
their origins
The pre-Cretaceous continental terranes of East
and SE Asia are shown in Figs. 1 and 2. Some of
these terranes may be composite, but where the
composite nature is not well established or is
controversial, they are treated as single units. The
Cretaceous and Cenozoic evolution of the small
terranes distributed around the South China Sea and
in eastern Indonesia (Fig. 2), including West
Irian Jaya, Buru-Seram, Buton, Banggai-Sula,
Obi-Bacan, North Palawan, Spratley Islands-
Dangerous Ground, Reed Bank, Luconia,
Macclesfield Bank, Paracel Islands, Kelabit-
Longbowan, Mangkalihat, Paternoster, West
Sulawesi, East Sulawesi and Sumba has been
discussed by Metcalfe (1990) and will not be
repeated here. The pre-Cretaceous terranes
discussed below include those which now
constitute mainland SE Asia. Some discussion of
other East Asian terranes which have a bearing
upon the evolution of SE Asia is also included.
South China
The South China block is regarded as composite by
some authors (e.g. Hsu et al. 1988, 1989, 1990).
Hsu et al. (1988; 1989) recognized component
Yangtze and Huanan Blocks separated by the
'Xiangganzhe Mesozoic suture'. This interpretation
has been challenged by Rogers (1989) and Rowley
et al. (1989), and geochronological data from
Anhui and Jiangxi Provinces suggest that the suture
recognized by Hsu is in fact of Proterozoic age
(Chen, J. et al. 1991). Hsu et al. (1990) and Chen et
al. (1993) recognize four component blocks, the
Yangtze, Xianggui, Cathaysia and Dongnanya
blocks separated by three melange zones (Fig. 3).
 SE ASIAN TERRANE EVOLUTION 99

TARIM
NORTH
CHINA

KT
SUTURES
SOUTH Q SongMa
CHINA Q Aibi-Xingxing
Q Xiliao-He
G Kunlun
G Qinling-Dabie
(~) Jinshajiang
HT Q Lancangjiang
O Banggong
O IndusYarlungZangbo
Q Nan-Uttaradit
G Raub-Bentong
ShahBoundary
Woyla
Meratus
WOYLA"
G Boyan
TERRANES G Changning-Menglian
(~ Ailaoshan
Fig. 1. Distribution of principal continental terranes and sutures of East and SE Asia. EM, East Malaya; WB, West
Burma; SWB, South West Borneo; S, Semitau Terrane; HT, Hainan Island terranes; L, Lhasa Terrane; QT, Qiangtang
Terrane; QS, Qamdo-Simao Terrane; SI= Simao Terrane, SG, Songpan Ganzi accretionary complex; KL, Kunlun
Terrane; QD, Qaidam Terrane; AL, Ala Shan Terrane; KT, Kurosegawa Terrane.

Palaeomagnetic data suggesting a separation of the
Yangtze and Huanan blocks by a latitude difference
of 8.3 ° in the Triassic (Chen et al. 1993) have been
used to support the pre-Triassic composite nature of
South China. However, the shallow inclinations
reported from the Yangtze and 'Huanan' blocks
(-11.8 and -13.0 respectively) are not significantly
different (even w i t h small o~95s), and do not
confirm a Triassic separation of these regions. The
occurrence of identical endemic Devonian fish
faunas from both the Yangtze region and southeast
South China, which define a South China Province
(Young 1993) also indicate a unified South China
block at that time. South China is here regarded as
a single tectonic unit since the Palaeozoic following
the arguments of Rogers (1989), Rowley et al.
(1989) and Chen, H. et al. (1991). Its boundaries
with North China and the Indochina block are the
Qin Lin-Dabei suture, and Ailaoshan and Song Ma
Sutures, respectively.
Lower Palaeozoic shallow marine faunas of
South China have affinities with those of northeast
Gondwanaland and belong to the Asia-Australian
and Austral realms in the Cambrian and Ordovician
respectively (Yang, J. 1994; Li 1994). These faunal
affinities, together with stratigraphic comparisons
suggest that South China had its origin on the
Himalaya-Iran region of the Gondwanaland margin
(Burrett et al. 1990; Nie et al. 1990; Nie 1991,
1994). The Lower Palaeozoic placement of South
China adjacent to the Himalaya-Iran region is
consistent with the palaeomagnetic evidence which
places South China in mid southern palaeolatitudes
in the Ordovician (Lin et al. 1985; Burrett et al.
1990; Fang et al. 1990; Metcalfe 1990). Hoffman
(1991) also placed the East and SE Asian blocks
100 i. METCALFE

PHILIPPINE zo.
SEA PLATE

0 500 km
SEA

LEGEND " "

X X X X Suture ~
,6. _¢~..,,x Thrust

.,L Inactive Trench

o ~ Active Trench
~!~i~iiii!i~!iii:i~i¸ ; i
Terrane Boundary ~ii ¸ ::~,i ~i i: ¸
.... Inferred Terrane Boundary
J J OceanIc Crust

~ AccretedCrust
~ 1 Extended Continental Crust

Continental Blocks and Fragments IOOE

Fig. 2. Distribution of continental blocks and fragments (terranes) and principle sutures of SE Asia (modified after
Metcalfe 1990). 1. South China; 2. Indochina; 3. Sibumasu; 4. East Malaya; 5. West Burma; 6. SW Borneo; 7.
Semitau; 8. Sikuleh; 9. Natal; 10. West Irian Jaya; 11. Buru-Seram; 12. Buton; 13. Bangai-Sula; 14. Obi-Bacan; 15.
North Palawan; 16. Spratley Islands-Dangerous Ground; 17. Reed Bank; 18. Luconia; 19. Macclesfield Bank; 20.
Paracel Islands; 21. Kelabit-Longbowan; 22. Mangkalihat; 23. Paternoster; 24. West Sulawesi; 25. East Sulawesi; 26.
Sumba; 27. Banda Allochthon; 28. Qiongzhong and Yaxian terranes of Hainan; 29. Simao terrane.

adjacent to the NW margin of Australia in his

Neoproterozoic supercontinent reconstruction.
es -~ -'--. ,, st" J
ss.j s a~'~'~. "",,,, _ S Recent comparisons of the Proterozoic basement of
South China with Australia and Laurentia led Li
et al. (1994) to propose an alternative model where-
by South China is placed between Eastern Australia
and Laurentia in the Late Proterozoic, forming part
.,,- of Rodinia. Such a position would be at variance
with an Early Palaeozoic position adjacent to
...-...... ./i Himalayan India, and Proterozoic palaeomagnetic
data are required to test this proposition. The Upper
Sinian and Palaeozoic tectonostratigraphy of South
China is also not comparable with Eastern
Australia. A northern Gondwanaland margin origin
for South China is favoured here.
[ . ~ _ . . _ ~ Island i i

Fig. 3. Component blocks and melange zones of South Sibumasu

China proposed by Chen et al. (1993). The Sibumasu block (Metcalfe 1984, 1988) is
bounded to the west by the Shah Boundary Fault
 SE A S I A N T E R R A N E E V O L U T I O N |01

and the Andaman Sea basin and to the southwest by • Deqin

,. / 'TIBET
t.. - ~ Late Permian-Triassic
the Woyla suture in Sumatra. Its eastern boundary Volcanic Arc
Lancangjlang
is formed by the Raub-Bentong suture in Peninsular Suture ' ~ Late Permian-Triassic
Granite
Malaysia, and the Changning-Menglian and i Ophiolitic rocks and
Lancangjian Sutures in western Yunnan. The f oceanic sediments
eastern boundary of Sibumasu in North Thailand MYANMAR
i
,i
~ Chongshan
metamorphic belt
and Burma is still not clear. The northwards J

extension of the Raub-Bentong suture has been

regarded by many authors to be the Nan-Uttaradit 120 km
I
D
suture in Thailand. This suture appears to bend
northeastwards to be truncated by the Red ~t SOUTH CHINA
i ~ s
River-Ailaoshan zone. In western Yunnan, the i Baoshan • TERRANE
Changning Menglian suture zone (Wu & Zhang
Changning
1987; Liu Benpei et al. 1991; Wu Genyao 1993; '; BAOSI-IAN Ailaoshan
Wu Haoruo et al. 1995) which separates the ".-- BLOCK / Suture
Baoshan/Tenchong blocks (disrupted parts of
Sibumasu) from the Qamdo-Simao terrane (Fig. 4) Menglian Suture
appears to continue southwards into Burma and SIMAO'
then possibly into the Chiang Rai region of north BLOCK
Thailand. Deep-marine ribbon-bedded cherts (Fang
• Simao --.\~,
Chert) in the Chiang Rai region have been dated by
radiolarians as Devonian to middle Permian v z/ ~,-
(Sashida et al. 1993). These are associated with ' ". VIETNAM
basic volcanics which may be subduction-related , I_AOS "-
MYANMAR ' ,
(Barr et al. 1990). Some authors have suggested i
i
f
i
that the Sibumasu terrane is composite and
have recognized a western 'Mergui Platelet'
Fig. 4. Principal blocks and suture zones of western
(Cameron et al. 1980), 'Phuket terrane' (Cooper
Yunnan (after Wu et al. 1995).
et al. 1989) or 'Mergui Group nappe' (Mitchell
1992) characterised by the presence of
Carboniferous-Permian diamictites. There
seems however to be little basis for separating the 1985; Metcalfe 1986, 1988, 1990, 1991, 1993;
diamictite-bearing 'terrane' from Sibumasu and the Burrett et al. 1990; Shi & Waterhouse 1991; Racey
implied thrust contact between the Mergui Group et al. 1994; Archbold & Shi 1995) strongly suggest
and equivalents and other Palaeozoic strata of a NW Australian origin for the Sibumasu terrane.
Sibumasu has not been demonstrated. In fact, in This is supported by the presence of Upper
northwest peninsular Malaysia, a stratigraphic Carboniferous-Lower Permian glacial-marine
contact can be unequivocally demonstrated diamictites (Stauffer & Mantajit 1981; Stauffer &
(Hutchison 1993). However, new geochronological Lee 1989, Jin 1994a, b; Wopfner 1994), Lower
data from western Thailand (Ahrendt et al. 1993, Permian cool-water faunas and 8180 cool-water
1994; Hansen et al. 1994) suggest that the high indicators (Ingavat & Douglass 1981; Waterhouse
grade metamorphic rocks previously regarded as 1982; Rao 1988; Fang & Yang 1991) which
Precambrian basement in NW Thailand, have indicate proximity to the Late Palaeozoic
depositional ages of less than 600 Ma. New K-Ar, Gondwanaland glaciated region. A Glossopteris
Rb-Sr and U-Pb ages and structural observations flora has also recently been reported south of
also indicate very young (Tertiary) cooling ages Baoshan in western Yunnan (Wang & Tan
for the metamorphic 'core complex' but older 1994) and typical Lower Permian Gondwanaland
(Mesozoic) ages for deformational ages in over- spores, including P r i m u s p o l l e n i t e s levius and
lying low grade metamorphic rocks. This suggests Schauringipollenites maximus, have been recov-
the possibility of eastwards directed thin-skinned ered from near Tenchong, western Yunnan (Yang,
tectonics in western Thailand during the Tertiary. If W. 1994). The palynoflora reported from the
this is the case, the geometry and boundaries of the Tenchong block (Sibumasu) is also remarkably
Sibumasu terrane may have been disrupted and the similar to palynofloras described from the Collie
Palaeozoic and Mesozoic strata may well have been basin of Western Australia (Yang, W. 1994). Gross
detached from their original basement. stratigraphical comparisons between Sibumasu and
Distinctive Gondwanaland faunas (Cambrian to NW Australia, especially with the Canning Basin
Lower Permian) with NW Australian affinities on (Metcalfe 1992, fig. 3), also show similarities
Sibumasu (Archbold et al. 1982; Burrett & Stait consistent with Sibumasu having been positioned
102
40
30,
LU
a 20.
E3
}-- 10.
}-
< o
._1
O -lo,
LM
< -20.
._1
,c~ -30
n Australia
-40
Devonianl CarboniferouslPermianlTriassicl
Fig. 5. Palaeolatitude versus time plots for the Sibumasu Block (after Van der Voo 1993).
outboard of NW Australian Gondwanaland in the
Palaeozoic. In addition, Palaeozoic palaeomagnetic
data indicate palaeolatitudes (Fig. 5) consistent
with a position off NW Australian Gondwanaland
in the Devonian, Carboniferous and Early Permian
(Bunopas 1982; Bunopas et al. 1989; Wu et al.
1989; Metcalfe 1990, 1994a, b; Huang & Opdyke
1991; Van der Voo 1993).
I n d o c h i n a - E a s t Malaya block
This block is bounded to the northeast by the Song
Ma and Ailaoshan suture zones and to the west by
the Changning-Menglian suture in Yunnan, the
Nan-Uttaradit suture in Thailand and the Raub-
Bentong suture in Malaysia. The block probably
extends northwards as the Qamdo-Simao or Simao
block (Wu 1993; C h e n & Xie 1994) which is
regarded as a separate, possibly South China-
derived, tectonic unit by Wu et al. (1995). The
Upper Palaeozoic and Mesozoic faunas and floras
are Cathaysian/Tethyan types, have affinities to
those of South and North China, and show no
relationship to Gondwanaland (Metcalfe 1986,
1988). Lower Palaeozoic faunas are poorly known,
and their affinities have yet to be clearly demon-
strated. The Precambrian core of the block
(Kontum Massif) comprises granulite facies rocks
exposed in Vietnam, and it has been suggested that
this may have originally formed part of the
Gondwanaland granulite belt (Katz 1993). It seems
likely that Indochina (including the East Malaya
and Simao terranes) had its origin on the northern
I. METCALFE
Sibumasu
Block
(Ref. at 18N, 95E)
-.~ Observations
::: E. Sumatra
E. Malaya
4 USTRALI~---A" --0- Predicted from
South China
--dr Predicted from
Jurassic I Cretaceous I Tertiary
Gondwanaland margin but this must remain largely
a speculation until we have constraining palaeobio-
geographical and palaeomagnetic data.
West Burma block
Formerly known as the 'Mount Victoria Land'
Block (Mitchell 1989), this terrane is here referred
to as the West Burma Block to avoid confusion with
Victoria Land in Antarctica. Mitchell (1993) has
reinterpreted Western Burma (together with the
Woyla Group of Sumatra and the Meratus ophiolite
of Borneo as a lower Cretaceous mafic arc,
produced by SW-directed subduction, which was
then thrust on to the Asian margin in the late Early
Cretaceous. Following Metcalfe (1990), the author
prefers to regard Western Burma as a probable con-
tinental fragment now largely obscured by younger
sediments and volcanic rocks. This block has a
pre-Mesozoic schist basement overlain by Triassic
turbidites and Cretaceous ammonite-bearing shales
and limestones in the Indoburman Ranges and by
an upper Mesozoic-Cenozoic arc association in the
Central Lowlands of Burma. There is as yet no
direct evidence for the origin of this block but it
sutured to Sibumasu in Cretaceous times. The West
Burma Block is considered (Metcalfe 1990) a good
candidate for part of the continental sliver that
provided a source for sediments derived from the
northwest in Timor during the Triassic, and which
must have rifted from Gondwanaland in the Late
Jurassic. This interpretation however remains
speculative.
 SE ASIAN TERRANE EVOLUTION
S W B o r n e o a n d S e m i t a u blocks
The origins of these two blocks, separated by the
Boyan suture, are poorly constrained. Upper
Carboniferous and Lower Permian fusulinid faunas
and other Lower Permian microfossils are similar
to those of South and North China and do not have
affinities with those of Malaya or Thailand
(Vachard 1990). Triassic and Jurassic floras and
faunas have affinities with South China, Indochina,
Japan and the Philippines (Hayami 1984; Kimura
1984; Kobayashi & Tamura 1984). There does not
appear to be any Gondwanaland connections in the
faunas and floras of these two blocks. A South
China/Indochina origin seems likely as proposed by
Rammlmair (1993) but reconciliation between sub-
stantial progressive counter-clockwise rotation of
SW Borneo during the Cenozoic and Cretaceous
and younger equatorial palaeolatitudes with the
various tectonic models is required.
H a i n a n I s l a n d terranes
Hainan Island appears to represent at least two
tectonostratigraphic terranes but the nature and
boundaries of these are still controversial. Yu
(1989) and Yang et al. (1989) recognise two small
terranes, the northern Quiongzhong and southern
Yaxian terranes bounded by the east-west
Jiushuo-Lingshui Fault. Chen et al. (1993) how-
ever, interpreted the Island as a Mesozoic orogenic
belt created by the collision of the Huannan (SE
China) and Dongnanya (SE Asia) blocks. Gt~rtir &
Sengrr (1992) interpreted the island as originally
forming a part of the Sibumasu terrane, being
separated from other Sibumasu components in the
Mesozoic by strike-slip faulting. The Lower
Palaeozoic faunas of the island indicate links
with northeast Gondwanaland (Sun 1963; Lin &
Jago 1993) and the presence of possible glacial-
marine diamictites and Permian faunas with
Gondwanaland affinities suggests a northeast
Gondwanaland origin (Yang et al. 1989; Yu 1989).
Recent field investigations in Hainan (Metcalfe et
al. 1994) have confirmed the probable presence of
two terranes on Hainan, but the boundary of these
is interpreted to be the SW-NE-oriented Baisha
Fault and not the east-west Jiushio-Lingshui Fault.
The Late Palaeozoic Gondwanaland origin of these
terranes remains problematic.
Woyla terranes
Several small continental fragments, including the
Sikuleh and Natal blocks, now located along the
southwest margin of Sumatra were accreted to
'Sundaland' in the Cretaceous and were called the
Woyla terranes by Cameron et al. (1980) and
Pulunggono & Cameron (1984). Stratigraphic
103
studies of part of the Woyla terranes in the Padang
area (Yancey & Alif 1977) reveal similarities with
the stratigraphy of the Exmouth Plateau of the NW
Australian shelf (Grriir & Seng6r 1992) and
limited palaeomagnetic data from the Sikuleh block
(Haile 1979) suggest palaeolatitudes of 26°S for the
Late Triassic and 10°S in the Late Mesozoic which
is consistent with a NW Australian origin of these
terranes.
K u r o s e g a w a terrane
The Japanese Islands are composed of a rifted
Asian basement and a huge accretionary complex,
produced by long-lived westwards subduction of
the Pacific and Philippine Sea plates beneath Asia,
over which the volcanic arc has migrated
eastwards. A number of accreted terranes has
been recognized within the accretionary complex
including oceanic (seamount) and continental
allochthonous terranes. The Kurosegawa con=
tinental terrane is a highly attenuated and disrupted
composite terrane of predominantly Palaeozoic age
now sandwiched between two Mesozoic terranes in
the Chichibu Belt of SW Japan. It comprises a
Lower Palaeozoic basement of igneous and
high-grade metamorphic rocks, Silurian-Devonian
volcaniclastic sediments with carbonate blocks,
serpentinite, a Upper Palaeozoic-Mesozoic chaotic
complex and well-bedded Upper Palaeozoic-
Lower Mesozoic cover sediments (Aitchison et al.
1991). Lower and Middle Palaeozoic faunas and
floras have affinities with Australia and suggest a
north Australian Gondwanaland margin origin for
this exotic terrane (Yoshikura et al. 1990; Aitchison
1993). Studies of chromium spinels from ser-
pentinites of the terrane (Hisada et al. 1994)
suggest that the serpentinites were emplaced in a
forearc setting during the Devonian. The precise
location of the terrane in Silurian-Devonian times
is at best speculative but a position adjacent to
eastern South China, on the Gondwanaland
margin, has been suggested (Saito 1992; Hisada
1994). The precise site of original attachment to
Gondwanaland and the drift history of this terrane,
however, remain largely unknown.
Rifting and separation of Asian terranes
from Gondwanaland
It is here proposed that three episodes of rifting
occurred on the northern margin of Gondwanaland
in the Devonian, Carboniferous to Early Permian
and Late Triassic-Late Jurassic. Three continental
slivers separated from Gondwanaland in the Late
Devonian, late Early Permian and Late Triassic to
Late Jurassic times (Fig. 6).
104 I. METCALFE
Fig. 6. Schematic diagram showing the three
continental slivers/collages of terranes, rifted from
Gondwanaland and translated northwards by the opening
and closing of three successive oceans, the Palaeo-
Tethys, Meso-Tethys and Ceno-Tethys.
D e v o n i a n rifting
Various Chinese continental blocks (North and
South China, Tarim, Indochina, Qaidam) that are
postulated to have been located along or close to
the northern margin of Gondwanaland in the Early
Palaeozoic, outboard of Sibumasu and the Tibetan
blocks, were by Carboniferous-Permian times
located equatorially, did not have Gondwanaland-
related faunas and floras, and could not have
formed part of Gondwanaland. Close proximity or
actual attachment of these blocks to Gondwanaland
in the Devonian is suggested by the distribution of
Devonian vertebrate faunas (Long & Burrett
1989; Young 1990, 1993; Ritchie et al. 1992). A
Late Devonian separation of these blocks from
Gondwanaland is therefore suggested. Late
Devonian clockwise rotation of these blocks away
from northern Gondwanaland would be consistent
with the contemporaneous counter-clockwise rota-
tion of Gondwanaland about an Euler pole located
in Australia (Fig. 7) proposed by Chen et al. (1993).
Nie (1994) has also suggested that the conspicuous
Devonian unconformity of South China, and
subsequent Devonian to Triassic passive margin
sequences along the southern margin of South
China record a Devonian rifting episode and
continental separation. In addition, a widespread
Late Devonian-early Carboniferous unconformity
is recognized on most of the Asian blocks and may
represent the breakup unconformity related to
continental separation.
Carboniferous to Early P e r m i a n rifting
There is now extensive evidence of extension and
rifting along the northern Gondwanaland margin in
the Carboniferous to Early Permian. Evidence of
Early Permian rifting in Northwest Australia has
long been known (Falvey & Mutter 1981; Powell
1976; Bird 1987). Recent deep seismic studies on
the Australian NW shelf have confirmed this and
suggest that in the mid-Carboniferous to Early
Permian a major episode of extension and crustal
thinning took place on the margin of NW
Australian Gondwanaland, including the present
Australian NW shelf, resulting in the development
of the Westralian Superbasin (Colwell et al. 1994).
Crustal thinning of the Australian NW shelf at this
time was most pronounced in the lower crust and
was concentrated northwest of the NW shelf
Megashear. Separation of the Sibumasu block (as
part of the Cimmerian continent) occurred in the
late Early Permian. Similar rifting events have been
recorded from Northern Pakistan and Afghanistan
(Pogue et al. 1992; Boulin 1988), the North Indian
margin (Baud 1994), the Malagassy Rift (Wopfner
1994), Oman (Pillevuit 1993) and Iran (St0cklin
1974).
Late Triassic to Late Jurassic rifting
Both Permian and Triassic ages have been
previously proposed for the separation of the Lhasa
block from Gondwanaland. A Permian separation
has been advocated, either as a part of the
Cimmerian continent (All~gre et al. 1984; Metcalfe
1988, 1990) or as a 'Mega-Lhasa' block which
included Iran and Afghanistan (Baud et al. 1993).
Evidence for Permian rifting on the North Indian
margin and in Tibet (Baud 1994) is here regarded as
being related to the separation of the Cimmerian
continental strip which included Iran, Afghanistan
and the Qiangtang block of Tibet, but not the Lhasa
block. A Triassic rifting and separation of the Lhasa
block from Gondwanaland has been proposed by
several authors including Chang et al. (1986) and
Dewey et al. (1988). Recent sedimentological and
stratigraphical studies in the Tibetan Himalayas and
 SE ASIAN TERRANE EVOLUTION
Fig. 7. Reconstructions showing the rapid counter-clockwise rotation of Gondwanaland in the Late Devonian to
Early Carboniferous (after Chen et al. 1993).
Nepal (Liu 1992; Liu & Einsele 1994; Ogg & von
Rad 1994; von Rad et al. 1994) have documented
the Triassic rifting and Late Triassic (Norian)
separation of the Lhasa Block from northern
Gondwanaland. This Late Triassic episode of
rifting is also recognized along the NW shelf of
Australia (Colwell et al. 1994) where it continued
into the Late Jurassic resulting in the separation of
Western Burma and the Woyla terranes (Metcalfe
1990, 1994a, b; G6rtir & Seng6r 1992).
SE Asian Pre-Cenozoic sutures
The formation of present-day SE Asia involved the
progressive suturing of terranes to each other
during Late Palaeozoic to Cenozoic times and their
subsequent disruption, principally caused by the
collision of India with Eurasia. The ages of sutures
in eastern and SE Asia become younger to the south
and southeast. Reconstruction of the history of
amalgamation and accretion in SE Asia requires
constraints on the ages of the various sutures.
Constraining data include: ophiolite obduction
105
ages; melange ages; ages of 'stitching' plutons;
ages of collisional or post-collisional plutons; ages
of volcanic arcs; major changes in arc chemistry
or isotope chemistry; convergence of apparent
polar wander paths (APWPs); loops or dis-
ruptions in APWPs; ages of blanketing strata;
palaeobiogeography; biostratigraphy (e.g. ages
of oceanic sediments now located in the suture);
stratigraphy/sedimentology (e.g. provenance
stitching); structural geology (eg. age of collision-
related thrusting). It is rarely possible to acquire all
types of constraining data for individual sutures.
The various SE Asian sutures are shown in Figs 1
and 2. The region formed by nucleation around the
South China block and the Palaeozoic and
Mesozoic sutures and constraints on their ages are
discussed below from oldest to youngest.
Song M a and Song Da sutures
The Indochina and South China blocks are separat-
ed by a NW-SE-oriented mobile belt in North
Vietnam. Two sutures are currently recognized
106 I. METCALFE

within this belt, the Song Ma and Song Da sutures in Vietnam or Laos, if this is contiguous with the
of Palaeozoic and Mesozoic age, respectively. The Song Ma suture and hence diachronous in nature, is
age of the collision and amalgamation of Indochina a major task for the future.
and South China is still controversial with some
authors favouring an early amalgamation along the
Song Ma suture in the Late Devonian-Early Ailaoshan suture
Carboniferous (Gatinsky & Hutchison 1987;
The Ailaoshan suture zone (Figs 1 and 4) is a
Hutchison 1989a, b) and others favouring a
narrow NW-SE orientated belt of ophiolitic rocks
Late Triassic amalgamation along the Song Da
and oceanic sediments which forms the boundary
suture (SengOr et al. 1988). Proponents of an
between the Simao (Indochina) and South China
early suturing have proposed that the Early
blocks. The ophiolitic rocks of the zone comprise
Carboniferous event represents the collision
peridotite, gabbro-diabase and basalt. The
between Indochina and South China along the Song
peridotite includes weakly depleted plagioclase
Ma suture to form the 'East Asia Continent' or
lherzolite and spinel harzburgite with Sm/Nd ratios
'Cathaysialand'. Alternatively, those advocating a
of mid-ocean ridge type (Zhang et al. 1984; Wang
Mesozoic suturing regard this event as the collision
& Zhong 1991). The ophiolitic rocks are associated
between an arc fragment and an already accreted
with deep-marine sedimentary rocks including
arc along a SW-dipping subduction zone and
ribbon-bedded cherts that have yielded some Lower
consider the final welding of Indochina to South
Carboniferous and Lower Permian radiolarians
China as having taken place in the Late Triassic
(Wang & Zhong 1991). Upper Triassic rocks
along the Song Da suture. The large-scale folding
(Norian limestones and Rhaetian sandstones)
and thrusting and nappe formation which took
appear to blanket the suture in this region. The
place in the Early to middle Carboniferous suggest
suture trends southeast into North Vietnam but its
continent-continent collision rather than an
extension there has not been demonstrated. If the
arc-continent collision and recognition of the Song
suture is Mesozoic in age it may well correlate with
Da Permo-Triassic sequences as probably
the Song Da suture. If it is Palaeozoic in age then it
representing rift basin sequences (Tran 1979;
probably correlates with the Song Ma suture.
Hutchison 1989a, b), supports the amalgamation
Another possibility is that it is Mesozoic but
of Indochina and South China as occurring in
correlates with the Song Ma suture, closure of the
the Early Carboniferous. In addition, middle
Palaeo-Tethys being diachronous, earlier in the east
Carboniferous shallow marine carbonates are
and younger in the west. It has also been suggested
reported to blanket the Song Ma suture in North
that the Ailaoshan ophiolitic zone possibly
Vietnam and palaeobiogeographical data show that
comprises two belts, one of which correlates
pre-middle Carboniferous faunas on each side of
with the Song Ma suture of Vietnam and one that
the Song Ma zone are distinctly different whilst the
correlates with the Nan-Uttaradit suture in
middle Carboniferous faunas are essentially similar
Thailand (Zhang et al. 1984). Further constraints
(Tran 1992). Newly discovered Carboniferous
are required on the nature and age of the Ailaoshan
floras on the Indochina block in Northeast Thailand
suture in order to resolve these various possibilities.
also indicate continental connection between
Indochina and South China in the Carboniferous
(Laveine et al. 1994). It is also interesting to note
Jinshajiang suture
that Middle Devonian fishes recently recorded from
central Vietnam (Indochina block) by Janvier et al. This suture, which probably correlates with the
(1994) do not include typical South China block Ailaoshan suture to the south (Figs 1 and 4), defines
forms (including the faunas from North Vietnam). the northeastern limit of the Qamdo-Simao block
This supports an Early Carboniferous amal- and forms the boundary between this block and the
gamation of the Indochina and South China blocks. Kunlun terrane and the Songpan Ganzi accretionary
The Song Ma suture is here regarded as the complex terrane. The suture has well developed
probable zone of welding of the South China and ophiolites, and melange comprises Devonian,
Indochina blocks in the Early Carboniferous. The Carboniferous and Permian exotic blocks in a
Song Da zone appears to represent a zone of Triassic matrix. The ophiolites are regarded as
Permo-Triassic rift basins which closed in the Late Upper Permian to Lower Triassic and the age of the
Triassic and which was reactivated as a major suture regarded as Late Triassic by Dewey et al.
strike-slip shear zone in the Late Mesozoic to (1988). An older age for this suture has however
Cenozoic. Much more work is required to unravel been suggested recently by Chen& Xie (1994) and
the very complex geology and tectonic history of Bian (1991) who recognize Upper Permian to
North Vietnam/Laos. Identification of the continu- Jurassic sediments unconformably overlying
ation of the Ailaoshan suture of Yunnan (see below) Lower Permian ophiolites in the Hoh Xil Range.
 SE ASIAN TERRANE EVOLUTION
Lancangjiang suture
This suture forms the boundary between the
Qamdo-Simao and Qiangtang terranes and is here
regarded as representing a segment of the main
Palaeo-Tethys ocean. The suture zone rocks include
Devonian and Carboniferous turbiditic 'flysch',
ultramafics, ocean-floor basic extrusives of
Permian age and Carboniferous-Permian melange.
These suture zone rocks are also blanketed by
Middle Triassic continental clastics ( C h e n & Xie
1994). Carboniferous-Permian island arc rocks are
developed along the west side of the suture (Fan &
Zhang 1994) and Upper Triassic collisional
granitoids are associated with the suture (Wang &
Tan 1994). In Tibet and Qinghai, the suture runs
along the Lancangjian Thrust and is covered by
huge thicknesses of Jurassic sediments in the
Tanggula Range. Scattered ultramafic rocks occur
in the Muta region of Tibet and in Qinghai (Chen &
Xie 1994). The suture appears to be similar in
character and age to the Changning-Menglian
suture of Western Yunnan and the Raub-Bentong
suture in Peninsular Malaysia here considered to
represent the main Palaeo-Tethyan ocean.
Banggong suture
This suture forms the boundary between the
Tibetan Lhasa and Qiangtang terranes. Well
developed ophiolites have been recorded which are
associated with radiolarian cherts and the age of the
ophiolites and their obduction in Tibet is well
constrained as Upper Jurassic (Girardeau et al.
1984; Dewey et al. 1988). The suture is blanketed
in Tibet by Cretaceous and Palaeogene rocks and
structural data indicate that collision took place
along this suture around the Jurassic-Cretaceous
boundary (Dewey et al. 1988; Fan & Zhang 1994).
The extension of this suture in Yunnan is not well
preserved due to shearing but some Triassic
'flysch' and ultramafic rocks are found in the Luxi
area and blanketing strata are of Middle Jurassic
age indicating an earlier closure of the suture in this
region.
Nan- Uttaradit suture
The Nan-Uttaradit suture zone in Thailand (Figs 1
and 2) has long been regarded as representing the
Palaeo-Tethys destroyed by continent-continent
collision between Sibumasu and Indochina. The
suture has also been generally regarded as being
contiguous with the Raub-Bentong suture in
Peninsular Malaysia (see below), and more
recently, the Changning-Shuangjiang (Changning-
Menglian of this paper) suture of SW China (Barr
& MacDonald 1987). The precise age of suturing
107
is still controversial with ages of Lower
Carboniferous (Helmke 1985b), Permian (Helmke
& Kraikhong 1982; Helmke 1983, 1985a; Helmke
& Lindenberg 1983), Early Triassic (Cooper et al.
1989; Mitchell 1989; Metcalfe 1990) and Late
Triassic (Ridd 1980; Mitchell 1981; Seng6r 1984)
being proposed. This wide range of ages reflected
paucity and poor quality of constraining data. The
suture includes a belt of pre-Permian ophiolitic
mafic and ultramafic rocks (Type II ophiolitic
associations) with associated blueschists distri-
buted northeast of Uttaradit (Barr et al. 1985; Barr
& Macdonald 1987). Hada et al. (1994) have
shown that the suture in southeast Thailand (Sra
Kaeo-Chanthaburi) comprises a western chert-
clastic belt and an eastern serpentinite melange
belt. The western chert-clastic sequence appears to
form a stack of imbricate thrust slices and is dated
as Middle Triassic by radiolarians. The serpentinite
melange belt includes a wide variety of rock pack-
ages including rocks of oceanic, island arc and
continental affinities of various ages. Mafic and
ultramafic blocks in the melange comprise ocean-
island basalts, backarc basin basalts and andesites,
island-arc basalts and andesites and supra-
subduction cumulates generated in Carboniferous
to Permo-Triassic times (Panjasawatwong &
Yaowanoiyothin 1993). Limestone blocks in the
melange range from upper Lower Permian to
middle Permian and a granitic lens has yielded a
zircon U-Pb age of 486 _ 5 Ma (Hada et al. 1994).
Structural studies of Carboniferous to Permian
meta-greywacke and phyllite, which enclose the
mafic-ultramafic sequence, in the Sirikit Dam area
(Singharajwarapan & Berry 1993) indicate that
these represent an accretionary complex. These
rocks are associated with Permo-Triassic volcanic
and volcaniclastic rocks of dacitic and rhyolitic
composition and a relatively unmetamorphosed
Lower Triassic sandstone-shale turbidite sequence.
Structures indicate east-directed accretionary
thrusting and hence westward directed subduction.
The suture zone rocks are overlain unconformably
by Jurassic redbeds and post-Triassic basaltic lavas
which disconformably overly the suture are
intraplate continental basalts (Panjasawatwong &
Yaowanoiyothin 1993). A Permo-Triassic age for
the suture is therefore indicated.
Changning-Menglian suture
The Changning-Menglian suture in western
Yunnan (Fig. 4) is a narrow north-south orientated
zone of dismembered ophiolites and associated
deep-marine sedimentary rocks that are interpreted
as representing a segment of the main Palaeo-
Tethys in East Asia (Huang et al. 1984; Zhang et al.
1985; Wu & Zhang 1987; Liu etal. 1991; Wu 1993;
108 I. METCALFE

I Chan~Belt have yielded fusulinids indicative of Lower

West Central
~,'~- • ~~ .. ~ - ~."~-.
" =0 II a . c- ~ - • .~ : ~ . . o. . . ~ O' . C L ' ~. C~
U Deqin (Fig. 4). The southern extension of the
~ ?
0 Bentong suture in
f,
121
Beddedchert/
Turbidites
F~ hallow-water
carbonates
~ Redbeds
~ Volcanics
Fig. 8. Devonian to Jurassic stratigraphy of the
Changning-Menglian suture zone belt (after Liu et al.
1991)
Fang et al. 1994; Wu et al. 1995). The suture can be
traced from Menglian northwards through
Laochong, Tongchangia to Changning and
ophiolitic melange includes blocks of harzburgite,
cumulate websterite, gabbro, basalt, limestone and
chert in a mud-silt grade matrix. Associated basalts
are of mid-ocean ridge and ocean-island types (Wu
et al. 1995) and remnants of limestone capped
seamounts have been identified in the zone (Liu et
al. 1991). Oceanic ribbon-bedded chert-shale
sequences have yielded graptolites, conodonts and
radiolarians indicating ages ranging from Lower
Devonian to Middle Triassic (Qin et al. 1980; Duan
et al. 1982; Wu & Zhang 1987; Wu & Li 1989; Liu
et al. 1991; see Fig. 8). Limestone blocks and
lenses dominantly found within the basalt sequence
of the suture and interpreted as seamount caps,
East Carboniferous to Upper Permian ages (Duan et al.
'~ .- ~:~ .o .'~ ' ~ .. c:~ ,o -~ 1982; Wu et al. 1995). The suture is truncated to
. .o . . ~ . o' .c . i c~ o . ~' . o
the north of Changning by the Chongshan meta-
morphic belt and almost certainly continues north-
wards as the Lancangjiang suture to the west of
Changning-Menglian suture is problematic. The
suture can be traced to the China-Myanmar border
but its extension into Myanmar has not been
documented. It may be offset by major left lateral
strike-slip faults then to continue down into Laos
along the Dien Bien Phu line of mafic and ultra-
mafic rocks in Laos and hence to connect with the
Nan-Uttaradit suture in Thailand and the Raub-
Peninsular Malaysia.
Alternatively, it has been proposed (Huang et al.
1984) that the Changning-Menglian suture extends
southwards as the Chiang Mai volcanic belt in
northern Thailand which had been interpreted as
the remnants of a subduction-related volcanic arc
by Bunopas & Vella (1978). The volcanics of this
zone have, however, been shown to have formed in
an extensional continental setting, possibly a
backarc setting related to westwards subduction
along the Nan-Uttaradit zone to the east
Bathyal (Macdonald & Barr 1978; Barr et al. 1990).
sediments Interestingly, mafic rocks in the Chiang Rai area
Shallow-water may possibly have formed in a subduction environ-
clastics ment and these are associated with a chert-clastic
sequence known as the 'Fang Chert' which has
Stratigraphic yielded Lower Devonian graptolites, Carboniferous
break conodonts and radiolarian faunas of Lower
Devonian, Carboniferous and Lower to middle
Permian ages. Could this be another small
disrupted segment of the main Palaeo-Tethyan
suture? Extension of the Changning-Menglian
suture to the south, rather than connecting it with
the Nan-Uttaradit zone which was considered to
terminate against the Ailaoshan suture, led Wu et
al. (1995) to propose a 'Simao Terrane' of
Cathaysian affinity, bounded by the Changning-
Menglian suture to the northwest, the Ailaoshan
suture to the northeast and the Uttaradit-Nan suture
to the southeast (Fig. 9). The similarity in both
nature and age of the Nan-Uttaradit and
Changning-Menglian sutures lead me to favour
their correlation and to here regard the 'Simao
Terrane' to be disrupted part or parts of the
Qamdo-Simao block which is regarded as an
extension of the Indochina terrane.
Raub-Bentong suture
The Raub-Bentong suture of Peninsular Malaysia
forms the boundary between the Sibumasu and
Indochina/East Malaya terrane and represents the
Palaeo-Tethys. The suture zone is exposed as a
 SE ASIAN TERRANE EVOLUTION 109

extension into Sumatra as suggested by Tjia (1989)

30N
and Hutchison (1993) is here favoured. The suture
is exposed as a c. 13 km wide zone in Peninsular
,' a Malaysia and comprises melange, oceanic
.~ t ? CHINA sediments (including ribbon-bedded cherts), schist,

"/ "
•
t
* Pi OKunming
and discontinuous, narrow, elongate bodies of
serpentinized mafic-ultramafic rocks that have
' j D O
been interpreted as representing ophiolite (Tjia
1987, 1989; Hutchison 1989b). The melange has a
,' BURMA fi "~ Pe*,
t" ".-I
,, mud/silt matrix and contains a variety of clasts
including ribbon-bedded cherts, limestones,
,' te, volcanic and volcaniclastic rocks. Basic sub-
: : Chiang duction-related volcanic clasts (which occur
commonly in the Nan-Uttaradit suture further
north) have not however been reported from the
AND "LADS Raub-Bentong suture. The precise age of suturing
is still poorly constrained but a latest Permian-
... - .... .,...'.- ..' earliest Triassic suturing age is here favoured.
~ , Harbury et al. (1990) presented structural evidence
.~CAMBODIA! which seems to preclude a Late Triassic collision.
Ages of limestone clasts in the melange include
Lower and Upper Permian (Metcalfe 1989) which
indicate a latest Permian or Triassic age for the
110E melange matrix. The Main Range 'collisional' 'S'
I
Type granites of peninsular Malaysia range from
Fig. 9. Map showing the Simao terrane proposed by Wu Upper Triassic (230 __+9 Ma) to lowermost Jurassic
et al. (1995) (207_+ 14 Ma) in age, with a peak of around
210Ma (Liew & Page 1985; Darbyshire 1988).
Ages of oceanic deep-marine bedded cherts within
narrow north-south-trending zone extending from the suture zone range from Upper Devonian to
Thailand through Raub and Bentong to the east of Upper Permian (Metcalfe 1992; Spiller & Metcalfe
Malacca. Its extension to the south and southeast is 1993, 1995; Metcalfe & Spiller 1994). These data
controversial (Fig. 10) but a southwards (disrupted) support the existence o f Palaeo-Tethys between
Sibumasu and Indochina/East Malaya from Late
Devonian to Late Permian times and suggest
collision of Sibumasu with Indochina in the latest
Permian or Triassic.

Qinling-Dabie suture

This suture forms the boundary between the North

Fig. 10. Map showing the southwards extension of the
Raub-Bentong suture as proposed by various authors.
and South China blocks. Estimates of the age of this
suture, based on geological data, range from the
Middle Palaeozoic (Zhang et al. 1984; Mattauer et
al. 1985, 1991; Wang 1991) to the Permo-Triassic
(Wang et al. 1982; Klimetz 1983; Hsu et al. 1987;
Seng/Jr 1987; Yin & Nie 1993; Nie & Rowley
1994). Palaeomagnetic data (Lin & Fuller 1990)
suggest that the amalgamation of North and South
China along the Qin Ling suture had occurred by
the Late Triassic-Early Jurassic. These data
comprise convergence of APWPs, rapid rotation of
the North China block in the Late Triassic-Early
Jurassic, convergence of palaeolatitudes and a
Jurassic-Cretaceous loop (stagnation point) in the
APWPs. The data also seem to be consistent with
stratigraphic data and ages of collision-related
110 ~. METCALFE

plutons in the Qinling foldbelt (Hsu et al. 1987). Woyla suture

Enkin et al. (1992) utilizing palaeomagnetic data,
proposed that the suturing could be as late as
Jurassic, a view challenged by Nie & Rowley
(1994) who cogently argue for a Permo-Triassic
collision of South and North China. A Permo-
Triassic suturing age is here favoured with closure
of the suture commencing in the Late Permian
in the east, and progressing westwards with
development of the Tan-Lu transform fault, into the
Middle Triassic.
Shan boundary suture
This suture, also known as the Sagaing or
Mandalay suture, forms the boundary between the
West Burma and Sibumasu terranes. An Early
Cretaceous age for the suture is indicated by
Cretaceous thrusts in the backarc belt (Mitchell
1990) and the Upper Cretaceous age of the Western
Belt tin-bearing granites.
This suture comprises Cretaceous ophiolites and
accretionary complex material along the southwest
margin of Sumatra and marks the zone of accretion
and collision of the small Woyla terranes with
Sibumasu in the Late Cretaceous (Wajzer et al.
1991).
Meratus suture
This suture, located in southeast Borneo, is
characterized by subduction melange and ophiolite
of middle Cretaceous age (Hamilton 1979) which
is overlain by Eocene strata. The ophiolite was
obducted in Cenomanian times during an arc-
continent collision (Sikumbang 1986).
Boyan suture
The Boyan suture forms the boundary between the
small Semitau terrane and the South West Borneo

50N

40N

30N

20N

1ON

10s

2os

3os

4os

50S

1 -~ 21-1 3 X 4A 5+ 6I 7® 8 © 9,& 10 • 11 ~ 12'~ 13~ 14~

15@ 16 © 17 O 18 ~ ) 19 [ ] 20 [ ] 21 O 22 X 23 o.*o 24 + 25'~ 26 •

Fig. 11. Palaeolatitude plots for the Tarim, North China, South China, Indochina, East Malaya and Sibumasu terranes
from Devonian/Carboniferousto Tertiary. Sources of data are 1. Lin (1987a, b); 2. Lin et al. (1985); 3. McElhinney
et al. (1981); 4. Opdyke et al. (1986); 5. Sasajima & Maenaka (1987); 6. Chan et al. (1984); 7. Achache et al. (1983);
8. Maranate & Vella (1986); 9. Achache & Courtillot (1985); 10. Bunopas et al. (1989); 11. McElhinney et al.
(1974); 12. E. A. Schmidtke & M. Fuller (pers. comm. 1994); 13. Haile & Khoo (1980); 14. Haile et al. (1983);
15. Bunopas (1982); 16. Sasajima et al. (1978); 17. Halle (1979); 18. Lin (1987b); 19. Zhao & Coe (1987);
20. Cheng et al. (1988); 21. Bai et al. (1987); 22. Li et al. (1988); 23. McFadden et al. (1988); 24. Zhai et al. (1988);
25. Huang & Opdyke (1991); 26. Enkin et al. (1992). Modified from Metcalfe (1993, 1994c).
 SE ASIAN TERRANE EVOLUTION 111

block. The suture zone rocks are Upper Cretaceous
melange (Williams et al. 1986; Williams &
Harahap 1987) produced during short-lived SW-
directed subduction which ceased after collision of
the Semitau terrane and then recommenced behind
this accreted terrane to produce the Lubok Antu
melange and the Sarawak accretionary complex.
Terrane evolution
Available data suggest that all the east and
SE Asian allochthonous terranes formed part of,
or were close to, the northern margin of
G o n d w a n a l a n d in the Early Palaeozoic. The
continental blocks that are placed adjacent to the

PALAEO-PAClFIC

LAURENTIA
_ SIBERIA

AUSTRALIA

CAMBRO-ORDOVICIAN
,~ (TREMADOC)

ANTARCTICA \,
)
|!!i!!!!!!~!!i

INDIA
t

30S

30N

LATE D E V O N I A N

Fig. 12. Reconstructions of eastern Gondwanaland for the Cambro-Ordovician and Late Devonian, showing the
postulated positions of the East and SE Asian terranes. NC, North China; SC, South China; T, Tarim; I, Indochina;
Q, Qaidam; WC, Western Cimmerian Continent; Qi, Qiangtang; L, Lhasa; S, Sibumasu; WB, West Burma;
GI, Greater India; SA, South America. Present day outlines are for reference only.
112 I. METCALFE

Greater India-NW Australian Gondwanaland
margin in the Early Palaeozoic include North &
South China, Indochina/East Malaya (which
includes the Qamdo-Simao block of western
China), Tarim (here regarded to include the Kunlun
and Ala Shan blocks), Qaidam, Sibumasu,
Qiangtang, Lhasa, NW & SE Hainan, West Burma,
Kurosegawa and Woyla terranes. Palaeomagnetic
and biogeographic evidence indicates that the
North and South China, Tarim, Indochina/East
Malaya and Qaidam terranes were in equatorial or
low northern palaeolatitudes by Carboniferous
times (Fig. 11) and their faunas had no affinities
with those of Gondwanaland. These terranes must
therefore be placed outboard of Sibumasu,
Qiangtang and Lhasa on the Gondwanaland margin
in the Early Palaeozoic and must have separated
from Gondwanaland in the Devonian (Fig. 12). The
separation of these terranes is interpreted to have
occurred in the Late Devonian when mainland
Gondwanaland rotated rapidly counter-clockwise
about an Euler pole in Australia (Chen et al. 1993),
and the Palaeo-Tethys opened between the terranes
and northern Gondwanaland (Fig. 7).
By Early Carboniferous times, South China and
Indochina had amalgamated along the Song Ma

EARLY EARLY
CARBONIFEROUS PERMIAN
(VISI~AN)

~ ~

~0
3o

~ LATE ~ LATE
PERMIAN 7, • ~ TRIASSIC

wBL

Fig. 13. Palaeogeographicreconstructions of the Tethyanregion for Early Carboniferous,Early Permian, Late
Permian and Late Triassic. Present day outlines are for reference only. Symbols as for Fig. 12.
 SE ASIANTERRANEEVOLUTION 113

suture to form 'Cathaysialand'. However, a narrow 80

ocean between the Qamdo-Simao portion of 70 TARIM Predicted from -
the Indochina terrane and South China (now
represented by the Ailaoshan suture) persisted until 60
Triassic times (Fig. 13). Palaeomagnetic data 50
indicate that South China was equatorial during the
Early Carboniferous (Lin et al. 1985; Lin 1987a; 40
see Fig. 11). No palaeomagnetic data are available
30
for the Carboniferous of Indochina but Middle and
Late Carboniferous and younger faunas and floras 20 ~ t t ObSerVed 1 i
have strong affinities with South China. 1
10
Palaeomagnetic and palaeobiogeographic data for
Irving & Irving (1982)
the North China block suggest that it was not 0 --- Van der Voo (1990)
attached to either South China or Siberia in the Besse & Courtillot (1991)
-10
Carboniferous and that it lay in low northern ~. Enkin et al. (1992)
palaeolatitudes (Lin 1987a; Lin & Fuller 1990; Nie -20 I I I I I I I I I I I I
1991; Fig. 11). In the Mid-Late Carboniferous, R N E K2 K1 J3 J2 J1 T3 T2 T1 P2

Gondwanaland rotated clockwise, collided with
Laurentia to form Pangaea, and eastern
Gondwanaland moved from low to high southem
palaeolatitudes. The predicted palaeolatitude for
Sibumasu, if placed adjacent to NW Australia,
would be about 40°S in the Late Carboniferous and
recent palaeomagnetic results from the Yunnan part
of Sibumasu give a palaeolatitude of 43°S for
the Late Carboniferous (Figs 5 and 11) which is
consistent with the proposed reconstruction (Fig.
13). Palaeomagnetic data for the Tarim block
indicate that it moved from southern to northern
palaeolatitudes in the Devonian and Carboniferous
(Fig. 11) and that since the Permian, observed
latitudes concur with those predicted from Eurasia
indicating that it had already sutured to Eurasia
(Nie & Rowley 1994; Fig. 14). Data to constrain
the palaeoposition of the Qaidam terrane are
largely lacking for this time.
A major rifting phase occurred on the margin of
NE Gondwanaland in the Early Permian (Powell
1976; Falvey & Mutter 1981; Bird 1987; Audley-
Charles 1988; Pillevuit et al. 1993; Pogue et al.
1992; Baud 1994; Wopfner 1994) indicating that a
substantial continental fragment or fragments
separated from Gondwanaland at that time.
Audley-Charles (1988) suggested that these
terranes were Iran, North Tibet (Qiangtang)
and Indochina but later (Audley-Charles 1991)
supported the view presented by the present author
(Metcalfe 1986, 1988, 1990, 1991, 1992, 1993,
1994a, b) that the rifting terranes were in fact
Sibumasu, and Qiangtang (along with other
terranes constituting the Cimmerian continent of
SengOr). Palaeomagnetic data indicate that
Sibumasu travelled rapidly from southern to north-
ern palaeolatitudes in the Permo-Triassic (Figs 5
and 11) supporting this view. Post-Middle Permian
faunas of Sibumasu are Tethyan in aspect and show
affinities with Indochina and South China and do
not contain any Gondwanaland elements (Metcalfe
Fig. 14. Palaeolatitude versus time plots for centre of
the Tarim block. Solid black circles are calculated from
measured data on the Tarim block as reported by Enkin
et al. (1992). Other plots are the palaeolatitudes
predicted from the Eurasian reference poles of Irving &
Irving (1982), Van der Voo (1990) and Besse &
Courtillot (1991), from Nie & Rowley (1994).
1988). Recent work on Permian brachiopod
palaeobiogeography (Shi & Archbold 1993,
1994a, b; Shi et al. 1995) has shown that terranes
of Sibumasu and other Cimmerian continent
components formed part of a Gondwanaland
province in the early Early Permian, but by the late
Early Permian there was a separate Cimmerian
province indicating separation of Cimmeria from
Gondwanaland. Upper Permian faunas belong to
the equatorial Cathaysian province. The rifting and
separation of the Cimmerian continent from
Gondwanaland was effected by opening of the
Meso-Tethys and northwards destruction of Palaeo-
Tethys beneath Eurasia and the North and South
China and Indochina blocks.
Continental palaeomagnetic data show that
Cathaysialand, characterized by its palaeo-
equatorial Cathaysian flora, remained in low
northern palaeolatitudes during the Permian
(Fig. 11). Sibumasu sutured to the Qamdo/
Simao/Indochina/East Malaya portion of
Cathaysialand along the Changning-Menglian,
Nan-Uttaradit and Raub-Bentong sutures in
Permo-Triassic times. The precise age of suturing
is still controversial and may well have been
diachronous but a latest Permian-Triassic suturing
age is here favoured (see above). Thus, by Late
Triassic times, Sibumasu had amalgamated with
Cathaysialand and this amalgamated superterrane
was located in low northern latitudes (Fig. 13).
Initial contact between South and North China
probably also occurred in the late Permian with
114 i. METCALFE

collision occurring first in the east with develop-
ment of the Tan-Lu transform fault and final
suturing along the Qinling-Dabei suture occurring
by Late Triassic times (Nie & Rowley 1994). The
presence of coesite- and diamond-bearing ultra-
high pressure metamorphic rocks in the Dabei and
Sulu regions (Wang et al. 1992; Okay e t al. 1993;
Maruyama et al. 1994) which were probably
exhumed during the Triassic (Nie et al. 1995)
indicate that the Qinling-Dabei orogenic system
produced vast quantities of sediment during the
Triassic to the Songpan Ganzi basin to the west
(Zhou & Graham 1993; Nie e t al. 1995).
It is not yet known if the small Gondwanaland
terranes of Hainan Island, which yield Lower
Palaeozoic trilobites and other invertebrates
endemic to Australia and Gondwanaland and which
have Lower Permian diamictites interpreted by

UINE~.

60~.

I iili ! ~ :i~:~•
PACIFIC OCEAN

c. s :
I ~. . . . '. . . . .: i ..'~. - Fig. 15. Palaeogeographic reconstructions for Eastern
Tethys in (a) Late Jurassic, (b) Early Cretaceous and
(c) Late Cretaceous times. SG, Songpan Gangzi
accretionary complex; SWB, South West Borneo;
SE, Semitau; Si, Sikuleh; N, Natal; M, Mangkalihat;
WS, West Sulawesi; Ba, Banda Allochthon; ES, East
Sulawesi; O, Obi-Bacan; Ba-Su, Bangai-Sula;
Bu, Buton; B-S, Buru-Seram; WIJ, West Irian Jaya;
Sm, Sumba; PA, Philippine Arc. M numbers represent
-60 { : Indian Ocean magnetic anomalies. Other terrane
symbols as in Figs 12 and 13. Modified from Metcalfe
(1990) and partly after Smith et al. (1981), Audley-
Charles (1988) and Audley-Charles et al. (1988).
c) : :;:::::: L ATE C R E T A C E O U S . . .'" ...r,. Present day outlines are for reference only.
 SE ASIAN TERRANE EVOLUTION
some workers as glacial-marine, were accreted to
South China in the Palaeozoic or Mesozoic.
The Late Triassic to Late Jurassic saw renewed
rifting on the northeast margin of Gondwanaland
and initial rifting of the third continental sliver from
G o n d w a n a l a n d opening the Ceno-Tethys. The
Lhasa terrane separated in the Late Triassic,
followed by the West Burma block and Woyla
terranes in the Late Jurassic. The Lhasa block
moved rapidly northwards in the Late Jurassic to
Early Cretaceous, separated from the West Burma
block and Woyla terranes by a transform fault
(Fig. 15). The Lhasa block sutured to Eurasia in the
Early Cretaceous and by Late Cretaceous times, the
References
ACHACHE, J. & COURTILLOT, V. 1985. A preliminary
Upper Triassic palaeomagnetic pole for the Khorat
plateau (Thailand): consequences for the accretion
of Indochina against Eurasia. Earth and Planetary
Science Letters, 73, 147-157.
- - , - - . 8z BESSE,J. 1983. Palaeomagnetic constraints
on the Late Cretaceous and Cenozoic tectonics of
Southeastern Asia. Earth and Planetary Science
Letters, 63, 123-136.
AHRENOT, H., CHONGLAKMANI, C., HANSEN, B. T. &
HELMKE, D. 1993. Geochronological cross section
through northern Thailand. Journal of Southeast
Asian Earth Sciences, 8, 207-217.
--, HANSEN, B. T., LUMJUAN, A., MICKEIN, A. &
WEMMER, K. 1994. Tertiary d6collement tectonics
in Western Thailand and its implications on
stratigraphy and paleogeographic reconstructions.
In: ANGSUWATHANA, P., WONGWANICH, T.,
TANSATHIAN,W., WONGSOMSAK,S. & TULYATID,J.
(eds) Proceedings of the International Symposium
on Stratigraphic Correlation of Southeast Asia.
Dept. Min. Res. Bangkok, Thailand, 291.
AITCHISON, J. C. 1993. Allochthonous Silurian volcani-
clastic rocks of the Kurosegawa terrane: Long-
dispersed early Gondwana fragments. In Third
International Symposium of 1GCP 321 Gondwana
dispersion and Asian accretion, Abstracts of
Papers. 15-17.
--, HADA, S. & YOSmKURA, S. 1991. Kurosegawa
terrane: disrupted remnants of a low latitude
Paleozoic terrane accreted to SW Japan. Journal of
Southeast Asian Earth Sciences, 6, 83-92.
ALLI~GRE,C. J. ETAL. 1984. Structure and evolution of the
Himalaya-Tibet orogenic belt. Nature, 307, 17-22.
ARCHBOLD,N. W. & SHI, G. R. 1995. Permian brachiopod
faunas of western Australia: Gondwanan-Asia
relationships and Permian climate. Journal of
Southeast Asian Earth Sciences, 11,207-215.
, PIGRAM, C. J, RATMAN, N. & HAKIM, S. 1982.
Indonesian Permian brachiopod fauna and
Gondwana-South-East Asia relationships. Nature,
296, 556-558.
AUDLEY-CHARLES,M. G. 1983. Reconstruction of eastern
Gondwanaland. Nature, 306, 48-50.
-- 1984. Cold Gondwana, warm Tethys and the Tibetan
Lhasa block. Nature, 310, 165.
115
West Burma block and the Woyla terranes had
accreted to proto-SE Asia (Fig. 15).
Continued work on the geological evolution of East and
SE Asia is currently supported by an Australian Research
Council Grant which is gratefully acknowledged. I also
thank the Department of Geology & Geophysics,
University of New England and the Department of
Geology, Australian National University for facilities
provided. Thanks go to M. G. Audley-Charles and R. W.
Murphy for constructive reviews of the manuscript. This
paper forms a contribution to IGCP Project 321 and
GSGP Project Pangaea. It was written whilst a Visiting
Fellow in the Department of Geology, Australian National
University, Canberra.
- - 1988. Evolution of the southern margin of Tethys
(North Australian region) from Early Permian to
Late Cretaceous. In: AUDLEY-CHARLES,M. G. 8z
HALLAM, A. (eds) Gondwana and Tethys.
Geological Society Special Publication, 37, 79-100.
- - 1991. Tectonics of the New Guinea area. Annual
Reviews Earth and Planetary Sciences, 19, 17-41.
- - - , BALLANTYNE,P. D. & HALL, R. 1988. Mesozoic-
Cenozoic rift-drift history of Asian fragments from
Gondwanaland. Tectonophysics, 155, 317-330.
BAI, Y., CHEN, G., SUN, Q., SuN, Y., LI, Y., DONG, Y. &
StJN, D. 1987. Late Paleozoic polar wander path for
the Tadm platform and its tectonic significance.
Tectonophysics, 139, 145-153.
BARR, S. M. & MACDONALD,A. S. 1987. Nan River suture
zone, northern Thailand. Geology, 15, 907-910.
- - - , MACDONALD, A. S., YAOWANOIYOTHIN, W. &
PANJASAWATWONG, Y. 1985. Occurrence of
blueschist in the Nan River mafic-ultramafic belt,
northern Thailand. Warta Geologi, 11, 47-50.
--, TANTISUKRIT, C., YAOWANOIYOTHIN, W.
MACDONALD, A. S. 1990. Petrology and tectonic
implications of Upper Paleozoic volcanic rocks of
the Chiang Mai belt, northern Thailand. Journal of
Southeast Asian Earth Sciences, 4, 37-47.
BAUD,A. 1994. Late Permian sequence stratigraphy of the
N Indian margin. In: International Symposium on
Permian Stratigraphy, Environments & Resources,
Guiyang, China, Abstracts. 1.
--, MARCOUX, J., GUIRAUD, R., RICOU, L. E. &
GAETANI, M. 1993. Late Murgabian (266 to 264
Ma). In: DERCOURT,J., RIcou, L. E. & VRIELYNCK,
B. (eds) Atlas Tethys Palaeoenvironmental Maps.
Explanatory Notes. Gauthier-Villars, Pads, 9-20.
BESSE, J. & COURTmLOT,V. 1991. Revised and synthetic
apparent polar wander paths of the African,
Eurasian, North American and Indian Plates, and
true polar wander since 200Ma. Journal of
Geophysical Research, 96, 4020-4050.
BrAN QIANTAO. 1991. On the tectonic characteristics and
evolution of the Hoh Xil region. In: REN, J. & Xm,
G. (eds) Proceedings of First International
Symposium on Gondwana Dispersion and Asian
Accretion - Geological Evolution of Eastern
Tethys. China University of Geosciences, Beijing,
305-306.
116 I. METCALFE
BIRD, E 1987. The geology of the Permo-Triassic rocks of
Kekneno, west Timor. PhD Thesis, University of
London.
BOULIN, J. 1988. Hercynian and Eocimmerian events in
Afghanistan and adjoining regions. Tectonophysics,
148, 253-278.
BUNOPAS, S. 1982. Palaeogeographic history of Western
Thailand and adjacent parts of Southeast Asia - a
plate tectonics interpretation. Geological Survey
paper, 5, Department of Mineral Resources,
Thailand.
-- & VELLA, E 1978. Late Paleozoic and Mesozoic
structural evolution of Thailand. In: NUTALAYA,P.
(ed.) Proceedings 3rd Regional Conference on
Geology and Mineral Resources of Southeast Asia.
Bangkok, Thailand. Asian Institute of Technology,
133-140.
, MARANTE, S. & VELLA, P. 1989. Palaeozoic and
Early Mesozoic rotation and drifting of Shan-Thai
from Gondwana-Australia. ln:4th International
Symposium on pre-Jurassic evolution of East Asia,
IGCP Project 224, Reports and Abstracts. 1, 63-64.
BURRETT, C. & STAIT, B. 1985. South-East Asia as part of
an Ordovician Gondwanaland. Earth and Planetary
Science Letters, 75, 184-190.
, LONG, J. & STAIT, B. 1990. Early-Middle
Palaeozoic biogeography of Asian terranes derived
from Gondwana. In: MCKERROW,W. S. & SCOTESE,
C. R. (eds) Palaeozoic Palaeogeography and
Biogeography. Geological Society, London,
Memoir, 12, 163-174.
CAMERON, N. R., CLARKE, N. C. G., ALDISS, D. T.,
ASPDEN, T. A. & DJUNUDDIN, A. 1980. The
geological evolution of northern Sumatra. Ninth
Indonesian Petroleum Convention, Jakarta.
CHANG CHENGFAET AL. 1986. Preliminary conclusions of
the Royal Society and Academia Sinica 1985
Geotraverse of Tibet. Nature, 323, 501-507.
CHAN, L. S., WANG, C. Y. & Wu, X. Y. 1984.
Palaeomagnetic results from some Permian-Triass
rocks from Southwestern China. Geophysical
Research Letters, 11, 1157-1160.
CHEN BINGWEI& XIE GUANGLIAN. 1994. Evolution of the
Tethys in Yunnan and Tibet. Journal of Southeast
Asian Earth Sciences, 9, 349-354.
CHEN HAIHONG,SUN SHU, LI JILIANG,HSU, K. J., HELLER,
E, HAAG, M. & DOBSON, J. 1991. Early Triassic
palaeomagnetism and tectonics of South China. In:
Seventh Regional Conference on Geology, Mineral
and Energy Resources of Southeast Asia (GEOSEA
VII), Bangkok. Abstract Volume. Geological Society
of Thailand, 24.
CHEN, J., FOLAND, K. A., XING, E, Xu, X. 8~ ZHOU, T.
1991. Magmatism along the southeast margin of the
Yangtze block: Precambrian collision of the Yangtze
and Cathysia blocks of China. Geology, 19,
815-818.
CHEN, Z., LI, Z. X., POWELL, C. McA. & BALME,
B. E. 1993. Palaeomagnetism of the Brewer
Conglomerate in central Australia & fast movement
of Gondwanaland during the Late Devonian.
Geophysical Journal International, 115, 564-574.
CHENG, G., BAI, Y. & SUN, Y. 1988. Palaeomagnetic study
on the tectonic evolution of the Ordos block, North
China [in Chinese with English abstract].
Seismological. Geology, 10, 81-87.
COLWELL, J. B., STAGG, H. M. J., SYMONDS, P. A.,
WlLLCOX, J. B. & O'BRIEN, G. W. (compilers for
AGSO North West Shelf Study Group). 1994. Deep
reflections on the North West Shelf: changing
perceptions of basin formation. In: PURCELL,R G. &
PURCELL, R. R. (eds) The Sedimentary Basins of
Western Australia. Petroleum Exploration Society
of Australia, 63-76.
COOPER, M. A., HERBERT, R. & HmL, G. S. 1989. The
structural evolution of Triassic Intermontane Basins
in Northeast Thailand. In: THAYASUTmPITAK,T. &
OUNCrtANUM, P. (eds) Proceedings of International
Symposium on Intermontane Basins: Geology and
Resources. Chiang Mai University, Thailand,
231-242,
DARBYSHIRE, D. P. E 1988. Geochronology of Malaysian
granites. Natural Environment Research Council,
Isotope Geology Centre, Open File Report 88/3.
DEWEY, J. E, SHACKLETON, R. M., CHANG CHENGFA
SUN YIYIN. 1988. The tectonic evolution of the
Tibetan Plateau. Philosophical Transactions of the
Royal Society of London, A327, 379-413.
DUAN YANXUE, XIAO YINWEN, HONG Yu, LIo YILAI &
ZHAO ZHANGMIN. 1982. 1:200,000 scale geological
map (Menglian) and technical report [in Chinese].
ENKIN, R. J., YANG,Z., CHEN, Y. & COURTILLOT,V. 1992.
Paleomagnetic constraints on the geodynamic
history of the major blocks of China from the
Permian to the Present. Journal of Geophysical
Research, 97, 13 953-13 989.
FALVEY, D. A. & MUTTER, J. C. 1981. Regional plate
tectonics and the evolution of Australia's passive
continental margins. BMR Journal of Australian
Geology & Geophysics, 6, 1-29.
FAN, C. & ZHANG,Y. 1994. The structure and tectonics of
western Yunnan. Journal of Southeast Asian Earth
Sciences, 9, 355-361.
FANG, N. & YANG, W. 1991. A study of the oxygen
and carbon isotope records from Upper
Carboniferous to Lower Permian in Western
Yunnan, China. In: REN, J. & XIE, G. (eds)
Proceedings of First International Symposium on
Gondwana Dispersion and Asian Accretion -
Geological Evolution of Eastern Tethys. China
University of Geosciences, Beijing, 35-36.
--, VAN DER Woo, R. & LIANG, Q. Z. 1990. Ordovician
palaeomagnetism of eastern Yunnan, China.
Geophysical Research Letters, 17, 953-956.
--, LIU, B., FENG., Q. & JIA, J. 1994. Late Palaeozoic
and Triassic deep-water deposits and tectonic evolu-
tion of the Palaeotethys in the Changning-Menglian
and Lacangjiang belts, southwestern Yunnan.
Journal of Southeast Asian Earth Sciences, 9,
363-374.
GATINSKY, Y. G. & HUTCHISON, C. S. 1987. Cathaysia,
Gondwanaland and the Palaeotethys in the
evolution of continental Southeast Asia. Geological
Society of Malaysia Bulletin, 20, 179-199.
GIRARDEAU,J., MARCOUX,J. ALLt~GRE,C. J., BASSOULLET,
J. P., TANG YOUKING, XIAO XUCHANG, ZAO
YOUGONG & WANG XmlN. 1984. Tectonic
environment and geodynamic significance of the
 SE ASIAN TERRANE EVOLUTION
Neo-Cimmerian Dongqiao ophiolite, Banggong-
Nujiang suture zone, Tibet. Nature, 307, 27-31.
GORI3R, N. & SENGOR, A. M. C. 1992. Palaeogeography
and tectonic evolution of the eastern Tethysides:
Implications for the northwest Australian margin
breakup history. In: yoN RAD, U., HAQ, B. U., et al.
Proceedings of the Ocean Drilling Program,
Scientific Results, 122, 83-106.
HADA, S., BUNOPAS,S., THITISAWAN,V., PASAJAWATWONG,
Y., YANOIYOTHIN,W., ISHII, K. & YOSHIKURA, S.
1994. Accretion tectonics in Thailand along the
Nan-Uttaradit Suture Zone. AAPG Bulletin, 78, 1143.
HAILE, N. S. 1979. Palaeomagnetic evidence for rotation
and northward drift of Sumatra. Journal of the
Geological Society, London, 136, 541-545.
- -& KHOO, H. P. 1980. Palaeomagnetic measurements
on Upper Jurassic and Lower Cretaceous sedimen-
tary rocks from Peninsular Malaysia. Bulletin of the
Geological Society of Malaysia, 12, 75-78.
, BECKINSALE, R. D., CHAKRABORTY,K. R., ABDUL,
H. H. & TJAHJO, H. 1983. Palaeomagnetism,
geochronology and petrology of the dolerite dykes
and basaltic lavas from Kuantan, West Malaysia.
Bulletin of the Geological Society of Malaysia, 16,
71-85.
HAMILTON, W. 1979. Tectonics of the Indonesian region.
US Geological Survey Professional Paper, 1078.
HANSEN, B. T., AHRENDT, H., LUMJUAN,A., MICKEIN, A.
~; WEMMER, K. 1994. Do the high grade meta-
morphic rocks in Thailand represent the basement
for the Paleozoic strata? In: ANGSUWATHANA,P.,
WONGWANICH, T., TANSATHIAN,W., WONGSOMSAK,
S. & TULYATID, J. (eds) Proceedings of the
International Symposium on Stratigraphic
Correlation of Southeast Asia. Dept. Min. Res.
Bangkok, Thailand, 55.
HARBURY,N. A., JONES, M. E., A UDLEY-CHARLES,M. G.,
METCALFE,'I. & MOHAMAD,K. R. 1990. Structural
evolution of Mesozoic Peninsular Malaysia.
Journal of the Geological Society, London, 147,
1 1 - 2 6 .
HAYAMI, I. 1984. Jurassic marine bivalve faunas and
biogeography in Southeast Asia. Geology and
Palaeontology of Southeast Asia, 25, 229-237.
HELMKE, D. 1983. On the Variscan evolution of central
mainland Southeast Asia. Earth Evolution Science,
4, 309-319.
- - 1985a. The Permo-Triassic 'Paleotethys' in main-
land Southeast-Asia and adjacent parts of China.
Geologische Rundschau, 74, 215-228.
- -1985b. The orogenic evolution (Permo-Triassic) of
central Thailand. Implications on palaeogeographic
models for mainland SE-Asia. Mdmoires de la
Socidtg Gdologique de France, 147, 83-91.
- -& I@,AIKHONG, C. 1982. On the geosynclinal and
orogenic evolution of central and northeastern
Thailand. Journal of the Geological Society of
Thailand, 5, 52-74.
& LINDENBERG, H. G. 1983. New data on the
'Indosinian' orogeny from Central Thailand.
Geologische Rundschau, 72, 317-328.
HISADA, K., ARAI, S. ~; NEGORO, A. 1994. Devonian
serpentinite protrusion confirmed by detrital
chromian spinels in outer zone od SW Japan. In:
117
ANGSUWATHANA,P., WONGWANICH,T., TANSATHIAN,
W., WONGSOMSAK, S. 8¢ TULYATID, J. (eds)
Proceedings of the International Symposium on
Stratigraphic Correlation of Southeast Asia. Dept.
Min. Res. Bangkok, Thailand, 76-80.
HOFFMAN, P. E 1991. Did the breakout of Laurentia turn
G ondwanaland inside out? Science 252, 1409-1412.
Hsu, K. J., SuN, S. & LI, J. 1989. Mesozoic suturing in the
Huanan Alps and the tectonic assembly of South
China. In: SENGOR, A. M. C. (ed.) Tectonic
Evolution of the Tethyan Region. Kluwer Academic
Publishers, 551-565.
--, WANG, Q., L1, J., ZHOU, D. & SUN, S. 1987.
Tectonic evolution of Qinting Mountains, China.
Eclogae Geologie Helveticae, 71,611-635.
--., L], J., CHEN, H., WANG, Q., SUN, S.
SENGOR,A. M. C. 1990. Tectonics of South China:
Key to understanding West Pacific geology.
Tectonophysics, 183, 9-39.
--, SUN, S., LL J., CHEN, H., PEN, H. & SENGOR, A. M.
C. 1988. Mesozoic overthrust tectonics in South
China. Geology, 16, 418-421.
HUANG, J., CHEN, G. & CHEN, B. 1984. Preliminary
analysis of the Tethys-Himalayan tectonic domain
[in Chinese with English abstract]. Acta Geologica
Sinica, 58, 1-17.
HUANG, K. & OPDYKE,N. D. 1991. Paleomagnetic results
from the Upper Carboniferous of the Shan-Thai-
Malay block of western Yunnan, China.
Tectonophysics, 192, 333-344.
HUTCmSON, C. S. 1975. Ophiolite in Southeast Asia.
Geological Society of America Bulletin, 86,
797-806.
1983. Multiple Mesozoic Sn-W-Sb granitoids of
Southeast Asia. In: RODDICK, J. A. (ed.) Circum-
Pacific Plutonic Terranes. Geological Society of
America Memoir, 159, 35-60.
- - 1989a. The Palaeo-Tethyan realm and Indosinian
orogenic system of Southeast Asia. In: SENGOR,
A. M. C. (ed.) Tectonic evolution of the Tethyan
region. Kluwer Academic Publishers, 585-643.
-- 1989b. Geological Evolution of South-East Asia.
Clarendon Press, Oxford.
1993. Gondwana and Cathaysian blocks,
Palaeotethys sutures and Cenozoic tectonics in
South-east Asia. Geologische Rundschau, 82,
388-405.
INGAVAT,R. & DOUGLASS,R. 1981. Fusuline fossils from
Thailand, Part XIV. The fusulinid genus
Monodiexodina from Northwest Thailand. Geology
and Palaeontology of Southeast Asia, 22, 23-34.
IRVING, E. & IRVING, G. A. 1982. Apparent polar wander
paths for Carboniferous through Cenozoic and the
assembly of Gondwana. Geolophyscial Surveys, 5,
141-188.
JANVIER, P., THANH TONG-DZUY & TRUONGDOAN NHAT.
1994. Devonian fishes from Vietnam: New data
from central Vietnam and their paleobiogeo-
graphical significance. In: ANGSLIWATHANA, P.,
WONGWANICH, T., TANSATHIAN,W., WONGSOMSAK,
S. & TULYATID, J. (eds) Proceedings of the
International Symposium on Stratigraphic
Correlation of Southeast Asia. Dept. Min. Res.
Bangkok, Thailand, 62-68.
 118 i. METCALFE
JIN, X. 1994a. Extent and timing of the Permo-
Carboniferous glacio-marine deposits bearing units
in southwestern China. In: CHO, M. & K ~ , J. H.
(eds) IGCP 321 Gondwana Dispersion and Asian
Accretion Fourth International Symposium and
Field Excursion. Abstract Volume. 43-47.
1994b. Sedimenmtary and paleogeographic
significance of Permo-Carboniferous sequences in
Western Yunnan, China. Geologisches Institut der
Universit~it zu KGln, Sonderveroeffentlichungen,
99.
KATZ, M. B. 1993. The Kannack complex of the Vietnam
Kontum Massif of the Indochina Block - An exotic
fragment of Precambrian Gondwanaland? In:
FINDLAY, R. H., UNRUG, R., BANKS, M. R. &
VEEVERS, J. J. (eds) Gondwana 8 - Assembly,
Evolution & Dispersal. (Proceedings Eighth
Gondwana Symposium, Hobart, 1991). A. A.
Balkema, Rotterdam, 161-164.
KIMURA, T. 1984. Mesozoic floras of East and South-
east Asia, with a short note on the Cenozoic floras
of Southeast Asia and China. Geology and
Palaeontology of Southeast Asia, 25, 325-350.
KLrMETZ, M. P. 1983. Speculations on the Mesozoic plate
tectonic evolution of eastern China. Tectonics, 2,
139-166.
- 1987. The Mesozoic Tectonostratigraphic Terranes
- Devonian of the Shan-Thai terrane indicate proxim-
and Accretionary Heritage of South-Eastern
Mainland Asia. In: LEITCH, E. C. & SCHEIBNER,E.
(eds) Terrane Accretion and Orogenic Belts.
American Geophysical Union Geodynamics Series,
19, 221-234.
KOBAYASHI,T. & TAMURA,M. 1984. The Triassic Bivalvia
of Malaysia, Thailand and adjacent areas. Geology
and Palaeontology of Southeast Asia, 25, 201-228.
LAVEINE, J. P., RATANASTHIEN,B. & SITHIRACH, S. 1994.
The Carboniferous flora of Na Duang coal
mine, Northeastern Thailand, Its paleogeographic
interest. In: ANGSUWATHANA,P., WONGWANICH,Y.,
TANSATHIAN, W., WONGSOMSAK,S. & TULYATID,J.
(eds) Proceedings of the International Symposium
on Stratigraphic Correlation of Southeast Asia.
Dept. Min. Res. Bangkok, Thailand. 83-90.
LI, Y., McWILLIAMS, M., Cox, A., SHARPS, R., Ll, Y., ET
AL. 1988. Late Permian palaeomagnetic pole from
dikes of the Tarim craton, China. Geology, 16,
275-278.
LI ZH1MING. 1994. Ordovician. In: YIN HONGFU(ed.) The
Palaeobiogeography of China. Clarendon Press,
Oxford, 64-87.
LI, Z. X., ZHANG, L. & POWELL, C. McA. 1994. Were the
East Asian continental blocks part of Rodinia? 12th
Australian Geological Convention, Perth, 1994.
Geological Society of Australia, Abstracts, 37,
248-249.
Lmw, T. C. & PAGE, R. W. 1985. U-Pb Zircon dating of
granitoid plutons from the West Coast Province of
Peninsular Malaysia. Journal of the Geological
Society, London, 142, 515-526.
LIN, J. 1987a. Late Carboniferous paleogeographic recon-
struction. In: llth International Congress of
Carboniferous Stratigraphy and Geology.
Abstracts. 1, 283.
- - 1987b. The apparent polar wander path for the South
China block and its geological significance [in
Chinese with English abstract]. Scientia Geologica
Sinica, 4, 306-315.
-- & FULLER,M. 1990. Palaeomagnetism, North China
and South China collision. & the Tan-Lu fault.
Philosophical Transactions of the Royal Society of
London, A331, 589-598.
, -- & ZHANG, W. 1985. Preliminary Phanerozoic
polar wander paths for the North and South China
Blocks. Nature, 313, 444--449.
LIN TIAN-RUI & JAGO, J. B. 1993. Xystridura and other
early Middle Cambrian trilobites from Yaxian,
Hainan Province, China. Transactions of the Royal
Society of South Australia, 117, 141-152.
LIu BENPEI, FENG QINGLAI & FANG NIANQIAO. 1991.
Tectonic evolution of the Palaeo-Tethys in
Changning-Menglian Belt and adjacent regions,
western Yunnan. Journal of China University of
Geosciences, 2, 18-28.
LIU GUANGHUA. 1992. Permian to Eocene sediments and
Indian passive margin evolution in the Tibetan
Himalayas. Tubinger Geowissenschaftliche
Arbeiten, Reihe A, 13.
& EINSELE, G. 1994. Sedimentary history of
the Tethyan basin in the Tibetan Himalayas.
Geologische Rundschau, 83, 32-61.
LONG, J. & BURRETT, C. 1989. Fish from the Upper
ity to east Gondwana and south China terranes.
Geology, 17, 811-813.
MACDONALD, A. S. & BARR, S. M. 1978. Tectonic
significance of a Late Carboniferous volcanic arc in
northern Thailand. In: NUTALAYA, P. (ed.)
Proceedings 3rd Regional Conference on Geology
and Mineral Resources of Southeast Asia. Asian
Institute of Technology, Bangkok, Thailand,
151-156.
MCELHINNY, M. W., HALLE, N. S. & CRAWFORD, A. R.
1974. Palaeomagnetic evidence shows Malay
Peninsula was not part of Gondwanaland. Nature,
252, 641-645.
--, EMBLETON, B. J. J., MA, X. H. & ZHANG, Z. K.
1981. Fragmentation of Asia in the Permian. Nature,
293, 212-216.
MCFADDEN, P, L., MA, X. H., MCELHINNY, M. W. &
ZHANG, Z. K. 1988. Permo-Triassic magneto-
stratigraphy in China: northern Tarim. Earth and
Planetary Science Letters, 87, 152-160.
MARANATE, S. • VELLA, P. 1986. Paleomagnetism of
the Khorat Group, Mesozoic, Northeast Thailand.
Journal of Southeast Asian Earth Sciences, 1, 23-31.
MARUYAMA, S., LIOU, J. G. & ZHANG, R. 1994. Tectonic
evolution of the ultrahigh-pressure (UHP) and high-
pressure (HP) metamorphic belts from central
China. The lslandArc, 3, 112-121.
MATTAUER, M., MATTE, P., MALUSKI, H., Xu, Z., ZHANG,
Q. W. & WANG, Y. M. 1991. Paleozoic and triassic
plate boundary between North and South China;
new structural and radiometric data on the Dabie
Shan (eastern China). Comptes rendus de
l'Academieies Sciences de Paris, Serie 2, 312,
1227-1223.
- - , MALAVIEILLE,J., TAPPONIER,P., MALUSKI, H.,
Xu, Z., Lu, Y. & TANG, Y. 1985. Tectonics of the
Qinling Belts, build-up and evolution of eastern
Asia. Nature, 327, 496-500.
 SE ASIAN TERRANE EVOLUTION
METCALFE, I. 1984. Stratigraphy, palaeontology and
palaeogeography of the Carboniferous of Southeast
Asia. Mdmoires de la Socidtd Gdologique de
France, 147, 107-118.
1986. Late Palaeozoic palaeogeography of
Southeast Asia: some stratigraphical, palaeonto-
logical and palaeomagnetic constraints. Geological
Society of Malaysia Bulletin, 19, 153-164.
1988. Origin and assembly of Southeast Asian
continental terranes. In: AUDLEY-CHARLES,M. G. &
HALLAM, A. (eds) Gondwana and Tethys.
Geological Society, London, Special Publication,
37, 101-118.
- 1989. Triassic sedimentation in the Central Basin of
- 165-178.
Peninsular Malaysia. In: THANASUTHIPITAK, T. &
OUNCHANUM, P. (eds) Proceedings of International
Symposium on Intermontane Basins: Geology and
Resources. Chiang Mai University, Thailand,
173-186.
- -1990. Allochthonous terrane processes in Southeast
Asia. Philosophical Transactions of the Royal
Society of London, A331, 625-640.
1991. Late Palaeozoic and Mesozoic palaeo-
geography of Southeast Asia. Palaeogeography,
Palaeoclimatology, Palaeoecology, 87, 211-221.
- -1992. Ordovician to Permian evolution of Southeast
Asian terranes: NW Australian Gondwana con-
nections. In: WEBBY, B. D. & LAURIE, J. R. (eds)
Global perspectives on Ordovician geology.
(Proceedings Sixth International Symposium on the
Ordovician System.). A. A. Balkema, Rotterdam,
293-3O5.
1993. Southeast Asian terranes: Gondwanaland
origins and evolution. In: FINDLAY,R. H., UNRUG,
R., BANKS, M. R. & VEEVERS,J. J. (eds) Gondwana
8 - Assembly, Evolution & Dispersal (Proceedings
Eighth Gondwana Symposium, Hobart, 1991.)
A. A. Balkema, Rotterdam, 181-200.
1994a. Gondwanaland origin, dispersion and
accretion of East and Southeast Asian continental
terranes. Journal of South American Earth Sciences,
7, 333-347.
- 1994b. Palaeomagnetic research in Southeast Asia:
- Thakkhola (Nepal). II: Paleolatitudes and compari-
progress, problems and prospects. Exploration
Geophysics, 24, 277-282.
1994c. Late Palaeozoic and Mesozoic
Palaeogeography of Eastern Pangea and Tethys. In:
EMBRY, A. F., BEAUCHAMP, B. & GLASS, D. J.
Pangea: Global Environments and Resources.
Canadian Society of Petroleum Geologists, Memoir,
17, 97-111.
- & SPILLER,E C. P. 1994. Correlation of the Permian
- Triassic of the Yangtze Platform. Journal of
and Triassic in Peninsular Malaysia: New data
from conodont and radiolarian studies. In:
ANGSUWATHANA,P., WONGWANICH,T., TANSATHIAN,
W., WONGSOMSAK, S. • TULYATID, J. (eds)
Proceedings of the International Symposium on
Stratigraphic Correlation of Southeast Asia. Dept.
Min. Res. Bangkok, Thailand, 129.
, SHERGOLD, J. H. & LI, Z. X. 1994. IGCP 321
Gondwana dispersion and Asian accretion:
fieIdwork on Hainan Island. Episodes, 16,
443-447.
MITCHELL, A. H. G. 1981. Phanerozoic plate boundaries
in mainland Southeast Asia, the Himalayas and
119
Tibet. Journal of the Geological Society of London,
138, 109-122.
1989. The Shan Plateau and Western Burma:
Mesozoic-Cenozoic plate boundaries and corre-
lation with Tibet. In: SENGOR, A. M. C. (ed.)
Tectonic Evolution of the Tethyan Region. Kluwer
Academic Publishers, 567-583.
-1990. Mesozoic and Cenozoic regional tectonics and
-
metallogenesis in Mainland SE Asia. Geological
Society of Malaysia Bulletin, 20, 221-239.
1992. Late Permian-Mesozoic events and the
Mergui Group Nappe in Myanmar and Thailand.
Journal of Southeast Asian Earth Sciences, 7,
- 1993. Cretaceous-Cenozoic tectonic events in the
-
western Myanmar (Burma)-Assam region. Journal
of the Geological Society of London, 150,
1089-1102.
Nm, Y. S. 1991. Paleoclimatic and paleomagnetic con-
straints on the Paleozoic reconstructions of south
China, north China and Tarim. Tectonophysics, 196,
279-308.
1994. Devonian rifting of South China from
Gondwana - a case study. In: 12th Australian
Geological Convention, Perth, 1994. Geological
Society of Australia, Abstracts, 37, 319.
& ROWLEY, D. B. 1994. Comment on
'Paleomagnetic constraints on the geodynamic
history of the major blocks of China from the
Permian to the Present' by R. J. ENKINet al. Journal
of Geo14hysical Research, 99, 18 035-18 042.
--, & ZmGLER, A. M. 1990. Constraints on the
locations of Asian microcontinents in Palaeo-Tethys
during the Late Palaeozoic. In: McKE~ow, W. S. &
SCOTESE, C. R. (eds) Palaeozoic Palaeogeography
and Biogeography. Geological Society, London,
Memoir, 12, 397-408.
, Y~, A., ROWLEY,D. B. & JrN, Y. 1995. Exhumation
of the Dabie Shan ultra-high pressure rocks and
accumulation of the Songpan-Ganzi flyusch
sequence, central China. Geology, 22, 999-1002.
OGG, J. G. & VON RAD, U. 1994. The Triassic of the
son with other Eastern Tethyan margins of
Gondwana. Geologische Rundschau, 83, 107-129.
OKAY, A. I., SENGOR, A. M. C. & SATIn, M. 1993.
Tectonics of an ultrahigh pressure metamorphic
terrane: The Dabie Shan/Tongbai Shan orogen,
China. Tectonics, 12, 1320-1334.
OPDYKE, N. D., HUANG, K., XU, G., ZHANG, W. Y. &
KENT, D. V. 1986. Paleomagnetic results from the
Geophysical Research, 91, 9553-9568.
PANJASAWATWONG, Y. & YAOWANO[YOTHIN, W. 1993.
Petrochemical study of post-Triassic basalts from
the Nan Suture, northern Thailand. Journal of
Southeast Asian Earth Sciences, 8, 147-158.
PILLEVUIT, A. 1993. Les Blocs Exotiques du Sultanat
d'Oman: Evolution paldogdographique d'une
marge passive flexurale. Mdmoires de Grologie
(Lausanne), 17.
POGUE, K. R., DIPIETRO, J. A., SAID RAHIM KHAN,
HUGHES, S. S., DILLES, J. H. & LAWRENCE, R. D.
1992. Late Paleozoic rifting in Northern Pakistan.
Tectonics, 11, 871-883.
 120 i. METCALFE
POWELL, D. E. 1976. The geological evolution and
hydrocarbon potential of the continental margin off
north-west Australia. Journal of the Australian
Petroleum Exploration Association, 16, 13-23.
PULUNGGONO, A. & CAMERON, N. R. 1984. Sumatran
microplates, their characteristics and their role in the
evolution of the Central and South Sumatra basins.
In: Proceedings of the Indonesian Petroleum
Association 13th Annual Convention. 121-143.
QIN DEHOU, HE CHANGXIANG,YANGJIAWEN,LI HANSONG,
CHEN KUNHONG, ET AL. 1980. 1:200,000 scale
geological map (Baoshan) and technical report
[in Chinese].
RACEY, A., SMITH, A. B. & DAWSON, O. 1994. Permian
echinoderms from Peninsular Thailand. In:
ANGSUWATHANA,P., WONGWANICH,T., TANSATHIAN,
W., WONGSOMSAK, S. 8z TULYATID, J. (eds)
Proceedings of the International Symposium on
Stratigraphic Correlation of Southeast Asia. Dept.
Min. Res. Bangkok, Thailand, 106-114.
RAO, C. P. 1988. Paleoclimate of some Permo-Triassic
carbonates of Malaysia. Sedimentary Geology, 60,
163-171.
RAMMLMAIR, D. 1993. The evolution of the Philippine
archipelago in time and space: a plate-tectonic
model. Geologisches Jahrbuch, 81, 1-48.
RIDD, M. E 1980. Possible Palaeozoic drift of S. E. Asia
and Triassic collision with China. Journal of the
Geological Society, London, 137, 635-640.
RITCHIE, A., WANG SHITAO, YOUNG, G. C. & ZHANG
GUORUI. 1992. The Sinolepidae, a Family of
Antiarchs (Placoderm Fishes) from the Devonian of
South China and Eastern Australia. Records of the
Australian Museum, 44, 319-370.
ROGERS, J. 1989. Comment on 'Mesozoic overthrust
tectonics in South China'. Geology, 17, 671-672.
ROWLEY, D. B., ZIEGLER, A. M. & GYOU, N. 1989.
Comment on 'Mesozoic overthrust tectonics in
South China'. Geology, 17, 384-386.
SAITO, Y. 1992. Reading geologic history of Japanese
Islands. Iwanami-shoten, Tokyo [in Japanese with
English abstract].
SASAJIMA, S. & MAENAKA, K. 1987. A paleomagnetic
aspect on the assemblage of East Asian fragmented
continents. IGCP Project 224, Pre-Jurassic
geological evolution of eastern continental margin
of Asia, Report, 2, 139-150.
, OTOFUJI, Y., HIROOKA, K. SUPARKA,SUWIJANTOt~
HEHUWAT, F. 1978. Palaeomagnetic studies on
Sumatra Island and the possibility of Sumatra
being part of Gondwanaland. Rock Magnetism
and Paleogeophysics, 5, 104-110.
SASHIDA, K., IGO, H., HISADA, K., NAKORNSRI, N. &
AMPORNMAHA, A. 1993. Occurrence of Paleozoic
and Early Mesozoic radiolaria in Thailand
(preliminary report). Journal of Southeast Asian
Earth Sciences, 8, 97-108.
SENGOR, A. M. C. 1979. Mid-Mesozoic closure of Permo-
Triassic Tethys and its implications. Nature, 279,
590-593.
1984. The Cimmeride orogenic system and the
tectonics of Eurasia. Geological Society of America
Special Paper, 195.
- -1987. Tectonic subdivisions and evolution of Asia.
Bulletin of the Technical University of lstanbul, 40,
355-435.
1989. The Tethyside Orogenic System: An
introduction. In: SENGOR, A. M. C. (ed.) Tectonic
Evolution of the Tethyan Region. Kluwer Academic
Publishers, 1-22.
- - - , ALTINER,D., CIN, A., USTAOMER,T. & Hsu, K. J.
1988. Origin and assembly of the Tethyside
orogenic collage at the expense of Gondwana Land.
In: AUDLEY-CHARLES, M. G. (~ HALLAM, A.
Gondwana and Tethys. Geological Society, London,
Special Publication, 37, 119-181.
SHI, G. R. & ARCHBOLD, N. W. 1993. Distribution of
Asselian to Tastubian (Early Permian) Circum-
Pacific brachiopod faunas. Memoir of the associ-
ation of Australasian Palaeontologists, 15, 343-351.
, 1995a. A quantitative analysis on the
distribution of Baigendzhinian-Early Kungurian
(Early Permian) brachiopod faunas in the western
Pacific region. Journal of Southeast Asian Earth
Sciences, 11, 189-206.
--. , 1995b. Permian brachiopod faunal sequence
of the Shan-Thai terrane: biostratigraphy, palaeo-
biogeographical affinities and plate tectonic/palaeo-
climatic implications. Journal of Southeast Asian
Earth Sciences, 11, 177-188.
t~ WATERHOUSE, J. B. 1991. Early Permian
brachiopods from Perak, west Malaysia. Journal of
Southeast Asian Earth Sciences, 6, 25-39.
- - - , ARCHBOLD,N. W. & ZHAN, L. P. 1995. Distribution
and characteristics of mixed (transitional)
mid-Permian (Late Artinskian-Ufimian) marine
faunas in Asia and their palaeobiogeographical
implications. Palaeogeography, Palaeoclimatology,
Palaeoecology, 114, 241-271.
S1KUMBANG, N. 1986. Geology and Tectonics of Pre-
Tertiary rocks in the Metalis Mountains, South-east
Kalimantan, Indonesia. PhD thesis, University of
London.
SINGHARAJWARAPAN,S. • BERRY, R. E 1993. Structural
analysis of the accretionary complex in Sirikit Dam
area, uttaradit, Northern Thailand. Journal of
Southeast Asian Earth Sciences, 8, 233-245.
SMITH, A. G., HURLEY, A. M. & BRIDEN, J. C. 1981.
Phanerozoic palaeocontinental world maps.
Cambridge University Press.
SPILLER, E C. P. & METCALFE, I. 1993. Late Palaeozoic
radiolarians from the Bentong-Raub suture zone,
Peninsular Malaysia - initial findings. In: Third
International Symposium of IGCP 321 Gondwana
dispersion and Asian accretion. Abstracts of
Papers. 67-69.
- & --- 1995. Late Palaeozoic radiolarians from the
Bentong-Raub suture zone and Semanggol
Formation of Peninsular Malaysia: -Initial Results.
Journal of Southeast Asian Earth Sciences, 11,
217-224.
STAUFFER, P. H. 1983. Unravelling the mosaic of
Palaeozoic crustal blocks in Southeast Asia.
Geologische Rundschau, 72, 1061-1080.
- -& LEE, C. E 1989. Late Palaeozoic glacial marine
facies in Southeast Asia and its implications.
Geological Society of Malaysia Bulletin, 20,
363-397.
 SE ASIAN TERRANE EVOLUTION
- & MANTAJIT, J. 1981. Late Palaeozoic tilloids of
- WANG ZUGUAN & TAN, X. 1994. Palaeozoic structural
Malaya, Thailand and Burma. In: HAMBREY,M. J. &
HARLAND, W. H. (eds) Earth's pre-Pleistocene
glacial record. Cambridge University Press, 331-337.
STOCKLIN,J. 1974. Possible ancient continental margins in
Iran. ln: BURK, C. A. & DRAKE, C. L. (eds) The
Geology of continental Margins. Springer-Verlag,
Berlin, 873-887.
SUN YUN-CHU 1963. On the occurrence of Xystridura
fauna from Middle Cambrian of Hainan Island and
its significance. Acta Palaeontoloica Sinica, 11,
608-610.
TJ1A, H. D. 1987. The Bentong Suture. In: SITUMONANG,
B. (ed.) Proceedings Regional Conference on
Mineral Hyd. Resources of SE Asia. 73-85.
- 1989. Tectonic history of the Bentong-Bengkalis
- Gondwana and Western Yunnan. In: CHO, M. &
Suture. Geologi Indonesia, 12, 89-111.
TRAN, V. T. 1979. Geology of Vietnam (North Part).
General Geological Department, Hanoi, Vietnam.
- 1992. Tectonic zonation of Indochina Peninsula. In:
- 127-131.
29th International Geological Congress, Kyoto,
Japan. Abstracts. 1, 276.
VACHARD, D. 1990. A new biozonation of the limestones
from Terbat area, Sarawak, Malaysia. CCOP
Technical Bulletin, 20, 183-208.
VAN DER VOO, R. 1990. Phanerozoic paleomagnetic poles
from Europe and North America and comparisons
with continental reconstructions. Reviews of
Geophysics, 28, 167-206.
-- 1993. Paleomagnetism of the Atlantic, Tethys and
Iapetus oceans. Cambridge University Press.
VON RAD, U., DI]RR, S. B., OGG, J. G. & WIEDMANN, J.
1994. The Triassic of the Thakkhola (Nepal). In:
Stratigraphy and paleoenvironment of a north-east
Gondwana rifted margin. Geologische Rundschau,
83, 76-106.
WAJZER, M. R., BARBER,A. J., HIDAYAT,S. & SUHARSONO.
1991. Accretion, collision and strike-slip faulting:
The Woyla Group as a key to the tectonic evolution
of North Sumatra. Journal of Southeast Asian Earth
Sciences, 6, 447-461.
WANG, H. Z., XU, C. Y. & ZHOU, Z. G. 1982. Tectonic
development of the continental margins on both
sides of the Qinling marine realm [in Chinese with
English abstract]. Acta Geologica. Sinica, 63,
270-280.
WANG, X. M., LIOU, J. G. & MARUYAMA, S. 1992.
Coesite-bearing eclogites from the dabie
Mountains, central China: Petrogenesis, P-T paths
and implications for regional tectonics. Journal of
Geology, 100, 231-252.
WANG YI & ZHON6 DALAI. 1991. Indo-Sinian deformation
in Sanjiang area, Western Yunnan Province, China.
In: Proceedings First International Symposium on
Gondwana dispersion and Asian accretion. China
University of Geosciences, 232-237.
WANG ZONGQI. 1991. The origins and their tectonic impli-
cations of Late Proterozoic and Paleozoic conglom-
erates in the northern Qinling Mountains. In: REN
JISHUN & XIE GUANGLIAN (eds) Proceedings First
International Symposium on Gondwana Dispersion
and Asian accretion - geological evolution of
Eastern Tethys, Kunming, China. China University
of Geosciences, Beijing, 249-250.
121
evolution of Yunnan. Journal of Southeast Asian
Earth Sciences, 9, 345-348.
WATERHOUSE, J. B. 1982. An early Permian cool-water
fauna from pebbly mudstones in South Thailand.
Geological Magazine, 119, 337-354.
WILLIAMS, P. R. & HARAHAP, B. H. 1987. Preliminary
geochemical and age data from postsubduction
intrusive rocks, northwest Borneo. Australian
Journal of Earth Sciences, 34, 405--415.
--, JOHNSON, C. R., ALMOND, R. A. & SIMAMORA,W.
H. 1986. Late Cretaceous to Ealy Tertiary structural
elements of West Kalimantan. Tectonophysics, 148,
279-297.
WOPFNER, H. 1994. Late Palaeozoic climates between
KIM, J. H. (eds) IGCP 321 Gondwana Dispersion
and Asian Accretion. Fourth International
Symposium and Field Excursion. Abstract Volume.
Wu, E, VAN DER VOO, R. & QIZHONG LIANG. 1989.
Devonian palaeomagnetism of Yunnan province
accross the Shan Thai-South China suture.
Tectonics, 8, 939-952.
Wu GENYAO. 1993. Late Paleozoic tectonic framework
and Paleotethyan evolution in Western Yunnan.
Scientia Geologica Sinica, 2, 129-149.
Wu HAORUO& LI HONGSHENG. 1989. Carboniferous and
Permian radiolaria in the Menglian area, western
Yunnan [in Chinese with English abstract]. Acta
Micropalaeontologica Sinica, 6, 337-343.
-- & ZHANG QI. 1987. Carboniferous and Permian
readiolarites of Western Yunnan - relict of the
Paleo-Tethys. Compte Rendu, 3, 90-96.
- - - , BOULTER, C. A., KE BAOJIA, STOW, D. A. V. &
WANG ZHONGCHENG. 1995. The Changning-
Menglian suture zone; A segment of the major
Cathaysian-Gondwana divide in Southeast Asia.
Tectonophysics, 242, 267-280.
YANCEY, T. E. & AHF, A. 1977. Upper Mesozoic strata
near Padang, West Sumatra. Geological Society of
Malaysia Bulletin, 8, 61-74.
YANG JIALU. 1994. Cambrian. In: YIN HONGFU (ed) The
Palaeobiogeography of China. Clarendon Press,
Oxford, 35-63.
YANG SHUFENG, YU ZI-YE, Guo LtNGZHI & SHI
YANGSHEN. 1989. The division and palaeo-
magnetism of the Hainan Island and plate tectonic
significance [in Chinese with English abstract].
Journal of Nanjing University (Earth Sciences
Edition), 1, 38-46.
YANG WEI-PING. 1994. Lower Permian palynological
studies of the Tenchong Block in Western Yunnan,
China. In: International Syposium on Permian
Stratigraphy, Environments & Resources, Guiyang,
China. Abstracts. 44-45.
Ym, A. & NIL, S. Y. 1993. An indentation model for
the North and South China collision and the
development of the Tan-Lu and Honam Fault
systems, eastern Asia. Tectonics, 12, 801-
813.
YOSHIKURA, S., HADA, S. & ISOZAKI, Y. 1990.
Kurosegawa terrane. In: ICHIKAWA, K., MIZUTANI,
S., HARA, I., HAOA, S. & YAO, A. (eds) Pre-
122 i. METCALFE
Cretaceous Terranes in Japan. IGCP Project, 224
Publication, Osaka, 185-201.
YOUNG, G. C. 1990. Devonian vertebrate distribution
patterns and cladistic analysis of palaeogeographic
hypotheses. In: MCKERROW, W. S. & SCOTESE,
C. R. (eds) Palaeozoic Palaeogeography and
Biogeography. Geological Society, London,
Memoir, 12, 243-255.
1993. Vertebrate faunal provinces in the Middle
Palaeozoic. In: LONG, J. A. (ed.) Palaeozoic
vertebrate biostratigraphy and biogeography.
Belhaven Press, London, 293-323.
Yu ZI-YE. 1989. Hainan Island - A terrane from
Gondwana. In: 4th International Symposium on
pre-Jurassic evolution of East Asia. IGCP Project
224 Reports and Abstracts. 1, 11-12.
ZHAI, Y. J., ZHANG,Z. K., LI, Y. A., LI, Q., LI, Y. P., ETAL.
1988. A study of Upper Carboniferous palaeo-
magnetism for the Tarim block. Geosciences, 2,
43-56 [in Chinese with English abstract].
ZHANG, Q., LI, D. & ZHANC, K. 1985. Preliminary
study on Tongchangjia ophiolite melange from
Yun County, Yunnan Province [in Chinese with
English abstract]. Acta Petrographica Sinica, 1,
1-14.
ZHANG, Z. M., LIOU, J. G. & COLEMAN,R. G. 1984. An
outline of plate tectonics of China. Geological
Society of America Bulletin, 95, 295-312.
ZHAO, X. t~£ COE, R. S. 1987. Palaeomagnetic constraints
on the collision and rotation of North and South
China. Nature, 327, 141-142.
ZHOU, D. & GRAHAM, S. A. 1993. Songpan-Ganzi
Triassic flysch complex as a remnant ocean
basin along diachronous Qinling collisional orogen,
central China. Geological Society of America,
Abstracts with Programs, 25, 118.
 Cenozoic SE Asia: reconstructing its aggregation
and reorganization
GORDON PACKHAM
Department of Geology and Geophysics, University of Sydney, NSW, Australia 2006

Abstract: Reconstructing SE Asia in the Cenozoic is dependent on being able to model three
major tectonic events and their consequences. They are the collision of India with Eurasia, the
rotational history of the Philippine Sea plate and the ongoing collision of Australia with eastern
Indonesia. Models of the Eocene India-Eurasia collision imply extrusion along major strike-slip
faults or crustal thickening and block rotation. Kinematic modelling of the present day tectonics
of central Asia indicates rotations and strike-slip faulting and an eastward displacement of crust.
Prior to middle Miocene time tectonic activity was closely associated with the collision zone and
extrusion of Indochina was supposed to have occurred. Analysis of the volume of the topography
formed by the collision indicates eastward displacement and the loss of some crust by expulsion
or into the mantle. Palaeomagnetic measurements have found northward displacement of Tibet
and strong clockwise rotation east of the eastern syntaxis but unequivocal evidence of southward
displacement is rare. Large Palaeogene to early Neogene displacements on strike-slip faults have
been documented and related with confidence to the rotations that formed at least the Oligocene
South China Sea floor. Early Oligocene docking of the Indoburman Flysch belt south of the
eastern syntaxis is difficult to explain in extrusion models. Whether the rotations that have been
documented are a response by internal deformation or block rotation is unresolved, as is
the amount of extrusion of SE Asia to the south. In the eastern region, rotation and associated
northward motion of the Philippine Sea plate as it developed puts constraints on possible models
because of a maintained close proximity of its southern margin to the northern edge of the
Australian craton. Assembling of continental fragments now in the Banda Sea and derived from
eastern New Guinea appears to have occurred by Late Cretaceous time. Formation of the Banda
arc may have isolated the Banda Sea from the Australia plate then collision of the Australian
margin possibly rotated the Banda Sea counter-clockwise. Although westward movement of the
Philippine Sea plate relative to Australia commenced in the early Miocene, convergent left-
lateral shear was not possible until the late Miocene thus limiting westward tectonic transport of
terranes and the transmission of the shear into the region west of the Banda Sea.

This paper reviews a n u m b e r of the tectonic argu-
ments which have been used to arrive at SE Asian
reconstructions. Ideally, reconstructions should be
based on the rigorous application of kinematic
parameters. While major plate motions provide the
large scale k i n e m a t i c f r a m e w o r k for deriving
reconstructions, simple rigid plate solutions have
limitations because of the uncertainties of the
locations of plate boundaries, non-rigid deforma-
tion and the progressive reordering of elements
within the complex. Two major but contrasted
C e n o z o i c collision events have f a s h i o n e d the
region. Namely, the collision between the cratonic
masses of India and Eurasia Asia in the west and
the collision o f arcs and small continental blocks
derived from Australia together with the Australian
craton in the east and southeast. Accordingly, this
review falls into two parts, one dealing with the
India-Asia collision and the formation of the South
China Sea where the emphasis is biased strongly
towards Palaeogene tectonic events and the second
with the P h i l i p p i n e - B a n d a - N e w Guinea region
where the focus of interest is on the Neogene. As
well as reconstructions covering the whole region,
other studies have focused on specific problem
areas. The four major regional subsets are: (a) the
India-Eurasia collision and its possible displace-
m e n t of the constituent blocks of Sundaland;
(b) development of the South China Sea; (c) motion
of the Philippine Sea plate and the Philippine
Islands; and (d) the d e v e l o p m e n t of the Banda Sea
region and northern New Guinea and arc collisions
with the Australian craton.
The India-Asia collision and
western SE Asia
While there is general agreement that the collision
has indented the margin of Eurasia and that strain
has been absorbed in substantial crustal thickening,
there is controversy about the amount and distri-
bution of the remaining strain and its effect on SE
Asian tectonics. In their well k n o w n extrusion

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, 123

Geological Society Special Publication No. 106, pp. 123-152.
124 G. PACKHAM

model, Tapponnier and his co-workers argued that

during the collision, continental blocks including
Sundaland were extruded possibly some 1000 km
southeast from the path of the Indian craton (e.g.
Peltzer & Tapponnier 1988). In contrast Dewey
et al. (1989) considered that residual strain was
confined to extension of the Tibetan Plateau east of
the eastern Himalayan syntaxis, accompanied by
oroclinal bending and block rotation in SE Asia and
west of the western Himalayan syntaxis (Fig. 1).
Two of the major ingredients of kinematic tectonic
solutions have been (a) estimations of rates of
slip on active faults and convergence in zones of
shortening, and (b) reconstructions, based on
forward modelling of the collision of the Indian
craton with Eurasia. This discussion reviews the
present day tectonics of Asia and then goes on
to examine kinematic models, the sequence of
tectonic events, the contribution of palaeomagnetic
studies, the implications of South China Sea
spreading and some geological relationships along
the western margin of SE Asia.

Fig. 1. Rotation model of the India-Asia collision from

Dewey et al. (1989). Zones of sinistral transpression Late Cenozoic tectonics of Asia
(oblique lines), zone of dextral shear (vertical lines),
maximum area of extrusion of Tibet (horizontal lines), Tectonic models
subduction accretion since 45 Ma (dotted) and motion A simplified kinematic model by Avouac &
of India with respect to Eurasia in cm a-~ (arrows). Tapponnier (1993) for central Asia (Fig. 2) which
applies for the last ten or so million years, was
based on estimates of slip-rates on faults from field

80*E IO0*E
Tibet/Siberl. I ~,, 2,, ' \~'~ ;
,t@~ (,0=0.86 + 0.1~k~\ ~II:$1~I"IIA .+:::+............. +t \%'~ All m o v e m l w ~ t s in r a m s -1

KAZAKHSTAN'.
N , TILM)I/TariM+":::! ....... % + ~ ....

~.. -~ .....................~:~i~#:~!+~i~i~ii~i!i~.':~.:~:~::: ,.... ~ ® , -

-..:. - .... .:.:.~ ..... :..:.:.x.:.:~:.:..:.:r...~;~ ~ .~.~.............++:-.~.....:.:.:-"
40ON
~i~:~:~:.~;:~!~.-.::~:.,.~:.. - r l ~ ; ~ + . ; ~ : ~ : . ~ . ~ 60 = 0.65 _+ 0.30 /'/ ,.~ (/+, 40* N
.......+:+,..~.~¢~.~'.:~-~;+~ ~ ~ ~ +.,--~ .+

,l.ll,
++++++++++:. ,~,.. J .... , il. ,

-~!+:~.,..,,.~:::,:.,..,~.~.
................,+ + + + + + : , , + ~ . ~o + o. -+ ~ ~ ~-' ~\ ' + ........................... : , . S . . ' - "--, ~ _~

++N+;+-...... ++o++
::-'::::::::;:::-."~ : : ~;~:~.~:~'~:~f'~,\
........ .~i,~'.z..'+.:....

N .+~+~+ ++'++:.
-.+ .....~ , ~

+ - . - .N
..,-.--

+: ~ o,,.
7 ~ct
~-.

~.
\
,%~
.+++
~
~ .'~
"b.
~
'
~
. ~
+
\ \
""~.~ ~.:.~...-~/
~,~"
++
. .+, +.
++++ / #
/_~.

30ON i
-:++:.+.~+.
...... '...~:. ..... . . . . . . . . . . . . . . . . . . . +~.++ +:m ~,,,+.'~ "t,,s+ z ~ \ +:.+.:.:
~ .+: :-:.,~ m
.,+S....
O U TH 300N

:: ,. ~ -.:::~:+.L+:+:+~..:~::.-.:+r~::::.+ • •.-:-:.:.:.:-:-:-:-:+>:....

i '::+~L.+~;--~::+@:':-:::m::::::::'~ "'- ~:~:5:::~

:::::::::::::~;k++~:~.:+:+:+:+jt:+:~(:.~:+~+~:+:+~++¢+:+~`++~:+:~:+:~:+:+:::~:+~+$~:+' %+ s /

.,--- -+-' :;,x +

s,,+,.-,,~ ,,,,,,,s + ,Et ,8,-,- r ~:::!+~]~i;~i~ +X,,~ ~ I
--<~ ~ + z~::~++++~_++++~!+:~[l""~;:'+LN,.D.
"~+:: O _ - "~ +" 500kin
INDIA , t::~.;';;:::ilg!;::i::::#~t' U H I N A 'O//,~,,~
I~!+i+t Z o n e s ol a c t i v e s h o r t e n i n 9 , It ~'~ili::!l~i~!ii+;::~il/ , ~%~. t
# - / i I • i -:.:.:+i ,:.:+.:'~:++:.:' i - ~ ,

Fig. 2. Kinematic model for central Asia modified from Avouac & Tapponnier (1993) showing slip rates on faults
and Euler poles with rotation rates. The values east of 95°E are not constrained by inversion.
 CENOZOIC SE ASIA RECONSTRUCTIONS
observations, compiled earthquake data and the
India-Eurasia convergence rate from the De Mets
et al. (1990) NUVEL-1 model. In the model, two
stable blocks (Tibet and Tarim) are rotating relative
to stable Siberia about poles indicated on Fig. 2.
Motion between the blocks is taken up along faults
and zones of active shortening although there are
large uncertainties in slip rates. The model transfers
the estimated 40 mm a-1 of northeastward move-
ment of Tibet into clockwise rotation and extrusion
of crustal blocks along curved left-lateral faults.
After allowing for shortening at the Longmen
Thrust, the South China Block is rotating clockwise
and moving southeast at 10-15 m m a -1 with
respect to Siberia.
Tibet, rather than being a simple rigid rotating
block as in the Avouac & Tapponnier (1993) model,
is according to Molnar & Lyon-Caen (1989)
extending east-west. They calculated from earth-
quake data that through strike-slip and normal
faulting, northern Tibet is shortening north-south
at roughly 5 mm a-1 and extending east-west at
roughly 1 0 m m a -1. In south Tibet and the
Himalayas north-south graben formation results in
east-west extension of 18 _ 9 mm a-1. The east-
west extension in southern Tibet and the Himalayas
is attributable to radial thrusting on the Himalaya
front and in northern Tibet to gravitational collapse.
Using radiometric age data, Harrison et al. (1992)
suggested that the graben could have originated
between 8 Ma and 4 Ma. Molnar and Lyon-Caen
(1989) concluded that eastern Tibet has an eastward
component of movement with respect to the Tarim
Basin and India of around 30-40 mm a-1 but
declined to estimate the effect on South China
because they could not evaluate the shortening rate
in Longmen Shan.
Avouac & Tapponnier (1993) estimated that 50%
of the motion between India and Siberia is absorbed
by the extrusion of Tibet to the northeast (at
40 mm a-1 at the eastern end). Molnar and Lyon-
Caen (1989) estimated it at about a third in their
analysis. Neither study took Altai into account.
Dewey et al. (1989) approached the problem by
adding the estimated present-day shortening rates
along slip lines for Eurasia-India convergence and
concluded that the residual available for extrusion,
which they ascribe to gravitational collapse of the
thickened Tibet crust along conjugate strike-slip
faults, was not significant. Updating their estimates
for the central Himalayas to Tienshan transect, the
total India-Siberia convergence is 45 mm a-l
(NUVEL-1) less the sum of 13 mm a-1 in Tienshan,
6 mm a-1 in west Kunlun, 5 mm a-1 in central Tibet
and 18 mm a -1 in the Himalayas; the remainder
is 3 mm a-t. A similar transect from the eastern
syntaxis to Altai starting with 55 mm a-1 from
NUVEL-1 and deducting the sum of 5 mm a-1 for
125
central Tibet, 3 mm a-1 in Nan Shan, 11 mm a -1
for the Altyn Shan fault and 18 m m a -1 for the
Himalayas, leaves a residual of 18 mm a-1.
In their topographic analysis Le Pichon et al.
(1992) demonstrated that Tienshan and Altai
(northeast of Tien Shan), together form an eastward
increasing topographic mass. It is therefore likely
that the Altai convergence rate is greater than the
13 mm a-1 estimated for Tienshan. Assuming a
value of 15 mm a-1 would again give a residual of
3 mm a -1. Although the estimates, apart from
NUVEL-1, have large uncertainties this method-
ology does not provide compelling evidence for
extrusion.
D u r a t i o n o f the p r e s e n t t e c t o n i c r e g i m e
The structures presently active may have originated
at various times from the early Miocene onwards.
The onset of the present tectonic regime in
Tienshan has been broadly established as early to
middle Miocene (16 +22/-9 Ma, i.e. 7-38 Ma) by
Avouac et al. (1993). The post-lower Mesozoic
displacement on the Altyn Tagh Fault was esti-
mated as 500 km by Peltzer & Tapponnier (1988).
In contrast, Dewey et al. (1989) consider that the
long major strike-slip faults such as the Altyn Tagh,
Kun Lun and Tang Ting (Xan Shui He), all have
displacements of 150 km. Depending on the
estimated displacement, the Altyn Tagh Fault the
fault is between 5 and 17 Ma old using an
estimated slip rate of 30 mm a-1 and ignoring the
uncertainty. The Red River Fault (Fig. 6) on the
boundary of the South China and Indochina Blocks
has been a right-lateral fault since 4.7 Ma but was
left-lateral until 17-19 Ma (Leloup et al. 1993).
Although not at present active, Quaternary right-
lateral slip rate was estimated by Allen et al. (1984)
at 2-5 mm a-1. The Pliocene to Recent fight-
lateral slip on the Red River Fault in the Avouac &
Tapponnier model is about 50-75 km, or as little
as 10-25 km using the slip rate of Allen et al.,
Longmen Shan and Qilian Shan presumably
absorbing any excess strain. Between 5 and 17-
19 Ma any residual extrusion of the South China
block would have also rotated Indochina because
the Red River Fault was inactive.
Kinematic models for the India-
Asia collision
The position of the Indian Craton from the onset of
collision, has been forward modelled from plate
tectonics by amongst others, Dewey et al. (1989)
and Le Pichon et al. (1992). The size of greater
India prior to the collision is a matter of conjecture,
126 G. PACKHAM

5~.~ ~ ~"~ ,-, ~o~'~ ,o~'~

~QAIDA.~.~OC~ <

ie~6 q
N. L H A S A
~\~= BLOCK

S O U T H CHINA
BLOCK
SHILLONG
HILLS

BENGAL

SHAN
PLATEAU
+6
N SHAN

--20'N

/
--15'N
INDOCHINA
BLOCK

--IO'N

MALAY
BASIN

-5*N

Fig. 3. Simplified tectonic map of western Southeast Asia and adjacent regions of central Asia modified from
Tapponnier et al. (1986) and Dewey et al. (1988) and unpublished sources, with Jurassic and Cretaceous palaeo-
magnetic deviations from expected palaeolatitudes and declinations. Data as in Fig. 8. Results from Jurassic rocks
are circled. Heavy lines are sutures, large arrows are early to middle Cenozoic fault slip and small arrows are late
Cenozoic fault slip.
 CENOZOIC SE ASIA RECONSTRUCTIONS
but for a given size and the time of collision, the
location of the boundary of Asia can be determined
within fairly small limits.
Time of collision
Besse & Courtillot (1991) pointed out that the
convergence rate between India and Eurasia drops
sharply between anomalies A22 and A21 (50.3-
47.8 Ma on the scale of Harland et al. 1990).
An Indian Ocean spreading ridge reorganization,
shown on Fig. 4 as the change from Phase 2 to
Phase 3 followed between A20 and A18 (43.9-
41.1 Ma). Both Dewey et al. and Le Pichon date
the collision at 45 Ma (early middle Eocene).
Subduction-related volcanism north of the suture
ends at about 50 Ma and related plutons range from
94 to 41 Ma (Dewey et al. 1988). Middle Eocene
Fig. 4. Sea floor spreading phases in the eastern Indian
ocean. Phase 1 is from Late Jurassic to mid-Cretaceous,
Phase 2 is from mid Cretaceous to middle Eocene
(anomaly 20) and Phase 3 is middle Eocene to the
present. Note the location of the Exmouth Plateau (EP)
and the Cape Range Fracture Zone (CR) which may
have been the northern limit of greater India before
rifting off Australia and Antarctica.
127
limestones are the youngest preserved Himalayan
marine sediments (Searle et al. 1987). Harrison
et al. (1992), Tapponnier et al. (1986) and Peltzer
& Tapponnier (1988) favoured a collision date of
50 Ma.
Greater India
On the basis of the fit of India against Australia and
Antarctica, and the shedding of continental clastics
north onto the Exmouth Plateau, Powell et al.
(1988) suggested that continental crust of greater
India extended north of the West Australian margin
to the Cape Range Fracture Zone (Fig. 5).
Spreading patterns in the Perth Abyssal Plain and
the Bay of Bengal indicate a common breakup date
of Valanginian (Powell et al. 1988; Ramana et al.
1994). This proposed extension of greater India
1300 km, north of the reconstructed location of
the eastern syntaxis, was used by Le Pichon
et a/.(1992) in their model which is discussed
below.
A smaller greater India has been proposed
Dewey et al. (1989), Harrison et al. (1992) and
Treloar & Coward (1991). Dewey et al. argue that
apart from erosion loss after collision the remainder
of the crust is stored in the Himalayas. On this
basis, Le Pichon et al. (1992) estimate a minimum
size for greater India as 600 km north of the suture,
whereas Dewey et al. (1989) give a range from
450-1000 km. In the Harrison et al. and Treloar
& Coward models, collision commenced near
the western syntaxis leaving a 700 km gap at the
eastern syntaxis which closed in the early
Oligocene over 10 Ma after the collision onset.
Le Pichon et al. (1992) assumed simultaneous
collision and located the collision boundary 1000 -
1200 km south of the Indus-Zangbo suture (Fig. 4).
Dewey et al. (1989) and Tapponnier et al. (1986)
also assumed simultaneous collision, placing the
Asia boundary fifteen degrees further south at its
eastern end.
The fit of pre-rift India against Antarctica makes
it unlikely that it extends east of 95°E. Its south-
eastern margin runs from east of the Shillong Hills
under the eastern Bengal Basin to the foot of the
slope parallel to the northeast trending coastline.
Some recent authors (e.g. Zhao & Morgan 1985;
Peltzer & Tapponnier 1988; Le Pichon et al. 1992)
show an eastward prolongation of the craton
including the Cretaceous to Palaeogene flysch
deposits of the Indo-Burman Ranges on the Indian
plate. This is not compatible with the pre-rift
reconstruction or the geology of the Bengal Basin
where a southeast deepening continent margin is
identifiable from the Early Cretaceous onwards
(Lindsay et al. 1991). It is argued later that the
flysch was part of the Asian margin.
128
I ' i t ~ I
60*E 70° ,,...,'1 80* 90" IO0"E
II-- " - ~
--10*N
-0"
i 1 I \~
Fig. 5. Estimated locations of the Asian margin at the
time of the Indian collision from Tapponnier et al.
(1986), Dewey et al. (1989) and Le Pichon et aL (1992).
Topographic analysis
Le Pichon et al. (1992) attempted to assess whether
significant extrusion occurred by estimating the
area of the crust lost during shortening, assuming
conservation of continental crust, and comparing
it with the volume of the topography that can be
attributed to the collision after adding an estimate
of the volume of eroded sediment. The model uses
the larger greater India indicated above, assumes
a simultaneous collision along the craton edge, and
extended further north than the Avouac &
Tapponnier (1993) model to include Altai.
Assuming an average pre-collision elevation of
500 m, Le Pichon et al. found there was a uniform
loss of mass west of the east syntaxis and an excess
east of it (Fig. 6). The excess accounted for one
third to half of the loss to the west. The loss of
non-thickened crust can be accounted for by some
combination of lateral extrusion and loss of lower
crust into the mantle, possibly by eclogitization.
The loss would be reduced if the crust was thinner
i.e. was at sea-level or below it. For a minimal
greater India in which there was 700 km of oceanic
crust north of the eastern syntaxis at the time of
collision, the loss of continental crust is reduced
by about 20%. The transfer of mass to the east of
103°E still remains, but the deficit is reduced to
between one third and one eighth, or even less if
the crust was thinner.
The minimum estimate of Le Pichon et al.
(1992) is not very different from the result obtained
by the kinematic approach of Dewey et al. (1989).
They summed the strain throughout the collision
event, along slip circle traverses through the
western syntaxis, central Himalayas and the eastern
syntaxis, and concluded that the crustal volumes
G. PACKHAM
nearly balanced. No extrusion was required beyond
I
4 the thickening of the Tibetan Plateau east of the
eastem syntaxis. The crust lost from the north of
the Indian Craton was assumed by Dewey et aL to
have been thin like that of the Exmouth Plateau off
Western Australia. The essential differences
between the models are the thickness, volume and
fate of the greater India crust. Le Pichon et al.
assumed that the upper crust was stored in the
Himalayas and partly eroded off, but the lower
crust was transferred to the mantle or emplaced
north of the Indus suture. If all the crust has been
transferred under the plateau, half of its elevation
would be due to the transfer and the remainder to
local shortening. The mechanism Le Pichon et al.
I favour is injection of Indian crust into the Tibetan
lower crust as a low viscosity fluid, as proposed by
Zhao & Morgan (1985). Le Pichon et aL rejected
the thin crust hypothesis of Dewey et al. because
there is no evidence of Eocene and Oligocene deep
water sediments overlying the Indian crust fed into
the system. A hint of thin (or thinned) crust comes
from the Upper Jurassic to Lower Cretaceous
section of the Tethyan Himalaya cover north of the
Main Central Thrust in Nepal. A clastic deltaic to
shelf clastic succession deposited on northward
palaeoslope is overlain by post-breakup pelagic
slope carbonates (Gibling et al. 1994).
3000-
2000-
ItVS y n t ~
o
IOO0
/o. ' Longitude
80" ;o. 100"
Fig. 6. Estimates of equivalent linear shortening
between India and Eurasia, determined by Le Pichon
et al. (1992), from the volume of the topography
and corrected for erosion assuming conservation of
continental crust. Average block elevations are plotted
for sea-level and 500 m. The upper curve with filled
triangles represents the linear shortening from
kinematics for a maximum greater India, the lower
curve (added here) with open triangles is for a minimum
greater India. Note the eastward displacement of
shortening determined from topography.
 CENOZOIC SE ASIA RECONSTRUCTIONS
Phases of the collision and their
expression in SE Asia
Rotation or extrusion of the part of SE Asia west of
the Red River Fault has been identified as a major
driving mechanism for sea floor spreading in the
South China Sea (e.g. Tapponnier et al. (1986) and
Briais et al. (1993) between 32 Ma and 16 Ma. This
event, resulting from India-Eurasia collision,
divides the Cenozoic SE Asian tectonic history into
three major phases, namely middle Eocene to early
Oligocene, early Oligocene to middle Miocene, and
middle Miocene to Recent.
M i d d l e E o c e n e to E a r l y O l i g o c e n e
Dewey et al. (1989) proposed that between 45
and 30 Ma, convergence was taken up by the
commencement of stacking of northern Himalayan
thrust sheets and thickening of the Tibet crust to
70 km elevating it to 3 km by 30 Ma, along with
progressive indentation of the Asian margin and
bending the older sutures (Fig. 1). Because the
Palaeogene and older strata in the Tibetan plateau
are strongly deformed and Neogene sediments are
essentially flat lying, Dewey et al. (1988) dated the
major deformation of the plateau as pre-Miocene
and estimated that the shortening was about 50%.
Harrison et al. (1992) accepted the degree of
deformation of some of the Palaeogene sediments
but questioned its generalization over the plateau.
They estimated shortening at 150 km (< 20%).
Using radiometric data to determine cooling
rates in the Quxu pluton near Lhasa, Harrison et al.
(1992) suggested thickening, uplift and unroofing
of the pluton commenced in the early Miocene
(21-18 Ma). Later work by Copeland et al. (1995)
related unroofing of granite bodies successively
towards the north to uplift on the Gangdese thrust
to the south commencing at 27 Ma. This does not
conflict with the argument of Dewey et al. (1988)
that Palaeogene crustal shortening elevated the
Tibetan Plateau. A review by Mercier et al. (1987)
of the climatological evidence for the date of
plateau uplift from palaeobotanical studies reveals
conflicting interpretations. Some authors consider
that steady uplift occurred in northern Tibet from
the end of the Cretaceous, and in the south
(Gangdese Belt) from the middle Eocene, and
suggest the Himalayan uplift was later. Other
authors consider that the uplift was predominantly
Plio-Pleistocene.
In their well known plasticine model experi-
ments, Tapponnier and his co-workers (Tapponnier
et al. 1982; Peltzer & Tapponnier 1988), partitioned
extrusion and crustal thickening equally throughout
the collision and conserved all crust. The indentor
representing the Indian craton deformed a striped
129
plasticine representing Asia, producing a remark-
ably good facsimile of the observed fault patterns.
In their tectonic evolutionary scheme for Asia,
Tapponnier et al. (1982, 1986) and Peltzer &
Tapponnier (1988) identified two major phases of
extrusion. The first phase between middle Eocene
and middle Miocene, had extruded the region south
of the Red River Fault a total of 800-1000 km in
two stages. Extrusion in the second phase from
middle Miocene to Recent employs a model similar
to that of Avouac & Tapponnier (1993) for the
present tectonic regime, in which extrusion
occurred north of the Red River Fault, as discussed
above.
Tapponnier et al. (1986) suggest thickening of
the Shan Plateau and the Lhasa Block may have
taken place before the formation of the South
China Sea. But in Peltzer & Tapponnier (1988), the
strain produced by convergence was apparently
partitioned between thickening and extrusion along
the Wang Chao and Three Pagodas Faults (Fig. 3).
The end of movement on the Wang Chao fault at
about 30 Ma has been supported by radiometric
data from Maluski et al. (quoted in Huchon et al.
1994). Peltzer & Tapponnier (1988) estimated
displacements on both the Wang Chao and the
Three Pagodas Faults at 300 km by matching the
displacement of the Uttaraditt suture and upper
Cretaceous magma belts respectively. The mag-
matic belts are not as clearly defined on geological
maps as portrayed by Peltzer & Tapponnier (1988)
and the displacements on the Three Pagodas Fault
could well be substantially less than their estimates
and may even be right-lateral if the displaced
extension of the Uttaraditt suture on the Gulf of
Thailand is aligned with the projected Bentong-
Raub suture (Fig. 3).
Harrison et al. (1992) explained the delay in
movement on the Red River Fault until early
Oligocene (35 Ma) by assuming a smaller greater
India than Le Pichon et al. (1992) and that collision
commenced at the western syntaxis at about 50 Ma
leaving a gap of 700 km of oceanic crust at the
eastern syntaxis. Movement on the Red River Fault
started after the early Oligocene closure of the gap
and collision in eastern Tibet. In this scheme, the
duration of displacement on the Wang Chao Fault
would have been short.
E a r l y O l i g o c e n e to E a r l y M i o c e n e
In the second phase, according to Dewey et aL
(1989), after 30 Ma when the Tibetan Plateau had
achieved considerably buoyancy, India began to
indent Asia, deflecting the Asian sutures and
commencing block rotation (Fig. 1). Wrenching
began along conjugate faults bounding the elevated
area and deformation spread north with diminishing
130 G. PACKHAM

intensity, into Tienshan and south into the

Himalayas elevating them along the Main Central
Thrust and thrusts in the Lesser Himalayas. Uplift
of the Gangdese belt in the early Miocene resulted
from underthrusting of the northern Himalayas.
In their model Peltzer & Tapponnier (1988)
proposed that movement on the Red River Fault
generated the South China Sea as the southern
part of Sundaland moved to the southeast. They
estimated the displacement on the fault as 500 km,
by matching the extrapolated Jinsha and Uttaraditt
Sutures (Fig. 3). The magnitude and direction of
displacement on the Red River Fault is supported
by Lacassin et al. (1993) who estimated the left-
lateral strain in mylonitic gneisses as 330 _+60 km
and radiometric dating by Leloup et al. (1993) has
found that in the Diancang Shan region in Yunnan,
left-lateral ductile shear terminated between 17 and
19 Ma, preceded by 7 km of unroofing at 23 and
19 Ma, due to large scale oblique shear. In the
Harrison et al. (1992) variant on the Tapponnier
model the main phase of movement on Red River
Fault was from 35 to 20 Ma. Half to two thirds of
the convergence was taken up by extrusion. The Fig, 7. Interpretation of South China Sea structure
end of the phase is based on evidence of unroofing modified from Tapponnier et al. (1986) showing:
of the Quxu Pluton near Lhasa, suggesting that oceanic crust (cross hatched) and shallow areas of
significant uplift of the Tibetan Plateau occurred continental crust (dotted). LP, Luconia Province;
at 18 to 21 Ma rather than after 17 Ma as suggested MAB, Macclesfield Bank; NCB, Nam Conson Basin;
by Tapponnier et al. (1986). The unroofing is PA, Palawan Block; PS, Paracel Shoals; RB, Reed Bank;
apparently related to uplift associated with move- SCAR, Scarborough Seamounts; WNB, West Natuna
ment on the Gangdese Thrust that commenced at Basin; YB, Yinggehai Basin.
27 Ma (Copeland et al. 1995).
Briais et al. (1993) made a kinematic study of the
spreading history of the South China Sea and the northeast Andaman Sea. At 30 Ma it would have
displacement on the Red River Fault. Spreading been located east of the eastern syntaxis but well
commenced at 32 Ma with a ridge jump at 26 Ma
south of the Indus suture. The closeness of the pole
and a change in ridge trend from east-west to to the Indochina Block suggests that rotation was
northeast-southwest at 24 Ma with spreading
more important than translation.
ending at about 15.5 Ma soon after the cessation of
Although the commencement of South China
movement on the Red River Fault at 17-19 Ma as Sea floor spreading at 32 Ma fits well with the
discussed above (Fig. 7). Using poles of opening initiation of movement on the Red River Fault,
they determined for the South China Sea and pre-drift rift basins containing Eocene sediments
assuming that the Indochina Block was attached to
occur on both margins of the South China Sea
the drifting margin Briais et al. (1993) determined
(Salvidar-Sali et aL 1982; P. Chen et al. 1993). In
the displacement on the Red River Fault as 555 km
the case of the Pearl River basin on the south China
for a point near the Uttaraditt suture; a good corre-
margin and Reed Bank on the drifted margin, there
lation with the estimate of Peltzer & Tapponnier. In
are both Eocene and Palaeocene sediments (Taylor
the reconstructions of Briais et al. convergence
& Hayes 1980; Yu 1990). That is, South China Sea
increased across the fault inland and there was margin extension commenced earlier than the
extension in the lower Red River region, corre- India-Asia collision and the supposed time of
lating well with the study of Leloup et aL (1993).
initiation of the Red River Fault.
The implication is that the Indochina Block and the
drifting South China Sea margin formed a single
small plate and that spreading was driven by
Early Miocene to R e c e n t
motion of the Indochina Block. This point will be
returned to later. A stage pole determined by The uncertainty in dating the initial elevation of the
Huchon et al. (1994) describing the early stage Tibet Plateau was discussed above. Rather than
of opening was located at 13°N and 97°E in the occurring by crustal shortening from late Eocene to
 CENOZOIC SE ASIA RECONSTRUCTIONS
mid-Oligocene time as proposed by Dewey et al.
(1989), Tapponnier et al. (1986) and Harrison et al.
(1992), consider that the bulk of the uplift took
place after movement on the Red River Fault. This
requires transfer of lower crust from the India plate
across the Indus suture because the Neogene
sediments of the plateau are relatively undeformed.
As described previously, the present tectonic
regime commenced in middle to late Miocene time.
During it some of the convergence was transferred
to eastward movement of Tibet relative to the
Himalayas and in part to South China as indicated
by the early Pliocene commencement of right-
lateral movement on the Red River Fault. Elevation
of the Tibet Plateau to its present height was
approximately coincident. Harrison et al. (1992)
suggested it occurred between 8 Ma and 5 Ma and
Dewey et al. at about 5 Ma. England & Houseman
(1988) explained this last uplift as the result of
delamination of the mantle lithosphere beneath the
plateau.
Palaeomagnetic data: central Asia
and mainland SE Asia
The objective of palaeomagnetic studies in central
and SE Asia has been to determine the amount
of northward movement of sites north of the
Himalayas and whether blocks in SE Asia have
been rotated and moved south due to extrusion.
The palaeomagnetic data have been recently
summarized by Y. Chen et al. (1993) and Huang
& Opdyke (1993) using different methodologies.
Rotations and change in palaeolatitude have been
determined by comparing measurements with the
Eurasian polar wander path compiled by Besse
& Courtillot (1991). The uncertainties taken into
account are statistical errors arising from the
measurements within and between sites, structural
corrections, estimates of rock age, date of
magnetiztion and uncertainty in the polar wander
path location. Fortunately, the Eurasia polar wander
path assembled by Besse & Courtillot (1991)
indicates very little movement in the pole location
in the Cretaceous between 130 and 80 Ma. Studies
have been biased towards Cretaceous sediments
but unfortunately many of them are in non-marine
successions and hence dating is imprecise.
Huang & Opdyke (1993) compared the results
from separate sites with the pole for that
magnetization age. The Y. Chen et al. (1993)
compilation compared the averages of individual
studies on Cretaceous rocks for each block with an
average Cretaceous pole. Although the results from
the two methods are in general comparable Y. Chen
et al. lost fine detail. The dataset for Middle
Jurassic to Upper Cretaceous sites from Huang &
131
Opdyke (1993) are replotted according to longitude
in Fig. 8 and individual sites in the eastern part of
the region have been plotted geographically on
Fig. 3. With one exception, the Pamir measure-
ments west of the western syntaxis show counter-
clockwise rotation and possible but minor south-
ward displacement. Tamir poles indicate significant
northward motion with greater displacement to the
west. The Tarim block data have large errors,
Huang & Opdyke summarize it as 1380 _+790 km
but suggest several hundred kilometres without
significant rotation is more likely. The Tibet data
(Lhasa Block) suggest about 1500 km of northward
displacement with counter-clockwise rotation
increasing to the east. East of the east syntaxis, the
Markam sites (eastern Qiangtang Block) are
rotated clockwise and unlike Tibet may be dis-
placed southward but the error bars overlap zero
displacement.
Sites on the northern Shan-Thai Block, south-
west of the Red River Fault and west of the
Uttaraditt suture, indicate clockwise rotation apart
from two locations close to the Red River Fault
which are rotated slightly counter clockwise (Figs
3 and 8). Five Shan-Thai sites indicate possible
northward displacement. The three Jurassic sites
(22, 23 and 25) apparently indicate a strong north-
ward displacement. These may not be significant
because the wander path of Besse & Courtillot
(1991) has a strong shift in the palaeopole corre-
sponding to a palaeolatitude change at these sites
from 49°N to 20°N between 170 Ma and 130 Ma
(Oxfordian to Valanginian) after being relatively
stationary from 210 Ma (Late Triassic). The fourth
site (R on Fig. 8) with possible northward shift is
from the Jurassic~Cretaceous Kalaw Red beds in
eastern Myanmar on the Shan-Thai Block (Richter
et al. 1993). For ages of magnetization between
130 and 80 Ma this sampling locality yields a
palaeolatitude displacement between 2.6 and 6.0
degrees northward within a _+6 ° error bar. If there
is a northward displacement it is only small. A
possible small northward displacement has also
been measured at Site 24. Possible southward
displacements were found at Cretaceous sites 17,
18 and 19 are (Fig. 8) but the errors make them
inconclusive.
Yang & Besse (1993) report a palaeolatitude
of 23.1 +_ 1.8°N for the medium temperature
magnetic component in sediments from the Khok
Kruat Formation (Albian-Aptian) in the Khorat
Basin on the Indochina Block. After including the
uncertainty of the palaeopole location Huang &
Opdyke (1993) quote the palaeolatitude change
as -8.0 _+5.3 ° northward with a clockwise rotation
of 14.4 _+6.1 °. No fold test was possible at the site
because of low dips but older formations in the
same sequence do give a positive fold test. The
132 G. PACKHAM

Indochina
2O
Moved North

°i
80 21,

60 dochina

Pami

01 /--I /'~ ~ 14~,Jl/ Markam (I,j-/~r 28 I II ~/

-20 8

-40 r t ' ' ' ' I . . . . Ill' 1' ' ' ! ' ' ' , I . . . . I . . . . i
70 80 90~ 100 110 120 130

Fig. 8. Plots of deviations of palaeomagnetic measurements from the expected palaeolatitude and declination from
the equivalent aged Eurasian pole from Besse & Courtillot (1991) plotted against latitude. Data and point numbers
are from Huang & Opdyke (1993) except for the point 'R' which is from Richter et al. (1993). The inner bar is the
uncertainty of determination from the individual dataset and the outer bar includes the uncertainty of pole
determination.

Cretaceous results south of the Red River Fault they are deflected around 70 ° clockwise from the
from the western side of the Shan-Thai Block Lhasa Block measurements (Figs 3 and 8).
suggest a small northward m o v e m e n t and a corre- At face value, these results could indicate
spondingly small southward m o v e m e n t on the east. rotation of the Shan-Thai Block about a pole
Declinations are generally rotated clockwise and located within it. The Khorat Plateau rotated further
 CENOZOIC SE ASIA RECONSTRUCTIONS
south could be part of the same domain. Variability
in disturbance of declinations suggest that the
Shan-Thai Block may not have behaved as a single
entity but was internally deformed by displacement
along the many faults through the region. Although
the rotation can only be dated as later than Early
Cretaceous, the event may be of middle to late
Tertiary age, since upper Cenozoic sites in central
and western Thailand are also disturbed and are
mostly but not universally rotated clockwise
(McCabe et al. 1993; Richter et al. 1993). Sites of
similar age from the Indochina Block in eastern
Thailand (Khorat) and Vietnam are not signifi-
cantly rotated.
The South China results indicate possible south-
ward displacement in the eastern part, but in the
west site 26 in the Sichuan Basin may have been
displaced to the north (6 _+6°). The magnetization
age is indicated by Huang & Opdyke as uncertain.
Sites 27, 28 and 33, between the Xian Shui He and
the Red River Faults, are in the central Hunnan
Basin are close to their expected palaeolatitude.
Site 33 in the central Yunnan Basin near the Red
River Fault is, like the Markam sites, rotated
significantly clockwise.
In summary, there is significant northward
translation of Tibet, counter-clockwise rotation in
Pamir, counter-clockwise rotation of eastern Tibet,
clockwise rotation of sites east of the eastern syn-
taxis but disturbed close to the syntaxis and near the
Red River Fault. One good Indochina Cretaceous
dataset indicates probable southward motion, the
remaining Cretaceous results spread from west of
the Dien Bien Phu Fault in the Shan-Thai Block
are ambiguous but could indicate clockwise block
rotation with little change of latitude.
The driving mechanism for
South China Sea spreading
In the discussion of early Oligocene to early
Miocene events, it was pointed out that South
China Sea spreading had been used to estimate
displacement on the Red River Fault on the
assumption that the spreading was driven by
extrusion of Indochina on the Red River Fault.
During the first stage of opening from 32 to 26 Ma,
spreading ridges were short and the Indochina
block had a long contact with the drifting side of
the basin, extending almost to Reed Bank (Fig. 7).
After a ridge jump between anomaly 7 and 6b at
26-24 Ma, spreading began to propagate rapidly
to the south. The ridge trend changed from 5-10°E
of N to about 60°E of N. The southwestward
extension stabilized at about anomaly 6 at 20 Ma.
As the spreading pattern changed, the drifting
block, consisting of an elongated strip of thinned
continental crust with oceanic crust outboard from
133
it, was moving at a higher velocity obliquely away
from the Indochina block, rapidly decreasing the
length of the contact between them. Thus after
26 Ma, spreading was not driven by motion of the
Indochina Block. Further, spreading ended about
15.5 Ma, later than the cessation of movement
on the Red River Fault (17-19 Ma according to
Leloup et al. 1993).
To invert the argument, using South China Sea
spreading to determine fault movement would
substantially overestimate the fault displacement
after 26 Ma. Of the 560 km of Red River Fault
displacement calculated by Briais et al. (1993),
only the 240 km which occurred between the
commencement of spreading and 26 Ma can be
attributed to driving South China Sea spreading.
While fault displacement may have continued,
spreading in the South China Sea does not provide
a means of determining it. The amount of strike-slip
movement may have been small in the last phase
of movement on the fault from 23-19 Ma when
oblique shear and unroofing occurred. The fault
could have been active before 32 Ma driving some
of the rifting which preceded South China Sea
spreading, although as noted earier this started in
Palaeocene time before the commencement of the
collision of India.
Extrusion of Indochina on the Red River Fault
fits well with the Oligocene phase of spreading in
the South China Sea, but is difficult to reconcile
with published basin histories in the region. In
the Nam Con Son Basin (Fig. 7) on the Indo-
china Block syn-depositional extension occurred
(Matthews & Todd 1993) while in the West Natuna
Basin to the south there was post extension
subsidence (Ginger et al. 1993).
The western margin of Sundaland
The geology of the western margin of Sundaland
provides some constraints on models of India-Asia
collision which have been largely ignored. Some
aspects are examined here.
The Java-Sumatra margin
In the Sumatra-Java margin, there is a Late
Cretaceous to middle Eocene hiatus in volcanic
activity. Possible explanations are that, prior to the
middle Eocene a north facing arc lay to the south
and collided with polarity reversal in the middle
Eocene or that there was little or no subduction
on the Sumatra-Java margin during the hiatus. The
latter is supported by the work of Wajzer et al.
(1991), who have interpreted the Upper Mesozoic
Woyla Group in the Natal area of western Sumatra
as a west-facing accretionary prism which has later
been disrupted by strike-slip faulting.
134
Packham (1990) suggested that between latest
Cretaceous and middle Eocene time, sea floor
spreading in the Indian Ocean (Phase 2 in Fig. 4),
extended east only as far as its present preserved
limit off southern Sumatra, 1300 km east of the
Ninetyeast Ridge. East of it lay the older sea floor
of the Australia plate, which at that time was
moving slowly away from Antarctica (Royer &
Sandwell 1989). With middle Eocene reorganiz-
ation of sea floor spreading, between anomalies
20 and 18 (44-41 Ma), subduction under the Sunda
margin commenced or increased from a low rate.
Radiometric ages of an andesite and a quartz
diorite from northern central Myanmar of 50 and
53 Ma respectively (Mitchell 1993) suggest that
subduction-related volcanism occurred there during
the Sumatra volcanic hiatus and that the India-
Australia plate boundary intersected the Asia
margin between Sumatra and Myanmar. The
Myanmar dates also fall in the time range of the
Linzizong Volcanics of the Lhasa Block (Dewey
et al. 1988) and presumably represent the eastern
end of the Paleogene volcanic arc.
Using the India-Asia reconstruction parameters
of Le Pichon et al. (1992), the extrapolated middle
Eocene position of the India-Australia plate
boundary would intersect the central Sumatra coast
500 km west of its present position (Fig. 9). It is in
a similar position in the 45 Ma reconstruction of
Peltzer & Tapponnier (1988). For the Sumatra
margin to be essentially passive until then requires
the India-Australia plate boundary to have been
west of it, i.e. Sumatra must have been about
500 km further east relative to South China at the
time collision commenced. The distance may be
reduced by about 100 km, subtracting movement
of western Sumatra on the Burma plate since the
inception of the Sumatra Fault.
M y a n m a r a n d the I n d i a - A s i a collision
The tectonic position of the Indoburman flysch
belt, the time of its docking and the distribution of
post-docking strain all have important implications
for reconstructions of western SE Asia. As indi-
cated in the discussion of greater India the eastern
boundary of the Indian craton runs from south of
the Shillong Hills to the Bay of Bengal (Fig. 3). The
Indoburman flysch belt has been accreted onto the
craton margin after collision. Palaeogeographically,
the flysch is an Asian margin accretionary wedge
and the proximal Central Basin of Myanmar is a
shelf deposit. The flysch is a thick sequence of
Senonian to middle Eocene feldspathic turbidites
(Bender 1983), which for the most part predate the
collision of India with Asia. Their distribution
down the Arakan coast to the west of the oceanic
G. PACKHAM
35 \ ,=
EURASIA 4 5 × _ _ X - -..~( ''~
INOIA " ~
PLATE
oo
......... ,5
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
20os PLATE
I
6O"E 80" 100*E
Fig. 9. Reconstruction of India and western SE Asia for
45 Ma from Peltzer & Tapponnier (1988). To it have
been added the proposed India-Australia plate boundary
prior to Phase 2 of Indian Ocean Spreading (Anomaly
20-44 Ma), and an X in Assam at the eastern syntaxis
and another on the Eurasia plate at the corresponding
point at the northern end of the Indoburman flysch belt
on the Eurasia plate from the same set of reconstructions
for 45, 35 and 15 Ma. Docking of the flysch belt at
Assam is thought to have occurred in the Oligocene
between 30 and 36 Ma.
crust of the Bengal Basin precludes their derivation
from the Indian margin. West of the flysch the
Central Basin of Myanmar contains a shelf to non-
marine Albian-Aptian to Pleistocene sedimentary
succession. In the Naga Hills near the eastern
syntaxis (Fig. 3), the Upper Cretaceous-Eocene
flysch is thrust over Palaeocene to upper Eocene
platform cover (Mitchell 1993) and both are over-
lain by the post-docking Oligocene to ?lower
Miocene, Barail Formation (Roy 1986). The Barail
Formation progrades south from fluvio-deltaic to
marine. The stratigraphic relations indicate an
early Oligocene docking of the flysch terrane on
the India plate, probably within 10 Ma of the
India-Asia collision, and movement north on the
craton since that time.
On a broader scale, the structural units of
western Myanmar appear to be an eastward
continuation of the Asian margin wrapped around
the eastern syntaxis in agreement with palaeo-
magnetic declination rotations (Fig. 3). The uplift
between the Central Basin and the flysch belt
corresponds to the Indus suture as indicated by
Dewey et al. (1988). At its northeastern extremity
in Myanmar, the suture appears to end against
a branch of the Sagiang Fault. It is apparently
 CENOZOIC SE ASIA RECONSTRUCTIONS
displaced 450 km to the south to a belt of ultra-
marie rocks 100 km north of Mandalay, thence it
trends north-northeast, curving north-northwest
around the northern end of the Central Basin where
it is separated from the Indus suture by the Lohit
Thrust. The presence of Triassic flysch in Myanmar
on the western margin of the Central Basin links it
with the southern Lhasa Block-Indus suture as does
the Upper Cretaceous flysch. The basement rocks
of Myanmar from the Shan Scarps westward form
part of the Cretaceous to Eocene magmatic belt
which Mitchell (1993) compares with the Gangdese
plutons.
From docking at the beginning of Oligocene
time, say 36 Ma, until the commencement of
Andaman Sea spreading at c. 13 Ma (Curray 1989),
the northward transport of the Indoburman flysch
belt on the Indian craton relative to fixed Asia
would be 1500 km, using the kinematics of Le
Pichon et al. (1992). If the Shan-Thai Block was
extruded the relative displacement would be that
much greater. Identification of such a dislocation is
very speculative. Mitchell (1993) suggested 1100
km of right-lateral displacement of the Cretaceous
outer magmatic belt exposed in highs within the
Central Myanmar basin, relative to the Shan-Thai
Block prior to Andaman Sea spreading. Assuming
coupling to the India plate this would take 17 Ma.
The fault would have to be located between the
Central Basin highs and the Sagiang Fault. This
would leave the remaining 400 km to be taken up
west of the Cretaceous magmatic belt in the
Indoburman Ranges, the Andaman arc and pre-rift
extension prior to Andaman spreading. A later time
of docking would reduce the residual by 64 km
per Ma.
The position of the northeast comer of the flysch
belt has been plotted on the 45 Ma reconstruction
of Peltzer & Tapponnier (1988) on Fig. 9 together
with its 35 and 15 Ma locations from their other
reconstructions. At 35 Ma, about the time of
docking according to the geological evidence cited
above, their reconstruction shows it about 2000 km
to the north. At 15 Ma it is still around 900 km
north of the corresponding docking point in the
Naga Hills. At 30 Ma when movement ended on
the Wang Chao Fault, relative to stationary Eurasia
the eastern syntaxis was at 1 I°N and the northern
edge of greater India as far north as 22 ° corre-
sponding to the present day location of the western
end of the fault. Andaman Sea spreading com-
menced at about 13 Ma (Curray 1989). Most of the
460 km of extension has been taken up onshore
along the westem margin of the Burma plate by
right-lateral slip on the Sagiang Fault on the
western margin of the Shan Plateau (Fig. 3). The
northward movement of the India plate relative
to stable Asia over the last 13 Ma was 790 km,
135
leaving a residual of 330 km between the India and
Burma plates which cannot be accounted for easily.
Summarizing the foregoing, the Indoburman
flysch belt was a forearc on the Asian margin at
the time of its collision. After docking early in the
Oligocene, the Indoburman flysch belt and possibly
central Myanmar, were wrapped around the syn-
taxis and transported north, possibly 2300 km on
the Indian Craton along the western SE Asian
margin. Attenuation of the crust in the Andaman
Sea region in early Miocene time, was followed by
sea floor spreading in the Andaman Sea from the
middle Miocene onwards accompanied by dis-
placement on the Sagiang Fault. Some of the north-
ward strain could have been translated into rotation
of the Shan-Thai and Indochina Blocks and west-
ward transport of Sumatra and possibly adjacent
parts of SE Asia. Movement on the Wang Chao and
Red River Faults was taking place as the flysch belt
and central Myanmar were being transported north.
Extrusion models would require the Shan-Thai and
Indochina blocks and the flysch belt to be extruded
from north or west of the eastern syntaxis in Eocene
time, prior to collision at the eastern syntaxis.
Alternatively, they could have been squeezed out
behind the SE Asian margin after the northern end
of the Myanmar flysch belt had docked in southern
Assam early in Oligocene time.
Western SE Asia: conclusions
Analysis of the present seismicity documents an
eastward motion of Tibet and sets the stage for the
interpretation of the earlier history. The model
proposed by Avouac & Tapponnier (1993) gives a
residual southeastward movement of South China
of 10-15 mm a-1. At this rate during the last 4.7 Ma
of fight-lateral displacement on the Red River
Fault, a displacement of only 50-75 km is
indicated. At a lower rate based on field studies by
Allen et al. (1984) it would be 10-24 km. These
latter are compatible with the Dewey et al. (1989)
suggestion that most of the eastward movement has
been absorbed in the Longmen Shan and that the
strike-slip displacements are a fraction of those
suggested by Tapponnier et al. (1986). Any pre-
4.7 Ma extrusion of South China would have
moved Indochina as well. Transition to the present
tectonic framework is likely to have occurred in
late early to middle Miocene time after left-lateral
movement on the Red River Fault ceased between
17 and 19 Ma (Leloup et al. 1993).
Forward modelling of the India-Eurasia
collision from its onset at about 45 Ma produces
different results depending on the estimated volume
of crust fed into the Himalayan system. Even
conservative models result in some extrusion of
crust into eastern Tibet. Dewey et al. (1989)
136 G. PACKHAM
assumed conservation of continental crust and a
greater India with thin continental crust north of the
present craton, and concluded that extrusion is
limited to the eastward extension of the Tibetan
Plateau east of the eastern syntaxis. With a larger
and thicker greater India, Le Pichon et al. (1992)
found that there was a deficit of topography
between the syntaxes and an excess to the east,
supporting the notion of eastward transport, but
between a half and one third of lost crust remains
unaccounted for. This can be explained by
partitioning the loss between extrusion and transfer
of lower crust to the mantle by eclogitization. If a
smaller greater India is assumed, the deficit reduces
to one eighth to one third and even less if the crust
was thinner, a result similar to that of Dewey et al.
Ahead of the advancing Indian Craton, Dewey
et al. (1988, 1989) suggested crustal thickening
elevated the Tibetan Plateau to 3 km between 45
and 30 Ma. Palaeogene deformation was suggested
because Neogene sediments on the Tibet Plateau
are relatively undeformed. India then began to
indent Asia, deflecting the old sutures and rotating
adjacent blocks as the Tibet crust became more
buoyant. After 30 Ma deformation spread north of
Tibet and south into the Himalayas. Elevation of
the southern margin of the plateau discussed by
Harrison et al. (1992) is related to thrusting on the
Ghagdese Thrust which commenced at 27 Ma
(Copeland et al. 1995) and is independent of dating
the elevation of the whole plateau. If plateau eleva-
tion did not occur until around 20-25 Ma, then
it is not possible to explain how over 20 Ma of
shortening was distributed without invoking
extrusion.
In their models Harrison et al. (1993) and Peltzer
& Tapponnier (1988) partitioned strain from con-
vergence between crustal thickening and extrusion.
They proposed that Indochina was extruded along
the Red River Fault before the uplift of plateau.
The latter authors proposed an Eocene to early
Oligocene extrusion of the region south of the
Wang Chao Fault. Left-lateral displacements of
the order of hundreds of kilometres have been
documented for the Wang Chao and Red River
Faults and radiometric data support an early end of
movement on the Wang Chao Fault. Harrison et al.
(1992) suggested that uplift of the plateau
commenced between 18-21 Ma. at the time of
oblique shear on the Red River Fault. Neogene
plateau elevation must be explained by some form
of crustal underplating by Indian Craton crust
across the Indus suture because Neogene sediments
of the plateau are mildly deformed. Elevation was
accompanied by movement on the Main Central
Thrust in the Himalayas and the commencement of
deformation to the north of Tibet, passing into the
present tectonic regime.
Relating spreading in the South China Sea to
determining displacement on the Red River Fault
by Briais et al. (1993), although elegant, may be
applicable only to the pre-Miocene fault displace-
ment since after that time, the southeastern drifting
side of the South China Sea was moving away from
Indochina and cannot have been driven by its
motion. Only about half of their estimated dis-
placement of 520 km can be demonstrated this way,
although some additional slip may have occurred in
the pre-spreading rift phase on the China margin.
The rift phase enigmatically had a Palaeocene
commencement, prior to the India-Asia collision.
Palaeomagnetic measurements have been quoted
in support of extrusion. A review of the data here
reveals that there is good evidence for counter-
clockwise rotation in the Pamir and northward
translation of the Lhasa Block with some counter-
clockwise rotation. Stronger clockwise rotation
predominates in the Indochina and Shan-Thai
Blocks but regional changes of latitude consistent
with extrusion have not been established.
Southward motion of the Khorat Plateau is the only
result in the east which falls outside the error bars.
The data are compatible with clockwise rotation of
the Shah-Thai block around a pole within the
block. The dispersion of declination deviations
could well be a response to complex deformation
such as thrusting on the western margin of the
Khorat Basin, and possible displacements on the
many faults which cross the region rather than
the simple extrusion of stable blocks.
Evidence for the location of some entities on
the western margin of Sundaland conflicts with
extrusion reconstructions by Peltzer & Tapponnier
(1988). A Late Cretaceous to middle Eocene hiatus
in volcanic activity in Sumatra has been interpreted
here to indicate that its middle Eocene position was
around 500 km further east adjacent to the slow
moving Australia plate (Fig. 9). For the
Indoburman flysch belt and the Central Basin of
Myanmar to dock at eastern Assam in early
Oligocene time, they also had to be substantially
east and south of the position reconstructed by
Peitzer & Tapponnier, unless they were extruded
with other parts of Sundaland prior to collision at
the eastern syntaxis. If the docking predated
Indochina extrusion, the extruded block was
extruded from northeast of the Central Basin
inboard from it. After docking, the flysch belt and
most of central Myanmar have been transported
possibly 2300 km north on the margin of the India
plate and rotated clockwise. There are corre-
sponding counter-clockwise rotations in the Lhasa
Block. A major dislocation may exist beneath the
western part of the Central Basin. Continued move-
ment resulted in the middle Miocene formation of
the Burma plate which is separated from the
 CENOZOIC SE ASIA RECONSTRUCTIONS 137

remainder of SE Asia by Andaman Sea spreading,

the Sumatra and Sagiang Fault. Some of the earlier
northward strain may have contributed to the
rotation of the Shan-Thai and Indochina Blocks
and suggested westward movement of Sumatra.
The India-Asia Collision has been responsible
for clockwise rotation of parts or all the Shan-Thai
and Indochina Blocks, left-lateral slip on the Red
River and probably for generation of at least the
Oligocene part of the South China Sea floor. After
accretion of the flysch belt onto the Indian craton it
was carried north while displacement was taking
place on the Wang Chao and Red River Faults. In
the absence of definitive evidence of southward
movement it is probable that, if there was extrusion,
it was on a smaller scale than advocated by
Tapponnier and his co-workers. While the present 10 Ma
day tectonics of Asia are indicative of the large- Rangin et al. (1990) .......
t ,l ._.._I--I- t
scale processes resulting from the India-Eurasia (a)
collision, determination of the time of elevation of
the Tibet Plateau is crucial to resolving the kine-
matics of contending general models. If, in the first
20 Ma after the collision, most of the shortening
was not taken up by thickening the crust of the
plateau then extrusion must have occurred.

The Philippine Sea plate and Australia-

SE Asia collisions
A number of regional reconstructions have treated
the eastern region in some detail. Some of the
diversity in interpretation is illustrated in the
Miocene to basal Pliocene reconstructions from
Rangin e t al. (1990), Daly e t al. (1991) and (b)
Packham (1990) in Fig. 10. The earlier regional
study of the western Pacific by Jolivet e t al. (1989) - ~ ~ ~ ~ ~'~ ~

has been largely incorporated in Rangin et al.

(1990). The kinematics of the model of Rangin
e t al. are illustrated in Fig. 1l. Reconstructions i:~?iii ~ k,,%~*iiiiiiiiiii~i!N~
by Hall (1996) have made a major advance in
the understanding of the regional tectonics. In the
following discussion, the history of rotation of the
Philippine Sea plate and the Philippine arc and
the problem of the Celebes Sea is reviewed to i;i!ii'
~/.~ o ',

provide background for the consideration of the

origin and dispersal of the microcontinental blocks
of Northern New Guinea and the Banda Sea, the
impact of Philippine Sea plate rotation on Northern
New Guinea and the Banda Sea, and the trans-
mission of Pacific plate motion to Northern New
Guinea and the Banda Sea.
,

The Philippine Sea plate and the
Philippine archipelago
Since 5 Ma the Philippine Sea plate has been
rotating clockwise relative to Eurasia about a pole
(c)
Fig. 10. Reconstructions of southeastern SE Asia:
(a) for 10 Ma from Rangin et al. (1990); (b) for late
Miocene time from Daly et aI. (1991); and (c) for 5 Ma
from Packham (1990).
138 G. PACKHAM
Fig. 11. Kinematic diagram showing trajectories of the various blocks of Southeast Asia digitized and plotted in
the reconstructions of Rangin et al. (1990). Plates are Australia (Aus), Caroline (Ca), Eurasia (Eu) and Philippine
Sea (Ph).
located north of Japan at 48°N and 157°E at
5.45°/Ma (Seno et al. 1993). The Philippine
archipelago straddles the convergent boundary
between the Eurasia and Philippine Sea plates
(Fig. 12). The left-lateral Philippine Fault which is
no older than late early Pliocene (Aurelio et al.
1991), takes up the motion between the essentially
anchored western flank of the islands and the
oceanic plate to the east. The youthful Philippine
Trench is a product of changing arc polarity.
Palaeomagnetic data from islands east of the
Philippines on the Philippine Sea plate and Ocean
Drilling cores have a coherent pattern indicating
a large scale clockwise rotation of the whole
Philippine Sea plate (Hall et al. 1995b). Because
the Philippine Trench is young, it is generally
accepted that much of the archipelago was
developed on the Philippine Sea plate rather the
Eurasian margin. In his reconstructions Hall (1996)
has separated Luzon and some of the islands to its
south from the southern part of the Philippines and
derived them from an extension of northern
Borneo, assembling the island arc chain in early
Miocene time by transferring Luzon to the
Philippine Sea plate. This interpretation is based on
the palaeomagnetic results of Fuller et al. (1991)
who found a large counter-clockwise rotation in
Luzon throughout most of the Tertiary with little
change of latitude.
Hall et al. (1995a) estimated that the Philippine
Sea plate has rotated c. 90 ° clockwise since 50 Ma.
The pre-Pliocene rotation is in two stages. A
rotation of 50 ° between 40 and 50 Ma about a pole
located at 10°N and 150°E, followed by an interval
with no significant rotation until 25 Ma when
rotation recommenced, amounting to 34 ° about a
pole at 15°N and 160°E (Fig. 13). These findings
put severe constraints on the reconstruction of north
New Guinea and the Banda Sea. The Miocene
rotation may have been initiated by the collision
of the northern edge of the Australian craton with
the Philippine Sea plate. The general location of
the Philippines has been plotted on the Fig. 13
on the assumption that they were on the Philippine
Sea plate.
The eastern part of the Philippine islands is
substantially comprised of rocks built up by island
arc processes from the Cretaceous onwards, on
oceanic basement dating back to Late Jurassic (e.g.
Geary et al. 1988). Continental basement of at least
Jurassic age is present in the west around Mindoro
and western Panay. It is an extension of the North
Palawan Block on the South China Sea drifted
margin. Pubellier et al. (1991) have identified a
terrane in western Mindanao, including Zamboanga
and the Daguma Range (Fig. 12), which may also
be continental. The continental blocks, part of the
Asia margin, were incorporated in the Philippines
 CENOZOIC SE ASIA RECONSTRUCTIONS 139

as the Philippine Sea plate rotated clockwise. The

Fig. 12. Major tectonic elements and sea floor
anomalies in the Philippine-Banda Sea region modified
from Hall et al. (1995b). Terranes of continental or
suspected continental origin are west of the line in the
Philippine Islands which includes Mindoro, west Panay
and west Mindanao.
essentially transform boundary brought the south
to southwest facing Philippine arc northward to
collide obliquely with the Asian margin. A middle
to late Miocene collision between the Palawan
Block on the southern flank of the South China Sea
has been interpreted differently by Holloway
(1982), Sarewitz & Karig (1986), Rangin (1991)
and Bird et al. (1993). Transpression and strike-slip
continued from latest Miocene time to the present
(Bird et al. 1993). The eastern and western terranes
of Mindanao were juxtaposed before early Pliocene
time by strike-slip movement (Pubellier et al.
1991).
Middle to upper Miocene volcanic rocks in
western Mindanao have been interpreted by
Pubellier et aL (1991) as an east-facing arc on the
western side of the Molucca Sea. The arc may
have continued south through the Sangihe arc to
similar age volcanic rocks in eastern SW Sulawesi
(Coffield et al. 1993). Rangin et al. (1990) and
Packham (1990) both show this as subduction of
the Philippine Sea plate under the Sangihe arc. Daly
et al. (1991) indicate subduction of Celebes Sea
crust from the west (Fig. 11). An alternative and
perhaps a better model, as discussed later, is that
before the Banda arc was initiated, the Molucca Sea
was part of a northward prolongation of the
Australia plate as illustrated by Daly et aI. (1991)
in their early Miocene reconstruction.

~ 1~,~
5M
'~ a ~ ~"
PII • ~'
• oo. ,0,
L ~ RotationPole
~ 40-50 Mo
EURASIA - - _

C.~s /IgWm~";~..~:: ~.I~

,k7 ~ S E A ~ BASIN m
? ~ y/ p l . ] ~ ~ ~ PLATE j
~'~,,~ ~ ~,. SOLOMON
- A U S T R A~L I A E 3 ~ ~ ~

1~ ~ . ~ . : _ ~ . ' ~ -,~-- ~ \1

Fig. 13. Schematic reconstructions of relationships in the Northern Australia region based on Philippine Sea plate
reconstructions from Hall et al. (1995b) for 40, 25 and 15 Ma. Locations are: MLA, the Melanesian arc; PA,a possible
Palaeogene Australia-Pacific arc between the Philippine Sea and Australia plates; MRA, the Maramuni arc; BA, the
Banda arc; PI, the location of the Philippine Islands; and CB, the Banda Sea Continental Blocks.
140 G. PACKHAM
The Celebes Sea
The Celebes Sea and the West Philippine Basin
have overlapping ages (Fig. 12). Both contain mid-
dle Eocene oceanic crust with similar spreading
rates. Weissel (1980) mapped magnetic anomalies
20 to 18 (46-52 Ma) which parallel the Sulu ridge
and young to the south, into the North Sulawesi
Trench. Anomaly trends differ because of the
clockwise rotation of the Philippine Sea plate and
possible counter-clockwise rotation of the Celebes
Sea (Weissel 1980; Shibuya et al. 1989). The
Celebes Sea could be part of a larger domain
including southern Borneo and parts of the Malay
Peninsular where Fuller et al. (1991) have also
documented counter-clockwise rotations. The
presence of contradictory results from Borneo
and the dating of the rotation make interpretation
controversial. Hall (1996) argues for the rotation of
Borneo on geological and palaeomagnetic grounds.
Silver & Rangin (1991) concluded that the
Celebes Sea was either a marginal basin formed
behind an arc or is a trapped fragment of a larger
entity, including the Molucca Sea now separated
from it by the narrow Sangihe arc. Similar sea floor
ages are suggested by the lower middle Eocene
cherts and limestones associated with ophiolites
embedded in what is possibly a lower Miocene
forearc terrane described by Moore et al. (1980)
from Talaud Island in the Molucca Sea.
Assuming the arc is not allochthonous the
Celebes Sea has a backarc relationship to
Palaeogene volcanic rocks in the north arm of
Sulawesi (Ratman 1976) and the eastern part of the
southwest arm (Sukamto 1982) where they have
been broadly dated as lower Eocene to upper
Oligocene. Hall (1996) has linked the Celebes and
Molucca Seas to the West Philippine Basin to form
a single basin behind a North Sulawesi-Philippine
arc. Marginal basin formation was probably
enhanced at anomaly 20 time when Australia
increased its velocity away from Antarctica from
c. 1 0 m m a-1 prior to anomaly 20, to about
30 mm a-I between anomalies 18 and 20 (Royer &
Sandwell 1989).
Continental blocks of the Banda Sea and
North New Guinea
One obvious feature of the terrane distribution in
northern New Guinea and the Banda Sea is that
oceanic and island arc terranes are to the north of
the continental terranes. Intervening between them
are the West New Guinea Composite Block and the
East Papua Composite Terrane. The Finisterre
Terrane continues offshore into the uncollided
Melanesian arc. East-west tectonic transport has
resulted in duplication of the arc so that the Manus
segment overlaps the New Britain segment (Fig.
14). Three categories of continental blocks are
found north of the Australian craton.
(1) In the southern Banda arc from Timor to
Tanimbar, parts of the Australian craton have
been incorporated into the upper plate during
collision.
(2) North of the Australia plate boundary, in the
Banda Sea region are rifted blocks including
the basement of Sumba, Buton, Banggai-Sula,
Buru, Seram, the Bird's Head, Obi and south-
ern Bacan and submarine ridges south of Buru
and Seram. This group could also include the
metamorphic belt in the north, central and
southeast arms of Sulawesi, and possibly some
of the higher tectonic units in Timor.
(3) Composite terranes comprising a mixture of
continental, arc and oceanic terranes have been
accreted onto New Guinea to form the West
New Guinea Composite Block and the East
New Guinea Composite Terrane.
Apart from Sumba and central Sulawesi which are
problematical the remaining blocks have an
Australian derivation, but there is disagreement
about their source, time of detachment from
Australia and the means of their tectonic transport.
S u m b a a n d the central S u l a w e s i
m e t a m o r p h i c belt
Sumba has a basement of Upper Cretaceous
turbidites overlain unconformably by gently
dipping Palaeogene shallow water sediments and
volcanic rocks (Burollet & Salle 1981), and
resembles the stratigraphy of the adjacent Asian
margin in southwest Sulawesi and offshore east
Java as described by Hasan (1991) and Bransden
& Matthews (1992). Sumba is separated from them
by the Flores Basin and was probably incorporated
in the Banda forearc in later Palaeogene time.
Van der Werff et al. (1994) and Rangin et al. (1990)
also consider that Sumba was rifted off SE Asia
and is a separate block from Timor.
Another problematical unit, not usually regarded
as one of the drifted blocks, is the metamorphic belt
of central Sulawesi which contains blueschists
and is overthrust on the east by the East Sulawesi
ophiolite belt. A lower Cretaceous metamorphic
date shows that these former continental margin
deposits are older than the Upper Cretaceous
turbidites of southwestern Sulawesi which overlie
an ophiolite basement. Two phases of Eocene-
Oligocene metamorphism have been identified in
the blueschists (Parkinson 1991). The metamorphic
belt extends down the west side of the southeast
arm of the island where Helmers et al. (1990) have
 I' I 1 500
•4°N• PHILIPPINE SEA krn I
CAROLINE PLATE L,
~.~ $ulawe~i Trerl~ _ PLATE
I-~(---./l~a'rnahera %%
~i
Ma
!o/ PACIFIC
1,5Ma / PLATE
Torricelli /25 Ma~
Terrane
~~,,ir ,,..~ OnactiV°'
~ i : : : : . ~ Finisterre E : ~ Manus
~::iii!ii~ ~ : : . : . . - . ~. Terrane ........

Tarera SOUTH
• BISMARCK
BandaSea iii:... PLATE

~J ~ West '
SE ASIA Tanimb Papuan
PLATE ~ ~, ~ ~J~ ,~_
Basin
ar•
<~umba " AUSTRALIA PLATE ~ - .".etr~i~iiii;i:!:!:!:.'%..':i:::'.%.'.'.

CoralSea ~:~:i:i:?:i:i:i:?::::.:......-..
~"s ~:i~:i!ii!i!::~i~:.
120OE 126= 138" ° 150" 156°E

Fig. 14. Plate tectonic framework of the New Guinea region modified from Struckmeyer et al. (1993) showing relative longitude of the former locations of the western edge of the
Philippine Sea plate as reconstructed by Hall et al. (1995b).

4~
142 G. PACKHAM
obtained an upper Oligocene metamorphic date.
Palaeogene volcanic rocks on the eastern side of the
southwest ann of the island reported by Ratman
(1976) and faulting events recorded in the late
Eocene and mid-Oligocene in the platform
carbonate Tonasa Limestone to the west of the
volcanic arc are compatible with a convergent
boundary and a mid-Oligocene collision. Helmers
et al. (1990) distinguish the event from the later
collision of the Banggai-Sula Block in northern
east Sulawesi which Davies (1990) dates as latest
Miocene. It can also be distinguished from the
mid-Miocene collision of Buton-Tukang Besi with
southeast Sulawesi (Davidson 1991). Hall (1996)
envisages transfer of the metamorphic belt and the
ophiolite from the India plate and to the Eurasia
plate at 25 Ma. The eastern and central Sulawesi
terrane has general similarity to the West New
Guinea Composite Block, which contains amongst
other things non-metamorphosed, continental slope
sediments, low to medium grade metamorphic
rocks of continental affinity, blueschists and
ophiolites (Struckmeyer et al. 1993).
The reconstruction of Daly et al. (1991) shows
eastern and central Sulawesi forming in the
oceanic domain, south of the Philippines and
colliding with western Sulawesi in the late Miocene
(Fig. 10). Although explaining the late Miocene
tectonic episode in western Sulawesi, it is incom-
patible with the continental margin origin and
tectonic history of the metamorphic belt.
Microcontinental blocks o f the Banda Sea
and the Bird's H e a d
Regionally, the northern Australian craton base-
ment can be divided into two types. In Papua New
Guinea there are deformed Palaeozoic sediments
and Palaeozoic to Triassic acid intrusive rocks.
West, in Irian Jaya and on the northern Australian
margin, undeformed Palaeozoic sediments overlie
Precambrian basement.
The microcontinental blocks of the northern
Banda Sea and the Bird's Head characteristically
have metamorphic pre-Mesozoic basements and
are of the eastern type. Above basement is a
Triassic clastic or carbonate shallow marine succes-
sion, overlain by deeper water Jurassic to
Palaeogene sediments, passing into Cretaceous to
lower Palaeogene bathyal to pelagic deposits. The
preserved upper Mesozoic to Neogene successions
are remarkably continuous, indicating little
tectonic disturbance. The stratigraphy of the blocks
has been illustrated by Audley-Charles (1988).
A lower Jurassic (Pliensbachian-Sinemurian)
breakup unconformity in the west Papuan Basin
indicates that Indian Ocean sea floor extended east
60N
30N
30S
60S-
120E 180E 240E
Fig. 15. Relative motions of the Australia, Pacific and
Eurasia plates in the early Palaeogene (64-56 Ma) from
Gordon & Jurdy 0988).
at least to eastern New Guinea at that time (Pigram
& Symonds 1991). Because of the divergence
between Australia and Pacific motions prior to
43 Ma (Fig. 15) determined by Gordon & Jurdy
(1988), it is unlikely that Pacific Ocean spreading
would have extended far west of Papua New
Guinea. Consequently the Australia plate boundary
probably ran west of north to the Asian margin or
the southern boundary of the spreading systems of
the Celebes and Molucca Seas and West Philippine
Basin. Since Phase 2 Indian Ocean sea floor
spreading (Late Cretaceous to middle Eocene) may
have extended east to western Sumatra (Fig. 4)
Jurassic to mid-Cretaceous crust of the Australia
plate may have extended from western New Guinea
to western Sumatra from the Late Cretaceous to the
middle Eocene.
Palaeomagnetic studies on the microcontinents
suggest that they were separated from the
Australian craton by Late Cretaceous time. Ali &
Hall (1995) concluded that palaeomagnetic results
from an Upper Cretaceous marl (86_ 3 Ma) at a
site on Sula indicated a formation palaeolatitude of
19 _+6°S, well to the north of Australia. Wensinck
et al. (1989) obtained similar results from palaeo-
magnetic measurements on Santonian and
Maastrichtian sediments from Misool. The palaeo-
latitudes obtained were 18.5 _ 3.2 ° and 21.5 _ 3.6 °
respectively. Reconstructions by Struckmeyer et al.
(1993) place the Late Cretaceous Papuan margin
at a palaeolatitude of 30°S, roughly 10°S of the
microcontinents at that time. The north-south
 CENOZOIC SE ASIA RECONSTRUCTIONS
relationship to Australia has been substantially
restored by shortening in the Banda subduction
zone. The Sula site and the Santonian Misool sites
both give counter-clockwise rotations of around
40 ° but the Maastrichtian site on Misool is not
rotated. Ali & Hall (1995) also found strong
counter-clockwise rotations in Upper Cretaceous,
upper Oligocene and middle Miocene volcanic
rocks from north Obi. They relate the rotations to
left-lateral shear on the left-lateral Strong Fault
Zone since as the sites are adjacent to, or between
splinters of the fault.
There is no clear evidence of latitudinal separa-
tion of the Kemum Terrane of the Bird's Head
from the Australian margin (Giddings et al. 1993).
By comparing palaeomagnetic results from the
Kemum Terrane with the Australian polar wander
path they suggested that the terrane, originally
located in eastern New Guinea, was rotated 55 °
counter-clockwise between Early Jurassic and
middle Eocene time, then in the late Eocene and
early Miocene interval moved westward to near its
present location. Finally there was 10° of post-
middle Miocene counter-clockwise rotation.
Unfortuiaately Cretaceous sediments did not yield
poles because of strong overprinting. Their data are
open to an alternative interpretation. The westward
movement of the terrane to the Bird's Head with a
decreased of palaeolatitude and counter-clockwise
rotation could have taken place between the Early
Jurassic and middle Eocene time followed by a 20 °
late Eocene and early Miocene clockwise rotation.
Because the microcontinental blocks and Bird's
Head terranes were were not involved in Oligocene
tectonism of the New Guinea orogen, they were
probably located to the west of the cratonic part of
Irian Jaya by early Oligocene time as indicated on
Fig. 13. From the foregoing discussion of palaeo-
magnetic data, Jurassic rifting may have removed
them from eastern New Guinea to reach the
indicated location by Late Cretaceous time. Hall
(1995), Struckmeyer et aL (1993) and Rangin et al.
(1990) all use a generally similar location for the
blocks in their reconstructions.
Using the palaeomagnetic interpretation of
Giddings et al. (1993), Struckmeyer et al. (1993)
moved the blocks westward by a complex Late
Cretaceous to Palaeocene spreading system,
followed by southwest movement on a major left-
lateral southwest strike-slip fault. It is difficult to
relate these events to known plate boundaries and
vectors.
In their reconstruction Daly et al. (1991) rotated
Seram and Buru counter-clockwise to their present
position from a location on the northwest
Australian margin adjacent to Timor extending
north to the Bird's Head of New Guinea.
Banggai-Sula derived from the Birdfs Head is
143
located immediately to the north (Fig. 10). In their
tectonic evolution, Banggai-Sula is transported
west in the late Miocene time. Southward directed
deformation in the Bird's Head and left-lateral
strike-slip on the Tarera Fault are responsible for
rotation of the Burn and Seram, arriving at their
present orientation as Pliocene underthrusting
occurred in the Seram Trough. This scheme avoids
the problem of moving the blocks west from the
Papua New Guinea margin but their pre-Triassic
metamorphic basement (Kemp & Mogg 1992)
indicates affinities with the eastern Australian
tectonic province. The location does not conform
to the Cretaceous palaeolatitude data but the sense
of rotation is in agreement with the measurements.
Packham (1990) assumed that initial rifting of
the blocks took place in the Mesozoic but that they
remained close to the New Guinea margin as
marginal plateaus away from clastic deposition.
About the beginning of Miocene time, they were
transferred to the Philippine Sea plate and started
to move west on the northern margin of the
developing New Guinea Orogen (Fig. 10). The
relative continuity of deposition non-clastic deposi-
tion makes this process problematical and in the
case of Buton an origin in northern New Guinea in
the vicinity of the Bird's Head is assumed. This
would involve an earlier episode of westward
movement. Again the palaeolatitude determinations
do not fit.
Models for North New Guinea and
the Banda Sea
Although the tectonics of the New Guinea-Banda
are driven principally by the west-northwest
convergence of the Pacific plate and the northward-
moving Australia plate, the Pacific plate and the
Australia plate are not in direct contact in the
region. They are separated by the southern end
of the Philippine Sea and Caroline plates. The
Caroline plate (Fig. 14) was formed in Oligocene
time by north-south spreading when, in the Hall
et al. (1995b) model, the Philippine Sea plate was
stationary. The Caroline plate has a slow counter-
clockwise motion relative to the Pacific plate about
a pole located about 5 ° north of it (Weissel &
Anderson 1978). It has been rotating away from the
Philippine Sea plate since about 12 Ma from near
the apex of the southward opening Ayu Trough and
is being slowly subducted under the Philippine Sea
plate in the Palau Trench (Fig. 11). Early Miocene
volcanism (20 Ma) reported from Palau by Haston
et aL (1988) suggests that it may have been part
of the Pacific plate before 12 Ma.
Caroline plate motion is transmitted westward
into the seismically active New Guinea Highland
144 G. PACKHAM
region (Abers & McCaffrey 1988) and thence
westward into the Banda Sea. In the south Banda
region the strain is principally absorbed by
thrusting into the Seram Trough but some of it is
transmitted westward along the Sorong Fault Zone
into Sulawesi and eventually to the North Sulawesi
Trench (Silver et al. 1983). West of the Bird's
Head, the northward motion of the Australia
plate is absorbed in the Tanimbar-Timor trough,
backarc thrusting north of Wetar and north-south
shortening of the Banda Sea (McCaffrey 1988).
The present tectonic configuration, dominated by
strong left-lateral shear has led to models in which
terranes were transported westward under its
influence. Hamilton (1979), Charlton (1986) and
Packham (1990) considered that the blocks were
driven west from northern New Guinea along with
the development of the Sorong Fault. As discussed
above models which propose that the micro-
continental blocks of the Banda Sea extended west
from the Bird's Head before the collision of arc
elements with north New Guinea are more
probable. Collision tectonics has dominated the
explanations for the formation of the New Guinea
Orogen (Pigram & Davies 1987; Struckmeyer et al.
1993) but recently extensional tectonics for the
Late Palaeogene phase has been invoked by Hill
et al. (1993).
The Philippine Sea p l a t e a n d N o r t h
N e w G u i n e a in the P a l a e o g e n e
In the Melanesian arc middle Eocene volcanic
rocks resting on oceanic basement mark the onset
of volcanism. The identified western limit of the arc
is the Finisterre Terrane of northeast New Guinea
(Fig. 14). Distinct from it to the west is the
Torricelli Terrane, in which Cretaceous to
Oligocene volcanic activity is recorded (Fig. 14). A
similar late Cretaceous commencement of arc
volcanism is also found nearby in Halmahera (Hall
et al. 1991; Ali & Hall 1995) and further afield in
the Philippines. Similar volcanic rocks may be pre-
served in the Arfak, Waigeo, Tosem, Biak/Superiori
and Gautier Terranes now located on and adjacent
to north New Guinea (Pigram & Davies 1987;
Charlton et al. 1991; Struckmeyer et al. 1993). This
volcanic assemblage, referred to here informally as
the northern arc, may have formed a south-facing
arc on the margin of the young Philippine Sea plate
as reconstructed by Hall et aI. (1995b) and Hall
(1996). The development of the arc commenced
before the Eocene rotational phase of the plate and
continued through the late Eocene and Oligocene
non-rotational phase.
The collision of the West New Guinea
Composite Block with the New Guinea margin in
mid-Oligocene time as Australia moved north has
been regarded as the first stage of the development
of the North New Guinea Orogen (Pigram &
Davies 1987, Struckmeyer et al. 1993). In their
models it had previously been rifted off and
collided over a north-dipping subduction zone. In
view of the proximity of the southern boundary of
the Philippine Sea plate to the northern New
Guinea margin at the time, as depicted by Hall et al.
(1995b), an alternative explanation is that emplace-
ment of the West New Guinea Composite Block
was the result of a collision between part of the
Philippine Sea plate and the New Guinea margin.
The Torricelli Terrane and the other volcanic
terranes to the west listed above may have
originated from the margin of the Philippine plate,
some of them from the gap between the present
Philippine Islands and Halmahera, being left
behind during Miocene clockwise rotation of the
Philippine Sea plate.
Hill et al. (1993), taking a very different
approach, interpreted the West New Guinea
Composite Block as part of the continental margin
in which extension has exposed mid-crustal rocks
as core complexes. The Sepik and Ramu Basins to
the north of the block were initiated as half-graben
in late Oligocene time. The model requires
continuous slow southward late Eocene to late
Oligocene subduction under the New Guinea
margin with some compressional deformation.
Most of the convergence was taken up to the north,
away from the New Guinea margin. This explains
the development of the composite block before
the late Oligocene docking of the arc terranes
commenced of which the first was the Torricelli
Terrane.
M i o c e n e B a n d a Sea tectonics
As a consequence of Miocene clockwise rotation
of the Philippine Sea plate, Ali & Hall (1995) have
determined that at its contact with the northern New
Guinea, the plate boundary was a left-lateral strike-
slip fault. The northward component of the
Philippine Sea plate vector approximately balances
the northward Australia plate vector leaving a
resultant left-lateral movement of the Philippine
Sea plate initiating westward movement of blocks
on the Sorong Fault. Until Philippine plate motion
changed at the end of the Miocene, strike-slip
relationship could not have persisted far west.
Westwards along the Sorong Fault, the distance
from the closely located Euler Pole of rotation
of the Philippine Sea plate increases hence the
velocity vector increases and has a more northerly
azimuth while the Australia plate vector is nearly
constant. The slip direction would swing towards
the north.
In a simple model of the Philippine Sea plate in
 CENOZOIC SE ASIA RECONSTRUCTIONS
which the boundary swung to the north, west of
Halmahera as envisaged by Rangin et aL (1990)
and Daly et al. (1991) the Molucca and Banda Sea
floors were on the Australia plate. They could be
ocean crust (Fig. 13) trapped when the Banda arc
formed by eastward extension of the Sunda arc. The
Banda Sea at least has heat flow levels and oceanic
sea floor depths appropriate for a late Mesozoic age
of the Molucca Sea as discussed above.
A rough minimum estimate of the age of the
Banda arc can be obtained from the 1000 km length
of the subducted slab in the vicinity of Timor
(McCaffrey 1989). Using a convergence rate
calculated from 7 9 m m a -1 calculated from
NUVEL-1, the age is over 12 Ma. Lavas on Wetar
range from 7.7 Ma to 3 Ma but granodiorite and
diorite intrusive rocks on the island are 12 Ma old
(Abbott & Chamalaun 1981). The maximum age is
possibly 15 Ma. Simple restoration of the 1000 km
of convergence of the Banda arc on the west and
200 km on the east would place the micro-
continental blocks in a line running west-northwest
from the Bird's Head at the time of the onset of sub-
duction (Fig. 13).
In the simple model proposed above, collision
of the Buton-Tukang Besi block in the middle
Miocene could have triggered the westward
extension of the Sunda arc, initiating the Banda
arc creating a transitory Banda-Molucca plate. As
subduction on its eastern extension was limited by
the Australian margin, the crust of the Banda Sea
was rotated counter-clockwise about a pole located
near the western end of the Timor Trough, isolating
the Molucca Sea. With two subduction boundaries
and probably a transform southern boundary, the
Molucca plate became a triangle of oceanic crust
larger than at present, able to adjust to changes
in convergence of the surrounding plates. Banda
Sea rotation provides an alternative explanation
for counter-clockwise palaeomagnetic rotations
observed on the microcontinental blocks. Seram
may have been further disturbed by thrusting
into the Seram Trough. Banda Sea rotation was
accompanied by subduction under eastern
Sulawesi, so that Sula collided with the Sulawesi
ophiolite belt in latest Miocene time, later than
Buton. A difficulty with this rotation model is that
the pre-rotation location of the microcontinental
blocks (Fig. 13) overlaps the Molucca Sea crust
shown to the west of Halmahera in the 25 Ma and
15 Ma reconstructions of Hall et al. (1995b). If the
Philippine Sea and Australia plates had been a
degree or so further apart or a slightly different
shape, the overlap would be removed. Alternatively
westward motion may have redistributed the blocks
to some degree. A further difficulty is that it does
not explain the upper Miocene volcanic rocks and
granites in central West Sulawesi.
145
Some of the wide diversity in interpretation of
the Banda Sea region is illustrated in Fig. 10.
Rangin et al. (1990) reconstructed the Banda arc as
an east-west arc, formed in the middle Miocene,
attached at the eastern end to western New Guinea
at the northern margin of the craton. The sub-
duction zone was south of a rifted continental block
which was incorporated in Timor. The western end
of the arc was joined onto the eastern Sunda arc in
the Pliocene as Australia moved north. In their
reconstructions Daly et al. (1991) subducted the old
crust Banda crust under eastern Sulawesi as it
moved westward, driven indirectly by Pacific plate
motion, to collide with western Sulawesi in late
Miocene time. In their model the Banda arc was
apparently formed after the collision. The change
of direction of Philippine Sea plate motion to close
to west at 5 Ma (Seno et al. 1993) introduced a new
tectonic regime which will be discussed in the
next section.
Caroline~Pacific plate motion and
eastern Indonesia
The collision of the Ontong Java Plateau on the
Pacific plate with the Melanesian arc in the
Solomon Islands (Packham 1973; Vedder 1986)
late in early Miocene time was an event remote
from eastern Indonesia but its ultimate conse-
quences extend east to the Banda Sea. The collision
reversed the polarity of the Melanesian arc, trans-
ferring it from the Australia plate to the Pacific
plate which transported it westward to collide with
northern New Guinea in late Miocene time (Pigram
& Davies 1987). With the western end of the
arc now identified as the Finisterre Terrane fixed,
westward transport ruptured the arc duplicating it
for about 1000 km between New Britain and
Marius (Fig. 14). This amount of overlap required
early late Miocene rupturing to have occurred.
Although the commencement of terrane docking
has not been dated precisely, deformed upper
Miocene sediments in the Ramu Basin are overlain
by a thick lower Pliocene to Pleistocene successor
basin deposit (Hill et al. 1993).
The progressive exposure of northern New
Guinea to convergence between the Caroline and
Australia plates is illustrated in Fig. 14. At 15 Ma,
the eastern edge of the Philippine Sea plate was
near the future site of docking of the Finisterre
Terrane. Before the Finisterre Terrain docked there
was no significant coupling between the Philippine
Sea and Australia plates except in the vicinity of
the boundary (the Sorong Fault). As already
mentioned, the northward component of the motion
of the Philippine Sea plate and the Australia plate
approximately balanced, leaving a residual left-
146 G. PACKHAM
lateral strike-slip. The motion of the Caroline (or
Pacific) plate could not be transferred to the
Australia plate until docking of the Finisterre
Terrane commenced (at about 7 Ma) because the
Melanesian arc was south-facing. From about 7 Ma
combined crustal shortening and lateral translation
extended f r o m the docked part of the Finisterre
Terrain to the eastern boundary of the Philippine
Sea plate. As the Philippine Sea plate moved west,
the interface increased in length from around
600-800 km as the terrane docked to its present
length of about 1200 km by 3.8 Ma when northern
segment of the arc started moving west with the
Caroline plate and extension commenced in the
Manus Basin, isolating the South Bismarck plate
from the Solomon plate (Taylor 1979). By 4.0 _+
0.5 Ma, thrusting and uplift had exposed sediments
in the Muller Anticline in the western part of the
Papuan Thrust Belt (Hill 1991).
Hill (1991) estimated the amount of shortening
in the Papuan fold belt at c. 100 km. From the
topography, Abers & McCaffrey (1988) suggested
it was 6 1 k m east of 140°E and 44kin in the
western highlands. These latter figures do not take
account of loss from erosion. The convergence rate
across the Australia-Caroline plate Boundary at
140°E and 2.5°S, is 97 mm a -1 towards 21°S of W
(calculated from NUVEL-1 and the Caroline-
Pacific pole of Weissel & Anderson 1978). This is
equivalent to north-south and east-west compo-
nents of 47 and 85 mm a-1 respectively. The
amount of north-south shortening in the east is
about 325 km. In a Finisterre-west Papuan Fold
Belt transect there is probably c. 200 km of
shortening across the fold belt and thrusting of the
Finisterre Terrain onto the margin identified by
Pegler et al. (1995). This accounts for about 4 Ma
of shortening. But much of the thrusting in the
eastern part had been completed by 4 Ma possibly
indicating a start at 6-7 Ma. This leaves over
100 km of shortening unaccounted for. In eastern
Irian Jaya shortening has been only possible for a
shorter time, probably less than 5 Ma, suggesting a
possible 240 km of shortening. As expected, the
thrust belt narrows westwards and the highlands
are of smaller mass. Even if the Abers &
McCaffrey estimate of 44 km of convergence is
doubled allowing for erosion, it still leaves about
150 km to be accounted for. The excess can
probably be accommodated in underthrusting into
the North New Guinea Trench and its former
extension in the east.
Seismic activity in the New Guinea orogen,
indicates that much of the strain is taken up in the
west by left-lateral strike-slip, especially near the
southern margin and while in the east, thrusting is
more common (Abers & McCaffrey 1988). In order
to account for the thrust direction in the Seram
Trough, McCaffrey (1988) concluded that the
Bird's Head is decoupled from the Australia plate
and is moving west or southwest with respect to it,
probably along the east-west left-lateral Tarera
Fault, south of the Bird's Head. The western end of
the fault forms a triple junction between the Bird's
Head, Banda Sea and the Australia plate where
it meets the Seram and Tanimbar Troughs.
The left-lateral component of convergence
between the Caroline and Australia plate distri-
buted across the orogen, is c. 600 km in the east,
if strain was effectively initiated at about 7 Ma and
possibly 400-450 km in the west if commence-
ment was c. 5 Ma. Some of the strain has been
absorbed in the eastern Bird's Head in the
Lengguru Thrust Belt where orogenic sedimen-
tation commenced in the Pliocene (post-N18,
c. 5 Ma) and thrusting continued till at least 2 Ma
(Moffat et al. 1991). The remainder of the Bird's
Head has behaved as a single entity throughout the
Neogene, bounded on the north and south by the
Sorong and Tarera left-lateral strike-slip faults.
This is kinematically consistent with a post-middle
Miocene 10° counter-clockwise rotation of Kemum
Terrane identified by Giddings et al. (1993).
Much of the strain was probably taken up by
westward or southwestward movement of the
Bird's Head into the Seram Trough where the
seismic zone beneath Seram extends to 300 km
(McCaffrey 1988). On stratigraphic evidence from
Seram deformation commenced at 5 Ma and
continued throughout Pleistocene time (Kemp &
Mogg 1992). Possibly 150km of the westward
movement of the Bird's Head is expressed in left-
lateral displacement of the thrust front on the Tarera
Fault. The displacement could have been achieved
in less than 2 Ma if the strain was concentrated on
the fault.
Apart from thrusting in Seram associated with
a strike-slip component on the Seram Trough, only
an insignificant amount to the convergence south-
west of the trough has been transmitted to the
southern part of the Banda Sea. Shortening of over
400 km, commencing at c. 5 Ma, can be accounted
for in the undeahrust slab beneath Seram plus the
shortening in the Lenggum Thrust Belt to which
could be added a contribution from oblique sub-
duction into the Noah New Guinea Trench. North
of the Seram Trough, shortening has been trans-
mitted to the west into Sulawesi between the Seram
Trough and within strands of the Sorong Fault,
absorption can be identified in the thrusting in the
Banggai-Sula block and East Sulawesi between
5.2 and 3.8 Ma (Davies 1990) and an estimated
250 km of left-lateral slip on the Matano-Palu
Faults Sulawesi (Silver et al. 1983).
Finally, it should be noted that large-scale
tectonic transport of terranes from eastern New
 CENOZOIC SE ASIA RECONSTRUCTIONS
Guinea to the Banda Sea region could not have
been driven by the Caroline plate interacting with
New Guinea because the amount of convergence
was too small. Further, the Buton-Tukang Besi
Block had already collided with eastern Sulawesi
and the Sula Block was close to doing so when
interaction commenced close to the end of Miocene
time.
Eastern SE Asia: conclusions
The history of rotation of the Philippine Sea plate
has been recently refined by Hall et al. (1995b). Its
initial southem latitude location places constraints
on reconstruction models of the Banda Sea and the
development of the New Guinea Orogen. Two
earlier phases of clockwise rotation have been
identified prior to the present one. Rotation about
the present pole, located northeast of Japan results
in westward motion at the southern end of the plate.
The rotations about the earlier poles located to the
east of the Philippine islands total 90 ° . The first
was between 50 and 40 Ma and the second between
25 and 5Ma. The second phase is the more
important in regional reconstructions.
The Miocene rotational phase has been
responsible for the middle to late Miocene collision
of the Philippine chain with the Asian Margin in
Mindoro and western Mindanao. The colliding arc,
with the possible exception of Luzon, developed
on the margin of the Philippine Sea plate before
Miocene rotation phase. Hall (1996) suggests
Luzon developed on the margin of the Eurasia plate
and was transferred to the Philippine Sea plate early
in Miocene time.
The crust of the Celebes and Molucca Sea may
have been formed as an eastward Eocene extension
of the West Philippine Sea spreading system, north
of an arc system linking Sulawesi and the
Philippines. There is inconclusive evidence that a
domain including the Celebes Sea, Borneo and the
Malay Peninsular underwent subsequent counter-
clockwise rotation.
Although most, if not all of the continental rocks
in the Timor-Tanimbar orogen have been sliced
from the Australian northwest shelf margin during
its Pliocene collision with the Banda arc, the base-
ment of Sumba straddling the Banda forearc has
probably been rifted off the adjacent Asian margin
in the Palaeogene time. The nearby composite
block of the central Sulawesi metamorphic belt and
the East Sulawesi ophiolite may have originated
on the Australia plate and been emplaced by an
Oligocene collision with West Sulawesi. Micro-
continental blocks of the Banda Sea and the Bird's
Head have an intruded and metamorphosed Permo-
Triassic basement which suggests an origin in
eastern New Guinea.
147
Late Cretaceous palaeolatitude differences of
around 10 ° between New Guinea and Sula, Seram
and Misool and an Early Jurassic breakup un-
conformity in the West Papuan Basin indicate the
blocks were probably rifted off at the time of
breakup. The breakup and the Mesozoic ocean
crust off western Australia suggest that Phase 2
Indian Ocean crust of the Australia plate probably
extended from eastern New Guinea to the Asian
margin and west to Sumatra, and the micro-
continental blocks were contained in it. The lack of
tectonic disturbance in the upper Palaeogene to
lower Neogene stratigraphy on the blocks indicates
they were assembled from the Bird's Head west-
ward before that time away from the developing
New Guinea Orogen. The close proximity of the
Philippine Sea plate to north New Guinea as recon-
structed by Hall et al. (1995b) makes it geo-
metrically improbable that the blocks were
transported to the southwest on the Pacific plate
after 43 Ma as proposed by Struckmeyer et aL
(1993).
The first tectonism in the New Guinea Orogen
occurred in mid-Oligocene time when the
Philippine Sea plate was stationary and well
after middle Eocene increase in northward velocity
of the Australia plate. According to Pigram &
Davies (1987) the West New Guinea Composite
Block comprising variously deformed Mesozoic
continental margin sediments, granitic intrusions,
metamorphic rocks, extrusive rocks and ophiolite
masses collided with the New Guinea continental
margin. Alternatively it could have been a collision
between the New Guinea continental margin and
the edge of the Philippine Sea plate. A third possi-
bility suggested by Hill et aL (1993) is that the
composite block is an extensional core complex
structure developed during Oligocene to early
Miocene time in rocks already located on the New
Guinea margin.
The Torricelli Terrane, north of the composite
block, is part of what is interpreted here as the
Northwest arc developed on the south facing edge
of the Philippine Sea plate. The arc which includes
the Philippines, Halmahera and the various terranes
of north New Guinea from the Torricelli Terrane
west has a volcanic record extending back to
Cretaceous time. It appears that docking the
Torricelli Terrane on the north New Guinea margin
was a late Oligocene event. It may have been
detached from the Philippine Sea plate when
Miocene clockwise rotation commenced and left-
lateral movement on the Sorong Fault was initiated.
The present author's preferred pre-15 Ma model
is for the Philippine Sea plate boundary to curve
northwest along the arc along the strike-slip
direction against the Australia plate extended north
to include the Molucca and Banda Seas prior to the
148 G. PACKHAM
formation of the Banda arc. The 1000 km long sub-
duction zone under the west Banda arc indicates
it was formed at c. 15 Ma presumably as an east-
ward extension of the East Sunda arc, transferring
the Molucca and Banda Seas to the Eurasia plate
from the Australia plate. Subduction may have
been initiated by the collision of Buton with south-
east Sulawesi. It would have reduced the latitude
differences between the microcontinents and
Australia. At about 12 Ma the Molucca plate
became a separate entity when eastward subduction
under Halmahera started. Australian continental
crust first impacted at the east end of the Banda
arc and has been subducted only 400 km or so.
Differential subduction would have rotated the
Banda Sea counter-clockwise, colliding Banggai-
Sula with western Sulawesi. This model does not
explain the late Miocene volcanic and plutonic
event in West Sulawesi.
The remote collision of the Ontong Java Plateau
with the north facing Melanesian arc north of the
Solomon Islands set in train events that spread
west to Sulawesi. The Melanesian arc has a much
shorter history than the western arc having com-
menced forming on oceanic crust in middle Eocene
time. Polarity reversal of the Melanesian arc trans-
ferred it to the Pacific plate. Its eastern end
eventually collided with northeastern New Guinea
to be emplaced as the Finisterre Terrane possibly
c. 7 Ma. With the elimination of the subduction
zone(s) between them, the convergence between
the Caroline and Australia plates was transmitted
across the New Guinea margin for the first time. At
7 Ma the eastern edge of the Philippine Sea plate
was at about 142°E and c. 600--800 km of the
margin was exposed to shortening. This distance
progressively increased to c. 1200 km at c. 4 Ma
when the Philippine Sea plate boundary was at
136°E. Combined strike-slip and spreading then
commenced in the Manus Basin carrying the
Manus arc segment west on the Caroline plate.
Folding and strike-slip faulting in the New Guinea
Orogen spread from east to west.
Assuming that convergence commenced at
c. 7 Ma and 5 Ma in the east and west respectively
the shortening components are c. 330 and 240 km
respectively. This can be accommodated in thrust-
ing under the Finisterre Terrane, thrusting in the
orogen and underthrusting into the North New
Guinea Trench. The strike-slip component in the
west is about 430 km. Some of this was taken up in
the Lengguru Thrust Belt and possibly by oblique
subduction into the North New Guinea Trench and
within the orogen. The remainder must have moved
the Bird's Head underthrusting it into the Seram
Trough and moving it west on the Tarera Fault.
McCaffrey (1988) concluded that Bird's Head is at
present moving west or southwest with respect to
Australia. North of the Seram Trough there may be
partial coupling with the Philippine Sea plate,
transmitting strain westward to the thrusts between
Banggai-Sula and East Sulawesi and to the Palu-
Matano fault system where Silver et al. (1983) have
suggested there is 250 km of left-lateral movement.
More important than explaining the distribution of
the strain is the observation that the total amount is
insufficient to transport blocks from eastern New
Guinea to the Banda Sea region.
Concluding remarks
The way in which the strain identified in the
neotectonics of central Asia is being distributed in
eastern Asia is still uncertain and hence its
applicability to the rather different Palaeogene to
Early Neogene tectonic regime is yet to be
resolved. Kinematics of models partitioning
Palaeogene shortening between crustal thickening
of Tibet and block extrusion will not be determined
until the time of elevation of the Tibet Plateau is
confidently known, although substantial displace-
ments have been documented on the Red River
and Wang Chat Faults. The style of regional
deformation indicated by the dispersal of palaeo-
magnetic results suggests that simple block rotation
does not apply in the northwest Indochina block but
it could well be an important process further east of
the syntaxis. The regional geology north and east of
the of the eastern syntaxis has not yet revealed the
link between the Shan-Thai block and the terranes
of eastern Tibet and the continuity of the eastern
Himalayas to northern Myanmar except in broad
terms. The related late Palaeogene palaeo-
geographical problem of the Indoburman flysch
belt could be resolved by a better definition of the
time of its docking south of the eastern syntaxis as
well as an assessment of its former location on the
Asian margin.
Although the two sections of this paper have
been treated separately, they are interdependent in
that reconstruction of the eastern part of SE Asia
requires knowledge of the location of the eastern
Asian margin from late Oligocene to early Miocene
time. One of the key pieces in the jigsaw is Borneo
which has been alluded to only in passing. The two
major unresolved problems are firstly, the relation-
ship between the South China Sea Cenozoic
spreading and deformation of the Rajang flysch
belt, and secondly, the implications of possible
counter-clockwise rotation of a domain including
the island and adjacent regions and it relationship
to Sundaland basin histories.
While significant advances in the elucidation of
the history of the rotation of the Philippine Sea
plate have been made, its position relative to Asia is
not well constrained as evidenced by the recent
 CENOZOIC SE ASIA RECONSTRUCTIONS
revision o f the location of the present Euler pole of
relative motion by SenD et al. (1993). Allied to the
relationship o f the Philippines to the Philippine Sea
plate and the Asian margin is the absence of a
detailed tectonic model for the Philippine arc from
the Cretaceous onwards. A lack of highly detailed
stratigraphic and palaeomagnetic studies in the
microcontinental blocks, volcanic and oceanic
terranes in the Banda Sea and along the northern
margin of N e w Guinea hampers a more complete
interpretation of the convergence and collision
References
ABBOTT, M. J. & CHAMALAUN, F. H. 1981.
Geochronology of some Banda Arc Volcanics. In:
BARBER, A. J. & WIRYOSUJONO, S. (eds) The
Geology and Tectonics of Eastern Indonesia.
Geological Research and Development Centre,
Bandung, Special Publication, 2, 253-268.
ABERS, G. & MCCAFFREY,R. 1988. Active deformation in
the New Guinea fold and thrust belt: seismological
evidence for strike-slip faulting and basement
involved thrusting. Journal of Geophysical
Research, 93, 13 332-13 354.
ALl, J. R. & HALL, R. 1995. Evolution of the boundary
between the Philippine Sea plate: palaeomagnetic
evidence from eastern Indonesia. Tectonophysics,
251, 251-275.
ALLEN, C. R., GILLESPIE,A. R., HAN YUAN, SLED,K. E.,
Zhuang Buchum & Zhu Chengnan. 1984. Red River
and associated faults, Yunnan Province, China.
Quaternary geology, slip rates and seismic hazard.
Bulletin of the Geological Society of America, 95,
686-700.
AUDLEY-CHARLES,M. G. 1988. Evolution of the southern
margin of Tethys (North Australian region) from
Early Permian to Late Cretaceous. In: AUDLEY-
CHARLES, M. G. & HALLAM, A. (eds) Gondwana
and Tethys. Geological Society, London, Special
Publication, 37, 79-100.
AURELIO,M. A., BARRIER,E., RANGIN,C. & MULLER,C.
1991. The Philippine Fault in the Late Cenozoic
tectonic evolution of the Bondoc-Masbate-
N. Leyte area, central Philippines. Journal of
Southeast Asian Earth Sciences, 6, 221-238.
AVOUAC, J. P. & TAPPO~IER, P. 1993. Kinematic model
of active deformation in central Asia. Geophysical
Research Letters, 20, 858-898.
- - , BAI, M., YoY, H. & WANG, G. 1993. Active
thrusting and folding along the northern Tien Shun
and the Late Cenozoic rotation of Tarim relative
to Dzunggaria and Kazakhstan. Journal of
Geophysical Research, 98, 6715-6744.
BENDER, F. 1983. Geology of Burma. Gebuder
Borntraeger, Bedim
BESSE, J. & COURTILLOT,V. 1991. Revised and synthetic
apparent polar wander paths of the African,
Eurasian, North American and Indian plates, and
time polar wander since 200Ma. Journal of
Geophysical Research, 96, 4029-4050.
BIRD, P. R., QUINTON,N. A., BEESON, M. N. & BRISTOW,
C. 1993. Mindoro: a rifted microcontinent in
149
history, as do the resolution o f the origin o f
the West N e w G u i n e a C o m p o s i t e Block, the
central Sulawesi Metamorphic Belt and the age
of the volcanic rocks of the Banda Sea and the
Sangihe arc.
I thank the University of London SE Asian Research
Group for sponsoring my attendance at the 1994
conference on the Tectonic Evolution of SE Asia and
R. Hall in particular for his assistance and the provision
of unpublished manuscripts. Petroconsultants (Australia)
Pty. Ltd. have funded the preparation of the text figures.
collision with the Philippines volcanic arc; basin
evolution and hydrocarbon potential. Journal of
Southeast Asian Earth Sciences, 8,449-468.
BRANSDEN, P. J. E. & MATTHEWS,S. J. 1992. Structural
and stratigraphic evolution of the East Java Sea,
Indonesia. In: Indonesian Petroleum Association,
Proceedings 21st Annual Convention. 417-454.
BRtAIS, A., PATRIAT,P. & TAPPONNIER,P. 1993. Updated
interpretation of magnetic anomalies and recon-
structions of the South China basin: implications
for the Tertiary evolution of Southeast Asia. Journal
of Geophysical Research, 98, 6299-6328.
BUROLLET, P. E & SALLE, C. 1981. Tectonic framework
of Eastern Indonesia. In: BARBER, A. J. &
WIRYOSUJONO, S. (eds) The Geology and Tectonics
of Eastern Indonesia. Geological Research and
Development Centre, Bandung, Special Publication,
2, 49-52.
CHARLTON, T. R. 1986. A plate tectonic model of the
eastern Indonesia collision zone. Nature, 319,
394-396.
--, HALL, R. & PARTOYO,E. 1991. The geology and
tectonic evolution of Waigeo Island, NE Indonesia.
Journal of Southeast Asian Earth Sciences, 6,
289-298.
CHEN, P., CHEN, Z. Y. & ZHANG, Q. M. 1993. Sequence
stratigraphy and continental margin development of
the northwest shelf of the South China Sea. AAPG
Bulletin, 77, 842-862.
CHEN, Y., COURTILLOT,V., COGNE, J-P., BESSE, J., YANG,
Z. & ENKIN, R. 1993. The configuration of Asia
prior to the collision of India: Cretaceous paleD-
magnetic constraints. Journal of Geophysical
Research, 98, 21,927-21,942.
COFFIELD, D. Q., BERGMAN, S. C. GARRARD, R. A.,
GuRrrNo, N., ROBINSON,N. M. & TALBOT,J. 1993.
Tectonic and stratigraphic evolution of the Kalosi
PSC area and associated development of a Tertiary
petroleum system, South Sulawesi. In: Indonesian
Petroleum Association, Proceedings 22nd Annual
Convention. 333-356.
COPELAND,P., HARRISON,T. M., PAN YUN, KIDD, W. S. E,
RODEN, M. & ZHANG YUQUAN. 1995. Thermal
evolution of the Gangdese batholiths, southern
Tibet: A history of episodic unroofing. Tectonics,
14, 223-236.
CURRAY,J. R. 1989. The Sunda Arc: a model for oblique
plate convergence. In: Proceedings of the Snellius H
Symposium, Theme: Geology and Geophysics of the
150 G. PACKHAM
Banda Arc and Adjacent Areas, Part 1. Netherlands
Journal of Sea Research, 24, 131-140.
DALY, M. C., COOPER, M. A., WILSON, I., SMITH, D. G. &
HOOPER, B. G. D. 1991. Cenozoic plate tectonics
and basin evolution in Indonesia. Marine and
Petroleum Geology, 8, 2-21.
DAVIDSON, J. W. 1991. The geology and prospectivity
of Buton Island, Southwest Sulawesi, Indonesia. In:
Indonesian Petroleum Association, Proceedings
20th Annual Convention. 209-234.
DAVIES, I. C. 1990. Geology and exploration review of
the Tomori PSC, Eastern Indonesia. In: Indonesian
Petroleum Association, Proceedings 19th Annual
Convention. 333-356.
DE METS, C., GORDON, R. G, ARGUS,D. E & STEIN, S.
1990. Current plate motions. Geophysical Journal
International, 101,425-478.
DEWEY, J. F., CANDE, S., PrrMAN, W. C. III. 1989.
Tectonic evolution of the India/Eurasia collision
Zone. Eclogae Geologicae Helvetiae, 82, 717-734.
, SHACKLETON,R. M., CHANGCHENGFA(~ SUN YIYIN.
1988. The tectonic evolution of the Tibetan Plateau.
Philosophical Transactions of the Royal Society of
London, Series A, 327, 379-413.
ENGLAND, P. & HOUSEMAN, G. 1988. The mechanics
of the Tibetan Plateau. Philosophical Transactions
of the Royal Society of London, Series A, 326,
301-320.
FULLER, M., HASTON, R., LIN, JIN-Lu, RICHTER, B.,
SCHMIDTr,E, E. & ALMASCO,J. 1991. Tertiary paleo-
magnetism around the S China Sea. Journal of
Southeast Asian Earth Sciences, 6, 161-181.
GEARV, E. E., HARRISON, M. T. & HIERTZLER, M. 1988.
Diverse ages and origins of basement complexes,
Luzon, Philippines. Geology, 16, 341-344.
GIBLING, M. R., GRADSTEIN, E M., KRISTIANSEN, I. L.,
NAGY, J., SARTI, M. & WEIDMANN, J. 1994. Early
Cretaceous strata of the Nepal Himalayas: conjugate
margins and rift volcanism during Gondwanaland
breakup. Journal of the Geological Society, London,
151, 269-290.
GIDDINGS, J. W., SANTANA, W. & PIGRAM, C. 1993.
Interpretation of palaeomagnetic results from the
Bird's Head, Irian Jaya: new constraints on the
drift history of the Kemum Terrane. Exploration
Geophysics, 24, 283-290.
GINGEr, D. C., ARDAKUSUMAH,W. O., HEDLEY, R. J. &
Pothecary, J. 1993. Inversion history of the West
Natuna Basin: examples from the Cumi-Cumi PSC.
In: Indonesian Petroleum Association, Proceedings
22nd Annual Convention. 635-678.
GORDON, R. G & JURDY, D. M. 1988. Cenozoic global
plate motions. Journal of Geophysical Research,
91, 12 389-12 406. , Tectonophysics, 160, 23-47.
HALL, R. 1996. Reconstructing Cenozoic SE Asia. This
volume.
--, ALl, J. R. & ANDERSON, C.D. 1995a. Cenozoic
motion of the Philippine Sea plate: palaeomagnetic
evidence from eastern Indonesia. Tectonics, 14,
1117-1132.
--, FULLER,M., ALl, J. R. & ANDERSON,C. D. 1995b.
The Philippine Sea Plate: Magnetism and
Reconstructions. In: TAYLOR,B. & NATLAND,J. H.
(eds) Active Margins and Marginal Basins: A
Synthesis of Western Pacific Drilling Results.
American Geophysical Union Monograph, 88,
371-404.
--, NICHOLS, G., BALLANTYNE, P., CHARLTON, T. &
ALl, J. 1991. The character and significance of
basement rocks of the southern Molucca Sea region.
Journal of Southeast Asian Earth Sciences, 6,
249-258.
HAMILTON, W. 1979. Tectonics of the Indonesian region.
United States Geological Survey, Professional
Paper, 1078.
HARLAND, W. B, ARMSTRONG, R. L., Cox, A. V.,
CRAIG, L. E., SMITH, A. G. & SMITH, D. G. 1990.
A geological time scale 1989. Cambridge University
Press.
HARRISON,T. M., COPELAND,P., KIDD, W. S. E & AN YIN.
1992. Raising Tibet. Science, 255, 1663-1670.
HASAN, K. 1991. Upper Cretaceous flysch succession of
the Balangbaru Formation, Southwest Sulawesi. In:
Indonesian Petroleum Association, Proceedings
20th Annual Convention. 183-208,
HASTON, R., FULLER, M. & SCHMIDTKE, E. 1988.
Paleomagnetic results from Palau, West Caroline
Islands: a constraint on Philippine Sea plate motion.
Geology, 16, 654-657.
HELMERS, H., HEBEDA, E. H. 8(: LUSTENHOUWER,W. J.
1990. The enigmatic relationship between subduc-
tion caused by metamorphism, ophiolitic emplace-
ment and collision on eastern Sulawesi, Indonesia
(abstract). In: Orogenesis in Action, Conference
Abstracts. Geological Society, London, 30.
HILL, K. C. 1991. Structure of the Papuan Fold Belt,
Papua New Guinea. AAPG Bulletin, 75, 857-872.
- - . , GREY, A., FOSTER, D. & BARRETT, R. 1993. An
alternative model for the Oligo-Miocene evolution
of northern PNG and the Sepik-Ramu Basin. In:
CARMAN, G. J. • CARMAN, Z. (eds) Petroleum
Exploration in Papua New Guinea. Proceedings of
the Second PNG Petroleum Convention. 241-259.
HOLLOWAY, N. n. 1982. North Palawan block - its
relation to Asian mainland and role in evolution of
South China Sea. AAPG Bulletin, 66, 1355-1383.
HUANG, K. (~ OPDYKE,N. D. 1993. Paleomagnetic results
from Cretaceous and Jurassic rocks of south and
southwest Yunnan: evidence for large clockwise
rotations in the Indochina and Shan-Thai-Malay
terranes. Earth and Planetary Science Letters, 117,
507-524.
HUCHON, P., LE PICHON,X. & RANGIN,C. 1994. Indochina
Peninsula and the collision of India and Eurasia.
Geology, 22, 27-30.
JOLIVET, L., HUCHON, P. & RANGIN, C. 1989. Tectonic
setting of Western Pacific marginal basins.
KEMP, G & Mock, W. 1992. A re-appraisal of the
geology, tectonics and prospectivity of Seram
Island eastern Indonesia. In: Indonesian Petroleum
Association, Proceedings 21st Annual Convention.
521-552.
LACASSIN, R., LELOUP,P. & TAPPONNIER,P. 1993. Bounds
on strain in large Tertiary shear zones of SE Asia
from boudinage restoration. Journal of Structural
Geology, 15, 677-692.
LE PICHON, X., FOURNIER, M. & JOLIVET, L. 1992.
 CENOZOIC SE ASIA RECONSTRUCTIONS
Kinematics, topography, shortening and extrusion
in the India-Eurasia collision. Tectonics, 11,
1085-1098.
LELOUP, P. H., HARRISON, T. M., RYERSON, E J, CHEN
WENJI, Li Qi, t~r AL. 1993. Structural, petrological
and thermal evolution of a ductile shear zone,
Diancang Shen Yunnan. Journal of Geophysical
Research, 98, 6715-6744.
LINDSAY, J. F., HOLLIDAY,D. W. & HULBERT,A. G. 1991.
Sequence stratigraphy and the evolution of the
Ganges-Bramaputra delta complex. AAPG Bulletin,
75, 1233-1254.
MCCABE, R., HARDER, S., COLE, J. T. & LUMADYO, E.
1993. The use of paleomagnetic studies in under-
standing the complex Tertiary history of East and
Southeast Asia. Journal of Southeast Asian Earth
Sciences, 8, 257-269.
MCCAFFREY, R. 1988. Active tectonics of the eastern
Sunda and Banda arcs. Journal of Geophysical
Research, 93, 15163-15182.
- - 1989. Seismological constraints and speculations
on Banda Arc tectonics. In: Proceedings of the
Snellius H Symposium, Theme: Geology and
Geophysics of the Banda Arc and Adjacent Areas,
part 1. Netherlands Journal of Sea Research, 24,
141-152.
MATTHEWS, S. J. & TODD, S. P. 1993. A tectonostrati-
graphic model for the southern Nam Con Son Basin,
offshore Vietnam. Warta GeologL Newsletter of the
Geological Society of Malaysia, 19, 276.
MERCIER, J-L, ARMIJO, R., TAPPONNIER, P., CAREY-
GAILHARDIS, R. & HAN TONG LIN. 1987. Change
from Late Tertiary compression to Quaternary
extension in southern Tibet during the India-Asia
collision. Tectonics, 6, 275-304.
MrrCHELL, A. H. G. 1993. Cretaceous-Cenozoic tectonic
events in the western Myanmar (Burma) - Assam
region. Journal of the Geological Society, London,
150, 1089-1102.
MOFFAT, D. T., HENAGE, L. E, BRASH, R. A., HARAHAP,
B. H. & TAUER, R. W. 1991. Lengguru, Irian Jaya:
prospect selection using field mapping, balanced
cross-section and gravity modelling. In: Indonesian
Petroleum Association, Proceedings 20th Annual
Convention. 85-106.
MOLNAR, P. & LYON-CAEN,H. 1989. Fault plane solutions
of earthquakes and active tectonics of the Tibetan
plateau and its margins. International Geophysical
Journal, 99, 132-153.
MOORE, G. E, KADARISMAN,D. & SUKAMTO, R. 1980.
New data on the geology of the Talaud Islands,
Molucca Sea. Bulletin of the Geological Research
and Development Centre, Bandung, 3, 5-12.
PACKHAM,G. H. 1973. A speculative Phanerozoic history
of the south-west Pacific. In: COLEMAN, E J. (ed.)
The Western Pacific: island arcs, marginal seas and
geochemistry. Western Australian University Press,
369-388.
1990. plate motions and Southeast Asia: some
consequences for basin development. Proceedings
of the Southeast Asian Petroleum Association, IX,
55-68.
PARKINSON, C. D. 1991. The petrology, structure and
geologic history of the metamorphic rocks of central
151
Sulawesi, Indonesia. PhD Thesis, University of
London.
PE~LER, G., DAS, S. & WOODnOUSE, J. H. 1995. A
seismological study of the eastern New Guinea and
the western Solomon Sea regions and its tectonic
implications. International Geophysical Journal,
122, 961-981.
PELTZER, G. & TAPPONNmR,P. 1988. Formation of strike-
slip faults, rifts, and basins during the India-Asia
collision: an experimental approach. Journal of
Geophysical Research, 93, 15 085-15 117.
PIGRAM, C. J. & DAVIES, H. L. 1987. Terranes and
accretion history of New Guinea. BMR Journal of
Australian Geology and Geophysics, 10, 193-212.
- - & Symonds, E A. 1991. A review of the timing
tectonic events in the New Guinea Orogen. Journal
of Southeast Asian Earth Sciences, 6, 307-318.
POWELL, C. McA., ROOTS, S. R. & VEEVERS,J. J. 1988.
Pre-breakup continental extension in East
Gondwanaland and the early opening of the eastern
Indian Ocean. Tectonophysics, 155, 261-283.
PUBELLIER, M., QUEBRAL,R., RANGIN, C., DEFFONTAINES,
B., MULLER, C., er AL. 1991. The Mindanao
Collision Zone: a soft collision event with a
continuous Neogene strike-slip setting. Journal of
Southeast Asian Earth Sciences, 6, 239-248.
RAMANA, M. V., NAIR, R. R., SARMA, K. V. L. N. S.,
RAMPRASAD, Z., KRISHNA, K. S., ET AL. 1994.
Mesozoic anomalies in the Bay of Bengal. Earth
and Planetary Science Letters, 121,469-475.
RANGrN, C. 1991. The Philippine Mobile Belt: a complex
plate boundary. Journal of Southeast Asian Earth
Sciences, 6, 209-220.
~, JOLIVET, L., PUBELLIER,M. 1990. A simple model
for the tectonic evolution of southeast Asia and
Indonesia region for the past 43 m.y. Bulletin de la
Socidtd gdologique de France, 8 VI, 889-905.
RATMAN, N. 1976. Geological map of the Tolitoli
Quadrangle, North Sulawesi, 1:250,000. Geological
Survey of Indonesia.
RICHTER, B., FULLER, M., SCHMIDTKE,E., TIN MYINT, U.,
TIN NGEW, U., MYA WIN, U. & BUNOPAS, S. 1993.
Paleomagnetic results from Thailand and Myanmar:
implications for tectonic rotations in Southeast
Asia. Journal of Southeast Asian Earth Sciences, 8,
247-256.
RoY, T. K. 1986. Petroleum prospects of the frontal belt
and subduction complex associated with the Indian
plate boundary in the northeast. In: Sixth Offshore
South East Asia Conference. 513-527.
ROYER, J-Y. & SANDWELL, D. T. 1989. Evolution of the
eastern Indian Ocean since the Late Cretaceous:
constraints on Geosat altimetry. Journal of
Geophysical Research, 94, 13 755-13 782.
SALVIDAR-SALI,A., OSTERLE, H. G. & BROWNLEE, D. N.
1982. The geology of offshore northwest Palawan,
Philippines. In: Proceedings of the 2nd Conference
of the Asian Council on Petroleum (ASCOPE)
Exhibition, Manila. 99-123.
SAREWrrz, D. R. & KARIG, D. E. 1986. Process of
allochthonous terrane evolution, Mindoro Island.
Philippines. Tectonics, 5, 525-552.
SEARLE,M. P., WINDLEY,B. E, COWARD,M. P., COOPER,
D. J. W., REX, A. J., ET AL. 1987. The closing of
152 G. PACKHAM
Tethys. Bulletin of the Geological Society of
America, 98, 679-701.
SENO, T., STEIN, S. A., GRIPP, A. E. & DEMETS, C. R.
1993. A model for the motion of the Philippine
Sea plate consistent with NUVEL-1. Journal of
Geophysical Research, 98, 17941-17948.
SHIBUYA,H., HSU, V. & MERRILL, D. 1989. Paleomagnetic
results of Leg 124: Celebes and Sulu Seas. Los, 70,
1365.
SILVER, E. A. & RANGIN, C. 199t. Development of the
Celebes Sea in the context of Western Pacific
marginal basin history. In: SILVER, E. A., RANGIN,
C., BREYMANN,M. YON (eds) Scientific Reports of
the Ocean Drilling Program, College Station Texas.
124, 39-49.
--, MCCAFr.~EY, R. & SMITH, R. B. 1983. Collision,
rotation and the initiation of subduction in the
evolution of Sulawesi, Indonesia. Journal of
Geophysical Research, 83, 1681-1691.
STRUCKMEYER,H. I. M., YUENG,M. & PIGRAM,C. J. 1993.
Mesozoic to Cainozoic plate tectonic evolution
and palaeogeography of the New Guinea Region.
In" CARMAN, G. J. & CARMAN, Z. (eds)'Petroleum
Exploration in Papua New Guinea. Proceedings of
the Second PNG Petroleum Convention. 261-290.
SUKAMTO, R. 1982. The geology of the Pangkajene and
western part of Watampone, Sulawesi 1:250,000.
Geological Research and Development Centre,
Indonesia.
TAPPONNIER, P., PELTZER,G. & ARMIJO,R. 1986. On the
mechanics of the collision between Indian and
Asia. In: COWARD, M. P. & Rms, A. C. (eds)
Collision Tectonics. Geological Society, London,
Special Publication, 19, 1154-157.
--,
9 LE DAIN, A. Y., ARMIJO, R. & COBBOLD, P.
1982. Propagating extrusion tectonics in Asia: new
insights from simple experiments with plasticine.
Geology, 10, 611-616.
TAYLOR, B. 1979. The Bismarck Sea: evolution of a
backarc basin. Geology, 7, 171-174.
- - & HAYES, D. E. 1980. Tectonic evolution of the
South China Sea Basin. In: HAYES, D. E. (ed.) The
Tectonic Evolution of Southeast Asian Seas and
Islands, Part 2. Geophysical Monograph Series,
American Geophysical Union, 27, 89-104.
TRELOAR, P. J. & COWARD, M. R 1991. Indian plate
motion and shape: constraints on the geometry
of the Himalayan orogen. Tectonophysics, 191,
198-198.
VAN DER WERFF, W., KUSNIDA, D. PRASETYO, H. & VAN
WEERING, T. C. E. 1994. Origin of the Sumba fore-
arc basement. Marine and Petroleum Geology, 11,
363-374.
VEDDER, J. G. 1986. Summary of the geology and off-
shore resources of the Solomon Islands. In: VEDDER,
J. G., POUND, K. S. & BOUNDY, S. Q. (eds) Geology
and offshore resources of Pacific island arcs -
central and western Solomon Islands. Circum-
Pacific Council for Energy and Mineral Resources,
Earth Science Series, 4, 295-306.
WAJZER,M. R., BARBER,A. J., HIDAYAT,S. • SUHARSONO.
1991. Accretion, collision and strike-slip faulting:
the Woyla Group as a key to the tectonic evolution
of North Sumatra. Journal of Southeast Asian Earth
Sciences, 6, 447-463.
WEISSEL, J. K. 1980. Evidence for Eocene oceanic crust
in the Celebes Basin.'In: HAYES, D. E. (ed.) The
Tectonic and Geologic Evolution of Southeast Asian
Seas and Islands. American Geophysical Union,
Geophysical Monograph Series, 23, 37-48.
& ANDERSON,R. N. 1978. Is there a Caroline plate?
Earth and Planetary Science Letters, 41, 143-159.
WENSINCK, H., HARTOSUKOHARDIO, S. & SURYANA, Y.
1989. Palaeomagnetism of Cretaceous sediments
from Misool, northeast Indonesia. Proceedings of
the Snellius H Symposium, Theme: Geology and
Geophysics of the Banda Arc and Adjacent Areas,
part 1. Netherlands Journal of Sea Research, 24,
287-301.
YANG, ZENYU & BESSE, J. t993. Paleomagnetic study
of Permian and Mesozoic sedimentary rocks from
Northern Thailand supports the extrusion model
for Indochina. Earth and Planetary Science Letters,
117, 525-552.
Yu, Ho-SING 1990. The Pearl River Mouth Basin: a rift
basin and its geodynamic relationship with the
southeast Eurasian margin. Tectonophysics, 183,
177-186.
ZHAO, W. & MORGAN, W. J. 1985. Uplift of the Tibetan
Plateau. Tectonics, 4, 359-369.
 Reconstructing Cenozoic SE Asia
ROBERT HALL
SE Asia Research Group, Department of Geology,
Royal Holloway, University of London, Egham TW20 OEX, UK

Abstract: Reconstructions of SE Asia at 5 Ma intervals for the past 50 Ma are presented. They
are constrained by new data from the Philippine Sea plate, which forms the eastern boundary of
the region, by recent interpretations of the South China Sea and Eurasian continental margin,
forming the western boundary, and by the known motions of the Indian-Australian plate to the
south. An attempt is made to satisfy geological and palaeomagnetic data from the region. The
implications of these reconstructions for the Tertiary evolution of SE Asia are discussed in the
light of other new data from the region. There are two regionally important periods of change
during the past 50 Ma. Both appear to be the expression of arc-continent collision and resulted
in major changes in the configuration of the region and in the character of plate boundaries. At
c. 25 Ma the collision of the Australian continent with the Philippine Sea plate arc caused major
effects which propagated westwards through the region. At c. 5 Ma collision of the Philippine arc
and the Eurasian continental margin occurred in Taiwan. This appears to be a key to the recent
tectonics of the region. Principal features of the model include the following interpretations.
Middle Tertiary counter-clockwise rotation of Borneo closed a large proto-South China Sea and
led to the development and destruction of marginal basins north of the Celebes Sea. The rotation
implies that much of the north Borneo margin was not a subduction, but a strike-slip boundary
for most of this period. It also suggests that the central West Philippine Sea, the Celebes Sea and
the Makassar Strait formed part of a single marginal basin which opened between late Eocene and
mid Oligocene, and narrowed westwards like the present South China Sea. Luzon is suggested to
have formed in an arc on the north side of the Celebes Sea-West Philippine Basin, whereas most
of the other Philippine islands probably formed part of an arc at the southern edge of the
Philippine Sea Plate before the Early Miocene. Arc-continent collision in the early Miocene
caused plate boundaries to change and initiated the clockwise rotation of the Philippine Sea plate.
Since then the Philippine fragments have moved in a very narrow zone, mainly as part of the
Philippine Sea plate, with significant strike-slip motion of fragments at the plate margin. Most
subduction under the Philippines was oblique, mainly at the western edge, and north of
Mindanao. The Molucca Sea was a very wide area which formed part of the Philippine Sea plate
before c. 15 Ma and originated as trapped Indian ocean lithosphere. It has been eliminated by
subduction on its east and west sides. The present-day double subduction system never extended
north of the present Molucca Sea into the Philippines. The Sulawesi ophiolite has an Indian ocean
origin and was emplaced on the west Sulawesi continental margin at the end of the Oligocene.
The major change in plate boundaries at the beginning of the Miocene following arc-continent
collision of the Australian margin with the Philippine Sea plate arc caused initiation of the Sorong
Fault system and led to westward movement of continental fragments which were accreted to
Sulawesi during the late Neogene. The Sula platform and Tukang Besi platform formed part of a
single large microcontinent with the Bird's Head before c. 15 Ma. They moved to their present
positions after slicing of fragments from this microcontinent at different times and each was
attached to the Philippine Sea plate for a few million years before collision. Most of the Banda
Sea is interpreted to have an extensional origin and to have opened during the late Neogene. The
reconstructions imply that there has been little convergence at the north Australian margin in Irian
Jaya since the early Miocene and most convergence has occurred during the last c. 5 Ma.
Movement of Philippine Sea arc fragments within the northern New Guinea margin along strike-
slip zones probably accounts for the terrane character of this orogenic belt.

Plate tectonic reconstructions o f SE Asia have understanding m o r e fundamental processes that

some obvious practical value for the region in
helping to understand the d e v e l o p m e n t o f sedi-
m e n t a r y basins, the history o f volcanic arc activity,
and similar processes w h i c h are linked through
tectonics to the distribution o f natural resources.
Reconstructions are also a necessary precursor to
have acted. W h a t are the important controls on the
tectonic d e v e l o p m e n t o f the region (e.g. the role o f
indentor tectonics), what are the critical events
(e.g. different types o f collision event), and what is
the nature o f deformation (e.g. rigid plate versus
distributed d e f o r m a t i o n ) ? H o w far can plate

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolutionof SoutheastAsia, 153

Geological Society Special Publication No. 106, pp. 153-184.
154 R. HALL
tectonics go in describing the development of the
region and the behaviour of crust and lithosphere
(e.g. what is the role of strike-slip faulting versus
contraction in the development of young orogenic
belts)? Similar, more general ~uestions have
relevance to collision processes and to the under-
standing of ancient orogenic belts for which parts
of SE Asia have often been used as analogues.
During the Cenozoic the region which now
forms SE Asia was bounded to the north and west
by a Eurasian plate, and to the south by the Indian-
Australian plate. The motions of these plates are
reasonably well known and their positions provide
limits to the zone within which the SE Asian
collage of microplates and sub-plate fragments can
be moved when attempting plate reconstructions.
The boundaries provided by these plate motions
have been used as limits in previous reconstructions
of Cenozoic SE Asia (e.g. Rangin et al. 1990; Daly
et al. 1991; Lee & Lawver 1994). There have been
considerable differences in dealing with the eastern
boundary of the region. At present, the eastern edge
of SE Asia is bounded by the Philippine Sea plate
but the motion of this plate has been difficult to link
to the global plate circuit because it is surrounded
by subduction zones. Its Tertiary movement history
has proved controversial, as illustrated by previous
reconstructions, and there are major uncertainties
in the position of the eastern edge of the region.
Rangin et al. (1990) accepted evidence for clock-
wise rotation of the plate whereas Daly et al. (1991)
and Lee & Lawver (1994) did not.
Hall et al. (1995b) used palaeomagnetic data
from east Indonesia to estimate Tertiary poles of
rotation for the Philippine Sea plate and made a
new reconstruction of this plate based on these
poles, incorporating the effects of marginal basin
opening within the plate. Discontinuous clockwise
rotation for the entire plate during the last 50 Ma
leads to palaeolatitude predictions which closely fit
all palaeomagnetic data and also satisfies con-
straints on rotation inferred from magnetic anomaly
skewness and seamount magnetization studies from
the Philippine Sea plate. This model has been used
to define the eastern margin of SE Asia as the basis
for reconstructing the region using the ATLAS
computer program (Cambridge Paleomap Services
1993) for the last 50 Ma. Figurel shows the present
geography of the region and identifies the principal
tectonic elements used in the reconstructions.
Approximately sixty fragments (the number
changes with age) have been used in reconstructing
the region. Mercator projections showing recon-
structions of the area bounded by latitudes 20°S
and 30°N, and longitudes 90°E and 160°E, are
presented for 5, 10, 15, 20, 25, 30, 35, 40, 45 and
50 Ma as Figs 2-11.
Details of rotation poles, and fragments used, are
listed in Tables 1 and 2 which are available as
Supplementary Publication No. SUP18101 (5 pp)
from the Society Library or the British Library
Document Supply Centre, Boston Spa, Wetherby,
W Yorks LS23 7BQ, UK. The reconstruction is
also available as an animation, on floppy discs,
which can be run on either a Windows-based PC
or a Mac with adequate hard disc space. Contact
the author for details.
Methods
ATLAS model
The ATLAS model uses a palaeomagnetic refer-
ence frame with Africa as the reference fragment
with its movement defined relative to magnetic
north. Movements of other major plates relative to
Africa are based on Cande & Leslie (1986),
Cochran (1981), Fisher & Sclater (1983), Klitgord
& Schouten (1986), Le Pichon & Francheteau
(1978), McKenzie et al. (1970), Royer & Sandwell
(1989), Sclater et al. (1981), S~goufin & Patriat
(1980) and Srivastava & Tapscott (1986). In these
reconstructions South China is used as a reference
and this is fixed to Eurasia for the period 50-0 Ma.
There has been little Cenozoic motion of Eurasia
whichever reference frame is used (e.g. Livermore
et al. 1984; Gordon & Jurdy 1986; Besse &
Courtillot 1991; Van der Voo 1993) and therefore it
remains in a similar position to the present-day in
all the reconstructions. In the ATLAS model there
are small movements of Eurasia owing to the plate
circuit used, particularly in the last 5 Ma, and there-
fore there are minor differences compared to recon-
structions which use a fixed Eurasia (Lee &Lawver
1994; Rangin et al. 1990).
Palaeomagnetic data
The model attempts to include the important con-
straints imposed by palaeomagnetism. Palaeo-
magnetic data can help to put some limits on
interpretation of geological data since in principle
they provide indications of palaeolatitudes and
rotations. Interpretation is not always simple, and
in SE Asia it is particularly difficult to reach un-
ambiguous solutions. Van der Voo (1993) discusses
in detail the use and problems of using palaeo-
magnetic results, and provides a particularly clear
summary of problems in SE Asia. Besides the
obvious drawbacks of collecting data in predomi-
nantly tropical, remote and often harsh terrain,
there are additional problems such as error limits,
remagnetization, and equatorial ambiguities. Not
least amongst these are the difficulties of deciding
 ..... ~%1: ............ EURASIA ...... ....

Taiwan
So~ China ......... ....

Indochtna

PamceVela
Basin

; ::,:: ' ~r"cles~e~d

<Bank
PHILIPPINE
SEA PLATE

Gulf of
Thailand
Palawan
SUNDA
Malay SHELF
Basins

Natuna I CAROLINE
Basins - PLATE _
f 1 / -

Sulawesi Sula
Borneo
' • Bm.t~
Java Sea
SOUth Basin ~/w
. . Besi Platform
,~-B~da,~ ,~'.' • "_-#// ' RECONSTRUCTIONS
~ " .~B,,,~_.:._..-.c--/
mJ MADE

INDIAN PLATE

Fig. 1. Simplified present-day tectonic configuration of SE Asia. Shaded areas represent mainly ophiolitic, arc and other accreted material added to Eurasian and Australian
margins during the Tertiary. Principal marine magnetic anomalies are shown schematically for the Indian ocean, Pacific ocean, South China Sea, Philippine Sea and Caroline Sea.
Circled Z identifies Zamboanga peninsula of Mindanao. Double lines represent active spreading centres. Narrow lines represent principal bathymetric features of the Philippine Sea k/I
plate and the Caroline Ridge; the margins of the oceanic parts of the South China Sea, the Celebes Sea and the Andaman Sea; and deeper parts of the Sulu Sea and the Makassar
Strait. Complexities in the Bismarck-Solomon Sea regions are not shown.
156 R. HALL
which interpretations of palaeomagnetic results
should be accepted.
One major problem with palaeomagnetic data,
often not emphasized by the palaeomagnetists, is
determining the motion history of a region. This is
particularly important in areas where, for example,
there are results from Mesozoic or older rocks, but
few or no results from Tertiary rocks, such as the
Malay peninsula and Thailand. Similarities in
declinations of similar age rocks may indicate a
common regional rotation history but in some cases
similar declinations may be the results of different
Tertiary motion histories for separate blocks
(cf. Celebes Sea counter-clockwise rotation and
age, v. Borneo counter-clockwise rotation and
age, which are discussed below). Sea floor
magnetic data from the South China Sea, for the
Celebes Sea and the West Philippine Sea
central basin also provide limits, although in
all cases there are uncertainties in anomaly ages
and correlation.
The distinction between local and regional
rotations is far from clear in SE Asia, particularly
because good palaeomagnetic data are sparsely
scattered in time and space, and modelling such as
that attempted here shows clearly the need for long
term, systematic and integrated palaeomagnetic-
geological studies of the region. It is possible to
avoid many difficulties with palaeomagnetic data
by explaining apparently anomalous or contro-
versial evidence of movements as a consequence of
local tectonics. This is often sensible, and Surmont
et al. (1994) provide an excellent SE Asian
example of how previously inferred large regional
rotations are better interpreted as very local conse-
quences of tectonics. However, at the same time
they show that smaller systematic regional rotations
can be recognized, and we must accept that much
essential evidence for movements, for palaeo-
geographic reconstructions, and certainly for
rotation about vertical axes, can only be acquired
from palaeomagnetism and therefore to ignore all
palaeomagnetic data is a position of despair. The
present author has tried to distinguish regional scale
movements from local movements but in several
cases, principally Sundaland-Borneo and the
Philippines, a choice has had to be made between
different views on the basis of inadequate data. In
these cases decisions were based on regional
geological arguments, recognizing that other
solutions are possible. But once a critical decision
has been made, for example in this model that
of accepting a large counter-clockwise rotation
of Borneo, the reconstructions become both a
development, and a partial test, of the decision; can
reasonable reconstructions then be made that are
consistent with the geological data set for the whole
region? Thus, the reconstructions presented here
should be seen as a way of distinguishing local and
regionally important datasets, identifying targets
for future work, and a possible model of the region
which provides a different, albeit sometimes
controversial, interpretation of the development
of SE Asia.
Principles and tests
The fragments were left at current size in all recon-
structions in order that they remain recognizable.
This is broadly satisfactory for Neogene recon-
structions, except for some areas of volcanic arcs.
The present shape of fragments has been main-
tained although this may contribute to overlap of
blocks if significant new crust has been created,
which must be the case in the Sunda, Banda and the
Philippine arcs. Before the Neogene many of the
fragments may have had quite different sizes and
shapes or simply may not have existed. This is true
for many areas in which the extent or age of the
basements is uncertain, for example, in parts of
Philippines, Sangihe arc, Sulu arc, Java and Banda
arc. In some cases, this has been incorporated by
omitting the fragment before a certain time. For
instance, most of the Bonin forearc did not exist
at c. 50 Ma and probably grew during a period of
rapid volcanism in the Eocene (Stem & Bloomer
1992; Taylor 1992). Thus, some fragments may
not appear on all reconstructions between Figs 2
and 11.
In moving fragments Occam's razor has been
used in an attempt to find the simplest possible
motion histories. Fragments were first attached, or
partially coupled to, a major plate with known
motion; most minor fragments can be linked in this
way to a major plate (Australia, Eurasia, Philippine
Sea). If more complex movements were required,
experiments were made with transferring fragments
from one major plate to another. During the recon-
struction process possible solutions were tested by
asking (1) are the palaeomagnetic data satisfied?
(2) are the geological data satisfied? and (3) is there
overlap of fragments during movements? As dis-
cussed below, both (1) and (2) require subjective
interpretations and judgements; readers will
inevitably have their own views of the value of
these. The most important sources are cited and the
reader may refer to regional compilations and
reviews such as Hamilton (1979); Hutchison (1989)
and Van der Voo (1993) for overviews of geology
and palaeomagnetism. Below are summarized the
main features of the data used and the recon-
structions made for principal sub-areas within SE
Asia. The implications of the judgements and
interpretations made for the development of the
region are then discussed.
 RECONSTRUCTING CENOZOIC SE ASIA 157

Regions considered south Borneo have been separated at the Lupar line,
which is a zone of Mesozoic ophiolites and south
Indochina and South China Sea Borneo has been treated as a single rigid fragment
back to 50 Ma. West Sulawesi separated from east
In the reconstructions South China is fixed to stable
Borneo in the Tertiary, resulting in opening of the
Eurasia. The Indochina block south of the Red
Makassar Strait and the development of large
River fault and north of the Malay peninsula has
sedimentary basins in east Kalimantan. West
been moved using the model of Briais et al. (1993).
Sulawesi has a long Tertiary history of igneous
They suggest c. 550 km of movement on the Red
activity and the present eastern margin of the
River fault system with left-lateral motion between
Sundaland block appears to have been an active
32-15 Ma and some dextral movement since the
margin for much of the Tertiary.
late Miocene. Their average small circle pole
There are two principal large-scale tectonic
(5.3°N, 66.3°E) results in significant overlap of
views of Borneo: one advocates a large counter-
Indochina and South China and this model there-
clockwise rotation of the island, the second argues
fore uses a small circle rotation pole (20.9°S,
for no rotation of Borneo. Palaeomagnetic results
61.5°E) further from the Red River fault, as Briais
are reported by Haile et al. (1977), Haile (1979),
et al. (1993) suggest should be the case, which
Schmidtke et al. (1990), Wahyono & Sunata (1987)
results in a smaller overlap but still provides the
and Lumadyo et al. (1993). These results are
history of contraction and extension that they
reviewed by Fuller et al. (1991) and Lee & Lawver
summarize. The movements on the fault estimated
(1994); the former favour a counter-clockwise
by them yield an approximately linear age-
rotation of the island and the latter favour no
displacement relationship and the model assumes
rotation. It is clear that the existing palaeomagnetic
regular displacement in the interval 15-32 Ma. The
data are inadequate to reach a conclusion and those
area between the Indochina coast and the present
who reject the rotation of Borneo (Lumadyo et al.
north Borneo coast is underlain by continental crust
1993; Lee & Lawver 1994; Rangin et al. 1990)
and the north Borneo margin has been fixed to
emphasize the problems with the data. However,
Indochina to provide an indication of the northern
although not all the evidence points in the same
edge of the proto-South China Sea created by
direction there are also regional geological argu-
removing the extrusion of the Indochina block.
ments that favour rotation and this model accepts a
Taylor & Hayes (1980, 1983) identified ocean
counter-clockwise rotation of southern Borneo. The
floor ~magnetic anomalies and used these to
major obstacle to incorporating the rotation in a
interpret the opening history of the South China
regional tectonic model is determining the position
Sea. This interpretation has been modified by
of the rotation pole. The chosen pole is close to
Briais et al. (1993) and their model of opening,
the northwest corner of Borneo (I°N, ll0°E). This
using their calculated poles of rotation, has been
allows Borneo to remain part of a Sunda block
used in the reconstructions without change except
while permitting the rotational movement to be
for reassigning anomaly ages to the Harland et al.
absorbed within the north Borneo accretionary
(1990) timescale. Palawan and Mindoro have been
complexes by closing a proto-South China Sea. It
moved with Reed Bank in the reconstructions. The
implies some extension between Borneo and the
islands of Palawan and Mindoro are considered to
Malay peninsula and allows the southern boundary
include continental crust of south China origin and
of Sundaland to rotate northwards. Because the
the bathymetric contour marking the north side of
pole is so close to the northwest corner of Borneo it
the Cagayan ridge is assumed to mark the southern
requires no major deformation of the Sunda shelf to
limit of continental crust. The northwest part of the
the northwest, although minor deformation would
Sulu Sea is thus considered to be underlain by
be expected. The earliest inversion event in the
continental crust (Hinz et al. 1991).
West Natuna Basin (Ginger et al. 1993) is Early
Miocene, which is consistent with the timing
chosen for the rotation (see below). The movement
Borneo
requires counter-clockwise motion of west
The basement of Borneo of western and interior Sulawesi with little latitude change, for which
Borneo consists of Palaeozoic and Mesozoic there is some evidence. It fails to account for the
igneous, sedimentary and metamorphic rocks and similarity of the Thailand and peninsula Malaysia
this area behaved more or less as a craton during counter-clockwise rotations reported by McElhinny
the middle and late Tertiary. To the north are et al. (1974) and Schmidtke et al. (1990). However,
younger additions to this continental core which poles further from Borneo which could account for
have been interpreted as subduction accretionary these data result in a much larger proto-South
complexes (Hamilton 1979) although this view is China Sea and also require large latitude changes
not universally accepted. In this work north and which are not seen in the palaeomagnetic data.
 L~
~ tw O0
! \ e ~ ' :: j} Okinawa trough "/

Clockwise
rotation
of Philippine
Sea Plate

trench

CAROLINE .~
PLATE ~\
/

Sula platform
Convergence begins
befween Caroline plate
•~ andNew Guinea

" Inner B*a'ndaarc

p.,,DIII~ ~dllb~l~4 j" Ambon
,pr°pagaresi
10°S

20°S
 RECONSTRUCTING CENOZOIC SE ASIA 159

Data from SW Sulawesi indicate that the SW
arm was close to its present latitude in the late
Jurassic (Haile 1978) and late Palaeogene
(Sasajima et al. 1980) but has since rotated counter-
clockwise by about 45 ° . Directions recorded by
late Miocene volcanic rocks of the SW arm are
indistinguishable from those of the present field.
The palaeomagnetic results from SW Sulawesi are
very similar to those from Cretaceous rocks of
the Malay Peninsula (McElhinny et al. 1974) and
in Borneo (Fuller et al. 1991) there are similar
counter-clockwise rotations since the late Mesozoic
and before the late Miocene. The limited evidence
from west Sulawesi and Borneo suggests that
counter-clockwise rotation occurred before the late
Miocene and sometime during the late Palaeogene
to middle Miocene. This model therefore uses a
rotation of 45 ° between 20 and 10 Ma.
Thai-Malay Peninsula
There are a number of faults at the southern
boundary of Indochina on which there is likely to
have been Tertiary movement although currently
the movement histories of these faults are not well
known. The Indochina block is separated from
Thailand and peninsula Malaysia on a line repre-
senting the Three Pagodas and Wang Chao faults.
This is undoubtedly an oversimplification and
movement of the Malay blocks in the reconstruc-
tions cannot account fully for the formation of the
sedimentary basins of the western part of the South
China Sea. The movement history of the Thai-
Malay region is difficult to determine. Offshore,
in the western South China Sea, are several large
sedimentary basins with complex trans-tensional
histories (e.g. Ngah et al. 1996; Tjia & Liew 1996).
The land area largely lacks Tertiary rocks and
palaeomagnetic results do not provide a clear
picture. Post-Cretaceous clockwise rotations are
recorded in Thailand and northern Malaysia
(Schmidtke et al. 1990; Fuller et al. 1991) whereas
counter-clockwise rotations (McElhinny et al.
1974; Halle et al. 1983; Schmidtke et al. 1990;
Fuller et al. 1991) are reported from Tertiary and
older rocks further south. Therefore a north Malaya
block has been separated from a south Malaya
block at the Khlong Marui fault. There is no
evidence for a major suture separating the Malay
peninsula from west Borneo and the counter-
clockwise rotations recorded are similar from both
regions. However, moving both the Malaya blocks
with south Borneo results in major overlap of
the peninsula and Indochina. Hutchison (1989)
observed, based on the compilation of Haile &
Briden (1982), that although the declinations
recorded in rocks from peninsula Malaysia and
Borneo are similar, there are large differences in
inclinations; these differences are also seen in more
recent results (Fuller et al. 1991). Thus, it is
possible that the counter-clockwise rotations of
Borneo and Malaysia are of different ages. The
model rotates the south Malaya block counter-
clockwise about the same pole as that used for
south Borneo but reduces its counter-clockwise
rotation to 15 ° . In contrast, clockwise rotations
could be explained by extrusion of Indochina. The
north Malaya block is rotated clockwise relative to
south China using a pole close to the fragment. This
keeps the Indochina and north and south Malaya
blocks close together and implies late Tertiary
trans-tensional faulting in the zone of the Three
Pagodas fault where the Indochina and north
Malaya fragments overlap before 20 Ma, Rotations
are assumed to have occurred between 20 and
10 Ma in order to maintain the separation of the
peninsula and south Borneo and to maintain the
northern and southern parts of the peninsula as a

Fig. 2. 5 Ma reconstruction of SE Asia. At this stage the Philippine Sea plate was rotating clockwise about a pole
close to its north edge (48°N, 157°E; outside the figure) and the Luzon arc collided with the Eurasian margin in
Taiwan. On this and the following figures Eurasian blocks and blocks forming part of Sundaland, with areas accreted
to its continental core before the early Tertiary, are shown in yellow. The Sunda Shelf and its extensions are shaded in
pale yellow. The present areas of the Indian and Pacific plates are coloured blue. Blocks of Australian continental-
origin are shown in red. Areas shaded in pink are shallow and deep parts of the Australian continental margin.
Submarine parts of Sula, Buton-Tukang Besi, and Bird's Head-related fragments are also shaded with pink. Areas
shown in green are mainly volcanic arc, ophiolite and accreted material of the Ryukyu islands, the Philippines, north
Moluccas, north Borneo, Sulawesi and northern New Guinea. The volcanic islands of the inner Banda arc are shown
in orange. Areas within the Philippine Sea plate filled with magenta are remnant arcs. Thin black lines are used to
show principal marine magnetic anomalies of the Indian ocean, Pacific ocean, South China Sea, Philippine Sea and
Caroline Sea. Thin light blue lines represent marine bathymetry outlining at different stages the present limits
of the Philippine Sea plate, Caroline Ridge, Caroline Sea, Sulu Sea, Andaman Sea, margins of the Makassar
Strait, and the Java-Sunda, the Izu-Bonin-Mariana-Yap-Palau, Negros and Manila trenches. Complexities in the
Bismarck-Solomon regions are not shown. Red lines with short paired arrows represent active spreading centres. Half
arrows represent strike-slip motion. Thick black and red lines represent major faults or fragment sutures used in the
reconstructions. Long arrows indicate motion directions of major plates. Circular arrows represent rotations.
160
broadly continuous fragment. Some extension is
also predicted by the model from about 32 Ma as
a result of Indochina extrusion.
S u m a t r a a n d the A n d a m a n Sea
North Sumatra is fixed to the south Malaya block
for all the reconstructions. Because of the rotation
of south Malaya discussed above, the Sumatran
margin before 20 Ma would have been closer to
N-S and sub-parallel to the motion vector for the
Indian plate. In this configuration it is possible that
the partitioning of convergence into an orthogonal
subduction component and a parallel strike-slip
component would not occur. South Sumatra is fixed
to north Sumatra before 15 Ma. As the counter-
clockwise rotation of the Malay peninsula, Sumatra
and Java proceeded, the angle between the
Sumatran margin and the Indian plate motion
vector would have become less oblique leading to
formation of the dextral strike-slip system. Dextral
motion along the Sumatran Fault zone is incor-
porated between 15 and 0 Ma. Rotating north
Sumatra with the Malay peninsula can also account
for extension in the Andaman Sea. The Andaman
region has been included in the reconstructions but
the model is probably over-simplified. The bathy-
metry of the Andaman Sea is complex (Curray
et al. 1979) and very simplified bathymetric
contours on the east and west sides of the sea are
used as markers, fixed to north and south Sumatra
respectively. This suggests that before c. 10 Ma
there was a small amount of orthogonal extension.
After c. 10 Ma extension was greater but highly
oblique. This is broadly consistent with the age
of the oldest oceanic crust in the Andaman Sea
(c. 11 Ma) and the recent pattern of opening
(Curray et al. 1979).
Java
Java is included largely for completion and has
been rotated counter-clockwise by 30 ° between
20 and 10 Ma. This is a compromise between the
rotations chosen for Sumatra and south Borneo.
There is no evidence for great extension of the
Java Sea during this period (e.g. Bishop 1980;
Van der Weerd & Armin 1992), but if Java is
rotated rigidly with south Borneo there is too much
overlap of Java and Sumatra. The difference in the
amounts of rotation for south Borneo and Java
would permit some extension, and probable strike-
slip faulting, of the east Java Sea between 20-
10 Ma. Because the pole of rotation is so close to
the Borneo, Java and the north Sumatra blocks
there is no significant change in their relative
positions, although there is some overlap of Java
and Sumatra. This could be accounted for by
R. HALL
assuming arc-parallel extension since both Java
and Sumatra are likely to have been smaller than
their present-day outlines, as indicated by the
extensional histories of Neogene sedimentary
basins of Sumatra, and the volume of Neogene arc
volcanic rocks.
The counter-clockwise rotation gives an assess-
ment of the likely pre-middle Miocene orientation
of the subduction zone south of Sumatra and Java
which is closer to NW-SE than present-day. The
rotation also has implications for the tectonic
history of Java (e.g. oblique subduction could have
caused strike-slip motion) and for the timing of
volcanism. The model predicts important changes
in volcanic and tectonic history beginning at
c. 20 Ma for both Java and Sumatra. The sub-
duction boundary south of Java must have changed
eastwards to a more complex link into the Pacific.
Sulawesi
Sulawesi consists of four principal tectonic belts:
the west Sulawesi volcano-plutonic arc, the central
Sulawesi metamorphic belt, the east Sulawesi
ophiolite belt, and the continental fragments of
Banggai-Sula, Tukang Besi and Buton. This
configuration has been widely interpreted in terms
of collision between the eastern micro-continental
fragments and the western volcanic arc (e.g.
Audley-Charles et al. 1972; Katili 1978; Hamilton
1979; Silver et al. 1983b) resulting in ophiolite
emplacement and metamorphism. However, more
recent work shows that this apparent simplicity is
partly a reflection of incomplete knowledge of the
region. There is evidence of several episodes of
subduction beneath the west arm of Sulawesi since
at least the late Cretaceous (Hamilton 1979) and
this formed part of the east Sunda margin since the
early Tertiary. Collision has played a significant
part in the Tertiary development of the island but
the micro-continental fragments arrived later than
initially thought (e.g. Davies 1990; Parkinson 1991;
Smith & Silver 1991).
There has been little palaeomagnetic work on
Sulawesi. The earliest results by Haile (1978) from
the SW and the SE arms indicated that these arms
originated in different regions during the late
Jurassic-early Cretaceous (Audley-Charles et al.
1972; Katili 1978). Data from SW Sulawesi indi-
cate that it was close to its present latitude in the
late Jurassic (Haile 1978) and late Palaeogene
(Sasajima et al. 1980) but rotated clockwise by
c. 45 ° between the late Paleogene and late Miocene
(Mubroto 1988). These results from SW Sulawesi
are very similar to those from Cretaceous rocks of
the Malay Peninsula (McElhinny et al. 1974) and
Borneo (Fuller et al. 1991). Sasajima et al. (1980)
reported Eocene-early Miocene clockwise rotation
 PACIFIC
PLATE

Subduction
at Manila
f
\ \
\
} -~~~ Clodo~ise
rotation ~ ,
iJ
i of Philippine
OUlU Sea Plate <
• Sea,,
rotationcomplete:
,. Suluarc
activity,,
Bomeo CAROLINE
rotation
complete f

MoluccaSea
, doublesubduction
(3 ~ f established
~ 1 "- ,e, Ayu trough ~ f
.,~.~%,,...=._ spreading
-_ N ~ t - . - -

Seram I ~ New Guineastrike-slipsystem

INDIAN PLATE ~ ..~ - • ~4

9U ~1- 1UO °1- l l U~1- l ZU °1- l ;JU~1- 140°~ 150°E 20°S J

Fig. 3. Reconstruction of the region at l0 Ma. Between 5-25 Ma the Philippine Sea plate rotated about a pole at 15°N, I60°E. Counter-clockwise rotation of Borneo and related
rotations of Sundaland were complete.
162 R. HALL
of the east part of the north arm whereas Otofuji
et al. (1981) suggested no significant latitude
change but clockwise rotation of more than 90 ° by
the north arm between the Eocene-early Miocene.
There are problems with interpretation of the
Otofuji etal. (1981) data (J. C. Briden, pers. comm.
1994), and Surmont et al. (1994) show that 20-25 °
clockwise rotation of the whole north arm has
occurred since the Miocene but that larger rotations
are related to local shear zones. These rotations are
consistent with the late Neogene tectonic history
of Sulawesi proposed by Silver et al. (1983a) who
reconstruct the island by removing the movement
on the Palu, Matano, Tolo, Lawanopo and
Kolondale faults. This paper follows their recon-
structions and assumes, as they suggest, that this
deformation occurred between 5 and 0 Ma.
Palaeomagnetic work shows that lavas of the
east Sulawesi ophiolite have a clear southern
hemisphere origin (Mubroto et al. 1994) and
formed at a latitude of 17 _+4°S. This is in marked
contrast to similar Cretaceous and Tertiary rocks
in the Halmahera islands where palaeomagnetic
results from ophiolitic and associated rocks indicate
sub-equatorial latitudes of formation (Hall et al.
1995a; Ali & Hall 1995). Palaeolatitudes of
Sulawesi rocks are north of Cretaceous and the
early Tertiary palaeolatitudes for the northern
Australian margin but similar to Cretaceous
palaeolatitudes for Sula (Ali & Hall 1995) and
Misool (Wensinck et al. 1989). The age and origin
of the east Sulawesi ophiolite is uncertain.
Simandjuntak (1986, 1992) interprets the ophiolite
as formed at an early Cretaceous spreading centre.
K-Ar dates reported by Simandjuntak (1986)
include 93--48 Ma gabbros and 54-38 Ma basalts.
K-Ar ages on lavas by Mubroto et al. (1994) range
from 79-16 Ma. Dating and geochemical studies
by Girardeau et al. (1995) suggest a 44 Ma age and
a backarc origin for part of the ophiolite suggesting
that the ophiolite could be composite. Since the
ophiolite and west arm were juxtaposed by the
early Miocene (Parkinson 1991), the ophiolite is
fixed to west Sulawesi from 25-0 Ma and before
25 Ma moved with the Indian plate.
Makassar Strait
Geological similarities of east Borneo and west
Sulawesi suggest that they have moved apart since
the middle Palaeogene (Hamilton 1979) although
the timing is not well constrained. The Makassar
Strait is thought to be underlain by attenuated
continental crust (Dtirbaum & Hinz 1982) and
stretching occurred between early Palaeogene and
Early Miocene (Situmorang 1982). The Makassar
Strait was closed by fitting bathymetric contours
on the west and east sides of the north and south
Makassar basins although this is only an approxi-
mation due to the thick Neogene sediments of the
Mahakam delta, and young deformation in west
Sulawesi (Bergman et al. 1996). The best fit is
achieved using a pole NE of the north end of the
strait at 6°N, 128°E requiring a rotation of 6 °.
Because the rotation of Borneo and the Philippine
Sea plate results in an alignment of the West
Philippine central basin, the Celebes Sea and the
Makassar Strait, the author suggests that these
basins opened as part of a single basin, probably by
spreading which propagated westwards from the
West Philippine central basin, as discussed further
below. The period of extension is assumed to be
the same (44-34 Ma) which is consistent with the
stratigraphic interpretation of Situmorang (1982,
1987).
Philippine Sea plate
The Philippine Sea plate provides an eastern
boundary for the reconstructions. At present the
plate is rotating clockwise about a pole near its
northern edge (Seno et al. 1993). However, the
Philippine Sea plate is now surrounded by sub-
duction zones which separate it from the oceanic
ridge system and consequently its earlier motion
with respect to other major plates is difficult to
determine. Subduction at the Philippine Trench is
young (Cardwell et aI. 1980) and hence much of the
Philippines must have been attached to the plate
before the late Neogene. At the southern edge of the
plate the Indonesian islands of the north Moluccas
still form part of the plate. Reconstruction of the
Philippine Sea plate and estimation of its past
position are discussed by Hall et al. (1995b) with
details of the data, the blocks, and the rotation poles
used in reconstructing opening of the marginal
basins. For the period between 5 and 0 Ma the
present Eurasia-Philippine Sea plate pole (Seno
et al. 1993) has been used. The rotation poles for
the plate used for the period 50-5 Ma are based on
palaeomagnetic data (Hall et al. 1995a) collected
from the east Indonesian islands of the Halmahera-
Waigeo region which contain a good Mesozoic and
Tertiary stratigraphic record and indicate a long arc
history. These palaeomagnetic data indicate that the
Philippine Sea plate has rotated clockwise in a dis-
continuous manner since the early Tertiary with
c. 35 ° clockwise rotation between 25 and 5 Ma,
no rotation between 40 and 25 Ma and c. 50 °
clockwise rotation between 50 and 40Ma.
Reconstructions based on these results, and
including opening of marginal basins within the
plate, show that other magnetic data from the plate
are consistent with this rotation model (Hall et al.
1995b).
 7 Spreadingends
EuRAsI, \ in Shikoku
"~ basin
15Ma \ PACIFIC
MiddleMiocene PLATE

20°N
f

Clockwise
rotation
of north
~, Malaya
©
i Cagayanfidge ~ ~ k OCIr~a~ti~ise# /
i collisionw#h
north
! Palawan ( Sulu

rotation ~ ,~ / .~--x activel,

<-4.
Sumatran
fault
active
MoluccaSea
subductionbegins
beneathHalmaheraL & ~ "
Tukang •~ AyuTrough ~..~ ~ ~ - ~ ~
~ ~ .~_~ %,. spreadingbegins --~

INDIAN Terranesmovingin north

Bird'sHead "~" q NewGuineastrike-slipsystem
microcontinent
in 'baconslicer'. ~ . • ~

150°E 20°S

Fig. 4. Reconstruction of the region at 15 Ma. Rotation of Borneo and parts of Malaya, Sumatra and Java were underway. Strike-slip motion at the southern boundary of the
Philippine Sea plate fragmented the Bird's Head microcontinent and moved blocks west in the plate boundary zone. Similar motions were occurring in the north Philippines.
o~
The backarc Sulu Sea began to close after collision of the Cagayan ridge with the Palawan margin,
164 R. HALL
Celebes Sea
ODP Leg 124 results indicate that the Celebes Sea
formed in the middle Eocene, probably not far from
its present latitude (Silver & Rangin 1991). The
palaeomagnetic inclinations indicate palaeo-
latitudes similar to the present-day latitude; no
errors are quoted by Shibuya et al. (1991) but the
data allow movement of up to 19 ° according to
Silver & Rangin (1991) implying relatively large
errors. The palaeomagnetic results of Shibuya et al.
(1991) also indicate a counter-clockwise rotation
of c. 60 ° between 42 and 20 Ma. The reconstruc-
tions in this paper imply a connection between the
Celebes Sea and the West Philippine central basin.
The suggestion that these two basins were linked
was discussed by Silver & Rangin (1991) who
argued against it without excluding it, based on
apparent inconsistencies of spreading history, rates
and palaeomagnetism between the two basins. In
fact, there is no major difference between spreading
rates estimated from spacing between anomalies 18
and 20 in the Celebes Sea (Weissel 1980) and the
West Philippine central basin (Hilde & Lee 1984),
and the small difference is consistent with a basin
narrowing westwards. The stratigraphy and sedi-
mentology of the two basins are similar (Nichols &
Hall 1995) based on drilling by DSDP Legs 31, 59
and ODP Leg 124. The reconstructions also show
that the apparently different rotation histories could
lead to a sub-parallel alignment of their magnetic
anomalies at the time of basin formation. In this
model the Celebes Sea basin is suggested to have
originally formed an extension of the West
Philippine central basin which opened between
44 and 34 Ma, based on the ages of anomalies
identified by Hilde & Lee (1984). The model there-
fore opens the basin and includes a 45 ° counter-
clockwise rotation between 44 and 34 Ma implying
that the rotation occurred during opening. The
magnetic anomalies identified by Weissel (1980)
indicate the spreading centre was south of the
present southern edge of the basin suggesting that
part of the ocean has been subducted at the north
Sulawesi trench since the late middle Miocene
(Rangin & Silver 1991). The model assumes
symmetrical spreading and eliminates the southern
half of the ocean between 10 and 0 Ma at the north
Sulawesi trench. The subduction is interpreted to
have resulted largely from rotation of the north arm
of Sulawesi (Surmont et al. 1994).
Sulu Sea-Cagayan ridge
The Sulu Sea is a marginal basin thought to have
opened as a backarc basin during the early Miocene
(Holloway 1982; Hinz et al. 1991; Rangin & Silver
1991) south of the Cagayan ridge, although the
northern part of the Sulu Sea is underlain by
continental crust as noted above. The Cagayan
ridge is interpreted as a volcanic arc active for a
short period in the early Miocene which collided
with the south China margin at the end of the early
Miocene (Rangin & Silver 1991). Part of this arc
may also be present in Mindoro and Tablas
(Marchadier & Rangin 1990). Rangin & Silver
(1991) offer two scenarios for the history of this
region. Their scenario A has been modelled by
southward subduction of a proto-South China Sea
beneath the Cagayan ridge between 20 and 15 Ma
forming the Sulu Sea. Collision of the Cagayan
ridge with Palawan at 15 Ma then resulted in
development of a new subduction zone and south-
ward subduction of part of the Sulu Sea beneath the
Sulu arc between 15 and 10 Ma.
Philippines
With the exception of Palawan, Mindoro,
Zamboanga and nearby parts of the west
Philippines, the Philippine archipelago is composed
of largely ophiolitic and arc rocks of Cretaceous
and Tertiary age. The present Philippine fault is
young, < 5 Ma, (Aurelio et al. 1991 ; Quebral et al.
1994) but there is evidence of older strike-slip
faulting in the northern Philippines (e.g. Rutland
1968; Karig 1983; Karig et al. 1986; Stephan et al.
1986) possibly dating from the early Miocene.
There is widespread evidence of volcanic activity
throughout the Neogene implying subduction. West
Mindanao east of Zamboanga is omitted before
5 Ma since almost all of this block consists of very
young arc material, although it may include some
basement of Eurasian continental affinities
(Ranneft et al. 1960; Pubellier et al. 1991). It is
likely that central Mindanao, although shown on
most of the reconstructions, was also much smaller.
The Philippines are widely considered to have
formed part of an arc system at the edge of the
Philippine Sea plate before the Pliocene (e.g.
Rangin et al. 1985, 1991; Rangin 1991). South of
Luzon palaeomagnetic results indicate clockwise
rotations consistent with movement as part of the
Philippine Sea plate. However, the Philippines are
still palaeomagnetically insufficiently known to
attempt a detailed plate tectonic model, even
assuming this were possible, since so many frag-
ments would be required. However, the general
plate tectonic evolution is known (Rangin et al.
1990) and indicates that most of the Philippines
have moved from the south and have collided with
the Eurasian margin during the Neogene, although
the relative importance of collision and strike-slip
tectonics is uncertain. The tectonic evolution of this
complex region has been simplified by localising
 RECONSTRUCTING CENOZOIC SE ASIA
all strike-slip movement on the present Philippine
Fault and by moving the Philippines south of Luzon
with the Philippine Sea plate. On the whole this
avoids overlap of fragments, except between 10-
5 Ma, and provides some limits on the large-scale
tectonic setting of the region. The motions for the
Philippines south of Luzon in the model are con-
sistent with the palaeomagnetic data which predict
clockwise rotations and northward movement
(McCabe & Cole 1989; Fuller et al. 1991).
The history of Luzon is more controversial.
McCabe & Cole (1989), Fuller et al. (1991) and
Van der Voo (1993) review the different interpreta-
tions of the palaeomagnetic data. This model
accepts the Fuller et al. (1991) preferred interpreta-
tion of a largely counter-clockwise rotation history
with a small latitudinal change for most of the
Tertiary. They interpret relatively young clockwise
rotations to indicate late Neogene movement with
the Philippine Sea plate. Fuller et al. (1983) suggest
that the palaeomagnetic results from Luzon indicate
tectonic models should involve counter-clockwise
rotation of Luzon since mid-Miocene, no important
northward motion (_+ 500 km) since then, north-
ward motion of the Zambales complex from
equatorial latitudes with counter-clockwise motion
since the Eocene, and no significant rotation in
the Plio-Pleistocene. If Luzon is moved with the
Philippine Sea plate and the rest of the Philippines
for the whole of the period 50-0 Ma, most of these
conditions are not met and there are major overlaps
of Luzon, Sulawesi and the Celebes Sea. However,
with the exception of the early counter-clockwise
motion of the Zambales complex (which could be
incorporated in a more detailed model) all of these
constraints are satisfied if Luzon is positioned on
the north side of the Celebes Sea before the
Neogene. For the Neogene, Luzon was moved
using the Philippine Sea plate rotation poles (Hall
et al. 1995b) but at slightly lower rates, which
implies a partial coupling to the plate, between 20-
0 Ma, and assumed 40 ° counter-clockwise rotation
between 25-20 Ma. The position of Luzon in the
reconstructions is different from that normally
assumed (e.g. Rangin et al. 1990) but can also
satisfy many of the geological data, as explained
below.
Halmahera
Reconstructions incorporating the rotation of the
Philippine Sea plate and removing the effects of
subduction at the Philippine Trench can be tested
in part against present-day observations in the
Molucca Sea region. The Philippine arcs terminate
in the south in the Molucca Sea collision zone
where the Halmahera and Sangihe arcs are actively
165
converging. West of Halmahera the Molucca Sea
plate has an inverted U-shaped configuration and is
dipping east under Halmahera and west under the
Sangihe arc. Regional seismicity indicates that
there are c. 200-300 km of subducted lithosphere
beneath Halmahera. On the other side of the
Molucca Sea, the Benioff zone associated with the
west-dipping slab can be identified to a depth of
c. 600 km beneath the Celebes Sea. In central
Halmahera a fold-thrust belt forms the boundary
between the ophiolitic eastern basement and arc
volcanic western basement. Balanced cross-
sections indicate at least 60 km east-west
shortening between east and west Halmahera in the
fold-thrust belt and within the east arms (Hall &
Nichols 1990). The age of thrusting is between
3 and 1 Ma. This movement is incorporated in the
model and probably reflects intra-plate deformation
associated with the change in motion of the plate.
The western arms of Halmahera are covered by late
Neogene to Recent volcanic .products related to
eastward subduction of the Molucca Sea Plate. The
volcanic arc was initiated at c. 12 Ma and sub-
duction probably began at c. 15 Ma. An older phase
of arc volcanic activity commenced in the late
Eocene and terminated in the early Miocene.
Stratigraphic similarities and the relative position
of the Halmahera islands and the east Philippines
suggest that they formed part of the same arc
system before 5 Ma. The Halmahera islands were
moved with the southern Philippine Sea plate in all
the reconstructions.
Bird's Head
The Neogene movement of the Bird's Head has
been estimated from constraints imposed by
reconstruction of the Molucca Sea. Progressively
restoring the oceanic crust subducted at the Sangihe
and Halmahera Trenches requires a wide ocean and
the Bird's Head needs to be moved further south
than the northern Australian margin in order that
the Bird's Head is south of this ocean. McCaffrey
(1996) suggests the Bird's Head is currently
moving south relative to Australia and this is
incorporated by a small movement in the past
0.5 Ma. Before this time the movement of the
Bird's Head has been modelled by assuming left-
lateral strike-slip motion along a boundary parallel
to the Aru Basin edge between 0.5 and 2 Ma. A
small counter-clockwise rotation of the Bird's Head
has been incorporated between 4 and 8 Ma based
on palaeomagnetic results of Giddings et aI. (1993).
A further small strike-slip motion is incorporated
between 8 and 12 Ma, again constrained by recon-
struction of the Molucca Sea. Before 12 Ma the
Bird's Head is fixed to Australia.
 I /I I I

E ~
PLATE
20 Ma " Spreadingin
Shikoku PACIFIC
Early Miocene \ basin PLATE
F~al
spreading
of South
China Sea 20°N__

of Philippine
Sea Plate

/
C" ., y, lSre''r'e lO.
Proto-
South China Sea
Cagayan ridge
separates f r o m /
basins
S~lu Sea
.__~ opens
/.~~y/ cA~o,,.~\ -~\ 1

crust thrust
'h Sulawesi Molucca Sea forms
part of
PhilippineSea Plate
10°S
INDIAN PLATE Sorong Fault
L~'---....~.. systeminitiated
Bird's Head
microconfinent ? ?
dismemberedby
Sorong fault splays . . . . o

20°S

Fig. 5. Reconstruction of the region at 20 Ma. Rotation of Borneo and parts of Malaya, Sumatra and Java began. Subduction of the Proto-South China Sea caused arc splitting
in the Sulu arc and the separation of the active arc of the Cagayan ridge. South China Sea opening propagated SW into the Sunda shelf.
 RECONSTRUCTING CENOZOIC SE ASIA
Sula
The Sula platform is a fragment of continental crust
widely considered to have been transported west
by the Sorong Fault. Its origin is uncertain. Many
authors consider it to be a piece of New Guinea
that was detached from western Irian Jaya in late
Cenozoic time (e.g. Visser & Hermes 1962;
Audley-Charles et al. 1972; Hamilton 1979; Silver
& Smith 1983). In contrast, Pigram et al. (1985)
have proposed that it originated c. 1000 km further
east, and was detached to form an independent
micro-continent from central Papua New Guinea
before the early Cretaceous. All these suggestions
are based on stratigraphic features discussed by
Pigram et al. (1985). The Sula platform is now
attached to east Sulawesi and has a thrust contact
with the east Sulawesi ophiolite. It was originally
suggested that the collision of the west Sulawesi
island arc and the Sula platform resulted in ophio-
lite emplacement in the east arm (Kiandig 1956;
Hamilton 1979; Silver et al. 1983b). However,
recent evidence indicates that this suggestion is
incorrect. The ophiolite was obducted westward
onto west Sulawesi at the end of the Oligocene
(Parkinson 1991), whereas thrusting of the ophio-
lite onto the western edge of the Sula platform
occurred in the latest Miocene (Davies 1990) indi-
cating collision of the Sula platform with east
Sulawesi must have occurred at c. 5 Ma. Sula is
therefore moved with east Sulawesi from 0--5 Ma.
Hamilton (1979) shows the Sula platform as a
fragment moving along the north side of the Sorong
Fault, implying it was attached to the Molucca Sea
or Philippine Sea plates. By moving the Sula
fragment with the Philippine Sea Plate before
5 Ma it fits closely to the Bird's Head at c. 10-
11 Ma. Charlton (1996) shows that the Tomori
and Salawati basins, both of which are sharply
truncated, would have formed the northern and
southern parts of a single large basin if the Sula
platform were moved back to this position. This
need not preclude a Mesozoic separation of a Bird's
Head microcontinent from further east on the north
Australian margin.
B u t o n - T u k a n g Besi
The Tukang Besi platform is a micro-continental
fragment that collided with Sulawesi during the
Miocene, although the timing of the collision is
interpreted differently by different authors.
Davidson (1991) suggested Buton to be a micro-
continental fragment which collided in the early
Miocene with SE Sulawesi, before the Tukang Besi
platform. Here, both have been treated as parts of a
single block, following Smith & Silver (1991), who
consider Buton as part of the Tukang Besi platform.
167
This difference may be explicable if both formed
part of an extended Birdfs Head microcontinent, as
discussed later. Smith & Silver (1991) argue that
collision of the Tukang Besi platform with
Sulawesi was complete before the late Middle
Miocene, since ophiolitic conglomerates and sand-
stones of this age (N13-N14) overlie deformed
basement rocks. They interpret Lower Miocene
conglomerates (N7-N9) that lack ophiolitic debris
to indicate uplift associated with initial detachment
of the platform from the northern New Guinea mar-
gin. If this interpretation is correct, initial uplift
associated with detachment from New Guinea
occurred between c. 15 and 17 Ma and collision
of Tukang Besi with Sulawesi must have been
complete by c. 11 Ma.
To model this history Tukang Besi was fixed to
west and central Sulawesi between 0 and 11 Ma. By
attaching the platform to the southern edge of the
Philippine Sea plate before that time Tukang Besi
returns to a position, west of, and adjacent to the
Sula fragment and the Bird's Head by 14 Ma. To
obtain the best fit requires rotation of the Tukang
Besi block, and this is incorporated by a clockwise
rotation of 40 ° . The author therefore interprets the
sequence of events to be: at c. 15 Ma a strand of
Sorong fault propagates west, south of Tukang
Besi; by 14 Ma Tukang Besi is fully attached to
Philippine Sea plate; at 11 MaTukang Besi collides
with Sulawesi, locking this strand of the Sorong
fault and requiring a development of a new fault
strand which caused the detachment of Sula. It is
interesting to note that the development of Molucca
Sea subduction beneath Halmahera begins at
c. 13-15 Ma, indicated by K-Ar ages of volcanic
rocks and reset ages (Baker & Malaihollo 1996), as
well as biostratigraphic ages. Thus, a locking of
subduction at the west side of the Molucca Sea,
requiring initiation of a new subduction system on
its east side, is temporally linked to development
of the Sorong fault splay.
Seram and Buru
Hamilton (1979) suggested there is only shallow
seismicity associated with the Seram Trough
whereas Cardwell & Isacks (1978) argued that a
deep slab was bent round the entire Banda arc.
McCaffrey (1989) attempted to reconcile these
differences on the basis of an increased number of
better located seismic events, and concluded that
there are two slabs, one subducted at the Timor
trough, and a second subducted at the Seram
trough. The Seram slab extends to no more than
300 km whereas the Indian ocean slab is con-
tinuous to over 600 km indicating that they record
different histories of subduction. The bend in the
Indian ocean slab implies that Seram has moved
168 R. HALL
eastwards while Australia has moved north. During
the late Neogene the Banda arc has migrated east
since the volcanoes become younger and the length
of the subducted slab decreases eastwards. This
configuration was modelled by moving Seram
northwards to its present position relative to the
Bird's Head between 4 and 0 Ma and moving
Seram eastwards relative to the Bird's Head from
12-4 Ma. This is consistent with, although simpli-
fied from, the present tectonic configuration in the
Banda Sea inferred by McCaffrey (1989).
Buru is currently situated between strands of the
Strong Fault system and Hamilton (1979) suggests
it may have moved westward from New Guinea.
The present author has allowed a small amount of
westward movement between 4 and 0 Ma which
fits Seram, Buru and the Bird's Head closer
together but this is merely a guess.
Caroline plate
The Ayu Trough opened during the middle and late
Miocene (Weissel & Anderson 1978) and spreading
may be continuing at the present day. There has
been no subduction at the Caroline-Philippine Sea
plate boundary and little convergence at the
Caroline-Pacific boundary c. 25 Ma. The model
uses the Caroline-Philippine Sea plate rotation pole
and rate of Sent et al. (1993) for the period 5-0 Ma
and then moves the Caroline plate with the east side
of Ayu Trough before this. The Caroline plate is
omitted from the reconstructions before 25 Ma.
Implications
Timing of major changes
The animated reconstructions show clearly that,
during the interval 50-0 Ma, although there are
important changes in movements and locally there
are significant manifestations of these changes,
there are two truly regionally important periods of
change. Both of these appear to be the expression of
arc-continent collision and resulted in major
changes in the configuration of the region and in
the character of plate boundaries. At c. 25 Ma the
collision of the Australian continent with the
Philippine Sea plate arc in New Guinea (Fig. 6)
caused major effects which propagated westwards
through the region. The Philippine Sea plate began
to rotate clockwise requiring development of new
subduction systems at its western edge. This led
to the assembly of the Philippines archipelago,
initiated new arc systems from north Sulawesi
through to the Philippines, and led to the growth
and partial destruction of marginal basins such as
the Sulu Sea. The continued rotation of the
Philippine Sea plate ultimately resulted in elimina-
tion of the Molucca Sea plate and accretion of
fragments from the northern Australian margin
into the SE Asian margin, notably in Sulawesi. At
c. 5 Ma collision of the Philippine arc and the
Eurasian continental margin occurred in Taiwan
(Fig. 2). This appears the key to the recent
tectonics of the region. Once again, the Philippine
Sea plate motion changed, and new subduction
systems were initiated, such as those currently
active on the east and west sides of the Philippines.
Most deformation now seems to be concentrated
in the region between the Banda Sea and Taiwan.
Summarized below are the major implications of
the reconstructions for different parts of SE Asia.
Borneo
Accepting rotation of Borneo has some important
consequences for the reconstruction of the region,
although the amount of rotation could be reduced
to about 30 ° without seriously affecting the model.
It means that there was initially a very wide proto-
South China Sea (Fig. 1 l) which began to close
from c. 44 Ma (Fig. 10). Alternative reconstruc-
tions (e.g. Rangin et al. 1990; Lee & Lawyer 1994)
which do not incorporate this counter-clockwise
rotation, appear to lack the space required for the
opening and closing of ocean basins such as that
north of the Cagayan ridge and the Sulu Sea.
Although the closure of the proto-South China Sea
may have been partly driven by extrusion of the
Indochina block the reconstructions suggest it was
achieved by subduction either at the southeast side
of the ocean or within the ocean (Figs 8 and 9).
Once subduction was established slab-pull force
would have caused extension of the Indochina-
South China margin leading to formation of
oceanic crust and extension of Eurasian continental
crust (Fig. 7). Thus, extension in this region
may not have been driven by strike-slip-related
extrusion. In fact, the model suggests that sub-
duction may have been underway before South
China Sea opening began. Opening of the Celebes
Sea required northward motion of Luzon because
of its position north of the spreading centre, thus
implying a subduction zone on the south side of
the proto-South China Sea (Fig. 9). The Oligocene
reconstructions are thus very similar to those of
Taylor & Hayes (1983) except that south Borneo
is rotated further.
Many of the differences between this model and
that of Rangin et al. (1990), which otherwise have
many resemblances, result from the new constraints
imposed by reconstruction of the Philippine Sea
plate. If Borneo is not rotated there are problems in
accounting for the evidence of continental crust in
west Sulawesi in the early Miocene (see below;
Bergman et al. 1996). The extended Bird's Head
 I "l f
EURASIAN ,b ~t b.

PLATE
25 Ma
L=eO,gocene
Openingof South
kChinaSeasouth
\ o f Macclesfield 20°N
~ ____1, \ Bank PACIFIC
PLATE

Extension ~ C ~
dueto slab-puff

10°N
Proto- Spreadingin
PareceVela
k basin

\ r

/
CAROLINE~ C
! PLATE

Ophiolite
collision
"'"~' .... Trapped
in.Sulawesi Indian ocean 10°S

Halmahera
withAustralia
? ?
Bird'sHead ~'
microcontinent

110OE 1120~E ~ ~ 1130°E~ 1 1 4 0 ° E ~ d5OOE 20oS

Fig, 6. Reconstruction of the region at 25 Ma. Collision of the Australian continental margin in New Guinea, and the Bird's Head microcontinental block in Sulawesi, with the arc
from north Sulawesi to Halmahera caused major reorganisation of plate boundaries. The active ridge jumped south in the South China Sea.
 |
I I f i

EURASIAN .4
PLATE ¸--,,3
30 Ma' 0
Mid Oligocene Openingof
Sin outh ChinaSea
movement on northof
Red R/vet fau/t Macclesfield
Bank 20°N_
/ / ~ Indochina PACIFIC
extruded PLATE
J ~ ~ /Pala
N°wa~
Dextral~\ Extensiondriven
motionot byslab-pufl
ree Pagodas ,andIndochina
system extrusion Proto-
South China Sea
10°N_.
Openingof
Parece Vela
Subduction basinbegins
decreases
westwards

0o

/
SulawesiwestArm
10°S_

INDIAN PLATE

??

Bird's Head
microcontinent
1100°E II10°E I1~ [150°E 20°S

Fig. 7. Reconstruction of the region at 30 Ma. South China Sea opening was driven by pull of the subducted Proto-South China Sea slab, and possibly by extrusion caused as India
indented Eurasia. Arc splitting began in the eastern margin of the Philippine Sea plate with spreading in the Parece Vela basin.
 RECONSTRUCTING CENOZOIC SE ASIA
microcontinent would have been much further
north than shown in the reconstructions and hence
would have occupied the site of the Molucca Sea.
For most of the Oligocene and early Miocene
(Figs 6 and 7) the Sundaland margin in Borneo
would have been orientated NW-SE and would
have been a strike-slip dominated region and there-
fore the sedimentary basins of northwest Borneo
should record an important component of this
strike-slip history. This configuration also suggests
that much of the sediment now found in north
Borneo may have been fed southwards across the
Sunda shelf from Indochina. This resolves a
problem of the sediments now found in north
Borneo which could not have a south Bomeo
provenance because of their volume and because
most of south Borneo was below sea-level during
much of this period. The timing of subduction
suggested by the model is similar to that deduced
from the north Borneo stratigraphic record by oil
company geologists (e.g. T a n & Lamy 1990) and
the present author suggests that their 'Deep
Regional Unconformity' is one manifestation of
Australia-Philippine Sea plate collision and conse-
quent reorganization of plate boundaries. Sub-
duction became increasingly important in NE
Borneo between c. 20 and 10 Ma (Figs 3-5) and the
proto-South China Sea was entirely eliminated
before 10 Ma.
For Malay blocks west of Borneo, the recon-
structions (Figs 4 and 5) can be regarded only as
an approximation. The model suggests greater
extension than that recorded in the Malay basins,
and the timing of the rotations in the model needs
to be improved. The similarities in rotations
recorded through this part of Sundaland may be
no more than coincidental, but if Bomeo and Java
have rotated, there must have been some movement
of Sumatra and the Malay peninsula, otherwise
there is major overlap of fragments. The implica-
tions of rotations of blocks for the wider region
are a problem which the palaeomagnetists often
neglect, and therefore the model has some value,
even if wrong, in attempting to face up to these
implications, which in turn emphasize the need for
more data to identify timing and to separate local
and regional rotations. There can be no doubt that
at present the palaeomagnetic data from Sundaland
do not on their own convincingly demonstrate
regional rotations.
Origin o f the Celebes Sea
If palaeomagnetic evidence for the rotation of
Borneo is accepted, there is a strong case to be
made that the central West Philippine Sea, the
Celebes Sea and the Makassar Strait formed part
of a single marginal basin (Figs 8 and 9) which
171
opened between late Eocene and mid Oligocene,
and widened eastwards like the present South
China Sea. This interpretation has been incorpor-
ated in the model and the palaeomagnetic data from
the Celebes Sea and Philippine Sea can i~e recon-
ciled in such a model. The author follows Hamilton
(1979) and many others in interpreting the
Makassar Strait as an extended area of crust, prob-
ably without oceanic crust, although this need not
preclude the Neogene convergent tectonics
suggested by Bergman et al. (1996) for its west
Sulawesi margin. The reconstructions are least
convincing for the period between 44 and 40 Ma
when the Philippine Sea plate was rotating rapidly
clockwise and the Celebes Sea was opening with
counter-clockwise rotation. However, this diffi-
culty may have more to do with the inadequacy of
the data on which the timings of rotations are based.
For the Philippine Sea plate all we can say is that
there was rapid rotation between 50 and 40 Ma
(Hall et al. 1995b) and more data from Palaeogene
rocks are needed to determine when and at what
rate the rotation occurred. It may well have been
fully complete by 45 Ma in which case there would
be no difficulty with smooth opening of a basin
narrowing west. For the Celebes Sea the position
of the basin is consistent with a backarc setting
related to northward subduction of Indian ocean
lithosphere beneath west and north Sulawesi.
Further east this setting appears less convincing
because of the very large distance between the
basin axis and the subduction zone (Figs 8 and 9).
There were at least three major basins which
opened in SE Asia with a similar east-widening
geometry: a Mesozoic proto-South China Sea, the
Eocene--Oligocene Philippine-Celebes basin and
the Oligo-Miocene South China Sea. The South
China Sea opening has been interpreted as linked to
India indentation (Tapponnier et al. 1982) and as
suggested here may be related to subduction, but
this explanation seems unlikely for the Philippine-
Celebes basin. Perhaps these three basins reflect
some other lithospheric mechanism related to the
long-term subduction of lithosphere east and south
of SE Asia.
Sulawesi collisions
The apparently simple tectonic configuration in
Sulawesi of arc-ophiolite-continent is not the
result of a single arc-continent collision (e.g. Silver
et al. 1983b) but is a consequence of multiple
collision events. The east Sulawesi ophiolite has
an Indian ocean origin (Mubroto et al. 1994).
Emplacement of the ophiolite on the west Sulawesi
continental margin occurred at the end of the
Oligocene (Parkinson 1991) and was followed by a
change in plate boundaries at the beginning of the
 I i / i I
• EURASIAN ~4
PLATE ....j
35 Ma
Early Oligocene

20°N_
PACIFIC
PLATE
Extensiondriven
bypullof
subducted
proto-SCSslab
Proto-
South China Sea 10 ° N
~, i ~ ~~l/estPhilippineSea
andCelebesSea
Trenchadvancewith~ ____I openingalmost
slabroll-backf . ~ ~ complete
Subducti°n'~7~.Z~'("
,:~/,~ ~ ~ , ~
dies o~
tow~st.,"X "£ 4-- 9
\ "/Z ...d, oo_

k Sea
/
Sulawesi
arc EastPhilippinesarc
10°S
~ f,,- P

~,,1~ INDIAN-AUSTRALIANPLATE l Halmahera

arc
? ?

I 90°E ~ i 100°E 140°E I150°E 20°S

Fig. 8. Reconstruction of the region at 35 Ma. Extension in the Sunda shelf and Eurasian continental margin was driven by pull of the subducted Proto-South China Sea slab.
Active spreading of the Celebes Sea-West Philippine Sea basin ended at 34 Ma.
 Eocene

~..,,.~.~.... PACIFIC
20°N__
PLATE
EURASIAN
PLATE

Izu
Proto-
10°N
South ChinaSea West Philippine Sea
spreading extends
to Celebes Sea
Subduction of

- / ~~,~~~/~'~,~1~ ~sea.
~'~~'le;es..~I'~~'
~lhA~
~,~
'~--"'~[~~~~tSea
" ~IL ~.a NorotatiOnphilippineseaPlateOf

Arc activity at south edge of

Philippine SeaPlate 10°S
.p

90°E 1100°E ]110°E 120°EL ~1111~,.--,.,,.,~ 1130°E 1140°E 1150°E 20°S

Fig. 9. Reconstruction of the region at 40 Ma. Subduction of the Proto-South China Sea began as a trench became active north of Zamboanga-Luzon, caused by rapid opening of
the Celebes Sea-West Philippine Sea. Between 40-50 Ma the Philippine Sea plate rotated clockwise about a pole at 10°N, 150°E.
t.,o
174 R. HALL
Miocene (Fig. 6). There is isotopic evidence from
geochemistry of igneous rocks for very old con-
tinental crust beneath west Sulawesi (Coffield et al.
1993; Priadi et al. 1993; Bergman et al. 1996),
probably of Australian origin, because of the
extreme isotopic compositions. The ages of the
oldest igneous rocks reported (Bergman et al. 1996)
imply early Miocene underthrusting of Australian
lithosphere. Since the early Miocene there have
been at least two further collisions, in SW and east
Sulawesi, as fragments of continental crust have
been sliced from the Bird's Head microcontinent
and transported west for brief periods on the
Philippine Sea plate or the Molucca Sea plate
which was partially coupled to the Philippine Sea
plate.
Bird's Head microcontinent
The reconstructions suggest that the Sula platform
and Tukang Besi platform formed part of a single
large microcontinent with the Bird's Head at
c. 15 Ma (Fig. 4). The model shows that movement
of small continental fragments to their present
positions can be explained easily if they were sliced
from this microcontinent at different times and each
moved with the Philippine Sea plate for a few
million years before collision (Figs 2-5). Thus, the
driving force for these motions was the Philippine
Sea plate. Temporary locking of the strike-slip
system at the southern edge of the Philippine Sea
plate required development of new splays of the
fault which resulted in the transfer of continental
fragments to the Philippine Sea plate. As each
fragment docked at the western end of the fault
system in Sulawesi, a new splay developed and a
new fragment began to move. The cartoons of
Hamilton (1979) for the Neogene development of
the region are remarkably similar to the predictions
of the model although the timing is somewhat
different, mainly because of new information from
east Indonesia. This model has been described
informally as a 'bacon slicer' and this does seem a
very appropriate description of the mechanics of
the process. The present author differs in one
important respect from many interpretations of this
region in suggesting that these fragments, although
colliders, are not the major causes of contractional
deformation associated with their docking. They
are merely passive participants riding on the
Philippine Sea plate whose major role is to lock a
fault strand on their arrival at the Sulawesi margin.
The reconstructions of Tukang Besi and Sula
with the Bird's Head do not account for the
evidence for Australian crust beneath west
Sulawesi by the late Early Miocene (Bergman et al.
1996). The extent of this continental crust was
estimated and used to outline an additional frag-
ment (Figs 5-9). If this is assumed to be beneath
west Sulawesi by 17 Ma but is moved with the
Philippine Sea plate from 21-17 Ma, it fits back to
the western edge of the Bird's Head microcontinent
(Fig. 6). It is suggested here that this fragment
records the earliest effect of the 'bacon slicer' and
was the first block detached from the Bird's Head
microcontinent. It separated as a result of the
development of the Sorong Fault system at the
southem edge of the Philippine Sea plate and, like
the other fragments, moved with the Philippine Sea
plate for about 5 Ma. On the basis of the 25 Ma
reconstruction, it can be speculated that the Bird's
Head microcontinent was a single block at the end
of the Oligocene formed by separation from
Australia during the Mesozoic, to which had been
added at least part of the east Sulawesi ophiolite.
The reconstructions do not answer the question
of the ultimate origin of the Bird's Head micro-
continent, but they do require that the micro-
continent has been moving with about the same
motion as, and has remained in a similar position
relative to, Australia for the past 25Ma.
Palaeomagnetic results suggest that the micro-
continent was at least 10°N of the north Australian
margin in the late Cretaceous (Wensinck et al.
1989; Ali & Hall 1995) but its early Tertiary
position is not well constrained in the model. The
pre-Neogene geology of the region suggests there
may have been several important events in the
development of the Banda Sea region (e.g. ophio-
lite emplacement and metamorphism described by
Sopaheluwakan (1990) from Timor), which could
record relative movement between Australia and
the microcontinent in passage to their 25 Ma
positions.
Molucca Sea
The model suggests that the Molucca Sea was a
very wide area formed by trapping of Indian ocean
lithosphere between the north Australian margin
and the Philippine Sea plate arc when collision
occurred at 25 Ma (Fig. 6). Thus, its age should be
pre-Tertiary. It was eliminated by subduction on its
east and west sides. Subduction probably began
soon after the 25 Ma change in motion of the
Philippine Sea plate on the west side of the
Molucca Sea in the north Sulawesi-Sangihe arc,
consistent with ages of arc volcanic rocks in north
Sulawesi (Dow 1976; Effendi 1976; Apandi 1977).
The Molucca Sea formed part of the Philippine Sea
plate up to c. 15 Ma (Figs 4 and 5) when east-
dipping subduction began beneath Halmahera
(Baker & Malaihollo 1996). Most of the subduction
occurred on the west side so the Molucca Sea has
remained partly coupled to the Philippine Sea plate.
The double subduction system of the Molucca Sea
 RECONSTRUCTING CENOZOIC SE ASIA
probably never extended north of its present
northern edge into the Philippines. However, there
was a oceanic area which was effectively con-
tinuous with, and north of, the Molucca Sea which
was formerly the central part of the Celebes Sea-
West Philippine central basin. In the reconstruc-
tions this northern part of the ocean is eliminated
between 25 and 12 Ma and the Panay sector of the
Philippine arc arrives at the western subduction
margin at about 12 Ma. Slightly earlier (15 Ma) the
Cagayan arc and Luzon arrive at the Palawan
margin (Fig. 4) resulting in collision in Mindoro
(Rangin et al. 1985). It may be one or both of these
events in the Philippines which caused initiation
of the east-dipping subduction system beneath
Halmahera and the concomitant development of
splays of the Sorong Fault system at the southern
edge of the Molucca Sea. The reconstructions of the
Philippines are too imperfect to be confident of the
relationships between cause and effect but the
timing of collision and age of the ophiolites in
Mindoro (Rangin et al. 1985) and events in Panay
(Rangin et al. 1991) are consistent with the model.
Philippines
The Philippines are difficult to reconstruct for
a number of reasons. Our understanding of
Philippines geology and evolution is still
insufficient, although considerable advances have
been made in recent years as a result of investi-
gations in the region (see for example, Sarewitz &
Karig 1986; Stephan et al. 1986; Rangin 1991; and
references therein). Reconstructions using the
outlines of present-day fragments can be only
approximate since much new crust has been
added by arc processes. In order to describe the
Philippines more precisely many more fragments
need to be used and this is currently beyond the
capacity of the ATLAS program. However, it seems
that rigid plate tectonics may be an inadequate tool
to describe the evolution of the area. At present this
is illustrated most clearly at the south end of the
archipelago where plate boundaries are ill-defined
probably because there is a significant amount of
within-plate deformation (Rangin et al. 1996).
Karig et al. (1986) have drawn attention to the way
in which at present only a small part of the
Philippines is completely coupled to the Philippine
Sea plate, and how in the past it is likely that
coupling was never complete across the entire
Philippine system because of strike-slip faulting.
Because of the situation of all the Philippine
fragments at the edge of plates it is probably unwise
to rely too heavily on the evidence of arc volcanism
to infer arc continuity. At present it is clear that
some volcanism is related to subduction on the west
side of the Philippines, some may be related to sub-
175
duction at the Philippine Trench, and some may not
be directly related to subduction. Switching of sub-
duction from one side of the Philippines to the other
has almost certainly occurred in the past (e.g.
Schweller etal. 1983; Karig e t a l . 1986). Some arcs
may also be eliminated completely during collision
as one arc overrides another, a process that is
underway in the Molucca Sea.
However, the reconstructions do provide some
useful limits on what is feasible. Most of the
Philippine islands probably formed part of an arc at
the southern edge of the Philippine Sea Plate before
the early Miocene '(Figs 7-10) as shown earlier in
the reconstructions of Rangin et al. (1990). The
choice here for the position of Luzon (Figs 6-10)
shows that the geological evidence for subduction
can be satisfied in ways other than by including all
the Philippines in this arc. Since the early Miocene
(Figs 2-6) the Philippine fragments have moved in
a very narrow zone, mainly as part of the Philippine
Sea plate, within which there appears to have been
a substantial component of strike-slip motion
(Sarewitz & Karig 1986). Most subduction under
the Philippines was oblique, mainly at the western
edge, and north of Mindanao. The Philippines are
an ephemeral feature. They will probably end up as
a composite arc terrane smeared onto the Eurasian
continental margin which will be impossible to
unravel. The great depths of young sedimentary
basins in close juxtaposition to areas of high emer-
gent topography suggest that lithospheric processes
operating in this region are not well described by
current plate tectonic concepts, and in some ways
the Philippines are reminiscent of Pre-Cambrian
greenstone belts. There seems no evidence that the
Philippines are the result of a collision of two
opposed arcs, progressively zipping up southwards
towards the present-day Molucca Sea as shown
in some reconstructions (e.g. Lewis et al. 1982;
Rammlmair 1993). In particular, there is no
evidence that the west-facing Halmahera arc
extended north of the present north edge of
the Molucca Sea into Mindanao (Quebral et al.
1995).
Banda Sea
The age and origin of the Banda Sea have long
been the subject of dispute. Several workers have
suggested it is relatively old, possibly Mesozoic,
with the Banda Sea representing trapped oceanic
crust (Katili 1975; Bowin et al. 1980; Lapouille
et al. 1986; Lee & McCabe 1986) whereas others
have preferred a much younger, late Tertiary, age
(Carter et al. 1976; Hamilton 1979; R6hault et al.
1995). Silver et al. (1985) suggested that the North
Banda Sea may include crust of Pacific origin
and identified the Banda ridges as continental
 k

o~
45 Ma
• Middle Eocene

\ 20°N
PACIFIC
PLATE
?
EURASIAN
PLATE
A

Subductionof IO°N
Proto- NNG-Pacific
South China Sea
Boninitemagmatism
,,~ in Izu-Boninand
Zambales~ •~ Mariana
ophioliteformation .= • \ forearcs

-, ~ ~ Lu~oo
PHILIPPINE
SEA .
PLATE
North Sulawesi ? ?
©
~,~1~ 4~mahera
INDIAN-AUSTRALIAN PLATE

90°E I 100°E I110°E 1120°E i130°E 1140°E I150°E 20°S

Fig. 10. Reconstruction of the region at 45 Ma. The North New Guinea-Pacific spreading centre was subducted causing forearc magmatism and massive extension of the
NE margin of the Philippine Sea plate.
 RECONSTRUCTING CENOZOIC SE ASIA
fragments based on dredging, a conclusion
supported by more recent dredging (Villeneuve
et al. 1994). Since some of the magnetic lineations
of supposed Mesozoic age cross the Banda ridges
the inferences based on magnetic anomaly ages
may be wrong. Many of the lineations are parallel
to the structural fabric of these continental areas
(Silver et al. 1985; Rthault et aI. 1991). Despite
this, there still appears to be a widespread accept-
ance that the Banda Sea is floored by old oceanic
crust (e.g. Hartono 1990).
This is not a conclusion supported by the
reconstructions in this paper. In these the Banda
volcanic arc is fixed to west Sulawesi during the
late Neogene and the arc-trench gap is assumed to
have maintained its present width. Timor and the
islands of the outer Banda arc are fixed to Australia
for the same period. The Birds' Head micro-
continent, including the island of Seram, has been
moved, as described above, to satisfy the con-
straints imposed by reconstruction of the Molucca
Sea and the distribution of the deep subducted slabs
inferred from seismicity. The reconstructions that
result show that before c. 10 Ma the gap between
Seram and Timor (Fig. 4) was filled by oceanic
crust, now subducted beneath the Banda arc, which
could indeed have been of Indian ocean origin and
Mesozoic age. Timor arrived at the trench at
c. 4-3 Ma consistent with geological data from
Timor (e.g. Carter et al. 1976; Audley-Charles
1986; Harris 1991). However, before c. 10 Ma the
relative distance between Timor and Seram is
maintained (Figs 4-6) suggesting there was no
subduction in the eastern Banda Sea area. The
reconstructions suggest subduction began at
c. 10 Ma (Fig. 3), resulting in the eastward
propagation of the volcanic arc consistent with the
very young age of the volcanoes in the inner Banda
arc (Abbot & Chamalaun 1981; McCaffrey 1989).
The north Banda basin extended as Seram moved
east and the arc propagated east. The eastward
movement of Seram is consistent with tectonic
inferences from present seismicity (McCaffrey
1989) and geological observations on Seram
(Linthout et al. 1991). The arc propagated east to
the longitude of Seram at c. 5 Ma (Fig. 2) which is
close to the age of the well known ambonites of this
area. The reconstructions also show the south
Banda basin extending rapidly between 5 and 0 Ma.
Therefore, both the north and south Banda Sea can
be interpreted as having an extensional origin and
to have opened during the late Neogene. These
interpretations are consistent with the age of young
volcanics dredged in the Banda Sea (R6hault et al.
1995). The great depth of the south Banda Sea is
still a problem, although it is known that Parsons
& Sclater's (1977) age-depth relationships often do
not hold in small ocean basins. Van Gool et al.
177
(1987) recorded high heat flows in several NW
Banda Sea basins and suggested that they are not
isostatically and thermally compensated.
Caroline plate and New Guinea
The consequences of fixing the Caroline plate to
the east side of Ayu Trough for Caroline-Pacific
boundary motion are that there has been predomi-
nantly left-lateral motion since 25 Ma, minor
oblique convergence between 15 and 5 Ma and
minor oblique extension between 5 and 0 Ma (Figs
2-6). The timing of convergence and extension are
partly dependent on the model since the period of
15-5 Ma was chosen as the interval of the main
opening of the Ayu Trough, based on Weissel &
Anderson (1978). However, these conclusions are
consistent with the evidence of young oblique
extension in the Sorol Trough (Weissel & Anderson
1978). The northern New Guinea arc terranes
have been omitted from the reconstructions for
simplicity and because there are insufficient data
to reconstruct them adequately. However, the
reconstructions do predict that there would have
been relatively little convergence between the
northern Australian margin and the Caroline plate
during the Neogene. Most of the convergence that
is required should have occurred in the past 5 Ma
(Fig. 2) and would have been oblique. At the
present day it appears that much of the con-
vergence is distributed (McCaffrey & Abers 1991;
Puntodewo et al. 1994) implying that only a pro-
portion need be absorbed by subduction. This is
consistent with the shallow seismicity associated
with the New Guinea trench and lack of active
volcanoes (Hamilton 1979; Cooper & Taylor
1987), and the absence of Neogene volcanicity in
Irian Jaya and the Bird's Head (Pieters et al. 1983;
Dow & Sukamto 1984). The reconstructions
require young (5-0 Ma) underthrusting on the
sector of the New Guinea trench east of the Bird's
Head as interpreted by Hamilton (1979) and
Cooper & Taylor (1987) but not north of the Bird's
Head, consistent with observations of Milsom et al.
(1992). The apparent complexity of the northern
New Guinea margin (cf. Pigram & Davies 1987)
appears to be due less to multiple collisions than
to strike-slip faulting and fragmentation of the arc
which collided with the Australian margin in the
early Miocene. No reconstruction has been
attempted here of the SW Pacific north and east
of New Guinea, and the reader is referred to Yan
& Kroenke (1993).
Final comments
The reconstructions presented here differ signifi-
cantly from earlier attempts to describe the
 I
Taiwan
SouthChina ...j
oo
50 Ma ~
End Early Eocene
(,,/ Mindom

'~/ ~ Indochina
North ..~

'\
Palawan/
t
EURASIAN
PLATE

Proto-
South China Sea
NORTH NEW GUINEA
PLATE

Zamboanga o

• West
Sulawesi

INDIAN-AUSTRALIAN PLATE

~ ~ Southand
ast Sulawesi
90°E 1100°E II10°E 1120°E 1130°E 1140°E 1150°E 20°S

Fig. 11. Reconstruction of the region at 50 Ma. The early rotation of the Philippine Sea plate began.
 RECONSTRUCTING CENOZOIC SE ASIA
development of SE Asia. Daly et "al. (1991) and
Lee & L a w v e r (1994) neglect the rotation of the
Philippine Sea plate and this accounts for major
differences in our interpretations of the eastern
parts of the region. Rangin et al. (1990) modelled
clockwise rotation of the Philippine Sea plate and
their reconstructions are quite similar to those here
for the Philippines region but the models differ on
the position of the Philippine Sea plate, the rotation
of Borneo and the position of Luzon. As noted
earlier, these differences are partly a consequence
of different interpretations of inadequate data.
Thus, even if these new reconstructions are
rejected they do at least serve to draw attention to
the need for many more good quality palaeo-
magnetic data from the region, in addition to
g e o l o g i c a l data, particularly on the timing,
chemistry and character of volcanic activity.
Biogeographical data could also be especially
useful in testing the different interpretations of the
region and there is a need for a joint approach from
botanists, zoologists and geologists. If the large
frame is used it also becomes clear how much one
is limited by the area, by the extent of possible
oceanic crust, etc.; these limitations are much less
obvious in local reconstructions where it is easier
to move problems outside the area of immediate
interest.
The reconstructions suggest that the indentation
of Eurasia by India has played a much less
important role in the development of SE Asia than
often assumed. They suggest that collision of the
Australian continent has driven major rotations,
and that the movement of smaller plates, such as the
Philippine Sea plate and the Borneo microplate, is
the result of the northward movement of Australia.
Events which are thus 'caused' by movements
of such plates may therefore be the consequence
of more fundamental regional movements. Thus,
caution is necessary in correlating events of similar
age and interpreting their causes. The difficulties of
reconstructing the Philippines are partly due to our
still inadequate knowledge of this region but may
References
ABBOTT,M. & CHAMALAUN,F. H. 1981. Geochronology
of some Banda Arc volcanics. In: BARBER,A. J. &
WmYosuJoNO, S. (eds) The Geology and Tectonics
of Eastern Indonesia. Geological Research and
Development Centre, Bandung, Indonesia, Special
Publication, 2, 253-268.
ALI, J. R. & HALL, R. 1995. Evolution of the boundary
between the Philippine Sea Plate and Australia:
palaeomagnetic evidence from eastern Indonesia.
Tectonophysics, 251, 251-275.
APANDI, T. 1977. Geologic map of the Kotamobagu
quadrangle, North Sulawesi. Geological Research
and Development Centre, Bandung, Indonesia.
179
also reflect our limited understanding of arc
processes, particularly at the lithospheric scale. The
intra-oceanic history of the region leaves a poor
record; arcs are ephemeral features on the geo-
logical timescale and may disappear completely,
a process currently underway in the Molucca Sea.
Probably the most difficult feature to incorporate
is the role of strike-slip faulting. Like Karig et al.
(1986) and Rangin et al. (1990) the present author
considers that strike-slip faulting has played a
major part in the development of the Philippines
and the reconstructions confirm that there is
limited space for the convergent motions often
inferred from the arc volcanic record. These
reconstructions also point to the importance of
strike-slip faulting in other parts of the region such
as Borneo and New Guinea, and suggest a need for
re-examination of data and interpretations based
on purely convergent collision models.
I am particularly grateful to Simon Baker who has pro-
vided me with invaluable support, not only with the work
involved in digitizing the fragment outlines used in the
reconstructions, but with discussion and ideas at many
stages. I also thank the numerous people with whom I
have discussed the geology of SE Asia over many years
who have contributed criticism, ideas and comments,
including D. A. Agustiyanto, J. R. Ali, C. D. Anderson,
M. G. Audley-Charles, P. D. Ballantyne, S. J. Baker, A. J.
Barber, J. C. Briden, T. R. Charlton, E. Forde, R. Garrard,
N. Haile, A. S. Hakim, S. Hidayat, T. W. C. Hilde,
M. Fuller, Kusnama, J. A. E Malaihollo, R. McCabe,
J. S. Milsom, A. H. G. Mitchell, S. J. Moss, R. Murphy,
G. J. Nichols, C. D. Parkinson, M. Pubellier, R. Quebral,
C. Rangin, H. Rickard, S. J. Roberts, N. Sikumbang, T. O.
Simandjuntak, M. J. de Smet, B. Taylor and S. L. Tobing.
Financial support has been provided at various stages by
NERC award GR3/7149, grants from the Royal Society,
the University of London SE Asia Research Group, and
the University of London Central Research Fund. Work
in Indonesia was facilitated by GRDC, Bandung and
Directors including H. M. S. Hartono, M. Untung,
R. Sukamto and I. Bahar. I thank R. J. Bailey, Gordon
Packham and Claude Rangin for reviews and many
helpful comments on the manuscript.
AUDLEY-CHARLES, M. G. 1986. Rates of Neogene and
Quaternary tectonic movements in the southern
Banda arc based on micropalaeontology. Journal of
the Geological Society, London, 143, 161-175.
- - - , CARTER, D. J. & MILSOM, J. S. 1972. Tectonic
development of Eastern Indonesia in relation to
Gondwanaland dispersal. Nature Physical Science,
239, 35-39.
AURELIO,M. A., BARRIER,E., RANGIN,C. & MULLER,C.
199 I. The Philippine Fault in the late Cenozoic tec-
tonic evolution of the Bondoc-Masbate-N. Leyte
area, central Philippines. Journal of SE Asian Earth
Sciences, 6, 221-238.
180 R. HALL
BAKER, S. & MALAIHOLLO, J. A. E 1996. Dating of
Neogene igneous rocks in the Halmahera region:
arc initiation and development. This volume.
BERGMAN, S. C., COFFIELD, D. Q., TALBOT, J. P. &
GARRARD, R. J. 1995. Tertiary tectonic and
magnetic evolution of western Sulawesi and the
Makassar Strait, Indonesia: evidence for a Miocene
continent-continent collision. This volume.
BESSE, J. & COURTrLLOT,V. 1991. Revised and synthetic
apparent polar wander paths of the African,
Eurasian, North American and Indian plates, and
true polar wander since 200Ma. Journal of
Geophysical Research, 96, 4029-4050.
BISHOP, W. E 1980. Structure, stratigraphy, and hydro-
carbons offshore southern Kalimantan, Indonesia.
AAPG Bulletin, 64, 37-58.
BowrN, C. O., PURDY,G. M., JOHNSTON,C., SHOR, G. G.,
LAWVER, L., er AL 1980. Arc-continent collision in
Banda Sea region. AAPG Bulletin, 64, 868-915.
BRIAIS, A., PATRIAT, P. 8z TAPPONNIER,P. 1993. Updated
interpretation of magnetic anomalies and sea-
floor spreading stages in the South China Sea:
implications for the Tertiary tectonics of Southeast
Asia. Journal of Geophysical Research, 98,
6299-6328.
CAMBRIDGE PALEOMAP SERVICES. 1993. ATLAS version
3.3. Cambridge Paleomap Services, CO. Box 246,
Cambridge, UK.
CANDE, S. C. 8z LESLIE, R. B. 1986. Late Cenozoic
tectonics of the southern Chile trench. Journal of
Geophysical Research, 91, 471-496.
CARDWELL,R. K. 8£ ISACKS, B. L. 1978. Geometry of the
subducted lithosphere beneath the Banda Sea in
Eastern Indonesia from seismicity and fault plane
solutions. Journal of Geophysical Research, 83,
2825-2838.
, - - & KARIG, D. E. 1980. The spatial distribution
of earthquakes, focal mechanism solutions and
subducted lithosphere in the Philippine and north-
eastern Indonesian islands. In: HAYES, D. E. (ed.)
The Tectonic and Geologic Evolution of South-east
Asian Seas and Islands. American Geophysical
Union Monograph, 23, 1-35.
CARTER, D . J . , AUDLEY-CHARLES,M. G. 8z BARBER,A. J.
1976. Stratigraphical analysis of island arc-
continental margin collision in Eastern Indonesia.
Journal of the Geological Society, London, 132,
179-198.
CHARLTON, Z. R. 1996. Correlation of Salawati and
Tomori basins, eastern Indonesia: a constraint on
left-lateral displacements of the Sorong Fault Zone.
This volume.
COCHRAN, J. R. 1981. The Gulf of Aden: structure and
evolution of a young ocean basin and continental
margin. Journal of Geophysical Research, 86,
263-287.
COFFIELD, D. Q., BERGMAN, S. C., GARRARD, R. A.,
GURITNO, N., ROBINSON,N. M. & TALBOT, J. 1993.
Tectonic and stratigraphic evolution of the Kalosi
PSC area and associated development of a Tertiary
petroleum system, south Sulawesi, Indonesia. In:
Proceedings Indonesian Petroleum Association
22nd Annual Convention 1993. 678-706.
COOPER, P. & TAYLOR, B. 1987. Seismotectonics of New
Guinea: a model for arc reversal following arc-
continent collision. Tectonics, 6, 53-67.
CURRAY, J. R., MOORE, D. G., LAWVER, L. A., EMMEL, F.
J., RArrT, R. W., e~r AL. 1979. Tectonics of the
Andaman Sea and Burma. In: WATKINS, J.,
MONTADERT, L. & DICKENSON, P. W. (eds)
Geological and Geophysical Investigations of
Continental Margins. AAPG, Memoir, 29, 189-198.
DALy, M. C., COOPER,M. A., WILSON, I., SMITH, D. G. &
HOOPLa, B. G. D. 1991. Cenozoic plate tectonics
and basin evolution in Indonesia. Marine and
Petroleum Geology, 8, 2-21.
DAVIDSON, J. W. 1991. The geology and prospectivity
of Buton island, S.C. Sulawesi, Indonesia. In:
Proceedings Indonesian Petroleum Association
20th Annual Convention 1991, 209-233.
DAVIES, I. C. 1990. Geological and exploration review of
the Tomori PSC, eastern Indonesia. In: Proceedings
Indonesian Petroleum Association 19th Annual
Convention 1990, 41-67.
Dow, D. B. & SUKAMTO, R. 1984. Western Irian Jaya:
the end-product of oblique plate convergence in the
late Tertiary. Tectonophysics, 106, 109-139.
Dow, J. 1976. Porphyry copper exploration in Block II,
Sulawesi. Geological Research and Development
Centre, Bandung Indonesia, Newsletter, 8 (33), 6.
Dt)RBAUM, H.-J. & HINZ, K. 1982. SEATAR-related
geophysical studies by BGR in the southwest
Pacific. In: Transactions of the Third Circum-
Pacific Energy and Mineral Resources Conference.
129-133.
EFFENDI, A. C. 1976. Geologic map of the Manado
quadrangle, North Sulawesi. Geological Research
and Development Centre, Bandung, Indonesia.
FISHER, R. L. & SCLATER,J. G. 1983. Tectonic evolution
of the southwest Indian ocean since the mid-
Cretaceous: plate motions and stability of the
pole of Antarctica/Africa for at least 80 Myr.
Geophysical Journal of the Royal Astronomical
Society, 73, 553-576.
FULLER, M., HASTON, R., LIN, J-L., RICHTEr, B.,
SCHMIDTKE, E. & ALMASCO, J. 1991. Tertiary
paleomagnetism of regions around the South China
Sea. Journal of SE Asian Earth Sciences, 6,
161-184.
--, MCCABE, R. WILLIAMS,I. S., ALMASCO,J., ENCINA,
R. Y., ZANORIA, A. S. & WOLFE, J. A. 1983.
Paleomagnetism of Luzon. In: HAVES, D. E. (ed.)
The Tectonic and Geologic Evolution of South-East
Asian Seas and Islands, Part 2. American
Geophysical Union Monograph, 27, 79-94.
GIDDINGS, J. W., SUNATA, W. & PIGRAM, C. 1993.
Reinterpretation of palaeomagnetic results from
the Bird's Head, Irian Jaya: new constraints on the
drift history of the Kemum Terrane. Exploration
Geophysics, 24, 283-290.
GINGER, D. C., ARDJAKUSUMAH,W. O., HEDLEY,R. J. &
POTHECARY, J. 1993. Inversion history of the West
Natuna basin: examples from the Cumi-Cumi PSC.
In: Proceedings of the Indonesian Petroleum
Association 22nd Annual Convention. 635-658.
GIRARDEAU,J., MONNIER,CH., MAURY, R., VILLENEUVE,
M., SOETISMA,D. • SAMODRA,H. 1995. Origin of
the east Sulawesi ophiolite. In: Abstracts, Eighth
 RECONSTRUCTING CENOZOIC SE ASIA
Regional Conference on Geology, Minerals and
Energy Resources of SE Asia (GEOSEA 95). 51-52.
GORDON, R. G. & JURDY, D. M. 1986. Cenozoic plate
motions. Journal of Geophysical Research, 91,
12389-12406.
Hamz, N. S. 1978. Reconnaissance palaeomagnetic
results from Sulawesi, Indonesia, and their bearing
on palaeogeographic reconstruction. Tectono-
physics, 46, 77-85.
--. 1979. Rotation of Borneo microplate complete by
Miocene: palaeomagnetic evidence. Warta Geologi,
Geological Society of Malaysia Newsletter, 5
19-22.
• & BRIDEN, J. C. 1982. Past and future palaeo-
magnetic research and the tectonic history of East
and Southeast Asia. In: Palaeomagnetic research in
Southeast and East Asia. UN/ESCAP, CCOP
Technical Publication Bangkok, 13, 25-46.
., MCELHINNY, M. W. & McDOUGALL, I. 1977.
Palaeomagnetic data and radiometric ages from the
Cretaceous of west Kalimantan (Borneo) and their
significance in interpreting regional structure.
Journal of the Geological Society of London, 133,
133-144.
, BECKINSALE,R. D., CHAKRABORTY,K. R., HUSSEIN,
A. H. & HARDJONO, T. 1983. Palaeomagnetism,
geochronology and petrology of the dolerite dykes
and basaltic flows from Kuantan, west Malaysia•
Bulletin of the Geological Society of Malaysia, 16,
71-85.
HALL, R. & NICHOLS, G. J. 1990. Terrane amalgamation
at the boundary of the Philippine Sea Plate.
Tectonophysics, 181, 207-222•
, ALI, J. R. ANDERSON,C. D. & BAKER, S. J. 1995a.
Origin and motion history of the Philippine Sea
Plate~ Tectonophysics, 251, 229-250.
, FULLER, M., ALI, J. R. & ANDERSON,C. D. 1995b.
The Philippine Sea Plate: Magnetism and
Reconstructions. In: TAYLOR, B. & NATLAND, J. H.
(eds) Active Margins and Marginal Basins: A
Synthesis of Western Pacific Drilling Results.
American Geophysical Union Monograph, 88,
371-404.
HAMILTON, W. 1979. Tectonics of the Indonesian region.
US Geological Survey Professional Paper, 1078.
HARLAND, W. B., ARMSTRONG,R. L., Cox, A. V., CRAIG,
L. E., Smith, A. G. & Smith, D. G. 1990.
A geologic time scale 1989. Cambridge University
Press.
HARRIS, R. A. 1991. Temporal distribution of strain in
the active Banda orogen: a reconciliation of rival
hypotheses. Journal of SE Asian Earth Sciences, 6,
373-386.
HARTONO, H. M. S. 1990. Late Cenozoic tectonic
development of the Southeast Asian continental
margin in the Banda Sea area. Tectonophysics, 181,
267-276.
HILDE, T. W. C. & LEE, C-S. 1984. Origin and evolution
of the West Philippine Basin: a new interpretation.
Tectonophysics, 102, 85-104.
HINZ, K., BLOCK,M., KUDRASS,H. R. 8z MEYER,H. 1991.
Structural elements of the Sulu Sea. Geologische
Jahrbuch, A127, 883-506.
HOLLOWAY, N. n. 1982. The stratigraphic and tectonic
181
evolution of Reed Bank, North Palawan and
Mindoro to the Asian mainland and its significance
in the evolution of the South China Sea. AAPG
Bulletin, 66, 1357-1383.
HUTCHtSON, C. S. 1989. Geological Evolution of South-
East Asia. Oxford University Press. Oxford
Monographs on Geology and Geophysics, 13.
KARIG, D. E. 1983. Accreted terranes in the northern
part of the Philippine archipelago. Tectonics, 2,
211-236.
--, SAREwrrz, D. R. & H~a~CK, G. D. 1986. Role of
strike-slip faulting in the evolution of allochthonous
terranes in the Philippines. Geology, 14, 852-855.
KaTILZ, J. A. 1975. Volcanism and plate tectonics in
the Indonesian island arcs. Tectonophysics, 26,
165-188.
1978. Past and present geotectonic position of
Sulawesi, Indonesia. Tectonophysics, 45, 289-322•
KLITGORD, K. D. & SCHOUTEN,H. 1986. Plate kinematics
of the central Atlantic. In: VOGT,P. R. • TUCHOLKE,
B. E. (eds) The Geology of North America. Volume
M. Geological Society of America, 351-378.
KT3NDIG, E. 1956. Geology and ophiolite problems
of east Celebes. Verhandelingen van het Konin-
klijk Nedertandsch Geologisch-Mijnbouwkundig
Genootschap, Geologische Serie, 16, 210-235
LAPOUILLE, A., HARTONO, H. M. S., LARUE, M.,
PRAMUNIJOYO, S. & LARDY, M. 1986. Age and
origin of the seafloor of the Banda Sea (Eastern
Indonesia). Oceanologica Acta, 81, 379-389.
LE PICHON, X. & FRANCHETEAU,J. 1978. A plate tectonic
analysis of the Red Sea-Gulf of Aden area.
Tectonophysics, 46, 369-406.
LEE, C. S. & MCC~E, [t. 1986. The Banda-Celebes-Sulu
Basin: a trapped piece of Cretaceous-Eocene
oceanic crust? Nature, 322, 51-54.
LEE, T-Y. & LAWVER,L. A. 1994. Cenozoic plate tectonic
reconstruction of the South China Sea region.
Tectonophysics, 235, 149-180.
LEWIS, S. D., HAYES, D. E. & MROZOWSKI, C. L. 1982.
The origin of the West Philippine Basin by inter-arc
spreading. In: BaLCE, G. R. & ZANORIA, A. S. (eds)
Geology and Tectonics of the Luzon-Marianas
Region. Philippines SEATAR Committee, Special
Publication, 1, 31-51.
LrNTHOUT, K., HELMERS, H. & ANDRIESSEN, P. A. M.
1991. Dextral strike-slip in central Seram and 3-
4.5 Ma Rb/Sr ages in pre-Triassic metamorphics
related to early Pliocene counter-clockwise rotation
of the Buru-Seram microplate (E. Indonesia).
Journal of SE Asian Earth Sciences, 6, 335-342.
LIVERMORE, R. A., VINE, E J. & SMITH,A. G. 1984. Plate
motions and the geomagnetic field, II, Jurassic
to Tertiary. Geophysical Journal of the Royal
Astronomical Society, 79, 939-961.
LUMADYO, E., MCCABE, R., HARDER, S. & LEE, T. 1993.
Borneo: a stable portion of the Eurasian margin
since the Eocene. Journal of SE Asian Earth
Sciences, 8, 225-231.
McC~a3E, R. & COLE, J. 1989. Speculations on the late
Mesozoic and Cenozoic evolution of the Southeast
Asian margin. In: BEN-AVRAHAM, Z. (ed.) The
Evolution of the Pacific Ocean Margins. Oxford
University Press.
 182 R. HALL
MCCAFFREY, R. 1989. Seismological constraints and
speculations on Banda arc tectonics. Netherlands
Journal of Sea Research, 24, 141-152.
1996. Slip partitioning at convergent plate
boundaries of SE Asia. This volume.
- - & ABERS, G. A. 1991. Orogeny in arc-continent
collision: the Banda arc and western New Guinea.
Geology, 19, 563-566.
MCELHINNY, M. W., HALLE, N. S. & CRAWFORD, A. S.
1974. Palaeomagnetic evidence shows Malay
Peninsula was not part of Gondwana. Nature, 252,
641-645.
MCKENZrE, D. P., DAVIES, D. & MOLNAR, P. 1970. Plate
tectonics of the Red Sea and east Africa. Nature,
226, 243-248.
MARCHADIER,Y. & RANGIN, C. 1990. Polyphase tectonics
at the southern tip of the Manila trench, Mindoro-
Tablas islands, Philippines. Tectonophysics, 183,
273-287.
MILSOM, J., MASSON, D., NICHOLS, G., SIKUMBANG, N.,
DWIYANTO, B., ET AL. 1992. The Manokwari trough
and the western end of the New Guinea trench.
Tectonics, 11, 145-153.
MUBROTO, B. 1988. A palaeomagnetic study of the East
and Southwest Arms of Sulawesi, Indonesia. DPhil.
Thesis, University of Oxford.
, BRIDEN, J. C., MCCLELLAND, E. & HALL, R. 1994.
Palaeomagnetism of the Balantak ophiolite,
Sulawesi. Earth and Planetary Science Letters, 125,
193-209.
NGAH, K, MAZLAN, M. & TJIA, H. D. 1996. Role of pre-
Tertiary fractures in formation and development
of the Malay and Penyu basins. This volume.
NICHOLS, G. J. & HALL, R. 1995. Stratigraphic and
sedimentological constraints on the origin and
tectonic history of the Celebes Sea basin. In:
Tectonics of SE Asia Conference abstracts.
Geological Society, London 1994, 40.
OTOFUJI, Y., SASAJIMA,S., NISHIMURA,S., DHARMA, A. &
HEHUWAT, F. 1981. Palaeomagnetic evidence for
clockwise rotation of the northern arm of Sulawesi,
Indonesia. Earth and Planetary Science Letters, 54,
272-280.
PARKINSON, C. D. 1991. The petrology, structure and
geologic history of the metamorphic rocks of central
Sulawesi, Indonesia. PhD Thesis, University of
London.
PARSONS, B. & SCLATER, J. G. 1977. An analysis of the
variation of ocean floor bathymetry and heat flow
with age. Journal of Geophysical Research, 82,
803-827.
PIETERS, P. E., PIGRAM, C. J., TRAIL, D. S., Dow, D. B.,
RATMAN, N. & SUKAMTO,R. 1983. The stratigraphy
of western Irian Jaya. Geological Research and
Development Centre Bulletin, 8, 14--48.
PIGRAM, C. J. & DAVIES, H. L. 1987. Terranes and the
accretion history of the New Guinea orogen. BMR
Journal of Australian Geology and Geophysics, 10
193-211.
--, SURONO & SUPANDJONO,J. B. 1985. Origin of the
Sula Platform, eastern Indonesia. Geology, 13,
246-248.
PRIADI, B., POLVI~, M., MAURY, R. M., BELLON, H.,
SOERIA-ATMADJA, R., E T AL. 1993. Tertiary and
Quaternary magmatism in central Sulawesi: chrono-
logical and petrological constraints. Journal of SE
Asian Earth Sciences, 9, 81-93.
PUBELLIER, M., QUEBRAL, R. D., RANGIN, C.,
DEFFONTAINES, B., MULLER, C., E T AL. 1991. The
Mindanao collision zone: a soft collision event
within a continuous Neogene strike-slip setting.
Journal of SE Asian Earth Sciences, 6, 239-248.
PUNTODEWO, S. S. O., MCCAFFREY, R., CALAIS, E., BOCK,
Y., RAIS, J., ET AL. 1994. GPS measurements of
crustal deformation within the Pacific-Australia
plate boundary zone in Irian Jaya, Indonesia.
Tectonophysics, 237, 141-153.
QUEBRAL, R. D., PUBELLIER,M. & RANGIN, C. 1995. The
onset of movement on the Philippine Fault in
eastern Mindanao: a transition from collision to
strike-slip environment. Tectonics, in press.
RAMMLMAIR, D. 1993. The evolution of the Philippine
archipelago in time and space: a plate tectonic
model. Geologisches Jahrbuch, B81, 3-48.
RANGIN, C. 1991. The Philippine Mobile Belt: a complex
plate boundary. Journal of SE Asian Earth Sciences,
6, 307-318.
- - & SILVER, E. A. 1991. Neogene tectonic evolution
of the Celebes-Sulu basins: new insights from
Leg 124 drilling. In: SILVER, E. A., RANGIN,
C., VON BREYMANN,M. T., er AL. Proceedings of the
Ocean Drilling Program, Scientific Results, 124,
51-63.
., JOLIVET,L. & PUBELLIER,M. 1990. A simple model
for the tectonic evolution of southeast Asia and
Indonesia region for the past 43 m.y. Bulletin de la
Socidtd gdologique de France, 8 VI, 889-905.
--., STEPHAN, J. F. & MI3LLER, C. 1985. Middle
Oligocene oceanic crust of the South China S e a
jammed in the Mindoro collision zone (Philippines).
Geology, 13, 425-428.
--, BUTTERLIN, J., BELLON, H., MI3LLER, C.,
CHOROW1CZ, J. & BALADAD, D. 1991. Collision
n6og~ne d'arcs volcaniques dans le centre des
Philippines: stratigraphie et structure de la chMne
d'Antique 01e de Panay). Bulletin de la Socidtd
g~ologique de France, 162, 465-477.
., DAHRIN, D., QUEBRAL, R. & THE MODEC
SCIENTIFIC PARTY. 1996. Collision and strike slip
faulting in the northern Molucca Sea, (Philippines
and Indonesia): preliminary results of a morpho-
tectonic study. This volume.
RANNEFT, T. S. M., HOPKINS, R. M., FROELICH, A. J. &
GWlNN, J. W. 1960. Reconnaissance geology and
oil possibilities of Mindanao. AAPG BuUetin, 44,
529-568.
RI~HAULT, J-P., MALOD, J. A., LARUE, M. BURHANUDDIN,
S. & SARMILI, t . 1991. A new sketch of the central
north Banda Sea, eastern Indonesia. Journal of SE
Asian Earth Sciences, 6, 329-334.
--, VILLENEUVE, M., HONTHAAS, C., BELLON, H.,
MALOD, J. A., er AL 1995. New geological
samplings along a transect across the southeast
Banda Sea basin (Indonesia). In: Tectonics of SE
Asia Conference abstracts. Geological Society,
London 1994, 49.
ROYER, J-Y. & SANDWELL,D. T. 1989. Evolution of the
eastern Indian Ocean since the Late Cretaceous:
 RECONSTRUCTING CENOZOIC SE ASIA
constraints from Geosat altimetry. Journal of
Geophysical Research, 94, 13,755-13,782.
RUTLAND, R. W. R. 1968. A tectonic study of part of
the Philippine Fault Zone. Quarterly Journal of
the Geological Society of London, 123, 293-325.
SAREWrrz, D. R. & KARIC, D. E. 1986. Geologic evolution
of western Mindoro island and the Mindoro suture
zone, Philippines. Journal of SE Asian Earth
Sciences, 1, 117-141.
SASAJ1MA, S., NISHIMURA,S., HIROOKA, K., OTOFUJI, Y.,
VAN LEEUVEN, T. & HEHUWAT, E 1980. Palaeo-
magnetic studies combined with fission track
datings on the western arc of Sulawesi, east
Indonesia. Tectonophysics, 64, 163-172.
SCHMIDTKE, E. A., FULLER, M. D. & HASTON, R. 1990.
Paleomagnetic data from Sarawak, Malaysian
Borneo, and the late Mesozoic and Cenozoic
tectonics of Sundaland. Tectonics, 9, 123-140.
SCHWELLER,W. J., KARIG, D. E. & BACHMAN,S. B. 1983.
Original setting and emplacement history of the
Zambales ophiolite, Luzon, Philippines, from
stratigraphic evidence. In: HAVES, D. E. (ed.) The
Tectonic and Geologic Evolution of South-East
Asian Seas and Islands, Part 2. American
Geophysical Union Monograph, 27, 124-138.
SCLATER,J. G., FISHER,R. L., PATRIAT,P., TAPSCOTT,C. &
PARSONS, B. 1981. Eocene to recent development
of the southwest Indian ridge, a consequence of the
evolution of the Indian ocean triple junction.
Geophysical Journal of the Royal Astronomical
Society, 64, 87-604.
S~GOUFIN, J. & PATRIAT,P. 1980. Existence d'anomalies
mrsozoiques dans le bassin de Somali. Implications
pour les relations Afrique-Antarctique-
Madagascar. Comptes Rendus de l'Acaddmie
Sciences, France, 291B, 85-88.
SENO, T., STEIN, S. A. & GR1PP, A. E. 1993. A model for
the motion of the Philippine Sea plate consistent
with NUVEL-1 and geological data. Journal of
Geophysical Research, 98, 17,941-17,948.
SHIBUYA,H., D. L./VIERRIL, V. HSU & LEG 124 SHIPBOARD
SCIENTIFIC PARTY. 1991. Paleogene counterclock-
wise rotation of the Celebes Sea - orientation of
ODP cores utilizing the secondary magnetisation.
In: SILVER, E. A., RANG1N, C., VON BREYMANN,
M. T., er AL. Proceedings of the Ocean Drilling
Program, Scientific Results, 124, 519-523.
SILVER, E. A. & RANGIN, C. 1991. Leg 124 tectonic
synthesis. In: SILVER, E. A., RANGIN, C., VON
BREYMANN,M. T., ET AL. Proceedings of the Ocean
Drilling Program, Scientific Results, 124, 3-9.
& SMITH, R. B. 1983. Comparison of terrane
accretion in modern Southeast Asia and the
Mesozoic North American cordillera. Geology, 11
198-202.
, MCCAFFREY, R. & SMITH, R. B. 1983a. Collision,
rotation and the initiation of subduction in the
evolution of Sulawesi, Indonesia. Journal of
Geophysical Research, 88, 9407-9418.
--, JOYODrWlRYO, Y. & STEVENS, S. 1983b.
Ophiolite emplacement by collision between the
Sula platform and the Sulawesi island arc,
Indonesia. Journal of Geophysical Research, 88,
9419-9435.
183
., GILL, J. B., SCHWARTZ, D., PRASETYO, H. &
DUNCAN, R. A. 1985. Evidence for a submerged and
displaced continental borderland, north Banda Sea,
Indonesia. Geology, 13, 687-691.
SIMANDJUNTAK,T. O. 1986. Sedimentology and tectonics
of the collision complex in the East Arm of Sulawesi,
Indonesia. PhD Thesis, University of London.
- - 1992. New data on the age of the ophiolite in
Eastern Sulawesi. Bulletin Geological Research
and Development Centre, Bandung, Indonesia, 15,
38--44.
SITUMORANG,B. 1982. Formation, evolution, and hydro-
carbon prospect of the Makassar basin, Indonesia.
In: Transactions of the Third Circum-Pacific Energy
and Mineral Resources Conference, 227-231.
- - 1987. Seismic stratigraphy of the Makassar basin.
Scientific Contribution on Petroleum Science and
Technology Lemigas Sumbangan Ilmiah, 1/87,
3-38.
SMITH, R. B. & SILVER,E. A. 1991. Geology of a Miocene
collision complex, Buton, eastern Indonesia.
Geological Society of America Bulletin, 103, 660-
678.
SOPAHELUWAKAN, J. 1990. Ophlolite Obduction in the
Mutis Complex, Timor, Eastern Indonesia: an
Example of Inverted, Isobaric, Medium-High
Pressure Metamorphism. Vrije University Press,
Amsterdam.
SRWASTAVA, S. P. & TAPSCOTT, C. R. 1986. Plate kine-
matics of the north Atlantic. In: VOGT, P. R. &
TUCHOLKE, B. E. (eds) The Geology of North
America. Volume M. Geological Society of
America, 379-404.
STEPHAN, J-E, BLANCHET,R., RANGIN, C., PELLETIER,B.,
LETOUZEY, J. & MULLER, C. 1986. Geodynamic
evolution of the Taiwan-Luzon-Mindoro belt since
the Late Eocene. Tectonophysics, 125, 245-268.
STERN, R. J. & BLOOMER, S. H. 1992. Subduction zone
infancy: examples from the Eocene Izu-Bonin-
Mariana and Jurassic California arcs. Bulletin of the
Geological Society of America, 104, 1621-1636.
SURMONT,J., LAJ, C., KISSEL,K., RANGIN,C., BELLON,H.
& PPdADt, B. 1994. New palaeomagnetic constraints
on the Cenozoic tectonic evolution of the north
Arm of Sulawesi, Indonesia. Earth and Planetary
Science Letters, 121, 629-638.
TAN, D. N. K. & LAMY, J. M. 1990. Tectonic evolution of
the NW Sabah continental margin since the Late
Eocene. Bulletin of the Geological Society of
Malaysia, 27, 241-260.
TAPPONNIER, P., PELTZER, G., LEDAIN, A., ARMIJO, R. &
COBBOLD, P. 1982. Propagating extrusion tectonics
in Asia: new insights from simple experiments with
plasticine. Geology, 10, 611-616.
TAYLOR, B. 1992. Rifting and the volcano-tectonic
evolution of the Izu-Bonin-Mariana arc. In:
TAYLOR, B. & FUIJIOKA, K., ET AL. Proceedings of
the Ocean Drilling Program, Scientific Results, 126,
627-651.
-& HAYES, D. E. 1980. The tectonic evolution of the
-
South China Basin. In: HAYES, D. E. (ed.) The
Tectonic and Geologic Evolution of South-east
Asian Seas and Islands. American Geophysical
Union Monograph, 23, 89-104.
184 R. HALL
& 1983. Origin and history of the South China
Basin. In: HAYES, D. E. (ed.) The Tectonic and
Geologic Evolution of South-east Asian Seas and
Islands, Part 2. American Geophysical Union
Monograph, 27, 23-56.
TJtA, H. D. & Lmw, K. K. 1996. Changes in tectonic stress
field in northern Sunda shelf basins. This volume.
VAN OER WEERD, A. A. & ARMIN, R. A. 1992. Origin and
evolution of the Tertiary hydrocarbon-bearing
basins in Kalimantan (Borneo), Indonesia. AAPG
Bulletin, 76, 1778-1803.
VAN DER Woo, R. 1993. Paleomagnetism of the Atlantic,
Tethys and lapetus Oceans. Cambridge University
Press.
VAN GOOL, M., HUSON, W. J., PRAWlRASASRA, R. &
OWEN, T. R. 1987. Heat flow and seismic
observations in the northwestern Banda arc. Journal
of Geophysical Research, 92, 2581-2586.
VILLENEUVE, M., CORNICE,J. J., MARTINI, R., ZANINETTI,
L., RI~HAULT, J-P., ET AL. 1994. Upper Triassic
shallow water limestones in the Sinta ridge (Banda
Sea, Indonesia). Geo-Marine Letters, 14, 29-35.
VISSER, W. A. & HERMES, J. J. 1962. Geological results
of the exploration for oil in Netherlands New
Guinea. Koniklijk Nederlands Mijnbouwkundig
Genootschaft Verhandlingen, Geologische Serie,
20.
WAHYONO, H. & SONATA, W. 1987. Palaeomagnetism
along transect VII. In: Joint CCOP/IOC Working
Group on SEATAR 1987: Preliminary Report of
the Jawa-Kalimantan Transect (Transect VII).
Geological Research and Development Centre,
Bandung, Indonesia, 1-10.
WEISSEL, J. K. 1980. Evidence for Eocene oceanic crust
in the Celebes Basin. In: HAYES, D. E. (ed.) The
Tectonic and Geologic Evolution of South-east
Asian Seas and Islands. American Geophysical
Union Monograph, 23, 37-48.
& ANDERSON, R. N. 1978. Is there a Caroline
Plate? Earth and Planetary Science Letters, 41,
143-158.
WENSINK, H., HARTOSUKOHARDJO, S. ~; SURYANA, Y.
1989. Palaeomagnetism of Cretaceous sediments
from Misool, northeastern Indonesia. Netherlands
Journal of Sea Research, 24, 287-301.
YAN, C. Y. & KgOENKE, L. W. 1993. A plate tectonic
reconstruction of the SW Pacific 0-100 Ma. In:
BERGER, T., KROENKE, L.W., MAYER, L., ET AL.
Proceedings of the Ocean Drilling Program,
Scient~c Results, 130, 697-709.
 Contrasting tectonic styles in the Neogene
orogenic belts of Indonesia
T. O. S I M A N D J U N T A K 1 & A. J. B A R B E R 2
1 Geological Research and Development Centre, Jalan Diponegoro 57,
Bandung 40122, Indonesia
2 Southeast Asia Research Group, Department of Geology, Royal Holloway,
University of London, Egham TW20 OEX, UK

Abstract: The recent compilation of a new tectonic map of Indonesia as part of the Geotectonic
Map Project of East Asia has prompted a reassessment of the contrasted tectonic styles repre-
sented by currently developing orogenic belts. These orogenic styles provide a range of models
illustrating the diversity and complexity of tectonic processes which may provide the key to the
interpretation of other orogenic belts elsewhere in the world. The following distinctive types of
orogenies have been recognized within Indonesia. 1. Sunda Orogeny in Java and Nusa Tenggara:
involving subduction of oceanic crust with normal convergence, producing an orogenic belt of
Andean type with trench, accretionary complex, forearc basin and Quaternary magmatic arc with
active volcanoes built on the margin of the Sundaland continent. 2. Barisan Orogeny in Sumatra:
with strongly oblique convergence and major strike slip transcurrent fault movement within the
magmatic arc, along which a segment of continental crust is being displaced northwards along
the western margin of Sundaland. 3. Talaud Orogeny in the northern Molucca Sea: convergence
of the Sangihe and Halmahera oceanic magmatic arcs as the Molucca Sea Plate subsides beneath
them. 4. Sulawesi Orogeny in eastern Sulawesi: collision of microcontinental blocks with sub-
duction systems along the eastern margin of Sundaland. 5. Banda Orogeny in the southern Banda
Arc between Sumba and Tanimbar: collision of the northern margin of the Australian continent
with the subduction system along the southern segment of the Banda Arc. 6. Melanesian Orogeny
in Irian Jaya and Papua New Guinea: a more advanced stage in the collision of the northern
margin of the Australian continent with a magmatic arc on the Philippine Sea plate, commencing
in the early Miocene with a partial reversal of polarity and the subduction of the Caroline plate
beneath the collision zone.
The present phase of orogenic activity in most of these occurrences commenced in mid-
Miocene times and orogenic processes are still in progress.

Since the commencement of the first five year
National Development Plan of Indonesia (Pelita I)
in 1970 the geology of the whole country has been
systematically and intensively investigated both by
government institutions and by the private sector.
These investigations have included the systematic
geological mapping of the Indonesia archipelago,
initially by the Geological Survey of Indonesia and
later by the Geological Research and Development
Centre (GRDC), partly in collaboration with the
United States Geological Survey (USGS), the
British Geological Survey (BGS), the Australian
Bureau of Mineral Resources (BMR) (now the
Australian Geological Survey Organisation:
AGSO) and the Japanese International Cooperation
Agency (JICA). At the same time geological and
geophysical research has been continually in
progress in many parts of Indonesia in collabora-
tion with geoscientists from the University of
London, UK; University of California, Santa Cruz,
USA; Institute of Geophysics, Hawaii, USA;
INSU, France; the Snellius Programme, Nether-
lands; and JICA, Japan. Geological and geo-
physical investigations have also been carried out
by the Indonesian government institutions and the
private sector, including both domestic and foreign
companies, in the exploration for energy and
mineral resources.
The result of these investigations has been an
immense increase in the knowledge and under-
standing of the geological and tectonic develop-
ment of the archipelago. The present paper is a
review of this current understanding of the tectonic
development of the Indonesian archipelago through
Neogene times to the present day. The review arose
out of the responsibilities of one author (TOS) in
the compilation of the new Geotectonic Map of
Indonesia as a member of the Working Group on
the Geotectonic Map of East Asia organized by the
Commission for the Geological Map of the World
(CGMW) and the Committee for Coordination of
Joint Prospecting for Mineral Resources in Asian

FromHall, R. & Blundell, D. (eds), 1996, TectonicEvolutionof SoutheastAsia, 185

Geological Society Special Publication No. 106, pp. 185-201.
186 T.O. SIMANDJUNTAK t~ A. J. BARBER

Offshore Areas (CCOP) as part of the NW Platforms; (4) Indian oceanic plate in the south-
Quadrant Circum-Pacific Map Project. west; (5) a transition zone, marking the zone of
current plate interaction with active seismicity and
volcanism extending through western Sumatra,
Tectonic setting of Indonesia
eastern Java and the Banda arcs to northern Irian
The present physiographic and tectonic configura- Jaya and through Sulawesi and the Moluccas to
tion of Indonesia is considered to have developed Mindanao in the Philippines in the central part of
since late Neogene times due to the interaction of the region. In this zone subduction is still active,
three of the Earth's major lithospheric plates: the with the development of thrust, transcurrent and
NNW-moving (c. 10 cm a-]) Philippine Sea plate; extensional faults. This region is also characterized
the NNE-moving (c. 8 c m a -1) Indo-Australian by allochthonous cohtinental microplates with
plate and the stationary, or slowly SE-moving Tertiary and Mesozoic sediments overlying Palaeo-
(c. 0.4 cm a-1) Eurasian plate (Minster & Jordan zoic basement, juxtaposed against Cretaceous and
1978). The Indonesian archipelago therefore repre- Tertiary terranes to form collision complexes.
sents an immensely complicated triple junction,
involving a complex pattern of small marginal
Tectonic development of Indonesia
ocean basins and microcontinental blocks bounded
by subduction zones, extensional margins and P r e - N e o g e n e tectonics
major transcurrent faults. On the basis of geological
and geophysical characteristics five regions of Western Indonesia. A concentric accretionary
crust of different origins can be distinguished in model for the tectonic evolution of western
the Indonesian region (Fig. 1). (1) Southeastern Indonesia, with the addition of successive orogenic
promontory of the Eurasian plate forming the belts to a continental core in Sumatra and
Sundaland continental craton in Sumatra, West Kalimantan from the Palaeozoic through the
Java and Western Kalimantan; (2) Philippine Sea Mesozoic and Tertiary to the present day, was
oceanic plate in the northeast; (3) Australian proposed by Katili (1973) and Gage & Wing
continental craton extending into the Indonesian (1980). Basaltic to andesitic volcanic rocks and
region in Irian Jaya and the Arafura and Sahul granites dated at 276-298 Ma, and volcaniclastic

I' 125° "

130 °

...... PHILIPPINE.
. "SEA
j
-.

..... I ~
~ CRATONS

MARGINALSEAS
I~ ............
EURASIAN -i ...... PLATE ....... i"]

~ OCEANPLATES
i.i South
iii China
::: Sea :.:.:.:..
ITRANSITIONCOMPLEX
.

EI=I:~SUNDALAND: iiiiiii
...... ............... iiiii .................
............... iiiiii
2o
[27;-iCelebes :I
ii:!:i:Sea.::::::;; :....... CAROLINE PLATE..........
.......... ---_ " .................. " ........... --_
t % ................ - ......... O°
,,; ,,; ,,: ,,." • .'i'
,,, ,, ; ,' ,,

5o,,' :,' : : ,. ~y~

,(.:.i.: Banda
,,.NewGuinea

ioo / INDIAN OCEAN Sumba% Ti

• , PLATE. ./' ,,:' ;:" •,', .', /~Java
,. T r ,"e /' n/' * .;~ -AUSTRALIAN

" ,y :'"' 95°'''' ;'"'/ ,::'"100o'': ,'": ,," Io5*'

,.' .,,,': .::',,,,""
,.'l,0o,,: ,,",,, ./." ]is°."
:: :",y",y
' ,?0o .'j

Fig. 1. The structure of the Indonesian archipelago and its tectonic setting in SE Asia.
 NEOGENE OROGENIC BELTS OF INDONESIA
sediments associated with limestones containing
a Permian fauna in northern and central Sumatra,
have been interpreted as representing a magmatic
arc with associated forearc or backarc regions
(Katili 1973; Silitonga & Kastowo 1975; Rosidi &
Santoso 1976; Cameron et al. 1980; Simandjuntak
et al. 1981). A similar assemblage of rocks was
described from West Kalimantan (Gage & Wing
1980; Pieters & Supriatna 1990; GRDC 1992),
from Sarawak (Tate 1991) and from the Main
Range of peninsular Malaysia (Hutchison 1973;
Gage & Wing 1980). Pupilli (1973) and Gage &
Wing (1980) suggested that these Permian mag-
matic arc systems were related to an Asia-ward
subduction system in Sumatra and an Indian
Ocean-ward subduction system in Kalimantan. A
similar opposed double subduction system is
proposed by Hutchison (1973) and Pupilli (1973)
for the Triassic, with the Malaysian-Indonesian
Tin Belt representing the magmatic belt related
to subduction from the Asian directed subduction
system, while the opposing system is represented
by the Serian Volcanics in Sarawak.
Cretaceous accretion complexes related to the
Asian-ward subduction system can be traced
through the melange wedges of this age in North
and West Sumatra and Bengkulu, in Ciletuh and
Karangsambung in Java, the Meratus Mountains of
SE Kalimantan and in Bantimala and Wasapondo
in South and central Sulawesi (Asikin 1974;
Simandjuntak 1980; Sukamto 1986; Wajzer et al.
1991; GRDC 1995). Cretaceous magmatic arcs
related to this system are represented by granites,
basaltic-andesitic volcanics and volcaniclastics at
Gumai and Garba Mountains in South Sumatra, the
Meratus Mountains of SE Kalimantan and in SW
Sulawesi (Musper 1937; Katili 1973; de Coster
1974; Sukamto 1975; Sikumbang & Heryanto
1986; GRDC 1992). These rocks together with
associated sediments may represent forearc, trench
slope, volcanic arc and backarc assemblages. The
opposing Indian Ocean-ward directed system is
represented by volcanic rocks and associated
granites in the Anambas, Tambelan and Natuna
islands in the South China Sea (Katili 1973; Pupilli
1973), the Schwaner Mountains of West
Kalimantan and the Kuching-Sibu zone in Sarawak
(Haile 1972).
Palaeogene subduction systems can be traced
by melanges through the outer arc islands of Nias,
Pagai and Sipora, offshore Sumatra and in S. Java
(Katili 1973; Karig et al. 1978; Hamilton 1979) and
by volcanics and intrusives, the 'Older Andesites'
of van Bemmelen (1949) associated with volcani-
clastic, volcanic and hemipelagic sediments in
Sumatra, Java and W Sulawesi (Katili 1973; de
Coster 1974; Djuri & Sudjatmiko 1975; Sukamto
1975; Koesoemadinata et al. 1978; Cameron et al.
187
1980; Sukamto & Simandjuntak 1983), extending
into N. Kalimantan, Sabah and Sarawak (Hutchison
1988; Pieters & Supriatna 1990; Tongkul 1991).
Palaeogene volcaniclastic, siliciclastic and
carbonate sediments on the eastern flanks of the
Barisan Mountains in Sumatra and in S. Java, SE
Kalimantan and South Sulawesi, followed by
shallow marine to terrestrial deposits with eco-
nomic coals and hydrocarbon source rocks, may
represent interarc and backarc sequences related to
the Palaeogene subduction systems of western
Indonesia (de Coster 1974; Koesoemadinata et al.
1978; Cameron et al. 1980; GRDC 1995).
Eastern I n d o n e s i a . In the Neogene a number
of microcontinents, e.g. Buton, Banggai-Sula,
collided with subduction systems along the eastern
margin of Sundaland to form the present tectonic
framework of eastern Indonesia. The pre-Neogene
history of eastern Indonesia relates therefore to the
origin and earlier history of these microcontinents
before these collisions took place. All the micro-
continents are considered originally to have formed
part of the northern margin of the Australian con-
tinent and to have separated during Mesozoic or
Palaeogene time. The history of break up and the
subsequent development of these microcontinents
is represented by sedimentary sequences on the
northern margin of Australia, in northern Irian
Jaya, the Arafura and Sahul platforms and within
the microcontinents themselves (Dow 1977;
Pieters et al. 1983; Pigram & Panggabean 1984;
Simandjuntak 1986; Dow et al. 1988; Garrard et al.
1988).
Thermal doming, volcanism, rifting and sub-
sidence to form sedimentary basins commenced
during the Palaeozoic on the northern margin of the
Australian continent, which at that time formed part
of Gondwanaland (Veevers 1984), and intensified
during the Permian and Triassic (cf. Bird & Cook
1991 from Timor). The exact timing of separation
of the microcontinents from Australia is still very
much in dispute. Pigram & Panggabean (1984),
based on the model of breakup sequences proposed
by Falvey & Mutter (1981), suggested that the
separation of microcontinents from central Papua
New Guinea occurred in the early Jurassic between
141 ° and 145°E longitude. Rifting with the
deposition of red beds in Irian Jaya and Papua New
Guinea was followed by the deposition of a passive
margin sequence which continued into the early
Tertiary.
It was earlier suggested that the microcontinents
were detached from Irian Jaya in the mid-Tertiary
and displaced westwards by movements along the
Sorong Fault and/or its subsidiary traces (Visser
& Hermes 1962; Krause 1965; Katili 1971; Gribi
1973; Hamilton 1979; Silver & Smith 1983).
188 T.O. SIMANDJUNTAK8z A. J. BARBER
However, the Jurassic passive margin sequence
preserved in the microcontinents is overlain para-
unconformably by Upper Cretaceous pelagic
calcilutites; the Early to mid-Cretaceous was a
period of non-deposition in the microcontinental
blocks (Pigram et al. 1985; Simandjuntak 1986;
Garrard et al. 1988). This depositional hiatus,
followed by the abrupt deepening of the margins
of the microcontinents in the Late Cretaceous, has
been used to argue that the microcontinents were
already detached and displaced from the northern
margin of Australia by this time (Simandjuntak
1986, 1994; R6hault et al. 1991).
Neogene orogenies in Indonesia
The present physiography of the Indonesian archi-
pelago can be attributed directly to Neogene
orogenic events. These events included plate
convergence with subduction beneath the margins
of Sundaland to produce a Cordilleran type of
orogeny, a unique arc-arc collision in the
Moluccas, collision of arcs and microcontinents
and major continental blocks with the construction
of mountain belts, transpression and transtension
along major transcurrent belts, the construction of
foreland fold and thrust belts, back-arc thrusting
and reversal of subduction polarity. Examples of
all of these types of orogenic event can be found
among the islands of the Indonesian archipelago.
These events will be described in terms of their
5° r - -
~ "P$ ~ 'S-~ )f~ ~ Fig6-
I°° INDIAN ]fig. 3 ~OC,EN¥ % [ ~
. . . . . | ~..__ OROGENY ~ c-q~J
- 95° l(~o 105° 0o 115° 1210°
Fig. 2. Location of Neogene orogenies in Indonesia. Detailed figs to illustrate the structure of these orogenies are
specified.
contribution to the development of five orogenic
types in Indonesia (Fig. 2).
Sunda orogeny
The Late Neogene Sunda orogeny affected the
segment of the Indonesian arc between West Java
and the islands of Nusa Tenggara as far east as
Flores (Fig. 3). In this segment of the arc con-
vergence between the Indian Ocean and SE Asian
plates is normal to the subduction trace in the Java
Trench with a rate of c. 7 cm a-1. The subduction
system comprises an accretionary complex com-
posed of offscraped Indian Ocean floor materials
in the Java forearc ridge, a forearc basin developed
on extended continental crust and containing late
Palaeogene to Recent sediments. The volcanic arc
which forms the backbone of Java and forms the
islands to the east is constructed on continental
crust, in West Java, on Mesozoic accretionary
complexes in Central and East Java and on oceanic
crust in Sumbawa and Flores. To the north of the
arc in Java and in the Java Sea Tertiary backarc
basins have developed on continental crust in the
Sunda Shelf and on oceanic crust north of Bali and
Flores. The basins on the Sunda Shelf formed in
the late Palaeogene as rift basins in a terrestrial
environment and subsided in the Neogene to be
transgressed and covered by marine sediments.
In the Late Neogene this system was affected by
compression associated with the Sunda orogeny.
TALAUD °
~ ~
BANDA[ ] Fig.8 ~-I0°~
q] ~
~,~,~_~ AUSTRALIA ~x. ,5o|
~25 ° ]:0 ° 135~ 11400/ 14; ° s.~.,,m..,,I~I.1
 NEOGENE OROGENIC BELTS OF INDONESIA
In N. Java Mio-Pliocene turbidites are deformed
into tight, locally isoclinal folds, readily observed
in the field and on aerial photographs, while in
S. Java-Nusa Tenggara older volcanic sequences
are folded, faulted and uplifted to form mountains
more than 3500 m above sea level. This phase of
folding is associated in time with the intrusion of
acid plutons, the uplift of the volcanic arc, the
development of a major thrust system in which the
volcanic arc is overthrust towards the Java Sea to
the north, and the subsidence of the backarc basins,
with the deposition of fine siliciclastic sediments,
marls and carbonates of Plio-Pleistocene age.
In N. Java the trace of a major backthrust, the
Barabis-Kendeng Thrust (Simandjuntak 1992) can
be traced from the Sunda Strait eastwards across
Java and through the Bali Basin into the Flores
Thrust, north of Flores (Prasetyo 1988). The thrust
may continue eastwards as the Wetar Thrust to the
north of Timor. The Barabis-Kendeng Thrust has
been imaged in seismic reflection profiles in the
northern part of West Java (Supryanto & Ibrahim
1993) and offshore north of Flores (Hamilton
1979; Prasetyo 1988), whilst the Bouguer gravity
anomaly pattern in the northern part of East Java
indicates the location of the Kendeng Thrust. In
Central Java the thrust is cut and disrupted by the
Cimandari and Citandui Faults which have wrench
components of movement (Dardji et al. 1994).
Earthquakes recorded at Majalengka, Brebes (West
/ Christmas convergoncoA-~j ava Tren,,h
~lz°s / [ / I Roo Rise
I ~/~~Lienti2f I ~' ~
/ I ~-~ ~ S ~--~
Fig. 3. Sunda orogeny in Java and Nusa Tenggara. Line of section illustrated in Fig. 9 is indicated.
189
Java) and Pekalongan (Central Java) show that
segments of the backarc thrusts are currently active
(Kertapati et al. 1972).
The causes of the Late Neogene Sunda orogeny,
with a change from an extensional to a com-
pressional regime across the subduction system, are
not clear. Compression may develop when smooth
subduction of the downgoing oceanic plate is
interrupted by topographic irregularities on the sea
floor. This explanation could apply in the East
Java-Bali segment of the subduction system where
the Roo Rise is beginning to enter the subduction
trench (Fig. 3), but does not explain compression
elsewhere in the arc. It may be that the compression
is localized along the northern margin of the
volcanic arc due to magmatic intrusion and the
consequent uplift of the arc.
Barisan orogeny
The Late Neogene Barisan orogeny affected the
segment of the Indonesian arc occupied by the
island of Sumatra (Fig. 4). Convergence in this
segment of the arc between the Indian Ocean and
Southeast Asian plates at a rate of c. 7 cm a-1,
is currently markedly oblique (50-65°). The sub-
duction system includes an accretionary complex,
exposed in the offshore islands such as Nias
(Moore & Karig 1980), and a forearc basin. The
volcanic arc is constructed on continental crust and
Sulawesi
NMD
• Activevolcanoesofthe VolcanicArc
Quaternaryvolcanicdeposits
I ExposedPre-TertiaryRocks
Cm(Cimandari)Fault;Ct (Citandui)Fault
/ S P SeribuPlatform, A-AD Asem-AsemDepression
~
WF E WestFlorenceEmbayment,A D ArdjunaDepres:
E F B EastFlorenceBasin, M D MasalemboDepression
N M D NoithMaduraDepression
J
190 T . O . SIMANDJUNTAK • A. J. BARBER

the volcanoes sit on an uplifted pre-Tertiary base-
ment terrane which forms the Barisan Mountains
running the whole length of Sumatra. Both the
basement and the volcanic arc are dissected by
the Barisan dextral transcurrent fault system.
Extensive Tertiary backarc basins lie to the north-
east, behind the arc.
Sumatra forms part of the Sundaland continental
craton. In the Palaeogene the region was affected
by extension and subsidence, resulting in the
formation of rift basins. Some of these basins, such
as the coal-bearing Ombilin Basin of West Sumatra,
lie within the Barisan Fault System and have
been interpreted as pull-apart basins formed by
transtensional movements along the fault
(Koesoemadinata et a/.1978; Koning & Aulia
1985; Wang et al. 1989; Situmorang et al. 1991).
Continuation of movements along the fault led to
the opening of the Andaman Sea (Curray et al.
1979; Harding 1983).
Commencement of the Barisan orogeny was
marked by the uplift of the Barisan Mountains
and arc volcanism, signalled by the influx of
volcaniclastic sediments and regressive sequences
in the Sumatran back-arc basin in the mid-Miocene.
Uplift was accompanied by intrusions in the
volcanic arc and transpressive movements along
the Barisan Fault System. Fault movement resulted
in the imbrication, duplexing and stacking of
crustal slices to form a large-scale positive flower

I
96°
I

98° 0 at'0km
N
~ ada~n~ '(~lt
panjang ~xSingkarak
Lake ~ rifts I
andEXtensi°nallakes
x~.~olok
\ P " , r ~t.~ n;t,..... J,

West
Sumatra
Lake Toba

, Mt. Kerinci
Jambi
~ ngaipenuh

Bengkulu

Line of
section

• Active volcano ()ff

4° ~ Quaternary-Recent Volcanics

Tertiary Sedimentary Basins

Pre-Tertiary Basement %
6°
0 100 200 300 400 500km • ,\\ • Krakatoa
n •

[ 96° 98° 100° NN 102° 104° 106° I

J ] I [ [ Steve Moss, J 995. I

Fig. 4. Barisan orogeny in Sumatra. Crossed lines in the back-arc basins indicate fold axial traces. Inset map shows
extensional features in the section of the Barisan Fault between Bukittinggi and Lake Kerinci (location indicated on
main map). Line of section illustrated in Fig. 9 is indicated.
 NEOGENE OROGENIC BELTS OF INDONESIA
structure, leading to the erosion of the cover and
exposure of the basement rocks which now rise
to nearly 4000 m above sea level. Uplift of the
Barisan Mountains was accompanied by sub-
sidence of the forearc and back-arc basins.
Plio-Pleistocene transpressive movements along
the Barisan Fault System are considered to have
induced fold structures in the sediments of the
back-arc basins on axes which trend at an angle of
20 ° to the main fault (de Coster 1974; Hamilton
1979). The folds are commonly associated with
small-scale wrench faults and may be related to
large-scale wrench faults in the underlying base-
ment (Tiltman 1990). Pleistocene movements along
the fault system have, at least locally been trans-
tensional with the opening of pull-apart basins,
often holding lakes, such as Laut Tawar, Toba,
Singkarak, Diatas, Dibawah, Kerinci (Fig. 4, inset),
Ranau or a Recent sedimentary fill as in the
Semangko Valley.
Recent earthquakes resulting in the cessation of
sulphuric hot springs and the disappearance of
ponds at Tarutung (North Sumatra), and at Padang
Panjang in Central Sumatra, indicate dextral trans-
pressional reactivation along the northern segment
of the Barisan Fault System. In Lampung Province,
S. Sumatra NW-SE extensional cracks formed
during an earthquake near Liwa in early 1994,
indicating sinistral transtensional movement along
this segment of the fault.
The Neogene Barisan orogeny is attributed to
variations in the rate of subduction of the Indian
Ocean plate and the response of SE Asia to the
continuing collision of India with the southern
margin of Asia and the adjustment of crustal blocks
by movements along major transcurrent faults
(Tapponnier et al. 1982). The obliquity of sub-
duction is considered to be responsible for the
development of the Barisan Fault and the detach-
ment of the Sumatran forearc to form a 'sliver
plate', which is partially coupled to the northward
movement of the Indian Ocean plate (Fitch 1972;
Curray 1989; McCaffrey 1991). Movements along
the Barisan Fault are considered to be responsible
for the uplift of the Barisan Mountains and the
transtensional and transpressional effects seen
along the fault trace. Results of break-out analysis
of sediments from boreholes in the Sumatran
backarc basins show that the effects of oblique
subduction are also being transmitted across the
Barisan Fault into the backarc (Mount & Suppe
1992).
Talaud o r o g e n y
The Neogene Talaud orogeny of the N. Moluccas
provides the only example of an active arc-arc
collision in the world. The Molucca Sea is bounded
191
to the west by the Sangihe volcanic arc and to the
east by the Halmahera Arc (Fig. 5). Seismicity
shows that the Sangihe Arc is underlain by a west-
dipping Benioff zone extending to a depth of
700 km, while an east-dipping Benioff zone
underlies the Halmahera arc to a depth of 200 km
(Silver & Moore 1978; Hamilton 1979; Sukamto
et al. 1981). The shallow Molucca Sea between
the two arcs is considered to be underlain by the
collided accretionary complexes of the two
opposing arcs uplifted to form the central Talaud-
Mayu Ridge, the underlying Molucca Sea plate
having subsided back into the mantle between
them. Hamilton (1979), on the basis of seismic
reflection profiles, suggested that the volcanic
aprons from the forearcs of the Sangihe and
Halmahera arcs had been thrust backwards across
their corresponding arcs as the result of the
collision. Recent Seabeam and seismic reflection
data from the northern part of the Molucca Sea
indicate that the floor of the sea consists entirely
of the Sangihe forearc which has been thrust east-
wards over the Halmahera forearc (Rangin et al.
1996). To the south the collision complex is
truncated by a major strand of the Sorong Fault
System. The Sangihe subduction system may swing
round to the west and continue as the Batui Thrust
in the East Arm of Sulawesi. Frequent earthquakes
show that all these structures are currently active
(McCaffrey et al. 1983; Simandjuntak 1989;
Kertapati et al. 1972).
Sulawesi orogeny
Sulawesi is made up of three structural units: pre-
Cretaceous accretionary material in the west upon
which later developed a Neogene volcanic arc;
Central Sulawesi and part of the SE Arm are
composed of metamorphic rocks; the E. Arm and
the remainder of the SE Arm are made up of a
major ophiolite complex. The metamorphic rocks
in Central Sulawesi and in the SE Arm include
materials of both continental and oceanic
derivation. These rocks are locally affected by
high pressure metamorphism forming blueschists,
developed in an east dipping Paleogene mid-
oceanic subduction zone. Subduction resulted in
the obduction of the East Sulawesi Ophiolite west-
wards across the metamorphic belt, accompanied
by the extrusion of ophiolitic melange (Parkinson
1991).
Neogene orogeny in Sulawesi was initiated by
the collision of the two microcontinental blocks
of Buton-Tukangbesi and Banggai-Sula with the
eastern part of the island (Fig. 6). These two micro-
continental blocks, having separated from the
northern continental margin of Australia, possibly
from the region of central New Guinea as has been
192 T. O. SIMANDJUNTAK~ A. J. BARBER

- - I
• Active Volcano

~ Quaternary Volcanic
Mindanao
Deposits PHILIPPINE
SEA
~ Molucca Sea
Collision Complex
6°N

CELEBES
SEA
4°N

lO00m
; / --MOLUC A
3000m

Line of
Section
Morotai
!°N

)00m .....

Sulawesi Halmahera

oo 7_7/

Batui Thrust

I '~ ~.~ ~ ~ ~ l t !{ S O R O N G FAUI~T S Y S T E M [

12~ ~ t ~ ~u~- ,~oE ,2U

Fig. 5. Talaud orogeny in the North Moluccas. Line of section illustrated in Fig. 9 is indicated.

previously discussed, were carried westwards along
the Sorong transcurrent fault zone by movements
of the Philippine Sea plate (Ali & Hall 1995) and
collided with the eastern margin of the ophiolite
complex. The collision caused the obduction of the
ophiolite onto the microcontinental blocks and
the shortening and thickening of the ophiolite by
imbrication. The leading edges of the Buton-
Tukangbesi and Banggai-Sula microcontinents
were thrust beneath the ophiolite, uplifting the
tightly folded, faulted and imbricated ophiolite and
its pelagic cover to heights of more than 3000 m
above sea-level (Smith 1983; Simandjuntak 1986;
Garrard et al. 1988; Fortuin et al. 1990; Davidson
 NEOGENE OROGENIC BELTS OF INDONESIA 193

T T
123 °E 125°E
2ON -
Minahassa T r e n c h - ~
~- lO00m \
/ ~ ~ ~ 3000m

~ ~/) North Arm :S~2~

MOLUCCA
SEA
3000m 0 o
s a'
(

Line o(

East Arm

ult

2°S
, Central

B A N D A SEA
Buru
Southeast Arm
4°S

Buton
-, i .k ng
/
I
\ i" A ,27°E

Recent Deposits ~;~ Continental Crust

~ Quaternary Volcanic Rocks ~ Metamorphic Rocks k , ~ Majene I
Neogene Volcanic Rocks ~ Ophiolitic Rocks \" Fold Belt I
S. Baker 19951

Fig. 6, Sulawesiorogeny. Line of section illustrated in Fig. 9 is indicated.

1991). Also as a result of the collision the meta-
morphic belt of Central Sulawesi was thrust west-
wards over West Sulawesi and uplifted to form
mountain ranges of nearly 3000 m. Overthrusting
resulted in the formation of a foreland fold and
thrust belt in Tertiary sediments, the Majene Fold
Belt, which continues to develop westwards to
the present day, affecting Recent sediments in the
Makassar Strait (Coffield et al. 1993; Bergman
et al. 1996).
Associated with the collision, or following
shortly after, was the development of the NNW-
SSE trending Palu-Koro sinistral transcurrent fault,
along which eastern Sulawesi has been displaced
northwards with respect to western Sulawesi.
Simandjuntak (1993a) has discussed the effects
of transtensional movements along this fault and
related displacement of the island of Sumba to its
present anomalous position in the Banda forearc.
More recent transtensional movements during the
194 T.O. SIMANDJUNTAK t~ A. J. BARBER

Quaternary, continuing to the present time, are the island of Timor (Barber et al. 1977; Charlton
responsible for opening pull-apart basins, such as et al. 1991; Harris 1991). The island is composed of
those of Lakes Poso, Matano and Towuti, as well as sediments of Permian to Pliocene age, of Australian
the Palu depression. Recent earthquakes along the affinity which have been folded, thrust and imbri-
Palu-Koro and related faults show that the system cated. The Australian sediments have been thrust
is currently active. beneath an ophiolite slab with an underlying meta-
morphic sole, representing the trace of the Benioff
zone prior to collision, but now uplifted to heights
Banda orogeny
of more than 3000 m above sea-level in Timor
The Neogene Banda orogeny is due to a continent- (Sopaheluwakan 1990). Most of this uplift occurred
arc collision where the northern margin of the in the Neogene, as Miocene, Pliocene and
Australian continent, moving NNE at c. 7 cm a-1 Pleistocene forearc sediments resting on the com-
is colliding with the subduction system along plex show a history of deposition shallowing
the southern side of the Banda Arcs (Fig. 7). The upwards from bathyal facies through shallow water
deformation front, marking the southern limit of the deposits to Pleistocene coral reefs (de Smet et al.
collision complex, lies along the 2 km deep Timor 1990). These facies changes indicate two periods
trough. Seismic reflection profiling across the of rapid uplift, one 2 Ma and the other c. 100 000
trough shows Australian passive margin sediments years ago, representing different stages in the
passing northwards down beneath the trough collision.
where, at the deformation front, the upper parts of The collision between the Australian margin and
the sequence have been uplifted on thrust surfaces the Banda Arc subduction system appears to have
to form fold ridges in an accretionary complex commenced in the region that is now East Timor,
(Karig et al. 1987). and the collision is at its most advanced stage in
The structure of the collision zone is exposed in this segment of the arc. Here the forearc between

,;o\ < ,;,,o

_FLORES SEA /" * *

C ACTI

SUMBA \

/ ................

I " ~ - "----~ " , L P - ~ " ~ t - ~ f

12o ~t . , ~ Y / ) ~ o HiberniaReef
/ ~ ~ f O AshmoreReef q i l l l ] I I I I l
O.~-v.-'l I I I I I I I I I J
. A c ~ O ~ : a qff+H-H -
I,'x,,3 ~#" I F I I I I I I I I I I I I I I I I
t.d III I I1111111 II
I~/"~]KTAI]IA De'l'q~ 91 I I I I I I I | I I I I ! I I
I.I'~.J.L~II"51d-/-~IX.I/~ n ~ I I I I t I 1 I I I I I t I I I
/[111 I I I 11 I I11 I I II I I
r~T TT 1"~ I l l l l l l l l l l l l l l l l l l l
~_ "tk I Y,,~x- |III1.1 ~ ¢11111111111111111ll i
PLATEAU/~ Scott Reef K.tr,.) , ¢ . ~ ~
" - ~ ~[-'~ , 4 " l - l - l - [ - l - l ~
v ~ - - I
J'l II |l IIII
1 l l l l l l l I I l l l l 4
IIII1111111 I
C

,"~ q~r-I" i i i i i i i . . . . . . . i ii ii ii11 i iii II II I I

o ,,oo/2oo 300 400 500 . ~ /'t./J-l"Ifl I II I 1 1 I 1 I /"~l./t,.,]l I I~"XL.tl/"II I I I I I I I I1 I I I II I I I I

t
~["11 I I I I I | I I I I I I I J I I"~ -kl.I I I I I I I I I I I I IJl I I I I
120° / 122° 124 ° ~t.,•, ~ I IdJ126°l
"~7 I I
I I
/1111111111LI
1 I 1280 . ~ c ~ .,,
ff I I I I X . . . ~ I ~
,.72130 ° I I ] i I I I
'xl III I 111
I 11 1132 ° I I
II11 I I
I I I I '~l¢~ t~"~ I | I I i 1 i i i i i i i i I t ' l 1 1 I I I )~'1"-~ I'T I t 1q I I I I I I I I I i i

Fig. 7. Banda orogeny: the collision zone between the southern Banda Arcs and Australia near Timor, based on
Hamilton (1979). Horizontal ruling, collision zone with islands in black; triangles, active volcanoes; small v,
associated with Quaternary volcanics; larger v pattern, Neogene volcanics. Line of section illustrated in Fig. 9
is indicated.
 NEOGENE OROGENIC BELTS OF INDONESIA 195

the volcanic arc and the deformation front has been
reduced to a width of less than 100 km, compared
with over 400 km for the forearc east of Sumba
where collision has commenced only recently.
Continued northward movement of the
Australian continent has been accommodated by
the uplift of the collision complex in Timor and
lateral extension along conjugate faults (Charlton
et al. 1991). Compression of the forearc has
virtually eliminated the forearc basin which is only
40 km wide to the north of East Timor. Further-
more, shallow seismic activity is absent beneath
East Timor suggesting that the downgoing plate
and the collision complex have become locked in
this segment of the arc. Volcanic activity in the
islands of Alor and Wetar to the north of East Timor
ceased about 3 Ma (Abbott & Chamalaun 1981)
and the volcanic arc now rests on a southward-
dipping thrust plane, the Wetar Thrust. The
development of a small accretionary complex to the
north of the thrust (Breen et al. 1989) suggests that
the whole of the forearc and the volcanic arc are
being carried northwards with the movement of
Australia, and are overthrusting the Banda Sea
floor to the north. These may be the first indications
of reversal of the polarity of the subduction system,
with the Banda Sea floor passing down southwards
beneath the northern margin of Australia. Recent
deep-seismic reflection profiles to depths of up to
100 km, a short distance to the east of East Tirnor,
show reflections representing thrust surfaces
dipping both northwards and southwards beneath
the collision complex which has been uplifted as
a wedge by the convergence of the Australian and
the Banda Sea lithosphere (Snyder et al. 1996).
Melanesian orogeny
The Neogene Melanesian orogeny in Irian Jaya
and Papua New Guinea is considered to be the
result of oblique convergence of the Australian and
Philippine Sea (and Caroline) plates (Dow &
Sukamto 1984). In the late Palaeogene or very
early Neogene, the northern promontory of the
Australian Continent collided with an oceanic
island arc constructed on the southern margin of
the Philippine Sea plate (Hall et al. 1995). The
remnants of this arc are now distributed through
the northern part of New Guinea (Pigram & Davies
1987), disrupted by transcurrent fault movements
to form discrete terranes (Fig. 8).
The Central Ranges of New Guinea are com-
posed of a major ophiolite complex, here inter-
preted as the oceanic lithosphere which underlay
the oceanic island arc, subsequently uplifted as the
result of the collision with Australia. To the south
of the ophiolite the Australian passive margin sedi-
ments are deformed by thin-skinned thrust tectonics
into a foreland fold and thrust belt, with folding,
imbrication and duplexing. At deeper levels
seismicity indicates that the underlying Australian
continental crust is also involved in the thrusting
(Abers & McCaffrey 1988). The whole collision
belt is being thrust southwards over the northern

t i i
~0. r ~ ~ G~nea PACIFIC OCEAN A

~
Active Volcanoes
Wai e°*~'~-SOR~Q~ FA UL~ ' ~ ~/~ ~ ~ ? b t t ~ ~ QuaternaryVolcanics
\~'-~~r~. ~ ~ PalaeogeneVolcanics
e a d ~ ~ ~ Ophiolites& related

-----,.,,.~%~ t...,.- /~ ~ ~LAEOGENF'~"..~~ "~ Thrusts

7~4"s, "~ " ~ "~. "~.\~"" " _ FORELAA,_ ~ "-,4N/'c~ . ~. k Nx&

\

,~~' ~ ~ / ~~f ~ / y / ~ / ~ rY pMERAUKE

1pM~LAFUR~
TRA Lineof~/FORELANI) 2 B A S I ~ ' ~ "
-~//~/// Section~....~-" ' ~ . -." . ~ I,
. jty~,aom
8s ~: .
m A~.(,Tanlmbar g------J~~ i ~ -x_
ARAFURA '--' ~ :
SHELF
1301*E 1321"E 13~'E 1361E 1381°E 1401. . . . ~E 1441~E I~4~'[Or~E

Fig. 8. Melanesian orogeny in New Guinea, based on Hamilton (1979), with modifications after BMR (1972) and
Dow et al. (1986). LFB, Lengguru Fold Belt; T-AF, Tarera-Aiduna Fault; WT, Waipona Trough. Line of section
illustrated in Fig. 9 is indicated.
196 T. O. SIMANDJUNTAK & A. J. BARBER

margin of the Australian continent. The southern
margin of the overthrust belt is marked by the
Asmat Thrust (Fig. 8).
The progress of the Banda and Melanesian
orogenies in eastern Indonesia has been controlled
by irrregularities in the northern margin of the
Australian continent (Charlton 1986). Collision
c o m m e n c e d in New Guinea. Using the k n o w n
rate of northward m o v e m e n t of the Australian
continent, Simandjuntak (1993b) postulated that
at the time of the original collision, the subduction
zone to the north of New Guinea was continuous
with the subduction zone to the south of the Banda
Arcs. Continued northward m o v e m e n t of Australia
since the collision has displaced the collision zone
northwards. Simundjuntak (1993b) also suggested
that the western m a r g i n of the N e w G u i n e a
promontory is marked by the Waipona Fault which
transects the Asmat Thrust. He speculated that the
northward m o v e m e n t of New Guinea relative to
the Banda Arcs is responsible for the northward
curvature of the Banda Arcs near Tanimbar and for
the curvature of the Lengguru Fold Belt of the
Bird's Head of Irian Jaya. Transcurrent m o v e m e n t
along this fault is transpressive near Tanimbar and
transtensional in the Aru Trough
Following the collision, continued northward
m o v e m e n t of Australia has thrust the Palaeogene
arc, now attached to the northern margin o f
Australia, over the Caroline plate to the north in a
m u c h quoted example of polarity reversal (Dewey
& Bird 1970; Johnson & Jaques 1980; Cooper &
Taylor 1987). Recent geophysical studies in the
Manokwari Trough immediately to the north of
Irian Jaya, however, show that subduction is no
longer in progress there (Milsom et al. 1992).
Extensive outcrops of andesitic volcanic rocks of
mid- to late Miocene age throughout northern New
Guinea indicate that southward subduction of the
Philippine Sea plate beneath Australia occurred at
that time. Quaternary and Recent volcanism is
largely restricted to Papua New Guinea, and the
potassic nature of the volcanic rocks suggests
that they are related to extension rather than to
subduction (Dow 1977).
The northern part of New Guinea is transected
by a major E - W strike-slip fault system, the New
Guinea Megashear, which continues westwards
into the Sorong Fault System of the Moluccas,
already mentioned. Some of the records of shallow
earthquake activity, especially concentrated in
central Irian Jaya, may be related to continued
m o v e m e n t on this fault system. Most records, how-
ever, show a thrust sense of movement related
to compression in the M o m b e r a m a Thrust Belt
(Abers & McCaffrey 1988). These thrust move-
ments may be the effects of transpression which
may also be responsible for the uplift of the
Central Ranges to a height of nearly 6000 m.
Transtensional movements in other parts of the
fault system may be responsible for the opening up
of rifts such as the Markham Valley in Papua New
Guinea (Hill & Gleadow 1989). Transcurrent fault
movements in northern New Guinea are attributed
to the c o n t i n u e d westward m o v e m e n t of the
C a r o l i n e / P h i l i p p i n e Sea plate at a rate of
12.5 cm a -l relative to the Australian plate.
Conclusions
The whole range of plate tectonic and orogenic
processes of subduction, accretion, construction
of volcanic arcs, back-arc thrusting, strike-slip
faulting, formation of sliver plates, arc-arc,

Fig. 9. Comparative interpretative cross-sections across the Neogene Orogenic Belts of Indonesia, all at the same
scale. (a) Sunda orogeny. Convergence between Indian Ocean and Eurasian plates is normal to the subduction trench.
The system consists of an accretionary complex, forearc basin, volcanic arc and backarc basin. Compression across
the system has resulted in backarc thrusting. (b) Barisan orogeny. Oblique convergence of the Indian Ocean and
Eurasian plates. The system is modified by the development of a 'sliver plate' where the forearc is driven northwards
along the Barisan strike-slip fault by the movement of the Indian Ocean plate. (c) Talaud orogeny. The Molucca Sea
plate is subsiding between the colliding Sangihe and Halmahera forearcs. The Sangihe forearc has been thrust over
the Halmahera forearc to form the Talaud Ridge. Subduction of the Sulawesi Sea plate has commenced only recently.
(d) Sulawesi orogeny. The Banggai-Sula microcontinent has collided with the ophiolite in the East Arm of Sulawesi.
Repercussions of this collision include strike-slip movements along the Palu-Koro Fault and overthrusting in the
Makassar Strait. (e) Banda orogeny. The Australian craton has been subducted ~.eneath the accretionary and collision
complex in the Timor Ridge. The ridge is composed of accreted Australian continental margin sediments and earlier
collided microcontinents. Subduction has ceased in this section of the collision zone, but the collision complex is
being driven over the volcanic arc, which is in turn being driven over the Banda Sea plate to the north, in the earliest
stages of subduction reversal. (f) Melanesian orogeny. The Australian Craton has been subducted beneath a
Palaeogene volcanic arc with the development of a foreland fold and thrust belt from its overlying sediments.
Ophiolite and arc volcanics in the central Ranges represent the uplifted hanging wall of the subduction zone. The
system has reversed its polarity with the subduction of the Caroline Sea plate but the system is truncated by move-
ment along the Sorong strike-slip fault zone due to the westward movement of the Caroline Sea plate. Data are drawn
from references quoted in the text of the paper.
 NEOGENE OROGENIC BELTS OF INDONESIA 197

SUNDA OROGENY JAWA JAWA SEA

INDIAN OCEAN Forearc Volcanic Arc Sunda Shelf
S RoD Rise Jawa Trench Ridge Basin __ ~.-_.d~,._~ Kendeng Thrust __ N
0 : -ml ___ - ~ ~ ~-..--
[ii:i:
.: ......................
;.
::::::::::::::::::::::::::::::::::::: ....
. ,
esozo,c Accret,onaryComplexes ~-~r/) ~ -Jr"
v- ~ F
X
..,11..
~- SUNDA~ND _,×7
Jr ' L
f
X

INDIAN OCE _-F__ "/'._I

~o :iii!!:!
(a)
BARISAN OROGENY NIAS BARISAN MOUNTAINS MALACCASTRAIT
INDIAN OCEAN Forearc Barisan Fault
SW Nicobar Fan Sunda Trench R dge flas n Toba Caldera Backam Basn NE
o . . . . .

~
~;~ ~# ~ < :~:~;~::::.,:::....~....................
s X 4-
i- X
4- Zs
F
u
NDA
LA
ND 4" 4-
~m ~+~t
:::i::ii~.i}i;::~:~: PLATE ~ i:~i:~:~ !~.:~iiiii~::iii!::~i{i~!:::.~.... :::~!ii:: !~!i ii}?i:~ili ::ii i ii ii i~:i::i~ i! ! EURAS AN PLATE i
~o :iii ~,,~,,,::,~:~::::~'':'~i!i!i!~i'{~ii,ii~!ii,i'iiii',!'~i~:i
ii~i!iiii'
i{i!i!ii!i!iiiiigi~!::~
(b)
TALAUD OROGENY MOLUCCA SEA HALMAHERA PACIFIC OCEAN
SULAWESI SEA NORTH SULAWESI Talaud Ridge Arc Halmahera-Waigeo
W Sang~e Arc thrust Mavu thrust Ophiolite Terrane E
O L ~ ,..- < .... ". . . . ". . . . - . . . . \\~-"- - - - 4
1 1 1 ~\[~'~.~" "~ " ,.. ... ~ .~'_'~. " ~ - 5 - ~ _ ' % \ \ \ \ . . . . . . . . . . . , \ \ \ \ \q
~ - ' ~ ~ .~:}.~ ~-~>\ \ Imbricatedophiolite . \ \ \ \ \ 1
\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ _ ~

....;:~:~::~ii~::::::ii::i!!:::::,i::: ii::ii!iiill~ili::i}~ii:.i!iiiii
i i::::ii::i~ii~7.if!~i:ii:~
:i i iii~ili!i!i!ii~ii:::iii:!i!~!i::ii~ili:.i::: i :.i!i iilliii:: :~:

,o '.iiii:,?i~:~)!:!iiiii~iiii!
iii':::i!
i
(c)
SULAWESI OROGENY CENTRALSULAWESI EASTARM SULAWESI
MAKASSAR STRAIT MagmaticArc MetamorphicBelt OphioliteBelt BANGGAI SULA
W thrust belt Palu-KoroFault Batui Thrust E

-- • 7L r--;l~z:-q-'--. " ~'",~L_-7L_T~--------~r., "-x -+- , ".~ 4- , -x 4-- .4. X ~ ./. X I

b++ sU.~A~D . ~ 9 C O . N T ~ ' ~ , N ~ x + + × + ~1
km ~" + ~" + % + ", ~. ' + ' ~ _~ ' + "'a % '+ ", "~ ' ~
50 ""~ .................................................. >:~' ' > : : : : : : : :1:~:+ :"~*:~;~"~ ~ .................1:...................." : 1 1 : . . . . . . . .

((1)
BANDA OROGENY VOLCANIC ARC
S TIMOR SEA TIMOR RIDGE (inactive) BANDA SEA
Sahul Shelf Timor Trough Kisar Thrust WetarThrust Gunung Api N

°b + .........

k4 . . . . . . . + 4. × 4- ./_
kml" + AUSTRALIANCRATON jr x + 4- x

50 +
(e)
MELANESIAN OROGENY CENTRAL RANGES PACIFIC OCEAN
Dolak Thrust B e l t Palaeogene
Arc Sorong Fault Zone
S MeraukePlatform ForelandBasin Asmat Thrust MeervlakteBasra New Guinea Trench N

L+ + ~+ + ,+ + ~+ + ~ ~ ~ ~ . ~ .
kml-4.+ x if'- + . X AUSTRALIAN cRATON x if" ~ if" +_, X ~
L -/_ "/- ~ ~ ,- ..,. .- ..L "/- -r- + ./. -r .~ ~,.,, ~ ~
.X'Y. -k'-- "/. k ' - - "/- ~ r - - "/. -k'-- "/. - X " - - × ' X

50
0 1O0 200 300 400 500 600 700 800
198 T.O. SIMANDJUNTAK t~ A. J. BARBER
micro-continental-arc and continental-arc
collision, foreland fold and thrust belt formation,
obduction and the uplift o f mountain belts m a y be
f o u n d currently in progress in the I n d o n e s i a n
archipelago. Our present understanding o f the
structure o f the N e o g e n e - R e c e n t orogenic belts o f
Indonesia is illustrated in a series o f cross-sections
in Fig. 9. These present-day e x a m p l e s o f orogenic
processes p r o v i d e models for the interpretation o f
older orogenic belts such as the A l p i n e - H i m a l a y a n
system, the H e r c y n i d e s and the Caledonides.
References
ABERS, G. & MCCAFFREY,R. 1988. Active deformation in
the New Guinea fold and thrust belt: seismological
evidence for strike-slip faulting and basement
involved thrusting. Journal of Geophysical
Research, 93,13 332-13 354.
ABBOTT, M. J. • CHAMALAUN, F. H. 1981.
Geochronology of some Banda Arc volcanics.
In: BARBER, A. J. & WIRYOSUJONO, S. (eds) The
Geology and Tectonics of Eastern Indonesia.
Geological Research and Development Centre,
Special Publication, 2, 253-268.
ALI, J. R. & HALL, R. 1995. Evolution of the boundary
between the Philippine Sea plate and Australia:
palaeomagnetic evidence from eastern Indonesia.
Tectonophysics, 251, 251-275.
ASIKIN, S. 1974. Evolusi tektonik Java Tengah dan
sekitarnya ditinjau dari segi toori tektonik dunia
yang baru. PhD Thesis, ITB Bandung.
BARBER,A. J., AUDLEY-CHARLES,M. G. & CARTER,D. G.
1977. Thrust tectonics in Timor, Journal of the
Geological Society of Australia, 24, 51-62.
BERGMAN, S. C., COFFIELD, D. Q., TALBOT, J. P. &
GARRARD, R. J. 1996. Tertiary tectonic and
magnetic evolution of western Sulawesi and the
Makassar Strait, Indonesia: evidence for a Miocene
continent-continent collision. This volume.
BIRD, P. R. & COOK, S. E. 1991. Permo-Triassic
successions of the Kekneno area, West Timor:
implications for palaeogeography and basin
evolution. Journal of Southeast Asian Earth
Sciences, 6, 359-371.
BMR, 1972. Geology of Papua New Guinea. Scale
1:1,000,000. Bureau of Mineral Resources,
Geology and Geophysics, Canberra.
BREEN, N. A., StaYER, E. A. & ROOF, S. 1989. The Wetar
back-arc thrust belt, Eastern Indonesia: effect of
accretion against an irregularly shaped arc.
Tectonics, 8, 803-820.
CAMERON, N. R. CLARKE, M. G. C., ALDtSS, D. T.,
ASPDEN, J. A. & DJUNUDDIN, A. 1980. The
geological evolution of northern Sumatra.
Proceedings of the Indonesian Petroleum
Association, 9, 149-188.
CHARLTON, T. R. 1986. A plate tectonic model of the
eastern Indonesian collision zone. Nature, 319,
394-396.
--, BARBER, A. J. & BARKHAM, S. T. 1991. The
structural evolution of the Timor collision complex,
The authors are indebted to the Director of the Geological
Research and Development Centre, Bandung, for
permission to publish this paper. H. Faeni and Sudarto
of GRDC, and Steve Moss, Simon Baker and Andy
McCarthy of the University of London assisted in the
preparation of the illustrations. Robert Hall and Derek
Blundell invited TOS to attend the meeting on the
'Tectonic Evolution of Southeast Asia' organized at
the Geological Society, and arranged sponsorship. Robert
Hall, Tim Charlton and Steve Moss made many
suggestions for the improvement of the paper.
eastern Indonesia. Journal of Structural Geology,
13, 489-500.
COFFIELD, D. Q., BERGMAN, S. C., GARRARD, R. A.,
GURITNO, N., ROBINSON,N. M. & TALBOT, J. 1993.
Tectonic and stratigraphic evolution of the Kalosi
PSC area and associated development of a Tertiary
petroleum system, South Sulawesi. Proceedings of
the Indonesian Petroleum Association, 22, 679-706.
COOPER, P. & TAYLOR, B. 1987. Seismotectonics of New
Guinea: a model for arc reversal following arc-
continent collision. Tectonics, 6, 53-67.
CURRAY, J. R. 1989. The Sunda Arc: a model for oblique
plate convergence. Netherlands Journal of Sea
Research, 24, 131-140.
, MOORE, D. G., LAWVER,L. A., EMMEL, F. J., RAITT,
R. W., ET AL 1979. Tectonics of the Andaman Sea
and Burma. In: WATKINS, J., MONTADERT, L. &
DICKENSON, P. W. (eds) Geological and Geophysical
Investigations of Continental Margins. AAPG
Memoir, 29, 189-198.
DARDJI, N., VILLEMIN, T & RAMPNOUX, J. P. 1994.
Palaeostress and strike-slip movements: the
Cimandiri Fault Zone, West Java, Indonesia.
Journal of Southeast Asian Earth Sciences, 9, 3-11.
DAVIDSON, J. W. 1991. The geology and prospectivity
of Buton Island, Southeast Sulawesi, Indonesia.
Proceedings of the Indonesian Petroleum
Association, 20, 209-234.
DE COSTER, G. L. 1974. The geology of Central and
South Sumatra Basins. Proceedings of the
Indonesian Petroleum Association, 3, 77-110.
DE SMET, M. E. M., FORTU1N,A. R., TROELSTRA,S. R.,
VAN MARLE, L. J., KARMINI, M., Er AL S. 1990.
Detection of collision-related vertical movements
in the outer Banda Arc (Timor, Indonesia) using
micropalaeontological data. Journal of Southeast
Asian Earth Sciences, 4, 337-356.
DEWEY, J. F. & BIRD, J. M. 1970. Mountain belts and
the new global tectonics. Journal of Geophysical
Research, 75, 2625-2647.
DJURI & SUDJATMIKO. 1975. Geological Map of the
Majene and western part of the Palopo
Quadrangles, Scale 1:250,000. Geological Survey
of Indonesia.
Dow, D. B. 1977. A geological synthesis of Papua New
Guinea. Bulletin Bureau of Mineral Resources,
Geology and Geophysics, 201, 1-41.
- - & SUKAMTO, R. 1984. Western Irian Jaya: the end
 NEOGENE OROGENIC BELTS OF INDONESIA
product of oblique plate convergence in the late
Tertiary. Tectonophysics, 106, 109-139.
--, ROBINSON, G. P., HARTONO, U. t~ RATMAN, N.
1988. Geology of Irian Jaya: Preliminary
Geological Report. Geological Research and
Development Centre - Bureau of Mineral
Resources, Canberra.
FALVEY, D. A. & MUTTER, J. C. 1981. Regional plate
tectonics and the evolution of Australia's passive
continental margins. Bureau of Mineral Resources
Journal of Australian Geology and Geophysics, 6,
1-24.
FITCH, T. J. 1972. Plate convergence, transcurrent faults
and internal deformation adjacent to Southeast Asia
and the Western Pacific. Journal of Geophysical
Research, 77, 4432-4460.
FORTUIN, A. R., DE SMET, M. E. M., HADIWASASTRA,S.,
VAN MARLE, L. J. TROELSTRA, S. R. &
TJOKROSAPOETRO, S. 1990. Late Cenozoic sedi-
mentary and tectonic history of South Buton,
Indonesia. Journal of Southeast Asian Earth
Sciences, 4, 107-124.
GAGE, M. S. & WING, R. S. 1980. Southeast Asian basin
types versus oil opportunities. Proceedings of the
Indonesian PetroleumAssociation, 9, 123-147.
GARRARD,R. A. SUPANDJONO,J. B. & SURONO,1988. The
geology of the Banggai-Sula Microcontinent,
Eastern Indonesia. Proceedings of the Indonesian
Petroleum Association, 17, 23-52.
GRDC. 1992. Geological Map of Indonesia, Scale
1:5,000,000. Geological Research and Develop-
ment Centre, Bandung.
1995. Geotectonic Map of Indonesia, Scale
1:5,000,000. Geological Research and Develop-
ment Centre, Bandung.
GRIBI, E. A. JR. 1973. Tectonics and oil prospects of the
Moluccas Eastern Indonesia. Bulletin Geological
Society of Malaysia, 6, 11-I6.
HAmE, N. S. 1972. Confirmation of a late Cretaceous age
for granite from Bunguran and Anambas Islands,
Sunda Shelf, Indonesia. Geological Society of
Malaysia Newsletter, 30, 6-8.
HALL, R., ALI, J. R., ANDERSON, C. D. & BAKER, S. J.
1995. Origin and motion history of the Philippine
Sea plate: evidence from eastern Indonesia.
Tectonophysics, 251, 229-250.
HAMILTON, W. 1979. Tectonics of the Indonesian Region.
United States Geological Survey Professional
paper, 1078.
HARDING,T. P. 1983. Divergent wrench fault and negative
flower structure, Andaman Sea. In: BALLY, A. W.
(ed.) Seismic Expression of Structural Styles.
AAPG, Studies in Geology, 15, 4.2, 1-8.
HARRIS, R. A. 1991. Temporal distribution of strain in
the active Banda orogen: a reconciliation of rival
hypotheses. Journal of Southeast Asian Earth
Sciences, 6, 373-386.
HILL, K. C. & GLEADOW, A. J. W. 1989. Uplift and
thermal history of the Papuan Fold Belt, Papua
New Guinea: apatite fission track analysis. AAPG
Bulletin, 75, 857-872.
HUTCHISON,C. S. 1973. Tectonic evolution of Sundaland:
a Phanerozoic synthesis. Bulletin of the Geological
Society of Malaysia, 6, 61-86..
199
1988. Stratigraphic-tectonic model for Eastern
Borneo. Bulletin Geological Society of Malaysia,
22, 135-151.
JOHNSON, R. W. & JAQUES, A. L. 1980. Continent-arc
collision and reversal of arc polarity: new
interpretations from a critical area. Tectonophysics,
63, 111-124.
KARIG, D. E., SUPARKA,MOORE, G. E & HEHANUSSA,P. E.
1978 Structure and Cainozoic evolution of the
Sunda Arc in the Central Sumatra Region. United
Nations, ESCAP-CCOP Technical Bulletin, 12,
87-108.
--, BARBER, A. J., CHARLTON,T. R., KLEMPERER, S
HUSSONG, D. M. 1987. Nature and distribution of
deformation across the Banda Arc-Australia
collision zone at Timor. Bulletin Geological Society
of America, 98, 18-32.
KATILI, J. A. 1971. A review of geotectonic theories and
tectonic map of Indonesia. Earth Science Reviews,
7, 143-163.
1973. On fitting certain geological and geo-
physical features of Indonesian island arcs to the
new global tectonics. In: Coleman, M. (ed.) The
Western Pacific: Island Arcs, Marginal Seas,
Geochemistry. University of Western Australia
Press, 287-305.
KERTAPATI, E. K., SOEHAMI,A. & DJUHANDA,A. 1972.
Seismotectonic Map of Indonesia. Geological
Research and Development Centre, Bandung.
KOESOEMADINATA, R. P., HARDJONO, USNA, I. &
SUMDIRDJA, H. 1978. Tertiary coal basins of
Indonesia. United Nations, ESCAP-CCOR
Technical Bulletin, 12, 43-86.
KONING, T. & AULIA, 1985. Petroleum geology of the
Ombilin intermontane basin, West Sumatra.
Proceedings of the Indonesian Petroleum
Association, 14, 117-137.
KRAUSE, D. C. 1965. Tectonics, marine geology and
bathymetry of the Celebes Sea-Sulu Sea regions.
Bulletin of the Geological Society of America, 17,
813-823.
MCCAFFREY, R. 1991. Slip vectors and stretching
of the Sumatran forearc. Geology, 19, 881-
884.
--, SUTARJO,R., Er AL. 1983 Micro earthquake survey
of the Molucca Sea and Sulawesi, Indonesia.
Bulletin Geological Research and Development
Centre, 7, 13-33.
MINSTER, J. B. & JORDAN, T. H. 1978. Present day
plate motion. Journal of Geophysical Research,
83, 5331-5334.
MILSOM, J., MASSON, D., NICHOLS, G., SIKUMBANG,N.,
DWIYANTO,B., ETAL. 1992. The Manokwari Trough
and the western end of the New Guinea Trench.
Tectonics, 11, 145-153.
MooRE, G. E & KARIG, D. E. 1980. Structural geology of
Nias Island, Indonesia: implications for subduction
zone tectonics. American Journal of Science, 280,
193-223.
MOUNT, V. S. & SUPPE, J. 1992. Present-day stress
orientations adjacent to active strike slip faults:
California, and Sumatra. Journal of Geophysical
Reasearch~ 97, 11 995-12 013.
MUSPER, K. A. E R. 1937. Toelichting bijblad 16 (Lahat).
200 T.O. SIMANDJUNTAK • A. J. BARBER
Geologic Map of Sumatra. Scale 1:2,000,000.
Geological Survey of Indonesia.
PARKINSON, C. D. 1991. The petrology, structure and
geologic history of the metamorphic rocks of
Central Sulawesi, Indonesia. PhD Thesis,
University of London.
~ETERS, P. E. & SUPRIATNA, S. 1990. Geological Map
of West, Central and East Kalimantan Area. Scale
1:1,000,000. Geological Research and Develop-
ment Centre, Bandung.
, PIGRAM, C. J., TRAIL, D. S., DOW, D. B., RATMAN,
N. & SUKAMTO, R. 1983. The stratigraphy of
western Irian Jaya. Bulletin Geological Research
and Development Centre, 8, 14-48.
PIGRAM, C. J. & DAVIES, H. L. 1987. Terranes and
accretion history of the New Guinea orogen. Bureau
of Mineral Resources Journal of Australian
Geology and Geophysics, 10, 193-212.
- - & PANGGABEAN,H. 1984. Rifting of the northern
margin of the Australian continent and the origin
of some microcontinents in Eastern Indonesia.
Tectonophysics, 107, 331-353.
, SURONO & SUPANDJONO,J. B. 1985. Geology and
regional significance of the Sula Platform, Eastern
Indonesia. Bulletin Geological Research and
Development Centre, 11, 1-13.
PRASETYO, H. 1988. Marine geology and tectonic
development of the Banda Sea region, Eastern
Indonesia: a model of an 'Indo-Borderland'
marginal basin. PhD Thesis, University of
California, Santa Cruz.
PUPILLI, M. 1973. Geological evolution of South China
Sea area: tentative reconstruction from border-
land geology and well data. Proceedings of the
Indonesian Petroleum Association, 2, 223-242.
RANGIN, C., DAHRIN, D., QUEBRAL, R. & THE MODEC
SCIENTIFIC PARTY 1996. Collision and strike-slip
faulting in the northern Molucca Sea (Philippines
and Indonesia): preliminary results of a morpho-
tectonic study. This volume.
REHAULT,J. P., MALOD,J. A., LARUE,M., BURHANUDDINN,
S. • SARM1LI,L. 1991. A new sketch of the Central
North Banda Sea, Eastern Indonesia. Journal of
Southeast Asian Earth Sciences, 6, 329-334.
ROSIDI, D. (~ SANTOSO,B. 1976. Geologic Map of Painan
Quadrangle, West Sumatera. Scale: 1:250,000.
Geological Survey of Indonesia.
SIKUMBANG, N. (~z HERYANTO, 1986. Geological Map of
the Banjarmasin Quadrangle, SE Kalimantan.
1:250,000. Geological Research and Development
Centre, Bandung.
SILITONGA, P. H. (~ KASTOWO, D. 1975. Geological Map
of the Solok Quadrangle, Sumatra. 1:250,000.
Geological Survey of Indonesia.
SILVER, E. A. & MOORE, J. C. 1978. The Molucca Sea
Collision Zone, Indonesia. Journal of Geophysical
Research, 83, 169 l - 1691.
& SMITH, R. B. 1983. Comparison of terrane
accretion in modern SE Asia and Mesozoic North
American Cordillera. Geology, 11, 198-202.
SIMUNDJUNTAK, T. O. 1980. Wasaponda Melanges.
Proceedings of the Indonesian Association of
Geologists, 8, 00-00.
-- 1986. Sedimentology and tectonics of the collision
complex in the East Arm of Sulawesi, Indonesia.
PhD Thesis, University of London.
1989. Tunjaman ganda melahirkan jalur gunung
api muda di Minahasa. Suara Pembaruan, 16 June
1989.
1992. Neogene tectonic development of the
Indonesian Archipelago. In: IGCP 246 Conference.
IGCP-IAGI-GRDC, Bandung.
- - 1993a. Tectonic origin of Sumba Island. Journal of
Geology and Mineral Resources, 22, 10-20.
1993b. Neogene tectonics and orogenesis of
Indonesia. Journal of Geology and Mineral
Resources, 20, 2-32.
1994. Tectonic development of the Indonesian
Archipelago and its beating on the occurrence of
hydrocarbons. AAPG Bulletin, 78, 1162.
• SURONO & BUHITRISNA, T. 1981. Geological
Map of West Central and East Kalimantan Area.
Scale 1:1,000,000. Geological Research and
Development Centre, Bandung (Open File).
SITUMORANG,B., GUNTUR,A., YULIHANTO,B., HIMAWAN,
R. & GAMAL JACOB, H. 1991. Structural develop-
ment of the Ombilin Basin, West Sumatra.
Proceedings of the Indonesian Petroleum
Association, 20, 1-16.
SMITH, R. E. 1983. Sedimentology and tectonics of a
Miocene collision complex and overlying Late
Neogene orogenic clastic strata, Buton Island,
Eastern Indonesia. PhD Thesis, University of
California, Santa Cruz.
SNYDER, D. B., PRASETYO,H., BLUNDELL,D. J., FhGRAM,
C. J., BARBER, A. J., ET AL. 1995. Style of crustal
deformation across the Banda Arc continent-arc
collision zone as observed on deep seismic
reflection profiles. Tectonics, in press.
SOPAHELUWAKAN, J 1990. Ophiolite obduction in the
Mutis Complex, Timor Eastern Indonesia: an
example of inverted isobaric medium/high pressure
metamorphism. PhD Thesis, Free University Press,
Amsterdam.
SUKAMTO, R. 1975. Geological Map of Indonesia Sheet
VIII, Ujung Pandang. Scale 1:1,000,000.
Geological Survey of Indonesia, Bandung.
1986. Tectonik Sulawesi Selatan dengan acuan
khusus, ciri-ciri himpunan batuan daerah
Bantimala. Doctoral Thesis, ITB, Bandung.
- - (~ SIMANDJUNTAK,T. O. 1983. Tectonic relation-
ship between geological provinces of Western
Sulawesi, Eastern Sulawesi and Banggai-Sula in
the light of sedimentological aspects. Bulletin
Geological Research and Development Centre, 7,
1-12.
, APANDI, Z., SUPRIATNA, S (~ YASIN, A. 1981. The
geology and tectonics of Halmahera Island and
surrounding areas. In: BARBER, A. J. &
WIRYOSUJONO, S. (eds) The Geology and Tectonics
of Eastern Indonesia. Geological Research and
Development Centre, Bandung, 2, 349-362.
SUPRYANTO (~ IBRAHIM, A. T. M. 1993. Pengembangan
kemampuan penambangan lepas pantai, masa
depan Indonesia. Proceedings of the Indonesian
Association of Geologists, 2, 1162-1174.
 NEOGENE OROGENIC BELTS OF INDONESIA
TAPPONNIER, P., PEWTZER,G., LE DAIN, A. Y., ARM1JO,R.
t~£ COBBOLD, P. 1982. Propagating extrusion
tectonics in Asia: new insights from simple experi-
ments with plasticine. Geology, 10, 611-616.
TATE, R. B. 1991. Cross-border correlation of geological
formations in Sarawak and Kalimantan. Bulletin of
the Geological Society of Malaysia, 28, 63-95.
TILTMAN, C. J. 1990. A structural model for North
Sumatra. In: Lemigas Scientific Contribution on
Petroleum Science and Technology. Special Issue,
24-44.
TONGKUL, F. 1991. Tectonic evolution of Sabah,
Malaysia. Journal of Southeast Asian Earth
Sciences, 6, 395-405.
VAN BEMMELEN,R. W. 1949. The Geology of Indonesia.
Government Printing Office, The Hague
VEEVERS, J. J. 1984. The Phanerozoic Earth History of
Australia. Oxford University Press.
201
VISSER, W. A. & HERMES,J. J. 1962. Geological results of
the exploration for oil in Netherlands New Guinea.
Verhandelingen Koninklijk Nederlands Geologisch
en Mijnbouwkundig Genootschap, Geologische
Serie, 20, 1-265.
WAJZER, M. R., BARBER, A. J., HIDAYAT,S. & SUHARSONO
1991. Accretion, collision and strike-slip faulting:
the Woyla Group as a key to the tectonic evolution
of North Sumatra. Journal of Southeast Asian Earth
Sciences, 6, 447--463.
WANG, J. E, BUDIJANTO, E M., JOHNSON, M. I. &
SIRINGORIN60, R. S. A. 1989. Neogene seismic
sequence and structural styles in 'B' and Peusangan
Blocks, North Sumatra Basin, Indonesia.
Proceedings of the Indonesian Petroleum
Association, 18, 339-361.
 Palaeomagnetism of the Sibumasu and Indochina blocks:
implications for the extrusion tectonic model

BRYAN RICHTER & MICHAEL FULLER

Department of Geological Sciences, University of California,
Santa Barbara CA 93106, USA. email: richter@magic.geol.ucsb.edu

Abstract: The Jurassic-Cretaceous Kalaw redbeds of Myanmar yield a prefolding Late

Cretaceous to Early Palaeogene magnetization which records c. 25-30 ° of clockwise (CW)
rotation relative to the South China block. This corresponds to 10-15 ° CW relative to the Lower
Cretaceous Khorat Plateau VGP. The data also show 5 ° of northward transport relative to the
100 Ma South China VGP or 12° relative to the Khorat Lower Cretaceous VGP. Similar CW
rotations are measured in remagnetized Palaeozoic carbonates in Peninsular Thailand and
Langkawi Island, Malaysia. These block motions most likely took place between the Late
Cretaceous and the Late Oligocene. These and other recently published data have several
implications for the extrusion tectonic model: (i) Sundaland has only rotated 25-30 ° CW
relative to South China during the Tertiary; (ii) southeastward translation is only 300-500 kin;
and (iii) Sundaland is composed of smaller sub-blocks, some of which have moved northward.
This is interpreted to indicate that deformation of the Sibumasu block is dominated by the oblique
subduction of the Indian Ocean Plate while deformation of the Indochina block is dominated by
extrusion, in turn driven by convergence between the Indian Craton and Eurasia.

SE Asia has been located at the intersection of
the Eurasian, Indian, Australian, Pacific and
Philippine Sea plates throughout much of the
Cenozoic (e.g. Hamilton 1979). These plate inter-
actions culminated in the Palaeogene to present
collision of India with Tibet and the Neogene to
present collision of Australia with Sulawesi, Java
and Borneo. The boundary conditions set up by the
relative motions between these plates have had a
profound impact upon the Cenozoic motions of the
terranes and micro-plates which comprise SE Asia.
Convergence between the Indo-Australia and
Eurasian plates continues today and ultimately the
terranes of SE Asia will form a broad orogenic belt
between these two plates.
It has become clear over the last decade that
SE Asia has moved with respect to Eurasia during
the Cenozoic. Field data which document these
motions have been difficult to obtain and often
more difficult to interpret. Thus, much of what is
postulated about Cenozoic motions is derived from
laboratory experiments (e.g. Tapponnier et al.
1986) suggesting that much of the deformation
associated with the India-Eurasia collision has
been accommodated by rigid plate-wide clockwise
(CW) rotation and southeastward translation of the
Indochina and Sibumasu blocks out of the collision
zone.
The available Mesozoic and Cenozoic palaeo-
magnetic data from the Indochina and Sibumasu
blocks qualitatively support the predictions of these
laboratory models. However, ambiguities regarding
the tectonic significance of these palaeomagnetic
data remain. Many of the tectonic conclusions are
based upon several key sampling localities or well
preserved chronostratigraphic successions and very
little is known about the intervening regions or time
periods. With regard to the actual palaeomagnetic
analyses, it has proved difficult to discriminate
between primary and secondary magnetizations. A
review of the literature on the Khorat Plateau, for
example, shows the same basic dataset interpreted
first as a primary magnetization, reinterpreted as a
secondary magnetization, then finally reinterpreted
as a primary magnetization (Achache et al. 1983;
Chen & Courtillot 1989; Yang & Besse 1993). It
is imperative, therefore, to reevaluate existing
datasets as new data become available.
Despite these difficulties, palaeomagnetism
remains one of the few direct records of ancient
plate motions available to us. Towards this end
we review the available palaeomagnetic data from
Sundaland and we present new palaeomagnetic
data collected from the Kalaw Redbeds of the Shan
Plateau of Myanmar and from Palaeozoic lime-
stones located in Central and Peninsular Thailand.
Moreover, we evaluate the tectonic significance of
these palaeomagnetic directions. For example, are
they caused by local or plate-wide rotations? Are
they controlled by structural deformation? Finally,
we conclude by evaluating existing tectonic models
and propose modifications which make the models

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, 203
Geological Society Special Publication No. 106, pp. 203-224.
204 B. RICHTER • M. FULLER
more consistent with the measured rotations and
translations.
Regional tectonic framework
The data presented in this paper have been
collected in Thailand and Myanmar from the
Tertiary geological 'plate' known as Sundaland
(Fig. 1). Sundaland is composed of several
Palaeozoic and Lower Mesozoic terranes (Fig. 2).
The origin of these blocks and their suturing ages
have, among others, been studied by Metcalfe
(1988, 1990). Hutchison (1989) provides an
excellent summary of additional studies.
The new data presented in this study were
collected from the elongate N-S trending Sibumasu
block (Shan-Thai). In our discussion we compare
these data to published data from the Indochina
block (Fig. 2).
Sibumasu block
The Sibumasu block consists of portions of
Myanmar, Thailand, northwest Malaysia, and
Sumatra (Fig. 2). It is bounded to the west by the
dextral Sagaing Fault, the Andaman Sea and
the dextral Sumatran Fault Zone; to the east by the
Bentong Raub line, the extension of the Bentong-
Raub line into the Gulf of Thailand, and finally
by the Nan-Uttaradit Ophiolite line in northern
30°N
20°N
10 ° N
90 * 100 ° 110 ° 120 °
Fig. 1. Regional geography of Southeast Asia. Light
circles show location of sites discussed in this study and
heavy circles indicate the sites of Yang & Besse (1993).
Thailand; and to the north by the Red River Fault
and the Himalayan Syntaxis (Metcalfe 1988). It
may have originally been continuous with the
Qiangtang or Lhasa blocks but intense deformation
makes it difficult to confirm this.
There is some agreement that the long axis of the
Sibumasu block was originally orientated approxi-
mately E-W and was attached to the northern
margin of Gondwanaland during the Early
Palaeozoic (e.g. Hutchison 1989). It rifted from
Gondwanaland in the Middle or Late Palaeozoic
and began travelling northward. Collision timing
along the Bentong-Raub line in Peninsular
Malaysia is widely thought to be marked by the
Late Triassic S-type granites of the Main Range
Province (e.g. Cobbing et al. 1986). In contrast,
Helmcke (1986), Barr & Macdonald (1991), and
Barr et al. (1990) have shown that the proposed
suture zones in Northern Thailand (Nan River Belt,
Phetchabun Fold belt) have a more complicated
collisional history which, although it includes the
Late Triassic deformation seen elsewhere, began in
the Upper or even Middle Palaeozoic.
These contrasting collisional ages can be
explained in several ways. (a) The Nan Suture may
not correlate directly with the Bentong Raub suture
and the Sibumasu block may have been broken into
northern and southern halves during the Late
Palaeozoic; (b) there may have been post-collision
lateral translations along these suture zones which
have obscured the original collisional relationships;
(c) the northern part of the Sibumasu block may
have collided with northern Indochina in the Late
Palaeozoic but the collision did not go to com-
pletion, forming a partially amalgamated super-
terrane with a small trapped oceanic basin which
finally closed in the Late Triassic; or (d) one or
more marginal 'back-arc' basins may have formed
and then been consumed prior to the final
collisional episode in the Late Triassic. The last
two hypotheses appear the most likely. Fortunately,
almost all researchers agree that the major blocks
which comprise the core of SE Asia had assembled
and become sutured to South China (Eurasia) by
the Late Triassic-Early Jurassic (Hutchison 1989
and references within).
Indochina block
The Indochina block is bounded to the north by the
Red River Fault (Song-Ma and Song-Da suture
zone) and to the east by the South China Sea
marginal basin (Fig. 2). The southern boundary
is poorly documented and it is not clear if the
Indochina block has always been continuous with
the Borneo block, which has similar Mesozoic
igneous sequences, or if there is an E-W trending
Late Palaeozoic suture zone near the Natuna and
 SIBUMASU BLOCK INDOCHINA BLOCK
Langkawi- Peninsular Shan Plateau, West Khorat Loei-Petchabun
NW Malaysia Thailand Myanmar Kampuchea Plateau Fold Belt
-- -- TI-1I- . . . . . . . . . . . . ~Kala-w ~~ng ~ "--' J . ~ ---- UEvap°rite ~ v~-~Orite
r.~-'-q Redbeds o[~~a--.~ K '5:-.'.5'F:Gres
Tembelinq_ t:-"--':-~n" L'L~a~'°~\~" South Jfl J :::::::::::::::::::::
Superieurs t~'~J ~~ ~ _ _
~ . ~ Group ~uZp~n "L:-':-':':I~--
[-':'[-'-'iLuang
Redbeds [~-.'-:-:--Lu ang I"' ~ _ ~ ~%:":"~..~ ,ndo- ~ "-'-'-'" ~a
. . . .
[ ~ F_.m
(~_F_m...=. .__ ~Nam
. . . . . . ( f ~t~'!, ,~ ~ C h i n a . [:-::---:lPhong Fro.
-.-.---Semanggol ~ -. V /~"% .-"P~%, E,st ~ H u a i Hin
Tr ~ Fm. -:.:......
--'.'." ::cCuupZg-E ~ Ratburi "'"'
" i:::::i,,s :i :i',,,s ,,',',
_ _ ..~ _ _ _ - _ L/S
',:i.:..:..-Kaen_g
_~ ~ ~ , ~ . ~T ~ ~ i~ i~ ~es~sph'~n :'"'"' nSai Fro.
C ~_~-~ga ~i~$il Krachan [ ~ -Le~yyin %" &~ ~ ;,','~ ;Saraburi L/S
Orouo
I "-'- Pindaya Khao Luak
[' "Y'.'"| Fm.
l) /s_
Fm. g i', ",", ~ Chert

Diamictite ~ Carbonate~ Marine ~Continental ]~E-] Chert ~ Congl. [-~T] Volcani- ~ Evaporite rTFTT]Stratigraphic
(Glacio- Shale, Redbeds clastics Break
Marine?) Mudstone (S/S +Cong.)
Fig. 2. Tectonostratigraphicterrane summaryfor SoutheastAsia showingPalaeozoicterrane boundaries and regional stratigraphiccorrelations.Tertiarystrata in peninsular
Thailand and the Shan Plateau of Myanmar are only locallypreserved in small Oligo-Mioceneintermontanebasins. Diagrammodified from Metcalfe(1988).

O
206 B. RICHTER & M. FULLER
Mekong basins. The western boundary has
traditionally been interpreted to be the Nan-
Uttaradit suture zone. This suture is thought to
extend southward, beneath the Central Thai Basin,
into the Gulf of Thailand and ultimately to the
Late Triassic Bentong-Raub line in Malaysia
(Metcalfe 1988; Hutchison 1989). In contrast, Barr
& Macdonald (1991) conclude that the Nan suture
actually marks a Permian suture between the
Indochina block and the recently identified
Sukhothai terrane. They place the Late Triassic
suture zone farther west, along the eastern margin
of the Northern Thai granite province. Unfortu-
nately, the Late Palaeogene to Neogene Central
Thai basin and the Pattani trough almost com-
pletely cover the suture zone from 16°N to 7°N and
it is not yet possible to correlate definitively sutures
along the length of the Indochina and Sibumasu
blocks.
Precambrian basement is exposed in the Kontum
Massif of Vietnam and presumably forms the base-
ment of the entire Indochina block. It may have
originated on Gondwanaland, possibly along the
eastern margin of Australia, but must have rifted
away by the Early Carboniferous (Metcalfe 1988).
Moderately to strongly deformed Palaeozoic sedi-
ments are widely distributed but the block is
dominated by thick, relatively undeformed
Mesozoic continental sediments of the post-
Indosinian Khorat Basin. Most palaeomagnetic
studies in the Khorat Plateau are concentrated in
this Mesozoic section.
Previous palaeomagnetic studies
Bunopas (1982) examined the palaeomagnetism
of a wide variety of rock types in Palaeozoic and
Mesozoic strata throughout both the Indochina and
Sibumasu terranes using orientated hand samples.
His reconnaissance study formed the foundation for
additional studies on Neogene basalts in Northern
and Central Thailand (Barr et al. 1976; Barr &
MacDonald 1979; McCabe et al. 1988); upon
Upper Palaeogene to Neogene intermontane basins
throughout Thailand (Richter et al. 1993); and upon
Upper Palaeozoic and Mesozoic continental strata
from the Khorat Plateau (Barr et al. 1978; Achache
et al. 1983; Achache & Courtillot 1985; Maranate
& Vella 1986; Chen & Courtillot 1989; Yang &
Besse 1993).
Upper Palaeozoic and Mesozoic o f the
Khorat Plateau
As noted above, the majority of palaeomagnetic
data has been collected from Upper Palaeozoic and
Mesozoic sediments of the Khorat Plateau. Many
of these early studies failed to recognize the
potential significance of secondary magnetizations
and their data are only presented in tectonically
corrected coordinates. Yang & Besse (1993) further
note that many of the demagnetization procedures
may not have been sufficient for discriminating
between primary and secondary magnetizations.
Despite these concerns, the consistency between
these studies is remarkable. Mean values of c. 30-
40 ° of CW deflection with positive inclinations
of 300-40 ° are found in almost all of the Khorat
studies. Furthermore, structural dips are low and
corrected and uncorrected directions commonly
overlap within their respective 0~95 error estimates.
The primary concern with these earlier data is
distinguishing between primary and secondary
magnetizations and then assigning an age to that
magnetization.
Yang & Besse (1993) present new data from
Upper Permian limestones, the Upper Triassic Huai
Hin Lat formation, the Lower Jurassic Nam-Phong
formation, the Upper Jurassic Sao Khua Formation,
and the Lower Cretaceous Khok Kruat Formation.
These data are of high enough resolution to allow
an approximate chronology of magnetizations and
rotations to be reconstructed. Their results are
reviewed below and summarized in Fig. 3 and
Table 2.
The Permian Limestone samples yield a high
coercivity CW deflected component which fails
a fold test at the 95% level (Fig. 3). Synfolding
analysis indicates that kappa is at a maximum at
30% unfolding and thus the magnetization is best
interpreted as a synfolding remagnetization. At this
locality, Upper Triassic continental sediments
unconformably overlie steeply dipping, strongly
deformed Permian units. Thus, the Late Triassic
Indosinian orogeny is a strong candidate for the
folding and initial remagnetization.
The Upper Triassic Huai Hin Lat formation also
gives a CW deflected direction. Based upon the
interpretation of a positive fold test in the overlying
Lower Jurassic Nam-Phong Fm they conclude that
this is a primary magnetization and they present a
tectonically corrected Late Triassic VGP at 52.1 °N,
169.8°E (o~95=7.3) (Table 2). This VGP is
indistinguishable from the remagnetized Permian
VGP.
The Lower Jurassic Nam-Phong formation also
yields a CW deflected magnetization. This is found
in both positive and reversed polarities and passes a
fold test at the 95% level (Table 2).The tectonically
corrected VGP for this unit is located at 54.4°N,
175.6°E ((x95 =4.9) and is statistically indistin-
guishable from the Permian and Late Triassic
VGPs. However, Yang & Besse (1993) caution that
the folding may have occurred anytime between the
mid-late Yanshanian orogeny and the Late Tertiary.
 PALAEOMAGNETISM OF THE SIBUMASU BLOCK
N The Aptian-Albian Khok Kruat Formation
N 26.0 ° and 35.3 ° , 29.8 ° respectively). Bedding
90 °
Ju and KI Redbeds
Khorat Plateau, Thailand
N Mesozoic of peninsular Thailand and
90°
+ In Situ
++ + + +++++1-90 °
Permian Carbonates
Khorat Plateau, Thailand
Fig. 3. Lower hemisphere equal area projections of site
mean directions measured in the Khorat Plateau by Yang
& Besse (1993). Solid symbols denote positive polarities
while open symbols denote negative polarities. Ellipses
are the 95% confidence interval about the mean
(alpha95) but these have been omitted from the Ju + K1
plot for clarity. See Table 1 for exact values and text
for discussion.
We do not follow them in assigning an Early
Jurassic age to this pole because the age of the
folding is so poorly understood. The interpretation
of a positive fold test is reasonable, but without
good age constraints on the age of folding we can
only say that this is a pre-folding VGP with a
maximum age of Early Jurassic.
The Upper Jurassic Sao Khua Formation yields
a high coercivity CW deflected direction (Table 2
and Fig. 3). This unit is poorly exposed and dips are
very low. As such, fold tests are inconclusive.
Polarity may be a useful argument in that one
would expect to see an even distribution between
normal and reversed polarities if the magnetization
truly formed in the Late Jurassic. The lack of a
conclusive fold test and the preponderance of
normal polarities leaves open the possibility that
this magnetization was acquired during the
Cretaceous long normal period.
207
yields medium and high coercivity components
which are quite close to each other (I, D = 37.2 °,
corrections are all less than 12° and the fold test is
inconclusive. Despite this, Yang & Besse (1993)
conclude that this is a primary Lower Cretaceous
VGP. As above, we urge caution in assuming at
Early Cretaceous age. The small ~95' the incon-
clusive fold test, and the lack of reversed polarities
leave open the possibility of a younger re-
magnetization event.
peninsular Malaysia
The southernmost extent of the CW rotations
described above is presently not clear and it is
useful to review the available palaeomagnetic data
from peninsular Malaysia. The principal work from
peninsular Malaysia is by McElhinny et al. (1974).
They detected two distinct populations of palaeo-
magnetic data: (1) the Pengerang rhyolites,
Singapore gabbro and dykes, Bentong Group,
Singa Formation and Sempah Fm. displayed
inclinations of 20-40 ° with CW rotated declina-
tions of 20-33°; and (2) the Kuantan-Massai dykes
and the Segamat basalts displayed inclinations of
30-40 ° with counter-clockwise (CCW) rotated
declinations of 40 °. Haile & Khoo (1980) detected
20-30 ° of CCW rotation in the Maran, Teka and
Kluang redbeds of Malaysia while Schmidtke
et al. (1990) confirmed the CCW rotated declina-
tions in mafic dikes at Kuantan and mafic flows
at Segamat.
McElhinny et al. (1974) argued that either the
poles or the field directions of three of their four
CW rotated sites passed a fold test. They estimated
the age of folding to be Late Triassic and thus
their CW rotated magnetization was assigned a pre-
Late Triassic age. In the context of new data and
statistical analyses, however, it is not clear that the
CW rotation is a pre-folding magnetization and
even if it is, it could be as young as Late
Cretaceous.
There are additional unpublished analyses con-
ducted at the UCSB palaeomagnetism laboratory
which bear on the review of Malaysian data.
Schmidtke (pers. comm. 1991) measured CCW
deflections of similar magnitude (30-45 ° ) in
samples collected from the poorly dated Singapore
and Lanchang dikes, the Jurassic-Cretaceous
Raub redbeds, and the Permo-Triassic Kodiang
Limestone. The Bukit Kemaman, Bukit Temiang,
Lanchang and Tanjung Penyabong dykes display
between 6 ° and 23 ° of CCW declination deflection
while sites from the Mesozoic Raub and
T a b l e 1. S u m m a r y o f the n e w p a l e o m a g n e t i c d a t a p r e s e n t e d in this p a p e r
to
Site Location N in situ in situ in situ in situ in situ Corr. Corr. Corr. Corr. Corr. in situ Corr. oo
(samples) I D tx95 k A ° Lat. I D ct95 k A ° Lat. VGP VGP

Juro-Cretaceous Kalaw redbeds of Shan Plateau (Shan-Thai)

B91-8 N20°48.37" E96°26.98 ' 7 -10.2 58.5 9.2 54.5 26.0 17.7 58.8 - - 11.7
B91-10 N20°47.98" E96°27.14 ' 5 13.8 24.1 11.9 42.0 13.8 9.4 28.0 - - 16.1
B91-11 N20°47.98 ' E96°27.14" 5 21.9 27.2 17.8 19.4 9.4 30.7 32.5 - - 4.3
B91-17 N20°38" E96°34 ' 8 22.8 42.9 8.5 51.1 8.8 30.4 50.6 - - 4.3
B91-18 N20°38.95 ' E96°33.02" 6 36.6 38.3 11.1 48.8 0.3 24.8 43.9 - - 7.6
B91-19 N20°42.24 ' E96°38.52 ' 7 77.0 352.9 12.4 24.8 --44.5 7.1 42.4 - - 17.1
B91-20 N20°45 ' E96°32" 8 -0.2 40.1 17.1 11.5 20.9 37.1 39.6 - - 0
B92-54 N20°47.06 ' E96°27.83 ' 4 -22.1 54.3 11.9 60.8 32.3 19.2 44.9 - - 10.9
B92-55 N20°45.40 ' E96°28.78 ' 5 27.9 40.4 11.5 45.1 5.9 29.1 45.8 - - 5.2
B92-57 N20°41.75 ' E96°30.16 ' 7 14.3 41.7 8.0 58.5 13.4 25.7 43.1 - - 7.2
B92-58 N20°42.24 ' E96°38.52 ' 5 -34.0 48.0 7.1 115.7 39.3 24.5 50.4 - - 7.9
B92-60 N20°38.27 ' E96°33.43 ' 6 21,6 44.8 10.3 43.3 9.4 28.4 51.1 - - 5.5
B92-81 N20°42.31 ' E96°30.55 ' 7 70.7 341.1 8.3 53.9 -34.3 15.3 40.4 - - 12.9
Mean N20°42 ' E96°33 ' 80 18.6 39.3 19.1 5.7 11.2 23.4 44.7 6.1 47.1 8.5 N50.6 ° E197.0 ° N46.4 ° E190.6 °

Late Palaeozoic Plateau Limestone, Shan Plateau, Myanmar

B 9 1 - 1 2 + 14 N20°48.89" E96°44.039 ' 6 17.6 317.3 14.9 21.2 11.7 9.5 311.6 - - 15.9 N47.3 ° E 3 5 7 . 4 ° N40.4 ° E354.9 °
B91-13 N20°40.89 ' E96°44.039 ' 3 26.5 32.1 24.1 27.3 6.7 45.7 21.3 - - -6.4 N58.7 ° E 1 9 3 . 8 ° N69.5 ° E164.3 °

Permian Ratburi Limestone, Peninsular Thailand

T90-10 N7°55 ' E99°55 ' 7 38.0 43.2 3.1 379.2 -13.4 35.4 66.4 - - -11.6 N46.3 ° E 1 6 7 . 2 ° N24.8 ° E172.0 °
T90-27 N10.95 ° E99.28 ° 9 46.3 48.3 11.4 21.3 -16.7 59.4 31.1 - - -29.3 N41.8 ° E161.9 ° N49.9 ° E137 °
T90-45 N9.9 ° E99.1 ° 7 37.6 27.1 8.8 48.5 -11.2 19.8 35.8 - - -0.3
T91-39 N7°36.43 ' E100°2.04 ' 6 22.6 37.2 18.9 13,5 -4.7 ~).3 42.7 - - 7.2 N53.1 ° E 1 8 0 . 6 ° N46.7 ° E 1 9 8 . 4 °

Permo-Triassic Lampang Limestone, Central Thailand

T91-8 N18°15.75 ' E99°36.92 ' 4 40.3 75.2 14.1 43.7 -4.7 67.9 142.4 - - -32.7 N20.2 ° E171.2 ° N13.4 ° E122.9 °
T91-9 N18°15.75 ' E99°36.92 ' 4 40.7 50.5 10.3 80.9 -5.0 62.7 31.8 - - -25.8 N42.7 ° E174.5 ° N52.9 ° E138.5 °
T91-10 N18°15.75" E99°36.92 ' 7 40.7 56.7 11.5 28.5 -5.0 -35.5 61.1 - - 37.9 N37.1 ° E173.8 ° N58.8 ° E12.4 °
T 9 1 - 8 , 9, 10 N18°15.75 ' E99°36.92 ' 15 41.0 60.8 14.9 69.3 -5.2 44.0 67.7 180.0 1.6 -7.5 N33.4 ° E173.0 ° N67.7 ° E27.4 °

Silurian Setul Limestone, Langkawi Island, Malaysia

SM-88-62 N6.2 ° E99.8 ° 9 19.2 33.5 4.1 157.0 -3.7 -10.9 35.6 - - 11.7 N56.7 ° E181.3 ° N 5 2 . 6 ° E207.3 °
SM-88-63 N6.2 ° E99.8 ° 5 17.8 36.1 7.5 105.0 -2.9 36.8 24.4 - - -14.3 N54.1 ° E 1 8 2 . 8 ° N 6 2 . 4 ° E156.3 °
SM88-62, 63 N6.2 ° E99.8 ° 14 18.5 34.8 6.2 1634.0 -3.3 13.0 30.6 150.2 6.0 -0.4 N55.4 ° E182.1 ° N59.6 ° E187.3 °

Limestone VGP mean 9 - - 7.9 59.9 - - - 21.8 8.6 - N49.2 ° E172.4 ° N60.5 ° E164.3 °

I, i n c l i n a t i o n (°); D, D e c l i n a t i o n (°); ¢~95, t h e 9 5 % c o n f i d e n c e i n t e r v a l a b o u t t h e m e a n d i r e c t i o n ; k, t h e p r e c i s i o n p a r a m e t e r k a p p a ; A L a t , p r e s e n t l a t i t u d e m i n u s p a l a e o l a t i t u d e , w i t h

n o r t h w a r d t r a n s l a t i o n s as p o s i t i v e a n d s o u t h w a r d t r a n s l a t i o n s as n e g a t i v e ; in s i t u , g e o g r a p h i c a l c o o r d i n a t e s ; C o r r . , s t r u c t u r a l l y c o r r e c t e d c o o r d i n a t e s . B o l d t y p e i n d i c a t e s m e a n
values a n d italics indicate the p r e f e r r e d direction b a s e d u p o n interpretation o f fold tests.
Table 2. Summary of published palaeomagnetic data

Age Formation N in situ in situ in situ in situ in situ Corr. Corr. Corr. Corr. Corr. in situ Corr.
(samples) I D ct95 k A° Lat. Inc. Dec. c~95 k A ° Lat. VGP VGP

Mesozoic of Western and Peninsular Thailand (Shan-Thai)

Mesozoic Barr et al. (1978) 19 . . . . . 38.8 35.6 12.0 31.7 -4.3 - N56.3 ° E!76.6 °
Mesozoic Barretal. (1978) 5 13.8 24.1 11.9 42.0 13.8 32.8 34.2 9.6 11.8 -0.3 - N57.5° E184.0 ° >
©
Mesozoic of Khorat Plateau (Indochina Block) 0 Chen & Courtillot (1989) $ Yang & Besse (1993) E
>
0 J + Tr 69 36.2 40.8 15.9 19.0 -3.1 50.0 43.6 14.1 23.0 -13.8 N51.3 ° E181.9 ° N48.1 ° E165.6 °
0 J + Tr 64 37.2 32.3 7.0 120.0 -3.8 46.9 36.4 9.3 70.0 -11.1 N59.3 ° E180.8 ° N54.8 ° E168.1 ° Z
$ Upper Triassic Huai Hin Lat F m 56 31.7 36.9 13.3 33.8 -0.5 44.4 39.5 8.3 85.0 -9.4 N54.7 ° E185.5 ° N52.1 ° E169.8 °
$ Lower Triassic Nam-Phong F m 61 25.4 39.0 15.3 14.0 3.3 40.1 37.2 6.6 71.6 -6.2 N52.2 ° E191.4 ° N54.4 ° E175.6 ° E
$ Tru + J1 Composite 117 27.9 38.2 9.8 18.8 41.8 38.1 4.7 77.5 -7.4 N53.2 ° E189.2 ° N53.6 ° E173.3 ° ©
$ Upper Jurassic Sao Khua F m 73 38.1 27.1 3.1 247.4 -4.8 37.3 26.6 2.6 339.2 -4.3 not given N64.8 ° E175.6 °
$ Lower Cretaceous Khok Kruat Fm 88 37.2 26.0 2.5 388.5 -4.6 40.5 28.1 2.4 398.5 -7.0 not given N62.7 ° E173.3 ° ~Z

Palaeozoic of Khorat Plateau (Indochina Block)

$ Upper Permian 9 39.0 24.8 12.6 17.6 -5.4 31.8 51.4 12.6 17.6 -0.6 N66.0 ° E174.9 ° N58.45 ° E176.15 °
$ Upper Permian 9 39.4 33.0 6.0 75.4 -5.7 39.4 33.0 6.0 75.4 -5.7 N48.5 ° E176.15 °
0 Permian Khorat Group 27 33.4 25.5 14.4 42.0 -1.2 31.4 60.1 15.4 36.0 0.0 N65.7 ° E186.2 ° N32.8 ° E183.4 °
0 Permian Khorat Group 25 37.9 27.6 15,0 68.0 -4.3 31.7 54.7 18.5 45.0 -0.2 N63.6 ° E179.3 ° N37.9 ° E184.2 °
©
<) Devonian pre-Khorat Group 3 46.2 39.5 11.3 120.0 -10.5 -12.1 62.0 11.3 120.0 23.1 N52.1 ° E169.8 ° N24.5 ° E208.2 °
0 Devonian pre-Khorat Group 3 -39.4 183.6 12.5 99.0 -5.3 8.0 192.6 31.6 16.0 21.0 N83.7 ° E134.9 ° N65.6 ° E251.3 °
OD+OP+OTr 102 38.2 29.7 3.1 22.0 -4.5 40.2 '~ 44.4 4.3 11.0 -5.9 N61.6 ° E179.1 ° N48.0 ° E177.4 °
OD+OP+OTr 13 38.3 32.3 8.1 25.0 -4.5 31.6 43.9 17.1 6.0 -0.1 N59.2 ° E179.3 ° N48.1 ° E186.1 °

F o r m a t is the s a m e as Table 1.

tO
210 B. RICHTER & M. FULLER
Tembeling redbeds reveal a broad spectrum of
directions between the CW and CCW rotated end
members identified by McElhinny et al. (1974).
These data have never been published because
it has been difficult to evaluate the age of the
magnetization and the influence of structural
deformation and present field chemical weathering
upon the magnetic signal. These unpublished data
are presently being reevaluated.
To summarize, the domain of CW rotations
observed in Mesozoic rocks of the Shan and Khorat
Plateaus only extends as far south as Langkawi
Island, Malaysia. South of this the Mesozoic
section shows a wide spectrum of CCW and CW
rotated directions. The CW end members are very
similar to the Shan/Khorat directions while the
CCW end members are very similar to directions
found in Borneo (Schmidtke et al. 1990). Thus,
it appears that peninsular Malaysia is a transition
zone between a CW rotating Indochina-Sibumasu
block and a CCW rotating Borneo block.
Cenozoic of Thailand
McCabe et al. (1988, 1993), building upon earlier
work by Barr et al. (1976) and Barr & Macdonald
(1979), detected small CW rotations in Neogene
basalts from northern and western Thailand but
detected no rotation of Neogene basalts located
on the Khorat Plateau. These basalts are part of a
tholeiitic to alkalic suite of Late Cenozoic basalts
that are exposed throughout the Indochina block
(Barr & Macdonald 1981). They conclude that
volcanic activity began by at least 12 Ma and that
there is no systematic relationship between the
age and geographical distribution of volcanism.
McCabe et al. (1988) initially interpreted the data
as a Neogene CW rotation of western Thailand
relative to both Eurasia and the Khorat Plateau
while McCabe et al. (1993) reinterpreted the data
as local block rotations driven by intracontinental
deformation.
Richter et al. (1993) presented reconnaissance
data from several Neogene intermontane basins in
Thailand. The Krabi and Nong Ya Pong basins of
peninsular Thailand show small CW declinations
comparable to those measured in Neogene basalts.
In contrast, the reconnaissance data from the Mae
Moh basin of north-central Thailand shows pre-
dominantly small CCW declinations. Although it
is possible that a 15 ° rigid body CW rotation of
peninsular Thailand took place during the Middle
to Late Neogene it is very unlikely that a 20 ° CCW
rotation of north-central Thailand took place
concurrently. Thus, local rotations are a strong
possibility.
In contrast to McCabe et al. (1993), however,
we conclude that the dispersion in Neogene palaeo-
magnetic data does not demonstrate that chaotic
Neogene local rotations have obscured all
Mesozoic motions. There is demonstrable con-
sistency in Mesozoic data collected over a large
region while all Neogene data have been collected
from small discontinuous exposures which, in
many cases, are structurally bounded. Thus, it
appears that regional block motions are preserved
in the Khorat and Shan Plateaus while data
collected from the fold belts which bound these
plateaus and from the Central Thai basin are
probably complicated by Neogene strike-slip
faulting.
Geology of palaeomagnetic sampling sites
A complete review of the regional stratigraphy of
the Sibumasu and Indochina blocks is beyond the
scope of this paper but regional stratigraphic
sections are presented in Fig. 2. The reader is
referred to Hutchison (1989) for regional sum-
maries; to Bunopas (1982), Snansieng et al. (1983)
and Nutalaya et al. (1983) for Thailand; and
Chibber (1934) and Bender (1983) for Burma for
a more detailed stratigraphic discussion.
M e s o z o i c continental s e d i m e n t s o f
Shan Plateau, M y a n m a r
The Kalaw redbeds are an extensive sequence of
continental sediments which are exposed along the
western margin of the Shan Plateau east of the Pan
Laung and Shan Boundary Faults (Fig. 4). They lie
near the northern end of a linear belt of Upper
Mesozoic continental deposits which include the
Raub and Tembling redbeds of peninsular Malaysia
and the Chumpon and Kanchanaburi redbeds of
Thailand (Metcalfe 1988; Bender 1983). Mitchell
(1992) indicates that they are coeval with the lower
portion of the Khorat Plateau redbed sequence.
Diagnostic fossils within the Kalaw sequence
itself are lacking and the age is poorly constrained
(Win Swe, pers. comm. 1992). The redbeds un-
conformably overlie the Lower-Middle Jurassic
Loi-An (Pan Laung) Series and are unconformably
overlain by Neogene to Quaternary non-marine
sediments (Bender 1983). Most authors agree that
the sequence ranges from Middle or Upper Jurassic
to Cretaceous (Bender 1983; Wolfart et al. 1984;
Myanmar Oil and Gas Enterprise, unpublished
geological map) but it must be emphasized that
geological mapping and dating within the Shan
Plateau are still at a reconnaissance stage.
The Kalaw sequence is moderately to strongly
folded about sub-horizontal NNW-SSE trending
axes. Folding is commonly thought to have taken
place during the Middle to Upper Cretaceous
 PALAEOMAGNETISM OF THE SIBUMASU BLOCK 211

95°E 15' 30' 45' 96°E 15' 30' 45' 97OE

21ON -- Mt. Popa / I \

Chauk ~ ~
tO Meiktilla ~ha
- _ . _ _ _ _.. , , o Taunggyi
45' a Q
Kalaw Pz
Exolanation "~
~k
30'
Q QuaternaryCover ~/\

Pt
Pz
_ _ _ ~ NeogeneBasalt "~/~
15'
_ ~ Cretaceousto Lower l/
_ Paleogene Intrusives ~, )lnle
20ON
_

Jurassic-Cretaceous
Kalaw Redbeds
45' Pz PaleozoicCarbonates
Mogok Series Gneiss, Marble, Schist~
30'

15'
fJ
Chaung Magyi Group
Faults (air photo lineaments)
\
/ Roads
19ON
0 20 40 60 80 100 km

Fig. 4. Geological sketch map of the western edge of the Shan Plateau at the latitude of the Kalaw redbeds (modified
from Bender 1983). Samples were collected at c. 400 m (0.25 mile) intervals along the Thazi-Kalaw road (dashed
line). The dextral Sagaing fault is just off the western boundary of the map.

(Wolfartet al. 1984; Win Swe, pers. comm. 1992)
but Palaeogene sediments are absent at Kalaw and
the age of the folding cannot be directly bracketed
by the age of undeformed overlying sediments.
Yang & Besse (1993) similarly caution that the
folding of Upper Jurassic-Lower Cretaceous units
in the Khorat Plateau may have occurred anytime
between the mid-late Yanshanian orogeny and the
Late Tertiary.
There are several pieces of data which point
towards a post-Cretaceous regional uplift and
folding event. First, MacDonald et al. (1993) show
that the Doi Inthanon metamorphic core complex in
Northwest Thailand was formed and then uplifted
between the Late Cretaceous (72 _ 1 Ma) and the
Latest Oligocene epochs (25 _ 4 Ma). Second,
Rangin (pers. comm. 1992) reports that Mesozoic
continental sediments of the Shan Plateau are thrust
over Eocene sediments of the Central Basin of
Thailand. Third, thrust-cored anticlines of the
Chauk Oil Field in the forearc basin of Myanmar
first formed during the Oligocene, as did the
E-W trending cross basin highs which separate the
forearc and back-arc basins into smaller sub-basins
(Bender 1983). Fourth, a significant Oligocene
unconformity is present throughout northern
Myanmar (Bender 1983). Finally, the intermontane
basins of Thailand as well as the Pattani trough
began subsiding rapidly in the Late Oligocene
(Polachan et al. 1991). Thus, there are several
regional geological events which point to a Tertiary
uplift and folding event.
Mesozoic and Palaeozoic carbonates
of Thailand
Upper Palaeozoic and Lower Mesozoic limestones
were sampled throughout peninsular Malaysia,
Thailand and Myanmar (Figs 1 and 2). These
212 B. RICHTER & M. FULLER
include the Ordovician to Silurian Setul Limestone
exposed on Langkawi Island (Jones 1978), the
Upper Palaeozoic Plateau Limestone of the Shan
region of Myanmar (Bender 1983), the Lower to
Middle Permian Ratburi Limestone exposed
throughout peninsular Thailand (Nutalaya et al.
1983; Baird & Bosence 1993), and the Triassic
Lampang Group of central Thailand (Nutalaya
et al. 1983). The Palaeozoic units represent sub-
tropical to tropical shallow marine shelf carbonates
which dominated the Sibumasu block prior to the
Indosinian orogeny while the more limited occur-
rences of Lower Mesozoic rocks record localized
shallow marine deposition (Hutchison 1989). These
limestones have almost certainly undergone the
Late Mesozoic to Palaeogene deformation seen in
the Kalaw redbeds but may also have experienced
an older Late Palaeozoic or Late Triassic deforma-
tion (Jones 1978). Baird & Bosence (1993) con-
clude that the Ratburi Limestone in northern
peninsular Thailand has undergone at least three
episodes of diagenetic alteration. They argue that
the high temperature isotopic signatures which
characterize the final diagenetic event is likely
the result of heating caused by the intrusion of
Cretaceous granites. Silicification, tensional calcite
veins and karstification characterize the Palaeogene
evolution of the unit. Thus, the folding recorded
in these carbonates could have taken place in either
the Late Triassic-Early Jurassic Indosinian orogeny
or subsequently during Late Mesozoic-Palaeogene
subduction.
Palaeomagnetic sampling and analyses
Sample locations are summarized in Table 1.
Between six and ten individually orientated cores
were drilled at each site using a portable gasoline
powered diamond coring drill. These were orien-
tated with a Pomeroy orientating platform and a
Brunton compass and no correction for secular
variation was required. None of the samples
required the use of a sun compass. If exposure
permitted, additional sites were collected several
tens of metres up or down section. Each site
collected during the 1991 and 1992 field seasons
was located with a hand held Magellan GPS
Navl000 Pro Global Positioning System unit and
locations are accurate to _+30 m. Sites collected
during the 1990 field season were located with
local base maps and location errors can be as
large as +_300 m.
Laboratory analyses
Palaeomagnetic analyses were conducted at the
University of California, Santa Barbara using either
a 2-G Enterprises 3-axis cryogenic magnetometer
or a Molyneux Molspin spinner magnetometer.
Three methods of demagnetization were employed
depending upon the magnetic behaviour of
individual samples: (a) alternating field de
magnetization (AFD), (b) continuous thermal
demagnetization (CTD) (Dunn & Fuller 1984),
and (c) stepwise thermal demagnetization.
Statistical procedures
Orthogonal vector plots (after Zijderveld 1967)
were constructed for each sample in order to
characterize its demagnetization behaviour.
Principal component analysis was conducted using
computer software written by S. L Gillette
(© 1985) which utilizes the least squares regression
technique of Kirschvink (1980). In addition, re-
magnetization circles were fitted to samples
displaying two component behaviour using the
Gillette software and directions were calculated
after the method of Halls (1978). In all cases, site
mean directions, confidence intervals (t~95), and
precision parameters (k), were calculated after the
method of Fisher (1953). Folded strata were
analyzed using the methods of McElhinny (1964)
and McFadden (1990).
New palaeomagnetic data
Kalaw redbeds, Myanmar
Palaeomagnetic samples were collected from 14
sites located along the Meiktila-Kalaw highway
(Fig. 4). The traverse crosses several parallel folds
and bedding attitudes vary considerably. Thus, it
was possible to conduct a detailed fold test.
Samples were collected primarily from medium
to fine grained red sandstones which vary from
weakly friable to well cemented. Hematite cements
are present in all samples while the strongly
cemented samples also contain as much as 30- 40%
carbonate cement. In the latter case, hematite dust
rims completely encircle the detrital quartz grains
and appear to predate carbonate cementation.
Conglomerates within the sequence were also
sampled in the hope of obtaining conglomerate
tests but the clasts are dominantly composed of
very weakly magnetized limestones which
exhibited variable behaviour during demagnetiz-
ation.
Orthogonal vector plots from demagnetization
experiments are presented in Fig. 5. The magnetic
behaviour of the medium to fine grained units
within the Kalaw redbeds is excellent and in almost
all cases, independent of grain size or degree of
cementation, samples display single component
behaviour above 150-250 ° C. The low temperature
component is less rotated than the stable endpoint
 PALAEOMAGNETISM OF THE SIBUMASUBLOCK 213

Thermal Demag AF Demag Thermal Demag

B92-58-4C B92-57-4B B92-57-4B
W Up WUp W Up
s I N
S [~L%I~36dC, , , N S : : : : : N -
~! "K"
:
4 °C
: : :

~\'%. oc
i \'%458oc 19 0 0 Oe,, 1300 Oe
I \- 77oc -,%
E Dn / ...-.'1 K "~50 Oe 0n] 1.o . \ ",.2.90 oC
EDn] I i ~ ) 9 5 " _ [C~24°C I.O'~,~ ~ , ~ 00e

1900Oe
|

500oC

Fig. 5. Characteristic orthogonal demagnetizationplots for samples collected from the Kalaw Redbeds, Myanmar.
Closed symbols represent declination data plotted in the horizontal plane while open symbols represent the projection
of inclination data upon the vertical N-S plane. Inset plots show the normalized intensity decay throughout a
demagnetization experiment. Demagnetization steps are in degrees celsius or Oersteds and initial intensities are in
Gcm3/g. Magnetic behaviour is excellent and after a soft present field component is removed the magnetization
decays univectorially to the origin. Samples which did not display this behaviour were not incorporated into the site
mean values.

and most likely reflects recent chemical weathering
in the present field direction. The remaining
magnetization is extremely stable to 550°C above
which intensity begins to drop off. In all samples
a stable direction remains until temperatures of
640°C at which point magnetic behavior becomes
erratic. These intensities have not been corrected
for measurement at high temperature but the
intensity drop at 580°C indicates that a fraction of
the magnetic signal resided in magnetite. Thus, a
combination of AF and thermal demagnetization
was employed to determine qualitatively the
relative importance of magnetite and hematite
carriers. Sample B92-57-4b (Fig. 5) was first
subjected to AF demagnetization in peak fields of
200 mT (2000 Oe) and was subsequently thermally
demagnetized to 630°C. Approximately 60% of
original intensity was still present after treatment in
200 mT fields, indicating that the majority of the
signal resides in a mineral which is very resistant
to AF demagnetization. In order to eliminate the
possibility of an AF resistant pyrrhotite component,
samples from sites B92-54, 55, 57, 58, 60 and 81
were thermally demagnetized in stepwise fashion to
400°C and then AF demagnetized in peak fields of
200 mT. The behaviour of these samples indicates
that goethite or pyrrhotite, which are very resistant
to AF demagnetization but not to thermal de-
magnetization, are not present. Hematite, there-
fore, is the primary magnetic carrier with a small
portion of the signal residing in magnetite. There
is no change in direction when the samples are AF
or thermally demagnetized and thus both magnetite
and hematite are interpreted to carry the same
magnetization.
Thirteen of the fourteen sites show CW declina-
tions of c. 40 ° with shallow positive inclinations
(Fig. 6 and Table 1). The in situ mean inclination
and declination are 18.6 ° and 39.3 ° respectively
while the structurally corrected values are 23.4 °
and 44.7 °. This gives a corrected VGP at 46.4°N
190.6°E. After application of individual bedding
corrections the o~95 decreases from 18.0 to 6.9
while kappa increases from 5.8 to 34.0. Statistical
analysis of the significance of this fold test indi-
cates that it is positive at the 99% level using the
test of McElhinny (1964) and is positive at the 95%
level using the SCOS1 procedure of McFadden
(1990).
In order to see if the Kalaw redbeds have a
magnetic fabric which might have affected the
remanence direction, we measured the anisotropy
of magnetic susceptibility (AMS) for 16 represen-
tative samples. These data show a weakly aniso-
tropic magnetic fabric with a strong clustering of
maximum directions parallel to the regional fold
axis and a girdle of minimum and intermediate
directions perpendicular to the maximum direction
(Fig. 7). The remanence direction is located
approximately 80-90 ° away from the maximum
anisotropy and does not appear to be controlled
by it. All of the available data, therefore, indicate
that the Kalaw redbeds possess a pre-folding
magnetization which has not been altered by
subsequent deformation.
Palaeozoic and Mesozoic carbonates
A variety of Palaeozoic and Lower Mesozoic
limestones also display c. 40 ° of CW deflection.
214 B. RICHTER & M. FULLER

Kalaw R +
Shan P l a t e a u , + ~ +
Myanmar + ~ ~ \ +
/ ~ ~ ~ + + Corrected / '
+ + + +++++Jr--- 90 °
/ ~'-~r / Cooroinales y+-
270°--~-++++++++'~++ +++++++~90 °
Fig. 6. Positive fold test at the 99% level in the Jurassic-Cretaceous Kalaw redbeds of the Shan Plateau of Myanmar.
The in situ and structurally corrected directions are 18.6°, 39.3° and 23.4 °, 44.7 °, respectively. Solid symbols denote
positive polarities while open symbols denote negative polarities. Ellipses are the 95% confidence interval about the
mean (alpha95). See Table 1 for individual site mean directions and fold test statistics.

Orthogonal vector plots for the Ratburi Limestones
are presented in Fig. 8 and directional data are
presented in Fig. 9 and Table 1. These samples
display three types of demagnetization behaviour:
(a) univectorial decay to the origin, (b) a strong low
temperature overprint followed by univectorial
decay of a high temperature component, and (c)
multicomponent behaviour in which a high
temperature signal cannot be isolated. In most
cases the intensity of the signal is swamped by
instrument noise before the Curie point is reached.
Nonh
270*
All three types of behaviour could be found at
the same site in samples collected less than 1 m
apart. This can be partially attributed to variable
mineralogy (magnetite v. pyrrhotite), variable
amounts of veining and recrystallization, and
extremely weak initial magnetizations. Samples
displaying multicomponent magnetizations have
not been used in site mean calculations but there
is still substantial dispersion in the mean directions.
These sites are distributed between latitudes 6°N
to 20°N (Fig. 9) and show a CW direction through-
out much of the peninsular Thailand. Fold tests in
the Setul Limestone of Langkawi Island, Malaysia,
and the Lampang Limestone of north-central
Thailand are negative (Table 1 and Fig. 10) and
indicate that the magnetism was acquired after the
youngest episode of folding (Fig. 11). Although the
errors are large it is clear that the in situ poles
cluster in a single region while the corrected poles
are dispersed. As in the prior two single fold tests,
this regional attitude test on the VGP'S indicates
90* that the magnetization was acquired after folding.

" AMSMinimum I+ ~// Age of magnetizations

,, ,,MS,,,a=um {;
• Axis
180 °
Fig. 7. Anisotropy of magnetic susceptibility (AMS)
data for the Kalaw redbeds. See Fig. 3 for an
explanation of symbols. The magnetic fabric is weakly
developed with the maximum susceptibility clustered
around the average fold axis for the region. However,
the remanence direction does not appear to be controlled
by it (see Fig. 6).
The magnetization measured in the Palaeozoic
and Mesozoic limestones is a secondary chemical
remanence which is carried by pyrrhotite and low
coercivity magnetite. There are two tectonic events
which could have driven this remagnetization:
the late Triassic Indosinian orogeny, and a
combination of Late Cretaceous magmatic activity
and the Late Cretaceous to Palaeogene? uplift event
discussed above. The lack of a Mesozoic age
progression of magnetizations makes it difficult to
discriminate between these two possibilities and
 PALAEOMAGNETISM OF THE SIBUMASU BLOCK 215

Thermal Demag
T90-45-2
W Up
S : : : : : N
'~78°C
- - ~ 85°C
'~, 121 °C

Dn
C

Thermal Demag
T90-27-6
W Up

t o.ok
E Dn
500°C

T90-27-9Thermal
Demag ( -. ,. i ,
/f _
w up ~%
Fig. 9. In situ palaeomagnetic directions of Palaeozoic
tu and Mesozoic limestones from Thailand and Malaysia.
23°C These show CW declination deflections similar to that
10 seen in the Kalaw redbeds. These directions are not
.~ o . - ~ - . - .~ -'.~ 7 ~ disturbed by the NE-SW trending Ranong strike-slip
. . . . 282o(~ / .~ fault zone (RFZ). This indicates that at least some fault
blocks in peninsular Thailand have not rotated with
E Dn respect to each other. However, these data cannot
500°C distinguish between a plate wide rotation or a systematic
'domino' style rotation.
Fig. 8. Orthogonal demagnetization plots for continuous
thermal demagnetization experiments conducted on
samples from the Ratburi Limestones. These samples
display three types of demagnetization behaviour: the errors overlap all of the VGPs reported from the
(a) Univectorial decay to the origin; (b) a strong low Mesozoic of the Khorat Plateau. Thus, we can only
temperature overprint followed by univectorial decay of confidently say that the pole measured in these
a high temperature component; and (c) multicomponent limestones is no older than L a t e Triassic and
behaviour in which a high temperature signal can not be probably no younger than Late Cretaceous. In view
isolated. Samples displaying Type (a) or Type (b)
of the fact that the mean VGP coincides with the
behaviour were used to calculate site mean values.
See Figs 3 and 5 for an explanation of symbols. Late Triassic pole (Fig. 11) calculated for the
Khorat Plateau by Yang & Besse (1993), a Late
Triassic remagnetization is more likely.
216 B. RICHTER & M. FULLER

N Conversely, the age of the magnetization found

in the Kalaw redbeds is clearly pre-folding. Thus
the magnetization could be a primary depositional/
N diagenetic magnetization which is as old as the
stratigraphic age (Late Jurassic to Cretaceous) or
it could be a much younger pre-folding chemical
magnetization which is tied to the Late Cretaceous
90° to Palaeogene uplift event described elsewhere in
this paper.
In order to refine the age of the Kalaw pole we
have prepared tectonic reconstructions of South
(a) China, Indochina, the Shan Plateau, and the Lhasa
and Qiangtang blocks at various Mesozoic time
N slices relative to the GAD and examined the
reconstructed position of the Shah Plateau (Fig.
12). The reconstructions were prepared on an
Evans and Sutherland workstation courtesy of
N G. Ramsayer and I. Norton of Exxon Exploration
+
Company. These were made by directly restoring
the palaeolatitude using the mean VGP position and
90° then by empirically rotating the terrane along lines
of palaeolatitude until space problems were
minimized and 300--400 km of sinistral offset was
present along the Red River fault (Table 3). A small
+ Corrected/ range of solutions is valid as long as the 95%
+~~+ + ++ +7 90° confidence ellipse for the VGP encloses the GAD
axis. The positions of the South China and North
China blocks are taken largely from Enkin et al.
(1992) but have been modified by Ian Norton (pers.
(b)
comm. 1994). The Lhasa and Qiangtang blocks are
Fig. 10. Negative fold tests in (a) Silurian Setul essentially held fixed to South China but com-
Limestones, Langkawi Island, Malaysia; and (b) the pressional deformation across these blocks has
Early Mesozoic Lampang Limestone, North Central been restored by using plate circuits to reconstruct
Thailand. See Table 1 for numerical data and Fig. 3 the position of India at the time of collision and
for an explanation of symbols.

r-----s /
~______~: :COn-6dti~d: : i " i :i: J

180 °

Fig. 11. Virtual geomagnetic poles for the Ratburi, Triassic and Setul limestones of Thailand and Malaysia. This plot
indicates a negative regional fold test but the errors are large. The corrected Khorat Tu, Ju, and K1 poles have been
reproduced from Yang & Besse (1993). The mean in situ value for this study (open circle) coincides with the Khorat
Tu pole. This suggests a Late Triassic remagnetization age for the carbonates measured during this study.
 PALAEOMAGNETISM OF THE SIBUMASU BLOCK 217

Table 3. Summary of the rotation poles, relative to the laps or underlaps which are inconsistent with the
GAD, used to reconstruct the Shan and Khorat Plateaus geological data. The Late Cretaceous and Late
against the reconstructed position of South China block Palaeocene reconstructions show the smallest space
(see Fig. 12) problems and also show the correct amount of
displacement along the Red River Fault. This
Time Shan Plateau Khorat Plateau
analysis suggests that the magnetization measured
slice (Shan-Thai Block) (Indochina Block)
(Ma) in the Kalaw redbeds is most likely to be Late
Cretaceous to Palaeocene? in age.
Lat. Long. Angle Lat. Long. Angle

60 -2.0 99.6 43.5 -17.4 83.6 29.3

Discussion of palaeomagnetic data
80 -4.1 100.5 43.8 -26.7 80.9 31.3
120 -13.1 95.6 44.7 -35.4 78.8 34.2 Regional rigid block rotations v. chaotic
140 -18.8 93.1 46.0 -40.4 75.4 36.8 local rotations
160 -30.7 86.9 50.6 -51.9 70.0 45.2
180 -29.2 88.3 50.4 -51.6 69.3 46.0 As d i s c u s s e d above, U p p e r P a l a e o g e n e and
Neogene palaeomagnetic data show a wide variety
of CW, unrotated, and occasionally C C W rotated
directions. McCabe et al. (1993) argue that the
then by retrodeforming Lhasa and Qiangtang out apparently chaotic N e o g e n e declinations in
to the reconstructed northern edge of India. Thailand record local block rotations and that older
This analysis shows that the Jurassic and Early terrane motions have been obscured. We present
Cretaceous reconstructions all have serious over- an alternative hypothesis. The Neogene rotations

Shan J-K pole I

Restored if held fixed to I
Khorat Plateau Indochina _1
JI-Ke Pole "t

Restored
Shan Plateau v~t~ x'--J I
00o

~ . ~ 60 ° N

120 ° E 120 ° E

120 Ma / ~60 Ma Late 1I

Early Cretaceous Paleocene |

Fig. 12. Plate reconstructions of South China (Eurasia) relative to the geocentric axial dipole (GAD) at various time
slices. For each time slice the VGPs for the Khorat and Shan Plateaus have been restored to the dipole axis and the
respective terranes then restored along lines of palaeolatitude until space problems were minimized. The white Shan
Plateau represents its position if held fixed to Indochina and then rotated CW relative to South China. The filled Shan
Plateau is the position obtained by restoring the Shan VGP directly to the GAD. The white lines are the coastlines of
South China and Indochina. Sinistral displacement along the Red River fault can be estimated by comparing these
coastlines. See Table 3 for rotation poles and the text for additional discussion.
218 B. RICHTER & M. FULLER
appear to be limited to extensional and trans-
tensional zones in the Central Thai province where
strike-slip faults are well documented (Polachan
et al.1991). The Central Thai province is bounded
on either side by moderately to highly deformed
fold belts and then by the relatively undeformed
Shan and Khorat plateaus. Although faults have
been mapped along the margins of these plateaus,
the interiors are remarkably undeformed (Bender
1983; Hutchison 1989). Thus, it appears that these
large crustal blocks have only deformed along their
margins. Data collected from the marginal fold
belts or the transtensional zones between these
blocks probably record local rotations driven by
transtension and strike-slip motion but data
collected from the interior of these blocks probably
record large-scale Late Mesozoic and Tertiary
terrane motions. Lamb's (1994) discussion of the
behaviour of the brittle crust in wide plate boundary
zones provides a theoretical basis for this
hypothesis. He proposes a simple two layer model
of the lithosphere in which the brittle upper crust
sits upon ductile lower crust (Fig. 13). The brittle
crust consists of rigid blocks of varying sizes which
can only deform by rotating and translating. Thus,
the fine scale velocity field will be discontinuous
with abrupt velocity changes across faults. When
viewed at a large scale, these velocity variations
are averaged out and match the velocity field of
the underlying ductilely deforming layer. Thus, an
overall CW rotation and southeastward translation
in the lower crust could be recorded in the upper
crust as a mixture of rotations. Sundaland is almost
certainly broken into smaller crustal blocks than
Rotating and
translating
blocks
Fig. 13. Velocity field model presented by Lamb (1994) in an attempt to explain brittle deformation in wide plate
boundary zones. This model for continental deformation may be able to explain the variety of palaeomagnetic
rotations observed in SE Asia. See text for discussion.
have previously been recognized but there is ample
evidence which indicates that terrane motions can
still be measured.
Rotations
The VGPs for the Khorat and Shan Plateaus and
the APW path for Eurasia-South China (Enkin
et al. 1992) are shown in Fig. 14. The mean Kalaw
VGP is rotated 47.6 ° CW relative to the GAD,
30.4 ° CW relative to the Upper Cretaceous
(100 Ma) Eurasia-South China reference pole, and
15.4 ° CW relative to the Khorat Lower Cretaceous
VGP presented by Yang & Besse (1993). A similar
CW rotation can be calculated relative to other
Cretaceous or Palaeogene segments of the Eurasia-
South China APW path. Thus, we conclude that the
Shan Plateau VGP is rotated c. 25-30 ° relative to
South China and c. 10-15 ° relative to the Khorat
Plateau (Indochina Block). Revisions in the ages
or error estimates of the Shan and Khorat VGPs
are unlikely to alter this basic conclusion. The
palaeomagnetic data do not uniquely indicate when
this rotation took place but the geological evidence
indicates that it occurred during the Late
Cretaceous and Palaeogene.
Latitude shifts
Palaeolatitudes have been calculated for all of the
data presented in Tables 1 and 2. These data show
a clear pattern with the Khorat Plateau travelling
slightly southward while the Shan Plateau has
travelled northward by 8.5 ° +/- 3 ° relative to the
Fine scale Large scale
velocity field velocity field
/.,,-J.,,,,..,, ,,;.".,., ,,
I"'-" "1- "" "" "
IP t J' • 1.
,,,'////
,r.......Koo°,°°o..,,o.
f; :.:r
 PALAEOMAGNETISM OF THE SIBUMASUBLOCK 219

.ur J,

rat~ ~Khora(Tu(corr.)~l "-~Xo,.2_ ~/ 9/

180*
Fig. 14. Composite VGP-APW plot which compares the motions of the South China block with the VGPs for the
Indochina (Khorat Plateau), and Sibumasu (Shan Plateau) blocks. The APW path for South China is reproduced
from Enkin et al. (1992) and the Khorat VGP positions are reproduced from Yang & Besse (1993). Despite the
ambiguity in the age of the Shan Plateau pole (K1 to Tpal?) it is clearly rotated CW relative to the equivalent range of
Khorat Plateau and South China positions. Likewise, it is clearly displaced northward. The exact amount of rotation
and northward translation is dependent upon the inferred age of the pole.

GAD over the same time period. This corresponds
to c. 12° northward motion relative to the Khorat
Plateau Lower Cretaceous VGP and c. 5 ° relative
to the 100 Ma Eurasia-South China VGE
AMS measurements (Fig. 7) show that a strong
bedding plane fabric is not present and thus shallow
inclinations do not seem to be related to a
depositional alignment of platy hematite. The
fold axis is almost horizontal and thus shallow
inclinations cannot be attributed to an overlooked
plunge correction. These observations indicate that
the small inclinations do not result from a problem
with the palaeomagnetic signal. However, there
are some potential structural complications. The
Kalaw redbeds are exposed along the very western
edge of the Shan Plateau within a zone of faulting
that is bounded to the west by the N-S trending
Sagaing-Shan Boundary-Pan Laung Fault system
and bounded to the east by unnamed NW-SE
trending faults (Fig. 4). It is possible that this
triangular sliver is partially detached from most of
the Shan Plateau. This leaves open the alternative
hypotheses that a separate 'Kalaw block' has been
tilted or has been translated northward relative to
the Shan Plateau. We have collected extensively
from Silurian, Devonian and Permian carbonates
exposed near Taunggyi (Fig. 4), well away from
the Pan Laung Fault. The inclinations from these
samples should indicate if there is a separate Kalaw
block which has moved northward relative to the
rest of the Shah Plateau.
Figure 12 shows the positions of the Shah
Plateau, relative to the Khorat Plateau, assuming
end member models of no relative northward
motion and the total palaeomagnetically deter-
mined northward motion. As a cautionary note, the
error estimates on the palaeomagnetic data do not
take into account external geological sources of
error such as poor age control, limited outcrop, the
uncertainty in recognizing remagnetizations and
other factors. Thus the real errors are likely to be
larger than shown. A more realistic reconstruction
would restore the Shan Plateau to a position
between the end members shown here.
There is little doubt that oblique convergence
has driven the Burma block almost 500 km north-
ward during the Neogene (Curray et al. 1979).
Furthermore, the Indian Ocean plate has been
obliquely subducting beneath the western margin
of Sundaland since the Late Cretaceous. Depending
upon the timing of the CW rotation of Sundaland,
this oblique subduction could have driven north-
ward translation of all or part of the Sibumasu
block during the Eocene and Oligocene epochs.
What is presently unresolved is the size of these
blocks and how displacement is distributed. Are
the blocks just small slivers within the western edge
of the Mergui shelf and Shan Plateau? Are they
larger blocks bounded by faults such as the Three
Pagodas and Wang Chao (Mae Ping) Faults? Is the
entire 'rigid' Sibumasu block moving? Is displace-
ment centralized on a master fault, similar to the
present Sagaing fault, or is it distributed as small
displacements along many faults? Additional
research is needed to choose between these
possibilities.
Review of tectonic models
Extrusion tectonics
The extrusion tectonic model has evolved from the
initial slip-line field theory experiments of Molnar
220 B. RICHTER & M. FULLER

& Tapponnier (1975) and Tapponnier & Molnar

(1976) to scaled down mechanical experiments
with plasticine (Tapponnier et al. 1982, 1986; see
also Jolivet et al. 1990). Peltzer & Tapponnier
(1988) added significantly to earlier studies in that
they examined the impact of distance between
indentor and the margin of the block as well as the
impact of vertical and horizontal inhomogeneities Time 2
upon extrusion.
The type, location and timing of structures which
developed during the plasticine extrusion experi- Time 3
ments are remarkably similar to those mapped in
SE Asia. These observations led Tapponnier et al. (a)
(1986) to conclude that the Eocene collision and
subsequent penetration of India into Eurasia was
accommodated primarily by more than 1000 km
of eastward expulsion and c. 20 ° of CW
rotation of Sundaland along the Red River fault
during the Oligocene and Early Miocene.
Beginning in the Middle Miocene the deformation
front shifted northward and the South China block
was likewise extruded eastward and rotated 20 °
CW between the Red River and Altyn Tagh faults.
Thus, Sundaland is thought to have undergone a
rigid body CW rotation of 40 ° since the Late
Palaeogene (Fig. 15). A more detailed review
(b)
of the extrusion model is presented in Richter
(1995).

D e x t r a l m e g a s h e a r tectonics
The dextral megashear model was elegantly
summarized by England & Molnar (1990) who
stated that 'the image of lateral transport on
(east-west trending) faults, known also as con-
tinental escape, extrusion, or expulsion, is an
illusion, and that instead the left-lateral slip on
east-striking planes in eastern Tibet is a mani-
festation of north-striking right-lateral shear'. As
this dextral shear evolves, the sinistral faults and
the N W - S E trending blocks which they bound
must rotate clockwise around local axes (Fig. 15).
An important consequence is internal deformation
which allows E - W shortening and N - S
lengthening.
Furthermore, Dewey et al. (1989) conclude that
the India-Eurasia collision has been accommo-
dated primarily by lithospheric thickening of
Eurasia in a widening east-west trending zone
which is bounded on the west by a sinistral mega-
shear extending from Makran to Baikal and to the
east by a dextral megashear which extends from
Sumatra to the Tanlu Fault system. The tectonic
evolution of eastern Tibet, Myanmar and Indo-
china would, therefore, be dominated by north-
south striking dextral shear and associated CW
rotations around local vertical axes with little to
no southeastward extrusion.
Extension in Central
Thai Basins? ~
(c)
Fig. 15. Schematic comparison of (a) the extrusion and
(b) megashear model at the beginning, middle and end
of motion. Compare this with (e) which attempts to
integrate internal deformation into the extrusion model.
In this model, Sundaland consists of large fault bounded
blocks which rotate CW as a consequence of extrusion.
But, as extrusion progresses, the blocks closest to the
collision zone are displaced farther northward along
N and NW-SE trending strike-slip faults (e.g. the
Sagaing, Uttaradit or Three Pagodas faults). As motion
continues, antithetic strike-slip faults will likely form
and local zones of extension and compression will be
created. Blocks caught within sinistral zones may rotate
CCW while those caught in dextral zones may rotate
CW. As the westernmost block (Shan Plateau) moves
northward it may rotate CW, driving rifting in the
Central Thai basins.
 PALAEOMAGNETISM OF THE SIBUMASU BLOCK
Polachan et al. (1991) have proposed a tectonic
model for the evolution of the Tertiary basins of
Thailand which is based upon extrapolation of the
dextral simple shear strain ellipsoid presented by
Wilcox et al. (1973) to the regional scale. This is
based upon the systematic geometrical relationship
between the north-south trending Tertiary basins
and the conjugate NW-SE and NE-SW trending
strike-slip faults of Thailand. The shapes and
spatial distribution of these basins suggest that they
formed as lazy-S and lazy-Z basins in direct
response to strike-slip faulting. Within this frame-
work the Red River, Mae Ping, Three Pagodas and
Sumatra faults, which all trend NW-SE, represent
the master dextral faults while the Northern
Thailand, Uttaradit, Ranong and Klong Marui
faults, which trend NE-SE, represent the conjugate
sinistral faults. A genetic relationship between
strike-slip faulting and basin formation is also
supported by more detailed studies conducted by
Remus et al. (1993) who have documented wrench
features within the Phetchuban graben.
The common theme of all three models is that
southeastward extrusion is minimal, CW rotations
are driven by dextral shear across plate boundaries,
and Sundaland is broken into smaller sub blocks
which have rotated in sympathy but which are
separated by strike-slip faults. However, a simple
dextral megashear model cannot explain northward
transport of blocks along the western boundary of
Sundaland.
Preferred tectonic model
The extrusion model is an elegant way to link many
of the geological features of Sundaland but data
collected over the last decade indicate that the
model must be refined (see Richter 1995). In
summary, the new model must: (1) allow for crustal
thickening and shortening within the collision zone
(e.g. Dewey et al. 1989; Le Pichon et al. 1992;
Houseman & England 1993); (2) recognize the
importance of dextral shear along the western
margin which may drive internal deformation
within the extruding block (e.g. Dewey et al.
1989; Polachan et al. 1991; Richter et al. 1993);
(3) reconcile palaeomagnetic data which show
20-30 ° of CW rotation of Indochina and Sibumasu
relative to South China, rather than the predicted
40-50 ° (Yang & Besse 1993 and this study);
(4) reconcile estimates of fault slip on the Red
River fault which are far smaller (300-400 km)
than the predicted thousands of kilometres
(Tapponnier et al. 1990; Scharer et al. 1990;
Lacassin et al. 1993; Leloup et a1.1993);
(5) integrate basin analysis data which indicate that
most or all of the fault motion on the Red River
Fault is accommodated by transtension in the
221
Yinggehai basin and is not transmitted into the
South China Sea (Qiming & Quanxing 1991; I. O.
Norton, pers. comm. 1994); (6) explain Cretaceous
and Palaeocene basin rifting ages along the South
China shelf which predate extrusion (e.g. Chen
& Pei 1993; G. Grabowski, pers. comm. 1994);
(7) explain palaeomagnetic data which show CCW
rotations in Borneo, Palawan, Sulawesi and the
Celebes Sea (Schmitdke et al.1990; Fuller et al.
1991); and (8) account for interaction with the
Australian and Philippine Sea plates along the
southeastern margin of Sundaland (Richter et al.
1992; Hall et al. 1995).
An alternative hypothesis is that Sundaland is
composed of smaller sub-blocks which have
rotated CW together but which have slid past
each other (Fig. 15). Deformation of the Sibumasu
block is dominated by the oblique subduction of the
Indian Ocean Plate, resulting in CW rotations and
regional dextral displacements along N-S and
NW-SE trending faults. At this stage, we have not
attempted a detailed structural analysis and it is
expected that motions between smaller sub-blocks
could be both dextral and sinistral. The model pro-
posed by Lamb (1994) provides a good analogy
(Fig. 15). The Indochina block, however, is too far
behind the plate boundary to feel this motion and it
deforms in response to extrusion, which is driven
by convergence between the Indian Craton and
Eurasia. These two modes of deformation meet in
the Central Thai Basin, and could drive extension
and uplift of the Doi Inthanon metamorphic core
complex (Macdonald et al. 1993), the formation
of intermontane basins, and the formation of the
Central Thai basin and the Gulf of Thailand
(Polachan et al. 1991). Finally, Sundaland is
bounded to the southeast by the Australian plate
and the Philippine Sea Plate (PSP). We hypothesize
that the extruding block cannot simply override
these plate boundaries and is partially or com-
pletely coupled to them. Thus, extrusion works
with northward subduction of Australian oceanic
crust and northward motion and CW rotation of
the PSP to drive the CCW rotation of Borneo.
Conclusions
(1) The Jurassic-Cretaceous Kalaw redbeds of the
Shan Plateau of Myanmar yield a mean prefolding
magnetization which is rotated CW 47.6 ° relative
to the GAD. The age of this magnetization is
interpreted to be Late Cretaceous or Early
Palaeogene. The exact amount of rotation relative
to South China varies with the inferred age but
is approximately 25-30 ° CW over the range of
possible ages. This rotation most likely took place
between the Late Cretaceous and Early Miocene.
222 B. RICHTER & M. FULLER
(2) The mean Kalaw VGP is rotated 15.4 ° CW
relative to the Khorat Lower Cretaceous VGE
When both palaeomagnetic and geological errors
are considered, this is consistent with a small CW
rotation between the Khorat and Shan Plateaus.
Such a rotation may be responsible for extension
in the Central Thai basin and uplift and unroofing
of the Doi Inthanon metamorphic complex in
Northern Thailand. This rotation most likely began
in the Middle or Late Oligocene.
(3) Neogene and Upper Palaeogene palaeo-
magnetic data from Thailand show a variety of CW,
CCW and unrotated declinations. We hypothesize
that Sundaland is composed of large rigid blocks
(the Shan and Khorat Plateaus, peninsular
Thailand) which are separated by extensional and
transtensional zones (Central Thai Basin, Pattani
trough, Intermontane basins). The young scattered
rotations appear to be limited to these extensional
and transtensional zones where a strike-slip com-
ponent of motion could easily drive local rotations.
Within the deformation model proposed by Lamb
(1994) the rotations recorded in the intermontane
basins and Neogene basalts may record the fine
scale velocity field while the rotations in the Shan
and Khorat plateaus may record the large-scale
velocity field.
(4) The palaeolatitude data from the Kalaw
redbeds show 8 ° of northward transport relative
to the GAD, 12° relative to the Khorat Lower
Cretaceous VGP, and 5 ° relative to the 100 Ma
South China VGE The exact amount of northward
motion varies with the inferred age of the respec-
tive VGPs. Part of this latitudinal motion can be
explained by the extrusion model but plate recon-
structions show that some or all of it must be
accommodated by overall dextral motion between
the Indochina and Sibumasu blocks or possibly
within the western margin of the Sibumasu block.
(5) CW rotations have been measured in
Palaeozoic carbonates in peninsular Thailand and
northern peninsular Malaysia. These were most
likely remagnetized during the Late Triassic
Indosinian orogeny but may have also been re-
magnetized during the Cretaceous. These data
establish that a domain of CW rotation extends at
least as far south as Langkawi Island, Malaysia.
References
ACHACHE, J. & COURTILLOT, V. 1985. A preliminary
Upper Triassic palaeomagnetic pole for the Khorat
plateau (Thailand): consequences for the accretion
of Indochina against Eurasia. Earth and Planetary
Science Letters, 73, 147-157.
--, --. & BESSE, J. 1983. Paleomagnetic constraints
on the Late Cretaceous and Cenozoic tectonics
of southeast Asia. Earth and Planetary Science
Letters, 63, 123-136.
South of this the pattern becomes more complex
and mixed CW and CCW declinations are found
in the same units. This zone of mixed CW and
CCW grades southeastward into Borneo where the
dominant sense of rotation is CCW.
(6) The palaeomagnetic and geological data from
Sundaland suggest several modifications to the
extrusion tectonic model. Northern Sundaland has
only rotated 25-30 ° CW with respect to South
China during the Tertiary. The total southeastward
translation of the Indochina block was only 300-
500 km. The extruding block is composed of
smaller rigid blocks (the Shan and Khorat Plateaus,
peninsular Thailand) which are separated by exten-
sional and transtensional zones (Central Thai
Basin, Pattani trough, Intermontane basins) which
have accommodated internal deformation. As such,
a regional CW 'domino' rotation is preserved in the
large blocks while a variety of mixed rotations is
found in the transtensional zones. Deformation of
the Sibumasu block resulted from oblique sub-
duction of the Indian Ocean Plate, whereas
deformation of the Indochina block was dominated
by extrusion driven by convergence between the
Indian Craton and Eurasia. The extruding block
could not simply override plate boundaries located
to the south and east and was partially or com-
pletely coupled to them. Thus, extrusion worked
with northward subduction of Australian oceanic
crust and northward motion and CW rotation of the
PSP to drive the CCW rotation of Borneo.
This study would have been impossible were it not for the
assistance of numerous colleagues throughout Thailand,
Myanmar, and Malaysia. We would like specifically to
thank S. Bunopas and V. Thitipawarn (DMR, Bangkok);
U. T. Myint (MOGE); J. Murphy, T. Acomb, and
N. Timble (Amoco Myanmar); C. Hutchison (University
of Malaya, Kuala Lumpur); N. Halle (Petronas); A. Rb.
Samsudin (University Kebangsan Malaysia); I. Norton,
D. Maughn, G. Ramseyer, M. Fitzgerald, D. Leary and
G. Grabowski (Exxon). Sample collection was assisted
by C. Anderson and S. Cisowski. B. Dunn assisted in data
analyses and interpretation. The study was funded by the
National Science Foundation as well as by contributions
by an industrial consortium consisting of Amoco,
Chevron, Exxon and Unocal. Finally, we thank J. Briden
and R. Hall for critical reviews which greatly enhanced
the quality of the manuscript.
BAIRD, A. & BOSENCE,D. 1993. The sedimentologica3and
diagenetic evolution of the Ratburi Limestone,
Peninsular Thailand. Journal of Southeast Asian
Earth Sciences, 8, 173-180.
BARR,S. M. & MACDONALD,A. S. 1979. Paleomagnetism,
age, and geochemistry of the Denchai basalts,
northern Thailand. Earth Planetary Science Letters,
46, 113-124.
& -- 1981. Geochemistry and Geochronologyof
 PALAEOMAGNETISM OF THE SIBUMASU BLOCK
Late Cenozoic basalts of Southeast Asia. Geological
Society of America Bulletin, 92, 1069-1142.
- - & -- 1991. Towards a Late Palaeozoic-Early
Mesozoic Tectonic Model for Thailand. Journal of
Thai Geosciences, 1, 11-22.
& Haile, N. S. 1978. Reconnaissance
Paleomagnetic Measurements on Triassic and
Jurassic Sedimentary rocks from Thailand.
Geological Society of Malaysia Bulletin, 10, 53-62.
, _ - & REYNOLDS,P. 1976. Paleomagnetism
- Reconstructions. In: Taylor, B. & Natland, J. (eds)
and Age of the Lampang Basalt (Northern Thailand)
and the Age of the Underlying Pebble Tools. Journal
of the Geological Society of Thailand, 2, 1-10.
~, TANTISUKRIT, C., YAOWANOIYOTHIN, W. &
MACDONALD, A. 1990. Petrology and Tectonic
implications of Upper Palaeozoic volcanic rocks of
the Chiang Mai belt, northern Thailand. Journal of
Southeast Asian Earth Science, 4, 37-47.
BENDER, E 1983. Geology of Burma - Beitrage zur
Regionalen Geologie der Erde. Gebruder
Borntraeger, Berlin.
BUNOPAS, S. 1982. Paleogeographic history of Western
Thailand and adjacent parts of Southeast Asia -
a plate tectonics interpretation. Geological Survey,
Paper 5, Department of Mineral Resources,
Thailand.
CHEN, S. & PEI, C. 1993. Geology and geochemistry of
source rocks of the Eastern Pearl River Mouth
Basin, South China Sea. Journal of Southeast Asian
Earth Sciences, 8, 1-4, 393-406.
CHEN, Y. & COURTILLOT, V. 1989. Widespread
Cenozoic(?) remagnetization in Thailand and its
implications for the India-Asia collision. Earth and
Planetary Science Letters, 93, 113-122.
CHIBBER, H. 1934. The Geology of Burma. MacMillan,
London.
fOBBING, E., MALLICK,I., PITFIELD, P. & TEOH, L. 1986.
The Granites of the Southeast Asian Tin Belt.
Journal of the Geological Society of London, 143,
537-550.
CURRA¥, J. R. et al. 1979. Tectonics of the Andaman
Sea and Burma. In: WATK~S, J., MOrCrADERT,L., &
DICKERSON,R W.(eds) Geological and Geophysical
Investigations of Continental Margins. AAPG
Memoir 29, 189-198.
DEWEY, J., CANDE, S. & PrrMAN, W. 1989. Tectonic
Evolution of the India/Eurasia Collision Zone.
Eclogae Geologica Helvitica, 82, 717-734.
DUNS, R. & FULLER,M. 1984. Thermal demagnetization
with measurement at high temperature (abs). LOS,
Transactions of the American Geophysical Union,
65, 863.
ENGLAND,P. & MOLNAR,P. 1990. Right lateral shear and
rotation as the explanation for strike-slip faulting
in eastern Tibet. Nature, 344, 140-142.
ENKIN, R., YANG, ZHENYU,CHEN, Y. & COURTILLOT,V.
1992, Paleomagnetic Constraints on the
Geodynamic History of the Major Blocks of China
From the Permian to the Present. Journal of
Geophysical Research, 97, 13 953-13 989.
FISHER, R. A. 1953. Dispersion on a sphere. Proceedings
Royal Society of London, 217, 295-305.
FULLER,M., HASTON,R., LIN, J., RiCHTER,B., SCHMIDTKE,
E. & ALMASCO,J. 1991. Tertiary Paleomagnetism
223
of regions around the South China Sea. Journal of
Southeast Asian Earth Sciences, 6, 161-184.
HALLE, N. S. & KHOO, H. P. 1980. Paleomagnetic
measurements on Upper Jurassic to Lower
Cretaceous sedimentary rocks from Peninsular
Malaysia. Bulletin of the Geological Society of
Malaysia, 12, 75-78.
HALL, R., FULLER, M., ALl, J. R. & ANDERSON, C. D.
1995. The Philippine Sea Plate: Magnetism and
Active Margins and Marginal Basins of the Western
Pacific. American Geophysical Union Monograph,
88, 371-404.
HALLS, H. 1978. The use of Remagnetization Circles in
Paleomagnetism. Physics of the Earth and
Planetary Interiors, 16, 1-11.
HAMILTON,W. 1979. Tectonics of the Indonesian Region.
United States Geological Survey Professional
Paper, 1078, 345.
HELMCKE, D. 1986. On the Geology of the Petchabun
Fold-Belt (Central Thailand) - implications for
the geodynamic evolution of Mainland S.E. Asia.
Geological Society of Malaysia Bulletin, 19, 79-85.
HOUSEMAN,G. & ENGLAND,P. 1993, Crustal Thickening
Versus Lateral Expulsion in the Indian-Asian
Continental Collision. Journal of Geophysical
Research, 98, 12 233-12 249.
HUTCHISON, C. S. 1989. Geological Evolution of South-
east Asia. Oxford University Press, Oxford
Monographs on Geology and Geophysics, 13, 368.
JOLIVET, L., DAVY,P. & COBBOLD,P. 1990. Right Lateral
Shear along the Northwest Pacific Margin and the
India-Eurasia Collision. Tectonics, 9, 1409-1419.
JONES, C. 1978. The Geology and Mineral Resources of
Perlis, North Kedah, and the Langkawi Islands.
Geological Survey of Malaysia, District Memoir,
17, 257.
KmSCHVrNK,J. 1980. The least-squares line and plane and
the analysis of palaeomagnetic data. Geophysical
Journal of the Royal Astronomical Society, 62,
699-718.
LACASSIN,R., LELOUP,P. & TAPPONNIER,P. 1993, Bounds
on strain in large Tertiary shear zones of SE Asia
from boudinage restoration. Journal of Structural
Geology, 15, 677-692. f
LAMB, S. 1994, Behavior of the brittle crust in wide plate
boundary zones. Journal of Geophysical Research,
99, 4457-4483.
LELOUP, P., HARRISON, T., RYERSON, F., WENJI, C.,
Qt,L., ET AL. 1993. Structural, Petrological and
Thermal Evolution of a Tertiary Ductile Strike-Slip
Shear Zone, Diancang Shan, Yunnan. Journal of
Geophysical Research, 98, 6715-6743.
LE PICHON, X~ FOURNIER, M. & JOLIVET, L. 1992,
Kinematics, Topography, Shortening, and Extrusion
in the India-Eurasia Collision. Tectonics, 11,
1085-1098.
MCCABE, R., HARDER, S., COLE, J. • LUMADYO,E. 1993,
The use of paleomagnetic studies in understanding
the complex Tertiary tectonic history of East and
Southeast Asia. Journal of Southeast Asian Earth
Sciences, 8, 257-268.
~ . , CELAYA,M., COLE, J., HAN, H., OHNSTAD,T., E T AL.
1988. Extension tectonics : The Neogene Opening
224 B. RICHTER & M. FULLER
of the North-South Trending Basins of Central
Thailand. Journal of Geophysical Research, 93,
11899-11910.
MACDONALD, A., BARR, S., DUNNING, G. &
YAOWANOIYOTHIN, W. 1993, The Doi Inthanon
metamorphic core complex in NW Thailand: age
and tectonic significance. Journal of Southeast
Asian Earth Sciences, 8, 117-125.
MCELHINNY, M. 1964. Statistical significance of the fold
test in paleomagnetism. Geophysical Journal of the
Royal Astronomical Society, 8, 338-340.
, HALLE, N. & CRAWFORD, A. 1974. Paleomagnetic
evidence shows Malay Peninsula was not a part of
Gondwanaland. Nature, 252, 641-645.
MCFADDEN, P. 1990. A new fold test for paleomagnetic
studies. Geophysical Journal International, 103,
163-169.
MARANATE,S. & VELLA, P. 1986. Paleomagnetism of the
Khorat Group, Mesozoic, Northeast Thailand.
Journal of Southeast Asian Earth Sciences, 1,
23-31.
METCALFE, I. 1988. Origin and assembly of south-east
Asian continental terranes. In: AUDLEY-CHARLES,
M. G. t~ HALLAM,A. (eds) Gondwana and Tethys.
Geological Society, London, Special Publication,
37, 101-118.
- - 1990. Allochthonous terrane processes in Southeast
Asia. Philosophical Transactions Royal Society of
London, A331, 625-640.
MITCHELL, A. 1992. Late Permian-Mesozoic events and
the Mergui Group Nappe in Myanmar and Thailand.
Journal of Southeast Asian Earth Sciences, 7,
165-178.
MOLNAR, E & TAPPONNIER, E 1975. Cenozoic tectonics
of Asia: effects of a continental collision. Science,
189, 419-426.
NUTALAYA,P. et al. 1983. Proceedings of the Workshop
on Stratigraphic Correlation of Thailand and
Malaysia, Volume 1, Technical Papers. Geological
Society of Thailand and Geological Society of
Malaysia.
PELTZER, G.& TAPPONNIER, P. 1988. Formation and
Evolution of Strike-slip Faults, Rifts, and Basins
during the India-Asia Collision: An Experimental
Approach. Journal of Geophysical Research, 93,
15 085-15 117.
POLACHAN, S., PRADIDTAN,S., TONGTAOW,C., JANMAHA,
S., [NTARAWIJITR, K. & SANGSUWAN, C. 1991.
Development of Cenozoic Basins in Thailand.
Marine and Petroleum Geology, 8, 84-97.
QIMING, Z. & QUANXING,Z. 1991, A distinctive hydro-
carbon basin - Yinggehai basin, South China Sea.
Journal of Southeast Asian Earth Sciences, 6,
69-74.
REMUS, D., WEBSTER, M. & KEAWKAN, K. 1993, Rift
Architechture and sedimentology of the Phetchabun
Intermontane Basin, central Thailand. Journal of
Southeast Asian Earth Sciences, 8, 421-432.
RICHTER, B. 1995. Tectonic Rotations of Southeast Asia -
The implications of New Paleomagnetic Data from
Myanmar and Thailand upon the Extrusion Tectonic
Model. MA Thesis, University of California, Santa
Barbara.
RICHTER, B., NORTON, I., SCHMIDTKE, E. & FULLER, M.
1992. Paleomagnetic Results from Southeast Asia-
Implications for Paleogeographic Reconstructions
(abs). LOS, Transactions of the American
Geophysical Union, 73, 146--147.
--, FULLER, M., SCHMIDTKE,E., U TIN MYINT, O TIN
NGWE, ET AL. 1993, Tertiary Paleomagnetism of
Myanmar and Thailand: Implications for the
tectonic evolution of Southeast Asia. Journal of
Southeast Asian Earth Sciences, 8, 1-4, 247-256.
SCHARER, O., TAPPONNIER,P., LACASSIN,R., LELOUP, P.,
DALAI, Z. & SHAOCHENG, J. 1990. Intraplate
tectonics in Asia: A precise age for large-scale
Miocene movement along the Ailao Shan-Red
River Shear Zone, China. Earth and Planetary
Science Letters, 97, 65-77.
SNANS1ENG, S., CHAODUMRONG, P., PRADIDTAN, S.
BORIPATKOSOI1983. The Tertiary sedimentary rocks
of Thailand. In: Proceedings of the Conference on
Geology and Mineral Resources of Thailand,
Bangkok, November 1983. 1-13.
SCHM1DTKE, E., FULLER, M. & HASTON, R. 1990.
Paleomagnetic data from Sarawak, Malaysian
Borneo, and the Late Mesozoic and Cenozoic
tectonics of Sundaland. Tectonics, 9, 1, 123-140.
TAPPONNIER, P. & MOLNAR, P. 1976. Slip line field theory
and large scale continental tectonics. Nature, 264,
319-324.
--, PELTZER,G. & ARMIJO, R. 1986. On the mechanics
of the collision between India and Asia In:COWARD,
M. P. & RIES, A. C. (eds) Collision Tectonics.
Geological Society, London, Special Publication,
19, 115-157.
, LEDAIN, A., ARMIJO, R. & COBBOLD, P.
1982. Propagating extrusion tectonics in Asia: new
insights from simple experiments with plasticine.
Geology, 10, 611-6.
- - , LACASSIN,R., LELOUP,P., SCHARER,U., DALAI, Z.,
ET AL. 1990. The Ailao Shan/Red River meta-
morphic belt: Tertiary left lateral shear between
Indochina and South China. Nature, 343,
431-437.
WILCOX, R. E., HARDING,T. P. & SEELY,D. R. 1973. Basic
wrench tectonics. AAPG Bulletin, 57, 74-96.
WOLFART, R., WIN, M., BOITEAU,S., WAI, M., UK CUNG,
P. & LWIN, T. 1984. The Tectonics of the Western
Shan Massif, Burma. Geologisches Jahrbuch, 75,
285-294.
YANG ZHENYU & BESSE, J. 1993. Paleomagnetic study
of Permian and Mesozoic sedimentary rocks from
Northern Thailand supports the extrusion model for
Indochina. Earth and Planetary Science Letters,
117, 525-552.
ZIJDERVELD, J. 1967. AC demagnetization of rocks:
analysis of results. In: COLLINSON, D. W., CREER,
K. M. & RUNCORN, S. K. (eds) Methods in Paleo-
magnetism. Elsevier, Amsterdam.
 Timing of the Shan-Thai-lndochina collision: new evidence
from the Pak Lay Foldbelt of the Lao PDR

R O B E R T B. STOKES 1, 4, PAUL F. LOVATT S M I T H 2 & KO S O U M P H O N P H A K D Y 3

1 Surbiton Geological Services, 14 Dennan Road, Surbiton KT6 7RY, UK
e Monument Resources (Overseas) Limited, 80 Petty France, London SW1H 9EX, UK
3 Department of Geology and Mines, Vientiane, Lao PDR
4 School of Geological Sciences, Kingston University, Penrhyn Road,
Kingston upon Thames KT1 2EE, UK

Abstract: The Pak Lay Foldbelt in the northwest of the Lao PDR is a product of the Indosinian
orogeny, resulting from the collision of Shan-Thai and Indochina along a suture now marked by
the Nan-Uttaradit ophiolite zone. Current hypotheses place the timing of this collision in either
the Permian or the Triassic. Recent field work and subsequent laboratory analyses suggest that,
in the area to the south of Pak Lay: pre-collision sediments are as young as Middle to Upper
Jurassic; arc-related volcanism continued from the Triassic into the Late Jurassic; all these sedi-
ments and volcanic rocks form an imbricate zone; and the Cretaceous Khorat Group rests with
marked unconformity above the imbricate wedge. From this we conclude that the Shan-Thai-
Indochina suturing occurred in the Late Jurassic.

This paper presents a reinterpretation of the strati-
graphy, structure and igneous activity in the
southern part of Sainabouli Province in the north-
west of the Lao PDR. It results from fieldwork,
sampling and subsequent analyses undertaken by
Monument Resources (Overseas) Limited and the
Shlapak Development Company in the hydro-
carbon exploration of their Contract Area. A total
of six weeks of fieldwork was carried out during
the months of November and December in 1992
and 1993 making a reconnaissance geological
survey of the Contract Area from the Muang
Kenthao region in the south to the Sainabouli Basin
in the north. The area of the present study lies to
the west of the Nam Khong (Mekong) River and
to the south of Muang Pak Lay.
Current interpretations
Tectonic setting
Workman (1972) described the Pak Lay region as
the exposed part of a NNE-trending Indosinian
linear foldbelt developed along the western margin
of the Indosinia (Annamia) cratonic block. The
folding was attributed to an early phase (approxi-
mately Late Triassic) of the Pacific Orogeny
(Workman & Page 1967).
In plate tectonic reconstructions, continental SE
Asia has been shown to be a composite of
allochthonous terranes which accreted to one
another during the Palaeozoic and Mesozoic
(Metcalfe 1988). There is general agreement that
they had all sutured to each other by the Late
Triassic. Of these major terranes, only the Shan-
Thai Block (also called Sibumasu) and the
Indochina Block (also called Indosinia) are directly
related to the Pak Lay Foldbelt. According to
Cooper et al. (1989), NE-dipping subduction was
initiated in the Permian beneath Indochina in
addition to a possible southwest-dipping sub-
duction beneath Shan-Thai. The two plates
collided in the Early Triassic creating (inter alia)
the Pak Lay-Petchabun Foldbelt. Extensional
collapse of the Indosinian orogen during the
Triassic allowed continental extensional basins to
develop in northeast Thailand and adjacent parts
of Lao, and then thermal subsidence, during the
Jurassic and Cretaceous, allowed the red-bed
Khorat Megasequence to be deposited over much
of northeast Thailand and the Lao PDR.
The suture between the Shan-Thai and Indo-
china blocks is generally accepted as being marked
by the Nan-Uttaradit ophiolitic zone (Hutchison
1989, Barr & Macdonald 1991). This narrow zone
extends northeast from Uttaradit along the Nan
River (Fig. 1). It is believed to extend through the
Lao PDR, past Luang Prabang, and into Vietnam
as the Lay Chau Fault. Its location within Lao is
uncertain.
Geological history of the Pak Lay Foldbelt
The first geological interpretation of the Pak Lay
area (Fig. 2) was produced from fieldwork by
Bourret (1925) who summarized the stratigraphy,

FromHall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, 225

Geological Society Special Publication No. 106, pp. 225-232.
226 R . B . STOKES ET AL.

~'.,F 106 ° E

~ VIETNAM
-, o

\ . "d'q

~it . ° i ' ~1~

SAKHON - NAKHON'~
BASIN

ET
BASIN

KHORAT BASIN
I
5
1t ...... T H A I L A N D .............................. ' ......... ,P

14°N

,1 C.i
0 250km
i i
r] Study Area i Nan-Uttaradit Ophiolite Zone ~
----- Subcrop against Quaternary Alluvium D Volcanics
r-] out crop of Indosinian Unconformity n Permo-Carboniferous Limestones
Triassic Huai Hin Lat Formation - ~ Himalayan Folding

Fig. 1. Regional geological summary of the Monument/SDC Contract Area of the Lao PDR showing the area
covered by this study.
 PAK LAY FOLDBELT, LAO PDR 227

,, 5 c.
18o15,N .......................... ........ •~ m ~
,~..K~.

~!i~ ~ ~<r~~i~¸~7~i~ii~!i;~!~¸~ i!~!ii~ ~7!!~!~! ~¸¸ ¸~<¸~

THAILAND

18°00'N
i~ii~i'~i~ !~¸~'~~
C-J
ED.R.

17°45'N

THAILAND

17°30'N

101°00'E 101 °15'E 101 °30'E

Phon Hong Group Metamorphic Aureole f Major Faults

E Khorat Group Granite Cretaceous f Thrusts

Undifferentiated Granite Triassic • (~) Locality

C-J

Phalat Formation mentioned in text

? Late Palaeozoic to
in East
Late Jurassic Argillites,
Triassic and Jurassic Volcanics,
with local development of
Mid-Permian Limestone in West 20km
I

Fig. 2. Simplified map of the geology of the Pak Lay Foldbelt including locations mentioned in the text.
228 R.B. STOKES ET
elucidated from superposition, as follows from
base to top: (1) uncrushed crystalline basement;
(2) terrain ancien: pre-Uralian shales and sand-
stones, Uralian limestones and cherts, very highly
folded on N-S trends; (3) andesitic volcanic rocks:
no trace of crushing in the border region where they
are unconformable on older strata; (4) terrain
rouge: gently dipping in the border mountain
region; (5) gr6s sup6fieurs: horizontal or hardly
dipping, except on the Nam Lay.
Bourret's reasoning for this stratigraphic history
can be summarized as follows. (i) The age of the
granites is Permian or Early Triassic because they
are post-tectonic and therefore younger than the
intense folding which affected the region in pre-
Triassic times. The granites pre-date the Mid
Triassic 'terrain rouge' even though no clasts of
granite could be found in the red-beds. (ii) The
oldest sediments (the 'terrain ancien') are highly
folded, often vertical, sandy shales and fine sand-
stones with calcareous shales, cut by numerous
quartz veins. In the southern part of the province
they occur only in the deepest parts of the valleys,
because the higher parts are all formed by 'roches
vertes'. The strike is always N-S and dips are
usually very steep and sometimes vertical. In the
area to the east of the Mekong (and to the west
of Chiang Khan) these sediments are not meta-
morphosed, but elsewhere they are. Bourret dates
these sediments as pre-Uralian and probably
mainly Devonian. (iii) A mass of andesitic rocks
forms an elevated plateau, between the Nam Sang
Ban Nam Pou Nong
Phuk Heo Ban Kok Ke
I
NW , Nam Hoi
I
[ ]
I I
L_ ' 'I
~ i ~ . ~ . ~:~ii~i~i~ '~i~:~ii~#=~ ~'. !~=~i' ',~.,
LEGEND
Gres superieurs Jurassic
Terrain rouge
Triassic volcanics Triassic
Granite intrusion
Palaeozoic sediments
Fig. 3. Cross-section interpretation of the Pak Lay Foldbelt taken from Bourret (1925) and modified from
information in Fontaine & Workman (1978).
AL.
and Pak Leng, which is cut by deep W-E fiver
valleys on the steep eastern scarp. According to
Bourret (p. 86) 'we can observe everywhere in
these valleys the superposition of the roches vertes
on the 'ancient' shales'. Bourret found no evidence
of bedding in the andesites. On the western margins
of the plateau, however, stratified tufts showed
moderate dips. (iv) The andesites are younger than
the Uralian because they have locally 'digested'
fusulinid limestones, and they are older than
Triassic because pebbles of the andesites are
included in breccias and conglomerates of the
'terrain rouge'. They can only be, therefore, either
Permian or Lower Triassic. Bourret states that they
are horizontally bedded in general.
Workman & Page (1967; repeated in Page &
Workman 1968) carried out a photogeological
study following a field traverse of the area and
established a stratigraphy which was very similar
to that of Bourret. Their summary is reproduced
here with ages taken from their text. (1) Deposition
of Palaeozoic sediments (Upper Carboniferous to
Permian argillites; thin beds of shale and siltstone,
with occasional fine-grained sandstones. The shales
are generally black or dark grey, often studded
with crystals of pyrite). (2) Cimmerian 1 folding
(approximately Upper Triassic). (3) Intrusion
of granitic rocks. (4) Triassic volcanic episode.
(5) Deposition of Khorat Group (Triassic to
Jurassic). (6) Gentle epeirogenic folding and
large-scale faulting.
The end of the older Cimmerian 1 folding was
Nam
Hueung
', SE
'I 'I, J
+ + 4-:4- 4- 4-
 PAK LAY FOLDBELT, LAO PDR
marked by post-tectonic granodiorite intrusion,
after which erosion and planation took place before
the Triassic volcanic episode and the deposition of
the Khorat group.
In his review of 1972 (repeated in Fontaine &
Workman 1978), Workman added (inter alia) the
following points to the geology of the Pak Lay area.
(1) The intrusive rocks, mostly with compositions
in the range of granodiorite to tonalite, were
emplaced in an early phase of the Indosinian
orogeny, probably in the Early Triassic. (2) The
thick volcanic formation in the Triassic at the base
of the continental Mesozoic series conformably
underlies the gently dipping red beds of the Upper
Triassic to Jurassic. The group consists of andesite
and trachyte lava flows interbedded with siliceous
and feldspathic crystal-tufts and agglomerates.
These are approximately Middle Triassic in age.
The currently-accepted history of this foldbelt
is that presented by Fontaine & Workman (1978)
which can be deduced from the cross section in
Fig. 3 and summarized as follows: (1) deposition
of Palaeozoic sediments; (2) Indosinian orogeny
(Early or Middle Triassic) resulting in isoclinal
folding of the Upper Palaeozoic formations along
N-S axes; (3) intrusion of granitic rocks (probably
Early Triassic); (4) period of erosion and planation;
(5) Triassic volcanic episode (approximately
Middle Triassic); (6) deposition of the continental
Khorat Group (Late Triassic to Jurassic) con-
formably overlying the thick volcanic formation;
(7) gentle epeirogenic folding and large-scale
faulting.
This view of the history of the foldbelt is largely
responsible for the acceptance of a Triassic age for
the Indosinian orogeny and the related collision of
Shan-Thai to Indochina.
New evidence from the Pak Lay Foldbelt
Biostratigraphy of the sediments
Our samples of the 'Uralian' limestones in the Pak
Lay region gave only Permian ages within the
range Murghabian to Dorashamian. These ages,
Table 1. Ages of volcanic and plutonic igneous rocksfrom the Pak Lay Foldbelt determinedfrom whole-rock K-Ar
analyses (localities indicatedon Fig. 2)
Locality Lithology Radiometric date ( M a )
number
6 volcaniclastic lithic wacke
7 crystal tuff
8 basalt
9 alkali basalt
10 granite
11 granite
229
based essentially on foraminifera, come from the
localities numbered 1-4 on Fig. 2. Their ages and
lithologies are as follows: (1) Upper Murgabian to
Lower Midian fossiliferous packstone; (2) Midian
(or possibly higher i.e. Dzhulfian) fossiliferous
boundstone; (3). Midian to Dzhulfian fossiliferous
packstone and boundstone at Ban Buamlao;
(4) yielded a wagenophyllid coral of probable
Murgabian age.
The remainder of the sedimentary rock samples
are almost universally barren of fossils, but one
exception is critical to our dating of the clastic
sequence. Thinly-bedded, grey to dark grey
claystones at Locality 5 have yielded a lean,
moderately well preserved, terrestrially derived
palynomorph assemblage most characteristic of a
Middle to Upper Jurassic age. Taxa present include
Classopollis spp., Calliatisporites dampierri,
Cerebropollenites mesozoicus, Uvaesporites sp.
and P. elatioides, along with some rare, reworked
striate bisaccate forms of Permian to Triassic
character.
Radiometric ages of the volcanic rocks
Volcanic rocks have long been known from
Permian and Triassic sequences in the Pak Lay and
Loei-Petchabun foldbelts. Our data show that this
vulcanicity continued into the Late Jurassic with
the youngest age being obtained from volcanic
rocks in the far west of the area. The field relation-
ships of the Bajocian basalt from Locality 8 are
unknown, but the Kimmeridgian alkali basalt from
Locality 9 is part of a volcanoclastic sequence
which dips 50°E with a strike of 175 °, conforming
to the regional structure seen in the sedimentary
sequences, and thus suggesting that it is extrusive.
Details of the dates obtained are summarized in
Table 1.
Radiometric ages of the granites
The Triassic age of the widespread granites in the
Pak Lay and Loei Foldbelts is confirmed by our
Chronostratigraphic Age
210.4 _ 5.2 Late Triassic (Norian)
226.7 _+5.4 Late Triassic (Carnian)
167.2 +_5.2 Middle Jurassic (Bajocian)
152.4 _+6.3 Late Jurassic (Kimmeridgian)
213.8 + 8.9 Late Triassic (Norian)
117.0 _+3.0 Mid-Cretaceous (Aptian)
230 R.B. STOKES ET AL.

dating of the Ban Pakthoun granite (Locality 10)
which intrudes the (?)Upper Permian Phalat
Formation in the east of the area. The dating of the
Ban Samxong granite (Locality 11), previously
regarded as Triassic, was made to test the Jurassic
age for part (Locality 5) of the stratified sequence
which it intrudes. The mid-Cretaceous age of the
granite supports the biostratigraphy of the argillites.
Details of the dates obtained are given in Table 1.
Structure
Relationship of the volcanic rocks to the Indosinian
unconformity. The present authors have been
unable to find any evidence in support of Bourret's
conclusions, followed by all subsequent workers,
that the volcanic rocks are unconformable above
the Palaeozoic sediments, that they are con-
formable beneath the Khorat Group, and that they
post-date the major episode of folding. On the
contrary, both theoretical considerations and field
evidence lead to the conclusion that the volcanic
activity preceded the Indosinian orogenic event.
The widespread basaltic to andesitic volcanic rocks
have been interpreted, in their southern extension
within the Petchabun belt in east central Thailand,
to have formed above an east-dipping subduction
zone (Bunopas & Vella 1983). This volcanism
would, therefore, have pre-dated the collision
event. Bourret's assertion that the volcanic rocks
occupy the high ground and the sediments the low
ground (and, therefore, that the volcanic rocks are
unconformable above the folded sediments) is
demonstrably untrue. For example, the road from
Ban Kengsao (on the Mekong) to Ban Bouamlao
(in the Nam Gnang valley) cuts through the major
NNE-trending ridge to the southwest of Pak Lay.
The crest of this ridge exposes a thick sequence of
claystones and fine to medium grained tufts. All the
rocks dip steeply to the east. Passing westwards
down from the ridge to the broad alluvial valley of
the Nam Gnang, volcanic rocks predominate in the
scattered outcrops.
Imbrication in the pre-Khorat sequence. Bourret's
assertion that the structure of this region is simple
seems false. Mapping the distribution of the
volcanic rocks is particularly difficult, as Whittle
(1970) pointed out. With few exceptions, dips in the
pre-Khorat rocks are relatively steep. The intimate
association of the volcanic rocks, claystones and
Permian limestones, and their apparent conformity
suggest that they have been strongly tectonized and
show either isoclinal folding or that the Pak Lay
Foldbelt is a zone of structural imbrication which
has juxtaposed sediments and volcanic rocks of
varying ages. On the ground it was found that
there were numerous unpredictable changes from
argillites to volcanic rocks and back along our
traverses. Where bedding is seen within the
volcanic rocks, it is steeply dipping and of the same
order of magnitude and orientation as is seen in
nearby sediments. At Locality 12, volcanic tufts
strike 036 ° and dip 75 ° to the SE; oil-stained silt-
stones and claystones roughly 1 km to the SSE
at Locality 13 strike at 014 ° and dip at 76 ° to the

NW SE

Imbricate zone of Late Palaeozoic to Late

Jurassic sediments and Triassic to Late
LEGEND Jurassic volcanics
Granite and its MetamorphicAureole
Khorat Group (Cretaceous)
Late Jurassic Sediments
Triassic to Late Jurassic Volcanics
. . . . . . Mid Permian Limestones

Fig. 4. Diagrammatic cross-section showing the revised interpretation of the Pak Lay Foldbelt.
 PAK LAY FOLDBELT, LAO PDR
ESE. Due to a lack of dating of most exposures,
imbricate thrusts can only be demonstrated where
Permian limestones appear to overlie Triassic or
Jurassic volcanic rocks in the northwest of the
study area.
Pre-Khorat Group (lndosinian) unconformity.
There is clearly a major angular unconformity
between the Upper Jurassic argillites and volcanic
rocks, and the overlying Cretaceous Khorat
Group. At Locality 5, the former dip 50°E with
a strike of 008 °. The Khorat Group shows a con-
sistent regional strike (140°-160 °) and gentle dip
(c. 20°SW) in the area of Locality 14.
History of the Pak Lay Foldbelt
Whilst individually none of the evidence above is
unequivocal, these obervations together lead to the
following conclusions concerning the geological
history of the Pak Lay Foldbelt.
The western part of the Pak Lay Foldbelt consists
of (?)Upper Carboniferous to Upper Jurassic sedi-
ments and associated Triassic to Upper Jurassic
lavas, agglomerates and tufts. The Triassic to
Jurassic volcanic activity is interpreted as arc-
related and to have occurred during a period of
subduction which preceded the collision event. The
latter can, therefore, be no older than Late Jurassic.
The Upper Palaeozoic to Jurassic sequence of the
western zone was deformed by a compressional
event at the end of the Jurassic which produced an
imbricate, accretionary wedge with steeply dipping
strata striking NNE. These structures result from
the collision of the Shan-Thai and Indochina
blocks during the Indosinian orogeny.
Following a period of considerable erosion, the
Cretaceous Khorat Group was deposited uncon-
formably, overstepping the truncated strata of
Carboniferous to Jurassic age. A major granite was
intruded later in the mid-Cretaceous, probably
associated with the basin inversion in the Khorat
region before the deposition of the Upper
Cretaceous Phon Hong Group (Lovatt Smith et al.
1996). Tilting to the SW and subsequent erosion
of the uplifted Paklay-Petchabun Foldbelt has
stripped the Cretaceous from much of the area
revealing the structurally complex foldbelt. Our
schematic reinterpretation of the line of section
given by Bourret (1925) is presented as Fig. 4.
References
BARR, S. M. & MACDONALD,A. S. 1991. Toward a
late Palaeozoic-early Mesozoic tectonic model
for Thailand. Journal of Thai Geosciences, 1,
11-22.
BOURRET, R. 1925. Etudes geologiques dans la region
231
Discussion
The results presented above relate to the area west
of the Nam Khong (Mekong) river. The area to the
east of the Mekong consists essentially of less-
tightly folded Palaeozoic strata which were
intruded by granites in the Late Triassic. The line
of separation between these two major tectono-
stratigraphic terranes is difficult to fix. It was
regarded as the Kenthao-Saiapoun fault zone by
Workman & Page (1967). We have drawn it some-
what arbitrarily along a number of faults along
the eastern margin of the C-J unit on Fig. 2. The
relationship of these terranes with each other is
obscure. The Late Triassic deformation and
plutonic activity in the east represent an earlier
event in the history of the western margin of the
Indochina Block.
The revised dating of the Ban Samxong granite
resolves two anomalies noted by earlier workers.
Firstly it explains why, contrary to his expectations,
Bourret found no pebbles of the granite in the basal
conglomerates of the Khorat Group. This is now to
be expected, since the granite is of a later age.
Secondly, the Cretaceous age explains why
Workman & Page (1967) found it to be 'unlike
any of the other intrusive rocks of the area .... ', since
all the other granites which they examined are of
Triassic age.
Cooper et al. (1989) suggested that the Khorat
Group was deposited during a phase of thermal
subsidence following continental collision and
extension during the Late Triassic. However, as
Racey et al. (1996) point out, since the majority
(if not the totality) of the Khorat Group is
Cretaceous in age, a different mechanism for basin
formation must be invoked. Our results suggest that
the continental collision was in the Late Jurassic
and that the thermal subsidence model is still
acceptable.
We thank the Directors of Monument Resources
(Overseas) Limited, the Shlapak Development Company
and the Department of Geology and Mines (Vientiane)
for permission to publish this work. We are particularly
pleased to thank the numerous villagers who gave help
and hospitality during the field work. The paper has
benefited from reviews by Charlie Bristow (Birkbeck
College, London) and Ian Metcalfe (University of New
England). All micropalaeontology, palynology and radio-
metric (K-As) datings were carried out by the Geochem
Group, Chester.
de Pak-Lay (Moyen Laos). Bulletin du Service
GEologique de l'Indochine, 24.
BUNOPAS, S. & VELLA, P. 1983. Tectonic and geologic
evolution of Thailand. In: Ntra'ALAYA, P. (ed.)
Proceedings of a workshop on stratigraphic
232 R.B. STOKES ET AL.
correlation of Thailand and Malaysia, Bangkok.
1, 307-322.
COOPER, M. A., HERBERT, R. & HILL, G. S. 1989. The
Structural Evolution of Triassic Intermontane
Basins in Northeastern Thailand. In:
THANASUTmPITAK, T. & OUNCHANUM, P. (eds)
Proceedings of the International Symposium on
Intermontane Basins: Geology & Resources,
Chiang Mai. 231-242.
FONTAINE, H. & WORKMAN, D. R. 1978. Review of the
Geology and Mineral Resources of Kampuchea,
Laos and Vietnam. In: NUTALAVA, P. (ed.)
Proceedings of the Third Regional Conference on
Geology and Mineral Resources of Southeast Asia,
Bangkok November 14-18, 1978. Asian Institute of
Technology, Bangkok, 539-603.
HUTCHISON, C. S. 1989. Geological Evolution of South-
East Asia. Oxford University Press.
LOVATT SMITH, P. E, STOKES, R. B., BRISTOW, C. &
CARTER, A. 1996. Mid-Cretaceous inversion in the
northern Khorat Plateau of Lao PDR and Thailand.
This volume.
METCALFE, I. 1988. Origin and assembly of south-east
Asian continental terranes. In: AUDLEY-CHARLES,
M. G. & HALLAM, A. (eds) Gondwana and Tethys.
Geological Society, London, Special Publication,
37, 101-118.
PAGE, B. G. N. & WORKMAN,D. R. 1968. Geological and
Geochemical Investigations in the Mekong Valley,
between Vientiane and Sayaboury and at Ban Houei
Sai. Institute of Geological Sciences, Overseas
Division, Report, 9.
RACEY, A., LOVE, M. A., CANHAM, A. C., GOODALL,
J. G. S., POLACHAN, S. & JONES, P. D. 1996.
Stratigraphy and reservoir potential of the Mesozoic
Khorat Group North Eastern Thailand: Part 1,
stratigraphy and sedimentary evolution. Journal of
Petroleum Geology, in press.
WHITTLE,G. 1970. Pak Lay Sheet. In: Notes on 1:250 000
Photogeological Reconnaissance Map of Laos.
Institute of Geological Sciences, Overseas Division.
WORKMAN, D. R. 1972. Geology of Laos, Cambodia,
South Vietnam and the eastern part of Thailand -
a review. Institute of Geological Sciences, Overseas
Division, Report, 19.
-- & PAGE, B. G. N. 1967. Interim Report on the
regional geology and geochemistry of the
Sanakham-Pak Lay area of western Laos. Institute
of Geological Sciences, Overseas Division, Report,
8.
 Mid-Cretaceous inversion in the Northern Khorat Plateau
of Lao PDR and Thailand

PAUL F. LOVATT SMITH l, R O B E R T B. STOKES 2, C H A R L I E B R I S T O W 3

& ANDREW CARTER 3
1 25 Tredegar Road, London E3 2EH, UK
2 Surbiton Geological Services, 14 Dennan Road, Surbiton, Surrey KT6 TRY UK
3 Department of Geology, Birkbeck College, University of London,
Malet Street, London WC1E 7HX, UK

Abstract: Evidence is presented from the study area for the occurrence of a regional com-
pressive tectonic event in the mid-Cretaceous(Aptian-Cenomanian). This is tentatively attributed
to the effects of a distant continent-continent collision to the west. Pre-existing structural trends
were reactivated parallel to palaeo-continental sutures to the northeast (Song Ma and Song Da)
and west (Nan). The event interrupted the latest Jurassic to earliest Palaeocene subsidence and
continental sedimentationof the Khorat Plateau Basin. The primary effects are shown by seismic
data to be regional tilting, compressive folding, reverse faulting and inversion of the basin
subsidence pattern. The data do not support the current view that such structuring was initiated
during the Tertiary. However they do show that there was some reactivation of mid-Cretaceous
structures during regional uplift in the Tertiary. This revision of timing of structuring suggests the
hydrocarbon potential of the area may be greater than previously anticipated.

The study area comprises the MonumenffShlapak
Contract Area in the Lao PDR and the western
margins of the Khorat Plateau in Thailand (Fig. 1).
The data presented in this paper results from
Monument's hydrocarbon exploration programme
in Lao PDR, and from sedimentological studies of
the Khorat Group in Thailand. Geological field-
work in the Lao PDR, totalling 19 weeks, was
undertaken by PLS, RBS and others during the dry
seasons of November-December 1991, January,
March-April and November-December 1992, and
November-December 1993. This resulted in the
production of 30 geological maps at a scale of
1:100,000. Samples were analysed in the labora-
tory for petrography, geochemistry, biostratigraphy
and isotopic dating. Geophysical surveys in the
Vientiane Basin were conducted in April 1992 and
March-May 1993 and resulted in 4 2 0 k m of
seismic, 2146 km of airborne gravity, some 500 km
of surface gravity and 6450 km of aeromagnetic
data. Fission track analysis on samples from the
Khorat Group cropping out in the Khorat mono-
cline of Thailand was carried out by CB and AC.
Regional setting
The Lao PDR study area is situated on the northern
and western margins of the Khorat Plateau Basin.
This is a large area of cropping out Mesozoic
(mainly Cretaceous) continental deposits covering
the northeast of Thailand, parts of Lao PDR and
Cambodia (Fig. 1). The basin system was once
more extensive but has been reduced in area and
thickness by significant regional Tertiary erosion.
Present-day sub-basins of the Khorat Plateau
Basin are defined by the distribution of Upper
Cretaceous sediments. The Vientiane and Pakxan
topographic plains are the Lao extensions of the
Upper Cretaceous Sakhon Nakhon Basin. Around
the northern margin of this basin the underlying
Khorat Group (uppermost Jurassic to Aptian) is
exposed in a pattern of eroded anticlines and out-
lying synclines (Fig. 2). The Phetchabun Foldbelt
separates this eastern area from the Nakhon-
Thai/Sayabouri Basin where Mesozoic sediments
belonging to both the Upper Cretaceous Phon Hong
Group and the Khorat Group are also preserved.
The similarity of the Cretaceous succession of this
basin to that further to the east indicates that both
areas can be considered as part of the same basin
system. The surrounding massifs consist mainly of
Precambrian to Upper Palaeozoic rocks, though
some Upper Palaeozoic to Upper Jurassic sedi-
ments and volcanics are exposed in the Phetchabun
Foldbelt.
Two structural trends are present in the study
area. An approximately north-south trend pre-
dominates in the west. In the Nakhon-Thai/
Sayabouri Basin, Cretaceous sediments are gently
folded and faulted on this trend. Underlying sedi-
ments in this area and in the Phetchabun Foldbelt
have been more severely tectonized by east-west

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolutionof SoutheastAsia, 233

Geological Society Special Publication No. 106, pp. 233-247.
234 P.F. LOVATT SMITH ET AL.

1 i
103~E 108:E
CHINA ~-~"N..
.~j'X..'-'-/'"
~ " " ""o 250km

\
t
°.
L-.j"
VIETNAM
0 Hanoi
\-,~.
- o,' %
~.°
Gu!f qt Tongkin
.? 20~S

Phrabang .'~. f

~.0 E D.

Uttaradit
0 South
~:LL~ ).' SAKHON NAKHON• • . "%'%O/k~ China
NAKHON THA ,'~..
),-I7
"1,¢:.. "0" :.BASIN• " T l l a l ~ h e d ~
L'udon••.~•i•i'••'.•.'. Sea
I. BASIN. '.•-~ t~2•.Thani. ~ • • • •. -. ~ ,,,

Phetchabuno :
i
~.J J..v......

•
.........
• : "
:•-.-.• .,,.,~
.,~/~:--•.... S4~tvanna~h~t~'

• ::::: ::::: K,.ORAT.BAS.~N.:.:.:.:.:...~

• i.l.i.i.i.i.i.i•i•i.i.i.i.i.i.i
• •

Ubon&i.i.i~'\."
~t

•
{

~.'~
•:':'.•.-..
~s,
z/c,),. • •
~'
Da Nang

15S i.i•i.i&i.i.i.i.i.i.~.i.i.i.i.i•i.i.i.i.i, ili~"~lPakse ,~'" '

THAILAND ~i
'.°.'.°•'." "." "°••• • '.'
. . i • : • : • : • i • : - i .; . ; - . - ~ " - : - : ~ ' - - . = - - " ~
....' - - ;- - . f L
•°• ".•.'

"
."
•

I"'"..
,-,"'/(: \ "
%. Bangkoko
%. "~

d •

'--" "I
X'\ (-o°

AMBODIA (~ ~' .. •.

~~:-:..-.::--,..- ~,0oml ,..j

__.,i
::,;•:~5.::: ;'~,
Gulf of Thailand ~,-
. . . . . .
.,~1..~ I
t..
",
r-
, . .j •
,.i .r... ... . . ~•
/ - "
103E~.x•/ F.'.~, ~ '~ ~. • . ",,108
I - /.' .' .~ ~. I "t I

LEGEND O Outcrop of Jurassic / Cretaceous

sediments
Axis of major foldbelts
/ Major Faults
• Granites with Cretaceous ,, - Subcrop of Jurassic / Cretaceous
K/Ar age dates against Tertiary / Quaternary Study Area

Fig. 1. Simplified geological map of the Khorat Plateau region•

compression. Along the east side of the foldbelt eastern part of this Basin a northwest--southeast
Mesozoic sediments reappear in the Khorat structural trend becomes more dominant, parallel to
Monocline, which defines the westem margin of the Annamitic Foldbelt. This trend is characterized
the Khorat Plateau Basin. Within the central and by large wavelength anticlines which have topo-
 CRETACEOUS INVERSION IN THE KHORAT PLATEAU 235

, , 1 ...) , , , f

PO. ,:.. 1
XIENG-KHOANG
---~MASS,F
"~"°'~.
"....
V ! ET NA M /!

~ :::::::::::::::::::::::::::
.J ~ : : : : : : : : : : : : : : : : : : : : : :

? lo~k~
.EGEND (~ Phon Hong Group Outcrop (~ Paleozoic or Basement Outcrop , ~ ' Major Syncline

Khorat Group Outcrop ,,If" Major Anticline J Major Lineament/Fault Line of Section

Fig. 2. Lao PDR study area showing the major structural elements and location of seismic lines of Figs 5-8.

graphic expression. These include the Nam Leuk,
Khammouan and Phu Phan Uplifts (Figs 1 and 2).
The structures are often bounded by deep-seated,
high angle reverse faults with large throws, often of
more than 1 km.
Stratigraphy
The stratigraphy of the Vientiane Basin is summa-
rized in Fig. 3 and is based on that proposed by
Ha Luong Tin & Do The Que (1991), being here
modified and further subdivided in order to facili-
tate correlation with the better known sequences
in Thailand. The maximum thicknesses indicated
in Fig. 3 are based mainly on geophysical data
supplemented by outcrop measurements. The
chronostratigraphic section in Fig. 4 is based on an
east-west seismic line across the Vientiane Plain.
The Permian is present as marine platform and
marginal facies, comprising limestones and clastic
sediments. The Triassic crops out only in the west
of the study area where it is present as extrusive
volcanics and volcaniclastics. To the east its facies
is assumed to be fluvio-lacustrine, similar to its
equivalents in Thailand. The Khorat Group com-
prises mainly fluvial sediments. The Lao Khorat
Group stratigraphy substantially follows that of
Thailand, recently updated by Racey et al. (1994),
except that the lowest unit, the Upper Norian to
Rhaetian Phulekphey Formation (Thai Nam Phong
Formation equivalent), has been assigned to the
pre-Khorat section. This is because a long hiatus
between this formation and the overlying Nam Set
Formation is indicated by the absence of Lower-
Middle Jurassic sediments.
The Phon Hong Group stratigraphic nomen-
clature is based on the results of potash exploration
carded out between 1983 and 1986 by Vietnamese
geologists, quoted by Ha Luong Tin & Do The Que
(1991). In the Vientiane area the Thangon
Formation reaches a maximum thickness of 700 m
and consists of three evaporitic sedimentary cycles.
A basal conglomerate has been noted in one
locality in the west of the Vientiane Basin (Howlett
1993). The basal evaporite cycle contains a halite
bed which reaches a maximum thickness of 340 m
in boreholes. In the southeast part of the Vientiane
Plain this salt has been mobilized into pillows and
small diapirs. The evidence for an Albian to
Cenomanian age for the Thangon (Maha Sarakham
equivalent) Formation evaporites is based on bio-
stratigraphic data from palynological analysis of
cores taken in the Sakhon Nakhon Basin of
Thailand (Sattayarak et aI. 1991). This formation
was identified by Utha-Aroon (1991) as being
deposited in a continental setting.
The Saysomboun Formation is up to 150 m thick
in the Vientiane area. It consists of predominantly
red-brown fluvial claystones grading upwards into
siltstones and fine sandstones. The claystones are
dated by palynology as Upper Cretaceous both in
the Vientiane by Ha Luong Tin & Do The Que
236 P.F. LOVATT SMITH ET AL.

MAX THICKNESS VIENTIANE ABBR. THAI

(KM) STRATIGRAPHY EQUIVALENT
0 .°o..•l.O.o°. A
SAYSOMBOUN FM. K2Sb PHU THOK FM.
A A A A A A A
LATE A A A A ^ ^ A ~ THANGON FM. K2Tg MAHASARAKHAM FM.

~ ~ -
W
BAN THALAT FM. K 1Bt KHOK KRUAT FM.
0 CHAMPA FM. K1C p PHU PHAN FM.
W ~ BAN ANG FM. K 1Ba SAO KHUA FM.
.,-. . .-~'-
. . . . . . o. .. ,. . . . . . . 0 PHU PHANANG FM. K 1Pp PHRA WlHAN FM.
o EARN . . . . .

711lllllll I ~
E
o NAM SET FM. KINS PHU KRADUNG FM.
z
1°o.......o... ~
? JUR.
-
o.oV"" • i . e V . " ".'4V."
PHULEKPHEY FM. TPI NAM PHONG FM.*
3
iTRIASSIC LATE ,,i ,,,,, ,-- HUAI HIN LAT FM. THh DAT FA SHALE
• V" "V" "V • •
LOWER TUFF / SST Tv PHU HI SST.

PHALAT FM. PPI U. PERMIAN CLASTICS

~- LATE
NALANG FM. PNI PHA NOK KHAO FM.
s

6 , LATEST
tl LOWER CLASTICS

NAMURIAN "-" 'm

rr" -VISEAN I K NAM THOM FM. CNt
7 O ,.--
~ ~ ~ .

MIDDLE i i J i i i t I NASAFM. DNs

DEV- J ] I [
ONIAN EARLY ~

z * The Nam Phong

,< Metamorphic
Basement NAM XOI FM. O/S formation is currently
included in the
O .J Khorat Group in
Thai stratigraphy.

Fig. 3. Generalized stratigraphy, Vientiane area, Lao PDR.

(1991) and in the Savannakhet Basin of Central Pakxan Plain. This represents the northern continu-
Laos (Huynh Xuan Dang and Boutheung ation of the Thai escarpment in which Phu Tok
Phengthavongsa 1989). In the Pakxan Plain the Formation sandstones were identified as having
upper part of the Saysomboun Formation is pre- aeolian characteristics by Sattayarak (1983). In the
served and contains massive beds of red-brown, Nakhon-Thai/Sayabouri Basin to the west this
cross-bedded sandstones. One of these sandstone formation is substantially thicker than in the
units is over 180 m in thickness and forms a Vientiane and Pakxan areas. Some authors have
scarped ridge (Sayphou Khout) in the centre of the assigned an Lower Tertiary age to this formation.
 CRETACEOUS INVERSION IN THE KHORAT PLATEAU 237

0 km 50
WEST I I EAST

AGE TECTONICS PHETCHABUN FOLDBELT [< KHORAT PLATEAU BASIN q,

EOC Transtension
and Uplift ~ KHORAT MONOCLINE ~ VIENTIANE (SAKHON NAKHON) BASIN =~ NAM LEUK UPLIFT

PAL
TERTIARY EROSION
Flexural
LATE Subsidence
K 2 ~ . . . . . . . . . . . . . . . . .^.A -- AA~
"'~
(J
Collision?

O
W
(5
EARLY
Flexural
I ~ , ~ , ~ . . . . --.._..= ...=
Subsidence K1p
K1Ns~_Ej~j_ -- -- ]- j
~ -- -- ...~.._J_-- -- i i .i ,i
. ~ ~ , I - - - - I I - - ~ 1 ] 1 i , - - - - Z I I - -

-- _ -- -- -- -- -- -- -- --o--o ~6"-(~ -- -- ---- o 0% -- -- --

DOGGERI
Quiescence JURASSIC HIATUS

LIAS
J PI
LATE Extension t h r u st ~ ' . . _ . . - , . ~ . . . . ~ ~ -
L~

~ MIDDLE INDOSINIAN OROGENY

Continental
Collision
EARLY
P~
LATE
Active.,'
Passive
LU EARLY Margin
o_

Fig. 4. Chronostratigraphic diagram for Vientiane area, Lao PDR.

The present authors prefer to place it in the Upper stress being fed to the Phetchabun Foldbelt via
Cretaceous, based on the palynological evidence the Mae Ping strike-slip fault zone. They inferred
cited above. a Late Cretaceous to Early Tertiary age for this
Neogene and Quaternary alluvial sediments inversion.
occur locally, infilling abandoned river channels. Most other authors interpret the inversion as
In boreholes they reach thicknesses of up to 80 m. beginning after the deposition of the Phon Hong
No older Cenozoic sediments are found in the Group, during the Tertiary. Mouret et al. (1993)
Mesozoic basinal areas. date the main inversion as beginning at c. 65 Ma
(earliest Palaeocene). This is based on fission track
data from outcrops in Thailand of the Phra Wihan
Current interpretations of post- Formation indicating that the maximum palaeo-
Early Cretaceous inversion history temperature, and therefore the onset of uplift,
There are currently two main interpretations of occurred at that time. These and other data are
the post-Early Cretaceous structural history of the interpreted to demonstrate uplift and erosion of
Khorat Plateau Basin, relating to the timing of 3000 m for surface outcrops of Khorat Group sedi-
inversion with respect to the deposition of the ments in the Phu Phra area of Thailand, which is
Upper Cretaceous Phon Hong Group. on the northem margin of the Phu Phan Uplift.
Cooper et al. (1989) interpreted the Khorat A similar age for the onset of inversion is also
megasequences as being deposited in a basin supported by Kozar et al. (1992).
formed by thermal subsidence during the Jurassic
and Cretaceous, which was inverted prior to the
Evidence of a two phase inversion history
deposition of the Maha Sarakham Formation
evaporites. This inversion created a rimmed While our fission track data are in close agreement
basin(s) within which the evaporites were with those of Mouret et al. (1993) indicating that
deposited. The inversion was stated to be caused substantial uplift and erosion began at 60 Ma (mid-
by the early stages of collision between India and Palaeocene), other data show that this was the
Eurasia in the region of central Burma (Myanmar), second phase of structuring to have affected the
238 P.F. LOVATT SMITH E T A L .
Khorat Group. The initial event occurred prior to
the deposition of the Phon Hong Group in the mid-
Cretaceous and was responsible for regional tilting,
compressive folding, reverse faulting and inversion
of the basin subsidence pattern. The Tertiary event
reactivated some of these trends but its main effect
was to produce regional uplift and substantial
erosion. This two phase model is consistent with
the evidence published from Thailand mentioned
above.~ The evidence in the study areas for this
interpretation is listed below.
Unconformity at the base of the
Phon Hong Group
The Upper Cretaceous sedimentary cover has been
more or less completely eroded in many areas of
the Khorat Plateau due to Tertiary uplift, obscuring
the evidence of mid-Cretaceous structuring.
However an unconformity at the base of the
Thangon (Maha Sarakham) Formation has previ-
ously been noted from seismic data in Thailand
(Sattayarak et al. 1991). Fortunately, in the
Vientiane area of Lao PDR. the preservation of a
relatively thick section of Phon Hong Group sedi-
ments means that the character of the unconformity
is readily discernible.
The unconformity has been noted from shallow
borehole data at the margin of the Vientiane Plain
of Lao PDR by Ha Luong Tin & Do The Que
(1991). A conglomerate containing clasts of silt-
stone, interpreted to be derived from the Khorat
Group, has also been identified at the base of the
Phon Hong Group at outcrop in the west of the
Vientiane Basin (Howlett 1993). Seismic data from
the Vientiane area show that this unconformity
represents a more significant tectonic event than
was previously realized. Four seismic lines (Figs
5-8) demonstrate this.
Figure 5 shows the aspect of the Phon Hong
Group towards the northern limit of its distribution.
Horizontal Thangon Formation reflectors onlap
uneroded Khorat Group both in the west (onto the
Khorat Monocline) and in the east (at the base of
the Nam Leuk Uplift). Although the shallowest
reflectors may be confused by recent weathering
effects it is apparent from these data that most of
the Nam Leuk Uplift and Khorat Monocline
structuring in this area occurred during the mid-
Cretaceous. The horizontal dip of the Phon Hong
Group reflectors shows that subsequent Tertiary
reactivation of these structures has been relatively
minor.
Figure 6 is a detail from a north-south line which
intersects the line of Fig. 5. It demonstrates
erosional truncation during the mid-Cretaceous of
the upper Khorat Group reflectors on a reverse
fault-bounded horst. There is clear onlap by
horizontal Thangon Formation reflectors over this
feature. The reverse faults moved during or before
the deposition of the Lower Salt Member of
the Thangon Formation evaporites (Albian-
Cenomanian), since only the highest part of the
Early Cretaceous horst has been eroded.
The large reverse fault-bounded anticline seen on
Fig. 7 is along a surface trend from the unfaulted
Nam Leuk Uplift anticline seen in Fig. 5. The relief
on both structures is of a similar order of between
1000 and 1400 m. The bulk of the reverse fault
movement is therefore inferred to have occurred
during the mid-Cretaceous event. It is interpreted
that the movement occurred both before and during
the deposition of the Thangon Formation, giving
rise to a fault scarp and localized development of
a syn-fault talus fan, against which the Thangon
Formation evaporites were deposited. Some
Tertiary reactivation is evident on this fault, shown
by dipping Phon Hong Group reflectors on the
upthrown side.
Finally, Fig. 8 is a detail from the western margin
of the Upper Cretaceous evaporite basin. Again, the
shallowest horizontal reflectors in the Phon Hong
Group may be confused by recent weathering
effects but westward onlap at the base of the Phon
Hong Group is strongly indicated. Together with
the onlap seen along trend on Fig. 5 this indicates
that the regional tilting of the Khorat Monocline
was initiated in the mid-Cretaceous.
Sediment thickness and the distribution of
halokinetic structures
Figure 9 illustrates three seismic lines from the
Vientiane Plain which demonstrate that an inverse
thickness relationship exists in this area between
the present day thicknesses of the Khorat Group
and Phon Hong Group. The data show the presence
of a complete section of Khorat Group sediments,
thickening towards the northeast from 1900 m on
line 1 to 4000 m on line 3. This demonstrates that
during the Early Cretaceous there was an increase
in the rate of subsidence and deposition from south-
east to northwest.
The subsidence pattern during the Late
Cretaceous can be interpreted by examining the
distribution of halokinetic structures in the
Thangon Formation. The salt pillows and diapirs
are assumed to have been generated by increasing
overburden pressure of Upper Cretaceous sedi-
ments up to the time of maximum burial during
the earliest Palaeocene. Line 1 shows that the most
intense salt movement has occurred in the south-
east, indicating that this was the area of thickest
Upper Cretaceous palaeo-overburden. However
 CRETACEOUS INVERSION IN THE KHORAT PLATEAU 239

TWO WAY TIME (SECONDS) TWO WAY TIME ( S E C O N D S )

LO o LO CO ~'~ o Lo
d -,~ ",~ d c5 ~" ',-
, I . I

LL
_J
B-

=)
III

,<
Z

~f

-2

c4

=
o

° ,,,,~

° o~

6
z
o

m I-

o o o. ~ ~ g g
o co
+ T ~ ,
r~

(W) H l d 3 0 31Vl~lXOHddV (w) Hld'::lO 3 1 V I N I X O H d d V

240 E F. LOVATT SMITH ET AL.

TWO WAY TIME (SECONDS) TWO WAY TIME (SECONDS)

~, d d d d d d d d
I I I I

2
",,'ca_

~g

c4
d~

>

d
Z
.~,

g,
m

E
• I I I
g g g g g ~ ~ g
r- co i'.- cq
+ ' ' 7. , , ,,7

(LU) H l d 3 0 :llVl/~lXOklddV (tu) H l d 3 a 31"v'lAIIXOt::lddV

 CRETACEOUS INVERSION IN THE KHORAT PLATEAU 241

TWO WAY TiME (SECONDS) TwO wAY TiME (SECONDS)

o 04 ~ qD 03
o o o o o d 6 c5 c5 c5
I I I I J

c4

6
Z
O

r~

I ! I I
~ o
+ cxl
, co
, ~_

(uu) H i d 9 ( ] :lLVl,gllXOk:lddV (w) H/d=IQ =I/V~IXOEIddV

ej
. ,..,.

l.z
P
'D
~. APPROXIMATE DEPTH (m) APPROXIMATE DEPTH (m)
-- _ +
-h ,,.b ..~ ~ +
o o o
~" I I I I I I I I I
3
o
g\
Z
9 i
4~
_c
t~ I
0
4~
I°
1",3
~i!ii!ii~iiil
I I t J
~, t:D o o o 9 o .o .o p
S Q N O 0 : : : I S ) 31~lJ_ A V M O N L L SGNOO3S) :::IIAII1 A r M O M /
YV , / 3 H.LIIAIS .LLVAO'-I "a "d ~
 CRETACEOUS INVERSION IN THE KHORAT PLATEAU 243

Seismic Line I
E

i] km VTN93-01

L o c a t i o n m a p of s e i s m i c L i n e s

Seismic Line 2

VIENTIANE BASIN,
LAO P.D.R.

10 lkm VTN93-04

Seismic Line 3
3000 2800
KEY

Post - lower salt sediments

(Phon Hong Group)

Lower salt, Thangon Fm.

(Phon Hong Group)

Khorat Group

~Km v i ~IU~

Fig. 9. Seismic montage showing relationship between Thangon salt diapirism and Khorat Group thickness.
244 P.F. LOVATT SMITH ET AL.
seismic lines 2 and 3 show that towards the north-
west no salt movement has occurred, indicating a
thinner Upper Cretaceous palaeo-overburden. The
opposing thickness trends between Khorat and
Phon Hong Groups are shown by this data to be
subsidence-related and demonstrate a mid-
Cretaceous inversion of the subsidence pattern.
It should be noted that in the Vientiane area
gentle salt pillows are currently present at very
shallow depths (Fig. 8), supporting fission track
data which show that a substantial amount of
Tertiary uplift and erosion occurred after maximum
burial in the earliest Palaeocene. In the area of
the Vientiane Plain the authors have calculated
maximum Tertiary erosion of c. 1000 m from the
character of the halokinetic structuring.
Abrupt facies change
At the level of the mid-Cretaceous unconformity
there is an abrupt change in facies from the Aptian
Ban Thalat Formation of fluvial claystones, silt-
stones and sandstones to the Albian-Cenomanian
Thangon Formation of restricted, continental
evaporites. This facies change occurs throughout
the Khorat Plateau Basin. It is difficult to envisage
any other way that such a facies change could have
been brought about other than by the effects of
tectonism. These evaporites were suggested by
Cooper et al. (1989) to have been deposited in
restricted basins created by the topographic
expression of syntectonic uplifts. Such potential
drainage barriers are exemplified by the marginal
uplifts visible on Figs 5-8. Three successively
thinner evaporite cycles are present in the Phon
Hong Group indicating three progressively more
minor tectonic episodes.
Presence of trapped hydrocarbons
The presence of thermally-derived gas in reservoirs
within closed structural highs has been success-
fully demonstrated by many exploration wells in
the Khorat Plateau of Thailand. Published seismic
data over these structures show that they are
defined by internally conformable, folded Khorat
Group sediments (Kozar et al. 1992). This indicates
that the structuring episode which formed these
traps post-dates the Early Cretaceous.
It can be assumed that the source rock for the gas
lies in the pre-Cretaceous section since organically-
rich sediments within the Khorat and Phon Hong
Groups are volumetrically insignificant. It is
generally agreed that the most likely source rocks
are within the Permian and Triassic section. The
authors' burial history modelling indicates that the
period of peak generation of hydrocarbons from
these source rocks occurred during the Late
Cretaceous burial. Thereafter only relatively minor
quantities of hydrocarbons are likely to have been
generated.
Since trap formation nearly always pre-dates
hydrocarbon generation these two timing con-
straints impose a mid-Cretaceous age of structural
development for such gas-bearing traps, further
supporting the evidence for mid-Cretaceous
tectonism.
Lack of structural dip in Phon Hong Group
At outcrop, the Phon Hong Group rarely dips at
more than a few degrees, whereas the Khorat Group
shows steeper dips and is even vertical or over-
turned adjacent to major faults. Although this could
be explained by erosion of steeply-dipping Phon
Hong Group sediments during the Tertiary, it is
interpreted by us as supporting evidence for a
mid-Cretaceous event.
Cretaceous granite intrusions
K-Ar dating on granite intrusions from various
parts of the Monument Contract Area gives ages of
78.3_+ 1.7 Ma (Campanian, Ban Boung Koang,
No. 3 on Fig. 1), 94.9 + 2.1 Ma (Cenomanian,
Keng Daeng, No. 1 on Fig. 1), l 1 7 + 3 . 0 M a
(Aptian, Muang Kenthao, No. 2 on Fig. 1), and
137.2 +_3.0 Ma (Valanginian, Ban Phontiou, No. 4
on Fig. 1). Although no detailed geochemical
analysis has been carried out, these granites are
inferred to indicate crustal melting following the
Indosinian continent-continent collision. The
timing of their intrusion may be coincident with
interruptions in the post-collision Cretaceous
thermal subsidence of the Khorat Plateau Basin
including, at c. 95 Ma, a mid-Cretaceous event.
Fission track data from Thailand
Data were obtained from seven outcrops of the Phra
Wihan and Khok Kruat Formations on the Khorat
Monocline in Thailand southwest of Khon Kaen.
The data show a spread of ages all of which are
younger than the depositional age (c. 145-112 Ma).
This indicates that the fission tracks have experi-
enced some annealing. Despite variations in mean
age from 84 + 8 to 40 + 5 Ma, statistical modelling
indicates that cooling began prior to 60 Ma and that
the rock has experienced temperatures of around
l l0°C. The model shows that the Cretaceous
thermal history is poorly constrained because it has
been overprinted by later heating. The dominant
thermal event was at 60Ma, but there is no
evidence to suggest that an episode of uplift did
not occur in the mid-Cretaceous.
 CRETACEOUS INVERSION IN THE KHORAT PLATEAU 245

Regional tectonic model

The pre-Cretaceous tectonic history of the Indo-
china region is still not well understood. However
• s c.,,A / J
there is agreement that a number of continental
fragments including the Indochina Plate, on whose
northern margin the Khorat Plateau Basin is
situated, rifted from the northern margin of
Gondwana during the Palaeozoic. These fragments
drifted northwards, eventually accreting to the
Eurasian margin during a series of continent-
continent collisions (Mouret 1994; Metcalfe 1996)
of which the India-Eurasia collision is the latest. i
The timing of the collision of the Indochina Plate
with its two neighbours, the South China Plate and
the Shan Thai Plate still requires more research.
Recent data from the Pak Lay Foldbelt show that
tectonism occurred during the Late Jurassic (Stokes
et al. 1996), indicating that the Shan-Thai-
Indochina collision may have occurred later than
previously thought. However it is evident that at the
onset of Late Jurassic-Cretaceous basin subsidence
in the Khorat Plateau and surrounding areas both 0 1000

of these sutures had already closed (Fig. 10). km

The Khorat Plateau and other continental sedi-

mentary basins in the region (Fig. 1) may have
formed as foreland basins as a result of flexural
subsidence in front of the orogenic belts created by
these sutures. Amongst the supporting evidence for
this theory is that the depocentre of the Lower
Cretaceous sedimentary basin in the Khorat Plateau
runs parallel to the Song Ma-Song Da suture and
Annamitic Foldbelt trend (N. Sattayarak, pers.
comm. 1994) and that palaeocurrent data in the
Khorat Group show a source from the Annamitic
Foldbelt (Howlett 1993; Mouret et al. 1993). Also,
there is no evidence of major syn-depositional
faulting in the Khorat Plateau Basin, indicating that
the stretching (13) factor was low.
The onset of structuring and inversion of the
Khorat Group as described in this paper can no
longer be attributed to the collision of the India
Plate with central Burma, as was proposed by
Cooper et al. (1989), since this collision began
during the Eocene (Daly et al. 1991). An alternative
tectonic model must therefore be invoked.
In the absence of a more convincing alternative,
we suggest that mid-Cretaceous tectonism is best
explained by a continental collision to the west,
such as that of Western Burma with Shan-Thai
which is stated by Metcalfe (1996) to have occurred
in the Cretaceous. The constraining forces created
by such a collision would have placed the entire
region in compression, leading to the structural
effects already described. The Nan suture runs
parallel to the north-south Phetchabun Foldbelt
trend, and the Song Ma-Song Da suture runs
parallel to the northwest-southeast inversion trend
Fig. 10. Pre-Cretaceous suture zones in SE Asia (after
Cooper et al. 1989).
described above (Fig. 10). Both of these trends
were compressionally reactivated during the
mid-Cretaceous event in the study area. Subsidence
resumed following this hiatus, initially in new
evaporitic basins created by the inversion, and
continued until the earliest Palaeocene. From this
time onwards the Indochina area was subjected to a
transtensional stress regime. Rifting occurred
during the Oligo-Miocene both to the east (South
China Sea) and west (Thailand Central Plain and
Gulf of Thailand) of the Khorat Plateau (Daly et al.
1991), however there is no evidence of rifting in the
Khorat Plateau itself. The evidence already cited
indicates that the study area underwent some
reactivation of mid-Cretaceous trends but that the
main effect was to produce regional uplift and
substantial erosion of much of the Upper
Cretaceous sedimentary cover.
Implications for petroleum exploration
Burial history models which assume a Tertiary age
for the timing of post-Early Cretaceous inversion
find it difficult to explain why, if trap formation
post-dated the main period of hydrocarbon
generation, most, if not all such inversion structures
which have been tested by drilling have proven to
contain gas. As discussed above the presence of
hydrocarbons in structural traps defined at the base
246 P.F. LOVATT SMITH ET AL.
of the Khorat Group is much better explained by
mid-Cretaceous timing of structural development.
Tertiary reactivation of mid-Cretaceous structures
has nevertheless enhanced the volume of some
traps, which might explain why some structures are
underfilled with gas (e.g. the Nam Phong gas field).
The implication is that the petroleum exploration
potential of the area is not only confirmed but may
be greater than previously anticipated.
References
COOPER, M. A., HERBERT, R. & HILL, G. S. 1989. The
structural evolution of Triassic intermontane basins
in northeastern Thailand. In" THANASUTHIPITAK,T.
(ed.) Proceedings of International Symposium on
Intermontane Basins: Geology and Resources,
Chiang Mai. 231-242.
DALY, M. C., COOPER,M. A., WILSON, ]., SMITH, D. G.,
HOOPER, B. G. D. 1991. Cenozoic plate tectonics
and basin evolution in Indonesia. Marine and
Petroleum Geology, 8, 1-21.
HA LUONG TIN & DO THE QUE. 1991. Determination
of salt-bearing stratum depth and its structure
in Vientiane region by geophysical data. In: TRAN
VAN TRI (ed.) Proceedings Second Conference on
Geology of lndochina, Hanoi. 2, 32-38.
HOWLETT, P. 1993. Sedimentology of the Khorat Group.
MSc Thesis, Birkbeck College, University of
London.
HUYNH XUAN DANG & BOUTHEUNG PHENGTHAVONGSA.
1989. Rock salt of the south of Laos. In: Geology of
Kampuchea, Laos and Vietnam. Dept of Mines and
Geology, Hanoi, 1, 28-30.
KOZAR, M. G., CRANDALL, G. E, & HALL, S. F. 1992.
Integrated structural and stratigraphic study of the
Khorat basin, Rat Buri Limestone (Permian)
Thailand. In: PIANCHAROEN,C. (ed.) Proceedings
National Conference on Geologic Resources of
Thailand: Potential for Future Development.
Bangkok, 692-736.
METCALFE,I. 1996. Pre-Cretaceous evolution of SE Asian
terranes. This volume.
MOURET, C. 1994. Geological history of northeastern
Thailand since the Carboniferous. Relations with
Indochina and Carboniferous to Early Cenozoic
We thank Monument Resources Overseas Limited and
Shlapak Development Company for permission to publish
as well as colleagues in Monument for their help and
advice. We also thank our numerous Lao and Thai
colleagues, particularly in the Department of Geology and
Mines, Ministry of Industry, Vientiane and the Mineral
Fuels Division, Department of Mineral Resources,
Bangkok, for their help, and the many Lao villagers for
their hospitality during the fieldwork.
evolution model. In: Proceedings International
Symposium on Stratigraphic Correlation of SE Asia,
Bangkok. 132-158.
--, HEGGEMANN,H., GOUADAIN,J. & KRISADASIMA,S.
1993. Geological history of the siliciclastic
Mesozoic strata of the Khorat Group in the Phu
Phan Range area, Northeastern Thailand. In:
THANASUTIPITAK,T. (ed.) Proceedings International
Symposium on Biostratigraphy of Mainland
Southeast Asia: Facies and Palaeontology, Chiang
Mai, 23-49.
RACEY, A., GOODALL,J. G. S., LOVE, M. A., POLACHAY,S.
& JONES, P. D. 1994. New age data for the
Mesozoic Khorat Group of northeast Thailand.
In: Proceedings International Symposium on
Stratigraphic Correlation of SE Asia, Bangkok.
245-252.
SATTAVARAK, N. 1983. Review of the continental
Mesozoic stratigraphy of Thailand. In: NUTALYA,P.
(ed.) Proceedings of Workshop on Stratigraphic
Correlation of Thailand and Malaysia, Bangkok. 1,
127-140.
--, POLACHAN, S. t~ CHARUSIRISAWAD, R. 1991.
Cretaceous rocksalt in the northeastern part of
Thailand. In: GEOSEA VII Abstracts. Bangkok,
36.
STOKES, R. B., LOVATTSMITH,P. F. & SOUMPHONPHAKDY,
K. 1996. Timing of the Shan-Thai-Indochina
collision: new evidence from the Pak Lay Foldbelt
of the Lao PDR. This volume.
UTHA-AROON, C. 1993, Continental origin of the Maha
Sarakham evaporites, northeastern Thailand.
Journal of Southeast Asian Earth Sciences, 8,
193-207.
 The 'Rajang accretionary prism' and 'Lupar Line'
problem of Borneo
CHARLES S. H U T C H I S O N
c/o Department of Geology, University of Malaya, 59100 Kuala Lumpur, Malaysia

Abstract: The Rajang Group, composed of the Belaga and Lupar Formations of Sarawak and the
Embaluh Group and Selangkai Formation (in part) in Kalimantan, had a turbidite sedimentation
history from Early Cretaceous to Late Eocene. These rocks generally young northwards. Inliers
within the eastern Miri Zone have been mapped as the Kelalan and Mulu Formations. The Rajang
Group was compressed into a steeply dipping quartz-veined phyllite-quartzite complex by
the Sarawak orogeny and unconformably overlain locally in the north by the Upper Eocene
continental to neritic Tatau Formation, extensively in the south by Middle to Upper Eocene basal
continental sequences of the Ketungau, Mandai and Melawi basins, and widely in the north by
the Upper Oligocene coastal to marine Nyalau Formation. In Sabah and East Kalimantan, the
Upper Cretaceous to Upper Eocene Mentarang, Sapulut, Trusmadi, and possibly the East Crocker
Formations, also belong to the Rajang Group. The West Crocker Formation demonstrates rapid
facies changes into the more shaly Temburong Formation, and was deposited as sandy turbidites
throughout the Oligocene. To the south their equivalents are the nearshore Kelabit and Long
Bawang Formations. The West Crocker Formation was folded and uplifted in several Miocene
pulses, resulting in regional unconformities and igneous events at Mount Kinabalu. The West
Crocker Formation has not been metamorphosed, and dips are shallower than in the Belaga and
Trusmadi Formations. Its provenance probably was from the uplifted Upper Cretaceous to
Eocene Lurah and Kelalan formations of NE Kalimantan and East Sarawak. It is therefore
proposed that the West Crocker and Temburong formations be excluded from the Rajang Group.
Middle to Late Miocene Crocker Formation uplift and deformation, herein called the Sabah
orogeny, was synchronous with spectacular basin inversion throughout the South China Sea and
in the Meratus Mountains of Kalimantan. Uplift ceased in the Late Miocene and undeformed
post-inversion formations unconformably overlie inverted folded structures.
The Rajang Group flysch-belt may be interpreted as a north-facing accretionary prism. The
Schwaner Mountains represent a subducfion-related Lower to Upper Cretaceous volcano-
plutonic arc. Scattered Eocene volcanism, Miocene Sintang intrusives and Pliocene Metalung
volcanics in Central Kalimantan and Sarawak post-date subduction. The Rajang Basin (Danau
Sea) rapidly narrowed and by Eocene time subduction was transformed to collision as the Rajang
Group was compressed between the Schwaner Mountain Zone and the Luconia-Balingian-Miri
block. The Ketungau Basin is in sharp contact with the Rajang Group along the bounding Lupar
Fault, which can be traced northwards into the East Natuna region. Palaeocurrents show that the
Upper Eocene basal sandstones have a provenance in the metamorphosed Sibu Zone. The Melawi
and Mandai basins of Kalimantan also unconformably overlie the flysch-belt, The basins are not
forearc and were formed after transformation of the accretionary prism to a landmass formed of
a collisional orogenic complex.

The models of tectonic evolution of the South
China Sea region conflict with the d o c u m e n t e d
geology of Borneo and especially with the Rajang
Group. These models have then been applied to
analyze the tectonic evolution of Borneo. The
m o d e l of Taylor & Hayes (1980, 1983) wrongly
shows active subduction of Mesozoic oceanic
lithosphere continuously through the Palaeogene,
b e c o m i n g extinct by the Middle Miocene when the
Luconia Shoals were interpreted to have collided
with the Northwest Borneo Trough. The model o f
Holloway (1981) is not significantly different. The
mistake can be traced back to Hamilton (1979),
w h o described extensive tracts o f B o r n e o as
'm61ange', interpreted as s y n o n y m o u s with sub-
duction-related accretionary wedges.
The m o d e l of Ru & Pigott (1986) ignores
the West Baram Line and wrongly extends the
Northwest Borneo (Palawan) Trough through the
Luconia and Balingian provinces (their fig. 1),
and has active Late Eocene, southeastwards sub-
duction. The model o f James (1984) perpetuated
this mistake by not closing in the Late Eocene the
Rajang Basin, or Danau Sea of Haile (1994). These
models wrongly concur that subduction continued
until the Early Miocene. The mistake of continuing
the Northwest Borneo (Palawan) Trough along the
southern margin o f the Luconia Province, ignoring

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolutionof SoutheastAsia, 247

Geological Society Special Publication No. 106, pp. 247-261.
248 C.S. HUTCHISON
the West Baram Line, is also made by Tongkul
(1990).
All these models are flawed because they also
ignore the peripheral, yet equally important region.
Thus, Ru & Pigott (1986) and Williams et al.
(1988) accept the palaeomagnetic evidence that
West Borneo has rotated anti-clockwise (Schmidtke
et al. 1990), but they ignore the evidence for clock-
wise rotation of Indochina (Achache et al. 1983).
Williams et al. (1988) bring peninsular Malaysia
and Sumatra together only in the Early Tertiary; an
impossibility because they share the same
Palaeozoic and Mesozoic stratigraphy and structure
without displacement (Hutchison 1993).
The Mid Oligocene to Early Miocene N-S
extensional opening of the east sub-basin of the
South China Sea, first detailed by Taylor & Hayes
(1980), has been widely accepted and forms the
basis of most subsequent models of the region,
including contiguous Borneo. Using the data of
Taylor & Hayes (1980, 1983) and Chen (1987),
Briais et al. (1993) have produced a refined
scenario of South China Sea spreading history,
composed of three episodes: NW-SE extension
from 32-30 Ma, followed by N-S extension from
30-26 Ma, then by NW-SE extension from 23-
16 Ma. The structural trends for the latest, Early
Miocene, episode have been resolved in the
Scarborough Seamounts region by Pautot et al.
(1986) and Briais et al. (1989) to be 50" (magnetic
anomalies) and 140° (transform faults). The choice
of magnetic anomaly designations for the main
eastern basin was influenced by the only direct
geological constraint of volcanics dredged from
the Scarborough Seamounts, radiometrically dated
10-15 Ma (Hrkinian et al. 1989). The volcanics are
thought to have been extruded along the still weak
relict spreading axis. Briais et al. (1993) admit that
their interpretation is not unique. Some of their
illustrated anomaly designations do not inspire
confidence. Their scenario should prevail until the
ODP programme can include the South China Sea
Basin. The practice of designating anomalies with-
out direct geological constraint has led to serious
misinterpretations in the nearby Sulu Sea (Lee &
McCabe 1986; Hutchison 1992a).
More carefully constructed evolutionary
scenarios of the Borneo margin of the South China
Sea were painted by Tan & Lamy (1990) and
Hazebroek & Tan (1993), which benefited from
comprehensive petroleum company data. However
their inferred subduction history from Late
Cretaceous through Late Eocene, using the Lupar
and Mersing lines, and from Late Eocene through
Middle Miocene, using the Northwest Borneo
(NW Sabah) trough, is not matched by identified
subduction-related post Palaeocene volcanic arcs
in Borneo.
The Rajang Group of Sarawak
The Rajang Group consists of the Belaga and Lupar
formations, which are isoclinally folded, with dips
usually between 80 and 90° (Fig. 1). It extends
across the border into Kalimantan as the Embaluh
Group (Tare 1991). This monotonous flysch group
is of interbedded sandstone and mudstone, is
locally metamorphosed to lower greenschist facies
phyllite and slate, and commonly veined by quartz
(Kirk 1957; Liechti et al. 1960; Wolfenden 1960).
The outcrop width of the group perpendicular to the
E-W strike from the Lupar Line to the Tatau Horst
is 200 km. Thickness cannot be determined and has
been guessed to be in the neighbourhood of 15 km
(Haile 1974), but the basis for his estimate is
unknown. The Rajang Group constitutes the Sibu
Zone of Sarawak (Fig. 1), outlined along the south
by the Lupar Line ophiolite and along the north by
the Bukit Mersing Line ophiolite (Hutchison 1989),
but it also crops out north of this in the Tatau Horst
and as various inliers within the Miri Zone, as the
Kelalan and Mulu Formations. The Lupar Line
ophiolite occurs as blocks in a 20-25 km wide
mrlange zone, known as the Lubok Antu Mrlange.
Along its SW side, the mrlange is in sharp contact
with the basal Silantek Formation beds of the
Ketungau Basin along the steep linear Lupar Fault.
The Rajang Group has been interpreted as a deep
marine fan deposit. Most outcrops have thin fine
to medium grained sand layers which may be
described as distal turbidite, but some outcrops
show thicker sandstones. Most are characterized
by prominent sole marks.
Adjacent to the Lubok Antu Mrlange, the 5 km
wide outcrop of the Lupar Formation is predomi-
nantly a turbidite sequence similar to the con-
tiguous Layar Member of the Belaga Formation
(Liechti et al. 1960; Tan 1979). It was mapped
separately because it incorporates large bodies of
ophiolitic rocks. Microfossils indicate an Upper
Cretaceous age of Santonian to Maastrichtian. The
Lupar Formation thus represents the oldest part of
the Rajang Group of Sarawak (Table 1). From the
south, adjacent to the Lupar Formation, to the Tatau
Horst in the north, the Belaga Formation has been
subdivided into the following members: Layar,
Kapit, Pelagus, Metah and Bawang. The fossils
indicate a progressive younging northwards from
Upper Cretaceous Layar to Upper Eocene Bawang.
The steep dips and the systematic northwards
younging and ophiolite and mrlange association
have led to the plausible interpretation of a north-
wards facing accretionary prism, related to Late
Cretaceous (85 Ma) to Late Eocene (45 Ma) south-
wards subduction beneath the Kuching Zone of
Sarawak of the Rajang Basin, named the Danau
Sea by Haile (1994).
 I I I !

Iio° 112" 114" 116"

Tertiary, predominantly BF = BELAGA Fm

continental basins TRS : TRUSMADI Fm
KOTA
BM : BOYAN MI~LANGE KINABALU
Oligocene to Lower Miocene
Crocker- Temburong flysch K = KAYAN SANDSTONE
M = MULU Fm
Lower Cretaceous to Upper Eocene LABUANCC3
KE : KELALAN Fm
orogenic flysch complex
(RAJANG 8 EMBALUH GROUPS -5 °.
SELANGKAI Fm)
SABAH
Jurassic- Cretaceous +BRUNEI%
PEDAWAN Fm ( in Sorowok) %+ ~,,,,3M
NIAH
FM

©
BINTULU // ~
-30 0 I00 200 PENIAN >
I I I
km
t"
(KUCHING ZONE) MERIT'-~ ~ --~-÷' '
_~
~----___-------- -PILA~_.J ~!%-!
ROU. --------
E
'---'----..~___~--~(SIBU ZONE) ~'+~--~ +
+ "- -------- ~_ + ~--~--+-. +++ ~0
f "L/~ I ~)PEDAWAN~ ~\,. ©
~. ~ E~MBALUH~ -
/f~l (~'" ~ O'~Z'- 5 Eocene P o s t - -
L-f KAPUAE~_.GROUP - IVy'4
Subduction Volcanism ©
• .. . . . . . . . . . . : M
~b i SE I = SERANTAK VOLCS.
• WEST KUTEI 2 = PI.YABUNG TUFF
3 = MULLER VOLCS.
- " • e ,.BM • •
. 4 = NYAAN VOLCS.
,~ ~ • PONTIANAK ".
f.... MELAWl SEEANGKAI --'"-' • • -- 5 = ARIP VOLCS.
.~__FM_~-") . . "UPI~ER . . . .
International
+ + Border
SCHWANER MOUNTAINS

Fig. 1. Sketch geological map of Sarawak, Sabah and contiguous Ka]imantan (Borneo) to emphasize the Rajang Group. The eastern Mid Zone contains several inliers of Kelalan t,O
and Mulu Formations, interpreted to be equivalent to the Rajang Group. The Sibu Zone contains several large Pliocene volcanic massifs. Within the Embaluh Group
outcrop they are known as the Metalung Volcanics. Based on Tan (1982), Lim (1985), Pieters et al. (1987) and Pieters & Supriatna (1989).
250 c . s . HUTCHISON

Table 1. Chronology of events discussed in the text

Schwaner Post-subduction Rajang Group Rajang Group N. W. Sabah South China Sea
Ma NERIE,~ Mountains Volcanism deposition deposition Crocker Ranges
volcano- & plutonism Sarawak& Sabah &
plutonic arc W. Kalimantan E. Kalimantan
5 ..Plioeene ....................... Metalung Voles.................................................................................................... Post-inversion strata .................
10 upper Mr. Kinabalu Shallow Regional
15 middle Deep Regional ~ SABAH OROGENYu"
20 Miocene Sintang
25 .... lower ............................. intrusives .... Deposition of the Melawi ........................................................ Deposition of the ....................
30 Oligocene Ketungau and Mandai Kelabit Fin. [W. Crocker Fm ] South China Sea
35 ..................................................................... basins ..................................... [ including the ] .............. basins ..........................
40 MiHler Tatau Fnt Batu Gadlng [ Temburong Fm l
45 Eocene ~ S A R A W A K OROGENY ~ ~ y ~ .
50 Nyaan, Piyubung /lk /l~ [ E. Crocker Fro. ]
55 ...................................................................... :~:. .............................................. "t'" .............. ~ ........ ~ ....... ~ .......................................................
60 Palaeocene (60 Ma) ? Kayan Sstn ? 0llelaga Fro] 1 Trusmadi Fm
........................... V ...................................... [Lupar F m ] / ~ ........... [ Sapulut Fm l ............... I"...................................................................
70 [Embaluh Group] [Mentarang F m ] |
75 Upper
80 Cretaceous
85 G [potassic granite] ]
90 P I A~
95 ........................... P ..................................... . ~ ........................... :4:" ..................................................................................................................
loo r / /
105 Lower P | [ . ,
110 Cretaceous P |- - /
~ 1/~ end ~ u n c o n f o r m i t y [
115 P /[Selaaghal Fm]""l I l time V = volcanism I
120 ........................... P ...................................... . ]-{Pedawan Fm] ...... I............................................. [ [duration P=plutonism [
125.. Valanginian ................................................. .[... .......................................................................... [ l o f turbidite G = granite ]
130 i I | sedimentati°n I
135 ..........................................................................................................................................................................................................................

Palaeocurrents in the Lupar Formation turbiditic

sandstones indicate a major current direction Relationship to younger formations
towards 50-60 °, with a subordinate maximum One of the most spectacular Malaysian angular
towards 80-90 ° (Tan 1979). The provenance of the unconformities is exposed on the east end of the
earliest part of the Rajang Group turbiditic sand- Tatau Horst (Wolfenden 1960; Halle & Ho 1991).
stones was from the Kuching Zone, lying to the SW The gently dipping basal conglomerate of the Tatau
and W. Most of this zone formed an integral part Formation rests on steeply dipping and complexly
of the extensive Late Cretaceous landmass known folded Upper Eocene Bawang Member turbidites
as Sundaland, upon which continental Eocene of the Belaga Formation. The unconformity is
formations were widely deposited (Hutchison palaeontologically dated as Upper Eocene. The
1992b). The fluvial system, which brought sedi- Tatau Formation is an Upper Eocene to Oligocene
ments from the interior of Sundaland, must have sequence of conglomerate, sandstone and shale. It
reached the Danau Sea in the neighbourhood of is continental to fluvio-marine, becoming neritic
what is now the Lupar Line. The geological record upwards, and includes the sub-aerial Arip
favours a narrow continental shelf as shown by Volcanics. From this informative outcrop, it may be
outcrops of the Jurassic-Cretaceous Bau Limestone inferred that the Danau Sea was eliminated and
rapidly giving way to turbiditic Pedawan Formation deformation, metamorphism,-uplift and peneplana-
of the Kuching Zone (Tang 1992), so it may be tion of the Belaga Formation to form a landmass
inferred that there was a rapid descent from land in was completed before the end of the Eocene. The
the Kuching Zone to deep marine conditions even author proposes to name this the Sarawak orogeny
before the Lupar Line was reached. The lower, (Table 1). This information invalidates the tectonic
Upper Cretaceous, part of the continental Kayan model of James (1984), reproduced by Hutchison
Sandstone (Tan 1986) may be the on-land (1989).
equivalent of the deep marine Lupar Formation An equally spectacular unconformity occurs at
and Layar Member of the Belaga Formation the Bukit Besungai area of Batu Gading (Fig. 1) in
(Fig. 1). However, Tate (1991) and R. J. Morley the Middle Baram region (Adams & Haak 1962).
(pers. comm. 1993) have expressed doubt about Highly fossiliferous Upper Eocene limestone over-
the older age palynological determinations by lies with spectacular angular unconformity an
Mtiller (in Tan 1986) of the Kayan Sandstone and Upper Cretaceous steeply dipping sequence of
regard it as Upper Eocene. turbiditic Kelalan Formation, the eastern equivalent
 RAJANG GROUP AND LUPAR LINE, BORNEO
of the Belaga Formation (Haile 1962). This
quarried location along the upper reaches of the
Baram River (Fig. 1) is of identical significance
to the Tatau Horst locality.
The Tatau Formation is overlain by a thin
(550 m) shale-sandstone sequence known as the
Buan Formation (Wolfenden 1960). However, the
major transgression over the landmass was accom-
plished by the Upper Oligocene to Lower Miocene
Nyalau Formation, estimated to be about 5 km
thick. It rests conformably on the Buan Formation
and with marked unconformity on the Belaga
Formation. The Nyalau Formation forms many
outliers over the main Belaga Formation outcrops
of the Sibu Zone (the largest at Merit Pila is shown
on Fig. 1) and the unconformity is documented by
Kirk (1957) and Wolfenden (1960). This is a very
sandy formation in the west around Bintulu,
becoming more muddy eastwards, where it is
known as the Setap Shale. Detailed analysis by
Shell indicates that a western landmass, known as
the Penian High (Fig. 1), progressively gave way
eastwards to a coastal plain rich in coal beds, then
an inner neritic continental shelf. The outer shelf
was characterized by a narrow NNW-SSE trending
carbonate platform (Subis Limestone), seen at
Niah. The edge of the shelf was marked by a
continental slope, known as the West Baram Line,
east of which bathyal conditions prevailed in a
region which became the Baram Delta in the Late
Miocene.
Southwards extension into Kalimantan
An understanding of the Rajang Group is
impossible without knowledge of its southwards
extension, perpendicular to strike, into Kalimantan.
The Lupar Formation and the Layar and Kapit
Members of the Belaga Formation are known in
northern Kalimantan as the Upper Cretaceous to
Eocene Embaluh Group (Fig. 1; Table 1), which is
not sub-divided (Tate 1991). This flysch sequence
is strongly deformed. Bedding is in places over-
turned, and two fold phases have been identified.
East of the Kapuas lakes, the main fold axis orien-
tation is horizontal, trending 280 ° . The folding is
asymmetrical, with southerly dips predominant
(Williams et al. 1988). However, overall strati-
graphic younging is towards the north, as in
Sarawak. South-dipping thrust faults have been
identified within the Embaluh Group. Southern out-
crops along the Upper Mahakam River are reported
to be unmetamorphosed (S. J. Moss pers. comm.
1995).
The Embaluh Group is succeeded southwards
in the Sintang District (Williams & Heryanto 1986)
by the Lower Cretaceous (Valanginian) to Upper
251
Cretaceous (Turonian) Selangkai Formation, which
youngs northwards and was first described by
Zeijlmans van Emmichoven (1939). It has been
restudied by Williams & Heryanto (1986), who
described it as predominantly of calcareous mud-
stone with intercalations of pebbly and bouldery
mudstone, graded sandstone and rare limestone and
conglomerate. However, the formation is litho-
logically variable and some stratigraphic sections
are dominated by turbiditic sandstone or other mass
flow deposits. The structural and stratigraphic com-
plexity has not been resolved, but Zeijlmans van
Emmichoven (1939) suggested a formation thick-
ness in excess of 3 km. The Selangkai Formation
occupies a 45 km wide belt in which the main fold
axes and outcrop trend WNW, with some cross
folding and some strongly cleaved and thrust
faulted zones. The lower part of the formation is
in fault contact with the Boyan Mrlange, where
it is tightly folded and cleaved. The dominant
calcareous mudstone is a dark grey strongly
sheared rock. It commonly contains, large blocks
of fossiliferous limestone. Williams & Heryanto
(1986) interpreted the sandstone beds, conglomer-
ate and calcareous mudstone containing dispersed
boulders as mass flow deposits. The sandstone beds
have prominent sole marks and are turbiditic. The
.-general appearance of the Selangkai Formation is
that of a deep basin submarine fan into which
blocks of limestone have been transported from a
more stable shelf. Abundant volcanic detritus
indicates a nearby contemporaneous volcanic arc,
probably in the Schwaner Mountains (Williams &
Heryanto 1986). The Upper Cretaceous (Turonian)
Ingar Formation is in fault contact with the
Selangkai Formation and both are folded and
unconformably overlain by the Dangkan
Sandstone, which forms the lower (Upper Eocene?)
sequence of the Melawi Basin (Williams &
Heryanto 1986). Unlike the Selangkai Formation,
the Ingar Formation is predominantly muddy with
an absence of turbiditic sandstone.
It seems reasonable to conclude that the Danau
Sea deep water fill is of Valanginian age in the
south (Selangkai Formation), younging northwards
through the Embaluh Group (= Lupar Formation
and lower Belaga Formation), and eventually
reaching its youngest (Late Eocene) age near the
Mersing Line at the Tatau Horst. Nevertheless, the
picture of a general northwards younging may be
a gross oversimplification of this poorly known
terrain. The Embaluh Group of the Upper
Mahakam and Boh rivers of Kalimantan has
yielded Middle Eocene planktonic foraminifera
(Carter & Yeats 1973). The Selangkai Formation
is similar both in its partial turbiditic character and
age to the Pedawan Formation, which lies west
across the Lupar Fault in the Kuching Zone.
252 C.S. HUTCHISON
The Lupar Fault appears from the Sarawak view-
point to represent the most southerly termination
of turbiditic rocks. However, the flysch belt extends
southwards into the Sintang region of Kalimantan.
The Lupar Line has therefore been given undue
importance. It is the prominent fault margin
between the Upper Eocene Silantek Formation of
the Ketungan Basin and the Sibu Zone flysch belt.
Indeed, within the Kuching Zone of Sarawak, there
are abundant Lower to Upper Cretaceous turbidites
known as the Pedawan Formation lying W and SW
of the Lupar Line (Ting 1992).
Ophiolitic m61ange zones are not confined to
the Lupar Line. They also occur along the Bukit
Mersing Line, and along the Semitau Ridge, known
as the Boyan M61ange, although clasts in the latter
are predominantly of quartz-rich sedimentary and
metamorphic rocks (Williams et al. 1988). All of
these m61ange belts lie close to and parallel to
major faults which bound superimposed Tertiary
basins. The Semitau Ridge margins are charac-
terized by north dipping thrusts post-dated by
Upper Oligocene sedimentary rocks.
Search for the paired volcano-plutonic arc
Most authors (e.g. Hamilton 1979; James 1984)
presume that the Upper Cretaceous through Upper
Eocene Rajang Group of Sarawak (now modified
to begin in the Lower Cretaceous by including
Kalimantan) represents a northwards-facing
accretionary prism, but there has been no clear
identification of an appropriately positioned con-
temporaneous subduction-related volcano-plutonic
arc. Gower (1990) concluded that the data from
Borneo 'support the postulate of southeasterly-
directed subduction of the South China Sea beneath
Borneo during the early Tertiary.' At the same time
he admitted that 'there are no convincing reports
of early Tertiary magmatism, coeval with the
postulated subduction.' The following discussion
may clarify the situation of the appropriately
distanced terrain lying 200-250 km to the south.
Convexity o f the arcs
The island of Borneo is dominated by a framework
of arcuate structures, all convex towards the south
and southwest (Fig. 1). This feature was empha-
sized by the earlier Dutch workers (Van Bemmelen
1949). Hamilton (1979) noted the importance of the
Schwaner Mountains volcano-plutonic arc thus
'Presumably these igneous rocks are paired to
the Cretaceous m61ange complexes that occur in
both northwestern and southeastern Borneo'. He
favoured subduction from the north, even though
the convexity would favour subduction from the
south.
The ophiolitic rocks of the Meratus Mountains
have been dated as Lower Aptian (116 Ma) by
Sikumbang (1986) and are associated with
calcareous mudstone, sandstone and limestone.
These ophiolites would seem to be ideal candidates
to be paired to the Early Cretaceous volcano-
plutonic arc of the Schwaner Mountains (Fig. 1).
However, Sikumbang (1986) considered that there
was no subduction relationship between them, and
he concluded that the Meratus Mountains contain
the record of a compressed marginal basin and
paired volcanic arc.
The Meratus ophiolite formed part of the Late
Cretaceous landmass upon which diamond-bearing
serpentinite conglomerate (the Pamali Breccia)
was fluvially deposited (Bergman et al. 1987).
However, the Meratus did not uplift to form the
present day mountains until the Late Miocene
(Hutchison 1992b). Before the uplift, the now
separate Barito, Asem-Asem and Pasir basins
were once one and the same.
The Schwaner Mountains
The rigid 'core' of Borneo is formed by the
Schwaner Mountains. Their northern boundary
forms the southern margin of the Tertiary Melawi
Basin (Fig. 1), and extends inland along an E-ESE
trend, which cuts the equator east of Pontianak. The
Schwaner Mountains extend eastwards more than
4 ° of longitude to disappear beneath the Cenozoic
NW Kutei Basin, and they extend over 2 ° of
latitude southwards to disappear beneath the
Quaternary sediments of the coastal plains, which
are underlain by the Billiton and Barito basins
of SE Kalimantan. The region was divided by
Van Bemmelen (1949) by an approximately E-W
line along latitude l°S into the 'Schwaner Zone' and
a 'southern region'. However, subsequent mapping
(Amiruddin & Trail 1989; Pieters & Supriatna
1989) has shown that both zones form part of
the Schwaner batholith, of which a major part is
the Sepauk tonalite, with the Sukadana granite of
the Ketapang batholith occurring in the SW corner.
The Schwaner Zone, east of Pontianak, is mostly
of 1-type (subduction-related) calc-alkaline
plutonic rocks ranging from granodiorite to gabbro,
dominated by the Sepauk tonalite, associated with
a variety of metamorphic rocks mapped by Pieters
& Supriatna (1989) as Palaeozoic, probably
Carboniferous to Permian, though their age is
indefinite. The 'southern region', lying east of
Ketapang, is composed of Jurassic-Cretaceous
flysch and contact metamorphosed limestone (the
Ketapang Complex of Van Bemmelen 1949), and
volcanic rocks (the Matan Complex of Van
 RAJANG GROUP AND LUPAR LINE, BORNEO
Bemmelen), intruded by granitic batholiths (Pieters
& Supriatna 1989). The Bunga basalt and Kerabai
volcanics, formerly included in the 'Matan
Complex', range from basalt, through andesite, to
dacite and rhyolite, and represent a mature volcanic
arc, which Tate (1991) paired with the Upper
Cretaceous part of the Rajang Group, representing
southwards subduction.
The plutonic complex of the Schwaner Zone
(Sepauk tonalite) has been dated by the whole-rock
K-Ar method by Williams et al. (1988) and Bladon
et al. (1989) at 104 (Albian) to 123 Ma (late
Valanginian). Earlier, McDougall (in Haile et al.
1977), using biotite and hornblende mineral
separates, determined K-At ages ranging from 75
(Campanian) to l l 5 M a (Barremian). The two
datasets overlap and extend over the Lower and
Upper Cretaceous.
Williams et aI. (1988) report only 4 K-Ar dates
from the potassic Sukadana granite of the 'southern
region', lying within the range 86-91 Ma. However
McDougall (in Haile et al. 1977) reported 10
mineral separate K-Ar dates from this province,
eight of which lie within the range 78-86 Ma,
but two granites gave considerably older ages of
127 and 153 Ma. Therefore this province contains
important Late Cretaceous plutonic activity, domi-
nated by alkaline granite, but there also occur some
Lower Cretaceous to Upper Jurassic granitoids.
The volcanic rocks of the 'southern region' were
considered too weathered by Haile et al. (1977)
for accurate dating. However, Pieters & Supriatna
(1989) reported Upper Cretaceous K-Ar ages of
65-75 Ma (Maastrichtian-Campanian).
In summary, most of the plutonic rocks of the
Schwaner Mountains may be paired to the
Selangkai Formation, which contains abundant
volcanic detritus. Volcanism and high-level
plutonism continued at least into the Maastrichtian,
when the Embaluh Group, Lupar Formation and
Layar Member of the Belaga Formation were being
deposited. The Cretaceous part of the Rajang
Group may therefore be reasonably interpreted as a
northwards-facing accretionary prism, paired to the
Cretaceous Schwaner Mountains volcano-plutonic
arc (Table 1 and Fig. 2).
N o r t h o f the S c h w a n e r M o u n t a i n s
A narrow E - W trending zone through the Upper
Kapuas region of central Kalimantan, with sporadic
occurrences of Early Eocene acid volcanic and
pyroclastic rocks (Fig. 1), of restricted 48-50 Ma
age, has been interpreted by several authors to
post-date subduction and be related to the early
rift stages of formation of the Ketungau, Mandai,
Melawi and West Kutei basins (Tate 1991; Doutch
253
1992; Van de Weerd & Armin 1992). They can be
demonstrated to post-date subduction, because in
several localities the Eocene volcanics uncon-
formably overlie either the Boyan M61ange or the
Embaluh Group.
The Piyabung Volcanics, of gently dipping sub-
aerial lithic and vitric tuff and agglomerate up
to 200 m thick, extend 40 km E-W from 112 to
112 ° 30' within the Semitau Ridge at 0 ° 18'N
(Pieters et al. 1987). A K-Ar whole rock age of
49.9 was obtained (Bladon et al. 1989).
The Nyaan Volcanics, of 200-900 m of acid
to intermediate tuff, agglomerate, ignimbrite and
dacite, occur much farther east in the Mahakam
area of the West Kutei Basin, at 114 ° 50% and
0 ° 50'N. They gave a K-Ar date of 48.6 Ma (Tate
1991) and 50 ___2.5 Ma (Pieters et al. 1987). In the
Kelian district of the Upper Kutei, Van Leeuwen
et al. (1990) describe Upper Eocene rhyolitic
volcanic and pyroclastic rocks in a region domi-
nated by Miocene volcanism. The Eocene igneous
rocks are exclusively silicic, and have not been
demonstrated to be subduction-related. In the same
general region, Wain & Berod (1989) show Upper
Cretaceous to Eocene volcanic rocks on their maps,
but do not describe them.
Near Singkawang, far to the west near the west
coast at l°N, the Bawang Dacite gave a K-At date
of 51.3 Ma. It is part of the Serantak Volcanics,
which are less than 300 m thick.
A much more extensive body of intermediate
to basic volcanics lies entirely within the Mandai
Basin in the Mtiller Mountains at 113 ° 20'E,
0 ° 30'N, K-Ar dated 40.9 Ma (Pieters et al. 1987).
These intermediate to basic tufts and flows, many
hundreds of metres thick, are interbedded with the
basal sandstones along the Mandai River. The
volcanic rocks were first documented by
Molengraaff (1902). Zeijlmans van Emmichoven
(1939) confirmed the existence of volcanic rocks,
predominantly of basic-intermediate to basic com-
position in the region of the Mandai Palaeogene
basin east of the Lake District (Kapuas), equivalent
to the Mtiller Mountains of Molengraaff (1902).
On the north side of the Rajang Group, the acidic
Arip Volcanics and associated Piring Granodiorite
of the Tatau Formation have not been dated but
post-date the metamorphism of the Rajang Group.
They are comparable to volcanics described above
from the southern side of the orogenic flysch belt.
It may therefore be concluded that there is no
Eocene subduction-related arc to the north of the
Schwaner Mountains, and that the predominantly
silicic and pyroclastic eruptions are an integral part
of the basin extensional process. It is therefore
doubtful if the term accretionary prism should be
applied to the whole of the Rajang Group, for only
its Cretaceous part can be paired to a volcanic arc,
254 c . s . HUTCHISON
within the Schwaner Mountains. The Palaeocene-
Eocene part should not be classified as an accre-
tionary prism, for subduction-related volcano-
plutonic activity had ceased, and the situation was
one of collision (Fig. 2).
A province of abundant small epizonal stocks,
sills and dykes extends from the Kuching Zone
throughout the Ketungau Basin, with maximum
development within the eastern Ketungau Basin
(Williams & Heryanto 1986). They have been
named the Sintang Intrusives (Williams & Harahap
1987). They range from pyroxene andesite to
microgranite and are I-type. K-Ar dates from
Sintang range from 30--23 Ma in the south and
18-16 Ma in the north. The stocks preferentially
intrude the basinal strata, and have metamorphosed
the Silantek Formation coals to anthracite (Tan
1979).
Volcanic rocks overlie the Belaga Formation in
the Usun Apau, Hose and other massifs (Kirk
1968). They are equated with the Metalung
Volcanics, which overlie the Embaluh Group of
Kalimantan (Doutch 1992). Although poorly
studied, they appear to be Pliocene for they overlie
outliers of the Nyalau Formation. In Kalimantan,
they have been dated by the K-Ar method within
the range 1.6-8.2 Ma (Bladon et al. 1989). They
have been extrapolated much farther east, where
they intrude the Mentarang Formation (BRGM
1982), and have yielded similar dates. However
K-Ar ages as old as 25 Ma may not be from
equivalents of the Metalung volcanics.
Ketungau Basin
In Sarawak, the basal formation of the Ketungau
Basin is known as the Silantek Formation. Only the
lower part, which outcrops adjacent to the Lupar
Line, can be dated by fossil content, but the age
range is generally considered to be Upper Eocene
to Miocene. The Upper Eocene Basal Sandstone
Member outcrops on the Marup Ridge and is
steeply dipping (Tan 1979). The strata flatten out
southwards away from the Lupar Fault.
The average down-dip palaeocurrent direction of
the lower member is 228 ° from ripple marks and
194 ° from cross bedding (Tan 1979) indicating that
the provenance was from the northeast. This is con-
sistent with the petrology of the conglomerate and
sandstone, which contain clasts of quartzite, schist,
meta-greywacke and mafic volcanics, indicating
provenance from uplifted and eroding Lupar
Formation and Layar Member of the Belaga
Formation. The base of the lower member is
nearshore to shallow marine, but the Silantek
Formation rapidly becomes continental upwards
and has red-beds in the upper part, followed by the
thick flat lying continental cross bedded Plateau
Sandstone. The Sibu Zone was therefore in the
process of inverting to form a landmass during the
early stages of Silantek Formation deposition. The
steep dips of the lower Silantek Formation adjacent
to the Lupar Line imply that the Rajang Group
orogenic belt continued to uplift after deposition
of the basal Silantek Formation, resulting in the
boundary becoming the Lupar Fault. The rapid
elimination of marine conditions from the Silantek
Formation indicates continuing inversion of the
Rajang Group flysch-belt.
Accordingly it is wrong to refer to the Silantek
Formation (Ketungau Basin) as forearc (Williams
et al. 1988), since the Rajang basin had been
eliminated and the Sibu Zone transformed from
an accretionary prism into an orogenic belt welding
together the Schwaner Mountains Zone of
Sundaland onto the Balingian-Luconia Province
(Miri Zone) (Fig. 2). The Rajang basin, or Danau
Sea, was not related to the younger South China
Sea and should not be referred to as the Proto South
China Sea (Haile 1994).
The Lupar Fault is the bounding margin of the
Ketungau Basin, which developed by extensional
tectonics affecting the Rajang Group orogenic belt.
Studies to the north in the East Natuna area have
shown that a similar situation persists, where
Tertiary basins are unconformably superimposed
on phyllitic basement (White & Wing 1978), which
can be equated to the Rajang Group. The basin
margins likewise commonly form prominent faults,
similar to the Lupar Fault, but the trend has been
bent into a N-S alignment northwards from the
coast of Sarawak.
During their early development, the Ketungau
and Melawi basins were one and the same (Doutch
1992). They have subsequently become separated
by uplift of the Semitau Ridge.
Rajang Group of Sabah and
East Kalimantan
The Rajang Group of Sabah, as presently defined,
is of much longer depositional duration (Table 1)
and its stratigraphy and structure have not yet been
satisfactorily resolved. However, it can be sub-
divided into a 'Lower Part', whose sedimentation
history extends from Late Cretaceous to Late
Eocene, and an 'Upper Part', extending from
Oligocene to Early Miocene (Lim 1985). It is the
Lower Part only which extends into Kalimantan
and Sarawak.
The Embaluh Group appears to continue east of
115° as the Upper Cretaceous to Eocene Mentarang
Formation, extensively exposed and of distal
turbidite character, strongly folded and displaying
slaty cleavage (BRGM 1982). Northwards along
 RAJANG GROUP AND LUPAR LINE, BORNEO 255

Schwaner Fore-arcbasin Accretionary Trench Proto

Barito Mountains Selangkai prism / south turbidites+
[marginal [active volcano- Formation (SelangkaiFm in part, / China Barremian
sea ?) plutonic arc) (in part) + Embaluh G r o u p ~ / Sea / chert
0 ~ ~/'~ sea level
,0

KM 30 -4 t ~ r .4~." Ophl'olitenow seen 30 KM

~ J[ ~ ~ ' ~ " ~ ~ inBoyan, Lupar and A,,
ou 11
i i , ons¢ or,.,.,'," ,,,,o,',o ,4°
50 "I parIiy @~ ascending , ~ " I 0
{uplifted--A magma + .@~. k,.... ~ 60
60 ]in Meratus ~volatiles , i f ~ ~r~" Mantle
70 ]...Mountains I ~'~'"Y I I 70

Present S 0 1O0 KM 200 300 PresentN

(a)

r- final position of 'trench'

volcanic arc EmbaluhGroup RajangGroup at subduction-collision
<~==extinct [Indonesia] (Sarawak) ¢ transition
0 . . . . .
---sea level
10 - ~ - - ~
- - - - - BALINGIAN _-- .... 10 ~
-"---------- : - : - - I - LUCONIA - -
2 0 - ~ ~ ~ _ - - _ - _ _ - - / C O N T I N E N T I - - .... 20
-- 30 KM

40
50 ~ "
.... 50
60 ~ i I I I 60
200 300 400 500 600
LEGEND: (b) KM ~ KM from volcano-
plutonic arc
~ ] deformed
turbidites Schwaner Early Early Lupar Sibu Zone Tatau Miri
Mountains Melawi Ketungau Fault Uplift Horst Zone
basal strata (buttress) Basin Basin ~ (Raiang Group)
overlying
unconformity
, .~.,,°n-.?~ ~..J,~ /,:...~/IA,'.4.~ °°,..°
I ~ volcanics

plutons
Piyabung ,,~,,-~ isostatic
..o.cc, r.,
oceanic volcanism
crust
,~'~" sinkingslab
continental I I I I I
lithosphere 0 1O0 200 300 400 500
KM
Present South Present North
(c)
Fig. 2. Schematic N-S tectonic cross-sections through the Schwaner Mountains and Rajang Group. Vertical and
horizontal scales are very approximate. (a) Late Cretaceous (c. 80 Ma) north-facing arc-trench system. (b) Palaeocene
(c. 60 Ma) cessation of subduction, forwards migration of the 'trench' and incipient collision. (e) Late Eocene
collision, compression and uplift of the Rajang Group between the Schwaner Mountains and the Balingian-Luconia
Continent, followed by normal faulting and deposition of Late Eocene strata unconformably upon peneplained
Rajang Group. The scenario of this figure is in strong contrast to that of James (1984), reproduced by Hutchison
(1989, p. 85).
256 C.S. HUTCHISON
strike in SW Sabah, it has been mapped as the
Sapulut Formation (Collenette 1965). No slaty
cleavage was mentioned by Collenette, who
described the mudstones as 'fissile and indurated',
although previous workers (see Reinhard & Wenk
1951) referred to the Sapulut, as well as the
Trusmadi Formation, as a "slate formation'. The
farthest extension into Sabah is known as the
Trusmadi Formation, well exposed in accessible
outcrops from Ranau (near Mount Kinabalu) along
the main road towards Telupid. The predominantly
Palaeocene-Eocene Trusmadi Formation is a
tightly folded, generally steeply-dipping, phyllitic
distal turbidite, characterized by quartz veining,
and identical in appearance to large parts of the
Lupar and Belaga Formations of Sarawak, but in
general more shaly.
The upper parts of the Mentarang and Sapulut
Formations show indications of shallowing of the
sea, and limestones and jasper conglomerates are
recorded in both (Hutchison 1989). In Kalimantan,
the Mentarang Formation is unconformably over-
lain by the Upper Eocene Sebuku Formation basal
conglomerate, followed by reefal limestone
(BRGM 1982). Turbiditic sedimentation was there-
fore eliminated by tectonic uplift by Late Eocene,
a situation identical to that in Sarawak (Table 1).
A folding and metamorphic event clearly must
separate the Trusmadi from the unmetamorphosed
and less deformed Crocker Formation. This event is
named here the Sarawak orogeny. However, only
fault contacts have been observed between the two
formations (Collenette 1965; Jacobson 1970). The
very sandy Crocker Formation turbidite, containing
several extremely thick sandstones, is usually con-
sidered as Palaeocene to Lower Miocene (Lim
1985). The upper western part has been mapped as
the turbiditic West Crocker and Temburong
Formations by Wilson (1964). The latter is very
fossiliferous and has an age of Oligocene to Lower
Miocene. The 'East Crocker Formation' is less well
known and its reported Palaeocene and Eocene age
is poorly constrained palaeontologically. Structural
relations between the "West' and 'East' Crocker
have not been resolved.
The Sarawak equivalent of the Oligocene to
Lower Miocene West Crocker and Temburong
Formations is the Setap Shale. The unconformity at
the top of the Trusmadi and Mentarang Formations
indicates that the West Crocker and Temburong
Formations should not be included as part of the
Rajang Group of Sarawak.
Palaeocurrent analysis on sole marks of the
West Crocker and Temburong formations (Stauffer
1968) indicate northwards directed currents. The
provenance is most likely to be the Lurah and Long
Bawang formations, which were uplifted before the
end of the Eocene; an interesting heavy mineral
provenance study remains to be carried out. These
shallow marine formations are considered the
time equivalent of the deep marine Mentarang
Formation (BRGM 1982).
If the Mentarang, Sapulut and Trusmadi
Formation turbidites represent an accretionary
prism, then the Late Cretaceous to Eocene paired
volcanic arc, which cannot be seen in Sabah, needs
to be sought in the northern arm of Sulawesi, which
was contiguous with eastern Sabah before the Late
Eocene opening of the Celebes Sea (Hutchison
1989). However, the volcanic arc activity of north
Sulawesi began in the Early Miocene. Older rocks
have been interpreted as ophiolite, with pre-
Tertiary metamorphic and intrusive rocks (Carlile
et al. 1990). The Upper Cretaceous to Eocene
Tinombo Formation in the north arm of Sulawesi
contains only minor calc-alkaline and tholeiitic
suites (Kavalieris et al. 1992).
An outstanding problem remains, and that is the
relationship between the shallow marine Upper
Cretaceous to Eocene Lurah Formation (BRGM
1982) and the West Crocker-Temburong
Formation. The areal distribution of formations
indicates that while sediments of the West Crocker
Formation were being transported from the south,
there lay a landmass predominantly made of the
Lurah Formation in that direction. Just across the
border in Sarawak, the Upper Cretaceous to Eocene
Kelalan Formation (Fig. 1) has been mapped by
Haile (1962). He correlated it with the Belaga
Formation, and outcrops at Batu Gading are of
turbidite identical to the Belaga Formation. In the
region south along strike from the West Crocker
and Temburong Formations, Haile (1962) mapped
the shallow marine coal-bearing Kelabit Formation,
which has yielded good Oligocene to Lower
Miocene fossils, and it is the proximal along-strike
equivalent of the deeper water West Crocker and
Temburong Formation turbidites. The Kelabit
Formation is probably equivalent to the Long
Bawang Formation, mapped in contiguous
Kalimantan in a region probably underlain by
evaporites, for all groundwater wells produce brine
(BRGM 1982).
Uplift o f the Crocker Formation
Deformation resulting in spectacular episodic uplift
of the Crocker Formation is well documented in
the offshore oil fields by several prominent un-
conformity surfaces (Levell 1987; Tan & Lamy
1990; Hazebroek & Tan 1993). The first is the Deep
Regional Unconformity (15 Ma, Middle Miocene)
and the latest is the Shallow Regional Un-
conformity (9 Ma, Upper Miocene). The tectonic
front, or northwestwards limit of tectonic deforma-
tion and uplift, progressively moved ocean-wards
 r
/" 116o / 115o j SLUMP SCAR ,,60 (b t~16° (C)
PLaTfORM (a) ~5 ° / ON SLOPE
/ / /
/ ~/ C = CARBONATE / /C /(, 9Mo ,..,. u SRU
s
APPROX. . ,' Ovf" 12Mo -T--~- U I U 7o_
.... PALAEO- ENVIRONMENT ."" ~_~'.,@
~ ' ,"~ ~ \ ~ 5 ~
BOUNDARIES ," 13Mo, , , LIU
*+ , - / ~ / , ~ ~": X ~ x Ii // 15MOv , v DRU
.... EXTENT OF
/" I" /P~" / EAST EROSION -/
BARAM / ' ,i dVt:: "I ~
,4 DELTA
,// / // /
>
l ~ ,? .
z

©
KINABALU 6° %,°
IIKINABALU >
--DRU ~:
[.-

t"
/ ¢-, SABAH × SABAH
/ o ©
I Oilfields .5}
©
- - - - - PROVfNCEBOUNDARY ~ ~ COUNTERREGIONAL
ANTPCLINE GROWTHFAULT
SYNCLINE ~ REVERSE/THRUSTFAULT
" " MAJORGROWTHFAULT , NORMALFAULT

Fig. 3. (a) The main structural elements of northwest Sabah (after Tan & Lamy 1990; Rice-Oxley 1991). (b) Late Miocene palaeo-environments for stages IV E/F (after
Rice-Oxley 1991), showing also the positions of identified slump scars along the Late Miocene continental slope (shelf edge), from Levell & Kusumajaya (1985) and Hutchison
(1995). (c) Sea-ward limits of the important regional unconformities, which successively migrated away from the tectonic front of the uplifting West Crocker Formation
fold-thrust belt (based on Levell 1987). Palaeo-environments: LCE lower coastal plain; C, coastal; HIN, holomarine inner neritic; HMN, holomarine middle neritic; HON,
holomarine outer neritic; BAT, bathyal. Unconformities: DRU, deep regional unconformity; LIU, lower intermediate; U1U, upper intermediate; SRU, shallow regional
unconformity. Oil fields: 1. Samarang; 2. Glayzer; 3. Erb West; 4. Ketam; 5. Tembungo; 6. St Joseph; 7. SW Emerald; 8. Furious; 9. Barton.
tO
258 C.S. HUTCHISON
(Fig. 3). The unconformities extended quite close
to the continental slope (tectonic front), so that the
fluvial system brought sands to the edge of the very
narrow continental shelf to spill over the slope into
the bathyal area as submarine slumps, such as the
one described by Hutchison (1995), and to feed
turbidite fans. The persistence of outer neritic to
bathyal marine conditions, without unconformities,
west and northwest of the tectonic front has been
documented by Rice-Oxley (1991). The Late
Miocene tectonic front of the West Crocker-
Temburong orogen approximately coincides with
the Jerudong Line, Morris Fault, and Transition
zone between the Inboard and Outboard belts of
Shell geologists (Fig. 3).
The Mid to Late Miocene inversion of the West
Crocker-Temburong basin to form the Crocker
Ranges of Sabah has never been given a name, but
the term Sabah Orogeny seems appropriate. Its
cause remains obscure. The spectacular Late
Miocene granitoid forming the core of Mount
Kinabalu (4101 m) is a product of this orogeny,
with whole rock K-Ar dates commonly as old as
10 Ma (Jacobson 1970).
The Sabah orogeny regionally
This orogenic event was widespread throughout
the South China Sea, but only in the Zhu Jiang
Kou (Pearl River Mouth) Basin has it been given
a name: Dongsha Movement ( S u e t al. 1989).
However, this is not a good regional name, since
the inversion there is not spectacular. The nearby
Song Hong-Yingge Hai Basin contains very
impressive Middle to Late Miocene inversion
structures, well displayed in regional seismic
records, but they have not been published.
The best published account of spectacular basin
inversion is by Ginger et al. (1993) in the West
Natuna area, where the Anambas Graben became
the Boundary High during Middle to Late Miocene
inversion. The eroded anticlines are unconformably
overlain by fiat lying latest Miocene-Pliocene
Muda Formation. In the contiguous Penyu and
Malay basins, inversion structures have not yet
been analyzed and described because of a lack of
seismic resolution at depth. However, the Upper
Miocene unconformity has been well documented
References
ACHACHE, J., COURTILLOT, V. & BESSE, J. 1983.
Paleomagnetic constraints on the Late Cretaceous
and Cenozoic tectonics of Southeast Asia. Earth
and Planetary Science Letters, 63, 123-136.
ADAMS, C. G. & HAAK, R. 1962. The stratigraphical
succession in the Batu Gading area, Middle Baram,
North Sarawak. In: HAILE, N. S. The geology and
mineral resources of the Suai-Baram area, north
in the southern end of the Malay Basin (Armitage
& Viotti 1977; Nik Ramli 1988). Even though the
Middle to Late Miocene inversion structures may
not have been identified and described, the Upper
Miocene unconformity is widely documented
throughout Southeast Asia.
All of the South China Sea basins rapidly
returned to post-inversion subsidence after erosion
of the inversion structures, so that the basins remain
today beneath sea-level.
Much farther south, the Late Miocene inversion
is well displayed in the Meratus Mountains, which
formed the basement to the Eocene through
Miocene sequence, then rose up only in the Late
Miocene to form a ridge which subdivided a
formerly much more extensive basin (Hutchison
1992b). The post-inversion unconformity is also
seen in Central Sumatra (Hutchison 1993). Both
in the Meratus and Central Sumatra the uplifted
basins have not subsequently subsided beneath
sea-level.
Summary
Although the interior of Borneo remains poorly
known, the following broad evolutionary scenario
may now be summarized.
The turbiditic Rajang Group of Sarawak may
be usefully extended into northern Kalimantan
and central Sabah. It may be interpreted as an
accretionary prism, compressed and uplifted
between the Schwaner Mountains volcano-plutonic
arc and the Luconia-Balingian-Miri micro-
continent of Sarawak by the Upper Eocene
Sarawak orogeny, to become an integral part of the
extensive Upper Eocene Sundaland landmass upon
which the Ketungau and other continental basins
were developed.
The Oligocene to Lower Miocene turbiditic West
Crocker and Temburong Formations of Sabah
should be excluded from the Rajang Group. Their
nearshore equivalent has been mapped to the south
as the Kelabit and Long Bawang Formations,
suggesting provenance from eroding Rajang Group
outcrops of the Sarawak Orogen. The West Crocker
and Temburong Formations were deformed and
uplifted by the Middle to Upper Miocene Sabah
orogeny.
Sarawak. Geological Survey, Department of the
British Territories in Borneo. Kuching, Memoir, 13,
141-150.
AMIRUDDIN. & TRAIL, D. S. 1989. Nangapinoh 1:25,000
quadrangle, West Kalimantan, Geological Data
Record. Geological Research and Development
Centre, Bandung.
ARMITAGE, J. H. & VIOTTI, C. 1977. Stratigraphic
 RAJANG GROUP AND LUPAR LINE, BORNEO
nomenclature-southern end Malay Basin. In:
Proceedings Indonesian Petroleum Association
6th annual convention. 69-94.
BERGMAN, S. C., TURNER, W. S. & KROL, L. G. 1987.
A reassessment of the diamondiferous Pemali
Breccia, southeast Kalimantan, Indonesia: intrusive
kimberlite breccia or sedimentary conglomerate?
Geological Society of America Special Paper,
215, 183-195.
BLADON, G. M., PIETERS, P. E. • SUPRIATNA, S. 1989.
Catalogue of isotopic ages commissioned by the
Indonesia-Australia Geological Mapping Project
for igneous and metamorphic rocks in Kalimantan,
Preliminary Report. Geological Research and
Development Centre, Bandung.
BRIAIS, A., PATRIAT,P. & TAPPONNIER,P. 1993. Updated
interpretation of magnetic anomalies and seafloor
spreading stages in the South China Sea: implica-
tions for the Tertiary tectonics of Southeast Asia.
Journal of Geophysical Research, 98, 6299-6328.
, TAPPONNIER,P. & PAUTOT,G. 1989. Constraints of
data on crustal fabrics and seafloor spreading in the
South China Sea. Earth and Planetary Science
Letters, 95, 307-320.
BRGM. 1982. Geological mapping and mineral
exploration in northeast Kalimantan 1979-1982,
Final Report. Rapport du Bureau de Recherches
Grologique et Mini~res, 82, RDM 007AO,
Orlrans, France.
CARLILE, J. C., DIGDOWIROGO, S. & DARIUS, K. 1990.
Geologic setting, characteristics and regional
exploration for gold in the volcanic arcs of North
Sulawesi, Indonesia. Journal of Geochemical
Exploration, 35, 105-140
CARTER, E. G. & YEATS, A. K. 1973. The Lower Tertiary
in Kattim Shell contract area, East Kalimantan:
Results of 1972-1975 field surveys. Kaltim Shell
NV Report.
CHEN, S. 1987. Magnetic profiles. In: Atlas of Geology
and Geophysics of the South China Sea, scale
1:2 000 000. Second Marine Geology Investigation
Brigade of the Ministry of Geology and Mineral
Resources, Guangdong, PR of China.
COLLENETTE,P. 1965. The geology and mineral resources
of the Pensiangan and Upper Kinabatangan area,
Sabah, Malaysia. Malaysian Geological Survey,
Borneo Region, Memoir, 12.
DOUTCH, H . E 1992. Aspects of the structural histories
of the Tertiary sedimentary basins of East, Central
and West Kalimantan and their margins. BMR.
Journal of Australian Geology and Geophysics, 13,
237-250.
GINGER, D. C., ARDJAKUSUWAH,W. O., HEDLEY, R. J. &
POT~tECARY, J. 1993. Inversion history of the West
Natuna Basin: examples from the Cumi-Cumi PSC.
In: Proceedings Indonesian Petroleum Association
22nd annual convention. 635-658.
GOWER, R. J. W. 1990. Early Tertiary plate reconstruc-
tions for the South China Sea region: constraints
from northwest Borneo. Journal of Southeast Asian
Earth Sciences, 4, 29-35.
HAILE, N. S. 1962. The geology and mineral resources of
the Suai-Baram area, north Sarawak. Geological
Survey Department of the British Territories in
Borneo, Kuching, Memoir, 13.
259
-- 1974. Borneo. In: Spencer, A. M. (ed.) Mesozoic-
Cenozoic Orogenic Belts: Data for Orogenic
Studies. Geological Society, London, Special
Publication, 4, 333-347.
- - 1994. The Rajang accretionary prism and the Trans-
Borneo Danau Suture. Abstracts Tectonic Evolution
of SE Asia Conference, December 1994. Geological
Society, London, 17.
-- & Ho C. K. 1991. Geological Field Guide: Sibu-
Miri traverse, Sarawak (24 September-1 October
1991). Petronas Petroleum Research Institute, Kuala
Lumpur.
--, MCELHINNY, M. W. & McDOUGALL, I. 1977.
Palaeomagnetic data and radiometric ages from the
Cretaceous of west Kalimantan (Borneo), and their
significance in interpreting regional structure.
Journal of the Geological Society, London, 133,
133-144.
HAMILTON, W. 1979. Tectonics of the Indonesian region.
US Geological Survey, Professional Paper, 1078.
HAZEBROEK, H. E & TAN, D. N. K. 1993. Tertiary
tectonic evolution of the NW Sabah continental
margin. Geological Society of Malaysia Bulletin,
33, 195-210
Ht]KINIAN, R., BONTt~, P., PAUTOT, G., JACQUES, D.,
LABEYRIE,L., ETAL. 1989. Volcanics from the South
China Sea ridge system. Oceanologica Acta, 12,
101-115.
HOLLOWAY,N. H. 1981. The stratigraphy and relationship
of Reed Bank, north Palawan, and Mindoro to the
Asian mainland and its significance in the evolution
of the South China Sea. AAPG Bulletin, 66,
1355-1383.
HUTCHISON, C. S. 1989. Geological evolution of South-
East Asia. Clarendon Press, Oxford, Oxford
Monographs on Geology and Geophysics, 13.
1992a. The Southeast Sulu Sea, a Neogene
marginal basin with outcropping extensions in
Sabah. Geological Society of Malaysia Bulletin,
32, 89-108.
1992b. The Eocene unconformity in Southeast
and East Sundaland. Geological Society of Malaysia
Bulletin, 32, 69-88.
1993. Gondwanaland and Cathaysian blocks,
Palaeotethys sutures and Cenozoic tectonics in
South-east Asia. Geologische Rundschau, 82,
388-405.
1995. Mrlange on the Jerudong Line, Brunei
Darussalam, and its regional significance.
Geological Society of Malaysia Bulletin, in press.
JACOBSON, G. 1970. Gunung Kinabalu area, Sabah,
Malaysia. Geological Survey of Malaysia, Report,
8.
JAMES, D. M. D. 1984. The geology and hydrocarbon
resources of Negara Brunei Darussalam. Brunei
Museum and Brunei Shell Petroleum Company.
KAVALIERIS,I., VAN LEEUWEN,T. M. & WILSON, M. 1992.
Geological setting and styles of mineralization,
North arm of Sulawesi, Indonesia. Journal of
Southeast Asian Earth Sciences, 7, 113-129.
KIRK, H. J. C. 1957. The geology and mineral resources of
the Upper Rajang and adjacent areas. Geological
Survey Department for the British Territories in
Borneo, Kuching, Memoir, 8.
-- 1968. The igneous rocks of Sarawak and Sabah.
260 c . s . HUTCHISON
Geological Survey of Borneo Region, Malaysia
Bulletin, 5.
LEE, CHAI-SHtNG & MCCABE, R. 1986. The Banda-
Celebes-Sulu Basin: a trapped piece of Cretaceous-
Eocene oceanic crust. Nature, 322, 51-54
LEVELL, B. K. 1987. The nature and significance
of regional unconformities in the hydrocarbon-
bearing Neogene sequences offshore West Sabah.
Geological Society of Malaysia Bulletin, 21, 55-90
- - & KASUMAJAYA, A. 1985. Slumping at the late
Miocene shelf-edge offshore West Sabah: a view
of a turbidite basin margin. Geological Society of
Malaysia Bulletin, 18, 1-29.
LIECHTI, P., ROE, E W. & HALLE,N. S. 1960. The geology
of Sarawak, Brunei and the western part of North
Borneo. Geological Survey Department for the
British Territories in Borneo, Kuching, Bulletin, 3.
LIM, P. S. 1985. Geological Map of Sabah, 1:500,000.
Geological Survey of Malaysia, Kota Kinabalu.
MOLENGRAA~, G. A. E 1902. Geological Explorations in
Central Borneo (1893-94). E. J. Brill, Leyden.
NIK RAMLI, 1988. Characteristics of J-sandstone (Tapis
Formation) reservoirs in the southeastern part of the
Malay Basin. In: Proceedings of the Southeast
Asian Petroleum Exploration Society (Seapex). 8,
239-248.
PAUTOT, P., RANGIN, C., BRIAIS, A., TAPPONNIER, P.,
BEUZART, P., ET AL. 1986. Spreading direction in the
Central South China Sea. Nature, 321, 150-154.
PIETERS, P. E. & SUPRIATNA, S. 1989. Preliminary
Geological Map of the West, Central and East
Kalimantan area, 1:1 000 000. Geological Research
and Development Centre, Bandung.
, TRAIL, D. S. & SUPR1ATNA, S. 1987. Correlation
of Early Tertiary rocks across Kalimantan. In:
Proceedings Indonesian Petroleum Association
16th annual convention. 291-306.
REINHARD,M. & WENK, E. 1951. Geology of the colony of
North Borneo. Geological Survey Department
of British Territories in Borneo, Kuching, Bulletin,
1.
RICE-OXLEY, E. D. 1991. Palaeoenvironments of the
Lower Miocene to Pliocene sediments in offshore
NW. Sabah area. Geological Society of Malaysia
Bulletin, 28, 165 194.
Ru, KE & PIGOTT, J. D. 1986. Episodic rifting and
subsidence in the South China Sea. AAPG Bulletin,
70, 1136--1155.
SCHMIDTKE, E., FULLER, M. & HASTON, R. 1990.
Paleomagnetic data from Sarawak, Malaysian
Borneo, and the Late Mesozoic and Cenozoic
tectonics of Sundaland. Tectonics, 9, 123-140.
SIKUMBANG, N. 1986. Geology and tectonics of pre-
Tertiary rocks in the Meratus Mountains, south-east
Kalimantan, Indonesia. PhD Thesis, University of
London.
STAUFFER, P. H. 1968. Studies in the Crocker Formation,
Sabah. Borneo Region of Malaysia, Geological
Survey Bulletin, 8, 1-13.
Su DAQUAN, WHITE, N. & MACKENZIE, D. 1989.
Extension and subsidence of the Pearl River Mouth
Basin, northern South China Sea. Basin Research,
2, 205-222.
TAN, D. N. K. 1979. Lupar Valley, West Sarawak,
Malaysia. Geological Survey of Malaysia, Report,
13, Kuching, Sarawak.
1982. Geological map of Sarawak, 1:500,000.
Geological Survey of Malaysia, Kuching.
1986. Palaeogeographic development of West
Sarawak. Geological Society of Malaysia Bulletin,
19, 39-49.
-- & LAMY, J. M. 1990. Tectonic evolution of the NW.
Sabah continental margin since the Late Eocene.
Geological Society of Malaysia Bulletin, 27,
241-260.
TATE, R. B. 1991. Cross-border correlation of geological
formations in Sarawak and Kalimantan. Geological
Society of Malaysia Bulletin, 28, 63-95
TAYLOR, B. & HAYES, D. E. 1980. The tectonic evolution
of the South China Basin. In: HAYES, D. E. (ed.) The
Tectonic and Geologic Evolution of Southeast Asian
Seas and Islands. American Geophysical Union,
Geophysical Monograph, 23, Washington,
89-104.
- - & -- 1983. Origin and history of the South China
Sea Basin. In: HAYES, D. E. (ed.) The Tectonic and
Geologic evolution of Southeast Asian Seas
and Islands, Part 2. American Geophysical
Union, Geophysical Monograph, 27, Washington,
23-56.
TING, C. S. 1992. Jurassic-Cretaceous palaeogeography
of the Jagoi-Serikin area as indicated by the
Bau Limestone Formation. Geological Society of
Malaysia Bulletin, 31, 21-38.
TONGKUL, E 1990. Structural style and tectonics of
Western and Northern Sabah. Geological Society of
Malaysia Bulletin, 27, 227-239.
VAN BEMMELEN, R. W. 1949. The geology of Indonesia.
Government Printing Office, The Hague.
VAN DE WEERD, A. A. & ARMIN, R. A. 1992. Origin and
evolution of the Tertiary hydrocarbon-bearing
basins in Kalimantan (Borneo), Indonesia. AAPG
Bulletin, 76, 1778-1803.
VAN LEEUWEN, T. M., LEACH, T., HAWKE, A. A. &
HAWKE, M. M. 1990. The Kelian disseminated
gold deposit, East Kalimantan, Indonesia. Journal
of Geochemical Exploration, 35, 1-6 I.
WAIN, T. & BEROD, B. 1989. The tectonic framework
and paleogeographic evolution of the Upper Kutei
Basin. In: Proceedings Indonesian Petroleum
Association 18th annual convention. 55-78.
WHITE, J. M. JR. & WING, R. S. 1978. Structural develop-
ment of the South China Sea with particular
reference to Indonesia. In: Proceedings Indonesian
Petroleum Association 7th annual convention.
159-164.
WILLIAMS, P. R. & HARAHAP, B. H. 1987. Preliminary
geochemical and age data for post-subduction
intrusive rocks, Northwest Borneo. Australian
Journal of Earth Sciences, 34, 405-4 15.
& HERYANTO, R. 1986. Sintang 1:250 000
Quadrangle, West Kalimantan: Geological Data
Record. Geological Research and Development
Centre, Bandung.
--, JOHNSTON, C. R., ALMOND, R. A. & SIMAMORA,
W. H. 1988. Late Cretaceous to Early Tertiary
structural elements of West Kalimantan. Tectono-
physics, 148, 279-297.
 RAJANG GROUP AND LUPAR LINE, BORNEO 261

WILSON, R. A. M. 1964. The geology and mineral
resources of the Labuan and Padas Valley area,
Sabah, Malaysia. Geological Survey of the Borneo
Region of Malaysia, Memoir, 17.
WOLFENDEN, E. B. 1960. The geology and mineral
resources of the Lower Rajang Valley and adjoining
areas, Sarawak. Geological Survey, Department for
the British Territories in Borneo, Kuching, Memoir,
11.
ZEIJLMANSVANEMMICHOVEN,C. P. A. 1939. De Geologie
van het centrale en oostelijke deel van de
Westerafdeeling van Borneo. Jaarboek van het
Mijnwezen in Nederlandsch-lndie Verlhandelingen,
7-186.
 Origin and tectonic significance of the metamorphic rocks
associated with the Darvel Bay Ophiolite, Sabah, Malaysia
S H A R I F F A. K. O M A N G 1 & A. J. B A R B E R 2
1 Department of Earth Sciences, Faculty of Science and Natural Resources,
Universiti Kebangsaan Malaysia, Sabah Campus, Locked Bag No 62,
88996 Kota Kinabalu, Sabah, Malaysia
2 SE Asia Research Group, Department of Geology, Royal Holloway,
University of London, Egham TW20 OEX, UK

Abstract: Banded hornblende gneiss, foliated amphibolite, hornblende, chlorite and siliceous
schist form lenses in an 8 km wide belt within the Darvel Bay Ophiolite Complex. Foliation in
the belt is generally steep to vertical, striking parallel to the trend of the belt and lineations are
sub-horizontal. Mineral and geochemical studies show that the metamorphic rocks represent
banded and isotropic gabbros, plagiogranites, doleritic and basaltic dykes, basaltic volcanics and
cherts formed at a spreadingridge in a supra-subductionzone environment, which were deformed
at high temperatures but low pressures along a transform fault. Incorporation of supracrustal
cherts indicates that the transform extended for hundreds of kilometres between spreading
centres. Garnet pyroxenites and amphibolites found as clasts in Miocene volcanic agglomerates
formed at high pressures, and temperatures are interpreted as derived from a metamorphic sole
underlying the complex, formed during subduction of ocean crust and the emplacement of the
ophiolite complex on Sabah.

Ophiolite outcrops are distributed throughout the
eastern part of Sabah, East Malaysia, from Banggi
Island in the north to Ranau and Telupid and the
Lahad Datu area in the south (Fig.l, inset). The
Darvel Bay Ophiolite Complex in the south is the
most extensive outcrop, extending 100 km west-
wards from Darvel Bay. The greater part of the
complex consists of peridotite, largely serpen-
tinized, but it also includes cumulate pyroxenites,
layered and massive gabbros and diorites, meta-
morphosed to varying degrees. The complex is also
cut by dolerite dykes, although these are never so
abundant to be termed a sheeted dyke complex.
These rock types are closely associated with out-
crops of the Chert-Spilite Formation, composed of
pillow basalt, banded ribbon chert, turbiditic sand-
stones, mainly volcaniclastic but with some rare
quartz sandstones, and a few occurrences of
massive limestone.
The ophiolite complex has been well described
in the publications of the Geological Survey of
Malaysia (Reinhard & Wenk 1955; Fitch 1955;
Dhonau & Hutchison 1966; Koopmans 1967). The
complex has been interpreted as a segment of ocean
floor, either of a Proto-South China Sea (Holloway
1981; Rangin et al. 1990) or of the Celebes Sea
(Hutchison 1988). A wide range of K-Ar age dates
has been obtained from the rocks of the ophiolite
complex from 210Ma (Leong 1971) to 137 Ma
(Rangin et al. 1990). Cherts from the Chert-Spilite
Formation have yielded radiolaria of Lower
Cretaceous age (Leong 1977; Rangin et al. 1990;
Aitchison 1994). Massive limestones associated
with the Chert-Spilite Formation contain
Cretaceous foraminifera (Leong 1974). The Chert-
Spilite Formation is interpreted as representing
ocean floor sediments which were deposited on top
of the ophiolite (Hutchison 1975), and carbonate
cappings to seamounts. Since the underlying
oceanic crust is unlikely to be much older than the
oldest overlying sediments, the Early Jurassic K-Ar
ages are regarded as spurious (Hutchison 1988).
Fragments of ophiolitic rocks are found as clasts
in Eocene sediments (Newton-Smith 1967; Rangin
et al. 1990), suggesting that the ophiolite complex
had been obducted onto Sabah either in the latest
Cretaceous or earliest Palaeogene. In the early
Miocene the ophiolite complex formed the base-
ment to a volcanic arc, possibly related to continued
subduction of the Proto-South China Sea. Mitchell
et al. (1986) and Rangin (1989) have suggested
that, as a result of continued compression, the
complex was backthrust over the Celebes Sea floor
to the south.
The ophiolite complex is surrounded by chaotic
melange deposits (Fig. 1) which contain fragments
of all the rock units represented in the complex
and the Chert-Spilite Formation, as well as

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, 263

Geological Society Special Publication No. 106, pp. 263-279.
264 s . A . K . OMANG (~ A. J. BARBER

0 5 10
r' Km Ophiolite Complex
E ~ Felsite ~ RecentAIluvium
~ ~'~] Metamorphic rocks E~ Chert-Spilite
• " '* '° "~"* "*"*"*"° "° "° ". "° " . . . - , ~ '. ". "° '. ". ". "° ". Formation
~ ' . ' ~ * * , ' i**: (foliated metabasites )
~ ~ : : * : i * " []'EI Sheared Serpentinite D Neogene
, ,, ,, , , , ,, ,, -° "4 "° . "° "° .o .° .° ...... sediments
, ', ', ', ', ', ', "°,:*.:*.:*.:°.:*,:*,:°,:*,:*, ~ Gabbros & Cumulates [ ~ M6langes
"°"°' ' °"°' l Peridotite ~ Thrust fault
(mainly serpentinised) - - " Fault
:!:!:!:!:i: l: 8" 20' E. :::::::::::::::::::::::::::::::::::::::::::::::::::::
~!~:i:i !i:i:i:!...i.i~.:i:i:i!!ii i!i!i i i i i i i i !i!i i i!i!i !i i!::ii!i!~ii
~~:i.i:i:i:i!ii!i!i!iii!iiiiii!iiiiiiii!!!iii!!!iliiiiiiiiiii
Gabbro
====================================================
*Dolerite dyke
PSI2
D~:: PS6

Itlr Pulau Sakar

! I ! rl~l ! ! ! ! |
, ~ ~PT3 Pulau "~;K4
~ , '
ruiau %,• 9~l Katung- /
~1 i i : • ~. Saddle/ % Kalungan-.L
l.ll.l~.l~.ll. 1.1~.13.11.1~.1~
I~. ~ ~ -.L Cumulate D
i i i i i i i i--.~.,1-.-~
p ,,.z.~ p C rocks
Darvel Bay
E t

Pulau
7 4* 58'N
SARAWAK 1 " ~
~, Bohayan Pulau Tabauwan
_/'f / ~,
P ~
f
V~Silumpat
/
~':*'i*' 118" 10' E. E --/- 118"20' E Dolerite dyke
I
Fig. 1. Geological map of the Darvel Bay Ophiolite Complex near Lahad Datu showing location of samples used in
this study. Inset map shows distribution of ophiolite complexes in Sabah. B, Banggi Island; R, Ranau; T, Telupid;
DB, Darvel Bay.

y o u n g e r sediments and volcanic rocks. Clennell

(1991) considered that the melanges were f o r m e d
by p r o c e s s e s o f s e d i m e n t a r y s l u m p i n g and
diapirism, coincident with the collision o f micro-
continental blocks with the north Sabah margin and
the extension o f the arc basement related to the
d e v e l o p m e n t o f the Sulu Sea in late early and early
mid M i o c e n e times. The ophiolite c o m p l e x and
the melanges are overlain u n c o n f o r m a b l y by Late
M i o c e n e to Pliocene sediments o f the 'circular
basins' and in the Quaternary again f o r m e d the
basement to a volcanic arc which extends from
M i n d a n a o in the Philippines, through the Sulu
archipelago to the Dent and S e m p o r n a peninsulas
o f Sabah.
Metamorphic rocks
A l t h o u g h the rocks f o r m i n g the Darvel B a y
Ophiolite Complex are variably m e t a m o r p h o s e d
throughout, d y n a m i c a l l y m e t a m o r p h o s e d rocks
are concentrated in an E - W belt c. 8 k m wide
extending westwards from Lahad Datu on the north
side o f Darvel Bay (Fig.l) which can be traced for
c. 40 k m to the west. Metamorphic rocks are well
exposed in coastal sections around the shores o f
Darvel Bay and particularly around the small
islands within the bay, w h e r e they have been
described previously by D h o n a u & H u t c h i s o n
(1966) and Hutchison & D h o n a u (1969, 1971) f r o m
the north side o f the bay, and by Koopmans (1967)
 DARVEL BAY OPHIOLITE, SABAH
from the south. Inland exposures are poor, apart
from road-cuts, quarries and scarce river sections.
The metamorphic rocks have been described as
metagabbros, amphibolites, hornblende gneisses
and schists. Reinhard & Wenk (1955) described
the dynamically metamorphosed rocks as
'Crystalline Schists' and Koopmans (1967) identi-
fied them as the 'Crystalline Basement Complex',
with the implication that they were the oldest rock
unit in Sabah. However, K-Ar dating reported by
Dhonau & Hutchison (1966) showed that these
metamorphic rocks are of Cretaceous age, the same
age as the Darvel Bay Ophiolite Complex, and form
an intrinsic part of the complex
The structure of the dynamically metamorphosed
belt has been fully described by Dhonau &
Hutchison (1966). The metamorphic rocks are
typically banded, schistose and lineated. Through
most of the belt the banding and schistosity are
vertical or steeply dipping and strike E-W, parallel
to the general trend of the belt, swinging round to a
more NW-SE trend in the islands of Darvel Bay.
The lineation is sub-horizontal, lying in the plane
of the schistosity and plunges at low angles either
to E or W. From their mapping of the structure
in the metamorphic rocks, Dhonau & Hutchison
(1966) identified two large-scale open post-
metamorphic monoclinal flexures on E-W axes
affecting the foliation in the islands of Pulau
Saddle, Pulau Bohayan and Pulau Silumpat.
This paper gives an account of the field relations,
petrography, mineral chemistry, geochemistry and
K-At isotopic dating of the metamorphic rocks
associated with the Darvel Bay Ophiolite Complex
and discusses their origins and significance in the
tectonic evolution of Sabah. A separate occurrence
of metamorphic rock is found among Miocene
volcanics at Tungku, 50 km to the east of Lahad
Datu (Reinhard & Wenk 1955). Garnet amphibolite
from this locality was also included in the present
study.
Banded hornblende gneiss
The dominant metamorphic rock type in the meta-
morphic belt within the Darvel Bay Ophiolite
Complex is a banded gneiss, the 'Silumpat Gneiss'
of Dhonau & Hutchison (1966) and Hutchison &
Dhonau (1971), with alternating bands of white
feldspathic and black or dark greenish ferro-
magnesian bands, well exposed on the north shore
of Darvel Bay, west of Lahad Datu and in road cuts
along the Silam Road. The bands range in thick-
ness from tens of centimetres to millimetres. The
banding is paralleled by a schistosity, with the
alignment of feldspathic and ferromagnesian aggre-
gates which may also be elongated to form a
rodding type of lineation. The intensity of the
265
schistosity increases in localized shear zones. The
banding is commonly folded into tight intrafolial
isoclinal folds which may show the closed eyed
outcrop pattern characteristic of sheath folds.
Where folds are present the schistosity is parallel to
the axial planes of the folds. In the more intensely
deformed shear zones aligned hornblende and
feldspar crystals define a mineral lineation, lying
in the plane of the schistosity.
Amphibolite
Amphibolites are commonly found associated with
the banded gneisses in the field. These may form
extensive outcrops, as on the northern side of Sakar
Island or as smaller bodies which may show cross-
cutting, intrusive relationships to the hornblende
gneisses. The grain size of these amphibolitic rocks
is variable. Coarse grained amphibolites show the
characteristic distribution of white feldspathic and
black or green ferromagnesian minerals which
indicate that they have been derived from gabbros
(metagabbro). Medium and finer grained
amphibolites may show relict ophitic textures in
thin section and large euhedral plagioclase pheno-
crysts or polycrystalline aggregates pseudo-
morphing plagioclase which indicate an origin as
hypabyssal intrusive rocks (metadolerite). Fine
grained amphibolites without these features may
represent basaltic volcanic rocks (metabasalts).
Like the banded gneisses the amphibolites show
varying degrees of deformation. Where they cross-
cut the banding in the gneisses, the amphibolites
generally show a less intensely developed
schistosity than the associated gneisses. In coarse
grained amphibolites the gabbroic texture may be
flattened in the foliation, in finer grained rocks
minerals may be recrystallized in the schistosity
to form hornblende schists. Where the coarser
amphibolites cross-cut the banding of the gneisses,
no clear chilled margins are evident and the
schistosity crosses the contact between the gneisses
into the amphibolites, although it is generally not
as intensively developed in the latter. Finer grained
amphibolites often have a dyke-like form, ranging
in width from a few metres to tens of centimetres
and show varying angular relationships with the
banding of the gneiss. Sometimes they are con-
cordant, and then generally have a well developed
schistosity with a similar intensity to that in the
adjacent gneisses. Where amphibolite dykes cut the
banding at a high angle, the schistosity is generally
less well developed than in the surrounding gneiss.
Evidently some of the dykes were intruded into
already deformed gabbros, but deformation con-
tinued after dyke emplacement, indicating that
the dykes were intruded into a zone of active
deformation.
266 s . A . K . OMANG t~ A. J. BARBER

twinning. Microprobe analysis shows that where

Mineralogy
In the less highly deformed and altered hornblende
gneisses and metagabbros the ferromagnesian
component includes relict crystals of pale green
or colourless pyroxene. These may form the cores
to brownish-green amphibole crystals and show
alteration along their margins and along internal
cracks to aggregates of small blue-green amphibole
crystals. Microprobe analysis shows that they are
clinopyroxenes of salite to augite composition
(En40.5 Fst3 Wo47-En45.5 Fs]2.7 Wo28),with Xrvlg
values of 90-97 and an Na20-content of 0.5 wt%.
Plagioclase crystals have a grain size up to
4 ram, and form tabular crystals with albite
plagioclase is associated with pyroxene it has a
composition of An68-An74 (labradorite to
bytownite) representing an igneous relic, but more
commonly is An 6 -An32 (oligoclase to andesine)
of metamorphic origin (Hutchison 1978). In the
amphibolites, and especially in highly deformed
rocks, the plagioclase is An 3 - A n 5 (albite). In
many samples of hornblende gneiss and amphibo-
lite the plagioclase has been completely pseudo-
morphed by low grade alteration products
(saussurite).
Epidote is a common constituent of the gneisses
and amphibolites as an alteration product of plagio-
clase feldspars, but in occurrences at the western

Tschermakite Pargasite

II Hornblende
Edenite

• Metagabbro
d- Metadolerite
Tremotite I
0 0.5 1.0
(a)

Glaucophane
2.0 1

1.8
]4)Si(IV) = Ca(M4)Ai(IV)
1.6 t Sodic-am~hibole
Na[M4] 1.4
..................................................... ....... II
1.2
Winchite a.0 Barroisite aramite Sodic-Calcicamphibole
0.8
. . . .

Calcic-amphibole
0.2 ~ + Metadolerite I
0.0
0 2
Actinolite Hornblende Tschermakite
AI[IV]
(b)
Fig. 2. (a) Composition of amphiboles in metagabbro and metadolerite from Darvel Bay on diagram of Deer etal.
(1966). (b) AI[IV] versus Na[M4] plot of amphibole compositions in metagabbro and metadolerite from Darvel Bay,
isobars after Brown (1977).
 DARVEL BAY OPHIOLITE, SABAH
end of Pulau Sakar and in Pulau Katung Kalungan
there are amphibolites in which epidote is a major
component of the rock. These rocks are distin-
guished in the field by their yellowish colour where
they occur as bands among more normal amphibo-
lites. In thin section they are foliated, with large
crystals of epidote, colourless or pale pink in a
matrix of blue-green amphibole and plagioclase.
The plagioclase is albite or is represented by an
aggregate of sericite and calcite. Microprobe
analysis of the epidote crystals show that it is an
iron-rich variety (Xps27_33). The Fe 3+ content is
very high ( 1 . 5 - 2.0 atoms per unit cell) and the
XFe 3+ (1.0) gives an estimated temperature of
formation of c. 400°C (Nakajima et al. 1977).
Amphiboles in the amphibolites occur either as
large brownish-green or as small blue-green
crystals which may be randomly oriented or aligned
to form foliar and linear structures. The larger
amphibole crystals are 2-4 mm in size and are
greenish-brown in colour. Some of the brown
hornblende has been interpreted by Hutchison
(1978) on textural and mineral chemical grounds
as of relict igneous origin. However, where they
enclose pyroxene relics, or where they are aligned
in the foliation and lineation, hornblende crystals
are clearly of metamorphic origin. Electron micro-
probe analyses plotted on the (Na+ K)/A1TM
diagram of Deer et al. (1966) show that the amphi-
bole composition lies between the hornblende and
pargasitic hornblende fields (Fig. 2a) (cf. Hutchison
1978, fig. 6). On the AlIV/Na (M4) plot of Brown
(1977) compositions extend from the calcic
amphibole towards the sodic-calcic amphibole
field, corresponding to pressures of about 5 kbar
(Fig.2b). On the AllY/A1vI plot of Fleet & Barnett
(1978) the amphiboles lie in the low pressure field
(Fig. 3a) and on the Na + K/A1TMplot of Jamieson
(1981) follow the high temperature/low pressure
trend (Fig. 3b). The Ti content of the amphiboles
(0.04-0.3 atoms per formula unit) indicates that
crystallization of the amphiboles took place in the
temperature range 550-650°C under low pressure
conditions in the lower amphibolite facies (Spear
1981). Using the semi-empirical geothermometer
proposed by Plyusnina (1982) for plagioclase/
hornblende pairs, core compositions of crystals
from the amphibolites give temperatures of
580-600°C and pressures of 3-4 kbars for R T
conditions at the time of crystallization.
Small blue-green amphibole crystals commonly
occur in the banded gneisses and amphibolites
marginal to pyroxene or brown hornblende crystals.
In the study by Hutchison (1978), microprobe
analyses of blue-green amphiboles from the Darvel
Bay Ophiolite Complex showed that their com-
position lay mainly in the actinolitic hornblende
field with some actinolite. In highly deformed
267
hornblende schists these actinolitic hornblendes
form a major component of the rock and are aligned
to form the schistosity and lineation.
Metamorphic histo:y
As far as may be determined the primary minera-
logical composition of the banded and isotropic
gabbros, now found as banded hornblende gneiss
and amphibolite, was clinopyroxene, hornblende
(cf. Hutchison 1978), labradorite/bytownite and
ilmenite. Olivine and orthopyroxene may have
been present in some rocks, as olivine and noritic
gabbros occur elsewhere in the Darvel Bay
Ophiolite Complex (Omang 1993), but all evidence
of their previous existence has been destroyed. The
gabbros were recrystallized in a high temperature,
low pressure metamorphic environment where
pyroxene was replaced by brownish-green horn-
blende and plagioclase by oligoclase/andesine.
Alignment of brown hornblende and plagioclase
crystals with foliar and linear structures indicate
that recrystallization took place in a high tempera-
ture dynamic environment; continued dynamic
metamorphism under lower temperature amphibo-
lite facies conditions are indicated where green
amphiboles are aligned with the schistosity and
lineation. Later recrystallisation under hydro-
thermal conditions at lower temperatures resulted
in the local replacement of plagioclase by
saussurite and the alteration of pyroxene and
hornblende to chlorite.
Geochemistry
Whole rock major element and trace element
analyses of hornblende gneisses and amphibolites
are shown in Table 1. On an AFM plot (Fig. 4)
these metabasites lie in the tholeiite and oceanic
gabbro (MORB) field of Kirst (1976). AlzO3]TiO2
ratios also fall in the MORB field (Fig. 5) of Sun
& Nesbitt (1978). Trace element analyses show that
high field strength (HFSE) elements (e.g. Zr, Y)
have a high concentration relative to large ion
lithophile (LILE) elements. The Cr-Y plot (Pearce
et al. 1984b) and Ti-Zr-Y diagram (Pearce & Cann
1973) also suggest a MORB-like character.
On the basis of spider diagrams the samples can
be divided into three groups. The metadolerite
dykes (Fig. 7a), are a suite of basaltic rocks with
moderate LILE enrichment but are with a clear
negative Nb anomaly relative to the light rare earth
elements (LREE). One group of metagabbros
(Fig. 7b) is very similar to the metadolerites, but
with more primitive compositions, indicated by
high Cr and Ni, and with much lower contents of
268 s.A.K. OMANG(~ A. J. BARBER

Pargasite Tschermakite
2

~-'l'/ High-pressure
~i~'l~/ CT"amphib°lez°ne
Ed~. .............Homilende ........................................
AI[IV] 1

Low-pressure
Actinolite ole zone
0 Glaucophane
0 2
(a) AI[Vl]
Glaucophane

2.01.51 High-pressuretrend

Na+K
Eckermakite 1
1 / Ede ite Pargasite

[ High-temperature
i + .=
d

0.0 Tremolite/A~tinolit? ~ , Tschermakite

0.0 0.5 1.0 1.5 2.0
(b) Ai[IV]
Fig. 3. (a) AI[VI] v. AI[IV], and (b) AI[IV] v. (Na + K) of amphibole compositions in metagabbro and metadolerite
from Darvel Bay; boundary between low and high pressure calcic amphiboles in (a) is from Fleet & Barnett (1978);
high and low pressure trends in (b) from Jamieson (1981).

K, Rb and Ba. The other group of metagabbros

(Fig. 7c) includes samples which have very spiky
profiles. They are extremely depleted in most
incompatible elements, particularly Nb, LREE, P
and Zr; K and Rb are conspicuously low, and only
Ba and Sr are at MORB levels in some samples.
In view of their textures and the field relation-
ships, the metabasic rocks are interpreted as part
of a single ophiolite suite, with geochemical data
indicating MORB-like compositions, but Nb
deficient and show a mild LILE enrichment. The
metadolerites represent an original liquid, with a
chemistry suggesting a supra-subduction zone
setting for their environment of formation. The
group of metagabbros which resemble the meta-
dolerites on spider diagrams are suggested to be
former isotropic gabbros, while the other group
of metagabbros probably represent cumulate rocks
which have lost an evolved inter-cumulus liquid
fraction enriched in incompatible elements.
Felsic rocks
The hornblende gneisses and amphibolites are
frequently cut by felsic veins on the outcrop scale.
These veins may be concordant, parallel to the
foliation and schistosity and are then boudinaged,
or are cross-cutting. Cross-cutting veins may be
straight or folded in a ptygmatic style, and un-
deformed veins may cut earlier folded veins.
Where the veins are folded, the schistosity in the
surrounding rocks has an axial plane relationship
to the folds. These relationships provide clear
evidence of multiple intrusion of felsic rocks into
a zone of active deformation.
At several localities within the metamorphic belt
larger bodies of felsic rock are found. On the north
shore of Pulau Sakar a felsic body, identified as
trondhjemite (Specimen PS12c) has been intruded,
with an irregular contact and without a chilled
margin, into metadolerites. A weak schistosity is
Table 1. Whole rock geochemical analyses for metagabbros and metadolerites of the Darvel Bay Metamorphic Complex

Metagabbro Metadolerite

Sample
Element JS4a JS L JSp KS2 PS2 PS 3a PS6 PS 12d PS 11 PS 11b PS 12a PS 12b PS 13d PT3a

SiO 2 44.22 48.07 47.55 49.73 46.42 48.82 50.11 48.52 51.42 51.24 49.28 48.82 51.43 51.42
TiO 2 1.70 1.63 1.55 0.62 0.97 0.23 0.57 1.28 1.61 1.64 1.34 1.63 1.61 1.55
A1203 15.86 15.25 15.62 14.17 17.87 19.99 15.49 15.05 14.87 14.30 15.64 15.40 14.67 15.32
Fe203* 12.38 11.70 11.55 8.58 9.83 5.58 6.57 10.60 12.10 12.07 10.96 11.37 11.27 13.94
MgO 9.13 8.60 8.59 10.53 8.83 8.50 9.96 6.52 6.44 6.86 8.45 9.29 7.22 4.64
MnO 0.22 0.18 0.19 0.15 0.18 0.14 0.13 0.18 0.20 0.19 0.17 0.19 0.20 0.23
CaO 15.01 11.40 11.95 13.41 13.52 14.28 14.85 14.74 8.45 9.00 10.22 9.52 8.18 8.18
N a2 0 0.99 2.94 2.75 2.38 2.01 1.96 2.07 2.26 4.46 3.84 3.58 3.12 4.62 4.10
K20 0.03 0.06 0.04 0.25 0.12 0.12 0.03 0.06 0.28 0.26 0.18 0.15 0.19 0.20
P205 0.13 0.15 0.13 0.02 0.02 0.01 0.02 0.12 0.16 0.15 0.17 0.14 0.15 0.21
Total 99.67 99.98 99.92 99.84 99.78 99.63 99.79 99.33 99.98 99.54 100.0 99.62 99.53 99.79
LOI 3.44 1.83 1.80 1.39 2.56 2.73 1.59 1.31 1.47 1.87 1.51 2.02 1.23 2.60

A1203/TiO2 9.33 9.36 10.08 22.86 18.42 86.91 27.18 11.76 9.24 8.72 11.67 9.45 9.11 9.88
CaO/TiO 2 8.83 6.99 7.71 21.63 13.94 62.09 26.05 11.52 5.25 5.49 7.63 5.84 5.08 5.28
mg c. 59 c. 59 c. 59 c. 71 c. 64 c. 75 c. 75 c. 55 c. 52 c. 53 c. 61 c. 62 c. 56 c. 40

Ni 115 71 105 127 206 159 128 103 52 45 113 136 92 11

Cr 338 364 368 487 413 243 886 269 80 74 282 296 273 5
V 355 336 331 228 392 120 214 303 354 345 308 314 295 386
Sc 56 57 53 60 69 41.6 63 41 49 49 43 43 46 35
Pb 0.7 n.d. n.d. n.d. 0.6 n.d. 0.3 0.6 1.1 0.4 1.3 n.d. 0.3 1.2
Sr 155.6 160.9 168.6 130.0 333.1 493.2 103.1 283.6 105.5 170.6 196.8 126.4 109.0 158.6
Rb 0.6 0.4 0.3 4.9 0.6 0.9 0.2 0.4 2.9 4.1 1.8 1.5 2.0 1.9
Ba 10 1 1 8 36 45 3 8 14 32 22 24 26 20
Th n.d. n.d. n.d. n.d. 0.2 n.d. n.d. 0.4 0.6 0.7 1.7 0.3 0.3 0.8
Zr 77.9 84.5 82.7 25.0 29.1 6.7 21.9 78.8 104.0 105.3 95.0 110.4 107.7 67.6
Nb 1.4 1.2 1.9 0.4 0.8 0.1 0.4 1.0 2.0 1.8 1.7 1.7 1.7 0.9
Y 38.8 34.6 34.7 15.6 19.3 8.8 16.7 27.4 35.9 36.0 29.6 32 35.1 35.1
La 2 2 2 0.3 1 1 1 2 3 2 7 3 2 1
Ce 10 9 11 0 4 1 2 9 12 6 20 10 11 5
Nd 12 10 12 2 6 4 3 7 10 8 13 10 11 8
Cu 60 100 64 5 8 32 60 17 35 64 46 18 83 43
Zn 142 97 158 53 90 42 48 69 102 93 99 98 99 121
C1 n.d. 219 161 359 33 n.d. 614 104 n.d. 44 121 n.d. 33 88
Ga 17 17 17 12 14 14 13 18 16 15 17 16 13 19

Oxides as wt%, trace elements in ppm. Data presented on a volatile-free basis; mg = M g [ M g +Fe2+], Mg =MgO/40; Fe203 ×0.9/72; total iron as Fe203 t,~
(Fe203* = Fe203 + FeOxI.III); n.d., below detection limit; LOI, loss on ignition at 1100°C.
270 s.A.K. OMANG• A. J. BARBER

/ L I+gabr°s
Me 1
M°ta"°l°rT

A M
I~ Oceanicgabbro(Kirst1976)~
AmphibolitesfromVema |
FractureZame ]
A M (blonnorezet at. 1984) J

Fig. 4. Metamorphic rocks from Darvel Bay plotted on an AFM diagram compared with gabbros and amphibolites
from the Vema Fracture Zone, equatorial Mid-Atlantic; the tholeiite calc-alkaline boundary is from Irvine & Baragar
(1971).

developed in the felsic body, parallel to the schis-
tosity in the surrounding metabasites. Thin section
study shows that the rock is fine grained with
elongated crystals of quartz and feldspar defining
the schistosity. Minor constituents are acicular
amphibole, epidote, apatite, chlorite, sphene, zircon
and Fe-Ti oxides.
A block (1 m × 1.2 m) of felsic rock, identified
as tonalite (Specimen PK4a), occurs enclosed in
amphibolite on the foreshore on the south coast
of Pulau Sakar. The block probably represents a
xenolith incorporated in a basic intrusion, but now
forms a boudin with a schistosity parallel to that
in the surrounding metabasite. Thin section study
shows that this rock is composed predominantly
of quartz and plagioclase feldspar with subsidiary
actinolitic hornblende. The rock has recrystallized,
with the growth of porphyroblastic feldspar and
hornblende crystals; granophyric texture in the
quartz-feldspar matrix may represent an original
igneous texture or be due to partial melting of the
tonalite. The hornblende porphyroblasts contain
trails of small inclusions of quartz, feldspar and
Ti-oxides defining an internal schistosity. The
porphyroclasts are enclosed in augen structures and
the external schistosity in the matrix cuts across the
internal schistosity of the crystal at a high angle,
indicating that the rocks have been subjected to
multiple deformation (Omang 1993, pl. 5.7D).
Whole rock major element and trace element
analyses of felsic rocks are listed in Table 2, and
are compared to ocean ridge granites (ORG) on
the spider diagram of Fig. 7d. These analytical data
show that although the felsic rocks are not true
granites they are comparable to ORG in their
incompatible element concentrations and resemble
acid rocks fore other ophiolites, interpreted as
co-genetic with basic rocks (Pearce et al. 1984a).
Metatuff
Chloritic schists are exposed at Km 134 north of
Kampong Silam on Jalan Silam, where they are
in contact with sheared serpentinite, and near
Kampong Lok Bikin on the coast of the mainland
opposite Pulau Sakar where they are interbedded
with metacherts. At the locality on Jalan Silam the
schists are folded by small-scale open folds on ENE
(060°-070 °) axes with a low angle of plunge
(10-12 °) and steep axial planes indicating a
southerly vergence.
 D A R V E L BAY O P H I O L I T E , S A B A H 271

100 In thin section the schistosity is defined by fine

grained chlorite flakes and small feldspar laths
80 diverging around larger euhedral to subhedral
ophiolite plagioclase crystals showing albite and Carlsbad
.C..•60 twinning (Fig. 8a).
These schists have the composition of tholeiitic
40 basalts and are therefore interpreted as tuffaceous
.< MORB rocks belonging to the ophiolite sequence, as
2O • hyaloclastic tufts forming part of the ocean floor
Arc assemblage, which has been intensely deformed
! ", !
1
and metamorphosed under greenschist facies
1 2 3 4 conditions.
(a) TiO2 (wt.%)
Metachert

Thin bedded red-brown ribbon cherts occur exten-

80
sively in the northern and southern areas of the
70 Darvel Bay Ophiolite Complex (Fig. 1). In the
60 northern part of the area cherts occur as blocks in
the Kuamut Melange (Clennell 1991; Aitchison
ophiolite 1994). In the southern area bedded cherts folded
'° and imbricated with basaltic pillow lavas, forming
~ 3a an accretionary complex, crop out along Jalan
20 ~ ~ Im MORB Silam.
A • 4-__ Near Kampong Diam, south of Lahad Datu, dis-
10
Arc ¥~I I continuous exposures of greenish-white chert with
o I ! a bedded appearance outcrop along the shore. In
o ~ 2 3 4 thin section these cherts are composed almost
(b) TiO2 (wt.%) entirely of microcrystalline quartz. The quartz
crystals are elongated and have a strong preferred
Fig. 5. (a).TiO2 v. A1203frio2; (b) TiO2 v. CaO/TiO2
plots of Darvel Bay metamorphic rocks; fields from orientation with a mylonitic texture (Fig. 8b). The
Sun & Nesbitt (1978). apparent bedded appearance is a schistosity pro-
duced by deformation. Concordant quartz and
epidotic veins extend along the schistosity which
is cut by calcite veins.

TfflO0
1000 .J.
JDI A MORB: Mid-ocean ridge basalts
Cr PK3a f / ~ IAATB~sla~d_alI~alt~°~e~iatesSalts
(ppm) i/ b

II \~ i
100 I
I
| } I : JD6
/ / \ I + Metad°'eri
e
Zrt / (~ ~o'nAT~AAmphib°lites,~,3
I h
~ IAT
. . . . . ! .i.'t , , ,

10 Y(ppm) 100
(a) (b)

Fig. 6. Tectonic discriminant diagrams. (a) Y/Cr after Pearce et al. (1984b); (b) Zr-Ti/100-Y after Pearce & Cann
(1973).
272 s.A.K. OMANG • A. J. BARBER

lO.00

1°°/ kssllsll
cR
1.0(3 0.00 PS12a

o.1( 1.

0.0 i 0 i ! i i i i ! 0 i a 0.10
Sr K Rb Ba Nb La Ce Nd P Zr Ti Y Sc Cr K Rb Ba Th Nb Ce Zr Y
(a) (a)
10.00 10.OO
Metagabbros
(Cumulates)

1.0£

0,10
OlO] ~ T •
V • ~_~ I
• X JSp I
&, PS12d I

0.01 o D i o o o I o a I u 0011 q i I I I I I I I I I I
Sr K Rb Ba Nb La Ce Nd P Zr Ti Y Sc Cr Sr K Rb Ba Nb La Ce Nd P Zr Ti Y Sc Cr
(b) (e)

Fig. 7. (a)-(c) MORB-normalized chemical plots for metadolerites and metagabbros from the Darvel Bay Ophiolite
listed in Table I; normalized values from Pearce et al. (1984b). (d) ORG-normalized ocean-ridge granite chemical
plot for felsic rocks listed in Table 2; normalized values from Pearce et al. (1984a).

Epidote crystals from the veins analysed by Datu (Reinhard & Wenk 1955). Garnet amphibo-
microprobe are of XpSl8_25. XFe 3+ (0.9) is rela- lites from the Tungku River were analysed during
tively high, corresponding to temperatures of the present study. In thin section the amphibolites
formation of 450-500°C (Nakajima et al. 1977), commonly show mylonitic textures with crystals of
i.e. in the greenschist facies. garnet, pyroxene and hornblende enclosed in augen
These mylonitic quartz schists have been formed structures, enclosed in a fine grained matrix of
by the deformation of bedded cherts in an active hornblende and plagioclase (Fig. 8c).
shear zone and affected by calcium metasomatism P, T estimates from amphiboles and from
in a hydrothermal system, in which the fluids were garnet-pyroxene pairs in the present study (Omang
enriched in calcium carbonate, under greenschist 1993), and an earlier study of garnet pyroxenite
facies metamorphic conditions. from the same locality by Morgan (1974), give the
conditions of formation of these rocks as
T > 850°C at P > 5 kbar.
Garnet amphibolite
Bulk rock geochemistry, rare earth and trace
Fragments of high grade metamorphic rocks element geochemistry of the garnet amphibolites
including garnet pyroxenites and garnet amphibo- show that they are MORB tholeiites and represent
lites occur in volcanic conglomerates in the oceanic crustal materials. These rocks were meta-
Tungku and Pungulupi Rivers 50 km east of Lahad morphosed as pyroxene granulites and garnet
 DARVEL BAY OPHIOLITE, SABAH 273

Table 2. Whole rock geochemical analyses for felsic and sediments. This association of rock types is
rocks from the Darvel Bay Metamorphic Complex characteristic of the ocean floor and its under-
lying mantle. The basic rocks have a tholeiitic
Sample No. LD5 PK4a PS 12c composition and fall predominantly in the MORB
field of discriminant diagrams, showing that they
Sit 2 60.14 61.46 67.27
originated by partial melting of mantle peridotite in
Tit 2 0.438 1.461 0.370
AI203 17.90 15.38 16.60 a mid-ocean ridge environment. Chemical analyses
Fe203* 5.16 9.43 2.89 show depletion in HFS elements, enrichment in
MgO 2.01 2.22 1.06 LIL and LREE, positive Sr and negative Nb and Ce
MnO 0.120 0.120 0.044 anomalies, characteristic of island arc tholeiites,
Cat 8.89 4.77 3.52 indicating that the mid-ocean ridge was generated
Na20 4.73 4.76 7.68 above an active subduction zone.
K20 0.333 0.091 0.044 In map view (Fig. 1), in cross-section (Fig. 9)
P205 0.250 0.308 0.149 and also on an outcrop scale the rocks are
Total 99.98 100.00 99.62
distributed in lenticular slivers elongated parallel to
LOI 4.73 1.30 0.73
mg c. 44 c. 32 c. 42 the E - W trend of the metamorphic belt. The slab of
mantle peridotite forming Silam Hill is juxtaposed
Ni 19 5 5 along the northern shore of Darvel Bay against the
Cr 13 2 2 'Silumpat Gneiss', representing deformed cumulate
V 75 94 26 gabbro, and with amphibolite and hornblende
Sc 10 24 4
Pb 3.4 3.4 0.3 schist, representing isotropic gabbro or doleritic
Sr 537 169 125 and basaltic dykes, from higher levels of the ocean
Rb 7.9 1.2 0.5 crust. On the northern side of Silam Hill mantle
Ba 206 23 188 peridotite is juxtaposed with felsic rocks, repre-
Th 2.9 0.3 2.9 senting K-deficient silicic differentiates from
Zr 273.3 169.6 188.0 tholeiitic magma which were intruded into oceanic
Nb 3.2 3.9 5.5 crust at a high structural level. Chloritic schists at
Y 22.9 53.9 25.1 Jalan Silam and Lok Bikin, identified as deformed
La 19 6 15 and recrystallized volcaniclastic rocks, represent
Ce 40 20 37
Nd 21 17 21 tufts or hyaloclastics, and mylonitic quartz-schists,
Cu 84 4 38 identified as banded cherts, represent the ocean
Zn 56 15 19 floor and its sedimentary cover. All these rocks
C1 8 35 179 are now at the same structural level as the mantle
Ga 22 19 13 peridotite.
Most of the metamorphic rocks in the belt are
Oxides in wt%, trace elements in ppm. Total iron as foliated, sometimes intensely, to form schists, and
Fe203(FeeO3 + FeOxI.III); mg = [100Mg/(Mg + Fe)], frequently show mineral lineations. The foliation or
where Mg = MgO/40 and Fe = F203 x 0.9/72. Data schistosity is generally steeply inclined or vertical
presented on a volatile-free basis; LOI, loss on ignition
at 1100°C. and the lineation is subhorizontal. Both foliation
and lineation are orientated parallel to the E - W
trend of the belt. Preferred orientation of minerals
in the gneisses and schists defining foliation, schis-
amphibolites at temperatures and pressures charac-
tosity and mineral lineation, and rotated inclusion
teristic of the upper mantle and were deformed
trails in hornblende porphyroblasts in the tonalite
and recrystallized with mylonitic textures in the
from Pulau Sakar, indicate that the metamorphic
amphibolite facies. These features are consistent
rocks recrystallized syntectonically. Compositions
with deformation and recrystallization of oceanic
of amphiboles and amphibole/plagioclase pairs
crustal rocks carried down in a subduction zone.
indicate that this recrystallization occurred under
high temperature but low pressure conditions in the
Origin of metamorphic rocks in the amphibolite and greenschist metamorphic facies.
Darvel Bay Ophiolite Complex No clear metamorphic gradation across the belt was
recognized in our study, different facies being
Protoliths of rocks which form the belt of meta- randomly juxtaposed (cf. Hutchison 1975).
morphic rocks in the Darvel Bay Ophiolite Basaltic dykes and felsic veins cutting the meta-
Complex extending westwards from Lahad Datu morphic rocks show varying relationships to the
can be identified as mantle peridotites, cumulate foliation and schistosity. Dykes and veins may be
pyroxenites and gabbros, isotropic gabbros, plagio- concordant to the foliation or schistosity in the
granites, doleritic and basaltic dykes, volcanics country rocks, or may cut across these structures
 t~

(a) ~)

1 mm
4

Fig.8. (a) Photomicrograph of chlorite schist, deformed crystal tuff, Jalan

Silam Km 134, Darvel Bay, Sabah. (b) Photomicrograph of siliceous schist
(metachert) with epidote bands and cross-cutting quartz veins, Kampong
Diam, Lahad Datu, Sabah. (e) Photomicrograph of garnet amphibolite
(Specimen EKc) showing a fractured garnet crystal with alteration to
hornblende enclosed in a fine grained mylonitic hornblende-plagioclase
matrix to form an augen structure. Pebble from Tungku River, Dent
Peninsula, Sabah. Scale bar 1 mm.
(c)
 DARVEL BAY OPHIOLITE, SABAH 275

I SECTION A-A' ]

Bukit Silam
South
3000
3000
2000 Feet
2000
1000 1000
0
0

Distance A-A'
~ 35 Km [

] SECTION B - B' I
South Bukit Silam North
3000 " ~ ~ ~ _ SilamDam 3000 Feet
2000- Silam Road ~ ~ Sg. Telewas ~ e . . . . . 2000
1000" ~ 1000
0
Distance B-B'
- 29 Km

I SECTION C-C' I SECTION D - D'[

South Kampong North South dolerite dyke North _

i o
1,,,~ P. Gifford Silam Kg. Sepagaya Pulau Sakar k, v~ Veer
tmo I000 L~.~" 1000
0 0 Feet 0

[ Distance C-C': - 17 Km] D-D': ~ 17 Kml

I Distance D-I3
South [ SECTION E" E'i North
10IN) 1000 Feet
0 0 Sea-level

[ Distance E-E' : ~ 11 Km I

Fig. 9. Schematic cross-sections across the metamorphic belt in the Darvel Bay Ophiolite Complex Sabah. Lines of
section and key to ornaments are shown on Fig. 1.

with varying degrees of obliquity. Hutchison &
D h o n a u (1971) reported a xenolith o f ' S i l u m p a t
Gneiss' (i.e. foliated banded gabbro) e n c l o s e d in a
basaltic 'sill' on Pulau Silumpat. Within the dykes,
schistosity m a y be d e v e l o p e d with varying degrees
of intensity; in general the m o r e concordant the
d y k e the m o r e intense the schistosity within it.
These features indicate that s o m e of the dykes
were intruded into rocks w h i c h had already been
d e f o r m e d , but that d e f o r m a t i o n c o n t i n u e d after
d y k e e m p l a c e m e n t . D e f o r m a t i o n o f the felsic veins
is indicated by b o u d i n a g e where they are con-
cordant, and folding w h e r e they are cross-cutting.
B o u d i n a g e d and folded veins are frequently cut
276 s.A.K. O M A N G • A. J. BARBER
by veins which are completely undeformed. The
relationships of basaltic dykes and felsic veins
to the foliation and schistosity demonstrate that
deformation and intrusion of igneous rocks were
going on continuously during the development
of the metamorphic belt.
Metamorphic rocks associated with ophiolite
complexes from many parts of the world have been
attributed either to processes of deformation and
metasomatism which affect ocean floor materials
shortly after their formation in the mid-ocean ridge
environment, or to the processes of subduction
which lead to the emplacement of the ophiolite
on a continental margin. It has been suggested that
foliated rocks may be formed at the crust-mantle
boundary in the region of a mid-ocean ridge by
gravitational spreading of the ridge (Smewing et al.
1984; Gibbons & Thompson 1991). Foliated meta-
morphic rocks within ophiolites have also been
interpreted as having originated along transform
fault zones (Karson & Dewey 1978; Saleeby 1978;
Simonian & Gass 1978; Prinzhofer & Nicholas
1980; Karson 1984). Foliated metamorphic rocks
which occur along the basal thrust of an ophiolite
complex have been interpreted as a 'metamorphic
sole' formed during subduction and emplacement
of the ophiolite on a continent (Davies 1971;
Williams & Smyth 1973; Jamieson 1980; Searle &
Malpas 1980; Spray & Williams 1980; Ghent &
Stout 1981; Moores 1982).
On the basis of the characteristics detailed above,
the rocks of the metamorphic belt within the Darvel
Bay Ophiolite are interpreted as having formed
along a transform fault zone. Throughout the
oceans, mid-ocean ridge systems are offset along
transform faults at intervals of a few hundred kilo-
metres. In these zones segments of ocean crust
are moving past each other between active ridge
segments. As the two crustal slices move past each
other, oceanic materials along the fault are
deformed, producing foliated peridotites, banded
gneisses from cumulate gabbros and foliated and
schistose amphibolites from gabbros and sheeted
dykes. In transform fault zones at the time of
their formation the foliation surfaces will be set
vertically, parallel to the trend of the fault zone and
lineation will be sub-horizontal representing the
direction of movement along the fault. In an active
fault zone crustal slices on either side of the fault
will be of different ages. On one side of the fault the
crust will have been formed recently at the ridge
axis and will be at a high temperature; on the other
the crust will be older and will have cooled and
subsided as it moved away from the ridge axis.
Deformation and shearing of the oceanic crustal
rocks along the fault, with the ingress of water, will
induce recrystallization of hydrous phases which
are aligned to form schistose and linear structures
parallel to the sense of movement within the fault
zone.
At the time of formation the grade of meta-
morphism will decrease across the fault zone, from
pyroxene and hornblende granulite against the
hotter younger crustal segment, declining to
greenschists and unmetamorphosed rocks towards
the older colder segment and vertically towards the
ocean floor. As has already been stated, no simple
gradation of metamorphic facies has been recog-
nized in the metamorphic belt at Darvel Bay.
However, in an active fault zone any such simple
arrangement is likely to be disrupted by continual
movement along the fault, juxtaposing unrelated
fault slices of different metamorphic grades and
from different crustal levels, as has been described
from Darvel Bay. No sedimentary infills, com-
posed of clastic debris flows formed by submarine
erosion, as reported from the Arakapas Transform
in Cyprus (Simonian & Gass 1978; McLeod &
Murton 1993) have been recognised as associated
with the Darvel Bay Transform, suggesting that
several kilometres have been removed by erosion
since the emplacement of the ophiolite on Sabah.
The mylonitic quartz-schists described from
Kampong Diam, and identified as deformed
banded cherts, are an unexpected component of
an assemblage of rocks incorporated in a fossil
transform fault zone. Cherts are deposited on the
ocean floor only after it has subsided beneath
carbonate compensation depth, some millions of
years after the ocean crust formed at the mid-ocean
ridge. The presence of chert among the deformed
rocks indicates that a much older segment of ocean
crust was juxtaposed against an active spreading
ridge, implying that the transform fault extended
over several hundreds of kilometres. Karson &
Dewey (1978) came to a similar conclusion
regarding the Coastal Complex transform associ-
ated with the Bay of Islands Ophiolite in
Newfoundland, and pointed to present day
examples where very long transform faults separate
short ridge segments, in the Gulf of California and
the Andaman Sea. Continued movements over a
long distance through transtensional and trans-
pressional fault segments increases the likelihood
of the juxtaposition of uplifted mantle peridotite,
with downfaulted crustal materials including
supracrustal volcanics and metasediments, as seen
in Silam Hill and along Jalan Silam.
The garnet pyroxenites and garnet amphibolites
from the Tungku and Pungulupi rivers described
earlier in this account are not compatible with a
transform fault origin. As previously concluded the
high pressures and temperatures of formation of
these rocks, combined with their mylonitic textures
are consistent with their formation in a subduction
zone. These rocks are therefore considered to
 DARVELBAY OPHIOLITE, SABAH
represent fragments of a metamorphic sole which
must be present at depth below the Darvel Bay
Ophiolite of the Dent Peninsula. A K-Ar age of
7 6 + 21 Ma obtained from garnet amphibolite
during the present study coincides with the Late
Cretaceous-Palaeogene age of subduction beneath
the Darvel Bay Ophiolite inferred from the
stratigraphic evidence.
These metamorphic sole rocks, which must
underlie the whole of the ophiolite complex, were
intersected by Miocene volcanics and carried to
the surface, to be incorporated with andesitic frag-
ments, ophiolitic rocks and sandstones in volcanic
breccias and conglomerates during the formation
of the Sulu Volcanic Arc.
Tectonic setting and emplacement of the
Darvel Bay Ophiolite
Evidence given in this account confirms the
interpretation of the Darvel Bay Ophiolite Complex
as a segment of oceanic crust and upper mantle
originating at a mid-ocean spreading ridge.
Evidence has also been given that the spreading
ridge was developed above a subduction zone in
a back-arc basin. The belt of metamorphic rocks,
which occurs within the complex and extends
westwards from Lahad Datu, is interpreted as a
fossil transform fault.
Radiometric ages and biostratigraphic ages from
radiolarian cherts and foraminiferal pelagic lime-
stones in the Chert-Spilite Formation indicate a
Lower Cretaceous age for the origin of the
ophiolite. Shallow-water limestones with Upper
Cretaceous fossils, interpreted as the carbonate
cappings to seamounts, indicate that the ophiolite
still formed part of the ocean floor at this time.
Hamilton (1979), Holloway (1981) and Rangin
et al. (1990) have suggested that the Darvel Bay
Ophiolite, together with other ophiolite fragments
in eastern Sabah, originally formed part of a Proto-
South China Sea crust. In this scenario, southward
References
AITCHISON, J. C. 1994. Early Cretaceous (pre-Albian)
radiolarians from blocks in Ayer Complex melange,
eastern Sabah, Malaysia, with comments on their
regional tectonic significance and the origins of
enveloping melanges. Journal of Southeast Asian
Earth Sciences, 9, 255-262.
BROWN,E. H. 1977. The crossite content of Ca-amphibole
as a guide to pressure of metamorphism. Journal of
Petrology, 18, 53-72.
CLENNELL, M. B. 1991. The origin and tectonic signifi-
cance of mrlanges in Eastern Sabah, Malaysia.
Journal of Southeast Asian Earth Sciences, 6,
407--429.
277
subduction of Proto-South China Sea crust led to
the obduction and uplift of the ophiolite by under-
thrusting of Upper Cretaceous and Lower
Palaeogene turbidites of the Crocker Formation.
Hutchison (1988), from the occurrence of
Oligocene granitoid intrusions in the Long Laai
area SW of Darvel Bay, suggests that the sub-
duction of continental crust may also have been
involved. Fragments of ophiolitic rocks in Eocene
conglomerates show that the ophiolite had been
obducted, uplifted and was subject to erosion by
Eocene times (Newton-Smith 1967).
It has been demonstrated in the foregoing
account that the major occurrence of metamorphic
rocks associated with the Darvel Bay Ophiolite
represents a fossil transform fault, while fragments
of high pressure metamorphic rocks in overlying
volcanics indicate that the ophiolite is underlain
by a metamorphic sole related to subduction and
the obduction of the ophiolite in Sabah.
The work described in this paper was presented as a
thesis by SAKO for the award of the PhD degree of
the University of London, sponsored by the Universiti
Kebangsaan Malaysia (UKM) and the Government of
Malaysia. Bulk rock geochemistry was determined in the
geochemical laboratories at Royal Holloway, University
of London using a Philips PW1480 XRF Spectrometer
under the supervision of Drs M F Thirlwall and G F
Marriner and Mr Ceil Jenkins. Mineral analyses were car-
ried out on a JEOL Superprobe 733 electron microprobe
at Birkbeck College under the supervision of Prof Robert
Hall, University College and mineral compositions were
recalculated using a suite of programs developed by Prof
Hall who also gave general advice on the interpretation of
the chemical data. K-Ar isotopic dating was carried out at
the NERC Isotope Geosciences Laboratory at Keyworth,
Nottingham under the supervision of Dr C. C. Rundle.
Mr David Lee, Director of the Geological Survey of
Malaysia, Kota Kinabalu, provided the samples of garnet
amphibolite from the Tungku River used in this study.
Attendance at the Conference on the 'Tectonic Evolution
of Southeast Asia' held at the Geological Society, London
was sponsored by UKM. Presentation has been greatly
improved by careful and constructive reviews by Prof.
C. S. Hutchison and Dr J. E. Dixon.
DAVIES,H. L. 1971. Peridotite-gabbro-basalt complex in
eastern Papua: an overthrust plate of oceanic mantle
and crust. Bulletin of the Australasian Bureau of
Mineral Resources, Geology and Geophysics, 128.
DEER, W. A., HOWIE, R. A. & ZUSSMAN, J. 1966. An
Introduction to the Rock-forming Minerals.
Longman, London.
DHONAU,T. J. & HUTCHISON,C. S. 1966. The Darvel Bay
area, east Sabah, Malaysia. In: Geological Survey of
Malaysia, Borneo Region, Annual Report for 1965.
141-160.
FITCH, E H. 1955. The geology and mineral resources of
part of the Segama Valley and Darvel Bay area,
 278 s . A . K . OMANG • A. J. BARBER
colony of North Borneo. Geological Survey,
Department of British Territories in Borneo,
Memoir, 4.
FLEET, M. E. & BARNETT, R. L. 1978. AliV/A1vi
partitioning in calciferous amphiboles from the
Frood Mine, Sudbury, Ontario. Canadian
Mineralogist, 16, 527-532.
GHENT, E. H. & STOUT,M. Z. 1981. Metamorphism at the
base of the Semail Ophiolite, southeastern Oman
Mountains. Journal of Geophysical Research, 86,
2557-2571.
GmBONS, W. & THOMPSON,L. 1991. Ophiolitic mylonites
in the Lizard Complex: ductile extension in the
lower crust. Geology, 19, 1009-1012.
HAMILTON,W. 1979. Tectonics of the Indonesian Region.
United States Geological Survey Professional
Paper, 1078.
HOLLOWAY, N. H. 1981. The North Palawan Block,
Philippines; its relation to the Asian mainland
and its role in the evolution of the South China
Sea. Geological Society of Malaysia Bulletin, 14,
19-58.
HONNOREZ, J., MEVEL, C. & MONTIGNY, R. 1984.
Occurrence and significance of gneissic amphibo-
lites in the Vernh Fracture Zone, equatorial Mid-
Atlantic Ridge. In: GASS, I. G., LIPPARD, S. J. &
SHELTON, A. W. (eds) Ophiolites and Oceanic
Lithosphere. Geological Society, London, Special
Publications, 16, 77-94.
HLrrCHISON, C. S. 1975. Ophiolite in Southeast Asia.
Geological Society of America, Bulletin, 86,
797-806.
- 1978. Ophiolite metamorphism in northeast Borneo.
- MOORES, E. M. 1982. Origin and emplacement of ophio-
Lithos, 11, 195-208.
1988. Stratigraphic-tectonic model for Eastern
Borneo. Geological Society of Malaysia Bulletin,
22, 135-151.
- & DHONAU, T. J. 1969. Deformation of an alpine
- Contributions to Mineralogy and Petrology, 48,
ultramafic association in Darvel Bay, East Sabah,
Malaysia. Geologie en Mijnbouw, 48, 481-494.
- & --
- 1971. An alpine association of metabasites
and ultrabasic rocks in Darvel Bay, East Sabah,
Borneo. Overseas Geology and Mineral Resources,
Institute of Geological Sciences, Report, 10,
289-308.
IRVINE, T. N. & BARAGAR,W. R. A. 1971. A guide to the
chemical classification of the common igneous
rocks. Canadian Journal of Earth Science, 8,
523-548.
JAMIESON, R. A. 1980. Formation of metamorphic
aureoles beneath ophiolites - evidence from the
St.Anthony Complex, Newfoundland. Geology, 8,
150-154.
- 1981. Metamorphism during ophiolite emplacement
- analysis. Earth and Planetary Science Letters, 19,
- t h e petrology of the St. Anthony Complex.
Journal of Petrology, 22, 397-443.
KARSON, J. A. 1984. Variations in structure and petrology
in the Coastal Complex, Newfoundland: anatomy of
an oceanic fracture zone. In: GASS, I. G., LIPPART,
S. J. ~¢ SHELTON, A. W. (eds) Ophiolites and
Oceanic Lithosphere. Geological Society, London,
Special Publication, 13, 131-143.
- & DEWEY, J. F. 1978. Coastal Complex, western
- M. E (eds) Marginal Basin Geology: Volcanic
Newfoundland: An Early Ordovician oceanic
fracture zone. Geological Society of America,
Bulletin, 89, 1037-1049.
KIRST, P. 1976. Petrology of the metamorphic rocksfrom
the equatorial Mid-Atlantic Ridge and Fracture
Zone. PhD Thesis, University of Miami.
KOOPMANS,B. N. 1967. Deformation of the metamorphic
rocks and Chert-Spilite Formation in the southern
part of the Darvel Bay area, Sabah. Geological
Survey of Malaysia, Borneo Region, Bulletin, 8,
14-24.
LEON6, K. M. 1971. Peridotite gabbro problems, with
special reference to the Segama Valley and Darvel
Bay area, Sabah, East Malaysia. Geological Society
of Malaysia, Newsletter, 28, 4-13.
-- 1974. The geology and mineral resources of the
Darvel Bay and Upper Segama area, Sabah.
Geological Survey of Malaysia, Memoir, 4
(revised).
- 1977. New ages from radiolarian cherts of the
-
Chert-Spilite Formation of Sabah. Bulletin of the
Geological Society of Malaysia, 8, 109-111.
MCLEOD, C. J. & MLrRTON, B. J. 1993. Structure and
tectonic evolution of the southern Troodos
Transform Fault Zone, Cyprus. In: PRITCHARD,
H. M., ALABASTER,T., HARRIS, N. B. W. & NEARY,
C. R. (eds) Magmatic Processes and Plate
Tectonics. Geological Society, London, Special
Publication, 76, 141-176.
MITCHELL, A. H. G., HERNANDEZ, E • DE LA CRUZ,
A. E 1986 Cenozoic evolution of the Philippine
Archipelago. Journal of Southeast Asian Earth
Sciences, 1, 3-22
lites. Review of Geophysics and Space Physics, 20,
735-760.
MORGAN, B. A. 1974. Chemistry and mineralogy of
garnet pyroxenites from Sabah, Malaysia.
301-314.
NAKAJIMa, T., BANNO, S. & SUZUKI, T. 1977. Reaction
leading to the disappearance of pumpellyite in low
grade metamorphic rocks of the Sanbagawa meta-
morphic belt in central Shikoku, Japan. Journal of
Petrology, 24, 263-284.
NEWtON-SMITH, J. 1967. Bidu-Bidu Hills, area, Sabah,
East Malaysia. Geological Survey of Malaysia,
Kuching, Report, 4.
OMANG, S. A. K. S. 1993. Petrology, geochemistry and
structural geology of the Darvel Bay Ophiolite,
Sabah, Malaysia. PhD Thesis, University of
London.
PEARCE, J. A. & CANN,J. R. 1973. Tectonic setting of
basic volcanic rocks determined using trace element
290-299.
, HARRIS, N. B. W. & TINDLE, A. G. 1984a. Trace
element discrimination diagrams for the tectonic
interpretation of granitic rocks. Journal of
Petrology, 25, 956-983.
- - , LIPPARD,S. J. & ROBERTS,S. 1984b. Characteristics
and tectonic significance of supra-subduction zone
ophiolites. In: KOKELAAR, B. P. t~ HOWELLS,
and Associated Sedimentary and Tectonic Processes
 DARVEL BAY OPHIOLITE, SABAH
in Modern and Ancient Marginal Basins.
Geological Society, London, Special Publication,
16, 77-94.
PLYUSNINA, L. E 1982. Geothermometry and geo-
barometry of plagioclase-hornblende bearing
assemblages. Contributions to Mineralogy and
Petrology, 69, 33--47.
PRINZHOFER, A. & NICHOLAS, A. 1980. The Bogota
Peninsula: a possible oceanic transform fault.
Journal of Geology, 88, 387-398.
RANGIN, C. 1989. The Sulu Sea, a back-arc basin setting
within a Neogene collision zone. Tectonophysics,
161, 119-141.
~, BELLON, H., BERNARD, E, LETOUZEY,J., MULLER,
C. & SANUDtY, T. 1990. Neogene arc-continent
collision in Sabah, Northern Borneo (Malaysia).
Tectonophysics, 183, 305-319.
REINHARD, M. & WENK, E, 1955. Geology of the Colony
of North Borneo. Geological Survey, Department
of the British Territories in Borneo, Bulletin, 1.
SALEEBY,J. 1978. Kings River ophiolite, southwest Sierra
Nevada foothills, California. Geological Society of
America, Bulletin, 89, 617-636.
SEARLE, M. P. & MALPAS,J. G. 1980. Structure and meta-
morphism of rocks beneath the Semail Ophiolite of
Oman and their significance in ophiolite obduction.
Transactions of the Royal Society of Edinburgh, 71,
247-262.
279
SIMONIAN, K. O. & GASS, I. G. 1978. Arakapas fault belt,
Cyprus: A fossil transform fault. Geological Society
of America, Bulletin, 89, 1220-1230.
SMEWING,J. D., CHRISTENSEN,N. I., BARTHOLOMEW,1. D.
8,: BROWNING, P. 1984. The structure of the oceanic
upper mantle and lower crust as deduced from
the northern section of the Oman ophiolite. In:
GASS, I. G., LIPPARD, S. J. & SHELTON, A. W.
(eds) Ophiolites and Oceanic Lithosphere.
Geological Society, London, Special Publication,
13, 41-53.
SPEAR, E S. 1981. An experimental study of hornblende
stability and composition variability in amphibole.
American Journal of Sciences, 281, 697-734.
SPRAY, J. G. • WILLIAMS,G. D. 1980. The sub-ophiolitic
rocks of the Ballantrae Igneous Complex, SW
Scotland. Journal of the Geological Society,
London, 137, 359-362.
SUN, S-S. & NESBITT, R. W. 1978. Geochemical
regularities and genetic significance of ophiolitic
basalts. Earth and Planetary Science Letters, 44,
119-138.
WILLIAMS,H. & SMYTH,R. 1973. Metamorphic aureoles
beneath ophiolite suites and Alpine peridotites:
tectonic implications with west Newfoundland
examples. American Journal of Sciences, 273,
594-621.
 Role of pre-Tertiary fractures in formation and development
of the Malay and Penyu basins
KHALID NGAH, MAZLAN MADON & H. D. T J I A
PETRONAS Research & Scientific Services, Lot 1026 PKNS Industrial Estate,
54200 Hulu Kelang, Malaysia

Abstract: Major faults in Sundaland trend NNW to NW, WNW, N and E. Some of the NNW
to NW and N-striking faults across Mesozoic areas of the Malay Peninsula were active until mid-
Eocene time. Small, fault-bounded Tertiary basins onshore may be pull-apart basins associated
with such faults. Mainly from seismic data, NNW to NW, N and E-striking faults have been
recognized in the pre-Tertiary basement of the Malay and Penyu basins off the east coast of the
peninsula. These faults were reactivated before the Late Oligocene and during the Middle to Late
Miocene. N-striking faults in pre-Tertiary areas are common throughout Sundaland. In the field,
these faults are found to be the oldest (possibly Jurassic) regional fractures. The regional
NNW-NW and WNW fractures are believed to have originated as strike-slip faults when the
peninsula was subjected to late Mesozoic deformation. The onshore E-W faults were probably
extensional fractures that developed as secondary structures associated with sinistral slip motions
along NNW-NW faults. Upper Cretaceous dolerite dykes fill some of the E-W fractures. NW-
striking basement faults of the Malay basin continue onshore SE Asia as the Three Pagodas fault
zone. Initially these were sinistral basement wrench faults creating secondary E-W extensional
fractures. In the Middle to Late Miocene the regional stress field changed, resulting in reversal of
slip movement along major wrench faults and structural inversion of the sedimentary basins. This
inversion is manifested as E-W anticlines located over half-graben. In the Penyu basin similarly
striking half-graben probably developed in the same fashion. There, the NW-striking Rumbia
fault divides the basin into two parts. Half-graben in the western part remained orientated E-W,
but those in the eastern part became rotated clockwise by continued left-lateral slip along the
Rumbia fault. After the Miocene the two basins continued to subside, developing an almost
undisturbed blanket of post-Miocene sediments. Locally, residual stress caused some of the
structures to grow.

The Malay and Penyu basins are Cenozoic sedi-
mentary basins located in the South China Sea and
are floored by continental lithosphere (Fig. 1). This
southern portion of the South China Sea belongs
to tectonically semi-cratonic Sundaland.
Geologically, Sundaland comprises the entire
Sunda Shelf and contiguous land areas of Borneo,
Java, Sumatra and peninsular Malaysia. The
margins of Sundaland have experienced young
crustal movements and are tectonically transitional
to the mobile belts bordering the region in the east,
south and west. The Malay basin is 500 km long
in a northwest direction, about 200 km wide, and
contains more than 12 km of Oligocene and
younger sediments. The older sediments are of
terrestrial origin with minor marine intervals; holo-
marine conditions have prevailed only since the
latest Miocene. Oil-prone areas are situated in the
southern and central parts of the Malay basin, while
the northern part is gas-prone. Hydrocarbon
prospects have also been located in the Penyu basin
and non-commercial oil discoveries have been
made. This basin has a roughly sub-circular shape
with a diameter of about 350 km and contains up
to 6 km of Cenozoic sediments that have been
interpreted to have been deposited in an environ-
ment generally similar to those of the Malay basin.
The N W to N N W elongation of the Malay basin
is roughly parallel to the elongation of the Malay
Peninsula to its west, while the more equant shape
of the Penyu basin suggests its shape to be strongly
controlled by fractures. The structural history of
these two basins comprised a tensional period
when the basins began subsiding before the Late
Oligocene. This subsidence continued during the
late Early Miocene to Late Miocene, accompanied
by regional compression during the Middle to Late
Miocene, and finally quiet tectonic subsidence in
the Plio-Pleistocene. These generalized events have
been described by Khalid Ngah (1975), Ng (1987)
and Md. Nazri Ramli (1988). Figure 2 is a corre-
lation table of the stratigraphy in the Malay, Penyu
and the West Natuna basins simplified from various
unpublished and published reports.
The oldest known sediments within the Malay
basin are Upper Oligocene. Their ages were deter-
m i n e d from pollen from cuttings from wells
located in the basin's border zones, while their

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, 281
Geological Society Special Publication No. 106, pp. 281-289.
282 K. NGAH ET AL.

VIETNAM
BASEMENT
NORTH
SUNDALAND

CON SON SWELL

• ~iiiiiiii
TERENGGANU~
1 2 3 4 5

900 O~iiiiiiiiiii!! BASIN

1200
I!i¸ iiiiiiiiiiiiiii
iiii!!;iililii!iiii!i!ii?' iiii:iii
i i i i~ii!i~i'!iiii~i
;:: i . ? : : :~::il~'ltl i ! ! : :: ! F-~.UkSIN IA!

SOUTH CHINA SEA

...... PENINSULAR ::
MALAYSIA i: : i DRM

Fig. 1. Outline of the Cenozoic Malay, Penyu and West Natuna basins of northern Sundaland, SE Asia. Pre-Tertiary
basement depths are shown in two-way travel time, for which an approximate conversion curve is given in the inset.
Heavy lines are major faults.

interpreted equivalents in the deeper part of the et al. (1992) believe that such younger intrusive
basin were extrapolated using seismic profiles. rocks may well exist offshore but have yet to be
Seismic profiles across the basins suggest that a discovered.
thick sequence of older, possibly mainly Cenozoic, Heatflow studies in the Malay basin and other
sediments still exists beneath the interpreted Upper well data from the West Natuna basin show above
Oligocene strata. normal values with certain areas possessing high
Felsic granitic rocks in the Natuna, Anambas and values over 140 mW/m 2 (Mohd Firdaus Abdul
Tambelan islands yield Upper Cretaceous radio- Halim 1993; Wan Ismail Wan Yusoff & Swarbrick
metric dates ranging between 90 Ma and 70 Ma 1994). One area of high heatflow coincides with
(Katili 1973). The ages of granitic (granites to the triple junction of the Malay, Penyu and West
granodiorites; Areshev et al. 1992) basement of Natuna basins. Pre-Tertiary basement is topo-
the Mekong basin and the adjacent shelf range graphically high in a wide region around the triple
from 9 7 M a to 178Ma, and most are around junction and has been postulated to represent the
100-110 Ma or Albian. This basement is also part remnants of a Malay Dome (new name; Tjia 1995).
of Sundaland. Upper Cretaceous-lower Palaeogene It has also been suggested that the centre of the
gramtes have been determined onshore in the Bana, dome was situated above a mantle plume. The
Ankroet and Deo-Ka complexes, and Areshev widespread occurrence of Upper Cretaceous felsic
 FORMATION AND DEVELOPMENT OF MALAY & PENYU BASINS 283

MALAY BASIN 1
AGE PENYU BASIN i WEST NATUNA
(Approximate) PULAI NO, 1 T CONOCO i PETRONAS I, BASIN
EPMI CARIGALI
I VIII
i
RECENT - [ PILONG i PILONG ! VII MUDA
PLIOCENE and " PILONG
Vl

MIOCENE .... J
J
UPPER PARI
J ! BEKOK [ SANDCOAL i V
IV
i BARAT
,
l TAPIS ~--LOWER SAND-COAL
K i --- [- TERENG~:3ANU liB TERENGGANU i UDANG
PULAI _-SHALE
................................... L GABUS •
TAPIS IIA
L [ SELIGI l PENYU
OLIGOCENE
u _i
J
N
i
i

Fig. 2. Stratigraphic correlation table of the Malay, Penyu and West Natuna basins from published and unpublished
data. EPMI; Esso Production Malaysia Inc.; Conoco, Continental Oil Company. Unpublished material is held by
Petronas.

granites and their regional distribution suggest that
the Malay Dome has a radius of c. 500 km.
The present paper addresses the role that pre-
Tertiary structures have played in the formation,
delineation, and development of the Malay and
Penyu basins.
Fracture patterns in the pre-Cenozoic
basement
Figure 3 shows the fracture patterns onshore in
peninsular Malaysia and offshore in the Malay and
Penyu basins. On land in peninsular Malaysia, the
north-south trending Bentong (-Raub) suture is the
limit to a western region of Gondwana origin. This
Gondwana fragment became attached to the eastern
portion of the peninsula by Early Triassic time
(Metcalfe 1988; Tjia 1989a). The eastern portion
of the peninsula was already part of the Asian
continent when the western region docked with it
along the Bentong suture. The pattern of detectable
strike-slip motions on the major onshore faults
indicates that the latest motions were in response to
regional compression with a maximum horizontal
stress orientated between ENE-WSW and E-W.
This is roughly perpendicular to the strike of
regional structures onshore. Approximately
N-NNE-striking faults (such as the Kelau fault in
Pahang, vertical faults within the Bentong Suture
zone, and two unnamed north-south faults on
Pinang Island) moved right-laterally. In contrast,
the NNW, NW and WNW-trending faults are left-
lateral strike-slip faults. These include the Lebir
fault in Kelantan; the Lepar fault, Pahang; the
Baubak fault, Kedah; the Bukit Tinggi and Kuala
Lumpur fault zones in Selangor/Perak; and the
Mersing fault bundle in Johor. Details of these
faults have been discussed elsewhere (Tjia 1989b).
Several small Tertiary basins in the peninsula are
located close to or within large strike-slip fault
zones. The Lawin basin in Perak seems to be
associated with the Baubak (older name: Bok Bak;
Burton 1965) fault zone and after detailed studies,
Syed Sheikh Almashoor (pets. comm. 1994) has
placed the basin within the fault zone. Another
basin, the Batu Arang basin in Selangor, is tangen-
tial to the Kuala Lumpur fault zone. Pollen studied
by PRSS biostratigraphers indicate Middle Eocene
basal beds in this basin. A recent study of major
peninsula faults by Zaiton Harun (1992) showed
that cataclastic to mylonitized material in certain
fault zones is Upper Cretaceous to Middle Eocene
(53-46 Ma). The regional map prepared by the
Geological Survey of Malaysia (1985) shows
uppermost Cretaceous (67 Ma) granites at the
Kelantan-Perak border, at Gunung Ledang (also
known as Mount Ophir), Johor and a few smaller
felsic intrusive bodies in the central belt of the
peninsula and in southern Johor. The evidence cited
284 K. NGAH ET AL.

$ I
102 105
ci
:!ii~:, .

Plnang
Island

FAULT PATTERNS l!'i ~'!' II "~,! ii'!!li[ ' ,

'~ .", ' ~ I ', ,i~ ' ' ~i ", , ' , iI
, ~' i '

Basement contours in km below sea level

loo lol lO2 ~ ~ ,= !~: ,li,,l~,~i~ lo3,, II1]~~11 L,h ~'.....
,04 ~rzo,'v~ lo5
I J , :,llii::Hi [~I;U,I~Ui!;U~ ] 1,,, ~ 1
Fig. 3. Fault patterns in the pre-Cenozoic basement of peninsular Malaysia and offshore in the Malay and Penyu
basins. Faults onshore are from Tjia (1989b); those in the offshore basins were derived from unpublished reports held
by Petronas. The dashed lines within the Malay basin represent major basement fractures which have been inferred
from structural styles in sediments overlying the basement and from trends of magnetic/gravity anomalies.

above strongly suggests that a regional diastrophic dominant. NNW and NW-trending fractures are
event took place in the region at about the evident along the borders of the Malay basin, with
Cretaceous-Tertiary boundary. At that time the some NE-trending fractures in its southeast corner.
entire peninsula, and most probably the entire Figure 3 shows that strikes of major fractures in
area of Sundaland to which it geologically belongs, the offshore regions mimic those mapped onshore
was already a cratonized basement. Upper the peninsula, and the authors conclude that the
Jurassic-Cretaceous sediments (Gagau Group) offshore fault strikes are of pre-Tertiary origin.
were not folded but were only tilted ten degrees
or less, and regional compressive stresses became
decoupled along major fractures. These major Major structures in the Malay
fractures moved mainly as strike-slip faults.
The fractures in the offshore area shown on and Penyu basins
Fig. 3 were compiled from various unpublished
sources held by PETRONAS. Major basement
Malay basin and vicinity
fractures within the Malay basin have been inferred Folds in the Malay basin generally trend east-west.
from structural styles in sediments overlying the Exceptions are along its southwest margin and in
basement and from trends of magnetic/gravity its northwest corner where anticlines trend NW-SE
anomalies. In general, different parts of the offshore or almost N-S, respectively. These strikes seem
area display different dominant fracture directions. associated with the NW-striking Western Hinge-
In the Penyu basin, and in the south end and axial line fault zone (WHL), and the north-striking
parts of the Malay basin, the east-west and NW-SE Kapal-Bergading tectonic line, respectively (Fig.
fractures are dominant, whereas in the north end 4). Known major fault zones are the WHL that
of the Malay basin, north-south fractures are marks the southwest boundary of the basin; the
 FORMATION AND DEVELOPMENT OF MALAY • PENYU BASINS 285

i I I

"%

o,.-' • %.
.. Ser gadlng ~,,, ~ESAH
fz 7"__
::...... ~'OO~,'ff6" ........
iI ° " ~% . ,outt zo~k ~uNm

nln )n,.,L • L.arut "'

1
i ~, ...... 4/,'%_
~ ........ ~ I:['~~i,, ~ wm_,, I
! ~,\ n,~,,~, :. I ~=_ b ~ /
......... ~ cC,,~. ! I W..W'... . . . . . . . . . . . . ~Pulai

; "%~, "q) . . . . . . . .. . . . . . . .
l\ i ~ .;/l'~- Angsi ~ i
............... !
/

gun ! Dungun i X ~ ' AA,~

n d idni ¢ng /:
+ o+.° +
++ v'ooL
iong ]
~1~ Hydrocarbonfields 7
+++ Fetsic grani toids /
in basement ..

ii o + ,ooo (, ,.-
104' 05"
I i
Fig. 4. Major fault zones and hydrocarbon areas in the Malay basin. Note the orientations of HC fields and their
apparent displacements across north-south fault zones. Known occurrences of felsic granites in the basement are
also shown.

K a p a l - B e r g a d i n g tectonic line; north-trending
Dulang, Bundi and Mesah zones; NNW-striking
D u n g u n zone that includes the D u n g u n Graben;
the N W Selambau zone; and an u n n a m e d NW-zone
along part of the basin's northeast margin.
There are half-graben and graben along many of
the offshore fault zones. Their shapes, orientations
of bounding fractures, and the positions of basins
along or within these major fault zones indicate
that the basins are pull-apart basins produced by
286 K. NGAHETAL.

wrenching. The pull-apart basins became depo-
centres of thick sedimentary sequences. Strike-slip
reversals along the major fault zones are demon-
strated by the fact that the sediments within the
pull-apart basins were compressed into en echelon
folds whose orientations are consistent with lateral
fault slip in directions opposite to those during the
formation of the pull-apart basins (see Tjia & Liew
1996). In cross sections, change in the regional
stress field from initially (trans)tensional to subse-
quent (trans)pressional is shown by inverted anti-
clines. The later stress field produced anticlinal
bulges and domes where the sedimentary columns
are thickest, that is, in the half-graben. In other
words, the orientation and position of the folds
were pre-determined by those of the basement
faults.
Seismic profiles show that the E-W anticlines in
the axial zone of the Malay basin are located over
E-W half-graben. Figure 5 indicates that many
of the half-graben are en echelon. Based on this
observation, it is suggested that these staggered
patterns are associated with NW-striking basement
faults. Furthermore, these faults are postulated to
represent a major NW-trending basement fault zone
along the axis of the basin. Towards the northwest,
this basement fault zone (named as the Axial Malay
fault; Tjia 1995) most probably continues as the

j ,<
0 t00 km 7~0~
: .\ I- I
\
\
\

\ 6%5
\
7 [ \\

Ma laysia I ~ ~ ~ '

- 1 \ I
Fig. 5. En echelon faulted sub-basins in the axial zone of the Malay basin are interpreted to have developed through
sinistral wrenching along the newly interpreted Axial Malay fault zone. Figure has been adapted from various sources
held by Petronas.
 FORMATION AND DEVELOPMENT OF MALAY & PENYU BASINS
Three Pagodas fault. Sinistral slip along the Axial
Malay fault produced the east-west half-graben
before the Late Oligocene. Strike-slip reversal
during the Middle to Late Miocene compressed
the sub-basin fills into east-west anticlines.
The major N-S faults of the Malay basin
delineate tectonic domains of different structural
styles and/or different deformation history, and
have laterally displaced crustal blocks. The distri-
butions of oil and gas fields (structural domes)
across the Kapal-Bergading Line and across each
of the other three N-S fault zones (Dulang, Bundi
and Mesah) appear displaced dextrally by several
tens of kilometres (Fig. 4). The position and orien-
tation of these domes have been determined by
those of the basement fractures. The stratigraphic
record of the basin suggests that compressive
tectonic deformation took place in the Middle to
Late Miocene, during which time right-lateral
wrench motion could have occurred along the
major N-S fault zones. The regional maximum
compressive stress was most probably orientated
between NE-SW and ENE-WSW. This orientation
of regional compression is also consistent with the
observed strike-slip reversals along the WHL and
Dungun fault zones.
WRENCH
,,~!~..~,,, DEPRESSION WITH
" ~ ~ DIRECTION OF
DEEPENING
FAULT SCARP
BASEMENT HIGH
PENYU BASIN
BASE OF TERTIARY ~ ....... ~.~'!!i~i
10 km
Fig. 6. Plan of major basins and faults at the base of the Tertiary in the Penyu basin. The Rumbia fault separates
E-W structures in the western part from WNW structures. En echelon sigmoidal fault pattern (top centre) suggests
wrenching. Simplified from an unpublished map by Texaco (1992) and held by Petronas.
287
Penyu basin
The pre-Tertiary basement highs that frame the
Penyu basin include (clockwise on Fig. 1): the
Tenggol Arch (N and NE), Johor Platform (S),
peninsular Malaysia (W) and Pahang Platform
(NW). The Tenggol Arch separates the Penyu basin
from the Malay and West Natuna basins.
Sedimentary cover over the basement highs is thin,
except for the Tenggol Arch that subsided some-
what and is capped by more than a kilometre of
Tertiary sediments. Basement faults in NW-SE and
E-W directions have compartmentalized the basin
into ten half-graben. The basin is further divided by
the NW-striking Rumbia fault into a west part and
an east part. Half-graben and fault scarps strike
E - W in the west, but strike WNW in the east part.
The Rumbia fault is over 100 km long and its heave
is more than 2 km. Left-lateral separation by the
fault of once-contiguous structures is believed to
exceed 20 kin. En echelon sigmoidal fractures near
the fault support this left-lateral slip interpretation
(Fig. 6). The E - W striking fractures and half-
graben are interpreted as secondary structures that
resulted from left-lateral slip along the Rumbia
fault and other, as yet undiscovered NW-trending
2;iii;iii;;;;(:~'El~Ttl~i;11i
i!i~iii!i!ii~ii!i!!ii!~.
PARI 1 0
ii.~';iiiiiii
i iiii iiii" .¢,: ~'~"~"" ~"~i:
~ _.~,,.
288 K. NGAH E T A L .
wrench faults in the basement. Continued left-slip
movement along the Rumbia fault is postulated to
have rotated the E-W faults and basins in the
eastern part of the basin into their present W N W -
ESE orientation, with rotation of up to 25 ° (Fig. 7).
There are anticlinal features over the deepest parts
of the corresponding half-graben but the effect of
compressional tectonics in the Penyu basin is less
obvious compared to that in the Malay basin. A
regional unconformity at the top of the Penyu
Formation (see Fig. 2) represents tectonic uplift at
the end of Oligocene time. Subsidence in Early
Miocene time began with deposition of the Pari
Formation consisting of alternating sandstone and
shale. Unpublished palaeontological information
from a study conducted by TEXACO in 1992 dates
a transpressional event in mid-Miocene.
Discussion and conclusions
The structural history and its relationship to
regional tectonics are postulated to be as follows.
(1) The Malay and Penyu basins originated
during the Late Cretaceous as two of the three rift
arms that developed above a continental crustal
dome located over a mantle plume. The West
Natuna basin is the third rift arm.
(2) The main strike of the Malay basin follows
the NNW-NW trending tectonic grain of this part
SOUTHWEST
Rumbia- 1 Rhu-
Fig. 7. Schematic cross-section of the Penyu basin. Note half-graben, mild inversion structures at Rumbia-1 and at
Cherating-1. From unpublished material by Texaco (1992) and held by Petronas.
of Sundaland. The northern part of the basin is
controlled by N-S fractures that became especially
dominant farther north in the Gulf of Thailand.
Geological evidence from onshore peninsular
Malaysia indicates that the NNW-NW tectonic
grain was established in Late Triassic-Early
Jurassic time. In the mid-Eocene, parts of con-
tinental SE Asia experienced expulsion as a conse-
quence of collision between the Indian plate with
the Eurasian plate. Fractures parallel and sub-
parallel to the tectonic grain in the basement of the
basin moved as wrench faults. A bundle of long
NW-SE fractures along the axis of the offshore
sub-basins (the Axial Malay fault zone) became an
extension of the Three Pagodas zone. Through
left-lateral motion of the Axial Malay fault zone,
en echelon east-west tension faults and half-graben
developed. At that time dextral wrench movement
along the WHL and its major fault splay, the
Dungun zone, resulted in the formation of smaller
pull-apart basins.
(3) The Penyu basin is dominated by NW-SE
and E - W structures. The NW-SE structures are
sub-parallel to the Late Triassic-Early Jurassic
tectonic grain of the Malay Peninsula. The Rumbia
fault zone consists of NW-SE fractures and divides
the basin into two parts. In the western part, E - W
trending half-graben and faults are dominant,
whereas on the east the dominant trend is W N W -
ESE. It is interpreted that this strike deviation is
EFt
NORTHEAST
Cherating- 1
E)
c~
IIIA Alu - Alu E- I
 FORMATION AND DEVELOPMENT OF MALAY & PENYU BASINS
the result of clockwise rotation of initially E - W
trending half-graben and other structures through
continued sinistral slip along the Rumbia fault.
These E - W structures are believed to have
developed within and across fault-bounded,
N W - S E elongated crustal blocks, which slipped
sinistrally during differential expulsion of SE Asia.
Expulsion began in mid-Eocene, when the Indian
plate started to push into the Eurasian plate.
(4) By the late Early Miocene, the Australian
continent had approached SE Asia sufficiently
closely as to obstruct free expulsion of its crustal
fragments. In addition, the effect of the westward
drive of the Pacific plate also began to be felt in this
region. Before that time, active N - S spreading in
the South China Sea basin and the Philippine basin
apparently prevented stress generated by the Pacific
plate motion to reach the SE Asian region. From the
late Early Miocene to Late Miocene, the combined
effect of northward and westward convergence of
the Indian-Australian and Pacific plates manifested
themselves in a regional compressive regime and
reversals of slip sense along several of the N W - S E
striking wrench faults. Some examples of the
resultant effect of this movement are in the Malay
basin, where dextral slip occurred along its Axial
References
ARESm~V, E. G., TRAN LE DONG, NGO THUONG SAN
SHNm, O. A. 1992. Reservoirs in fractured basement
on the continental shelf of southern Vietnam.
Journal of Petroleum Geology, 15, 451-464.
BURTON,C. K. 1965. Wrench faulting in Malaya. Journal
of Geology, 73, 781-798.
GEOLOGICAL SURVEY OF MALAYSIA. 1985. Geological
Map of Peninsular Malaysia, 1:500,000. Ipoh,
Geological Survey of Malaysia, 8th edition.
KATILI,J. A. 1973. Geochronology of west Indonesia and
its implications on plate tectonics. Tectonophysics,
19, 195-212.
KHALID NGAH. 1975. Stratigraphic and Structural
Analyses of the Penyu Basin, Malaysia. MSc Thesis,
Oklahoma State University.
MD. NAZRI RAMLI. 1988. Stratigraphy and paleofacies
development of Carigali's operating areas in the
Malay Basin, South China Sea. Bulletin Geological
Society of Malaysia, 22, 153-187.
MOHD. FIRDAUSABDUL HALIM. 1993. Heat flow in the
Malaysia sedimentary basins. In: Petronas Research
& Scientific Services Sdn. Bhd., Proceedings
Exploration Research Seminar. I, Progress Reports
and Findings. 11-25.
METCALFE, I. 1988. Origin and assembly of Southeast
289
Malay basement faults, accompanied by com-
pression that developed anticlinal features forming
oil- and gas-bearing structures. Unpublished data
from EPMI-acreage in the southeastern part of the
basin suggest thrusting towards the south. These
thrust faults and the anticlines trend E-W, that is,
the orientation determined by the basement
fractures. Slip reversals along the Dungun fault
zone and other N - S to N W - S E trending regional
faults which inverted the pull-apart sedimentary
in fills are also indicative of the movement. The
axes of the inversion were orientated in a sense
consistent with the lateral slip sense of the associat-
ed wrench fault. The Rumbia fault does not seem to
have experienced slip reversal. On the contrary,
continued sinistral slip appears to have rotated
structures in its eastern part clockwise for 25 ° .
Since the mid-Miocene, continued sinistral strike-
slip on the Rumbia fault most probably negated
the effect of most of the compressive stress exerted
by the convergence of the plates. This stress-
decoupling explains the milder deformation
observed in the Penyu basin.
We are grateful to C. S. Hutchison and R. E. Swarbrick for
constructive comments on an earlier version of the paper.
Asian continental terranes. In: AUDLEY-CHARLES,
M. G. & HALLAM,A. (eds) Gondwana and Tethys.
Geological Society, London, Special Publication,
37, 101-118.
NG. T. S. 1987. Trap styles of the Tenggol Arch and
the southern part of the Malay Basin. Geological
Society of Malaysia Bulletin, 21, 177-193.
TJIA, H. D. 1989a. Tectonic history of the Bentong-
Bengkalis suture. Geologi Indonesia, 12, 89-111.
1989b. Major faults of Peninsular Malaysia on
remotely-sensed images. Sains Malaysiana, 18,
101-114.
1995. Inversion tectonics in the Malay Basin:
evidence and timing of events. Geological Society
of Malaysia Bulletin, in press.
& L~EW, K. K. 1996. Changes in tectonic stress
field in northern Sunda Shelf basins. This volume.
WAN ISMML WAN YUSOFF & SWARBRICK,R. E. 1994.
Thermal and pressure histories of the Malay Basin,
offshore Malaysia [abstract]. AAPG Bulletin, 78,
1171.
ZAITON HARUN. 1992. Anatomi Sesar-sesar Utama
Semenanjung Malaysia. PhD Thesis, Universiti
Kebangsaan Malaysia, Bangi.
 Changes in tectonic stress field in northern Sunda Shelf basins

H. D. T J I A & K. K. L I E W
PETRONAS Research & Scientific Services, Sdn. Bhd., Lot 1026 PKNS Industrial Estate,
54200 Hulu Kelang, Malaysia

Abstract: The Tertiary basins of the northern Sunda Shelf are underlain by normal and
attenuated continental crust that is characterized by moderate to high average geothermal
gradients in excess of 5°C/100 m. In the Malay basin, upper Oligocene and younger sediments
are more than 12 km thick; in the other basins, such sediments are between 4 and 8 km thick.
The Malay, Penyu and West Natuna basins are aulacogens meeting at a triple junction that marks
a Late Cretaceous hot spot in the centre of the Malay Dome. Sub-basins (commonly haif-graben)
developed as pull-apart basins within regional, north to northwest-striking, wrench fault zones.
The NW-striking Three Pagodas fault most probably extends farther southeast as the Axial Malay
fault in the basement along the length of the Malay basin. Pre-late Oligocene sinistral slip along
this fault developed east-west half-graben that fixed the position and orientation of inverted
anticlines that later developed in the basin-filling sediments through strike-slip reversal along
the fault. Initial basin subsidence took place during Eocene-Oligocene time. Regional tensional
conditions prevailed until the early Miocene. During the middle to late Miocene, regional
compressionai stresses caused reversals of the sense of motion on the major wrench faults, and
structural inversion of the basin-filling sediments. On some of the north-striking wrench faults
there are indications of up to 45 km of right-lateral displacement which is possibly post-
Miocene. The regional wrench faults have acted as domain boundaries with each tectonic domain
characterized by different stress fields. The stress systems evolving during the Cenozoic are
attributed to varying degrees of interference of plates coupled with changes in convergent
directions and/or rates of motion of the Pacific plate, the Indian Ocean-Australian plate, and
continued, differential extrusion of SE Asian crust following the collision of the Indian plate with
the Eurasian plate.

Sundaland is the geological province in SE Asia
which comprises pre-Tertiary basement rocks of the
Sunda Shelf and contiguous land areas of western
Borneo (or West Kalimantan plus westernmost
Sarawak) and peninsular Malaysia. The northern
Sunda Shelf extends south to the Johor Platform,
and north to a line connecting Kota Bharu with the
SW tip of Vietnam (Fig. 1). Most of Sundaland
is tectonically stable, but possible Late Cenozoic
vertical crustal movements have been detected in
certain areas. One such area is onshore Johor where
uplift reached several scores of metres and where
block tilting has been recognized (Anizan Isahak
1993). The western edge beneath the petroleum
basins of central and south Sumatra have experi-
enced downwarping. On the shelf proper, net sub-
sidence of several kilometres created the Malay,
Penyu, West Natuna, Nam Con Son and Mekong
basins. This subsidence was accompanied by
stretching and attenuation of the continental crust.
Average geothermal gradients of 51.8°C/km in the
Malay basin are consistent with thinner than normal
crust (Mohd. Firdaus Abdul Halim 1993).
Tapponnier et al. (1982) suggested that the
formation of the basins in continental SE Asia
resulted from collision of the Indian plate with the
Eurasian plate. The collision that began 45 Ma ago
has pushed out crustal blocks of SE Asia towards
the southeast along major wrench faults. This
extrusion took place in stages with one block
moving earlier than the other. It is suggested that
this extrusion and associated wrench-faulting
created a tensional environment in continental SE
Asia which resulted in the formation of basins
floored by continental crust. By mid-Miocene time,
the stress field changed into a compressional
regime. The reason was, according to one
suggestion, that the northward progression of the
Indian-Australian plate blocked further expulsion.
Harder et al. (1992) agreed that the continuing
northward push by India had been one of the
causes that changed the stress system in SE Asia
throughout most of the Tertiary period. They
further believed that changes also occurred through
rotation of the extruded fault-bounded crustal
blocks. We propose that the change from a
tensional regime into a compressional stress field
could also be attributed to the influence of the
westward-moving Pacific plate. In the Eocene (at
43 Ma), the Pacific plate changed its direction from
northwestward to westward convergence. It seems
that the effect of the latter on the SE Asian region

From Hail, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, 291
Geological Society Special Publication No. 106, pp. 291-306.
292 ~. D. TJIA & K. K. LIEW

1 ,

l'°i I
I0~
~',~

!
INDOS NIA
i sT~R",
g .¢ .. .•

10"II ! ,'
! a I,]/ KHMER
I I It , ~
l llii. ....

II L . ~ : : " " '~°

.:--.,: -, . . . .o¢, •
, , .---~
blAM" CON SON'
;~i'.' ". '.. , /" #'-
• ..: • " . . " . - o .

,,j , '--lJj
I "~-SONGKHLA ~, , ", ,. .," ' . . "• '
| / •

I ~ ( s
• . . - . ,.•..
( pt "J2 - /"
l -.':' [
. . ...~:.. • ! ~

I Peninsular
q j , , • . , :;: ["-"
. Malaysia ".. I<C
;.Tenggol ' ' i..
~
w e s t • '
r':..Arch "- ' UNA I <
• " t " . . • , .

' ' ~4 -
" PENYI ,

,.-- iOi i' '-"i " Not u( • . ".: .. x

Fig. 1. The major Cenozoic basins o f the northern Sunda Shelf. The shelf area is stippled; its northern boundaries are
only approximate• Basin outlines drawn after a map by C C O P (1991)•

only became felt by mid-Miocene time. Before that

Malay basin
time, the westward push of the Pacific plate may
have been buffered by north-south spreading in Among the Cenozoic basins in the northern Sunda
the West Philippine, Caroline and South China Sea Shelf, the Malay basin is a major hydrocarbon-
basins. bearing basin on the continental shelf of the South
The present paper deals with the geometry of China Sea. The basin strikes northwest over a
structures and tectonic domains in the northern distance of 500 km and is 200 km wide (Fig. 1).
Sunda Shelf. Known faults and new evidence for Oligocene and younger sediments within the Malay
major faults and their kinematics serve to define basin are at least 12 km thick• Average long-term
the tectonic domains more accurately and establish subsidence rates are therefore in the order of 0.3-
the structural development of the Tertiary basins 0.4 m/1000 years, which is one order lower than
of the region. rates of vertical crustal movements in tectonically
 STRESS FIELD CHANGES, NORTHERN SUNDA SHELF
mobile regions. One such region is eastern
Indonesia where long-term uplift and subsidence
rates of 1-4 m/1000 years are common (Tjia et al.
1974; Fortuin & De Smet 1991). The northwest
strike of the Malay basin is of restricted extent.
Towards the northwest, the basin adjoins north-
trending basins of the Malaysia-Thailand Joint
Development Area (MTJDA) and of the Gulf of
Thailand. Towards the southeast the Malay basin is
intercepted by the West Natuna basin that strikes
east to north-northeast. In the southwest corner,
the Tenggol Arch, a basement high, separates the
Malay basin from the almost east-trending Penyu
basin (Fig. 1).
Stratigraphy
The Malay basin sedimentary sequences are most
commonly classified using the EPMI (Esso
Production [Malaysia] Incorporated) seismo-
stratigraphic scheme which is based on information
from the northeastern and southeastern parts of the
basin. For the northern and southwestern parts,
PETRONAS Carigali's nomenclature is preferred.
Carigali's stratigraphic units are based on the
recognition of regional shale and/or marine trans-
gressive pulses, supported by seismic correlation
(Md. Nazri Ramli 1988). Most EPMI seismic
sequence boundaries correspond to erosional
unconformities at the basin margins, except the top
of Group I (uppermost lower Miocene) which is
a maximum flooding surface. The groups are tied
to the EPMI Tertiary eustatic sea-level chart.
Recently, based on micropalaeontological markers,
the EPMI stratigraphic scheme has been revised
and this is used in the following explanation. The
initial basin fill (Group M to Group J, Lower
Oligocene-lowermost Miocene) was mainly non-
marine. It was followed by deposition of Group I to
Group H (Lower Miocene-lower Middle Miocene)
in a stable geological environment. Severe
regression occurred in mid-Group H time, and this
was succeeded by a major marine transgression
during upper Group H and Group F times (Middle
Miocene). Group E (upper Middle Miocene) was
dominated by general regression. During Group D
(uppermost Middle Miocene) times, transgressive
conditions prevailed. A major unconformity
marked the end of Group D time, but transgression
continued into Group B and Group A (Upper
Miocene-Pleistocene). A comparison of the
various stratigraphic schemes used in the Malay,
Penyu and West Natuna basins is shown in Khalid
Ngah et al. (1996).
The pre-Tertiary geology of the deeper part of
the basin is poorly known. The basement is thought
to consist mainly of Jurassic-Cretaceous meta-
sediments, some upper Palaeozoic metasedi-
293
mentary rocks and igneous bodies. The centre of
the basin contains more than 12 km of sediments
which progressively onlap onto the basin margins.
Wells are generally less than 5 km deep. The oldest
dated sediments are upper Oligocene. Recently
acquired deep (12 seconds) seismic sections
suggest that in the axial zone of the basin, layered
sequences possessing seismic signatures similar
to the oldest known sediments extend several
kilometres deeper below the Upper Oligocene
strata. In other words, the basin was initiated well
before the early Oligocene.
Structures in the Malay basin
Fractures and f o l d s
The five main fracture directions in the Malay basin
are north, NW, NNW, NNE and east. The east-
trending fractures are confined to the pre-Tertiary
basement (Fig. 2). The NW-striking fractures are
along the basin margins with a single example, the
Selambau fault (9 on Fig. 2) in the axial zone of the
basin. Northward, the Western Hinge-line (WHL: 5
on Fig. 2) fault zone curves to become parallel to
the dominant north-striking faults of the MTJDA
and the Gulf of Thailand regions. The Kuda-Ular
fault (4) is one of north-trending zones. Another
north-trending fracture is the Kapal-Bergading
fault zone (5) that separates the Malay basin into
a southern and a northern domain. The northern
domain is marked by north-trending faults and
folds, while the southern domain has east-west
folds and variably striking major faults. Work in
progress onshore peninsular Malaysia has identi-
fied southern extensions of these northerly striking
faults. Onshore the Ping-Teris fault is aligned along
longitude 102053 ' and offshore can be traced into
the northerly-striking fault bundle on the western
margin of the Malay basin. The Ping-Teris fault
zone consists of several, up to 10 m wide, bands of
contorted phyllonite alternating with parallel bands
of less deformed phyllite totalling 30-100 m wide
in outcrops near the Ping River and Teris Village in
Terengganu State. The phyllonite and the phyllite
bands strike 345-350 °, with foliations dipping
west or east at more than 75 °. Slickensides on
fault planes are subhorizontal and associated small-
scale fault-plane features such as bruised steps,
accretion spalls, and pluck steps (see Tjia 1972a,
for descriptions and definitions) indicate fight-
lateral wrenching. Proprietary data indicate that
in the MTJDA, a north-striking fault that may be
the offshore extension of the Ping-Teris fault is
associated with fight-stepping en echelon faults
in Miocene sediments and suggests, therefore,
sinistral wrenching of younger age. Three
north-south fault zones are also prominent in the
294 H . D . TJIA •
/
A ,- Folds in Oligocene-
6
eeoooeo
D I
ODOQQ
PENINSULAR[~
0 5O 100
I /r , I
ILLI,IIIIIIII
I IIIIIII I
1o,,.
Fig. 2. Major structures in the Malay basin. Named faults are: 1. Dungun; 2. Western Hinge-line or WHL;
3. Tenggol; 4. Ular-Kuda; 5. Kapal-Bergading; 6, Dulang; 7, Bundi; 8. Mesah; and 9. Selambau. Note the apparent
dextral separations of east-west fold axes in the central and southern part of the basin. MTJDA, Malaysia-Thailand
Joint Development Area. Information derived from unpublished reports held by PETRONAS.
southern domain: (6) Dulang, (7) Bundi and (8)
Mesah fault zones. These three fault zones appear
to have dextrally offset the hydrocarbon fields of
the Malay basin. The Dulang fault exhibits 45 km,
and together the Mesah and Bundi faults causes
30 km of dextral separation (see Khalid Ngah et al.
1996).
Within the southern tectonic domain, most of the
east-west folds are interpreted to be associated
with east-west basement faults that usually form
the southern side of half-graben. These deeper
structural trends have been based on gravity,
magnetic and occasionally also on seismic
interpretation. Another fracture trend in the basin is
NNE, which is confined to the northeast and east
sides of the Malay basin. The WHL is associated
with NNW-striking fault zones: the (1) Dungun and
(3) Tenggol fault zones (Fig. 2).
The Upper Oligocene (and most probably older
Tertiary strata) to Miocene basin-filling sediments
are deformed into large open folds with limbs that
commonly dip less than 15 ° . Steeper dips are
clearly associated with faults. The Sepat anticline is
at 35 km the longest. Within the central part of the
Malay basin, fold axes strike east-west; they are
north-south in the northem part of the basin, and
K. K. LIEW
- Faults
-- Faults in basement
Miocene sediments
....
/ °e=
a'eoeeeooG ,
eooeoo
\" los.1 /
I I.
NW-SE to NNW-SSE along the basin margins. In
the southern part of the basin the east-west folds
are transected by shallow, N-S crestal faults.
Seismic profiles show that the basinal east-west
folds are located over half-graben and deeper parts
of the main basin (Fig. 3). Three to four east-west
trending parallel belts of anticlines and associated
hydrocarbon fields can be distinguished (Fig. 2).
Adjacent to and within the Western Hinge-line fault
zone, the anticlines are parallel or almost-parallel
to the fault zone. The major part of this fault zone
strikes northwest. The fold trends resemble drag-
fold orientations caused by left-lateral wrenching.
North to north-northwest trending folds occur to
the west of the Kapal-Bergading north-south
trending tectonic zone. Some rounded domal and
basinal structures straddle this tectonic zone. These
more rounded fold structures in the border zone
most probably reflect the influences of different
stress fields in adjacent tectonic domains.
Pull-apart depressions
Distinct pull-apart basins in the pre-upper
Oligocene basement occur along the Western
Hinge-line fault zone in the northwest, and in the
 STRESS FIELD CHANGES, NORTHERN SUNDA SHELF 295

SW NE
Inversi on ~-Plio- Pleistocene
!

,~ ~ Miocene __ ~ :

/ IIIIIIIIII r i4. I I 11111111 '

/ iJJiilllll
)re-Tertiory [i1| pre-U. OLigocene
11 '/ 111111111
I I /11"IIIIIIIIIIIIIIIIIIIIIII

°SilTliit [11 li lllltlllllillllr!!l]!!tlll

Fig. 3. Compressed composite cross-section across the Malay basin interpreted from seismic lines. Even at 12 km
depth no basement is apparent in the axial zone which also has a large inverted anticline. Another structural inversion
in the SW part of the section represents pull-apart basins within the Western Hinge-line fault zone (2 on Fig. 2). The
deeper parts were interpreted from a seismic line reprocessed by PRSS.

south-central part of the Malay basin. Along the
WHL fault zone, the long axes of the pull-apart
basins trend north to NNW (Fig. 4). This orien-
tation, combined with the general strike of the fault
zone, suggests right-lateral slip along the WHL
as the cause of the pull-apart basins, This trans-
tensional stress field pre-dates the Upper Oligocene
sediments. The anticlines formed in the Upper
Oligocene-Miocene sedimentary fills of the basins
trend sub-parallel to the elongation of the basins.
This fold pattern resembles that of drag folds due
to left-lateral slip on the WHL fault zone. The
sense of this later wrench motion was opposite
to that which produced the pull-apart basins, and
developed when the region was subjected to trans-
pression during the middle to late Miocene (Tjia
1995). On the basis of existing information the
WHL fault zone can be traced southeastward only
as far as the Penyu basin.
In the area of the Laris fields, within the north-
striking Mesah Fault zone, are at least three
probable pull-apart basins (Fig. 5). The orientation
of their faulted boundaries with respect to the
regional strike of the main fault zone suggests left-
lateral wrenching as the cause. However, the fault
pattern within the northwest-striking Selambau
fault zone consists of two medium-sized pull-apart
structures that suggest dextral wrenching along the
major fault zone (Fig. 5). The major east-west
anticlines consisting of Miocene and older sedi-
ments of the Malay basin are offset by the Mesah
fault zone in a right-lateral sense by about 30 km
(Fig. 2). This suggests that dextral fault displace-
ments occurred in Late Miocene or even younger
times. On the other hand, colleagues from EPMI
(pets. comm. 1994) report that they have not
encountered indications of large-scale horizontal
displacements along the north-south faults they
are familiar with. They further suggest that major
strike-slip displacements on N-S faults probably
took place before the Malay basin became filled
with Tertiary sediments.
D u n g u n Graben
The Dungun Graben is located to the northwest of
the Tenggol Arch. This graben has rhomboid shape
in plan view; its two longer sides trend northwest,
while the two shorter sides trend north (Fig. 6).
Liew (1993) studied the structural development
of this graben. It is elongated in a NNW-SSE
direction and continues towards the NNW as a
40 km long and 5-6 km wide fault zone. Fracture
sets in the pre-Tertiary basement of the graben also
distinctly display a pattern of longer NNW faults
interconnected by shorter north-south faults. The
basement deepens towards the northeast where a
large NNW-fault forms its boundary. Along this
fault, heave is 1.8 km and throw is more than
2 km. The west side of the graben has many NNW-
striking step-faults. There is a NNW-plunging base-
ment antiforrn in the south corner of the graben.
The Tertiary base of this graben consists of three
elongated structural units (west, central and east;
296 H . D . TJIA • K. K. LIEW

~, .
104 O0
lo
o
E
r-

,
!'(I
? ...... r-~,

o ........ ..
r-
t--
--4

"%°, _~ I ,..

°°. ...
°"° !

. k ........ , ..... ~ . . . . .

-6

°'- - . . . . . . . . o... ....

~ . .... / / . . . . .... .~
Isl. ...
\'~ '~ ........ ' .......... ,JjI
% f i/I

w ! ! ~ ' ".
,.. , , , b . . . "~'...
........... • C ~
........ d
.... -:z .... • 430.
~'x'~ f I 2 3 X
A to F Inverted pull-apart basins "~(X 104"00
I
Fig. 4. Structures in uppermost Oligocene to Upper Miocene sediments of the northwestern part of the Malay basin.
Explanation of symbols: a. major fault, hachures on downthrown side where determined; b. smaller fault; c. axis
of plunging anticline; d. fold axis, type not determined; e. synclinal axis; f. sense of slip on wrench faults;
1. pre-Oligocene; 2. middle-late Miocene; 3. probably post-Miocene. Faults and folds based on an EPMI map
of 1991; newly interpreted are the wrench faults and their timing, and the pull-apart basins. Note that the Kapal-
Bergading tectonic line separates east-trending from north-trending folds.
 STRESS FIELD CHANGES, NORTHERN SUNDA SHELF 297

104" 10/.,°30
. ..,'..

f(": i:: DULANG BUNDI i "-;."'"'"'" """ I

.....=FAULTZONE .-'---,FAULT ZONE
.k..,

MESAH "": 'r

FAULTZONE,:" l 6'30

(...

' "-----FAULT
FOLDAXIS
\ i's
"~'..~i':: HCFIELDof PROSPECT
b04" ~ P U L L APART
\ f
Fig. 5. Structures in the Central Malay basin. The interpreted pull-apart basins within the Mesah and Selambau fault
zones possess patterns suggesting right-lateral wrenching as their cause. Structures were interpreted from a 1991
EPMI map held by PETRONAS.

Fig. 6). The east unit contains the large NW fault
set that forms the graben boundary. There the base-
ment dips steeply west and is transected by a few
NNW faults. The central unit is characterized by
left-stepping en echelon WNW-striking faults and a
basement that is gently inclined towards
the NNW. Bifurcating NNW- and N-trending faults
dominate the west unit of the graben. There, the
basement deepens towards the ENE.
The rhomboid shape of the Dungun Graben and
its location along the NNW Dungun fault zone
suggest that it was formed by dextral wrenching
along the fault zone (Fig. 6). The Dungun-B and
Dungun-C graben may be dextrally-displaced
portions of a single basin with an offset of c.
14 km. However, the NNW-trending fold axes of
the Tertiary basement and overlying Tertiary sedi-
ments (not shown on Fig. 6) are interpreted as drag
folds and suggest a later sinistral wrench motion
along the Dungun fault zone.
Penyu basin
The Penyu basin lies offshore of Kuantan, Pahang.
It consists of two parts separated by the major NW-
striking Rumbia fault (new name; see Khalid Ngah
298 H.D. TJIA & K. K. LIEW
Fig. 6. The Dungun fault zone runs NNW and contains
the Dungun Graben and Dungun-B graben, Dungun-A
and Dungun-C half-graben. Structures based on
unpublished 1992 map of Western Mining Co., held
by PETRONAS.
et al. 1996). It is bordered by the Tenggol Arch to
the northeast, the Pahang Platform to the northwest,
and the Johor Platform to the south (Fig. 1).
Khalid Ngah (1975) analysed the stratigraphy
and structure of the basin. Proprietary data avail-
able since 1992 reveal further information. The
western part is elongated east-west and possesses
half-graben of similar strikes. The basement of
these half-graben dips northward, except in the
southeast corner, where the basement is inclined
towards the southeast. The main east-west fault set
of the western basin abuts against the NW-striking
Rumbia fault. Half-graben in the eastern part of the
Penyu basin trend WNW to NW. The basement
in most of these half-graben inclines towards the
southwest. Most fold axes within the Penyu basin
are parallel to major faults. This suggests that these
folds may be rollovers.
The NW Rumbia fault separates the E - W
structures in the western basin from the W N W -
ESE structures to the east. This fault is more than
100 km long and possesses an average heave of
2 km. The structure also separates areas with
different tilting directions. In some areas en echelon
faults, sometimes sigmoidal in shape, suggest
wrench fault movement. The Rumbia fault runs
parallel to a zone of probable left-lateral faulting
(see Khalid N g a h e t al. 1996). In other words, the
Rumbia fault is perhaps also a left-lateral wrench.
The structural history of the Penyu basin is as
follows. Prior to the Late Oligocene, north-south
extension/transtension formed east-west striking
half-graben in the western and eastern Penyu basin.
Alternatively, these half-graben may have formed
as consequence of left-lateral wrenching along
major NW-trending faults parallel to the Rumbia
fault but located outside the Penyu basin proper.
We suggest that during Late Oligocene to Middle
Miocene continued left-lateral motion on the
Rumbia fault zone rotated the half-graben of the
eastern Penyu basin clockwise some 25 ° into
WNW-striking structures.
West Natuna basin
The West Natuna basin broadly curves from an
east-northeast trend in the southwest to a northeast
direction farther towards the east and north. The
structural trend of the Malay basin continues into
the west-central part of the West Natuna area (Figs
1 and 7).The following geological information on
the West Natuna basin is from Wongsosantiko &
Wirojudo (1984), Daines (1985), and Sujanto et al.
(1986).
The basin began to develop during the Early
Oligocene when rifting and/or pull-apart processes
produced NE-SW half-graben that began to be
filled with non-marine sediments. Upper
Oligocene-Lower Miocene sediments are of
fluvial-lacustrine character and are considered
hydrocarbon source rocks. In late Early Miocene,
the area became subjected to compression and this
resulted in reversal of slip sense on existing normal
faults and inverted structures in the half-graben.
Right-lateral displacements took place along east
and ESE-striking fault zones, while left-lateral
offsets developed along north-striking faults (see
Daines 1985). Folding of sediments filling half-
graben produced anticlines with ENE trends. This
fault pattern implies a maximum horizontal stress
directed NW-SE. During the remainder of the
Miocene, paralic-marine sedimentation took place
in quiet tectonic surroundings. Late Miocene to
Early Pliocene tectonic deformation affected the
entire basin and developed an unconformity,
especially on anticlinal crests. Overlying Pliocene
to Recent marine sediments are relatively un-
 STRESS FIELD CHANGES, NORTHERN SUNDA SHELF
--2" ~(~ 108"
Fig. 7. Isopachs (in km) of Cenozoic sediments in
some of the northern Sunda Shelf basins. North-south
lineaments of isopach contours probably represent
major faults. The Vietnam Shear is also indicated by the
linear shelf edge. Note the en echelon arrangement of
sub-basins within the Nam Con Son basin. Structures
were interpreted from a CCOP map (1991).
disturbed. Pliocene faulting is essentially absent.
Locally, total sediment thicknesses reach 4500 m.
It is believed that except in the deeper parts,
maturation and migration of hydrocarbons occurred
only very recently.
East Natuna basin
The north-south elongated East Natuna basin, also
known as the Sokang/Soikang basin, lies to the east
of the Natuna Arch and occupies the eastern margin
299
of the Sunda Shelf (Figs 1 and 7). Wongsosantiko
& Wirojudo (1984) pointed out that this basin has a
different structural history compared with the West
Natuna basin. Oligocene half-graben occur only in
the northern part. The graben fills are thinner and
suggest a smaller amount of subsidence, and conse-
quently also a weaker rifting effect. Outside the
graben area, the larger part of the East Natuna basin
is covered by relatively thin (less than 1 km)
Oligocene sediments. Miocene sediments are
locally thick and are bounded by growth faults
which began to develop in Early Miocene time.
Most of the Miocene faults are extensional and can
be grouped into NNW and NE-striking fractures.
Only minor folds exist and these are associated
with NNW-SSE faults. Some of these NNW faults
suggest dextral-slip reflecting some lateral tectonic
stress. In the East Natuna basin the dominant stress
regime during the Miocene was (ex)tensional.
During the Plio-Pleistocene, the basin was
tectonically inactive except in its southern part
where Pliocene faults are common, and in its
southeast part which is characterized by clay
diapirism. Marked thickening of Plio-Pleistocene
sediments towards east is interpreted to indicate
rapid subsidence coupled with sedimentary load-
ing. These processes probably caused faulting and
clay diapirism. NE-striking faults occur in the
northernmost part and in the centre of the East
Natuna basin. Elsewhere, faults strike NNW to
NW. The boundary with the West Natuna basin is
marked by two regional north-striking, weakly
curved normal faults; the southern fault down-
throws west and is concave in that direction, the
northern fault downthrows east.
Mekong basin
The Mekong and Nam Con Son basins are located
in the northeast comer of the Sunda Shelf that
adjoins the Indochina craton (Figs 1 and 7). The
basins were formed in fractured and subsided pre-
Tertiary basement and are filled by more than 4 km
and more than 6 km of continental sediments
respectively, and locally also by peri-marine sedi-
ments (CCOP 1991). New proprietary data indicate
that a large part of the basement includes upper
Mesozoic granites that serve as important hydro-
carbon reservoirs.
The Mekong or Cuu Long basin strikes ENE and
is located offshore of the Mekong Delta. It consists
of a main basin where the maximum sediment
thickness exceeds 6 km, and several smaller basins.
All are within a 100 km wide zone trending NE
over a distance of 450 km. In the basement rocks of
the main basin, longitudinal faults dominate and are
associated with most of the half-graben. Other half-
graben are associated with shorter east-west faults,
300 H . D . TJIA & K. K. LIEW
especially in the SW sector and in the NE part of
the basin. Along the NE-orientated faults, down-
throw directions are mainly symmetrical about the
basin axis. The shortest faults are north-south. In
the White Tiger field in the centre of the basin,
these north-south faults seem to be associated with
the NE-trending faults, suggesting left-lateral
motion along the latter. Many faults with throws
between 200 m and 1000 m occur in the northeast
part of the basin (CCOP 1991). Other structural
elements include basement highs, basins and
monoclines. Unpublished cross-sections of the
White Tiger and Dragon fields show that deforma-
tion intensity decreases from the Lower Oligocene
to the Middle Miocene beds. Post-Miocene beds
are essentially undisturbed (Areshev et al. 1992).
The basin fill consists mainly of siliciclastic
sediments intercalated with some basalt and
andesite horizons in the Oligocene and Lower
Miocene. The Palaeogene is generally of con-
tinental character, while the younger beds represent
deltaic to shallow marine environments. Regional
unconformities were mapped at the base of
Palaeocene to Oligocene, in the Miocene, and in
the Pliocene (CCOP 1991). The basement may be
mainly block-faulted granitic rocks with K-Ar ages
of 97-178 Ma (T. Minh et al. (1991) cited by Mauri
et al. 1993). Geothermal gradients are moderate
to high (3.9°C/100 m) and Recent volcanism in
the northeastern part of the basin (Fontaine &
Workman 1978) may be expected to contribute
to the high heat flow.
Nam Con Son basin
On its south side, the Mekong basin is separated
by the NE-striking Con Son Swell from a basin
complex that is collectively known as the Saigon
basin, or Ho Chi Minh basin, or Nam Con Son
basin (Figs 1 and 7). The Con Son Swell may be
a regional horst bounded by NE-striking normal
faults and is composed of Cretaceous granites
(Mauri et al. 1993). Fault throws average 800 m
(CCOP 1991). The Nam Con Son basin may be
divided into an eastern part and a western part. The
division is effected by a north-trending basement
high that extends from the Natuna Arch to the Con
Son Swell. The eastern part of the basin strikes
NNE and contains two deeper en echelon troughs
trending NE. Its eastern border is linear and strikes
north (Fig. 7). The CCOP (1991) map shows that
this lineament extends along the entire edge of the
Sunda Shelf basement. We provisionally name it
the Vietnam Shear and interpret it as a major fault
zone with probable strike-slip motion. The pattern
of NE-striking troughs in the eastern Nam Con Son
basin in association with the Vietnam Shear
suggests that the basins may be the result of right-
lateral wrenching along the north-trending Vietnam
Shear. The western part of the Nam Con Son basin
consists of several smaller troughs that in the south
strike north and swing to NNE trends in the north.
The change in strike can probably be attributed
to the fact that the western Nam Con Son basin
straddles a north-trending fault zone that also forms
the western margin of the basement high that sub-
divides the greater Nam Con Son basin. North-
trending faults border the troughs of the western
basin. The CCOP (1991) map shows 7 km of
sediments in the basin.
Matthews & Todd (1993) classify the basin as
a Tertiary rift complex and recognize four major
tectonostratigraphic units. The oldest unit was
interpreted from seismic features and comprises
Palaeogene rift-fill deposited in a gently divergent
setting; seismic signatures of lower Miocene post-
rift sediments suggest regional thermal subsidence.
A second pulse of extension is represented by
middle-Miocene rift-fill that is capped by an
onlapping marine sequence. Gentle compression
associated with uplift of depocentres formed a
major late-middle Miocene truncational uncon-
formity. A third pulse of mild extension is indi-
cated by minor faulting and major regional
subsidence resulting in very rapid flooding during
the late Miocene and Plio-Pleistocene.
Changes in stress field
Recently, Tjia (1994) proposed that the Malay,
Penyu and West Natuna basins originated as rift
arms upon an upper Mesozoic crustal dome in the
northern Sunda platform. The updoming was
probably caused by a rising mantle plume. Major
arguments for that concept include the following.
(1) The three basins form a three-armed star and
meet at a triple junction. (2) The triple junction is
characterized by one of the higher heat flow areas
in a region that already possesses above-average
geothermal gradients. (3) The relatively high
positions of pre-Tertiary basement represented
by the Con Son Swell, The Natuna-Paus-Ranai
High, the Johor Platform, the Tenggol Arch and
peninsular Malaysia may be remnants of the crustal
dome. (4) Late Cretaceous granitic intrusions are
widespread in this region and define the initial age
of the domal structure. (5) The tensional stresses
that appear to have acted contemporaneously to
create the three basins of varying strikes are most
simply explained as result of rifting atop a litho-
spheric dome. In other words, the Malay, Penyu
and Natuna basins are interpreted as aulacogens
that developed upon a late Cretaceous Malay Dome
(new name) possessing a diameter of 1000 km.
At one time or another, previous researchers on
the Cenozoic basins of this region have attributed
 STRESS FIELD CHANGES, NORTHERN SUNDA SHELF
the basin origins to subsidence by crustal thinning,
or to rifting, or to wrench-fault related processes
without further explanation. From this study it has
been concluded that these basins experienced
multiple deformational processes of contrasting
nature. Superimposition of deformational signa-
tures has contributed to the confusion about their
origin. For instance, the Penyu, West Natuna and
East Natuna basins have been suggested to be
extensional basins. However, en echelon arrange-
ment of fractures and basins in their basement rocks
also indicates that transtension caused by wrench-
ing must have been important modifiers in basin
development. Another example: east-west half-
graben in the western Penyu basin seem related to
left-lateral wrenching along the NW-striking
Rumbia Fault and/or other major faults of similar
strikes. Similarly, the NE-trending half-graben in
the northern part of the East Natuna basin may be
associated with right-lateral wrenching along the
major, north-striking Vietnam Shear that marks
the remarkably linear western margin of the South
China Sea basin (Figs 1 and 7).
Kinematic: Cretaceous-Eocene
At the transition from the Mesozoic into the
Cainozoic, the regional stress field for onshore
Peninsular Malaysia consisted of a maximum
horizontal stress orientated between E - W and
ENE-WSW. This regional compression caused
strike-slip motions in sinistral sense along WNW,
NW and NNW major faults, and in dextral sense
along faults striking between N-S and ENE-
WNW (Tjia 1972b, 1989). The Malay Dome, a late
Cretaceous crustal dome of c. 1000 km diameter,
developed near the present triple junction of the
Malay, Penyu and West Natuna basins. These
basins are interpreted to have developed as rift arms
on that dome. Figure 8 shows the offshore faults in
the northern Sunda Shelf that may have been active
during this period and the strain ellipses illustrate
how the regional compression direction was
NE-SW in the northeast and changed gradually to
an ENE-WSW orientation in the southwest.
Kinematics: Middle Eocene-
Middle Oligocene
The stress fields in the domains bordered by the
major wrench faults are schematically shown by
strain ellipses on Fig. 9. Opposed slips on sub-
parallel wrench faults suggest that the fault-
bounded crustal blocks moved in directions indi-
cated by the large arrows. The differential block
movements are interpreted to have resulted from
continued indentation of the Indian plate into Asia
301
combined with interference by local obstacles
blocking or retarding certain block movements.
The obstacles could consist of massifs or of faults
locking where fault strikes changed. The Axial
Malay fault is interpreted to exist in the deep base-
ment in the Central Malay basin and to represent an
extension of the Three Pagodas fault. Sinistral slip
along the Axial Malay fault created roughly east-
west orientated half-graben and graben in the
pre-mid Eocene basement. The Vietnam Shear
most probably moved dextrally as a result of N-S
spreading of the South China Sea basin. North-
south faults bordering the western Nam Con Son
basins may have moved in the same sense.
Subsidiary structures generated by these move-
ments could be the en echelon NE-trending sub-
basins in the Nam Con Son basins (see also Fig. 7).
Kinematics: end Miocene
By the beginning of the Middle Miocene (17 Ma;
Taylor & Hayes 1983) the South China Sea basin
stopped spreading. As a result, the authors suggest
that from then on the westward convergence of the
Pacific plate was able to influence the Sunda Shelf
region. By the end of the Miocene, the continental
crust of the Indian-Australian plate had moved
closer to continental SE Asia and thus also began
to form an important barrier to southeastward
extrusion of the crustal blocks. These external
barriers combined with internal obstacles (massifs,
locking faults) most probably caused slip reversals
along several major wrench faults. The Ping-Teris
fault and the Western Hinge-line fault became left-
lateral, and the Axial Malay fault became right-
lateral. The Mae Ping Fault may have become
left-lateral again through the convergence of the
Pacific plate. The stress fields consistent with these
wrench movements and the directions of block
extrusions are shown on Fig. 10. Slip reversal along
the Axial Malay fault caused folding of the sedi-
ment fills of the east-west half-graben into east-
west anticlines and probably also dextral motions
along the Dulang and Mesah faults.
Conclusions
During the Late Cretaceous a hot mantle plume
is interpreted to have risen, centred at the triple
junction of the Malay-Penyu-West Natuna basins.
This formed the Malay Dome, with a diameter of
c. 1000 km, of which relics are represented by
relatively high regions of basement rocks such as
most of peninsular Malaysia, Natuna Islands and
Con Son Swell. The Malay-Penyu-West Natuna
basins and perhaps also the western Nam Con Son
basin (see Fig. 1) began to develop as aulacogens.
302 H.D. TJIA • K. K. LIEW

lU .':'-k !

• . . . • . -o
J ,,, Strike-slip fault 10"--
Strain ellipse

VIETNAM

5 8'

~o Z
" " ;:'2"":'" ". 0

4_ N : ! " "': 0
I
100
I
200
Ikm
6"

WEST '
NA T U NA

IlIiI%~IIN..,,~ a°t'I 11'

~,9 ~ 11 ... 1..
Fig. 8. Kinematics in the northern Sunda Shelf during the Cretaceous to Eocene. Strike-slip faults onshore peninsular
Malaysia were determined through fieldwork; the offshore faults are discussed in the text. Strain ellipses show
gradual change of regional compression (parallel to the short axes of the ellipses): in the northeast striking NE-SW
and in the southwest trending ENE-WSW. The orientations of strain ellipses have been determined by aligning the
short (parallel to compressional stress) ellipse axis at about 45 ° to the adjacent wrench fault and taking the slip sense
into consideration. Inset shows the Mekong basin and part of the Nam Con Son basin. Faults are: KL, Kuala
Lumpur; B, Bukit Tinggi; BA, Baubak; BB, Bentong-Bengkalis Suture; L, Lebir; P, Ping-Teris; M, Mersing zones;
MO, Ma' Okil; D, Dungun; Du, Dulang; Me, Mesah.

The origin of the other basins, Mekong, eastern
N a m Con Son and East Natuna basins, is not yet
clear. The presence of thick sediments in these
basins indicates that tensional stresses caused
basins to form in the early Oligocene or earlier in
the area of the M e k o n g and East Natuna basins, and
by the early Miocene also in the N a m Con Son area.
The middle Eocene collision of the Indian plate
with the Eurasian plate began p u s h i n g out
elongated crustal slabs of continental SE Asia
 STRESS FIELD CHANGES, NORTHERN SUNDA SHELF 303

I I

Strike slip faul t

G Strain ellipse

r, :.'.:- (:i:i;'...
• .,." •
, ~'." , . '

!.i: ::::
....4"-
NA 1"::i;.':",
NATUNA ~ \'~"
I. ""

0 100 20O
I I Ikrn
104" 106'
,11 I I
Fig. 9. Kinematics between the Mid-Eocene to Mid-Oligocene in the northern Sunda Shelf. Faults: WHL, Western
Hinge-line; DR, Dungun-Rumbia. Graben: Dg, Dungun; Tb, Tok Bidan. The shorter strain ellipse axis is parallel to
horizontal compressional stress, consistent with sense of wrench motion on the adjacent fault. Dashed bold lines are
N-S faults interpreted from lineaments of the Nam Con Son basin. Half-tipped arrows indicate sense of wrenching.
Large arrows indicate direction of motion of fault-bounded crustal blocks. Basins are stippled.

towards southeast consistent with the Tapponnier
et al. (1982) model of extrusion tectonics. These
crustal blocks are bounded by NW-trending wrench
faults, of which the Mae Ping and the Three
Pagodas faults and their splays play important roles
in the tectonic development of the northern Sunda
Shelf region. Kinematic analysis of structures in
pre-upper Oligocene rocks of the basins shows that
until late-early Miocene, regional tensional stress
prevailed, and that the major faults and their splays
acted as wrench faults. Alternate crustal blocks
bounded by these faults moved in opposite
directions (Fig. 9). The Ping-Teris, Western Hinge-
line and Mae Ping faults had a dextral slip sense
whereas the Dungun-Rumbia and the Axial Malay
faults had a sinistral slip sense. Pull-apart basins
developed along the WHL, and east-west half-
graben formed in the Malay basin. The eastern part
of the northern Sunda Shelf was influenced by
dextral slip along the Vietnam Shear as result a
of north-south spreading of the South China Sea
basin. This major shear and other N-striking faults
on the shelf probably created the NE-trending
sub-basins in the Nam Con Son basin.
By the middle Miocene, structural inversion
began and the slip sense along wrench faults
304 H.D. TJIA • K. K. LIEW

J
Strike slip fault
S t r a i n ellipse

g e n c e at
,•Conver
.'-~';.-
.--' ".~.PLATE.....
-:. -. -.: ~:.'""
:'1 i'.":C

....
.':..,.........."
"i
. ;
,;: --: ...::. .~ : : • .... :~|::'..'~

?:,,.~...:..... ,-~.• -.... : .;,/ ;T/

•~.:'. :..." • ":~...;'.':
, , . : . . ' , . ' . . . -.;

WEST ,
NATUNA I

NATUNA I.
' 25' Rotation
of half groben
in eastern
Penyu Basin 0 IO0 2oo
I ,, ~ km
106"

Fig. 10. Kinematics at the end of the Miocene in the northern Sunda Shelf. Abbreviations as on Fig. 8 and 9. Other
faults: D, Dulang; Me, Mesah. Continued sinistral slip along the Dungun-Rumbai fault caused 25 ° clockwise rotation
of structures in the eastern Penyu basin. Half-tipped arrows indicate sense of wrenching. The stress fields consistent
with these wrench movements are represented by strain ellipses. Shorter axes of strain ellipses are parallel to
compressive stress; large arrows indicate directions of motions of fault-bounded crustal blocks.

reversed. We propose that the change into a
general compressional stress regime for the region
was the result of convergence by the Indian-
Australian plate from the south, by the Pacific plate
from the east, and by continued indentation of the
Eurasian plate onto the Indian plate. The latter had
continuously pushed out SE Asian crustal blocks
towards the southeast, but by middle Miocene time
this tendency became impaired. First, the Indian-
Australian plate had reached a proximal position
with respect to SE Asia. Second, before middle
Miocene, the influence of westward convergence of
the Pacific plate was absorbed or greatly reduced
by north-south spreading of the West Philippine
basin (Eocene) and by that of the South China Sea
basin (mid Oligocene-early Miocene). Reversal to
dextral slip on the Axial Malay fault compressed
the basin-filling sediments into large east-west
anticlines, whose position and orientation were
fixed by the pattern of half-graben. Sinistral slip
along the W H L compressed the sedimentary fills
of the pull-apart basins into anticlines trending sub-
parallel to the long axes of the pull-apart basins.
The crustal block bounded by the Dungun-Rumbai
and Axial Malay faults together with the crustal
block to the northeast of the Mae Ping fault moved
 STRESS FIELD CHANGES, NORTHERN SUNDA SHELF
NW-ward, while the other crustal blocks were
extruded SE-ward (Fig. 10). After the Miocene, the
development of shallow crestal faults orientated
normal to the anticlines probably attests to further
warping of the structures. This deformation may
be a tectonic process or may represent differential
isostatic adjustment in the Malay basin.
A major outstanding problem is the timing of
wrenching along the northerly trending Dulang,
Bundi and Mesah faults. In the Malay basin, dextral
offsets of folds associated with hydrocarbon fields
suggest post-folding movement. The youngest
folded sediments are upper Miocene. But
colleagues from EPMI have shown that a north-
south fault which has not appreciably displaced
References
ANIZAN ISAHAK. 1993. Geomorphology and soils of
Southeast Johore. PhD Thesis, Universiti
Kebangsaan Malaysia, Bangi.
ARESHEV, E. G., TRAN LE DONG, NGO THUONG SAN
& SHNtP, O. A. 1992. Reservoirs in fractured
basement on the continental shelf of southern
Vietnam. Journal of Petroleum Geology, 15,
451464.
CCOE 1991. Total sedimentary isopach maps offshore
East Asia. (6 maps with explanatory text) CCOP
Technical Bulletin, 23.
DAINES, S. R. 1985. Structural history of the West Natuna
basin and the tectonic evolution of the Sunda region.
In: Proceedings of the Indonesian Petroleum
Association 14th Annual Convention, Jakarta.
39-61.
FONTAINE, H. & WORKMAN,D. R. 1978. Review of the
geology and mineral resources of Kampuchea, Laos
and Vietnam. In: Geology and Mineral Resources of
Southeast Asia, 3rd GEOSEA, Bangkok, Thailand.
538-603.
FORTUrN, A. R. & DE SMET, M. E. M.1991. Rates and
magnitudes of Late Cenozoic vertical movements
in the Indonesian Banda Arcs and the distinction
of eustatic effects. In: MACDONALD, D. (ed.)
Sedimentation, Tectonics and Eustasy: Sea-level
Changes at Active Margins. International
Association of Sedimentologists, Special
Publication, 12, 77-89.
HARDER, S. H., MCCABE, R. J. & FLOWER,M. E J. 1992.
A single mechanism for Cenozoic extension in
and around Indonesia. In: Symposium on Tectonic
Framework and Energy Resources of the Western
Margin of the Pacific Basin, Kuala Lumpur
Nov-Dec 1992. Programme & Abstracts. 34.
KHALID NGAH. 1975. Stratigraphic and Structural
Analyses of the Penyu Basin, Malaysia. MSc Thesis,
Oklahoma State University.
--, MADON, M. & TJ~A, H. D. 1996. Role of pre-
Tertiary fractures in formation and development
of the Malay and Penyu basins. This volume.
LIEW KIT KONG. 1993. Structural development at the
west-central margin of the Malay basin (basement
of Blocks PM 2 and PM 7). Geological Society of
305
sediments is as old as middle Miocene (Md. Nor
Mansor & K.W. Rudolph at the Petroleum Geology
Seminar, Geological Society of Malaysia in Kuala
Lumpur, 7 & 8 December 1993).
Most of the unpublished information referred to in this
article was obtained from PETRONAS Exploration
Management Department. We thank the managements
of PETRONAS EMD and PETRONAS Research and
Scientific Services for permission to publish this article.
Thorough reviews of an earlier version by R. W. Murphy
and B. Clennell are very much appreciated and led to this
rewritten paper. A travel grant from the organizers of
the SE Asia Tectonic Evolution Conference enabled Tjia
to participate in the meeting. Liew's attendance was
supported by PRSS.
Malaysia, Annual Geological Conference 1993,
Warta Geologi, 19, 131-132.
MATTHEWS, S. J. & TODD, S. E 1993. A tectono-
stratigraphic model for the southern Nam Con Son
basin, offshore Vietnam. In: Geological Society of
Malaysia, Petroleum Geology Seminar 1993,
Programme and Abstracts. 22-23.
MAURI, S., HARDER, S. & MCCABE, R. 1993. Gravity
modelling of Cenozoic extensional basins, offshore
Vietnam. In: FLOWER, M. E J. et al. (eds) 1992
Report to the Indochina Research Consortium
(January 1993).
MD. NAZR1 RAMLI. 1988. Stratigraphy and paleofacies
development of Carigali's operating areas in the
Malay basin, South China Sea. Bulletin Geological
Society of Malaysia, 22, 153-187.
MOHD. FIRDAUS ABDUL HALIM. 1993. Heat flow in the
Malaysia sedimentary basins. Petronas Research
& Scientific Services Sdn. Bhd., Proceedings
Exploration Research Seminar. 1, Progress Reports
and Findings. 11-25.
SUJANTO, E X., HARTOYO, P. • SUGIHARTO, H. 1986.
Hydrocarbon geology of producing basins in
Indonesia and future exploration for stratigraphic
traps. In: CCOP Technical Publication 17/ASCOPE
Technical Publication, 7, 5-24.
TAPPONNIER, P., PELTZER,G., LEDAIN, A., ARMIJO, R. &
COBBOLD,P. 1982. Propagating extrusion tectonics
in Asia: new insights from simple experiments with
plasticine. Geology, 10, 611-616.
TAYLOR, B. & HAVES, D. E. 1983. Origin and history of
the South China basin. In: HAYES, D. E. (ed.) The
Tectonic and Geologic Evolution of South-east
Asian Seas and Islands, Part 2. American
Geophysical Union Monograph, 27, 23-56.
TJIA, H. D. 1972a. Fault movement, reoriented stress
field and subsidiary structures. Pacific Geology, 5,
49-70.
1972b. Strike-slip faults in West Malaysia. In:
International Geological Congress, 24th Session,
Montreal, Section 3. 255-262.
1989. Major faults of Peninsular Malaysia on
remotely-sensed images. Sains Malaysiana, 18,
101-114.
 306 n . D . TJIA & K. K. LIEW

- - 1994. Origin and tectonic development of Malay-
Penyu-West Natuna basins. In: PETRONAS
Research & Scientific Services Research for
Business Excellence Seminar, Kuala Lumpur, June
20-21, 1994.
1995. Inversion tectonics in the Malay basin:
evidence and timing of events. Geological Society
of Malaysia Bulletin, in press.
, FUJII, S., KIGOSHI, K. & SUGIMURA,A. 1974. Late
Quaternary uplift in eastern Indonesia.
Tectonophysics, 23, 427-433.
WONGSOSANTIKO, A. & WIROJUDO, G. K. 1984. Tertiary
tectonic evolution and related hydrocarbon
potential in the Natuna area. In: Proceedings
Indonesian Petroleum Association 13th Annual
Convention, Jakarta. 161-183.
 Far-field and gravity tectonics in Miocene basins
of Sabah, Malaysia
BEN CLENNELL
Dept of Earth Sciences, University of Leeds, Leeds LS2 9JT, UK

Abstract: At the end of the early Miocene, at c. 17 Ma, oceanic spreading ceased in the South
China Sea Basin, as a series of collisions were initiated between continental blocks derived from
the Asian mainland and the northwestern margins of Borneo and Palawan. In the NW Sabah
Basin this tectonic event caused a major period of uplift and erosion which produced the so-called
'Deep Regional Unconformity'. While these mainly compressional events were occurring at
the southern margin of the South China Sea, the Sulu Sea was undergoing extension, with rifting
in the northwest and oceanic spreading in the southeast.
Recent seismic surveys in the Sabah offshore area show that the Deep Regional Unconformity
can be traced through from the South China Sea into the Sulu Sea as a boundary between two
contrasting depositional systems that are both represented in sedimentary sections onshore
eastern Sabah. Eastern Sabah changed from an environment of deep marine clastic deposition
in the Oligocene and early Miocene, to shallow marine and terrestrial sedimentation in the mid
to late Miocene, with a major period of sedimentary mrlange formation occurring at the time of
the Deep Regional Unconformity.
Inversion of the Miocene sequence in eastern Sabah appears to be limited to the edges of
basement blocks, which were moved by far-field tectonic stresses. Post Middle-Miocene basin
evolution in the onshore (Central Sabah Basin) and adjacent parts of the Sulu Sea offshore
(Sandakan Basin) has been strongly influenced by mud diapirism and gravitational sagging of
progradational sand-rich sediments into underlying muds and mrlange units. These sequences
appear to have been decoupled from later tectonic events by the great thickness of underlying
muds and mrlanges.

SE Asia has been an area of diverse and rapidly
changing tectonic and sedimentary environments
throughout the Tertiary. It is generally agreed that
large-scale plate motions provided the ultimate
driving force of the main periods of Tertiary basin
formation in SE Asia (e.g. Hamilton 1979;
Hutchison 1989). For this reason, major tectonic
events, such as phases of rifting or inversion, may
be expressed synchronously in several basins
across the region. Nevertheless, the timing and
kinematics of basement movements vary from
basin to basin. Within individual basins, the effects
of far-field tectonic forces on the sedimentary
architecture are augmented by gravity tectonics and
autocyclic sedimentary processes. Only rarely can
eustatic signals be resolved amid the noise of these
competing processes, so that classical sequence
stratigraphy can only play a limited role in the
understanding of SE Asian basins.
Sabah is a state of Malaysia that lies in North
Borneo, at the geological junction between the
cratonic core of Sundaland, the Sulu, Celebes and
South China Sea marginal basins, and the volcanic
arc systems of the Philippines. This paper focuses
only on one time interval in the geological evolu-
tion of this complex region; namely the Early
to Middle Miocene transition. A simplified inter-
pretation of the Early Miocene tectonics of the
Sabah region is shown in Fig. 1.
Areas discussed here are the Northwest Borneo
Basin, the Southwest Sulu Sea Basin, and its onland
extension into eastern Sabah, known as the Central
Sabah Basin (Hutchison 1992). The offshore areas
are k n o w n from recent deep-seismic surveys
(Hinz et al. 1989, 1991), and commercial seismic
surveys and drilling for petroleum exploration and
production. The data from onland Sabah come
principally from road cuttings associated with a
major road improvement scheme that traversed
eastem Sabah from 1988 to 1991 (Clennell 1992).
Western and northern Sabah and adjacent
parts of the South China Sea
Proceeding from N W to SE, the west of Sabah can
be divided into several structural provinces,
originally delineated by Hinz et al. (1989) and
Hazebroek & Tan (1993); these are shown in a
modified form in Fig. 2. A simplified stratigraphy
of Sabah is shown in Fig. 3.
The outermost part of the northwestern Sabah
margin is floored by several blocks of attenuated

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, 307
Geological Society Special Publication No. 106, pp. 307-320.
308 B. CLE~'mELL

Fig. 1. Suggested Early Miocene (18 Ma) tectonics of the Sabah region. The following should be noted: active
submarine and subaerial vulcanicity along the Cagayan Ridge extending into Sabah in the Sandakan area; cessation
of spreading in the South China Sea Basin is linked to onset of compression in Northwest Borneo Basin, Western
Sabah, Northwest Sulu Basin and Palawan; incipient spreading in the Southeast Sulu Sea, and associated rift-basin
initiation in the Central Sabah Basin. The geological evidence suggests compression and extension were underway
synchronously in the same region, while between these domains strike slip zones are postulated between the major
lithospheric blocks (positions of inter-block transform faults are guides only). Abbreviations: NWSS, Northwest Sulu
Sea; SESS, Southeast Sulu Sea; RB, Reed Bank; NP, North Palawan; SP, South Palawan; CAG, Cagayan Ridge;
MI, Miri Platform; LS, Luconia Shoals; DG, Dangerous Grounds; ZG, Zamboanga; NWBB, Northwest Borneo
Basin; PA, Panay; LU, Luzon; BSF, Balabac Strait Fault (postulated); WSSB, West Sulu Sea Basins; CSB, Central
Sabah Basin. Data from Daly et al. (1991); Rangin et al. (1990a, b, c, 1991); Mascle & Biscarrat (1979).

continental lithosphere, which separated from the
south Asian mainland during Palaeocene rifting and
moved southwards as oceanic spreading occurred
in the South China Sea from c. 4 5 - 1 7 M a
(Holloway 1982; Ru & Piggott 1986; Briais et al.
1989; Ben-Avraham 1989). The continental frag-
ment adjacent to Sabah is known variously as the
Dangerous Grounds Block or the Northwest Sabah
Platform (Zone I in Fig. 2).
The NW Borneo-Palawan trough (Zone II) is a
SE-facing half-graben at the edge of the NW Sabah
platform that is overridden to the SE by shortened
sediments of Neogene age (Zone III). Zone III is
variously interpreted as the compressive toe of a
subsided delta (e.g. Tan & Lamy 1990; Hazebroek
& Tan 1993, in analogy with the Niger Delta
described by Damuth 1994), or a tectonic thrust
system in its own right (Hinz et al.. 1989). Details
of the structure suggest that the former inter-
pretation applies, at least to the south, where Zone
III passes gradually into the East Baram Delta
proper (Zone VI). To the north in Zone IV, the
lower compressive belt is overridden by a major
thrust sheet system of mainly Palaeogene sediments
that can be traced onland as the Crocker Formation
and equivalents (Fig. 3).
The D e e p Regional Unconformity
Figure 4 (A-A' on Fig. 2) is an interpretation of the
seismic line BGR 86-06 (Hinz et al. 1989) showing
the relationships between these two compressive
belts. The top of the upper thrust slice is marked
by a major unconformity, known as Horizon 'C' by
Hinz et al. (1989, 1991), that is believed to mark
the time of collision of the Dangerous Grounds
 MIOCENE TECTONICS OF SABAH, MALAYSIA 309

115° }116° I117°

Balabac
IVa ~" Strait
/
J

VIII t/.= B~ggi

7°
KUDAT

/~ V,,J">-'-'-""
• SOUTH; ~
I~ CHINA / p-Y'" /
fz_
,,~ S E A'~,,~ .t~-" KOTABELUD'
•
r" I _( ,,I'Xb' "-{
VII Mt. Kinabalu

.-~ I, : •"r~:" ..'KOTA'"

~i "~ "-.~I[I~, RANAU ~il~'-j,,~.
~, / :KINABALU • \ I~l
.. #.f,,,"
/ .,
4
/ ,
yr t
Baramg. Kiias
Delta Labu~. f
, ,,: ..... ,, , ,

,~.......
i ,,)~/&:/ .... .. .... ..,,~..- 1~ -.:~--
t S :"::..--/
50 km i
5° l BRUNEk .. ,, ,i
BAY
- I 115 ° °,, ,
KEY STRUCTURAL PROVINCES
......... Major ..,I~ " Major
lineament Overlhrust I II III
N.W. Sabah Platform NW Borneo- Palawan Lower Thrust Sheets &
M~langes Synformal "~.% Reverse
(Dangerous Grounds) Trough (Sabah Trough) Compressive Delta Toe
axis t~, fault IV V VI
""'x" Major Thrust Sheet RedeformedMajor Thrust East Baram Delta:
""'x" •.~:~.~ Strike- System (Imbricated Sheet System pierced by growth faulting, diapirs
Ophiolite Palaeogene & ? Ophiolite) mud dia.oirs wrench tectonics
slip
Anti formal
~....~... axis .. • Normal VII VIII IX
Inboard Belt: reverse Banggi-Belabac Zone: Crocker Belt (Palaeogene
Granitoids " fault and wrench faulting thrust/wrench oomplex accreted complex)

......... Structural Province boundary X Central Sabah Basin (onland extension of Sotu Sea Basin)

Fig. 2. Tectonic map of western Sabah and adjacent offshore area. Onland data from Lim & Yin (1985), B6nard et al.
(1990) and Tongkul (1991). Offshore data from Hinz et al. (1989) and Rice-Oxley (1991).

platform with northern Borneo. This unconformity belt by listric extensional detachments (Levell &
is extensive, and is more usually referred to as the Kasumajaya 1985), and the entire sequence of
'Deep Regional Unconformity' (Bol & Van Hoorn structural units is pierced by NE-trending walls
1980; Levell 1987). of diapiric shale. In the south, the sediments of
While the nature and origin of this unconformity zone VII and slivers of the underlying basement
can be appreciated in sections such as Fig. 4, it have been buckled upwards to form Labuan island
becomes more deformed, and less easy to interpret, and the Klias Peninsula, where mud-cored folds
further landwards. The overprint to the west in and mud volcanoes are found (Leichti et al. 1960;
Zone IVa is predominantly compressional, while Wilford 1967; and personal observations). Trans-
later extensional structures are dominant in Zone pression has affected this zone, inboard parts of
IVb. To the far north, in Zone V, the overlying Zone IVb and V, and is the dominant structural
sediments are decoupled from the underlying thrust style in Zone VIII.
310 B. CLENNELL

Time N.W. Sabah Offshore W. & N. Sabah E. Sabah & W. Sulu Sea Basins
units N W Outboard Belt Inboard Belt S E [ W SandakanBasin S.E. Sabah
I
E

~ ~ ' . ~ /'.~ _ ~ . ~ G a n d u m a n & T o g o p i Fms.

[:'::::-"-.-~~ it,,, [ - i" ~ ueep I-legional /" i -- i ~ i-
,,, u. f : : : : : : - ~ : i : : : : : \ fC::k Unconformity/ /'~;~:,,~:~
z I [ ! - - 3 - : ' : : ~ ~ Horizon"C",~ ~ ~ ~ . . - . . ' . ' . ' . . ~
°1
0
~~_::.:a..
M. ============================= 0 ~¢/
:::::: : : : : : :""r:":["::" " !" :" :" :" :"'"

,,, | :"
Z I U. Youn er shalowmarnecastcs£_ ' :,i<.=": KS".:', ;i..";~.
uJ ~ nc ~andakan Fm& Tanjong I--m "::: :::.; ... .":::: :"!i..:..:
Ne ene and Ouatemary: ......',!::::::::::::::::::: "

"o8| M ~ :::!ii::>:::
,,, | L .

, ,^ ~ Crocker Formation and equivalents:

~ [~ ~',~%,1 Palaeogenedeep marine dastics ..............................
,~ 0 ~ Cretaceousophiolite and related ~ ~ C " ~ _ _ ~

Fig. 3. Simplified stratigraphic section across Sabah from NW to SE. Modified from Tan & Lamy (1990) and
Clennell (1992).

NORTHWEST SOUTHEAST P
A A
~--II ~I-
~' III J" IVa
1- TROUGH COMPRESSIONAL TOE SLOPE: EXTENDING COVER OVERLIES
BURIED THUST BELT
2 m
_ Neogene~
- ~ ~ , e , , ' ' , ~ - A
Plio-Pleistocene ,~ _ ~ \ "" ^
3-
Seafloor ......., ...... ", ; . . ......... .~¢~.o "."%%%. :
: i,"./.-"?"" ............ " C " . . " . . ". " ""
-- A ~ ....... "---=., ~ , ~ " ....... " - , ~ : x ",, Major Thrust System (Lower Tertiary
t--; 5 _ ..: j:..."~; . . . . . . . . . ",,, o ~'~,. "'-,., ",, "Crocker"sediments & ? Ophiolite)

• ':" ,'6
" J " " --.'~
"" . . . . . .". Lower Thrust" Shee
........... - " - . . . . . . . . . . . . . . . . . . . . . . . . . .
6- ~ ? L.-M. Miocene . . . . . . . . . . t System (Neo-en ..
"" ~ • sea~ments)
D'-'-~ - " " ~ -- D ~
7-- I 0 kn'l Northwest Sabah Platform: rifted continentalmargin
[ ] basement and ? Mesozoicto Palaeogenesediment

Fig, 4. Structural cross-section A-A" across northwest Sabah Margin, re-drawn from unmigrated deep seismic line
BGR 86-06 in Hinz et aL (1989). Vertical exaggeration is c. 10 x. Horizon 'C', at c. 17 Ma, corresponds to the end
of an early Miocene compressional phase in the South China Sea. Abbreviations: A, Early Pliocene unconformity;
B, Late Miocene uncf.-Shallow Regional Unconformity of Levell (1987); C, Latest Early Miocene uncf. (Deep
Regional Unconformity); D, top NW Sabah Platform (?Oligocene to L. Miocene carbonate); 5, Miocene/Pliocene
boundary; 6, Base Upper Miocene.
 MIOCENE TECTONICS OF SABAH, MALAYSIA
Onland and regional structure
of western Borneo
The onland structure in the west of Sabah is
dominated by an uplifted belt of folded and faulted
turbiditic sediments (comprising Crocker, Sapulut
and Trusmadi Formations) believed to have been
accreted to Borneo since early Tertiary times,
during some form of plate convergence, (Hamilton
1979; Hutchison 1988, i989; Rangin et al 1990b;
B6nard et al. 1990). Tongkul (1987) presents a
detailed structural and sedimentological model
explaining the Crocker Formation of the Kota
Kinabalu area as a packet of trench sediments
accreted to northern Borneo in Palaeogene to early
Miocene times during subduction of an oceanic
slab ('Proto-South China Sea'; Fig. 1). It remains
unclear how much oceanic lithosphere, if any, was
subducted during this accretionary/compressive
episode (compare Holloway 1982; Taylor & Hayes
1983; Hutchison 1988; Rangin et al. 1990b;
Clennell 1992) but field outcrops in the Mount
Kinabalu and Telupid areas suggest that the
Crocker sediments are floored by ophiolitic
basement.
Certainly, the Neogene history of the North
Borneo and Palawan area has involved only
collision of continental material rifted from south
Asia rather than subduction of any oceanic litho-
sphere. The rifled continental material is divided
into a series of blocks by what may be through-
going structures linking to transforms at the
South China Sea spreading centre (Tongkul 1990,
1994). Although the trajectories of the individual
blocks were broadly from the north, channelled
by the lineaments, they were slightly divergent in
rate and direction. The main period of convergence
at the Borneo margin occurred during the Early to
Middle Miocene, at which time the Palaeogene
sediments were uplifted to form the mountains
of the Crocker Range, which reach heights of
1-2 km above sea-level, even in their present
eroded state.
The Deep Regional Unconformity (DRU) is not
a time-line of regional extent, but rather marks the
period during which collision had the maximum
impact on the sedimentary record in a particular
place. In some areas, there is a marked angular
discordance and a considerable section of missing
stratigraphy, while in other places, the DRU is
represented by a brief hiatus. The controls on this
pattern depend not only upon local rates and
degrees of tectonic uplift, but also on tectonic and
eustatic events that influenced the source areas and
distribution paths of the sediments (Tan & Lamy
1990; Rice-Oxley 1991). Thus the architecture of
the unconformity depends on the combined effects
311
of erosion and sediment dispersal patterns. There
was a second culmination in convergence in the
mid Late Miocene which buckled the unconformity
surface in many parts of Zones III and IV
(Hazebroek & Tan 1993).
The Central Sabah-Western Sulu Sea
basin system
According to Hutchison (1992) the Central Sabah
Basin can be regarded as an onland extension of
the Southeast Sulu Sea Basin (Fig. 5). Results of
scientific drilling (Nichols et al. 1990; Rangin &
Silver 1990a, b) suggest that the Southeast Sulu
Sea Basin was initiated in latest Oligocene or
earliest Miocene time, by rifting of a pre-existing
island arc and ophiolitic terrain (Fig. 2) whose
onland extensions are now exposed in the Upper
Segama area of Sabah, and in the Philippine island
of Panay (Hinz et al. 1991).
Subsequently, a limited amount of oceanic
spreading occurred along a NE-SW axis, opening
up a marginal basin bounded to the north by a
rifted and subsided arc (Cagayan Ridge) and
ophiolite terrain (the Northwest Sulu Sea Basin and
Palawan Island), and to the south by the volcanic
arc of the Sulu islands. The close proximity of the
pole of rotation to the marginal basin meant that
eastern Sabah lay at the hinge zone of the spreading
(Hutchison 1992). Lithospheric stretching near this
pole was presumably not sufficient to generate new
oceanic crust (White & McKenzie 1989), but there
is scattered evidence of basaltic magmatism along
the axis of the Central Sabah basin in the relevant
time interval, i.e. from the late Oligocene to the
middle of the Miocene (Kirk 1968; Tamesis 1990;
Clennell 1992). It is evident that the eastern Sabah
to southeastern Sulu Sea region experienced
mainly NW-SE extension from Early to Middle
Miocene time.
At some time in the late Miocene, opening of the
Southeast Sulu Sea Basin was extinguished by the
subduction of the spreading ridge southwards and
eastwards beneath the Sulu Archipelago, and the
larger Philippine islands of Negros and Panay
(Hinz & Block 1990; Hinz et al. 1991). At this time
the tectonic development of eastern Sabah became
disconnected from that of the Sulu Sea, probably
by the growth of a transform fault system running
from the Sulu archipelago towards the Balabac
Strait (Hinz et al. 1991). The geometry and
kinematic linkage of this system remains obscure,
but wrench faulting has apparently affected a series
of upper Neogene basins running from the Balabac
Basin (Beddoes 1976) through to the Sandakan
Basin (Fig. 5).
312 B. CLENNELL

D ~ BALABAC 1118° t
I 119° E
/
KEY /
SEDIMENTS
~ ] Sandakan & D'
Tanjong BALABA C i/
Formations STRAIT , .... /
,;
/
Kulapis Formation /
~7~ Labang Formation

IGNEOUS ROCKS
[ ~ Granitoids Miocene
~; anggi I. ........ ....

/
""--....,.

/
",,.
/
t
/

to BANCAUAN /
Volcanics Quaternary BASIN /
I ~ Late Oligocene to /
Middle Miocene
pyroclastics and lavas
I'-,, I ",~"" I
I ~1",,. "" .I
OTHER UNITS i. "-::.2-, ,,,,,.
• /
~l~ Earlyto Middle ...,S A 3
Miocene melange units
Manlang
~ j Ophioliticbasement
and oceanic cover
sediments SANDAKAN
B, i I
1

Neogene oceanic / J
lithosphere f

, 7 Cf j /
/
.---" Lineament :. C. E .N T. R. A.L " . .•. . .,,"
. . Bukit
.. Garam
.~,~
// /" /
I /,, ~ ' , ~ "v" ",,~::: 5:-::-:-~:;::G .:::.- <,'-" • ~_-" " l~ ~t'~- - . .~.. . .-. . . , . . .'

5°
m

".... ~. I , , , . . . ,,),IBASIN . DARVEL BAY

""~':""::~":::'"":" ~" ~ ~ C E L E B E S
• ,~ ;:((I Bp!n :::I ~ O S EA
,/ " ~¢^^^~^%~2- -',^^^,
" "..........
~ it'*- i~tTAWAU
^ ' ^ ' ^ ~ ~ - - ' v~ SEMPORNA

I KALIMANTAN [ 117° ~ ' J , "~~ - ~ ' ~~"~ 1 ~1118°

'~ [ 119°E
Fig. 5. Simplified geological map eastern of Sabah, based on Lim & Yin 1985, Tongkul (1991, 1993) and Clennell
(1991, 1992). Offshore information from Hinz & Block (1990), Hinz et al. (1991) and Wong (1993).

Earlier sediments in eastern Sabah, The Labang Formation (Collenette 1965)

the Labang and Kulapis
In the Central Sabah Basin, Oligocene to Lower
Miocene sedimentation produced two contrasting
units known as the Labang and Kulapis formations.
The relationships between these units are largely
obscured by later deformation but their general
nature can be assessed from the available
exposures.
consists of grey-coloured deep marine turbidites
formations
(full Bouma a-e sequences, and c--d-e sequences
are commonplace) and interbedded hemi-pelagic
shales. The entire Labang unit is deformed by
folding and faulting, but even so, a lower shaly part,
and an upper part with thicker and more abundant
sandstones, can be distinguished. The foraminiferal
and trace fossil assemblages both suggest that the
sediments were deposited in a bathyal environment.
 MIOCENE TECTONICS OF SABAH, MALAYSIA
The Kulapis Formation (Collenette 1965) is a
particularly unusual unit, as it consists of pre-
dominantly red-coloured shales and sandstones that
can be shown by occasional, well preserved
foraminiferal and nannofossil assemblages to have
been deposited in a fully marine environment
(L. Gallagher pers. comm. 1991). The most distinc-
tive facies consists of thickly bedded sandstones,
up to 10 m thick, that occasionally show graded
bedding and convoluted flute marks at the base, but
more usually are structureless, and may have been
deposited by large grain-flows. Elsewhere the
Kulapis Formation consists of thick mudstones
with packets of more thinly bedded sandstones and
siltstones that show normal grading and flute
marks. The facies associations and fauna together
suggest that the Kulapis Formation was deposited
in a marine outer shelf environment with unstable
slopes.
East Sabah melanges
A remarkable feature of eastern Sabah is the great
extent of mud-matrix melange units and associated
chaotic rocks (dark stipple on Fig. 5). The source
of the melange matrix, and most of the sandstone
blocks, can be traced to disaggregated Labang
Formation sediments (the grey matrix melanges).
Subordinate volumes of melange and broken
formation are derived from slumping of Kulapis
Formation sediments (red matrix melanges), and
these include large intact rafts of the massively
bedded sandstones. There is a small amount of
ophiolitic material within the melanges, and to the
south of Sabah, both the matrix and blocks are
sourced increasingly by volcaniclastics.
The nature, origin and tectonic significance of
the East Sabah melanges have been described in
considerable detail by the author (Clennell 1991,
1992), and it was concluded that melanges were
mostly produced by submarine slope failures that
were triggered by tectonic rearrangement of the
Central Sabah Basin at the end of the Early
Miocene. Parts of the melanges are demonstrably
of olistostromal origin, such as those associated
with slump folds. Large volumes of weakly
foliated scaly clay melange may have originated
as submarine debris flows. Tracts of the melange
which have a broadly horizontal fabric of more
intense shearing, marked by zones of scaly clay and
cataclasites, are believed to have been produced
by tectonic shearing and gravitational sliding.
Relatively rigid blocks of sand-rich Kulapis
Formation sediments slid over a substratum of
muddier Labang Formation sediments as the whole
sequence underwent stratal extension that detached
along melange horizons. A relatively small volume
of melange occupies piercement structures
313
emplaced into Middle Miocene and younger
sediments. These melanges are demonstrably of
diapiric origin, and are associated with effusive
mud volcanism which continues to the present day
in eastern Sabah and in the Sandakan Basin.
Successor basins, the Tanjong
and Sandakan formations
Sedimentation in the Central Sabah Basin con-
tinued throughout the Miocene, with input of
quartz-rich sediments from the uplifting Crocker
Formation to the west and volcaniclastic material
from active magmatic arcs to the south. These
sediments spilled into the axial part of the Central
Sabah Basin, forming a system whose NE-SW
trend parallels that of the Sulu Sea. It is therefore
widely presumed that the sediments were funnelled
into rifted basins associated with extension and
spreading of the Sulu Sea (e.g. Hutchison 1988,
1992; Tjia et al. 1990). This appears to be the most
likely explanation of the sedimentation patterns,
but direct evidence of extensional structures onland
over much of eastern Sabah is lacking. Some
normal faults with the correct trend have been
traced in the field and in remote-sensing images by
Tongkul (1993), but he points out that many of the
larger NE-SW trending faults have a considerable
wrench component. The association with basaltic
magmatism and melange formation, and the trend
of the later basins and their later structural style,
provides indirect evidence that the successor basins
are rift-related.
In the south and central part of Sabah, the quartz-
rich sedimentary sequence is known as the Tanjong
Formation, and it consists of mainly shallow
marine facies with some thickly-bedded sandstones
intercalated with thinner-bedded sandstones, silt-
stones and shales. Coarsening-upwards and
thickening-upwards sequences are common,
occasionally culminating in terrestrial deposits,
including coals. In places the underlying Kulapis
Formation sediments grade upwards into the
Tanjong Formation, the red colour of the earlier
sediments gradually giving way to the grey and
yellow colours of the Tanjong. Where the Tanjong
Formation overlies the Labang Formation or the
melange units, there is generally a marked un-
conformity, examples of which can be seen on the
Kinabatangan River near Bukit Garam. The earliest
Tanjong sediments are upper Lower Miocene, and
extend up into the Middle Miocene. This means
that the Tanjong must have been deposited in
quiescent areas while melanges were still being
formed in other parts of the basin.
In the Sandakan Peninsula, the Tanjong
depositional system is represented by slightly
314 B. CLENNELL

WSW ENE ~
B -] ...... ~ - ' .....~. . . . . . . ~ . - ......._:: ' .--~---. --,-' ..........'...... ..~........ ~., =_-_.' .:.......~'....,-,~--,-::.~-x,,~,~_..--
. . . . . 1 ,=: ~ ..... ,.~-.--,~-:
.......... B
..... ,~,~, . - ...... _ ~ ..... ..,

1
:;"~4.'~',.~:~-~;:'~.~:~':_~-YZ:~ - " ...... ~'~';~-'~-~TW.~';~.'~*,~,~""4~.~ ' " ~ ' ~ . , ~ ~ : -"~:'~--~:-='-¢~°!!~:~:,~'=~'~'-~..~="

~. ~z~-~ -~ . ~ - - ~ , ~ ~

4 ~:...... ' ..... ~ _ ' Y ~ U - . ~ - - ~ . ~

(a)

WSW ENE
i -
B,
v

~ " Unit 3 ( P l i o - P l e i s t o c e n e ) ~ _ - -

1 --

.~ Unit lm(Base M i i ~ i o c ~ ~.1% _~__..~

¢9

3--

Lower Miocene shales, ~ 5 km

volcaniclastics and m61ange I |
4 m

(b)

younger sediments, whose oldest fossils date from

Subsidence, diapirism and circular basins
the early part of the Middle Miocene (Lee 1970;
Ujiie 1970, 1977; Clennell 1992). For this reason Tjia et al. (1990) suggested, on the basis of facies
they are known as the Sandakan Formation, but it and palaeocurrent correlations, that the Tanjong
is thought that the Tanjong and Sandakan were sedimentation was originally extensive throughout
probably contiguous at one time, as the minera- the Central Sabah Basin. From facies analysis, the
logical composition of the sands in the two pattern emerges of a prograding sequence of shore-
formations are petrographically very similar line and terrestrial deposits, which originated in
(Clennell 1992). The Sandakan Formation is every- the south and become younger to the NE (towards
where unconformable on older rocks: either the Sulu Sea).
volcaniclastics, tuffaceous sediments that can be The Tanjong deposits later became segmented
correlated with the Labang Formation, or melanges. during subsidence that produced a series of circular
The youngest fossils in these melanges are them- sub-basins, three of which (the Maliau, Malibau
selves of earliest Middle Miocene age, so the and Bukit Garam basins) are shown on Fig. 5. The
unconformity at the base of the Sandakan structure of these circular basins cannot be
Formation can be dated to the early Middle explained by a far-field stress with a single orien-
Miocene. tation of the principal components. Normal faults
 MIOCENE TECTONICS OF SABAH, MALAYSIA 315

WNW ESE
C :- C '

(c)
WNW ESE
C -C'
~ Unit 3 (Plio-Pleistocene) ~ -------~1 ~,.~
.--""""--~~ Unit2 ? . -----
~- ~-_.~-~'~ / \ -4. Q
1--
[..a
N ~ ~Unit 1 (Base MiddleMiocene J / /
\ ~ to mid UpperM i o c e n e ) j / ' / \ \Q.

~ 2 - - Lower Miocene ~N. ,,~"~ ~ . . ~ ~ Mud diapir

volcarnMi2~n: ~ -~ , ~
8
3- , 5 km i

LowerMioceneshalesand m~lange
4--
(d)
Fig. 6. (a, b) Excerpt of commercial seismic line SB6-27 orientated down dip across part of the Sandakan basin
immediately offshore the Kinabatangan estuary (section B-B' on Fig 5.). (c, d) Part of line SB6-01 orientated across
the strike of the Kinabatangan circular basin (section C-C' on Fig. 5). Unit l, a prograding wedge of Middle and
Upper Miocene sediments (Sandakan Formation equivalents and younger) builds out over thick wedge of shales and
?melanges which extrude diapirically at the basin edges. The overlying Plio-Pleistocene sequence (Unit 3) is
generally aggragational and shows onlap to the southwest. Unit 2, a second progradational sequence deposited
during a lower sea-level stand at the end of the Miocene, is preserved only in central parts of the basin. Stratigraphic
information from Wong (1993).

generally trend tangentially to the basin structure
and wrench faults are often orientated radially.
Tham (1984) and Lee & Tham (1989) explain
this as the result of differential subsidence of
neighbouring sectors of the basin. The basin edges
have tilted more than the basin interior, leading
to higher stratal dips at the periphery than at the
centre, where the sediments are almost flat lying.
There is evidence that some of this subsidence and
tilting was syn-sedimentary, as slump structures,
mud injections and convoluted flute marks attest to
rapid and unstable deposition (Lim 1990; Clennell
1992). Also, some fault block rotations generated
topographic highs within the Tanjong basin system
316 B. CLENNELL
that were exploited by reef-building organisms. The
subsidence is greatest where the Tanjong sediments
overly a great thickness of mud-rich sediments. For
example, the Bukit Baram basin has a central high
area where it overlies Kulapis Formation sedi-
ments, and two curved lobes to the NW, where the
Tanjong sediments have subsided into melanges,
and to the south where the Tanjong overlies the
Labang Formation.
To the north, the Sandakan Formation shows
some of the same features as the Tanjong basins.
The main outcrop on the Sandakan Peninsula is
founded on volcanic basement that has not
permitted great subsidence, but has extended to
form a series of gently tilted fault blocks. Outlying
basins have mostly been eroded, but around the
Manjang river a small circular basin of Sandakan
Formation sediments can be identified (Collenette
1966; Clennell 1992). Although this is less than
5 km across, field exposures and aerial photographs
show that it has the same arcuate shape and internal
structure as the larger circular basins.
In the adjacent offshore area, the Middle to
Upper Miocene progradational sequence continues
into the Sandakan Basin (Wong 1993). A seismic
dip line from offshore the Kinabatangan area (Fig.
6a, b) shows the main depocentre, with a wedge
of sandy sediment (Unit 1, base Middle Miocene
to mid Late Miocene in age) prograding over a sub-
stratum with chaotic internal reflectors, interpreted
as thick shales or possibly melanges, which are
pushed up ahead of the sand to form a diapiric
wall. The corresponding strike line (Fig. 6c, d)
shows the arcuate section of the basin, with the
reflectors dipping more steeply at the basin
margins than in the centre. The lateral margins
of the basin are interpreted as being constrained
by walls of mud-rich melange or shale pushed up
as the sandy sediments sank into them. If exhumed
by later uplift, this basin would presumably
resemble the circular basins currently exposed
onshore.
A number of other sagged features can be identi-
fied in the Sandakan basin, including some large
basins several kilometres across and at least one
smaller circular depression only c. 1 km across, that
is, on the scale of the Manjang basin. Widespread
mud diapirism is evident in the offshore area where
mud volcanic islands and 'mud lump' islands can
be found north of Sandakan, and also onshore, in
the Dent Peninsula, where there are several large
mud volcanoes (Rheinhard & Wenk 1951; Haile &
Wong 1965; Wilford 1967; Tongkul 1989). The
clay-mineral composition of the muds erupted
from eastern Sabah mud volcanoes shows that
they are derived from the Labang Formation or
from the clay matrix of the melange units (Clennell
1992).
Relationship with Neogene tectonic events
Both the Kulapis and Labang Formations are
affected by intense tectonism, which produced
large-scale folding and faulting across the Central
Sabah Basin. Since the overlying Tanjong sedi-
ments are essentially undeformed except for syn-
sedimentary structures (e.g. Tongkul 1993), the
main deformational episode must have occurred at
around the time of the melange formation, i.e. from
late Early Miocene to early Middle Miocene. From
direct geological and biostratigraphic correlation,
the foregoing events can be tied to the opening and
spreading of the Southeast Sulu Sea marginal basin
(see Hutchison 1992). Therefore, the main sense of
lithospheric strain in the region during the Lower
to Middle Miocene could be assumed to be exten-
sional and to operate in a NE-SW direction.
However, exactly the same time interval is
characterized in the northwest Borneo offshore by
the major shortening and uplift that produced the
Deep Regional Unconformity.
The DRU can be traced continuously through the
Balabac strait from the South China Sea into the
Sulu sea, on the seismic data of Hinz et al. (1989,
1991); and on unpublished commercial data. It has
already been argued that this unconformity is not a
time line, but rather an 'event' line that marks the
propagation of a major phase of structure formation
across the region. To the west the structures pro-
duced are mainly compressional, with wrenching
becoming increasingly important in the northeast.
In the Balabac basin, the northernmost sub-basin
of the Sulu sea (Fig. 7), the DRU is clearly distin-
guishable as the boundary between unreflective
Lower Miocene sediments and the overlying
stratified sequence of middle to late Miocene age.
The basin is floored by oceanic crust which has
been faulted and imbricated. The structure of the
Balabac basin suggests that it may have formed
originally as a rift basin, which was subsequently
inverted by NW-SE compression (Hinz & Block
1990; Hinz et al. 1991). The DRU and overlying
reflectors of up to Late Miocene age are buckled
upwards in the immediate vicinity of the overriding
basement, and draped by undeformed Plio-
Pleistocene sediments. This brackets the age of a
second phase of tectonism to the later Miocene. To
the south of the Balabac Basin, the basement of the
Banggi Ridge has been pushed over the Miocene
sediments, so that it forms a 'pop-up' structure that
verges to the NW in the north and to the SE into the
Bancauan Basin.
Contemporaneous shortening in eastern Sabah is
illustrated by two roadside exposures. At Telupid
ophiolitic basement overrides grey-matrix melange
and, further west, sediments of the Kulapis
Formation, in a high-angle imbricate sequence.
 MIOCENE TECTONICS OF SABAH, MALAYSIA 317

NORTHWEST SOUTHEAST
< PALAWAN J.,, BALABAC BASIN ~l ~ BANGGI RIDGE ?
D HIGH rl ~ rl" D
Seafloor Plio-Pleistocene Mid Miocene / U. Miocene- L. Pliocene

.......-./; ,;v,:..~-~~--..::. B ~";'.-:'.~,, "'-. "','.'., ? Palaeo~ene and

.~.,~ _.~.-" .," . . . . . . . . )-.-- ._, . -. . . . - - . - - - - - ~ C a = . ~ ~ - v , ""-..":'- L-M. Miocene
~ N ~ ~ Palaeo.gene and Lower Miocene deep marine clayey sediments " , , ~ . . ~
~4

~6

I I ? Palaeogene sediment

Fig. 7. Structural cross-section D-D' across Northwestern Sulu Sea (location shown on Fig. 5), redrawn from
unmigrated deep seismic line BGR 86-32 in Hinz et al. (1991). Vertical exaggeration is c. 10 ×. The Balabac Basin
is bounded to the northwest and southeast by compressional ridges of imbricated ophiolitic basement. The initial
tectonic phase ceased in the latest Early Miocene and is marked by Horizon 'C', the equivalent of the Deep Regional
Unconformity. Later compression, inverting the basin, occurred in early to mid Upper Miocene time. Abbreviations:
A, Early Pliocene unconformity; B, base Late Miocene unconformity; C, Late Early Miocene unconformity (Deep
Regional Unconformity of Levell 1987); D, top oceanic crust (Cretaceous ophiolite).

The transport direction is towards the NNW. The
amount of shortening here is modest (probably less
than 1 km), and dies out towards the east. The
timing of this compression seems to tie with the
age of possible inversion noted in the offshore;
certainly it post-dates the melange formation and
deposition of the Kulapis Formation. From south of
the Segama River almost to the Kinabatangan
River, the original unconformity surface between
the Labang depositional system and the overlying
sediments is almost completely lost among later
compressional structures. Early to Middle Miocene
sediments, which are the stratigraphical equivalent
of the Tanjong Formation, but richer in volcani-
clastic detritus, are imbricated with slices of
Labang Formation sediments and melanges. This
area is structurally and stratigraphically analogous
to the higher levels of the seismic section in Fig. 7,
where sediments overlying the DRU are involved
in the shortening.
While some areas of the Central Sabah Basin
have experienced this later compressional phase,
most of the Tanjong, Sandakan and equivalent
sediments that were deposited north of the
Kinabatangan River and in the Malibau and Meliau
Basins are essentially undeformed. Minor inversion
and wrench structures of mid-Late Miocene age
are evident in the Sandakan and Bancauan basins
(Bell & Jessop 1974; Tamesis 1990; Wong 1993;
and unpublished commercial data). These move-
ments are very modest compared with the struc-
tures shown in Fig. 7. It is suggested that these
later events are expressions of far-field tectonic
forces that are transmitted through rigid crustal
blocks, but only expressed strongly near to block
margins (e.g. around the Telupid and Segama
ophiolite blocks, and along the Banggi and
Bancauan ridges shown on Fig. 5).
As an additional factor, in the interior of the
Central Sabah Basin, the cover sediments may
have been effectively decoupled from the effects
of tectonism by the thick shales and melanges
which form the main substratum to the Tanjong
depositional system. Tanjong/Sandakan sedimen-
tation was therefore able to prograde continuously
northeastwards along the axis of the Central Sabah
Basin without any major interruptions, while some
surrounding areas were being uplifted. In the sedi-
mentary record this history is expressed as con-
tinuing sedimentation from the Middle Miocene
to lower Pliocene without any marked tectonically
produced unconformities. Indeed, the post Middle
Miocene sediments of the Sandakan Basin are
amenable to sequence stratigraphic analysis (Wong
1993) demonstrating that eustatic sea-level changes
were large compared with tectonic movements.
Many of the main structural features in the basin
developed as a consequence of local gravitational
instabilities rather than regional tectonic events
(c.f. Talbot 1992).
Summary
Both 'far field tectonics', i.e. stresses ultimately
driven by motions of major plates, and gravity
tectonics (diapirism and slumping) have interacted
to influence basin process in the Sabah region.
A major unconformity surface, the Deep Regional
318 B. CLENNELL
Unconformity, can be traced across the northwest
Sabah offshore and into the onshore of eastern
Sabah. The Deep Regional Unconformity marks a
major reorganization of the regional stress and
strain field that produced a widespread interruption
in deposition. Rather than being a time line this
surface is diachronous, and was formed largely by
c o m p r e s s i o n in the n o r t h w e s t and mainly by
extension in the southeast (Central Sabah and
Sandakan Basins). In eastern Sabah this upheaval
was marked by the generation of extensive mud-
matrix melanges and a change from deep marine to
shallow marine and terrestrial deposition.
While basin initiation in Sabah was controlled
by the main regional tectonic events, the internal
architecture that developed through the Miocene
was controlled by local patterns of subsidence and
uplift of the substratum. Where this material is
relatively rigid, the effects of tectonically-driven
compression, extension and wrenching were trans-
mitted directly to the basin sediments. This pro-
duced inversion structures that are localized to the
margins of the main basement blocks, such as the
Banggi Ridge and the Telupid ophiolite block.
References
BEDDOES, L. R. 1976. The Balabac sub-basin, south-
western Sulu Sea, Philippines. SEAPEX Program,
South East Asia Conference, Paper 15, 1-22.
BELL, R. M. & JESSOP, R. G. C. 1974. Exploration and
Geology of the west Sulu Basin, Philippines.
Journal of Australian Petroleum Exploration
Association, 14, 21-28.
BI~NARD, E, MULLER, C., LETOUZEY,J., RANGIN, C. &
Tahir, S. 1990. Evidence of multiphase deformation
in the Rajang-Crocker Range (Northern Borneo)
from Landsat imagery interpretation: Geodynamic
implications. Tectonophysics, 183, 321-339.
BEN-AVRAHAM,Z. 1989. Multiple opening and closing of
the Eastern Mediterranean and South China Basins.
Tectonics, 8, 351-362.
BOL, A. J. & VAN HOORN, B. 1980. Structural style in
western Sabah offshore. Bulletin of the Geological
Society of Malaysia,12, 1-16.
BRIAIS, A., TAPPONNIER, P. & PAUTOT, G. 1989.
Constraints of Sea Beam data on crustal fabrics
and seafloor spreading in the South China Sea.
Earth and Planetary Science Letters, 95, 307-
320.
CLENNELL, M. B. 1991. The origin and tectonic signifi-
cance of mdlanges in Eastern Sabah, Malaysia.
Journal of South East Asian Earth Sciences, 6,
407-429.
1992. The Mdlanges of Sabah, Malaysia. PhD
Thesis, University of London.
COLLENETTE, P. 1965. The Geology and Mineral
Resources of the Pensiangan and Upper
Kinabatangan area, Sabah. Malaysian Geological
Survey, Borneo Region, Memoir, 12.
- - 1966. The Garinono Formation, Sabah, Malaysia.
Where the succeeding sediments prograded over
thick shale-rich sequences or m u d - m a t r i x
melanges, the basin interiors are largely decoupled
from far-field tectonic influences, but have under-
gone gravity-driven sagging and m u d diapirism,
producing a series of remarkable circular basins in
the Central Sabah and Sandakan basins.
This paper led from a PhD project supervised by Dr Tony
Barber at Royal Holloway, University of London and
Dr Felix Tongkul of Universiti Kebangsaan Malaysia,
Sabah. Their continued help has been invaluable. The
project was fully funded by Shell International Petroleum
Co. Ltd. Further support came from the University of
London SE Asia Research Group: sincere thanks go to its
administrator, Diane Cameron.
Professors K. Hinz and Claude Rangin provided much
useful information. Seismic lines of the Sandakan Basin
were provided by W.M.C. (Malaysia) SDN BHD.
Particular thanks go to Terry Walker, and to Petronas SDN
BHD for allowing access to, and publication of, this data.
Additional help and ideas came from Herman Soediono,
Lim Pen Siong, Michiel de Smet and Chris Talbot. This
paper has benefited greatly from reviews by Chris Elders
and Felix Tongkul.
In: Malaysia Geological Survey, Borneo Region,
Annual Report for 1965. 161-167.
DALY,M. C., COOPER,M. A., WILSON, I., SMITH,D. G. &
HOOPER, B. G .D. 1991. Cenozoic plate tectonics
and basin evolution in Indonesia. Marine &
Petroleum Geology, 8, 3-21.
DAMUTH, J. E. 1994. Neogene gravity tectonics and
depositional processes on the deep Niger Delta
continental margin. Marine & Petroleum Geology,
11,320-346.
HALLE, N. S. & WONG, N. E Y. 1965. The Geology and
Mineral Resources of the Dent Peninsula, Sabah.
Malaysia Geological Survey, Borneo Region,
Memoir, 16.
HAMILTON,W. 1979. Tectonics of the Indonesian Region.
United States Geological Survey, Professional
Paper, 1078.
HAZEBROEK, H. P. & TAN, D. N. K. 1993. Tertiary
tectonic evolution of the NW Sabah continental
margin. Bulletin of the Geological Society of
Malaysia, 33, 195-210.
HINZ, K. & BLOCK, M. 1990. Summary of geophysical
data from the Sulu and Celebes Seas. In: RANGIN,
C. StaYER, E. A. & VON BREYMANN, M. T. (eds)
Proceedings of the ODP, Initial Report, Ocean
Drilling Program, 124, 87-91.
, KUDRASS, H. R., & MEYER, H. 1991.
Structural elements of the Sulu Sea, Philippines.
Geologisches Jahrbuch, A127, 483-506.
- - - , FRISCH,J., KEMPTER,E. H. K., MANAFMOHAMMAD,
A., MEYER, J., Er AL 1989. Thrust tectonics along
the continental margin of Sabah, Northwest Borneo.
Geologische Rundschau, 78, 705-730.
HOLLOWAY, N. H. 1982. North Palawan block - its
 MIOCENE TECTONICS OF SABAH, MALAYSIA
relation to Asian mainland and role in evolution of
South China Sea. AAPG Bulletin, 66, 1355-1383.
HtrrCHISON, C. S. 1988. Stratigraphic-tectonic model for
E. Borneo. Bulletin of the Geological Society of
Malaysia, 22, 135-151.
1989. The Geological Evolution of S.E. Asia.
Oxford University Press.
- - 1992. The Southeast Sulu Sea, a Neogene marginal
basin with outcropping extensions in Sabah.
Bulletin of the Geological Society of Malaysia, 32,
89-108.
KrRK, H. J. C. 1968. The Igneous Rocks of Sarawak and
Sabah. Malaysian Geological Survey Bulletin,
Borneo Region, 5.
LEE, C. P. & THAM, K. C. 1989. Circular Basins of Sabah.
Proceedings Geological Society of Malaysia
Petroleum Geology Seminar, Kuala Lumpur, 1989
(abs.). Bulletin of the Geological Society of
Malaysia, 29, 54.
LEE, D. T. C. 1970. Sandakan Peninsula, Eastern Sabah,
Eastern Malaysia. In: Geological Survey of
Malaysia, Borneo Region, Report, 6, 1-75.
LEICHTI, P., ROE, F. W. & HAILE,N. S. 1960. The Geology
of Sarawak, Brnnei and the western part of North
Borneo. British Territories of Borneo, Geological
Survey Department, Bulletin, 3.
LEVELL, B. K. 1987. The nature and significance of
regional unconformities in the hydrocarbon-bearing
Neogene sequence offshore West Sabah. Bulletin of
the Geological Society of Malaysia, 21, 55-90.
-- • KASUMAJAYA, A. 1985. Slumping of the late
Miocene shelf-edge offshore West Sabah: a view of
a turbidite basin margin. Bulletin of the Geological
Society of Malaysia, 18, 1-30.
LIM, E.H. 1990. Geologi am kawasan timur Kota
Kinabatangan, Sabah. BSc Thesis, Universiti
Kebangsaan Malaysia, Sabah [in Malay].
LtM, P. S. & YIN EE HENG. 1985. Geological Map of
Sabah, Scale 1:500 000. 3rd. Edition. Geological
Survey of Malaysia.
MASCLE, A. & BISCARRAT, P. A. 1979. The Sulu Sea:
a marginal basin in South East Asia. In: WATKrNS,
J. S., MONTADERT, L. & DICKERSON, P. W. (eds)
Geological and Geophysical Investigations of
Continental Margins. AAPG Memoir, 29,
373-381.
NICHOLS, G. J., BETZLER, C., BRASS, G. W., HUANG, Z.,
L/NSLEY, B., ET AL. 1990. Depositional History of
the Sulu Sea from ODP sites 768,769 and 771.
Geophysical Research Letters, 17, 2065-2068.
RANG/N, C. & SILVER, E. 1990a. Geological setting of
the Celebes and Sulu Seas. In: RANG/N, C., SILVER,
E. A. & VON BREYMAYN,M. T. (eds) Proceedings of
the ODP, Initial Report. Ocean Drilling Program,
124, 35-41.
- - & -- 1990b. Neogene tectonic evolution of the
Celebes-Sulu Basins: New insights from Leg 124
drilling. In: RAYG/N, C., StaYER, E. A., & YON
BREYMANN, M. T. (eds) Proceedings of the ODP,
Initial Reports. Ocean Drilling Program, 124,
123-144
--, JOLIVET, L., PUBELLIER, M. & the Tethys-Pacific
working group. 1990a. A simple model for the
tectonic evolution of SE Asia and Indonesia region
319
for the last 43 m.y. Bulletin de la Socidt~ g~ologique
de France, 8, 889-905.
- - - , BELLON, H., BENARD, F., LETOUZEY,J., MOLLER, C.
& SANUDIN, Y. 1990b. Neogene arc-continent
collision in Sabah, Northern Borneo (Malaysia).
Tectonophysics, 183, 305-319.
, SILVER, E. A., HINZ, K., KUDRASS,H. & ODP Leg
124 Scientific Party 1991. Structural setting of the
Sulu Basin: New insights from drilling. ODP
Scientific Results, 146. Ocean Drilling Program.
, PUBELLIER,M., AZt~MA,J., BRIA1S, A., CHOTIN, P.,
eT AL. 1990C. The quest for Tethys in the western
Pacific. 8 palaeogeodynamic maps for Cenozoic
time. Bulletin de la Socidtd gdologique de France,
8, 907-913.
RHEINHARD, M. & WENK, E. 1951. The Geology of the
Colony of North Borneo. British Borneo Region
Geological Survey Bulletin, 1, 7-11.
RIcE-OXLEY, E. D. 1991. Palaeoenvironments of the
Lower Miocene to Pliocene sediments in offshore
NW Sabah Area. Bulletin of the Geological Society
of Malaysia, 28, 165-194.
Ru, K. & PIGOTT, J. D. 1986. Episodic rifting and
subsidence in the South China Sea. AAPG Bulletin,
70, 1136-1155.
TALBOT, C. J. 1992. Quo Vadis Tectonophysics? With a
pinch of salt! Tectonophysics, 16, 1-20.
TAMESIS, E. V. 1990. Petroleum geology of the Sulu Sea
Basin, Philippines. SEAPEX Proceedings, Volume
IX, (Eighth Offshore Southeast Asia Conference,
Singapore). 45-54.
TAN, D. N. K. & LAMY, J. M. 1990. Tectonic evolution
of the NW Sabah continental margin since the Late
Eocene. Bulletin of the Geological Society of
Malaysia, 27, 241-260.
TAYLOR, B. & HAYES, D. E. 1983. Origin and history
of the South China Sea Basin. In: HAYES,D. E. (ed.)
The tectonic and geologic evolution of SE Asian
seas and islands, Part 2. American Geophysical
Union, Geophysical Monographs Series, 27,
23-56.
THAM, K.C. 1984. Geology of the Bukit Garam Area,
Sabah, East Malaysia. BSc Thesis, University of
Malaya.
TJIA, H. D., KOMOO,I., LIM, P. S. ~¢ TUNGAHSURAT. 1990.
The Maliau Basin, Sabah: geology and tectonic
setting. Bulletin of the Geological Society of
Malaysia, 27, 261-292.
TONGKUL, F. 1987. Sedimentology and structure of the
Crocker Formation in the Kota Kinabalu Area,
Sabah, East Malaysia. PhD Thesis, University of
London.
- - 1989. Geological control on the birth of the Pulau
BatH Hairan mud volcano, Kudat, Sabah. Warta
Geologi, 14, 153-165.
- - 1990. Structural style and tectonics of western and
northern Sabah. Bulletin of the Geological Society
of Malaysia, 27, 227-239.
1991. Tectonic evolution of Sabah, Malaysia.
Journal of South East Asian Earth Sciences, 6,
393-406.
1993. Tectonic control on the development of
Neogene basins in Sabah, East Malaysia. Bulletin of
the Geological Society of Malaysia, 27, 95-103.
320 B. CLENNELL
-- 1994. The Geology of northem Sabah, Malaysia:
its relationship to the opening of the South China
Sea Basin. Tectonophysics, 235, 131-137.
UJHE, H. 1970. Miocene foraminiferal faunas from
the Sandakan formation, north Borneo. In:
Contributions to the Geology and Palaeontology of
S.E. Asia, University of Tokyo Press, 8, 165-185.
1977. New species and subspecies of benthonic
foraminifera from the Miocene Sandakan
Formation, north Borneo. In: Contributions to the
Geology and Palaeontology of S.E. Asia, University
of Tokyo Press, 18, 87-102.
WHITE, R. & MACKENZIE, D. E 1989. Magmatism at
rift zones: the generation of volcanic continental
margins and flood basalts. Journal of Geophysical
Research, 94, 7685-7729.
WILFORD, G. E. 1967. The mud volcanoes of Sabah.
Journal of the Sabah Society, 3, 12-21. Sabah
Museums Service, Kota Kinabalu.
WONG, R. H. E 1993. Sequence stratigraphy of the
Middle Miocene to Pliocene southern offshore
Sandakan Basin, E. Sabah, Malaysia. Bulletin of the
Geological Society of Malaysia, 33, 129-142.
 Mesozoic and Cenozoic plutonic evolution of SE Asia:
evidence from Sumatra, Indonesia

W. J. M c C O U R T 1, M. J. C R O W 1, E. J. C O B B I N G 2 & T. C. A M I N 3
1 British Geological Survey, International Division, Keyworth, Nottingham NG12 5GG, UK
2 Consultant Geologist, 25 Main Street, Radcliffe-on-Trent, Nottingham, UK
3 Geological Research and Development Centre, Bandung, Indonesia

Abstract: New K/Ar mineral ages from the Barisan Mountains of southern Sumatra suggest
four main periods of plutonic activity: Miocene-Pliocene (20-5 Ma), Early Eocene (60-50 Ma),
Mid-Late Cretaceous (l17-80Ma) and Jurassic-Early Cretaceous (203-130Ma). These and
all other published ages from exposed plutons in western Sumatra indicate a further period of
plutonic activity in the Permian (287-256 Ma). They also suggest either that Early Mesozoic
activity began in the Late Triassic, or that there were two distinct magmatic cycles, one in the
Late Triassic to Early Jurassic (220-190 Ma) and one extending from the Mid-Jurassic to Early
Cretaceous (170-130 Ma). In addition, poorly controlled ages from eastern Sumatra indicate that
the important Triassic to Early Jurassic (240-195 Ma) tin-belt magmatism of the peninsular
Malaysia Main Range Province extends into that area.
Preliminary geochemical studies on the Mesozoic granitoids of the Barisan Range of southern
Sumatra confirm that they are calc-alkaline, I-type, metaluminous, subduction-related volcanic
arc granites (VAG). They broadly correspond to the southerly extension of a combination of the
Central Valley and Western Granite Provinces of Thailand and Burma, and underline the fact
that there has been a history of subduction-related magmatism along the southwestern edge
of Sundaland since earliest Mesozoic times. The plutonic suites are crudely arranged in sub-
parallel, locally overlapping, NW-SE trending belts, focused along deep-seated faults that
have acted as magmatic conduits. It is proposed as a preliminary model that breaks in plutonic
activity broadly correspond to changes in approach angle and/or rate of subduction, and that
in some instances at least they relate to periods of collision and accretion of allochthonous
material (terranes, slivers or blocks) of both oceanic and continental character. At least two such
events seem to have occurred during the Mesozoic-Cenozoic tectono-plutonic evolution of
Sumatra. One in the early Middle Cretaceous reflects collision and accretion of the oceanic Woyla
terranes, and one in the latest Cretaceous is possibly related to collision of a continental
sliver/block, the West Sumatra terrane, to the Sundaland margin.

Plutonic rocks are widely exposed throughout the
Barisan Mountains of western Sumatra, and locally
exposed, through Tertiary-Quaternary cover
sequences, in eastern Sumatra. In general terms the
granitoid rocks of the Barisan Mountains, define
a series of sub-parallel, N W - S E trending, south-
westerly migrating plutonic belts of Triassic-
Jurassic, Cretaceous and Late Tertiary age (Katili
1973; Gafoer & Purbo-Hadiwidjoyo 1986) and a
less well-defined Eocene belt (McCourt & Cobbing
1993). The presence of these plutonic belts and
their assumed calc-alkaline composition, has been
extensively quoted as evidence that Sumatra has
been located in a region of plate convergence,
marked by continental margin subduction, since the
Early Mesozoic and perhaps earlier (Katili 1973;
Hamilton 1979; Cameron et al. 1980).
As part of the Southern Sumatra Geological and
Mineral Exploration Project (SSGMEP, 1989-94)
a reconnaissance field study and sampling
programme was undertaken in 1992 of the main
granitoid plutons of the Barisan Mountains of
southern Sumatra. The aim of this study was to
establish the geological and geochemical charac-
teristics and the isotopic ages of the plutons in order
to evaluate the role of plutonism in the tectonic
evolution of this part of Sundaland. During this
investigation thirteen intrusive bodies ranging from
individual plutons to batholiths were examined; all
exposed major lithologies from each were sampled
for whole rock geochemical analysis, and selected
samples dated using the K-Ar method on mineral
separates. The geochemical and geochronological
data, comprising major element oxides and trace
element compositions, new K-Ar mineral ages
and a summary of all published ages from plutonic
rocks of Sumatra and the tin-islands, quoted in this
paper are available as Supplementary Publication
No. SUP 18098 (8 pp) from the Society Library or
the British Library D o c u m e n t Supply Centre,
Boston Spa, Wetherby, W. Yorks LS23 7BQ, UK.
The study has established that the Barisan

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, 321

Geological Society Special Publication No. 106, pp. 321-335.
322 W. J. McCOURT
Mountains granitoids range in age from Late
Triassic to Pliocene, and has confirmed that they
are I-type in character.
Tectonic setting and regional
geological framework
Sumatra lies along the southwest margin of the
Sundaland cratonic block, the SE Asian continental
extension of the Eurasia Plate (Fig. 1). To the west,
oceanic crust of the India plate is being obliquely
subducted beneath Sumatra along the Sunda
trench in a N20(E direction at a rate of between
6-7 c m a -1 (Curray et al. 1979; Hamilton 1979).
This zone of oblique convergence forms part of the
Sunda arc-trench system which extends for more
than 5000 km from Burma to eastern Indonesia.
The Tertiary-Recent Barisan volcanic arc of
western Sumatra and the pre-Tertiary sequences
are variably broken up into a number of NW-SE
Ocean
Plate /
lOO 11o ~
I I /
Fig. 1. Tectonic setting of Sumatra on the SW margin of Sundaland (from Hutchison 1989).
ET AL.
trending slivers by major faults which are strands of
the Sumatra Fault System (SFS), some of which are
Mesozoic in age. The SFS as defined in this paper,
incorporates all those (NW-SE) faults along and
west of the eastern foothills of the Barisan
Mountains and includes the Great Sumatra Fault
(van Es 1919; van Bemmelen 1949; Westerveld
1952). Many of the faults of the SFS have a history
of multiple movement and reactivation and are
interpreted as deep-seated, subduction related,
continental margin structures. To the east of the
magmatic arc and the SFS, are the back-arc basins
containing thick sequences of Tertiary to
Quaternary sediments, underlain by probable
Upper Palaeozoic continental basement (Gafoer
1990).
The pre-Tertiary continental core of SE Asia
is interpreted as a complex assembly of various
tectono-stratigraphic terranes including allo-
chthonous micro-continental blocks, island arcs
and accretionary complexes bounded by regional
Subduction zones
'~X Transcurrent Fault
Direction of Plate
movement
20-
Philippine
Sea Plate
10 m
12¢
 MESOZOIC--CENOZOIC PLUTONISM IN SUMATRA
faults, some of which are interpreted as sutures,
marked by slivers of ophiolite, melanges and
volcanic-plutonic arcs (Mitchell 1981; Stauffer
1983; Audley-Charles et al. 1988; Metcalfe 1988,
1990, among others). The pre-Tertiary framework
of Sumatra is similarly interpreted as a mosaic of
oceanic and continental terranes (Fig. 2) assembled
through processes of microplate convergence and
accretion since the Middle Permian (Pulunggono
& Cameron 1984; McCourt et al. 1993).
Much of central and eastern Sumatra is believed
to be underlain by continental crust and the oldest
exposed units are Upper Palaeozoic (Permo-
Carboniferous) metasediments. Pulunggono &
Cameron (1984) proposed that the basement
comprised two continental terranes, the Mergui
and Malacca microplates, separated by a Triassic
suture complex, the Mutus Assemblage (Fig. 2).
The dominantly Palaeozoic Mergui microplate
comprises, 'older granitic basement', Permo-
Carboniferous clastic and carbonate metasedi-
mentary sequences, Lower to Middle Permian
volcanics and associated sediments and Middle to
Upper Triassic marine sediments. The Malacca
microplate, which is confined to subcrop in
Sumatra, consists predominantly of quartzites
(Eubank & Makki 198t; Koning & Darmono
1984). It is considered to be of Carboniferous to
Middle Permian age based on lithological and
structural correlations with the Kenny Hill
Formation of peninsular Malaysia (Fontaine &
Gafoer 1989) and accordingly is part of the
Gondwana affinity terrane of peninsular Malaysia
(Hutchison 1994). The Mutus Assemblage,
described from well records by Eubank & Makki
(1981), comprises a mixture of radiolarian chert,
red-mauve shales, thinly bedded limestone, a sand-
stone-shale sequence and basalt. It was assumed to
be Triassic by Pulunggono & Cameron (1984), but
Hutchison (1994) has proposed that it represents a
Palaeozoic subduction-accretion complex.
The Permo-Carboniferous metasedimentary
basement sequences of the Mergui microplate are
exposed along the eastern flank of the Barisan
Mountains and through the Tertiary cover of the
present day back-arc zone. They comprise in broad
terms, glacio-marine coarse clastic pebbly units,
finer-grained turbiditic slates interpreted as distal
equivalents, and warmer water marine shelf and
slope deposits including massive limestones
(Simandjuntak et al. 1991). The contact between
the sequences of glacio-marine sediments
(Bohorok-Mentulu and Kluet-Gangsal formations)
and warmer water marine-shelf sediments
(Kuantan-Terantam formations) is interpreted to
be tectonic and the two are considered to represent
separate terranes with differing Late Palaeozoic
tectonic histories that probably united in the Middle
323
Permian (Gafoer 1990; McCourt et al. 1993). In
south Sumatra this contact corresponds to the
regional, NW-SE striking, fault separating (or
juxtaposing) the basement sequences of the
Tigapuluh and Duabelas Mountains (Fig. 2),
structurally corresponding to the Medial Sumatra
Line of Hutchison (1994).
The Permo-Triassic basement units are exposed
along the axial region of the Barisan Mountains.
They comprise a lower sequence of Lower to
Middle Permian andesitic volcanics, reefal lime-
stones and carbonaceous sediments (Palepat,
Silungkang and Mengkarang Formations) and an
upper one of Middle to Upper Triassic pelagic
sediments and limestones (Kuala and Tuhur
Formations). In north Sumatra, the Permian
volcanic arc association has recently been inter-
preted as a separate terrane from the Permo-
Carboniferous sequences (Wajzer et al. 1991) and
this model is followed here. The contact of the
Triassic units, where seen, is disconformable on the
Permian volcanic association and unconformable
on the Permo-Carboniferous sequences.
Upper Mesozoic sequences occur along the
western side of the Barisan Mountains in southern
Sumatra, in tectonic contact with older units to the
east. They comprise varying proportions of oceanic
basalts, ultrabasic rocks interpreted as ophiolitic,
andesitic lavas and volcaniclastics, pelagic sedi-
ments and cherts, fine-grained turbiditic units,
shallow marine clastics, limestones, phyllitic rocks
and schists. The majority of these are (?Mid- to)
Upper Jurassic to Lower Cretaceous and comprise
a series of allochthonous slivers that correspond in
part to the Woyla terranes of Wajzer et al. (1991)
and belong to the Woyla microplate of western
Sumatra (McCourt et al. 1993).
In broad terms the Tertiary sequences of Sumatra
correspond to three distinct depositional environ-
ments: sediments and minor volcanics of the
forearc zone, calc-alkaline volcanics and sediments
of the magmatic arc zone and sediments of the
back-arc zone. During the Quaternary, Sumatra
was the scene of tremendous volcanic activity
throughout the Barisan Range, and the modern
physiography was established. Tectonism and uplift
in the latest Miocene and Plio-Pleistocene caused
inversion of the Cenozoic sedimentary basins,
development of widespread NW-SE trending
fold structures, and reactivation of deep-seated
basement faults.
Granitoid plutonism of Sumatra
Granitoid rocks are present throughout the Barisan
Mountains and are particularly common along
and close to the junction of the Mergui and Woyla
microplates (Figs 2 and 3). Permian magmatism
324 W. J. McCOURT ET AL.

1. MALACCA MICROLATE
2, MERGUI MICROPLATE
A. Bohorok Tigapuluh Terrane

B. Kluet-Kuantan.Duabelas-
Terrane
C, Palepat Terrane
D. Kuala FM

3. WOYLA MICROPLATE A. Woyla terrane

B. Pasamar~ terrane
C. Gumai-Garba terrane

4. CONTINENTAL A. Sikuleh
FRAGMENTS B. Natal
C Bengkulu

: 2A

EQUATOR

0 250km
I J
v V v '~

v vv~

Fig. 2. Simplified pre-Tertiary microplate configuration of Sumatra (modified from Pulunggono & Cameron 1984).

is indicated by Rb-Sr ages of 264 + 6 (Sibolga
granite) and 256 +_6 Ma (Ombilin granite), a K-Ar
muscovite age of 287 _ 3 Ma (Ombilin granite) and
by restricted exposures of andesitic volcanics. The
Sibolga granite is a composite batholith recording
several intrusive phases that intrudes metasedi-
ments of the Kluet Formation (Aspden et al. 1982).
The main rock types are porphyritic K-feldspar
biotite-hornblende granite, monzogranite, quartz
diorite and diorite. In addition to the Permian age
several younger dates are available suggesting
mainly Late Triassic (219-206Ma) and Late
Jurassic (147-144Ma) magmatic activity. The
Ombilin granite is a foliated muscovite granite
without clear field relationships, although Silitonga
& Kastowo (1975) imply that it intrudes the meta-
sedimentary Kuantan Formation. The Permian
volcanic arc appears to be restricted to a fault slice
along the western margin of the pre-Mesozoic
metasedimentary basement.
Evidence for widespread 'Triassic' plutonism is
seen in the abundance of intrusive ages ranging
from 230-195 Ma from Sumatra and the adjacent
tin-islands. This phase of plutonism appears to
correspond to two separate belts: one in eastern
Sumatra and the tin-islands, the southerly extension
of the Main Range and Eastern Granite Provinces
of peninsular Malaysia, and the other in the Barisan
Mountains of western Sumatra. The latter com-
prises hornblende-bearing biotite-granitoids of
1-type character with a wide compositional range
which are predominantly quartz diorites to grano-
 MESOZOIC--CENOZOIC PLUTONISM IN SUMATRA
SIKULEH
GRANITE
SIBOLGA
NITE •HATAPANG~X
GR NIT
MANUNGGAL BATHOLITH.~
GRANITE
OMBILIN
GRANITE
Fig. 3. Distribution of the principal plutonic belts of (western) Sumatra and location of the main plutons.
diorites. This western Sumatra Triassic to Early
Jurassic plutonic arc, is located close to the edge of
the Mergui microplate intruding metasediments of
the Kluet-Kuantan terrane and its trace coincides
with a compositionally similar but more extensive
Middle Jurassic-Early Cretaceous (170-130 Ma)
arc.
A Mid to Upper Cretaceous (117-80Ma)
plutonic arc extends along the length of the Barisan
Mountains, intruding oceanic rocks of the Woyla
microplate and broadly contemporaneous Upper
Mesozoic continental foreland sequences. This arc
is focused along a major NW-SE striking fault
zone interpreted to approximate to the Early
Cretaceous Woyla Suture, and dominated by
dioritic to granodioritic hornblende-bearing
granitoids with subordinate K-feldspar megacrystic
biotite granites of I-type character. Scattered Lower
Eocene plutons (60-50 Ma) intrude earlier plutonic
arcs, mainly the Upper Cretaceous arc. The
Miocene arc (20-5 Ma) consists of granodioritic
325
D Tertiary/(Eocene and Mid-Pliocene)
Plutonic Arc
Mid-late Cretaceous
Plutonic Arc
EZ] ( Late ) Triassic-Jurassic/Early
Cretaceous Plutonic Arc
[ ~ Permian Plutonic-Volcanic Arc
BATHOLITH
PLUTON
and granitic plutons which commonly intrude
Upper Oligocene to Miocene andesitic volcanics
to the west of the Semangko Fault.
Geochemistry
In terms of their overall major element oxide
geochemistry (Table 1 & Fig. 4), the granitoids
are typical subduction-related, I-type granites
(Chappell & White 1974). Plutonic rocks range in
composition from gabbro to granite on all common
classification-nomenclature plots, with a concen-
tration of samples in the fields of quartz diorite-
tonalite, granodiorite and monzogranite. On the
Streckeisen QAP plot they occupy the same field as
granitoids from the Lima segment of the Coastal
Batholith of Peru (Pitcher et al. 1985). They are
calc-alkaline (Fig. 5a, b) cafemic, and dominantly
metaluminous with the most highly differentiated
felsic derivatives ranging to slightly peraluminous
(McCourt & Cobbing 1993). Finally although
326 W.J. McCOURT ET AL.

Table 1. Summary of new K-Ar plutonic ages from the Barisan Mountains, southern Sumatra

INTRUSION

Jurassic to Early Cretaceous

Sulit Air Suite 203 _+6 183 +_ 13 149 _+5 141 ± 5 138 ± 5
Bungo Batholith 169 ± 5 156 ± 6 154 +_7 153 ± 4 148 ± 4 131 _+7 129 ± 4
Way Sulan Gabbro 151 __4
Middle to Late Cretaceous
Garba Pluton 117 ±3 117 __ 2 115 ± 4 86 ± 3 8223
Sulan Pluton 113 ±3 111 __ 3
W Sekampung Diorite 89 --!-3 89 ± 2 89 ± 2
Branti Granodiorite 86 _+3
Padean Pluton 82 ±3 82 ± 2 8l ± 2 82 _+2
Early Eocene
Jatibaru Microgranite 60 ± 3
Lassi Pluton 57 ± 2 54 ± 2 55 __ 2 53 ± 2 53 + 2
Bungo Batholith 54 _ 2 54 ± 2
Miocene
Way Bambang Pluton 20 _ 1 19 ± 2
Lolo Pluton ll ± 1 5 ± 0.2

discussed here as a single population, the granitoids
are not interpreted to be co-magmatic, since they
r e p r e s e n t m a g m a t i s m e x t e n d i n g o v e r almost
200 Ma, but are considered to be products of a
similar and l o n g - l i v e d m e c h a n i s m o f m a g m a
genesis.
On the Nb-Y and Rb-(Nb + Y) plots o f Pearce
et al. (1984) the volcanic arc character o f the
granitoids is clear (Fig. 6a, b) with only four
samples falling outside of the VAG field, all o f
which are highly differentiated and anomalous with
respect to the bulk population (McCourt & Cobbing
1993). The VAG character o f the Barisan granitoids
is also evident w h e n the granitic rocks (SiO 2
> 70%) are plotted on the ORG-norrnalised plots
o f Pearce et al. (1984). A c c o r d i n g to these authors,
LIL e n r i c h m e n t is a c o m m o n feature o f both
s u b d u c t i o n - r e l a t e d volcanic arc granites and
within-plate rifted granites. Within-plate granites,
however, also typically show an enrichment in Ta
and Nb, and have Zr, Y and Ce normalized values
greater than or close to 1; volcanic arc granites on

I I I // I 1

5-

4-
~
/ 00 0-
g °0
eo
3-
O ~ %" •
~o • 0° g • •
2-

QQ
1-

/.
0 I . t _t
0 1 2 3 4 5 6

Na20 ( w t % )

Fig. 4. Alkali variation diagram for the south Sumatra granitoids. Line separates the fields of I- & S-type granites of
Chappell & White (1974).
 MESOZOIC--CENOZOIC PLUTONISM IN SUMATRA
Na20 + K20
(a)
6 f
//.
/
jJ
II
J • o11 j
4 • f J
j J J
Z
J
o I
50
(b)
Fig. 5. (a) AFM plot for the south Sumatra granitoids. (b) Total alkalis v. silica variation diagram for the south
Sumatra granitoids; dashed line denotes calc-alkaline field.
the other hand have Ta and Nb normalized values
close to 1 and Zr, Y and Ce values less than 1,
depending on the nature of the arc setting. The
entire population the Barisan granites are geo-
chemically similar and Fig. 7 confirms their VAG
characteristics of LIL enrichment, depletion in Nb,
Zr and Y, and Ce values close to 1. Collectively
these patterns closely resemble those of sub-
duction-related, Andean I-type granites from Peru
and Chile (cf. Pearce et al. 1984; Pitfield et al.
1986). On primitive mantle- or MORB-normalized
spider diagrams, the subduction-related character
of the Sumatra rocks, excluding the granites sensu
stricto, is once again evident and the strongly
negative Nb N and, to a lesser extent, TiN anomalies
of the entire population are emphasized (cf.
McCourt & Cobbing 1993; Fig. 7b). This
327
MgO
~ • 0 0
ooo
oee ee~J ~
/
L
60 70 80
SiO 2 (wt %)
Ta-Nb-Ti, and associated Y-Yb depletion is a
typical geochemical signature of calc-alkaline
intermediate magmas formed in a subduction
setting, most probably reflecting amphibole and
garnet retention at the site of partial melting
(Briqueu et al. 1984; Foley & Wheller 1990), and
thus characteristic of volcanic arc granitoids. Low
values of Rb/M (M = HREEs, Y, Zr, etc.) and high
to very high values of K/Rb of the granitoids
(McCourt & Cobbing 1993) underline their
continental-arc, I-type character, and Rb/Zr ratios
generally less than 1 and often less than 0.70, in
conjunction with low Nb contents, suggest a
primitive to normal arc setting (Brown et al. 1984).
In summary the geochemical data clearly con-
firm the volcanic arc character of the Barisan
granitoids and provide evidence for the existence
328 W . J. M c C O U R T ET AL.

1000 z

WPG

E lOO

v
: "

3 syn- COLG ee o ~

10
= ;+ /
. . - - - ./
• • eeo I • / ORG

/
(a) 1 , , , ,, ,,=l ~ , I ~ ~,+1 ~ , , = ~l~,l
100 1000 2000
Log Y ( ppm )
2000 i i i i i t I i I + ~ ,','''~ J /Y' '~'~l t

1000 __

syn - C O L G WPG
_
_

Et~
Q"
v
100
-- • qbo • oo
--
_ • • O o •
• •0

VAG ORG -

(b) 1 I I I I I Ill+ I I I I Illl I I I I I 1}1]

10 100 1000 2000
Log Y + Nb (ppm)

Fig. 6. Tectonic setting-classification plots for the south Sumatra granitoids, following Pearce et al. (1984). VAG,
Fields of volcanic arc granite; syn-COLG, syn-collisionalgranite; WPG, within-plate granite and ORG, oceanic ridge
granite as defined by Pearce et al. (1984).

of an active margin and subduction along the arcs of western Sumatra, although this remains
Sumatran continental edge. A further important unproven in the absence of isotopic data for our
feature brought out by the data is the compo- dataset.
sitional similarity of the granitoids of different
ages, as can be seen from the way that the entire
sample population, representing Late Triassic to
Geochronology
Pliocene plutonism, plots very close together on the
normalized spider diagrams (Fig. 7a, b). The best Material was collected for both Rb/Sr and
interpretation of this similarity is that the mech- K-Ar analysis but the granitoids whole rock
anism of magma genesis has been a consistent and geochemistry precludes their use for Rb/Sr geo-
long-lived one throughout the last 200 Ma. It is chronology and consequently the dating pro-
considered that this mechanism can only be batch gramme was restricted to the K-Ar method.
or partial melting in a subduction zone and mantle Approximately 40 new mineral ages were deter-
wedge environment below the continental margin. mined on samples collected from the Barisan
A crustal contribution does not seem to have been a plutonic belt. All were fresh, undeformed, non-
factor in the geochemical evolution of the plutonic mineralized material collected at outcrop away
 MESOZOIC--CENOZOIC PLUTONISM IN SUMATRA 329

Norm: OGR Incorporating all other published ages from

100 , , , ! i , , i i , ! i w
exposed plutons in western Sumatra (Fig. 8), a
"ALL" SAMPLES
further plutonic event in the Permian (?287-
256 Ma) is indicated. Furthermore, either the
Jurassic plutonism began in the Late Triassic, or
,& alternatively there were two distinct magmatic
episodes, one in the Late Triassic to Early Jurassic
d (220-183 Ma) and one in the Mid-Jurassic to Early
Cretaceous (170-130Ma). The latter suggestion
is preferred because the ages correspond to
similar cycles of plutonism in peninsular Malaysia,
peninsular Thailand and Burma (Cobbing et al.
1992).
.01 I I I I I 1 t I i I I I |

(a) Sr KzO Fib Ba Th Ta Nb Oe HI Zr Sm Y Yb

Episodic plutonism in Sumatra:
Norm: MOFIB
1000
i i w , i i i , I ! ! 1 , , r , , ! ! ! i ~ a plate tectonic model
"ALL" SAMPLE S
Episode A (?287-256 Ma)
100
K A discontinuous Lower-Middle Permian volcanic
arc is present as an elongate fault-bounded strip
along the western edge of the Mergui microplate
"-- x
1D -.~ and is interpreted to be subduction related (Katili
1973; Pulunggono & Cameron 1984). Palaeonto-
logical evidence from sediments associated with
the andesitic volcanics indicates a warm climate
and Cathaysian affinities, in contrast to the
Gondwana characteristics of the remainder of
the Mergui microplate sedimentary sequences
.01 i i t i i i l i l i i l i , i l | | i i t i

R b B a T h U K N b L a C e S r N d H f Z r S m E u Ti G d B y Y E r Y b Lu
(Fomaine & Gafoer 1989). The volcanics and
(b) associated sediments are accordingly interpreted as
Fig. 7. Multi-element spider diagrams for the south a separate terrane, probably an oceanic arc, that was
Sumatra Barisan granitoids, normalized to (a) oceanic subsequently accreted to the Mergui continental
ridge granite and (b) mid-ocean ridge basalt. Area in margin (cf. Wajzer et al. 1991) through northerly
between solid lines shows compositional field of all directed subduction and the closure of a marginal
samples (n = 54). ocean basin in the Late Permian or, more likely,
Early Triassic. A single K-At age of 248 _+ 10 Ma
from these Lower to Middle Permian volcanics in
Sumatra (Nishimura et al. 1978) may approximate
from any obvious faults, shears, solution planes to the age of collision and accretion. The Sibolga
or sizeable dykes. The limitations and pitfalls of and Ombilin granites, with ages of 287-256 Ma,
K-Ar dating are well known to the authors, in may represent plutonism associated with this Late
particular that plutons or intrusive bodies dated Palaeozoic subduction but their present position
by a single sample should not be considered to be is probably exotic.
reliably dated. Most of the quoted ages are either
from duplicate samples and/or mineral pairs and are
Episode B (224-180 Ma)
considered reliable and interpreted as intrusive
ages. In addition, the plutonic episodes discussed The collision and accretion of the Permian
below (A-F) have ages which are in close agree- (Peusangan-Palepat) volcanic arc is interpreted
ment with those from other regional studies using to be part of a major terrane amalgamation event,
Rb-Sr mineral and whole rock isochron dating that included the collision of the continental East
techniques in addition to K-Ar methods (Cobbing Malaya and West Malaya (Sibumasu) Blocks along
et al. 1992). the Bentong-Raub Line (suture) in the earliest
At least four periods of plutonic activity have Triassic. The extensive granitic, S- and I-type,
been identified: Jurassic to Early Cretaceous magmatism of peninsular Malaysia, Thailand and
(203-130 Ma), Middle to Late Cretaceous (117- the Indonesian tin-islands was a direct consequence
80 Ma), Early Eocene (?60-50 Ma) and Miocene- of this collision (Mitchell 1977). Granite ages range
Pliocene (20-5 Ma) as outlined in Table 1. from 250-195 Ma but the majority are c. 220 Ma
330 W.J. McCOURT E T A L .
A MID - LATE PERMIAN ARC
B ( LATE ) TRIASSIC - EARLY JURASSIC ARC
C MID - JURASSIC - EARLY CRETACEOUS ARC
D MID- LATE CRETACEOUS ARC
E EOCENE ARC
F NEOGENI5ARC
30--
O
_Z
5: (E)
2o
c~ (F)
~ o )
~o
I I
1o 30 50 70 100
PLUTON AGES ( M a )
Fig. 8. Histogram of available plutonic ages from Sumatra, incorporating this study and previously published data.
(Cobbing et al. 1992) suggesting that most of the
plutonism was post-orogenic. In addition, a partly
coeval plutonic arc of subduction related I-type
granitoids can now be identified in western
Sumatra (219 _+4 to 183 _+ 13 Ma). Plutons of this
subduction-related arc intrude Palaeozoic meta-
sediments of the West Malaya (Sibumasu) terrane
deformed by the collision, supporting an Early
Triassic suturing age. Thus it is tentatively
suggested that the Early Mesozoic plutonism may
in fact be made up of two separate magmatic
events: an Early Triassic event dominated by
collision-related crustal S- and I-type granites in
peninsular Malaysia, Thailand and the Indonesian
tin-islands, and a Late Triassic to Early Jurassic
(220-180 Ma) post-collisional event. The latter is
represented by an I-type plutonic arc in western
Sumatra (Episode B) and coeval S-type and crustal
I-type granitic magmatism in the Main Range
Province, Indonesian tin-islands and eastern
Sumatra (Episode B1). This magmatism was
probably related to tectonic release and adiabatic
decompression, with resulting anatexis, in the
back-arc region, with granites channelled along
deep-seated faults. A postulated change in the con-
vergence angle of the oceanic plate resulted in a
more oblique subduction regime in the Early
Jurassic that brought the Episode (B-B 1) plutonism
to an end and resulted in transpressional strike-slip
along the Sundaland continental margin which was
taken up along older fault structures.
c> ::iiiiiiiiiiiYiiiSiiiiiJJ
(A)
~i iiiiiiiiiiJ :::::::::::::::::::::::
130 150 200 250
Episode C ( 1 6 9 - 1 2 9 Ma)
Middle Jurassic to Early Cretaceous plutonism
in Sumatra is represented by an extensive I-type,
subduction-related belt (Bungo batholith, Sulit Air
suite) focused along the western edge of the Mergui
microplate broadly coincident with, but laterally
more extensive than, the Episode B plutonic belt.
The plutonism appears to have been channelled
along the junction between the Permian volcanic
arc and the Palaeozoic continental margin meta-
sediments, interpreted as an Early Triassic suture.
This period of plutonism is correlated with north-
west-directed subduction beneath the Sundaland
continental margin, in line with postulated north-
westward spreading based on identified sea floor
magnetic anomalies in the eastern Indian Ocean
(Patriat & Achache 1984). This plutonic arc may
also extend north into the Shan scarp region of
Burma on the basis of limited geochronological
evidence (Cobbing et al. 1992). Episode C pluton-
ism terminated in the latest Early Cretaceous (or
early Middle Cretaceous) following the collision,
accretion and local obduction of the allochthonous
Woyla terranes of southern Sumatra, some 125 Ma
ago (McCourt et al. 1993). These ophiolitic rocks
can be correlated with similar rocks in western
Burma (Mawgyi andesites), Tibet (Donqiao ophio-
lite) and possibly SE Kalimantan (Alino Formation
and Meratus ophiolite), and it is likely that they
represent fragments of an oceanic arc system that
 MESOZOIC---CENOZOIC PLUTONISM IN SUMATRA
collided with, and was thrust over, the continental
margin of Sundaland at this time (cf. Mitchell
1993).
Episode D (120-75 Ma)
Subsequent to the accretion event, possible north-
directed subduction was reestablished and I-type
granitoid magmas were emplaced into the now
cratonized Woyla microplate, with the majority
of the plutons focused along the main suture line
and related faults (Sikuleh granite, Manunggal
batholith, Ulai intrusion, Garba pluton, Sulan
pluton). Based on Mitchell (1993), it is probable
that this subduction-related arc extends north into
Burma west of the Sagaing Fault, where the oldest
dated plutons are of mid-Cretaceous age. This
western Burma Arc (?equivalent to the Central
Valley Province of Cobbing et al. (1992)) is made
up of I-type granodioritic to tonalitic plutons, with
K-Ar ages of 106+7, 103+4, 9 8 + 4 , 9 4 + 4
and 91 __.8 Ma, and like its Sumatran equivalent
intrudes a sequence of deformed oceanic rocks,
basaltic andesites and basalt pillow lavas, the
Mawgyi andesites (Mitchell 1993). Broadly con-
temporaneous plutonism is also recorded from the
Western Province of Thailand and Burma (Cobbing
et al. 1992) and corresponds to a mixed population
of I- and S-type granites with high initial ratios
indicative of a significant crustal component in
most cases. Clarke & Beddoe-Stephens (1987)
proposed that this belt of Upper Cretaceous S- &
I-type crustal granites also extends into eastern
Sumatra, as indicated by the 80 Ma Hatapang
granite. It is suggested that this plutonism was
related to anatexis, the result of crustal thickening
accompanying thrusting that was contemporaneous
with subduction and VAG, I-type magmatism in the
Central Valley Province and western Sumatra.
Middle to Late Cretaceous magmatism continued
northwards through the Mogok Belt into Assam
(Mitchell 1993) and reported ages of 113-82 Ma on
the Gandise batholith in Tibet (Debon et al. 1986)
may indicate a further extension of this plutonism.
The general absence of plutonic rocks with ages
in the range 75-60 Ma, coincides with the well
documented latest Cretaceous deformation
throughout this region, including Sumatra (de
Coster 1974; Hamilton 1979; Cameron et al. 1980;
Pulunggono & Cameron 1984). Exactly why
plutonism ceased is not clear. The model proposed
here involves a suggested change from high angle
to oblique subduction along the continental margin,
related to a change in oceanic spreading patterns
and plate configurations, that resulted in the
accretion of a continental sliver, the West Sumatra
terrane. This terrane is now present as a series
of fragments such as Sikuleh, Natal and possibly
331
Bengkulu (Fig. 2) that correspond to the southerly
extension of the West Burma terrane of Metcalfe
(1994). Much of the evidence for this event, how-
ever, has since been destroyed during tectonism and
disruption related to the Early Tertiary collision
of India and Eurasia.
Episode E (60-?50 Ma)
Following the Late Cretaceous deformation event,
a new subduction regime was established along
the continental margin of Sumatra as evidenced by
a short-lived but extensive plutonic episode from
60-50 Ma (Episode E). This I-type, VAG plutonic
arc (Lassi pluton, Nagan granodiorite etc.) was
superimposed on the earlier Cretaceous and
Jurassic arcs via deep-seated older fault structures
in the continental margin which acted as magma
conduits. Limited regional evidence suggests that
this mainly Early Eocene (57-52 Ma) plutonism
extends into Burma and Thailand where it is of
combined I- & S-type character (Cobbing et al.
1992; Mitchell 1993). It is suggested that this
plutonic episode was brought to an end by the
Middle Eocene collision of India and Eurasia at
about 50 Ma, approximating to the timing of
proposed ophiolite emplacement in the Indo-
Burman Ranges of western Burma (Mitchell 1993).
A further important consequence of the collision of
India and Eurasia was the indentation and related
deformation of the Lower Tertiary margin of Asia
and the probable extrusion and clockwise rotation
of much of SE Asia, including Sumatra. The shape
of this margin prior to collision was, as suggested
by Tapponnier et al. (1986), a simple slightly
convex line extending from Sumatra to the western
Makran. As a preliminary model it is proposed that
this margin was characterized by a series of sub-
parallel, outwardly younging plutonic belts repre-
senting prolonged convergence and subduction-
related plutonism, along the margin, from the Early
Mesozoic to the Early Tertiary.
Episode F (30-0 Ma)
Subsequent to the India-Eurasia collision, and a
related major reorganization of plate motions
and spreading patterns in the Indian Ocean, NNE
directed subduction was established along the
Sundaland margin. Available plutonic ages from
Burma, 38 + 1 Ma (Mitchell 1993) suggest that
subduction-related activity was taking place along
the margin by the Early Oligocene, although
plutonism in Sumatra was apparently not estab-
lished until the Early Miocene (Episode F). Wajzer
et al. (1991) reported Late Oligocene ages (30-
28 Ma) from the Air Bangis granite of central
Sumatra, but concluded that these plutons, and
332 W.J. McCOURT ET AL.

/'),
J !
./
/
~ • (
/

\
v x
v
'/+~L .3
. + )

I
I

GRANITE PROVINCES

EASTERN, I - type of
Permian - Triassic age
r~ MAIN RANGE, S - type
of dominantly Triassic age
WESTERN, I & S-type of
Jurassic - Cretaceous & Tertiary age
~] ENTRAL VALLEY, I - type
Mid - Late Cretaceous age

q 0 ~

0 400km
L I

Fig. 9. The main granite provinces of SE Asia, modified from Cobbing et al. (1992) for Sumatra. Inset shows the
principal tectonostratigraphic terranes of SE Asia based on Metcalfe (1988, 1990).
 MESOZOIC---CENOZOIC PLUTONISM IN SUMATRA 333

contemporaneous volcanics of the Langsat volcanic
arc, had formed elsewhere along the Sundaland
margin and were tectonically juxtaposed against
the Woyla Group of Sumatra sometime prior to the
Middle Miocene. Rock et al. (1983) proposed that
the Langsat Volcanics were of Palaeogene age, a
conclusion confirmed by Wajzer et al. (1991), who
assigned them a Late Eocene to Early Oligocene
age on the basis of whole rock K-Ar dates (40-
38 _+ 1 Ma). Thus the proposed age of the Air
Bangis plutonism is almost identical to that noted
above from Burma. The younger dates (30-28 Ma)
from the Air Bangis granites could therefore reflect
their collision with and accretion to the Sumatran
margin and this could relate to the proposed mid-
Oligocene collisional event responsible for the
recorded inversion in the forearc basins of Sumatra
and Java as proposed by Daly et al. (1991).
Following this mid-Oligocene event, widespread
andesitic volcanism was established in Sumatra and
the main Neogene magmatism of the Barisan arc
was initiated. Subduction-related VAG plutonism
was widespread by the end of the Early Miocene
and in the Middle Miocene the entire Barisan
arc became volcanic. Middle Miocene to Early
Pliocene, I-type granitoid plutons are essentially
confined to the western Barisan Ranges in Sumatra,
focused along the Semangko Fault strand of the
SFS, approximately coinciding with the exposed
western limits of the Upper Mesozoic Woyla
terranes and the main axis of the Upper Oligocene
volcanic arc. The Neogene plutonic episode is also
recorded from the Burma Arc and possibly the
Mogok Belt (Mitchell 1993).
Conclusions
It is proposed that the Mesozoic-Cenozoic
plutonic evolution of Sumatra was episodic and
that five plutonic episodes can be recognized: Late
Triassic-Early Jurassic, Middle Jurassic-Early
Cretaceous, Middle Cretaceous-Late Cretaceous,
Early Eocene and Miocene-Pliocene. Each episode
was dominated by I-type, subduction-related
plutonism. Granitoids from different episodes are
very similar petrographically and geochemically
and are of VAG character. This is interpreted to
mean that Sumatra has been the site of an active
convergent margin and of long-lived subduction,
albeit episodic, since the Early Mesozoic. The
majority of the plutonic episodes identified from

N [~ MID- LATE
(EUR)ASIA , . ~ X __ ~ "J CRETACEOUS ARC

._4_ ,~• ~ ,~ -----<._×

- - - . . - , < ,~ ;"..... :"- .>----------~.. |
~, --.<.:-.>..'a.:~.: . : .-. , ~ ...........
• ",": ~ " . : '.:.' '.... I

, ,-~!...:.;.,.....,

,/f \k "

Fig. 10. Cartoon, not to scale, illustrating a possible plate-tectonic setting along part of the Asian margin prior to
the Eocene collision of India (modified from Metcalfe (1990). SE Asian terranes as follows: SC, South China;
C, Qiangtang; I, Indochina/East Malaya; L, Lhasa; S, Sibumasu; WB, West Burma; WY, Woyla.
334 W.J. McCOURT ET AL,
Sumatra can be recognized, albeit in disrupted
form, throughout m u c h of SE Asia (Fig. 9). Thus
the Triassic to Early Jurassic plutonism correlates
with the Eastern and Main Range Granite Provinces
of Thailand and Malaysia, whereas the Middle
Jurassic and Cretaceous plutonism can be corre-
lated with a combination of the Western and
Central Valley Provinces of Thailand and Burma.
It is suggested as a preliminary model that prior to
the Eocene collision of India the Sundaland margin
was orientated approximately W N W and made up
of a series of outwardly younging subduction-
related plutonic arcs (Fig. 10), some of which
probably extended along the southern margin of the
Asian plate. The present distribution and geometry
of these arcs in SE Asia is the result of the effects
of the collision of India and Eurasia, i.e. inden-
tation, extrusion and strike-slip faulting, as pre-
dicted by the model of Tapponnier et al. (1982,
1986). The more recent major dextral strike-slip
m o v e m e n t s along the SFS and related master faults
outside of Sumatra, the onset of which coincided
with the opening of the A n d a m a n Sea c. 11 Ma ago
References
ASPDEN, J. A., KARTAWA,W., ALDISS, D. T., DJUNUDDIN,
A., WHANDOYO, R., er AL. 1982. The geology of
the Padangsidempaun and Sibolga Quadrangle,
Sumatra (1:250,000). Geological Research and
Development Centre, Bandung.
AUDLEY-CHARLES,M. G., BALLANTYNE,P. D. & HALL, R.
1988. Mesozoic-Cenozoic rift-drift sequence
of Asian fragments from Gondwanaland.
Tectonophysics, 155, 317-330.
BR1QUEU, L., BOUGAULT, n. & JORON, J. L. 1984.
Quantification of Nb, Ta, Ti and V anomalies in
magmas associated with subduction zones: petro-
genetic implications. Earth and Planetary Science
Letters, 68, 297-308.
BROWN, G. C., THORPE, R. S. & WEBB, P. C. 1984. The
geochemical characteristics of granitoids in con-
trasting arcs and comments on magma sources.
Journal of the Geological Society, London, 141,
413-426.
CAMERON, N. R., CLARKE, M. C. G., ALDISS, D. T.,
ASPDEN, J. A. & DJUNUDDIN, A. 1980. The geo-
logical evolution of northern Sumatra. In:
Proceedings Indonesian Petroleum Association 9th
Annual Convention. 149-187.
CHAPPELL, B. W. & WHITE, A. J. R. 1974. Two con-
trasting granite types. Pacific Geology, 8, 173-174.
CLARKE, M. C. G. & BEDDOE-STEPHENS, B. 1987.
Geochemistry, mineralogy and plate tectonic setting
of a Late Cretaceous Sn-W Granite from Sumatra,
Indonesia. Mineralogical Magazine, 51, 371-387.
COBBING, E. J., PITEIELD,P. E. J., DARBYSHIRE,D. E E
& MALLICK, D. I. J. 1992. The granites of the
Southeast Asian tin belt. British Geological Survey,
Overseas Memoir, 10.
CURRA¥, J. R., MOORE, D. G., Lawver, L. A., EMMEL,
E J., RArrr, R. W., er AL. 1979. Tectonics of
(Curray et al. 1979), has further complicated this
scenario.
We also propose that breaks in plutonic activity
correspond to periods of oblique approach, that, in
some instances, relate to the collision and accretion
of allochthonous terranes. We further suggest that
one of the underlying factors that controls the
development and siting of the various plutonic
arcs is the availability of deep-seated faults along
the continental margin, that probably extend down
to the site of m a g m a generation at or close to the
subduction zone.
This paper is published with the permission of the
Directors of the Geological Research and Development
Centre, Bandung, and the British Geological Survey,
Nottingham. The work in Sumatra was carried out as part
of a bilateral technical cooperation project between the
governments of Indonesia and the United Kingdom and
funded jointly by the Indonesian Directorate General of
Geology and Mineral Resources (DGGMR) and the
Overseas Development Administration (ODA) of the
British Foreign Office. We thank A. H. G. Mitchell and
S. J. Moss for suggestions which improved the text.
the Andaman Sea and Burma. In: WATKINS, J.,
MONTADERT, L. & DICKENSON, P. W. (eds)
Geological and Geophysical Investigations of
Continental Margins. AAPG Memoir, 29, 189-198.
DALY, M. C., COOPER,M. A., WILSON, I., SMITH,D. G. t~
HOOPER, B. G. D. 1991. Cenozoic plate tectonics
and basin evolution in Indonesia. Marine and
Petroleum Geology, 8, 1-21.
DEBON, E., LE FORT, P., SHEPPARD,S. M. E & SONET, J.
1986. The four plutonic belts of the Transhimalaya-
Himalaya: a chemical, mineralogical, isotopic and
chronological synthesis along a Tibet-Nepal section.
Journal of Petrology, 27, 219-250.
DE COSTER, G. L. 1974. The geology of the Central
and Southern Sumatra Basins. In: Proceedings
Indonesian Petroleum Association 3rd Annual
Convention. 77-110.
EUBANK, R. T. & MAKK1,A. Ch. 1981. Structural geology
of the Central Sumatra Back-Arc Basin. In:
Proceedings Indonesian Petroleum Association,
lOth Annual Convention. 153-196.
FOLEY, S. E • WHELLER, G. E. 1990. Parallels in the
origin of the geochemical signatures of island arc
volcanics and continentaI potassic igneous rocks:
the role of residual titanates. Chemical Geology, 85,
1 18.
FONTAINE,H. & GAFOER,S. 1989. The pre- Tertiary fossils
of Sumatra and their environments. CCOP
Technical Secretariat Publication, Bangkok,
Thailand.
GAFOER, S. 1990. Tinjauan Kembali Tataantratigrafi
Pratersier Sumatra Bagian Selatan. [A review of
the pre-Tertiary sequences of Southern Sumatra.
Abstract in English.] In: Prosiding Persidangan
Sains Bumi dan Masyarakat Universiti Kebangsaan
Malaysia, 9-10 July 1990.
 MESOZOIC--CENOZOIC PLUTONISM IN SUMATRA
-- & PURBO-HADIWIDJOYO,M. M. 1986. The geology
of southern Sumatra and its beating on the occur-
rence of mineral deposits. Bulletin of the Geological
Research and Development Centre Bandung, 12,
15-30.
HAMILTON,W. 1979. Tectonics of the Indonesian Region.
USGS Professional Paper, 1078.
HUTCHISON,C. S. 1989. Geological evolution of Southeast
Asia. Clarendon Press, Oxford.
1994. Gondwana and Cathaysian blocks,
Palaeotethys sutures and Cenozoic tectonics in
Southeast Asia. Geologische Rundschau, 82,
388-405.
KATILI, J. A. 1973. Geochronology of west Indonesia and
its implications on plate tectonics. Tectonophysics,
19, 195-212.
KONING, T. & DARMONO, F. X. 1984. The geology of
the Beruk Northeast field, central Sumatra: oil
production from pre-Tertiary basement rocks. In:
Proceedings Indonesian Petroleum Association
13th Annual Convention, 385-406.
McCOURT, W. J. & COBBING, E. J. 1993. The geo-
chemistry, geochronology and tectonic setting of
granitoid rocks from southern Sumatra, western
Indonesia. Southern Sumatra Geological and
Mineral Exploration Project, Report Series, 9,
Directorate of Mineral Resources/Geological
Research and Development Centre, Bandung,
Indonesia.
, GAFOER, S., AMIN, Z. C., ANDI MANGGA, S.,
KUSNAMA,ET AL. 1993. The geological evolution of
Southern Sumatra. Southern Sumatra Geological
and Mineral Exploration Project, Report Series,
13, Directorate of Mineral Resources/Geological
Research and Development Centre, Bandung,
Indonesia.
METCALFE, I. 1988. Origin and assembly of Southeast
Asian continental terranes. In: AUDLEY-CHARLES,
M. G. & HALLAM, A. (eds) Gondwana and Tethys.
Geological Society, London, Special Publication,
37, 101-118.
- 1990. Allochthonous terrane processes in Southeast
- Geology, 10, 611-616.
Asia. Philosophical Transactions of the Royal
Society, 331A, 625-640.
1994. Gondwanaland origin, dispersion and
accretion of East and Southeast Asian continental
terranes. Journal of South American Earth Sciences,
7, 333-347.
MrrCHELL, A. H. G. 1977. Tectonic settings for emplace-
ment of Southeast Asian tin granites. Bulletin of the
Geological Society of Malaysia, 9, 123-140.
1981. Phanerozoic plate boundaries in mainland
SE Asia, the Himalayas and Tibet. Journal of the
Geological Society, London, 138, 109-122.
- 1993. Cretaceous-Cenozoic tectonic events in the
- 1991. Accretion, collision and strike-slip faulting:
western Myanmar (Burma)-Assam region. Journal
of the Geological Society, London, 150, 1089-1102.
NISHIMURA,S., SASAJIMA,S., HIROOKA,K., THIO, K. H. &
HEHUWAT, E 1978. Radiometric ages of volcanic
products in the Sunda Arc. In: SASAJIMA, S. (ed.)
Studies of the physical geology of the Sunda Arc.
University Press, Kyoto, 34-37.
335
PATRIAT, P. & ACHACHE, J. 1984. India-Eurasia collision
chronology has implications for crustal shortening
and driving mechanisms of plates. Nature, 311,
615~521.
PEARCE, J. A., HAm~tS, N. B. W. & TINDLE, A. G. 1984.
Trace element discrimination diagrams for the
tectonic interpretation of granitic rocks. Journal of
Petrology, 25, 956-983.
PITCHER, W. S., ATHERTON, M. R, COBBING, E. J. &
BECKINSALE, R. D. 1985. Magmatism at a plate
edge-the Peruvian Andes. Blackie, Glasgow &
Halstead Press, New York.
PITFIELD, P. E. J., COBBING, E. J., CLARKE, M. C. G.,
MALLICK, D. I. J. & TEOH, L. H. 1986. Granitic
provinces in the Southeast Asia Tin Belt. In:
Transactions of the Fourth Circum-Pacific Energy
and Mineral Resources Conference Singapore.
575-589.
PULUNGGONO, A. & CAMERON, N. R. 1984. Sumatran
microplates, their characteristics and their role in the
evolution of the Central and South Sumatra Basins.
In: Proceedings Indonesian Petroleum Association
13th Annual Convention. 121-143.
ROCK, N. M. S., ALDISS, D. T., ASPDEN, J. A., CLARKE,
M. C. G., DJUNUDDIN,A., ETAL. 1983. The geology
of the Lubusikaping Quadrangle Sumatra.
Geological Research and Development Centre,
Bandung.
SILITONGA, P. H. & KASTOWO. 1975. Geological map of
the Solok Quadrangle, Sumatra. Geological Survey
of Indonesia, Bandung.
SIMANDJUNTAK,T. O., SURONO,GAFOER, S. & AMIN, T. C.
1991. The geology of the Muarabungo Quadrangle,
Sumatra. Geological Research and Development
Centre, Bandung.
STAUFFER, P. H. 1983. Unravelling the mosaic of
Palaeozoic crustal blocks in Southeast Asia.
Geologische Rundschau, 72, 1061-1080.
--, LE DAIN, A. Y., ARMUO, R. & COBBOLD,P.
1982. Propagating extrusion tectonics in Asia, new
insights from simple experiments with plasticene.
TAPPONNIER, P., PELTZER, G. & ARMIJO, R. 1986. On the
mechanics of the collision between India and Asia.
In: COWARD, M. P. & RIES, A. C. (eds) Collision
Tectonics. Geological Society, London, Special
Publication, 19, 115-157.
VAN BEMMELEN, R. W. 1949. The Geology of Indonesia.
Martinus Nijhoff, The Hague.
vAN Es, L. J. C. 1919. De tectoniek van het westelijk helft
van den Nederlandsch Oost Indischen Archipel.
Jaarboek Mijnwezen Nederlandsch Oost Indies,
Verhandelingen, 5-143.
WAJZER, M. R., BARBER,A. J., HIDAYAT,S. & SUHARSONO.
the Woyla Group as a key to the tectonic evolution
of north Sumatra. Journal of Southeast Asian Earth
Sciences, 6, 447--461.
WESTERVELD,J. 1952. Phases of mountain building and
mineral provinces in the East Indies. Report of the
18th International Geological Congress, London,
13, 245-255.
 The Mentawai fault zone and deformation
of the Sumatran Forearc in the Nias area

M . A. S A M U E L & N. A. H A R B U R Y
SE Asia Research Group, Research School of Geological and Geophysical Sciences,
Birkbeck College and University College London, Gower Street, London W C I E 6BT, UK.

Abstract: The Sumatran Forearc is the site of oblique plate convergence between the Indo-
Australian and SE Asian-Eurasian Plates. It is generally accepted that the forearc sliver is not
behaving as a rigid plate and that the rate of slip increases along the fight-lateral Sumatran Fault
System from southeast to northwest. Two contrasting hypotheses have been invoked to explain
this pattern of decoupling: by using slip vectors to suggest arc-parallel stretching; and a second
major right-lateral strike slip zone, parallel to the Sumatran Fault System, to accommodate
the oblique subduction. This zone, termed the Mentawai fault zone, lies just to the east of the
outer-arc ridge and is indicated as intersecting with other faults on Nias Island. In this paper data
are presented from Nias where a thick forearc succession is excellently exposed in three sub-
basins with half-graben geometries and broadly comparable Palaeogene-Recent histories. Mud
diapirism has generated melanges which cut the Oligocene to Recent strata. Outcrop data,
LANDSAT, Synthetic Aperture Radar and aerial photographs were used to determine fault offset
patterns, and marked extension of the island along north-striking right lateral faults and ESE
striking left-lateral faults is revealed. Extension was further accommodated along a set of ENE
striking normal faults. In the region where the proposed Mentawai fault zone comes onto Nias
there is an important, long-lived, basin-bounding fault. This structure had in excess of 5 km
normal throw during the Oligocene-Miocene whilst minor contractional reactivation of the fault
occurred during the Pliocene phase of uplift and deformation which affected all of Nias. Data
from onshore and offshore the Nias region suggest that all the features visible from seismic
sections collected over the Mentawai fault zone to the south of Nias can be explained in terms of
inversion of originally extensional structures and mud diapirism. It is suggested that strike-slip
motion is of limited importance along the 600 km long Mentawai fault zone. Rather this highly
structured zone represents a deformation front whose origin can be explained by Pliocene to
Recent subduction-driven inversion at the outer margin of the forearc.

The Sumat~an Forearc has been recognized as a to n o r t h - s o u t h ( N e w c o m b & M c C a n n 1987;

type example of an obliquely convergent margin for
over twenty years (e.g. Fitch 1972). In recent years
two contrasting models of deformation of the
Sumatran Forearc have been proposed (McCaffrey
1991; Diament et al. 1992). This paper outlines
these m o d e l s and presents new data on the
Sumatran Forearc, from the area of Nias, which
allow a critical assessment of these two models.
Furthermore the results lead to an improved under-
standing of the d e f o r m a t i o n processes acting
within, and at the margin of, the forearc.
Deformation of the Sumatran Forearc
The Sumatran Forearc forms part of the Sunda
subduction system which extends from Sumba in
the east to B u r m a in the north (Figs 1 and 2; Moore
et al. 1980b; Curray 1989). Plate tectonic models
predict convergence rates varying from 7.8 c m a -1
near S u m b a w a to 6 c m a -1 near the A n d a m a n
Islands (Minster & Jordan 1978; DeMets et al.
1990) with the direction of convergence close
McCaffrey 1991).
The Sumatran arc has a classic morphology of
trench, accretionary prism, outer-arc ridge, forearc
and volcanic chain with active andesitic volcanism
(Karig et al. 1979) and seismicity displays a
distinct Benioff zone (Page et al. 1979). Recent
work has shown that the outer-arc ridge along
Sumatra comprises the outer part of the forearc
rather than forming part of the accretionary prism
which is located further to the southwest (Samuel
1994; Samuel et al. 1995). The outer-arc ridge is
wholly submerged off Java whereas the ridge gains
sub-aerial expression in the Sumatran Forearc. A
further important difference between the Javan and
Sumatran forearcs is the subduction direction
which varies from nearly orthogonal off Java to
oblique west of Sumatra where the trench strikes
N140°E.
Fitch (1972) proposed that oblique convergence
in the Sumatran area could be a c c o m m o d a t e d by
the right-lateral Sumatran fault zone. This fault
zone connects to the A n d a m a n Sea in the northwest

From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, 337
Geological Society Special Publication No. 106, pp. 337-351.
338 M.A. SAMUEL • N. A. HARBURY
where extension and sea floor spreading is an
extreme effect of oblique convergence (Fig. 2;
Curray et al. 1979). To the southeast the zone
extends to the Sunda Strait where further extension
is occurring (Huchon & Le Pichon 1984; Harjono
et al. 1991) and where it probably crosses the
forearc (Curray 1989). A sliver plate (the Burma
Sliver Plate; Curray et al. 1979) is thus decoupled
from the Eurasian and Indo-Australian plates and
moves northwest with respect to the Eurasian Plate.
This plate has been called the Sumatran Sliver Plate
or forearc sliver in the Sumatran area (Jarrard 1986;
Diament et al. 1992).
McCaffrey (1991) presented slip vector data that
suggested the oblique convergence in the Sumatran
area was not completely decoupled. The results
indicated increasing decoupling northwestwards
(Fig. l a). Modelling of the pattern of slip vectors
by McCaffrey (1991) suggests that the forearc
is not behaving as a rigid plate but is subject to
arc-parallel stretching at a uniform strain rate of
3--4 x 10-8 per year. He showed that the north-
westward motion of the forearc relative to the SE
Asian plate should increase from near zero at the
Sunda Strait to 4 5 - 6 0 m m a -1 in northwest
Sumatra. McCaffrey (1991) acknowledged that the
deformation processes acting to stretch the forearc
were poorly understood. Shallow strike-slip earth-
quakes in the forearc have nodal planes that strike
north and east across the forearc rather than
parallel to it (McCaffrey 1991) and it was
suggested the deformation of the forearc was
most likely to occur on strike-slip faults crossing
the margin than on arc-parallel strike-slip faults
or arc-perpendicular normal faults.
There are a number of geological indications that
support the concept that deformation must be
occurring in the Sumatran Forearc. For example
there does not appear to have been sufficient move-
ment along the Sumatran Fault Zone to account for
the 460 km of opening of the Andaman Sea in the
last 13 Ma (Curray et al. 1979). Indeed Bellier et al.
(1991) suggest that the rate of slip along the
southern end of the Sumatran Fault zone is only
6 mm a-1 which is considerably less than the
40 mm a-1 rate of opening of the Andaman Sea
(Curray et al. 1979).
Diament et al. (1992) proposed an alternative
solution to explain the decrease in rate of slip to
the southeast along the Sumatran Fault system. A
single channel seismic reflection survey led to the
identification of a complex set of structures which
they correlated for 600 km along the outer margin
of the Sumatran forearc and named the Mentawai
fault zone (Fig. l b, c). These features were
apparent on all seismic sections from southern
Sumatra to Siberut although the detailed structure
of the zone was highly variable; from north to
south, faulted anticlines, faulted blocks, horst and
graben systems and flexures were all described.
Diament et al. (1992) interpreted the structures as
comprising a right-lateral strike slip zone (Fig. lc).
They based their arguments on three points: (1) it is
possible to interpret some features, identified on
seismic lines, as positive flower structures; (2) the
fault zone is straight and appears to run con-
tinuously for hundreds of kilometres; (3) a pro-
nounced difference is commonly found in the depth
to acoustic basement on either side of the fault
zone. Diament et al. (1992) further suggested that
displacement on the Mentawai fault zone may be
relayed to the Batee fault which they tentatively
map as linking with the Mentawai fault zone on
Nias (Fig. lc). Nias is therefore of prime interest as
it potentially lies at the intersection of two major
fault zones. Furthermore it is the only island along
the outer-arc ridge which is cut by the Mentawai
fault zone (Fig. lc).
Deformation in the Nias area
Previous work by Moore & Karig (1980) suggested
that Nias comprised part of an uplifted accretionary
complex. Two main stratigraphic units were
defined. The lower unit, the 'Oyo Complex' was
interpreted as a tectonic melange that formed prior
to the Early Miocene, when deep marine sediments
were unconformably deposited above it in trench
slope basins (Moore et al. 1980a). The Miocene to
Pliocene sedimentary succession was reported as
a shallowing-up sequence. Uplift was continuous,
with compressional deformation occurring largely
along arcward-dipping thrust and reverse faults.
More recently Pubellier et al. (1992) have
suggested that Nias comprises part of an Eocene
Tethyan suture zone. Their interpretation is based
largely on the apparent recognition of in situ
Eocene rocks forming part of the non-ophiolitic
sedimentary succession on the island. The research
of Samuel (1994) and many previous workers (e.g.
Douville 1912; Moore et al. 1980a) shows how-
ever that the rocks in question are of Upper
Oligocene and Lower Miocene; they contain both
a reworked Eocene fauna and younger microfossils.
As part of an integrated study of the Sumatran
Forearc by the University of London SE Asia
Research Group the structural and stratigraphic
evolution of the whole of Nias has recently been
studied in detail (Samuel 1994; Samuel et al. 1995).
The studies have led to a substantial reinterpret-
ation of the sedimentological and structural setting
of the island and the work has shown that Nias
consists of three main sub-basins and a basement
high (Fig. 2). These detailed sedimentological,
biostratigraphic and palaeobathymetric studies of
the successions on Nias have led to the construction
 SUMATRAN FOREARC & MENTAWAI FAULT ZONE 339

Bengal
Fan

.................
~ Stretching vectors
(Forearc relative to SEA
Plate)
Slip vectors
---4> Plate vectors
-- IO°N --e

S.E. Asian
Plate

3- j

~-~ Borneo
-- 0o
L

Sunda
Strait
East Java Sea
Sumbawa
Indian Ocean

- - 10os ~

90OE IO0OE 110OE , ~. ,;^1

t J p Ao .,ha,,
~ I \ Sin;kel Fault~ t I--Eql I-'~'~'11
0-"~_~ [Singkel"%~.~ "~ - I t - ) II \ I ,'-, II
~anyakl% ,[ basra ~ "-" . X " _ ' '1 \ ' '1

% / 5" I~__~ \ ~,,~'e,> Sumatra I \ X \ \ I

I \ \ }~,

~ "~"~ Quaternaryandl -
~a ~ activev°lcan°es-AI • \ ~ \ t l

India cean
o oeoo, o I I
Fig. 1. (a) Location of the Sunda Arc and major structural features in the region. Tectonic features, slip vectors and
plate motion vectors after McCaffrey (1991). Only the azimuth of individual slip vectors is shown. These slip vectors
show a clockwise rotation from the Sunda Strait northwestwards along Sumatra. Modelling of the slip vectors
indicates that the forearc is subject to arc-parallel extension as shown by the stretching vectors. (b) Structural features
in the Nias area. The location of the 'Mentawai fault zone' suggested by Diament et al. (1992) is shown and other
offshore faults have been mapped from seismic profiles by Karig et al. (1980) and Matson & Moore (1992). The
faults onshore Sumatra have been mapped during field studies (e.g. Page et al. 1979). (c) Model of Diament et aL
(1992). Right-lateral motion is taken up both along the Sumatran Fault System and the parallel-orientated Mentawai
fault zone (Mfz). Right-lateral movement is transferred via the Batee Fault (BF) which transects the forearc.
340 M . A . SAMUEL 8Z N. A. HARBURY

L [ •

t
Recent alluvium

1 °30'N ~ 1 II [i~ Elevated reef compl. . . .

Upper Pliocene-Pleistocene

Middle Miocene-Lower Pliocene

Y Lower-Middle Miocene shallow marine sediments
Oligocene-Lower Miocene deep marine sediments

Sediment & Melange Complex

Basement (B)

-~unungsitoli

Gido

\ Line of
\° X-section
\ (Fig. 4 )
\\ ",.\
Major anticline
_ 1ON \,
N Major synciine

Faults
J
Active mud
volcanoes

"B

N
A

30'E
0 ~ 20 km

Fig. 2. Simplified geological map of Nias showing major structures identified on the island. The map was compiled
from field traverse data, aerial photographs, SAR (synthetic aperture radar) and LANDSAT images. The line of the
'flexure' in southeastern Nias is highlighted and the line of cross-section, shown in Fig. 4, is indicated. Key localities
referred to in the text are also plotted. A large area of the Mujoi Sub-basin has been mapped as 'sediment and melange
complex'. This complex is composed of deformed sedimentary sections, predominantly of Oligocene and Lower
Miocene age, intruded by melanges. The individual melange intrusions cannot be shown accur~/tely on maps
of this scale. Two active mud volcanoes on Nias are also located. The inset shows the position of the three sub-basins
and the basement high on Nias.
 SUMATRAN FOREARC 8,~ MENTAWAI FAULT ZONE 341

of a new stratigraphy for Nias (Fig. 3). When these One of the fundamental differences in the new
studies are combined with structural data collected stratigraphy for the forearc compared, for instance,
over the entire island a new picture of the geo- with the scheme of Moore e t al. (1980a), is the
logical evolution of Nias emerges (Fig. 3). There recognition that the Oligocene and Lower Miocene
is strong evidence that the Oligocene to Early sediments on Nias form part of a single, strati-
Miocene history of the Sumatran Forearc was graphically continuous, sedimentary succession
extensional/transtensional and below we outline (Fig. 3). Workers such as Moore & Karig (1980)
some of the key points of this history before believed that the Oligocene deep marine sedi-
concentrating on specific aspects of the deforma- mentary rocks only occurred in the melanges
tion of Nias. whereas this study, particularly in the central part of

N NN T Lahewa Mujoi Gomo Mola I Structural

AGE c) ~,,D I Sub-basin Basement
Sub-basin I Sub-basin I High I events

Quaternary (7) Rapid uplift

continues with

(6) Pliocene uplift

(5) Regional
subsidence

(4) Early M i o c e n e
uplift

(3) Extension of
basement
& development
of basins
under transtension

~
P18 NP22

P16 Imr'~Lu I
.~ ~ :. ~. . . .~.. . +,.~ ~. /.". ~. . . . .~ ..~ ~......

. . . . . . . . .
~ ~ ," -%
.~ . ./ . . "~ = s.:~..,,, j..:-...~....
~ .................

"" -- -- "
<..~'
.................

(2) Emplacement
& uplift
of b a s e m e n t

I~
o Mid
P" .......... i~;');~i:i~i!i~i~,,.i~!~!i
i, ii~~?,'!i~'~/~i~ii,~,,~!~:~,, ;L,E! !i ~;;!; i;:!,/~,
¸i,,~' ~ i,ii! ;)i;~ i/i!~i
........ ii Formation
,1,hydrothermal
: , ~! ! . : : i:'~ O,UnKnownor,9,n alteration of
I ~ leo ~ ~ ocean,corust
and mantle

~[ = Ophio[itic basement, [I = Ohgocene-Lower Miocene deep marine sedimentary rocks, I I ] = Tufts, I V = Oligocene-Lower

1
Miocene conglomerate marker, V = L o w e r - M i d d l e Miocene shallow marine limestone-dominated sediments, V[ = M i d d l e
Miocene-lowest Pliocene shelfaI to upper bathya] sedimentary rocks, V I I = diatomites, V I I I = M i d d l e Miocene-lowest Pliocene
tuff marker, IX = M i d d l e Miocene-lowest Pliocene limestone marker, X = U p p e r Pliocene-Pleistocene shallow marine
siliciclastics, XI = U p p e r P]iocene-P]eistocene elevated reef complexes.

Fig. 3. Chronostratigraphy for Nias. The scheme is based on detailed sedimentological, palaeobathymetric and
biostratigraphic data collected during four field seasons. Over 350 samples from across the island were dated. N, P,
planktic foraminiferal zones (Bolli et al. 1985); NN, NP, nannofossil zones (Bolli et aL 1985); and T, far eastern
benthic foraminiferal letter stages (Adams 1984). The large numbers refer to the structural events detailed in the right
hand column. The Roman numerals refer to the relevant formations and marker horizons. Sub-basin bounding faults
are marked by double vertical lines and the single vertical lines represent faults which transect the sub-basins. Both
the SE-striking sub-basin bounding faults and the transecting north- and ESE-striking faults were important in
controlling sub-basinal sedimentation, particularly during the Early Miocene.
342 M.A. SAMUEL (~ N. A. HARBURY
Nias, has identified sections where Oligocene rocks
pass conformably up into the Lower Miocene, as
suggested earlier by Elber (1939) and Paul (1941).
Contrary to previous interpretations, detailed
studies of the melanges on Nias have revealed that
they intrude and include strata of Oligocene to
Pleistocene age (Samuel et al. 1995). The melanges
formed by mud diapirism, which continues to the
present day as shown by active mud vulcanism in
the north of the island (Fig. 2). The melanges do not
therefore form basement to the sedimentary succes-
sions on Nias. The basement, in at least western and
central parts of Nias, is ophiolitic in nature. This
is deduced from the presence of apparently intact
sections of ophiolitic material along the western
coast of Nias and the occurrence of ophiolitic
material in many of the diapiric melanges across
the island (Fig. 2; Samuel et al. 1995). The
youngest age for the ophiolite complex is Mid-
Eocene, as dated by radiolarian cherts. The ophio-
lite complex was therefore emplaced between the
Mid-Eocene and Oligocene when deep marine
sedimentary rocks were unconformably deposited
above it (Fig. 3). The earliest structures recorded in
these sediments are extensional and sedimento-
logical, and stratigraphic criteria reveal that the
Oligocene and Lower Miocene sediments were
deposited in a set of half-grabens on a slope leading
down to the Oligocene to Early Miocene trench to
the southwest (Samuel et al. 1995). As extension/
transtension proceeded during the Early Miocene,
topographic variations were produced by trans-
tensional faults that cut across the strike of the
sub-basins. This is recorded in the stratigraphy
of Nias, since shallow marine Lower Miocene
carbonates are found along-strike of coeval deep
marine facies, particularly in the Gomo Sub-basin
(Fig. 3). Facies in the carbonates such as P o r i t e s
bindstones and floatstones, algal bindstones and
oncolitic packstones prove that these carbonates
must have been deposited in inner to mid-shelfal
environments as suggested by a number of previous
workers (e.g. Elber 1939; Paul 1941; Burrough &
Power 1968; Bradley 1973; Harbury & Kallagher
1991).
Localized uplift of some western parts of Nias
occurred during the Early Miocene whilst sedi-
mentation was uninterrupted in other areas (Fig. 3;
Samuel et al. 1995). The uplift was driven by
reverse movements on originally extensional and
transtensional faults. Regional subsidence com-
menced during the Mid-Miocene; deposition was
initiated above the Mola Basement High and
carbonate sedimentation, in areas such as the Gomo
Sub-basin, gave way to shelfal and upper bathyal
siliciclastic-dominated sedimentation despite a
global fall in sea-level (Fig. 3; Haq et al. 1987).
The main phase of uplift and deformation on Nias
occurred during the Pliocene (Fig. 3). The majority
of the faults that had controlled the earlier
extensional/transtensional geometry on Nias were
reactivated (Fig. 4).
The depth to basement in easternmost Nias
(the Mola Basement High) is about 2 km (Fig. 5).
Carbonates, sitting directly above basement, can be
identified on a number of seismic sections across
the Mola Basement High and they can be tied to
Middle Miocene carbonates recovered from the
Suma Well (Fig. lb; Karig et al. 1980). Equivalent
carbonates can also be examined in rivers such as
the Gawo in eastern Nias. The Upper Miocene
succession thickens to the southwest (Fig. 5) and
both the Upper Miocene and the Pliocene un-
conformity can be directly traced to the surface
(Fig. 6). Workers such as Moore & Karig (1980)
have remarked on the steepening of the Upper and
Middle Miocene sedimentary rocks from the
northeast to the southwest (from the Mola Base-
ment High to the Gomo Sub-basin; Figs 4 & 6) and
have termed the feature a large homocline or
flexure. This flexure has been interpreted by Moore
& Karig (1980) as the surface expression of a
reverse fault with a steep west dip. In contrast,
Diament et al. (1992) suggest that the 'flexure' is a
continuation of the Mentawai strike-slip fault zone
they identified to the southeast.
The results of this study strongly support the
observations made by Moore & Karig (1980) that
the feature is the surface expression of a major
southwesterly dipping fault. Rather than having a
net reverse throw, however, this fault must have a
large extensional throw to accommodate the thick
succession of Lower Miocene and Oligocene sedi-
ments to the southwest; a stratigraphic thickness
of at least 3.3 km of monoclinally tilted northeast
dipping Lower Miocene-Oligocene sediments is
exposed in the Gawo River (Fig. 6). These
sediments are underlain by a further c. 2 km of
Oligocene sediments (Fig. 4; Samuel & Harbury
1996). This thick Oligocene to Lower Miocene
succession is clearly not present above the Mola
Basement High and this high does not continue into
the region of the Gomo Sub-basin (Fig. 4).
The fault, which bounds the sub-basins on Nias
from the Mola Basement High, is a long-lived
structure; considerable normal throw occurred
across it during the Oligocene and Miocene exten-
sional phase. There is a large difference between
the depth to basement either side of the fault (2 km
on the footwall and an estimated maximum of
5-6 km on the hanging wall). The fault was
mildly reactivated in a contractional sense during
the Pliocene phase of uplift and deformation. This
reactivation is manifested by minor folding of
Middle and Upper Miocene sedimentary rocks
which can be traced across the flexure in
 Lahewa Sub-basin Mujoi Sub-basin Gomo Sub-basin Mola Basement High-----

coast N17 N12 N9 P21-22 N4 P22 , N4 N4.5 N4-5 P22 N4 N7 N l l N17 N22 Coast ;Z

t'rl

Upper Pliocene-Pleistocene ~ MiddleMiocene-LowerPliocene Miocene shallowmarine sediments

Lower-Middle
Oligocene-LowerMiocene deep marine sediments Basement ~ Sediment& melange complex 0 V=H 10 Km,

Fig. 4. Cross-section over Nias Island. The line of cross-section is portrayed on Fig. 2. The section is drawn at true scale from near continuous traverse information and the
position of critical dates is shown using the planktic foraminiferal zonation scheme (Bolli et al. 1985). This cross-section is necessarily interpretative at depth. A full justification
for these interpretations is given in Samuel & Harbury (1996).

4~
 Southwest Coast Shelf edge Northeast
250 240 230 220 250 210 20(3 190 80 170 160 150 40 30 0 ()O 90 80 70 6O 50 4O 30 20 10 20 30 40 50 60 70 80 90 00 10 20 30 40 150 60 70 180 190 200 210 220 230 240 250 260 270 280 290 300 310 Sholpoints

NOISE -~ . _ Pliocene unconformity ~ ------- . txeer ~ ~ _ ~

II ~ ! ?Reef? ~ ~ ~ - ~ - ~ : ~ " ~ ' ~ . ~ - 1 .>
Two- ~ ~ ~ -J~ -- ~ ~ ~ - ~ _ _ _ - - - - - -- Two-
;>
. , ~ " 2tl(t~e
'"21 Prominent reflection reflectlons~ ~ ~ -
.z
.>
3 f-,~ i<~/, ~ Gawo river
;>
7~
4 "~-.....,.)J-'~", 2 "~' <-Y -4
| \, ,-.'~ ,'
25 km "

Fig. 5. Line drawing of two seismic lines crossing easternmost Nias and running offshore. The vertical exaggeration is c. × 2 and the lines are unmigrated. These profiles are tied
by further lines to the S u m a Well (Fig. Ib).
 SUMATRAN FOREARC t~ MENTAWAI FAULT ZONE 345

50 K m

2 Km

: .... : :iiliiiiii:!i~i:i~i:
)iI~ ~i~:~i::~iii!~:~ .... ili!

97°45'E

L o w e r - M i d d l e Miocene
L Upper pliocen ,eistocono shallow marine sediments
E with limestone marker horizon
L •. . . . . .
-. -., -.

G • L o w e r Miocene d e e p marine sediments

E ~ U p p e r Miocene-lowest Pliocene with conglomerate marker horizon
N with limestone a n d tuff marker
horizons Oligocene d e e p marine sediments
D
C o n f o r m a b l e Contact
Major Faults D ~ m ~ m ~ ~ Unconformity

Fig. 6. Traverse map of the Gawo river in eastern Nias. 1:50 000 topographic maps were used as base maps and a
contour interval of 25 m is shown. A 'flexural' steepening is apparent through the Pleistocene and Middle Miocene.
Note also the localized folding of the Upper Miocene. Both these features are related to the reactivation of the main
basin-bounding fault (Fig. 4).

southeastern Nias (Moore & Karig 1980; Samuel
1994). The folding can be seen in a number of
along-strike traverse sections including that of the
Gawo river (Fig. 6).
Whilst it is possible to correlate broadly the
Mentawai fault zone with the region where the
'flexure' and the associated basin-bounding fault
occur on Nias, it is clear that, although of variable
width, the Mentawai fault zone is not confined to
a single fault. The Mentawai fault zone is repre-
sented by a number of faults within a zone of
variable width (4-25 km). Onshore Nias there are
a set of southwest dipping faults, synthetic to the
main basin-bounding fault (Figs 2 and 4). These
faults have a similar history to the main basin-
bounding fault; they were extensional during the
Oligocene to Early Miocene, became passive
features during the Mid-Miocene and Late Miocene
and were variably reactivated during the Pliocene.
Fault zones associated with faults, such as that
separating the Gomo Sub-basin from the Mujoi
Sub-basin, are exposed in some parts of the island,
346 M.A. SAMUEL 8z N. A. HARBURY
such as in the Me river (Fig. 2; Samuel & Harbury
1996). Deformation in these sections is exclusively
contractional with no indication of any strike-slip
movements.
Although the Mentawai fault zone may be
identifiable as a distinct feature for 45 km along the
strike of southeastern Nias, its location to the north
of the Gido river is less clear (Fig. 2). The trace of
the main basin-bounding fault would appear to run
offshore and there is a marked change in the style
of deformation to the north of the Gido river (Fig.
2). The flexural steepening is no longer apparent.
Instead the sedimentary rocks are deformed into
broad, upright and NE verging folds, whose
geometry is apparent from the trace of feature-
forming lithologies (Fig. 2). Discrete areas of
folding are separated by north and ESE striking
faults which controlled the development of these
structures. The orientation of these structures is the
same as that of faults such as the Batee and Singkel
faults that cut across the present-day forearc basin
(Fig. 1B).
Fault patterns and displacements on Nias
Examination of aerial photographs, LANDSAT
and particularly SAR (Synthetic Aperture Radar)
images (Fig. 7) has allowed the identification of
the following four distinct sets of faults on Nias
(Fig. 8).
SE-striking. Four major SE striking faults are
identifiable. They are readily distinguishable from
the stratigraphic ridges which generally run south-
east to northwest (Figs 2 and 6) as they bound
separate areas of Nias with markedly different
morphological expression (Figs 7 and 8). These
geomorphologically distinct areas, which corre-
spond to the sub-basins, have been subjected to
different intensities of deformation (Fig. 4). The
easternmost, SE-striking fault, is the least well
defined as it is only marked by its flexural surface
expression.
North-striking faults are clearly identifiable over
large areas of Nias (Fig. 8). These faults are
observed in many cases to cross-cut and offset the
southeast striking faults with a right-lateral sense
of displacement.
ESE-striking faults display a similar character to
the north-striking faults (Fig. 8). Many of them
clearly offset the SE-striking faults with a left-
lateral sense of displacement. Some also offset the
north-striking faults with a left-lateral sense of
displacement. Conversely some of the ESE-striking
faults are themselves offset, with a right-lateral
sense of displacement, by north-striking faults
(Fig. 8).
ENE-striking. A network of ENE striking faults
is identifiable over all of Nias and is particularly
pronounced in southeastern areas of the island
(Fig. 8).
Although excellent outcrops can be found on
Nias within incised fiver valleys, the faults them-
selves are rarely exposed as the gouges are subject
to rapid weathering and therefore direct kinematic
indicators are generally obscured. The senses of
displacement on these faults can however be deter-
mined from stratigraphic offsets and the relation-
ships between the different faults (Fig. 8). The
SE-striking faults controlled the geometry of the
sub-basins during the Oligocene and Early Miocene
extension/transtension when they developed as
a SW-dipping synthetic set of extensional faults.
The north-striking and ESE-striking faults also
developed during the Oligocene and Early Miocene
as oblique-slip faults. Differential throws on these
faults led to the along-strike changes in sedimen-
tation recorded in the stratigraphic record on Nias
(Samuel et al. 1995; Fig. 3). Further movements
occurred on these faults during the Pliocene phase
of uplift and deformation as they offset some of the
SE-striking faults. The apparent offsets indicate
that the north-striking faults have a right-lateral
sense of movement and the ESE faults have a left-
lateral sense of movement although the amounts of
lateral movement on individual faults are generally
small (less than 5 km).
The combination of the right-lateral movement
on the north-striking faults and the left-lateral
movement on the ESE-striking faults produces
marked extension of the island parallel to the
trench-forearc system. The deformation is one of
pure shear, rather than simple shear. In some parts
of Nias, sections of Oligocene and Lower Miocene
sedimentary rocks have been rotated away from
the strike of the island (Samuel et al. 1995). These
rotations are confined to discrete fault bounded
segments and do not exhibit any consistent rotation
about vertical axes. There is no indication that they
can be associated with a simple shear deformation
although palaeomagnetic studies are planned to
investigate this further.
NW-SE extension across Nias is not only taken
up by the north and ESE-striking faults but also
by the ENE-striking faults which form part of an
extensional fracture set (Fig. 8). Small-scale normal
faults, generally with throws of less than 5 m, are
exceptionally common on Nias and many of them
have led to NW-SE extension.
Discussion and conclusions
Although it is possible to identify a major struc-
tural feature on Nias that is in alignment with the
Mentawai fault zone, there is a complete lack of
 SUMATRAN FOREARC • MENTAWAI FAULT ZONE 347

Fig. 7. Detail of a 1:100 000 SAR (Synthetic Aperture Radar) image of Nias. The image has been used to map the
network of faults over the island (Fig. 8).

evidence for significant strike-slip motion along it.
The linear structure on Nias, previously referred
to as a flexure, is the surface expression of an
extensional fault that has been mildly reactivated in
a contractional sense. Further synthetic Oligocene
to Miocene extensional faults have been subject
to varying degrees of contraction however it is
clear that these faults have not been subject to any
considerable strike-slip displacement.
The structural studies rule out the possibility that
the Mentawai fault zone in the Nias area has a
strike-slip origin. The evidence from Nias alone
cannot however rule out the possibility that sub-
stantial strike-slip movements occur further to the
southeast along the zone. Study of the seismic data
available tends to suggest to us however that this
is unlikely; although the structure of the Mentawai
fault zone is variable it shows two characteristics
that are most atypical of strike-slip faults zones.
Firstly, sections to the northeast and southwest of
348 M.A. SAMUEL t~ N. A. HARBURY

\
\
L

0 10 20 30
i

krn

Southeast striking
faults

" ~ North& east-southeast

striking faults
East-northeast striking iiii!iiii~i~:i~i~i~:
faults
!/~ili~~ilii~¸

Fig. 8. Map of the fault network identified over Nias. The faults were traced directly from the 1:100 000 SAR image
(e.g. Fig. 7). Fault movement since the Pliocene has extended the island parallel to the arc.

the zone are always downthrown relative to the
central regions. One of the characteristics of strike-
slip fault zones is that the magnitude and senses of
displacement change markedly along the length of
the fault system (Underhill et al. 1988). Secondly
the Mentawai fault zone is both atypically narrow
(c. 25 km) and atypically straight for a major strike-
slip system.
Strike-slip faults with small amounts of move-
ment (< 5 km) are identifiable on Nias but they cut
across the strike of the island. Movements on these
structures effectively resulted in extension of the
island in a direction parallel to the arc. Furthermore
extension is accommodated on a well developed
arc-perpendicular network of extensional faults.
The strike-slip faults on Nias can be matched, in
terms of orientation and sense of movement, to
larger-scale strike-slip faults, such as the Batee and
Singkel faults which cut across the present-day
forearc basin (Fig. lb). Clearly the forearc is not
behaving as a rigid plate but is being subjected, at
least in the Nias area, to arc-parallel stretching as
suggested by McCaffrey (1991).
This study provides an indication of how arc-
parallel extension may be accommodated in fore-
arcs. The importance of strike-slip faults such as the
 SUMATRAN FOREARC • MENTAWAI FAULT ZONE 349

Batee and Singkel faults has been highlighted by
Matson & Moore (1992) who have shown that
these faults controlled the development of basins
within the present-day forearc basin (Fig. lb;
Matson & Moore 1992). In addition to motion on
relatively large strike-slip faults our work suggests
that the extension that must occur along the forearc
is taken up over a network of relatively small-scale
strike slip and normal faults distributed across the
forearc. Any vertical throws on these faults may be
so small as to preclude identification on seismic
sections.
It was suggested by Diament et al. (1992) that
the Mentawai fault zone lies on the boundary
between the forearc ridge, which they believed was
part of the accretionary prism, and the forearc basin
which they supposed was located in a continental
domain. The very straightness of the fault zone
makes this unlikely. The studies on Nias have
shown that, although the sub-basins developed on
ophiolitic material, they evolved in an extensional
environment and not as part of an accretionary
prism. Basement to the west of the main basin-
bounding fault is ophiolitic although it is con-
ceivable that the Mola Basement High may be
composed of continental material. On a larger scale
however gravity studies have failed to identify
any obvious oceanic-continental crustal boundary
(Kieckhefer et al. 1981; Milsom 1993). Instead
these studies have shown that the basement beneath
the forearc basin is most likely to be extremely
heterogeneous. Studies of melanges on Siberut,
which lies to the southwest of the Mentawai fault
zone, have identified blocks of continental base-
ment material in the melanges on the island (Andi
Mannga & Burhan 1986). The melanges have
circular geometries and both include and intrude
Neogene material suggesting that they have a
similar origin to the diapiric melanges on Nias.
Basement beneath parts of Siberut may therefore
contain continental material that has been struc-
turally covered, and brought up, in diapiric
melanges.
Although we suggest that the Mentawai fault
zone is not a major strike slip zone defining
a marked rheological boundary, this zone of
deformation is an important feature whose origin
needs to be understood. Onshore seismic lines
across the Mola Basement High can be tied to
offshore lines which run across the present-day
forearc basin. These lines (e.g. Fig. 6) reveal the
presence of extensional/transtensional basins
actually within the present-day forearc basins and
the Mola Basement High can therefore be viewed

Southwest Northeast
Basementfragment Oligoceneto recent
rich in ophiolitic material sediments Continental crust
\
// Sumatran fault zone /
Indian Ocean Trench Accretionarywedge \ r--Ni, ;---~ Forearc basin & magmatic arc

Descending ~ocean;~c~ii~t

Crust of
\ ~Ol1~
unknown nature
Overriding
continental
lithosphere

Fig. 9. Cross-section from the Indian Ocean and trench across Nias to the Sumatran mainland. Nias island lies at the
outer-edge of the Sumatran Forearc and the deformation seen over the island is comparable with that seen on seismic
to the southeast of Nias along the Mentawai fault zone. The Mentawai fault zone represents a line of deformation at
the outer edge of the Sumatran forearc.
350 M.A. SAMUEL & N. A. HARBURY
as a large horst block (Fig. 9). Gravity measure-
ments (Milsom 1996) and seismic reflection studies
(Matson & Moore 1992) reveal that the Mola
Basement High has a north-south trend. Features
such as the Mola Basement High have led to the
variations in structural style along the Mentawai
fault zone.
The structure from Nias into the present-day
forearc basin (Fig. 9) is readily comparable with
that seen on some of the seismic sections crossing
the Mentawai fault zone further to the southeast.
The underlying feature is that central regions of the
fault zone are structurally higher than the margins.
It is striking that Diament et al. (1992) report on the
identification of faulted blocks, horst and graben
systems, and flexures within the Mentawai fault
zone. These are all features that have been identi-
fied on Nias and need not require strike-slip
motion. Structural studies on Nias, and the other
forearc islands, have revealed the importance of
mud diapirism in deformation along the outer-edge
of the forearc (Samuel 1994). In addition diapiric
activity, particularly along faults within the present-
day forearc basin, has been recognized (Rose 1983;
Milsom 1996). Features on several of the seismic
sections crossing the Mentawai fault zone can be
explained in terms of mud diapirism and these
diapirs appear to produce substantial sea floor
topography. In addition structures such as primary-
rim synclines, which are diagnostic of diapirism
(Talbot & Jackson 1987), are readily apparent.
The authors suggest that features visible in seismic
sections across the Mentawai fault zone can be
explained by the inversion of originally exten-
sional structures and by mud diapiric processes.
This is entirely consistent with the geological data
on Nias (Figs 4 & 9). The structure of the Sumatran
outer-arc ridge is wholly comparable with that of
the Mentawai fault zone.
Diament et al. (1992) have suggested that the
Mentawai fault zone may connect with faults such
as the Batee Fault. Given, however, that there has
not been extensive strike-slip movement along
the Mentawai fault zone, where it runs onshore
References
ADAMS, C. G. 1984. Neogene Larger Foraminifera.
Evolutionary and Geological Events in the Context
of Datum Planes. In: IKEBE,N. & TSVCHI,T. (eds)
Pacific Neogene Planes. University of Tokyo Press,
Tokyo, 47-67.
ANDIMANNGA,S. & BURHAN,G. 1986. Laporan Geologi
Lembar Siberut, Sumatra. Sekala 1;250,000.
Geological Research and Development Centre,
Bandung.
BELLIER, O., St~BRIER,M. & PRAMUMUIJOYO,S. 1991. La
grande faille de Sumatra: Gromrtrie, cinrmatique
et quantit6 de drplacement mises en 6vidence par
southeastern Nias, linkage is not required.
Structures such as the Singkel Fault and the Batee
Fault appear to have had marked effects on the
styles of deformation along the outer-arc ridge and
may have led to the indentation into the forearc
of islands such as the Banyak Group (Fig. 1). From
the evidence on Nias it is clear that deformation
along the flexure or Mentawai fault zone initiated
during the Pliocene (Fig. 2; Karig et al. 1980;
Samuel et al. 1995). This deformation may be a
response to an increase in the rate of subduction
(Samuel et al. 1995). The arc-parallel extension on
Nias also dates from this time whilst spreading in
the Andaman Sea initiated at 13 Ma (Curray et al.
1979). This implies that there may have been a time
lag as the new stress regime propagated along the
forearc.
Karig et al. (1980) suggested that the flexure on
Nias, and its associated continuation along the
edge of the forearc, marked the arcward limit of
accretionary prism deformation. We agree with
Karig et al. (1980) that the Mentawai fault zone
represents a deformation front. Our evidence
indicates that the deformation, as seen on Nias, did
not however occur in an accretionary prism and
the Mentawai fault zone should not be viewed
as a backthrust. Rather this subduction-driven
deformation is manifested in the inversion of sedi-
mentary basins which originally formed during the
Oligocene to Early Miocene at the outer edge of
the Sumatran Forearc. The great straightness of
the Mentawai fault zone, and its position parallel to
the trench, suggests that its location is controlled by
parameters such as the angle and rate of subduction
of the Indian Oceanic plate.
We thank B. Situmorang of LEMIGAS for his support
of our field programmes in Indonesia. E Banner,A. Bakri,
L. Hartono, K. Bartram, A. Mitlehner and J. Ling are
thanked for dating so many of our samples and M. Jones,
G. Roberts and J. Milsom are thanked for their helpful
comments. We are grateful to J. Malod and G. Moore
for their constructive reviews. This work was sponsored
by British Petroleum and the University of London SE
Asia Research Group.
l'imagerie satellitaire. Comptes Rendues de
l'Academie des sciences, Serie II, 321, 1219-
1226.
BOLLI, H. M., SAUNDERS, J. B. 8/: PERCH-NIELSON, K.
1985. Plankton Stratigraphy. Cambridge University
Press, Cambridge.
BRADLEY, K. 1973. Geology of S.E. Nias Island, 1971
Geological Survey and relationship of the Nias
Section to offshore seismic data. Union Oil
Company of Indonesia.
BURROUGH,H. C. & POWER, P. E. 1968. Field Survey,
Southern Part of North West Sumatra Contract
 SUMATRAN FOREARC t~¢ MENTAWAI FAULT ZONE
Area. Union Oil Company of Indonesia, Report,
RGE 43.
CtW,RAY, J. R. 1989. The Sunda Arc: A model for oblique
plate convergence. Netherlands Journal of Sea
Research, 24, 131-140.
- - , MOORE, D. G., LAWVER,L. A., EMMEL,E J., RAITT,
R. W., ET AL. 1979. Tectonics of the Andaman Sea
and Burma. In: WATKINS, J., MONTADERT, L. &
WOOD-DICKERSON, P. W. (eds) Geological and geo-
physical investigations of continental margins.
AAPG Memoir, 29, 189-198.
DEMETS, C., GORDON, R. G., ARGUS, D. E & STEIN, S.
1990. Current plate motions. Geophysical Journal
International, 101, 425-478.
DIAMENT, M., HARJONO, H., KARTA, K., DEPLUS, C.,
DAHRIN, E, er AL. 1992. Mentawai fault zone off
Sumatra: A new key to the geodynamics of western
Indonesia. Geology, 20, 259-262.
DOUVILLE, H. 1912. Les foraminif'eres de File de Nias.
Sammlungen Geologisches Reichs-Museums,
Leiden, 8, 253-278.
ELBER, R. 1939. Report on a Regional Geological Survey
ofNias. Permina NW Sumatra Report, 40102.
F~TCH, T. J. 1972. Hate convergence, transcurrent faults
and internal deformation adjacent to southeast Asia
and the Western Pacific. Journal of Geophysical
Research, 77, 4432-4462.
HAQ, B. U., HARDENBOL, J. & VAIL, P. R. 1987.
Chronology of fluctuating sea levels since the
Triassic. Science, 235, 1156-1167.
HARaLrRY, N. A. & KALLAGHER,H. J. 1991. The Sunda
Outer-Arc Ridge, North Sumatra, Indonesia.
Journal of SE Asian Sciences, 6, 463-476.
HARJONO, H., DIAMENT, M., DuBo~s, J. & LARUE, M.
1991. Seismicity of the Sunda Strait: Evidence for
crustal extension and volcanological implications.
Tectonics, 10, 17-30.
HUCHON, P. & LE PICHON, X. 1984. Sunda Strait and
central Sumatra fault. Geology, 12, 668-672.
JARRARD, R. D. 1986. Terrane motion by strike-slip
faulting of fore-arc slivers. Geology, 14, 780-783.
KARIG, D. E., LAWRENCE,M. B., MOORE, G. E & CURRAY,
J. R. 1980. Structural framework of the fore-arc
basin, NW Sumatra. Journal of the Geological
Society, London, 137, 77-91.
, SUPARKA, S., MOORE, G. E & HEHANUSSA,P. E.
1979. Structure and Cenozoic evolution of the
Sunda Forearc in the Central Sumatra region. In:
WATKINS, J. S., MONTADERT, L. & WOOD-
DICKERSON, P. (eds), Geological and geophysical
investigations of continental margins. AAPG
Memoir, 29, 223-237.
KIECKHEFER, R. M., MOORE, G. E, EMMEL, E J. &
SUGIARTA,W. 1981. Crustal structure of the Sunda
forearc region west of central Sumatra from
gravity data. Journal of Geophysical Research, 86,
7003-7012.
MCCAFFREV, R. 1991. Slip vectors and stretching of the
Sumatran forearc. Geology, 19, 881-884.
MATSON, R. G. & MOORE, G. F. 1992. Structural
Influences on Neogene subsidence in the central
Sumatra fore-arc basin. In: WATKINS,J. S., FENGZHI
QIANG & MCMILLEN, K. J. (eds) Geology and
Geophysics of Continental Margins. AAPG
Memoir, 53, 157-181.
351
MILSOM, J. S. 1993. Interpretations of gravity data from
the vicinity of Nias. University of London
Geological Research in SE Asia, Report 119.
1996. Basin Formation in the Nias Area of the
Sumatra Fore Arc, Western Indonesia. Geological
Society of Malaysia, Special Publication, in press.
MINSTER, J. B. & JORDAN, T. 1978. Present-day plate
motions. Journal of Geophysical Research, 83,
5331-5354.
MOORE, G. E & KARIG, D. E. 1980. Structural geology of
Nias Island, Indonesia: implications for subduction
zone tectonics. American Journal of Science, 280,
193-223.
, BILLMAN,H. G., HEHANUSSA,P. E. & KARIG, D. E.
1980a. Sedimentology and paleobathymetry of
Neogene trench-slope deposits, Nias Island,
Indonesia. Journal of Geology, 88, 161-180.
, CURRAY, J. R., MOORE, D. G. & KARIG, D. E.
1980b. Variations in geologic structure along the
Sunda Fore Arc, northeastern Indian Ocean. In:
HAYES, D. E. (ed.) The tectonic and geologic
evolution of southeast Asian seas and islands.
American Geophysical Union Monograph, 23,
145-160.
NEWCOMB, K. R & MCCANN, W. R. 1987. Seismic
history and seismotectonics of the Sunda Arc.
Journal of Geophysical Research, 92, 421-439.
PAGE, B. G. N., BENNETT,J. D., CAMERON,N. R., BRIDGE,
D. MC., JEFFREY,D. H., ETAL. 1979. A review of the
main structural and magmatic features of northern
Sumatra. Journal of the Geological Society, London,
136, 569-579.
PAUL, E 1941. Supplementary (Commitment) Survey,
Nias. Permina NW Sumatra Report, 40103.
PUBELLIER, M., RANGIN, C., CADET, J-P., TJASHURI, I.
BtrrrERLIN, J. & MULLER, C. 1992. L'ile de Nias,
un edifice polyphase sur la bordure interne de la
fosse de la sonde (Archipel de Mentawai,
Indonesie). Comptes Rendues de l'Academie des
sciences, Serie II, 315, 1019-1026.
ROSE, R. 1983. Miocene carbonate rocks of Sibolga
basin, northwest Sumatra. Proceedings Indonesian
Petroleum Association, 12, 107-125.
SAMUEL, M. A. 1994. The structural and stratigraphic
evolution of islands at the active margin of the
Sumatran Forearc, Indonesia. PhD Thesis,
University of London.
-- & HARBURY,N. A. 1996. Basin Development and
Uplift at an Oblique-slip Convergent Margin: Nias
Island, Indonesia. Geological Society of Malaysia,
Special Publication, in press.
--, JONES, M. E. & MATTrmWS, S. J. 1995.
Inversion-controlled uplift of an outer-arc ridge:
Nias Island, offshore Sumatra. In: BUCHANAN,J. G.
& BUCHANAN, P. G. (eds) Basin Inversion.
Geological Society, London, Special Publication,
88, 473-492.
TALBOT, C. J. & JACKSON,M. P. A. 1987. Salt Tectonics.
Scientific American, 256, 70-79.
UNDERHILL, J. R., GAYER, R. A., WOODCOCK, N. H.,
DONNELLY, R., JOLLEY, E. J. & STIMPSON, I. G.
1988. The Dent Fault System, northern England-
reinterpreted as a major oblique-slip fault zone.
Journal of the Geological Society, London, 145,
303-316.
 Tectonic evolution of the Bantimala Complex,
South Sulawesi, Indonesia
KOJI WAKITA 1, J A N S O P A H E L U W A K A N 2, K A Z U H I R O M I Y A Z A K I 1,
I S K A N D A R Z U L K A R N A I N 2 & M U N A S R I 2' 3
1 Geological Survey of Japan, 1-1-3 Higashi, Tsukuba, Ibaraki 305, Japan
2 Research and Development Centre for Geotechnology, Jl. Cisitu, 21/154D,
Bandung, 40135, Indonesia
3 Tsukuba University, 1-1-1 Tennodai, Tsukuba, Ibaraki 305, Japan

Abstract: The Bantimala Complex of South Sulawesi, Indonesia is an assemblage of northeast-

dipping tectonically stacked slices. The slices consist mainly of high pressure metamorphicrocks,
radiolarian chert, breccia, sandstone and shale, and melange. In order to understand the tectonic
evolution of the Bantimala Complex, we have investigated the lithology, age, stratigraphy,
structure and relationships of the components. The K-Ar ages of high P-low T metamorphicrocks
suggest that an oceanic plate subducted beneath the Sundaland continent during the Late Jurassic
or earliest Cretaceous. The subduction ceased during the Albian, and the high pressure schists
were exhumed and eroded at the surface before and during the deposition of middle Cretaceous
radiolarian chert. The exhumation of the schists was related to the collision of microcontinents
derived from Gondwanaland. The Jurassic shallow marine sedimentary rocks in the Bantimala
Complex are possibly remnant fragments of the collided microcontinent. Tectonic stacking of the
Bantimala Complex was caused by Neogene subduction and collision of another continental
fragment further to the east.

Melange complexes are widely distributed in the
Indonesian and east Malaysian regions. Tectonic,
sedimentary and diapiric processes have been
proposed as possible mechanisms for melange
formation in these regions. Asikin (1974),
Hamilton (1979) and Hehuwat (1986) stressed
tectonic processes occurring during interaction of
continental and subducting plates in the origin of
melanges of the Indonesian region. Barber et al.
(1986) proposed a diapiric origin for melange of the
Bobonaro Scaly Clay in Timor, Indonesia. Clennell
(1991) proposed that sedimentary processes such as
submarine slumps and slides were the major mode
of melange formation in east Sabah, Malaysia.
These three processes are not mutually exclusive,
but rather are end-member driving mechanisms
which can operate together or sequentially (Lash
1987; Clennell 1991). It is important to note that
multiple processes occurring at different stages in
their history caused the chaotic mixing of many
melange complexes.
In order to understand the multiple processes
of chaotic disruption in the Cretaceous melange
complexes in Java and South Sulawesi, the authors
have tried to determine in detail the tectonic
history of the melange complexes. Recently,
Wakita et al. (1994a) determined the history of the
accretionary process which formed the Luk-Ulo
Melange Complex of the Karangsambung area,
central Java on the basis of radiolarians extracted
from siliceous and argillaceous rocks. The detailed
dating of sedimentary rocks suggests that sub-
duction occurred between the Early and Late
Cretaceous. Oceanic materials such as chert, lime-
stone and pillow basalt travelled on the oceanic
plate, and were accreted with terrigenous materials
at the 'Karangsambung Trench'. Cretaceous
melange complexes are found in Sulawesi and
south Kalimantan as well as in central Java (Fig. 1).
The Bantimala Complex components are very
similar to those of the Luk-Ulo Melange Complex
in central Java but differences of structure and
lithology suggest a different tectonic evolution in
each area. Described here is the geology of the
melange complex of the Bantimala area along with
discussion on the tectonic evolution of this area
during Mesozoic and Cenozoic time.
Tectonic setting
The southeastern part of pre-Cretaceous continental
basement of Sundaland extends into west
Kalimantan (Hutchison 1989). Cretaceous granites
were intruded into the basement of central and
western Kalimantan, Sumatra and the western part
of the Java Sea (Hamilton 1979). Cretaceous
melange complexes surround the southeastern
margin of Sundaland. As they are widely covered

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, 353

Geological Society Special Publication No. 106, pp. 353-364.
354
by Cenozoic sedimentary and volcanic rocks, the
complexes are exposed only in a few localities
(Fig. 1) such as Bantimala (South Sulawesi),
Karangsambung (central Java) and Meratus (south
Kalimantan). The distribution of Jurassic-
Cretaceous granites and the direction of the con-
tinental growth by accretion suggest the oceanic
plate subducted mainly towards the Sundaland
continent during Cretaceous time. The Cretaceous
complexes of the Meratus and Bantimala areas
formed a single complex prior to Neogene foreland
subsidence of the Makassar Strait (Bergman et al.
1995). Tertiary subduction complexes and
obducted ophiolite are distributed to the east in
central and east Sulawesi (Simandjuntak 1990;
Parkinson 1991; Coffield et al. 1993; Bergman
et al. 1996), and the microcontinent of the
Banggai-Sula has pushed them westward (Fig. 1).
Outline of the geology
The Bantimala area is located c. 40 km northeast
of Ujung Pandang, South Sulawesi (Fig. 2). The
detailed geology of this area has been investigated
by Sukamto (1975, 1978, 1982, 1986) and Hasan
(1990, 1991).
The Bantimala area is underlain by the
Bantimala Complex (Jurassic-Cretaceous),
propylitized volcanic rocks (Palaeocene), the
Malawa Formation (Eocene), the Tonasa Formation
__,, I.
"i!i!!!!!i!i!i!i!i
,~~..:.::!:i~lIMeratus
~~," e~,',',',',',',',',',',',',',',',"
ti al c~ e- o~ u~""s
~ ~ n g s ~
/ ~
]-lds
/ ~po
Fig. 1. Tectonicsetting of the Bantimala Complex (modified from Hamilton 1979)
K. WAKITAET AL.
(Eocene-Middle Miocene), the Camba Formation
(Middle to Late Miocene), and Quaternary sedi-
mentary cover in ascending order (Fig. 3). The
Tonasa Formation rests conformably on the
Malawa Formation. The other Cenozoic formations
lie unconformably on the older formations. The
Bantimala Complex is a tectonic assemblage of
slices and blocks consisting of sandstone, shale,
conglomerate, chert, siliceous shale, basalt, ultra-
mafic rocks, schist and 'schist breccia'. The ages
of components range from Jurassic to middle
Cretaceous. Propylitized volcanic rocks similar to
Lamasi Complex rocks to the north consist of
breccia, lava and tuff mainly of andesitic and
partly of basaltic and trachytic composition. K-Ar
and fission track dating indicate a Palaeocene age
(Sukamto 1982). The Malawa Formation is com-
posed of sandstone, conglomerate, siltstone and
marl with layers or lenses of limestone and coal
or lignite. This formation is Upper Palaeocene to
Lower Eocene on the basis of large foraminifers,
palynology and ostracods (Sukamto 1982; Hasan
1990). The Tonasa Formation consists of limestone
with subsidiary tuffaceous marl, siltstone and sand-
stone and contains foraminifera of Lower Eocene
to Middle Miocene age.
The Camba Formation is composed of marine
volcanic and sedimentary rocks. The sedimentary
rocks are sandstone, siltstone and claystone, inter-
calated with marl, limestone and coal. Volcanic
A) ~ ,~ I . N g~G--
#
nt
,.J .~ ~ %----T,,r:~-rv "
s u b d u~ c t i o n ~ Q .i.:.~:.!..~.:.i.:.i.:..-:.~
Continental
BANDA SEA
c~ ~ fragment
Timor /
~ e~o ~oo 4oo ,oo ~oo ~ooo ,,.
l?s 1~o" -~ ~ - , . . . . ,
 BANTIMALA COMPLEX EVOLUTION, SULAWESI
. . . . . . . . . .
-Pankajene I I
B a n t J m a l a~ ) x
Fig. 2. Locality and distribution map of the Bantimala Complex. The exposures near Barru are equivalent to the
Bantimala Complex (after Sukamto 1975).
rocks of basaltic and andesitic composition include
tuff, breccia, volcanic conglomerate and lava.
Fossils of the Camba Formation indicate it ranges
from Middle to Upper Miocene. K-Ar ages of
Camba Formation volcanic rocks from Buakayu,
western Sulawesi range from 6-14 Ma (Bergman
et al. 1996). Diorite stocks intrude the Camba
Formation as well as the Bantimala Complex and
the Tonasa Formation. Quaternary alluvial deposits
in the western part of the Bantimala area consist
of gravel, sand and clay.
Bantimala Complex
The Bantimala Complex is a tectonic assemblage
of various lithologies ranging in age from Jurassic
to Cretaceous. The complex is unconformably
overlain by, or in fault contact with, Tertiary and
Quaternary formations.
Lithology
The main components of the Bantimala Complex
are sandstone, shale, conglomerate, chert, siliceous
shale, basalt, ultramafic rocks, schist, and 'schist
breccia'. The stratigraphic relationships among the
355
Diorite (Miocene)
Bantimala Complex
~ Balangbaru Formation
D Melange
~.v#~/~l ~ Schist
I~ 5 10 15 20 km
I , i i I
1 1914EE
components are mostly unclear because of their
fault contacts with each other, except for the
following. Schist is unconformably overlain by
'schist breccia' (A. J. Barber pers. comm. 1994).
The 'schist breccia' grades into sandstone that is
overlain by chert. The chert is conformably over-
lain by siliceous shale interbedded with sandstone
and is locally intercalated with conglomerate.
Melange is one of the characteristic units in the
Bantimala Complex, and includes rock types such
as sandstone, shale, siliceous shale, chert, basalt,
schist, and felsic igneous rocks within a sheared
shale matrix. Jurassic sedimentary rocks are known
in two tectonic slices and are quite different from
Cretaceous sandstone, shale and conglomerate of
the complex (see below).
Structure
The tectonic slices of the complex are elongated
and strike NW-SE. The dip of strata ranges from
20-80 ° towards the northeast or east (Figs 3 and 4).
Palaeocene propylitized volcanic rocks uncon-
formably cover the faults between tectonic slices
of the Bantimala Complex, and the boundary
faults between the Balangbaru Formation and the
356 K. WAKITAET AL.

Quaternary

I--"] alluvium
ooooo o o o o c 4 _4 5_ _
, oooooo oooooooooo
~:;.. o o o o , ooooooooo M. Miocene P l i o c e n e
-

i:.... o o o ,~. o oooooooo

Camba Formation
oooo .,?;?;%tsr~rsr = o o o o o o o o
ooo o
ooo oo
.;.*;,%.*;,%t;.%.~;... o o o o o o o o o
,?;.*;?;?;?;?;?;~;.~; ' : o o o o o o o o o
dionte
ooo oo ~,;.';,~;.~;~;%?;?;." .=:~:=. o o o o oo
ooo ooo ;.*;..~;,%.*;,%.*;,*' "::;:~:;.... o o Eocene - E. Miocene
oo ooo , ;?;.*;,*;,~;.';.~;.';?;.';~;?;~;?;." =..=. o o Malawa & Tonasa Formations
~ ~ o "o ~. i o ! oi i o' ?;,%.*;,,~;,*;,.%,%.%,*;t;,,~;.%?;?;t;
i ~ :;:~:;:~,: o
Paleocene - Eocene
Propyfltized volcanic rocks

dacite & diorite

Bantimala Complex
'~ Balangbaru Formation
'=:¢~:=,:¢~:: ". . . . . . :;:.... o ..... ..¢
Melange
I- "":':: o o o, ::,:~'= " =.:~:~:
bedded chert
.... o o o o % 0 0 0 0 o'"o'~:~.... . schist breccia
ooo oooo o ".-.=i
o o o ooo o .;%.*;?;~ .: schist

ultramafic rocks

Jurassic shallow

SW NE' marine formation

lOOO Fault
500
_o2 _Ol~: 0

Fig. 3. Geological map of Bantimala area (after Sukamto 1986).

Bantimala Complex. NW-SE thrust faults cut the (Fig. 6a). Some include fossils such as hexacorals,
Miocene Camba Formation (Fig. 3) and Plio- foraminifers, calcareous algae and sponges. As
Pleistocene wrench faults occur locally. stromatoporids are not present, the limestones are
younger than Jurassic and are considered to be
Cretaceous (K. Mori pers. comm. 1994). Highly
Melange sheared polymictic rocks are distributed along the
The most distinct outcrop of melange occurs near Pateteyang River. Brecciated diorite and schist as
the junction of the rivers Pangkajene, Pateteyang well as clasts of sandstone and chert occur in
and Cempaga (Figs 4 and 5). The melange includes severely sheared shale (Figs 5 and 6b). Diorite
clasts and blocks of chert, sandstone, basalt, lime- forms Tertiary intrusions in this area and occurs as
stone and schist embedded within a sheared shale clasts. This mixture is not typical of melange in
matrix. Major clasts are sandstone, chert and the Bantimala Complex. Berry & Grady (1987)
siliceous shale. Basalt and limestone are locally attributed the chaotic features to Plio-Pleistocene
dominant. Fragments of metamorphic rocks are faulting.
very rare. The shale matrix is usually sheared to
some degree, and is severely sheared along the Ultramafic rocks
western part of the Pateteyang River (Fig. 5). Ultramafic rocks of the Bantimala Complex are
The clasts are subrounded to subangular, and mainly distributed along the western margin of the
rhomboidal, spherical, blocky and irregular in Balangbaru Formation (Fig. 3). Very small outcrops
shape. Clast sizes range from several millimetres of serpentinite also occur along some of the
to several hundred metres. Along the Pateteyang southwestern boundaries of the metamorphic
River, clasts of siliceous shale and sandstone are tectonic slices. The ultramafic rocks are mostly
elongated in the highly sheared matrix. Sandstone serpentinized peridotite, with local chromite lenses.
clasts are usually less than 1 m in length, but some- They are unconformably overlain by sandstone
times reach several metres. These are usually of the Balangbaru Formation (Hasan, 1990).
poorly sorted and fine to coarse grained arkoses,
containing mineral (quartz, feldspars, micas, etc.)
and lithic fragments (metamorphic rocks, felsic Schist
tufts, etc.). Limestone clasts are locally dominant in The metamorphic grade of schists in the Bantimala
the melange near the mouth of the Cempaga River Complex ranges mainly from greenschist to
 I I
119 ° 40' .~ 119"41'

>
7~
,.-]

>
>
©
c--'--~~a.lone2
~P.la
~,~/erg
ka.lolae Siliceous shal "~\o/~,.e
v' °"'
° ' sc"~'
"~ ~l,~ .~.".bm
4*46'30" fi
t /
)) mX
x ~~ MeJange matrix
Melange m atrax sscc ~ Schist ~ sc
"~.,)~2c'~'~-~_ <
©
t"

0 , 100 , 200m ~ ch ©
r~ 7
ss~ Sandstone bm ~ Breccia ;,o
wilh metamorphic
bs~ Basalt rock fragments

~Tonasa Formation -,-~--. Thrust ---F Fault

To (Teariary
Limestone)
, ss mx
To
L
Fig. 4. Route map along the Pangkajene River, east of the Mangilu village. The box shows the area of Fig. 5.

t~
t.~
"-,3
358 K. WAKITA E T A L .

i
I 1
119° 40' 55" mx 119 ° 41' 00" 119" 41' 05"
SC

t~t

mx Melange matrix
( S h e a r e d s h a l e w i t h clasts)

Pangkajene ch Radiolarian chert

~X
River
ss Sandstone
'SO
6"- br Breccia,
b
SS ch w i t h basalt, l i m e s t o n e
and c h e r t f r a g m e n t s
ch ~ bs Basalt 4"46'15"-
sc ~ .... '~', ', bs sc Schist
x
bm Breccia,
bs with metamorphic fragments
1.,(t~ u Ultramafic rock

Fault
i"t
/
o Riverside
,brn
":* @"2
bn
N
SC'~

/T/X t
ch

lt~,
4°46'30"

o 5ore : • •

Fig. 5. Route map along the Cempaga and Pateteyang Rivers. This is the same location as that shown in Berry &
Grady (1987, fig. 5). Most of the slices and blocks are in tectonic contact with each other. Only major faults are
shown on the map. Breccia (bm) includes diorite clasts as well as metamorphic fragments within sheared red shale.
Eclogite and glaucophane schist occur as boulders at the base of schist in the northern part of the map.
 BANTIMALA COMPLEX EVOLUTION, SULAWESI
Table 1. K-At ages of muscovites from metamorphic rocks of the Bantimala Complex.
Sample no. Rock type Mineral
BT-11b eclogite muscovite
Mg-47 eclogite muscovite
L-01B eclogite muscovite
BT-08e pelitic schist muscovite
BT-17 pelitic schist muscovite
amphibolite facies. In a tectonic slice, the meta-
morphic grade increases toward the northeast
(tectonically towards the top). Glaucophane schist
and eclogite associated with serpentinite are locally
present as tectonic blocks and slices and occur
along the boundary fault of one tectonic slice
consisting of metamorphic rocks. Most of the meta-
morphic rocks of greenschist to amphibolite facies
were originally eclogite and have suffered retro-
gressive metamorphism. Wakita et al. (1994b)
reported K-Ar ages of micas from schists ranging
from 113-132 Ma (Table 1). The ages of eclogite
vary widely, although the ages of greenschists are
concentrated c. 114-115 Ma.
Schist breccia
'Schist breccia' is one of the main lithologies of the
Bantimala Complex. Most are sedimentary breccias
consisting mainly of schist fragments, although it
is sometimes difficult to distinguish tectonically
brecciated schist from sedimentary breccia. The
metamorphic grades of the schist fragments are the
same as that of schists in the tectonic slices of the
Bantimala Complex. The relationship between
chert and 'schist breccia' of the Bantimala
Complex has been described as an 'unusual uncon-
formity' in the Paring River (Haile et al. 1979;
Wakita et al. 1994b). The 'brecciated schist' grades
into sandstone, which is overlain by radiolarian
chert. As the chert is intercalated with thick beds
of 'schist breccia' at several horizons, as well as
with sandstone layers of various thicknesses, the
three rock types, i.e. chert, sandstone and 'schist
breccia' are considered to be contemporaneous
deposits of an unstable sedimentary basin.
Chert and siliceous shale
Chert layers range from 1-20 cm thick and are
interbedded with thinner shale layers less than 1 cm
thick. The bedded chert is mostly red or reddish
brown, and sometimes pale green or grey in colour.
It is composed mainly of skeletons and fragments
of radiolarians, and a small amount of shale. The
chert sometimes includes well preserved middle
359
Age (Ma) 4°mr %40Ar %K
132 + 7 3.21 93.9 5.99
124 _ 6 2.94 90.4 5.94
113 _ 6 3.76 96.6 8.34
114 _+6 2.5 92.2 5.48
115 _ 6 2.37 94.5 4.98
Cretaceous radiolarians (upper Albian to lower
Cenomanian) including Holocryptocanium barbui,
Thanarla conica, Archaeodictypomitra vulgaris
and Rhopalosyringium majuroensis. The chert is
underlain by 'schist breccia' and coarse grained
sandstone. It is intercalated with sandstone beds
and laminae in the lower part of the succession. The
chert is intercalated with rhyolite tuff layers along
the Pateteyang River (Fig. 6e). The rhyolite is pale
green in colour, is usually fine but sometimes
coarse grained. The chert locally grades into
siliceous shale towards the stratigraphic top in
some localities. The shale is grey or reddish-brown
in colour, and composed of radiolarian skeletons,
terrigenous fragments and other detrital materials.
One of the authors (Munasri) extracted radiolarians
including Holocryptocanium barbui, Pseudo-
dictyomitra pseudomacrocephala and Thanarla
veneta from a siliceous shale collected in the
Pangkajene River.
Jurassic Paremba Sandstone
Jurassic shallow marine sedimentary rocks, called
the Paremba Sandstone (Sukamto & Westermann
1992), are incorporated as tectonic slices in the
Bantimala Complex. The lower part of the Paremba
Sandstone along the Bontolio River is composed
of thin bedded sandstone and shale, intercalated
with thin limestone layers. Some shallow marine
sedimentary structures such as ripple and convolute
laminations are recognized (Fig. 6c). The upper
part of the formation is rich in conglomerate
(Fig. 6d) which includes pebbles mainly of basalt
and schist. Ammonites (e.g. middle Liassic
Fuciniceras), gastropods and brachiopods of the
Lower and Middle Jurassic are reported from the
Paremba Sandstone (Sukamto & Westermann 1992).
Cretaceous Balangbaru Formation
Cretaceous flysch sequences assigned to the
Balangbaru Formation are widely distributed in the
northeast and north in the Bantimala area. Detailed
descriptions were given by Hasan (1990, 1991) and
the following is based on Hasan (1990).
360 K. WAKITA ET AL.

(a) ~)

(c)- (a)

(e) (13

Fig. 6. Outcrop photographs of the Bantimala Complex. (a) Cretaceous limestone blocks embedded within a sheared
shale matrix along the Cempaga River. (b) Highly sheared shale matrix including various kinds of rocks as clasts,
along the Pateteyang River. Plio-Pleistocene faults cut the shale matrix of the Cretaceous melange. (c) Jurassic
Paremba Sandstone, showing ripple marks, interbedded with shale, along the Bontorio River. (d) Jurassic Paremba
Sandstone with intercalations of conglomerate, along the Bontorio River. (e) Rhyolite layers intercalated in
radiolarian chert, along the Pateteyang River, near the mouth of the Sanggi River. (f) Turbidite of the Balangbaru
Formation, in the upstream part of the Balangbaru River.

The formation is subdivided into three members,
in ascending order: the Allup, Panggalungan and
Bua Members. The Allup Member is composed
of pebbly sandstone, conglomeratic breccia and
interbedded sandstone with shale. The Pangga-
lungan Member consists of interbedded sandstone
and shale and chaotic breccia deposits. The Bua
Member is composed of interbedded sandstone,
shale (Fig. 6f) and conglomerate. Hasan (1990)
interpreted the facies associations of the Allup,
Panggalungan and Bua members as inner fan, outer
fan to basin plain and middle fan respectively. The
 BANTIMALA COMPLEX EVOLUTION, SULAWESI
mineral composition of the sandstones varies from
the bottom to top of the formation. Garnet, spinel,
glaucophane and chloritoid are characteristic of the
lower parts of the Balangbaru Formation, while
zircon, apatite and tourmaline increase in pro-
portion towards the stratigraphic top. The mineral
assemblages suggest that the lower sediments were
derived principally from metamorphic rocks of the
Bantimala Complex, and the upper sediments have
a continental or magmatic arc provenance. Hasan
(1990) reported planktonic foraminifera, such as
Globotruncana helvetica, G. arca, G. foricata and
Heterohelix globulosa, from the fine grained sand-
stone and siltstone of the Panggalungan and Bua
members of the Balangbaru Formation indicating
a lower Turonian to upper Maastrichtian range.
One of the authors (Wakita) extracted radio-
larians from shale and siliceous shale of all
members of the Balangbaru Formation. Archaeo-
dictyomitra spp., Pseudodictyomitra spp., Rhopalo-
syringium majuroensis, Thanarla spp., Praecono-
caryomma sp., Stichomitra sp. Archaeodic~omitra
and Pseudodictyomitra are most dominant genera
among them. The assemblage remains very similar
from the stratigraphic bottom to top of the
formation, and is also very similar to that from
chert of the Bantimala Complex reported by Wakita
et al. (1994b). Rhopalosyringium majuroensis
ranges from upper Albian to upper Turonian
(Schaaf 1984) and the common occurrence of
Pseudodictyomitra spp. indicates that assemblage
ranges from Lower to middle Cretaceous. No
species indicating an age younger than Coniacian
has been found in samples of the Balangbaru
Formation. The radiolaria indicate the age of the
Balangbaru Formation is between upper Albian
and Turonian.
Sukamto (1975, 1978, 1982, 1986) and Hasan
(1990) excluded the Balangbaru Formation from
the Bantimala Complex. However, the new radio-
larian age data suggest that the age of the
Balangbaru Formation is not different from the age
of chert in the B antimala Complex. The Balangbaru
Formation also occurs as tectonic slices stacked
with the other members of the Bantimala Complex.
The authors therefore propose, on the basis of new
radiolarian data, that the Balangbaru Formation is
part of the Bantimala Complex. The formation is
almost contemporaneous with sandstones deposited
on chert of the Bantimala Complex, and on the
ultramafic rocks.
Tectonic evolution
The following lithologies and structures of the
Bantimala Complex are critical elements in the
tectonic evolution of the Bantimala area: high
pressure metamorphic rocks, Jurassic shallow
361
marine sedimentary rocks, schist breccia overlain
by radiolarian chert, melanges including blocks of
various kinds of rocks, and tectonically stacked
slices of various rocks unconformably overlain
by Palaeogene volcanic rocks. To explain the
l i t h o l o g i e s and structures of the Bantimala
Complex, the authors propose the following hypo-
thetical tectonic evolution: Cretaceous subduction,
collision and accretion of a microcontinent, and
Neogene tectonic stacking of slices caused by
westward collision of another microcontinent.
Figure 7 shows a possible tectonic evolution of the
Bantimala area based on this hypothesis.
K-Ar ages of high pressure metamorphic rocks
range from 132-113 Ma, and indicate that there
was subduction from Jurassic to early Cretaceous
towards the 'West Kalimantan Continent' as indi-
cated by the distribution of Jurassic to Cretaceous
granites. Jurassic shallow marine sedimentary
rocks are the most important piece of evidence for
a microcontinent that subducted, collided and
accreted in the early Cretaceous 'Bantimala
trench'. Shallow marine clastic formations con-
temporaneous with the Paremba Sandstone of the
Bantimala area are found further to the east in
central and SE Sulawesi, and in the Banggai-Sula
area (e.g. the Buya Formation; Surono & Sukarna
1993). On the other hand, there is no indication of
unmetamorphosed Jurassic rocks to the west of the
Bantimala area.
Various sizes of continental fragments drifted
northward and accreted along the Asian continental
margin since the break-up of the Gondwanaland
(Nur & Ben-Avraham 1983; Maruyama et al.
1989). The microcontinent on which the Jurassic
Paremba Sandstone was deposited is thought to be
one of them. Subduction of the oceanic plate caused
the formation of high pressure metamorphic rocks
at least from Late Jurassic to Early Cretaceous
times, and brought the microcontinent with its
overlying Jurassic shallow marine sedimentary
rocks into the 'Bantimala trench'. After the arrival
at the trench, the microcontinent was subducted,
collided and accreted within the accretionary
wedge. After the collision and accretion of the
microcontinental block, subduction ceased at the
'Bantimala trench'. Underthrusting of the light and
buoyant continental fragment caused the rapid
uplift and exhumation of high pressure meta-
morphic rocks. Their 113-132Ma K-Ar ages
indicate the time of cooling during exhumation.
After metamorphic rocks of the Bantimala
Complex appeared at the surface, they were eroded
and provided 'schist breccia' and sandstone to a
sedimentary basin in which radiolarian remains
were deposited at a relatively high rate during the
late Albian to early Cenomanian (Wakita et al.
1994b).
362 K. WAKITA ET AL.
Jurassic
Shallow marine
West Kalimant an
sediment at ion
Cont inent
High P/T
met amorphic rocks
C e n o m a n i a n - early T u r o n i a n
olist ost rome BalanclbaruFormation
chert
Fig. 7. Tectonicevolution of the Bantimala area, South Sulawesi.
The radiolarian biostratigraphy and lithostrati-
graphic relationships reveal that the ages of 'schist
breccia', chert, and the Balangbaru Formation are
similar. The occurrences of slump deposits and
'schist breccia' layers within chert suggested that
they were deposited in an unstable sedimentary
basin. Successive exhumation of metamorphic
rocks caused the uplift of the basement of the basin.
The basement provided fragments of the other
members of the Bantimala Complex into melanges
which were originally formed as olistostromal
deposits.
Ultramafic rocks of the Bantimala Complex
were exhumed at almost the same time as the
high pressure schists. The Balangbaru Formation
was deposited unconformably on the ultramafic
rocks. After deposition, the tectonic slices of the
Balangbaru Formation and its ultramafic basement
were tectonically juxtaposed with other slices of the
Bantimala Complex.
The most distinct feature of the Bantimala
Complex is tectonic stacking of slices. The struc-
ture is very similar to that of accretionary prisms
and accretionary complexes elsewhere in the
world. The stacking of structures dipping to the east
are, however, opposite to that expected from a
westward-dipping, oceanic plate subduction toward
the Sundaland continent during Cretaceous time.
Sukamto (1982) suggested that tectonic stacking
and mixing in the Bantimala Complex were formed
prior to Late Cretaceous, because the Balangbaru
late Albian
schist breccia radiolarian chert
Exhumat ion
Collision &
accret ion
O l i g o c e n e - Pliocene
sla
Formation was considered to overlie the Bantimala
Complex unconformably. However, as the
Balangbaru Formation is found as tectonic slices in
the Bantimala Complex, the stacking must have
happened after its deposition.
Although the geological map of Sukamto (1986)
indicates that Palaeogene propylitized volcanic
rocks unconformably cover the tectonic slices and
their boundaries (Fig. 3), we doubt that the stacking
of slices occurred in before the Palaeogene. We
need to check the nature of boundaries and
relationship between members of the Bantimala
Complex and Palaeogene propylitized volcanic
rocks. Thrust faults locally cut the Miocene
Camba Formation. Some of the faults obviously
moved later than Miocene time. Coffield et al.
(1993) and Bergman et al. (1996) argued that
stacking of the Lamasi Complex of South
Sulawesi was caused by westward obduction
of the ophiolite in Oligocene time and Miocene
to Pliocene collision of microcontinent. A
similar scenario is acceptable for the tectonic
stacking of the Bantimala Complex. Obduction of
an oceanic plate and subsequent collision of a
microcontinent caused westward-thrusting during
Oligocene to Pliocene time (Parkinson 1991;
Coffield et al. 1993; Bergman et al. 1996).
Tectonic stacking of the Bantimala Complex is
similarly interpreted to result from ophiolite
obduction followed by microcontinent collision in
Tertiary time.
 BANTIMALA COMPLEX EVOLUTION, SULAWESI
Conclusions
The Bantimala C o m p l e x consists of tectonic slices
of sandstone, shale, conglomerate, chert, siliceous
shale, basalt, ultramafic rocks, schist, and 'schist
breccia' The rock components of the complex
range in age from Jurassic to middle Cretaceous.
The c o m p l e x was u n c o n f o r m a b l y overlain by
Cenozoic sedimentary and volcanic formations
and was intruded by Tertiary igneous rocks.
Schists in the Bantimala C o m p l e x range mainly
from greenschist to amphibolite facies and locally
include glaucophane schist and eclogite. The K-Ar
ages of the schists range from 113-132 Ma.
Schist was unconformably overlain by radio-
larian chert and clastic rocks during the late Albian
to early Cenomanian.
Radiolarian biostratigraphy reveals that the
Balangbaru Formation ranges from upper Albian to
Turonian and belongs to the Bantimala Complex.
Jurassic shallow marine sedimentary rocks are
the most important pieces of evidence for a lost
m i c r o c o n t i n e n t that subducted, collided and
accreted in the early Cretaceous ' B a n t i m a l a
trench'.
References
ASlKIN, S. 1974. The geological evolution of central Java
and vicinity in the light of the new global tectonics.
PhD Thesis, Bandung Institute of Technology
[in Indonesian with English abstract].
BARBER, A. J., TJOKROSAPOETRO,S. & CHARLTON,T. R.
1986 Mud volcanoes, shale diapirs, wrench faults,
and melanges in accretionary complexes, eastern
Indonesia. AAPG Bulletin, 70, 1729-1741.
BERGMAN, S. C., COFFIELD, D. Q., TALBOT, J. P. &
GARRARD, R. J. 1995. Tertiary tectonic and
magnetic evolution of western Sulawesi and the
Makassar Strait, Indonesia: evidence for a Miocene
continent-continent collision. This volume.
BERRY,R. F. & GRADY,A. E. 1987. Mesoscopic structures
produced by Plio-Pleistocene wrench faulting in
South Sulawesi, Indonesia. Journal of Structural
Geology, 9, 563-571.
CLENNELL,B. 1991. The origin and tectonic significance
of melanges in Eastern Sabah, Malaysia. Journal of
Southeast Asian Earth Sciences, 6, 407-429.
COFFIELD, D. Q., BERGMAN, S. C., GARRARD, R. A.,
GURITNO, N., ROBINSON, N. M. & TALBOT, J. P.
1993. Tectonic and stratigraphic evolution of the
Kalosi PSC area and associated development of a
Tertiary petroleum system, south Sulawesi,
Indonesia. In: Proceedings Indonesian Petroleum
Association 22nd Annual Convention. 678-706.
HAILE, N. S., BARBER, A. J. & CARTER, D. J. 1979.
Mesozoic cherts on crystalline schist in Sulawesi
and Timor. Journal of the Geological Society,
London, 136, 65-70.
HAMILTON,W. 1979. Tectonics of the Indonesian region.
US Geological Survey, Professional Paper, 1078.
363
Underthrusting o f a light and buoyant con-
tinental f r a g m e n t c a u s e d the rapid uplift and
exhumation of high pressure metamorphic rocks
including eclogite.
Tectonic stacking of slices in the Bantimala
Complex was mainly caused by Oligocene ophio-
lite obduction and M i o c e n e - P l i o c e n e collision of
a microcontinent.
This paper is one of the results of the joint project between
the Research and Development Centre for Geotechnology
(RDCG) and the Geological Survey of Japan (GSJ) under
the ITIT programme 'Research on Mineral Resources
Assessment of Oceanic Plate Fragments'. The authors
thank Drs A. J. Barber and C. D. Parkinson of Royal
Holloway, University of London for their critical review,
effective suggestions and discussion of the geology of this
area. We also express thanks to Dr Ir. S. Suparka, director
of RDCG for his helpful support during our geological
survey. We are grateful to Dr R. Sukamto of the
Geological Research and Development Centre for his
kind offer of unpublished data and information on the
field area. Thanks are also extended to Drs K. Mori,
T. Nakamori and Y. Iryu of Tohoku University and
Dr Y. Sato of the Geological Survey of Japan for the
identification of fossils in limestone clasts of the melange
of the Bantimala Complex.
HASAN, K. 1990. The Upper Cretaceous Flysch
Succession of the Balangbaru Formation,
Southwest Sulawesi, Indonesia. PhD Thesis,
University of London.
-- 1991. The Upper Cretaceous Flysch succession of
the Balangbaru Formation, Southwest Sulawesi. In:
Proceedings Indonesian Petroleum Association
20th Annual Convention. 183-208.
HEHUWAT, E H. A. 1986. An overview of some
Indonesian melange complexes - a contribution
to the geology of melange. Memoir of Geological
Society of China, 7, 283-300.
HUTCHISON, C. S. 1989. Geological Evolution of South-
East Asia. Oxford University Press, Oxford
Monographs on Geology and Geophysics, 13.
LASH, G. G. 1987. Diverse melanges of an ancient
subduction complex. Geology, 15, 652-655.
MARUYAMA,S., LIU, J. G. & SENO, T. 1989 Mesozoic and
Cenozoic Evolution of Asia. In: BEN-AVRAHAM,Z.
(ed.) The Evolution of the Pacific Ocean Margins.
Clarendon Press, Oxford, Oxford Monographs on
Geology and Geophysics, 75-99.
NUR, A. & BEN-AVRAHAM,Z. 1983. Break-up and accre-
tion tectonics. In: HASHIMOTO,M. & UYEDA, S.
(eds) Accretion Tectonics in the Circum-Pacific
Regions. Terrapub, Tokyo, 3-18.
PARKINSON, C. D. 1991. The Petrology, Structure and
Geologic History of the Metamorphic Rocks of
Central Sulawesi, Indonesia. PhD Thesis,
University of London.
SCHAAF, A. 1984. Les radiolaires du Cr~tacd Infdrieur
et Moyen: Biologie et Systematique. Sciences
G6ologiques Memoire, 75.
364 K. WAKITA ET AL.
SIMANDJUNTAK,T. O. 1990. Sedimentology and Tectonics
of the Collision Complex in the East Arm of
Sulawesi, Indonesia. Geology of Indonesia, 13,
1-35.
SUKAMTO, R. 1975. Geologic Map of Indonesia, Ujung
Pandang Sheet. 1:1,000,000. Geological Survey of
Indonesia.
- - 1978. The structure of Sulawesi in the light of plate
tectonics. In: Proceedings of Regional Conference
on Geology and Mineral Resources of SE Asia.
121-141.
-- 1982. The geology of the Pangkajene and Western
part of Watampone, South Sulawesi, scale
1:250,000. Geological Research and Development
Centre, Bandung.
1986. Tectonik Sulawesi Selatan dengan acuan
khusus ciri-ciri himpunan batuan daerah
Bantimala. Dissertation, ITB, Bandung.
8Z WESTERMANN, G. E. G. 1992. Indonesia and
Papua New Guinea. In: WESTERMANN,G. E. G. (ed.)
The Jurassic of the Circum-Pacific. Cambridge
University Press, USA, 181-193.
SURONO & SUKARNA,D. 1993. Geology of the Sanana
Sheet, Maluku, 1:250,000. Geological Research and
Development Centre, Bandung.
WAKITA, K., MUNASRI ,~ BAMBANG, W. 1994a.
Cretaceous radiolarians from the Luk-Ulo Melange
Complex in the Karangsambung area, central Java,
Indonesia. Journal of SE Asian Earth Sciences, 9,
29-43.
--, MUNASRI,SOPAHELUWAKAN,J., ZULKAm~AaN,I. &
MIYAZAKI, K. 1994b. Early Cretaceous tectonic
events implied in the time-lag between the age of
radiolarian chert and its metamorphic basement in
Bantimala area, South Sulawesi, Indonesia. Island
Arc, 3, 90-102.
 The Tertiary evolution of South Sulawesi: a record in redeposited
carbonates of the Tonasa Limestone Formation

M O Y R A E. J. W I L S O N & D A N W. J. B O S E N C E
SE Asia Research Group, Department of Geology, Royal Holloway University of London,
Egham TW20 OEX, UK

Abstract: South Sulawesi, situated at the junction of three major plates and with an almost
complete Tertiary sequence, is an ideal location in which to study syntectonic sedimentation.
Redeposited carbonate facies of the lower/middle Eocene to middle Miocene Tonasa Limestone
Formation in the Barru area prove to be reliable indicators of tectonic activity. South of the Barru
area contemporaneouscarbonate sediments formed on a relatively stable shallow-waterplatform,
known as the Tonasa Carbonate Platform. Redeposited carbonate facies and interbedded marls
from the Barru area are described and interpreted in this study. The immaturity and provenance
of clasts indicate that the redeposited facies were derived from the faulted northern margin of the
Tonasa Carbonate Platform. A relay ramp between at least two major NW-SE trending faults
is the inferred configuration of this margin. Three main phases of faulting are indicated by the
redeposited facies: late Eocene to early Oligocene, middle Oligocene and early to middle
Miocene. This is consistent with other outcrop and seismic data from the region and with the
inferred plate tectonic situation during the Tertiary.

Sulawesi is located in an exceedingly complex
tectonic region, where three major plates have been
interacting since the Mesozoic. With reference to
the hotspot frame the Pacific-Philippine plate is
moving WNW, the Indo-Australian plate NNE and
both are colliding with the relatively stable
Eurasian plate (Hamilton 1979; Daly et al. 1987,
1991). The convergence zone of this triple junction
is a composite domain of micro-continental frag-
ments, accretionary complexes, m61ange terrains,
island arcs and ophiolites. Successive accretion
from the east of oceanic and microcontinental
material, and the associated development of island
arcs, have all controlled the stratigraphic develop-
ment of Sulawesi.
South Sulawesi (Fig. 1), located on the eastern
margin of Eurasia, has an almost complete strati-
graphic sequence representing the period between
the late Cretaceous and the present day (Fig. 2;
Sukamto 1975; Hamilton 1979; Van Leeuwen
1981). South Sulawesi is therefore an ideal location
in which to study the effects of local or regional
tectonics preserved within a sedimentary sequence.
Carbonate deposits of the middle Eocene to middle
Miocene Tonasa Limestone Formation comprise a
major part of the Tertiary succession in the western
part of South Sulawesi (Fig. 2). The aim of this
paper is to document platform, slope and deep
water carbonate lithologies in the vicinity of
Barru (Fig. 1), which preserve evidence of local,
contemporaneous tectonic activity.
Sulawesi is formed of distinct north-south
trending tectonic provinces (Sukamto 1975). In the
west, the north and south arms of Sulawesi are
composed of thick Tertiary sedimentary and
volcanic sequences overlying pre-Tertiary base-
ment complexes (Sukamto 1975; Van Leeuwen
1981). The Barru area lies within this western arc or
province (Sukamto 1975; Hamilton 1979). Central
Sulawesi is composed of sheared metamorphic
lithologies and in the east a highly tectonized
melange complex is present (Sukamto 1975;
Hamilton 1979; Parkinson 1991). The eastern
periphery of this melange has been overthrust by
a dismembered and imbricated ophiolite sequence
(Sukamto 1975; Silver et al. 1978; Simandjuntak
1990; Parkinson 1991). Emplacement of this
ophiolite and resulting formation of the melange
occurred during the middle Oligocene (Parkinson
1991). The microcontinental fragments of the
Buton-Tukang Besi Block and Banggai-Sula are
thought to have collided with the eastern part of
Sulawesi during the early-middle Miocene
(Fortuin et al. 1990; Davidson 1991; Smith &
Silver 1991) and late Miocene-early Pliocene
respectively (Garrard et al. 1988; Smith & Silver
1991). The Tertiary stratigraphy of western
Sulawesi is comparable with many of the Tertiary
basins in neighbouring east Kalimantan. West
Sulawesi, the East Java Sea and east Kalimantan
are thought to have comprised a widespread basinal
area, the formation of which commenced during the
early-middle Eocene (Van de Weerd & Armin
1992).

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, 365
Geological Society Special Publication No. 106, pp. 365-389.
366 M.E.J. WILSON 8~ D. W. J. BOSENCE

Geology and stratigraphy of Two inliers of the pre-upper Cretaceous base-

South Sulawesi ment complex of Sulawesi are exposed west of the
Walanae Depression (Fig. 1). They comprise
South Sulawesi is structurally separated from the tectonic slices of metamorphic, ultrabasic and sedi-
rest of the western arc of Sulawesi by a NW-SE mentary lithologies (Hamilton 1979; Sukamto
trending depression which passes through the 1982). Deep marine clastics of the upper
Sengkang Basin (Fig. 1; Van Leeuwen 1981). Cretaceous Balangbaru and laterally equivalent
Geologically and geomorphologically South Marada Formation overlie the basement complexes
Sulawesi is divided by a present day N-S trending unconformably (Fig. 2; Van Leeuwen 1981;
depression known as the Walanae Depression (Fig. Sukamto 1982; Hasan 1991). Palaeocene-Eocene
1). The Walanae Depression has been described as volcanics of the Langi Formation and marginal
a major left-lateral strike-slip zone (Sukamto 1975; marine siliciclastics, shales and coals of the Eocene
Van Leeuwen 1981). Seismic (Grainge & Davies Malawa Formation overlie with an angular un-
1983) and present-day outcrop constraints suggest conformity the Balangbaru Formation in the eastern
significant normal displacement on basin-bounding and western parts of west South Sulawesi respec-
faults occurred during the Tertiary. The Barru area tively (Sukamto 1982). The upper part of the
of this study includes the northernmost outcrops Malawa Formation interdigitates with shallow
of the Tonasa Limestone Formation and is located marine carbonates of the middle Eocene to middle
on the southern margin of the depression passing Miocene Tonasa Limestone Formation to form a
through the Sengkang Basin, to the west of the transgressive sequence.
Walanae Depression (Fig. 1). Deposition of at least 400 m of shallow marine
carbonates occurred in the area between Maros
and Tonasa II (Fig. 1; Garrard et al. 1989; Crotty &
Engelhardt 1993). These sediments of the Tonasa
Limestone Formation are thought to have formed
on a relatively stable, gently subsiding, large-scale
" 7 ::!i:l - o platform c. 80 km across (personal observation;
Garrard et al. 1989), named here as the Tonasa
( /
Carbonate Platform. An intra-formational mid-
Oligocene unconformity occurs within shallow-
water carbonates of the Tonasa Limestone
Formation in the eastern Biru area (Figs 1 and 2,
.... ., .-.-. .
Van Leeuwen 1981). From the early Miocene
- ::::: :~:~::. :: :: :::. onwards in the Biru area there was a deepening
of the environment, and carbonate sedimentation
t T°nasallctdil ':::~:5,-~::::~:IV~iiiiill~ i ~ .7 :I was strongly influenced by local tectonism (Van
\ "1 l-".11r:: :~:3:.~:~ iiii{iiiiiiiiii7 - Leeuwen 1981). This study describes and interprets
~~ i I:: i::: ~ '".::.~iSl!]!
" '~i!!!i!!!i!i~ deep marine late Eocene to mid-Miocene carbon-
ates of the Tonasa Limestone Formation, from the
Barru area to the north of the Tonasa Carbonate
Platform. During the middle to late Miocene
UpJUI~G
carbonate production was terminated by the
influx of volcaniclastic deposits of the Camba
Formation. These volcaniclastics were derived
from a N-S trending volcanic arc which developed
in South Sulawesi (Sukamto 1982; Yuwono et al.
1985).
East of the Walanae Depression, lithologies are
quite distinct from those to the west and the oldest
l I lithologies are of Eocene age (Figs 1 and 2;
KEY I TONASA LIMESTONE Fm. Sukamto 1975). The lithologies are dominated by
Alluvium ~ KalamisengFm.
LompobatangVolcanics ~ SaloKalupangFm. volcanics and volcaniclastics of the Salo Kalupang,
Walanae Fro. ~ MalawaFro.
Tacipi/SalayerLimestone U LangiVolcanics Kalamiseng and Camba Formations (Sukamto
Camba& othervolcanicFrns, ~ 13alangbaruFro. 1982; Yuwono et al. 1985). Eocene shallow-marine
Intrusives ~ BasementComplex
carbonates of the Tonasa Limestone Formation out-
Fig. 1. Geological map of the South Sulawesi, showing crop only as a fault-bounded sliver at the eastern
the location of the Barru area. Modified after Sukamto margin of the Walanae Depression (Fig. 1; Sukamto
(1982) and Sukamto & Supriatna (1982). 1982).
 TERTIARY EVOLUTION OF S SULAWESI 367

WESTERN SOUTHSULAWESI EASTERNSOUTHSULAWESI localities between the Tonasa Limestone and

Age W Formationnames. Uthologies E W Formation names, lilhologies E
and thicknesses and thicknesses Camba Formations. The Tertiary lithologies dip
'" " ~ ~ TACIP~IW?~.LANAE" Fro.
Late
v%.
~v~
UPPERCAMBA .%%Y,
FORMATION& . % ~ ,
I T ......... eastwards at 10-25 ° .
Miocene 1- az~,~OTHER VOLCANICS . % ~ ,
v~ ~ North of the Bantimala Block the Tertiary
Midale ~.~ - : - ~ %'.'.'~':" FORMATION .~:%..-:~: lithologies are folded into a WSW-verging,
Mk)cene ~=-~-~-, LOWER .
_ _ ~ CAMBA Fm' regional-scale N N W - S S E trending anticline,

.... I ~"-'~ LIMESTONE Iq:;~
1 ~":-:~ FORMATION~ = ~
1 ~:£i~ -3oom-
C~te ~
~...-.-~
~ . _ . ~ ; ; . . . ,.,_,_, ,.,~
I ____ =====================
i Z.:.':E~>:'> > > >:+',!
:.--..---~,E;.I.I.I.I.I.:.I.bN
Oligecer
Late
Eocene
Middle
Lady
Eocene
PaleeocqLate
Eally
Palaeoc~ = ~ .... .~,~-~"~
'] Deep marine clastics &
C.....
Late ~. F ORMA_TIO_N-.~ FORMATION,, r ~ volcaniclastics
•~
Middle " ~ - ~ / 1 3 A S E M E N T ~>,',>,~
Fig. 2. Stratigraphic correlation chart of the units to the
west and east of the Walanae Depression in South
Sulawesi. Based on Van Leeuwen (1981), Sukamto
(1982), Sukamto & Supriatna (1982) and Grainge &
Davies (1983).
Structure and stratigraphy of the Tonasa
Limestone Formation in the Barru area
The Tonasa Limestone Formation in the Barru area
is bounded to the south by pre-Tertiary meta-
morphic and ultrabasic basement lithologies of the
Bantimala Block (Fig. 3; Berry & Grady 1987).
A smaller inlier (8 km across) of metamorphic
basement lithologies, known as the Barru Block
(Berry & Grady 1987), is located to the southeast of
Barru (Fig. 1). The overall structure of these base-
ment complexes is of relatively rigid blocks tilted
to the east and bounded on the remaining sides by
eastward dipping thrusts or sinistral wrench faults
(Sukamto 1982; Berry & Grady 1989). A number
of generally NW-SE trending faults also cut the
Tertiary sequence in the Barru area (Fig. 3).
The eastern and southeastern flanks of both
basement blocks are unconformably overlain by an
almost continuous stratigraphic sequence from the
Balangbaru Formation through to the Camba
Formation. Tertiary angular unconformities occur
at the base of the Malawa Formation and in some
I!
named here the Rala anticline (Fig. 3). On the
~1=1;;;~1~~ northeast limb of the Rala anticline there is a
complete stratigraphic sequence from the Malawa
Formation through to the Camba Formation, with
no apparent unconformities. On the southwestern
limb of this anticline Tertiary lithologies dip
...,,
20--40 ° to the WSW and the carbonate sequence
~ TOI LSA ~.*~ is considerably condensed or absent. An angular
unconformity separates the Camba Formation from
the older underlying lithologies on this western
limb.
Igneous bodies composed of diorite-granodiorite
and trachyte (Sukamto 1982) intrude the Tonasa
Limestone Formation and older lithologies in a
Key to lithologies: number of localities (Fig. 3). Although the age
~ Shallow marine limestone of the intrusives in the Barru area is not known,
Redeposited carbonate facies
Basinal marls
similar lithologies from other areas have been dated
D
(Marginal) marina clastics using K-Ar techniques as middle to upper Miocene
(Sukamto 1982).
Lavas & volcaniclastics
The Tonasa Limestone Formation in the Barru
~ Deep marine clastics
~ Metamorphic and ultrabasic area reaches a maximum thickness of 1100 m in the
Rala section (Fig. 4). The basal few metres of the
carbonate sequence are interbedded with silici-
clastics of the Malawa Formation. The earliest
carbonate sediments are composed of wackestones,
packstones and floatstones (Fig. 4). Some beds
contain a rather limited fauna of miliolids,
gastropods and occasional pseudomorphs after
gypsum or anhydrite. Other facies contain large
alveolinids or broken branching corals as well
as miliolids. These lower to middle Eocene facies
(T a, T. Wonders 1993 pers. comm.) indicate a
normal marine back reef-barrier environment with
minor restriction. Usually the initial carbonate
sediments pass upwards into thick successions
of metre-scale bedded packstones and rudstones
composed of abundant large Nummulites and
coralline algae. The coralline algae often encrust
the large benthic foraminifera to form rhodoliths,
suggesting shallow marine, relatively turbulent
conditions prevailed (Fig. 5a).
In the Rala section only, a subsequent deepening
of the environment is indicated by marls and
wackestones containing abundant, monospecific,
large, flat Discocyclina typical of the lower limits
of the photic zone (T. Wonders 1993 pers. comm.).
Most areas contain decimetre-scale planar-bedded
packages of bioclastic packstones, which include
planktonic foraminifera and show occasional
evidence of current or wave reworking (Fig. 4).
This facies shows an overall fining-upward trend
and is interpreted as outer shelf or slope deposits
368 M . E . J . WILSON (~ D. W. J. BOSENCE

" -? ! ecent =,uv, um

,..
Overlap of' ÷"i'.l,'r
. i Igneous intrusion
,- B a r r u / " ' " ' upper volcaniclastic
Bantimala . . . . ~ ~.=
~edeposited . . . . Lower marine
member (CF)
.- F a c i e s " ,-..,-,.ov~nm~reaet=o~
: . . 2 ' ~ ," ",
~/ Marls + redeposited
;;;;gd 552 • _ _ facies (TLF)
Bioclastic
", ", ", ", B l o c k ' , ", ", ", ", ", ", ", ", . . . . packstones (TLF)
~ Shalk)w-water
carbonat~ tacles (TLF}
iiiiiiii Malawa Formation
.... ~ r U
•% - - - - " Formation
" ' Basement complex
I "., "a
CF - Camba Fm.
TLF - Ton,s, Limestone I
v ~ v v v ~ v v v

v ~ v v v ~ v v v

v ~ v v v v v v v v

• ~ • ~ - - - - - ,

v ,., v v ,,, -., v ... -., ,., ,.,

v v v ~, v

v v ~

0 1 2 3 4
I
km
v v ~ v ~ v

v v v v v v ~
-. KEY
v v v v v v v • Village
v v v v v v v Dip and strike
~ v v v v v v ~- 2 5 direction of
bedding, with
v ~ v v v v v
angle of dip
v v v ~ v ~ v
~ Fault(tickmarks
v ~. ,, ~, downthrown block
v ~ v v v v

v v v ~ v
B. ,,v vv ,, ~ Anticline
~ ! ~ Unconformity
v ~ v v v

v v v ~ ~'~++~' V" Z Line of cross

~ I v A ~ ~ B section
v v v v v ,
Lineof measured
section
Rala S~¢ if
Z anticUne Line oL~"j Z' E o~- o oo
5~ + secti°ln Y'Y' ~-~'/~~- - 1
250 ~----------....~ . . I v ,,- ,,, v v v
SL
-250
SW
~. v v ~.
-~
NE
t
Fig. 3. Geological map of the Barru area with cross-sections (Z-Z', Y-Y'). Palaeocurrent data from the redeposited
facies are shown as arrows and rose diagrams. The cross-sections are drawn to scale. The locations of the measured
sections (A-B, C-D, E-F) in Fig. 6 and the cross-section (M-M') in Fig. 9 are shown. Measured sections are
composite sections along stream cuttings and tracks.
 TERTIARY EVOLUTION OF S SULAWESI 369

Lithology Age Facies description in intervening areas this facies is considered to be

- - -(3oarsev~canic/astic.s of tl~e laterally continuous between sections.
upper Camba Fm. The marls are pale green-grey in colour, poorly
---Lower-mbmber-of-the-Camba Fm.
cemented and often fissile. Planar lamination on a
1105120~0[:i . . . . . . . . . . . - d e e p marine shales interbedded
_ _ _willa yotc~ielastic_s ......... millimetre scale is a common feature of the marls.

1001
~ .....
NN2-
NN6
e/m
Miocene
In some localities the marls appear more homo-
geneous, and water-worn surfaces reveal mottling
caused by the presence of randomly orientated
Redeposited carbonate
~ . . . . '~ -
. .
facies interbedded with burrows on a millimetre to centimetre scale. A
804 ................. marls pelagic biota, including planktonic foraminifera
. __^ This section shows the and nannofossils is well preserved and occurs in
795- r~-z~, trends in the redeposited
m'°hgfacies in the Rala section. abundance in this facies.
743- ~ The number of beds and E The fine grained laminated nature of the marls
715 their thickness are not u.
drawn to scale. See Fig. m indicates that this facies was deposited in a low
649 . . . . . . . . . . . . . . 19 for a more accurate z energy environment largely from suspension. The
representation. O
presence of abundant, well preserved tests of
LU
P17 pelagic organisms indicates deposition in a deep
I.Eoc -i marine basinal setting above the carbonate
4"/0 ............. - . . . . . . . . . . . . . . . . . . . . . . . . . ~l:

4 5 0 ~ '="d"d N15 Decimetre-bedded bioclastic compensation depth. Burrowing in some localities

400 --k~-,-,-,-~, I.Eoc packstones z suggests an aerobic environment.
375 ~ .......................... ~

Discocyclina rich marls &

251 - ~ wackestones
2 0 7 ~ ...........................
Decimetre-bedded bioclastic
M packstones
- - -Num-rhulit~&fi?J-c-o~M,66~fg-al
bioclastic packstones & rudstones
m'-Milio/icl-& a/veo/inid . . . . . . . .
C M S f rn c G P C "-. wackestones & packstones
. . . . . . . .
Fig. 4. Simplified composite stratigraphic section of
the Tonasa Limestone Formation in the Rala measured
section. Location of section C-D is shown on Fig. 3.
with a upper Eocene age (N15, T. Wonders 1993
pers. comm.). In the northern and eastern parts of
the Barru area (Fig. 3) these packstones are overlain
by a thick sequence of upper Eocene to middle
Miocene marls interbedded with redeposited
carbonate facies (Fig. 5b), which are the main focus
of this paper.
Redeposited carbonate and marl facies
of the Barru area
Marls
This easily weathered facies is interbedded with
coarser, cemented carbonate facies, and is
especially abundant in northern and eastern parts
of the Barru area. The percentage thickness of
the marls in the Tonasa Limestone Formation
decreases from the Bangabangae section (E-F; 69%
thickness) moving southwards towards the
Bantimala Block (Figs 3 & 6). Marls are best
exposed in stream cuttings, especially in the
Bangabangae section, and although poorly exposed
Bioclastic packstone facies
This facies is seen in the upper parts of most
sections in the Barru area interbedded with marls
and other coarser carbonate facies. The bioclastic
packstone facies occurs most abundantly in the
. . . . . . . . . . . . . . .
Rala section (13.8% thickness of total deeper water
facies) and is best exposed in stream sections at this
locality (Figs 3 & 6). Although only well exposed
in stream cuttings, beds of this facies are thought to
be laterally extensive on a scale of tens of metres.
In this facies beds are planar-bedded with bed
thicknesses varying from 10-90 cm. Bed contacts
are usually planar and sharp. The grain size varies
from fine to coarse sand grade between adjacent
beds. Sedimentary structures were not observed
and burrows with millimetre to centimetre
diameters are common in this facies. Although
delicate tests of planktonic foraminifera are
frequently well preserved, more robust fossils
such as coralline algae, large and small benthic
foraminifera and echinoids are common as
fragments. Based on 'bundling' of beds and grain
composition, this facies can be subdivided into two.
(a) Packages of medium bedded packstone which
generally lack intervening marly units were
found in only one locality in the Rala section
(see below, Fig. 12). Faunal elements contained
in a single bed are consistently of one age. This
subfacies is very similar to the bedded pack-
stones described above, which underlie the
marls (Figs 3 & 4).
(b) Single packstone beds interbedded with marls
commonly contain both grains which are con-
temporaneous with and older than the inter-
vening marls. This subfacies often includes
370 M. E. J. WILSON & D. W. J. BOSENCE

Fig. 5. Carbonate facies of the Tonasa Limestone Formation from the (a) lower and (b) upper parts of the Rala
Section. (a) Shallow-marine Nummulites and coralline algal rudstone. The coralline algae encrusts the large benthic
foraminifera forming rhodoliths. Scale is in centimetres. (b) Typical outcrop of marls interbedded with redeposited
carbonate facies. The lower resistant, bioclastic packstone bed is separated from a clast-supported breccia unit by
rather easily weathered green-grey marls. Vertical field of view is 1.5 m.

sand grade, angular clasts from all the forma- are interbedded with marls. Coarser breccia facies
tions underlying the Tonasa Limestone are almost invariably overlain by this facies (see
Formation. below). The percentage thickness of this facies
decreases northwards away from the northern
A redeposited origin is inferred for this facies
margins of the basement blocks (Fig. 6). Beds of
because it is interbedded with basinal marls and
this facies are often laterally continuous on a scale
contains fragmented shallow-water bioclasts.
of tens of metres, although a 10% decrease in bed
Sedimentary structures which might indicate the
thickness over 5 m has been observed.
mode of reworking are absent. The packages of
Bed thicknesses of this facies vary from 5 cm up
bioclastic packstones may perhaps have a similar
to 110 cm (Fig. 7). The basal bed contacts are sharp
origin to the outer shelf-slope packstones under-
and often planar. Less commonly this facies has an
lying the marls (Fig. 4). In terms of clast content the
erosive base and exhibits rare groove or flute casts
single beds of bioclastic packstones are comparable
(see below for palaeocurrent data). Beds are graded
with the graded bioclastic packstone facies
both in terms of grain size and composition; fining
described below. Non-graded packstone beds are
upwards from pebble-coarse sand grade to fine
often documented in sequences containing abun-
sand-silt grade (Fig. 8a). Heavier gravel or sand
dant graded packstones and have been interpreted
grade schist, ultramafic or sandstone clasts tend to
as grain flow or modified grain flow deposits
be concentrated in the lower parts of beds. Rare
(Lowe 1976; Cook & Mullins 1983; Gawthorpe
water escape structures and imbrication of planar
1986; Eberli 1987). However, criteria indicating
clasts are present. Planar lamination on a milli-
grain flow deposits such as dish structures, inverse
metre scale is a common sedimentary structure in
grading, diffuse lamination and outsized clasts
the upper part of beds (Fig. 8a). The upper bed
(Middleton & Hampton 1976; Cook & Mullins
contact may be planar or slightly undulose and
1983) have not been identified in this facies.
sharp or gradational into marls.
This facies incorporates a large variety of often
fragmented bioclasts, including large and small
Graded bioclastic pack-grainstone facies
benthic foraminifera, coralline algae, echinoderm
This facies is well exposed and common through- and very rare coral debris. It contains up to 25%
out the upper part of most sections in the Barru well preserved planktonic foraminifera, especially
area. Graded, bioclastic pack-grainstone facies in the upper finer part of beds. Lithic clasts include
 TERTIARY EVOLUTIONOF S SULAWESI 371

NF
,~
Bangabangae section
10km

Upper coarse volcaniclastic

~

Dm!
5km
Rala section
B
.~ S
Doi-doi section
m i"~"~"~"~"~'~ member of the Camba Fm. ~,,~,,~,-~ m
775 ~'1 Lower member of the Camba 1200 ['l 520.
Fro. - Deep madne shales ~4n= I.'.~ 425
interbeclded with volcaniclastics ..,,,-'-'------------ ................
.°_ . . . . - . . . . -~- . . . . -~- . . . . _---

1
488 ~ ' " - ' " ' - ~~~-~'""~-~"
"~"~ 207,
, , , ,
Redeposited carbonate ,,'A MSfmcGPC
facies interbeddedwith /'
~ ~ marls of the Tonasa ~_ . . . . . '
"~ ,o Limestone Fm. ,,'"

78 ~ ~ - - -
...... aliow-wa ei acl;soit -;- ona .....
-t::::! .........
,,

E ~ MS f m c G PC Limestone Formation C I I~ s f' m c G P C:

Ban~labangae section Rala section Doi doi section
Total thickness of made & redeposited facies 408 m 635 rn 218 m
Total number of redeposited beds 76 173 66
Clast-supported breccia 0.93 39.64 58.51
Graded bioclastic packstones topping the above 0.47 11.85 20.14
Marl supported breccia 0.40
Graded bioclastic packstone 8.82 12.63 4.35
Bioclastic packstone 6.17 13.77 7.26
Planktonic foraminifera wacke/packstone 7.82 1.10 5.62
Mad 69.02 20.22 4.13
Other 6.76 0.40

% % of limestone conglomerates, packstones and marls

100 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ~. . . . . . . . . . . . . . . . . . . .
I"'::::::::.Marl : : : : : ~ ........ .... -_.,~.v.y.-......,..d..~~. r~ ~
50 '- ............. ¢" ¢'" "'~¢" " "

0
% % Amalgamation of redeposited carbonate beds
100
5O
0
DISTAL ~ ' PROXIMAL
BASlNWARD
Fig. 6. Three composite stratigraphic sections through the redeposited facies interbedded with marls showing
northwards proximal to distal and basinward trends. See Fig. 3 for location of the stratigraphic sections (A-B,
C-D, E-F).

metamorphic (see below) and quartzose sandstone clasts, as nodules and as irregular but apparently
lithologies, a wide range of limestone facies and continuous 'beds'. Silica nodules are parallel to
marl clasts. Both packstone and grainstone textures bedding, up to 15 cm thick and 5 0 c m across
occur, although in the majority of beds the matrix is and sometimes follow centimetre diameter
marly or composed of finely fragmented bioclasts. burrows.
Trace fossils on a millimetre to centimetre scale, Because this facies is interbedded with basinal
including Taphrhelminthopsis are common at bed marls and contains fragmented lithoclasts and
contacts and may be parallel or perpendicular to shallow marine bioclasts, it is considered to be
bedding. Silica replacement is frequent within redeposited. Sedimentary structures such as an
this facies and occurs in three main forms: as erosive base, normal grading and parallel lamina-
chalcedony within individual bioclasts or litho- tion are typical of calciturbidites (Crevello &
372 M.E.J. W I L S O N ( ~ D, W, J, B O S E N C E

Graded bioclastic pacidgrainstone facies Planktonic foraminifera wacke/packstone facies

6) U p p e r bed contact is planar or slightly
5) Planar sharp or transitional
iiiiiiiiil overlyingUndul°se
iand
n sharp
t O or
m gradational
a r l s
l----
'k 5) Millimetre scale parallel l:i::i::i::3 ,~~ , ~
contact into marls
4) Grain- or matrix-supported

T
8
'i~i~i~i~i~i~i~
~
%.4)
laminations

'..~.~.~,.~.~.~.~.~.~.~.~.
i~i~i~i~i~i~i~'i~i~
.<~<~<~<.<~<~<~<~<~<~
Rare water escape structures
3) Normally graded
packstone fining upward
from g r a v e l / c o a r s e s a n d to
~ ~
~%%~
~%~s=~%~,
~;%%%%%~
in a marly matrix
3) Well preserved planktonic
foraminifera comprise 7 0 - 9 0
% of the grain types.
2) Normal grading from
<~~<~<
-i~ ~-~ ~'~"l~'~'~"l~"l~'~
~<~<~ ( ~ < ~ < . < ~ < ~ •
fine sand/slit, ~ ~ ' ~ m e d i u m sand ~lrade to silt
,.~-~.~.~.~.~.-~.~.-i~.~.~.~-~
~_~.~.~-_.~ grade is occas,onally
m
..~ •{'~•~ •~ •~ •~. ~; ~.'.~,'-~,'-~
.,~.~.~-~-i~.
~ ~',,'~ imbrication
.~'1~-~'1
I ~_~_~_~
-~-~-~ ~-~
seen. Other beds appear
ungraded

,~,.,~,~,~;,:,~ ,, sharp o . a n ,~=~-~;~]~ 1 ) Planar sharp or

.
'¢t:<
. . . t
.
~ |......~...:
. <
. .E.
..... .¢. .
E . .
d . . ~
. .~. .
( ~ E ( ~ ~ ~._ p lanar base to bed ~-~L~.I transitional basal contact
•-'"'""-":~ with rare groove or
I i i i
............."1 f's gs (~ flute casts. M silt fs ms cs

Clast-supported breccia facies topped by a Marl-supported breccia facies

graded bioclastic pack/grainstone facies

7) A graded packstone facies

~ii iilil topped by a graded bioclastic
pacWgrainstone facies
'<~.~!~-~ (see above) almost invariably ~-,,~-,,~ 5) Sharply overlain by graded
'~<'~,.~
'~
overlies the coarse limestone
conglomerate
~i~2
~;~;~:~:~x packstone facies (see above left).
"~!'-'!~!"!'-'!~'-3.::: =....... .d
'i~i~ 6) Relatively sharp contact ~.--.~.-.-~"~"~ 4) Chaotic poorly
"'"~7 T"'I.~.,'[ 1:%1 sorted facies which
::'"',~.,..'L.'._"'_" _ "'~ ,1:.":.1 lacks any signs of

~ ¢ 1 ....
~i;~: ~ ";~'~:~:~:~i~
.o. ¢1~. ~ m. . 5). Shght
. fining upward of
~(~,~:;~;: °.'~:?.o:Z :~.~.~ angular clast-supported
:;:.~.~{6.,~.,'~.~:U~~:~!~; ~ :$ breccia. ~"~/~~)!~!,~)(.~(.~t packst0ne channel,
~r~t..~.~.~(.~,l which includes cross
::_~DS;;(S:m;:~;~".~:~!:~ 4) Clast s i z e varies t ~ ~ lamination and clast
?~!6!:'o!~:~9'.::~!~:'~:~:~ from 4 m across /~...:~::.~ imbrication.
i~:~:i~'~:~)~),~:~ down to fine sand
!~.~:~/:?~;~i:~!~::~ i~ :~:~fi~ grade ~ - - : : : : I =~Sub-angular
!i~;'.~i~:~,:~i~'~<~-~:~i~i~ii~ l 3) C o m m o n circum-clast ~ - ' - J ~:'-'1 clasts up tO 1 m
~~:~_:~9.i~i~I~:~~:~_:~(~. ~ stylolitesa n d rare clast ---7-/~ "N-..7 ~'-": I across mostly
i~.:.~;~':-~.~'!~.:~,'~ ~ ~l imbrication. 7" ~.-I J':':l supported by a
~.~b~£)~!6~:2:~
,~§~'/~i6-.:~?0?~b:~!:~
!] 2) Coarsening upward at
I base m a y occur.
1 ~ ~narp, planar or
irregular base
~ M fs cs G C
1 ~) Sa~a~gr~t~tctdulOse

M fs cs G C

Fig. 7. Schematic diagrams showing the main features of four of the redeposited facies. The bioclastic packstone
facies are not shown as they are planar bedded and lack other visible sedimentary structures. Diagram not to scale.

Schlager 1980; Cook & Mullins 1983; McIlreath & of the silica are sponge spicules and microfossils,
James 1984). A complete Ta__e Bouma sequence present in both the marls and packstones, and litho-
(Bouma 1962) characteristic of siliciclastic clasts such as quartzose sandstones present in the
turbidites is not seen in this facies. Incomplete redeposited carbonate facies. The Tonasa
Bouma cycles are a common feature of other Limestone Formation is similar to a number of
modern (Schlager & Chermak 1979) and ancient other ancient carbonate deposits in which it is
redeposited carbonates interpreted as calci- reported that redeposited facies are the only
turbidites (Pfeil & Read 1980; Mcllreath & James carbonate facies in which silica replacement is
1984; Gawthorpe 1986; Eberli 1987; Braithwaite common (Gawthorpe 1986; Bustillo & Ruiz-Ortiz
& Heath 1992; Herbig & Bender 1992). Since 1987; Coniglio 1987; Eberli 1987; Reijmer &
skeletal carbonate grains are ofter~ porous or Everaars 1991; Herbig & Bender 1992).
irregular in shape, grain size distribution and sedi- Silicification is thought to be prevalent only in the
mentary structures in calciturbidites are commonly redeposited facies because the rapid burial of trans-
less regular than in siliciclastic turbidites (Tucker & ported siliceous tests and clasts prevents the dis-
Wright 1990). solving silica passing into the overlying seawater
The radial chalcedony preserving 'ghosts' of (Bustillo & Ruiz-Ortiz 1987). Certain horizons or
earlier bioclasts and the smooth irregular shape of burrows perhaps had a higher original porosity and
nodules both indicate the secondary replacement the silica has preferentially remobilized to these
nature of the silica in this facies. Probable sources areas during diagenesis.
 TERTIARY EVOLUTION OF S SULAWESI 373

Clast-supported breccia facies which up to 30 cm of the lowermost part of the bed

This sedimentary facies is interbedded with marls
and bioclastic packstones and occurs particularly
abundantly in localities close to the northern
margins of the basement blocks (Fig. 3). In the
Doi-doi section (A-B) this facies comprises 58.5%
of the thickness of the deeper water carbonate
facies (Fig. 6). Beds are laterally continuous on a
scale of tens to hundreds of metres. This facies is
almost invariably capped by graded bioclastic
pack-grainstone facies (described above) and the
contact may be sharp or transitional over about
10 cm.
Usually the sequence of clast-supported breccia
topped by a graded bioclastic pack-grainstone
facies is between 70 cm and 5 m thick, though units
up to 25 m thick do occur (Fig. 7). The basal con-
tact of this breccia facies is usually sharp and may
be planar or more irregular. In rare cases, this facies
exhibits a transitional basal contact with marls, in
contains up to 25% of the green-grey marl as a
matrix. The lower part of the sequence is a coarse,
clast-supported breccia with angular-sub-angular
clasts usually up to 60 cm across. In thicker units
the maximum clast size seen is 4 m across. The bio-
clastic packstone matrix of this facies is rarely seen
due to circum-clast stylolites which give a fitted
fabric to the breccia. Locally, this facies may
coarsen upward over the lower 40 cm (Fig. 7).
More usually there is a slight fining upward of clast
size through the breccia facies. Rare imbrication
of planar clasts does occur. Replacement or partial
replacement of clasts by silica is sometimes seen
within this facies (see above). A wide variety of
lithic clast types is present in this facies (Fig. 8c).
These include metamorphic, igneous (see below)
and quartzose sandstones lithologies, and a range
of shallow-water wackestone, packstone and grain-
stone and deeper-water marl clasts. Matrix com-
prises a relatively low percentage (2-8%) by

IIO
!iii~iiiiiiiiiii~ii!i~
~ ~i~ ¸¸~¸~ ...... ,i~ii!i ¸~ii~i ~...........~..............~~i!i~~ % ~

Fig. 8. Photographs of the redeposited carbonate facies. (a) Graded bioclastic packstone bed, showing fining upward
and parallel lamination at the top of the bed. Darker grains are formed of older non-carbonate lithologies. Scale is in
centimetres. (b) Amalgamation of clast-supported breccias topped by graded bioclastic packstone unit. The lower
coarse, angular clast-supported breccia is sharply overlain by a graded bioclastic packstone unit. Not that the parallel
lamination in the graded bioclastic packstone becomes closer together as the unit fines upwards. The bioclastic pack-
stone is sharply overlain by another clast-supported breccia layer. (c) Outcrop of angular clast-supported breccia
showing the variety of clast types present within the redeposited carbonate facies. The dark clasts are schists and
ultrabasics and were derived from the older formations in South Sulawesi. Clasts from a range of shallow-water
carbonate facies and small rip-up clasts of basinal marls occur in abundance. Scale is in centimetres. (d) Photo-
micrograph of a bed of planktonic foraminifera wacke-packstone. Scale bar is 1 mm across.
374 M . E . J . WILSON t~ D. W. J. BOSENCE
volume of the breccia layer and is composed of
marl or finely fragmented clasts or shallow-water
biota. Although this sequence of units is inter-
bedded with marls, amalgamation of both the
breccia and overlying graded packstone facies
frequently occurs, especially adjacent to the
northern margins of basement blocks (Fig. 8b).
Sometimes a coarse breccia directly and sharply
overlies another breccia and the packstone facies
is missing.
This facies is redeposited because it is inter-
bedded vdith basinal marls and contains a variety
of coarse lithoclasts and fragmented shallow-water
biota. Clast-support, poor sorting, occasional
inverse grading and clast imbrication suggest trans-
port and deposition from a lower traction layer
(Lowe 1982; Pickering et al. 1986; Tucker &
Wright 1990). Circumclast dissolution has removed
much of the matrix (cf. Davies 1977). Debris flows,
which occasionally show inverse grading at the
base, can transport coarse material with as little as
5% matrix (Rodine & Johnson 1976; Picketing
et al. 1986). However, normal grading of clasts
towards the top of breccia units argues against a
debris flow as the mode of origin. A high density
turbidity current with a lower inversely graded R 2
layer and upper normally-graded R 3 layer (cf.
Lowe 1982; Tucker & Wright 1990) is inferred
for the transport and depositional origin of this
facies.
After deposition of the coarser breccia material,
the graded bioclastic packstone capping this facies
is thought to have been deposited out of suspension
from turbulent flow. An analogous experimental
example of a high density turbidite topped by a
lower density turbidite has been described by
Postma et al. (1988). Many examples of two layer
carbonate sediment gravity flows have been
described from ancient (Cook & Taylor 1977;
Hubert et al. 1977; Cook 1979; Johns et al. 1981;
Kepper 1981; Hiscott & James 1985; Gawthorpe
1986; Eberli 1987; Watts 1987) and modern
(Mullins & van Buren 1979; Schlager & Chermak
1979) settings. Turbidity flows are often entrained
on top of coarse gravity flows and can be generated
by reverse shear at the flow-water interface in the
head region of the coarse flows (Hampton 1972). A
genetic relationship is therefore suggested between
the clast-supported breccia topped by the graded
bioclastic pack-grainstone facies and the graded
bioclastic pack-grainstone facies (see below).
M a r l - s u p p o r t e d breccia f a c i e s
This facies occurs in only two localities in stream
cuttings in the Rala section (Figs 6 & 12). This
facies is laterally continuous on an outcrop scale
(m). Graded bioclastic pack-grainstone facies
sharply overlie the marl-supported breccia facies
in both outcrops.
Bed thicknesses of this facies are up to 2 m, with
sharp and undulose bed contacts (Fig. 7). In this
poorly sorted facies sub-angular clasts up to 1 m
in diameter are present (Fig. 7). The matrix is
composed of marl, and both matrix and clast-
supported fabrics occur, although the former
support mechanism is more common. One bed
contains a N-S orientated, flat-topped channel
60 cm deep and 2.5 m wide, composed of breccia
fining upward to a grainstone. Internal sedimentary
structures within this lensoid body include planar
cross-lamination and a layer with abundant marl
clasts, some of which show imbrication. A large
variety of sub-angular limestone clasts, including
wackestones, packstones and float-rudstones occur
within this facies. Large rounded clasts of
green-grey marls and clasts of graded bioclastic
packstone and clast-supported limestone breccia
occur.
The massive, chaotic, poorly sorted, nature of
this facies and presence of large sub-angular clasts
floating in a marly matrix are features typical of
debris flows (Cook et aI. 1972; Middleton &
Hampton 1976; Lowe 1979, 1982; Cook & Mullins
1983). In debris flows the cohesive strength of a
fine sediment-water matrix is sufficient to support
and transport large clasts on slopes as low as 1°
(Cook et al. 1972; Prior & Coleman 1982; Hiscott
& James 1985). The range of clast types indicates
that both shallow and deeper water carbonate
lithologies were incorporated into this facies. Since
clasts of reworked breccia or packstones were not
seen in other redeposited facies, a deeper water
origin for this facies is suggested. Modern mud-
supported and clast-supported debris flows occur
in proximal and distal parts of the lower slope of
the Bahamas respectively (Mullins et al. 1984).
Graded bioclastic packstones capping debris
flow deposits have been described from other
modern and ancient deposits (Crevello & Schlager
1980; Mullins & Cook 1986). Channels containing
graded lenses of carbonate sands within debris
flows are less commonly reported (Braithwaite &
Heath 1992). It has been suggested that turbulence
may have played an important role in supporting
grains at the top of flows (Cook et al. 1972;
Hampton 1972). The channelized normally graded
body within this facies perhaps indicates
amalgamation of the breccia.
Planktonic foraminifera wacke-
packstone facies
Although this facies occurs in the upper part of all
sections in the Barru area, it is best exposed and
 TERTIARY EVOLUTION OF S SULAWESI
occurs most commonly in the Bangabangae section
(E-F) where it comprises 7.8% of the thickness
of the deeper water facies (Fig. 6). This facies out-
crops in stream cuttings and is only known to be
laterally continuous on an outcrop scale (m).
Planar beds of this facies between 20 and 60 cm
thick are interbedded with marls (Fig. 7). This
facies differs from the marls in that it is cemented
and contains a higher percentage of planktonic
foraminifera and fragmented shallow-water biota.
Upper and lower bed contacts are planar and may
be transitional or sharp with interbedded marls.
Normal grading, fining from medium to silt-size, is
occasionally seen (Fig. 7). Well preserved plank-
tonic foraminifera comprise 70-90% of the bio-
clasts present in a micritic or marly matrix (Fig.
8d). Fragmented echinoderms, benthic foraminifera
and coralline algae form the remaining bioclasts.
Silica nodules parallel to bedding are present
within this facies.
The normal grading sometimes seen in this
facies suggests deposition by waning current flow.
Reworking could also account for the increased
concentration of planktonic foraminifera and
presence of fragmented shallow-water bioclasts,
relative to the intervening marls. Although sedi-
mentary structures which could differentiate
between storm, contour currents or mass flow
processes as the mode of reworking are absent, the
location of this facies relative to other redeposited
facies suggests a mass flow depositional origin (see
below).
Origin of the redeposited carbonate facies
Redeposited carbonate facies have been related to
slope instability (Hopkins 1977; Schlager &
Camber 1986), eustatic sea-level changes (Watts
1987; Reijmer et al. 1988, 1991, 1992; Burchell
et al. 1990; Reijmer & Everaars 1991; Herbig
& Bender 1992) or tectonics (Bosellini 1989;
Fernandez-Mendiola & Garcia-Mond6jar 1989;
James et al. 1989; Playford et al. 1989; Elmi 1990;
Garcia-Mond6jar 1990; Kenter et al. 1990; Watts
& Blome 1990; Everts 1991). In this section it is
shown that the majority of the redeposited facies in
the Barru area are derived from a tectonically active
faulted platform margin. Downcurrent changes in
the organization of the redeposited facies are also
discussed below.
Subdivision of the redeposited facies
Redeposited carbonate facies interbedded with
basinal marls are located in the upper parts of
sections in the northern and eastern parts of the
Barru area (Fig. 3). West of the Rala anticline
375
redeposited facies interbedded with marls are
considerably reduced in thickness or are absent
(Fig. 3). In terms of spatial distribution and base-
ment clast content the redeposited facies can be
subdivided into two groups (Fig. 3). (1) The
Bantimala redeposited facies are located to the
north and east of the Bantimala and Barru Blocks
respectively. This facies group contains basement
clasts of lithologies, such as quartz-mica schists
and ultrabasics, which outcrop in the northern part
of the Bantimala Block. (2) The Barru redeposited
facies crop out only to the north of the Barru Block.
These facies contain basement clasts of lithologies
such as serpentinite and granite-granodiorite,
which are present in outcrops within the northern
margin of the Barru Block.
Proximal to distal changes
In the Bantimala redeposited facies the total thick-
ness of marls and redeposited facies thickens and
then thins towards the north over a distance of
15 km (Fig. 6)~ When the thickness of redeposited
units is expressed as a percentage of total strati-
graphic thickness there is a clear decrease in re-
deposited carbonate units, relative to marls,
towards the north (Fig. 6). Correspondingly the
percentage of amalgamated redeposited carbonate
beds also decreases towards the north (Fig. 6).
These factors indicate proximal to distal and basin-
ward trends towards the north.
The occurrence of a clast-supported breccia
topped by a graded bioclastic pack-grainstone
facies suggests a genetic relationship between this
sequence and single beds of the graded bioclastic
pack-grainstone facies. The spatial distribution of
facies (Fig. 6) might also suggest a downcurrent
change from a two layer into a single layer sedi-
ment gravity flow deposit. Downcurrent or lateral
changes of two component gravity flows into single
layer flows have been reported from both modern
and ancient carbonate deposits (Davies 1977; Cook
1979; Krause & Oldershaw 1979). The planktonic
foraminifera wacke-packstone facies may be the
correlative distal deposit of the graded bioclastic
pack-grainstone facies (cf. Davies 1977). Outcrop
constraints preclude the possibility of proving this
inferred proximal-distal facies relationship.
Nature o f clasts
The majority of lithoclasts in the redeposited facies
are angular to sub-angular, often with the clast
margin cutting across the grain fabric. A continuum
of clast sizes ranging from 4 m down to fine sand
grade are present. A wide variety of clast types
occur in both groups of redeposited facies (Fig. 8c).
376 M.E.J. WILSON & D. W. J. BOSENCE
Schist, serpentinite, ultrabasic, quartzose sandstone
and shale clasts were derived from the underlying
formations. A range of limestone clasts which
vary both in terms of age and environment of
deposition are also present in the redeposited
facies. Often a redeposited bed will contain lime-
stone clasts which are both contemporaneous with
and older than the intervening marls. Rip-up clasts
of marl are often found within the redeposited
facies.
Tectonic versus eustatic origin
The deposition of the redeposited facies may have
been triggered by eustatic or tectonic events. The
variety of clast types excludes other possible
causes, such as gas charging, tidal or surface wave
action, rapid deposition or patchy shallow marine
cementation resulting in unstable slopes. A number
of factors, outlined below, exclude the possibility
of a eustatic mechanism as the main cause of the
majority of the redeposited facies.
Shallow-water carbonate lithologies in the Barru
area and on the Tonasa Carbonate Platform to the
south, with the exception of the lowermost beds
in the carbonate sequence and some beds in the
Doi-doi section, lack lithic grains. It is therefore
suggested that no major areas of basement highs
existed during the deposition of Eocene shallow-
marine carbonates. Erosion of basement lithologies
or siliciclastic grains from carbonate deposits due
to sub-aerial exposure would have required the
removal of between 60-470 m of shallow-water
carbonate deposits. This is much greater than, or
close to, the inferred maximum for, short-term
sea-level falls for the Tertiary, the value of which is
thought to vary from 75-100 m (Haq et al. 1987;
Sarg 1988). Limestone clasts in the Bantimala
redeposited facies contain no evidence for sub-
aerial exposure until the latest Oligocene. Prior to
the deposition of marls during the late Eocene, both
the Barru area and the area of the Tonasa Carbonate
Platform were the sites of widespread accumulation
of shallow-water carbonates. During the late
Eocene rapid deepening occurred in the Barru area,
whilst shallow marine sedimentation with only
localized evidence for Oligocene karstification
continued on the Tonasa Carbonate Platform.
The range and immaturity of clast types within
the redeposited units suggest derivation from a
tectonically active faulted margin of a carbonate
platform (cf. Kepper 1981; Hurst & Surlyk 1984;
Martini et al. 1986; Eberli 1987; Burchette 1988;
Eaton & Robertson 1993). Carbonate slopes
become unstable at angles of 30-40 ° (Kenter
1990). Derivation of the redeposited carbonates
is thought to be related to oversteepening of the
platform margin caused by faulting. The underlying
formations are relatively easy to erode and under-
cutting of the platform margin may also have
contributed to the oversteepening. Provenance
studies, the spatial distribution of redeposited facies
and field evidence outlined below, indicate that the
redeposited facies were derived from the northern
faulted margins of the Bantimala and Barru blocks.
A eustatic control or enhancement of a tectonic
control cannot be dismissed for some of the
redeposited facies. Increased shallow-water
carbonate production and 'shedding' into deeper
water areas, caused by increased accommodation
space and more favourable conditions may be
related to eustatic changes. For example, packages
of decimetre bedded bioclastic packstones, which
only contain fossils contemporaneous with the
adjacent marls, may be related to eustatic changes
(cf. Reijmer et al. 1988; 1991; 1992; Watts 1987;
Burchell et al. 1990; Reijmer & Everaars 1991;
Herbig & Bender 1992).
Field evidence f o r derivation from a faulted
platform margin
No complete sections through the Barru re-
deposited facies are exposed. In general, the maxi-
mum clast size decreases and the bed thickness
of these redeposited facies thins towards the north
(Fig. 9). These factors indicate basinward and
proximal to distal trends towards the north away
from the Barru Block within the Barru redeposited
facies.
The Barru Block is bounded to the north by a
fault downthrowing and dipping steeply (60-70 °)
towards the NE (Fig. 3). This fault juxtaposes
basement lithologies in the footwall against coarse
redeposited facies of the Tonasa Limestone
Formation in the hanging wall (Fig. 9). Although
this suggests normal fault movement, kinematic
indicators were not seen and the fault surface itself
is not exposed, so it is not clear if oblique-slip was
also involved. All the basement and limestone
lithologies reworked into the Barru Redeposited
Facies outcrop in situ within or adjacent to the
margins of the Barru Block (Fig. 3), suggesting
local derivation. These facts strongly suggest that
the area of the Barru Block was the site of shallow-
water carbonate accumulation and the Barru
redeposited facies were derived from an active
syn-depositional faulted margin to the Barru Block.
No similar fault is exposed juxtaposing lith-
ologies of the Bantimala Block against redeposited
facies south of Rala. In part this may be due to
outcrop constraints, with the upper volcaniclastic
member of the Camba Formation unconformably
overlying much of the northern edge of the
 TERTIARY EVOLUTION OF S SULAWESI 377

SSE NNW

~ ...... Erosion of footwall

,, %" ,, ~' ,,:~.--~--k ,,-~ %" ,~ <--~-~-~--~'i lithologies into the
, ,'%',', ,', ,', t, s, s%',', , ' ~ , ' % ' ~ t % ' t ~ s % ' ~ 1 ~,~ hangingwall area Present day
,, z...~.,~,, ,, ,, ,, ,, ,, ,,~¢...,¢_.~ ,, ,, ,,1 l i b , ............. land surface ,
% %" %'~1,~ % % Y + 4~ % % % % % % % % % | ~ .................... .
,',',~',','4°,%~,',g$,,',',',',~, I V ~ . . : . . : - . = . :
%
%
.,.
% %
%"
,,-
'%
%
%
.,,
%
,,,
%
%
i~%
%"
.,. ,., ke -
" ..
= :
.. ':':.:'.~""%"::i:(.:.!!-:i::~:~i)!~ ::....
::
. .::..
•.............................................................................
<: .:"~;4.~.~.:::: .I:.::+-. , , , , , o , . , , , , , . . . . . . . . . . . . . . . . . . . . . . . . .
. ., ':~:~,?i!:~i~!:~::?~i~!i?i!~.~...
~ . . . . . . . . . . . . . .
,,,, .......................
/
i ~% ~ ~ %,% ~ @ @ @ i ~ ~ ~ ! % % % %. .::. .~ ..... L : : ::~:~.?~:;-:?:.:~-¢...,
.... • ...........!ili,........
. . . ,o,o,,
. . . ........ ~:~E

• ". " . ~ ". ~ [ ~ " ~ +. ÷ Y . ". "'1~. "..Z'. ' ~ . ~ ~ ....

" , , %' , , '%, , ' ,%, ' ~%, , %~ , ,%' ~ k%~~ ÷ ' ÷@' ~ ÷, 1' ,%, " , ,%~ ,%1~l,~
%
' , , ~ , "% ~ % ~ ~ '
~ii~!~i::~.~i~!~
~
--- - . . -. . .-. .-. .-. . -. . .-. .-. .= -----------"""'""
................................
II
~ ~.%'.~ % % ~ . ~ ÷ +~. ' ~ . ~ . ~ ' . ".~ ~ ~ ................................ I
% % % % %" % % % %~+ + ~ % % % % % ~ ~ ~ ............... I
% % % % % % % % % ~ @ ~ % % % % % % % %1"...~ :2 . . ; . . , . ' . . :; :::.. ,.. i::.. , ; . . * ~ . . . . . . . . l.ii i I]'l i"" "l. . . . . . . . . . ii.

310/60-70 ° Underlying lithologies not exposed

................................
',KEY ; Main fault

M~::ld;posited packstonesand I ~ i oE 200"1

matrix to the limestone bm~ia/~ ~ ~I 100"t
Shallow marine carbonates / ~ E ~1 E o50-1
Intrusive - pre-Tertiary ' -~u'! .E_=N104 " ~ " ~ ' ~ ~ - ~
[~-~ Calcite-veined metamorphic ' , : ~ 11
,_____bas_~_m_~_n_t-_pr_~-_T_e_rtia~_. . . . . . . . . ~ 0 160 250 560 1 6 0 0 2000 4 6 0 0 8000 -~
Distance awayfromfault(metres)

Fig. 9. Schematic cross-section through the northern bounding fault to the Barru Block, showing redeposited facies
thinning and fining towards the north away from the fault. See Fig. 3 for location. Not to scale.

Bantimala Block (Fig. 3). Faults south of Rala with

Derivation
similar trends as the north-bounding fault to the
Barru Block may have acted as sources for the On the eastern limb of the Rala anticline the lower
redeposited facies. However, movements on these marine member of the Camba Formation con-
faults contemporaneous with deposition of the formably overlies the Tonasa Limestone Formation
Bantimala redeposited facies cannot be proved. (Fig. 3). On the western limb of this anticline the
lower marine member is absent and an angular
unconformity separates a condensed sequence of
Configuration of the faulted platform the Tonasa Limestone Formation, or older forma-
tions, and the upper volcanic member of the Camba
margin
Formation. The lack of abundant limestone clasts
Redeposited facies are inferred to have been in the volcaniclastics directly above this uncon-
sourced from at least two faulted northern margins formity suggests that major erosion had occurred
of the Barru and Bantimala blocks. The lack of prior to the deposition of the Camba Formation.
channelization and facies distribution suggests that This angular unconformity may therefore have
both groups of redeposited facies were derived been generated during the Eocene to Miocene and
as sheets of carbonate debris from a faulted line- supplied material for the Bantimala redeposited
source forming base-of-slope carbonate aprons facies.
(Mullins & Cook 1986). The Barru redeposited Two kilometres south of Doi-doi is a fault with
facies are inferred to have been derived from the a trend similar to the north-bounding fault of the
fault which today delimits the northern edge of Barru Block (Fig. 3). This normal fault dips to the
the Barru Block. The precise area and fault for NE and juxtaposes basement lithologies against
derivation of redeposited facies from the northern the Balangbaru Formation. Southwest of this fault a
edge of the Bantimala Basement Complex is less thin layer of the Malawa Formation and upper
clear. Regional geological observations and Eocene packstones of the Tonasa Limestone
palaeocurrent data can be used to define possible Formation unconformably overlie metamorphic
configurations of the platform margin and to relate lithologies of the Bantimala Block. The packstones
the two source areas. are in turn separated from the upper volcanic
378 M.E.J. WILSON & D. W. J. BOSENCE

member of the Camba Formation by an angular
unconformity. The eastern part of this fault, which
does not penetrate the Camba Formation, is
intruded by a later diorite stock. The offset of the
N-S trending contact between the Malawa and
Tonasa Limestone Formations north of this
intrusion indicate normal or possibly strike-slip
movement on the fault (Fig. 3). South of the
intrusion the Balangbaru, Malawa and Tonasa
Limestone formations are all present. The Tonasa
Limestone Formation in this location is composed
of 3 1 0 m of shallow-water Eocene-Oligocene
carbonates overlain by 50 m of deeper water
carbonates dated as Oligocene (E. Finch 1993 pers.
comm.). This block bounding fault and/or other
similar trending faults in the vicinity or under the
Camba Formation are inferred to be the northern
faulted margin of the Bantimala Block, which shed
material forming the Bantimala redeposited facies.
Palaeocurrent data
The Bantimala redeposited facies thin and fine
northwards away from the Bantimala Block (Fig.
6). Rare clast imbrication and scour structures (cf.
Hubert et al. 1977; Hiscott & James 1985), indicate
that the transport direction of the redeposited facies
had strong northerly, northnorthwesterly and east-
erly trends (Fig. 3). Although palaeocurrent data
are scarce, the southern area shows a dominant
transport direction towards the east (Fig. 3). In the
northern area finer redeposited facies tended to be
north-directed and are inferred to have been derived
from the Bantimala Block which lay to the south.
Coarser redeposited units in the north tend to be
easterly directed and are thought to have been
derived from the northern margin of the Ban'u
Block, an inference supported by clast provenance
data (Fig. 3).
Evidence for Tertiary exposure
Carbonate lithologies are poorly exposed along the
margins of the basement blocks close to the north-
bounding faults of the Bantimala and Barru blocks.
Limestone clasts reworked from these footwall
blocks, however reveal evidence of karstification
and subaerial exposure. Limestone clasts derived
from the Barru Block during the late Eocene
(NP17-19, E. Finch 1993 pers. comm.) contain
features such as irregular dissolution hollows
lined by blocky cement and then infilled by fine
carbonate silts and micritic glaebules (Fig. 10a).
The margins of clasts cut across all of these early
diagenetic features, suggesting subaerial exposure
of the footwall area of the Barru Block. Limestone
clasts in the Bantimala redeposited facies do not
show corresponding evidence for sub-aerial
exposure of the Bantimala Block until the latest
Oligocene.
Regional tilting
A number of factors suggest that the whole of the
Barru area was tilted towards the east during the
deposition of the Tonasa Limestone Formation. A
thinned carbonate sequence, the angular uncon-
formity with the overlying Camba Formation and
absence of the Balangbaru Formation in the
western part of the Bantimala footwall block all

Fig. 10. Photomicrographs showing features of the redeposited facies. Scale bar is 1 mm across. (a) Shallow-water
carbonate clast showing evidence for karstification. An irregular dissolution hollow has been coated in a blocky
cement, subsequently fine sediment has infilled the void and there is a final phase of sparry calcite cement at the top
of the cavity. (b) Quartz mica schist derived from the pre-late Cretaceous basement complex adjacent to an Eocene
large benthic Discocyclina foraminifera. Sample from a graded bioclastic packstone bed from the Bantimala
redeposited facies.

suggest that erosion was more effective to the
west. Only the Barru Block underwent sub-aerial
exposure from the late Eocene. From the late
Eocene onwards, west of the Rala anticline, re-
deposited facies are considerably reduced in thick-
ness (< 45 m) or absent compared with those to the
east (up to 635 m). This indicates much greater
accommodation space in the east during deposition
of the Bantimala redeposited facies. In certain
areas, palaeoflow towards the east and the overlap
of redeposited facies from the two source areas
in the northern palaeocurrent grouping imply a
surface sloping towards the east. If this inferred
eastward palaeotilt existed, slopes would have been
very low since the two northern palaeocurrent
groupings indicate that palaeoflow was mainly
north directed.
Morphology of the faulted margin
Two configurations of the faulted northern area are
feasible from the late Eocene onwards. The first is
domino style faulting, with the Barru Block being
downthrown relative to the Bantimala Block. The
second scenario is an east-dipping relay ramp
between two major parallel faults; the north-
bounding faults to the Barru and Bantimala blocks.
Horizontalscale
0 5 10
km ' ..-/""
Vedical exaggeration /
~"/"/
: "'" Block
', , ~", •", ," ~.' .......::::::::::::::::::::::::::::::::::::::::::::::.
--
Key ~
[ ] Basinal marl
• .,. ,,.
~_~ Redeposited facies
Shallow water platform carbonate
Sandstones, shale and coal
~ . " Shales, sandstones and conglomerates
I'.1 Metamorphic, volcanic and sedimentary lithologies
~ Areal extent of the redeposited facies
Overlap of redeposited facies
Fig. 11. Block diagram showing the preferred reconstruction of the northern margin of the Tonasa Carbonate
Platform in the Barru area during the late Eocene. Two tectonically active, major NW-SE trending normal faults are
separated by an east-dipping relay ramp. The area is thought to have been tilted gently towards the east.
TERTIARY EVOLUTIONOF S SULAWESI 379
An east-dipping relay ramp between two major
parallel faults is the preferred configuration of the
northern area for the reasons outlined below and
illustrated in Fig. 11.
(a) The abundance of low energy marl facies
east of the north-bounding fault to the Barru Block
suggests that the Barru Block was not undergoing
erosion and did not have topographic relief east of
its present day outcrop. The general lack of coarse
redeposited facies north of the easternmost outcrop
of the Barru Block suggests the north-bounding
fault to the Barru Block did not continue south-
eastwards. The Camba Formation is not displaced
east of the Barru Block, which also suggests the
block bounding fault dies out towards the SE.
(b) Much of the palaeocurrent data from re-
deposited facies in the two northern palaeocurrent
groupings is north-directed. This indicates that
sedimentation rates were high enough to overcome
the dip slope of the hanging wall block. Lower
angles of dip on the hanging wall block would be
expected in the relay ramp situation. In the domino
fault model axial flow between the two major faults
towards the east, the direction of regional tilt,
would be more likely. The Bantimala redeposited
facies were transported slightly further north than
the northern margin of the Barru Block. This
/
.
Barru
i ;
Block ~ ,
~ Z :':':"7 - - Sea,eve,
~'-::-/~,. Fault and direction
i:::'7 ~ of movement
. ~ . . ~ Karstification of subaerially
~","" exposed limestone
~ Transpod directioof
redeposited facies
n
380 M.E.J. WILSON • D. W. J. BOSENCE
situation would be most unlikely in a domino fault
block model.
(c) With domino style faulting and tilting, foot-
wall exposure of both major faults might be
expected. In the relay ramp situation the two main
faults border different parts of the same large block.
With tilting of a large single block, the northern
Ban'u Block area would have been more likely to
undergo emergence compared with the more
southerly area of the fault bounding the Bantimala
Block. Earlier subaerial exposure of the Barru
Basement area did occur, suggesting a relay ramp
model is more appropriate (Fig. 11).
The lack of a SE continuation of the north-
bounding fault to the Barru Block, redeposited
facies distribution, and patterns and timing of sub-
aerial exposure of footwall areas all suggest that a
relay ramp is the more appropriate configuration
for the platform margin.
History of the platform margin
Changes in the bed-thickness, clast-type and con-
tent of the redeposited facies through time can be
used to constrain the history of the platform margin
(cf. Everts 1991; Reijmer et al. 1991). Figure 12
is a plot of the bed thicknesses of the redeposited
facies against the percentage of clast types and
ages, based on point counting thin sections. The
plots on the right hand side of Fig. 12 highlight the
percentage of clasts older than the background marl
sedimentation and the percentage of clasts which
are lithic fragments.
Ages for the background marl sedimentation are
derived from nannofossil (E. Finch pers. comm.
1993) and planktonic foraminifera identification
(T. Wonders pers comm. 1992; E Banner pers.
comm. 1994). Provenance studies and a knowledge
of the stratigraphic sequence in South Sulawesi can
be used to infer an age for non-carbonate clasts. For
example, metamorphic and quartzose sandstone
lithologies could only have been derived from
pre-upper Cretaceous and upper Cretaceous-
Eocene non-carbonate formations respectively
(Fig. 10b). Although it may be possible to date
limestone clasts containing large benthic
foraminifera (cf. Adams 1970), many of the
carbonate grains cannot be dated. The presence of
Discocyclina, Pellatispira or Biplanispira indicates
Eocene ages, whilst Lepidocyclina indicates an
Oligocene or younger age. Unbroken delicate
planktonic foraminifera or rip-up marl clasts are
considered to be contemporaneous with the back-
ground basinal marl sedimentation. Clasts of plank-
tonic foraminifera bioclastic packstones occur very
rarely.
Redeposited carbonate units are thickest during
three periods of time: late Eocene to early
Oligocene; latest early-early late Oligocene and
early to middle Miocene (Fig. 12). When the re-
deposited facies are thickest, thick-bedded, clast-
supported breccias are the dominant facies and
marls occur rarely. During these times between 40-
90% of grains or clasts are lithic fragments and
correspondingly a significant proportion of the
clasts are older than the background marl sedimen-
tation. In comparison, when the redeposited facies
beds are thinner, marls are more abundant, there
is a decrease in the amount of lithic clasts, and the
majority of clasts are contemporaneous with the
background sedimentation (Fig. 12).
It has been inferred above that the main cause
of the redeposited facies was tectonic activity on
a faulted carbonate platform margin. Derivation
of lithic clasts and lithologies older than contem-
poraneous marl sedimentation would have occurred
during periods of faulting. At these times the foot-
wall block would have had relative relief leading to
'exposure' and erosion of the block margins.
Earlier lithified material from these elevated areas
was prone to reworking into hangingwall depo-
centre areas. Since karstification of the Bantimala
footwall area is only inferred from the late
Oligocene, it is important to note that for much of
the time erosion probably occurred in a sub-marine
rather than a sub-aerial realm. Submarine down-
slope redeposition of shallow-water carbonate
material is common along the margins of many
modern carbonate banks due to slope instabilities
(James & Ginsburg 1979; Mullins 1983). Three
main phases of tectonic activity on the platform
margin are inferred. These occurred during the late
Eocene to early Oligocene, latest early/early late
Oligocene and the early to middle Miocene (Fig.
12). These three phases can now be compared with
more regional tectonic events.
Regional comparison
The Tertiary sequence in the western part of South
Sulawesi is similar to sequences in neighbouring
Kalimantan and the east Java Sea. It has been
postulated that the whole region was part of a wide-
spread basin; the formation of which commenced
in the early-middle Eocene (Daly et al. 1987, 1991;
Pieters et al. 1987; Van de Weerd & Armin 1992).
The Eocene was a period of widespread plate
reorganization in SE Asia during which India
collided with Eurasia (Dewey 1980; Patriat &
Achache 1984; Daly et al. 1987, 1991). The bend
in the Hawaii-Emperor sea mount chain also
developed at this time, indicating a possible change
in the motion of the Pacific plate (Daly et al. 1987,
1991). The Australian-Eurasian spreading centre
shifted to its present southeast Indian location,
during the Eocene (Packham 1990). Although this
 TERTIARY EVOLUTION OF S SULAWESI 381

Thickness of redeposited
~I~ "~ ~ carbonate units (m)
1, 2, 3, 4, 5, 1, 0 1,5 2 0, 2 5, % Clast age Sample % Clast type
0 % 100 No. 0 % 100
~f--'~, NN2-NN6 early/middle J106 t ' ~ , , ' , , ' , .'_- ,'., ,'., , ' , J !
1100 _ I1--\ Miocene \ 'Envelope' 'Envelope' i
" showing trend ,. showing trend
\ of increasing \ of increased
\.~-___ percentage of percentage of i
\ clasts older li~ic fragments.i
\ than
- background
•, marl
\ sedimentation.
!:_-_-_-_:-~
• ..,
=
1000
379 ~,="=='==-=="=="=="=="=~-. %-, II
PN6-N7 /" 1
middle
Miocene " 1
J69
i ..
early Miocene
~~2,v,¢>.~ x.~.~ =¢ J52

900 ~ J30
~ J28
J22
J15
~ J12
J3
J2
D97

D85
NP23 latest ~ /
early/early late D79
800 Oligocene i

i
! - -
///
~.~ NP23 latest early/early late D70 L-,'::..:..:..-..-..-...,
Oligocene
~ =lcwI 062 I~'d. , " , ; . - . - . - .'.7
~'Carf:16n-at-e ~ r a c i e s - t y l ~ s . . . . .
700 / ~= BP - Bioclastic packstone
/ ~ ~ MSB- Marl supported breccia•
CSB - Clast supported breccm I
\
\
GBP - Graded b oc ast c packstone
i~ PFW - Planktonic foraminifera
wackestone
-_A_rk_os_ic packst_one_ J
i' i
D60 L-~-_, . ; , . _ . . _ . ,_..-...'...-...1 i
D52 I;..==-_J~J~.J"--'3i-,,-J~: ; .- ; : ; i I
" ! _ Early Oligocene D49 i . . . . . . . . . . .:I ....... i

i" ~ . . . .

i
D42 t--,L--=L'..;'--';'--':'-'.=!;.
.
i
:, ;, I

D28 .k-=_;=_;=_;=_;=_;,j,'_-.:-._, .-...-=!

D25 I . . . . ~ ; ~ " ,." ~ 11
~ - - ' ~ / ' ~ E n v e l o p e of bed D23 ".=: - " -'.-'_" -~:. r-.'.-r.;~C •
- ~' thickness of the J / \.\
600 _i _~.~_[~
---- _carbonate
redepositedfacies
D2 'h=:-=:-=;-=;-=:-=;-=-~_ :. , :. , J i

I ~ ~ NP18-NP20 late Eocene DR23 ~ - - . - " = - - . , ' : ' ; . • ".. ". • ". • ". • i
R

500 =...___)1=17 late Eocene J an36 k-.,-_ . . . . . :_.:':.,"_'U-.-.,i

F ~,e-,re?; . . . . . . . . . . . . . . . . . . ',',-&e-,; ~ . . . . . .
I mini Pre late Cretaceous [ ] Miocene carbonate ,, F~ Lithic clas '~,~
i [r~
t ~ ?Palaeocene/Eocene rI.--I
- I Undated carbonate ii ~ .UlOClaSIs
... -
470 ~ ") D 0 - - sandstone ma Contemporaneous iI ~ . fraomented
nn~E] 1 2 3 4 5 10152025 = [ ] Eocene carbonate m basinal sediments H r ~ _. ° . . . .
~ o ~ Thickness of redeposited I [ ] O h"goce n e carbo n ate [ ] V olcanic casts
I == I 1151oclasIs WhOle
i jr.
Carbonate facies carbonate units (m)

Fig. 12. Point-counting data plotted against the bed thicknesses for the redeposited facies in the Rala measured
section. Note the increased percentage of clasts derived from earlier formations or reworked from older shallow
marine carbonates of the Tonasa Limestone Formation when the redeposited beds are thickest.
382 M.E.J. WILSON & D. W. J. BOSENCE

plate reorganization is thought to have resulted in NW TT-2 TT-1 SE

Eocene basin formation, in detail the mechanics of
this basin formation are as yet unclear. Initiation 1~ -~:--~ I ~~ z , ~ "-~r-__-S,-_z' % t
and formation of different parts of the basin have
variously been related to extensional back-arc
formation (Hamilton 1979; Daly et al. 1987, 1991)
and a foreland-fore-arc flexural origin (Williams km
et al. 1988; Hutchison 1989). ~ 4
m TT-2 TT-1 m
Two major fault trends are apparent from seismic
0 mid-Miocene to Recent
data in the Makassar Strait offshore Sulawesi (Fig. 192
13). Approximately NE-SW trending faults have
been related to back-arc extension due to roll-back 62 early to mid- i KEY
of a plate subducting eastwards under eastern 821 .~ Mi°cene i [] Deepmarine
" , shales+ marls
Sulawesi (Daly et al. 1987, 1991; Letouzey et al.
1238. . . . . . .
1990). Although Eocene-Oligocene volcanics in :: ,,mestonetaces
Redeposited

eastern South Sulawesi are thought to delimit the 1 5 ~ ~ Shallowmarine

plate margin, detailed analysis has not been under-
taken and it is not clear if these volcanics are ,\\ ~'~]
rq~,~, Oligocene
~ elastics(Marginal).
marine
~ "~, ~ 2182 to early ~ Non marine
related to subduction (Letouzey et al. 1990; Van de \ ~i~~ Miocene ~ clastics
Weerd & Armin 1992). In the Kalosi area upper ~\ \\. ~] ~ Basement
Miocene-Pliocene igneous rocks have been related ', ,\ ~ ...................
to subduction of continental crust (Bergman et al. ~ 2892. . . . . . . . . .
~. ~ 3055 Eocene
1996). Another possibility is that the Makassar ~~ 3238
Strait extension may represent a failed arm- ~J,~i no recovery
extremity of sea floor spreading in the Celebes Sea 3560 . . . . . . . . . .
]~P'--7] Cretaceous-inferred
to the northeast (Hutchison 1988; Moss 1994). The Fig. 14. Seismicsection and borehole data from the
second group of faults trends NW-SE and includes inverted Taku Talu fault, Makassar Straits (after
the Adang and Sangkulurang faults (Fig. 13). Situmorang 1987). See Fig. 13 for location of
boreholes.

Offshore analogue o f the northern

faulted margin
The late Eocene to middle Miocene redeposited
carbonate facies in the Barru area are inferred to
have been derived from the faulted northern margin
of the Tonasa Carbonate Platform. An analogous
example of syn-tectonic carbonate sedimentation
is revealed from seismic and borehole data in the
Makassar Straits, from the vicinity of the NE-SW
trending Taku Talu fault (Fig. 14). Eocene to
middle Miocene carbonate breccias containing
clasts from the underlying basement were
deposited as a thickened hanging wall sequence,
whilst shallow-water carbonate sedimentation
continued on the footwall block of the Taku Talu
fault (Fig. 14). Thus the Taku Talu fault was a
major active Eocene to early Miocene normal fault
which underwent later inversion during the middle
Miocene (Situmorang 1987).

~ Outcropping Eocene to Holocene.TT-2 Boreholelocation(seeFig. 14) [

sedimentary basin fill A d a n g fault
pre-Cretaceous continentalcrust ..-,.--
j Subduct-iotick
nFaUlt downthrowside
markSzone
The Adang fault is a major NW-SE trending
structure separating the North and South Makassar
Fig. 13. Map to show the location of Tertiary basins in
Kalimantan and western Sulawesi. Modified after Basins (Fig. 13). This lineament is thought to
Van de Weerd & Armin (1992). extend westwards onshore forming the southern
 TERTIARY EVOLUTION OF S SULAWESI
boundary of the Kuta Basin and has been linked
across Borneo with the Lupar Line forming a
'trans-Borneo shear' (Woods 1985; Kusuma &
Darin 1989; Wain & Berod 1989; Bransden &
Matthews 1992; Van de Weerd & Armin 1992).
Both strike-slip (Woods 1985; Hutchison 1989;
Biantoro et al. 1992) and normal displacements
(Biantoro et al. 1992; Rangin et al. 1990) on the
Adang fault have been inferred. It has been
suggested that the Adang fault was an important
Eocene to early Miocene normal fault down-
throwing to the north, which subsequently under-
went strike-slip reactivation during the middle-late
Miocene (Kusuma & Darin 1989; Rangin et al.
1990; Biantoro et al. 1992). During the late Eocene
to Miocene the Adang fault had a strong influence
on sedimentation patterns. The Paternoster
Platform and Barito Basin were the sites of
shallow-water carbonate development, whilst
deeper sedimentation occurred in the Kuta Basin
to the north (Van de Weerd et al. 1987; Kusuma
& Darin 1989; Wain & Berod 1989).
Seismic data across the Makassar Straits
(D. Coffield pers. comm. 1995) suggest there was
no fault linkage between the NW-SE trending
Adang Fault and the main north-bounding fault to
the Tonasa Carbonate Platform. It is suggested that
NW-SE trending structures influenced sedimen-
tation patterns in Sulawesi and Kalimantan. NNE-
SSE trending faults such as the Walanae Fault also
had an effect on Tertiary sedimentation patterns in
western Sulawesi (Van Leeuwen 1981; D. Coffield
pers. comm. 1995). Seismic data N-S across the
Sengkang Basin reveal normal block-faulting, with
blocks being downthrown to the centre of the basin
(A. Ngakan pers. comm. 1994). Indeed, the north-
bounding faults to the Tonasa Carbonate Platform
and faults with a similar trend to the east may mark
the southern boundary of the NW-SE trending
depression, which runs through the Sengkang Basin
and appears structurally to separate South Sulawesi
from the rest of the western arc of Sulawesi
(Fig. 1).
Timing
Redeposited carbonate facies in the Barru area
suggest three phases of tectonic activity in the
area during the late Eocene-early Oligocene, early-
late Oligocene and early-middle Miocene. Other
evidence for possible tectonic activity in the region
is documented below and possible causes are
discussed (Fig. 15).
E a r l y - m i d Eocene. A basal angular unconformity
to many initially transgressive Tertiary sequences
marks widespread basin initiation (Cater 1981;
Pieters et al. 1987; Van de Weerd et al. 1987;
383
Kusuma & Darin 1989; Wain & Berod 1989;
Letouzey et al. 1990; Bransden & Matthews 1992;
Hutchison 1992; Van de Weerd & Armin 1992).
The angular unconformity between the Malawa
Formation and underlying Balangbaru Formation is
one such contact. Basin initiation possibly occurred
as early as the Palaeocene in some localities
(Bishop 1980; Cater 1981; Wain & Berod 1989).
Carbonate production, including the deposition of
the Tonasa Limestone Formation, had begun in
many areas by the middle to late Eocene. Active
faulting and graben formation has been inferred
from seismic (van de Weerd et al. 1987; Bransden
& Matthews 1992) and outcrop data (Tyrrel et al.
1986; Kusuma & Darin 1989). Basin formation is
thought to be a response to widespread Eocene
plate reorganisation.
Late E o c e n e - e a r l y Oligocene. This period corre-
sponds with the first phase of coarse redeposited
facies in the Barru area, indicating the initiation
(or reactivation) of faulting on the margin of the
Tonasa Carbonate Platform. The geometry and
timing on major faults from regional seismic data
suggest that localized extension was underway by
the early Eocene (P9/P10), and rifting was wide-
spread by the late Eocene (Van de Weerd et al.
1987; Letouzey et al. 1990; Bransden & Matthews
1992). The major variation in preserved thick-
nesses across faults, and hence syn-tectonic
deposition is observed during the late Eocene and
early Oligocene (Bishop 1980; Bransden &
Matthews 1992). During the late Eocene a signifi-
cant tectonic event caused widespread high angle
faulting and subsequent erosional truncation
affecting a number of areas (Bransden & Matthews
1992). This period is thought to represent the main
extensional phase in the region (Letouzey et al.
1990; Bransden & Matthews 1992). The middle-
late Eocene is also the age of the oldest oceanic
crust in the Celebes Sea (Weissel 1980).
Middle Oligocene. A mid-Oligocene unconformity
has been reported from many offshore (Bransden &
Matthews 1992) and onshore areas (Van Leeuwen
1981; Kusuma & Darin 1989; Van de Weerd &
Armin 1992). This unconformity, the major shift
in facies belts, localized channelling and thickness
variations have been related both to the mid
Oligocene eustatic lowstand (Bransden &
Matthews 1992) and tectonic activity (Cater 1981;
Van de Weerd et al. 1987; Wain & Berod 1989;
Sailer et al. 1992; Van de Weerd & Armin 1992).
Sailer et al. (1992) reported exposure and subse-
quent deepening prior to the middle Oligocene
(29.5-30 Ma) sea-level fall of Haq et al. (1987). In
Sulawesi the second phase of coarse redeposited
facies in the Barru area, an unconformity in the
384 M.E.J. WILSON (~ D. W. J. BOSENCE

W.Kalimantan Upper Kutai SE.Kalimantan E.Java Sea / Southwest

Age Basins Basin Basins Kangean Sulawesi Events

:'. •. •. -7"~- T r 7:

L '" : :" : : : : : : : : : : [ : : : : : : : : : : : : : :[vv",,"v"vvv"vvv'

• " " [ : : : : : : : ' : ' : : : : 1 % % % % % % " , , '

• -.-.-.-.-.-.- ......... Collision of Australia
.............. r. ". ". ". ". ". ".-,~, ,,. . . . . . . with Indonesia and
M .. • • ~--;'~.~.'.'.' :"... "=-.. " - ~ microcontinent
_ .~.. ;~ . ; . . - - . _ ; T . . . - ' . ' _ . " " " • " • " .- • _- _._ _,_,- • • ---_._ -.. fragments impinge
- , - - . ' ~ l - T r r . - - . - . l T r r " • " * " • " • " • " • •. .~?.= " " " m m m .... = " °.
•.'.'.'.'.'.'.'.'.'.'...'." . . . . . . . • • . • • •. . • . • •. • • • • • • • , , • " • on east Sulawesi

. . . . . . . . ~o , ; i i i ."'.',',',"',',",',','.',',','"
• .-,-o'.-.-,-.'.','.'.','.'." . . . . . . . ~:.~. = = = m i m " " "
I.............................[-,-.- .... -....... -.-];,:,~:_. • • • " ~ t ~ , ~ . ~, • • l l : ~ . • •
E I'.'.'.'.'.'.'.'.'.'.'.'.'.'.'$'-'-'.'.'.'.'.'.'.'-'.'.'.'.I~Ir~ ~ , " " ~ " - ~ . " " " i i , I:~L.;~rL " "
I'.'.'.'.'.'.'.'.'.'-'-'.'-'.T.'.'.'.'.'.'.'.'.'.'.'.'-'-,, , , ~,%%'~" " " l I ' , 1;"~.~" " I l ' I'~r~.~_" " "
r.'.'.'.'.'.'.'.'.'.'.'.'.'.'L'.'.'.'.'.'.'.'.'.'.'.'.'.'-, m m =~= = " r m m m m.,~." " " m m m ~:':~" " =
.:.:.:.:.:.:.:.:.:.:.:.:.:.:.r ::.- - , . :..-.-.,--~.:.:.:
,. . . . . . . . . . . . . . . . . . . . . . . . . I , , ~ .~ ". . . . ".
÷: " - . ., . . , ' "~,~" -_,. . . ." ' ' ~...........
'~'~.~.." " "

o
,
. i • ".'.-.'.-.-.'.:.:.:.:.:.:.:,
"'"'""':'"":':"':'"""
::::::::::::::::::::::::::: : : : : : : : : : " :-:
.:.:.:.:.:....................
I.:.:.:.:.:.::.:.:.:.:.:.:.iil!.:.:
r- .-- . ' . '.. -.. -.. '.. _. ' ..' _. . '.. ' .. ' ..' . ' . ' l .
1',
": . :- • :- . ": . :- . :" • :~. ~- ~ ~, .~~
•. •. •. •.
:.:.:.:.
• : • : • .:. . . . . .
-- - I. " ' - ' ":
-.-.-
•
.:.
•
. •.
. : .

. N•.
.j
-
• ...~
~. -. ', .-
.

•
.'. .

•
.

,
,,,,,
, ,
i -. . . . . . .- ' - -
,~..,
~,%
• .
• Sea floor spreading
' '• . ~ '". : • : • ~• - • • • •" t " r' - r--.F' .~ : . L " ' ~ i ~ - ~" " ~'.~- -" . " ~ ' ~• "• "-'='~'=-
-" "--" -"- - _-" , ~ ' , ~ , . , ~ ' - - - " . 'I. i n t h e S.China Sea

,-..,~.. .... .~,.:,~,_. • • ., .... ~.~. • ~ , , ,,~, , .... .~.~ Emplacement of the
I l
EI:::::...Y'~',~ ', , , ~"'~""~
~ . ". . ". " l l l
, . . . .
~,"..:t, "
.~.~,, . .
"
,
m
,
m
,
m .~,..w;--- " " " I
,~. • •
l I
, , , .::~
l V.. " "
• • EastSulawesi
~'.'.'.---.'~--':~, , ~.~.- " " , , , ~'~" - , , ,-.:.~" " ", . . . . :&~" " Ophiolite
~--~'~ ',1,1 I I I ~ ' ~ : : ' ~ " . ~.'#'~, | I I I l I ' ~'.~.~.~%~'~ ~ ' l , ' ~=--.--I : , ~ Y I ~l : ~ . - - .
- : -

~.z.~--'T'T'-.'--..']:~T'T'~'-.'~i~:'.'.'.','.'-[.'.'.'-:-: : : I m i m , , ,
I'.'.'.'.'.'-"".".'-',,.;r'.. " • "1. " . " • " • " • " • " • " 4" • " • " • " • " • " • " • [ • " • " • " . . . . n m n n n n
i.................~.-..~.. •. •. •. •. • ~..4~.. •. •. •. •. r. •. -. •. •. • • • • • l m n n m m l

M " .................,~.,v~.. . ..... > ......................... Collisionof Indiawith

========================================= :::: ::::::: C.an0e,n
om -." " . '•. ,' .. .' ...". "." " . '-..'... '-.. . . . . . '... ',.. ', .. ,' ...' ...'...'.. • ~ ; . , = ."- : -•, - -"- ~•, ~" • "• •. "• " " • •. •. . , , plate motion of
. . . . . . . . . . . . . . • • Pacific plate.
" - ~ ~ ~ " ~ ~ J ~ Development of
E spreading centre in
Indian Ocean

J ~ (Marginal) marine ~ Non marine

KEY [ 7 t Deep marine
shales + roads r ~ Volcaniclastics
~
~
Redeposited
carbonate facies [~ Shallow marine limestones [,_'1 clastms I',','i clastics

Fig. 15. Tertiary stratigraphic correlation chart, showing similarities between the sequences in Kalimantan, East Java
Sea and Sulawesi. Data taken from Bishop (1980), Cater (1981), Van de Weerd et al. (1987), Kusuma & Darin
(1989), Wain & Berod (1989), Bransden & Matthews (1992) and Van de Weerd & Armin (1992).

Biru area (Van Leeuwen 1981) and minor localized up to the present-day, is widely attributed to the
evidence for exposure on the Tonasa Carbonate collision of microcontinental fragments onto
Platform (personal observation) during the mid eastern Sulawesi during the middle to late Miocene
Oligocene, all suggest a tectonic cause. During the (Daly et al. 1987, 1991; Van de Weerd & Armin
Oligocene sea floor spreading began in the South 1992). The middle to late Miocene is also the time
China Sea (Holloway 1982; Ru & Pigott 1986; when a volcanic arc developed in western Sulawesi
Daly et al. 1991) and the east Sulawesi ophiolite (Sukamto 1975; Yuwono et al. 1985; Coffield et al.
was emplaced (Parkinson 1991). 1993; Bergman et al. 1996).
Similar to the major bounding faults of the north
Early to middle Miocene. In the Barru area west- margin of the Tonasa Carbonate Platform, many
verging folding affects both the Tonasa Limestone of the faults mentioned above had major normal
Formation and the overlying middle-late Miocene displacements. Inferring an additional strike-slip
volcaniclastics. The latest Oligocene to middle motion would depend on along-strike linkage of
Miocene phase of the Bantimala Redeposited faults, a factor which is difficult to ascertain from
Facies therefore records a period of faulting and the available seismic and outcrop information. A
uplift prior to folding (and possible inversion). transtensional Palaeogene history for the region has
Based on seismic evidence from the east Java Sea been suggested on the basis of the length of faults,
and Makassar Strait, contemporaneous faulting and apparent reversal and observed variability of fault
syn-tectonic graben fills have been inferred until trends (Bishop 1980; Bransden & Matthews 1992).
the early Miocene (Bishop 1980; Cater 1981,
Letouzey et al. 1990). Inversion of many of the
earlier faults with normal displacement occurred
C o n c l u s i o n s
slightly later during the early to middle Miocene
(Fig. 14; Letouzey et al. 1990; Bransden & Late Eocene to middle Miocene redeposited facies
Matthews 1992). This compressive regime, active of the Tonasa Limestone Formation in the Barn]
 TERTIARY EVOLUTION OF S SULAWESI
area provide a unique example of the use of
carbonate sedimentology in comparing local and
regional tectonic events. In the shallow-water
carbonates of the Tonasa Carbonate Platform to the
south, tectonic effects are impossible to distinguish
from eustatic effects. However, clast composition
and facies types in the redeposited carbonates
reveal that tectonic and not eustatic changes were
the main controls on sedimentation. In fact, re-
deposited carbonates in the Barru area prove to
be a remarkable natural seismograph, recording
regional tectonic changes and a number of local and
regional conclusions can be inferred from them.
(1) The spatial distribution, b a s e m e n t clast
content and p r o x i m a l - d i s t a l trends within re-
deposited facies suggest derivation from two main
source areas. These areas were from the northern
margins of the Bantimala and Barru Blocks.
(2) The textural immaturity and provenance of
clasts indicate that the redeposited facies were
derived from the faulted margins of a carbonate
platform. Redeposited facies were derived from at
least two faulted line-sources and were deposited as
sheet flows forming carbonate aprons at the base
of the slope.
(3) An east-dipping relay ramp between two
main N W - S E trending faults is the preferred con-
figuration of the northern margin of the Tonasa
Carbonate Platform.
(4) The redeposited facies indicate three phases
of tectonic activity. These occurred during the late
References
ADAMS, C. G. 1970. Reconsideration of the East Indian
Letter Classification of the Tertiary, British Museum
of Natural History Geology Bulletin, 19, 87-137.
BERGMAN, S. C., COFFIELD, D. Q., TALBOT, J. P. &
GARRARD, R. J. 1996. Tertiary tectonic and
magmatic evolution of western Sulawesi and the
Makassar Strait, Indonesia: evidence for a Miocene
continentcontinent collision. This volume.
BERRY,R. E & GRADY,A. E. 1987. Mesoscopic structures
produced by Plio-Pleistocene wrench faulting in
South Sulawesi, Indonesia. Journal of Structural
Geology, 9, 563-571.
BIANTORO, E., MUR1TNO, B. E & MAMUAYA, J. M. B.
1992. Inversion faults as the major structural control
in the northern part of the Kutai Basin, east
Kalimantan. In: Proceedings of the Indonesian
Petroleum Association 21st Annual Convention.
45-89.
BISHOP, W. E 1980. Structure, stratigraphy, and hydro-
carbons offshore southern Kalimantan, Indonesia.
AAPG Bulletin, 64, 37-58.
BOSELLINI, A. 1989. Dynamics of Tethyan carbonate
platforms. In: CREVELLO,P. D., WILSON,J. L., SARG,
E & READ, J. E (eds) Controls on Carbonate
Platform and Basin Development. Society of
Economic Paleontologists and Mineralogists,
Special Publication, 44, 3-13.
385
Eocene to early Oligocene, the middle Oligocene
and the early to middle Miocene. This is consistent
with other similar sequences in Kalimantan and
the East Java Sea and with inferred regional plate
tectonic changes.
(5) The northern faulted margin of the Tonasa
Carbonate Platform has a similar trend to N W - S E
orientated structures in the Makassar Straits and
Kalimantan. These structures are inferred to have
i n f l u e n c e d s e d i m e n t a t i o n patterns during the
Eocene to middle Miocene.
The senior author gratefully acknowledges BP
Exploration, UK for their generous financial support
during the course of her PhD study, of which this work
forms a part. The SE Asia Research Group, Royal
Holloway, University of London, especially Dr Tony
Barber and Diane Cameron, are thanked for their
administrative and technical support. In Indonesia,
Alexander Limbong, the senior author's counterpart from
GRDC, Bandung, during three gruelling 'non-stop' field
seasons deserves special thanks. GRDC, Bandung,
Kanwil, South Sulawesi, particularly Darwis Falah and
family, BP offices in Jakarta and Ujung Pandang and LIPI
all provided technical and practical support. Dr Ted Finch
and Prof. Fred Banner at University College London,
and Dr Toine Wonders, Consultant, UK, are thanked for
their excellent biostratigraphic work. The constructive
comments from referees Dr Dana Coffield and Dr Neil
Harbury and those by Dr Tony Barber, Rob Bond, Nigel
Deeks and Dr Steve Moss towards improving this paper
were much appreciated. Keith Denyer is thanked for
producing the photographic plates.
BOUMA, A. H. 1962. Sedimentology of Some Flysch
Deposits: A Graphic Approach to Facies
Interpretation. Elsevier, Amsterdam.
BRAITHWAITE, C. J. R. & HEATH, R. A. 1992. Deposition
and diagenesis of debris flows in Upper Ordovician
limestones, Hadeland, Norway. Sedimentology, 39,
753-767.
BRANSDEN, P. J. E. & MATTHEWS, S. J. 1992. Structural
and stratigraphic evolution of the east Java Sea,
Indonesia. In: Proceedings of the Indonesian
Petroleum Association 21st Annual Convention.
417-453.
BURCHELL, M. T., STEFANI, M. t~ MASETTI, D. 1990.
Cyclic sedimentation in the southern Alpine
Rhaetic: the importance of climate and eustasy
in controlling platform-basin interactions.
Sedimentology, 37, 795-815.
BURCHETTE, Z. P. 1988. Tectonic control on carbonate
platform facies distribution and sequence develop-
ment: Miocene, Gulf of Suez. Sedimentary Geology,
59, 179-204.
BUSTILLO, M. A. & RUIZ-ORTIZ,P. A. 1987. Chert occur-
rences in carbonate turbidites: examples from the
Upper Jurassic of the Betic Mountains (southern
Spain). Sedimentology, 34, 611-621.
CATER, M. C. 1981. Stratigraphy of the offshore area
south of Kalimantan, Indonesia. In: Proceedings of
386 M . E . J . WILSON ,~¢ D. W. J. BOSENCE
the Indonesian Petroleum Association lOth Annual
Convention. 269-284•
COFFIELD, D. Q., BERGMAN, S. C., GARRARD, R. A.,
GURITNO, N., ROBINSON,N. M. & TALBOT,J. 1993.
Tectonic and stratigraphic evolution of the Kalosi
PSC area and associated development of a Tertiary
petroleum system, South Sulawesi, Indonesia. In:
Proceedings of the Indonesian Petroleum
Association 22nd Annual Convention. 679-706.
CONIGLIO, M. 1987. Biogenic chert in the Cow Head
Group (Cambro-Ordovician), western Newfound-
land. Sedimentology, 34, 813-823.
COOK, H. E. 1979. Ancient continental slope sequences
and their value in understanding modem slope
development. In: DOYLE, L. J. & PILKEY, O. H.
(eds) Geology of Continental Slopes. Society of
Economic Paleontologists and Mineralogists,
Special Publication, 27, 287-305.
- -& MULLINS,n. T. 1983. Basin margin. In: SCHOLLE,
E A., BEBOUT, D. G. & MOORE, C. H. (eds)
Carbonate Depositional Environments. AAPG
Memoir, 33, 539-612.
• & TAYLOR,M. E. 1977. Comparison of continental
slope and shelf environments in the Upper
Cambrian and lowest Ordovician of Nevada. In:
COOK, H. E. & ENOS, P. (eds) Deep-water
Carbonate Environments. Society of Economic
Paleontologists and Mineralogists, Special
Publication, 25, 51-81
, MCDANIEL, P. N., MOUNTJOY,E. W. & PRAY,L. C.
1972. Allochthonous carbonate debris flows at
Devonian bank ('reef') margins, Alberta, Canada.
Bulletin of Canadian Petroleum Geology, 20,
439-497.
CREVELLO, P. D. & SCHLAGER,W. 1980. Carbonate debris
sheets and turbidites, Exuma Sound, Bahamas.
Journal of Sedimentary Petrology, 50, 1121-
1148.
CROTTY, K. J. & ENGELHARDT, D. W. 1993. Larger
foraminifera and palynomorphs of the upper
Malawa and lower Tonasa Formations, south-
western Sulawesi Island, Indonesia. In:
THANASUTHIPITAK, T. (ed.) Proceedings of the
International Symposium on Biostratigraphy of
Mainland Southeast Asia: Facies & Paleontology.
31 January-5 February, Chiang Mai, Thailand.
71-82.
DALY, M. C., HOOPER, B. G. D. & SMITH, D. G. 1987.
Tertiary plate tectonics and basin evolution in
Indonesia. Proceedings of the Indonesian Petroleum
Association, 16th Annual Convention. 399-427.
, COOPER, M. A., WILSON, I., SMITH, D. G. &
HOOPER, B. G. D. 1991. Cenozoic plate tectonics
and basin evolution in Indonesia. Marine and
Petroleum Geology, 8, 2-21.
DAVIDSON, J. W. 1991. The geology and prospectivity of
Buton island, S.E. Sulawesi, Indonesia. In'.
Proceedings of the Indonesian Petroleum
Association, 20th Annual Convention. 209-233.
DAVIES, G. R. 1977. Turbidites, debris sheets, and
truncation structures in upper Paleozoic deep-
water carbonates of the Sverdrup basin, Arctic
archipelago. In: COOK, H.E. & ENos, E (eds)
Deep-water Carbonate Environments. Society of
Economic Paleontologists and Mineralogists,
Special Publication, 25, 221-247.
DEWEY, J. E 1980. Episodicity, sequence and style at
convergent plate boundaries. In: STRANGEWAY,
D.W. (ed.) The Continental Crust and its Mineral
Deposits. Geological Association of Canada,
Special Paper, 20, 553-576.
EATON, S. & ROBERTSON, A. H. E 1993. The Miocene
Pakhna Formation, southern Cyprus and its relation-
ship to the Neogene tectonic evolution of the
Eastern Mediterranean. Sedimentary Geology, 86,
273-296.
EBERLI, G. P. 1987. Carbonate turbidite sequences
deposited in tiff-basins of the Jurassic Tethys Ocean
(eastern Alps, Switzerland). Sedimentology, 34,
363-388.
ELMI, S. 1990. Stages in the evolution of late Triassic and
Jurassic carbonate platforms: the western margin of
the Subalpine Basin (Ard~che, France). In: TUCKER,
M. E., WILSON, J. L., CREVELLO, P. D., SARG, J. R.
& READ, J. F. (eds) Carbonate Platforms: Facies,
Sequences and Evolution. International Association
of Sedimentologists, Special Publication, 9,
109-144.
EVERTS, A. J. W. 1991. Interpreting compositional varia-
tions of calciturbidites in relation to platform-
stratigraphy: an example from the Paleogene of
SE Spain. Sedimentary Geology, 71, 231-242.
FERNANDEZ-MENDIOLA, P. A. & GARCIA-MONDt~JAR, J.
1989. Evolution of a Mid-Cretaceous carbonate
platform, Gorbea (northern Spain). Sedimentary
Geology, 64, 111-126.
FORTUIN, A. R., DE SMET, M. E. M, HADIWASASTRA,S.,
VAN MARLE, L. J., TROELSTRA, S. R. &
TJOKROSAPOETRO, S. 1990. Late Cenozoic sedi-
mentary and tectonic history of south Buton,
Indonesia. Journal of Southeast Asian Earth
Sciences, 4, 107-124.
GARCIA-MONDI~JAR, J. 1990. The Aptian-Albian
carbonate episode of the Basque-Cantabrian Basin
(northern Spain): general characteristics, control
and evolution. In: TUCKER, M. E., WILSON, J. L.,
CREVELLO, P. D., SARG, J. R. & READ, J. E (eds)
Carbonate Platforms: Facies, sequences and
evolution. International Association of Sedimen-
tologists, Special Publication, 9, 257-290.
GARRARD,D., SILALAHI,D., SCHILLER,D. & MAHODIM,P.
1989. Sengkang basin, South Sulawesi. Indonesian
Petroleum Association post-convention field trip,
1989.
GARRARD,R. A., SUPANDJONO,J. B. & SURONO. 1988. The
geology of the Banggai-Sula microcontinent,
eastern Indonesia. In: Proceedings of the Indonesian
Petroleum Association, 17th Annual Convention.
23-52.
GAWTHORPE,R. L. 1986. Sedimentation during carbonate
ramp-to-slope evolution in a tectonically active
area: Bowland Basin (Dinantian), northern England.
Sedimentology, 33, 185-206.
GRAINGE, A. M. & DAVIES,K. G. 1983. Reef exploration
in the east Sengkang basin, Sulawesi, Indonesia.
In: Proceedings of the Indonesian Petroleum
Association, 12th Annual Convention. 207-
227.
 TERTIARY EVOLUTION OF S SULAWESI
HAMILTON, W. 1979. Tectonics of the Indonesian region.
United States Geological Survey Professional
Paper, 1078.
HAMPTON, M. A. 1972. The role of subaqueous debris
flow in generating turbidity currents. Journal of
Sedimentary Petrology, 42, 775-793.
HAQ, B. U., HARDENBOL, J. & VAIL, P. R. 1987.
Chronology of fluctuating sea levels since the
Triassic. Science, 235, 1156-1165.
HASAN, K. 1991. The Upper Cretaceous flysch succession
of the Balangbaru Formation, Southwest-Sulawesi.
In: Proceedings of the Indonesian Petroleum
Association, 20th Annual Convention. 183-208.
HERBIG, H. G. & BENDER, P. 1992. A eustatically driven
calciturbidite sequence from the Dinantian II of the
Eastern Rheinisches Schiefergebirge. Facies, 27,
245-262.
HISCOTT, R. N. & JAMES, N. P. 1985. Carbonate debris
flows, Cow Head Group, western Newfoundland.
Journal of Sedimentary Petrology, 55, 735-745.
HOLLOWAY,N. H. 1982. North Palawan block, Philippines
its relations to Asian mainland and role in
evolution of the South China Sea. AAPG Bulletin,
66, 1355-1383.
HOPKINS, J. C. 1977. Production of foreslope breccia by
differential submarine cementation and down-
slope displacement of carbonate sands, Miette
and ancient wall buildups, Devonian, Canada.
In: COOK, H. E. & ENOS, P. (eds) Deep-
water Carbonate Environments. Society of
Economic Paleontologists and Mineralogists,
Special Publication, 25, 155-170.
HUBERT, J. E, SUCHECKI, R. K. & CALLAHAN,R. K. M.
1977. The Cow head breccia: sedimentology of
the Cambro-Ordovician continental margin,
Newfoundland. In: COOK, H.E. & ENOS, E (eds)
Deep-water Carbonate environments. Society of
Economic Paleontologists and Mineralogists,
Special Publication, 25, 125-154.
HURST, J. M. & SURLYK, E 1984. Tectonic control of
Silurian carbonate-shelf margin morphology and
facies, North Greenland. AAPG Bulletin, 68,
1-17.
HUTCHISON, C. S. 1988. Stratigraphic-tectonic model for
eastern Borneo. Geological Society of Malaysia
Bulletin, 22, 135-151.
1989. Geological Evolution of Southeast Asia.
Oxford Monograph on Geology and Geophysics,
13. Oxford Science Publications.
-- 1992. The Eocene unconformity on southeast and
east Sundaland. Geological Society of Malaysia
Bulletin, 32, 69-88.
JAMES, N. P. & GINSBURG, R. N. 1979. The seaward
margin of the Belize barrier and atoll reefs.
International Association of Sedimentologists,
Special Publication, 3.
, STEVENS,R. K., BARNES, C. R. & KNIGHT, I. 1989.
Evolution of a lower Paleozoic continental-
margin carbonate platform, northern Canadian
Appalachians. In: CREVELLO, P. D., WILSON, J. L.,
SARG, F. & READ, J. E (eds) Controls on Carbonate
Platform and Basin Development. Society of
Economic Paleontologists and Mineralogists,
Special Publication, 44, 123-146.
387
JOHNS, D. R., MUTTI, E., ROSELL, J. & SI~GURET,M. 1981.
Origin of a thick, redeposited carbonate bed in
Eocene turbidites of the Hecho Group, south-central
Pyrenees, Spain. Geology, 9, 161-164.
KE~rrER, J. A. M. 1990. Carbonate platform flanks:
slope angle and sediment fabric. Sedimentology,
37, 777-794.
--, REYMER, J. J. G., VAN DER STRAATEN, H. C. &
PEPER, T. 1990. Facies patterns and subsidence
history of the Jumilla-Cieza region (southeastern
Spain). Sedimentary Geology, 67, 263-280.
KEPPER, J. C. 1981. Sedimentology of a middle Cambrian
outer shelf margin with evidence for syn-
depositional faulting, eastern California and western
Nevada. Journal of Sedimentary Petrology, 51,
807-821.
KRAUSE, t E & OLDERSHAW, A. E. 1979. Submarine
carbonate breccia beds - a depositional model
for two-layer sediment gravity flows from the
Sekwi Formation (Lower Cambrian), Mackenzie
Mountains, Northwest Territories, Canada.
Canadian Journal of Earth Sciences, 16, 189-199.
KUSUMA, I. & DARtN, T. 1989. The hydrocarbon potential
of the lower Tanjung Formation, Barito Basin, SE
Kalimantan. In: Proceedings of the Indonesian
Petroleum Association, 18th Annual Convention.
107-138.
LETOUZEY, J., WERNER, P. & MARTY, A. 1990. Fault
reactivation and structural inversion. Backarc and
intraplate compressive deformations. Examples of
the eastern Sunda shelf (Indonesia). Tectonophysics,
183, 341-362.
LOWE, D. R. 1976. Grain flow and grain flow deposits.
Journal of Sedimentary Petrology, 46, 188-199.
1979. Sediment gravity flows: their classification
and some problems of application to natural flows
and deposits. In: DOYLE, L.J. & PILKEY, O.H.
(eds) Geology of Continental Slopes. Society of
Economic Paleontologists and Mineralogists,
Special Publication, 27, 75-82.
- -1982. Sedimentary Gravity Flows: II. Depositional
models with special reference to the deposits of
high-density turbidity currents. Journal of
Sedimentary Petrology, 52, 279-297.
MCILREATH, I. A. & JAMES,N. P. 1984. Carbonate Slopes.
In: WALKER,R. G. (ed.) Facies Models. Geoscience
Canada, 12, 245-257.
MARTINI, I. P., RAU, A. & TONGIORGI, M. 1986.
Syntectonic sedimentation in a middle Triassic rift,
northern Apennines, Italy. Sedimentary Geology,
47, 191-219.
MIDDLETON, G. V. & HAMPTON,M. A. 1976. Subaqueous
sediment transport and deposition by sediment
gravity flows. In: STANLEY,D. J. & SWaFT, D. J. P.
(eds) Marine Sediment Transport and Environ-
mental Management. John Wiley and Sons, New
York, 245-257.
Moss, S. 1994. Tertiary basins of Kalimantan - A pre-
survey report. University of London Southeast
Asia Research Group, Report, 127.
MULLINS, H. T. 1983. Modem carbonate slopes and basins
of the Bahamas. In: COOK, H. E., HINE, A. C. &
MULLINS, H. T. (eds) Platform Margins and
Deepwater Carbonates. Society of Economic
388 M . E . J . WILSON • D. W. J. BOSENCE
Paleontologists and Mineralogists, Short Course
Notes, 12, 4.1-4.138.
- - & COOK, H. E. 1986. Carbonate apron models:
alternatives to the submarine fan model for
paleoenvironmental analysis and hydrocarbon
exploration. Sedimentary Geology, 48, 37-79.
-- & VAN BUREN, H. M. 1979. Modem modified
carbonate grain flow deposit. Journal of
Sedimentary Petrology, 49, 747-752.
--, HEATH, K. C., VAN BUREN, H. M. & HINE, A. C.
1984. Anatomy of modem open-ocean carbonate
slope: northern Little Bahama Bank. Sedimentology,
31, 141-168.
PACKHAM,G. H. 1990. Plate motions and Southeast Asia:
some tectonic consequences for basin development.
Southeast Asia Petroleum Exploration Society
Proceedings, 9, 55-68.
PARKINSON, C. D. 1991. The Petrology, Structure and
Geologic History of the Metamorphic rocks of
Central Sulawesi, Indonesia. PhD Thesis,
University of London.
PATRIAT, P. & ACHACHE, J. 1984. India-Eurasia collision
chronology has implications for crustal shortening
and driving mechanisms of plates. Nature, 311,
615-621.
PIETERS, P. E., TRAIL, D. S. & SUPRIATNA, S. 1987.
Correlation of early Tertiary rocks across
Kalimantan. In: Proceedings of the Indonesian
Petroleum Association, 16th Annual Convention.
291-305.
PFEIL, R. W. & READ, J. E 1980. Cambrian carbonate
platform margin facies, Shady dolomite, south-
western Virginia, U.S.A. Journal of Sedimentary
Petrology, 50, 91-116.
PICKERING, K., STOW, O., WATSON, M. t~ HISCOTT, R.
1986. Deep-water facies, processes and models:
A review and classification scheme for modem
and ancient sediments. Earth Science Reviews, 23,
75-174.
PLAYFORD, P. E., HURLEY, N. F., KERANS, C. t~
MIDDLETON, M. F. 1989. Reefal platform develop-
ment, Devonian of the Canning basin, western
Australia. In: CREVELLO,P. D., WILSON, J. L., SARG,
F. & READ, J. E (eds) Controls on Carbonate
Platform and Basin Development. Society of
Economic Paleontologists and Mineralogists,
Special Publication, 44, 188-202.
POSTMA,G., NEMEC,W. & KLEINSPEHN,K. L. 1988. Large
floating clasts in turbidites: a mechanism for their
emplacement. Sedimentary Geology, 58, 47-61.
PRIOR, D. B. & COLEMAN,J. M. 1982. Active slides and
flows in underconsolidated marine sediments on
the slopes of the Mississippi Delta. In: SAXOV, S.
8¢ NIEUWENHUIS, J. K. (eds) Marine Slides and
Other Mass Movements. NATO Scientific Affairs
Division, 37-79.
RANGIN, C., JOLIVET,L. & PUBELLIER,M. 1990. A simple
model for the tectonic evolution of Southeast Asia
and Indonesian region for the past 43 Ma. Bulletin
de la Societig Ggologique de France, 8, 889-905.
REIJMER, J. J. G. & EVERAARS,J. S. L. 1991. Carbonate
platform facies reflected in carbonate basin facies
(Triassic, Northern Calcareous Alps, Austria).
Facies, 25, 253-278.
--, SCHLAGER,W. & DROXLER, A. W. 1988. Site 632:
Pliocene-Pleistocene sedimentation cycles in a
Bahamian basin. Proceedings of the Ocean Drilling
Program, Scientific Results, 101,213-220.
--, --, BOSSCHER, H., BEETS, C. J. & MCNEILL,
D. E 1992. Pliocene/Pleistocene platform facies
transition recorded in calciturbidites (Exuma Sound,
Bahamas). Sedimentary Geology, 78, 171-179.
, TEN KATE,W. G. H. Z., SPRENGER,A. & SCHLAGER,
W. 1991. Calciturbidite composition related to
exposure and flooding of a carbonate platform
(Triassic, Eastern Alps). Sedimentology, 38, 1059-
1074.
RODINE, J. & JOHNSON,A. M. 1976. The ability of debris,
heavily freighted with coarse clastic materials, to
flow on gentle slopes. Sedimentology, 23, 213-234.
Ru, K. & PIGOTT, J. D. 1986. Episodic rifting and sub-
sidence in the South China Sea. AAPG Bulletin,
70, 1136-1155.
SALLER, A., ARMIN. R., ICHRAM, L. O. & GLENN-
SULLIVAN, C. 1992. Sequence stratigraphy of upper
Eocene and Oligocene limestones, Teweh area,
central Kalimantan. In: Proceedings of the
Indonesian Petroleum Association, 21st Annual
Convention. 69-92.
SARG, J. E 1988. Carbonate sequence stratigraphy. In:
WmGUS, C. K., HASTINGS, B. S., KENDALL,C. G.
ST. C., POSAMENTIER, H. W., Ross, C. A. &
VAN WAGONER, J. C. (eds) Sea-level Changes -
An Integrated Approach. Society of Economic
Paleontologists and Mineralogists, Special
Publication, 42, 155-181.
SCHLAGER, W. t~ CAMBER, O. 1986. Submarine slope
angles, drowning unconformities, and self-erosion
of limestone escarpments. Geology, 14, 762-765.
-- • CHERMAK,A. 1979. Sediment facies of platform-
basin transition, Tongue of the ocean, Bahamas.
In: DOYLE, L. J. & PILKEY, O. H. (eds) Geology
of Continental Slopes. Society of Economic
Paleontologists and Mineralogists, Special
Publication, 27, 193-208.
SILVER, E. A., JOYODIWIRYO, Y. & MCCAFFREY,R. 1978.
Gravity results and emplacement geometry of the
Sulawesi ultramafic belt, Indonesia. Geology, 6,
537-531.
S1MANDJUNTAK,Z. O. 1990. Sedimentology and tectonics
of the collision complex in the East Arm of
Sulawesi, Indonesia. Geologi Indonesia: Journal
of the Indonesian Association of Geologists, 13,
1-35.
SITUMORANG, B. 1987. Seismic stratigraphy of
the Makassar basin. In: Lemigas Scientific
Contribution, 3-37.
SMITH, R. B. & SILVER,E. A. 1991. Geology of a Miocene
collision complex, Buton, eastern Indonesia.
Geological Society of America Bulletin, 103,
660-678.
SUKAMTO,R. 1975. The structure of Sulawesi in the light
of plate tectonics. In: Proceedings of the Regional
Conference on the Geology and Mineral Resources
of SE Asia, Jakarta. 1-25.
-- 1982. Geologi lembar Pangkajene dan Watampone
bagian barat, Sulawesi. Geological Research and
Development Centre, Bandung.
 TERTIARY EVOLUTION OF S SULAWESI
- - & SUPRIATNA, S. 1982. Geologi lembar Ujung
Pandang, Benteng dan Sinjai quadrangles,
Sulawesi. Geological Research and Development
Centre, Bandung.
TUCKER, M. E. & WRIGHT, V. E 1990. Carbonate
Sedimentology. Blackwell Scientific Publications.
T W ~ L , W. W., DAVIS, R. G & McDOWELL, H. G. 1986.
Miocene carbonate shelf margin, Bali-Flores Sea,
Indonesia. In: Proceedings of the Indonesian
Petroleum Association, 15th Annual Convention,
124-140.
VAN DE WEERD, A. A. & ARMIN, R. A. 1992. Origin and
evolution of the Tertiary hydrocarbon-bearing
basins in Kalimantan (Borneo), Indonesia. AAPG
Bulletin, 76, 1778-1803.
, --, MAHADI, S. & WARE, P. L. B. 1987. Geologic
setting of the Kerendan gas and condensate
discovery, Tertiary sedimentation and paleo-
geography of the northwestern part of the Kutei
basin, Kalimantan, Indonesia. In: Proceedings of
the Indonesian Petroleum Association, 16th Annual
Convention. 317-338.
VAN LEEUWEN, T. M. 1981. The geology of southwest
Sulawesi with special reference to the Biru area. In:
BARBER, A. J. & WmYOSUJONO, S. (eds) The
Geology and Tectonics of Eastern Indonesia.
Geological Research and Development Centre,
Bandung, Special Publication, 2, 277-304.
WAIN, T. & BEROD, B. 1989. The tectonic framework
and paleogeographic evolution of the upper Kutei
389
basin. In: Proceedings of the Indonesian Petroleum
Association, 18th Annual Convention. 55-78.
WATTS, K. F. 1987. Triassic carbonate submarine fans
along the Arabian platform margin, Sumeini Group,
Oman. Sedimentology, 34, 43-71.
- - & BLOME, C. D. 1990. Evolution of the Arabian
carbonate platform margin slope and its response to
orogenic closing of a Cretaceous ocean basin,
Oman. In: TUCKER,M. E., WILSON,J. L., CREVELLO,
P. D., SARG, J. R. & READ, J. E (eds) Carbonate
Platforms: Facies, Sequences and Evolution.
International Association of Sedimentologists,
Special Publication, 9, 291-323.
WEISSEL, J. K. 1980. Evidence for Eocene oceanic crust
in the Celebes Basin. In: HAYES, D. E. (ed.) The
Tectonic and Geologic Evolution of Southeast Asian
Seas and Islands. American Geophysical Union,
Geophysical Monograph, 23, 37-47.
WILLIAMS, P. R., JOHNSTON, C. R., ALMOND, R. A. &
SIMAMORA, W. H. 1988. Late Cretaceous to early
Tertiary structural elements of west Kalimantan.
Tectonophysics, 148, 279-297.
WOODS, B. G. M. 1985. The mechanics of progressive
deformation in crustal plates - A working model for
Southeast Asia. Geological Society of Malaysia
Bulletin, 18, 55-99.
YUWONO, W. S, BELLON, H., SOERIA-ATMADJA, V. &
MAURY, R. C. 1985. Neogene and Pleistocene
volcanism in South Sulawesi. Proceedings Ikatan
Ahli Geologi Indonesia, 14, 169-179.
 Tertiary Tectonic and magmatic evolution of western Sulawesi
and the Makassar Strait, Indonesia: evidence for a Miocene
continent-continent collision

STEVEN C. B E R G M A N 1, D A N A Q. C O F F I E L D 2, J A M E S E T A L B O T l, 3
& RICHARD A. G A R R A R D 4
1ARCO Exploration & Production Technology, Exploration Research & Technical Services,
2300 W. Plano Pkwy., Plano TX 75075-8499, USA. e-mail dprscb@er.arco.com
2 Atlantic Richfield Indonesia, The Landmark Center, Tower B, Jl. Jend Sudirman,
Kav. 70A, Jakarta, Indonesia 12910. e-mail jnvdqc@ariie.arco.com
3Now at: KT Geoservices, 661 N. Plano Rd, Suite 317, Richardson TX 75081, USA
4 ARCO International Oil & Gas Co., Asia-Pacific New Ventures Exploration,
2300 Plano Pkwy., Plano TX 75075-8499, USA. Now at: ARCO Alaska, 700 G Street,
Anchorage AK 99701, USA

Abstract: New field and laboratory data from western Sulawesi, Indonesia, integrated with
available data establish its Late Cenozoic igneous framework and a new model for its tectonic
evolution. Western Sulawesi contains three major Neogene N-S-trending tectonic domains (from
W to E): (1) an active foldbelt, in which Pliocene and Miocene volcanogenic rocks are involved
in W-vergent thrusting which extends into the Makassar Strait; (2) a central belt comprised of
a deformed submarine Miocene volcanoplutonic arc built on an Oligocene-Eocene clastic and
carbonate platform with Latimojong Mesozoic basement metamorphic and sedimentary rocks
thrust over its eastern margin on W-vergent faults; and (3) an accreted Cretaceous-Palaeogene(?)
ophiolite (Lamasi Complex) between the Latimojong basement block and Bone Bay. The Lamasi
Complex ophiolite includes dioritic plutons, basaltic sheeted dykes, pillow lavas, greenstones,
tufts and volcanic agglomerates with depleted (MORB-like) Sr & Nd isotope and REE charac-
teristics of probable normal oceanic crust with possible subduction-related or back-arc affinity.
New K-Ar, 4°Ar-39Ar, Rb-Sr, and Nd-Sm isotope data suggest Cretaceous to Eocene crystalliz-
ation and Oligocene to Miocene obduction. Late Miocene to Pliocene extrusive and intrusive
rocks form a cogenetic volcanoplutonic complex of calc-alkalic to mildly alkalic, potassic, and
shoshonitic felsic and mafic magmatic rocks of bimodal composition which were erupted and
intruded during a short episode of Middle Miocene to Pliocene (3-18 Ma) lithospheric melting.
Based on new Rb-Sr, Nd-Sm, and U-Pb isotope, and major and trace element geochemical data,
parental source rocks of the Miocene melts were Late Proterozoic to Early Palaeozoic crustal and
mantle lithospheric assemblages which became heated and melted owing to a continent-continent
collision in which west-vergent continental lithosphere derived from the Australian-New Guinea
plate was subducted beneath eastern-most Sundaland. The timing of this magmatism and subse-
quent cooling and denudation history are constrained by 113 new K-Ar, 4°Ar-39Ar, and fission
track ages. The new tectonic model differs significantly from previous models: 'the Makassar
Strait is now interpreted as a foreland basin bound on both sides by converging Neogene thrust
belts, in contrast to previous models suggesting Late Tertiary oceanic spreading or continental
rifting. West-vergent obduction of a pre-Eocene oceanic, primitive arc, or back-arc crust onto
western Sulawesi occurred during late Oligocene to Miocene times. The Late Miocene western
Sulawesi magmatic arc is envisioned as a continent-continent collision product, in contrast
to previous models involving a normal ocean--continent or ocean-ocean subduction-related
magmatic arc (west or east vergent) or post-subduction rifting. The east Sulawesi ophiolite
extends into western Sulawesi, suggesting that Bone Bay resulted from collapse of the over-
thickened Miocene orogen. The new tectonic model illustrates the central role western Sulawesi
plays in unlocking the complex evolution of Indonesia as well as the temporal and magmatic
details of a continent-continent collision zone.

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, 391

Geological Society Special Publication No. t06, pp. 391-429.
392 s.C. BERGMAN ET AL.
The objective of this paper is to integrate new and
published geological, petrological, isotopic, strati-
graphic, structural and geochemical data for
western Sulawesi in order to test available models
and formulate a new tectonic model for the region.
Sulawesi is located in a complex tectonic position,
at the intersection of three major lithospheric
plates: the westward-moving Pacific Plate, the
northward-moving Australian Plate and the rela-
tively stationary Eurasian Plate (Fig. 1). The
complex plate tectonic position of Sulawesi is
manifested in a varied Tertiary structural and strati-
graphic record (Figs 1-3). The south arm of
Sulawesi is dominated by Miocene and younger
volcanic and plutonic rocks forming a magmatic
belt that many workers have regarded as a sub-
duction-related volcanic arc involving a westward-
dipping (Sukamto 1978; Hamilton 1979), or east-
dipping (Katili 1978) oceanic plate. Some recent
workers interpret the magmatic arc as a post-
collisional rift-related magmatic belt (Yuwono
et al. 1988b; Leterrier et al. 1990; Kalavieris
et al. 1992; Priadi et al. 1994). The present study
presents new field and laboratory evidence based
on three seasons of geological research (1990-
1993) associated with the exploration for hydro-
carbons in the Kalosi Production Sharing Contract
(PSC) (see Coffield et al. 1993 for a summary of
the petroleum systems framework). The results
demonstrate that previous 'conventional' tectonic
models for the Late Tertiary evolution of western
Sulawesi and the adjacent Makassar Strait require
significant modification. The authors propose that
the Miocene to Recent tectonic framework of
western Sulawesi and Makassar Strait was
dominated by compressional continent-continent
collision processes.
Geological and tectonic framework
Sulawesi is located at the southeast limit of the
Sunda Platform crustal domain and north and west
of the Australian Continental Platform, its probably
derivative fragments forming Irian Jaya, Sula,
Buru, Seram and the Tukang Besi Platform.
Sulawesi formed along the Oligocene-Miocene
collision zone between the Eurasian Plate and
micro-continental fragments derived from the
Indian-Australian Plate (Hamilton 1979;
Hutchison 1989a; Rangin et al. 1990; Daly et al.
1991). The four arms of Sulawesi and adjacent
islands form four distinct megatectonic provinces.
The northern arm is composed of late Palaeogene
to Neogene subduction related volcanic arc rocks
resulting from the west-dipping subduction of the
Molucca Sea Plate (Jezek et al. 1981; Bellon &
Rangin 1991). The east and southeast arms are
composed of Mesozoic and younger allochthonous
metamorphic and ophiolitic rocks which were
obducted during the Oligocene epoch (Parkinson
1991). The south arm is dominated by Miocene and
younger volcanic and plutonic rocks which form
a magmatic belt superimposed on the Mesozoic
basement of the southeastern margin of Sundaland
(Katili 1978; Silver et al. 1983a, b). The fourth
megatectonic province contains Late Palaeozoic
and Mesozoic Australian-derived microcontinents
which have been accreted to the eastern margin
of Sulawesi, comprising Banggai, Sula, Buton,
Kabaena and Tukang Besi, among other islands.
The present work concerns parts of the south and
north arms of Sulawesi near their boundary in an
area approximately bound by Mamuju, Palopo,
Parepare and Barulatong (Fig: 2), which will be
referred to as 'western Sulawesi' for simplicity.
The present-day tectonic framework of western
Sulawesi is dominated by active uplift, and com-
pressional and strike-slip faulting. The region is
seismically active and nearly all historic earth-
quake hypocentres are < 50 km below the surface
(cf. USGS National Earthquake Information
Center, Boulder, Colorado, PDE database which
shows > 40 shallow events between November
1929 and July 1994 with Richter magnitudes 3-5
in the Mamasa-Parepare areas). The study area is
500-800 km north of the Sunda-Banda volcanic
arc/subduction zone in which the Australian Plate
is being subducted beneath Sumbawa, Flores and
the Flores Sea. Miocene and younger collision of
Australian continental crust has resulted in the
accretion of Timor to the upper plate of Eurasia and
a south-dipping thrust zone has developed north of
Flores in response to the continued compression.
Several major NNW trending strike-slip faults
cut through the south arm; these faults have
traditionally been interpreted as sinistral from
field-based structural observations (Berry & Grady
1987) and LandSat imagery interpretations
(M. Crawford pers. comm. 1991), although there
is evidence for right-lateral offset on some faults.
Our recent work has shown that some of these
strike-slip faults are transpressional and possess
recent compressional offset. Late Tertiary com-
pressional structures include oblique-trending folds
and shallow-angle thrust faults. The presence of
several Quaternary basins within and adjacent to
the south arm of Sulawesi indicate the presence o f
Late Tertiary subsidence, possibly representing
piggy-back basins adjacent to thrust sheets, or due
to transtensional processes.
Geological data relevant to the study area include
published 1:1 000000 and 1:250000 scale geo-
logical maps (Djuri & Sudjatmiko 1974; Ratman
1976; Sukamto 1975, 1982; Ratman & Atmawinata
1988) and various literature information (e.g. van
Fig. 1. Present-day tectonic elements of Indonesia showing the distribution of active trenches, Quaternary subduction-related volcanic arcs, and major structures.
L~
394 S.C. BERGMAN ET AL.
Bemmelen 1949; Audley-Charles 1977; Hamilton
1979; Sasajima et al. 1980; Van Leeuwen 1981;
Garrard 1989; Hasan et al. 1991; Coffield et al.
1993), which are summarized in a geological
sketch map in Fig. 2 and in a schematic strati-
graphic section in Fig. 3. Geological studies in
western Sulawesi include early Dutch work
initiated during the first part of this century
(Abendanon 1915; t'Hoen & Ziegler 1917; Reyzer
1920; Sax 1931a, b; Sung 1948). More recent
pubfished information on geochronology and
petrology of volcanic and plutonic rocks of western
Sulawesi include Sasajima et al. (1980), Van
Leeuwen ( 1981 ), Sukamto & Simandjuntak (1983),
Leterrier et al. (1990), Kalavieris et al. (1992),
Yuwono 1987, Yuwono et al. (1988), Van Leeuwen
et al. (1994) and Priadi et al. (1993, 1994).
The stratigraphy of western Sulawesi is domi-
nated by Neogene rocks but also includes forma-
tions as old as Jurassic (Fig. 3). Five main litho-
stratigraphic sequences are recognized (Garrard
et al. 1992; Coffield et al. 1993): (1) Mesozoic
(mainly Late Cretaceous) metasedimentary rocks
of the Latimojong, Balangbaru and Marada
Formations, which include flysch deposits believed
to have formed in a forearc basin setting, and
ophiolites of the Lamasi Complex Cretaceous-
Palaeogene?); (2) a Palaeogene (Middle to Late
Eocene) syn-rift sequence composed of silici-
clastic, coal, volcanic and carbonate sedimentary
deposits of the Toraja and Malawa Formations;
(3) an Eocene to Middle Miocene carbonate
post-rift sequence including the Makale and Tonasa
Formations; (4) a Middle Miocene to Pliocene
basaltic to dacitic volcano-plutonic and epiclastic
sedimentary sequence including the Enrekang and
Camba Volcanic series and the Buakayu Formation,
and consanguineous granitic to gabbroic intrusives;
and (5) Pliocene and younger synorogenic, non-
marine to upper bathyal sedimentary deposits
including the Walanae Formation, formed during
widespread thrusting, regional uplift and erosion.
These five sequences occur in a variety of
structural positions. The older sequences are
typically structurally highest in the sections and in
the hinterland to the east; the younger sequences
occur in structurally lower positions than in the
sequences towards the west or foreland.
Importantly, the western promontory of the south
arm of Sulawesi (near 3°S latitude) is a relatively
recent geomorphological product, resulting from
uplift due to Neogene thrusting which extends for
200-300 km from the Majene foldbelt in the west
to the Kalosi foldbelt in the east.
Makassar Strait
The Makassar Strait is characterized by a central
deep-water trough with depths as great as 1500-
2500 m. In contrast to most of the Makassar Strait
margins which contain a broad shelf, the western
Sulawesi margin north of Majene possesses a very
narrow shelf (< 10-20 km) and steep slope. The
tectonic evolution of the Makassar Strait has been
the subject of considerable debate since Alfred
Russell Wallace delineated his Australian-Asian
faunal boundary along its axial trend in 1858 (for
a Wallace Line review, see George 1981). Most
workers have incorporated phases of Palaeogene,
Miocene or Quaternary rifting which resulted from
NW-SE extension along the Makassar Strait (Katili
1971, 1975, 1978; Carey 1976; Audley-Charles
1977; Hamilton 1978, 1979; Burollet & Salle 1981;
Situmorang 1982a, b, 1984; Faugeres et al. 1989;
Shaw & Packham 1992; Effendi 1993, among
many others)i Rifting is thought to have involved
oceanic (Hamilton 1979) or continental crust
(Burollet & Salle 1981). Malecek et al. (1993)
proposed that the Strait is underlain by trapped
Cretaceous oceanic crust. Alternative models
involving a foredeep to the western Sulawesi
orogen have been advanced by van Bemmelen
(1949), although this interpretation has largely
been neglected by recent workers, until now.
Most recent interpretations tend to draw long
continuous normal faults down the axis of the
Strait, or attempt to close up the basin by matching
the Sulawesi western coastline with the eastern
coastline of Borneo (Katili 1978). Whereas this
procedure may work well for passive margins such
as West Africa and east South America, the
recently active (15 Ma to present) fold and thrust
belt of western Sulawesi has produced a rapidly
evolving Neogene coastline which does not lend
itself to coastline matching.
Seismic sections orientated NW-SE show the
salient tectonostratigraphic sequences and
structural styles. Whereas Eocene extensional
faulting is recorded in certain deeper parts of the
sections, Oligocene passive platform conditions
are present, and Late Miocene to Pliocene fold
and thrust belt development is apparent (Fig. 4).
The leading edge of the thrust front is evident on
the seismic section, as is a series of stacked thrust
sheets with a series of overlying piggy-back basins.
These piggy-back basins contain distinct strati-
graphic packages which record the movement of
the underlying thrust sheets. Contemporaneous
sedimentary packages in front of the leading thrust
are entirely parallel bedded and undeformed
whereas those in the piggy-back basins contain
diverging reflection packages which onlap the
adjacent growing ramp anticlines. The Samarinda
anticlinorium, a fold and thrust belt on the opposite
(western) side of the northern Makassar Strait
mirrors the Majene foldbelt. The Samarinda anti-
 MIOCENE W SULAWESI CONTINENT COLLISION 395

120" 00'

KALOSI PSC.

03*

03 ° 30'

0 10 20 30 Kilt.
i ..... !

II II i i111 ii i ii1,1 iii

t2~ t25"
LEGEND : ~i
04* O0' ~PARE PARE
QUATERNARYALL~ 1 VOLCANICS "1

PMOCENE WALANAE FM.

1 1 GRANITIC OTHER
VOLC/LNICS& CLAt~I'ICS INTRUSIVES
L EOCENE - M. MIOCENE
0
It¢ - L EOCENETORAJA FM. INDEX MAP
llllllll,i i I I i~ i
AGE ? LAMASl
i COMPLEX
OPHIOLITE
PRE TERTIARYBASEIkrcNT
LATIMOJOI~

I .......... I ,,~ I ,,i

Fig. 2. Generalized geological map of western Sulawesi.

clinorium is a Neogene east-vergent compressional is controversial and calls upon inversion of pre-
trend which has developed in the Kutei Basin along existing rift structures (Biantoro et al. 1992) or
strike to the north of the west-vergent Meratus gravity sliding from the hinterland of the Kutei
Mountains orogen. The origin of these structures Basin (Ott 1987; van de Weerd & Armin 1992).
396 S.C. BERGMAN ET AL.

STRATIGRAPHY
NEW K-Ar ages
FORMATION LITHOLOGY TECTONICS 0 10
Ma
,==. , , . . . . , .
40 :)CENE
~- sroc. SYN-
WALANAE FM.
p, ~C~.Et F .__q O R O G E N I C
•

PEAK

FORELAND
15-
BASIN/
FOLD & THRUST
BELT INITIATION "!

20-
OPHIOLITE I~, , , , , , , , ,

OBDUCTION
>-
25- IT" POST-RIFT
<
m
1--
30- rr

UJ
1--
35.

TORAJA FM
( MALAWA )
40.
SYN-RIFT

4,5'

50.

55

60 LATE
PRE-RIFT
m ~ P
~L..--"~
J__---
VV V J ~,r~
65.
VVV .........
VVV

+ +'1~

Fig. 3. Stratigraphic column of western Sulawesi.

Fig. 4. Interpreted E-W seismic section 201 through the Makassar Strait (after Coffield et al. ]993).
 ~T

'i I
~- /
°F-
~
,.,.~ °ILl
I

Z
_1
 MIOCENE W SULAWESICONTINENT COLLISION
Herein, the Makassar Strait is interpreted as a
present-day foreland basin bound on both sides by
converging Neogene fold and thrust belts.
Nature and distribution of Cretaceous-
Palaeogene ? and Neogene igneous rocks
Two groups of igneous rocks are distinguished
on the basis of field relations, age, composition,
petrology and isotope and trace element geo-
chemistry: Cretaceous-Palaeogene(?) ophiolitic
rocks of the Lamasi Complex and Miocene-
Pliocene igneous rocks of the Camba-Enrekang-
Mamasa Complex.
Lamasi Complex ophiolitic rocks are intensely
deformed, typically metamorphosed and mainly
bounded by thrust faults. Associated sedimentary
deposits include red cherts and mudstones. The
Lamasi Complex ophiolites include a variety of
intrusive and extrusive lithologies. Intrusives such
as sheared diorite plutons (> 20 km diameter),
amphibolite gneisses (metagabbros) and sheeted
basalt dyke complexes have been recognized.
Extrusives include basaltic pillow lavas, basalt to
andesite lava flows and agglomerates, and andesite,
dacite and rhyolitic pyroclastic deposits comprising
block breccias, tuff breccias and tuffs,
Neogene igneous rocks of the Camba-
Enrekang-Mamasa Complex and derivative
volcaniclastic deposits cover more~han 75% of
the surface of western Sulawesi (Fig. 2). Volcanic
outcrops consist of massive lavas and pyroclastic
sections with > 5 km thicknesses; plutons and
laccoliths of batholithic dimensions (> 50-100 km
diameter) are intimately associated with this
volcanic assemblage. Volcanic and intrusive
sequences are variably deformed; many volcanic
sequences are steeply dipping to vertical and
batholithic-scale laccoliths are detached and
bounded by thrust faults. The volcanic rocks range
from basalt to rhyolite in composition, but are
dominated by basaltic to dacitic stratovolcano
complexes, lava domes and flows, and pyroclastic
sequences. Plutonic rocks range from gabbroic and
dioritic stocks, sills and dykes to granitic-quartz
monzonitic to monzodioritic stocks, plutons and
large laccoliths. Lamprophyre dykes are common.
The age of the Camba-Enrekang-Mamasa igneous
sequences is mainly Middle to Late Miocene
(averaging 8 _+2 Ma), although the total age range
observed is 3-18 Ma (see below).
Analytical results and interpretation
Petrology
A summary of lithological and location data for
intrusive and extrusive rocks included in the
397
present study is provided in Table 1. (For more
details see table 1 of Supplementary Publication
No. SUP 18099 (11 pp) available from the Society
Library or the British Library Document Supply
Centre, Boston Spa, Wetherby, W. Yorks LS23
7BQ, UK.) Analytical methods and details are
provided in the Appendix. The locations of various
types of analytical data are summarized in Figs 5-7.
Modal analyses of eight selected phaneritic
Miocene intrusives of the Camba-Enrekang-
Mamasa Complex (Table 2) range from quartz
monzonite and monzonite to diorite in the IUGC
classification scheme.
The Miocene Camba-Enrekang-Mamasa
Complex comprises basaltic and gabbroic lith-
ologies rich in augite, through andesites/diorites
rich in augite and amphibole with rare enstatite, to
granitic-monzonitic lithologies relatively poor in
quartz and rich in K-feldspar, albite and biotite. The
most primitive basaltic lithologies are phlogopite-
bearing lamprophyres (olivine minettes) and alkali
gabbros. Volcanic rocks are dominantly sub-
alkaline to shoshonitic, with leucite or analcime
observed in many trachytes or trachyandesites.
Palaeogene ophiolites of the Lamasi CoMplex
include a variety of intrusive and extrusive lith-
ologies. Intrusive rocks include basalt, diorite and
gabbro of varying degrees of metamorphism
(greenschist-amphibolite grade). Extrusives are
dominantly basalt, andesite and dacite.
Major and trace element geochemistry
Major and trace element compositions of 59
samples define specific compositional groups and
trends for the Lamasi and Miocene complexes,
illustrated in Harker variation diagrams and
compositional plots (Figs 8-14) and REE and
spider diagrams (Figs. 11-12) (selected averages
are listed in Table 3; detailed analytical data are in
table 2 of Supplementary Publication SUP 18099
(see above)). The Miocene igneous rocks range
from basalt to rhyolite in the TAS classification
scheme (Fig. 8), and are dominated by high K
calc-alkalic and shoshonitic series (see Bergman
1987 for an overview of K-rich alkaline igneous
rocks). On an AFM diagram, the Miocene igneous
suite defines a calc-alkaline trend (Fig. 10). The
presence of basaltic trachyandesites, trachy-
andesite, and borderline trachyte-dacite compo-
sitions indicates an alkaline affinity of the intrusive
and extrusive units. Compositions are dominated
by relatively alkaline basaltic and dacitic rocks,
with a conspicuous lack of andesites. Miocene
intrusive and extrusive lithologies are geo-
chemically similar and define the same trends on
major element cross plots (Fig. 9). Compositions
of Miocene intrusive and extrusive rocks from a
 ~D

Table 1. Sample location summary

lat S long E Analyses

elev
sample lithology locality age (m) deg min deg min K A G I

90SUL1 phlogopite olivine basalt dyke Awang R. Miocene 290 3 11.9 119 43.9 i
90SUL2 arkosic sandstone Awang R. Miocene 295 3 11.8 119 43.8
90SUL3 alkali gabbro sill Sipe Hill Miocene 290 3 11.7 119 43.6 b i sn
90SUIAa syenogabbro sill Tou R. Miocene 290 3 11.3 119 43.7 b i sn a
90SUL5 basalt crystal lithic lapilli tuff Subing R. Miocene 730 3 9.9 119 44.7 i sn
90SUL6a basalt crystal lithic lapilli tuff Sadang R. Miocene 830 3 9.9 119 45.8 i sn z
90SUL6b graded basalt crystal lithic lapilli tuff Sadang R. Miocene 830 3 9.9 119 45.8 i sn
90SUL7 trachyandesite crystal lithic lapilli tuff Sadang R. Miocene 830 3 7.6 119 46.9 b i sn z
90SUL8 monzodiorite pluton Palopo pluton Miocene 1070 2 57.4 120 4.2 bh i sn azs
90SUL9 quartz monzodiorite pluton Palopo pluton Miocene 1050 2 56.8 120 4.9 bh i sn z
90SUL10a andalusite siltstone homfels Palopo pluton aureole Cretaceous 250 2 57.1 120 7.5 sn a
90SUL 11 basalt block in flow breccia Bambulu R. Palaeog./Cret 10 2 50.6 120 7.2 1 sn a
90SUL12 quartz monzodiorite pluton Palopo pluton Miocene 475 2 57.5 120 5.7 bh 1 snp az
90SULI3 monzodiorite stock Rantepao stock Miocene 835 2 57.2 119 58.9 h 1 sn z
90SUL14 dacite porphyry lava Mt. Tambakkuku Miocene 400 3 30.7 119 47.0 b 1 sn z
90SUL15 dacite porphyry lava Mt Tambakkuku Miocene 365 3 28.9 119 47.1 bs 1 P a
90SUL16a coarse crystal lithic dacite tuff Mt Tambakkuku Miocene 400 3 28.9 119 47.6 bp 1
90SUL 18a crystal lithic dacite lapilli tuff Batu R. Miocene 100 3 33.2 119 43.5 1
90SUL18b welded crystal lithic dacite lapilli tuff Batu R. Miocene 100 3 33.2 119 43.5 bp l sn
90SUL20 dacite porphyry lava Loca R. Miocene 120 3 34.0 119 42.3 b 1 sn Z
90SUL21 a quartz monzodiorite pluton Mamasa Miocene 1100 2 56.4 119 22.7 b snp az
90SUL21 b quartz monzodiorite pluton Mamasa Miocene 1100 2 56.4 119 22.7 a
90SUL22a monzonite pluton Balakalua Miocene 1035 59.1 119 20.4 b
90SUL22b fluvial alluvium draining pluton Balakalua Recent 1035 59.1 119 20.4 b
90SUL23 dacite porphyry lava Buntubuntu Miocene 975 0.9 119 18.9 b sn
90SUL24 vitric rhyolite lava Sumarorong Miocene 850 10.1 119 19.8 sn
90SUL25 quartz monzodiorite pluton Polewali pluton Miocene 70 24.2 119 20.8 bh snp aZS
90SUL26a andesite crystal lithic lapilli tuff Pare Pare stratavolcano Miocene 60 55.2 119 42.5 b Z
90SUL26b andesite crystal lithic lapilli tuff Pare Pare stratavolcano Miocene 60 55.2 119 42.5 Z
90SUL26c andesite block in tuff breccia Pare Pare stratavolcano Miocene 60 55.2 119 42.5 b snp a

90SUL26d andesite block in tuff breccia Pare Pare stratavolcano Miocene 60 55.2 119 42.5 b sn z
90SUL27 trachyte porphyry dome Allakuang quarry Miocene 100 59.4 119 47.8 b sn a

90SUL28 quartz monzonite gneiss clast in Walanae Enrekang Pliocene 200 33.1 119 45.0 b snp a

RAG/90/73 fine grained volcaniclastic sandstone Sadang River Miocene 300 12.7 119 43.7 a

RAG/90/74 siliciclastic sandstone Sadang River Miocene 300 12.6 119 43.8 a

92SUL1 vertical dacite lava Biloka River Miocene 189 34.6 119 40.6 b
92SUL2 diorite/gabbro xenolith in dacite lava Biloka River Miocene 190 33.9 119 41.9
92SUL3 vertical dacite lava thrust sheet Biloka River Miocene 185 34.4 119 40.5 b
92SUIA horizontal dacite lava Biloka River gorge Miocene 160 38.2 119 41.7
92SUL5 horizontal dacite lava Benteng dam Miocene 30 41.3 119 4O.0
92SUL6 horizontal dacite tuff Waru Miocene 20 34.3 119 34.8 b
92SUL7 dacite pyroclastic flow lapilli tuff Balu Miocene 55 42.1 119 37.3 b
92SUL8 vertical Eocene sandstones Lambou/Kalosi Eocene 270 19.8 119 49.4 az
ND92/29 dacite lava dome Tirasa Mtn/Benteng Miocene 425 37.4 119 37.9 b
RAG92-004 tuffaceous sandstone end of New Rangas Rd Miocene 400 37.6 118 49.2 s
RAG92-006 trachyte/shoshonite tuff breccia end of New Rangas Rd Miocene 5 37.6 118 49.2 s
RAG92-008 leucite shoshonite lapilli tuff end of New Rangas Rd Miocene 5 37.6 118 49.2
400 s . c . BERGMAN ET AL.

Table 2. Modal compositions of selected Miocene intrusive rocks

sample quartz plagioclase K-feldspar biotite hornblende augite

(vol%)

90SUL8 4 61 17 6
90SUL9 7 50 24 3
90SUL12 11 55 14 8
90SUL13 2 63 10 9
90SUL21 a 11 49 26
90SUL22a 4 45 33
90SUL25 5 38 27 19
90SUL28 22 41 32

Normalized volume percentages based on c. 500 points using a 1 mm grid.

200-300 km wide area define continuous trends exhibit flat to LREE depleted patterns similar to
on these variation diagrams, suggesting a cogenetic N-type MORB and possess spider diagram patterns
relationship. similar to MORB and supra-subduction zone rocks
Lamas• Complex ophiolites are compositionally (SSZ; arc or back-arc) defined by slight negative
distinct from Miocene intrusives and extrusives in Nb anomalies.
terms of major and trace element contents. Lamas•
ophiolitic rocks range from basalt to rhyolite in Conventional K - A r and 4°Ar-39Ar
major element composition (Fig. 9), are marginally
geochronology
thole••tic on an AFM plot (Fig. 10), and are
depleted in most trace elements compared with Fifty-three conventional K-Ar ages of biotite, horn-
the Miocene magmatic units. Lamas• ophiolites blende, plagioclase, san•dine, phlogopite, white

118 `>30' 119 ° 120 ° 120 ° 30'

2 ° 30' I I 2 ° 30'

RAG92-0~

90-11 J
93-3093-33
• qP93-34b
~ + • 90-21a,b 90-1390-9 90-10aJ Palopo
93-210 • 90.8~OA93.2728
3°-
9022a'0bq~Mamasa , 90-12 " ~ ~ ' Bone
90-23- 93.2o• -t- 93-3~ ~ 93~.~4o
93-363 II - ~,~
_ 3°

Makassar (- 93-15• 93-18b 90-7 93~193,39 ~,

Strait \ 93-13 • 9090"~O090-6a,b
90-24• • •
..... 93-17
~ 93RSI79A 93-11 RAG/90/73,74
92-9• 92-8 93-45 93-~
93-10#93.9 KaJosi • • O• •93-43
90"2,~1L~93"5 93-3 93-53•93~23~b-48a,b 93-44
~,93Rs2°1.~~e
A J 93"63~'~ 09~1 90-15093"5-03~9933"~7
-- "~, " 90-14•90-16aRAG92-018
927~2°_.09o-28 93-46
]"92.6• 92.111~2.2~' En rekang
/ • • 9U-1Ua,D
ND92/29 92-4
• %.~
0 • 251 _ 501km ~ 92-7

I0 25 mi \
90.2%b,0,,
Pare Pare/ . 90-27
4° t ~ , ~ 9 , I 4°
118 ° 30' 119 ° 120 ° 120 ° 30'

Fig. 5. Sample location map for all samples included in the present study.
 118 ° 30' 119 ° 120 ° 120 ° 30'
2 ° 30' I / I 2 ° 30'
b 7.3.+0.2 /
b 2.4.+0.1 , /
sn 5 . 3 _ + 0 . ~

Mamuju /I
L p120.0-.+1.0 ~1

] b77+03 b 6.2_+0.2 h 10.7.+1.1 / Hal•DO I

~ " b 72±0.3 • ' - ' ,, ~+,, " ~ = b 5.4_+0.2, h 9.6_+0.9' I
t/ , b 7.0±0.3 i • h I~'R°+fl~u'~I~ n4-n P 4 b N A p172.4tf,18+2.1plat" am56.9,
') --'-
i b 7 3+0 3 • " a a m asa . . . . . . . . ' ~- I' -~ ' ~ A m" • NA ~am........ ;~--' '~'-]'
203t~one ~1_ 3 °
- ' " -" wr9 7_904 -- / "-p123.6+1.6, Nl~20.4tl~.2+0.2ptat
Makassar (~" • b 9.8_+0.3 " -" p1120-25 NA p1137.3tf # a y l
Strait ~ + b 10.0-~-0.4 ' "~-
b 7.6-+0.20 wr 4~7_3.1 I b 11_+0.4
b 12.5_+0.5 &
I • b9.6_+0.3 p1201_+9(0.10%K) [
b 12.8_+0.3 NA p1312tf; am 123.2tf /
. phi 17.5_+0.4~ Kalosi• "+^ • • I
• b 7.7-+0.2 . . . . ~ b 9.8_+0.3 wm ~4t_z A/A p 126.4tf; am 371tf |
• . , , - ' - - ' - " - ' ~ h ~ l m q / ~ n+n-~ b 7.2_0.3, sn 6.7_+0.3 pl 7.8_+0.8, b 7.3_+0.3
~A~T^~ u o. r_-lJ.,~,r~.o~-u.~ b 8.8_+0.3, pl 7.7_+0.8• b 7.1-+0.3
"~'~ } b7.6.+0.2 _-"I•.b8.1_+0.3 ii
b 8.05-_0.31 b 7.4~_0.'~2b~8.1_+0.3 p162_+13 (0.02%K)
sn Enrekang

0 25 50 km ~ b 8.5-+0.3 /
I I i I
0 25 mi
b 6.8+0.3

Pare P ~ ~ 7.8_+0.39b4.8_+0.2 I
4° I 4°
118 ° 30' 119 ° 120 ° 120 ° 30'

Fig. 6. K-Ar and 4°Ar-39Ar (A/A) age results location map. Age _+ 10 (Ma) follows phase dated: wm = white mica,
pl = plagioclase, see Table 4 for other abbreviations.

118 ° 30' 119 ° 120 ° 1 2 0 ° 30'

2 o30' .,~ ~j I 2 o30'

a 7.3+1.2, z 13.0.+0.6 a 2 1+0 5 z 6 4+0 4 s4 4+0 5 ~ _ . I

a 7.2.+1.0 ...... t ...... ; / " D ~ i,-,,-,,-, I
e ....... z 10.8+~6 -~-ea 8.4_+4.3 . . . . ~,v ,
D ivlu[ll~u +
3 ° -- / + 7RR+t3a _mea46.+0.6
.¥ zS~0.~ ••a8.7-+8.8 -B o n_e _
. . . . J ........ z 8.8_0.5 T a 6.2_+1.1, z 6.3_+0.4 - -
Makassar (,- z8.6_+0.6 | Bay
• ~ •
Strait ~ , a 20.4_+5.2 . z 4.3_+0.6 - 1
a 4.0+_2.9 a 3.8_+1.6 k %
/ a 9.2+_1.4, z 9.8+_0.5 a 2~.6+_6.5 t "
t • a 12.9.+3.8, z 92-+10 /
n 7 (~+1 9 7 . . . . . a 5.8.+0.9 Kalosi • • /
\ ,t~ u_~,iL, LZ.U_TV.'~ • a4.1-+1.0 a 8.2+9.5 z 93-+9
\ " f-l,,Ja 2.4_+1.8, z 6.8_+0.5 • '..a 5.5_+1.6, z 67.+8 l
a4.2_+06 z 5 0 + 0 3 s4.5_+0.9 ~ _~11+~~ Wjl~a6.3+~.3, z128_+12 /
\ -A. . . . . . . . . . . . . . . . . . . . ~ a, ,z, ,~ a 4.5+0 9 • / "z 123_+9
\,"v f- ~ z67_-',~'4• a9.9.+1.6 z69-+5

? a• \
25'r.i z##:, /

118 ° 30' 119 ° 120 ° 1200 30'

Fig. 7. Fission track age results location map. Central FT age + ]~ (Ma) fo|]ows p h a s e d a t e d : see T a b l e 4 f o r
abbreviations.
 402 S . C . B E R G M A N ET AL.

Table 3. Summary of igneous, metamorphic and sedimentary rock major and trace element geochemical data averages

number of (wt%)
Lithology samples SiO 2 TiO 2 A1203 Fe203 MgO CaO FeO FeO*

Miocene volcanic rocks 36 mean 57.05 0.70 13.83 3.42 4.02 6.43 2.15 5.23
std. dev. 9.70 0.20 2.50 1.67 2.60 4.72 1.41 2.06
Miocene intrusive rocks 19 mean 59.35 0.65 14.97 1.87 3.94 5.48 3.24 4.92
std. dev. 8.51 0.27 2.00 1.45 2.26 3.22 1.42 2.39
Cretaceous metamorphic rock 1 38.60 0.34 7.49 3.12 1.15 25.70 0.80 3.61
Lamasi Ophiolite 13 mean 53.14 1.11 15.52 3.60 5.05 8.30 6.05 9.29
std. dev. 5.36 0.49 1.86 1.11 1.68 3.72 1.80 2.24
Palaeogene clastic rocks 5 mean 63.88 0.48 11.91 2.69 1.43 6.54 2.40 4.36
std. dev. 18.94 0.19 4.21 1.70 0.97 10.50 1.45 1.58

(ppm)
Li Be Rb Cs Sr Ba Y La
Miocene volcanic rocks 36 mean 19 6 169 16 610 1225 24 56
std. dev. 10 3 104 19 372 824 9 35
Miocene intrusive rocks 19 mean 29 5 179 7 553 996 19 41
std. dev. 12 1 99 4 266 740 7 17
Cretaceous metamorphic rock 1 20 3 21 2 581 132 12 13
Lamasi Ophiolite 13 mean 6 4 10 2 191 104 28 5
std. dev. 5 1 13 1 182 139 15 4
Palaeogene clastic rocks 5 mean 30 4 35 6 277 221 14 12
std. dev. 20 1 22 4 261 134 11 5

(ppm)
Hf Nb Ta W Sc V Cr Co
Miocene volcanic rocks 36 mean 6.4 16.9 1.4 5.5 20 129 142 18
std. dev. 3.5 9.0 0.8 4.6 11 69 107 11
Miocene intrusive rocks 19 mean 4.8 13.5 1.0 2.3 18 125 124 20
std. dev. 1.5 5.0 0.2 1.2 13 87 99 13
Cretaceous metamorphic rock 1 1.5 7.0 < 0.5 < 1 10 92 34 9
Lamasi Ophiolite 13 mean 3.2 5.4 < 0.6 3.3 32 264 135 28
std. dev. 1.9 1.4 0.6 10 127 164 13
Palaeogene clastic rocks 5 mean 3.7 7.4 < 0.7 2.0 11 92 181 14
std. dev. 1.2 1.8 1.0 6 52 214 9

For complete data see table 2 of Supplementary Publication SUP 18099 (see text).

mica or whole rock separates define the cooling
histories of the major igneous complexes (sum-
marized in Table 4 and illustrated on a map in
Fig. 6; detailed data provided in table 3 of
Supplementary Publication SUP 18099 (see
above)). High temperature cooling ages (> 300°C)
of 38 Miocene and Pliocene Camba-Enrekang-
Mamasa Complex intrusive and extrusive rocks
overlap and range from 2.4-17.5 _+0.2-0.4Ma
(average 8 +_2.5 Ma). Where multiple phases were
dated, concordant ages are generally observed with
the exception of two samples which contain horn-
blende ages significantly older than their biotite
ages. This discordance may reflect the higher
blocking temperature of hornblende (500°C) than
biotite (300°C), suggesting protracted cooling in
the interval 500-300°C (see below).
Six conventional K-At plagioclase or whole rock
ages of seven samples of the Lamasi Complex
range from 20-201 Ma. The rocks possessing older
ages (162-201 Ma) contain 0.02-0.10 wt% K in
mineral separates or whole-rock splits and are
interpreted as anomalously high due to excess
radiogenic 4°Ar. In contrast, some Lamasi Complex
samples with higher K contents (0.2-0.5 wt%) give
younger K-Ar ages in the range 20-24 Ma and are
interpreted as reset cooling ages thus dating the
ophiolite emplacement. Two plagioclase separates
with moderate K contents (0.10-0.14wt%) give
Palaeocene or early Cretaceous ages (46_ 3 and
120 _+5 Ma) which are interpreted as a partially
reset higher temperature cooling events. Ten
4°Ar-39Ar ages of plagioclase or amphibole
separates from five rocks of the Lamasi Complex
are summarized in Table 4 (Bergman, unpublished
data). Unfortunately, the plagioclase and amphi-
boles possess very low K contents, yielding a
variety of ambiguous ages with a bewildering
and poor interpretation potential. Nevertheless,
three groups of ages are distinguished: excessively
old ages due to excess 4°At, high T (> 400°C) or
partially reset cooling ages, and low T (< 300°C)
cooling ages. The separates with among the lowest
K contents (0.001-0.06 wt%) possess the oldest
ages in the range 203-371 Ma. These are regarded
as erroneously old due to excess 4°At contamin-
 MIOCENE W SULAWESI CONTINENT COLLISION 403

MnO Na20 K20 P205 CO2 SO 4 S F C1 H 2 0 + H20- LOI Total Mg#

0.13 2.78 3.18 0.40 1.04 0.44 0.07 0.08 0.02 2.78 1.84 5.32 99.79 54.20
0.20 1.37 1.56 0.32 3.92 0.49 0.13 0.04 0.01 1.90 2.14 4.91 0.54 12.39
0.10 2.98 3.81 0.29 0.94 0.05 0.02 0.07 0.03 1.67 0.33 2.52 99.55 56.91
0.05 1.04 1.66 0.21 1.89 0.05 0.02 0.03 0.02 1.15 0.29 2.62 0.47 7.61
0.16 1.40 0.90 0.07 19.00 <0.1 0.02 0.02 < 0.01 1.90 0.70 20.60 101.37 36.22
0.18 3.55 0.35 0.17 0.05 0.16 0.07 0.02 0.02 2.30 0.35 2.21 99.80 48.43
0.04 1.30 0.56 0.13 0.04 0.10 0.05 0.02 0.01 0.83 0.27 1.08 0.27 9.08
0.18 0.73 0.92 0.07 4.45 0.20 0.21 0.03 < 0.01 3.32 1.16 8.83 99.94 34.45
0.14 0.97 0.62 0.02 8.15 0.10 0.36 0.02 < 0.01 1.85 0.94 9.00 0.88 13.92

Ce Pr Nd Sm Eu Gd ~ Dy Ho Er Tm Yb Lu ~ U Zr
108 12 44.9 8.7 1.8 7.3 0.9 4.9 0.9 2.4 0.3 2.2 0.3 32 7.8 230
68 7 25.9 4.7 0.9 3.8 0.3 1.3 0.2 0.6 0.1 0.5 0.1 24 5.0 127
79 8.4 32.1 6.0 1.3 5.1 0.7 4.0 0.7 2.1 0.3 1.8 0.3 20 5.4 1~
31 3.1 10.9 1.8 0.4 1.7 0.2 1.3 0.3 0.8 0.1 0.8 0.1 10 3.0 45
22 3 12.1 2.8 0.8 2.7 0.4 2.3 0.4 1.1 0.1 0.9 0.1 3 1.4 85
14 2.3 11.6 3.9 1.3 4.5 0.8 5.0 1.0 3.0 0.4 2.7 0.4 1 1.3 109
9 1.3 6.8 2.2 0.6 2.5 0.4 2.8 0.6 1.7 0.2 1.4 0.2 1 0.9 62
27 3.3 13.8 3.6 1.0 3.3 0.5 2.8 0.5 1.4 0.2 1.2 0.2 5 2.2 143
9 1.1 3.9 0.9 0.3 1.3 0.3 1.7 0.3 1.0 0.1 0.7 0.1 2 0.6 31

(ppb)
Ni Cu Zn As Pb Sn T1 Ag Sb B G Ge Au Pd Pt Hg
61 47 70 8.2 35.0 14.0 1.4 0.6 0.8 44 19 14 3.1 14.4 11.5 20
52 69 13 7.2 27.2 8.5 0.8 0.3 0.5 25 3 5 2.4 16.3 2.9 13
61 33 65 4.0 28.6 <9 1.3 0.5 0.5 30 20 11 4.1 21.7 11.0 52
45 39 12 4.0 16.5 0.8 0.1 0.7 21 2 2 3.4 15.7 2.9 97
19 30 62 8.1 2.0 9.0 0.1 0.4 0.4 43 9 14 <2 3.0 < 10 57
30 58 85 1.3 <2 10.8 < 0.3 0.3 0.3 24 20 16 <3 3.3 < 10 43
25 42 27 1.4 5.0 0.2 0.2 4 4 5 0.7 0.6 26
47 39 70 8.0 5.5 12.0 0.3 0.4 0.6 56 15 15 <2 3.6 < 18 82
40 46 21 6.4 3.0 5.9 0.1 0.1 0.1 18 5 2 1.5 78

ation b e c a u s e fresh a m p h i b o l e f r o m s a m p l e W h i t e m i c a f r o m a sandy m u d s t o n e o f Oligocene

93SUL43 yield total fusion and total gas ages o f
122-123 Ma, whereas plagioclase separates from
the same rock possesses a m u c h older set o f ages
(294-312 Ma). The high T (> 400°C) or partially
reset cooling ages fall into two groups: Early
Cretaceous (121-137 Ma) and Late Cretaceous to
P a l a e o c e n e ( 5 7 - 9 0 Ma) and are interpreted to
indicate the t i m i n g o f Cretaceous to Palaeogene
o p h i o l i t e c r y s t a l l i z a t i o n , and in s o m e cases
partially reset b y subsequent e x h u m a t i o n or thermal
events. A third group is interpreted as low T
(< 300°C) cooling ages ranging f r o m M i o c e n e to
P l i o c e n e ( 5 - 2 0 M a ) and p r e s u m a b l y date the
cooling f o l l o w i n g the obduction event. Therefore,
although the samples are a m o n g the m o s t difficult
for u n a m b i g u o u s 4°mr-39Ar age interpretation due
to variable alteration/metamorphism and very low
K-contents, yielding excess 4°Ar problems, we
believe that some limited useful age information
h a s b e e n p r o d u c e d , i n d i c a t i n g C r e t a c e o u s to
Palaeogene ophiolite crystallization f o l l o w e d by
E o c e n e to O l i g o c e n e o b d u c t i o n and M i o c e n e
exhumation.
d e p o s i t i o n a l age ( s a m p l e 9 3 S U L 5 3 ) y i e l d s a
114 +_ 2 M a K - A r age, indicating a M i d to Late
Cretaceous p r o v e n a n c e o f detrital white mica. This
114 M a age is virtually identical to those for the
P o m p a n g e o Schist C o m p l e x o f central Sulawesi
and the B a n t i m a l a C o m p l e x o f S W S u l a w e s i
(Parkinson 1991), which are likely candidates for
the p r o v e n a n c e c o m p o n e n t s o f the E o c e n e -
Oligocene clastic deposits o f Sulawesi. These and
other red marine or fluvial sandstones and mud-
stones o f O l i g o c e n e - E o c e n e age f r o m the Kalosi
area contain a m i x e d h e a v y mineral assemblage
including both granitic (rutile, ilmenite, zircon) and
ophiolitic-accretionary provenances (chromite)
( B e r g m a n u n p u b l i s h e d data), i n d i c a t i n g the
presence o f both continental granitic and oceanic
ophiolitic sources in the uplands o f the O l i g o c e n e -
Eocene depocentres. T h e distinctive red colour is
interpreted to result f r o m oxidation o f r e d u c e d iron
(probably derived f r o m the ophiolitic provenance)
which has been p r o m o t e d by weathering in a
tropical climate.
404 S.C. BERGMAN ET AL.

14- ~ olite

10 . . . . . . . .

/ /~kJTr~ite F•~.~ "~,,-~ '~'0 k \ // Rhyolite

• ......
K,O 6- Trachybasalt• " x~.._O~Z ~• ." ,~
(wt%) Tephrite/ I~1[ & I ~1~
Basanite In [ ,I I~
4- •,,~ Dacite ~ • •
\
2- • \
• Basaltic Andesite
'icrobasalt Basalt [ Andesite t ~.
0 I I I I I I I
40 45 50 55 60 65 70 75 80
O Miocene Intrusive Igneous Rocks SiC 2 (wt%) Pumices and Shards,
• Miocene Extrusive Igneous Rocks Electron Microprobe data
• Paleogene? Lamas• Complex Rocks Wt% data on a volatile free basis.

Fig. 8. Western Sulawesi igneous rock total alkalis v. silica (TAS) plot showing the nomenclature of common
volcanic lithologies (after Cox et al. 1979).

either the time of exhumation if the intrusives were

Fission track analysis
emplaced at depths > 3-4 km, or the late stage
Apatite (n = 28 samples), sphene (n = 2) and zircon crystallization history if they were emplaced near
(n = 23) fission track (FT) analyses of represen- the surface (< 1-2 km depth). For the volcanic
tative Mesozoic and Tertiary igneous, meta- rocks, these ages reflect the late stage crystalliz-
morphic and sedimentary rocks define their ation history or subsequent exhumation-related
provenance and thermal history (summarized in thermal history if they were buried to depths
Table 4, and detailed in table 4 of Supplementary > 3-4 km following deposition. Nearly concordant
Publication SUP 18099 (see above)). Zircon grains apatite, zircon and sphene FT age histograms and
in the igneous and sedimentary rocks are euhedral age spectra are illustrated for the Polewali pluton
to subhedral and primarily yellow-brown, but in Fig. 15. Extremely rapid cooling ( > 5 0 -
range from yellow to brown and red-brown. 100°C Ma -1) is indicated for the interval 60-250°C
Sphene grains are yellow-brown and subhedral. for this and other samples (Figs 15-16).
Apatite grains from the igneous rocks are mainly Apatite FT ages of Miocene intrusive and
euhedral, whereas those from the sedimentary extrusive units range from 1.9-11 Ma and average
rocks are euhedral and anhedral with rare rounded 6 -+ 2 Ma. Some vitric-crystal tuffs (e.g. samples
grains. The FT ages reflect the time elapsed since 93RS201 and 93RS279) contain identical apatite
cooling through the respective closure temperatures and zircon FT ages, defining the age of deposition
for sphene (250 _+50°C), zircon (200_ 25°C) and of the host tufts and therefore the age of the host
apatite (120 _+25°C) (Naeser & Faul 1969; Dodson sedimentary sections. Mean track lengths of
1973; Dodson & McClelland-Brown 1985; Green samples with statistically meaningful populations
et al. 1989). Some samples for which FT ages were (n > 30) range from 12.3-15.5 tam with widely
determined on several minerals show concordant varying standard deviations of 1.1-3.8 pm, mainly
FT ages, whereas others exhibit discordance, indicating rapid cooling (> 10°C Ma -t) through the
interpreted as rapid and protracted cooling, respec- temperature interval 60-130°C. Apatite uranium
tively. For intrusive rocks, these FT ages represent contents also vary widely from 2-71 ppm U.
the approximate time of cooling through 225°C Sphene FT ages range from 4.4_+0.5 to
(zircon) or 60-130°C (apatite) and may represent 4.5 _+0.9 Ma with U - 50-69 ppm. Zircon fission
 MIOCENE W SULAWESICONTINENT COLLISION 405

i , ,
10
o 8 A

O M i o c e n e Intrusive Igneous R o c k s ShoshoniteSeries g

• M i o c e n e Extrusive Igneous R o c k s o 1
• P a l a e o g e n e ? Lamasi I g n e o u s R o c k s o •" t.=A,•.%~-" o
4 o
~.-<: • --- ~-o Oa,c~,~a~,oeSeres
16 , , , , , ,
0
Alkaline Series ' _ ~ - -
O • ll . ....-
iA
O~o • • "" • 0 8,~

9 8 o ~A,mi s-~" UA lhA

t.. i~l.--12.~ • ,, I ~ A 4 +
(') 4 o • •
• •
mmm= No • l' Subl.alkaline or Tholeiitic Series •
0 0
,~,12
LI• • A

• (3 •
0.8
~ 0_,-~ 0
•~ •
O~oO, v

• IA
A~AI~., • -- 0 0.4 2:
• • _.mlI.
u A L~,rr~ "
l.m m~ i 0 •
l__l_ • A~ • o
0.0
' o . . . . . •

g ,o
• "o 2 = ._c•
1AAA
A=mo •
A
0.2
I~" • • ~ • llli-,l_'U •
A0 All _ml
l = m~ ll~ •
m~''° mlm "lm 0 "lmlO r-- -- 0.1 ~
g ~o i i)miil~ i
i
•

• • • l0
~ 5 0.0 ~
12
~ 1.5- ~,~m-° O~ :

"~1.0 ~ 0 ~ • I -- I
&li_illq) • i
m ,ln'-ii_
0.5 • ~-~l~ l~ • • r~imw--
llcm~91 • •
4~
ol m • m0
0.0 , , , , , , , i
0
45 50 55 60 65 70 75 80 45 50 55 60 65 70 75 80
SiO 2 ( w t % ) Si0 2(wt%)

Fig. 9. Western Sulawesi igneous rock Harker variation diagrams showing various major element oxide compositions
plotted as a function of silica content.
406 S . C . BERGMAN ET AL.

FeO* (FeO + 0.9 Fe203 )

~ TholeSeriesl~
ite
• \

y" - o ,, \

• ....- .., ..

• 1~ Calc-AlkSeri
alinees __~
\
0 MioceneIntrusiveIgneousRocks \
• MioceneExtrusiveIgneousRocks _.~
kVI A, Pal:eogene?L:masi ComplexRocks kv/ ~/ ~./

~/ k/
Na20 + K20 MgO
Fig. 10. Western Sulawesi igneous rock AFM ternary diagram (wt%) showing the trends defined by a tholeiitic
series, represented by the Tonga-Marianas-S Sandwich Island suites and a calc-alkaline series, represented by rocks
from Cascades, North Chile, and New Guinea (after Brown, 1982). Dashed line shows boundary between tholeiitic
and calc-alkaline (and alkaline) series (after Kuno, 1968).

track ages for Miocene igneous rocks range from
4.3-10.8 Ma; U = 171-582 ppm. Apatite electron
microprobe compositions vary widely in anion
abundance; C1 contents in several samples
approach 2-3 wt% (table 4 of Supplementary
Publication SUP 18099 (see above)), indicating
higher than normal apatite blocking temperatures
(140 ° v. 120°C) are probably applicable (Green
et al. 1989).
Apatite FT track ages of five sandy non-marine
or marine mudstones with Oligocene-Eocene
depositional ages from the Kalosi area are in the
range 6-13 Ma, indicating the Palaeogene section
now exposed at the surface in the Kalosi area was
buried to temperatures > 120°C and cooled through
60-120°C during the Late Miocene. Zircon FT ages
of these five Oligocene-Eocene mudstones are in
the range 67-93 Ma, suggesting a Late Cretaceous
provenance and that these samples had not been
heated to > 225°C since deposition. Zircon FT ages
of two Latimojong Complex metasedimentary
rocks are 123-128 Ma, with a 6.8 Ma apatite FT
age, indicating an early Cretaceous provenance
and Miocene cooling event due to thrust-related
exhumation. The mostly Late Miocene apatite
fission track ages of igneous, metamorphic, and
sedimentary rocks from western Sulawesi are
interpreted to suggest the time of rapid
(> 10°CMa -]) cooling through the temperature
interval 130-60°C, probably due to thrust-related
exhumation associated with compressional
tectonism.
Rb-Sr and Nd-Sm isotope data
87Sr/86Sr and 143Nd/144Nd ratios and Rb, Sr, Nd
 MIOCENE W SULAWESI CONTINENT COLLISION
and Sm concentrations were determined on 32
representative whole rock samples with the purpose
of constraining the nature of crustal or mantle
magma provenance and the possible presence of
earlier melting events than the most recent Miocene
and older Lamasi magmatism (summarized in
Table 5, and detailed in table 5 of Supplementary
Publication SUP 18099 (see above)). In addition,
three Palaeogene and Cretaceous sedimentary
samples were analyzed to determine provenance
and test the possibility that near surface sedi-
mentary units represent a possible contaminant to
the igneous rocks. 87Sr/86Srttnltlal)
....... values of the
Miocene intrusive and extrusive rocks overlap,
both groups ranging from 0.703-0.74, and are
in general positively correlated with Rb/Sr ratios
(Fig. 17). The lack of strong positive correlation
between 87Sr/86Sr and 1/86Sr values indicates that
the correlated 87Sr/86Sr and Rb/Sr values are not
the result of mixing of two unique end members
(Fig. 18). Palaeogene to Cretaceous sedimentary
rocks can also be eliminated as potential contami-
nants of the igneous rocks because the sedimentary
and metamorphic samples do not form isotopic
end-members of the trends defined by the Miocene
igneous rocks. The similarity in Miocene intrusive
and extrusive isotopic ratios and ages indicates that
the two groups are members of a cogenetic suite of
lithospheric melts. 87Sr/86Sr values are not corre-
lated with major and trace element composition,
such as silica and alkalis, further suggesting that
mixing of two compositional end members can be
eliminated in explaining the major element and
isotopic compositional variation. The entire
Miocene suite representing samples from a 500 ×
300 km area defines a Rb-Sr pseudo-isochron
(Fig. 17) with an age of 305 Ma (initial 87Sr/86Sr =
0.7094, excepting the anomalous sample
90SUL28) whereas four separate samples from the
Palopo pluton define a whole-rock isochron of
506 _+21 Ma (initial 87Sr/86Sr = 0.7073). The rela-
tively high initial 87Sr/86Sr values suggest melting
of relatively old, enriched mantle and crustal litho-
spheric material. Because the Palopo pluton
possesses Late Miocene biotite and hornblende
K-Ar ages (300-500°C cooling ages = 6-11 Ma),
the 506 _+21 Ma isochron indicates early Palaeo-
zoic crustal rocks were partially melted, yet not
totally homogenized, by a Miocene melting event.
U-Pb zircon data (see below) substantiate this
conclusion of Miocene recycling of Palaeozoic or
older crustal materials.
The material with the most depleted 87Sr/86Sr
values and enriched 143Nd/144Nd values are from
the Lamasi Complex samples and are distinct from
the Miocene igneous rocks. On a Rb-Sr isochron
plot, a slight positive correlation is observed and
a maximum homogenization age of 50 Ma can be
407
inferred from the limited range in 87Sr/86Sr with
a large range in Rb/Sr.
143Nd/144Nd(initial) ratios of Miocene igneous
rocks range from 0.5119-0.5131 and Sm/Nd values
are in the range 0.094-0.25, 143Nd/144Nd values
exhibit a strong negative correlation with 87Sr/86Sr
values and fall mainly in the field of continental
lithospheric melts (Fig. 19). The western Sulawesi
data form a trend ranging from the most depleted
ratios of the Palaeogene or older Lamasi Complex
ophiolites, which are similar to MORB and nearby
Celebes Sea and Sulu Sea Eocene ocean floor
basalts, to the more enriched ratios of the Mamasa
granite and related rocks. The Late Miocene
Buakayu alkali gabbro sill (sample 90SUL3)
possesses a slightly depleted set of Sr & Nd iso-
topic ratios which indicate a contribution from
asthenospheric mantle. Besides this gabbro and
the samples of Lamasi Complex, all other samples
indicate derivation from relatively enriched litho-
spheric parental materials with Nd-model ages of
1-2 Ga. In contrast, Lamasi Complex rocks were
derived from melting of an extremely LIL-element
depleted, time integrated low Rb/Sr and high
Sm/Nd asthenospheric source mantle, typical of
mid-ocean ridge basalts or back-arc basin basalts.
143Nd/144Nd values of Miocene igneous rocks
exhibit a slight positive correlation with
147Sm/144Nd values (Fig. 20), defining a Nd
pseudo-isochron age of c. 0.6 Ga. 143Nd/144Nd
values are not correlated with 1/Nd, other trace
element, silica or alkali contents, suggesting that
mixing of two compositional end members can be
eliminated in explaining the major element and Nd
isotopic compositional variation. Depleted-mantle
Nd model ages vary from 0.8-2.3 Ga and are
not correlated with silica content or other major
element composition. The 143Nd/144Nd and
147Sm/144Nd values of Lamasi Complex rocks are
positively correlated and define a 178 Ma Nd
pseudo-isochron (initial 143Nd/144Nd ratio =
0.51278). On a Rb-Sr isochron plot, a slight
positive correlation is observed and a maximum
homogenization age of 50 Ma can be inferred from
the limited range in 87Sr/86Sr ratios despite a large
Rb/Sr variation, possibly indicating Paleogene
pertubation of the Rb-Sr system.
The Sr and Nd isotopic ratios of the three Eocene
to Cretaceous sedimentary and metamorphic
samples are in the ranges 87Sr/86Sr = 0.705-0.707
and 143Nd/144Nd = 0.5123-0.5126, overlapping
the western Sulawesi Miocene igneous rock trend.
These compositions are consistent with the view
that the Palaeogene to Cretaceous sedimentary and
metamorphic rocks possess a mixed provenance
which includes depleted mantle array-like oceanic
or calc-alkaline volcanic rocks and more enriched
crustal components.
Table 4. K-Ar, 4°Ar-39Ar and fission track age data summary

sample location/area lithology K-Ar phase & age (Ma) FT phase & age (Ma) Ar-Ar phase & age (Ma)

90SUL-2 Bua Kayu arkosic sandstone a 3.8__.1.6

90SUL-3 Bua Kayu alkali gabbro sill b 12.5___0.5
90SUL-4a Bua Kayu syenogabbro sill b 11__.0.4 a 20.4_+5.2
90SUL-6a Bua Kayu basalt crystal-lithic lapilli tuff z 4.3_+0.6
90SUL-7 Bua Kayu trachyandesite tuff b 10.0+_0.4 z 8.6_+0.6 .~
90SUL-8 Palopo monzodiorite pluton b 6.8_+0.3, h 6.0-+0.3 a 2.1-+0.5, z 6.4_+0.4, s 4.4_+0.5
90SUL-9 Palopo qz monzodiorite pluton b 6.2-+0.2, h 10.7_+1.1 z 5.7_+0.4
90SUL-10a Palopo hornfelsed andalus siltstone a 8.4_+4.3
90SUL-11 Palopo basalt block in tuff breccia a 7.7_+1.2
90SUL-12 Palopo qz monzodiorite pluton b 5.4_+0.2, h 9.6_+0.9 a 6.2_+1.1, z 6.3_+0.4
90SUL-13 Palopo monzodiorite pluton h 8.7_+0.9 z 10.8_+0.6
90SUL-14 Enrekang trachyandesite lava flow b 7.1-+0.3 z 6.7-+0.4
90SUL- 15 Enrekang trachyandesite lava flow b 7.2_+0.3, sn 6.7-+0.3 a 4.5_+0.9
90SUL-16a Enrekang dacite tuff p 7.8_+0.8, b 7.3-+0.3 "~
90SUL-18b Enrekang dacite tuff b 8.8_+0.3, p 7.7_+0.8
90SUL-20 Enrekang trachyandesite lava flow b 8.1_+0.3 z 6.6_+0.5
90SUL-21a Mamasa qz monzodiorite pluton b 7.7_+0.3 a 7.3+1.2, z 13.0-+0.6
90SUL-21b Mamasa Quat alluv/Mamasa pluton a 7.2_+1.0
90SUL-22a Mamasa qz monzodiorite pluton b 7.2_+0.3 z 6.6_+0.5
90SUL-22b Mamasa diorite in pluton b 7.0_+0.3 a 4.6__.0.6
90SUL-23 Mamasa dacite porphyry b 7.3_+0.3 z 6.8_+0.4
90SUL-25 Polewali qz monzonite pluton b 6.1_+0.3, h 5.8_+0.6 a 4.2__.0.6, z 5.0-+0.3, s 4.5-+0.9
90SUL-26a Pare Pare andesite tuff breccia b 8.7-+0.9 z 6.1_+0.4
90SUL-26b Pare Pare crystal lithic andesite tuff z 6.2_+0.4
90SUL-26c Pare Pare andesite clast in tuff breccia b 6.8_+0.3 a 5.9_+0.8
90SUL-26d Pare Pare andesite clast in tuff breccia b 7.8-.+0.3 z 5.8_+0.4
90SUL-27 Pare Pare trachyte flow/dome b 4.8___0.2 a 6.9___1.5
90SUL-28 Enrekang gran. orthogneiss clast b 8.1-+0.3 a 1.9-+0.5
RAG90-73 Bua Kayu volcaniclastic sandstone a 24.6_+6.5
RAG90-74 Bua Kayu siliciclastic siltstone a 4.0_+2.9
92SUL-1 Biloka River trachyandesite lava b 7.6_+0.2
92SUL-3 Biloka River trachyandesite lava b 7.4_+0.2
92SUL-6 Waru trachyandesite tuff b 8.0_+0.3
92SUL-7 Balu trachyandesite lapilli tuff b 8.5_+0.3
92SUL-8 Lambou/Kalosi sandstone a 12.9_+3.8, z 92_+10
RAG92-4 Mamuju trachyte tuff sn 5.4-+0.2
RAG92-6 Mamuju leucite shoshonite lava sn 5.3-+0.2
RAG92-9 Mamuju leucite shosh, vitric crystal tuff b 2.4-+0.1
ND92-29 Tirasa/Benteng trachyandesite lava b 7.3_+0.2
93SUL3 Loca divide basaltic lapilli tuff breccia all_+l.2
93SUL5 Kanang/Polewali equigranular granodiorite b 9.8-+0.3 a 4.1-+1.0
93SUL6a Kanang/Polewali silicified mudstone a 2.4-+1.8, z 6.8_+0.5
93SUL9 Polewali lamprophyre phi 17.5_+0.4 a 5.8_+0.9
93SUL10 Polewali basaltic tuff block breccia b 12.8_+0.3
93SUL13 Minanga equigranular granite b 7.6_+0.2
93SUL15 Busu andesite lava b 9.8_+0.3
93SUL17 Pappang andesite blocks/tuff breccia wr 45.7+_3.1
93SUL20 Lemo basaltic sill wr 9.7-+0.4
93SUL27 Lembang amphibolite/gabbro a 8.7_+8.8 p 72.4tf,18_+2.1plat; am
56.9tf
93SUL28 Buntu Saragi basaltic andesite lava am 203tf
93SUL30 Bua basaltic andesite lava p 20.0-+1.0 p 15.6tf,12.2_+0.2plat
93SUL35 Bakabaka massive pillow basalt lavas p 23.8-+1.6 p 20.4tf, 5.2_+0.2plat
93SUL38 Batupapan sheared basaltic greenstone p 120-+5 p 137.3tf
93SUL42 Cilallang rhyolite block in breccia wr 23.8-+1.2
93SUL43 Bonelemo sheeted gabbro dikes p 126.4tf; am 371tf
93SUL45 Salu Ranteballa sheared equigranular diorite p 201_+9 (0.10%K) p 312tf; am 123.2tf,
93SUL46 Barulatong rhyolitic clasts in breccia p 162_+13 (0.02%K)
93SUL47 Langae chloritic phyllite z 123_+9
93SUL48a Dea calcareous augen/greenschist a 6.3_+2.3, z 128_+12
93SUL49 Raada light green sandstone a 9.9__.1.6, z 69_+5
93SUL52a Panjura coarse grained sandstone a 5.5_+1.6, z 67-+8
93SUL53 Saruran sandy mudstone wm 114+_2 a 8.2-+2.5, z 93-+9
93RS201A Matama trachyte tuff b 7.7+_0.2 a 7.0_+1.2, z 7.0-+0.4
93RS279A Limboro trachyte tuff b 9.6_+0.3 a 9.2-+1.4, z 9.8_+0.5

b, biot; p, plag; h, hbl; a, apatite; z, zircon; am, amphibole; sn, sanidine; wr, whole rock; s, sphene; tf, total fusion; plat, plateau; phi, phlog.
410 s . c . BERGMAN ET AL.

Table 5. Summary of Rb-Sr and Nd-Sm isotope data for I I I I I I I I I I I I

igneous, metamorphic and sedimentary rocks OIB
N-Type MORB
E-Type MORB
sample lithology 87Sr 143Nd
code 86Sr*i 144Nd*i average REE abundances
100 ~ normalized
90SUL3 Mi(m) 0.70643 0.51274
90SUL4A Mi(m) 0.71234 0.51212
90SUL5 Mv(m) 0.71391 0.51192
90SUL6A Mv(m) 0.71235 0.51207
90SUL7 Mv(i) 0.7 1551 0.51206 10
90SUL8 Mi(i) 0.71449 0.51205
90SUL9 Mi(f) 0.71610 0.51202
90SUL10A Km(p) 0.70787 0.51253
90SULll Pv(m) 0.70335 0.51301
Miocene Intrusive Igneous Rocks
90SUL 12 Mi(f) 0.71094 0.51220
90SUL13 Mi(f) 0.71200 0.51227 1
La Is" Ig l'y I
90SUL14 My(f) 0.72126 0.51214 0.315
Ce
0,1
Nd
0.192
Eu
0.259
Tb
0.325
He
0.213
Tm
0.209
Lu
90SUL18B Mv(f) 0.71679 0.51218 0.813 0.597 0.0722 0.049 0.072 0.032 0.0323
90SUL20 Mv(f) 0.71217 0.51229
90SUL21a Mi(f) 0.71973 0.51215
90SUL22a Mi(f) 0.71921 0.51194
90SUL23 Mv(f) 0.72213 0.51215
90SUL24 Mv(f) 0.72210 0.51216 100 chondrite normalized
90SUL25 Mi(f) 0.71324 0.51215
90SUL26C My(i) 0.71066 0.51225
90SUL26D Mv(i) 0.70996 0.51226
90SUL27 Mv(i) 0.70717 0.51251
90SUL28 Mv(i) 0.73909 0.51221
93SUL1 Mv(m) 0.70852 0.51246 10
93SUL9 Mi(m) 0.71645 0.51217
93SUL27 Pi(m) 0.70300 0.51304
93SUL28 Pv(m) 0.70317 0.51305
93SUL30 Pv(i) 0.70342 0.51294
93SUL43 Pi(m) 0.70360 - 1
93SUL45 Pi(i) 0.70316 0.51310 La I Pr I Sm I Od I Dy I Er I Yb
0.315 0.1 0.192 0.259 0.325 0.213 0.209
93SUL49 Ps(p) 0.70662 0.51262 Ce Nd Eu Tb He Tm Lu
93SUL52A Ps(p) 0.70559 0.51230 0.813 0.597 0.0722 0.049 0.072 0.032 0 . 0 3 2 3
t I I I I I I I I I I I I
....... OIB
Age: M, Miocene; P, Palaeogene/Cretaceous; K, N-Type MORB
Cretaceous. Rock type: v, volcanic; s, sediment; E-Type MORB
i, intrusive; m, metamorphic. Composition: (i) inter- ~ average REE abundances
mediate; (m) mafic; (f) felsic; (p) pelitic. *i, initial value; 100 "-,, chondrite normalized
whole rock samples in all cases.

Zircon U/Pb data

10
U/Pb dating was attempted on zircon and sphene
separates from six representative rocks with the
purpose of testing the hypothesis that an inherited
crustal component is present in the Miocene
magmatic rocks (Table 6). All samples exhibit Palaeogene? Lamasi Complex Rocks

varying degrees of discordance, several (samples 1

90SUL12, 15, 21, 25 and 26c) less so than others
•a I
0.315 O, 1
I0.192
I o' I'y
0.259 0.325 0.213
I
0.209
I'b
Co Nd Eu Tb He Tm Lu
(sample 90SUL28). Zircons from samples 0.813 0.597 0.0722 0.049 0.072 0.032 0.0323

90SUL12, 15, 21, 25 and 26C possess similar

2°6pb/238U and 2°Tpb/235U ages of 5-13 Ma,
broadly similar to the K-Ar and fission track
ages on the same samples. In contrast, 2°Tpb/2°rpb
ages range from 88-1073 Ma for this material.
Sample 90SUL28 displays Palaeozoic to
Proterozoic U/Pb and Pb/Pb ages and defines a
Fig. 11. Western Sulawesi igneous rock rare earth
element (REE) profiles, normalized to chondrite
compositions (Hanson 1978) showing trends observed
for ocean island alkali basalts (OIB), normal mid-ocean
ridge tholeiitic basalts (N-MORB) and enriched mid-
ocean ridge tholeiitic basalts (E-type MORB) (after Sun
& McDonough 1989).
 MIOCENE W SULAWESI CONTINENT COLLISION
chord on the concordia diagram that possesses
an upper intercept of 391 Ma, similar to the Rb-Sr
whole rock model ages for the same sample and the
entire suite of Sulawesi Miocene granitoids and
andesites. Sphene U/Pb data indicate concordant
2°6pb/238U and 2°Tpb/235U ages for sample
90SUL25 (6 Ma), yet much younger U/Pb ages
of 1.0-1.3 Ma compared to its biotite K-Ar age
(8.1_+0.3 Ma) and similar to its apatite fission
track age (1.9 _+0.5 Ma).
Discussion
Petrogenesis
Mid to Late Miocene mafic, intermediate and felsic
volcanic and intrusive rocks are abundant and
widespread in western Sulawesi, indicating exten-
sive lower crust and upper mantle melting during
the interval 2-18 Ma (mean and mode 8 _+4 Ma).
The Pliocene to Miocene intrusives and extrusives
possess overlapping age and major, trace element
and Sr-Nd isotopic compositions, indicating a
cogenetic volcanoplutonic belt or province. These
Miocene igneous rocks form a bimodal compo-
sitional suite, with basaltic and dacitic compo-
sitional subsets predominating, and generally
lacking intermediate andesites. This suggests two
distinct sources of magmas: a more mafic, probably
peridotitic parent for the basaltic rocks and less
mafic, probably basaltic parental material for the
dacites and compositionally similar intrusives. The
sources of these Miocene magmatic rocks, there-
fore, range from enriched (in LIL and other incom-
patible elements) ancient upper mantle peridotite
lithosphere, which melted to produce the alkaline
basaltic sills, dykes, and pyroclastic sequences in
the region, to Palaeozoic or older lower crustal
rocks of basaltic to intermediate composition,
which melted to produce the rhyolite-dacite tufts
and quartz monzonite to monzodiorite stocks,
plugs and plutons. Because an asthenospheric
mantle source does not appear to be important in
the petrogenesis of the Miocene basaltic melts
(except sample 90SUL3), based on the radiogenic
isotopic and enriched trace element compositions,
classic subduction-related melting processes were
probably not involved in the formation of the
Miocene basalts. The Miocene granitic rocks
(quartz monzonite to monzodiorite) are relatively
alkalic and silica-poor compared with granodiorites
of subduction-related arcs, and the most likely
mechanism f o r their formation involves litho-
spheric melting instead of subduction-related
melting of a mantle wedge overlying a Benioff
zone. The western Sulawesi igneous rocks are
notably distinct in major element and isotopic
composition from arc-continent collision-related
411
magmatism in the East Sunda-Banda Arc (van
Bergen et al. 1993). Furthermore, andesitic compo-
sitions form a minor component of the Miocene
suite of western Sulawesi. Instead, a bimodal
assemblage of alkaline basaltic and dacitic compo-
sitions predominate. The few andesites (e.g.
Parepare) are relatively enriched in 87Sr
(87Sr/86Sri = 0.710) compared with typical conti-
nental subduction-related andesites (87Sr/86Sr =
0.705-0.707; Gill 1981; Ewart 1982) and are more
compositionally allied to the group III late
collisional melts of Harris et al. (1986) in terms of
radiogenic 87Sr/86Sr, low Rb/Zr, and presence of
mafic and felsic melts. The most likely mechanism
for generation of both basaltic and granitic material
in western Sulawesi involves mantle and crustal
lithospheric thickening resulting from accretionary
processes, which produced progressive heating
of the lithosphere, and finally melting during the
Miocene (see Jaupart & Provost (1985), Delisle
(1986), England & Thompson (1986) and
Karabinos & Ketcham (1988) for relevant details
of collisional magmatism thermal modelling). It is
also possible that a lithospheric delamination
process (Bird 1979; Kay & Kay 1991, 1993; Davies
& van Blanckenburg 1995) was involved in the
melting process.
Trace element geochemistry of western Sulawesi
Miocene igneous rocks indicates that they are not
allied directly with normal subduction-related
processes. Spider diagrams suggest similarity to
subduction-related rocks from volcanic arcs
associated with compressional boundaries in which
oceanic crust is subducted beneath continental crust
(Fig. 12). Trace element discrimination diagrams
proposed for granitic rocks (Fig. 13) yield
ambiguous results: intermediate between syn-
collisional granites and volcanic arc granites. The
compositions of western Sulawesi intrusives
largely cluster near the granite minimum melt
eutectic at 650-670°C, in equilibrium with rela-
tively Ca-poor plagioclase (An4_10). This lends
further support to the view that the granitic melts
represent partial melts of continental crust. Using
discrimination criteria based on mineral abun-
dances (classification of Lameyre & Bowden
1982), the western Sulawesi granitic rocks are
intermediate between normal alkaline and calc-
alkaline series granites (calc-alkaline granodioritic
and monzonitic series). Harris et al. (1986) divided
collision zone magmatism into four types based on
the relative timing of magmatism and tectonism:
groups I to IV for pre-, syn-, late- and post-collision
intrusions. In their classification, the western
Sulawesi Miocene magmatic rocks are group III
late-collision calc-alkaline to alkaline intrusions,
based on structural relations and compositional
traits. The granitic melts are distinct from typical
412 S. C. BERGMAN ET AL.

10000

Western Sulawesi Miocene Extrusive Rocks

•~ 1000

.~ 100

E
o
Z 10
I
c

1
I OIB
. m
I E-Type MORB • High K Andesite
E t N-Type MORB Calc-Alkaline Andesite
•= 0.1
I • Low K Andesite
average abundances
[] Collision Granite
primitive mantle normalized
© Volcanic Arc Granite
0.01

1oo0oj I
Cs Rb Ba Th U K Ta Nb La Ce Sr Nd P Hf Zr Sm Ti Tb Y
o.188o.6356.989o.085 0.021 250 o.014 0.7130.6871.775 21.1 1.354 95 0.1066 11.2 0.4441300 0.108 4.55

Miocene Intrusive Rocks

._o 1000
gc
N 100 t~ , ,,-
• \

Z 10

.t- O. 1
n t • Low K Andesite

0.01
Cs Rb Ba Th U K Ta Nb La Ce Sr Nd P Hf Zr Sm Ti Tb Y
0.1880.6356.9890.085 o.o21 250 0.o14 0.7130.6871.775 21.1 1.354 95 o.1066 11.2 0.4441300 0.108 4.55

continental crust melts associated with continental
collision, such as the Himalayan belt (Dietrich &
Gansser 1981), which are formed by leucogranites
significantly more enriched in SiO 2 and A1203,
although of similar 87Sr/86Sr, relative to the western
Sulawesi granitoids. This contrast may be a
function of contrasting lower crust lithologies in the
two collision zones: the Himalayan source is pelitic
or granitic, whereas the western Sulawesi Miocene
igneous parental rocks are intermediate to mafic
 MIOCENE W SULAWESI CONTINENT COLLISION 413

10000 l I I I I I I I I I I I I I I I I I L
F
Palaeogene? Lamasi Complex Rocks I
1000

", ,~ a'- ~',,~ t'-, I

100 QcE.~'" /LV',,~., q -, ~-

10

I i • High K Andesite I:
0.1 I/ 'I~ORB ~ Calc-Alkaline Andesite
• Low K Andesite
average abundances [] Collision Granite
0.01 I

Cs Rb Ba Th
primitive mantle normalized
I I I

U
I

K
I

Ta
I

Nb La
I I

Ce Sr
o
I
Volcanic Arc Granite
!

Nd
I

P
I

Hf
I

Zr Sm Ti Tb Y
! I I F
0.1880.6356.9890.085 0.021 250 0.014 0.7130.6871.775 21.1 1.354 95 0.1066 11.2 0.4441300 0.108 4.55
Fig. 12. Western Sulawesi igneous rock 'spider' diagrams, normalized to primitive mantle trace element abundances,
showing trends observed for ocean island alkali basalts (OIB), normal mid-ocean ridge tholeiitic basalts (N-MORB),
enriched mid-ocean ridge tholeiitic basalts (E-type MORB), and various representative andesite and granite
compositions (after Sun & McDonough 1989; Pearce et al. 1984; Ewart 1982).

meta-igneous assemblages. Priadi et al. (1993)
presented trace element and St, Pb, and Nd isotopic
data on basement peridotites, granulites and
amphibolites from Palu (Central Sulawesi, north
of the present study area) which demonstrated
an Australian lithospheric character. However, the
Palu Sr and Nd isotope data fall in distinct fields
and define contrasting trends compared with the
Miocene magmatic rocks of the present study and
the Palu basement rocks lack the isotopic and trace
element compositions required to form a significant
contaminant or parental contributor to the Miocene
igneous rocks of western Sulawesi.
The Lamasi Complex ophiolites form a distinct
compositional group and should be considered
separately from the Miocene igneous rocks. The
Lamasi rocks are tholeiitic in composition with
relatively flat REE to LREE-depleted patterns,
in contrast to Miocene calc-alkaline and alkaline
assemblages which possess strongly LREE-
enriched signatures. The geochemistry of the
Lamasi ophiolites are similar to ocean floor basalts
and tholeiitic subduction-related (primitive arc) and
back-arc basalts in major, trace element, 87Sr/86Sr
and 143Nd/144Nd composition. These Lamasi
Complex ophiolites were obducted during the Late
Oligocene and Early Miocene based on new K-Ar
and 40Ar-39Ar data and probably indicate that the
eastern Sulawesi ophiolite (Silver et al. 1983a, b)
extends into western Sulawesi. Other members
of the Lamasi Complex, otherwise known as the
Palaeocene-Eocene Langi Volcanic Formation or
the 'Older Andesites' of van Bemmelen (1949)
and the Papayato Volcanic Formation of Ratman
(1976) may represent part of these Cretaceous-
Palaeogene(?) ophiolitic assemblages. Mubroto
et al. (1994) presented K-Ar data on the Balantak
ophiolite of East Sulawesi which indicate
Cretaceous to Eocene crystallization ages, similar
to the Lamasi Complex volcanics in the present
study. Nishimura et al. (1980) presented gravity
data for South Sulawesi which show the existence
of a gravity high east of the Bone Mountains,
consistent with the area underlain by ophiolite. The
Lamasi volcanic complex could represent a
fragment of Indian Ocean crust, similar to that
thought to be trapped in the Banda Sea (Lapouille
et al. 1986), or, less likely, more proximal forearc
or back-arc oceanic crust related to the Sunda
subduction zone.
414 S.C. BERGMAN ETAL.

i i i i lllll i i i i iiiii [ i i i iiiii i l i i I i i i i I I I I I I I t I I I i i i i I I I I I I I I I I

Syn-Collision 1.0 -- • 0 Miocene Intrusive Igneous Rocks --

1000 -
Granites • Miocene Extrusive Igneous Rocks
L a m a s i Complex Rocks
at P a l a e o g e n e ?

• Extrusive O
O Intrusive O ¢ WithinPlate/
~1~{ 'I Gra~ o.s- ~-,, ,,,~. • -
100 ",4;. "4 "o

n-
.$3
Q.

lO
Volcanic
Granites
Arc

O n Ridge
~00_ i o ",

o •",,i
", ,,

c "'.,,"',
•

Granites o

"4 •

-05 - V@~ , , --

' ' ''""1 ' ' '""1 ' ' '''"'1

10 1O0 1000 Field of Normal ",,
Y + Nb (ppm)
, , ,,,,,,I , , ,,,,,,I , , ,,,,,,I

1000 -'J \
k • Extrusive Within Plate -1,0 45
I I I I
&
I I I I
5/5
I I I I I 1 I I
60 IJ51 ' '
I I
70 .... 715
I I I I
80

1 ~ ntrusive Granites SiO2(wt%)

Fig. 14. Western Sulawesi igneous rock Cat/alkali

v. S i t 2 plot showing the composition of SW Sulawesi
rocks relative to typical calc-alkaline subduction-related
melts (after Pearce 1983).

Q.

10 IVo,cani rc .
Granitesand
Syn-Collision
Granites I1~ I/
I
I
o ~"

1 . . . . . . . .
I I0
Fig. 13. Western Sulawesi igneous rock trace element
discrimination diagrams for granitic rocks, showing the
compositional fields defined by granitic rocks of various
tectonic associations (after Brown 1982).
Thermal history
The K-Ar and fission track geochronology data
place firm constraints on the thermal history of
western Sulawesi. The age data are plotted on
time-temperature diagrams in Fig. 16, in which the
blocking temperature of each phase is used. The
cooling histories of different rocks in each area are
similar, and the five areas possess nearly identical
cooling histories, with the exception of several
sandstones from Buakayu, which possess much
older fission track ages. The cooling trajectories
form two distinct gradients: an earlier phase of
rapid cooling (> 100°CMa -1) in the interval
> 500-100°C, and a later phase of more protracted
cooling (at 5-20°CMa -1) from 100-20°C. The
OceanRidge
Granites
earlier cooling phase is probably associated with
magmatic cooling, which typically occurs at
> 100°C Ma -], and the latter phase with denuda-
tion, typically at rates of 5-20°C Ma -]. The
thermal history of the area is one of rapid emplace-
ment of magmatic rocks and eruption of volcanic
I . . . . . . . . I . . . . . . . . I
rocks during a short term episode in the Mid to Late
1 O0 1000
Y (ppm) Miocene (3-18 Ma) followed by rapid cooling of
these melts in the upper crust. Late Miocene to
Pliocene cooling was more gradual (5-20°C Ma-1),
although rapid enough to produce relatively long
confined apatite fission track length distributions
(> 14 ~tm mean track lengths), and was probably
associated with thrust-induced uplift and erosion.
Several sandstones from the Buakayu area possess
some of the oldest apatite FT ages (11-25 Ma,
relative to depositional ages) of all samples studied.
The detrital apatites possess significantly older FT
ages than their corresponding depositional ages,
suggesting the rocks have not been heated to
> 120-150°C since deposition.
The best constrained crystallization, cooling and
unroofing history of an individual pluton occurs
at Palopo. Four samples from 475 m, 835 m,
1050 m, and 1070 m elevations (samples
90SUL12, 13, 9 and 8) exhibit a range in horn-
blende A1 contents (averaging 5.7, 12.9. 4.9 and
10.4 wt% A1203; Bergman unpublished data) and
predicted crystallization pressures, averaging 1.7,
7.6, 1.0 and 5.7 kbar, respectively (calibration of
Schmidt 1992; Hammarstrom & Zen 1986;
 MIOCENE W SULAWESI CONTINENT COLLISION 415

elevation) crystallized at pressures of 1.7 _ 0.5 kbar

ZIRCON (5-6 km subsurface depths). Combining these
90SUL25 inferences with the fission track and K-Ar
>`6 Polewali Pluton geochronology data for the Palopo pluton and its
t- contact aureole, its cooling and uplift history can be
~4 constrained. The Palopo parental granitic magma
O"
was formed in the lower crust during a Miocene
I-i-2. melting event, ascended toward the surface, and
intruded the Cretaceous deposits at depths of
3-5 km during 6-10 Ma and rapidly cooled through
500-300°C by 6-8 Ma. The pluton finally cooled
i I through 60-120°C and ascended through 2-3 km
0 5 10 15 20
Apparent Fission Track Age (Ma) depths by 2-3 Ma, presumably during denudation-
related uplift and erosion associated with thin-
SPHENE skinned thrusting.
90SUL25
>,6"
c.) Polewali Pluton Tectonic framework
t-"

The Miocene plate tectonic framework of Sulawesi

has been the subject of ongoing vigorous debate
(Katili 1978; Sukamto 1978; Hamilton 1979, 1989;
U- 2 .
Van Leeuwen 1981; Hutchison 1982, 1989a, b;
Silver et al. 1983a, b; Wood 1985; Nishimura 1986;
Leterrier et al. 1990; Letouzey et al. 1990; Rangin
I i
0 5 10 15 20 et al. 1990; Audley-Charles 1991; Audley-Charles
Apparent Fission Track Age (Ma) & Harris 1991; Daly et al. 1991; Rangin & Silver
1991). Most workers envision a Miocene west-
APATITE ward-dipping subduction zone to produce the
>6 m
90SUL25 widespread Miocene volcanic and plutonic rocks
Polewali Pluton in western Sulawesi. The present study, however,
indicates Palaeozoic to Proterozoic lithospheric
~4
O"
parental rocks unlike any known in Sundaland (see
Hutchison 1989a for details of Sundaland) were
melted to produce the Miocene igneous rocks. Such
Palaeozoic to Proterozoic lithospheric assemblages
characterize the northern Australian Plate, such as
in western Irian Jaya (Pieters et al. 1983; Pigram
I I
0 5 10 15 20 & Panggabean 1984) and derivative continental
Apparent Fission Track Age (Ma) blocks including Sula, Buru, Seram and the Tukang
Besi Platform. Silurian or Permian granitic rocks
Fig. 15. Fission track age histograms for apatite, sphene form part of an Early Palaeozoic continental
and zircon from the Polewali pluton. assemblage and extensive Miocene collision of arc
terranes and associated crustal melting are known
along the northern margin of NE Australia-New
Hollister et al. 1987; Rutter et aL 1989). The Guinea (Audley-Charles 1991). This similarity to
presence of andalusite porphyroblasts within the the inferred crustal parental material of the
Palopo pluton contact metamorphic aureole Miocene magmatic rocks of western Sulawesi
(sample 90SUL10, elevation 250m) limits the suggests that a crustal fragment derived from the
intrusion pressure to < 3.8 kbar, the aluminosilicate northern Australian plate was accreted onto SE
triple point. This indicates that the elevated Sundaland during the Miocene due to westward
pressures inferred for samples 90SUL13 and 8 vergent plate motion of the Pacific plate or was
probably reflect earlier phases of crystallization in transported along and juxtaposed against SE
the cooling history of the pluton, possibly during Sundaland by a transform margin. It is also possible
magma ascent through the lower and middle crust that the accreted crustal block was one of the
(15-23 km). The top of the pluton (sample 90SUL9 allochthonous Pacific microplates which charac-
from 1050 m elevation) crystallized at pressures terize eastern Indonesia (Hutchison 1989a, b;
of 1.0 __.0.5 kbar (3-4 km subsurface depths), Audley-Charles 1991; Audley-Charles & Harris
whereas the lower-most sample (90SUL12, 0.5 km 1991). Early Miocene emplacement of the East
416 S . C . BERGMAN E T A L

AGE (Ma)
25 20 15 10 5 0
20
E °4e°e° ~
rein100 ~ ~

AGE (MliI
25, 20 15. 10. 5. 0 /~ AGE (Ms) 0
20 ~ ,,.,,:.~400~ J 25 20 15 10 5

100 ~" ~ o.Conll~ P_hy!Ute • "...........

... oo o,gn.~,. ~ooLT .."'"'~"" T T
Ip i i
~',,.,0e + PALOPO 200~ lind Turblditel
<~ PLUTON (a +
/ -
Ul

M .AMAS.A BA'I:H O LI';'H s00 o Igneous

SmdztonnRocks

PAREPARE IGNEOUS ' BUA I(AYU '

ENREKANG

a o

,oo o"~" ," ". • I i.

0 50 100rni
.SO0 0 SarDdmon~
Igneous Rocks
2s io l's 1'o ~ o 2s ~o l's 1'o ~ o
AGE (Ma) AGE (Ma)

Fig. 16. Sulawesi igneous and sedimentary rock time-temperature plots summarizing new K-Ar and fission track age
data and thermal history trajectories.

Sulawesi ophiolite marked the onset of the Sula
Platform accretion event (Audley-Charles 1987;
Silver et al. 1983a, b; Parkinson 1991). Litho-
spheric thickening following obduction of a crustal
block produces melting at the base of the crust and
formation of granitic magma at c. l0 Ma after
initial crustal thickening, based on thermal
modelling by Delisle (1986), which nicely matches
the obduction and magmatism age relations
observed in Sulawesi. Continued west-vergent
compression led to crustal thickening beneath the
south arm of Sulawesi, culminating in Late
Miocene lithospheric melting, extensive volcanism,
and Late Miocene to Pliocene thin-skinned thrust
faulting (see Fig. 21) in a manner broadly similar
(albeit of a smaller scale) to the Eocene accretion of
the Indian plate onto Eurasian plate and the subse-
quent development of the Himalayas. The Pliocene
framework was characterized by regional deforma-
tion and uplift along a 10 000 km long belt on
the northern margin of Australia and in much
of Indonesia. Fold and thrust belts of Alpine-
Himalayan scale developed during the last 5 Ma in
New Guinea, Sulawesi, the Moluccas and Banda
Arc. Interestingly, the rapid uplift inferred for
South Sulawesi since 5-8 Ma on the basis of apatite
FF data is also suggested for two other major
fold-thrust belts, the Papuan fold belt (Hill &
Gleadow 1989) and the Mt Everest area of the
Himalayan orogen (Bergman et al. 1993) using
 MIOCENE W SULAWESICONTINENT COLLISION 417

0.725 I I I
A Palaeogene or Mesozoic? Lamasi Ophiolitic Rocks . . - " ""
+ Miocene Palopo Intrusives I-'"'i
O Miocene Intrusive Rocks
• Miocene Extrusive Rocks ..-"'""
0.720 Palaeogene and Cretaceous (~) ~(3~°~ ' c )
SedimentaryRocks '~"~J .-'" ~.~K.~./.ce~-"

0 2 4 6 8 10
, , , , ,
0.715 -
.... ;4,~,- ,~--',/
87Sr/ 86Sr ~ O ~~ 'e ' - m~;~''" 0.74 - .,""
present 0 -
'~/ ....-""'" ~.~
0.710 -
(0.70937 i) - oCxf77 f" f
(0.7073i)
0 HI m
0.705 -
A o7,-

0.700 I i
1 2 3
87Rb/ 8eSr
Fig. 17. Western Sulawesi igneous rock isotope plot: 87Sr/86Srv. Rb/Sr.

apatite and zircon b-rI' data. The Pliocene structural
evolution of South Sulawesi is probably not com-
posed of compressive deformation alone; Miocene
collision-related nappe emplacement in Timor and
Seram produced post-collisional uplift and nappe
attenuation by low and high angle normal faulting
(Harris 1989). The presence of the Lamasi ophiolite
complex of western Sulawesi and analagous easterri
arm ophiolite separated by the deep intervening
Bone Bay suggests orogenic collapse may be
occurring here as well. Seismic reflection data
across the Bone Bay show the existence of an N-S
trending, east-dipping normal fault system
accommodating > 4 km of post-Miocene clastic
sedimentary fill.
The lithospheric melts formed in the lower crust
were transported to the upper crust along well
established zones of crustal weakness, probably
reactivated fault zones (Fig. 21). Crystallization
took place in the upper crust within 5-10 km of
the surface as intrusive masses of various sizes and
shapes, including narrow dykes and sills 2-50 m
across, plugs and stocks 0.1-1 km in diameter, and
large plutons, laccoliths, and batholiths 20-100 km
in diameter. Some batholithic-scale laccoliths (e.g.
Mamasa) were later detached from their site of
crystallization and structurally transported west-
wards to the surface during the last 6-8 Ma along
low angle thrust faults.
The thermal consequences of crustal accretion,
magmatism and thrust faulting are variable and
offsetting. These include a reduction in subsurface
heat flow due to attempted subduction of relatively
cool continental crust, increased heat flow due to
the emplacement of large plutons and batholiths
into the upper crust, and the development of
inverted thermal gradients due to thrusting of hot
sheets over cold sheets (see England & Thompson
1986 and references therein for more details of the
latter process). Present-day heat flow values are
largely lacking for Sulawesi (Thamrin 1985, 1986;
Matsubayashi & Nagao 1991); the only available
data are for the western coastal area of western
Sulawesi and range .from 10-48 mW/m 2. The
authors suggest an elevated regional thermal
regime during Late Miocene and Early Pliocene
magmatism and thrusting, with presumed regional
heat flows in the range 50-90 mW/m 2. The large
volume (> 100-1000km 3) of granitic melts
emplaced in the upper crust would produce the
largest thermal effects of regional importance
due to conductive cooling of 650-700°C melts,
accentuated by convective and advective heat
transfer by fluids.
The new geochronology and structural data place
f'trm constraints on the Neogene tectonic evolution
of the Makassar Strait. Instead of most previous
models which define it as a rift-related tectonic
418 s.c. BERGMAN ET AL.

0.725

© m m 072°73
Jto'~o"l....
J' ~ .....
0.720 -
© 0 0.71
1• •
t,•~ ~: :~
0.70 1 , , , I . . . . . . . .
mm 0.000 0.008 0.016
0.715 -
87Sr / 86Sr
© <so>
00 m
initial m
0.710 - Omm O Miocene Intrusive Igneous Rocks
• Miocene Extrusive Igneous Rocks
m • Palaeogene? Lamasi Complex Rock:
m o Palaeogene and Cretaceous
Sedimentary Rocks
0.705 -
A A A A

0.700 t I I
0.000 0.002 0.004 0.006 0.( )08
1/Sr

Fig. 18. Sulawesi igneous rock isotope plot: 87Sr/86Srv. l/Sr.

MORBasalts I I I I I I I
(asthenosphere) I ~ ~.2~- -lO
o.o by crustal residenceage

0 •5130 -~ ,-d~
sulu- k ~ ' , ,
O Miocene Intrusive Igneous Rocks
Celebes lx-I ', • Miocene Extrusive Igneous Rocks
^ M ',~,,.i Java -5
_~ea ',, f-~ • Palaeogene? Lamasi Complex Rocks
Mantle',, I",.l(-') ~ Palaeogene and Cretaceous Sedimentary Rocks
Array " " ~ " t ~
Z 0 -o
z
"~" 0.5126 -_ Bu'k_Eah_
rt~""_',,~l
',~a nda Arc - C

- O
"~ 1.0 by crustal residenceage
Z Q.
- -5 IJ.i

C-
oj -10

m (3
2.o by crustal residenceage -15
0.5118 t t t t I I
0.700 0.710 0.720 0.730 0.740
87Sr/86Sr initial

Fig. 19. Western Sulawesi igneous rock isotope plot: 143Nd/144Ndv. 87Sr/86Sr (fields after Whitford & Jezek 1982;
Serri et al. 1991).
 MIOCENE W SULAWESI CONTINENT COLLISION 419

0.5132 I I I I I I I
,. omptex Rocks ~k
, o ene Larnas~C . .j&~

0.5127 -
•
Z
~ ~,oC%~,~so,~
T--
-o~ 0.5122-
Z
a-
-o m
0.5117]
J
O Miocene Intrusive Igneous Rocks
• Miocene Extrusive Igneous Rocks
• Palaeogene? Lamasi Complex Rocks
Palaeogene and Cretaceous Sedimentary Rocks
0.5112 I I I I I I
0.00 0.04 0.08 0.12 0.16 0.20 0.24
147Sm/144Nd

Fig. 20. Western Sulawesi igneous rock isotope plot: 143Nd/144Nd v. Sm/144Ndisochron plot showing two
pseudoisochrons for Cretaceous-Palaeogene Lamasi ophiolites and Miocene igneous rocks.

element (e.g. Hamilton 1979), the central Makassar
Strait appears to be a Neogene foreland basin,
bounded to the west by the E-vergent Kutei Basin
fold belt and to the east by the W-vergent western
Sulawesi thrust belt (Fig. 22). In the western part of
the study area, the Majene foldbelt extends offshore
into the Makassar Strait where regional multi-
channel seismic data show frontal folds and thrusts
actively developing below 2000 m of water (Fig.
4). Further west, within the deepest parts of the
Makassar Strait, undeformed, horizontally strati-
fied Late Neogene sediments extend to the easterly-
directed toe thrusts associated with the Mahakam
Delta.
In many parts of western Indonesia, the sites of
major Palaeogene sediment accumulation (many
related to rifting) were inverted during Late
Miocene and Pliocene times in response to the
Palawan, Bangga-Sula and Australian continental
plate collisions. The resulting post-collisional
Neogene sedimentary sequences were deposited
within more tectonically quiet platform regions
preserved between the zones of active inversion.
Such structural inversions have been well docu-
mented especially along the eastern Sunda Shelf
at the southern entrance to the Makassar Strait
(Letouzey et al. 1990).
The inversion episode in western Sulawesi was
particularly intense, mainly because of its location
perpendicular and close to the main centre of
collision. Extensive imbrication of individual
thrust sheets collectively resulted in thrust loading
which depressed the adjacent lithosphere of
easternmost Sundaland to produce a major fore-
land basin. The erosion product of the western
Sulawesi uplift (termed Walanae Formation in this
study or 'Celebes Molasse' by van Bemmelen
1949) is actively being transported westwards
and deposited in the central Makassar Strait (North
Makassar Basin). To date, this foreland basin
remains only partially filled. Preliminary flexural
modelling of the cross section shown in Fig. 22
(Steve Crews pers. comm. 1994) demonstrates
that the lithospheric load of the western Sulawesi
thrust belt and the Mahakam-Delta is sufficient
to explain the general topographic framework
of the North Makassar Basin and western
Sulawesi.
In eastern Kalimantan, a similar although
easterly directed Late Neogene sediment-thrust
loading event took place in association with the
development of the Mahakam Delta, which
includes additional compressional structures
related to toe-of-slope.gravity sliding (Ott 1987;
van de Weerd & Armin 1992). The dominant,
present-day structural trend includes a series of
tightly folded NNE-SSW trending anticlines and
synclines observable onshore at the surface and
Table 6. Zircon and sphene U-Pb isotopic and age data

age
ratio ratio ratio ratio
weight (mg) 206pb 2o6pb 207pb 2o7pb 206pb 207pb 207pb
sample # % loss Pb U Pb
initial final (ng) (ppm) (ppm) 2o4pb 238U 235U 2o6pb 238U 235U 206pb
(Ma) (Ma) (Ma)

Zircon
90SUL-12 54.69 41.88 23 37 853 0.9 412 8.4174e-4 6.0236e-3 5.1901e-2 5.42 6.1 281
90SUL-15 49.95 38.84 22 40 764 1.1 542 1.2770e-3 8.5312e-3 4.8450e-2 8.22 8.63 121 t,~
90SUL-21A 57.11 38.02 34 38 890 1 888 6.85 7.48 217
90SUL-25 31.61 23.19 27 37 962 1.6 733 7.84 12.73 1073
90SUL-26C 18.37 10.41 43 6 432 0.6 212 1.0015e-3 6.5980e-3 4.7780e-2 6.45 6.68 88.4
90SUL-28 29.33 17.09 42 646 1279 38 1233 2.9190e-2 2.2290e- 1 5.5389e-2 186 204 428
unsized+M3 ° 9.68 4.78 49 196 1335 41 534 178 194 398
M4 ° + M 6 ° 5.78 1.48 76 30 958 29 298 178 171 -2894
391 Ma = upper intercept concordia age
Sphene
90SUL-25 57 285 241 4.9 22 6.1 6.19 40
90SUL-28 42 63 1240 1.5 256 1.27 1.15 -248
Fig. 21. Schematic E - W cross-section (present-day geography) illustrating the important points of the new tectonic model for western Sulawesi (after Coffield et al. 1993). Note
that the westward-verging orogen of South Sulawesi is formed over a back-thrust or retro-arc thrust system in the hanging wall of a westward-dipping subduction zone. The t,~
Miocene volcanics are derived from melting of the subducted continental lithosphere of Australian affinity. See Fig. 1 for location of sections.
422 S . C . BERGMAN ET AL.

i)(
÷

T O M I N I BAY

4:" fl
/

EAST

LEGEND

ORMATION8 THEMAHAJ(AM
mm ~ = ~ ,
DELTA

OP}~:)UTESOF FOLD& THRUST qq~ I~RUSl~

BELT80Ir W.
'~ NORMALFAULTS
WRENCHFAULT8
PLATFORM OTHERGRANITI~R ~. FOLDAXEtOFTHE
SEDIMI~rm
KtrrB 8ASlH
r ~? 'r 'r T ~ IIA1~METRY

KUTEO.ASl. MAKASSA. S T . A I T WESTE..SULAWESl .ONS A'

& Thrust B,Ito)
Miihelmm Delta)
~n km
5

i°
T
H -S

-10

l"om~v~l SIMPLIFIED STRUCTURAL CROSS-SECTION | ~ ~ !

Fig. 22. Simplified geological map and NW-SE cross-section through the Makassar Strait showing the main
Neogene stratigraphic and tectonic elements.
 MIOCENE W SULAWESI CONTINENT COLLISION
identifiable offshore as far east as the 200 m
isobath (Courteney et al. 1991).
The combined crustal effect of these two con-
verging fold and thrust belts, together with their
associated erosion products, is presently manifested
as the deep-water Makassar Strait. The North
Makassar Basin is, therefore, not the site of a
Tertiary spreading centre, but the result of NW to
SE shortening in response to a Neogene continent-
continent collision that took place in western
Sulawesi (Fig. 22). A unified Late Cenozoic
tectonic evolution of western Sulawesi should
include a phase of obduction of the Lamasi
ophiolite during the Palaeogene, followed by a
Late Miocene continent-continent collision of an
Australian Craton-derived block with Sundaland
with associated fold and thrust-belt development
and extensive lithospheric melting-related
magmatism.
Conclusions
The integration of existing data with new geo-
chemistry, petrology, geochronology and field data
lead to the following major conclusions.
(1) Miocene calc-alkaline to mildly alkaline and
shoshonitic felsic to mafic plutonic and
cogenetic volcanic rocks dominate the surface
and shallow subsurface geology of western
Sulawesi, forming over 70% of the area.
Intrusive rocks include biotite-hornblende
granitic rocks ranging from monzonite and
quartz monzonite, to quartz monzo-
diorite and gabbro.
(2) Dominant lithofacies include airfall and sub-
aqueous tufts, tuff turbidites, lapilli tufts,
tuff breccias, and volcanic conglomerates.
These lithotypes form stratovolcano deposit
sequences and volcanic apron successions
> 5 km in thickness. Domes and lavas rarely
occur. Intrusive bodies form sills, dykes,
stocks, plugs, plutons and small laccolithic
batholiths ranging from several metres to
100 km in maximum plan dimension.
(3) Intrusive rocks are temporally, geochemically,
and isotopically indistinguishable from
volcanic rocks in the region and both are
considered members of the same cogenetic
suites. These Miocene igneous rocks form a
bimodal compositional suite, with basaltic and
dacitic compositional subsets, with minor
intermediate composition andesites.
(4) Tectonic framework classifications based on
trace element discrimination diagrams indicate
the western Sulawesi Miocene igneous rocks
are transitional between volcanic arc and
423
collisional granitoids. However, the authors
urge caution in application of these classi-
fication schemes owing to lithospheric source
compositional heterogeneities.
(5) New K-Ar ages of mineral separates from
volcanic and intrusive rocks are indistinguish-
able and range from 2-18 Ma, averaging
8 _+4 Ma. Apatite, zircon and sphene fission
track ages exhibit the same general relationship
and are slightly younger than the K-Ar ages
(2-14 Ma), and average 5 _+2 Ma. A two stage
cooling history consisting of an initial
magmatic cooling phase at 100-400°C Ma -1,
followed by a secondary cooling phase related
to denudation at c. 5-20°C km -1 is indicated
by the geochronology data. Assuming geo-
thermal gradients of 30°C km -1 for the most
recent denudation cooling phase, these cooling
rates correspond to uplift rates of 200-
700 m Ma -1, rates typical of fold and thrust
belts such as Taiwan and the Himalayan chain.
These values are confirmed by mineral equi-
libria geobarometry determinations of selected
plutons, which indicate relatively shallow-
level intrusion of Miocene granitic laccoliths
and plutons at 3-10 km depths.
(6) 87Sr/86Srinitial values of Miocene igenous rocks
(range 0.7035-0.790; n = 23) exhibit a positive
correlation with 87Rb/86Sr, here interpreted
with age significance. Whole rock Rb-Sr
isochrons for an individual pluton, such as
Palopo (506 Ma) are distinct from the best
fit 'errorchron' or isochron of the entire suite
(305 Ma), yet both can be interpreted to
indicate a Palaeozoic 'closure' or 'homogeniz-
ation' event of the lithospheric source region of
the magmas. Whole rock 143Nd/144Nd values
range from 0.511-0.525 and are correlated
with Sm/Nd, indicating a possible isochronal
relationship with errorchron ages of c. 0.6 Ga.
Model ages of time since separation from
depleted mantle range from 0.9-2.5 Ga. Zircon
U/Pb age systematics of six representative
granitic intrusive and andesitic extrusive
samples possess nearly concordant 2°6pb/238U
and 2°7pb/235U ages in the range 5-12Ma,
similar to K-Ar biotite and hornblende ages
on the same rocks. In contrast, 2°7pb/2°6pb
ages are grossly discordant and are in the range
88-1073 Ma, suggesting the presence of
recycled ancient zircons derived from
Proterozoic to Palaeozoic lithospheric source
rocks in the Miocene granitic magmas.
Lithosphere of this character is absent in SE
Sundaland and is abundant in the northern
Australian plate; the most likely source of the
Miocene Sulawesi igneous rocks was
Australian plate lithosphere which became
424 s . c . BERGMAN ET AL.
accreted onto Sundaland during an Oligocene
to Miocene collision event.
(7) In contrast to the Miocene rocks, the Lamasi
C o m p l e x ophiolites possess the most depleted
87Sr/86Sr and enriched 143Nd/ln4Nd values, and
trace element systematics typical of mid-ocean
ridge basalts or back-arc basin basalts.
(8) The fission track, K-Ar and U-Pb ages, major
and trace e l e m e n t geochemistry, mineral
chemistry geobarometry and Rb-Sr/Nd-Sm/
U-Pb isotope c o m p o s i t i o n s of western
Sulawesi M i o c e n e m a g m a t i c rocks form
c o m p l e m e n t a r y and mutually i n d e p e n d e n t
datasets which constrain the following petro-
genetic features: western Sulawesi M i o c e n e -
Pliocene extrusive and intrusive rocks form
cogenetic volcanoplutonic complexes of calc-
alkalic to mildly alkalic and potassic, felsic
to mafic magmatic rocks which were erupted
and intruded during a relatively short episode
(2-18 Ma) of lithospheric t h i c k e n i n g and
melting, possibly involving lithosphere
delamination. Parental source rocks of the
M i o c e n e melts were Late Proterozoic to
Palaeozoic crustal and mantle lithospheric
assemblages which became heated and melted
due to a continent-continent collision in which
west vergent crust of the A u s t r a l i a n - N e w
Guinea plate was subducted beneath eastern-
most Sundaland. Regional e a s t - w e s t com-
pression continued into Pliocene to Recent
times and resulted in the development of a Late
Neogene fold and thrust belt with ramp-related
imbrication p r o v i d i n g e n h a n c e d regional
uplift rates of 200-700 m Ma -1 during the last
2-18 Ma.
(9) The effects of Late Neogene thrust loading in
western Sulawesi combined with sediment
loading in eastern Kalimantan, associated with
the d e v e l o p m e n t of the M a h a k a m Delta,
resulted in crustal subsidence along much of
the Makassar Strait. The North Makassar Basin
is interpreted as an actively forming foreland
basin located immediately in front of the
Majene foldbelt. The Neogene initiation of the
Makassar Strait subsidence represents an
important surface manifestation of the western
Sulawesi continent-continent collision
episode.
GRDC geologists Amarudin, Supandjono, Sukido, Trisna
and their colleagues provided much assistance in the
field. We thank Theo Van Leeuwen, Imans Kavalieris,
Maury Cucci, Scot Krueger, Charles Hutchison and
Mike Crawford for comments, discussions and insights.
R. Maury and C. Rangin provided preprints of publica-
tions during an early phase of this study. We thank Jeff
Corrigan and Marty Robinson for reviews of an earlier
version of the manuscript. We appreciate the meticulous
analytical work provided by Shaft Kelley, Ken Foland,
Fritz Hubacher, Long Liang, Tom Bills, Jenny Hopps,
Paul Green, Ian Duddy and Paul Mueller. We thank Steve
Crews for timely flexural modelling of the thrust belt.
Christopher Parkinson, Robert McCaffrey and Martin
Menzies provided most useful reviews, for which we are
extremely grateful. David Nicklin, Gene Richards, Jamie
Robertson, Suherman and John Duncan are thanked for
their support of this project, and we thank Pertamina,
ARCO and KG Kalosi for permission to publish this
research.
Appendix: analytical methods
Analytical work was performed at ARCO Exploration
Research Laboratories in Plano, Texas and at several
university and industrial laboratories. Detailed analytical
data are available from the senior author upon submission
of a formal written request.
Sample details. Samples, weighing 2-10 kg each, were
trimmed from Sulawesi surface outcrops. Representative
splits were subjected to a variety of analyses and mineral
separation. Modal analyses were performed on represen-
tative coarse-grained plutonic lithologies using a 0.4 mm
grid and an entire 27 × 46 mm thin section (n > 500
points).
Analytical techniques. Samples were crushed using a
conventional jaw crusher and disk mill, and biotite,
hornblende, sanidine, plagioclase, apatite, sphene and
zircon were concentrated using conventional heavy
liquid and magnetic methods at Krueger Enterprises,
Cambridge, MA.
Fission track analysis. The external detector method
(Naeser 1979; Gleadow et al. 1983, 1986) was used for
individual grain dating of apatite, sphene and zircon.
Apatite fission track age and length measurements were
made in the laboratories of Geotrack International,
Melbourne, Australia (G. Laslett & D. Arne, analysts).
Zircon, sphene, and some apatite FT analyses were
performed by S. Kelley in the laboratories of Southern
Methodist University Geology Department. Apatite
grains were mounted in epoxy, polished to expose grain
interiors, and etched for 25 seconds in 5 molar HNO 3
to reveal spontaneous fission tracks. Nonmagnetic,
> 3.1 gm/cc splits containing lexcessive pyrite were
treated with aqua regia to concentrate zircon. Zircon
grains were mounted in melted FEP Teflon tape, polished
and etched in a KOH-NaOH eutectic at 230°C tbr
3-6 hours. Since most samples contained multiple zircon
populations (based on morphology and etching character),
two mounts were made, and each etched for different
times to achieve the optimum etch for the full range of
grains. Sphene grains were mounted in FEP teflon tape,
polished, and etched in a 5.0 M NaOH solution at 125°C
for 4 hours.
Apatite, sphene and zircon mounts were covered with
low U muscovite detectors, apatites sandwiched between
standards including Durango apatite, NBS glass SRM962
and Corning CN-6 glass, and sphenes and zircons
between Fish Canyon and Myalla Road syenite zircon age
standards as well as the two glasses. Both packages were
 MIOCENE W SULAWESI CONTINENT COLLISION
irradiated at the Texas A & M reactor (sphene and zircon)
or the HIFAR reactor, Australia (apatite). Zircons and
sphenes were irradiated at a flux of 6.4-7.0x 1015
neutrons/cm 2, and apatites at .7-1.4 x 1016 neutrons/cm 2.
Reactor neutron fluence was calculated using accepted
ages of 27.9, 172.8 and 31.4 Ma for Fish Canyon, Myalla
Road and Durango standards and flux gradients were
accounted for using the glass standards. Muscovite
detectors were etched for 13 minutes in 48% HF to reveal
fission tracks resulting from induced fission of 235U
during neutron irradiation of apatite and zircon.
Individual grain ages were calculated using the
formulation of Price & Walker (1963); grain age un-
certainty calculations are those of Hurford et al. (1984).
Grain age uncertainties are quoted at the 1 sigma
confidence interval. Apatite confined track length
measurements were made using a Leitz microscope with
a 100x air lense by Geotrack, camera-lucida tube and
a digitizing tablet. Uncertainties are _.+0.2 pm. Only
horizontal tracks with well developed tips in grains with
prismatic faces were measured. Major and minor element
contents of dated apatites from selected samples were
determined to better interpret the FT age and length data
using wavelength-dispersive spectroscopic analyses on a
1983 JEOL 733 Superprobe at the ARCO Advanced
Analytical Services Laboratory in Piano, TX (L. Liang,
analyst). Natural mineral standards were used to convert
X-ray count rates to elemental concentrations; operating
conditions were: 15 kV potential, 15 nA beam current,
rastered beam 10-20 lain square and t0--30 second count
times; an interlayered Si-W crystal was used for F
determinations. Analytical uncertainties are _+ 1-3
relative % for major elements (> 10 wt%) and _ 5-10
relative % for minor elements (1-10 wt%) and > _+ 10-20
relative % for elements present in concentrations
< 1 wt%. OH was calculated by difference assuming
anion stoichiometry. Thermal modelling of apatite FT
data was performed using the kinetic models developed
by Laslett et al. (1987), Carlson (1990) and Corrigan
(1991) which are capable of predicting or inverting the
degree of length and age reduction as a function of time
and temperature.
References
ABENDANON, E. C. 1915. Geologische en Geographische
Doorkruisingen van Midden-Celebes (1909-1910).
E. J. Brill, Leiden, 1.
AVDLEV-CHARLES, M. G. 1977. Mesozoic evolution
of the margins of Tethys in Indonesia and the
Philippines. In: Proceedings 5th Indonesia
Petroleum Association Convention. 2, 25-52.
- 1987. Dispersal of Gondwanaland: relevance to the
- BERGMAN, S. C. 1987. Lamproites and other Potassium-
dispersal of the angiosperms. In: WHITMORE, T. C.
(ed.) Biogeographical Evolution of the Malay
Archipelago. Clarendon Press, Oxford, Oxford
Monograph, 4, 91-102.
- -1991. Tectonics of the New Guinea area. Annual
Reviews of Earth and Planetary Sciences, 19,
17-41.
- &HARRtS, R. A. 1991. Allochthonous terranes of
- pressional and extensional cooling and exhumation
the southwest Pacific and Indonesia. Philosophical
Transactions of the Royal Society of London, A331,
115-131.
425
K-Ar and 4°Ar-39Ar analyses. Conventional 4°K-4°Ar
dating was performed on whole rock or mineral concen-
trate samples in the laboratory of Geochron Inc.,
Cambridge, MA (Tom Bills and Jenny Hopps, analysts)
using conventional techniques (flame photometry for K
determinations and gas source mass spectrometry for 4°Ar
measurements). Mineral concentrates were seived to
-80/+200 mesh and whole rock and feldspar splits were
washed in dilute HF and HNO 3 to remove alteration
phases. Uncertainties were calculated using an empirical
formulation dependent on K-content and age. Two or
three splits were analyzed and averaged to generate the
observed ages. 4°Ar-39Ar thermal release analyses were
performed by Ken Foland and coworkers at the Ohio State
Geochronology Laboratories, Columbus, OH, using
conventional techniques.
Rb-Sr and Nd-Sm isotope analyses. Conventional
isotope dilution solid source mass spectroscopy was used
to determine 87Sr/86Sr and 143Nd/t44Nd ratios and Rb, Sr,
Nd and Sm contents in the geochronology laboratory of
Ohio State Dept Geology and Mineralogy, Columbus,
Ohio (F. Schubacher and K. Foland, analysts).
U-Pb zircon analyses. Zircon and sphene concentrates
prepared by Geochron were hand picked for purification
and further separated into magnetic splits using a Frantz
magnetic barrier separator. Conventional U-Pb solid
source mass spectroscopy was used to determine U-Pb
ratios, and U and Pb contents in the Geochronology
Laboratories of Univ. Florida, Dept Geology, Gainesville,
(P. Mueller, analyst).
Major and trace element geochemistry. Major element
contents were determined using conventional X-ray
fluorescence spectroscopy, minor and trace elements
using wet chemistry, ICP-MS, DCP, XRF, INAA, and
other techniques in the laboratories of X-ray Assay
Laboratories, Toronto. Analytical uncertainties are < 1-5
relative % for major elements (present in concentrations
> 10 wt%) and < 5-20 relative % for minor and trace
elements.
BELLON, H. & RANGIN, C. 1991. Geochemistry and
isotopic dating of Cenozoic volcanic arc
sequences around Celebes and Sulu Seas. In:
SILVER, E. A., RANGIN, C., VON BREYMANN,
M. Z., ET AL. (eds) Proceedings O c e a n
Drilling Program, Scientific Results, 124, 321-
338.
rich igneous rocks: a review of their occurrence,
mineralogy and geochemistry. In: FITTON, J.G. &
UPTON, B. J. G. (eds) Alkaline Igneous Rocks.
Geological Society, London, Special Publications
30, 103-190.
--, COFFIELD, D. Q., DONELICK, R., CORRIGAN, J.,
TALBOT, J., ET AL. 1993. Late Cenozoic com-
of the Qomolangma (Mt. Everest) region, Nepal.
Geological Society America Abstracts with
Programs, 25, A243.
426 S. C, BERGMAN ET AL.
BERRY, R. E & GRADY,A. E. 1987. Mesoscopic structures
produced by Plio-Pleistocene wrench faulting in
South Sulawesi, Indonesia. Journal Structural
Geology, 9, 563-571.
BIANTORO, E., MURITNO, B. P. & MAMUAYA,J. M. B.
1992. Inversion faults as the major structural
control in the northern part of the Kutei Basin,
East Kalimantan. In: Proceedings 21st
Indonesia Petroleum Association Convention. 1,
45-68.
BIRD, P. 1979. Continental delamination and the Colorado
Plateau. Journal of Geophysical Research, 84,
7561-7571.
BROWN, G. C. 1982. Calc-alkaline intrusive rocks: their
diversity, evolution, and relation to volcanic arcs.
In: THORPE, R. S. (ed.) Andesites, orogenic
andesites and related rocks. J. Wiley and Sons, New
York, 437-461.
BUROLLET, P. E & SALLE, C. 1981. Seismic reflection
profiles in Makassar Strait. In: BARBER, A. J. &
Wiryosujono, S. (eds) Geology and tectonics of the
eastern Indonesia. GRDC, Special Publication, 2,
273-276.
CAREY, S. W. 1976. Tectonic evolution of South East
Asia. In: Proceedings 4th Indonesia Petroleum
Association Convention, 2, 3-23.
CARLSON,W. D. 1990. Mechanisms and kinetics of apatite
fission track annealing. American Mineralogist, 75,
1120-1139.
COFFIELD, D. Q., BERGMAN, S. C., GARRARD, R. A.,
GURITNO, N., ROBINSON, N. M. & TALBOT, J. P.
1993. Tectonic and stratigraphic evolution of the
Kalosi PSC area, and associated development of a
Tertiary petroleum system. In: Proceedings 22nd
Indonesia Petroleum Association Convention.
679-706.
CORRIGAN, J. D. 1991. Inversion of apatite fission track
data for thermal history information. Journal
Geophysical Research, 96, 10347-10360.
COURTENEY, S., COCKCROFT,P., LORENZ,R., MILLER, R.,
OTT, H. L., ET AL. 1991. Indonesian Oil and Gas
Field Atlas, vol. V. Kalimantan, Indonesian
Petroleum Association, Jakarta.
Cox, K. G., BELL, J. D. & PANKHURST,R. J. 1979. The
interpretation of igneous rocks. Allen & Unwin,
London.
DALY, M. C., COOPER,M. A., WILSON, I., SMITH,D. G. &
HOOPER, B. G. D. 1991. Cenozoic plate tectonics
and basin evolution in Indonesia. Marine and
Petroleum Geology, 8, 2-21.
DAVIES, J. H. & VON BLACKENBURG, E 1995. Slab
breakoff: a model of lithosphere detachment and
its test in the magmatism and deformation of
collisional orogens. Earth and Planetary Science
Letters, 129, 85-102.
DELISLE, G. 1986. The subsurface temperature field
resulting from obduction of a crustal segment.
Geologische Jahrbuch, E34, 77-85.
DIETRICH, V. & GANSSER, A. 1981. The leucogranites
of the Bhutan Himalaya (Crustal anatexis versus
mantle melting). Schweizerische Mineralogische
und Petrogropische Mitteilungen, 61, 177-202.
DJURI & SUDJATMIKO1974. Geologic map of the Majene
and western part of Palopo quadrangles, South
Sulawesi, scale 1:250,000. Geological Research and
Development Centre, Bandung.
DODSON, M. H. 1973. Closure temperature in cooling
geochronological and petrological systems.
Contributions Mineralogy and Petrology, 40,
259-274.
& MCCLELLAND-BROWN, E. 1985. Isotopic and
paleomagnetic evidence for rates of cooling,
uplift, and erosion. In: SNELLING,. N. J. (ed.) The
chronology of the geological record. Geological
Society, London, Memoir, 10, 315-325.
EFFENDI, L. 1993. Selat Makasar merupukan wilayah
kompleks antara perairan Bagian Barat dan
Timur. In: Proceedings 22rid Annual Convention
Indonesian Association Geologists, 2, 950-961
(in Bahasa).
ENGLAND, P. C. & THOMPSON, A. 1986. Some thermal
and tectonic models for crustal melting in
continental collision zones. In: COWARD,M. P. &
RIES, A. C. (eds) Collision Tectonics. Geological
Society, London, Special Publications, 19,
83-94.
EWART, A. J. 1982. The mineralogy and petrology of
Tertiary-Recent orogenic volcanic rocks: with
special reference to the andesite-basalt compo-
sitional range. In: THORPE, R. S. (ed.) Andesites,
orogenic andesites and related rocks. J. Wiley and
Sons, New York, 25-87.
FAt;GERES, J.-C., GAYET, J. & GONTHIER, E. 1989.
Microphysiographie des depots quaternaires darts le
detroit de Makassar (ocean Indien). Opposition
entre une marge stable (Borneo, Kalimantan) et une
marge active (Celebes, Sulawesi). Bulletin Soci~t~
gdologique France, 8, 807-818.
GARRARD, R. A. 1989. Sengkang Basin, South Sulawesi.
Post-convention field trip guidebook, Indonesian
Petroleum Association, Jakarta.
--, NUSATRIYO, G. & COFFIELD, D. Q. 1992. The
prospectivity of early Tertiary rift sequences in the
Neogene foldbelts of South Sulawesi. Extended
Abstract. The Eastern Indonesia Symposium.
Petramina and Simon Petroleum Technology,
Jakarta.
GEORGE, W. 1981. Wallace and his line. In: WHITMORE,
T. C. (ed.) Wallace's Line and plate tectonics.
Clarendon Press, Oxford, 3-8.
GILL, J. B. 1981. Orogenic andesites and plate tectonics.
Springer-Verlag, New York.
GLEADOW,A. J. W., DUDDY,I. R. & LOVERING,J. E 1983.
Fission track analysis: a new tool for the evaluation
of thermal histories and hydrocarbon potential.
Australian Petroleum Association Journal, 23,
93-102.
--, GREEN, P. F. & LOVERING, J. E 1986.
Confined fission track lengths in apatite - a
diagnostic tool for thermal history analysis.
Contributions to Mineralogy and Petrology, 94,
405-415.
GREEN, P. E, DUDDY, I. R., LASLETT, G. M., HEGARTY,
K. A., GLEADOW,A. J. W. & LOVERING,J. E 1989.
Thermal annealing of fission tracks in apatite: 4.
Quantitative modelling techniques and extension to
geological timescales. Chemical Geology (Isotope
Geoscience Section), 79, 155-182.
 MIOCENE W SULAWESI CONTINENT COLLISION
HAMILTON, W. 1978. Tectonic map of the Indonesian
region. US Geological Survey, Map 1-875-D.
1979. Tectonics of the Indonesian region. US
Geological Survey, Professional Paper, 1078.
- 1989. Convergent-plate tectonics viewed from the
- KAVALIERIS,I., VAN LEEUWEN,T. M. & WILSON, M. 1992.
Indonesian region. In: SENCOR, A. M. C. (ed.)
Tectonic evolution of the Tethyan region. Kluwer
Academic Publishers, Dordrecht, 655-698.
HAMMARSTROM,J. M. & ZEN, E-AN 1986. Aluminum in
hornblende: an empirical igneous barometer.
American Mineralogist, 71, 1297-1313.
HARRIS, N. B. W., PEARCE, J. A. & T1HOLLE,A. G. 1986.
Geochemical characteristics of collision-zone
magmatism. In: COWARD,M. P. & REIS, A. C. (eds)
Collision tectonics, Geological Society, London,
Special Publications, 19, 67-81.
HARRIS, R. A. 1989. Processes of allochthon emplace-
ment, with special reference to the Brooks Range
ophiolite, Alaska and Timor, Indonesia. PhD Thesis,
University of London.
HASAN, K., GARRARD, R. & MAHODIM, P. 1991. South
Sulawesi, Post-convention field trip guidebook,
Indonesian Petroleum Association, Jakarta.
H1LL, K. C. & GLEADOW, A. J. W. 1989. Uplift and
thermal history of the Papuan fold belt, Papua New
Guinea: apatite fission track analysis. Australian
Journal Earth Sciences, 36, 515-539.
HOLLISTER, L. S., GRISSOM, G. C., PETERS, E. K.,
STOWELL,H. H. & SISSON,V. B. 1987. Confirmation
of the empirical correlation of A1 in hornblende with
pressure of solidification of calc-alkaline plutons.
American Mineralogist, 72, 231-239.
HURFORD, A. J., FITCH, E J. & CLARKE, A. 1984.
Resolution of the two age structure of the detrital
zircon populations of the Lower Cretaceous sand-
stones from the Weald of England by fission track
dating. Geological Magazine, 121, 269-296.
HUTCHISON, C. S. 1982. Indonesia: regional distribution
and characteristics. In: THORPE, R. S. (ed.)
Andesites, orogenic andesites and related rocks.
J. Wiley, London, 207-224.
- 1989a. Geological evolution of South-east Asia.
- Basin. Proceedings 22nd Indonesia Petroleum
Oxford University Press.
- 1989b. Displaced terranes of the southwest Pacific.
- MATSUBAYASHI, O. & NAGAO, T. 1991. Compilation of
In: BEN-AVRAHAM, Z. (ed.) The Evolution of the
Pacific Ocean Margins. Oxford University Press,
161-174.
JAUPART, C. & PROVOST, A. 1985. Heat focusing, granite
genesis and inverted metamorphic gradients in
continental collision zones. Earth and Planetary
Science Letters, 73, 385-397.
JEZEK, P. A., WHITFORD, D. J. & GILL, J. B. 1981.
Geochemistry of recent lavas from the Sangihe-
Sulawesi arc, Indonesia. In: BARBER, A. J. &
WmYOSUJONO, S. (eds) Geology and Tectonics of
Eastern Indonesia. Geological Research and
Development Centre, Bandung, Special Publication,
2, 383-389.
KARABtNOS, R & KETCHAM, R. 1988. Thermal structure
of active thrust belts. Journal of Metamorphic
Geology, 6, 559-570.
KATILI, J. A. 1971. A review of the geotectonic theories
and tectonic maps of Indonesia. Earth Science
Reviews, 7, 143-163.
427
1975. Volcanism and plate tectonics in the
Indonesian islands. Tectonophysics, 26, 165-168.
1978. Past and present geotectonic position of
Sulawesi, Indonesia. Tectonophysics, 45, 289-322.
Geologic setting and styles of mineralization, north
arm of Sulawesi, Indonesia. Journal SE Asia Earth
Sciences, 7, 113-129.
KAY, R. W. & KAY, S. M. 1991. Creation and destruction
of the lower continental crust. Geologische
Rundschau, 80, 259-278.
& 1993. Delamination and delamination
magmatism. Tectonophysics, 219, 177-189.
KUNO, H. 1968. Origin of andesite and its bearing on
island arc structure. Bulletin Volcanologique, 32,
141-17.
LASLETT, G. M., GREEN, E E, DUDDY, I. R. & GLEADOW,
A. J. W. 1987. Thermal modelling of fission tracks
in apatite: 2. A quantitative analysis. Chemical
Geology, 65, 1-13.
LAMEYRE, J. & BOWDEN, P. 1982. Plutonic rock types
series: discrimination of various granitoid series and
related rocks. Journal Volcanology and Geothermal
Research, 14, 169-186.
LAPOUILLE, A., HARTONO, H. M. S., LAROU, M.,
PRAMUNIJOYO, S. & LARDY, M. 1986. Age and
origin of the seafloor of the Banda Sea (Eastern
Indonesia). Oceanolica Acta, 81, 379-389.
LETERRIER, J., YUWONO, Y. S., SOERIA-ATMADJA, R. &
MAURY, R. C. 1990. Potassic volcanism in central
Java and South Sulawesi, Indonesia. Journal SE
Asia Earth Sciences, 4, 171-187.
LETOUZEY, J., WERNER, R & MARTY, A. 1990. Fault
reactivation and structural inversion. Backarc and
intraplate compressive deformations. Example of
the eastern Sunda shelf (Indonesia). Tectonophysics,
183, 341-362.
MALECEK, S. J., REAVES, C. M., SOERIA-ATMADJA,R. &
WIDIANTARA, K. O. 1993. Seismic stratigraphy of
Miocene and Pliocene age outer shelf and slope sed-
imentation in the Makassar PSC, offshore Kutei
Association Convention. 345-371.
heat flow data in SE Asia and its marginal seas. In:
CERMAK, V. & RYBACH, L. (eds) Terrestrial heat
flow and the lithosphere structure. Springer-Verlag,
New York, 444-456.
MUBROTO,B., BRIDEN,J. C., MCCLELLAND,E. & HALL, R.
1994. Paleomagnetism of the Balantak ophiolite,
Sulawesi. Earth and Planetary Science Letters, 125,
193-209.
NAESER, C. W. 1979. Fission track dating and geological
annealing of fission tracks. In: JAGER, E. &
HUNZIKER,J. C. (eds.) Lectures in Isotope Geology.
Springer Verlag, N.Y. 154-169.
- -& FAUL, H. 1969. Fission track annealing in apatite
and sphene. Journal of Geophysical Research, 74,
705-710.
NISH~URA, S. 1986. Neo-tectonics of East Indonesia. In:
Geological Society of China Memoirs, 7, 107-123.
--, YOKOYAMA,T. & HERRY 1980. Gravity measure-
ments at South Sulawesi. Phys. Geol. Indonesian
Island Arcs, 35-41.
428 S, C. BERGMAN ET AL.
OTT, H. L. 1987. The Kutei Basin-A unique structural
history. Proceedings 16th Indonesia Petroleum
Association Convention, 307-316.
PARKINSON, C. D. 1991. The petrology, structure, and
geologic history of the metamorphic rocks of central
Sulawesi. PhD Thesis, University of London.
PEARCE, J. A., HARRIS, N. B. W. & TINDLE, A. G. 1984.
Trace element discrimination diagrams for the
tectonic interpretation of granitic rocks. Journal of
Petrology, 25, 956-983.
PIETERS, P. E., PIGRAM, C. J., TRIAL, D. S., Dow, D. B,
RATMAN,N. & SUKAMTO,R. 1983. The stratigraphy
of W Irian Jaya. Bulletin Geological Research
Centre, Bandung, Indonesia, 8, 14--48.
PIGRAM, C. J. & Panggabean, H. 1984. Rifting of the
northern margin of the Australian continent and the
origin of some microcontinents in eastern Indonesia.
Tectonophysics, 107, 331-353.
PRIAD1, B., POLVE, M., MAURY, R. C., SOERIA-ATMADJA,
R. & BELLON, H. 1993. Geodynamic implications
of Neogene potassic calc-alkaline magmatism in
Central of Sulawesi: geochemical and isotopic
constraints. Proceedings 22nd Annual Convention
Indonesian Association Geologists, 1, 59-81.
--, BELLON, H., SOERIA-ATMADJA, R.,
JORON, J. L. & COYrEN, J. 1994. Tertiary and
Quaternary magmatism in Central Sulawesi:
chronological and petrological constraints. Journal
SE Asia Earth Sciences, 9, 81-93.
PRICE, P. B. & WALKER, R. M. 1963. Fossil tracks of
charged particles in gneiss and the age of minerals.
Journal of Geophysical Research, 68, 4847-4862.
RANGIN, C. & SILVER, E. A. 1991. Neogene tectonic
evolution of the Celebes-Sulu basins: new insights
from leg 124 drilling. In: SILVER,E. A., RANGIN,C.,
VON BREYMAYN, M. T., Er AL. (eds) Proceedings
Ocean Drilling Program Scientific Results. 124,
51--63.
, PUBELLIER,M., AZEMA, J., BRIAIS, A., ET AL. 1990.
The quest for Tethys in the western Pacific. 8 paleo-
geodynamic maps for Cenozoic time. Bulletin de la
Socigtg Gdologique de France, 6, 907-913.
RATMAN, N. 1976. Geologic map of the Toli-Toli
quadrangle. North Central Sulawesi, Geological
Survey Indonesia.
-- & ATMAWINATA, S. 1988. Geological report of
Mamuju Quadrangle, South Sulawesi. Geologic
Research and Development Center, Bandung, Open
File Report, 1:250 000.
REYZER, J. 1920. Geologische aantekeningen betreffende
de Zuidelijke Toraja-landen, verzameld uit de
verslagen der mijnbouwkundige onderzoekingen
in Midden-Celebes. In: Jaarboek van het
Mijnwezen Nederlandsch Oost-lndie 1918.
Weltevreden, Gov't Printing Office, 154-209.
RUTTER, M. J., VAN DER LAAN, S. R. & WYLLIE, P. J.
1989. Experimental data for a proposed empirical
igneous geobarometer: Aluminum in hornblende
at 10 kbar pressure. Geology, 17, 897-900.
SASAJIMA, S., NISHIMURA, S., HIROOKA, K., OTOFIJI, Y.,
VAN LEEUWEN, T. & HEHUWAT, F. 1980. Paleo-
magnetic studies combined with fission-track
datings on the western arc of Sulawesi, East
Indonesia. Tectonophysics, 64, 163-172.
SAx, H. G. J. 1931a. Geological Research South West
Celebes. Confidential BPM Report, Balikpapan,
Indonesia.
1931b. Temporary Report of the Geological
Research of the Pare-Pare and Loewoe Permits.
Confidential BPM Report, Balikpapan,
Indonesia.
SCHMIDT,M. W. 1992. Amphibole composition in tonalite
as a function of pressure: an experimental
calibration of the Al-hornblende barometer.
Contributions to Mineralogy and Petrology, 110,
304-310.
SERRI, G., SPADEA, P., BECCALUVA, L., CIVETTA, L.,
COLTORT1,M., DOSTAL,J., Er AL 1991. Petrology of
igneous rocks from the Celebes Sea basement. In:
SILVER, E.A., RANGIN, C., VON BREYMANN,M. T.,
ET AL. (eds) Proceedings Ocean Drilling Program
Scientific Results. 124, 271-296.
SHAW,R. D. & PACKHAM,G. H. 1992. The tectonic setting
of sedimentary basins of eastern Indonesia:
Implications for hydrocarbon prospectivity, The
APEA Journal, 195-213.
SILVER, E. A., MCCAFFREY, R. & SMITH, R. B. 1983a.
Collision, rotation, and the initiation of subduction
in the evolution of Sulawesi, Indonesia, Journal of
Geophysical Research, 88, 9407-9418.
--, JOYODrW~RYO, Y. & STEVENS, S. 1983b.
Ophiolite emplacement by collision between the
Sula platform and the Sulawesi island arc,
Indonesia. Journal of Geophysical Research, 88,
9419-9435.
SITUMORANG, B. 1982a. The formation and evolution
of the Makassar Basin, Indonesia. PhD Thesis,
University of London.
1982b. The formation of the Makassar Basin
as determined from subsidence curves. In:
Proceedings llth Indonesia Petroleum Association
Convention. 383-107.
1984. Formation, evolution, and hydrocarbon
prospects of the Makassar Basin, Indonesia. In:
Proceedings of the 3rd Circum Pacific Energy and
Minerals Research Conference. 227-231.
SUKAMTO, R. 1975. Ujung Pandang, Sheet VIII,
1:1,000,000. Geologic Map of Indonesia, Bandung.
- 1978. The structure of Sulawesi in the light of plate
-
tectonics. Proceedings Regional Conference
Geological Mineral Resources SE Asia, 1975.
Jakarta, 121-141.
--. 1982. Geologic map of the Pangkajene and western
part of Watampone quadrangles, South Sulawesi,
scale 1:250,000. Geological Research and Develop-
ment Centre, Bandung.
- & SIMINDJUNTAK,T. O. 1983. Tectonic relationship
-
between geologic provinces of western Sulawesi,
eastern Sulawesi, and Banggai-Sula in the light of
sedimentological aspects. Bulletin Geological
Research Centre, Bandung, 7, 1-12.
SUN, S.-S. & MeDONOU~H, W. E 1989. Chemical
and isotope systematics of oceanic basalts:
implications for mantle composition and
processes. In: SAUNDERS, A. D. & NORRY, M. J.
(eds) Magmatism in the ocean basins. Geological
Society, London, Special Publications, 42,
313-329.
 MIOCENE W SULAWESI CONTINENT COLLISION
SUNG, G. L. 1948. Samenvatting van belangrijkere
geologische gegevens over Celebes. GL.A. Report
No 22575, Pertamina.
THAMRIN, M. 1985. An investigation of the relationship
between the geology of Indonesian sedimentary
basins and heat flow density. Tectonophysics, 121,
45-62.
1986. Terrestrial heat flow map of Indonesian
basins. Indonesian Petroleum Association
Proceedings, 33-70.
T'HOEN, C. W. & ZmGLZR, K.J. 1917. Verslag over
de resultan van geologisch-mijnbouwkundige
verkenningen en opsporingen in Zuidwest Celebes.
Jaarb. Mijnw. Nederlandsch Oost-lndie, 1915, 44,
237-363.
VAN BEMMELEN, R. W. 1949. The geology of Indonesia.
The Hague, Netherlands, Govt. Printing Office.
VAN BERGEN, M. J., VROON, P. Z. & HOOGEWE~'F, J. A.
1993. Geochemical and tectonic relationships in the
east Indonesian arc-continent collision region:
implications for the subduction of the Australian
passive margin. Tectonophysics, 223, 97-116.
vAN DE WEEe,D, A. A. & At~l~, R. A. 1992. Origin and
evolution of the Tertiary hydrocarbon basins in
Kalimantan (Borneo), Indonesia. AAPG Bulletin,
76, 1778-1803.
VAN LZEUWEN, T.M. 1981. The geology of South
Sulawesi with special reference to the Biru area.
In: BARBER,A. J. & WtRYOSUJONO,S. (eds) Geology
and Tectonics of Eastern Indonesia. Geological
429
Research Development Centre, Special Publication,
2, 277-304.
, TAYLOR, R., COOTE, A. & LONSTA~E, E J. 1994.
Porphyry molybdenum mineralization in a
continental collision setting at Malala, northwest
Sulawesi, Indonesia. Journal Geochemical
Exploration, 50, 279-315.
WHITFORD, D. J. & JEZEK, P. A. 1982. Isotopic constraints
on the role of subducted sialic material in
Indonesian island-arc magmatism. Geological
Society of America Bulletin, 93, 504-513.
WOOD, B. G. M. 1985. The mechanics of progressive
deformation in crustal plates-a working model for
SE Asia. Geological Society Malaysia Bulletin, 18,
55-99.
YUWONO, Y. S. 1987. Contributions gt l'~tude du
volcanisme potassique de l'Indondsie. Exemples du
sud-ouest de Sulawesi et du Volcan Muria (Java).
Doctorat Th~se de l'Universit6 de Bretagne
Occidentale, Brest.
- - . , MAURY, R. C., SOERIA-ATMADJA,R. & BELLON, H.
1988a. Tertiary and Quaternary geodynamic
evolution of S Sulawesi: constraints from the study
of volcanic units. Geologi Indonesia, 13, 32-48.
- - . , PRIYOMARSONO,S., MAURY,R. C., RAMPNOUX,J. P.,
SOERIA-ATMADJA, R., ET AL. 1988. Petrology of the
Cretaceous magmatic rocks from Meratus Range,
southeast Kalimantan. Journal SE Asian Earth
Sciences, 2, 15-22.
 SE Sundaland accretion: palaeomagnetic evidence of large
Piio-Pleistocene thin-skin rotations in Buton

J A S O N R. A L I 1, J O H N M I L S O M 2, E D W A R D M. F I N C H 2 & B U N D A N M U B R O T O 3
1 Department of Oceanography, The University, Southampton, UK.
2 Department of Geological Sciences, University College, London, UK.
3 Geological Research and Development Centre, Bandung, Indonesia.

Abstract: The Tukang Besi Platform, an Australian microcontinental fragment, began docking
with Sundaland in the Pliocene, impacting on east Buton (SE Sulawesi). Fortuin et al. (1989;
Journal of SE Asian Earth Sciences, 4, 107-124) postulated that south Bnton had rotated clock-
wise through about 60 ° relative to central/north Buton in response to the Tukang Besi collision.
A palaeomagnetic investigation was carried out to test this model. Some 41 (of 72) palaeo-
magnetic sites from the upper Neogene Tondo and Sampolakosa Formations on Buton yielded
interpretable data. Sites from south Buton record locally consistent directions, but at sampling
localities < 25 km apart deflections are between 0 ° and 35 ° clockwise. There is no trend in these
data with respect to the broad geotectonic setting. In central Buton declination offsets are
negligible. In north Buton localized (kilometre scale) large (30-60 °) clockwise and counter-
clockwise declination offsets are observed. Effectively the Buton data provide spot markers on
cover sequences that have been locally deformed as 'thin-skin' sheets. The underlying basement
may have experienced the large relative motions proposed by Fortuin et al. (1989) but the upper
Cenozoic cover has not been deformed in such a simple way. This study demonstrates that thin-
skin sheets associated with continental collision may undergo 30-60 ° rotations within very short
intervals (< 2-3 Ma).

The Indonesian archipelago has been a zone of Buton

active convergence throughout the Tertiary. Since
at least the early Miocene, the tectonic develop-
ment of the area has been dominated by the inter-
actions between the Sundaland margin of Eurasia
and the Indo-Australian and Philippine Sea plates.
Convergence of these plates on their triple junction
in eastern Indonesia has led to obduction, sub-
duction, fragmentation and shuffling of numerous
microplates, typically covering 104 to 105 km 2,
within an ever-decreasing area. The boundaries
between these fragments have probably been
generally short-lived (< 5 Ma) and over the next
10-20 Ma many will be incorporated into either
Sundaland or Australia. This process is already
well advanced in the western Banda Sea where
Australian margin fragments such as the Banggai
and Sula Islands, Buton and the Tukang Besi
Platform have collided with the Sundaland margin
represented by the large island of Sulawesi. A
quantitative understanding of the development
of this margin will provide valuable insights
conceming the processes (and rates) involved in
terrane accretion at continental margins. In this
paper new palaeomagnetic data are presented from
upper Neogene rocks on Buton and the results
related to the Plio-Pleistocene collision processes in
the region.
Buton measures 150 km from north to south and
approximately 60 km across at its widest point
(Fig. 1). Topographically, it differs markedly from
Muna Island to the west and the islands of the
Tukang Besi platform to the southeast, being
generally mountainous, with peaks in excess of
1000m. Geologically, the island attracted early
interest because of the occurrence of deposits of
asphalt which are worked commercially at a
number of localities. Because of this, the geology
of the island is well known, at least by comparison
with most of the other islands of eastern Indonesia.
Present understanding is based on the published
work of Wiryosujono & Hainim (1978), Smith
(1983), Fortuin et al. (1989) and Smith & Silver
(1991), as well as unpublished work by M. E. M.
de Smet (pers. comm. 1992). In addition, a con-
siderable amount of land and marine seismic work
has been carded out by oil companies exploring for
hydrocarbons in the region (Davidson 1991). Other
marine seismic, gravity and magnetic surveys have
formed parts of broader scale investigations of the
Banda Sea by US oceanographic institutes. There
have been onshore gravity surveys on Buton, Muna
and one of the Tukang Besi islands and in the
extreme southeast of Sulawesi by a joint University

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, 431
Geological Society Special Publication No. 106, pp. 431-443.
432 J.R. ALI ET AL.

r- --

SULAWESI 41,
"N
l I,

4o30'

Tanjung
Batu

BUTON

5o30'

Traverse

O 5 O k m

Quaternary o~ [ 1 ~ ] Alluvium
post- . , . , .: WapolakaFormation (reef limestone)
collision [U Pliocenet
units pper Miocene~ SampolakosaFormation (chalk and marl)
[Middleto Upper~ TondoFormation (fine - coarse clastics)
W Miocene
~ Pre-Neogeneformations / (, ~ 250km

Fig. 1. Geological map of Buton, SE Sulawesi (based on Smith & Silver 1991) with palaeomagnetic sample sites:
1, Bungi River; 2, Sampolakosa River (no reliable palaeomagnetic data obtained, ND); 3, Kemberu River;
4, Sampolakosa Bridge, (ND); 5, Warumbia River, Lapodi; 6, Wabiau River; 7, Wakalumbe River, (ND); 8, Waulala
River; 9, Kawau River; 10, Longito River; 11, Loangkumbe River; 12, Siloi River; 13, Walue River.
 SE SUNDALAND ACCRETION 433

of London/Geological Research and Development
Centre team (Milsom 1992) and these link to
earlier work by members of UC Santa Cruz in SE
Sulawesi (Silver et al. 1983).
The style and setting of sedimentation on Buton
has been described within a framework of rift-drift
followed by collision and amalgamation (e.g.
Fortuin et al. 1989). The pre-Neogene section
records sedimentation on a micro-continental
fragment which was rifted away from northern
Australia in the Mesozoic. The block drifted ahead
of the Australian continent and encountered the SE
Sulawesi subduction zone in the early Miocene,
undergoing deformation as the zone 'choked' and
subduction ceased (Fig. 2). Convergence continued
between Sundaland and other Australian fragments
in the Banda Sea, accommodated by subduction
east of Buton until the mid-Pliocene, when the
Tukang Besi platform entered the trench off
southern Buton, giving rise to a second deform-
ational event (Fig. 2). Fortuin et al. (1989)
suggested that continued convergence following
this second event could explain the elevated
Quaternary reefs in southern Buton, the pattern of
major faults, folds and thrusts across the island, and
the 50-60°change in regional strike between its
northern and southern halves. Clockwise rotation of
the south with respect to the north would inevitably
lead to extension seaward of central Buton, and an
extensional basin does exist in approximately the
expected position. The pattern of Bouguer gravity
contours, which 'wrap around' southern Buton and
Muna, is also persuasive evidence for differential
rotation.
In addition to the sedimentary rocks, outcrops of
a dismembered ophiolite are scattered through
western Buton, their presence being generally
attributed to the first of the two collisions (e.g.
Fortuin et al. 1989). Gravity surveys have failed,
however, to locate any major anomalies associated
with these rocks, or any sign of a high density root
zone (Milsom 1992). A gravity map of the
Buton-Tukang Besi region is shown in Fig. 3.
These observations are evidence in favour of the
interpretation of the structure and deformation of
Buton in terms of a thin-skinned thrust model;
such a model has also been proposed on purely
geological grounds (H. Manur, pers. comm. 1993).

IwNwt
Sea Level EOCENE - OLIGOCENE
÷~ ~ .......... :......................
++ + + + + ~ l ~ t ~ t ~ ~ ~ . --~,..~----,x.~..~...'.~E~~
+ + + + + + ~~1~
~ ~~- ~ ~' ' ~~~ / - - - - ~
+~ / BUTONMICRO- CONTINENT
MUNA/ SE .ff'~'//
SULAWESI - /
CONTINENTALCRUST
BUTON KAPANTOREH
STRAITS OPHIOLITE
BASIN .. ~ BULU BASIN MIOCENE

~ _ . . ~ _ _ _ _ / ~ ~ - " T U K A N G BESI MICRO - C O N T I N E N T

UPLIFTED
• ~ ~ FOLD & T H R U S T BELT
M U N A / S E SULA WESI
CONTINENTAL CRUST

PLIOCENE / PLEISTOCENE
-- ,,

ACCRETED B UTON ACTIVE FOLD & THRUST BELT

/ / MICRO - C O N T I N E N T
I /

Fig. 2. Cartoons to illustrate the mid to late Cenozoic development of the SE Sulawesi region, based on Fortuin et al.
(1989) and Davidson (1991).
434 J.R. ALI ET AL.
5"S
123°E 124°E
Fig. 3. Bouguer anomalies, Buton, Muna and the Tukang
Besi Platform. Reduction density = 2.67 Mg m-3.
Muna and the Tukang Besi platform
Muna, which lies to the west of Buton, from which
it is separated by a narrow seaway, is an uplifted
carbonate platform of generally low relief. A
narrow limestone ridge which runs close to the
eastern coastline is interpreted as a former barrier
reef. Uplift in the extreme south has been stronger
than in the remainder of the island, although less
strong than in the adjacent parts of Buton, where
flights of coral terraces are impressive features of
the landscape. Basement rocks are exposed only in
a tiny area of schist outcrop at Tanjung Batu, on the
west coast.
The Tukang Besi platform, east of southern
Buton, is largely submerged but some moderately
large islands rise a few metres above sea-level.
These seem to be entirely built up of upper
Neogene or Quaternary limestone. Gravity data
from Wangi Wangi, the largest of the islands, and
from marine surveys (Fig. 3), suggest that the
crustal thickness in the area approaches that of a
normal continent, supporting the view expressed by
Hamilton (1979) that the plateau is another
fragment of Australian-related continental crust.
Seismic reflection data obtained using relatively
low power sources on oceanographic cruises have
imaged only very thin sedimentary sections and it
seems unlikely that there are thick Neogene
sediments analogous to those exposed on Buton
anywhere on the plateau. The same seismic data
suggest a clear structural break between the plateau
and Buton itself, the intervening seaway being
considerably deeper than the strait between Buton
and Muna.
Formations sampled for palaeomagnetism
The aim of the project was to use palaeomagnetism
to quantify the proposed differential rotation
between north and south Buton (Fortuin et al.
1989). As the collision of Tukang Besi Platform
with Buton is believed to have started in the
mid-Pliocene, sampling was confined to two upper
Neogene formations exposed across much of Buton
(the stratigraphy of Buton is summarized in Fig. 1).
Older formations were not sampled as they were
considered to have complex pre-Miocene motion
histories: isolating their record of Plio-Pleistocene
deformation would be difficult, if not impossible.
The upper Miocene Tondo Formation comprises
terrigenous clastic rocks ranging from breccias to
mudstones. The sediments were eroded following
the uplift associated with the Buton-Sundaland
collision. By Mio-Pliocene times the region had
stabilized, and pelagic limestones and marls of the
Sampolakosa Formation were being deposited.
The Tondo and Sampolakosa Formations are
unconformably overlain by Quaternary reef
deposits of the Wapulaka Formation (Sikumbang &
Sanyoto 1981). Reefal limestones are usually very
weakly magnetized, and because of this the
Wapulaka Formation was not sampled for this
palaeomagnetic study.
In addition to the palaeomagnetic studies,
nannofossils were examined from 32 of the palaeo-
magnetic sites, thus providing information on the
maximum age of the magnetic remanence. The
nannofossils have been assigned to the Martini
(1971) zones which in turn have been linked to the
geomagnetic timescale of Cande & Kent (1992)
using the magnetochron-biozone correlations of
Berggren et al. (1985). Unfortunately some of the
assemblages from this study lack the nominate
taxon for the Martini scheme, however the total
recorded assemblages are sufficient to enable
assignment to a particular zone(s).
Palaeomagnetism
Methods
The majority of specimens were obtained using a
gasoline-powered rock-drill which was used to cut
2.5 cm diameter mini-cores. The cores were
orientated to _+ 2 ° using a magnetic compass-
inclinometer. A palaeomagnetic site comprises 6-9
orientated mini-cores spanning 10-30cm of
section within a bed. A number of orientated
blocks (c. 300 cm 3) were also retrieved. Wherever
possible, sites were collected with the aim of
utilizing the fold and reversal tests to constrain the
age of magnetization.
The stability of magnetization of each specimen
was assessed after using either stepwise alternating
 SE SUNDALAND ACCRETION
field (AF) demagnetization or continuous thermal
demagnetization to isolate the various components
of magnetization held within the rocks. The
AF-processed specimens were analysed at
Southampton, using either a 'Molspin' spinner
magnetometer in tandem with a 'Molspin' de-
magnetizer, or a '2G Enterprises' cryogenic
magnetometer which has an in-line 3-axis
demagnetizing unit. Continuous thermal de-
magnetizations were performed on a number of
specimens in the laboratory of M. Fuller at Santa
Barbara, using a '2G Enterprises' cryogenic mag-
netometer with an in-line thermal demagnetizing
unit (Dunn & Fuller 1984). In practically all cases
the remanence of the Tondo and Sampolakosa
Formations was found to be dominated by a single
component suggesting a simple magnetization
history. Characteristic components of magnetisa-
tion were identified from vector end-point plots
(Zijderveld 1967), and calculated using a
Kirschvink (1980) based software package. Site
mean directions (Table 1) have been calculated
using the statistics of Fisher (1953). The statistics
of McFadden & Reid (1982) have also been used to
calculate mean inclinations; in a number of cases
the sites are from isolated outcrops and averaging
site directions using Fisher (1953) statistics may
have been incorrect. It is worth noting, however,
that in all cases the difference between the Fisher
(1953) and McFadden & Reid (1982) values is less
than 0.5 ° (see Table 1).
Isothermal remanent magnetization
Isothermal remanent magnetization (IRM) analysis
was carried out on representative specimens from
sites yielding reliable data to provide information
on their principal remanence carriers. The IRM was
generated using a 'Molspin' pulse magnetiser (up
to 0.86 T) and was measured between steps using
a 'Molspin' magnetometer. The shape of the IRM
curve (as well as the peak IRM value) was used to
evaluate the characteristic remanence carrier(s).
The IRM ratio, defined by Ali (1989), is the ratio
of the IRM at 0.3 T to the IRM at 0.86T. As
a general guide, values above 0.9 indicate that
magnetite is the dominant remanence carrier,
whereas values below 0.9 indicate that the
remanence may be due to other minerals. All of
the specimens have IRM ratios greater than 0.95
(Table 1) suggesting that the remanence of the
Tondo Formation and the basal Sampolokosa
Formation is carried by magnetite.
Palaeomagnetic results: south Buton
Bungi River, Bungi
Four drill sites and one hand specimen were
sampled from the lower part of the Sampolakosa
435
Formation at five outcrops exposed along the Bungi
River, about 5 km ESE of Bungi (Fig. 1; 1). The
sites were from beds which dip to the north and
west at 10-15 ° . The formation is dominated by
compacted marls in beds typically 30-80 cm thick.
Samples from three sites were analysed for nanno-
fossils, but the faunas were too poorly preserved to
yield a reliable age; a Mio-Pliocene age is assumed.
Specimens from Sites BU1, 3, 4 and 5 have
NRM intensities between 0.07 and 1.4 mA m -1
(Table 1). AF demagnetization of these sites
indicates that the remanence is carried by a single
high-coercivity component. Site BU2.H, from the
Tondo-Sampolakosa transition, revealed quite
different magnetic behaviour; the remanence being
carried by a relatively soft single component of
magnetization. Specimens from Sites BU 1, 2 and 5
were the subject of IRM studies (Table 1). All three
specimens yielded IRM ratios >0.95 suggesting
that the remanence is probably carried by mag-
netite. The five sites had a mean in situ direction of
D = 357.9 °, I = -23.5 °, a95 = 7.7 °, K = 98.4 (Table
1, Fig. 4a), and a tilt corrected direction of
D = 2.7 °, I = -31.3 ° a95 = 9.6 °, K = 63.8 (Fig. 4b).
A normal polarity magnetization is assumed.
The remanence data can be interpreted in two
ways. Assuming the individual site corrections
account for all of the deformation experienced by
each site, then the in situ and tilt corrected a95 and
K values suggest that the remanence post-dates
tilting. However, the present-day geocentric axial
dipole inclination for the Bungi River area is
-10.7 °, which suggests that the remanence pre-
dates deformation. The age of magnetization of the
Sampolakosa Formation in the Bungi River is con-
sidered ambiguous. The declination data, whether
interpreted as either pre- or post-folding, indicate
negligible rotation.
Kemberu River, Masiri
Numerous outcrops of the Tondo Formation were
exposed in the dry bed of the Kemberu River, near
Masiri (Fig. 1; 3). The formation is dominated by
massively bedded conglomerates and immature
sandstones; lithologies not particularly suited for
palaeomagnetic investigations. The beds are folded
about axes aligned NNE-SSW. Dips are typically
20-40 ° , although values of 60-70 ° were locally
recorded. Eight drill and two hand sites were
sampled from 10-20 cm thick mudstones and
siltstones. Three sites yielded data with a mean in
situ direction of D = 355.6, I = -14.2 °, a95 = 18.4 °,
K = 4 5 . 8 , and a tilt corrected direction of
D = 3 5 9 . 6 ° , I = - 1 4 . 8 ° a95=15.3 ° , K = 6 5 . 4
(Table 1, Fig. 4c/d). This direction is interpreted
as a primary normal polarity magnetization with
negligible net rotation.
Table 1. Summary of palaeomagnetic data at sites named on Fig. 1 with latitude and longitudes of each
tao
NRM in situ Strike tilt corr. IRM peak c7~
TRAVERSE Exp SITE Fm Nanno. age (Ma) Np Nc (mA m -1) DEC INC & dip DEC INC a95 K Pol ratio IRM

BUNGI a BU1 S indet. ? c. 5 6 5 0.07-0.3 358.8 -34.3 290/10 N 355.7 -43.5 7.4 108.1 N 0.99 28.9
(5.42°S/122.70°E) b BU2.H T-S indet. ? c. 5 4 4 c. 2.0 357.1 -14.7 237/18 NW 0.8 -29.5 6.2 219.4 N 0.96 348.1
c BU3 S 7 7 0.5-1.0 4.3 -26.2 210/18 NW 13.9 -32.7 5.3 128.9 N 0.95 57.2
d BU4 S 6 6 0.14-0.25 354.1 -23.0 165/14 W 359.5 -20.1 9.5 51.1 N
e BU5 S 6 6 0.8-1.4 356.3 -19.4 216/18 NW 3.0 -30.1 5.1 172.2 N
IS (F) 5 357.9 -23.5 7.7 98.4
IS (M&R) 5 NA -23.8 9.2 60.2
tilt (F) 5 2.7 -31.3 9.6 63.8
tilt (M&R) 5 NA -31.7 10.4 46.8

KEMBERU a BUll T indet. ? 9.5-5.3 6 6 10-50 1.5 -16.3 200/08 W 3.9 -18.7 12.9 28.0 N 0.99 21393
(5.67°S/122.67°E) b BU12 T indet. ? 9.5-5.3 6 5 22-53 342.0 -16.4 160/24 W 348.6 -14.1 9.3 68.9 N 0.99 17067
c BUI3 T NPI7-21 reworked 5 5 0.4-1.6 3.0 -9.6 190/18 W 6.3 -11.3 14.1 30.6 N
IS (F) 3 355.6 -14.2 18.4 45.8
IS (M&R) 3 NA -14.1 9.7 216.1
tilt (F) 3 359.6 -14.8 15.3 65.4
tilt (M&R) 3 NA -14.7 9.3 235.2

WARUMBIA a BU21.H T 1 1 29,45 58.4 -27.7 062/35 S 42.6 -20.4 NA NA N

(5.47°S/122.83°E) b BU24 T indet. ? c. 5 6 6 2.0--4.0 78,4 -56.2 090/27 S 48.8 -43.6 9.0 56.9 N 0.96 1521
c BU27.H T 1 1 91.78 21.6 -34.8 080/18 S 17.1 -19.1 NA NA N
d BU29 S indet. ? c. 5 6 6 30-100 16.0 -45.1 120/12 SW 18.2 -33.4 6.4 110.7 N 0.98 8115
IS (F) 4 41.7 -43.6 28.7 11.1
IS (M&R) 4 NA -42.7 20.0 21.0
tilt (F) 4 31.0 -29.8 21.0 19.9
tilt (M&R) 4 NA -30.1 18.6 24.4

WABIAU a BU30 T NNI0 9.5-8.2 6 6 1.5-5.0 8.4 -4.6 075/15 SE 8.6 9.2 6.2 116.6 N
(5.35°S/123.03°E) a BU31 T 6 5 5-10 0.7 -15.7 075/15 SE 0.1 -1.2 5.9 167.2 N
b BU32 T 6 5 0.2--0.5 7.7 -10.8 060/12 SE 6.9 -1.2 6.0 165.7 N 1.00 377.6
c BU33 T 5 5 0.1-0.25 345.5 -18.7 120/14 SW 347.8 -8.5 12.4 38.8 N
d BU34 T NNll-15 8.2-3.6 6 6 3-6 345.2 -10.7 065/21 SE 345.2 10.0 7.1 90.1 N
d BU35 T NNll-15 8.2-3.6 5 5 2.5-4.8 346.5 -11.6 060/16 SE 346.2 3.8 6.4 142.2 N 0.99 2247
e BU36 T 6 5 1.3-7.0 355.7 -13.6 210/10 NW 358.2 -19.1 17.7 19.7 N
e BU37 T NN10 9.5-8.2 6 6 5-12 358.1 -8.6 210/10 NW 359.8 -13.8 8.1 68.7 N
e BU38 T 6 6 c. 2.4 6.0 -5.6 200/15 W 7.9 -9.0 8.1 69.5 N
f BU39 S 4 4 c. 0.3 357.3 -17.1 140/08 SW 355.0 -21.8 11.5 64.9 N
f BU40 S NN10-11 9.5-5.3 6 6 0.1-0.3 353.8 -13.0 130/08 SW 351.4 -18.4 6.8 99.5 N 0.99 47.1
IS (F) 11 356.8 -11.9 5.4 72.9
IS (M&R) 11 NA -11.8 3.0 165.3
 tilt (F) 11 357.0 -6.4 8.1 32.8
tilt (M&R) 11 NA ~6.6 7.6 25.8

WAULALA a BU46.H T 1 1 2.93 214.9 -4.0 235/25 NW 215.1 4.7 NA NA R

(5.23°S/122.95°E)

KAWAU a BU49 T 5 4 0.2-0.35 354.8 -23.9 083/15 S 354.7 -8.5 6.9 178.4 N
(4.95°S/122.97°E) a BU50 T 5 4 0.16-0.5 353.1 -18.9 070/12 S 352.5 -7.2 17.3 29.0 N
a BU51 T 5 5 0.4-0.9 190.5 33.4 080/16 SE 188.0 18.3 8.7 79.0 R 1.00 1268
a BU52 T indet. ? 9.5-5.3 6 6 0.9-3.6 352.4 -16.9 037/37SE 350.8 9.9 4.2 259.5 N
a BU53 T 6 5 0.5-1.3 4.6 -22.5 130/10 SW 6.5 -14.2 10.1 57.8 N
a BU54 T 6 6 1.5-2.5 157.6 21.1 110/12 S 160.0 12.1 4.6 213.2 R
a BU55 T 6 6 0.4-0.9 354.3 19.6 085/20 S 354.3 -0.4 7.6 77.9 N 0.99 154.1
IS (F) 7 355.0 -17.3 14.6 17.8 M
tilt (F) 7 355.1 -7.4 10.0 37.2 M

LONGITO a BU56 T NNll-15 8.2-3.6 4 3 0.1-0.2 176.1 19.2 150/12 SW 179.3 13.6 26.5 22.6 R
(4.97°S/122.92°E) a BU58 T 4 4 0.1-0.2 348.9 -12.7 170/34 W 356.3 -11.1 22.0 18.3 N 0.99 106.7
a BU59 T 3 3 0.1-0.3 14.7 -14.4 030/31 SE 9.6 -4.7 27.8 20.7 N
IS (F) 359.8 -15.7 20.4 37.3 M
tilt (F) 1.7 -9.8 12.7 95.2 M

LOANGKUMBE a BU62 T 6 6 4-12 147.1 14.9 140/32 SW 153.9 8.9 8.3 66.5 R 1.00 9481
(4.70°S/123.08°E) a BU63 T indet. ? 8.2-3.6 6 4 0.1-0.35 334.1 -10.9 140/28 SW 337.5 3.1 17.5 28.8 N
b BU66 T 6 6 10-17 226.9 13.9 145/32 SW 226.7 -17.8 8.4 65.3 R 1.00 17415
c BU68 T 5 3 c. 7 352.4 -38.4 125/22 SW 0.4 -21.1 23.1 29.4 R
c BU69H.a T 1 1 37.68 39.5 -31.0 126/26 SW 39.0 -5.0 NA NA N
c BU69H.c T 1 1 54.23 32.3 -15.9 126/26 SW 32.3 4.1 NA NA N
d BU70 T NNll-15 8.2-3.6 7 5 8-59 193.6 -4.3 140/30 SW 188.0 -27.8 10.3 56.0 R 1.00 17756
IS (M&R) 7 NA 18.1 13.0 17.1 M
tilt (M&R) 7 NA 9.0 12.9 17.2 M

SILOI a BU71 T indet. ? 9.5-5.3 5 4 13-46 355.1 0.9 280/18 N 354.5 -16.5 18.1 26.6 N 1.00 56538
(4.72°S/123.00°E)

WALUE a BU72 T indet. ? 9.5-5.3 6 6 6-17 7.0 -33.7 350/80 E 317.6 -19.6 12.7 28.6 N 0.99 12216
(4.53°S/122.92°E)

Abbreviations after site number are: F = Fisher (1953) statistics used to calculate mean direction; M & R = McFadden & Reid (1982) statistics used to calculate mean inclination.
Other abbreviations are Fm = Formation, T = Tondo, S = Sampolakosa; Nanno. = nannofossil zones; Np = Specimens processed; Nc = Specimens used to calculate the site direction;
NRM = initial intensity; in situ = site mean direction before application o f tectonic correction: DEC = declination in degrees; 1NC = inclination in degrees; NA = not applicable;
Strike & Dip = bedding orientation; tilt corr. = site mean direction after tectonic correction; a95 = circle of 95% confidence about the site mean; K = precision parameter;
Pol = polarity: N, normal, R, reversed; M = statistics calculated by inverting reverse polarity directions;IRM ratio = IRM at 0.3 T/IRM at 0.86 T; peak IRM expressed in m A m 2.
4~
438 J.R. ALI E T A L .

N N N N N N

E
b c ~ d
............... i......................... t .............. !Bu:4 ........e t
N N N N

: i

g E h

N N
®®
i ®
~ - j j
N

t~uov.l-la [ i BU68

........................i.......................................... i...................
BU66 :: I i
i BU62[ @ i BU62
~BU70 ~ BU66 @ ~

Fig. 4. Site mean data from (a, b) Sampolakosa Fm, Bungi River (a = in situ, IS, b = tilt corrected, TC); (c, d) Tondo
Formation, Kemberu River (c = IS, d = TC); (e, f) Tondo Fm and basal Sampolakosa Fm, Warumbia River (e = IS,
f = TC); (g, h) Tondo Fm and basal Sampolakosa Fm, Wabiau River (g = IS, h = TC); (i, j) Tondo Fm, Kawau River
(i = IS, j = TC); (k, !) Tondo Fm, Longito River (k = IS, 1 = TC); (m, n) Tondo Fm, Loangkumbe River (m = IS,
n = TC). Open and closed circles represent downward and upward dipping vectors respectively. Site directions based
on mini-core samples are shown with their 95% confidence circle. Hand specimens do not have associated
confidence circles.

Warumbia River, Lapodi
Three hand specimens and four drill sites were
collected from the Warumbia River, about 2 km
north o f Lapodi (Fig. 1; 5). Drill site B U 2 9 was col-
lected f r o m a calcilutite at the base o f the
S a m p o l a k o s a Formation. The remaining sites were
from mudstones and fine sandstone in the upper
part o f the Tondo Formation. Reliable data were
obtained from two hand specimens, B U 2 1 . H and
B U 2 7 . H , and two drill core sites, B U 2 4 and B U 2 9 .
The mean in situ direction is D = 41.7 °, I = - 4 3 . 6 °,
a95 = 28.7 °, K = 11.1 (Fig. 4e). F o l l o w i n g the
application o f the tilt corrections the mean direction
 SE SUNDALAND ACCRETION
is D = 3 1 . 0 °, I = - 2 9 . 8 ° a95=21.0°, K = 1 9 . 9
(Table 1, Fig. 4f). Although a95 and K improve with
the correction for dip, the scatter of the vectors
suggests that the local tectonics are more complex
than simple S - S W tilting.
Wabiau River, Walumpo
East Buton is close to the supposed Tukang Besi
Platform-Buton Block suture. The Wabiau River
(Fig. 1; 6) is adjacent to the Tondo Formation type
area and contains splendid exposures of both the
Tondo Formation and the basal part of the
Sampolakosa Formation. The rocks are mildly
deformed only with dips generally <15 ° . Nine
Tondo Formation and two Sampolakosa Formation
drill sites were sampled at six outcrops. The eleven
sites yielded data with a mean in situ direction of
D = 356.8 °, I = -11.9 °, a95 = 5.4 °, K = 72.9, and a
tilt corrected direction of D = 357.0 °, I = - 6 . 4 °,
a95 = 8.1 °, K = 32.8 (Table 1, Fig. 4g/h). Although
the in situ directions are slightly more clustered
than the tilt corrected vectors, suggesting that
the remanence post-dates deformation, the magnet-
ization of the Wabiau River Tondo-Sampolakosa
Formation sequence is considered to be primary.
Demagnetization isolated essentially single
component low-coercivity remanences (also see
Table 1 ; IRM data). Field observations suggest that
the rocks have undergone nothing more dramatic
than burial and slight structural upheaval. It is
possible that the assumption of a tilt corrected mean
direction based on simply restoring the beds to
horizontal is incorrect, and that faults between each
of the outcrops may have accommodated local
small-scale deformation. The tilt corrected mean
direction suggests negligible rotation of the site
since the start of the Pliocene.
Waulala River, Lawele
Superb exposures of the Tondo and Sampolakosa
Formations are present in the Waulala and Loko
Rivers, about 5 km south of Lawele (Fig. 1; 8). The
geology is dominated by a number of major
N E - S W trending faults limiting the area to which
palaeomagnetic declination data obtained might be
applied. Four drill sites and one hand specimen
were sampled from the Tondo Formation, together
with one drill site from the Sampolakosa
Formation. Only site BU46.H, the hand specimen
from the Tondo Formation, yielded reliable data
with a reverse polarity tilt corrected direction of
D = 215.1 °, I =4.7 ° (Table 1), suggesting clock-
wise rotation in excess of 30 ° . However, as the site
is adjacent to a major fault system, it is suggested
that this deflection reflects local deformation.
439
Palaeomagnetic results: central Buton
Kawau River, Bubu
The Tondo Formation is exposed over a kilometre
of continuous outcrop in the Kawau River, 5 km
inland of Baubau (Fig. 1; 9). Seven drill sites were
sampled from beds that dip towards the east and
south at between 10 ° and 30 °. Sites BU49, 50, 52,
53 and 55 record normal polarity magnetizations;
BU51 and 54 record reverse polarity magnetiz-
ations (Table 1). The in situ mean direction is
D = 355.0 °, I = -17.3 °, a95 = 14.6 °, K = 17.8 (Table
1, Fig. 4i). Application of the tilt corrections
produces a direction of D = 355.1 °, I = - 7 . 4 °,
a95 = 10.0 °, K = 37.2 (Fig. 4j), which suggests
that the remanence pre-dates folding. The presence
of both normal and reverse polarity sites in a
single outcrop is highly suggestive of a primary
magnetization. The mean direction indicates
negligible rotation of the site since formation.
Longito River, Bubu
The Kamunte River, a tributary of the Longito
River, was accessed from the road linking Bubu to
Pure (Fig. 1; 10). Five sites were sampled from a
600 m long exposure of the Tondo Formation in the
dry river bed. The rocks, mainly mudstones and
siltstones in beds 10-30 cm thick, are deformed
into an open synform. The Tondo Formation
samples are weakly magnetized with initial NRM
intensities between 0.06 and 0.20 mA m -1. Reliable
data were obtained for three sites. Scatter for
specimens from these site is large with values for
a95 and K typically 25 ° and 20 respectively.
However, the outcrop contains both normal (BU58
and BU59) and reverse (BU56) polarity sites,
which suggests that the remanence is primary.
The in situ mean direction is D = 3 5 9 . 8 °,
I = -15.7 °, a95 = 20.4 °, K = 37.3 (Table 1, Fig. 4k).
Application of tilt corrections produces a mean
direction of D = 1.7 °, I = -9.8 °, a95 = 12.7 °,
K = 9 5 . 2 (Fig. 41), which suggests that the
magnetization pre-dates folding, with no rotation.
Palaeomagnetic results: north Buton
Loangkumbe River, Loangkumbe
The Tondo and Sampolakosa Formations are well
exposed along the Loangkumbe River (Fig. 1; 11).
This is one of the most important areas of Buton
for the collection of palaeomagnetic samples,
potentially providing the Mio-Pliocene reference
direction for the northern part of the island.
Unfortunately the stratigraphic interval from which
palaeomagnetic sites can be sampled is limited
because the Loangkumbe River flows parallel to
440 J.R. ALI ET AL.
the general strike of rocks in the area. The lower
reaches of the fiver expose marls from the lower
part of the Sampolakosa Formation but only a
single site was sampled; determining accurate dips
and strikes in the formation is difficult because
obvious signs of bedding have been obliterated by
erosion. Eight drill sites and one hand specimen
were sampled from the Tondo Formation.
Seven Tondo Formation sites, from three
outcrops, have yielded data (Table 1, Fig. 4rn/n).
Sites BU62 and BU63 from an outcrop c. 200 m
upstream of the L o a n g k u m b e - M o u s o Cabang
yielded tilt corrected directions of D = 153.9 °,
I = 8.9 °, a95 = 8.3 °, K = 66.5, N = 6 and D = 337.5 °
I = 3.1 ° a95 = 17.5 °, K = 17.5, N = 4. The presence
of reverse and normal polarity sites in one outcrop
suggests that the remanence is primary. The
declinations indicate about 25 ° counter-clockwise
rotation.
Five sites, between 3 and 5 km upstream of Sites
BU62 and 63, yielded directional data. BU66 has a
tilt corrected mean direction of D = 226.7 °,
I = -17.8 °, a95 = 8.4 °, K = 65.3, N = 6, indicating
about 45 ° of clockwise rotation. Site BU70, about
600 m upstream, also carries reverse polarity
magnetization, with D = 188.0 °, I = - 2 7 . 8 °,
a95 = 10.3 °, K = 56.0, N = 5, indicating negligible
rotation. Site BU68 is normally magnetized with no
declination offset. From the same outcrop, hand
samples BU69H.a and BU69H.c record c. 35 ° of
clockwise rotation.
Declination data from the Loangkumbe River
suggest that rotations for this part of north Buton
are significant, but appear to be local (sub-
kilometre scale). Using the statistics of McFadden
& Reid (1982) the mean tilt corrected inclination is
+9.0 °, assuming a normal polarity, suggesting
deposition at a latitude c. 4.5°N. However, because
of the large a95 confidence circle (12.9 °) associated
with this inclination, it can only be stated with
confidence that this part of Buton was close to the
equator during Mio-Pliocene times.
Siloi River, Ronta
The Siloi River, accessed from the road which links
Ronta to Maligano, exposes the Tondo Formation
(Fig. 1; 12). The geology of the area is complicated
by a series of major faults juxtaposing these upper
Miocene rocks with a number of pre-Neogene
formations. During fieldwork, exposures of the
Tondo Formation were seen only in the river bed,
seriously hampering the palaeomagnetic sampling.
However, a site in the Tondo Formation was
sampled 4 km downstream from the road bridge
over the Siloi River. The site (BU71) is in a
medium-grained sandstone adjacent to a N E - S W
striking fault that juxtaposes the formation with the
Ogena Formation (lower Jurassic; Davidson 1991).
The in situ site mean direction is D = 3 5 5 . 1 °,
1 = 0 . 9 °, a95 = 18.1 °, K = 26.6, N = 4. (Table 1).
Application of the tilt correction produces a
direction of D = 354.5 °, I = -16.5 °. The remanence
is carried as a normal polarity, but the age of the
remanence (depositional/post deformation) is not
known.
Walue River, Kobakoba
Exposures of the Tondo Formation in the Walue
River are scarce and just one drill site, BU72, was
sampled (Fig. 1; 13). The minicores were obtained
from adjacent sandstone and siltstone beds which
dip 80 ° to the east. The specimens from both the
sandstone and siltstone had similar initial NRM
intensities of about 10 mA m -1. The site has an
in situ mean direction of D = 7.0 °, I = - 3 3 . 7 °,
a95 = 12.7 °, K = 28.6, N = 6 (Table 1) and a tilt
corrected direction of D = 317.6 °, I = -19.6 °. It is
assumed that the remanence is primary because the
in situ direction is much steeper than the predicted
geocentric axial dipole inclination for the Walue
River area (-8.9°). The steep dip of these rocks, and
hence the amount of deformation that site BU72
has undergone, must limit the areal extent for which
the declination offset can confidently be used to
unravel rotations relative to north. However, the
tilt corrected inclination angle can be used to
help define the latitude of formation ( 1 0 _ 7.5 ° S,
assuming that the remanence was acquired during a
normal polarity geomagnetic field).
Summary of the palaeomagnetic data
Latitudinal information from all of the palaeo-
magnetic sites yielding interpretable data is sum-
marized in Fig. 5. They suggest negligible (c. 1o N)
northward motion of Buton and, by implication, of
SE Sulawesi since the late Miocene.
The palaeomagnetic declination data from both
the upper Miocene Tondo Formation and the
Mio-Pliocene Sampolokosa Formation on Buton
are summarised in Fig. 6. Sites from south Buton
record locally consistent declination offsets, but at
sampling localities less than 25 km apart the
deflections vary between 0 ° and 35 ° clockwise.
There appears to be no simple pattern to these data
with respect to the regional tectonic setting. The
age of magnetization of the Sampolakosa
Formation sites from the Bungi River is uncertain.
The five sites show marginally tighter in situ (rather
than tilt corrected) clustering. However the in situ
inclination o f - 2 3 . 8 ° is too steep to be the result of
a recent weathering induced chemical remanence,
where an inclination of around -10.7 ° would be
predicted. Regardless of the age of the remanence,
 SE SUNDALAND ACCRETION 441

10 - MEAN NORTHWARD the Bungi River area has experienced negligible

8- DRIFT = + O ~ rotation since the magnetization was acquired. In
the Kemberu River, 25 km to the south of Bungi,
the Tondo Formation sites also record negligible
rotation. In the Warumbia River, 25 km to the east
of Bungi, four upper Tondo Formation sites record
clockwise rotations of c. 30 °. East Buton was the
one part of the island where large rotations might
have been predicted because it is closest to the
Z~4-©
2- L Tukang Besi Platform-Buton impact point.
However, the large number of sites from the
0 I ' I ' I I ' I ' I '
B
I '
Wabiau River revealed negligible declination
offsets. The single reliable Tondo Formation site in
-18 -12 -6 0 6 12 18 the Waulala River has a large declination offset
Apparent Northward Drift (c) (35 ° clockwise). Large rotations might be expected
here as this site is from a block located in the
Fig. 5. Apparent formation latitude shift for the Tondo middle of several NE-SW striking faults.
and Sampolakosa Formation sites, Buton. In central Buton, declinations for the Tondo
Formation sites in both the Kawau and Longito
Rivers indicate negligible rotation. The two
sections include both normal and reverse polarity
sites suggesting that the remanence is primary. Also
in both sections there is a clear clustering of the site
vectors following application of the tilt corrections
to the in situ data.
In north Buton, declination data from the
Loangkumbe River suggest local (< km) large-
scale, clockwise and counter-clockwise rotations. A
single site from the Tondo Formation in the Siloi
River records negligible rotation, but the age of the
magnetization is uncertain. One Tondo Formation
site from the Walue River, records a counter-
clockwise rotation of over 40 ° . However, the site
was sampled from steeply dipping beds, and the
declination deflection probably reflects localized
deformation.

Fig. 6. Summary of the palaeomagnetic declination data
from the Tondo and Sampolakosa Formations, Buton.
The orientations of the arrows illustrate the declination
offsets. Note each arrow has an error typically 10-13°
(see Table 1).
Discussion
The Buton region contains a record of the
relatively recent successive accretion of two micro-
continents to the edge of a continent, and a
quantitative understanding of the processes
involved provides a valuable analogue for older
systems where the rock record is less complete.
Prior to the present study no palaeomagnetic
information was available for the area and the
initial aim of the research was to test the model of
Fortuin et aL (1989) which proposed large-scale
(c. 60 °) Pliocene clockwise rotation of south Buton
relative to central and north Buton as a result of the
Tukang Besi Platform collision. Although it
remains possible that the underlying basement
experienced these large relative motions, the study
has shown that the upper Cenozoic cover has not
been deformed in such a simple way (Fig. 6).
In recent publications (e.g. Davidson 1991) the
geology of Buton has been interpreted in terms of
442 J.R. ALI ET AL.
thin-skinned deformation processes but no insights
have been provided into the rotations which the thin
sheet overthrusts might have experienced. Strictly
speaking, the palaeomagnetic study has no bearing
on the question of thin-skin versus thick-skin
tectonics but it is perhaps easier to envisage large
relative rotations of small blocks taking place in a
thin-skin rather than a thick-skin context.
Resolving the fine scale detail of the tectonics of
Buton is beyond the scope of this or indeed any
project, because of the lack of exposure, but it has
been demonstrated by the palaeomagnetic results
that in a collision setting thin overthrusts may well
be rotated by 30-60 ° within 2-3 Ma.
The palaeomagnetic study has also provided
support for the two-stage model for collision,
which is still controversial; if the Tukang Besi
Platform were part of the Buton block prior to its
collision with SE Sulawesi in the Miocene, there
would be little reason for the large Pliocene to
Recent rotations which are actually observed.
However, the amount of thin-skinned rotation
does not seem to have been a simple function of
proximity to the microcontinental suture. Rotations
in north Buton, which is further from Tukang Besi,
are greater than rotations at sites, such as the
Wabiau River, which are closer to it. The difference
in deformation style between north and south is
also the reverse of expectation. In the south,
each individual traverse section, up to 15 km in
length, shows a consistent declination deflection,
suggesting the presence of coherent sheets at least
15 km across, whereas in the north rotations
appear to have occurred on a much more local
(c. 1 km) scale.
Finally, the results presented here emphasize the
importance of comprehensive sampling in any
References
ALI, J. R. 1989. Magnetostratigraphy of early Palaeogene
sediments from NW Europe. PhD thesis, University
of Southampton.
BERGGREN, W. A., KENT, D. V., FLYNN, J. J. & VAN
COUVERING, J. A. 1985. Cenozoic geochronology.
Bulletin of the Geological Society of America, 96,
1407-1418.
CANDE,S. • KENT,D. V. 1992. A new geomagnetic time
scale for the Late Cretaceous and Tertiary. Journal
of Geophysical Research, 97, 13917-13951.
DAVIDSON,J. W. 1991. The geology and prospectivity of
Buton Island, SE Sulawesi, Indonesia. Proceedings
of the Indonesian Petroleum Association 20th
Annual Convention, 209-231.
DUNN, J. R. & FULLER,M. 1984. Thermal demagnetiz-
ation with measurements at high temperature using
a SQUID magnetometer. LOS Transactions AGU,
65, 863.
FISHER, R. A. 1953. Dispersion on a sphere. Proceedings
of the Royal Society of London, Series A, 217,
295-305.
palaeomagnetic programme in SE Asia. If just the
Waulala and Warumbia sites had been sampled in
southern Buton, the Fortuin et al. (1989) rotation
hypothesis would probably have been regarded as
confirmed, although with rotations somewhat
smaller than those originally suggested. If, on the
other hand, these sites had not been visited and
work had been confined to the Bungi, Kemberu and
Wabiau Rivers, it would have been concluded that
rotations were absent in southern Buton. Neither
conclusion is supported by the full dataset actually
obtained.
Conclusions
This palaeomagnetic study of the SE Sulawesi/
Buton/Tukang Besi accretion complex has
provided a quantitative insight into the process of
crustal deformation at a continent-microcontinent
accretion site. In the Buton region, a significant
component of the deformation has taken place as
thin-skin overthrusts. The palaeomagnetic dataset
suggests that the rotation of the thin-skin sheets is
complex; there is no obvious relationship between
the rotation of a sheet and its proximity to the
microcontinent impact point. Rotations of up to
30-60 ° may be generated. Rotation rates are
considerable; the Tukang Besi Platform-Buton
collision began in the Pliocene, 2-3 Ma.
This study was funded by the University of London
Geological Research in SE Asia Consortium. Special
thanks are given to Tony Barber, John Davidson, Kate
Davis, Michael de Smet, Mike Fuller, Ernie Hailwood,
Hendry Manur and Graham Rose for their input to this
project. The constructive comments of H. Wensink, A. R.
Fortuin and an anonymous referee greatly improved this
manuscript.
FORTUIN, A. R., DE SMET, M. E. M., HADIWASASTRA,S.,
VAN MARLE, L. J., TROELSTRA, S. R. &
TJOKROSAPOETRO, S. 1989. Late Cenozoic sedi-
mentary and tectonic history of south Buton.
Journal of SE Asian Earth Sciences, 4, 107-124.
HAMILTON,W. 1979. Tectonics of the Indonesian region.
US Geol. Survey Professional Paper, 1078.
KIRSCHVlNK,J. L. 1980. The least squares line and plane
analysis of paleomagnetic data. Geophysical
Journal of the Royal Astronomical Society, 62,
699-718.
MARTINI, E. 1971. Standard Tertiary and Quaternary cal-
careous nannoplankton zonation. In: FARINACCI,A.
(ed.) Proceedings of the H Planktonic Conference,
Roma 1970. Edizioni Tecnoscienza, Rome, 739-85
MCFADDEN,P. L. & REID, A. B. 1982. Analysis of palaeo-
magnetic inclination data. Geophysical Journal of
the Royal Astronomical Society, 69, 307-319.
MtLSOM, J. 1992. Structure and collision history of the
Buton continental fragment, eastern Indonesia
(abstract). AAPG Bulletin, 76, 1117.
 SE SUNDALAND ACCRETION
SIKUMBANG,N. & SANYOTO,P. 1981. Geologic map of the
Buton and Muna Quadrangle, southeast Sulawesi,
Scale 1:250,000. Geological Research and
Development Centre, Bandung.
SILVER, E. A., MCCAFFRE¥, R. & SMITH, R. B. 1983.
Collision, rotation and the initiation of subduction in
the evolution of Sulawesi, Indonesia. Journal of
Geophysical Research, 88, 9407-9418.
SMITH, R. B. 1983. Sedimentology and tectonics of a
Miocene collision complex, and overlying late
orogenic clastic strata, Buton Island eastern
Indonesia. PhD thesis University of California
Santa Cruz, USA.
-- & SILVER, E. A. 1991. Geology of a Miocene
443
collision complex, Buton, eastern Indonesia.
Bulletin of the Geological Society of America, 103,
660-678.
WIRYOSUJONO, W. & HAINIM, J. A. 1978. Cainozoic
sedimentation in Buton Island. In: WIRYOSUJONO,
W. & SUDRAJAT, A. (eds) Proceedings Regional
Conference on the Geology and Mineral Resources
of Southeast Asia, Jakarta, August 1975. Indonesian
Association of Geologists, Jakarta, Indonesia,
109-119.
ZUDERVELD,J. D. A. 1967. AC demagnetization of rocks:
analysis of results. In: COLLINSON,D. W., CREER,K.
M. & RUNCORN, S. K. (eds). Methods in
Palaeomagnetism. Elsevier, New York, 254-286.
 Pb and Nd isotope constraints on the provenance of tectonically
dispersed continental fragments in east Indonesia
P Z. V R O O N 1,2, M. J. V A N B E R G E N 1 & E. J. F O R D E 2
1 Faculty of Earth Sciences, University of Utrecht, PO Box 80.021, 3508 TA Utrecht,
The Netherlands.
2 SE Asia Research Group, Department of Geology, Royal HoIloway University of London,
Egham Hill, Egham, Surrey TW20 OEX, UK.

Abstract: Microcontinental fragments, tectonically dispersed at the triple junction of the

converging SE Asian, Indo-Australian and Pacific plates, are a widespread phenomenon in the
east Indonesian region. This paper investigates the potential of Pb-Nd isotope characteristics of
igneous and (meta) sedimentary rocks from these fragments as indicators of their provenance.
From the correspondence of isotopic signatures, it appears possible to identify provenance areas
of four selected microcontinents, which were emplaced in different settings and regions of
east Indonesia since late Mesozoic times. According to their isotopic affinities, the blocks of
Ambon-Seram, Bacan, the Banda Ridges and Sumba originally formed part of southern New
Guinea, north Australia, 'Pacific' New Guinea and Sundaland respectively. These conclusions are
in line with geological constraints on the palaeopositions of the fragments. The results thus
indicate that isotopic signatures are potentially a powerful tool in reconstructions of the tectonic
history of east Indonesia.

Relatively small continental and oceanic fragments

form a conspicuous feature in the late Cenozoic
tectonic evolution of east Indonesia. They represent
tectonically dispersed elements which are thought
to be composed of: (1) Australian-New-Guinean
continental crust; (2) SE Asian arcs; (3) Pacific
oceanic crust; (4) Indian oceanic crust (Hamilton
1979; Bowin et al. 1980; Silver et al. 1985;
Nishimura & Suparka 1986; Audley-Charles &
Harris 1990; Hartono 1990). Tectonic transport of
these fragments sometimes occurred over large
distances (e.g. Hamilton 1979; Silver et al. 1985).
Hence, their origin is difficult to assess and has
been widely debated.
Many of the microcontinental fragments
identified so far (Fig. 1) have been considered to be
related to the Mesozoic break-up of Gondwana
which created the northern margins of the
Australian and New Guinean continent (e.g. Pigram
& Panggabean 1984). Subsequent convergent
movements of the Indian-Australian and Pacific
plates towards SE Asia led to the involvement of
Fig. 1. Distribution of continental basement in
these margins in collisions, and to the accretion of eastern Indonesia (after Hutchison 1989). Australia-
the allochthonous continental fragments in recent New Guinea form a large continental plate in which the
times (e.g. Audley-Charles & Harris 1990). northern edge is currently subducted beneath the Banda
As an alternative mechanism, detachment of arc. In the area northwest of the Australian plate lie
continental blocks may have occurred by strike-slip numerous continental fragments. The larger
faulting induced by plate convergence. Based on are: (1) western arm of Sulawesi; (2) Banggai-Sula
this model, it has been proposed that many spur; (3) Buton island; (4) Sumba; (5) Bacan;
continental fragments (e.g. the Banda Ridges, (6) Ambon-W. Seram islands; (7) Banda Ridges.
Banggai-Sula, Obi, Bacan, eastern Sulawesi,

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolutionof SoutheastAsia, 445

Geological Society Special Publication No. 106, pp. 445--453.
446 P.Z. VROON E T AL.
Buton, Buru, Seram) originated from northern New
Guinea and moved westward to their present
position along left-lateral strike slip faults such as
the Sorong fault system (see Visser & Hermes
1962; Hamilton 1979; Silver et al. 1985; Hall et al.
1991).
Tectonic reconstructions (Daly et al. 1987;
Audley-Charles et al. 1988; Metcalfe 1988; Rangin
et al. 1990) account for the large-scale evolution of
continental masses, but the precise history of
the small fragments remains largely obscure. As
pointed out by Hartono (1990), the microcontinents
probably do not all have the same origin. Some may
not have drifted far and remained relatively close to
a northwest Australian origin (Timor, Tanimbar,
Seram), whereas others (Buton, Sula, Banda
Ridges) may originate from northern New Guinea
and have moved over considerable distances along
the Sorong fault system (Fig. 1).
In this paper, the applicability of isotopic data is
tested in assessing the provenance of four micro-
continents: Ambon-W. Seram, Bacan, the Banda
Ridges and Sumba. The evidence is based on a
comparison of Pb-Nd isotopic signatures between
(meta-) sedimentary or volcanic rocks from the
microcontinents and possible provenance areas.
Analytical techniques
Fine-ground samples were dissolved in a HF:HNO 3
(4 : 1) mixture at 180°C for at least a week. Nd was
separated using a standard ion exchange technique
and measured on a variable multicollector VG354
mass spectrometer at RHUL. A quadruple jump
experiment was used (Thirlwall 1991) and 21 runs
of the Nd in house standard ('Aldrich') yielded
0.511422_+ 11 (2 s.d.). The average whole
procedure Nd blank was 250 pg.
The Pb separation procedure was similar to the
one described by White & Dupr6 (1986). The Pb
isotopes were measured on a variable multi-
collector VG354 mass spectrometer at RHUL in
static mode. The results were corrected for mass
fractionation (1.3%o per amu) based on replicate
runs of NBS SRM-981. The average total
procedure Pb blank was 300 pg.
Isotopic characteristics of continental
provenance regions
The microcontinents originate from the former
continental margins of Australia-New Guinea or
Sundaland. In terms of Pb-Nd isotopic compo-
sitions, significant variation exists within the conti-
nental masses of northern Australia and New
Guinea, largely due to major variations in the age of
outcropping rocks. Ancient formations of Middle
Proterozoic age (2200-1400Ma) are dominant
in northern Australia, while Phanerozoic ages
(<600 Ma) are characteristic of western New
Guinea. Palaeozoic rocks occur in the Bird's Head
and the Snow Mountains, and may be overlain by
formations up to Miocene in age. The northern part
of New Guinea consists mainly of island-arc
volcanics and subordinate oceanic basalts which
collided with southern New Guinea during the
Miocene (Hamilton 1979). As a result of this
arc-continent collision, the central mountains of
the island were uplifted, which triggered sediment
deposition in the basins of southern New Guinea.
The terrigenous fractions of surface sediments
present on the shelf around the Banda arc carry the
isotopic signatures of the adjacent continental
areas. From a geochemical survey along and across
the arc, the following major domains have been
distinguished (cf. Vroon 1992; Vroon et al. 1995):
(1) north Australia, with very high 2°6pb/2°4pb (up
to 19.57) and low J43Nd/144Nd (=0.51190-
0.51200), corresponding to the Proterozoic
complexes on the mainland.
(2) western New Guinea, with low 2°6pbfl°4pb
(--18.6-19.0), relatively high 143Nd/144Nd
(--0.51218-0.51225), representing probably
Palaeozoic/Mesozoic complexes of New
Guinea.
The Pb average composition of Bird's Head (Fig. 1)
has a lower 2°6pb/2°4pb (18.60-18.75) than
southern New Guinea (18.75-19.0), allowing a
distinction between northern New Guinea (Bird's
Head) and southern New Guinea provenances
(Figs 2 & 3).
As will be discussed below, Sundaland and the
Pacific realm of New Guinea are two other
domains, which should be considered as possible
origins for the fragments studied here. Their
isotopic characteristics are less well defined than in
the case of northern Australia and west New
Guinea. Sundaland cratonized in Late Triassic
times and is composed of welded cratonic terranes
and younger orogenic belts (Metcalfe 1988;
Hutchison 1989). It is thus reasonable to assume
that Pb isotopes are on average less radiogenic than
in the much older Proterozoic domain of northern
Australia. The isotopic signatures of sediments
from the Pacific Ocean (Figs 2 & 3) can be used
as reference for a 'Pacific' origin.
Isotopic affinity and origin of
microcontinents
The isotopic characteristics of sediments and
volcanics from the continental fragments can be
used to relate their provenance to one of the above-
mentioned domains. For terrigenous sediments
 PROVENANCE OF CONTINENTAL FRAGMENTS 447

which were deposited prior to, or simultaneous

A Bacan
dacites with, the detachment of a fragment, it is assumed
AUSTRALIA that their Pb-Nd isotopic signatures correspond to
15.8.
Ambon
those of the formerly adjacent continental region.
g_ The isotopic ratios of volcanics provide an indirect
means to assess the affinity of a fragment. Given
15.7 Sumba ~,. ~ ' ~ ~ Ind. . . . ian
the large contrast in Pb concentrations between
Sulawes ts

o,1
mantle and continent-derived material, the Pb
isotopes are particularly sensitive to contamination,
15.6.
such that ratios observed in volcanic rocks tend
to be dominated by the 'continental' signature,
New BritaintlJ~ Bna da ridges meta sediments ff Error even if the degree of contamination is small.
15.5
18.5 i i
18.0 190 19.5 20.0 Contamination of a subduction-related magma may
206 P b / 2 0 4 p b
occur either in the mantle source through addition
of subducted sediments, or by assimilation of
B Bacan , ~ crustal rocks from the overriding plate during
dacites

AUSTRALIA ascent. Hence, assimilation (as well as crustal

remelting) yields information on the isotopic
#_ Ambon characteristics of the basement. On the other hand,
39.5"
~o
P,o,,,oo. . . . \/ esian
source contamination provides a signature of the
crustal material which entered the subduction
#_ . . . . . sediments
39.0 - system when volcanism was active. The volcanic
sour, Banda arc, where the Pb-Nd isotopic signatures
.............. NEWGUINEA
38.5- :: NORTH
reflect the recent input of sediments from the
surrounding passive Australian margin, represents
New Britain Sulawesi Sumba ZTError a clear example of this case (Vroon et el. 1993; Van
38.0 i , p
18.0 11~.5 19.0 19.5 Bergen et al. 1993).
206 P b / 2 ° 4 p b New and published isotope data for volcanic
Fig. 2. (a) 2°6pb/2°4pb vs. 207pb]2°4pb and (b) and sedimentary rocks from Ambon, Bacan, the
2°6pb]2°4pb vs. 2°8pb/2°4pb showing Pb isotopic Banda Ridges and Sumba are given in Table 1. A
composition of Bacan, the Banda Ridges, Ambon and comparison between the respective micro-
Sumba, and possible provenance areas (north New continents and potential provenance regions is
Guinea, south New Guinea and Australia). Data sources presented in Figs 2 & 3.
are given in Table 1 - New Britain volcanic arc:
Woodhead & Johnson (1993), Pacific Ocean sediments
and east Indonesian sediments: Vroon et al. (1993; A m b o n ( a n d W. S e r a m )
1995) and references cited therein.
The island of Ambon forms part of the extinct
northern end of the volcanic Banda arc. Plio-
Pleistocene volcanic rocks, which cover most of the
island, and associated granitic intrusives have been
I - M O ~ ~ New Britain Arc dated at 3.4-4.5 Ma (Priem et al. 1978; Abbott &
0.5130 ~ volcanics
Chamalaun 1981). Exposed basement rocks are
unmetamorphosed Triassic sediments but the wide-
0.5128 IB~anda-- ridges
mete sediments spread presence of polymetamorphic xenoliths in
~.~Z o.5126-" Sul . . . . i .........~.~. "~11 the volcanics points to the presence of a folded
Pacific ~ Sumba metamorphic deeper basement similar to form-
Z 0.5124 Ocean ---..... ~
sediments ations of southern Seram (Van Bergen et al. 1989;
~ O ~ G u I NSOUTH
E A Bacanl unpublished data Utrecht). According to Pigram &
,m,J
NORTH ~ ~ ~
...... NEWGUINEA/..,,~ ~ ~J1 Panggabean (1984), Ambon forms part of a
...... Ambon "~. ~ [ Buru-Seram microcontinent, which was derived
0.5118 i
East ,rid,. . . . ian sedimen~,AUST,RAL~I from the northeast sector of the former Australian
17.o 17'.s 1~.0 185 ~.o 19.5 ~.o continental margin and rifted away during the
206pb/204pb Middle Jurassic. However, different views exist on
Fig. 3. 2°6pb/2°4pbvs. 143Nd/144Ndshowing the how much of Seram can be considered as a micro-
possible provenance areas and the Bacan-Ambon crustal continental block, and on whether Buru may have a
contaminated volcanics and Banda Ridges structurally separate origin (see Hamilton 1979;
metasediments. I-MORB = Indian Mid-Oceanic Ridge Audley-Charles et al. 1979; Tjokrosapoetro &
Basalt. References as in Fig. 2. See text for discussion. Budhitrisna 1982). In this paper we use the isotopic
448 p . z . VROON ET AL.

Table 1. Nd-Pb isotopic compositions of sediments and volcanics of east Indonesian microcontinents

Sample Type 143Nd/144Nd 2°6pb/2°4pb 2°Tpbfl°4pb 2°8pb/2°4pb

Ambon
AM3A granite 0.512099 -+ 05 18.893 15.695 39.172
AM7 granite 0.512124 -+ 05 18.874 15.681 39.114
18.878 15.683 39.119
AM26B andesite 0.512168 + 06 18.871 15.676 39.098
18.861 15.663 39.053
AM32I rhyolite 0.512111 _+21 18.865 15.676 39.094
AM48I dacite 0.512080 ___07 19.001 15.709 39.387
19.017 15.725 39.435
AM93Al-I andesite 0.512181 + 07 18.886 15.678 39.128
AM 104A 1 andesite 0.512131 -+ 05 18.842 15.676 39.083

Bacan
BE2 dacite 0.512125 _ 05 19.934 15.797 40.200
BE6 dacite 0.512150 _ 05 19.865 15.807 40.165
BE7 dacite 0.512104 -+ 08 19.911 15.795 40.172
BE9 dacite 0.512161 -+ 06 19.907 15.805 40.193
BC-3 dacite 0.51215 19.88 15.80 40.17
BC-6 dacite 0.51217 19.94 15.83 40.33

SW Sulawesi (Balangburu Formation)

CJ30.5 shale 0.512463 +_05 18.738 15.647 38.880
CJ68.5 shale 0.512502 +_09 18.668 15.628 38.733
CJ97.4 sandstone 0.512548 _+05 18.730 15.638 38.818

Sumba (Lasipu Formation)

SL.L 1 shale 0.512476 +_05 18.767 15.645 38.883
SL.P1 shale 0.512459 _+04 18.767 15.645 38.883
SL.R3 shale 0.512439 _+04 18.743 15.635 38.846

Banda Ridges
- metasediment 0.51290 18.49 15.55 38.39
- metasediment 0.51260 18.97 15.58 39.17

Data of this study except Bacan (BC-3 and BC-6) and the Banda Ridge samples, which are from Morris et al. (1983)
and Morris et al. (1984), respectively. Errors for 2°6pb/2°npb, 2°Tpb/2°4pb and 2°8pb/2°4pb are _+12, ___12and _+40,
respectively.

data of the A m b o n volcanics to assess the origin of
a continental fragment represented by the basement
of A m b o n and the adjacent western part of Seram
where Palaeozoic m e t a m o r p h i c c o m p l e x e s are
widespread (Linthout et al. 1989). Evidence exists
for a significant anti-clockwise rotation of Seram
during the Late Cenozoic (Haile 1981; Nishimura
& Suparka 1986), but its palaeoposition is still
uncertain.
The acidic volcanic rocks of A m b o n
display typical values of crustally contaminated
rocks: 143Nd/144Nd -- 0.51208-0.51218, 87Sr/86Sr =
0.7105-0.7188 and 2°6pb/2°4pb = 18.7-19.0 (Table
1 and unpublished data RHUL). These data are
consistent with the preliminary results of Morris
(1984). The polymetamorphic xenoliths and the
existence of a crude positive correlation between
SiO 2 and 2°6pb/2°4pb-87Sr/86Sr provide evidence
for assimilation of basement rocks. The most
contaminated rocks have high 87Sr/86Sr (>0.715)
and 2°6pb/2°4pb (>18.80) ratios, the latter over-
lapping the south New Guinea fields (cf. Figs 2 &
3). Pre-Triassic metapelitic and granitic rocks from
the Tehoru Formation and Kaibobo Complex of W.
Seram (cf. Linthout et al. 1989) have Nd and
Pb isotopic ratios which are virtually identical
to the A m b o n volcanics (e.g. 143Nd/144Nd=
0.51204-0.51220 and 2°6pb/2°4pb = 18.74-18.80;
Tommasini et al. 1994). These data support the
supposition that A m b o n and W. Seram share the
same basement and belong to one microcontinental
block derived from south New Guinea.
Bacan
The island of Bacan is situated in the southern part
of the Halmahera arc (Fig. 1). In contrast to the rest
 PROVENANCE OF CONTINENTAL FRAGMENTS
of the arc, high-grade metamorphic continental
basement rocks are exposed in central Bacan
(Van Bemmelen 1949) and may also underlie the
southern part of the island and Obi (Hall et al.
1991). Both islands have been considered to form
one microcontinent (e.g. Pigram & Panggabean
1984), but they may represent different blocks
separated by a splay of the Sorong fault system. It
has been suggested that the continental basement
rocks of Bacan are derived from New Guinea (e.g.
Hamilton 1979).
Dacites of Quaternary age from Bacan have
similar 87Sr/86Sr (0.7098-0.7248) and 143Nd/144Nd
(0.51210-0.51217) as the Ambon volcanics, and
are also considered to represent crustally con-
taminated rocks (cf. Morris et al. 1983). The Sr
isotopes form two different groups (=0.719 and
=0.725) and tend to show a positive correlation
with SiO 2, which provides evidence for assim-
ilation. Pb isotopes are much more radiogenic
than in the case of Ambon, and suggest that
the assimilated component has 2°6pb/2°4pb close
to 19.9. However, outcropping metamorphic
basement rocks of Bacan have lower 2°6pb/Z°4pb
(18.64-19.40) and higher 143Nd/la4Nd
(0.51227-0.51291; unpublished data, E. Forde).
These signatures make these rocks unlikely assimi-
lants. If other formations with highly radiogenic Pb
isotopes are representative of (part) of the Bacan
crust, it is likely that this block originates from
northern Australia, unless ancient Precambrian
basement is present in the basement of the northern
parts of west New Guinea.
Two different options for the timing of the
arrival of continental crust in this area have been
suggested (Hall et al. 1991). Either late Neogene
westward transport of Australian continental crust
along the Sorong fault system resulted in the
juxtaposition with Pacific arc-ophiolitic terranes
along strike-slip faults, or both types of terranes
were first amalgamated by earlier (Oligocene?)
arc-continent collision at the northern 'Australian'
margin before they were incorporated as complex
units in the Sorong fault system. However, a
'Precambrian' Australian, rather than a New
Guinean origin of the Bacan fragment is difficult
to reconcile with these models, as both imply a
provenance in north New Guinea. An alternative
may therefore be that the block had already arrived
in the region before the Sorong fault system
became active.
Banda Ridges
The Banda Ridges (Lucipara Ridges) are relatively
shallow structural platforms in the northern part of
the Banda Sea (Fig. 1). Clastic sedimentary and
449
metamorphic dredge samples, two of which were
dated at 10.8 and 22.5 Ma, revealed the continental
character of the ridges (Silver et al. 1985).
According to these authors, the lithology and age
resemble those of rocks on other microcontinents in
this region, and are particularly similar to the
Tamrau Formation on the Bird's Head, which
contains Jurassic and Cretaceous fossils. Hence, the
Banda Ridges are considered to represent
fragments of the northern margin of New Guinea
which were displaced along a left-slip fault system.
The dredged metasedimentary rocks display the
characteristics of arc volcaniclastics and have
87Sr/86Sr = 0.7047-0.7073 and 143Nd/144Nd =
0.5126-0.5129 (Morris et al. 1984). Their
a°6pb/e°4pb and e°8pb/2°4pb ratios largely overlap
with the New Guinea values, but the a°Tpb/a°4pb
ratios are much lower than observed in any of the
provenance regions defined above. Based on
Pb-Nd signatures (Fig. 3), it seems more plausible
that the origin of the Banda Ridges is related to the
Cenozoic part of New Guinea composed of accret-
ed Pacific Ocean sediments and island arcs (e.g.
Hamilton 1979) rather than to its 'continental'
Palaeozoic-Mesozoic area. The Banda Ridges plot
between Pacific Ocean sediments and the
island-arc volcanics (as represented by the New
Britain arc (Woodhead & Johnson 1993) further to
the east), thus the isotopic data favour a 'Pacific'
origin outside the Palaeozoic or older domains of
New Guinea and Australia (Fig. 3).
Sumba
The island of Sumba is a continental fragment in
the fore-arc region between the East Sunda
volcanic arc and the Java-Timor trench (Fig. 1).
Based on stratigraphic and palaeomagnetic
considerations, numerous studies have related the
provenance of Sumba to different regions: (1) the
Australian continent to the south; (2) the margin of
southeast Sundaland to the north; or (3) an isolated
position within the Tethys region (e.g. Hamilton
1979; Otofuji et al. 1981; Von der Borch et al.
1983; Audley-Charles 1985; Rangin et al. 1990;
Wensink 1994).
Marine sediments were analysed from the Late
Cretaceous Lasipu Formation which are the oldest
rocks on Sumba, and were deposited prior to or
simultaneous with the initial stages of the island's
drift history. They display limited variations in
143Nd/144Nd (0.51244-0.51248) and Pb isotopes
(2°6pb/e°4pb = 18.74-18.77). These isotopic
signatures do not correspond to the Australian or
New Guinean continental domains, and thus favour
a northern rather than a southern origin. Because of
stratigraphic indications for a palaeoposition of
450 P.Z. VROON ET AL.
Sumba near SW Sulawesi (Simandjuntak 1993),
Upper Cretaceous flysch sediments from the
Balangburu Formation of SW Sulawesi were
analysed for comparison (Table 1). They yielded
143Nd/144Nd = 0.51246-0.51255 and 2°6pb/2°4pb =
18.67-18.74, which implies a close isotopic
similarity with the Lasipu Formation. In the
Mesozoic prior to opening of the Makassar Strait,
SW Sulawesi was probably joined with SE
Kalimantan (e.g. Katili 1978; Hamilton 1979) on
the border of Sundaland. The isotope results thus
provide evidence for Sumba's affinity with
Sundaland, and corroborate the same conclusion
from recent paleomagnetic and geochemical work
(Wensink 1994; Wensink & Van Bergen 1995).
Discussion and conclusions
The Pb-Nd isotopic signatures of rocks from the
microcontinents studied here provide new evidence
on their pre-drift origin. Ambon-W. Seram has
affinities with south New Guinea, Bacan with
Precambrian northwest Australia, the Banda Ridges
with 'Pacific' north New Guinea and Sumba with
southeast Sundaland. These findings are generally
compatible with previous hypotheses from
geological correlations, and pose constraints on
various options for palaeopositions and trajectories
that have remained open so far. An important
implication of the findings is that no single
mechanism can be invoked to reconstruct the origin
and emplacement of all of the fragments (cf.
Hartono 1990).
For Ambon-W. Seram only a small north-
westward shift relative to western New Guinea,
possibly accompanied with a significant anti-
clockwise rotation (cf. Nishimura & Suparka 1986;
De Smet 1989; Linthout et al. 1991), is required to
account for the observed south New Guinea-like
isotopic signatures. The 'tectonic shaving of north
New Guinea' mechanism seems less likely
because it would imply a too northern origin. The
interpretation presented here thus concurs with the
suggestion of Pigram & Panggabean (1984), who
proposed that the origin of Seram may have been
along the former margin of the Australian continent
somewhere east of the Joseph Bonaparte Gulf
(Fig. 1).
The inferred origin of Bacan is also difficult to
reconcile with a dominant westward movement
along the Sorong fault system. If its affinity with
Precambrian domains is correct, a major northward
drift of this block, starting along the NW Australian
margin, seems a more plausible direction of
movement. Early dislocation from the provenance
area, perhaps during the Mesozoic rifting of the
Australian margin, may have led to the arrival at its
northern position prior to the initiation of the
Neogene Sorong fault system. Amalgamation with
adjacent blocks is then a relatively young feature,
related to the westward movement of north New
Guinean and Pacific terranes.
The provenance of the Banda Ridges from
northernmost 'Pacific' New Guinea fits with the
mode of emplacement suggested by Silver et al.
(1985), who envisaged the ridges as fragments
of New Guinea which were displaced along a
left-lateral transform system. Although the isotopic
composition of the andesites suggests that
Australian craton material has contributed to the
andesite genesis (Silver et al. 1985), it probably
signals source contamination from subducted
Australian margin sediments, rather than major
assimilation of basement rocks. The present
position of the Banda Ridges south of the
Ambon-W. Seram block has consequences for the
relative timing of emplacement. Irrespective of the
options for Seram's origin, its emplacement must
have occurred after the Banda Ridges had arrived in
the Banda Sea area, perhaps some 5-10 Ma ago (cf.
Silver et al. 1985; Hartono 1990). Any traces of the
fault system along which the ridges moved may
have been obliterated by later tectonic events
associated with the closure of the Banda arc.
Significant movement and rotation of Seram may
still have taken place during the last 3 Ma (Linthout
et al. 1991).
The palaeoposition of Sumba near the border of
Sundaland also has implications for the timing of
the drift history of this microcontinent. Critics of
this northern origin have pointed at the difficulty of
how the microcontinent could have passed the
Tertiary Sunda arc (e.g. Katili 1989). However,
recent palaeomagnetic evidence indicates that
Sumba started to drift in the Late Cretaceous and
had already arrived at or near its present position in
the Early Miocene (Wensink 1994; Wensink & Van
Bergen 1995). Hence, its emplacement may have
been contemporaneous with the initial stages of
development of the adjacent easternmost volcanic
sector of the Sunda arc. Alternatively, the volcanic
front of the proto-East Sunda Arc in this region
may have had a more southerly position (cf. Fortuin
et al. 1994), and subsequently shifted northward.
More detailed age dating of the Tertiary magmatic
rocks is required to distinguish these alternatives.
Based on present plate movements, McCaffrey &
Abers (1991) presented an evolutionary scenario
which illustrates the structural complexity of the
east Indonesian orogenic belt as it will appear after
the next 10 Ma. They emphasized the role of strike-
slip faulting in juxtaposing the continental and
oceanic elements. Their model can also be used to
predict the future complexity of this region in terms
of isotopic heterogeneity, as an example of a
 PROVENANCE OF CONTINENTAL FRAGMENTS 451

The Banda Orogen

in 10 Ma Bacan

~iiiiiiiiiiiiiiiiiiiii~ New Guinea

" • ..... :~:'"~ii'~i ~ .~!~ii!~i~i~i~!~ii~i!~i~i~iii~i~iiiWi~i@~i~i~i~i~ii~i~i~!i~#ii
~ ..
~:::::::iBandai~rc i i i i::i::iii::i::i::i::i::ii ! i:ii:ii i!i i i:

.[.:.:•[.:
~ ~ Australian continental crust
SundaArc
[ ~ Mixture (dominant Australian)
~ Sunda Land continental crust
•

i ~ New Guineacontinentalcrust
. , . , . . . . . . . - .

Pacific arc rocks

~ i
Yiiii!iii }iEiiiiiili;iiiiii:ii!i!ii
~ ~ i !iAusiralia: :::ii}i::i:::i::: SE Asian arc rocks
[5~ Ophiolites
Fig. 4. Hypothetical construction of the Banda arc in 10 Ma time, after McCaffrey & Abers (1991). Shown are some
of the provenance areas of microcontinental fragments. Note that the orogen contains many 'exotic' terranes, which
are difficult to relate to their original provenance areas. See text for discussion. AWS = Ambon-W. Seram.

collision belt containing drifted continental blocks
of different provenances. Figure 4 shows the
predicted configuration based on the scenario of
McCaffrey & Abers (1991). The most striking
feature is that relatively small isotopic enclaves will
be generated: Sumba with a 'Sundaland' isotopic
signature (e.g. relatively unradiogenic Pb) will
become part of a dominantly Australian domain
(radiogenic Pb); the Banda Ridges will be a Pacific-
like element surrounded by (a) arc rocks with Pb
isotope signatures which are largely determined by
subducted passive-margin sediments with New
Guinean and Australian affinities (cf. Vroon et al.
1993), and (b) dominant Australia-type as well as
New Guinea-type domains; Bacan will yield a
pocket of highly radiogenic Pb in a predominantly
Pacific environment. Thus, the positioning of
various terranes, together with superimposed
effects of ageing and denudation, will finally yield
an isotopically highly complex picture when
orogenesis and future cratonization proceed.
It is concluded that isotopic fingerprinting is a
promising tool to delineate the origin of continental
fragments in east Indonesia. In this paper, Pb
and Nd isotopes of (terrigenous) sediments and
crustally contaminated volcanic rocks are used to
relate microcontinents with relatively large source
domains. The apparent success of this attempt can
be ascribed to the large isotopic differences
between the continental masses involved. Further
detailed correlations may yield a greater spatial
precision, particularly if well-dated stratigraphic
intervals in pre-drift deposits of the fragments can
be compared with analogues in continental regions
of smaller size. In addition to isotopic signatures,
trace-element ratios can provide extra constraints
to define the geochemical characteristics of
provenance areas (cf. Vroon et al. 1995). Similar
techniques may also be applicable to study the
origin of accumulated material in the accretionary
wedges, and to distinguish between an Indian
Ocean and Pacific origin of the oceanic terranes.
We would like to thank Dr. H. Wensink and Dr. J. Ali
for providing samples from Sumba and Sulawesi,
respectively. Samples from Ambon were collected during
the 1984-1985 Snellius H Expedition which was jointly
organized by the SOZ (Netherlands Council of Oceanic
Research) and the Indonesian Institute of Science (LIPI).
Fieldwork on Ambon was done in collaboration with
R. Erfan, T. Sriwana, A. D. Wirakusumah, K. Suharyono,
R. E E. Poorter and J. C. Varekamp. Samples from Bacan
were collected during fieldwork (1989-1992 with
assistance of R. Hall, J. Mahaihollo, S. Baker and
D. Agustiyanto) funded by NERC, Royal Society and the
London University SE Asia Research Group. We thank
Simone Tommasini for allowing us to quote isotope data
from unpublished work. Discussions on the geology and
tectonics of east Indonesia with A. Barber and R. Hall
helped to improve some of the ideas in this paper. R. M.
Ellam and M. A. Menzies provided helpful reviews.
452 P.Z. VROON ET AL.
References
ABBOTT,M. J. & CHAMALAUN,E H. 1981. Geochronology
of some Banda Arc volcanics. In: BARBER,A. J. &
WIRYOSUJONO, S. (eds) The Geology and Tectonics
of Eastern Indonesia. Geological Research and
Development Centre, Bandung, Special Publication,
2, 253-272.
AUDLEY-CHARLES, M. G. 1985. The Sumba enigma: is
Sumba a diapiric fore-arc nappe in process of
formation? Tectonophysics, 119, 435-449.
- - & HARRIS, R. A. 1990. Allochtonous terranes of the
Southwest Pacific and Indonesia. Philosophical
Transactions of the Royal Society of London, A331,
571-587.
, BALLANTYNE,P. D. & HALL, R. 1988. Mesozoic-
Cenozoic rift-drift sequence of Asian frag-
ments from Gondwanaland. Tectonophysics, 155,
317-330.
--, CARTER, D. J., BARBER, A. J., NORVICK, M. S. &
TJOKROSAPOETRO, S. 1979. Reinterpretation of the
geology of Seram: implications for the Banda Arcs
and Northern Australia. Journal of the Geological
Society, London, 136, 547-566.
BOWlN, C., PURDY, G. M., JOHNSTON, CH., SHOR, G.,
LAWVER, L., HARTONO,n. M. S. & JEZEK, P. 1980.
Arc continent collision in Banda Sea Region.
AAPG Bulletin, 64, 868-915.
DALY, M. C., HOOPER, B. G. D. & SMITH, D. G. 1987.
Tertiary plate tectonics and basin evolution in
Indonesia. Proceedings of the Indonesian Petroleum
Association. 16th Annual Convention, October
1987, 399-428.
DE SMET,M. E. M. 1989. A geometrically consistent plate
tectonic model for eastern Indonesia. Netherlands
Journal of Sea Research, 24, 173-183.
FORTUIN, A. R., ROEP,T. B. & SUMOSUSASTRO,P. A. 1994.
The Neogene sediments of East Sumba, Indonesia -
products of a lost arc? Journal of Southeast Asian
Earth Sciences, 9, 67-80.
HAILE, N. S. 1981. Paleomagnetic evidence on the
geotectonic history and paleography of Eastern
Indonesia. In: BARBER, A. J. & WIRYOSUJONO, S.
(eds) The Geology and Tectonics of Eastern
Indonesia. Geological Research and Development
Centre, Bandung, Special Publication, 2, 81-87.
HALL, R., NICHOLS, G., BALLANTYNE,P., CHARLTON,T. &
ALI, J. 1991. The character and significance of
basement rocks of the southern Molucca Sea region.
Journal of Southeast Asian Earth Sciences, 6,
249-258.
HAMILTON, W. 1979. Tectonics of the Indonesian region.
US Geological Survey Profesional. Paper, 1078.
HARTONO, H. M. S. 1990. Late Cenozoic tectonic
development of the Southeast Asian continental
margin of the Banda Sea area. Tectonophysics, 181,
267-276.
HtYrcHISON, C. S. 1989. Geological evolution of South-
east Asia. Clarendon Press, Oxford.
KATILI, J. 1978. Past and present geotectonic position of
Sulawesi, Indonesia. Tectonophysics, 45, 289-322.
1989. Review of past and present geotectonic
concepts of eastern Indonesia. Netherlands Journal
of Sea Research, 24, 103-129.
LINTHOUT,K., HELMERS,H. & ANDRIESSEN,P. A. M. 1991.
Dextral strike-slip in Central Seram and 3-4.5 Ma
Rb/Sr ages in pre-Triassic metamorphics related to
Early Pliocene counterclockwise rotation of the
Buru-Seram microplate (E. Indonesia). Journal of
Southeast Asian Earth Sciences, 6, 335-342.
- - , SOPAHELUWAKAN,J. & SURYANILA, E. 1989.
Metamorphic complexes in Buru and Seram,
Northern Banda Arc. Netherlands Journal of Sea
Research, 24, 345-356.
M¢CAFFREV, R. & ABERS, G. A. 1991. Orogeny in
arc-continent collision: the Banda arc and western
New Guinea. Geology, 19, 563-566.
METCALFE, I. 1988. Origin and assembly of south-east
Asian continental terranes. In: AUDLEY-CHARLES,
M. G. & HALLAM, A. (eds) Gondwana and Tethys.
Geological Society, London, Special Publications,
37, 101-118.
MORRIS, J. D. 1984. Enriched geochemical signatures
in Aleutian and Indonesian arc lavas: an isotopic
and trace element investigation. PhD thesis,
Massachusetts Institute of Technology, Cambridge,
MA.
- - - , GILL, J. B., SCHWARTZ,D. & SILVER, E. A. 1984.
Late Miocene to Recent Banda Sea volcanism, III:
Isotopic compositions [abstract]. LOS Transactions,
AGU, 65, 1135.
--, JEZEK, P .A., HART, S. R. & GILL, J. B. 1983. The
Halmahera island arc, Molucca Sea collision zone,
Indonesia: a geochemical survey. In: HAVES, D. E.
(ed.) The Tectonic and Geologic Evolution of
Southeast Asian Seas and Islands, Part 2. American
Geophysical Union, Geophysical Monograph
Series, 27, 373-387.
NISHIMURA, S. & SUPARKA, S. 1986. Tectonic develop-
ment of East Indonesia. Journal of Southeast Asian
Earth Sciences, 1, 45-57.
OTOFUJI, g., SASAJIMA, S., NISHIMURA, S., YOKOYAMA,
T., HADIWISASTRA, S. & HEHUWAT, F. 1981.
Paleomagnetic evidence for the paleoposition of
Sumba Island, Indonesia. Earth and Planetary
Science Letters, 52, 93-100.
PIGRAM, C. J. & PANGGABEAN,H. 1984. Rifting of the
northern margin of the Australian continent and the
origin of some microcontinents in eastern Indonesia.
Tectonophysics, 107, 331-353.
PRIEM, H. N. A., ANDRIESSEN, P. A. M., BOELRtJK,
N. A. M., HEBEDA, E. H., HUTCHISON, C. S.,
VERDURMEN, E. A. TH & VERSCHURE,R. H. 1978.
Isotopic evidence for a middle to late Pliocene age
of the cordierite granite on Ambon, Indonesia.
Geologie en Mijnbouw, 57, 441-443.
RANGIN, C., JOLIVET, L., PUBELLIER, M. & THE TETHYS
WORKINC GROUP. 1990. A simple model for the
tectonic evolution of southeast Asia and Indonesia
for the past 43 m.y. Bulletin Societe Geologique de
France, 8, 889-905.
SILVER, E. A., GILL, J. B., SCHWARTZ,D., PRASETYO,H. &
DUNCAN, R. A. 1985. Evidence for a submerged and
displaced continental borderland, north Banda Sea,
Indonesia. Geology, 13, 687-691.
SIMANDJUNTAK, T. O. 1993. Tectonic origin of Sumba
Island. Jurnal Geologi dan Sumberdaya Mineral,
3, 10-20.
 PROVENANCE OF CONTINENTAL FRAGMENTS
THIRLWALL, M. E 1991. Long term reproducibility of
multicollector Sr and Nd isotope analyses. Chemical
Geology, 108, 85-104.
TJOKROSAPOETRO, S. & BUDHITRISNA,T. 1982. Geology
and tectonics of the Northern Banda Arc. Bulletin of
the Geological Research and Development Centre,
6, 1-17.
TOMMASINI, S., DAVIES, G. R., STAUDIGEL,H. & MANETTI,
P. 1994. Sr, Nd and Pb isotopic data of crustal melts
from the island of Seram, Indonesia: evidence for
disequilibrium melting during anatexis (abstract).
US Geological Survey Circular, 1107, 325.
VAN BEMMELEN,R. W. 1949. The Geology of Indonesia,
vol. 1A. Government Printing Office, The Hague.
VAN BERGEN, M. J., ERFAN, R. D., SRIWANA, T.,
SUHARYONO,K., POORTER,R. E E., VAREKAMP,J. C.
& VROON, P. Z. 1989. Spatial geochemical varia-
tions of arc volcanism around the Banda Sea.
Netherlands Journal of Sea Research, 24, 313-322.
VAN BERGEN, M. J., VROON P. Z., & HOOGEWERFF,J. A.
1993. Geochemical and tectonic relationships in the
East Indonesian arc-continent collision region:
Implications for the subduction of the Australian
passive margin. Tectonophysics, 223, 97-116.
VISSER, W. A. & HERMES,J. J. 1962. Geological results of
the exploration for oil in the Netherlands New
Guinea. Koninklijk Nederlands Geologisch
Mijnbouwkundig Genootschap, Verhandelingen
Geologische Series, 20, 1-265.
WON DER BORCH, C. C., GRADY, A. E., HARDJOPRAWIRO,
S., PRASETYO, H. & HADIWISASTRA, S. 1983.
Mesozoic and Late Tertiary submarine fan
sequences and their tectonic significance, Sumba,
Indonesia. Sedimentary Geology, 37, 113- 132.
453
VROON, P. Z. 1992. Subduction of continental material in
the Banda Arc, Eastern Indonesia: S r - N d - P b
isotope and trace-element evidence from volcanics
and sediments. PhD thesis University of Utrecht,
Geol. Ultraiectina, 90.
---, VAN BERGEN, M. J., KLAVER, G. & WHITE, W. M.
1995. St, Nd and Pb isotopic and trace-element
signatures of the East Indonesian sediments:
Provenance and implications for Banda Arc magma
genesis. Geochimica et Cosmochimica Acta,
59, 2573-2598.
--, WHITE, W. M. & VAREKAMP, J. C. 1993.
Sr-Nd-Pb isotope systematics of the Banda Arc,
Indonesia: Combined subduction and assimmilation
of continental material. Journal of Geophysical
Research, 98, 22349-22366.
WENSINK, H. 1994. Paleomagnetism of rocks from
Sumba: tectonic implications since the late
Cretaceous. Journal of Southeast Asian Earth
Sciences, 9, 51-65.
& VAN BERGEN, M. J. 1995. The tectonic
emplacement of Sumba in the Sunda-Banda Arc:
paleomagnetic and geochemical evidence from the
early Miocene Jawila volcanics. Tectonophysics,
in press.
WHITE, W. M. & DUPRI~, B. 1986. Sediment subduction
and magma genesis in the Lesser Antilles: isotopic
and trace element constraints. Journal of
Geophysical Research, 91, 5927-5941.
WOODHEAD, J. D. & JOHNSON, R. W. 1993. Isotopic and
trace-element profiles across the New Britain island
arc, Papua New Guinea. Contributions to
Mineralogy and Petrology, 113, 479-491.
 4°Ar/39Ar constraints on obduction of the Seram
ultramafic complex: consequences for the evolution
of the southern Banda Sea

KEES L I N T H O U T l, H E N K H E L M E R S 1, J A N R. W I J B R A N S 1 &
J A N D I E D E R I K A. M. VAN W E E S 2
Departments o f Petrology & Isotope Geology I and Sedimentary Geology 2,
Vrije Universiteit, De Boelelaan 1085, 1081HVAmsterdam, The Netherlands

Abstract: On Kaibobo (W Seram), obduction of hot oceanic lithosphere produced high-grade

metamorphism and granite in overthrust continental crust. 4°Ar/39Ar plateau ages of 5.65-
6.0 Ma and 5.4 Ma of muscovite and biotite from the sole, among the youngest so far measured
on single crystals, indicate complete resetting of the Ar-isotope system. Chloritized biotite
crystals gave disturbed spectra and younger ages. Single fusion on biotite from the granite gave
an age of 5.5 Ma. Time limits for the obduction-exhumationhistory of the Kaibobo complex and
sole are approximated by simple 1D forward P-T-t modelling, constrained by PT data and
4°Ar/39Arthermochronology(c. 400°C at 6 Ma; c. 320°C at 5.5 Ma). Best fit results indicate that
the decelerating post-emplacementexhumation of ophiolite and sole began < 8 Ma ago. The rate
during the first 1 Ma of this uplift exceeded 7.5 mm a-1. Undoing 8 Ma of migration in the SW
Pacific and Australian regimes of motion back-tracks Kaibobo to the site where obduction-
emplacement ended: near the southeastern corner of the Banda Sea plate. Similar compositions
and time-settings of the Kaibobo and N Timor ophiolites lead us to postulate that, during the
Early Miocene, slow-rate spreading occurred in the oceanic lithosphere of the southern Banda
Sea, south of the current volcanic arc.

Ultramafic complexes on islands and peninsulas
around SW Seram (Figs 1 & 2) occupy the highest
Neogene thrust sheets, overlying Palaeozoic,
continental, regional-metamorphic rocks (Audley-
Charles et al. 1979; Linthout et al. 1989;
Sopaheluwakan et al. 1992). The ultramafic rock
on Kaibobo is finely banded and isoclinally folded.
At five localities, in sole and ophiolite, crossing
folds in the latter, ENE-directed (062°___25 °)
stretching lineations were measured. Widespread
serpentinization is locally associated with
tremolite-talc-chlorite schist. Net-veining by very
coarse-grained hornblende gabbro and sills of fine-
grained ophitic diabase is common. Pillow basalts,
of unknown petrological affiliation, reported from
Kelang (60 km west of Kaibobo) are 7.6 +_ 1.4 Ma
old (Beckinsale & Nakapadungrat 1979). On the
basis of geochemistry, petrography and thermo-
barometry, Linthout & Helmers (1994) interpreted
the ultramafic rocks as remnants of weakly
depleted, lherzolitic lithosphere, representing an
early stage in oceanic development, about 10 Ma
old at the time of obduction.
Obduction-induced metamorphism in the sole
reached at least 740°C, at 0.4-0.5 GPa, changing
low- to medium-grade metamorphic rocks into
high-grade phyllonites, mylonitic gneisses and
amphibolites. Coincidence of the retrograde part
of the sole's PT-curve with the cooling path of the
ophiolite indicates a joint exhumation history.
Cordierite-bearing granite, intruded into the
ophiolite, contains a significant metapelitic restite
component and its melt formed under the PT
conditions of the metamorphic peak in the sole. On
these grounds Linthout & Helmers (1994) argued
that its melt was generated by obduction-induced
partial anatexis in that sole. Identical Nd and Pb
isotope ratios in the Kaibobo granite and meta-
pelites from the sole (Tommasini et al. 1994)
confirm this conclusion.
The vestige of important mantle-crust interaction
on Seram should be understood within the frame-
work of the intricate dynamics of the Neogene
geotectonic setting of the Banda arc region, where
three major plates interact: the slowly moving
Eurasian plate, of which the Banda Sea plate
currently forms the southeasternmost part; the
Australian plate moving northward at
c. 70 km Ma-1; and the SW Pacific plates in
westward motion at c. 100 km Ma -t (DeMets et al.
1990; Smith et al. 1990). Exotic microblocks can
be attributed to one or the other major plate and
their translations, rotations and deformations have
to be understood as a result of the interaction of the
three large domains of directions of motion (cf.
Rangin et al. 1990). The E-W elongated island of

From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, 455
Geological Society Special Publication No. 106, pp. 455-464.
456 K. LINTHOUTET AL.

Seram, currently in the SW Pacific plates' regime
of motion (Fig. 1), is a continental fragment of
Australian derivation (Hamilton 1979; Pigram &
Panggabean 1984; Vroon 1992). In accord with the
rates and directions of the large plates, and
sustained by palaeomagnetic work by Haile (1979),
suggesting that since the Late Miocene Seram
moved 2 ° northward and rotated 74 ° anticlockwise,
Linthout & Helmers (1994), much like Johnston
(1981), proposed a model in which Seram, as part
of the Australian plate, initially migrated along the
'NNW transform', marking the eastern boundary of
the Banda Sea plate (Fig. 1). Approaching the SW
Pacific plates, Seram underwent ever stronger
westerly-directed stress, exerted on the Australian
plate by the strong westward motion at its northern
boundary; in the Early Pliocene, during transition
from the Australian plate into the regime of the
SW Pacific plates, Seram was rotated (and block-
faulted, Linthout et al. 1989, 1991) into an E-W
orientation, after which it migrated westward to
its current location.
In this scenario, the provenance of Kaibobo's
ophiolite is constrained by the time of its obduc-
tion. Linthout & Helmers (1994) considered rocks
generated by obduction-induced partial anatexis
as time markers for obduction and subsequent
cooling. To this end they used Rb/Sr ages 4.4 Ma
and 3.3 Ma of cordierite-bearing dacite and granite,
respectively, given by Abbott & Chamalaun (1981)
and Priem et al. (1978). These young ages, how-
ever, are not reliable since, in a study comprising
cordierite-bearing Kaibobo granite, featuring partly
resorbed andesine cores in oligoclase phenocrysts
(Linthout & Helmers 1994), evidence was given
for disequilibrium melting with respect to the Sr
isotopes during anatexis in the sole (Tommasini
et al. 1994).

Fault zo..~4
' / /

C/~.ru

_...~.~.-\ o ° o ~/~

0 \~'~ °

AUSTRALIAN
PLATE 10 ° -

5 0 0 km
! I

1 4o , 128o
Fig. 1. Reference map of the Banda arc region. Arrows correspond to plate motion directions. Dashed lines surround
areas exposing ultramafic rock, associated basalts and high-grade metamorphic rocks on Seram (N Banda arc) and
on N Timor and some smaller islands (S Banda arc). Circle indicates the restored site of obduction of the Kaibobo
ophiolite, NE of Tanimbar, where the 'NNW transform' (dotted line) between the Australian and the Banda Sea plates
intersects the outer Banda arc, as marked by the incision in the 1000 m depth contour.
 40AR--39AR DATING OF SERAM OPHIOLITE OBDUCTION 457

To better constrain the Neogene thermal 9.47 wt% K20 and a total dry wt% of 94.59,
evolution of the rocks of Kaibobo, 4°Ar/39Ar ages confirms its freshness; the annite content is 33%.
of biotite and muscovite crystals were measured In BK21D, with muscovite and biotite crystals up
from the sole's high-grade phyllonites and from to 3 mm, the greenish brown biotite apparently is
cordierite-bearing granite. The accurate age data chloritized, having only 8.85 to 6.65 wt% K20 and
prompted a model of the joint exhumation history dry totals of 93.46-90.75 wt%. In the mylonitic
of the ophiolite and its sole, on the basis of Linthout gneiss BK210 muscovite crystals are up to 7 mm
& Helmers' (1994) well-documented PT trajec- and biotite is chloritized. The fine-grained granite
tories, in order to approximate the time and BK18 contains medium-sized crystals of anti-
implicitly the site of obduction. perthitic oligoclase, with partly resorbed cores of
An55_35, and cordierite. The granular matrix of
quartz, poikilitic microcline perthite and dispersed,
Petrography mm-sized, fresh, brown biotite flakes, displays
The sole samples studied, from c. 50 m below the a gradual transition into gneissose xenoliths of
contact with the ophiolite, are sillimanite and spinel-sillimanite-garnet-(corundum)-cordierite-
garnet-bearing phyllonite and gneiss with 'mica biotite gneiss and quartzite.
fish' crystals, typical of mylonitic deformation.
Fibrolitic sillimanite occurs in microfolded and
boudinaged bundles. Since the mica fish show
Analytical results and discussion
inclusions of microfolded trails of fihrolite, their Muscovite and biotite crystals of c. 250-1000 mm
blastesis followed formation and deformation of were separated by hand picking after gentle
sillimanite. As sillimanite was formed by obduction crushing and sieving of rock samples. Twenty
metamorphism only (Linthout et al. 1989), mica crystals were loaded on a 22 mm diameter A1 tray
blastesis must be related to obduction as well. In in 3 mm diameter, 2 mm deep holes. Four positions
BK21A, layers of lattice orientated quartz crystals were loaded with a mineral standard (TCR
contain strongly altered andalusite and feldspar sanidine, K-Ar age 27.92 Ma). The J factor for
clasts. These are interpreted as pre-phyllonitic these experiments was 0.003275 with an un-
relicts of the medium-grade regional metamorph- certainty of 0.4% (lcy). The remaining holes were
ism, characteristic of the Palaeozoic continental filled with single crystals of unknowns. The experi-
rocks of Seram (Linthout et aL 1989). Biotite mental techniques are described more fully by
crystals are typically fresh, red-brown and 2 mm Wijbrans et aL (1995). Following overnight bake-
long, and muscovite crystals are 0.7 mm long. out, the measurement routine started with three
Electron microprobe analysis of biotite, indicating replicate analyses of air argon, and system blanks
were measured following every block of five runs
of unknowns. Laser heating at increasing laser
power levels for one minute were followed by four
q minutes of clean up on the Fe-V-Zr and Zr-A1
alloy getter cartridges and isotopic analysis on an
MAP 215-50 mass spectrometer. All isotopes
Can ~ ...':-.-.=:-:-:-:-:..:::_
_::-_...:::-2..:::-:...Vv-..=
-::-
were measured using an electron multiplier
collector. System blanks were typically m/e:40
4 x 10-17 moles, m/e:39 4 x 10-19 moles, m/e:38
6 × 10-19 moles, m/e:37 1 × 10-17 moles, m/e:36
1 × 10-18 moles, based on an estimated system
~~'~i*~/~
~ [~ obduct ~ioi~n-ii~tenducecl
c;°rldai;rani
sensitivity of 2 × 10-17 moles/mV.
The experiments were carried out as single

~/~;~/~.~__~
~ rnetamorphi rockc ~fusion analyses, and for the samples from the sole
also as incremental heating analyses on single

~"~//***~**** i i Triassic limestone

crystals. The latter technique permits the identi-
fication of excess 4°Ar and of contaminating
phases. This was considered necessary since the

KV~AGE~-'t O----~.thrust,2 Neogene obduction metamorphic sole rocks do

have a Palaeozoic metamorphic history (Linthout
et al. 1989), and some of the biotite samples were
somewhat chloritized. As the incremental heating
Fig. 2. Geological sketch map of Kaibobo, partly after experiments were carried out with rather low ion
Valk (1945); sites of the samples used for this study are beam intensities in the mass spectrometer, on single
indicated. flakes of mica < 5 mg and < 6.0 Ma old, they were
458 I<. L I N T H O U T ET AL.

Table 1. Data age spectrum and single fusion experiments

36Ar(Atm) 37A1 38Ar 39Ar(K) 4°Ar(cor) % 39Ar 4°Ar*/39Ar K % 4°Ar* Age Sd Age

BK21D age spectrum

Ms 0.000066 0.000002 0.000436 0.037242 0.063397 1.4 1.175 69.2 6.93 1.71
0.001266 0.000002 0.013308 1.102730 1.493429 42.9 1,015 75.0 5.99 0.07
0.000102 0.000323 0.004007 0.346613 0.371159 55.9 0,984 91.9 5.80 0.18
0.000039 0.000002 0.001376 0.123768 0.133951 60.5 0.989 91.4 5.83 0.51
0.000242 0.000371 0.004322 0.366767 0.426120 74.3 0.967 83.2 5.70 0.17
0.000158 0.000002 0.003570 0.305908 0.344558 85.8 0.974 86.4 4.74 0.21
0.000101 0.000002 0.004488 0.377270 0.392930 100.0 0,962 92.4 5.68 0.17
Mean 5.90 0.14
BK21D age spectrum
0.026913 0.011127 0.011792 0.973922 8.372794 0.4 25.93 5.0 2,55 0.18
0.007246 0.004526 0.006230 0.548332 2.389034 0.5 40.53 10.4 2.67 0.17
0.006074 0.003972 0.005512 0.452996 2.037368 0.5 52.59 11.9 3.16 0.21
0.005077 0.005041 0.004377 0.382368 1.691693 0.5 62.77 11.3 2.95 0.25
0.003888 0.003587 0.003415 0.291488 1.302205 0.5 70.53 11.8 3.11 0.28
0.006815 0.005940 0.006352 0.536643 2.275960 0.5 84.82 11.5 2.88 0.15
0,001659 0.001115 0.001652 0.137284 0,556490 0.5 88.48 11.9 2.85 0.48
0.004730 0.003458 0.005067 0.432808 1.657361 0.6 100 15.7 3.54 0.23
Mean 2,91 0.21
BK210 age spectrum
Ms 0.000960 0.000002 0.000161 0.011407 0.298540 0.8 1.302 5.0 7.68 6.08
0.000531 0.000002 0.000111 0.009516 0.165896 1.4 0.955 5.4 5.63 6.64
0.000578 0.000002 0.001032 0.089212 0.259008 7.2 0.989 34.1 5,84 0.73
0.000103 0.000002 0.005127 0.425172 0.447379 35.1 0.98 93.2 5,78 0.15
0.000058 0.000002 0.004612 0.389430 0.393628 60.6 0.967 95.6 5.70 0.16
0.000052 0.000023 0.001928 0.160388 0. t 63667 71.1 0.926 90.6 5.46 0.40
0.000049 0.000482 0.001021 0.089340 0.141548 77.0 1.423 89.8 8.39 0.71
0.000026 0.000002 0.004103 0.350976 0.364035 100.0 1.015 97.9 5.99 0.19
Mean 5.83 0.26
BK21A age spectrum
Ms 0.000073 0.000245 0.000322 0.028395 0.043715 8.0 0.777 50.7 4,59 2.05
0.000059 0.000002 0.000434 0.036074 0.043225 18.1 0.716 59.7 4.23 1,62
0.000031 0.000654 0.000254 0.021331 0.023741 24.1 0.682 61.4 4,02 2.72
0.000049 0.000026 0.000373 0,031575 0.044948 33.0 0.963 67.8 5.69 1.88
0.000046 0.000980 0.000820 0.068610 0.124600 52.3 1.617 89.1 9.53 0.84
0.000080 0.000002 0.000724 0.063145 0.085900 70.0 0.985 72.5 5.81 0.96
0.000027 0.000002 0.000423 0.035239 0.038036 79.9 0.851 79.0 5.02 1.68
0.000022 0.000002 0.000821 0.071514 0.082849 100.0 1.066 92.2 6.29 0.82
Mean 6.64 1.22
BK21A age spectrum
0.000490 0.000002 0.000602 0.050376 0.193229 5.6 0.962 25.1 5.68 1.38
0.000192 0.000086 0.001939 0.155976 0.206207 22.9 0.958 72,5 5.65 0.42
0.000061 0.000002 0.001252 0.101022 0.109721 34.1 0.907 83.6 5.35 0.65
0.000076 0.000002 0.001827 0.152322 0.161633 51.0 0.913 86.1 5.39 0.43
0.000081 0.000002 0.002316 0,184141 0.191071 71.4 0.908 87.5 5.35 0.36
0.000050 0.000105 0.002201 0.178189 0.180763 91.2 0.932 91.8 5.50 0.37
0.000065 0.001058 0.000949 0.079223 0.078179 100.0 0.743 75.4 4,38 0.83
Mean 5.41 0.48
BK21A single fusion
Ms 0,000523 0.000002 0.004680 0.400443 0.538321 100.0 0.958 71.3 5.65 0.15
Bi 0.000521 0,000645 0.006704 0.557240 0.673589 100.0 0.932 77.9 5.50 0.11
BK210 single fusion
Ms 0.000881 0.000002 0.012380 1.043467 1.298813 100.0 0.995 80.0 5.87 0.06
Bi 0.014522 0.004796 0.004295 0.357160 4.468706 100.0 0.497 4.0 2.94 0.37
BK21D single fusion
Ms 0.003423 0.000002 0.013080 1.129049 2.159830 100.0 1.017 53.2 6.00 0.08
Bi 0.024772 0.010005 0,013873 1.145038 7,992747 100.0 0.587 8,4 3.47 0.08
B K I 8 single fusion
Bi 0.007065 0.000177 0.000230 0.191014 4.422786 100.0 12,225 52.8 5.50 0.04
Bi 0.016010 0.000498 0.000466 0.429655 10.008335 100.0 12,283 52.7 5.52 0.02
Bi 0.007363 0.000427 0.000293 0.269728 5.477904 100.0 12.243 60.3 5.51 0.02
Mean 5.51 0.02

Intensities of Ar isotopes are given in beam intensity x 10-I 3 amps; 36Atand 39Arare corrected for contribution of a calcium-derivedcomponent, 37Arand 38Ar
for radioactive decay after irradiation, and 4°Ar for the contribution of nucleogenic 4°Ar; the error is given at 1 s level; % 4oAr* represents the proportion
of radiogenic 4OArin the signal. See text for J factor. For the incremental heating experiments a weighted mean age has been calculated, using the inverse of
the variance (l/s)2 of the 40 Ar* /39 ArK as the weighting factor. A standard error of the mean is calculated by dividing the standard deviation by (n) 1/2 where n

is the number of steps contributing to the mean age. The ages were calculated using a decay constant of 5.543 x 10-l° a-1 for 4°K.
 40AR--39AR DATING OF SERAM OPHIOLITE OBDUCTION 459

very sensitive to relatively small disturbances in the The muscovite crystals yielded flat age spectra
data used for correction of the signal such as the of 5.65-6.0 Ma. Muscovite BK21A gave a larger
system blank values or the beam intensity of 36mr. error (6.65 _+ 1.2 Ma, Fig. 3a), but is equivalent
However, despite some obvious outliers, which within its analytical uncertainty. One well-
were attributed to experimental procedure rather preserved biotite crystal yielded an undisturbed age
than to sample inhomogeneity, it has been shown spectrum, indicating an age of 5.4 Ma (Fig. 3d),
(Table 1, Fig. 3) that resetting of the Ar system in comparable to muscovite, but chloritized biotite
both muscovite and biotite was complete, and that crystals contained substantially more atmospheric
chloritized biotite crystals contain large amounts argon and showed increased scatter in apparent
of atmospheric Ar, precluding calculation of their ages and much lower integrated ages (Fig. 3e,
precise ages (Table 1). Table 1). As no significant inhomogeneities were
found in the age spectra, the integrated ages and the
single fusion results may be considered equivalent.
However, the higher beam intensities obtained for
12
the single fusion experiments should yield the most
Age (Ma)
lo
precise ages (Table 1). Muscovite ages vary from
6-5.65 Ma and fresh biotite from the sole yielded
8 an age of 5.5 Ma. Three replicate analyses of
fresh biotite crystals from granite BK18 gave

L
6
5.51 __+0.02 Ma (weighted mean).
4

2 Ms A, 6 . 6 :I: 1 . 2 M a Closure temperatures

0
m
The use of 4°Ar/39Ar data of micas for thermo-
chronology is well established (McDougall &
8
Harrison 1988; see Spear 1993 for recent refer-
6 /-! ences); it requires, however, the determination or
the assumption of the argon closure temperatures
4
(To) for muscovite and biotite. Based on laboratory
Ms O, 5 . 8 3 + 0.26 (Vla and field calibrations, for conditions of slow to
2 intermediate cooling rates, 350°C has been
assumed as Tc for muscovite (Purdy & J~iger 1976;
8 Wagner et al. 1977) and 280°C for biotite with low
to intermediate annite contents (Harrison et al.
6 l ,,
I"-1
I I | | 1985). Tc exceeding 400°C has been estimated for
muscovite of the eastern Alps and 320°C for
4
Ms D, 5 . 9 0 + 0 . 1 4 Ma coexisting biotite (Blanckenburg et al. 1989).
2
Typically, TcS for coexisting muscovite and
biotite from various areas differ by 70-90°C
8 (Blanckenburg et al. 1989; Cosca et al. 1991). The
Bt A, 5.41 _+ 0 . 4 8 Ma crystals used in this study were quite large and their
6 cooling went fast. Since these factors both tend to
shift TcS to higher values (Dodson 1973), we argue
4 that the TcS for the Kaibobo micas should be at the
high end of the range estimated.
2

4
B t D, 2 . 9 1 + 0 . 2 1 Ma Obduction exhumation history and
its P - T - t modelling
As shown in Fig. 4, results of thermobarometry
reveal distinct stages in the uplift of the Kaibobo
ophiolite (Linthout & Helmers 1994). Early
i
e recrystallizations indicate isobaric cooling of the
o 2o 40 6o- 80 io, lherzolite from c. 1050 to c. 950°C at a depth of
%39Ar
25 km (boxes A-B); then the oceanic lithosphere
Fig. 3 a-e. Age spectra of single grains of muscovite was obducted, undergoing isothermal exhumation
(Ms) and biotite (Bt) from samples BK21A-O. (B-E). By the end of obduction emplacement, the
460 K. LINTHOUT ET AL.

' ' ' ,]]:!:~! ,i[!~1 , , , , , , , , , i . . . . . . . . . i . . . . . . . . . i . . . .

iiiiii~ii ~iiii~i~iil G
5.5 ~ i :il;i~i~,_ _ III

iii! ii!i!..... 7 Ma

iiiiiiii iiiiiiiiiiii ._.x

r~ Q.
8 Ma (D
"O a,,,I,
8.25 Ma
8.25 Ma = r
(n iiiiii!~i i!!i!!!i~i~i~ A

ii~i!i iii~iiiiil PC

Q. 3
~i~iii iii~iiiiii
!i~iii iii_>iiiiii
0
m!

9.25 Ma

. . . . i ~L~:i , , . . . . . ~ , I , i i i , , i , i I i i J L I I ' ' I I I t , ,

400 600 800 1000

temperature (oc)

Fig. 4. Best fit ID forward P-T-t model of the Kaibobo obduction and exhumation path. For conditions and
parameters see text. Shaded boxes for the ophiolite and open boxes for the sole after thermobarometry by Linthout
& Helmers (1994); shaded bars mark the thermochronologybased on Ar dating of micas from the sole and obduction-
induced granite. Filled and open circles mark modelled timing for ophiolite and sole, respectively.Model results
indicate a 2 Ma period of post-emplacementexhumation between thermal relaxation of the ophiolite and sole and the
closure temperature of muscovite for the Ar system (c. 400°C). However, cooling by granite melt displacement,
serpentinizationand hydrothermal activity may allow a shorter period (c. 1.5 Ma) of exhumation.

sole and the juxtaposed hot ophiolite equilibrated to of geohistory analysis (De Smet et al. 1989).
a temperature of c. 740°C (E, I-II). After burial by Assuming a similar thickness of sediments
the obducting slab at a depth of 14 km, the'sole was covering the metamorphic basement as found on
lifted again, following a joint post-emplacement nearby Australian fragments Buru and Misool, both
exhumation trajectory with the ophiolite, charac- much less disturbed by Neogene thrusting (Pigram
terized by fast uplift at first (II-F) and by more et al. 1982; Tjokrosapoetro & Budhitrisna 1982),
isobaric cooling later (F-III-G). the sole's metapelites were at a pre-obduction depth
Age data for the early part of the ophiolite's of c. 5 km and rather cool. So, obduction replaced
history are not available but some time constraints a sedimentary pile of c. 5 km by c. 14 km of ophio-
for this period of emplacement may be inferred by lite, effectively burying the sole rocks to a 9 km
considering that obduction took place in a setting deeper level, which is of the same order as that
of collision between the Australian and the Banda inferred for the simultaneous exhumation of the
Sea plates, converging at a rate 6f 70 km Ma -1. The obducted slab.
probable slow spreading, responsible for creating Clearly, a model, searching to fit both the joint
the obducted lherzolite, if still active during obduc- exhumation trajectory (II-F-III-G) and the
tion, may have had an additional effect, but accurate 4°Ar/39Ar thermochronology (c. 400°C
emplacement faster than 90 km Ma -1 is not likely. at 6 Ma and c. 320°C at 5.5 Ma), will constrain
At this rate, considering that the width of the the timing of the sole's reversal from burial to
area covered by remains of ultramafics around uplift, thereby closely approximating the end of
SW Seram (in the ENE direction of the stretching obduction-emplacement. For this purpose we use a
lineations) is c. 100 km (Fig. 1), the emplacement 1D forward P-T-t model, adopting simple kine-
could most probably not have been accomplished in matic scenarios of England & Thompson (1986)
less than about 1.25 Ma. The exhumation of about and Davy & Gillet (1986), in which obduction is
11 km (B-E) of the ophiolite during this 1.25 Ma is modelled by instantaneous saw-tooth shaped geo-
similar to the highest uplift rates measured around thermal perturbation, followed by isobaric thermal
the Banda arc by the palaeontological method relaxation and exhumation is incorporated by
 40AR-39AR DATING OF SERAM OPHIOLITE OBDUCTION
transient uplift assuming instantaneous erosion.
The heat conduction equation is solved using a
finite difference technique (grid spacing 250 m,
cf. Wees et aI. 1992). Heat conduction parameters
of ultramafic and continental rocks are taken as
equal (conductivity = 3.1 W m -1 °C-l; diffusivity =
8.95 x 10-1 m 2 s-l); radiogenic heat production is
neglected. The initial temperature conditions of the
model correspond to a linear geotherm constrained
by surface conditions and 880°C at 25 km (B, D).
The base of the model (viz. the base of the litho-
sphere, 1325°C), is characterized by a constant heat
flow condition of 110 mW m -1, allowing the base
of the lithosphere to migrate.
The dynamic and thermal processes involved in
obduction-emplacement (B-C-D-E, cool sole-I-II)
are too complicated to incorporate in a 1D model.
Therefore, the model was simplified by adopting
three stages for the obduction-emplacement in
1.25 Ma: 1 Ma of isothermal exhumation, uplifting
the oceanic lithosphere from depths of 25 to 14 km;
subsequent, instantaneous, isobaric overthrusting
onto the sole that was already heated to 600°C;
concluded by 0.25 Ma of thermal relaxation at the
contact of the juxtaposed ophiolite and sole to
740°C. Although this part of the tectonic history is
subject to relatively large uncertainties in age and
kinematics, possible variations will not signifi-
cantly influence the estimate of timing of post-
emplacement exhumation - the main objective. The
best fit results for the latter indicate 2 Ma of
decelerating post-emplacement exhumation to
400°C (Fig. 4). The fin'st million years is marked
by rapid uplift of 7.5 mm a-1 (II-F). The slower
uplift of 3.2 mm a-I, from 7 to 6 Ma, is followed by
an isobaric cooling of c. 80°C from 6 to 5.5 Ma,
in perfect accord with the thermochronology of the
micas.
Longer .and shorter histories were also modelled.
Longer exhumation requires a more pronounced
deceleration of exhumation in order to meet the
PT constraints of boxes II-F. A maximum of
2.5 Ma seemed possible, by adopting a 1 Ma phase
of isothermal exhumation followed by isobaric
cooling. However, the cooling rates between 6 and
5.5 Ma, predicted under these assumptions, are
extremely low compared to those inferred from
thermochronology. Only by incorporating renewed
fast uplift in the 6-5.5 Ma interval, can the cooling
rate increase; for this, however, there is no
supporting geological evidence. For shorter periods
of exhumation during the cooling from 740 to
400°C a minimum of c. 1.5 Ma is indicated, again
constrained by boxes II-F. In this case, the pre-
dicted cooling from 6 to 5.5 Ma is extremely fast
(c. 150°C) and clearly not in accord with the
thermochronology. However, granite-melt dis-
placement, metamorphic reactions, serpentinization
461
and hydrothermal activities, active in the 740-
400°C interval, will all tend to speed the cooling.
If these effects were considered in the model, they
would have reduced the predicted cooling rates
in the 400-300°C interval. Consequently a 1.5 Ma
exhumation period seems likely.
Summarizing, this new data and the results of
P-T-t modelling indicate that the post-emplace-
ment exhumation of the Kaibobo complex and
its sole started between 7.5 and 8 Ma ago. The
implied rates of exhumation, prevailing around
7.5 Ma ago, at 7 . 5 - 1 0 m m a -1, are among the
highest reported from the Banda arc (cf. De Smet
et al. 1989).
Site of obduction of the Kaibobo
ultramafic complex
Undoing westward migration in the SW Pacific
regime back-tracks Kaibobo in c. 4 Ma to the
junction of the 'NNW transform' and the Tarara-
Aiduna fault zone, viz. the area where it was being
rotated in the transition from the Australian into the
SW Pacific regime (Fig. 1). The remaining 3.5 to
4 Ma spent in Australian motion brings Kaibobo
back, over between 245 and 280 kin, to where
the sole started its uplift 7.5 to 8 Ma ago. Thus the
inferred locus of obduction is NE of the current
position of Tanimbar, on the bathymetric incision
in the outer Banda arc, marking the 'NNW trans-
form' between the Banda Sea and Australian plate
(Fig. 1).
It is noteworthy that 600 km to the WSW, on
N Timor and also on lesser islands between Timor
and Tanimbar, ultramafites associated with high-
grade metamorphic rocks occur (Bemmelen 1949;
Pertamina-Beicip 1982). The ultramafic rocks
obducted on the N coast of Timor, near Atapupu
and Dili, are of particular interest since the high-
grade metamorphic rocks in the sole near Dill give
4°hr/39Ar ages of 8 and 5.4 Ma for hornblende and
muscovite, respectively (Berry & MeY)ougall
1986). Considering that at Atapupu cooling went
slower since the overriding hot slab was c. 6 km
thicker (Helmers et al. 1989) we have remarkably
similar time settings of obduction for the N Timor
and Kaibobo ultramafic complexes. Moreover, like
in Kaibobo, the Atapupu ultramafites are lherzolitic
and net-veined by coarse-grained hornblende
gabbro (Helmers et al. 1989). Lherzolitic compo-
sition, association with a sole of metamorphic
continental crust, rather than one of oceanic crustal
descent, abundance of dykes and sills, antigorite as
the serpentine mineral formed in a tectonic setting
together with tremolite and talc, and absence of
chromite pods, as observed in Kaibobo as well as
in Atapupu, are all typical features of the lherzolite
462 K. LINTHOUT ET AL.
ophiolite type, corresponding to very slow
spreading, as may be achieved by juvenile
spreading in a transform environment (Boudier &
Nicolas 1985). On N Timor also, there are Upper
Miocene basalts. For pillow basalts of the
Manamas Formation, west of and adjacent to the
Atapupu complex, K/At dating suggests a
minimum age of 6 Ma (Abbott & Chamalaun
1981); a palaeontological age was tentatively
estimated at 7 Ma (Carter et al. 1976); K/At data
provide an age of 10 Ma for Manamas basalts and
dioritic dykes (H. Bellon, pers. comm. 1995). Trace
element analyses of Manamas basalts indicate
compositions close to MORB, not far from OIB
(Vroon 1992).
Early Miocene spreading in the
southern Banda Sea
From the foregoing it appears that during the Early
Miocene (c. 18 Ma ago, i.e. 10 Ma before obduc-
tion) one or several spreading basins were being
formed in the ocean floor in the southem Banda
Sea, to the south of the current volcanic arc.
The origin and age of the Banda Sea have been a
matter of debate. Hamilton (1979) considered the
south Banda Sea an active marginal basin; accord-
ing to R6hault et al. (1994a, b), 6-9 Ma old
volcanic rocks, recently dredged from the north
References
ABBOTT,M. J. & CHAMALAUN,E H. 1981. Geochronology
of some Banda Arc volcanics. In: BARBER,A. J. &
WIRYOSUJONO,S. (eds) The Geology and Tectonics
of Eastern Indonesia. Indonesian Geological
Research and Development Centre, Special
Publication, 2, 253-268.
AUDLEY-CHARLES,M. G., CARTER,D. J., BARBER,A. J.,
NORVICK, M. S. & TJOKROSAPOETRO,S. 1979. Re-
interpretation of the geology of Seram: implications
for the Banda Arcs and northern Australia. Journal
of the Geological Society, London, 136, 547-568.
BECKINSALE,R. D. & NAKAPADUNGRAT,S. 1979. A Late
Miocene K-Ar age for the lavas of Pulau Kelang and
Seram, Indonesia. In: UYEDA,S., MURPHY,R. W. &
KOBAYASHI, K. (eds) Geodynamics of the Western
Pacific. Advances in Earth and Planetary Sciences,
6, 199-201.
BEMMELEN,R. W. VAN 1949. The Geology of Indonesia.
Government Printing Office, Nijhoff, The Hague.
BERRY, R. E & McDOUGALL, I. 1986. Interpretation of
4°Ar/39Ar and K/At dating evidence from the Aileu
Formation, East Timor, Indonesia. Chemical
Geology, 59, 43-58.
BLANCKENBURG, F. VON, VILLA, I. M., BAUR, H.,
MORTEANI, G. & STEIGER, R. H. 1989. Time
calibration of a PT-path from the Western Tauern
Window, Eastem Alps: the problem of closure
temperatures. Contributions to Mineralogy and
Petrology, 101, 1-11.
Banda Sea and from various locations on ridges
in the Banda Sea (north of the volcanic arc) have
geochemical signatures suggesting a process of
backarc opening. Bowin et al. (1980) and Lee &
McCabe (1986), on the other hand, suggested the
Banda Sea is a trapped, Cretaceous-Paleogene
piece of Indian Ocean crust.
Accepting the evidence in favour of an old
oceanic crust, i.e. low heat flow, depth of basement
and interpretation of magnetic lineations, but also
considering the evidence from the obduction
complexes of Seram and Timor, the authors cannot
dismiss r e n e w e d spreading during the Early
Miocene in the Banda Sea, south of the later
developed volcanic arc.
Audley-Charles et al. (1979) noted that the
geology of Seram is in many respects the mirror
image of the geology of Timor. The results here
indicate that the apparent mirror-image relation-
ship between both islands on opposite sides of the
Banda Sea is the fortuitous outcome of a series
of Late Neogene plate-tectonic translations and
rotations of plate fragments caused by the interplay
of three major plates.
We are grateful for stimulating and constructive reviews
by H. Bellon, S. Kelly and J.-P. R6hault. This is publica-
tion NSG 950409 of the Netherlands Research School
of Sedimentology. Reactions to lint@geo.vu.nl (fax
020-6462457) are welcome.
BOUDIER, F. & NICOLAS, A. 1985. Harzburgite and
lherzolite subtypes in ophiolitic and oceanic
environments. Earth and Planetary Science Letters,
76, 84-92.
BOWIN, C. O., PURDY, G. M., LAWYER,L., HARTONO,H.
M. S. & JEZEK, E 1980. Arc-continent collision
in the Banda Sea region. APPG Bulletin, 64,
868-915.
CARTER,D. J., AUDLEY-CHARLES,M. G. & BARBER,A. J.
1976. Stratigraphical analysis of island arc-
continental margin collision in eastern Indonesia.
Journal of the Geological Society London, 132,
179-198.
COSCA, M. A., SUTTER, J. F. & ESSENE, E. J. 1991.
Cooling and inferred uplift/erosion history of the
Grenville Orogen, Ontario: constraints from
4°Ar/39Ar thermochronology. Tectonophysics, 10,
959-977.
DAVY, PH. & PH. GILLET. 1986. The stacking of thrust
slices in collision zones and its thermal conse-
quences. Tectonics, 5, 931-929.
DEMETS, C., GORDON, R. G., ARGUS, D. E & STEIN, S.
1990. Current plate motions, Geophysical Journal
International, 101,425-478.
DE SMET,M. E. M., FORTUtN,A. R., TJOKROSAPOETRO,S.
& VAN HINTE, J. E. 1989. Cenozoic vertical move-
ments of non volcanic islands in the Banda Arc
area. Netherlands Journal of Sea Research, 24,
263-275.
 40AR--39AR DATING OF SERAM OPHIOLITE OBDUCTION
DODSON, M. H. 1973. Closure temperature in cooling
geochronological and petrological systems.
Contributions to Mineralogy and Petrology, 40,
259-274.
ENGLAND, P. C. & THOMPSON, A. B. 1986. Pressure-
temperature-time paths of regional metamorphism
I. Heat transfer during the evolution of regions
of thickened crust. Journal of Petrology, 25,
894-928
HAILE, N. S. 1979. Palaeomagnetic evidence for the rota-
tion of Seram, Indonesia. In: UYEDA, S., MURPHY,
R. W. & KOBAYASHI,K. (eds) Geodynamics of the
Western Pacific. Advances in Earth and Planetary
Sciences, 6, 191-198.
HAMILTON,W. 1979. Tectonics of the Indonesian Region.
United States Geological Survey Professional
Paper, 1078.
HARRISON, T. M., DUNCAN, I. & McDOUGALL, I. 1985.
Diffusion of 4oAr in biotite: Temperature, pressure
and compositional effects. Geochimica et Cosmo-
chimica Acta, 49, 2461-2468.
HELMERS, H., SOPAHELUAKAN,J., TJOKROSAPOETRO,S. &
SURYA NILA, E. 1989. High-grade metamorphism
related to peridotite emplacement near Atapupu,
Timor, with reference to the Kaibobo peridotite
mass on Seram, Indonesia. Netherlands Journal of
Sea Research, 24, 357-371.
JOHNSTON, C. R. 1981. A review of Timor tectonics, with
implications for the development of the Banda
Arc. In: BARBER, A. J. & WIRYOSUJONO, S. (eds.)
The Geology and Tectonics of Eastern Indonesia.
Indonesian Geological Research and Development
Centre, Special Publication, 2, 199-216.
LEE, C. S. & MCCABE, R. 1986. The Banda-Celebes-Sula
basin: a trapped piece of Cretaceous-Eocene
oceanic crust? Nature, 322, 51-54.
L1NTHOUT, K., & HELMERS, H. 1994. Pliocene obducted,
rotated and migrated ultramafic rocks and
obduction-induced anatectic granite, SW Seram
and Ambon, Eastern Indonesia. Journal of
Southeast Asian Earth Sciences, 9, 95-109.
., HELMERS,H. & ANDRIESSEN,P. A. M. 1991. Dextral
strike-slip in central Seram and 3-4. 5 Ma Rb/Sr
ages in pre-Triassic metamorphics related to Early
Pliocene counterclockwise rotation of the Buru-
Seram microplate (E Indonesia). Journal of
Southeast Asian Earth Sciences, 6, 335-342.
- - , SOPAHELUWAKAN,J. & SURYANILA, E. 1989.
Metamorphic complexes in Buru and Seram,
northern Banda Arc. Netherlands Journal of Sea
Research, 24, 345-365.
McDOUGALL,I. & HARRISON,T. M. 1988. Geochronology
and Thermochronology by the 4°Ar/e9Ar Method.
Oxford University Press, New York.
PERTAMINA-BEICIP 1982. Geological Map of Eastern
Indonesia, scale 1: 2,000,000.
PmRAM, C. J. & PANGGABEAN,H. 1984. Rifting of the
eastern margin of the Australian continent and
the origin of some microcontinents in Indonesia.
Tectonophysics, 107, 331-353.
, CHALLINOR,A. B., HASSIBUAN,E, RUSMANA,E. &
HARTONO, U. 1982. Lithostratigraphy of the Misool
archipelago, Irian Jaya, Indonesia. Geologie en
Mijnbouw, 61, 265-279.
463
PRIEM, H. N. A., ANDRIESSEN,P. A. M., BOELRIJK, N. A.
I. M., HEBEDA, E. H., HUTCHISON, C. S.,
VERDURMEN,E. A. TH. & VERSCHURE,R. H. 1978.
Isotopic evidence for a middle to late Pliocene age
of the cordierite granite on Ambon, Indonesia.
Geologie en Mijnbouw, 57, 441-443.
PURDY, J. W. • JAGER, E. 1976. K-At ages of on rock-
forming minerals from the Central Alps. Institute
of Geology and Mineralogy Padova University
Memoirs, 30.
RANGIN, C., JOLIVET, L. t~ PUBELLIER,M. 1990. A simple
model for the tectonic evolution of southeast Asia
and Indonesia region for the past 43 m.y. Bulletin
de la Socidt( gdologique de France (8), XXVI,
889-905.
Rt~HAULT, J.-e., MAURY, R. C., BELLON, H., SARMILI, L.,
BURHANUDDIN,S., ET AL. 1994a. La Mer de Banda
Nord (Indon6sie): un bassin arribre-arc du Miocene
sup~rieur. Comptes rendus de l'Acaddmie Sciences
Paris, 318, s6rie II, 969-976.
, VILLENEUVE, M., HONTHAAS, C., BELLON, H.,
MALOD, J.-A., ET AL. 1994b. New geological
sampling along a transect across the South East
Banda Sea Basin (Indonesia). Tectonic Evolution
of SE Asia Abstracts, December 7-8 1994, The
Geological Society London, 49.
SMITH, D. E. & NINE OTHERS. 1990. Tectonic motion
and deformation from satellite laser ranging to
LAGEOS. Journal of Geophysical Research, 95,
22 013-22 041.
SOPAHELUWAKAN, J., LINTHOUT, K., HELMERS, H. &
PERMANA, H. 1992. Peridotite-metamorphite
relation in West Seram: Constraints to the vertical
movements of the North Banda Arc. Proceedings of
the Indonesian Association of Geology, XXI Annual
Scientific Meeting, Yogyakarta, December 7-10
1992, 599-609.
SPEAR, E S. 1993. Metamorphic phase equilibria and
pressure-temperature-time paths, Mineralogical
Society of America, Washington D. C.
TJOKROSAPOETRO, S. & BUDHITRISNA,T. 1982. Geology
and tectonics of the northern Banda Arc. Bulletin
Indonesian Geological Research and Development
Centre, 6, 1-17.
TOMMASINI, S., DAVIES, G. R., STAUDIGEL,H., HELMERS,
H., MANETTI, P. & POLl, P. 1994. Sr, Nd, and Pb
isotope data of crustal melts from the island of
Seram, Indonesia: evidence for disequilibrium
melting during anatexis. US Geological Survey
Circular 1107, Abstracts of the Eighth International
Conference on Geochronology, Cosmochronology
and Isotope Geology, 325.
VALK, W. 1945. Contributions to the geology of West
Seram. PhD Thesis, University of Utrecht, De
Bussy, Amsterdam.
VROON, P. Z. 1992. Subduction of continental material in
the Banda Arc, Eastern Indonesia. PhD Thesis,
University of Utrecht, Geologia Ultraiectina, 90.
WAGNER, G. A., REIMER, G. M. & J.~GER, E. 1977.
Cooling ages derived by apatite fission track, mica
Rb/Sr and K/Ar dating: the uplift and cooling
history of the Central Alps. Institute of Geology
and Mineralogy, Padova University Memoirs, 30,
1-27.
464 K. LINTHOUT ET AL.

WEES, J. D. VAN, DE JONG, K., & CLOETINGH, S. 1992.
Two-dimensional P - T - t modelling and the
dynamics of extension and inversion in the
Betic Zone (SE Spain). Tectonophysics, 203,
305-324.
WIJBRANS, J. R., PRINGLE, M. S., KOPPERS, A. A. E &
SCHEVEERS, R. 1995. Argon geochronology of
small samples by laser heating. Proceedings
van de Koninklijke Nederlandse Academie van
Wetenschappen, 98 (2), 185-218.
 Correlation of the Salawati and Tomori Basins, eastern Indonesia:
a constraint on left-lateral displacements of the Sorong fault zone

TIM R. C H A R L T O N
Department of Geology, Royal Holloway University of London,
Egham, Surrey TW20 OEX, UK
Present address: Ridge House, 1 St. Omer Ridge, Guildford, Surrey GU1 2DD, UK

Abstract: The Salawati Basin of western New Guinea and the Tomori Basin of eastern Sulawesi,
Indonesia, are two sedimentary basins located either side of the main fault strands in the Sorong
fault zone strike-slip system. It is suggested that prior to displacement on the Sorong system the
two formed a single sedimentary basin. Movement on the Sorong system occurred largely during
the latest Miocene-Quaternary, contemporaneous with deposition of a clastic sedimentary
succession. An older basinal sequence, essentially Miocene in age, is composed predominantly
of carbonate sediments, and this may have formed part of the foreland basin sequence related to
the east Sulawesi orogenic belt. Correlation of the Salawati and Tomori Basins implies a left-
lateral displacement of about 900 km on part of the Sorong fault zone.

The Salawati Basin of northwestern Irian Jaya a broad zone of regional left-lateral shear usually
(Indonesian western New Guinea) is the most known as the Sorong fault zone (Figs 1 & 2). The
important petroleum producing basin in eastern Sorong fault zone extends between the Bird's Head
Indonesia. The northern edge of the basin is region of Irian Jaya in the east and the island of
truncated by the Sorong fault, one fault strand in Sulawesi in the west, a distance in excess of

Continental shelves: 120°E 130°E 140°E

Eurasian
PACIFIC OCEAN
Australian

M.S.C.Z. Molucca Sea III

Collision Zone ~...
/ .4":

Sorong ,, r~

1BUTON NEW GUINEA

• ~ . " ° ° °

JAVA % " ,~
~ I d 8~b -'~ ' . ° . ° . . . , •

"%°°s -..-....,~ . . . . . . . _~ At

INDIANOCEAN /
/.'.'.'.'.'~
/ . . . . . AUSTRALIA EU
I / " I . . . . 7_ "" : :1

Fig. 1. Regional tectonic setting of the Sorong fault zone. Inset: relative convergence vectors for the Eurasian (EU),
Indo-Australian (AU) and Pacific (PA) plates in the eastern Indonesia region.

From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, 465
Geological Society Special Publication No. 106, pp. 465-481.
466 T.R. CHARLTON

[-~ Platform
Blank onshore: Post-orogenic basins ~ Imbricated Terranes of 'Australian'
stratigraphic affinity
Terranes of non-Australian origin ('allochthon') ~ Basement
0 200 400 km
m m m m m

ang ai islands I1~ "~ o. S ....... .

• "" " ........'~::::': W T ....... Inan Jaya
~'kj ~ Bird's Head')

' iso

"'~-SO Ut'' \-Su,abesi __,7_r ' - " " .- ~- ~_ S e r a m ~ " ' ~ ~

Fig. 2. Structural setting of the Salawati and Tomori Basins and the Sorong fault zone. Bathymetry shaded at 1000 rn
intervals adapted from R6hault et al. (1991), Smith & Silver (1991), Milson et al. (1992) and Hamilton (1979).

1500 km. The total offset on this fault system is
poorly constrained, but most estimates are in the
range of many hundreds to several thousands of
kilometres (e.g. Visser & Hermes 1962; Tjia 1973;
Pigott e t al. 1982; Pigram & Panggabean 1984).
This poor constraint results primarily from the
absence of unambiguous displacement markers
either side of the fault zone. The main aim of this
paper is to suggest that prior to the development of
the Sorong fault system the Tomori Basin, located
off the east coast of Sulawesi, formed the northern
half of a single sedimentary basin with the Salawati
Basin. This implies a displacement on the Sorong
system of approximately 900 km. This interpret-
ation has important implications for regional
tectonic evolution and hydrocarbon exploration in
eastern Indonesia, both of which will be considered
in this paper.
Regional tectonic setting
Eastern Indonesia is situated in the zone of inter-
action between three of the Earth's major tectonic
plates (Fig. 1): the Eurasian, Indo-Australian and
Pacific plates (the Philippine Sea and Caroline
plates, which separate the Pacific plate proper from
the eastern Indonesia region, have only small
motion vectors relative to the Pacific in this region,
and for the purposes of the present paper can be
treated as sub-plates of the main Pacific plate).
Relative to the mantle, Eurasia is nearly static,
whilst the Indo-Australian plate is moving towards
the NNE, and the Pacific plate is approaching
Eurasia in a WNW direction (Fig. 1). The Sorong
Fault Zone is a transcurrent boundary between the
Pacific and Indo-Australian plates, with a
WSW-ENE orientation sub-parallel to the relative
movement vector between these two plates.
One important effect of the Sorong transcurrent
system has been to strip continental terranes from
the northern margin of the Australian continental
block and transfer them westward as elements of
the Pacific plate. Subsequently these allochthonous
terranes have collided with and been accreted into
the western margin of Eurasia. Such displaced
continental terranes with Australian stratigraphic
affinity include Buton, southeast and east Sulawesi,
the Banggai-Sula block and southwest Obi island
(Figs 1 & 2; e.g. Pigram & Panggabean 1984). The
Tomori Basin, one of the primary focuses of the
present paper, is a successor basin developed on the
Banggai-Sula continental fragment.
 SALAWATI& TOMORI BASIN CORRELATION
The geology of the Sorong fault zone is rela-
tively poorly known (Hall et al. 1991), and various
lineaments have been proposed as the main fault
strands in this system. The fault pattern preferred
by the present author is shown in Fig. 2. In the east,
the Sorong fault sensu strictu forms a well-defined
east-west fault zone some 10-20 km broad through
the Bird's Head. West of the Bird's Head, the author
interprets the main fault strands to be the direct
westward extension of the Sorong fault south of
Obi island to immediately east of Sulabesi island
(here named the Sorong-Sulabesi fault), and the
South Sula fault (the western half of the South
Sula-Sorong fault of Hamilton 1979) west of
Sulabesi island. Present-day seismicity in the
Sorong fault zone is primarily associated with the
Sorong-Sulabesi fault (Kertapati et al. 1992). At its
western end, this seismically active belt connects
via Sulabesi island southeastwards to a NW-SE
trending belt of shallow level seismicity through
Seram island, possibly the right-lateral fault system
identified by Linthout et al. (1991); and northward
into seismicity associated with the eastern margin
of the Molucca Sea collision zone (Fig. 1). The
island of Sulabesi has a structure controlled by N-S
trending normal faults, and marks a right step
between the presently active and inactive strands
of the main Sorong fault system. Sulabesi has
probably been rotated some 90 ° anticlockwise rela-
tive to the main Sula islands to the north as a result
of transtension. Other fault strands, such as the
North Sula-Sorong fault and the Molucca-Sorong
fault proposed by Hamilton (1979) and others may
or may not have had significant displacements in
the past, but are not active at the present day. These
faults, however, lie north of the Banggai-Sula
displaced terrane which is the primary concern of
the present paper, and any movement on these
faults would be additional to the strike-slip offsets
interpreted here.
A number of estimates have been made as to
when the Sorong fault zone became an active
feature. These include the Oligocene (Pigott et al.
1982), Early Miocene c. 25 Ma (Hermes 1968),
Early Miocene (Tjia 1973), Early-Mid Miocene
(Hamilton 1979), post Mid Miocene (Visser &
Hermes 1962), Early Pliocene (Dow & Sukamto
1984) and mid Pliocene (Froidevaux 1977). As will
be discussed in more detail subsequently, it is
suggested that the fault system has been active
since the Late Miocene (c. 6-8 Ma).
In the following sections, the geology of the
Salawati and Tomori basins will be outlined, and
then the points of similarity that lead to the con-
clusion of a common origin will be discussed. This
will be followed by discussion of the regional
tectonic and hydrocarbon exploration implications
of this interpretation.
467
The Salawati Basin
The Salawati Basin, located between the western
margin of the Bird's Head and the island of Misool
(Figs 2 & 3), is essentially a Neogene feature,
although the older basinal succession is concordant
with an underlying Palaeogene shelf succession.
Two distinct phases of basinal development are
apparent: a Miocene phase in which carbonate
sedimentary environments predominated, and a
Pliocene-Recent phase dominated by clastic
sedimentation.
Structural setting
Total Tertiary isopachs for the Salawati Basin
define a NW-SE trending trough which is trun-
cated in the north by the Sorong-Sulabesi fault
(Fig. 3). The eastern margin of the basin is found in
the island of Salawati and the northwest corner of
the Bird's Head, whilst the western margin is
marked by the island of Misool. As mentioned
above, the basin is readily divisible into Miocene
and Plio-Quaternary sub-basins. The Plio-
Quaternary basinal depocentre is situated midway
between Salawati and Misool islands, whilst the
Miocene depocentre was located beneath the
present-day Sele Straits separating Salawati island
from the Bird's Head (Fig. 4).
The Miocene carbonate basin beneath Salawati
island and the NW corner of the Bird's Head had
a semi-enclosed horseshoe shape open to the NW,
with deep-water environments surrounded to the
NE, east and south by shallower carbonate shelf
(e.g. Vincelette & Soepardjadi 1976; Gibson-
Robinson et al. 1990; see Fig. 10). No significant
faulting is recognized in association with this
earlier basinal phase, which seems to have formed
by essentially passive downwarping after
deposition of the shallow marine Oligocene Sirga
Formation (described subsequently). The main
fault-related structural elements of the present-
day basin developed during accumulation of the
younger, clastic basinal sequence. According to
Cockcroft et al. (1984), regional tilting of the basin
occurred after deposition of the Miocene Kais and
Klasafet formations, and this was followed during
the mid Pliocene-Pleistocene by development of
N-S trending normal faults. E-W trending folds,
found in particular near the Sorong fault, are also
of Pliocene age (Gibson-Robinson et al. 1990).
Pre- Tertiary stratigraphy
Pre-Tertiary stratigraphy, ranging in age from
Palaeozoic to Upper Cretaceous, is exposed on the
flanks of the Salawati Basin in the northern Bird's
Head of Irian Jaya and in Misool island (Fig. 5).
468 T.R. CHARLTON
130°E
Batanta,
0 5O km
I I I I
1os
Kofiau Salawati
SALAWATi\
BASIN \ ,
~ -
Misool
Fig. 3. The Salawati Basin. Tertiary isopach contours are in km, interpolated from Imperial (feet) contouring by
Redmond & Koesoemadinata (1976).
In the northern Bird's Head, continental basement
comprises Siluro-Devonian shales and turbidites
(the Kemum Formation) which were deformed,
metamorphosed and intruded by granitic rocks (e.g.
the Melaiurna Granite) during the Late Devonian-
Early Carboniferous (Visser & Hermes 1962;
Pieters et al. 1983). Lower Carboniferous(?) syn-
orogenic sediments of the Aisajur Formation are
folded but not metamorphosed.
The mid Palaeozoic basement is succeeded
unconformably by a platform sequence com-
mencing with the Upper Carboniferous-Permian
Aifam Group. Further granitic intrusion into the
Kemum basement occurred during the Permo-
Triassic (the Anggi Granite Suite), contem-
poraneous with the deposition of the Aifam Group
and the succeeding Tipuma Formation, which is
a red bed sequence of Triassic-Lower Jurassic age.
In Irian Jaya generally the Tipuma Formation is
succeeded by the Kembelangan Group of upper
Lower Jurassic-Upper Cretaceous age. In the
northern Bird's Head the only element of the
Kembelangan Group recognized at outcrop is
the Cretaceous Jass Formation (Pigram & Sukanta
1982). This comprises calcareous and micaceous
mudstone with minor limestone and sandstone.
131°E
rian
. . . . . . *..
Plio-Quaternary
Miocene
~ ~ Mesozoic-ealaeogene
11--~Contni ent
I-~
basaelment
Tertiaryvolcanics on
continental crust
Non-continentalelements
A third phase of igneous intrusion also occurred
during the Late Cretaceous, with granitic core
samples from the base of the Salawati K-IX well
yielding Campanian-Maastrichtian K-Ar radio-
metric ages (71 _+ 1.5 Ma from biotite and
79.3_ 1.1 Ma from amphibole: Lunt & Djaafar
1991).
In Misool, metaturbidites apparently similar to
the Kemum Formation of the Bird's Head form
the oldest recognized stratigraphic unit (the Ligu
Metamorphics: Pigram et al. 1982). Their age,
however, is uncertain, and Siluro-Devonian
(Pigram et al. 1982), Permian (Froidevaux 1974)
and Triassic (Simbolon et al. 1984) ages have been
suggested for the sedimentary protolith. The oldest
unmetamorphosed sequences in Misool are Triassic
turbidites of the Keskain Formation, and partly
contemporaneous Upper Triassic reefal limestones
of the Bogal Limestone Formation. The Bogal
Limestone is overlain unconformably by Lower-
Middle Jurassic shales (Yefbie Shale), and these
are in turn succeeded by Upper Jurassic shelf
carbonates of the Ligu Formation and contem-
poraneous(?) shales of the Lelinta Shale Formation.
These pass up conformably into Cretaceous bathyal
limestones of the Facet Limestone Group (Gamta
 Change of direction

SW NE'W
Misool Salawati island Sele Bird's Head >
Straits Klamogun Klamumuk
(TBF-lX) TBO-lX K-lX J e f Lio-1 I Klamonol Klawilis
R~
. . . . . . . . . . . . . . . . . . . . ,'7. , ,"7 ,'7",72. :'7,,~, ~ :'7,,'7",",,, ,'~, 7":, :'7,,,, ,~ i " ~..:.::::i:iiii::!iiii~!~:i:::::
.....
©
2 Une "" I q""~-~..-.-... - . - - - .................... -...- -.- ...... -.- - iiiiiiiiiii:;......... ........ ~:~.......... • " ""
©
3 volcanics I :-":
4 km / .......... "- -' ' : ~ , , .........................................
" " n ::::::P:P:P:]:::::::::::::::::::::::::::::"
~°~'.,.~" r-.'~ ~ ° ~ ' " ;>
E/3

u , ~ , ~ =u Km in Salawati-N-1X i ~'.~ ,~,~ :Z

Vertical exaggeration x5 ~ ©
i

lililiiiiiiiiiiiii!]Pa,aeogen e [ ~ ~ Miocene she,f ~ Miocene deepwater ! I Miocenebasina' ~P,io-

carbonate with limestone facies facies t - . - . = .-] Quaternary
pinnacle reef (Klamogun Fm.) 0
i Z
Fig. 4. Cross-section through the Salawati Basin showing Tertiary geology and pre-Tertiary subcrop (eastern part after Vincelette & Soepardjadi 1976). The Miocene Salawati
Basin had a depocentre beneath the present-day Sele Straits, distinct from the main Plio-Quatemary depocentre between Salawati and Misool islands.

4~
470 T.R. CHARLTON

Salawati Basin Tomori Basin

SYSTEM/
SERIES North Bird's Head Misool Banggai-Sula East Sulawesi
Pliocene ' • ' , - " - S e e Fiqure 6 ' ' - ' - I Atkari Formation .'•~" ". S u l a w e s i ' G r o u p .••
I I I
>_ Miocene
I Neff~-in'ea I Ooe~taEat. I ~ ~ ~ ' • . • . ' ' . .
n" 'l "r I , h-..--~ .-.. ~ S a l o d i k L / 1 ~ ~
~
F-- •
Oligocene
. - L i m e s t o n e G r o u p - I - ~ ~F~rmatiorl ; ~ ; ; ~ S a l o d i k G r o u - p Z
rr
- - r - e t c . (see Figure 6 ) - - - r - , ,
~__ Eocene / I . I / I~aag Limestone, ~

Palaeocene ~ ' b ~ ~ ~ ~ i i i i i i"'~="~=~'=~'~'~

[~~'-'---'~--~.-~-.-'-:]-I- -r'Tanamu F°rmati°n-I- "-""-~t-'-r""-'-r"'=~"
I--:-Fafanlap Formation---
Upper =~'_:-~ F_or.m_at.io_n:'~. ~...,- I . . .i. I I-i-I- -I- I I ' I ' I J " , d& " l' -I~ormation~.~l~l
& -' r ' Lk , &u~ =O k -t"
o •-. --. - - - - :..- - - . ~,1 I
O •
W
O
3 . . . . . . . . . M t__r0Lno=:
<
I- .5
Lower ~ --~
o
o
E Ill I I I ~
,--.-:-.-.- -:---
. i i i i i . . . . .

Upper ~, ,13emu F m , . . ~ " ~ e ~ n ~ S-h~-e: ---.=Buya F o r m a t i o n ' . . " - -Tetambahu/Nambo'Fms,-

' .... .-- I I I I
..r" . . . . . . ~ . . . . " .............. I ........ '
Middle - --Yefbie S h a l e - ' - - - - - - ~ - - - - -%-- /Nanaka Formation - -
. . . . . . . . . . . ~ ~ ...... ! . . . . . . . . . . . .

• • . " . . . . . . . . . . . .':Bobor~g Formatior~.;

Lower .'I'~R:~'I'~':~'I'~':~R"~'-~" ': ".'~,'-'R~" " "

o : "•- T.p. .u.m a F o. .r .m• .a t i o. .n. . . . ::' - ---•-=-'-: ~,? -I+.B.uo ~ T o k ~ a l"a q"~ "
.. ___.
Upper I."• '6" "-' '. ' " ".'• "" "-':. ~"
- - , ..•LLime,
s I o. .n. .e.-. -. :. -. .' -. -. .-.-.-.~. .- : - ' : . .~ N o f a n ~ Volcanics
< .r' ", .'"...'... ,~
Middle
• " ' " ' " " "fl" " " ' "" 4- "J"~"" K e ~ k ~ i f f Formation. +
Lower "', , • "= , • "- • " • • '
i

Z Upper - _A!nlm
• : - - E o r m a ~• ~ " 4- (Relationship with Ligu (Age of m e t a m o r p h i c 4- ( B a s e m e n t not seen)
,~ • 'e~ . . . . . . . . . M e t a m o r p h i c s uncertain) b a s e m e n t protolith
"'I --
+ unknown)
n-" ~ - - A i f a t Formation- 4-

i~ Lower L9 . . . . . . . . . ---
n E. . . . . . . . . .
Shelf limestone R Red beds
" "" i Coal
0 Upper .<_"-_'-Aimau Formation--.:_.
D e e p w a t e r limestone -• Chert
iiUJ - --='-R
. ."' "--='-:-R:-:
. . . "-" '-" R: -- - . Shale Marl
T~ o o Conglomerate
O .~.._?~_~1Aisajur 4- Siltstone V Acid volcanics
< Lower :~.~;2;~;,~,. Formation 4-
O Sandstone/shale r~ Periods of acid
igneous intrusion
SlLURO ............................. Sandstone
-•_ K e m u m F o r m a t i o n : .
DEVONIAN

Fig. 5. Simplified stratigraphy of basement terranes beneath the Salawati and Tomori Basins• Primary data sources
for the Northern Bird's Head: Visser & Hermes (1962), Pieters et al. (1983) and Pigram & Sukanta (1989); Misool:
Pigram et al. (1982) and Simbolon et al. (1984); Banggai-Sula: Garrard et al. (1988), Surono & Sukarna (1993) and
Supandjono & Haryono (1993); East Sulawesi: Simandjuntak (1990), Corn6e et al. (1994).

Limestone and Waaf Formation). The uppermost A number of petroleum exploration wells
Cretaceous is represented by shales passing located between the Bird's Head and Misool have
upwards into shelf limestones (the Fafanlap encountered the pre-basinal succession beneath the
Formation), which are succeeded conformably Salawati Basin. Broadly comparable sequences to
by Lower Palaeogene sandstones (the Daram those found in the Bird's Head and Misool were
Sandstone Formation). A shelf limestone sequence encountered, but significant differences include
equivalent to the New Guinea Limestone Group the presence of Jurassic section (post-Toarcian) in
of the Bird's Head follows conformably; this equivalents of the Kembelangan Group, and the
will be described in more detail in the following occurrence of volcanic breccias, volcaniclastic
section. sediments and tuffs in the Upper Cretaceous.
 SALAWATI & TOMORI BASIN CORRELATION 471

comprises a thin sequence of basal clastics

Tertiary stratigraphy
succeeded by shelf carbonates. The overlying
The older Tertiary succession of the Bird's Head- Oligocene Sirga Formation is predominantly a
Misool region, which overlies Mesozoic and older clastic sequence consisting of sandstone and shale
rocks unconformably, is primarily developed in with minor limestone. The Sirga Formation was
limestone facies. In the Bird's Head this sequence deposited under neritic to epineritic conditions, and
forms the New Guinea Limestone Group (Visser & is up to about 200 m thick.
Hermes 1962; Pieters et al. 1983). The Tertiary of The oldest part of the truly basinal succession
this region is best known from the NW Bird's Head comprises a series of contemporaneously
and Salawati island, where it has been intensively developed Lower-Middle Miocene carbonate
studied during hydocarbon exploration. However, lithofacies, corresponding to a SE-NW transition
the Tertiary section of Misool also forms part of from shelf to deep-water environments (Figs 5
the marginal Salawati Basin succession. & 6). The shelf carbonates are assigned to the Kais
The Tertiary succession in the western Bird's Formation, and deep-water time equivalents form
Head and Salawati island (Fig. 6) commences the Klamogun Formation. The latter comprises
with the Faumai Formation, dated as upper Lower- interbedded marly and flaggy limestones with
Upper Eocene (Gibson-Robinson et al. 1990), abundant planktonic foraminifera. The Kais
although lithological equivalents extend up into the Formation persisted into the Upper Miocene, but
Oligocene (Fig. 6; Pieters et al. 1983). The Faumai the Klamogun Formation was progressively
Formation, which is up to about 250 m thick, replaced by a shallowing-upward sequence of

~,1~ TOMORI BASIN SALAWATI BASIN

~ ~0 I ~ ) NORTH /CENTRAL I SOUTH NW SE
~uaternal~ K~lomba Fro. o ' L m _ ,LuwukFormafiort , m~_O_..~ o Sele Con Iomerate o o o
,2 ~ "Jppe, giaF-m.o 1 ,. . . . . . . . . . . . . . . . . .-'..~.-.~'.f .... - ..... i ....
•4 ~ .o-~," " " ^" " " " • - ._. ... i:-.--.-.Klasaman Formation:.-.-.I
= . w, - . . . . . . . . . . . ~ ~.Kintorn t-ormadonl-- . . . . . . . T . . . . . . . . .'11
a. " "" "" " - - ' @ K o l 0 Ubr. ' , ' I ' t'''" .... "'" ..... "J;; - - - ~ - - - ' - - " ~ ' ' ' " .... "''"
" ~ ]~U~' I I [ I I I I ~[.Klas.a.fet. :q-LF~.~__~ I , x ' L ~ ."" " - - l
•8 ~- Mantawa Mbr. I ' I ' , ' . Formation/ ~ ~ r - ~
•10 vlnahak F o r m a t on ~ .t, r •.. ._ • .
, , " " " ~ r ~ a s t-ormat on.
•~2 ~ l t l I l -- • J4~i~L-- J J l
'14 ~ "O -° ~ ~ .'7 ~ .... I,m" . . . . . . . . . I
® ~ ~i-.Matindok Formation ~ . ~
• 16 ._~ ~ ~ ~
, • • . , q ~,40 ~ ' , r,K l a m o ~ u n , , : : ~ ~" I I
-Upper Tomon -- I r ', ,,-ormation, I I I
•18
2C a~ i Formation-'-- _rFormation i I " I I
i i I ',
-3° I Midd!eTom.ori I Lr(Undiff.),_ ,', : , : , ,',,',,',,~1~ I I I I
L24
~ .......... _ ..............

.... f . . . . . . ;;:i i:: '1

j_ ,
34' ~o L wer--rl : :
38, - TTomqr : :;
~! ~ .For~na.tiorq ~1

~11~14244462, .~
"~ ............ :

52
54 ] I :
Ig n e o u s r o c k s - IM i c {2241~9Ma}
~ts(~fiis{ .......................... r- . . . . . . . . . . . . . . . . . . . . - ~ -
Pre-Tertiary (Silurian-U. Cretaceous)

Fig. 6. Stratigraphic comparison of the Salawati and Tomori Basins (Salawati Basin after Gibson-Robinson et al.
1990; Tomori Basin after Davies 1990, Handiwiria 1990 and Abimanyu 1990). The simpler, more laterally continuous
appearance of the Tomori Basin is probably an artefact of less intense exploration in that basin.
472
marl and siltstone (the Klasafet Formation). On
the basinal flanks the shelfal Kais Formation is
typically 350 m thick, whilst the basinal Klamogun
Formation is 1159 m thick and the Klasafet
Formation about 1925 m thick in their respective
type sections (Visser & Hermes 1962). Pinnacle
reefs developed locally near the shelf-slope break
during the Middle-Upper Miocene, and these
extend the typical thickness of the Kais Formation
by up to 533 m (Gibson-Robinson et al. 1990).
The second, clastic, basinal sequence is repre-
sented primarily by the Pliocene Klasaman
Formation. This consists of interbedded sandy
calcareous mudstone and argillaceous sandstone
with minor conglomerate and lignite. This
sequence has a maximum thickness of about
4500 m (Visser & Hermes 1962; Pieters et aI.
1983). It is succeeded unconformably by the Sele
Conglomerate of Quaternary age.
In Misool (Fig. 5) the Upper Palaeocene-
Oligocene(?) Zaag Limestone, which is approxi-
mately the equivalent of the Faumai Formation in
the Bird's Head, succeeds the Daram Sandstone
......
Syn- to post-
orogenic sediment
East Sulawesi ~ Continental
Ophiolite
Fig. 7. The Tomori Basin, east Sulawesi. Structure of Tolo Gulf after Davies (1990). Basin isopachs in km adapted
from Hamilton (1979) taking into account later drilling data. Onshore geology simplified from GRDC mapping.
T.R. CHARLTON
conformably (Pigram et al. 1982; Simbolon et al.
1984). No equivalent of the Oligocene Sirga
Formation is recognized, this time period being
represented by an unconformity possibly associated
with minor folding. As in the Bird's Head, the
Lower Miocene is represented by contemporaneous
shelf carbonates (the Openta Limestone) and a
somewhat deeper water sequence (the Kasim Marl)
which consists of well bedded marly and sandy
limestones containing planktonic foraminifera. The
Openta Limestone extends up into the Middle
Miocene, but the Upper Miocene is absent. The
Pliocene-Quaternary Atkari Formation overlies the
Openta Limestone unconformably. In contrast to
time equivalents in the Bird's Head, the Atkari
Formation is predominantly a shelf carbonate
sequence.
The Tomori Basin
The Tomori Basin is located off the east coast of
Sulawesi island (Fig. 7). Unlike the intensively
explored Salawati Basin, the Tomori Basin is a
an0 a .n0:
Continental
cover ,.." Subsurface limit of
oO" East Sulawesi
basement °° Ophiolite
 SALAWATI t~ TOMORI BASIN CORRELATION 473

region of frontier hydrocarbon exploration with
only ten wells drilled to date. These are all located
along the western margin of the basin, either on
the coastal plains of east Sulawesi or immediately
offshore.
Structural setting
Figure 7 shows a Tertiary isopach map for the
Tomori Basin based on the contouring of Hamilton
(1979), but modified in the light of the more recent
well data. The basin has a triangular or T-shaped
map plan, probably reflecting two stages in the
basinal evolution. In the north, the N E - S W
elongation of the basin corresponds closely to the
thrust front of the east Sulawesi fold-and-thrust
belt, and it has been suggested (e.g. Davies 1990;
Handiwiria 1990) that this portion of the Tomori
Basin originated as the foreland basin to this
orogenic belt. The NW-SE prolongation of the
basin does not have such an obvious origin. It will
be suggested later that this trend marks the former
connection between the Tomori and Salawati
Basins.
Figure 8 shows two cross-sections through
the western margin of the Tomori Basin; one in the
north through the Minahaki well, the other in the
south through the Tiaka field (Fig. 7). The northern
and southern halves of the basin have very different
structural styles, and also significantly different
timing of deformation. The southern part of the
basin is characterized by thrust faults, whilst the
northern part is characterized by normal and
wrench faulting (Davies 1990). There is also a
marked change in the strike of thrust structures,
from SW-NE in the east Sulawesi orogenic belt
west of the northern basin, to nearly N-S in the
south, offshore in the Tolo Gulf (Fig. 7). In the
southern foldbelt, section as young as the Lower
Pliocene Kintom Formation (described later) is
affected by thrusting (Fig. 8b), whilst in the north
there is no clear evidence for active thrusting
during deposition of this unit (Fig. 8a). In the north
the dip of sediments in the Kintom and Bia
Formations away from the orogenic front suggests
post-orogenic molasse-type deposition. The only
faulting recognized on the northern cross-section
is normal faulting downthrowing west toward the

NW o, , , , , 5km
, Minahaki-1
.O. SE a!
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0

' ~ --2.
+ ..I- + + + + ...i- ..i- + -I- + + ..1..
, i i , , , i , , , , • -

N~^I 0 5kin Tiaka-3 Tiaka-1 ~,,,.._~. ~ ¢.,t- 0

"' . _.t_., ........... o._ b

I ~:~Kaiomba~n...
" : ~ . . . . . .~

I 0

Fig. 8. Sketch cross-sections through the westem margin of the Tomori Basin based on seismic sections figured by
Davies (1990), Abimanyu (1990) and Handiwiria (1990). (a) Section through the Minahaki-1 well in the north of the
basin. (b) Through the Tiaka field in the south of the basin. Note different horizontal and vertical scales. Horizontal
scales were not included in the original figures and are therefore approximate. See Fig. 7 for locations.
474 T.R. CHARLTON
mountain front. This only affects rocks as young as
the Miocene (and older) Salodik Group, and could
be interpreted as downfaulting in the foreland
associated with development of the east Sulawesi
orogenic belt. This is consistent with the usual
dating of orogenesis in this region as approxi-
mately Middle Miocene (e.g. Davies 1990).
However, the younger, Pliocene, deformation seen
in the south does not seem to have significantly
affected the northern region. The timing and
significance of this deformation history will be
discussed in more detail later.
As in the Salawati Basin, most of the variation in
total stratigraphic thickness of the Tomori Basin is
accounted for by the Plio-Quaternary, predomi-
nantly clastic, sequences. Thickness variations in
the Miocene succession are less marked, but there
is a general southward thickening from about
600-650 m in the northern wells to more than
1000 m in the south (where the foldbelt deforma-
tion makes precise estimation of true stratigraphic
thickness from well sections more difficult). Pre-
Miocene sedimentary section is absent from the
north of the basin, and is only 60-90 m thick in
the central basin. The pre-Miocene has not been
penetrated by drilling in the south of the basin,
although it is recognized in the subsurface from
seismic data (R.A. Garrard, pers. comm.). Thus,
as in the Salawati Basin, basinal subsidence
commenced at the beginning of the Lower
Miocene, at least in the northern and central
portions of the Tomori Basin.
Pre-basinal stratigraphy
Pre-basinal sequences forming the 'basement' of
the Tomori Basin are exposed on the Banggai and
Sula islands to the east, and are imbricated in the
east Sulawesi fold-and-thrust belt to the west. In
Banggai and Sula (Fig. 5) the oldest rocks recog-
nized include unnamed mica schist and gneiss,
lower grade metasediments (the Menanga
Formation) and granitic intrusives (the Banggai
Granites). The age of the sedimentary protolith is
unknown, but schists on Peleng island have yielded
Carboniferous radiometric ages (305 -- 6 Ma:
Sukamto 1975), interpreted as the age of meta-
morphism. The granites have yielded Late
Permian-Triassic radiometric ages (Pigram et al.
1985), and are probably cogenetic with acid
volcanics (the Mangole Volcanics) of probable
Triassic age (radiometric ages of 210 __ 25 Ma
and 330 _ 90 Ma: Sukamto 1975).
The oldest unmetamorphosed sedimentary rocks
in the Banggai and Sula islands are probably reefal
limestones of the Nofanini Formation. These were
interpreted by Garrard et al. (1988) as Upper
Triassic, although they have not been dated
accurately (Garrard, pers. comm.). The present
author has also examined outcrops of the Nofanini
Limestone, and they appear directly comparable to
reef limestones also provisionally dated as Upper
Triassic from the Tanimbar islands.
The Nofanini Formation is succeeded by or is
partly contemporaneous with red beds of the
Bobong Formation. This is dated as Lower-Middle
Jurassic by Surono & Sukarna (1993) and
Supandjono & Haryono (1993), but it may range as
old as Upper Triassic (Garrard et al. 1988). Marine
transgression occurred during the Toarcian (e.g.
Sato et al. 1978), and subsequently during the rest
of the Jurassic and the early part of the Cretaceous
(at least to the Hauterivian) a sequence of claystone
and marl with minor limestone accumulated (the
Buya Formation). Above this, apparently after a
mid Cretaceous hiatus, are deep-water chalky lime-
stones of the Tanamu Formation. This ranges in
age from at least the Santonian until the Upper
Palaeocene (Garrard et al. 1988). Succeeding this
in the Banggai islands are Eocene-Middle Miocene
shelf carbonates of the Salodik Formation. Further
east, in the Sula islands, the Palaeogene is absent
according to the mapping of Surono & Sukarna
(1993), although Garrard (pers. comm.) has
obtained Palaeogene ages from samples collected
in Mangole island. Shelf limestones equivalent to
the upper part of the Salodik Formation were
deposited in this area during the Lower-Middle
Miocene.
Dolerite dykes have been recorded locally
intruding the Banggai Granite and the Jurassic
Bobong Formation. Their age has not as yet been
better constrained than post-dating deposition of
the Bobong Formation (Surono & Sukarna 1993;
Supandjono & Haryono 1993).
In the east Sulawesi fold-and-thrust belt, two
distinct stratigraphic sequences are imbricated
together: a parautochthonous sequence of con-
tinental origin with affinities to the Banggai-Sula
sequence described above; and an allochthonous
sequence related to the obducted east Sulawesi
ophiolite. Only the parautochthonous sequence is
described here (Fig. 5) as the allochthonous
sequence has only limited relevance to the Tomori
Basin.
In the parautochthonous sequence basement has
not been recognized. The oldest sedimentary rocks
are of Triassic age, and comprise contemporaneous
hemipelagic carbonates (the Tokala Formation) and
turbiditic sandstone-shale sequences (the Bunta
Formation) (Katili 1978; Rusmana et al. 1986).
Recently Upper Triassic reefal facies have also
been identified (Corn6e et al. 1994). These are
succeeded by Lower Jurassic conglomeratic sand-
stones with shale and coal (the Nanaka Formation)
and Middle-Upper Jurassic limestones and marls
 SALAWATI• TOMORI BASIN CORRELATION
(the Nambo and Tetambahu formations) (Surono
et aI. 1987; Simandjuntak 1990).
In the Cretaceous two formations (the Luok and
Matano Formations) have been recognized through
regional geological mapping. Both consist of
relatively deep-water carbonate-shale-chert
sequences, and the two have been interpreted as
more proximal and more distal facies within a low
energy sedimentary environment (T. Simandjuntak,
pers. comm.). The Luok Formation is dated as
Turonian-Maastrichtian in age (Rusmana et al.
1984). It is succeeded unconformably by shelf
carbonates of Lower Eocene-Lower Miocene age
which, as in the Banggai islands, are assigned to
the Salodik Formation (Simandjuntak 1990). In
the Eocene and Oligocene, deeper water marl with
limestone intercalations has been mapped as the
Poh Formation (Rusmana et al. 1984; Surono et al.
1987). The Salodik and Poh formations have been
placed within a re-defined Salodik Group by
Handiwiria (1990). The Salodik Group is
succeeded unconformably by molassic sediments
of the Sulawesi Group (Abimanyu 1990). The
Salodik and Sulawesi groups as recognized in the
offshore Tomori Basin will be described in more
detail in the following section.
Pre-basinal basement has been penetrated in
several exploration wells drilled through the
Tomori Basin (Davies 1990; Handiwiria 1990).
Three wells in the north of the basin (Matindok-1,
Minahaki-1, Mantawa-1) encountered intrusive
igneous rocks, whilst the Tiaka-2 well in the central
portion of the basin penetrated quartz-mica schist,
dated radiometrically as Middle-Late Triassic
(224 _+9 Ma: Handiwiria 1990). Wells in the south
of the basin did not reach basement except for
Dongkala-1 which penetrated an ophiolitic
sequence. This is believed to be part of the over-
thrust and allochthonous east Sulawesi ophiolite
rather than pre-basinal basement related to the
Banggai-Sula microcontinent (Davies 1990). No
unmetamorphosed pre-Tertiary sedimentary rocks
have yet been encountered beneath the Tomori
Basin.
Basinal stratigraphy
The oldest sedimentary rocks recognized through
exploration drilling in the Tomori Basin are
assigned to the lower member of the Tomori
Formation (Fig. 6). This consists of a thin basal
clastic sequence succeeded by shelf carbonates,
dated as Upper Eocene-Lower Oligocene in age
(Handiwiria 1990). The unconformably succeeding
middle and upper members of the Tomori
Formation are essentially Lower Miocene in age,
although they include Upper Oligocene sediments
in one well (Minahaki-1). In the north and centre
475
of the basin this sequence comprises shelfal
argillaceous limestones passing up into dolomitized
limestones with coal interbeds. In the south the
undifferentiated Tomori Formation comprises
deeper water carbonates lacking coal interbeds.
The Middle Miocene Matindok Formation,
comprising claystone and sandstone with minor
limestone and coal, overlies the Tomori Formation
conformably (Handiwiria 1990). This is in turn
succeeded conformably by the essentially Upper
Miocene Minahaki Formation which marks a return
to shelf carbonate deposition. Pinnacle reefs occur
in the north of the basin (the Mantawa Member)
and claystone interbeds in the south, again
suggesting deepening of the palaeoenvironment
to the south.
The Pliocene-Recent succession is composed
predominantly of clastics assigned to the Sulawesi
Group (Abimanyu 1990). The Lower Pliocene
Kintom Formation, which apparently succeeds the
Minahaki Formation without significant hiatus,
consists of sandstone and claystone grading up into
conglomerate and sandstone. This is succeeded
conformably by the Upper Pliocene Bia (or Biak)
Formation, composed of conglomerate and sand-
stone. The Quaternary Kalomba Formation, which
is also composed largely of conglomerate, follows
unconformably. This developed contempor-
aneously with the Luwuk Formation which is a
reef limestone sequence.
Correlation of the Tomori and
Salawati Basins
The following points of correlation can be recog-
nized between the Salawati and Tomori Basins:
• Both basins are underlain by continental base-
ment consisting of Palaeozoic(?) metasediments
intruded by Carboniferous-Triassic granites.
• Comparable unmetamorphosed Mesozoic sedi-
mentary sequences subcrop the basins or outcrop
nearby. Common stratigraphic features include
the occurrence of reef limestones of probable
Upper Triassic age in east Sulawesi, Sula and
Misool; Upper Triassic turbidites in east
Sulawesi and Misool; Triassic-Lower Jurassic
red beds in Banggai-Sula and the Bird's Head;
Toarcian transgression followed by Middle-
Upper Jurassic shales in Misool and Banggai-
Sula; Upper Jurassic limestones in east Sulawesi
and Misool; mid Cretaceous hiatuses in Banggai-
Sula, the Bird's Head and possibly east Sulawesi;
and deep-water Cretaceous limestones in Misool,
Banggai-Sula and east Sulawesi. In addition, it is
possible that the Late Cretaceous granite of the
Bird's Head and volcanic/volcaniclasticrocks in
476 T . R . CHARLTON
Misool and offshore to the east might correlate
with the as yet undated but post-Middle Jurassic
dolerites of Banggai-Sula.
• Tertiary sedimentation commenced in the Tomori
Basin and the eastern Salawati Basin (Salawati
island and the western Bird's Head) during the
Eocene with thin transgressive elastics succeeded
by shelf carbonates. In both basins, however, the
base of the Tertiary is strongly diachronous,
ranging from upper Lower Eocene-Lower
Miocene in the eastern Salawati Basin, and at
least Upper Eocene-Lower Miocene in the
Tomori Basin (Fig. 6). In Misool the basal
elastics (the Daram Sandstone) are thicker
(about 50 m: Pigram et al. 1982) and as old as
Palaeocene. The succeeding shelf carbonates
(Zaag Limestone) may also be locally as old as
Palaeocene. Older Tertiary section may underlie
the southern part of the Tomori Basin, but this
has not yet been penetrated by exploration
drilling.
• A break in predominant Tertiary carbonate sedi-
mentation occurred in both basins in the mid
Oligocene. In the eastern Salawati Basin this was
marked by deposition of the elastic Sirga
Formation, whilst to the west in Misool there is
an unconformity possibly associated with minor
folding. In the Tomori Basin a mid Oligocene
unconformity separates the lower and middle
members of the Tomori Formation.
• Prior to the mid Oligocene break in carbonate
sedimentation both the Salawati and Tomori
regions were accumulating relatively thin shelf
sequences: the Faumai Formation is about 250 m
thick in the eastern Salawati Basin (Visser &
Hermes 1962; Pieters et al. 1983), and the lower
member of the Tomori Formation is up to 100 m
thick (Handiwiria 1990). Truly basinal succes-
sions did not begin to accumulate in both areas
until the Early Miocene. In the eastern Salawati
Basin the Lower-Middle Miocene consists of
contemporaneous shelf (Kais Formation) and
basinal (Klamogun Formation) limestones, with
basinal deepening to the NW, towards the Sorong
fault. In the Tomori Basin, shelf carbonates are
found in the north, with more basinal limestones
in the south, towards the South Sula fault. The
Miocene section in the northwestern part of the
eastern Salawati Basin is more than 2000 m thick
(e.g. Vincelette & Soepardjadi 1976), and is more
than 1000 m thick in the southern Tomori Basin
(Handiwiria 1990).
• In the Middle Miocene fine-grained siliciclastics
began to accumulate in both basins. In the eastern
Salawati Basin the Middle-Upper Miocene
partly clastic Klasafet Formation progressively
replaces the deep-water carbonate Klamogun
Formation. The proportion of siltstone in the
Klasafet Formation decreases southward away
from the Sorong fault (Visser & Hermes 1962;
Pieters et al. 1983), suggesting a elastic sedi-
mentary source to the north or west. In the
Tomori Basin the Middle Miocene Matindok
Formation is recognized in all parts of the basin
so far drilled, but it thickens considerably and the
palaeoenvironment deepens to the south towards
the South Sula fault.
• During the Late Miocene there was a return to
shelf carbonate sedimentation in the Tomori
Basin. On the southern flank of the eastern
Salawati Basin carbonate shelf sedimentation had
continued without significant interruption from
the Early-Middle Miocene. Isolated patch and
pinnacle reefs are recorded in both basins at this
time.
• Carbonate sedimentation and growth of the
economically important patch and pinnacle reefs
became area!ly restricted in both the Tomori and
Salawati basins near the Miocene-Pliocene
boundary. The Pliocene Klasaman Formation of
the eastern Salawati Basin corresponds approxi-
mately with the Kintom and Bia formations of
the Tomori Basin. Both sequences show an
overall coarsening-upward, from predominant
claystone at the base to more sandy and con-
glomeratic facies at the top. Shelf carbonate
facies persisted into the Plio-Quaternary of
Misool (the Atkari Formation) and the Luwuk
Formation of the southern Tomori Basin.
• A strongly erosional base Pleistocene uncon-
formity is recognized in both basins. A succeed-
ing conglomeratic unit is also recognised in both
areas: the Sele Formation in the Salawati Basin,
and the Kalomba Formation in the Tomori Basin.
• In addition to stratigraphic similarities, the
Salawati and Tomori Basins share common
hydrocarbon kerogen characteristics (SPT/
Pertamina 1992). In both cases an anoxic,
restricted marine carbonate or calcareous shale
sequence is the suspected source rock (Robinson
1987). In the Salawati Basin this is either the
Lower Miocene Klamogun Formation (Robinson
1987) or the Upper Miocene Klasafet Formation
(Livingstone 1992). In the Tomori Basin the
source sequence is probably the Lower Miocene
middle member of the Tomori Formation (Davies
! 990; Handiwiria 1990).
Regional tectonic implications
Figure 9 shows a possible reconstruction of the
eastern Sulawesi-western New Guinea (Bird's
Head) region prior to development of the Sorong
fault zone. The primary control on this recon-
struction is the alignment of the Salawati and
 SALAWATI (~; TOMOR! BASIN CORRELATION 477

• / Banggai
r Islands
.L •
, / ,- .
~ Sula islands _

"-:~
,soo,
WATt,
=2:::::i:i~i:~ii::iii~BASlNili!i~i: ]
:!!i!i!i!iii!i!iii!i!i!i!i!:
"-- 3 ~ Tertiaryisopach (km) ~ Australiancover ..........
I ~ AIIochthonoUSbasement I

i
~ Basin >2km sediment ~ Australian basement

Fig. 9. Reconstruction of the Bird's Head-Sula-Banggai-East Sulawesi region prior to development of the Sorong
I
fault. The Bird's Heads block is drawn in its present-day orientation with respect to north. In order to keep SW Obi as
an identifiable body, the inferred wrench faults bounding this terrane have not beeen restored.

Tomori basinal depocentres. However, as the
Salawati Basin has a double depocentre, there is
some ambiguity as to which depocentres should
be aligned• Both alignments give a reasonable
correlation of the basin isopachs, but placing the
Tomori Basin north of the eastern (Sele Straits)
depocentre of the Salawati Basin (Fig. 9) gives
a more plausible basemap for palaeogeographic
reconstructions (such as that for the Miocene
described below - see Fig. 10). The net offset on
the Soron~ fault zone based on this reconstruction
is about 900 km.
The Salawati and Tomori Basins show strong
stratigraphic similarities up until fairly recent
times, and this probably indicates that offset of the
two half-basins occurred relatively recently. The
restored composite basin in Fig. 9 is roughly
elliptical in shape, with an oblique orientation
suggestive of a strain ellipse resulting from left-
lateral shear parallel to the Sorong fault zone. This
probably suggests that the basin was formed
largely as a result of transtension on the Sorong
system. As most of the Tertiary stratigraphic thick-
ness that gives rise to the elliptical shape is
accounted for by the Plio-Quaternary clastic
sequences, and additionally as the Salawati Basin
in particular shows little evidence of fault-related
deformation prior to the Early Pliocene, it is likely
that movement on the Sorong fault zone occurred
primarily during the Pliocene and Quaternary
periods. The precise age of the onset of clastic
sedimentation in the Salawati Basin has not been
stated in the published literature, but a stratigraphic
column illustrated by Livingstone (1992) indicates
the earliest deposition of the clastic Klasafet
Formation in the Upper Miocene. Livingstone
(1992) also showed an idealized burial plot for the
Waipili-1 well in the northern part of Salawati
island which indicates that the top of the Kais
platform sequence began to subside rapidly
between 6-7 Ma ago. This might indicate the
time at which the Sorong fault system began to
develop.
A further constraint on the age of the Sorong
fault zone comes from regional plate tectonics. It
was suggested earlier that the orientation of the
fault system parallel to the relative convergence
vector between the Pacific and Indo-Australian
plates indicates that the present-day Sorong system
acts as a transcurrent boundary between these two
plates (Fig. 1). The motion of Australia relative to
the Pacific in this region is at a rate of 128 km Ma -1
478 x.R. CHARLTON
on a bearing of 068 ° (using the Euler poles of
Minster & Jordan 1978). At this rate the 900 km
displacement would be achieved in a period of
7 Ma. Assuming that plate motion has been
constant through this period, this provides a
minimum age for the Sorong system. This estimate
is very similar to the commencement of subsidence
in the Salawati Basin indicated by Livingstone
(1992).
The absence of significant pre-Pliocene faulting,
together with the largely non-clastic nature of the
Miocene basin fill, suggests that the Miocene
carbonate basin pre-dated (and is therefore
genetically unrelated to) the development of the
Sorong fault system. Figure 10 shows a possible
interpretation of the Miocene basinal palaeo-
geography based on extrapolating the well estab-
lished palaeoenvironments of the Salawati Basin
into the Tomori Basin, and taking into account the
limited geological data from Banggai-Sula and east
Sulawesi. It would seem reasonable from Fig. 10
that the combined Salawati-Tomori Basin origin-
ated at least in part as the foreland basin to the east
Sulawesi orogenic belt. If this interpretation is
correct, then the clear evidence from the Salawati
Basin for strong basinal subsidence in the Lower
Miocene suggests that orogenesis was underway
in east Sulawesi by this time. This is somewhat
Fig. 10. Miocene palaeogeography of the eastern
Salawati Basin (after Gibson-Robinson et al. 1990)
extrapolated into the repositioned Tomori Basin.
earlier than the usually accepted Middle Miocene
age (e.g. Davies 1990).
There is, however, a discrepancy between the age
of the east Sulawesi orogen as inferred from
onshore geology (which suggests important
orogenesis during the Middle Miocene or earlier)
and the offshore seismic and well data which
indicates fold-and-thrust belt development in the
Tolo Gulf during the Pliocene (Davies 1990). As
already described, the strike of the Tolo Gulf fold-
and-thrust belt is markedly oblique to that of the
orogenic belt onshore in east Sulawesi. It appears
that there were two distinct phases of development
in the east Sulawesi orogen: an earlier pre-Middle
Miocene phase which gave rise to a foreland basin
on its eastern flank (the Miocene Salawati-Tomori
Basin), and a later phase of fold-and-thrust belt
development during the Pliocene. The earlier phase
of deformation arose from arc-continent collision
(Davies 1990), whilst the Pliocene deformation is
probably related to the westward translation of
eastern Sulawesi in the Sorong fault system. The
Southeast Arm of Sulawesi shows clear structural
and stratigraphic similarities with the East Arm,
and probably formed a direct continuation of the
the east Sulawesi orogenic belt prior to the
Pliocene. The Tolo Gulf foldbelt probably
developed contemporaneously with oroclinal
bending of eastern Sulawesi, with the Southeast
Arm rotating anticlockwise with respect to the
East Arm.
In addition to the Bird's Head and eastern
Sulawesi, three displaced terranes are provisionally
repositioned in Fig. 9. These terranes are relocated
to give the tightest geographic fit consistent with
the regional geology. Banggai and most of the Sula
islands are treated as a single terrane which is fixed
relative to eastem Sulawesi and the Tomori Basin.
The positioning of Banggai-Sula is thus largely
defined by the re-positioning of the Tomori Basin
north of the Salawati Basin. However, some
freedom remains as to body rotation, and in Fig. 9
Banggai-Sula (and eastern Sulawesi) have been
rotated about 15 ° anticlockwise relative to the
Bird's Head in order to achieve the tightest
geographic fit. If, as is likely, the Bird's Head-
Misool block has itself undergone body rotations
relative to Australia during the Neogene, the
inferred 15 ° rotation of Banggai-Sula would be
additional to that in the Bird's Head.
The second displaced terrane in Fig. 9 comprises
Sulabesi island, eastern Mangole and the south-
western part of Obi island. The boundaries of this
terrane may be a simplification: Mangole island is
cut obliquely by a number of ENE-WSW trending
steep faults which may be left-lateral faults (as
suggested in the restoration) or could be normal
faults. In Fig. 9 Sulabesi is restored by reversing
 SALAWATI • TOMORI BASIN CORRELATION 479

an assumed anticlockwise rotation of about 90 ° Implications for hydrocarbon exploration

relative to Banggai-Sula and also relative to eastern
The reconstructions in Figs 9 & 10 may also have
Mangole. Sulabesi and eastern Mangole are then
some implications for future oil exploration in
translated about 175 km eastward relative to
this region. In the Salawati Basin, nearly all oil
Banggai-Sula (cf. 180 km displacement without
production is from pinnacle reefs which developed
rotation estimated by Garrard et al. 1988, based
near the shelf-slope break around a semi-enclosed
on the similarity of structural style and the nature
Miocene basin. If this palaeogeography can be
of Jurassic sediments in Sulabesi and the Taliabu
extrapolated into the repositioned Tomori Basin as
Shelf immediately west of the Sula islands).
suggested in Fig. 10, then the belt of pinnacle reefs
The third displaced terrane in Fig. 9 is the
should follow the eastern margin of the basin to
Tamrau Terrane of the northern Bird's Head. Most
the south and west of the Banggai island group. In
authors (e.g. Pigram & Davies 1987) have
addition, it now seems widely agreed that the
interpreted very large displacements between the
primary source rocks for the Salawati oils are
Tamrau Terrane and the Bird's Head, in part based
Miocene deep marine and poorly oxygenated
on the apparent necessity of accommodating the
calcareous mudstones and/or marly limestones
large strike-slip offsets of the Sorong fault zone to
which were deposited either contemporaneously
the west on the single fault strand of the Sorong
with or immediately above the Kais pinnacle reefs.
fault. However, the geology of the Tamrau Terrane
It is probably more than coincidence that these
is not greatly different from that of the Bird's Head
restricted basinal sediments were deposited in the
(cf. Visser & Hermes 1962), and it is suggested that
semi-enclosed Miocene Salawati embayment (Fig.
the Tamrau Terrane can be repositioned by a rela-
10). The palaeogeographic map also indicates that
tively small eastward translation of about 60 km.
a similar embayment would be expected at the
This would place the Netoni Igneous Complex
northern end of the Tomori Basin, and such a
(Pieters et al. 1981) north of the Anggi Granites of
restricted basin might well be the source for oil and
the eastern Bird's Head, and re-aligns the eastem
gas discoveries along the western flank of the
end of the Tamrau Terrane with the eastern edge
Tomori Basin. However, the predominant structure
of the continental Bird's Head block. (It has been
of the northern Tomori Basin is that of a foreland
suggested by several authors that the Netoni Block
basin, and any oil generated in this region would be
is itself a distinct 'mini terrane' within the Sorong
more likely to migrate updip towards the foreland;
fault; however, the overlap of the Cretaceous Amiri
that is towards the postulated pinnacle reef trend
Sandstone across the supposed Tamrau-Netoni
on the eastern flank of the Tomori Basin.
terrane boundary (Hartono et al. 1989) does not
Unfortunately, most of the extrapolated pinnacle
support this distinction).
reef trend lies under water depths of 1000-
A consequence of this relatively limited dis-
2000 m, and is therefore unlikely to be a com-
placement on the Sorong fault through the Bird's
mercial prospect with present-day economics.
Head is that most of the movement on the Sorong
system must be taken up north of the Tamrau
Terrane. In Fig. 9 most of the inferred left-lateral This work was initiated as part of the London University
movement is taken up along the northern edge of study of the Sorong fault zone under the leadership of
the Tamrau Terrane such that Sulabesi island is Prof. Robert Hall. Fieldwork in this area was sponsored
by the Royal Society, NERC grant GR3/7149 and the
re-positioned immediately northwest of the Tamrau
London University Consortium for Geological Research
Terrane. This aligns N-S trending normal faults in in Southeast Asia. Thanks to Chris Gibson-Robinson and
the Sula islands with similar faulting on the eastern Peter Lunt (Petromer Trend Corp.) and Andy Livsey
flank of the Salawati Basin, and also juxtaposes the (Simon Petroleum Technology, Jakarta) for useful
Jurassic-Cretaceous shale sequences of the Sula discussions, and to Tony Barber (Royal Holloway
islands with comparable sequences in the Tamrau University of London) and Dick Garrard (ARCO) for
Terrane. useful comments in review.

References
AB1MANYU,R. 1990. The stratigraphy of the Sulawesi
Group in the Tomori PSC, East Arm of Sulawesi.
Paper presented at the 19th Annual Convention of
the Indonesian Association of Geologists.
COCKCROFT,P. J., GAMBER,D. A. & HERMAWAN,H. M.
1984. Fracture detection in the Salawati Basin of
Irian Jaya. Proceedings of the Indonesian Petroleum
Association, 13, 125-133.
CORNICE,J.-J., VILLENEUVE,M., MARTINI,R., ZANINETFI,
L., VACHARD, D., E~r AL. 1994. Une plate-forme
carbonatEe d'~ge rhEtien au centre-est de Sulawesi
(rEgion de Kolonodale, CEl~bes, IndonEsie).Comtes
Rendus Acad~mie des Sciences de Paris, 318(II),
809-814.
DAVIES,I. C. 1990. Geological and exploration review of
the Tomori PSC, eastern Indonesia. Proceedings
of the Indonesian Petroleum Association, 19,
41-67.
480 T.R. CHARLTON
Dow, D. B. & SUKAMTO,R. 1984. Western Irian Jaya: the
end-product of oblique plate convergence in the
late Tertiary. Tectonophysics, 106, 109-139.
FROIDEVAUX,C. M. 1974. Geology of Misool island (Man
Jaya). Proceedings of the Indonesian Petroleum
Association, 3, 189-196.
• 1977. Tertiary tectonic history of the Salawati area,
lrian Jaya, Indonesia. Proceedings of the Indonesian
Petroleum Association, 6, 199-220.
GARRARD, R. A., SUPANDJONO, J. B. & SURONO. 1988.
The geology of the Banggai-Sula microcontinent,
eastern Indonesia. Proceedings of the Indonesian
Petroleum Association, 17, 23-52.
GIBSON-ROBINSON, C., HENRY, N. M., THOMPSON,
S. J. & HARYONO TRI RAHARIO 1990. Kasim
and Walio fields - Indonesia. In: BEAUMONT,E. A.
(compiler) Stratigraphic Traps, I, AAPG Treatise
of Petroleum Geology, 257-295.
HALL, R., NICHOLS, G. J., BALLANTYNE,P., CHARLTON,T.
& ALl, J. 1991. The character and significance of
basement rocks of the southern Molucca Sea region.
Journal of SE Asian Earth Sciences, 6, 249 258.
HAMILTON, W. 1979. Tectonics of the Indonesian Region.
US Geological Survey Professional Papers, 1078.
HANDIWmlA, Y. E. 1990. The stratigraphy and hydro-
carbon occurrences of the Salodik Group, Tomori
PSC area, East Arm of Sulawesi. Paper presented at
the 19th Annual Convention of the Indonesian
Association of Geologists.
HARTONO, U., AMRI, C. & PIETERS, P. E. 1989. Geological
map of the Mar Sheet, Irian Jaya. 1:250,000.
Geological Research & Development Centre,
Indonesia.
HERMES, J. J. 1968. The Papuan geosyncline and the
concept of geosynclines. Geologie en Mijnbouw,
4 7 , 81-97.
KATILI, J. A. 1978. Past and present geotectonic position
of Sulawesi, Indonesia. Tectonophysics, 45,
289-322.
KERTAPATI, E. K., SOEHAIMI, A. & DJUHANDA, A. 1992.
Seismotectonic map of Indonesia. Geological
Research & Development Centre, Indonesia.
LINTHOUT, K, HELMERS,H & ANDRIESSEN,P. A. M. 1991.
Dextral strike-slip in Central Seram and 3-4.5 Ma
Rb/Sr ages in pre-Triassic metamorphics related
to Early Pliocene clockwise rotation of the Buru-
Seram microplate (E. Indonesia). Journal of SE
Asian Earth Sciences, 6, 335-342.
LIVINGSTONE, H. J. 1992. Hydrocarbon source and
migration, Salawati Basin, Irian Jaya. Eastern
Indonesia Exploration Symposium, Simon
Petroleum Technology/Pertamina.
LUNT, P. & DJAAFAR,R. 1991. Aspects of the stratigraphy
of western Irian Jaya and implications for the
development of sandy facies. Proceedings of the
Indonesian Petroleum Association, 20, 107-124.
MILSOM, J., MASSON, D. 8,: NICHOLS, G. 1992. Three
trench endings in eastern Indonesia. Marine
Geology, 104, 227-241.
MINSTER, J. B. & JORDAN, T. 1978. Present-day plate
motions. Journal of Geophysical Research, 83,
5331-5354.
PIETERS, P. E., HARTONO, U. & AMRI, C. 1981.
Preliminary geologic map of the Mar Quadrangle,
lrian Jaya. 1:250,000. Geological Research &
Development Centre, Indonesia.
--, PIGRAM, C. J., TRAIL, D. S., DOW, D. B., RATMAN,
N. & SUKAMTO, R. 1983. The stratigraphy of
western Irian Jaya. Bulletin of the Geological
Research & Development Centre, Indonesia, 8,
14--48.
PIGOTT, J. D., TRUMBLY, N. I. & O'NEAL, M. V. 1982.
Northern New Guinea wrench fault system: a
manifestation of late Cenozoic interactions between
Australian and Pacific plates. In: Watson, B. (ed.)
Transactions of the Third Circum-Pacific Energy
and Mineral Resources Conference, 613--620.
PIGRAM, C. J. & DAVIES, H. L. 1987. Terranes and
accretion history of the New Guinea orogen. BMR
Journal of Australian Geology & Geophysics, 10,
193-211.
- - & PANGGABEAN,H. 1984. Rifting of the northern
margin of the Australian continent and the origin
of some microcontinents in eastern Indonesia.
Tectonophysics, 107, 331-353.
-- & SUKANTA,U. 1982. Geological data record of the
Taminabuan 1:250,000 sheet area, Irian Jaya. Open
file report, Geological Research & Development
Centre, Indonesia.
-- & -- 1989. Geology of the Taminabuan Sheet
area, lrian Jaya. 1:250,000. Geological Research
& Development Centre, Indonesia.
, CHALLINOR,A. B., HASIBUAN, E, RUSMANA, E. &
HARTONO, U. 1982. Lithostratigraphy of the Misool
Archipelago, Irian Jaya, Indonesia. Geologie en
Mijnbouw, 61,245-279.
, SURONO & SUPANDJONO, J. B. 1985. Origin of
the Sula Platform, eastern Indonesia. Geology, 13,
246-248.
REDMOND,J. L. & KOESOEMADINATA,R. P. 1976. Walio oil
field and the Miocene carbonates of Salawati Basin,
lrian Jaya. Proceedings of the Indonesian Petroleum
Association, 5, 41-57.
RI~HAULT,J. P., MALOD, J. A., LARUE,M., BURHANUDDINN,
S. ~ SARMILI,L. 1991. A new sketch of the central
North Banda Sea, Eastern Indonesia. Journal of SE
Asian Earth Sciences, 6, 329-334.
ROBINSON, K. M. 1987. An overview of source rocks and
oils in Indonesia. Proceedings of the Indonesian
Petroleum Association, 16, 97-122.
RUSMANA,E., KOSWARA,A. & SIMANDJUNTAK,T. O. 1984.
[Report on the geology of the Luwuk Quadrangle,
1:250,000]. Open file report, Geological Research
& Development Centre, Indonesia• [In Indonesian].
--, SUKIDO, SUKARNA, D., HARYONO, E. &
SIMANDJUNTAK, T. O. 1986. Preliminary geological
map of the Lasusua-Kendari Quadrangle, Sulawesi.
1:250,000. Geological Research & Development
Centre, Indonesia.
SATO, T., WESTERMANN, G. E. G., SKWARKO, S. K. &
HASIBUAN, E 1978. Jurassic biostratigraphy of the
Sula islands, Indonesia. Bulletin of the Geological
Research & Development Centre, Indonesia, 4,
1-28.
SIMANDJUNTAK,T. O. 1990. Sedimentology and tectonics
of the collision complex in the East Arm of
Sulawesi, Indonesia. Geologi Indonesia, 13(1),
1-35.
 SALAWATI & TOMORI BASIN CORRELATION
SIMBOLON, B., MARTODJOJO, S. & GUNAWAN,R. 1984.
Geology and hydrocarbon prospects of the pre-
Tertiary system of Misool. Proceedings of the
Indonesian Petroleum Association, 13, 317-340.
SMITH, R. B. & SILVER,E. A. 1991. Geology of a Miocene
collision complex, Buton, eastern Indonesia.
Bulletin of the Geological Society of America, 103,
660-678.
SPT (SIMONPETROLEUMTECHNOLOGY)/PERTAMINA.1992.
Eastern Indonesia: Biostratigraphy, Geochemistry
and Petroleum Geology. Unpublished non-exclusive
consultancy study.
SUKAMTO,R. 1975. Geologic map of Indonesia, sheet 8:
Ujung Pandang. 1:1,000,000. Geological Survey of
Indonesia.
SUPANDJONO,J. B. & HARYONO,E. 1993. Geology of the
Banggai Sheet, Sulawesi-Maluku. 1:250,000.
Geological Research & Development Centre,
Indonesia.
481
SURONO & SUKARNA, D. 1993. Geology of the Sanana
Sheet, Maluku. 1:250,000. Geological Research &
Development Centre, Indonesia.
--, SIMANDJUNTAK, T. O., SITUMORANG, R. L. &
HADIWIJOYO, S. 1987. Geologic map of the Batui
Quadrangle, Sulawesi. 1:250,000. Geological
Research & Development Centre, Indonesia.
TJIA, H. D. 1973. Irian fault zone and Sorong melange.
Sains Malaysiana, 2(1), 13-30.
VINCELETTE,R. R. & SOEPARJADI,R. A. 1976. Oil-bearing
reefs in Salawati Basin of Irian Jaya, Indonesia.
AAPG Bulletin, 60, 1448-1462.
VISSER, W. A. & HERMES,J. J. 1962. Geological results of
the exploration for oil in Netherlands New Guinea.
Koninklijk Nederlands Geologie en. Mijnbouw
Genootschaap Verhandlingen, Geologische Serie,
20.
 The geology and tectonic evolution of the Bacan region,
east Indonesia

JEFFREY E A. M A L A I H O L L O & ROBERT HALL

SE Asia Research Group, Department of Geological Sciences,
University College London, Gower Street, London WCIE 6BT, UK

Abstract: Bacan is located in the zone of convergence between the Eurasian, Philippine Sea
and Australian plates. The oldest rocks in Bacan belong to the Sibela Continental Suite and are
probably of Precambrian age. The complex includes continental phyllites, schists and gneisses of
upper amphibolite facies. Isotopic dating yielded extremely young ages due to interaction with
hydrothermal fluids. Juxtaposed against the continental rocks is the mostly unmetamorphosed,
arc-related Sibela ophiolite, probably derived from the Philippine Sea plate. Isotopic dating
yielded a Cretaceous age with an Oligocene-Miocene overprint. In north Bacan, the oldest
formation is the Upper Eocene Bacan Formation which comprises interbedded arc volcanic and
turbiditic volcaniclastic rocks, metamorphosed under conditions between the prehnite-
pumpellyite and greenschist facies. A similar Lower Miocene sequence, assigned to the same
formation, is exposed in south Bacan. The Oligocene Tawali Formation on Kasiruta, NW of
Bacan, consists of arc basalts and volcaniclastic turbidites, metamorphosed to zeolite facies. The
Bacan and Tawali Formations represent different parts of an arc, active from Late Eocene until
Early Miocene, resulting from northward subduction of the Australian plate under the Philippine
Sea plate. There is a major Lower Miocene unconformity, representing collision of the Australian
continent with the Philippine Sea plate, above which shallow marine limestones of the Lower-
Middle Miocene Ruta Formation were deposited. This deposition was interrupted by sudden
influxes of volcaniclastic sands, forming the Amasing Formation. The Upper Miocene-
Pleistocene Kaputusan Formation, rests locally unconformably on older rocks, and includes three
members. The Goro-goro Member consists of arc andesites, originating from four eruption
centres, which erupted from Late Miocene to Pleistocene, with the oldest in south and the
youngest in north Bacan. The Pacitak Member consists of shallow marine pyroclastic rocks. The
Mandioli Member formed fringing coastal reef limestones. The volcanic rocks of this formation
were produced by eastward subduction of the Molucca Sea plate. Quaternary basalts are related
to movement along the Sorong fault. It is concluded that most of the Bacan region has been part
of the Philippine Sea plate since the Cretaceous; volcanic rocks of different ages all have an arc
character and chemistry, and lithological variations reflect different positions within the volcanic
arc; and there is evidence for continental crust of Australian origin in the Bacan area by the Early
Miocene.

Three major plates, the Eurasian, Philippine Sea
and Australian plates, converge in the Bacan region
(Fig. 1). The Eurasian plate has a complex eastern
margin which includes continental and volcanic arc
crusts accreted during collision with the Philippine
Sea plate (e.g. Rangin 1991). The Philippine Sea
plate has a long arc history (Hall et al. 1995a),
extending back at least to the Cretaceous, and is
currently converging with Eurasia in the
Philippines, and overriding the Molucca Sea plate
in the south. The Australian plate has been moving
northwards since the Cretaceous. During much of
the Palaeogene, oceanic crust of the Australian
plate was subducted under the southern edge of
the Philippine Sea plate (Hall et al. 1995b). A
strike-slip fault system, the Sorong fault, subse-
quently developed at this plate boundary.
Bacan contains rocks of continental, ophiolitic
and arc affinities (Fig. 2), and therefore an under-
standing of the tectonic evolution of Bacan should
greatly enhance our knowledge of the development
of this area of complex plate boundaries. This
paper briefly describes and summarizes the
geology of each of the formations in the Bacan
region, discusses them in the light of regional
tectonic events, and proposes a synthesis of the
tectonic evolution of this region. A detailed
description of the lithostratigraphy of Bacan
(Fig. 3), with chemical and palaeoenvironmental
+analyses can be found in Malaihollo (1993).
Geology and tectonic evolution of Bacan
Sibela continental suite
The oldest rocks in the Bacan region form part of
the Sibela Continental Suite. The complex includes
continental phyllites, schists and gneisses of upper

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of Southeast Asia, 483
Geological Society Special Publication No. 106, pp. 483-497.
484 J.F.A. MALAIHOLLO 8z R. HALL
500 krn
-4°N
CELEBESSEA
GULF OF TOMINI
MOLUCCA
SEA
I
I BACAN
FLORES SEA
o
Fig. 1. Present-day tectonic setting of east Indonesia based on Hamilton (1979).
amphibolite facies (c. 600°C, c. 5 kbar), with strong
penetrative fabrics indicative of a polyphase
deformation and recrystallization history, typical
of Barrovian type dynamo-thermal metamorphism
(Brouwer 1923; Hall et al. 1988). There is also
evidence for retrograde metamorphism associated
with post peak-metamorphism deformation. The
protoliths were pelites with minor amounts of
sandy and carbonate/marly horizons. Whole rock
chemistry of the pelites suggests that they were
derived from a cratonic area and were deposited on
an active margin underlain by continental crust.
Based on regional stratigraphical arguments
(Hamilton 1979) and the Pb isotopic signatures
of Quatemary volcanic rocks (Vroon 1992), these
metamorphic rocks are postulated to be of
Precambrian or Palaeozoic age. This suggestion
has not been confirmed isotopically, as K-Ar and
Ar-Ar step heating results yielded extremely young
ages (< 0.21 Ma; Malaihollo 1993), interpreted to
be a result of interaction with a hydrothermal fluid
carrying fractionated argon which were subse-
quently concentrated between the sheet silicate
layers, causing an apparent young age. This mech-
anism could account for other unusually young
ages from high-grade metamorphic rocks in the
region (cf. Linthout et al. 1991).
There are highly foliated metasedimentary rocks
PHILIPPINE
SEA PACIFIC
f~ OCEAN
I 2
Morotai
~ LMAHERA ~ano'kwari 1"to,.
o~.. Waigeo "
-.,-... o, Gag ~ ~ ~$orang Fau, "~0.~(
~ Bird's
Head
IRIAN JAYA
BANDA SEA
ARAFURA SEA
D
in the south Saleh Islands and the SE comer of
north Bacan. These rocks have a similar character
to the Sibela Continental Suite, particularly in
having brown metamorphic biotite which through-
out the whole region is found only in the Sibela
Continental Suite, but are of lower metamorphic
grade.
This paper interprets the Sibela Continental Suite
to be derived from the Australian plate, where these
rocks formed either part of the Mesozoic-Tertiary
passive margin of north Australia or fragments
rifted from Australia during Gondwana break-up.
Many authors have suggested or implied that the
continental rocks in Bacan were introduced into the
region via strike-slip motion of the Sorong fault
system from the Bird's Head region of Irian Jaya
(e.g. Hamilton 1979; Silver et al. 1985) or central
Papua New Guinea (Pigram & Panggabean 1984).
87Sr/86Sr isotopic compositions of volcanic rocks
(E. Forde, pers.comm. 1993) suggest that the Upper
Miocene Kaputusan Formation on south Bacan was
erupted through continental crust, indicating the
presence of continental material under the Bacan
region by the Late Miocene. The presence of a
Lower Miocene post-collisional intrusion (the Nusa
Babi Monzodiorite) in Bacan suggests the presence
of continental crust by the Early Miocene (as
discussed below).
 "• J127°15'E I ~27°3°'E 1127"45'E
s' I LEGEND •
[--1 Alluvium&
t 7 ,,' Limestones
(~ Volcanics
~ Kaputusan Fm
I ~ Mandioli Mb
! o~ 0"~5"~
~_ Kaputusan Fm
,,s ® Pacitak Mb
Nusa Babi ~ Kaputusan Fm
Saleh - Goro-Goro Mb
~ Saleh
Isles~, Intrusions ©
Goro-goro v.' ~ Amasing Fm ©
~ ~,,%- i ~ Ruta Fm
~ NusaBobi
Intrusions
8 Tawali Fm ('1
I'7~_-~'~ Bacan Fm
Kasiruta ~e--- , ~ ~ , ~
0,30,'~
o ©
(TawaliBesar) - =~-~. . "
.I~" K,apuTusan ~, o~
o Sibela Ophiolite
L)

i t a k ~
z
Continental
® Suite
Z
0
:.Sibela Mts::
>
m
0"45'S

i i

0 10 km 20 30

Fig. 2. New simplified geological map of the Bacan region, based on aerial photographic interpretation and field mapping (Malaihollo 1993). 4~
Oo
C~h
486 J.F.A. MALAIHOLLO 8Z R. HALL

Ma EPOCH Kasiruta Bacan Possible Tectonic Events

0- Alluvium Reef ~ Quaternary volcanism related to movements on Sarong Fault.Upliftof Sibela Block.
Pleistocene Thrustingin Halmahera, movement on splaysof Sarong Fault.

aputusan m Kapulusan FB Associated subaerial and shallow marine pyroclasticdeposits,and fringingreefs,

Pliocene
Goro-goro Mbr Goro-goro Mbr Four eruptive centres which migrated northwards with time.
5.2 Pacitak Mbr Pacltak Mbr Start of volcanism in Bacan (-7 Ma), Halmahera and Obi (- 12 Ma).
Mandioli Mbr Mandioli Mbr

10,4 Initiationof eastward subducfion of Molucca Sea Plate under Halmahera,

Miocene

16.3 m
Ruta Shallow marine carbonate platform [open platform, patch reef,tidalbar and foreslope facies).
Formation Local shallow marine volcaniclasticand littoraldeposit,

Collisionof Australiaand arc on the PhilippineSea plate.

23.3
Arc volcanic and turbidites,
Tawali Fm Repeated, high concentration, proximal volcaniclasticturbidites, i Metamorphosed to Prehnite-
Inner and middle fan facies, Pumpellylte to Lower
Marikapal Mbr
Bacon I
Greenschist facies.

Oligocene Formation Volcanism related to the

northward movement of
Tawali Fm Pillowlavas with volcanic arc chemistry. Australiaunder the
Zeolitefaciesmetamorphism.
Jojok Mbr PhilippineSea plate,
I
I Similarsequence in Bird's
35.4 Head and southern
Philippines,
Eocene

Fig. 3. Cenozoic stratigraphy of the Bacan region with related local tectonic events. Timescale after Harland et al.
(1990).

Sibela ophiolite
Juxtaposed against the continental suite is the
Sibela ophiolite (Hall et al. 1988). There are two
types of material recognized in this suite: ophiolitic
rocks, most of which are of lower crustal plutonic
and mantle origin (microgabbros, gabbros and
serpentinised harzburgites) with few volcanic
rocks, and spatially-related rocks including unusual
amphibole cumulates and many rocks with
magmatic-tectonic fabrics. This dismembered,
incomplete ophiolitic complex is mostly un-
metamorphosed, although locally there are
mylonitic rocks which have been affected by upper
amphibolite-lower granulite facies metamorphism
(c. 1000oc, c. 5 kbar) related to ductile deformation
of hot rocks at or near their place of formation, and
local recrystallization in shear zones (Malaihollo
1993).
Chrome spinel compositions suggest that the
peridotites may have formed in an arc setting. This
is supported by geochemical evidence from
cogenetic metagabbros. Petrographic study reveals
that most of the cumulate rocks consist mainly of
cumulus amphibole with minor pyroxene and inter-
cumulate plagioclase, indicating crystallization
under hydrous conditions, implying an arc-related
setting. The cumulate rocks in the Sibela ophiolite
are very different from those of the Halmahera
ophiolite, which consist of olivine, orthopyroxene
and clinopyroxene (Ballantyne 1992). The
Halmahera cumulates are genetically related to the
ophiolite which is interpreted to have formed in a
supra-subduction zone, normally associated with
the initiation of a subduction zone, from the intra-
Pacific subduction (Ballantyne & Hall 1990;
Ballantyne 1991). Preliminary Nd-Sm ages on the
Halmahera cumulates are Jurassic (M. E Thirlwall,
pers. comm. 1992). Ar-Ar and K-Ar dating of
the Sibela cumulates and metagabbro (Malaihollo
1993) yielded Cretaceous (97-94 Ma) and
Oligocene-Miocene (37-25 Ma) ages, although
amphiboles analysed contain excess argon and
therefore plateaux ages will be slightly older than
closure ages (Lanphere & Dalrymple 1976). The
Cretaceous age may be linked to volcanic activity
recorded in the east Halmahera ophiolite (94-
80Ma; Ballantyne 1992), suggesting a link
between the Sibela ophiolite and the east
Halmahera ophiolite. The Oligocene-Miocene ages
are related to the Oligocene volcanism (discussed
below).
 GEOLOGY • TECTONICS OF BACAN, E INDONESIA
There are two possible interpretations for the
association of ophiolitic and cumulate rocks. The
first possibility is that the ophiolitic rocks represent
older crust of possibly Jurassic age (cf. east
Halmahera-Gag ophiolite) intruded by Cretaceous
arc cumulates related to intra-oceanic subduction.
This interpretation is supported by chemical
similarities between the cumulates and zoned
ultramafic complexes, normally associated with arc
magmatism. Alternatively, both the ophiolitic and
arc cumulate rocks may be related to the same
intra-oceanic subduction during the Cretaceous.
The age and nature of the original crust upon which
the arc was built is unknown.
Juxtaposition of continental and
ophiolitic rocks
The continental and ophiolitic rocks are juxtaposed
in the Sibela Mountains. Differences in meta-
morphic character indicate that metamorphism
occurred before amalgamation of continental and
ophiolitic rocks. The continental suite is interpreted
to be part of the Australian continental basement,
whereas the ophiolitic rocks represent the
Philippine Sea plate. Regional mapping shows that
since the Early Miocene, rocks from these two
plates have the same geological history, indicating
an Early Miocene collision (Hall et al. 1995a).
There are three possibilities for the mode of juxta-
position of the continental and ophiolitic rocks:
(1) the Sibela Mountains represent a suture zone
resulting from collision between the Australian and
Philippine Sea plates; (2) the continental rocks
collided with the Philippine Sea plate and were
subsequently translated by strike-slip faulting as
part of a composite fragment to their present
position; (3) the continental rocks remained part of
the north Australian Margin until the late Neogene
and were then translated by strike-slip faulting on
the Strong fault into the region. These possibilities
are discussed below.
The Bacan Formation
In north Bacan the oldest rocks exposed belong
to the Upper Eocene Bacan Formation. This
comprises interbedded basic-intermediate volcanic
and volcaniclastic turbiditic rocks. The Bacan
Formation was folded and subsequently meta-
morphosed under conditions transitional between
the prehnite-pumpellyite and lower greenschist
facies (250-330°C, c. 2 kbar upwards), with
characteristics of burial metamorphism. Hydro-
thermal alteration may be related to the Neogene
Kaputusan volcanism. Mineral and whole rock
geochemistry indicates that the volcanic rocks were
487
erupted in an arc setting. Turbiditic rocks were
deposited by dilute, low concentration, possibly
distal or low energy currents. The Bacan Formation
was affected by a thermal event at c. 15 Ma.
In south Bacan the oldest rocks dated are a
very similar sequence of Early Miocene age.
These interbedded volcanic and volcaniclastic
turbidite rocks were metamorphosed to prehnite-
pumpellyite facies (c. 240-330°C, c. 2 kbar) due
to burial metamorphism, with local hydrothermal
alteration. Whole rock and mineral chemistry
indicates an arc origin for the volcanic rocks.
Graded volcaniclastic arenites indicate deposition
by turbidity currents which dissipated their energy
as they travelled down slope. Isotopic dating shows
that this formation was affected by a thermal event
atc. 8Ma.
Undated metabasites in the Saleh Islands were
metamorphosed at conditions between upper
prehnite-pumpellyite and lower greenschist facies
(c. 250-360°C, c. 4 kbar). The metamorphic
character of these rocks suggests static, regional
metamorphism. Geochemical and lithological
evidence from the metabasites indicates that they
represent an arc-related calc-alkaline sequence,
possibly formed in a backarc, and they have similar
chemical characteristics to the Upper Eocene and
Lower Miocene sequences.
The Upper Eocene, Lower Miocene and the
Saleh Island sequences are similar in all aspects,
except age. More dating is needed, although this
may prove difficult, since despite examination of
> 200 thin sections, only four samples were found
to be suitable for biostratigraphic or isotopic dating.
This is attributed to lack of fossils in the original
sediments, metamorphism which destroyed fossils,
and alteration by thermal overprints of the volcanic
rocks. The rocks assigned to the Bacan Formation
could represent either different parts of a single,
continuously active arc, or separate arcs active at
different times between the Late Eocene and Early
Miocene.
The Tawali Formation
The oldest rocks on Kasiruta belong to the newly
defined Tawali Formation, named after the island
of Kasiruta (also known as Tawali Besar). Rocks
forming part of this formation were originally
assigned to the Bacan Formation (Yasin 1980), but
here they are assigned to a new formation because
of differences in lithology and the character of
metamorphism and deformation. The lower part of
the Tawali Formation consists of basaltic pillow
lavas and interbedded sediments (Jojok Member),
and it is overlain by fossiliferous volcaniclastic
turbidites (Marikapal Member). This formation
is affected by burial metamorphism transitional
488 J.F.A. MALAIHOLLO & R. HALL
between low and high temperature zeolite facies
(c. 180°C, < 2 kbar). Whole rock chemistry indi-
cates that the Lower Oligocene basalts are highly
differentiated arc lavas, erupted in a deep, open
marine environment above the CCD. The Upper
Oligocene volcaniclastic rocks are products of
repeated, high concentration, proximal turbidity
currents with associated slumped deposits.
The Tawali Formation is interpreted as an
equivalent of the Bacan Formation although there
are several differences between them. There is an
apparent lack of Oligocene rocks on Bacan,
although as noted above, only a few samples of
the Bacan Formation have proved dateable. There
are also differences in lithology and metamorphic
character (particularly the lower grade of meta-
morphism suffered by the Tawali Formation rela-
tive to the Bacan Formation), and trace element
differences between the Tawali and the Bacan
Formations. Rocks of similar lithology and age
to the Tawali Formation have now been identified
throughout the Halmahera-Waigeo region (the
Tawali Formation in Halmahera, the Anggai River
Formation in Obi and the Rumai Formation in
Waigeo; Hall et al. 1991). The Oha Formation of
Halmahera is probably not of Upper Cretaceous-
Eocene age as previously suggested (Hakim & Hall
1991), but is interpreted as part of the same arc as
the Bacan Formation. Despite differences, all these
formations range in age from Upper Eocene to
Upper Miocene and they all pre-date the c. 22 Ma
unconformity (discussed below) interpreted to be
related to the arrival of the continental crust in the
region. Differences in lithology could be explained
if the different formations represent different parts
of the arc (e.g. arc slopes, backarc basin, forearc
slopes). Differences in the grades of metamorphism
TawaliFm,,MarikapalMb.&
RumaiFm.(turbidites) BacanFm.arc BacanFm.&AnggaiRiverFm,
~ ~
(back-arcpillowbasalts)
. X~4
Philippine Sea plate
Fig. 4. Different positions, in the pre-collisional stack, of the volcanic formations in the Bacan-Halmahera-Waigeo
region which were part of the Late Eocene-Early Miocene arc.
could be a function of different structural positions
during collision with the Australian continent or
different thermal environments due to different
positions in the pre-collision arc.
Thus, the simplest explanation for all these
Upper Eocene, Oligocene and Lower Miocene
volcanic arc formations is that they represent arc
volcanism at the edge of the Philippine Sea plate
due to the subduction of the Indo-Australian plate
under the Philippine Sea plate (Fig.4). This is
supported by the fact that all of these formations
record similar palaeomagnetic declinations and
inclinations which are characteristic of the
Philippine Sea plate (Hall et al. 1995a; Ali & Hall
1995).
Early Miocene unconformity
Above the Bacan and Tawali Formations, there is a
major regional unconformity. This is characterized
by a major change in lithological, metamorphic and
structural character. Rocks below the unconformity
are folded arc volcanic rocks and turbidites, meta-
morphosed up to greenschist facies, whereas the
rocks immediately above it are mostly unfolded,
unmetamorphosed carbonates. The age of the
unconformity is estimated to be c. 22 Ma (Early
Miocene) throughout the Halmahera region (Hall
et al. 1995a).
Nusa Babi intrusive rocks
The Nusa Babi Monzodiorite (NBM) includes
quartz-monzodiorite dykes and plugs with associ-
ated aplite dykes of monzogranite composition.
These intrude the Bacan Formation and have been
reported to intrude the Sibela Complex (Yasin
(forearcturbidites
SibelaContinentalSuite
× ×X XX ,
Australian plate
No Scale Intended
 GEOLOGY & TECTONICS OF BACAN, E INDONESIA
1980), but do not intrude the Ruta and Amasing
Formations. The NBM is locally altered, possibly
due to auto-metasmorphism. Magma evolution was
primarily controlled by fractionation and the addi-
tion of water. Whole rock geochemistry indicates
that the NBM is of plutonic arc or collisional
magmatism origin. K-Ar dating yields an Early
Miocene age (Baker & Malaihollo 1996). The geo-
chemistry, isotopic age and the fact that the NBM
intrudes the Bacan Formation indicate a post-
collisional origin, probably related to the arrival
of the Sibela Continental Suite in the region.
The Ruta and A m a s i n g Formations
The Lower-Middle Miocene Ruta Limestone lies
unconformably above the Bacan and Tawali
Formations. There are four microfacies recognized
in the Ruta Limestone: skeletal wackestones-
packstones, algal boundstones, foraminiferal
packstones and bioclastic-lithoclastic packstones.
These represent deposition on an open platform,
a platform margin build-up (patch reef), a tidal bar
downslope from a patch reef (open platform) and
foreslope talus respectively, all forming part of a
shallow marine carbonate platform. The deposition
of the Ruta Formation began in the Early Miocene,
predominantly as open platform, platform margin
build-up and foreslope talus facies• Between the
Early~ Miocene and Middle Miocene, deposition
was locally interrupted by sudden and high
influxes of volcaniclastic material forming sand-
stones of the Amasing Formation. Three facies in
the Amasing Formation have been recognized:
shallow marine with storm horizons, shoal or
estuarine, and beach deposits. These facies
represent shallowing upwards from an open
carbonate platform to a beach environment.
Deposition of the Ruta Formation continued until
the Middle Miocene and its upper part is domi-
nated by the tidal bar facies. There was widespread
development of carbonate platforms at this time
throughout the Halmahera-Waigeo region (Hall
et al. 1995a), in north Irian Jaya (Pigram & Davies
1987; Pigram et al. 1990) and in the Philippines
(Mitchell et al. 1986). Widespread carbonate
platform development suggests that most of the
Bacan-Halmahera-Waigeo region was in a quiet
tectonic setting, in an equatorial position and was
part of an extensive shallow marine area.
In north Irian Jaya, the Middle Miocene Moon
Volcanics (Pieters et al. 1989) may be linked with
the Maramuni arc of Papua New Guinea which
Dow (1977) suggested represented post-collision
subduction reversal. However, the character and
distribution of this volcanism is still poorly known
and if an arc existed, it may have been limited to
the eastern part of New Guinea (Ali & Hall 1995).
489
The Saleh Diorite
The Saleh Diorite intrudes the Sibela Continental
Suite in the Saleh Islands and is juxtaposed against
it in central Bacan. This intrusion consists of
amphibole-bearing diorite and micro-diorite.
Locally these rocks have suffered low temperature
alteration, probably due to auto-metamorphism,
and subsequent hydrothermal alteration. Whole
rock geochemistry indicates that the Saleh Diorite
is different from the NBM and is of arc or post-
collisional origin. K-Ar dates suggest the c. 15 Ma
age is related to the initiation of the Halmahera arc.
The Kaputusan Formation
The Kaputusan Formation was deposited above
the Ruta and Amasing Formations. It consists of
three members: the Goro-goro Volcanic, the Pacitak
Volcaniclastic, and the Mandioli Limestone
Members• Field mapping and aerial photographic
interpretation indicates that the Kaputusan
Formation rests directly upon the Bacan Formation
in some areas implying an unconformable contact
at its base.
The Goro-goro Member consists of two-
pyroxene andesites (TPAN), hornblende-pyroxene
andesites (HPAN), hornblende andesites (HBAN),
and hornblende-biotite andesites (HBIAN).
Associated with the volcanic rocks are subaqueous
pyroclastic flows with related base surge deposits.
The petrography, mineral chemistry, whole rock
major and trace element chemistry of these rocks
are typical of island-arc volcanic rocks. They are
mostly fresh, akhough locally they are metamor-
phosed to zeolite facies, interpreted as due to
hydrothermal activity. Magma diversification of
the Goro-goro Member was achieved mainly by
fractionation of plagioclase, pyroxene, Fe-Ti oxide,
amphibole and biotite, with evidence for replenish-
ment, immiscibility, resorption and assimilation.
Magmatic conditions were < 8 kbar and 2-10 wt.%
H20 with PH2'-' < Ptotal" The Goro-goro Member
• U
includes at least four eruption centres (as defined
by bulk trace element ratios): in south Bacan
(mostly HBIAN), the Goro-goro area (mostly
TPAN), and the north Mandioli and the Kaputusan
areas (largely HPAN and HBAN). The Mandioli
group are shoshonitic rocks, which may indicate
eruption in an extensional setting within a domi-
nantly convergent zone, such as a strike-slip zone
where similar rocks commonly occur (e.g. Ellam
et al. 1988). K-Ar ages indicate that these centres ~
were erupting from the Late Miocene (c. 7.5 Ma)
to the Pliocene (c. 2.2 Ma). Each of these centres
appears to have been active for c. 2 Ma, with a
history of multiple eruptions. The Upper Miocene-
Upper Pliocene Pacitak Member consists of
490 J.F.A. MALAIHOLLO ,~ R. HALL

reworked pyroclastic and volcaniclastic material Discussion

from the Goro-goro Member, deposited in an
The history and motion of the Philippine Sea plate
oxygen-rich, nearshore, shallow marine environ-
has been complex, although recent palaeomagnetic
ment. The Upper Miocene-Lower Pliocene
results are contributing to a clearer picture. The
Mandioli Member consists of wackestones and
work of Hall et al. (1995a, b) suggests that between
packstones which formed fringing coastal reefs.
c. 50-40 Ma the plate experienced a c. 50 ° clock-
The Kaputusan Formation is interpreted to be the
wise rotation with a southward translation; no
product of the eastward subduction of the Molucca
rotation between 40 to 25 Ma, and from 25 to 0 Ma
Sea plate under Halmahera, and can be correlated
the plate rotated clockwise 35 ° with northward
with the Weda Group of Halmahera and the Woi
translation. The first rotation may be related to the
Formation of Obi (Hall et al. 1995a; A l i & Hall
change of motion in the Pacific plate (Clague &
1995).
Jarrard 1973) at about 42 Ma. Australia was
moving north throughout the Tertiary and we
Q u a t e r n a r y deposits interpret the regional unconformity at c. 22 Ma as
the result of collision of Australia with a volcanic
Although currently there are no active volcanoes
arc at the southern edge of the Philippine Sea
on Bacan, Quaternary volcanic rocks are present.
plate (discussed below). This may have caused or
These are fresh olivine-, pyroxene- and plagio-
contributed to renewed clockwise rotation of the
clase-phyric basalts. Mineral and whole rock geo-
Philippine Sea plate.
chemistry indicate that these are arc basalts which
erupted through a quartz-rich basement (Sibela
Continental Suite). There is evidence of a within-
plate basalt component from whole rock and Pre-Tertiary and early Tertiary
mineral chemistry, possibly indicating magmatism evolution of Bacan
related to movements along the sinistral Sorong
fault. This is supported by the linear distribution There are two types of pre-Tertiary rocks in Bacan:
of the volcanic centres, and in accordance with continental crust and ophiolitic/arc material. The
the view of Silitonga et al. (1981). Detritus from continental crust is presumably of Australian
the Sibela Continental Suite can be found solely margin or micro-continental origin; north of this
in the Quaternary deposits, indicating that it margin was oceanic crust. Ophiolite/arc material in
has become available for erosion only recently, Bacan records Cretaceous and Oligo-Miocene ages,
suggesting extremely fast rates of uplift (c. both of which can be correlated with ages recorded
1 mm a-l). in Halmahera which now forms part of the

I / / I / I / I / / / / / / / I ~
,~/, ,~//I Philippine [~ /, // /, /, /, 1 i/i/ill/i/i///// 4 / / ~ P h i l i p p i n e i~,
// JPhilippi ne I ii i,i i,i ii i i i , / 4 Sea Plate I /
[Se 0 .Plgte b / / / / / ] , ~ / L./ / / / ///
/I/I/////~ / e a Plate / / / / / / / eurasia I = , i # / ~ / / / /
IIIIIIIIIII/I// ~ J l l l l l ~
i i i / / / / / / 1 1 1 / / i - " I II i
IIIIIIIIIIlilll
.~111111~-~1
~ ~
:::.::::.
/ / /
i
'Iregion
. . . . .
[ /
I
::::::'*::::,. i .

i:iSi:i:!:!:!:!.. ::'.::::

Late Eocene Oligocene Early Miocene

Fig. 5. Simplified tectonic setting of the Bacan region during the Late Palaeogene-Early Miocene. The oceanic crust
of the Australian plate was being subducted under the Philippine Sea plate. The Bacan region moved from arc-forearc
(Late Eocene) to a backarc (Oligocene) setting and returned to an arc-forearc position (Early Miocene).
 GEOLOGY ~ TECTONICS OF BACAN, E INDONESIA
Philippine Sea plate. This implies that the ophiolitic
rocks of Bacan were part of the Philippine Sea plate
since at least the Cretaceous. Plate tectonic recon-
structions show that almost all the oceanic floor
of the Mesozoic west Pacific has long since been
subducted. The ophiolites of Halmahera and
Waigeo, which represent the oldest parts of the
Philippine Sea plate, were formed in an intra-
oceanic arc setting (Ballantyne 1991, 1992) and the
Cretaceous Halmahera volcanic arc was situated at
sub-equatorial latitudes (Hall et al. 1995a). There
is no lithostratigraphic evidence on Bacan for the
history of the region between the Cretaceous and
Late Eocene.
45-22 Ma
Between the Late Eocene and Early Miocene there
was extensive arc activity throughout the Bacan-
Halmahera region. Sukamto et al. (1981) suggested
that Oligocene volcanism in the Bacan-Halmahera
region was the result of west-dipping subduction
from the Pacific side, implying that the region lay
on the Eurasian plate. Recent palaeomagnetic
evidence (Hall et al. 1995a) from the Tawali
Formation shows that the region was part of the
Philippine Sea plate. Oceanic crust of the
Australian plate was subducted northwards,
forming an arc at the southern margin of the
Philippine Sea plate and fragments of this arc
extend from northern New Guinea, through the
Bacan-Halmahera-Waigeo region, into the east
Philippines (Rangin et al. 1990; Hall 1995; Hall
et al. 1995a). The Oligocene Tawali Formation is
interpreted as part of this volcanic arc.
The Upper Eocene to Lower Miocene Bacan
Formation is genetically related to the Tawali
Formation and represents temporal and spatial
variations of the same arc (Fig. 5). There are two
possible correlatable volcanic arcs in the east
Indonesia-west Pacific area: the Oligocene-
Miocene volcanic rocks of Irian Jaya and Papua
New Guinea (e.g. Pigram & Davies 1987) or the
Eocene-Oligocene Palau-Kyushu remnant arc
(Karig 1975; Sutter & Snee 1980). The Palau-
Kyushu and the West Mariana ridges are associated
with opening of the Parece Vela Basin (at 30-
17 Ma) and these remnant arcs are attributed to
subduction of Pacific ocean (e.g. Uyeda & Ben-
Avraham 1972; Seno & Maruyama 1984) and are
therefore different from the arc represented by the
Tawali Formation. Correlation of volcanic rocks
in Bacan and the Bird's Head was suggested by
Van Bemmelen (1949) and Verstappen (1960). The
arc volcanic rocks of north Irian Jaya, for example
the Arfak (Ratman & Robinson 1981; Pieters et al.
1982), Batanta (Sanyoto et al. 1985), Yapen
(Atmawinata et al. 1989) and Mandi Formations
491
(Pieters et al. 1989) and Papua New Guinea, for
example the Bismarck Volcanic Province (Dow
1977), are interpreted to be related to subduction
between the Australian and Pacific plates (Dow
1977; Pigram & Davies 1987) and may be a
continuation of the Bacan-Tawali Formation
volcanic arc.
22 Ma unconformity
A collision between the Australian margin and an
island-arc has been implied or suggested by many
authors (e.g. Dow 1977; Jaques & Robinson 1977;
Pieters et al. 1983; Pigram & Davies 1987),
although suggestions of the age of this collision
vary. Pigram & Symonds (1991) review estimates
of its age ranging from Eocene to Late Miocene; for
east New Guinea they argue for a collision un-
conformity at c. 30Ma. Charlton et al. (1991)
tentatively suggested a mid-Oligocene age for
collision on Waigeo, based on a poorly defined
Upper Oligocene age of the Mayalibit Formation.
However, new isotopic dating of volcanic rocks
from this formation indicates a narrow age range,
of uppermost Oligocene-lowermost Miocene (our
unpublished results). At the south end of Mayalibit
Bay on Waigeo these volcanic rocks dip at up to
40 ° beneath Miocene limestones, suggesting they
are below the unconformity.
The presence of the Australian continental rocks
on Bacan could be due to: (1) Early Miocene
collision; (2) Early Miocene collision followed by
Neogene strike-slip translation; or (3) Pliocene or
younger pure strike-slip translation. Although there
is no indisputable evidence, the proximity of the
Sibela Continental Suite to the arc, interpreted to be
the result of subduction of the Indo-Australian plate
under the Philippine Sea plate, and the juxta-
position of the continental rocks with the ophiolite
interpreted to represent Philippine Sea plate
material favours a collisional interpretation (1 or 2).
The character of the Nusa Babi Monzodiorite,
suggesting post-collisional melting of a continental
source, and the evidence of a continental crustal
contribution to Late Miocene volcanic rocks also
suggest the presence of a continental basement
beneath central and south Bacan by the early
Neogene.
On Bacan there is a regional unconformity of
Early Miocene age which is interpreted as resulting
from the collision of Australian and Philippine
Sea plates (Fig. 6). There is a change from folded,
metamorphosed arc rocks to unmetamorphosed
carbonates and a possible stitching intrusion of
Philippine Sea plate rocks (the Bacan and Tawali
Formations) and Australian plate rocks (Sibela
Continental Suite) by the Nusa Babi Monzodiorite
which is of Early Miocene age. This interpretation
492 J.F.A. MALAIHOLLO • R. HALL
Fig. 6. Simplified tectonic setting of the Bacan region during the Early Miocene. Continuation of subduction
culminated in collision and thrusting of the continental crust, under the arc-forearc region of the Philippine Sea plate.
Following collision, the Philippine Sea plate rotated westward.
is supported by a similar lithostratigraphy through-
out the Bacan-Halmahera region (Hall et al. 1988
1991), where most of the Lower to Middle Miocene
is represented by carbonate rocks deposited above
Upper Eocene-Lower Miocene arc rocks of
Philippine Sea plate affinity and rocks of Australian
continental affinity. It is therefore concluded that
the collision in east Indonesia occurred in the Early
Miocene and the authors agree with Ali & Hall
(1995) who suggest that much of the evidence from
east Indonesia and New Guinea can be understood
in terms of pre-Miocene intra-oceanic arc tectonics,
an arc-continent collision at about 22 Ma, and
Neogene strike-slip tectonics.
19-15 Ma
After the Early Miocene collision, the region was
uplifted and deposition of shallow marine
carbonates followed. A tectonically quiescent
period is envisaged at this time. Localized uplift
contributed influxes of volcaniclastic material onto
the carbonate platform (Fig. 7), probably derived
from the pre-Miocene formations.
Initiation of Molucca Sea subduction and
s u b s e q u e n t arc v o l c a n i s m
Northward movement of Australia during the
Neogene occurred without subduction at the
boundary of the Australian and Philippine Sea
plates, which was the left-lateral strike-slip Sorong
fault system. However, arc volcanism did result
from the eastward subduction of the Molucca Sea
////////////////////
/ ///~/ /~/~/ /~l ~P~h'i'l' ip i e ] p n
~ / / / / /, ,
//// Platel/////
// / // / // /// /// /// /// /// /~Sea
,/
~-.i~/Iregion ," !
.:.:....:.:.:.:.::..:.
,...................
+
1~ : ! : ! $t . - '.-.mi!~!ii!!~iiiiiiiiir,~+
: i~: ! : ! : +! : ! : ! :i -:
,:-:-.*.:o:-.'.;..'-,'..'.." ~r ................ ===================== ,',.'.
':':':':':':':':':':':';'. + x "1";'.":
.-'.:.:.:.;-'.:.-'.'..'.'..:.:.:.:.;.. + ;.;.:.:.,.;
plate under Halmahera (the Philippine Sea plate) at
the Halmahera trench (Fig. 8). On Bacan there is
a local unconformity of Late Miocene age, and in
places the Upper Miocene Kaputusan Formation
rests directly on the Bacan Formation. The
Kaputusan Formation is the equivalent of other
Upper Miocene arc sequences on Halmahera and
Obi. Although the oldest isotopic age obtained
from the Kaputusan Formation is c. 7.5 Ma, there
was a thermal event at c. 15 Ma recorded by wide-
spread resetting of isotopic ages in the Bacan
Formation and rocks on Obi. The Saleh Diorite,
interpreted as a precursor to the Kaputusan
volcanics, was intruded at c. 15 Ma. The oldest
isotopic ages obtained from volcanic sequences on
Obi are c. 12 Ma (Baker & Malaihollo 1996). Thus,
initiation of subduction possibly started at c. 15 Ma
and the first volcanic products were erupted at
c. 12Ma.
Late Neogene-Recent
At c. 3 Ma there was more than 60 km shortening
between east and west Halmahera, attributed to
movement along the Sorong fault (Nichols & Hall
1991). Splays of the Sorong fault running through
the Bacan region may have contributed to the
ending of Kaputusan volcanism. Quaternary
volcanism was later reactivated along the fault
splays (Fig. 9). These faults are also responsible for
the shaping of Bacan coastlines and the creation of
deep basins surrounding Bacan. The culmination
of collision processes in the Molucca Sea will
eventually result in the transfer of the Bacan-
 GEOLOGY & TECTONICS OF BACAN, E INDONESIA 493

~ / / / / / / / / / / /
I;~:::::::::::::::~////l Philippine 1/
/ / / / ] S e a Plate[ /
I-.----------4////) ...... /
1::74=. . . . . ~ , . , I : = ~ / / / / " I I . Z _ ~ / / /
#,! ~ ~ ~, , ...... 1 : ~ ~ I I / / / / / / / / / / / / /
r~ ~"q,','- ~ 2,',',',',','," l---~--'---J/..,~../////////

,1~,,~ I I I I I~ } t I I I I I 1 i i i Ie

~. " ~ ~'~I.~L~ ° " "'""-41 I I I I I ~'~-~.# I I I I ['- I ' ~ _ t ~ + ~ % +~J~l-~,~_,_%~_~o?

¢- ~ ~ ~ - -,,a,~ • ~ ~ ~ ~ ~ , ~ ~ ~ ~' I ~ " ~ " ~-",ae,-".,. ~':':.:.:,
~ ~ i, , , , , ,_.t , , , , , ~ { i : ~ . . , a k T d ] ~ i ' % ~ : ; : ; : ; :

, i i i i I .%1 i i i ..if .%:.:

LOCQI ~ ~ ~ + ~ ! i
volcaniclastic ~ I ~ ~ + ~ ' ~
ioput - " ::iiii

Fig. 7. Simplified tectonic setting of the Bacan region during the Middle Miocene. Northward movement of Australia
was accommodated by a transform boundary, which moved north with Australia (Hall e t al. 1995a, b). The Philippine
Sea plate was rotating clockwise due to subduction on all sides. There was widespread development of platform
carbonates in the Bacan region, away from the plate boundary.

H a l m a h e r a region f r o m the edge of the Philippine region. In a regional context, it has contributed to
Sea plate to the Eurasian margin (Hall & Nichols an u n d e r s t a n d i n g o f the tectonic history o f the
1990). convergent zone b e t w e e n the Australian, Philippine
Sea and Eurasian plates in east Indonesia. S o m e o f
the m o s t important implications of this w o r k are:
Implications for regional tectonics (1) M o s t of Bacan is interpreted as part o f the
This study has p r o v i d e d n e w geological data and Philippine Sea plate. Cretaceous links are indi-
elucidated the tectonic d e v e l o p m e n t of the Bacan cated by the presence of the Sibela ophiolite,

'~ F / ]Philippine I
./ I¢ " 4seQ PIQtel
! E / / "" " "" "
~lZl irn~in ~ F / / / / / / / /

A:~:i:i&. . . . . . . . .:
~. ,&:~!~F~i~."i!~!##!~7 I " VolcQniccentres
~, "~;~i~i;ii;;i;!!;!!!;/' I~ ~g,%%o~,oMb
!!i!i.:i!i.:i::"..i!i!i!i!i!i!i!i!i!:====================================================

!:!:!:~:!:..':!:!:..':!:!:..':!:!:!:!:!:
~iiii~iiiiiiiiiiiiiiiii~-
::;:::::;St::::::;:::::;::::::::::, + : :::::::::::::::::"

iiiiiii i!ii!!!iiiiiii :,:,:,:,:,:,;,:°:,:,:°:,:,:,:,:,:°:,:,,

~ i:'.i~i!i!iiiii~ii~i~ii~i~i:..
~!~!~!~!i
+ +

~x?-~~' +' 3~

Fig. 8. Simplified tectonic setting of the Bacan region during the Late Miocene. Westward subduction of the Molucca
Sea plate under the Philippine Sea plate started at c. 15 Ma, probably from the south and moved northward.
Volcanicity started in Bacan at c. 7.5 Ma. Shallow marine sequences and fringing reefs formed around volcanic
centres.
494 J.F.A. MALAIHOLLO ~; R. HALL

Fig. 9. Simplified tectonic setting of the Bacan region during the Late Pliocene-Quaternary. The Sorong fault system
transform boundary cut through the Bacan region, contributing to the cessation of Miocene-Pliocene volcanism.
Quaternary volcanism erupted through continental crust along strands of the Sorong fault. Faults controlled coast
lines and basins developed between strands of the fault.

and younger connections are indicated by the
similarity of Eocene-Oligocene arc sequences
which also record a similar palaeomagnetic
history (Hall et al. 1995a).
(2) The Bacan and Tawali Formations are most
likely to be arc-related sequences produced as
a result of subduction of oceanic crust of the
Australian plate, probably directly linked to the
Indian Ocean, under the Philippine Sea plate.
(3) The Sibela Continental Suite arrived in the
Bacan region after collision of Australia or a
rifted fragment of Australia with the Philippine
Sea plate.
(4) The collision of Australia with the Philippine
Sea plate can be dated at c. 22 Ma in the Bacan
region.
(5) Younger Neogene-Recent movements on
strands of the Sorong fault have modified the
geology of the region and permitted emplace-
ment of younger volcanic rocks.
Several different models for the tectonic develop-
ment of NE Indonesia have been proposed, e.g. the
indentor model of Charlton (1986); the terrane
model of Silver et al. (1985); and the marginal
basin model of Ben-Avraham (1978), and all were
based upon limited geological data. The Ben-
Avraham model considers Bacan-Halmahera as
part of the Australian continent, which is wrong
as almost all of the region has been part of
the Philippine Sea plate since Cretaceous. Both the
Charlton and Silver et al. models imply that the
Sibela continental block was translated by post-
Miocene strike-slip motion to the Bacan region,
which has been disputed by results of this study. Ali
& Hall (1995) and Hall (1995) offer alternative
interpretations of the region which incorporate the
results of this and other recent geological and geo-
physical studies.
In a wider context, if the terrane definition of
Howell et al. (1985) is followed, the Sibela
Continental Suite, the Sibela ophiolite, the Bacan,
Tawali and Kaputusan Formations could be con-
sidered as five different tectono-stratigraphic
terranes. This may lead to interpretations such as
those of Struckmeyer et al. 1993, who apparently
assign separate histories to each of these 'terranes'.
Karig et al. (1986) argued that rocks in the north
Philippines, which have a similar character to the
Sibela Continental Suite, Sibela ophiolite, Bacan
and Kaputusan Formations, were separate allo-
chthonous terranes juxtaposed by strike-slip fault-
ing. Whilst not arguing against their interpretation,
this study has shown that the juxtaposition of rocks
of different character does not necessitate the
notion of allochthonous terranes. In the Bacan
region, the only fragment that may genuinely be an
allochthonous terrane is represented by the Sibela
Continental Suite. For the most part, the region has
always been at the edge of the Philippine Sea
plate, and different 'tectono-stratigraphic terranes'
reflect different tectonic regimes at the edge of the
plate.
The use of stratigraphical similarities is often
used in the literature to correlate tectono-strati-
graphic terranes and ultimately tectonic history
(e.g. Hamilton 1979; Pigram & Panggabean 1984).
Using this technique, the Bacan region could be
 GEOLOGY • TECTONICS OF BACAN, E INDONESIA
correlated with parts of Papua N e w G u i n e a
(Oligocene pillow lavas against continental base-
ment; Dow 1977), the Zamboanga Peninsula of
Mindanao (high grade continental rocks juxtaposed
against meta-ophiolite unconformably overlain by
Middle Miocene volcanic rocks; Rangin 1991) and
north Philippines (see above; Karig et al. 1986).
This type of correlation may lead to erroneous
interpretations since each of these regions has
different tectonic affinities (Bacan was and is still
part of the Philippine Sea plate; North Papua New
Guinea was part of the Philippine Sea plate, but
now is part of the Australian plate; Z a m b o a n g a and
north Philippines were and are still part of the
Eurasian plate). Similarities in the stratigraphy are
consequences of similar tectonic histories (areas
recording the collision of continental with oceanic
plate), and one should be cautious in correlation
and interpretation of m o v e m e n t of terranes along
strike-slip fault. All too often tectonic reconstruc-
References
ALl, J. R. & HALL, R. 1995. Evolution of the boundary
between the Philippine Sea Hate and Australia:
palaeomagnetic evidence from eastern Indonesia.
Tectonophysics, 251, 251-275.
ATMAWINATA, S., RATMAN, N. & PIETERS, P. E. 1989.
Geological map of the Yapen Sheet, lrian Jaya.
Geological Research and Development Centre,
Bandung, Indonesia.
BAKER, S. J. & MALAIHOLLO,J. E A. 1996. Dating of
Neogene igneous rocks in the Halmahera region:
arc initiation and development. This volume.
BALLANTYNE, P. D. 1991. Petrological constraints upon
the provenance and genesis of the East Halmahera
Ophiolite. Journal of Southeast Asian Earth
Sciences, 6, 259-269.
- - 1992. Petrology and geochemistry of the plutonic
rocks of Halmahera ophiolite, eastern Indonesia, an
analogue of modern oceanic forearcs. In: PARSON,
L. M., MURTON, B. J. & BROWNING, E (eds)
Ophiolites and their Modern Oceanic Analogues.
Geological Society, London, Special Publication,
60, 179-202.
- - & H A L L , R. 1990. The Petrology of the Halmahera
Ophiolite, Indonesia: an early Tertiary forearc. In:
MALPAS, J., MOORES, E. M., PANAYIOTOU,A. &
XENOPHONTOS,C. (eds) Ophiolites: Oceanic Crustal
Analogue. Proceedings of the Symposium Troodos
1987. The Geological Survey Department: Ministry
of Agriculture and Natural Resources, Nicosia,
Cyprus, 461-476.
BEN-AVRAHAM,Z. 1978. The evolution of marginal basins
and adjacent shelves in east and Southeast Asia.
Tectonophysics, 45, 269-288.
BROUWER, H. A. 1923. Bijdrage tot de geologie van het
eiland Batjan. Jaarboek Mijnwesen Nederlandsch
Oost-lndie 1921, 50, 73-105.
CHARLTON, T. R. 1986. A plate tectonic model of the
eastern Indonesia collision zone. Nature, 319,
394-396.
495
tion of a region is based on meagre geological data,
such as in the Bacan region, resulting in a diversity
of interpretations. Although this study has provided
the most comprehensive geological dataset from
the Bacan region, the history of the region in
the Early M i o c e n e and before the E o c e n e is
still unclear. Bearing this in mind, one should be
cautious in interpreting the tectonic evolution of
other, similarly complex, older regions.
This work was supported by NERC award GR3/7149,
grants from the Royal Society and the University of
London SE Asia Geological Research Group, and
financial assistance from Amoco Production Company.
We thank S. J. Baker, P. D. Ballantyne, E T. Banner, T. R.
Charlton, E. M. Finch, G. J. Nichols and S. J. Roberts for
their contributions and discussion, and C. C. Rundle and
D. C. Rex for guidance and help with isotopic dating.
Logistical assistance was provided by GRDC, Bandung
and the Director, R. Sukamto with field support by D. A.
Agustiyanto, A. Haryono, and S. Pandjaitan.
--, HALL, R. & PARTOYO, E. 1991. The geology
and tectonic evolution of Waigeo Island, NE
Indonesia. Journal of SE Asian Earth Sciences, 6,
289-297.
CLAGUE, D. A. • JARRARD, R. I. 1973. Tertiary Pacific
plate motion deduced from Hawaiian-Emperor
chain. Geological Society of America Bulletin, 84,
1135-1154.
Dow, D. B. 1977. Geological synthesis of Papua New
Guinea. Bureau of Mineral Resources Australia
Bulletin, 201.
ELLAM, R. M., MENZIES, M. A., HAWKESWORTH,C. J.,
LEEMAN, W. P., RosI, M. & SERRI, G. 1988. The
transition from calc-alkaline to potassic orogenic
magmatism in the Aeolian Islands, Southern Italy.
Bulletin of Volcanology, 50, 386-398.
HAKIM, A. S. & HALL, R. 1991. Tertiary Volcanic rocks
from the Halmahera Arc, Eastern Indonesia.
Journal of SE Asian Earth Sciences, 6, 271-287.
HALL, R. 1996. Reconstructing Cenozoic SE Asia. This
volume.
-- & NICHOLS, G. J. 1990. Terrane Amalgamation in
the Philippine Sea Margin. Tectonophysics, 181,
207-222.
- - , ALI, J. R., ANDERSON,C. D. & BAKER, S. J. 1995a.
Origin and motion history of the Philippine Sea
Plate. Tectonophysics, 251, 229-250.
--, AUDLEY-CHARLES,M. G., BANNER,F. T., HIDAYAT,
S. & TOBING, S. L. 1988. The basement rocks of
the Halmahera region, east Indonesia: a Late
Cretaceous-Early Tertiary forearc. Journal of the
Geological Society, London, 145, 65-84.
, FULLER,M., ALl, J. R. & ANDERSON,C. D. 1995b.
The Philippine Sea Plate: Magnetism and
Reconstructions. In: TAYLOR,B. & NATLAND,J. H.
(eds) Active Margins and Marginal Basins: A
Synthesis of Western Pacific Drilling Results.
American Geophysical Union Monograph, 88,
371-404.
496 J . F . A . MALAIHOLLO & R. HALL
--, NICHOLS, G. J., BALLANTYNE,P. D., CHARLTON,T.
& ALl, J. 1991. The character and significance of
basement rocks of the southern Molucca Sea region.
Journal of SE Asian Earth Sciences, 6, 249-258.
HAMILTON, W. 1979. Tectonics of the Indonesian Region.
US Geological Survey Professional Paper, 1078.
HARLAND, W. B., ARMSTRONG,R,. L., COX, A. V., CRAIG,
L. E., SMITH, A. G. & SMITH, D. G. 1990. A
geologic time scale 1989. Cambridge University
Press.
HOWELL, D. G., JONES, D. L. & SCHERMER,E. R. 1985.
Tectonostratigraphic Terranes of the Circum-Pacific
Region. In: HOWELL, D. G. (ed.) Tectonostrati-
graphic Terranes of the Circum-Pacific Region.
Circum-Pacific Council for Energy and Mineral
Resources Earth Science Series, 1, 3-30.
JAQUES, A. L. & ROBINSON, G. P. 1977. The continent/
island-arc collision in northern Papua New Guinea.
BMR Journal of Australian Geology & Geophysics,
2, 289-303.
KARIG, D. E. 1975. Basin genesis in the Philippine Sea.
Initial Reports of the Deep Sea Drilling Program,
31, 857-879.
, SAREWITZ, D. R. & HAECK, G. D. 1986. Role of
strike-slip faulting in the evolution of allochtonous
terranes in the Philippines. Geology, 14, 852-855.
LANPHERE, M. A. & DALRYMPLE, G. B. 1976. Identi-
fication of excess 4°Ar by the 4°Ar/39Ar age
spectrum technique. Earth and Planetary Science
Letters, 32, 141-148.
LfNTHOUT, K., HELMERS, H. & ANDRIESSEN, P. A. 1991.
Dextral strike-slip in Central Seram and 3-4. 5 Ma
Rb/Sr ages in pre-Tertiary metamorphic related
to Early Pliocene counterclockwise rotation of the
Buru-Seram microplate (E. Indonesia). Journal of
SE Asian Earth Sciences, 6, 477-493.
MALAIHOLLO,J. F. A. 1993. The Geology and Tectonics of
the Bacan Region, Eastern Indonesia. PhD thesis,
University of London.
MITCHELL, A. H. G., HERNANDEZ,E & DELA CRUZ, A. P.
1986. Cenozoic evolution of the Philippine
Archipelago. Journal of SE Asian Earth Sciences,
l, 3-22.
NICHOLS, G. J. & HALL, R. 1991. Basin formation and
Neogene sedimentation in a back arc setting,
Halmahera, eastern Indonesia. Marine and
Petroleum Geology, 8, 50-61.
PIETERS, P. E., HAKIM, A. S. & ATMAWlNATA, S. 1982.
Preliminary Geological Map of the Ransiki
Quadrangle, lrian Jaya. Geological Research and
Development Centre, Bandung, Indonesia.
, HARTONO,U. & AMRI, CH. 1989. Geological Map of
the Mar Sheet, lrian Jaya. Geological Research and
Development Centre, Bandung, Indonesia.
.... , PIGRAM, C. J., TRAm, D. S., Dow, D. B., RATMAN,N.
& SUKAMTO, R. 1983. The stratigraphy of western
Irian Jaya. Geological Research and Development
Centre Bulletin, 8, 14-48.
PIGRAM, C. J. & DAVIES, H. L. 1987. Terranes and the
accretion history of the New Guinea orogen. BMR
Journal of Australian Geology & Geophysics, 10,
193-211.
- - & PANGGABEAN,H. 1984. Rifting of the northern
margin of the Australian Continent and the origin of
some microcontinents in Eastern Indonesia.
Tectonophysics, 107, 331-353.
- - & SYMONDS, P. A. 1991. A review of the timing
of the major tectonic events in the New Guinea
Orogen. Journal of SE Asian Earth Sciences, 6,
307-318.
--, DAVIES, P. J., FREARY, D. A., SYMONDS, P. A. &
CHAPRONIERE, G. C. U. 1990. Controls on Tertiary
Carbonate Platform evolution in the Papuan Basin:
new play concepts. In: CARMAN, G. J. & Z. (eds).
Petroleum Exploration in Papua New Guinea.
Proceedings of the First PNG Petroleum
Convention, 185-195.
RANGIN, C. 1991. The Philippine Mobile Belt: a complex
plate boundary. Journal of SE Asian Earth Sciences,
6, 209-220.
--, JOLIVET,L. & PUBELLIER,M. 1990. A simple model
for the tectonic evolution of southeast Asia and
Indonesia region for the past 43 m.y. Bulletin de la
Socidtd ggologique de France, 8 VI, 889-905.
RATMAN, N. & ROBINSON,G. P. 1981. Geological Map of
the Manokwari Quadrangle, lrian Jaya. Geological
Research and Development Centre, Bandung,
Indonesia.
SANYOTO, P., PIETERS, P. E., AMRI, CH., SIMANDJUNTAK,
W. & SUPRIATNA,S. 1985. Preliminary Geological
Map of Parts of the Sorong, Kasim, west Waigeo
and Misool Quadrangles, lrian Jaya. Geological
Research and Development Centre, Bandung,
Indonesia.
SENO, T. & MARUYAMA, S. 1984. Paleogeographic
reconstruction and origin of the Philippine Sea.
Tectonophysics, 102, 53-84.
SILITONGA, P. H., PUDJOWALUJO,H. & MOLLAT, H. 1981.
Geological reconnaissance and mineral prospecting
on Bacan island (Moluccas, Indonesia). In: BARBER,
A. J. & WmYOSUYONO, S. (eds) The Geology and
Tectonics of Eastern Indonesia. Geological
Research and Development Centre, Bandung,
Indonesia, Special Publication, 2, 373-81.
SILVER, E. A., GILL, J. B., SCHWARTZ,H., PRASETYO,H. &
DUNCAN, R. A. 1985. Evidence for a submerged and
displaced continental borderland, north Banda Sea,
Indonesia. Geology, 13, 687-691.
STRUCKMEYER,H. I. M, YEUNG, M. & PIGRAM,C. J. 1993.
Mesozoic to Cainozoic plate tectonic and palaeo-
geographic evolution of the New Guinea Region.
in: CARMAN, G. J. & CARMAN, Z. (eds) Petroleum
Exploration and Development in Papua New
Guinea. Proceedings of the Second PNG Petroleum
Convention, 261-290.
SUKAMTO, R., APANDI, T., SUPR1ATNA, S. & YASIN, A.
1981. The geology and tectonics of Halmahera
Island and surrounding areas. In: BARBER, A. J. &
WIRYOSUYONO,S. (eds.) The Geology and Tectonics
of Eastern Indonesia. Geological Research and
Development Centre, Bandung, Indonesia, Special
Publication, 2, 349-62.
SUTTER, J. F. & SNEE, L. W. 1980. K/Ar and 4°Ar/39Ar
Dating of Basaltic Rocks from Deep Sea Drilling
Project Leg 59. Initial Reports of the Deep Sea
Drilling Project, 59, 729-734.
 GEOLOGY • TECTONICS OF BACAN, E INDONESIA 497

UYEDA, S. & BEN-AVRAHAM, Z. 1972. Origin and
development of the Philippine Sea. Nature Physical
Sciences, 240, 176-178.
VAN BEMMELEN, R. W. 1949. The Geology of Indonesia.
Volume Ia, General Geology. Government Printing
Office, The Hague.
VERSTAPPEN, H. TH. 1960. Geomorphological observa-
tions on the north Moluccan - northern Vogelkop
Island Arcs. Nova Guinea Geology, 3, 11-38.
VROON, P. Z. 1992. Subduction of Continental Material in
the Banda Arc, Eastern Indonesia. PhD thesis,
University of Utrecht.
YASIN, A. 1980. Geologic Map of the Bacan Quadrangle,
North Maluku. Geological Research and Develop-
ment Centre, Bandung, Indonesia.
 Dating of Neogene igneous rocks in the Halmahera region:
arc initiation and development
SIMON BAKER & JEFFREY MALAIHOLLO
SE Asia Research Group, Department of Geological Sciences,
University College, London WC1E 6BT, UK

Abstract: Potassium-argon ages of Neogene to Recent igneous rocks from the Halmahera region
record a history of intra-oceanic arc development since the late Middle Miocene following an
earlier phase of collisional plutonism. Arc formation from the Middle Miocene onwards was
due to the east-directed subduction of the Molucca Sea plate beneath the Philippine Sea plate as
it arrived at the Eurasian margin. The distribution of ages within the Neogene arc indicates
a northward migration of volcanic activity during the Late Miocene to Pliocene. Results of the
dating work show that after collision with the Australian margin at c. 22 Ma there was a period
of volcanic quiescence and limestone deposition before a new arc formed. This arc began
erupting at around 11 Ma on Obi as a result of subduction of the Molucca Sea plate. Initiation
of subduction is thought to have occurred around 15-17 Ma and may have been responsible for
disturbing potassium-argon ages of pre-Neogene rocks. Dates from fresh rocks show that the
volcanic front migrated northwards through Bacan and Halmahera throughout the Late Miocene
to Early Pliocene. Limestone deposition was curtailed as arc activity migrated north while
volcanism died out from the south. No Neogene volcanism younger than 8 Ma is observed in the
Obi area while on Bacan subduction-related volcanism ceased at c. 2 Ma. Late Pliocene crustal
deformation caused a 30-40 km westward shift of the volcanic front. Quaternary volcanic rocks
exposed in Bacan and the extreme south of Halmahera are not direct products of subduction but,
rather, display geochemical characteristics of both subduction and fault-related magmatism.
These volcanic rocks are distributed along splays of the Sorong fault system. The formation and
propagation of the Halmahera arc is a consequence of the clockwise rotation of the Philippine Sea
plate as the southern edge moved across the northern Australian margin and impinged on the east
Eurasian margin. The ages of initiation of volcanism and subduction track the developing plate
boundary as subduction propagated northwards.

The principal islands of the Halmahera group
(Halmahera, Bacan and Obi) lie in northeastern
Indonesia in the province of Maluku, straddling the
equator between 127°E and 129°E. These islands
lie at the junction of three major plates (the
Philippine Sea plate, the Australian plate and the
Eurasian plate) where the A l p i n e - H i m a l a y a n and
the C i r c u m - P a c i f i c orogenic belts meet. The
present-day tectonics are therefore complex (Fig.
1). These belts meet in the region of the Sorong
fault system which is responsible for transferring
crustal fragments of Philippine Sea and Australian
origin into the complex of island arcs and small
ocean basins that forms the Eurasian margin.
Present-day tectonics are the result of the north-
ward m o v e m e n t of continental Australia (Australia
and New Guinea) into the Pacific region throughout
the Tertiary. Clockwise rotation of the Philippine
Sea plate since c. 25 Ma led to the development
of the Sorong Fault Zone as it collided with the
northern Australian margin (Hall et al. 1995a, b).
Current motion between these two plates is taken
up by sinistral m o v e m e n t on the Sorong fault
system in northern New Guinea.
The currently active H a l m a h e r a arc, at the
eastern edge of the Molucca Sea (Hamilton 1979;
Moore & Silver 1983), formed as a consequence
of eastward subduction of the Molucca Sea plate
beneath the Philippine Sea plate as the latter
rotated clockwise. Geophysical evidence from
earthquake data indicates a seismic zone dipping
at c. 45 ° to the east to depths of about 200 k m
(Cardwell et al. 1980). To the south the subduction
zone appears to be terminated by a strand of the
Sorong fault system just north of the eastern tip of
Mangole (Sula Platform). Opposing the Halmahera
arc is the Sangihe-north Sulawesi arc approxi-
mately 250 k m to the west which is the product
of westward subduction o f the Molucca Sea plate;
beneath the Sangihe arc the slab dips at 55-65 ° and
reaches a depth of 600 k m (Cardwell et al. 1980).
The oldest rocks known from the Sangihe arc are
of early Middle Miocene age and particularly
voluminous arc activity occurred between 5 and
14 Ma (Hamilton 1979).
The present Halmahera arc lies north of the
equator and is built upon a Neogene arc that
extended from Obi northwards through Bacan to

From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, 499
Geological Society Special Publication No. 106, pp. 499-509.
500 S. BAKER & J. MALAIHOLLO

I[ " .+.~. " ~+.! 7Z'l/'TZ'

,/\ ,... / ;

Euraeian -,- MOROTAI

Margnn A
tN
+

~_. • . +
• +
.

t
• ~
. +

~_."
•

¥. # ;,- # O 0 N.E .~ I A
. . . . . . . . . O

,'-',',-',',
HALMAHEF, A.-I -' ;'."
Phnlnnnnne -',',-
+/:# ~ ~= % / <
r r .: . ' - -
Sea Rate-; :'\-;
~ ~#" Molucca \-~ "-; -~-~-; :-i :w~,J~o.o°:_
- Sea
Collision
Complex

-, "- z - -. "- " ~ - \ \ ~.÷~'~"~9~~b~C~UINP_A

~ustra
-1,50 km

Fig. 1. Locationmap showingthe principaltectonic elements and islands of the study area.

north Halmahera (Hall et al. 1988b; Hall & Nichols
1990). This paper describes the Neogene to
Pliocene development of this arc based on
potassium-argon and petrographic studies of the
volcanic rocks. Other Neogene rocks dated by the
potassium-argon method are described and placed
in a plate tectonic context.
Stratigraphy
The Tertiary stratigraphy of the region is summa-
rized in Fig. 2. The pre-Neogene geology of the
region consists of a basement of ophiolitic and arc-
related rocks (Hall et al. 1988a; 1991; Ballantyne
1990) overlain by and imbricated with, arc
volcanic, plutonic and sedimentary rocks of late
Mesozoic to Eocene age (Hall et al. 1988b; Hakim
1989; Ballantyne 1990; Hakim & Hall 1991).
Following an Eocene unconformity (possibly
linked with the Pacific plate reorganization at this
time) a Late Eocene arc developed with associated
basaltic pillow lavas and volcaniclastic turbidites
(Hall et al. 1991; Malaihollo 1993); this arc ceased
activity in the earliest Miocene and its termination
is marked by a regional unconformity. Middle
Miocene limestones were deposited throughout
the region during a period of stability before the
development of the Neogene arc and associated
sedimentary rocks. Deformation in the Pliocene
is marked by another unconformity. Continental
metamorphic rocks representing Australian crust
and possibly of Palaeozoic or younger age
(Hamilton 1979; Malaihollo 1993) are juxtaposed
against ophiolitic rocks in the southern part of the
region. Detailed stratigraphic descriptions of pre-
Neogene and Neogene rocks are given in Hall
et al. (1988a, b, 1991), Hakim & Hall (1991) and
Roberts (1993).
Analytical methods and data presentation
Samples were selected for potassium-argon dating
primarily on the basis of their freshness and the
presence of suitable potassium-bearing phases.
Mineral separates were used; amphiboles and micas
were the preferred minerals. Whole rock analyses
on fresh volcanic rocks provided good quality
results (confirmed by biostratigraphic dating in
many areas) with small errors. Biostratigraphic
dating gives confidence that young, fresh rocks
with a potassium rich, glassy residuum provide
reliable isotopic ages. Sample locations are given
 ARC DEVELOPMENT IN THE HALMAHERA REGION
HALMAHERA
(& KAYOA)
vvvvvvvvvvvvvvvv
i¢.aya(~a
Formal;ion
Wcd~ Group
(rn=intyW~e~) KaFul;u~a~
Format..ion
For'm=l:ion
(E=n~)
~ubaim Lim¢~t,onc
Formation (Eal~t.) Rut,a
LIm~B1;one
Formal;ion
Tawali
Forma¢ion
Tawali Forma'cion
Oha Volcanic. Bacan
Formal;Ion (lh'¢~l;) Format.ion
Wayamli
Formal;Ion
(Ea~.)
Fig. 2. A simplified Tertiary stratigraphy for the three principal islands of the region studied.
in Fig. 3 whilst analytical results are presented in
Table 1.
Samples for potassium-argon analysis were
crushed and sieved to yield a fraction between
300 l.tm or 250 l.tm and 125 ~m (if the sample was
small the sieve fraction 500-125 pm was used);
this was then washed in deionized water. The
washed fraction was either used directly for
potassium-argon analysis or further processed to
produce mineral separates by a combination of
heavy liquid and magnetic separation techniques.
The samples were then split for separate potassium
and argon analyses. All potassium-argon analyses
were carried out at the NERC Isotope Geoscience
Laboratory (NIGL) at the BGS site in Keyworth,
Nottinghamshire. Potassium analyses were con-
ducted in duplicate using flame photometry against
standard potassium solutions and a pure water
blank after sample digestion in excess hydrofluoric
and perchloric acids. Argon analyses on a separate
aliquot were performed by complete sample fusion
in a gas extraction line; the sample gases were
mixed with an enriched 38Ar spike (isotope dilution
method) and cleaned up using a titanium getter and
a liquid nitrogen cold trap. The extraction line was
coupled to a gas source Micromass 1200 mass
spectrometer with a magnetic deflection of 60 ° and
sensitive to 3 x 104 Amps/Torr; calculations and a
direct printout of the results were provided by VG
software using the constants recommended by
Steiger & J~tger (1977). The mass spectrometer
and spike volume were regularly calibrated by
comparison with the International Standard G1-O
glauconite (reproducibility results available from
the authors on request).
501
BACAN OBI Ma
c;. ~ . ~ . ~.,c:~..
0
Form,l;,o° ~ ~ d 2
Quaternary
q .L Vo[can'mla~t.ic~' z- "1 Format.ion ~ ~
Pliocene
,4 Ar~ V.olc,,~nic~ ,4 . / " I
Woi 6, Gu~t.l ]< , volc~nlcla~r.lc~ mI 5
Forrnat.ione
Miocene
Fluk ~ LIm¢~'f.on~e
Fo~,l;~o~ ~ 23
Oligocene
F....... ~ 35
Air I~,,a~i !lIHI)'//.l'it~-~l~ !...... ,!lh
• ~ri~ ~J I J E~SM8
? ? ?
56
I~leocene
~5
Error calculation and data reduction
methods
All isotopic ages are quoted with errors at the 95%
confidence level (2o). The major factor affecting
the error is connected with the correction for atmos-
pheric argon. If the atmospheric component
(40Aratm) is small, uncertainties in its measurement
will be minor compared to the total 4°At. However,
if 4°Aratm is the major component then the same
degree of uncertainty will cause large errors in
the smaller radiogenic component (40Arrad). For
samples with low atmospheric contamination a
minimum error of 1% is assumed based on the
uncertainty in the calculation of the spike volume.
Potassium errors were calculated from the average
of at least two analyses and 10 error calculated
from the deviation of the average from the actual
values obtained, l o error in % was used for age
calculation and then converted to 2o error for
reporting purposes (see Table 1). For some samples
K error is relatively high. This is due to a number
of factors namely: matrix effects affecting the flame
photometer reading for some samples; undigested
sample (opaques) and sample heterogeneity
exacerbated by the small size of aliquots that must
be used for samples with high K content. All efforts
have been made to minimize these problems
including triplicate analyses if necessary.
For some very young samples 40Arrad is below
the limits of detection for the spectrometer so the
VG software at NIGL allows an maximum age
calculation for such samples by assuming there is
less than or equal to 1% radiogenic argon present
(the limit of precision for argon analysis). These
502 s. BAKER & J. MALAIHOLLO
Kahatola
HB16,
A HB93
HB94
HB95
Ternate~aAt147fA
mid0re~lklH2H~m!260
M~aki~~H ~ results from volcanic rocks are plotted in geo-
BACAN ~ ~
HR209
0R303
~°~isa
~ ~D42~OJD140
BI ,~J58
Fig. 3. Sample location map. Black triangles represent
currently active arc volcanoes.
ages are of limited use and in some cases the use
of a larger sample for argon analysis (or of a more
K-rich fraction) yields sufficient radiogenic argon
for analysis.
Isochron plots provided weighted averages for
suites of ages in some cases. Two types of isochron
plots were used, 4°Ar vs. 4°K and 4°Ar/36Ar vs.
4°K/36Ar, and ages determined from the slope of
the line. The intercept of the line with the y-axis
qualitatively indicates loss/gain of argon and initial
4°Ar/36Ar ratio respectively. Where potassium-
argon ages from a single formation are within error
of one another a site or formation mean is quoted
along with the standard error on the mean (Table 1)
and in these cases this is accepted as the best
estimate of the age.
Potassium-argon data were also tested statistic-
ally to try to identify periods of volcanism and
probable gaps in the spectrum of Neogene ages;
iterative cluster analysis of the data and their
associated uncertainties was used. First, a density
trace was obtained by computing the fraction of
data points that fall into a certain interval or
window (0.7 Ma was chosen in this case) using a
distance weighted kernel estimator (to assess the
contribution of each data point in the current
window) and assuming a Gaussian distribution for
the errors about the data points. Following this
the probability of gaps in the suite of ages was
calculated by sampling one point from equally
spaced locations in the range 0-12 Ma (including
points from uncertainty ranges) to produce a trace
of 'gap ages' (see below).
All age data discussed here are presented in
Table 1; sample localities are shown in Fig. 3;
graphical order in Fig. 4. Ages are described by
area, working from south to north. The time scale
of Harland et al. (1990) is used.
Pre-arc Neogene ages
Bacan
The oldest Neogene rocks from the region are
quartz-monzodiorites of the Nusa Babi intrusions
from Bacan. These plugs are distributed throughout
eastern Bacan and petrographically are charac-
terized by plagioclase with lesser orthoclase, green
amphibole, biotite, quartz and opaques; some
samples show chlorite and/or epidote alteration.
Trace and major element tectonic discrimination
diagrams suggest either an arc or a post-collisional
magmatic origin (Malaihollo 1993). Dating of
biotite from the type locality (BR80) gave an age
of 19.8 _+ 1.6 Ma interpreted to be a crystallization
age which is supported by the field observation that
these intrusions cut only rocks older than Middle
Miocene. Samples from other localities show the
effects of later alteration and yield younger
potassium-argon ages.
Rocks belonging to the Saleh intrusions from
Pulau Saleh Kecil (SR6) and south Bacan (B66) are
amphibole-rich diorites and chemically distinct
from the Nusa Babi intrusions. Petrographic studies
indicate an order of crystallization typical of arc
rocks; tectonic discrimination diagrams are
ambiguous, suggesting either a post-collisional or
arc origin. These samples yield Early to Middle
Miocene ages (9.3_ 1.5 Ma and 15.1 _+ 1.6 Ma
respectively) on amphibole separates interpreted
as crystallization ages.
Halmahera
In NE Halmahera a minor syeno-gabbro sill (c. 2 m
thick, lateral extent probably less than 100 m) was
Table 1. K-Ar Dates from Halmahera, Obi and Bacan

Formation Sample Location Rock type F Material Grain size % K Wt for Ar 4°Ar atm (hi g q ) 4°At* (nl g-t) Age Ma
(for localities see Fig. 3) (~tm) (2(y error) (g) (%) (1~ error) (2(~ error)

Halmahera
Quaternary HR209 SW Halmahera Bi tuff A Bi 250-125 3.79 ± 0,08 0.1484 98.76 <LOD <0.4

Kayasa H260 Central Halmahera 2px A n d (f) B Wr 500-125 0.91 ± 0,02 1.0018 76.42 0.076 _+ 373 2.1 ± 0.2
HB16 N W Halmahera Welded tuff A Wr 250-125 2.68 _+ 0,05 2.0039 93.26 0,228 ± 15.13 2.2 ± 0.7
H258 Central Halmahera Hb A n d (f) A Hb 500-125 0.94 ± 0,02 1.3406 91.18 0.085 ± 10,57 2.3 ± 0.5
HR717 N W Halmahera Mdiorite O Wr 250-125 0.94 ± 0.02 0.519 84.34 0.109 ± 5.98 3.0 ± 0.4

Weda Group HR705 Central Halmahera 2px A n d A Wr 500-125 1,34 ± 0,13 1.1727 55.38 0.220 ± 1.67 4.2 ± 0.5
HR703 Central Halmahera Hb px A n d A Wr 500-125 1.12 _+ 0,02 1.3133 79.97 0.203 ± 4,16 4.7 ± 0.4
HA147A Central Halmahera 2px A n d (f) A Wr 500-125 1.60 _+ 0.11 1.3356 61.31 0.301 ± 1,94 4.9 _+ 0.4
H256 Central Halmahera Hb A n d (f) B Hb 500-125 0.30 _+ 0.01 1.5009 82.84 0.065 ± 5,09 5.7 ± 0.6
KT5b dp Kayoa Island Px A n d B Wr 300-125 1.07 ± 0.02 1.1572 86.42 0.239 ± 6,54 5.8 _+ 0.8
H275 Central Halmahera Hb A n d (f) A Wr 500-125 1.12 ± 0.03 1.2911 83.11 0.273 ± 5.10 6.3 ± 0.7
HR711 Central Halmahera Hb A n d A Hb 500-125 0.66 ± 0.01 1.0282 86.14 0.196 ± 6.38 7.7 _+ 1.0

Alkaline Sills HB93 NE Halmahera S-gabbro A W 250-125 2.77 _+ 0.06 0.5295 72.58 1.297 + 2.88 12.0 ± 0.7
HB94 NE Halmahera S-gabbro A Bi 250-125 3.42 ± 0.07 0.4168 56.22 1.723 _+ 1,74 12.9 _+ 0.5
HB97 NE Halmahera S-gabbro A Hb 250-125 0.88 ± 0.02 0.4074 89.73 0.359 ± 8.93 10.5 ± 1.9
Site Mean for 11.8 ± 0.7
Alkaline Sills
Obi
Woi and Guyuti OJ58 Central Obi Px And (f) B Wr, 250-125 1.79 ± 0.04 0.733 75.1 0.744 ± 3.2l 10.7 ± 0.7
OJ40 dp NE Obi Bi px And A Bi, Px 250-125 0.83 ± 0.02 1.0418 65.33 0.347 _+ 2.16 10.7 ± 0.5
OD194 Central Obi And (f) A Wr 300-125 1,64 _+0,3 1.353 35.46 0.708 +_ 1.16 11.1 +__0.3
OD15 NE Obi Crystal tuff B Bi 250-125 4,76 ± 0.26 0,1403 67.48 2.100 ± 2.35 11.3 ± 1.3
OD42V N Obi Px And (f) B Wr 300-125 1.50 _+0.3 0.9353 42.39 0,669 _+ 1,27 11,5 + 0.5
Formation Mean for 11,1 _+ 0.2
Woi and Guyuti Fms
Bisa OR303 P. Bisa Hb A n d C Wr 500-125 1.63 + 0,05 1.0283 86.41 0.557 ± 6,63 8.8 ± 1.2
OR306dp P. Bisa Hb A n d C Wr 500-125 2,01 ± 0,03 1.232 78.96 0.748 ± 3,95 9.5 ± 0.8
Site Mean for 9.2 ± 0.4
Bisa Volcanic Rocks
Baean
Kaputusan BR82 dp N Bacan Hb A n d A Wr 125-250 1.36 _+ 0.04 1.0238 66.66 0.118 ± 2.42 2.2 ± 0.1
BP30 dp Nr. Labuha Hb A n d A Wr 250-500 1.09 ± 0.02 1.2044 87.57 0.107 _+ 7.23 2.5 _+ 0.4
BM131dp Ra R. cent. Bacan Hb And B Wr 250-425 1.54 _+ 0.09 2.4686 84.74 0.153 ± 5.74 2.5 ± 0.3
BM453dp P. Mamalaya Bi px A n d B Wr 250425 1.38 ± 0.04 2.2700 81.91 0.117 _ 4.69 2.7 _+ 0.2
BR203 SW Bacan Hb A n d B Hb, px 125-250 1.89 ± 0.09 1.0420 92.59 0.311 ± 3,13 4.2 ± 1.1
BM256 Bibinoi R. Welded tuff A Wr 250-500 1.63 _+ 0.03 0.3493 72.16 0.304 ± 2,98 4.8 ± 0.3
BM26 P. Mandioli S Bacan Hb A n d A Hb, px 125-250 1.10 _+0.06 0.6240 77.58 0.212 _+ 3.73 5.0 ± 0.5
BR174 S W Bacan Hb A n d A Hb 125-250 2.26 ± 0,05 1.0276 81.35 0.445 ± 4.57 5.1 ± 0.5
BM361 NE Bacan Px And A Wr 125-250 0.97 _+ 0,03 0.3237 86.48 0.207 ± 6,75 5.5 ± 0.8
BR16 tp NE coast Bacan Px And A Wr 250-425 1.50 ± 0.16 1.0052 80.11 0.341 +_ 4.19 5,8 +_ 0.8
BR58 dp N coast Bacan Px And A Wr 250-425 1.16 _+ 0,01 2.2149 58.28 0.304 ± 1.78 6.7 ± 0.3
BR49 SE Bacan Hb And A Wr 250-500 2.10 ± 0.04 0.0584 78.64 0.612 ± 6.42 7.5 ± 1.0

Nusa Babi Intrusion BR80 dp E Nusa Babi Monzonite C Bi 125-250 4.29 _+ 0.09 0.1024 68.74 3.325 ± 3,87 19.8 ± 1.6

Saleh Diorite SR6 P. Saleh Kecil Diorite A Amp 125-250 0.36 ± 0.01 1.0611 89.20 0.132 _+ 8.49 9,3 ± 1.6
B66 Sibela Mtns. Diorite B Wr 125-250 0.65 -+ 0.01 2.110 75.85 0.131 ± 3,33 15.1 ± 1.6

Abbreviations. Sample: dp = duplicate; tp = triplicate. Rock type: A n d = andesite; Bas = basalt; Dac = dacite; Tuff = tuff; Bas and = basaltic andesite; Mdiorite = microdiorite; S-gabbro = syeno-gabbro; f = float,
c = clast from conglomerate or similar, if neither then in situ. Freshness F: A = fresh (no secondary minerals, glass fresh); B = mainly fresh (< 5% modal secondary minerals); C = slightly altered (5-10% modal secondary minerals);
O = opacitic alteration of mafic minerals (usually hornblende). Material: Wr = whole rock; Bi = biotite; Musc = muscovite; Hb = hornblende; Px = pyroxene. Others: L O D = limit of detection.
504 s. BAKER & J. M A L A I H O L L O

Ma
M.Miocerle NORTH 50UTH

~)
T [_ <v~v ~', I~i~a
vvvvvvv,
1
I I kA; . . . . . ' ...................... ,~v . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ,, . . . . vJz . . . . I VolGani6~ot;k~
............ ~ W j ~ ~ ~ , i . " ~ ~ ~ ~ ~ u g u e a n I;m. ',T,~:I~

................................. ~ . . . . . . . . . . . . . . I¢~ u'r,u e a n F ...............

. . . . . Q u a t e r n a r y Art; V'ol~ania~ . . . . V o l ~ a n i ~ acr, ivit,y a l o n o f a u l t , e l a y ~

o I;;;;; ............................. ;;;: P
rc~

Halmahera 15a6an Obi

Fig. 4. Primary potassium-argon ages (and their errors) for Neogene volcanic rocks plotted in geographical order
from north to south. Shaded areas illustrate periods of volcanism within the named formations derived from the
statistical treatment described in the text and shown graphically in Fig. 5. Note the younging trend from south to
north is particularly marked from Obi, via Bisa, to Bacan. Quaternary volcanic activity is also shown.

found intruding redeposited limestones of Eocene
age. The sill is fresh and has two distinct grain
sizes, coarse and medium. Analcite occurs as a
primary constituent of the groundmass together
with alkali and plagioclase feldspar; titanaugite,
sodic amphibole and biotite are all intimately inter-
grown along with titanomagnetite. Dating of three
samples (whole rock, biotite and amphibole) yield
similar ages which can be reduced to a site mean of
11.8 _ 0.7 Ma interpreted as primary crystallization
ages.
Arc-related Neogene
Obi
The Woi Formation in central and northwest Obi
consists predominantly of volcanic rocks (lava
breccias and flows with pyroclastic rocks) with
interbedded volcaniclastic sandstones and breccio-
conglomerates. Chemically the volcanic rocks are
sub-alkalic calc-alkaline arc-related rocks and the
majority are andesitic although there is a significant
spread into the basaltic andesite and dacite fields
(Miyashiro 1978; Cox et al. 1979). Petro-
graphically they are typically plagioclase- and
clinopyroxene-phyric, minor phenocryst phases are
biotite and rarely amphibole (both commonly show
opacitic rims). Feldspar and clinopyroxene are
often associated as glomerocrysts. Phenocrysts are
set in a fine-grained groundmass of small clino-
pyroxenes, opaque grains and plagioclase micro-
lites showing felty, intersertal and, occasionally,
trachytic textures. Variable amounts of glass occur
in the groundmass, amygdales occur in some areas
and are infilled with opaline silica and a variety of
fibrous minerals. The lateral deep-water equivalent
of the Woi Formation is the Guyuti Formation
which consists of debris flow deposits containing
andesite and limestone clasts representing a deep
marine slope forearc environment (R. Hall &
D. Atmo, pets. comm., 1994).
ages Potassium-argon results on rocks from the Woi
and Guyuti Formations in central Obi are similar,
with small errors. The formation mean for Woi and
Guyuti Formation samples is 11.1 _+0.2 Ma. This
age is supported by nannofossil and foraminiferal
ages from sedimentary rocks within the Woi
Formation as well as from limestone clasts within
the Guyuti Formation (foraminiferal and nanno-
plankton dates indicate a late Middle to Late
Miocene age range).
Neogene volcanic rocks also occur just north
west of Obi island on Pulau Bisa. These are pre-
dominantly fine-grained aphyric volcanic rocks
(although small opacitic brown hornblendes may
form a minor phenocryst phase) and contain xeno-
liths of ultramafic material probably derived from
 ARC DEVELOPMENT IN THE HALMAHERAREGION 505

the Tapas ophiolite. Two samples from this island, 0.~50

OR303 and OR306, yield ages of 9.5_ 0.8 and 0.25
8.8_ 1.2 Ma respectively, giving a site mean of
0.20
9.2 ___0.4 Ma - notably younger than ages from
the Woi Formation. "~ o.18

I~ 0.10

Bacan 0.05

0.0
The Kaputusan Formation is widely distributed vvvvvvv ¸ vvvvvvvwvvv~v vvw vvv, ~vvvvvvv,

throughout Bacan and consists of volcaniclastic o ~ 2 ~

Age Ma
sedimentary rocks, lavas and associated pyroclastic
rocks deposited in a subaerial or shallow marine "/~5 8:.& 10.5 11.E

environment. The Goro-Goro Member is the e,,, O.20_

volcanic member of the Kaputusan Formation and
is divided petrographically into: two pyroxene 0,15-
andesites, hornblende andesites, hornblende
pyroxene andesites and hornblende biotite 0.10 i
andesites. All these rocks display typical andesitic
0.06.
features such as strongly zoned plagioclase
and mafic phenocrysts, glomerocrysts of early o.o

crystallising minerals, abundant opaques o t 4 5 ~ 7 b 9 io h

(magnetite, titanomagnetite), microcrystalline to AgeMa

glassy groundmasses and occasional trachytic

textures with the addition of the named phenocryst
phases. Chemically the volcanic rocks are typical
calc-alkaline arc rocks, predominantly basaltic
andesites to andesites with minor basalts and
dacites; some samples plot in the high-K to
shoshonitic fields. Their trace element ratios are
also typical of arc rocks (Malaihollo 1993).
Twelve andesite samples from the Goro-Goro
Member of the Kaputusan Formation yielded good
quality potassium-argon results. Biostratigraphic
dates (nannofossils) from sedimentary rocks of the
Pacitak Member of the Kaputusan Formation
indicate a Pliocene age ranging from N19 to N21.
Isotopic ages obtained from rocks in the Goro-Goro
Member fall between 7.5 Ma and 2.2 Ma but
within this range there appear to be two concen-
trations of ages. This apparent division may be an
artefact of incomplete sampling. The ages form a
broad continuum (Figs 4 & 5, Bacan area), but
there does appear to be a gap at 3.5 Ma. This gap
corresponds to an overall lack of volcanism at this
time in both Halmahera and Bacan. The two groups
of rocks show minor chemical differences, based on
trace element data (Malaihollo 1993), and a broad
difference in geographical distribution; the older
rocks are found in south Bacan.
Halmahera and Kayoa
Neogene volcanic rocks on Halmahera and Pulau
Kayoa belong to the Volcanic Member of the Weda
Group and the younger Kayasa Formation. These
rocks occur in thick sequences that show abundant
evidence of sub-aerial eruption from large volcanic
edifices in central and northwest Halmahera whilst
Fig. 5. Gaussian kernel smoothed density estimates for
observed Neogene ages (using a window of 0.7 Ma).
For description of the technique see text. (a) Densities
of ages at certain times suggests that volcanic activity
is pulsed rather than continuous over the lifetime of the
arc. (b) Calculated probability of gaps in the volcanic
record implying lack of volcanic activity at intervals
of high probability.
in the southwest arm volcanic rocks are found in
redeposited volcaniclastic sequences. These were
built on an early to mid Tertiary volcanic arc that
deposited the Oha Volcanic Formation on
Halmahera and the Bacan Formation on Bacan.
Petrographically the rocks are predominantly
porphyritic pyroxene and/or hornblende andesites
with strongly zoned plagioclase in a micro-
crystalline to glassy matrix; hornblendes frequently
show opacitic or highly included rims. Alteration is
rare and the vast majority of the samples are fresh.
Chemically the rocks are basaltic andesites to
dacites and are medium to high-K rocks of the
calc-alkaline series with typical arc volcanic REE
patterns (Hakim 1989; Hakim & Hall 1991) that are
directly comparable to volcanic rocks from the
Sangihe Arc (Morris et al. 1983; Morrice & Gill
1986). Samples analysed for potassium-argon were
collected from Pulau Kayoa, midway between
Bacan and Makian, and from central Halmahera;
most are float samples as they are invariably the
least weathered.
Potassium-argon ages for Weda Group volcanic
rocks on Halmahera and Kayoa range between
4.2-7 Ma whilst those from the Kayasa Formation
506 s. BAKER •
are restricted to around 2 to 3 Ma. All these
samples were included in the computation of the
density trace and 'gap ages' along with those from
Bacan. The results of this statistical test suggest one
fairly long period of continuous volcanism (7.8 to
4 M a ) a n d a second period from 3 to 1.8 Ma on
both Bacan and Halmahera.
Quaternary volcanic activity
Current volcanic activity in the region is restricted
to an area north of, and including, Makian Island
as far as Galela town in northwest Halmahera.
Evidence of Quaternary volcanism on Bacan is
witnessed by the presence of (now extinct) volcanic
edifices that retain much of their original form;
eruptive rocks from these are olivine and pyroxene
phyric basalts. These have not been dated but
chemically they show characteristics of both arc
and within-plate volcanism (Malaihollo 1993).
Off the extreme southern tip of Halmahera is
the small island of Pulau Doworalama; a biotite
separate from a biotite tuff collected here did not
yield sufficient radiogenic argon for analysis
(below the limit of detection) indicating that it is
very young. Calculation, assuming < 1% radio-
genic argon, yields an age of less than 0.5 Ma.
A number of other small islands lie off the south
of Halmahera but none were visited by the
University of London group. Morris et al. (1983)
visited the islands of Woka, Kekik and Pisang and
sampled the Quaternary volcanic rocks exposed
there. They found them to have some shoshonitic
characteristics with higher alkali contents than are
found in their 'oceanic segment' of the arc, with
dramatic enrichments of Ba and St.
Neogene reset ages from
pre-Neogene rocks
Potassium-argon data relating to thermal overprints
on older rocks are notoriously difficult to interpret
and meaningless 'mixed' ages can be the result.
The problems of incomplete resetting under the
localized low grade metamorphic and fluid-rich
conditions obtaining within island arcs are encoun-
tered in the Halmahera region. This is observed in
areas where rocks with disturbed potassium-argon
ages are found within a few metres of rocks with
unaffected, primary ages (also reflected in the
petrography of the samples). A number of demon-
strably pre-Neogene rocks (based on stratigraphic
and biostratigraphic evidence) yield Neogene ages,
probably as a result of enhanced heat flow during
initiation and development of the Neogene arc.
These reset ages appear to show a systematic
variation, generally younging from south to north.
J. MALAIHOLLO
Discussion
Post-collisional Neogene igneous activity
The oldest Neogene age from the region is
associated with intrusive rocks of the Nusa Babi
intrusions (BR80) whose age is 19.8 _+ 1.6 Ma. As
noted earlier, tectonic discrimination diagrams
suggest that the Nusa Babi Intrusive may have a
collisional or an arc magmatic origin. The regional
unconformity of Early Miocene age (c. 22 Ma) is
interpreted by Hall et al. (1995a) to be the result of
Australia-Philippine Sea plate collision. The Nusa
Babi Intrusive rocks post-date collision by several
million years and are therefore considered to be the
result of magmatic activity associated with heating
and stacking of the young orogenic belt in the
Bacan area (Malaihollo 1993). The Nusa Babi
monzodiorites are the most acidic plutons in the
region and this supports the possibility that
continental crust was involved in their formation.
The sills found in NE Halmahera are dated
c. 12 Ma;. The geographical position of these
rocks suggests they are due to isolated, shallow
magmatic activity in a region well behind the
developing arc front. Their La/Nb ratios are less
than 1 (0.86 for HB93 and 0.72 for HB97, un-
published ICP data) indicating a true alkaline
chemistry (M. E Thirlwall, pers. comm. 1995) that
strongly suggests they are unrelated to the early
arc activity occurring in Obi at this time.
Neogene arc initiation and development
Arc precursors: The two diorite samples dated
from Bacan (Saleh intrusions, SR6, B66) display
mineral chemical and geochemical characteristics
that indicate an arc origin. The interpretation of
their isotopic ages as primary crystallization ages
implies a period of plutonic activity preceding the
development of an eruptive arc in the Bacan region,
perhaps in response to subduction initiation (no
volcanic rocks of this age range have been found
in Bacan).
Age of arc inception: The earliest primary ages
from arc volcanic rocks are from samples collected
from the Woi and Guyuti Formations in central Obi.
None of these ages deviate by more than half a
million years from 11 Ma, indicating a short but
prolific eruption of andesites at this time. These
samples were collected from a wide range of geo-
graphical and stratigraphic locations and are
thought to be representative of the earliest volcanic
episode on Obi; no younger dates have yet been
obtained from this region. These rocks are
interpreted as the first effusive products resulting
from subduction of the Molucca Sea plate and
indicate that the slab must have reached depths
 ARC DEVELOPMENT IN THE HALMAHERAREGION
sufficient to have caused melting (100-110kin,
Tatsumi 1986).
Age variation within the arc
Ages obtained from acid lavas on Pulau Bisa, just
north of Obi, indicate volcanic activity between
9.5-8.8 Ma. The gap in ages between Obi and Bisa
volcanic rocks suggests either a hiatus in volcanic
activity between 11 and 9.5 Ma or erosion causing
removal of younger arc products on Obj.
Significant erosion of the arc is unlikely since ages
of volcanic clasts from Neogene sedimentary
formations on Obi (notably the Guyuti Formation)
are also within the age range of Woi Formation
volcanism. The rocks from Bisa may represent late
stage products of the waning arc in the Obi region
as arc activity moved northwards.
The oldest volcanic rocks found in Bacan are of
early Late Miocene age (7 to 8 Ma) and activity
appears to have been fairly continuous until the
Late Pliocene-Pleistocene (c. 2.5 Ma). Within this
time span there are greater densities of ages at 7-8,
c. 5, c. 2.5 and 0.5Ma. These groupings are
reflected in concordant chemical and fractionation
trends indicating each group evolved from a
separate magma chamber (thus representing
different volcanic centres). Within the Bacan group
of islands there is no clear northward younging
of ages but rather a vague suggestion of general
younging to the north and west. Neogene and
probable pre-Neogene rocks outcrop on Pulau
Kayoa (c. 30 km north of Bacan) and KT5B has an
age of c. 5.5 Ma.
Ages of Halmahera volcanic rocks lie in a
spectrum from c. 8.7 Ma to c. 2 Ma. This is
divided into two groups corresponding to the two
divisions on Bacan (c. 2 to 3 Ma and 4 to 7.8 Ma)
but it has not been possible to relate these to
discrete volcanic centres as on Bacan. Volcanic
rocks from Halmahera, like Bacan, do not display a
clear younging of ages within this area and the
effect is only seen on a regional scale. A coarse
northward younging trend is illustrated in Fig. 4
where the ages of samples are plotted against their
locations from north to south.
The present-day active arc that stretches from
Makian through Tidore, Ternate and into the north-
west arm of Halmahera is located approximately
5 0 k m west of the Neogene arc in central
Halmahera. This shift must have occurred at or very
near the time of the Pliocene shortening event that
thrust together the eastern and western halves of
Halmahera and overthrust arc rocks of the Woi
Formation onto the forearc Guyuti Formation on
Obi. It also coincides with a probable gap in the age
spectrum indicating no volcanic activity around
3.6 Ma (Fig. 5).
507
Significance of reset ages
As indicated earlier, apparent Neogene potassium-
argon ages have been obtained from pre-Neogene
rocks. They demonstrate a systematic northward
younging whilst also consistently pre-dating the
ages of Neogene volcanic rocks typically by 3 Ma
within a given area.
Fault-related volcanic activity
The very young volcanic rocks found in the Bacan
and southern Halmahera area of the region (Pulau
Doworalama, Pulau Pisang, Pulau Woka and Pulau
Kekik) are thought to be related to volcanism
occurring along splays of the Strong fault system.
Chemical analyses of these rocks show them to
have an arc-like character but with a significant
within-plate component (Morris et al. 1983) that
supports the notion that they are fault-related.
Quaternary volcanoes in Bacan lie along a lineation
interpreted to be a fault strand whilst the southern
islands lie away from the line of the arc in an
area thought to contain several fault strands. It is
possible that these faults have provided lines of
weakness through which small partial melts,
related to subduction processes, could pass. If these
fault strands do provide conduits for magma
migration this indicates that they must be deep
crustal discontinuities.
Conclusions
Using a conservative estimate of the rate of conver-
gence between the Molucca Sea plate and the
Philippine Sea plate of c. 20 km Ma -1 (R. Hall,
pers. comm. 1994) in the Neogene and assuming a
depth of subducted slab of 100 km beneath the Obi
sector of the arc at the time of the first volcanic
products, the time of initiation of subduction can be
estimated. It is suggested here that the Molucca Sea
slab began to sink beneath the Philippine Sea plate
at around 15 to 17 Ma (corresponding to reset ages
of older rocks from the Obi region). This resetting
reflects locally high geothermal gradients as a
result of crustal heating and magma rise prior to
volcanic activity. Models for the thermal structure
of island arcs indicate a higher heat flow and raised
isotherms over the volcanic portion of the arc
(Oxburgh & Turcotte 1970; Toksoz et al. 1971)
especially in very young arcs. Oxburgh & Turcotte
(1970) suggest that heat transferred high in the
crust by rising magmas could produce near-
surface vertical gradients of c. 100°Ckm -1.
Convective hydrothermal cooling through fractures
and permeable rocks could lead to even higher
gradients (Parmentier & Scheidl 1981). The
patchiness of the alteration and resetting of these
508 S. BAKER • J. MALAIHOLLO
rocks suggests that the m e c h a n i s m for these
changes is predominantly fluid based rather than a
regional raising of isotherm levels. The very
localized nature suggests that fluid migration
occurs along fractures rather than permeating the
rock pile and there is abundant evidence of faulting
and fracturing with associated mineralized veins
within pre-Neogene formations, notably the Bacan
and Oha Formations.
Arc activity in the Halmahera region began on
Obi at 11.5 Ma causing uplift and cessation of lime-
stone deposition. By 9 Ma volcanic activity had
shifted north to Pulau Bisa. Meanwhile, intrusion of
diorites was occurring in Bacan prior to eruptive
v o l c a n i s m at c. 7 Ma. Arc d e v e l o p m e n t on
Halmahera was more or less synchronous with that
on Bacan and limestone deposition in these areas
ceased (except in the backarc region of NE
Halmahera where it continued into the Pliocene).
Pliocene deformation caused a westward shift of
the arc and at c. 2 Ma arc volcanism ceased south of
the equator (Pliocene to Recent volcanic activity in
this region was restricted along strands of the
Sorong fault system).
A large volume of volcanic products was erupted
in the first 1 Ma on Obi with little evidence of
subsequent volcanic activity. These oldest lavas are
calc-alkaline and there is no indication of unusual
boninitic or tholeiitic melts as observed in the early
stages o f some intra-oceanic systems, particularly
the I z u - B o n i n - M a r i a n a System (Hawkins et al.
1984; Stern & Bloomer 1992). Subsequent volcanic
References
BALLANTYNE,P. D. 1990. The Petrology of the Ophiolitic
Rocks of Eastern Halmahera, Indonesia. PhD
thesis, University of London.
Cox, K. G., BELL, J. D. & PANKHURST,R. J. 1979. The
Interpretation of Igneous Rocks. Allen and Unwin,
London
CARDWELL,R. K., [SACKS,B. L. & KARIG,D. E. 1980. The
spatial distribution of earthquakes, focal mechanism
solutions, and subducted lithosphere in the
Philippine and northeastern Indonesian islands. In:
HAYES, D. E. (ed.) The Tectonic and Geologic
Evolution of Southeast Asian Seas and Islands.
American Geophysical Union Monograph, 23,
1-35.
HAKIM, A. S. 1989. Tertiary volcanic rocks from the
Halmahera Arc, Indonesia. MPhil Thesis,
University of London.
& HALL, R. 1991. Tertiary volcanic rocks from the
Halmahera Arc, eastern Indonesia. Journal of
Southeast Asian Earth Sciences, 6, 271-287.
HALL, R. & NICHOLS,G. J. 1990. Terrane Amalgamation
in the Philippine Sea Margin. Tectonophysics, 181,
207-222.
, FULLER,M., ALl, J. R. & ANDERSON,C. D. 1995b.
The Philippine Sea Plate: Magnetism and
rocks from other parts of the arc are also normal
calc-alkaline types. Any unusual chemical and
isotopic c o m p o s i t i o n s are related to crustal
contamination of subduction melts or fault-related
volcanism. On Bacan and Halmahera, volcanism
also appears to have occurred in pulses of 1 to 2 M a
duration followed by similar periods of relative
volcanic quiescence. Subduction-related volcanic
activity did not occur along the whole length of arc
simultaneously but migrated northwards with time
probably as a c o n s e q u e n c e of the c l o c k w i s e
rotation of the Philippine Sea plate (Hall et al.
1995b). Using the oldest age from Obi ( l l . 5 Ma)
and that from central Halmahera (7.7 Ma), volcanic
activity in the arc has m o v e d north at c.
65 k m Ma -~.
This work was supported by NERC award GR3/7149, and
grants from the Royal Society and the University of
London SE Asia Geological Research Group to R. Hall,
and financial assistance from Amoco Production
Company. We thank J. R. Aii, C. D. Anderson, E D.
Ballantyne, E T. Banner, T. R. Charlton, E. M. Finch,
E. Forde, R. Hall, G. J. Nichols and S. J. Roberts for
their help and discussion, C. C. Rundle of NIGL for
guidance and help with isotopic dating and R. Howarth
for suggestions and help with statistical analysis.
Logistical assistance was provided by GRDC, Bandung
and the Director, R. Sukamto with field support by D. A.
Agustiyanto, S. Atmawinata, A. Haryono, S. Pandjaitan
and Sardjono. We also thank reviewers M. E Thirlwall
and T. R. Charlton for their helpful comments and
suggestions in preparation of this paper.
Reconstructions. In: TAYLOR,B. & NATLAND,J. H.
(eds) Active Margins and Marginal Basins: A
Synthesis of Western Pacific Drilling Results.
American Geophysical Union Monograph, 88,
371-404.
, AUDLEY-CHARLES, M. G., BANNER, F. T., HtDAY•r,
S. & TOBING, S. L. 1988a. The basement rocks
of the Halmahera region, east Indonesia: a Late
Cretaceous - Early Tertiary forearc. Journal of the
Geological Society, London, 145, 65-84.
, , , - - &--. 1988b. Late Paleogene-
Quaternary Geology of Halmahera, Eastern
Indonesia: initiation of a volcanic island arc.
Journal ~" the Geological Society of London 145,
577-590.
- - . , ALI, J. R., ANDERSON,C. D. & BAKER,S. J. 1995a.
Origin and motion history of the Philippine Sea
Plate, Tectonophysics, 251,229-250.
, NICHOLS, G . J., BALLANTYNE, P. D., CHARLTON, Z. &
ALI, J. 1991. The character and significance of
basement rocks of the southern Molucca Sea region.
Journal of Southeast Asian Earth Sciences, 6,
249-258.
HAMILTON, W. 1979. Tectonics of the Indonesian region.
US Geological Survey Professional Paper, 1078.
 ARC DEVELOPMENT IN THE HALMAHERA REGION
HARLAND,W. B., ARMSTRONG,R. L., Cox, A. V., CRAIG,
L. E., SMITH, A. G. & SMITH, D. G. 1990. A
Geologic Time Scale 1989. Cambridge University
Press.
HAWKINS, J. W., BLOOMER, S. H., EVANS, C. A. &
MELCHIOR, J. T. 1984. Evolution of intra-oceanic
arc-trench systems. Tectonophysics, 102, 175-205
MALAIHOLLO,J. E A. 1993. The Geology and Tectonics of
the Bacan region , East Indonesia. PhD thesis
University of London.
MIYASHIRO, A. 1978. Nature of alkalic volcanic rock
series. Contributions to Mineralogy and Petrology,
66, 91-104
MOORE, G. F. & SILVER, E. A. 1983. Collision processes
in the Northern Molucca Sea. In: HAYES, D. E. (ed.)
The Tectonic and Geologic Evolution of Southeast
Asian Seas and Islands, Part 2. American
Geophysical Union Monograph, 27, 360-372.
MORRICE, M. G. & GILL, J. B. 1986. Spatial patterns in the
mineralogy of island arc magma series: Sangihe
Arc, Indonesia. Journal of Volcanology and
Geothermal Research, 29, 311-353.
MORRIS, J. D., JEZEK, P. A., HART, S. R. & GILL, J. B.
1983. The Halmahera island arc, Molucca Sea
collision zone, Indonesia: a geochemical survey. In:
HAYES, D. E. (ed.) The Tectonic and Geologic
Evolution of Southeast Asian Seas and Islands, Part
509
2. American Geophysical Union Monograph, 23,
373-387.
OXBURGH, E. R. & TURCOTTE, D. L. 1970. Thermal
structure of island arcs. Geological Society of
America Bulletin, 81, 1665-1688.
PARMENTIER, R. D. & SCHEIDL, E 198t. Thermal aureoles
of igneous intrusions: some possible indications of
hydrothermal convective cooling. Journal of
Geology, 89, 1-22.
ROBERTS, S. J. 1993. The Neogene Stratigraphy of the
Halmahera Region, East Indonesia. Ph.D. Thesis
University of London.
STEIGER, R. H. • J.~GER, E. 1977. Subcommission on
Geochronology: Convention on the use of decay
constants in geo- and cosmochronology. Earth and
Planetary Science Letters, 36, 359-362.
STERN, R. J. & BLOOMER, S. H. 1992. Subduction zone
infancy: Examples from the Eocene Izu-Bonin-
Mariana and Jurassic California arc. Geological
Society of America Bulletin, 104, 1621-1636.
TATSUMI, Y. 1986. Formation of the volcanic front in
subduction zones. Geophysical Research Letters, 8,
717-720
TOKSOZ, M. N., MmEAR, J. W. & JULIAn, B. R. 1971.
Temperature field and geophysical effects of a
downgoing slab. Journal of Geophysical Research,
76, 1113-1138.
 Docking and post-docking escape tectonics in the
southern Philippines
M. P U B E L L I E R 1, R. Q U E B R A L 2, M. A U R E L I O 2 & C. R A N G I N 1
1 C N R S URA 1759, D d p a r t e m e n t de G~otectonique, Universit~ P & M Curie,
75252 Paris, F r a n c e
2 M i n e s a n d G e o s c i e n c e s Bureau, Manila, P h i l i p p i n e s

Abstract: The structure of the Philippine archipelago results from the juxtaposition, between the
Late Miocene and the present, of a volcanic belt against fragments of the Eurasian margin and
associated marginal basins. The southern Philippines offers the opportunity of studying the
mechanics of the deformation from active contraction to a more complex post-docking setting.
The docking period is characterized by a compression which began during the early Late Miocene
in the central Philippines and has been recently studied in the island of Mindanao. There,
deformation initiated in the Late Miocene-Early Pliocene on a NW-trending wrench zone, and
continued until the Late Pliocene with thrusting on west-verging fiats and ramps within the arc
and east-verging thrusts within the Sangihe forearc. This deformation is still active to the south
in the Molucca Sea. The post-docking period began in the Early Pleistocene in northern
Mindanao and is represented by a new geodynamic framework with a paired subduction zone and
strike-slip fault. However, convergence is still active in the Manila and the Negros trenches,
although the Philippine fault is not offset. Large wrench faults, which reactivate the ramp faults
of the collision stage, transfer strain from the Philippine fault to the Manila and Negros trenches.
These observations imply active intra-arc extension and fragmentation within the Philippine
mobile belt.

The Philippine archipelago is composed of frag-
ments of the Eurasian margin, which have been
rifted away from mainland Eurasia (Fig. 1), and a
large volcanic belt, referred to hereafter as the
Philippine arc, whose history is linked to that of the
Philippine plate. The juxtaposition of the Philippine
arc against the margin occurred during the late
Neogene but in detail varies from north to south.
The tectonic features of the collision include strike-
slip faults and thrust tectonics. The post-collision
setting involves two subduction zones of opposite
polarity which define beween them a varied assem-
blage of continental, older volcanic and oceanic
crust fragments upon which the active volcanic arcs
are built (Cardwell et al. 1980; Lewis & Hayes
1984; Fig. 1).
The intra-oceanic belt is built upon an Eocene-
Oligocene volcanic arc and is in tectonic contact
with the Eurasian margin by means of subduction
zones in front of the marginal basins and by means
of collision zones in front of the continental frag-
ments. The subduction zones are arched toward the
marginal basins and anchored at the collision zones
(Fig. 1). The collision zones are from north to
south: the Taiwan area (Davis et al. 1983; Barrier
1985; Pelletier et al. 1985), the Mindoro-Panay
collision zone (Rangin et al. 1985; Marchadier &
Rangin 1990), and the Mindanao collision zone
(Moore & Silver 1983; Hawkins et al. 1985;
Mitchell et al. 1986; Pubellier et al. 1991, 1994).
The re-entry zones within the belt imply active
convergence at the Manila, Negros, and Sulu
trenches. However, the Philippine fault does not
respect such geometry. The elongated Philippine
mobile belt is traversed by the 1200 km long left-
lateral Philippine fault which roughly parallels the
Philippine trench (Willis 1937; Allen 1962; Aurelio
et al. 1991; Barrier et al. 1991). It has been demon-
strated that the onset of m o v e m e n t on the
Philippine fault post-dates the arc-continent
collision in the northern Philippines (Pinet &
Stephan 1990; Ringenbach 1992), in the central
Philippines (Aurelio et al. 1991; Aurelio 1992), and
in Mindanao (Pubellier et al. 1991, 1994; Quebral
et al. 1995).
In addition, active extensional tectonics have
been documented in various parts of the Philippines
(Fig. 1). Several basins have been described in the
island of Luzon, some of which are not directly
linked with motion along the the Philippine fault.
Some extensional lineaments, such as the Macolod
Corridor (Fig. 1), are still controversial. Extension
also occurs in the Visayas (central Philippines)
controlled by faults which parallel the fold axes.

From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, 511
Geological Society Special Publication No. 106, pp. 511-523.
512 M. PUBELLIER ET AL.
Fig. 1. Geodynamic setting of the plate boundary
between the Philippine arc and the Eurasian margin,
and major extensional zones of the Philippine Mobile
Belt. Map distinguishes kinds of faults (thick lines for
subduction and strike-slip, thin lines for normal faults)
and nature of lithosphere (see key). Lithosphere is
classified as continental margin (cross-hatching), intra-
oceanic volcanic arc (fine stipple) or marginal basin
floored by oceanic crust (blank). Labelled tectonic units
are: West Mindanao continental block (WM), Palawan
continental block (PB), Palawan-Sulu trenches (PST),
Sulu arc continental block (SA), SVPF; Sibuyan-Verde
Passage fault (SVPF), Daguma Cotabato fault (DCF),
Saranggani Peninsula block (SP), Zambales Massif
block of oceanic crust material (Z). South China Sea
(SCS), Philippine Sea Basin (PSB), Siargao fault (SF),
Lianga fault (LF), Bislig fault (BF), Macolod Corridor
(MC), Visayan basins (VB), Bohol Sea (BS), Lanao
Lake (LL).
A recent marine survey also indicated an active
basin floored with oceanic crust SE of Marinduque
(Sarewitz & Lewis 1991).
In this paper, the deformation associated with the
docking of the Philippine arc is distinguished
from that associated with the post-docking setting.
It is shown that most of this late deformation is
extensional on the basis of field and an escape
model is proposed for the deformation of the
mobile belt of the Philippines (Fig. 2).
The docking phase
The Neogene collisional setting involves large
strike-slip fault zones (Karig et al. 1986; Sarewitz
& Karig 1986), and local collision zones (Holloway
1982; Bachman et al. 1983; Haeck & Karig
1985; Hawkins et al. 1985; Rangin et al. 1985;
Marchadier & Rangin 1990; Pinet & Stephan 1990;
Pubellier et al. 1991), depending on whether the
Philippine arc was in contact with a continental
fragment or oceanic crust. In Mindanao, wrenching
began at the Late Miocene-Early Pliocene
boundary as a NW trending wrench zone, and
extended untill the Late Pliocene with movement
on west-verging flats and ramps within the arc and
east-verging thrusts within the Sangihe forearc.
Mindanao: wrenching and thrust tectonics
No detailed investigation of collision tectonics has
previously been conducted in Mindanao because
the recognition and the dating of the tectonic
elements is complex and the access is difficult. In
addition, Miocene compressional features (Fig. 2)
were overprinted by Pleistocene to recent extension
and wrenching. A synthesis of the results of the
recent tectonics is compiled on the neotectonic map
of Mindanao (Pubellier et al. 1993; simplified as
Fig. 2).
Deformation began by the latest Miocene in
western Mindanao and is marked by wrench
faulting and folding (Fig. 3, top). The westernmost
part of Mindanao (west Cotabato Basin and Illana
Bay) underwent deformation on parallel NW-
trending fault-bounded elongated pressure ridges.
These ridges are known from seismic lines (Fig. 3,
bottom) and their surface geometry was deduced
from drainage anomaly mapping (Deffontaines
et al. 1993; Pubellier et al. 1994). Some lie along
the eastern flank of the Daguma Ridge near Lake
Sebu or emerge from the Quaternary sediments of
the basin (Roxas Ranges). Synthetic cross-sections
drawn from a compilation of unpublished seismic
lines by Letouzey et al. (1987) show that Pliocene
reflectors unconformably overlie these structures,
which are associated with west-dipping low angle
faults. Their 'en echelon' geometry also suggests
a strike-slip environment which affected Upper
Miocene to Lower Pliocene strata (Pubellier et al.
1991). The largest of these ridges is buried and
known from drilling and seismic lines only (un-
published PNOC report 1986). Seismic lines show
NW-SE faults with jogs resulting from a combi-
nation of reverse and strike-slip faults trending NW.
Correlation of seismic lines show these structures
to be at 90°to the basement reverse faults, and
to connect with the Flecha Peninsula of northern
Zamboanga. Reflection seismic data (Fig. 3)
 DOCKING 8Z ESCAPE TECTONICS, S PHILIPPINES 513

Bohol

iiiiiiiiiiilsilandiiiiiii
Sea

Cagayan

.... ~ ~ Cotabato
~ ~ i ~ Basin t , .

:iiig~ i:iii ....:ii ii!ili~:i!::~,• , ':iliiilii!!:

Pre ple,stocene rocks ........ ~ ii::! iiiil;

:::::: :i~;/~I],,
,, I' ~ii::M::
~:::~'.~!
::~:.~:~:!!
........ .~

I I Phil. Arc basement ~ . . . . . "W,~r.,~fl / i!i:i:i"i:i .~:-

~.:::~:~::..~.:
i............
,. |~:~:~:~"::::::::::::::|
Talaud• ridge + Mindanao oe,~"nic
- - - series -'---
--,,,x -. t • ~ ~:::~:~:-
~.: !:!:!:~::.?.::..

~iiilt Sang,hewedge - ~ ~" ! ~!iiiiiiii~:

inferred Eurasian Margin basement 6 o

Fig. 2. Structural sketch map of the Neogene compressional features of Mindanao island prior to the Pleistocene.

indicate that the deformation is sealed by thick
Pliocene and possibly uppermost Miocene sedi-
ments. In the field, the Opol Formation, whose base
is dated as uppermost Miocene ( N N l l ) by nanno-
plankton (C. Muller, pers. comm. 1994), is uncon-
formable on the basement, although substantially
uplifted and affected by more recent (Late
Pliocene) open folds.
By the end of the Pliocene, deformation was
more intense in central Mindanao. It is marked by
folds axes and thrusts whose orientation varies
from N-S to ENE (080°), which die out along-
strike as they branch into NW-trending faults (Figs
2 & 4). South and west of Cagayan de Oro city,
overturned folds and large thrusts have brought the
metamorphics and the ultramafic complex with its
overlying arc series onto upper Oligocene or Lower
Miocene foraminifera-bearing limestones. The
Opol Formation, dated at its base as uppermost
Miocene (NN11, Pubellier et al. 1991), is uncon-
formable on these structures. The northern part of
the Cotabato Basin also includes two narrow ranges
(north of Pigkawayan) composed of peridotites,
pillows and radiolarites where similar structural
observations were made (Figs 2 & 4). The folds are
sealed by a sequence of rocks including columnar
basalt flows and thick tufts, which have yielded
K - A r ages of 2 Ma (Sajona et al. 1993). This
papwe therefore separates a N W - S E strip (Fig. 4)
with very few outcrops sandwiched between the
Philippine arc and the continental margin on which
the Sangihe forearc rests. This central strip is
interpreted as a zone of wrench deformation
along-strike from the Molucca Sea collision
complex, deduced from the bathymetry observed
during the MODEC cruise (Rangin et al. 1995), and
514 M. PUBELLIER E T AL.

Sindan~/t~r'~urorari~'dge~a.~kjl~,A . . . . . / ~"i~ariyl~iiocenetopre:seni~

=.// Interpretedsectionfromvar~us
~b~~l~ ,pil~ ~ datain centralCotab~y

Illana l N ~ Cotabato /

, ,.-i

12~Ramn~e M~arbel~~
6N

SSW Flechalanticline IllanaBay NNEj

Auroraridge
~.....~_ . ~lt~pA ~ ~ ~ 1sac
. ~,' .,

I SSW / IllanaBay / NN~_EL

FlechaAnticline ~ 1sec
_ I IllanaAnticline

' ''=~2...... "~ " "" ' ~" -: :'-.-~-'~~'~-.'~" ~ ~'~~ '~" ~ "'~L~~ . . . . . . . . . . -. . . . ~---~"4sec

Fig. 3. Structural map of westem Mindanao and seismic lines within the Illana Bay. Similar features in the Cotabato
basin are deduced from Letouzey et al. (1987) and drainage anomalies (Pubellier et al. (1994). Upper map also shows
the Aurora ridge (dark pattern), c 3 values from fault slip analyses in Pleistocene sediments and focal mechanisms
from shallow earthquakes.

with rare outcrops of ophiolites and melanges in
central Mindanao (Fig. 2).
Ramp faults are associated with this event, the
largest of which is the 50 km long left-lateral
Tagoloan fault (Figs. 2 & 4), SE of Cagayan de
Oro, but several others appear on Landsat images.
The Tagoloan fault affects affects the Upper
Miocene to at least Upper Pliocene clastic Opol
Formation overlying serpentinized peridotites. The
fault is capped by 50 m of columnar basalts. The
flows were not dated in this area but samples from
the huge volcanic field of central Mindanao yield
ages of 0.4 Ma and younger (Sajona et al. 1993)
and the volcanic ridge of northern Zamboanga is
intercalated with lower Pleistocene (NN19) marine
sediments. It is suggested that the wrench tectonic
phase was coeval with the thrusting in northern
Mindanao and is of late Pliocene age. The most
 DOCKING & ESCAPE TECTONICS, S PHILIPPINES
• ,.. .....
lind '~ --~2=,"i
Eurasian basement ~ 1 -- ( ' ; ; ] ] ~
F-!7;!~Collision complex of the "~'~.._ :~..... ~,.J,~;; !
L ~ Molucca Sea (Talaud ridge ~ . ~ . . . . . . . . ~" )
TI I ltl & wedge and cent. Mindanao ",~=:~...
JLWPhilippine Arc ~-----1,~
Fig. 4. Schematic map of the structural units of the
transfer zone during the Pliocene. Folds and thrusts
are east-verging in the Sangihe forearc and the margin
(SW), and are west-verging in the Philippine arc
(horizontal pattern) and the central strip (dark pattern).
The central strip is interpreted as the extension of the
Molucca Sea collision complex deduced from the
bathymetry of the MODEC cruise (Rangin et al. 1995)
and with scarce outcrops of ophiolites and melanges in
central Mindanao.
accurate dating of this Late Pliocene compres-
sional event comes from shallow marine sediments
of the Agusan-Davao basin in eastern Mindanao.
There, tight 050 ° to 070°-trending folds affect
Middle Pliocene (NN15) to Upper Pliocene (NN16)
clays and greywackes. These folds are sealed by
Upper Pleistocene (NN20-21) sediments (Quebral
et al. 1995). To the west, the basin rests on the
eastern side of the Central Cordillera, which in an
assymetric west-verging thrust-anticline.
In the southern part of eastern Mindanao, serial
seismic lines onland and offshore of Davao Gulf
(Quebral et al. 1995; Rangin et al. 1995) show the
ridges to be very young asymmetrical anticlines,
most of them being bounded on the eastern side by
reverse faults. Unconformable reflectors overlying
the folds are rare except for locally seismically
transparent detrital fans eroded by channels. These
anticlinal structures are very tight east and north of
Samal island which is formed by the emergence of
such a ridge.
Only the southwestern part of Mindanao shows a
difference in tectonic style because duplexes are
thrust to the east (Fig. 4). The most remarkable
structure is the Saranggani Peninsula which is an
anticline cored with middle Miocene limestone
underlying a thick unit of volcanic and volcani-
clastic upper Miocene rocks (whole rock K-Ar
515
dating, Pubellier et al. 1991 ). This series is overlain
by a sandy and marly unit whose base is of lower
Pliocene age and whose upper section includes
upper Pleistocene sediments. This again indicates
that the beginning of the deformation began by the
LF Late Miocene-Early Pliocene boundary. According
to Quebral et al. (1995), the Pujada Peninsula at the
southernmost end of the Pacific Cordillera, is one
of these east-verging thrusts. These interpretations
were later confirmed by the MODEC cruise
IIIIIIIIII (Rangin et al. 1995).
Jll Luzon: M i d d l e M i o c e n e c o m p r e s s i o n
Coarse upper Oligocene-lower Miocene clastic
sediments (Klondyke Formation) are interpreted to
have come from mass wasting along the arc and are
covered by upper Lower Miocene to lower Middle
Miocene platform limestones. This is interpreted to
record a tectonic event related to the early accretion
of the Philippine arc against the Eurasian margin
with sedimentary detritus carried as far as Ta'iwan
(Florendo 1994; Pelletier et al. 1985). The tectonic
features of this event are folds, thrusts, and poorly
understood strike-slip faults in the Philippine arc as
in Catanduanes island (Geary & Kay 1989) or in
the Zambales massif (westernmost Luzon; Hawkins
& Evans 1983; Karig 1983). Stratigraphic evidence
of strike-slip motion is found within upper Middle
Miocene rocks (Maleterre et al. 1988; Pinet &
Stephan 1990). However, these faults reactivated
previous structures that controlled sedimentation
during the Oligocene. This contraction, which
involved wrenching, was closely followed by a
renewal of volcanic activity attributed to the
present Manila trench (Defant et al. 1990). The
second arc started activity by the late Middle
Miocene and its products are clearly recorded in
sedimentary basins (Ilocos and Cagayan basins) on
both sides of the Cordillera (Pinet & Stephan
1990).
N E Visayas a n d central Visayas:
Late M i o c e n e - E a r l y P l i o c e n e accretion
The Visayan archipelago in the central Philippines
(Fig. 1) is composed of a Cretaceous to Palaeogene
ultramafic, metasedimentary and Eocene to middle
Oligocene volcanic and plutonic basement (Irving
1950; Gervasio 1966). Most important is the
presence of an unconformity at the middle and
upper Miocene boundary (9 Ma; Rangin et al.
1991). This unconformity is typical of the central
Philippines; it is known in Palawan (Fricaud
1984), and Panay-Mindoro (Rangin et al. 1985;
Marchadier & Rangin 1990). It was also found in
oil wells in the Bantayan graben NW of Cebu. The
516 M. PUBELLIER ET AL.
upper Miocene-Pliocene calcareous Carcar
Formation is unconformable on older rocks,
although moderately folded.
Fault set analyses in Bondoc, Masbate and Leyte
indicate that stress tensors correspondingly shifted
from a collision-related direction to a strike-
slip-related one between the latest Miocene to the
Pliocene (Aurelio et al. 1991; Aurelio 1992). This
shift in tectonic regime is best manifested in
Bondoc Peninsula, where collisional ~1 orien-
tations are generally N-S (incompatible with left-
lateral motion along the Philippine Fault) while the
Philippine Fault sl orientations are between 050
and 080 ° . It should however be noted that traces of
older strike-slip faults can be found in the region,
the most conspicuous of which are in Mariduque
island (Fig. 5). Here, a NNW-SSE trending left-
lateral fault affects Middle Miocene sedimentary
formations but is sealed by Plio-Pleistocene
formations. Furthermore, a Quaternary volcano
built on the southern tip of the island does not
appear to be truncated by the strike-slip fault
(Aurelio 1992).
The post-docking phase: Late Piiocene to
Present strike-slip, extension and escape
In contrast, and with the exception of the active
thrusts described in Zamboanga and in NW
Mindoro, most of the active deformation of the
Philippine Mobile Belt is presently extensional and
strike-slip (Fig. 1). Rather than a classical post-
orogenic collapse in the sense of Dewey (1988), the
extensional features observed seem closely linked
to the tectonic setting of oblique convergence and
the geometry of the suture zone. In addition to the
pull-apart basins associated with the Philippine
fault, two fabrics are observed in the the active
grabens on either side of the Philippine fault. A set
of extensional basins situated within the strip
between the Philippine fault and the Philippine
trench is attributed to a process referred to here-
Fig. 5. Structural map of Marinduque island showing
the Late Miocene left-lateral wrench fault zone
(modified from Aurelio 1992).
an
azpi-
qon
3asin
,rgao strait
ianga Bay
3islig Bay
ateel Fault
•Mati Fault
Fig. 6. Sketch diagram showing the dispersion of the
tectonic sliver between the Philippine fault and the
Philippine trench, as a result of the southward migration
of the trench-fault pair,
after as a post-docking dispersion since it removes
fragments of the Philippine arc (Figs 6 & 7). On the
other hand, grabens located in the middle of the
Philippine arc seem to be controlled by the sub-
duction zones in front of the marginal basins to the
west, which generate differential displacements
inside the belt, a process referred to hereafter as
lithospheric escape.
Post-docking dispersion
Dispersion between the Philippine fault and the
Philippine trench (Figs 6 & 7) results from the
oblique convergence along the trench and the
southward propagating trench-fault system. Strike-
slip motion along the Philippine fault separated and
defined a strip of volcanic material between the the
Philippine fault and the trench (Barrier et al. 1991).
Field data show that the fault is older in the
northern Philippines (Pinet & Stephan 1990) than it
is in the central (Aurelio et al. 1991; Aurelio 1992)
and the southern Philippines (Quebral et al. 1995).
This implies that the total offset of the fault
increases from south to north in the manner
 DOCKING ~ ESCAPE TECTONICS, S PHILIPPINES
Fragmentation Dispersion
Fig. 7. Simplified map of the major post-early
Pleistocene extensional faults of Mindanao and the two
main strike-slip fault zones, with labels. Light pattern
for the Eurasian margin and dark pattern for the accreted
Philippine arc undergoing dispersion and fragmentation
(see text). Visayan basins (VB), Tandag fault (TF),
Lianga fault (LF), Bislig fault (BF), Cateel fault (CF),
Daguma fault (DF), Cotabato fault (DCF), Valencia
(East Cotabato fault, VF).
described by McCaffrey (1992) for the forearc of
Sumatra. Differential displacement of discrete
blocks due to a southward decrease in the total
displacement along the Philippine fault accounts
for extension in the region (Fig. 6).
Lianga and Bislig Bays in Mindanao
The northern Pacific Cordillera (Figs 1, 2 & 7) is
considered to be a large ramp anticline guided
along its southern limit by the NNW-SSE Lianga
fault. During the early deformation which affected
upper Pliocene (NN15-16) rocks, the Lianga fault
acted as a left-lateral strike-slip fault. However,
truncation of the northern Pacific Cordillera by the
Philippine fault and its subsequent northward
strike-slip displacement resulted in the reactivation
of the Lianga fault as a normal fault (Fig. 7). The
youngest slickensides observed in upper
Pleistocene (NN20-21) marls therefore correspond
to normal faulting. A NNE-SSW structural fabric,
observed throughout the Pacific Cordillera,
strongly influences the coastline of eastern
Mindanao (Bislig, Cateel Bays, Fig. 7). Bislig Bay,
for example, is interpreted as a half-graben.
Sediments along its northern coast dip gently
517
towards the bay; whereas its southern coastline
is controlled by a major NNE-SSW escarpment
which exposes the Oligocene volcanic basement of
the cordillera. Between the Lianga fault and Bislig
Bay is a carbonate platform whose morphology is
interrupted by normal faults expressed as north-
facing escarpments with NNE-SSW and NNW-
SSE segments which affect rocks as young as Late
Pleistocene.
Northern Philippines
Several other areas in the Philippines display
features which are compatible with this mech-
anism. The Baler-Casiguran basin (Fig. 8) is
located along the southern coast of the Sierra
Madre of Luzon, and was studied on the basis of
seismic lines between the San Ildefonso Ridge and
the Casiguran Coast by Ringenbach et al. (1993).
The elongated graben is considered to have under-
gone minor right-lateral slip according to
Ringenbach et al. (1993) and therefore to be the
result of oblique rifting. However, the overall
geometry on the seismic lines is consistent with a
basically extentional motion. The active faulting
reactivated pre-existing 050°-trending basement
faults.
The Legaspi Lineament (Fig. 8) is a left-lateral
strike-slip fault trending WNW-ESE in the south-
eastern Luzon area. It intersects both the Philippine
Fault in the Ragay Gulf and the Philippine trench
east of Catanduanes Island. Aside from displacing
the Philippine trench left-laterally by about 40 km,
its activity is demonstrated by the occurrence of
earthquakes on its southeastern segment. Onshore,
the Legaspi Lineament is responsible for the
formation of an elongated depression represented
by Lake Bato, probably situated in a local pull-apart
setting in a restricted segment of the lineament. It
also appears to partly control the activity of Mayon
Volcano, the most active of all the Philippine
volcanoes. Into the Ragay Gulf, the lineament
occurs as a steeply SW-dipping fault cutting
through the entirety of the sedimentary fill (seismic
profile). Simultaneous activity along the Legaspi
Lineament and the Philippine fault gives rise to the
elongated geometry of the Ragay Gulf depression
(Fig. 8).
Post-docking escape
Extensional transfer faults
The term transfer faults is used for large wrench
faults which connect the collision zone or the
Philippine fault to the intraplate deformation front
(e.g. the Negros and Manila trenches, or the
Mindoro-Panay thrusts). Geologic data and seismic
518 M. PUBELLIERET AL.
\
\
\ -0
S W
Fig. 8. Structure of the Legazpi Lineament and location of seismic lines at the termination of the Sibuyan-Verde
Passage fault with major focal mechanisms.
lines across these faults reveal that an extensional
component is important (Figs 1 & 7).
Cotabato fault and Aurora volcanic ridge
This ridge is located fight on the suture zone and is
presently an elongated bayonet-shaped horst which
obliquely separates northern Illana Bay from the
southern Dipolog basins. The left-lateral motion
on the bordering NW faults is indicated by several
focal mechanisms determined on shallow earth-
quakes (Fig. 3). Microtectonic studies along the
Sindangan-Cotabato-Daguma lineament clearly
indicate E-W compression generating left-lateral
faulting. Microtectonic measurements on Pleisto-
cene sediments and overlying basalts indicate
N-S extension. E-W normal faults accommodate
the strike-slip motion and have created pull-apart
basins west of Aurora in northern Zamboanga.
-0
\ " -' ® NE
\
\
y d . Slbuyan Sea Fault
\ -0
Daguma fault
Both edges of the Cotabato Basin are fault-
bounded with tilted blocks (Fig. 7). On the eastern
side, a N-S diffuse fault zone exists in central
Mindanao. Tilted blocks and small volcanic cones,
some of them dated between 1.15 and 0.25 Ma
(K-Ar), occur there and underline normal faults
which are thought to reactivate the short limb of
the Central Cordillera ramp anticline. On the
western side, Pliocene thrust-cored folds orientated
NW (N130 °) along the Daguma range have been
reactivated as normal faults with considerable off-
set southeast of Cotabato (Daguma or Tiruray
fault). Some of these normal faults cross-cut the
Parker volcano in southern Kudarat Province and
cut the Quaternary reefal terrace south of General
Santos. Although this extensional pattern can be
related to backarc extension behind the Cotabato
 DOCKING ~; ESCAPE TECTONICS, S PHILIPPINES 519

trench, the authors favour a wrench-related
interpretation following the accretion.
Other occurrences in the central and
northern Philippines
Gabaldon Basin (Ringenbach 1992; Ringenbach
et al. 1993) is a Pliocene basin controlled by the
Cascades fault zone which is presently associated
with the Philippine fault in central Luzon (Fig. 9).
The basin in its present stage is represented by
fluvial deposits of Pliocene age which post-date
the folded lower Miocene sediments. The present
configuration is a N W - S E elongated graben
bounded by listric faults as indicated by tilted
terrace deposits on the basin's southern rim. This
basin has been interpreted to be the result of move-
ment at a releasing bend of the Philippine fault
(Ringenbach et al. 1993), but it also fits well the
model of an escape of the Baguio-Zambales block
from a fixed Philippine fault (Fig. 10). Similar
observations were made by Ringenbach (1992) in
the Pantabangan-Carranglan basins which are
marked by the deposition of middle Miocene
sequences and presently undergo transtension as
well as extension with subsidence.
Presently, the western-central Philippine area
is characterized by the presence of NW-trending
strike-slip faults such as the Sibuyan Sea Verde
Passage faults (Fig. 8). To the south, near Masbate
Island, the Sibuyan Sea fault trends WNW-ESE
and cuts across the southeastern extension of the
Marinduque Basin (Sibuyan Sea) filled with recent
sedimentary deposits. Seismic reflection profiles
clearly indicate extensional structures best
exemplified by tilted blocks and graben structures.
Recent activity along this fault is manifested both
in terms of minor earthquakes (magnitude < 5 Mb)
and variations in stress orientations in Masbate
where the fault intersects the Philippine fault
(Aurelio 1992).
To the NW, the Verde Passage fault trends in the
same direction as the Sibuyan Sea fault as it passes
between Mindoro Island and Luzon towards the
Manila trencE Its termination into the trench is,

t Cretaceous
~to Eocene
Basement

~ Miocene

Philippine Arc

Basement

150Kin
_ _ 1

NE
SW Cascades FZ Philippine FZ #

Northern Sierra Madre + ~ + So rn

• \~" +/Madre Sierra

W Baler Casiguran Basin San IldefonsoRidgc

r'--~'~--~-- = -:--~" - - - - - -. . . . --':-~ " -- . . . . . . ~ _ ~ _ ~ Jl~

Fig. 9. Location and structure of the Baler Casiguran Basin of Luzon and the Mocalod corridor modified from
Ringenbach (1992) and Wolfe & Self (1983). Manila fault (MnF), Marikina fault (MkF).
520
FRAGMENTATIO~
~ Sol,~niab.
L ~ l l l V l ~ " ~ i i ! i : Gabaldon
Jr" ii~Bl~.~:~i~i!::~ f Pantapangan
South [ ~ i i i : : : , Carranglanb.
China ~' ~ k ~ ~ i i i i : : : i : : i ! i i .
S e a ~ ~ ' ~ d
~u,u ~1 ~I!:i:i:!:i3:::
nao
Sea _ ~~ ~ ,=~1
...... ~ Buluan
V Sea = ;:!i!ii!!!!:i:ii oasm
Fig. 10. Sketch diagram and map showing
fragmentation and escape west of the Philippine
fault due to the trench pull in front of the marginal
basins.
however, unclear. It appears to die out into the
accretionary prism which is well developed at this
latitude as observed during the POP2 cruise
(Rangin et al. t988). Seismically, minor to
moderate earthquakes are known to have occurred
along this structure for the past 100 years (Aurelio
1992). Its connection or non-connection with the
Sibuyan Sea fault, however, still remains to be
understood. Bathymetrically, these two strike-slip
faults are in fact linked by the Sibuyan Sea-
Marinduque Basin.
M. PUBELLIERET AL.
Extension resulting from subduction at
the Negros and Manila trenches
L a n a o extensional field, B o h o l Sea a n d
Visayan basins
Extension almost parallel to the convergence
direction was first observed by Rangin et al. (1989)
in the central Philippines where a radial set of
graben post-date the Late Miocene-Early Pliocene
docking phase and tend to parallel the Negros
trench (Figs 1 & 7). The basins were explored for
oil (Glocke 1980) and some of them are filled
with up to 6 km of Miocene to Recent sediments.
Normal faults of the grabens reactivate the short
limbs of the folds and affect the Pliocene to
Pleistocene Carcar Formation (Rangin et al. 1989).
The southernmost basin is the Bohol (or Mindanao)
Sea which is a poorly known basin of 2.5 s
reflection seismic floor with 0.5 s mean sediment
thickness (Hamilton 1979). In central Mindanao,
the 200 km long Lanao volcanic field is a huge
plateau situated around Lanao lake. Nb-enriched
calc-alkaline and potassic lava samples have been
Coridor
dated between 2.31 and 0.29 Ma (Sajona et al.
1993) in this area. Extension is marked by 20 to
50 km long arcuate normal faults (Ranneft et al.
1960) which appear on Landsat images and control
the geometry of the lake. Although the faults do
affect the volcanic aprons, the active or dormant
volcanic cones are aligned in the same direction, as
pointed out by Hamilton (1979). The normal faults
connect with the active transtensile Cotabato fault
field
and Aurora ridge as part of a single tectonic system
with ~3 striking N-S (Pubellier et al. 1994).
'''n
Symmetrically, on the opposite side of the Cotabato
fault, an extensional zone affects the Quaternary
volcanics and the Quaternary reefal terraces of
Mt Blick in the northern Daguma Range. The
050°-trending normal fault provide the structural
control for the tiny Bongo island.
M a c o l o d corridor
The Macolod Corridor (Figs 1 & 8) is a complex
graben structure which has been interpreted as a
pull-apart basin characterized by the occurrence of
intense volcanism (Defant et al. 1990). However,
the bounding faults are not clearly identified and
the mechanism debatable (F"rster et al. 1988).
Wolfe & Self (1983) suspected that this feature
may be a link between the northwestern shelf of
Palawan and the Philippine fault of central Luzon.
Regardless, the distribution and abundance of
volcanoes suggest lithospheric fractures striking
mainly NE-SW but also NW-SE. The large Taal
volcanic centre is located at the intersection of
these directions. Clearly the Macolod corridor
 DOCKING 8Z ESCAPE TECTONICS, S PHILIPPINES
accommodates a differential displacement between
the central and northern parts of Luzon, in response
to forces induced between the Manila trench and
colliding Benham Plateau, and the encroachment of
the southern part of Luzon on the Palawan Block
(Fig. 1).
Marinduque Basin
Recently, incipient sea floor spreading has been
documented in the central Philippines (Sarewitz &
Lewis 1991), in connection with a wrench environ-
ment. The Marinduque Basin, located east of
Marinduque Island and Bondoc Peninsula, has a
NW-SE bathymetric axis. Offshore data, however,
suggest the presence of an E-W trending bathy-
metric ridge which, according to Sarewitz & Lewis
(1991), might correspond to a N-S spreading axis,
based upon the observation that the ridge is made
up of volcanic rocks whose magnetic signature
is comparable to that of rocks produced during
oceanic accretion. In this model, the system, now
inactive, could have operated along N-S trending
strike-slip zones. Such structures are presently
observed as thrust faults lining up the western
edges of the Bondoc Peninsula compressive ridge,
associated with the Philippine fault. In view of its
location between the Sibuyan Sea fault and the
Verde Passage fault, it is tempting to argue that the
Marinduque Basin is an extensional relay zone
between strike-slip faults.
Interpretation and discussion
One of the best recorded Tertiary tectonic events in
the Philippines is the collision of the continental
block of north Palawan with island arc units in
the central Philippines. Manifestations of this
collisional event are readily observable in the
islands of northern Palawan, Mindoro, Panay and
other neighboring islands (McCabc et al. 1983;
Rangin et al. 1985; Mitchell et aL 1986), and
recently mapped in Mindanao (Pubellier et al.
1993, 1994; Quebral et al. 1995). Folds and thrusts
produced by this event are associated with lateral
ramps striking NW-SE. It is believed that the
lateral ramps are part of a large transfer zone
with follows the Eurasian margin geometry and
truncates the Molucca Sea (Rangin et al. 1995). In
Mindanao, this wrench tectonic phase began at the
Late Miocene-Early Pliocene boundary and lasted
References
ALLEN• C. R. 1962. Circum-Pacific faulting in the
Philippine-Taiwan region. Journal of Geophysical
Research, 67, 4795-4812.
AURELIO, M. 1992. Tectonics of the central segment of
the Philippine Fault: structures, kinematics and
521
until the end of the Pliocene resulting in deforma-
tion of the Agusan-Davao Basin. In the northern
Philippines this event is dated as Middle Miocene.
The post-docking period is now represented by
a new geodynamic framework with paired sub-
duction zones and strike-slip faults, due to con-
tinued convergence at the Manila and Negros
trenches. Large transfer wrench faults (Fig. 4),
which reactivate the lateral ramps of the collision
stage, transfer strain from the Philippine fault to the
Manila and Negros trenches. If the Philippine
Mobile Belt east of the Philippine Fault is con-
sidered to be fixed on the basis that the 1200 km
long Philippine fault is not offset nor disrupted by
the transfer faults, it can be inferred that the
western half is undergoing extension driven by
slab-pull at the Manila and Negros trenches. In
response to the extension, this part of the belt is
fragmented and the fragments tend to escape
toward the marginal basins. The escape concept,
which refers to the motion of a package of rocks
towards a free edge within a compressional
environment, has been much described as a by-
product of frontal collision (e.g. the Indochina
block: Tapponnier et al. 1982; Davy and Cobbold
1988; or the Alps: Ratschbacher et al. 1991), or
from the combination of movement of the
Anatolian block triggered by Arabian-Eurasian
convergence and extensional forces due to slab-
pull at the Aegean trench (McKenzie 1972; Sengtr
et al. 1985). This paper proposes that oblique con-
vergence along an irregular continental buttress
is also a likely setting for escape of continental
fragments. The escape motion affects the con-
tinental fragments dragged by the northward
motion of the Philippine plate to a smaller extent.
This paper is a result of five years of co-operation
between the Mines and Geosciences Bureau of the
Philippines, and the Universitl~E et M. Curie in Paris in
the southern Philippines. We are very grateful to the
Philippine National Oil Company (PNOC) and the Bureau
of Energy for allowing access to unpublished data. This
work has been funded by the MinistEre des Affaires
EtrangEres. Laboratory facilities were supplied by the
Universit6 P. et M. Curie in Paris, and some palaeonto-
logical dating were provided by C. Muller (Institut
Francais du Pttrole) and J. Butterlin (Inst. Phys. Globe de
Paris). Two of the authors (M.P and C.R.) belong to the
Centre National de la Recherche Scientifique, and two
others (R.Q. and M.A.) are from the Mines and
Geosciences Bureau of the Philippines.
geodynamic evolution. Th~se de Doctorat de
l'Universite Paris 6, France.
• BARRmR,E., RANOrN,C. & Mt3LLER,C. 1991. The
Philippine Fault in the late Cenozoic evolution
of the Bondoc-Masbate-N. Leyte area, Central
522 M. PUBELLIER ET AL.
Philippines. Journal of SE Asian Earth Sciences,
6, 221-238.
BACHMAN, S. B., LEWIS, S. D. & SCHWELLER,W. J. 1983.
Evolution of a forearc basin, Luzon Central Valley,
Philippines. AAPG Bulletin, 67, 1143-1162.
BARRmR, E. 1985. Tectonique d'une chaine de collision
active: Taiwan. Thbse de Doctorat d'Etat, 85-29,
Paris, France.
BARRIER, E., HUCHON, P. & ACP,EUO, M. 1991. The
Philippine fault: A key for Philippines kinematics.
Geology, 19, 32-35.
CARDWELL,R. K., ISACKS,B. L. & KARIG, D. E. 1980. The
spatial distribution of earthquakes, focal mechanism
solutions and subducted lithosphere in the
Philippine and northeastern Indonesian islands. In:
HAYES, D. E. (ed.) The Tectonic and Geologic
Evolution of South-east Asian Seas and Islands.
American Geophysical Union Monograph, 23,
1-35.
DAWS, D., SUPPE, J. & DAHLEN, F. 1983. Mechanics of
fold and thrust belts and accretionary wedges.
Journal of Geophysical Research, 88, 1153-1162.
DAVY, P. & COBBOLD, P. R. 1988. Indentation tectonics
in nature and experiment. Bulletin of the Geoligical
Institute of Uppsala, N. S. 14, 129-141.
DEFANT, M. J., MAURY,R. C., JORON,J. L., FEIGENSON,M.
D., LETERRIER, J., ET AL. 1990. The geochemistry
and tectonic setting of the northern section of the
Luzon (the Philippines and Ta'iwan). Tectono-
physics, 183, 187-205.
DEFEONTAINES, B., PUBELLIER, M., RANGIN, C. &
QUEBRAL, R. 1993. Discovery of an intra-arc
transform zone in Mindanao (Philippines) using
morphotectonic data. Zeitschrift fiir Geo-
morphologie, 94, 261-273.
DEWEY, J. E 1988. Extensional collapse of orogens.
Tectonics, 7, 1123-1139.
FLORENDO, E E 1994. Tertiary arc rifting in northern
Luzon, Philippines. Tectonics, 13, 623-640.
FORSTER, H., OLES, D. & DEFANT, M. J. 1988. The
Macolod Corridor: a rift crossing the Philippine
island arc. International Symposium on the
Geodynamic Evolution of Eastern Eurasian
Margins, Abstracts. Paris, 53.
FRICAUD,L. 1984. Etude g~ologique et structurale de la
marge Ouest Palawan (Mer de Chine m~ridionale).
Thesis, Universit6 Paris.
GEARY, E. E. & KAY, R. W. 1989. Identification of an
Early Cretaceous ophiolite in the Camarines Norte-
Calaguas islands basement complex, Eastern Luzon,
Philippines. Tectonophysics, 168, 109-126.
GERVASIO,E C. 1966. Age and nature of orogenesis of the
Philippines. Tectonophysics, 4, 379-402.
GLOCKE, A. 1980. Geology and hydrocarbon prospects of
the Visayan Basin. Appendix V: Results of seismic
interpretation. 77, BED unpublished report.
HAECK, G. D. & KARIG, D. E. 1985. Strike-slip genesis
of the Ilocos Norte melange. EOS, 64, 871.
HAMILTON,W. 1979. Tectonics o f t he Indonesian Region.
U. S. Geological Survey Professional Paper, 1078.
HAWKINS, J. W. & EVANS, C. 1983. Geology of the
Zambales range, Philippine islands: ophiolites
derived from an island arc-back arc basin pair. In:
HAYES, D. E. (ed.) The Tectonic and Geologic
Evolution of South-east Asian Seas and Islands.
Part 2. American Geophysical Union Monograph,
27, 96-123
- - , MOORE, G. E, VILLAMOR,R., EVANS,C. & WRIGHT,
E. 1985. Geology of the composite terrane of East
and Central Mindanao. In: HOWELL, D. (ed.)
Tectonostratigraphic Terranes of the Circum-
Pacific Region. Circum-Pacific Council for Energy
and Mineral Resources, Earth Sciences Series, 1,
437-463.
HOLLOWA¥, N. H. 1982. The stratigraphic and tectonic
evolution of Reed Bank, North Palawan and
Mindoro to the Asian mainland and its significance
in the evolution of the South China Sea. AAPG
Bulletin, 66, 1357-1383.
IRVING, E. M. 1950. Review of Philippine basement
and its problems. Philippine Journal Science, 79,
267-307.
KAR~G, D. E. 1983. Accreted terranes in the northern
part of the Philippine archipelago. Tectonics, 2,
211-232.
--, SAREWrrz, D. R. & HAECK, G. D. 1986. Role of
strike-slip faulting in the evolution of allochtonous
terranes in the Philippines. Geology, 14, 852-855.
LETOUZEY,J. E, SAGE, MULLER, C. 1987. Geological and
structuctural maps of Eastern Asia. Institut Francais
du Petrole, Paris.
LEWIS, S. D. & HAYES, D. E. 1984. A geophysical study
of the Manila Trench, Luzon, Philippines; forearc
basin structural and stratigraphic evolution. Journal
of Geophysical Research, 89, 9196-9214.
MALETERRE,P., STEPHAN,J. E, ANDREIEFF,P., BELLON,H.,
CHOROWICZ,J., BOmAT,J. M. & BALCE,G. R. 1988.
The southern Central Cordillera of Luzon: a multi-
stage Upper Eocene to Pleistocene arc deformed on
the northern end of the Philippine strike-slip fault.
International Symposium on the Geodynamic
Evolution of Eastern Eurasian Margin (Abstracts),
Paris 13-20 September 1988.
MARCHADIER,Y. & RANG~N,C. 1990. Polyphase tectonics
at the southern tip of the Manila trench: Mindoro-
Tablas islands, Philippines. Tectonophysics, 183,
273-288.
MCCABE, R., ALMASCO,J. & DIEGOR, W. 1983. Geologic
and paleomagnetic evidence for a possible Miocene
collision in Western Panay, central Philippines.
Geology, 10, 325-329.
MCCAFFREY, R. 1992, Slip vectors and streching of the
Sumatran fore arc. Geology, 19, 881-884.
MCKENZIE, D. E 1972. Active tectonics of the
Mediterranean region. Journal of Geophysical
Research, 30, 109-185.
MITCHELL, A. n. G., HERNANDEZ, E & DE LA CRUZ,
A. E 1986, Cenozoic evolution of the Philippine
archipelago. Journal of SE Asian Earth Sciences,
1, 3-22.
MOORE, G. E & SILVER,E. A. 1983. Collision Processes
in the northern Molucca Sea In: HAYES,D. E. (ed.)
The Tectonic and Geologic Evolution of South-east
Asian Seas and Islands. Part 2. American
Geophysical Union Monograph, 27, 360-372
PELLETIER, B., STEPHAN,J. E, BLANCHET,R., MULLER,C.
& HU, H. N. 1985. L'rmergence d'une zone de
collision active tl la pointe sud de Taiwan (Prninsule
 DOCKING & ESCAPE TECTONICS, S PHILIPPINES
d'Hengshun): tectoniques superpos6es et mise en
6vidence d'une tectonique miocene moyen. Bulletin
de la SocMtg gdologique de France, 8, I, 161-171.
PIm~T, N. & STEPHAN,J. E 1990. The Philippine wrench
fault system in the Ilocos Foothills, Northwestern
Luzon, Philippines. Tectonophysics, 183, 207-224.
PUBELLIER, M., DEFFONTAINES, B., QUEBRAL, R. &
RANG~N, C. 1994. Drainage network analysis and
tectonics of Mindanao, Southern Philippines.
Geomorphology, 9, 325-342.
, QUEBRAL, R., DEFFONTAINES, B. & RANGIN, C.
1993. Neotectonic map of Mindanao (Philippines).
1:800,000 with explanatory notes. Asia Geodyne
Co. Manila.
, RANGIN, C., DEFFONTAINES,B., MULLER, C.,
BUTTERLIN,J. & MANZANO,J. 1991. The Mindanao,
collision zone, a soft collision event within a
continuous strike-slip setting. Journal of Southeast
Asian Earth Sciences, 6, 239-248.
QUEBRAL, R., PUBELLIER,M. & RANGIN, C. 1995. Eastern
Mindanao, Philippines: transition zone from a
collision to strike-slip environment. Tectonics, in
press.
RANGIN, C., DAHRIN, D., QUEBRAL, R. & THE MODEC
SCIEYrIFIC PARTY 1995. Collision and strike-slip
faulting in the northern Molucca Sea, (Philippines
and Indonesia): preliminary results of a morpho-
tectonic study. This volume.
RANGIN, C., MULLER, C. & PORTH, H. 1989. Neogene
Geodynamic Evolution of the Visayan Region. In:
PORTH, H. & VON DANmLS, C. H. (eds) On the
geology and hydrocarbon prospects of the Visayan
Basin, Philippines. Geologisches Jahrbuch, B 70,
7-28.
, STEPHAN, J. E, BLANCHEr, R., BALADAD, D.,
BOUYSSE, P., ET AL. 1988. Sea-Beam survey at the
southern end of the Manila trench. Transition
between subduction and collision processes, off-
shore Mindoro Island, Philippines. Tectonophysics,
146, 261-278.
., BUTTERLIN, J., BELLON, H., MOLLER, C.,
CHOROWICZ, J. (~ BALADAD, D. 1991. Collision
n6og6ne d'arcs volcaniques dans le centre des
Philippines: stratigraphie et structure de la chaine
d'Antique 01e de Panay). Bulletin de la Soci6t6
g6ologique de France, 162, 465--477.
, & Muller, C. 1985. Middle Oligocene
oceanic crust of the South China Sea. jammed into
Mindoro Collision Zone, Philippines. Geology, 13,
425-428.
523
RANNEFT, T. S. M., HOPKINS, R. M. J., FROELICH,A. J. &
GWlNN, J. W. 1960. Reconnaissance geology and
oil possibilities of Mindanao. AAPG Bulletin, 44,
529-568.
RATSCHBACHER,J., FRISCH, W., LINZER, H. G. & MERLE,
O. 1991. Lateral extrusion in the Eastern Alps.
Tectonics, 10, 245-271.
RINGENBACH, J. C. 1992. La Faille Philippine e t l e s
cha~nes en ddcrochement assocides (Centre et
Nord de Luzon). These de Doctorat et Travaux
de l'Institut de G6odynamique de Nice, 16.
~, STEPHAN, J. E, MALETERRE, PH. • BELLON, H.
1993. Structure and geological history of the
Lepanto-Cervantes releasing bend on the Abra
River Fault, Luzon Central Cordillera, Philippines.
Tectonophysics, 183, 225-242.
SAJONA, F., MAURY, R., BELLON, H., COTTEN, J., DEFANT,
i . & PUBELLIER,M. 1993. Initiation of subduction
and the generation of slab melts in Western and
Eastern Mindanao, Philippines. Geology, 21, 1007-
1010.
SAREWITZ, D. & KARIG, D. E. 1986. Processes of
allochthonous terrane evolution in Mindoro Island,
Philippines. TECTONICS,5, 525--552.
& LEWIS, S. D. 1991. The Marinduque intra-
arc basin, Philippines: basin genesis and in situ
ophiolite development in a strike-slip setting.
Geological Society of America Bulletin, 103,
597-614.
SENGOR, A. M. C., GORUR, N. & SAROGLU, E 1985.
Strike-slip faulting and related basin formation
in zones of tectonic escape; Turkey as a case
study. In: BIDDLE, K. T. & CHRISTIE-BLICK, N.
(eds) Strike-Slip Deformation, Basin Formation
and Sedimentation. Society of Econonomic
Paleontologists and Mineralogists Special
Publication, 37, 227-264.
TAPPONN/ER, P., PELTZER, G., LEDAIN, A., ARMIJO, R. &
COBBOLD, P. 1982. Propagating extrusion tectonics
in Asia: new insights from simple experiments
with plasticine. Geology, 10, 611-616.
WILLIS, B. 1937. Geologic observations in the Philippine
islands. Natural Resources Council of the
Philippines Bulletin, 13.
WOLFE, J. A. & SELF, S. 1983. Structural lineaments and
Neogene volcanism in southwestern Luzon. Sea. In:
HAYES, D. E. (ed.) The Tectonic and Geologic
Evolution of South-east Asian Seas and Islands.
Part 2. American Geophysical Union Monograph,
27, 157-172.
 Thermochronological and geochemical constraints on the tectonic
evolution of northern Papua New Guinea

R V. C R O W H U R S T , K. C. HILL, D. A. F O S T E R & A. R B E N N E T T
Victorian Institute of Earth and Planetary Sciences, School of Earth Sciences,
La Trobe University, Melbourne, Victoria, 3083, Australia

Abstract: The Bewani-Torricelli-Prince Alexander Mountains, along the northern margin of

Papua New Guinea, probably formed as a tholeiitic island arc in the Late Eocene-Early
Oligocene. The arc is interpreted to have accreted to the margin by the Late Oligocene, but may
have formed on a ribbon of extended continental crust along the New Guinea margin. Inferred
roll-back of the subducting slab beneath New Guinea in the Early Miocene placed the margin into
extension, creating starved graben in northern New Guinea and causing regional subsidence.
Near the graben, metamorphiccore complexes resulted as the upper crust was pulled off the lower
crust along low angle detachments such that lower crustal rocks cooled rapidly from temperatures
>500°C. The two inferred core complexes that have been dated show rapid cooling from
27-23 Ma and 20-18 Ma. Continued subduction beneath New Guinea resulted in formation of
the Maramuni arc in the Middle Miocene and the end of extension. In the Late Miocene,
collision of the Melanesian arc caused regional uplift of all basement of northern Papua New
Guinea, mainly from 8-5 Ma, causing at least 3-4 km of denudation. The compressional
deformation propagated south causing uplift, denudation and cooling in the Papuan Fold Belt at
c. 4 Ma, but is continuing at the present.

New Guinea is a prime example of arc-continent
collisions, one of which continues today. The recent
and ongoing events make New Guinea an important
area to study in order to understand arc-continent
collision, terrane accretion, the consumption of
microplates and the role of extension in the
collision process. The collision zone in northern
Papua New Guinea is the New Guinea Mobile Belt
(Fig. 1), variously interpreted to be underlain by
Palaeozoic continental crust or accreted Tertiary
oceanic microplates. The Mobile Belt is seismic-
ally active and the neotectonics and Mio-Pliocene
collision of the Melanesian arc with the Australian
continent are moderately well understood.
However, several models have been proposed for
previous collision, accretion or docking events of
arcs/terranes, particularly in the Eocene and the
Late Oligocene-Early Miocene. Pigram & Davies
(1987) and Struckmeyer et al. (1993) proposed
docking of the Sepik or North New Guinea terrane
in the Late Oligocene causing a compressional
orogeny resulting in exposure of Lower Cretaceous
medium to high grade metamorphic rocks in the
Mobile Belt, with the Sepik-Ramu Basins as fore-
land basins (Fig. 2). In contrast, Hill et al.
(1993) suggested that the Late Oligocene-
Early Miocene was a time of extension in
northern New Guinea, creating the Sepik-
Ramu Basins as rift basins. They proposed that
the exposure of the metamorphic rocks was
due to tectonic delamination of the crust creating a
metamorphic core complex.
Critical areas to resolve these conflicting models
are the Bewani-Torricelli Mountains separating the
Sepik and Aitape Basins, and the metamorphic
provinces immediately south of the Sepik Basin
(Fig. 2). Although along-strike and largely basic in
nature, the Bewani-Torricelli Mountains may not
be akin to the accreted Palaeogene Melanesian arc
in the Adelbert and Finisterre Ranges (Fig. 2) as
they include Upper Cretaceous igneous and meta-
morphic rocks. In order to test the models and
determine the tectonic history, this paper presents
the results of field mapping, apatite fission track
thermochronology and geochemical analyses of
basement rocks in the collision zone, from around
the Sepik Basin (Fig. 1). Field mapping defines the
present structure and stratigraphy, fission track
analyses delineate the Neogene events and
geochemical analyses help to constrain the
Palaeogene tectonogenesis.
Stratigraphy and structure of northern
Papua New Guinea
The study area around the Sepik Basin can be
divided into four main structural-stratigraphic
units: the Mobile Belt metamorphic and ophiolitic

From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, 525
Geological Society Special Publication No. 106, pp. 525-537.
526 r'. v. CROWHURST ET AL.

140° I Caroline Plate

I 500 km I"ml~v, . ~ ~ "
A_AdelbertArc ~ h ~ "~ ~ 0
Arc
F - Finisterre J ~/ ~ ' - ' - " ~ ~ ~ i t a ~r"~pe Basin ~ ...... -B"i s m a r c k " -ea
SSP- Solomon Sea Plate ~ / / / ~ ~a~in~s~. . . . . . Plate

~ ~ ° t ~ P ~ °~ite .
120°~ 140°~ 160° 20°. ~ / ~ O ; ~ L
-'0,~ ~, : - j racitic .~,~"~. -'~ l~,~ f .7 ~ r/ . r 3 ! q e ~ ~"-~ .

0°_ PapuanUltramafica , ' ~

( '%127 " \ ..
A
NG- NewGuinea SP- Solomon SeaPlate Indo-/1 Australian Plate
BP - Bismark Plate O - Melanesian Arc I 150 °
i

Fly Platform I Papuan Fold Belt New Guinea Mobile Belt Melanesian Arc
I
Komewu Darai Kubor Marum Adelbert-Finisterre-
Strike-slip faults?
A 1& 2 Plateau Gobe Anticline ®1× OlX Ophiolite Huon Block A l
SW NE

V.E. approx, x 2 \
Continental Basement Ramu Basin Oceanic Basement
i Approx. 100 km i (Early Miocene)

Fig. 1. Plate tectonic setting of New Guinea showing the major structural zones, after Dow (1977) and the area of
this study, around the Sepik-Ramu and Aitape Basins and the Bewani-Torricelli Mountains. The inset shows the
present plate convergence vector and the location of the Melanesian arc. The schematic and stylized section illustrates
the complex structure of folding and thrusting which has formed in response to the Late Miocene to Recent collision
of the Papua New Guinea margin with the Melanesian arc.

rocks; the Sepik Basin, the Bewani-Torricelli-
Prince Alexander Mountains and the Aitape Basin
(Fig. 2).
Mobile Belt metamorphic and ophiolitic
rocks
Between the Papuan Fold Belt, formed in the
Pliocene (Hill 1991), and the Sepik Basin is a
highly faulted zone of Cretaceous and Palaeogene
metasediments and medium-high pressure meta-
morphic schists, gneisses and amphibolites with
Early Miocene cooling ages. These are interpreted
to have been thrust faulted against Middle Miocene
volcanics and to be overlain by an uppermost thrust
sheet, preserved in mountain tops, comprising the
April ophiolite, probably of Palaeogene age (Fig. 2;
Rogerson et al. 1987).
The Sepik Basin
Francis (1990) describes the Sepik Basin as a
Pliocene to Pleistocene successor basin which
unconformably overlies three Miocene infra-
basins. Reflection seismic data suggest that these
are rift-graben or pull-apart basins bordered by
Lower Miocene carbonates, but inferred to have a
condensed Lower Miocene basinal sequence
(Doust 1990; Hill et al. 1993). The basins are filled
with mainly bathyal, but shallowing upwards,
Middle to Upper Miocene volcaniclastic rocks. The
basins are probably underlain by Cretaceous to
Lower Tertiary basic igneous and metamorphic
rocks of oceanic origin (Doust 1990), but may
include areas of crystalline Palaeozoic continental
crust (Rogerson et at. 1987).
Bewani-Torricelli-Prince Alexander
Mountains
These mountains expose sediments of the Sepik
and Aitape Basins on the flanks, including
deformed Pliocene sediments, but the mountain
core comprises Palaeocene to Lower Miocene
Bliri Volcanics and the Cretaceous-Palaeogene
 TECTONIC EVOLUTION OF NORTHERN PAPUANEW GUINEA 527

P l ~ c e ~ - Q u t ernary
[ I depeat~ (sed, 'volt,)
Be~la~o'~T~ee/l i ~ 50, , 100kml
uuas F.ocene-Mi~.ene arc
n ~ t e r i ~ (sed, vok)
Eocene,-Mioeene sediments,
7+1 intru~ives and e x t r u v ~

~1:1 b~. Meltable-Miocene metamorphica

mid ledimenla of the Mobile Belt
,17.6bt-19.3hbMa

Basin

6t4

Radiometric dates axe by : ~ff"~¢.~

Potassium-Argon analyses on...
bt -"biotite Rubidium-Strontium analyses on...
hb - amphibole
In - muscovite Rb - whole rock
ph - pheugite Rbm- muscovite
gl - glauco~aue
ZFTA - Zircon Fission Track Age ~ Apatite Fission Track Ages
BTFZ - Bewa~i-Tomce]ti Fault Zone 6~1 (dosure temp. - ll0+l-10oc)
Ultramafics south of Sepik Basin are termed April Ophiolites (Rogerson et al, 1987)

Fig. 2. The general geology of northern Papua New Guinea, showing the Mobile Belt to the south, the
Bewani-Torricelli-Prince Alexander Mountains to the north and the Finisterre-Adelbert (Melanesian) arcs to the east.
Also shown are K-Ar and Rb-Sr metamorphic cooling ages from Rogerson et al. (1987), Hutchison & Norvick
(1980) and Page (1976), which indicate that cooling from high temperatures occurred during the Cretaceous and in
the Late Oligocene to Early Miocene. Thirty-four apatite fission track ages overlay the geology, indicating that
regional cooling from >110-<60°C occurred during the Late Miocene-Early Pliocene. The single zircon fission track
age indicates an earlier cooling period through c. 200--250°C. The ultramafic rocks south of the Sepik basin are
termed the April ophiolite after Rogerson et al. (1987). Most intrusive and extrusive rocks in the Mobile Belt are
from the Miocene Maramuni arc (Rogerson et al. 1987).

Torricelli Intrusive Complex (Hutchison & Norvick
1980; Rogerson et al. 1987). The Torricelli
Intrnsives consist of dolerites, gabbros and diorites,
whereas the Bliri Volcanics comprise basalts,
andesites and volcaniclastic rocks. The Prince
Alexander Mountains in the east include amphi-
bolite and gneiss with Mesozoic and Early Miocene
metamorphic cooling ages, together with igneous
intrusions in metapelites, the latter extending
beneath the Sepik Basin (Hutchison & Norvick
1980; Rogerson et al. 1987; Doust 1990).
Aitape Basin
Kugler (1990) described the Aitape Basin as a
mildly deformed half-graben which deepens rapid-
ly to the south towards the Bewani-Torricelli fault
zone (Fig. 2). Using reflection seismic data, he
showed preserved normal faults within the basin
and suggested that the Aitape Basin was connected
to the Sepik Basin prior to c. 10 km of uplift of the
Bewani-Torricelli Mountains. Recent detailed
fieldwork (Hilyard et al. 1994) has better defined
the main stratigraphic units of the southern margin
of the Aitape Basin (Fig. 3). Restricted marine
deposition is inferred before the Early Miocene
followed by deposition of the coarse volcaniclastic
Wogamush Beds, correlated with subduction
beneath PNG initiating the continental margin
Maramuni arc (Rogerson et al. 1987; Hill et al.
1993). This was followed in the Late Miocene-
Pliocene by deep-water turbidite deposition of the
Barida Beds, synchronous with thrust-related uplift
of the Bewani-Torricelli Mountains, from which
most of the shoaling sediments of the Aitape Basin
were derived (Fig. 3).
528 P.V. CROWHURST ET AL.

B Apatitefission track age(Ma) recording

S coolingfrom >110°C to <60°C ] 0 2 4 ~'
...._ ~lt kilometres 500o

f[ ~+~ 5 ,- ... -- "" ~ - - "- .. Bewani-TorriceUi Fault Zone 40o0

~/11_+6 ( -'"" ---- \ J / t ,,
/11
',,/ 17' ./y
3

.-V'"/
,,,,'"~
.
BewamMountmns
.

I \ ~
""
... ~ - -
/Z'"--''"/'--'', O~-~
~000.

• ,,,,,,,,,,,',',','," ~..~ ~ ,~

older basement [.;.~.,~.q k volcanics ." ..... , ,.,.,',',.'..',',L..',',.",',',','..',

Fig. 3. An interpretative structural cross-section trending N - S from the southern margin of the Aitape Basin, through
the Bewani Mountains and into the Sepik Basin (see Fig. 2 for location). The boundary between the Bewani
Mountains and the Aitape Basin, known as the Bewani-Torricelli fault zone, is dominated by an imbricate fan
structure with south over north folding and thrusting, probably with a component of sinistral strike-slip movement.
The fission track data from the Bewani Mountains indicate that the peak of denudation of the basement, and hence the
major deformation, probably occurred during the Late Miocene.

Structure of the north Bewani The pattern of steep-dipping imbricate thrust

Mountains-Aitape Basin
The southern boundary of the Aitape Basin is
marked by the east-west trending Bewani-
Torricelli fault zone. Stream traverse mapping of
the northern Bewani and Torricelli Mountains
demonstrated basement rocks uplifted to the south
across the fault zone, locally overturning the
sedimentary section. Faults and brecciated zones
were observed to be generally steep-dipping, where
seen in the mountains, but no estimation of slip
direction was possible due to poor outcrop.
Multiple faults are inferred within the
Bewani-Torricelli fault zone, but it is difficult to
correlate these and the intervening thrust sheets
between widely spaced stream sections. Therefore,
the structural style is illustrated on a north-south
section across the Bewani Mountains (Fig. 3). The
mountains are interpreted to have been upthrust to
the north on a series of steep-dipping imbricate
thrusts which may coalesce downwards into a
shallow-dipping detachment or a steep-dipping
strike-slip fault. The tightly to isoclinally folded
nature of both the basement and the cover suggests
that folding may have occurred before the steep
thrust faults broke through. However, the lack of
any cleavage in the sediments and the absence
of any metamorphism (except hydrothermal
alteration) indicates that the rocks were relatively
cool when deformed.
faults shown in Fig. 3 resembles that of a positive
flower structure formed in a transpressional regime,
as proposed by Doust (1990) for the Bewani-
Torricelli-Prince Alexander Mountains. This is
consistent with the narrow, ribbon-like nature of the
mountains and their obliquity to the direction of
Indo-Australian and Pacific plate convergence
(Figs 1 & 2). It is also consistent with the first
motion solutions from earthquakes in the region
which record both overthrust and transcurrent
movement (Hutchison & Norvick 1980). The
faulting shown in Fig. 3 clearly deforms the
Neogene section. Therefore, the Bewani-
Torricelli-Prince Alexander Mountains are inter-
preted as positive flower structures due to left-
lateral transpressive motion along the Bewani-
Torricelli fault zone during the Late Miocene-
Pliocene. This is consistent with the time of
accretion of the Adelbert and Finisterre arcs,
c. 500 km along-strike to the east (Figs 1 & 2). The
estimate of the time and amount of uplift and
denudation can be tested by apatite fission track
analysis.
Apatite fission track analysis
A total of 51 surface samples were processed from
intrusive and metamorphic complexes in northern
Papua New Guinea and 34 yielded apatite for
fission track analysis (Table 1). Zircon yields were
Table 1. Apatite fission track analytical results and details from northern Papua New Guinea

Sample Longitude Latitude Elevation Lithology Chronostrat, Number Standard Fossil Induced Chi Fission
Number (m) age (Ma)$ of track track track square track
92PNG- grains density density density probability age~[
(×106 cm -2) (xl05 cm -2) (×106 cm -2) (%) (Ma)

17 141-09-30 3-20-05 320* diorite >38 2 2.943(3973) 0.185(2) 0.861(93) 68.7 10.9 + 7.8
8# 141-11-45 4-18-30 250* gneiss >38 20 2.943(3973) 0.032(4) 2.780(343) 95.4 5.9 ± 3.0
127 141-25-30 4-17-45 320 diorite 42.6-20.1 20 2.943(3973) 1.841(85) 132.6(6121) 100 7.0 ± 0.8
13t 141-14-05 3-36-40 360 diorite >38 20 2.943(3973) 2.288(84) 158.9(5835) 99.8 7.3 ± 0.8
23? 141-11-20 3-14-30 600 granite 97.5-14.4 11 2.943(3973) 0.101(6) 10.30(609) 87.2 5.0 + 2.0 c~
24~ 141-11-20 3-14-30 605 granite 97.5-14.4 12 2.943(3973) 0.055(2) 12.91(467) 99,1 2.2 ± 1.5 O
28t 142-46-35 4-26-15 180 granodiorite >38 6 2.943(3973) 1.107(27) 82.98(2024) 77.4 6.7 ± 1.3 Z
29t 142-46-30 4-26-20 150 diorite >38 13 2.943(3973) 1.683(133) 118.1 (9329) 99,7 7.2 ± 0.6 c~
r~
32t 142-47-10 4-25-20 -- meta-felsic >38 7 2.943(3973) 1.518(47) 61.89(1916) 97.9 12.4 ± 1.8 <
34t 142-46-45 4-25-40 -- granite >38 20 2.943(3973) 1.815(82) 135.3(6113) 100 6.8 + 0.8
35t 142-47-30 4-25-15 -- diorite >38 20 2.943(3973) 0.4817(62) 40.18(5171) 99.7 6.1 ± 0.8
36? 142-47-30 4-25-15 -- granodiorite >38 21 2.943(3973) 0.383(28) 35.28(2578) 1t30 5.5 ± 1.0 ~-©
42t 143-00-10 4-15-45 -- diorite 19.7-22.8 6 2.943(3973) 1.975(40) 68.84(1394) 76.3 14.5 ± 2.3
43~ 143-00-10 4-15-45 -- diorite 19.7-22.8 5 2.943(3973) 0.638(12) 21.69(408) 98.8 12.5 + 3.7 ©
45? 142-47-10 4-47-15 200* andesite >38 16 2.943(3973) 0.107(15) 18.18(2536) 99.7 2.5 ± 0.7
Z
52? 142-09-30 3-20-50 310 granite 97.5-14.4 6 2.943(3973) 0.7136(7) 53.41(524) 96.1 5.7 ± 2.2 O
60t 142-08-45 3-25-35 780 ultramafic 97.5-14.4 8 2.943(3973) 1.671(63) 103.2(3890) 100 6.9 ± 0.9
617 142-08-45 3-25-35 780 ultramafic 97.5-14.4 2 2.943(3973) 1.389(10) 69.17(498) 4.9 8.6 ± 2.7 ~=~
=
69? 142-09-00 3-24-10 1180 meta-sed.? 97.5-14.4 27 2.943(3973) 1.069(82) 70.87(5434) 99.8 6.4 ± 0.7
L
79? 142-54-30 3-26-30 300* gneiss 97.5-38.0 26 2.943(3973) 0.521(74) 35.82(5.87) 98.3 6.2 ± 0.7
081at 144-49-20 5-21-15 150 volc. sandstone -- 7 2.943(3973) 0.373(8) 37.58(805) 60.4 4.2 ± 1.5
2# 141-10-00 3-20-50 -- diorite >38 5 1.397(6285) 0.025(4) 0.575(91) 15.8 10.8 ± 5.5
4# 141-11-45 4-18-25 230 amphibolite >38 20 1.397(6285) 0.017(21) 1.239(1504) 76.4 3.4 ± 0.8 >
011a# 141-09-30 4-16-20 200* garnet gneiss >38 21 1.397(6285) 0.138(212) 4.606(7068) 35.4 7.4 + 0.5
14# 141-14-05 3-36-40 360 diorite >38 20 1.397(6285) 0.020(11) 0.924(498) 8.9 5.4 ± 1.9
25# 141-11-20 3-14-30 600 granite 97.5-14.4 16 1.397(6285) 0.042(21) 1.480(738) 5.1 7.2 ± 2.0
37# 143-13-10 4-39-30 -- gneiss >38 20 1.397(6285) 0.005(3) 0.208(119) 19.5 6.3 ± 3.7
39# 143-09-10 4-39-30 -- meta-felsic >38 20 1.397(6285) 0.002(1) 0.132(87) 99.4 2.8 ± 2.8
053+54# 142-09-20 3-21-30 460 granite 97.5-14.4 35 1.397(6285) 0.005(5) 0.434(399) 99.1 3.1 ± 1.4 >
58# 141-09-45 3-21-40 570 granite 97.5-14.4 20 1.397(6285) 0.005(3) 0.303(196) 62.7 3.8 ± 2.2
59# 141-09-45 3-21-40 650 granite 97.5-14.4 20 1.397(6285) 0.012(6) 0.400(193) 66 7.6 ± 3.2
68# 142-09-00 3-24-20 1140 meta-mafic 97.5-14.4 30 1.397(6285) 0.018(40) 0.304(672) 4 15.0 + 3.0
102# 142-25-00 5-00-30 280 greywacke -- 30 1,397(6285) 0.015(19) 0.437(553) 20.2 8.4 + 2.0
K8-22# 145-18-30 5-45-30 1700" diorite c. 60 30 1.212(5488) 0.015(14) 0.638(615) 9.9 5.1 ± 1.5
K8-11# 145-43-45 6-15-00 600 gneiss-granite c. 150-190 10 0.784(3723) 14.02(2184) 37.90(5906) 0 17.5 ± 1.3

t counted by Peter Crowhurst (Zeta value = 343.5 ± 5), # counted by Kevin Hill (Zeta value = 350.0±5) ; * Denotes estimation of elevation
$ Chronostraigraphic ages are estimated from Rogerson et al. (1987), Hutchinson & Norvick (1980) and Page (1976); Ix#
Fission track age is central age after Galbraith & Laslett (1993); note K8-11 is a zircon fission track analysis; Track densities in ( ) are the number of tracks counted.
530 P.v. CROWHURST ET AL.

generally poor, and only one zircon fission track
analysis is reported. Methods followed those
recently reviewed by Foster & Gleadow (1992).
Fission track ages were calculated using the zeta
calibration method and standard fission track age
equation (Hurford & Green 1982). Errors were
calculated using the techniques of Green (1981).
The zircon sample was analysed similarly,
following the methods of Gleadow & Lovering
(1978).
Sample results
The 51 rocks processed comprised 29 igneous, 19
metamorphic and 3 sedimentary samples. Apatite
yields for the 34 samples counted were variable
from poor (1-10grains) to very good recovery
(>20 datable grains): Due to the consistency
of young fission track ages and low uranium con-
centrations of many samples very few confined
tracks were present and none were measured.
Similarly, the number of tracks counted was
usually low, resulting in low precision on single
grain ages, such that statistical tests were used with
caution. The age calculations of all but two samples
pass a Chi2 test at the 95% confidence level,
indicating that the single grain ages for each
sample can be treated as a statistically valid single
population. The pooled age is presented for
those samples. Two samples failed the Chi 2 test
suggesting that the dated grains represent multiple
populations. For these samples the central fission
track age is presented (Galbraith & Laslett 1993)
Interpretation
The 34 samples listed in Table 1 show a spectrum
of ages from 2.2_+1.5Ma to 16.9_+3.6Ma.
Twenty-one of the samples lie within the range
5-8 Ma with a weighted mean age of 6.0 _ 0.3 Ma
which indicates a major cooling event at this time.
The distribution of ages is shown on the map of
northern Papua New Guinea in Fig. 2 showing no
apparent trend to the data. Indeed the ages are
consistent throughout a very large area (c. 500 km
E-W and c. 200 km N-S).
Eight of the samples analysed were collected
from a north-south traverse across the Torricelli
Mountains over a wide spread of elevations. The
data are shown on a smoothed topographic profile
in Fig. 4. Most notable on this traverse is the single
older sample at a higher elevation. Elsewhere in the
Bewani-Torricelli Mountains samples with Late
Miocene fission track ages are present at similar
elevations suggesting relative fault displacement of
hundreds of metres, uplifting the rocks with young
ages (Fig. 4). This is consistent with the 'flower' or
'pop up' structural interpretation presented above.
In addition, the 5-8 Ma apatite fission track ages
throughout the mountains are consistent with
uplift in the Late Miocene-Pliocene, dividing the
previously continuous Sepik and Aitape Basins.
Fission track ages are plotted against elevation in
Fig. 5 in which the concentration of ages between
5-8 Ma is apparent, indicating a period of rapid
cooling during the Late Miocene-Early Pliocene.
As the samples are currently at surface, the authors
interpret this rapid cooling event to be related to
regional uplift and denudation of northern Papua
New Guinea. Samples from the Marum ophiolite
and Adelbert accreted arc record ages of 5 +_ 1 and
8 _ 2 Ma respectively, suggesting that both were
denuded in the Late Miocene (cross-section of Fig.
1 and Fig. 2). Although largely constrained by only
one sample, the elevation vs. age plot in Fig. 5
shows a possible slight trend of increasing fission

Ct
- 2000

- 1500

- ~ ~ ~
-1000 r

- - +tl~....~ 6+2
Sample # from Table I from right to left (prefix 92PNG): 052, 053+054, 0 5 8 , 0 5 9 , 0 6 9 . 0 6 8 , 0 6 1 , 0 6 0 . 3-.~ ~-...~
t I I I ~1 I I , , ' 5 -
1 2 3 4 5 6 7 8 9
142o08'30" DISTANCE (kilometres) 142009'30 ''
3026,00 ,, 3°21 '00"

F i g . 4 . A north-south topographic profile across the Torricelli Mountains with fission track ages and interpreted
thrust faulting. See Fig. 2 for location. The bulk of the ages are under 10 Ma and reiterate the Late Miocene-Early
Pliocene regional cooling event. The single anomalous sample recording an age of 17 Ma could represent an earlier
stage of denudation in the area and is now closely juxtaposed by a younger cooling age due to later faulting. This
interpretation is in general agreement with earlier discussion of thrusting and 'pop-up'/flower structures (Fig. 3).
 TECTONIC EVOLUTION OF NORTHERN PAPUA NEW GUINEA
1200
1000
800
- -- ; ,,"m
I []
600
I D I
m
400
I, []
200
0 '
0 5
Fission Track Age (Ma)
Fig. 5. Fission track age versus elevation plot of all
samples with true recorded elevation, excluding float, all
showing 1 ~ error. The bulk of the samples lie
within 2-10 Ma with a concentration around 5-8 Ma,
which suggests this was the time of maximum cooling
and denudation, due to regional uplift. It is possible that
a slight increase in age occurs with elevation, which is
typical of upthrusted rocks; i.e. samples at the top cool
first thus recording the oldest age which gradually
decreases with depth.
track age with elevation typical of fission track data
sets from mountain ranges. Also, the similarity of
ages (c. 5-8 Ma) from 150-1200 m supports the
hypothesis of rapid cooling at that time. The one
sample from high elevation was probably slightly
below the temperature of total track annealing and
records a mixed age with fission tracks preserved
from both before and after the Late Miocene
cooling event. Rocks at lower elevations were more
deeply buried and hence hotter, with no fission
tracks present prior to uplift and cooling, so record
the age of cooling.
In general, apatite fission track analyses of 34
basement samples from northern Papua New
Guinea demonstrate regional rapid cooling from
temperatures of >100°C in the Late Miocene-
Pliocene. The cooling was due to denudation
caused by regional uplift which resulted from
continent-arc collision starting in the Late
531
Miocene. This collision caused deformation
: "9 !
throughout the Mobile Belt and the Papuan Fold
Belt (cross-section of Fig. 1). Thus the thermal
history of northern Papua New Guinea is entirely
consistent with that of the Papuan Fold Belt to the
south, where Hill & Gleadow (1989) recorded
Pliocene uplift and denudation from apatite fission
track analyses. No evidence for older orogenesis is
I
preserved in the apatite fission track data, so the
higher temperature thermal history must be
assessed by studying the cooling of igneous and
metamorphic rocks.
Thermochronology
Mobile Belt
Page (1976) and Rogerson et al. (1987) presented
K-At and Rb-Sr ages from metamorphic rocks in
northern Papua New Guinea. In the Mobile Belt,
K-At apparent ages of hornblende and biotite are
clustered around 27 and 23 Ma respectively. The
exposed basement around Amanab has Permian
' 's K-Ar biotite ages in strong contrast to the
10 1 20 Oligo-Miocene ages of the Mobile Belt. The horn-
blende and biotite data from the Mobile Belt
(Landslip Range, Fig. 2) indicate relatively
rapid cooling through the temperature interval of
c. 500°C to 300°C (see McDougall & Harrison
1988 for a review of closure temperatures for argon
in these phases) at a rate of_>50°C Ma -1 between 27
and 23 Ma. A single zircon fission track age from
the Karmantina Gneiss, 400 km along-strike, is
consistent with such cooling (Fig. 2 and Table 1).
The zircon fission track age of 18 _+ 1 Ma records
the time of cooling below c. 200-250°C.
These data have been combined with the apatite
fission track data to construct a cooling history for
the metamorphic rocks of the New Guinea Mobile
Belt (Fig. 6). The graph shows rapid cooling from
>500°C to c. 150°C in the Early Miocene and a
second cooling to surface temperatures in the Late
Miocene to Pliocene. Regional subsidence of the
Papua New Guinea margin occurred during the
Middle Miocene with deposition of 1-2 km of
limestone and/or thick volcaniclastic rocks (Home
et al. 1990; Hill et al. 1993). Thus the geology
indicates stable or increasing temperatures in the
Middle Miocene rather than ongoing uplift or
cooling (Fig. 6).
Bewani-Torricelli-Prince Alexander
Mountains
Hutchison & Norvick (1980) reported 36 K-Ar
dates of rocks from the Torricelli Intrusive
Complex. Twelve K-At hornblende analyses on
dioritic-granodioritic rocks yielded Early
532 P . V . C R O W H U R S T ET AL.

Time (Ma)
35 30 25 20 15 10 5 0

D ewani-Torricelli Mountains
100 (Apatite
Track Fission
Age) ~ a Si[iii~!!!ii
i~:~i?
M ~
New Guinea Mobile Belt

~ apid cooling due to

L)
O
200 ( Zircon Fission Track Age) 20 9 9 erosional denudation
caused by arc collision
Rapid cooling possibly
due to extensional --"
¢D 300 tectonics and denudation-" / ~ 5 : : :!B,q~.__ . . . . . . . . .
• 151via [r~-e~r age - olOtlte)
of the metamorphxc core
Ma (K-Ar age - biotite)
,® ?
/ ~ 'A~N~ ~ " ~ C o o l i n g history of metamorphic rocks
/ [ , .,~'~ I from the Prince Alexander Mountains
500 -.--,/~ [Mid-crustal k'K~N,~X7 m
fJ rocks ~ .~p.~
7'----~l>15kmdepth S e . , ~ . . . . . . . . ? 20Ma(K-Arage-hornblende)
37Ma(K-Ar age - hornblende) 27Ma (K-Ar age - hornblende)
(Crystallization Age ?)

Fig. 6. Cooling histories of the mid-crustal rocks now exposed in the Mobile Belt and of the intrusive and
metamorphic rocks from the Bewani-Torricelli-Prince Alexander Mountains constrained by fission track and K-At
dating. K-Ar ages are from Rogerson et al. (1987), Hutchison & Norvick (1980) and Page (1976) as shown in Fig. 2.
Closure temperatures used to generate the curves are: K-Ar hornblende indicates a closure temperature of c. 500°C;
K-Ar biotite indicates a closure temperature of c. 300°C; zircon fission track ages indicate a closure temperature of
c. 200-250°C and apatite fission track ages indicate a closure temperature of c. 120°C.

Oligocene ages of 30-41 Ma, with a mean of
37 Ma. Two other analyses yielded Late Cretaceous
ages. Five K-Ar biotite analyses from these rocks
yielded Early Miocene ages (16-23 Ma). In
contrast, six K-Ar hornblende analyses of
amphibolites from the Prince Alexander Mountains
yielded Early Cretaceous ages (Fig. 2; Hutchison &
Norvick 1980), but six metamorphic rocks from the
same area dated by Rogerson e t al. (1987) yielded
K-Ar biotite ages of 18 Ma and two K-Ar
hornblende ages of 20 Ma. Thus the Torricelli
Intrusive Complex appears to comprise mainly
Oligocene intrusions into Cretaceous or older
rocks, but there may have been a later minor phase
of Early Miocene intrusions and local contact
metamorphism. Alternatively, the Oligocene
intrusions may have resided at temperatures of
c. 400°C until the Early Miocene, below the
blocking temperature of hornblende but above that
of biotite. In the Early Miocene they then cooled
below c. 300°C, giving an Early Miocene K-Ar
biotite cooling age consistent with cooling of the
metamorphic rocks.
These data have been combined with the apatite
fission track analyses to construct cooling histories
for the metamorphic and intrusive rocks of the
Bewani-Torricelli Mountains (Fig. 6). The graph
shows Oligocene crystallization and cooling then
further rapid cooling in the Early Miocene to
temperatures of c. 150°C. The rocks then under-
went cooling to surface temperatures in the latest
Miocene.
Geochemistry
Sixteen surface samples collected from the
Bewani-Torricelli Mountains, consisting of
tonalites, diorites and gabbro were analysed for
major and trace elements. The details of
procedures, petrogenesis, figures, analytical and
petrographic tables are available as Supplementary
Publication No. SUP 18100 (11 pp) from the Society
Library or the British Library Document Supply
Centre, Boston Spa, Wetherby, N. Yorks, LS23
7BQ, UK. Interpretation of the results follows
Pearce et al. (1984) and is outlined below.
 TECTONIC EVOLUTION OF NORTHERN PAPUA NEW GUINEA 533

The results of the geochemical analyses can be Oligocene age, similar to the Adelbert and
summarized as follows: Finisterre (Melanesian) arcs to the east (Jaques
1976). However, the arc rocks appear to be overlain
(1) the major elements (alkalis) suggest a sub-alka-
by the Miocene-Pliocene Sepik and Aitape Basins
lic magma;
which received 'Wogamush' volcaniclastic rocks
(2) the chemistry of clinopyroxenes from mafic
from the Maramuni arc within the Mobile Belt
samples suggest a tholeiitic magma from an
(Fig. 3). Therefore, unlike the Melanesian arc, the
active margin setting;
Bewani-Torricelli arc probably accreted prior to
(3) Nb depletion in the tonalites suggests volcanic
the Miocene or was formed on the northern margin
arc and collision-generated granites;
of the Australian plate with subduction to the south.
(4) trace element plots for tonalite and mafic sam-
The structural and fission track data show that
pies are most akin to those from a tholeiitic to
the present Bewani-Torricelli Mountains were
calc-alkaline volcanic arc;
upthrust in the Late Miocene-Pliocene, such that
(5) MORB normalized plots suggest a subduction-
the Sepik and Aitape Basins were probably
related setting, possibly with partial melting of
continuous in the Early Miocene, implying that the
subducted sediment and fractionation of
Bewani-Torficelli arc was then largely submerged.
clinopyroxenes;
Reflection seismic data from both basins have
(6) although the precise sample ages are currently
been interpreted to show extensional faulting of
unknown, the geochemical trends are consis-
basement (Kugler 1990; Hill et al. 1993).
tent with a simple genetic relationship between
The cooling histories in Fig. 6 show two
the samples.
regional cooling events in northern Papua New
Guinea occurring between 27-18 Ma and also at
Summary of main results c. 6 Ma. The tectonic history of northern Papua
New Guinea must include these two cooling events
1. The geochemical analyses presented show
and account for the nature and amount of
a tholeiitic island-arc signature for the Bewani-
denudation involved. The rapid cooling of >300°C
Torricelli samples and indicate a simple genetic
in the Early Miocene implies c. 10km of
relationship between the samples. Diorites and
denudation assuming normal temperature grad-
granodiorites dated by Hutchison & Norvick (1980)
ients. The Early Miocene cooling ages have been
suggest a Late Eocene to Early Oligocene
interpreted to result from accretion/collision,
crystallization age.
causing subsidence of the New Guinea margin and
2. The Bewani-Torricelli-Prince Alexander
creation of the Sepik-Ramu basins as foreland
Mountains are interpreted as positive flower
Basins (Pigram & Davies 1987; Struckmeyer et al.
structures that owe their present relief to left-lateral
1993). However, such rapid cooling is unusual in
transpressive motion along the Bewani-Torricelli
compressional orogenies and, as the metamorphic
fault zone, during the Late Miocene-Pliocene. This
rocks cover thousands of square kilometres, the
is consistent with the time of accretion of the
metamorphic detritus would have inundated
Melanesian arc.
adjacent basins. To the contrary, the Sepik Basin
3. Apatite fission track analysis shows regional
was a deep and probably starved basin with
cooling of northern Papua New Guinea from
marginal carbonates in the Early Miocene (e.g.
temperatures of > 100°C during the Late Miocene-
Doust 1990), despite having Early Miocene meta-
Pliocene. This indicates regional denudation
morphic terrains on both sides. Regionally, in the
contemporaneous with compressional deformation
Mobile Belt and northern Fold Belt the inundation
and Melanesian arc collision.
of sediment was in the Middle Miocene and
4. K-Ar thermochronology suggests rapid
comprised mainly volcanogenic material from the
cooling of metamorphic rocks in the Early
Maramuni arc (Hill et al. 1990, 1993).
Miocene, from >500°C to c. 150°C. The cooling
Therefore, it is inferred that the rapid Early
occurred from 25-20 Ma in the Mobile Belt and
Miocene cooling resulted from tectonic denudation
from 20-18 Ma in the Prince Alexander Mountains.
during regional extension, which also created the
In both cases these medium to high grade
Sepik graben and regional subsidence of the Papua
metamorphic rocks are in close proximity to
New Guinea margin (Hill et al. 1993). During
similar rocks with Cretaceous to Permian K-At
extension the upper crust was pulled off the lower
cooling ages.
crust, such that hot lower crustal rocks rose
rapidly towards the surface as metamorphic core
complexes. However, these lower crustal rocks did
Discussion
not reach the sur-face, but remained covered by up
The interpreted tholeiitic island-arc of the to 5 km veneer of extended upper crustal rocks,
Bewani-Torricelli Ranges is probably of Early with low relief, causing little denudation. The
534 r'. v. CROWHURSTET AL.

Cretaceous to Permian K-Ar cooling ages for
adjacent metamorphic rocks around Amanab and in
the Prince Alexander Mountains suggest that these
areas remained part of the upper crust during
extension, and therefore show no Tertiary cooling.
Following arc-continent collision in the Late
Miocene, the rocks were thrusted, denuded and
cooled giving the 6 _+2 Ma apatite fission track
cooling ages.
The metamorphic rocks of the Mobile Belt are in
thrust contact with the unmetamorphosed April
ophiolite (Fig. 2) which Rogerson et al. (1987)
interpret as the uppermost thrust sheet cropping out
in mountain tops. The ophiolite is probably
Palaeogene in age (Rogerson et al. 1987) and
outcrops occur over thousands of square kilo-
metres, indicating that it was previously a
significant area of oceanic crust. Like Rogerson
et al. (1987) it is inferred that the ophiolite was
thrust over the metamorphic rocks in the Late
Miocene-Pliocene compressional deformation,
consistent with the regional apatite fission track
cooling ages and particularly the 5 Ma age for the
Marum ophiolite. Therefore, prior to the Late
Miocene there must have been a significant area of
oceanic crust adjacent to the Mobile Belt and
separating it from the Sepik Basin and Bewani-
Torricelli arc. Such a marginal basin must be
included in any tectonic model. Note that an
alternative model under consideration is that
denudation of the metamorphic rocks could be due
to sinistral transpression along wrench faults,
similar to the Alpine fault of New Zealand. The
detritus could have been deposited on the marginal
basin and then been eroded during Late Miocene
thrusting. However, an extensional model is
preferred due to the rapid cooling rates, the
apparently extensional graben and the regional
subsidence of the entire margin.
Tectonic model
As shown in this paper and elsewhere, the factual
geologic data for northern Papua New Guinea
are limited, so that any model is necessarily
speculative. However, these new data further
constrain tectonic models, so we present a
model that is consistent with the present dataset,
illustrated in Fig. 7.
Oligocene
In the Early Oligocene the New Guinea margin
comprised continental crust as far north as the
Mobile Belt, then a narrow marginal basin of
Palaeogene oceanic crust bound by a ribbon of
extended Cretaceous continental and oceanic crust.
The marginal basin probably opened in the Late
Cretaceous-Palaeocene, contemporaneous with
Coral Sea spreading, as suggested by Pigram &
Symonds (1991). Oceanic crust north of New
Guinea was subducted to the north beneath an
Oligocene tholeiitic volcanic arc, but slow sub-
duction may also have occurred beneath New
Guinea (Fig. 7; Hill et al. 1993) as suggested by
Upper Oligocene-Lower Miocene intrusive rocks
and Wogamush Volcanics in the Sepik Headwaters
region (Rogerson et al. 1987).
By the Late Oligocene, the arc had accreted to
the ribbon of possible continental crust along the
margin of New Guinea and had been partially sub-
ducted such that temperatures remained high. A
new south-dipping subducting slab developed, with
the trench to the north of the accreted arc.
Early Miocene
Extension occurred in the New Guinea margin
above the newly developed slab, probably due to
roll-back of the slab hinge. The extension was
focused in the relatively thicker and weaker
continental crust and occurred along low angle
faults through the crust (Fig. 7). As the upper crust
was pulled off, the lower crust rose to within a few
kilometres of the surface and cooled rapidly as a
metamorphic core complex. This occurred in the
Mobile Belt from c. 27-20 Ma and in the accreted
arc (Prince Alexander Mountains) from 20-18 Ma.
Adjacent rocks in the upper crust retained their
Mesozoic or older age.
As part of the regional extension, the ribbon of
possible continental crust was extended to create

Fig. 7. Sequential cross-sections illustrating the Early Miocene extensional model for the evolution of northern Papua
New Guinea (refer to D-D' on Fig. 1 for location). The Late Eocene passive margin of Papua New Guinea was
disrupted by the onset of rapid convergence between the Australian and Pacific plates. The resultant subduction
remote from Papua New Guinea, created the Caroline backarc basins and the Finisterre-Adelbert arcs. In the Late
Oligocene accretion of arc material jammed the subducting plate, thus producing another southerly dipping
subduction zone further to the north. Extension occurred in the continental crust above the newly formed slab during
the Early Miocene, due to either roll-back of the slab hinge or slowing of convergence following collision of the
Indo-Australian plate with Asia to build the Himalaya. During the Middle Miocene, partial melting of the subducting
slab created the Maramuni arc and saw the onset of obduction of oceanic crust. The Pliocene saw the onset of major
compression which produced large-scale thrusting and deformation in the Mobile Belt and caused the upthrust of the
Bewani-Torricelli Mountains, which now separate the Sepik and Aitape Basins.
 TECTONIC EVOLUTION OF NORTHERN PAPUA NEW GUINEA 535

D D
Early Oligocene Late Eocene I

P a p u a N e w Guinea Pacific Plate

Island arc
I Palaeogene
marginal basin
(Sontlnental crust
. . . . . . . . . . . . . . . . . . . . . .

.... Mantle .... y

. . . . . . . . . . . . . . . .

diorites
I ,
granodiorite
i?

Late Oligocene
ccreted arc
Palaeogene
marginal basin

N e w subduction
.-?
zone forming

Early Miocene Prince Alexander

Sepik Graben
'metamorDhic rocks
27-20 Ma 20-18 Ma

Caroline
Incipient low-angle / Plate
detached normal fault Rollback
Metamorphic core
complex

Middle Miocene
Maramuni Arc Incipient
/ obduction Volcaniclastics
/-- \

Pliocene
N e w Guinea
Papuan M o b i l e Belt
Sepik Bewani-Torricelli
Fold
Obducted ophiolite Basin Mountains
Belt
_/ \ _ / Aitape Basin
536 P.V. CROWHURST ET AL.
the Sepik and Aitape Basins, which were con-
tinuous at that time. There was little uplift and
erosion, so the basins were starved of sediment and
carbonates were deposited around the margins
(Wilson et al. 1993). The regional extension also
resulted in Miocene subsidence of the whole New
Guinea margin leading to deposition of thick
Miocene carbonates.
Middle Miocene
Partial melting occurred above the south-dipping
subducting slab resulting in extensive intrusions
and volcanism to create the Maramuni arc.
Abundant volcanogenic sediment was shed into the
Sepik and Aitape Basins and into the deep basins
abutting the Miocene platform carbonates to the
south (Hill et al. 1990).
Late Miocene-Pliocene
Accretion of the Melanesian arc in eastern New
Guinea and attempted subduction of the young and
buoyant Caroline plate (Oligocene; Hegarty &
Weissel 1988) placed the New Guinea margin into
References
ABERS, G. & MCCAFFREY,R. 1988. Active deformation in
the New Guinea fold-and thrust-belt: seismological
evidence for strike-slip faulting and basement
involved thrusting. Journal of Geophysical
Research, 93, 13 332-13 354.
DOUST, H. 1990. Geology of the Sepik Basin, Papua New
Guinea. In: CARMAN,G. J & CARMAN,Z. (eds)
Petroleum Exploration in PNG: Proceedings of the
First PNG Petroleum Convention, Port Moresby.
PNG Chamber of Mines and Petroleum, Port
Moresby, 461-478.
Dow, D. B. 1977. A geological synthesis of Papua New
Guinea. Bureau of Mineral Resources Australia
Bulletin, 201.
FOSTER, D. A. & GLEADOW,A. J. W. 1992. The morpho-
tectonic evolution of rift-margin mountains in
central Kenya: Constraints from apatite fission-
track thermochronology. Earth and Planetary
Science Letters, ll3, 157-171.
FRANOS, G. 1990. The North New Guinea Basin and
associated infra-basins. In: CARMAN, G. J &
CARMAN, Z. (eds) Petroleum Exploration in PNG:
Proceedings of the First PNG Petroleum
Convention, Port Moresby. PNG Chamber of Mines
and Petroleum, Port Moresby, 445-460.
GALBRA1TH, R. E ~: LASLETT, G. M. 1993. Statistical
models for mixed fission track ages. Nuclear Tracks
and Radiation Measurements, 21, 459-470.
GLEADOW, A. J. W. & LOVERING,J. E 1978. Thermal
history of granitic rocks from western Victoria: a
compression. In the Sepik area the compression
was initially taken up by obduction of a sliver of
oceanic crust from the Palaeogene marginal basin,
thrust over the Lower Miocene metamorphic rocks
of the Mobile Belt. Continued compression caused
more intense thrusting throughout northem and
central New Guinea, resulting in regional uplift,
denudation and cooling from 8-5 MR. The
deformation propagated south into the Papuan
Fold Belt in the Pliocene and the Sepik Basin and
particularly the Ramu Basin underwent regional
subsidence as foreland basins to the Melanesian
arc. The oblique plate convergence was partitioned
into transpression along the northern margin
causing uplift, denudation and cooling of the
Bewani-Torricelli-Prince Alexander mountains
and N E - S W compression in the Mobile Belt and
Fold Belt to the south (Abets & McCaffrey 1988).
This work has been sponsored by LL&E Sepik Pty
Limited, Mobil New Exploration Ventures Company,
Mobil Exploration Niugini and Esso Australia Ltd.
Fission track neutron irradiations are supported by a grant
from the Australian Institute of Nuclear Science and
Engineering. The views presented here are entirely those
of the authors and do not necessarily correspond to those
of any company.
fission track dating study. Journal of the Geological
Society of Australia, 25, 323-340.
GREEN, P. E 1981. A new look at statistics in fission track
dating. Nuclear Tracks, 5, 77-86.
HEGARTY, K. A. & WEISSEL, J. K. 1988. Complexities
in the development of the Caroline Hate region,
western equatorial Pacific. In: NAIRN, A. E. M.,
STEHLI, E G. & UYEDA, S. (eds) The Ocean Basins
and Margins 7B, The Pacific Ocean. Plenum Press,
New York and London, 277-301.
HILL, K. C. 1991. Structure of the Papuan Fold Belt,
Papua New Guinea. AAPG Bulletin, 75, 857-872.
& GLEADOW, A. J. W. 1989. Uplift and
thermal history of the Papuan Fold Belt, Papua New
Guinea: Apatite Fission Track Analysis. Australian
Journal of Earth Sciences, 36, 515-539.
- - , GREYA., FOSTER,D. A. 8z BARRETTR. A. 1993. An
alternative model for the Oligo-Miocene evolution
of northern PNG and the Sepik-Ramu Basins. In:
CARMAN, G. J & CARMAN, Z. (eds) Petroleum
Exploration in PNG: Proceedings of the First PNG
Petroleum Convention, Port Moresby. PNG
Chamber of Mines and Petroleum, Port Moresby,
241-259.
, MEDD, D. & DARVALL,P. 1990. Structure, strati-
graphy, geochemistry and hydrocarbons in the
Kagua-Kubor area, Papua New Guinea. In:
CARMAN, G. J & CARMAN, Z. (eds) Petroleum
Exploration in PNG: Proceedings of the First PNG
Petroleum Convention, Port Moresby. PNG
 TECTONIC EVOLUTION OF NORTHERN PAPUA NEW GUINEA
Chamber of Mines and Petroleum, Port Moresby,
351-366.
HtLYARD, D. B., FORD, C. & MCDONALD S. 1994. Aitape
Basin, PPL 150, Papua New Guinea: A report for
LL&E Sepik Pty Ltd. Ford Geoconsultancy Pty Ltd.
HOME, P. C., DALTON, D. G. AND BRANNAN, J. 1990.
Geological evolution of Western Papuan Basin. In:
CARMAN, G. J & CARMAN, Z. (eds) Petroleum
Exploration in PNG: Proceedings of the First PNG
Petroleum Convention, Port Moresby. PNG
Chamber of Mines and Petroleum, Port Moresby,
107-118.
HURFORD, A. J. & GREEN, P. E 1982. A users' guide to
fission track dating calibration. Earth and Planetary
Science Letters, 59, 343-354.
HtrrCHISON, D. S. & NORVICKM. S. 1980. Geology of the
North Sepik region, Papua New Guinea. BMR
Record 1980/24.
JAQUES, A. L. 1976. High K20 island-arc volcanic rocks
from the Adelbert Ranges and Finisterre Ranges,
northern Papua New Guinea. GEOLOGICALSOCIETY
OF AMERICABULLETrN,87, 861--867.
KUGLER, A. JR. 1990. Geology and petroleum plays of the
Aitape Basin, New Guinea. In: CARMAN, G. J &
CARMAN, Z. (eds) Petroleum Exploration in PNG:
Proceedings of the First PNG Petroleum
Convention, Port Moresby. PNG Chamber of Mines
and Petroleum, Port Moresby, 479-490.
McDOUGALL,I. & HARRISON,T. M. 1988. Geochronology
and thermochronology by the 4°Ar/39Ar method.
Oxford University Press. New York.
PAGE, R. W. 1976. Geochronology of igneous and meta-
morphic rocks in the New Guinea Highlands.
Australian Bureau of Mineral Resources, Bulletin,
162.
537
PEARCE, J. A., HARRIS, N. B. W. & TINDLE, A. G. 1984.
Trace element discrimination diagrams for the tec-
tonic interpretation of granitic rocks. Journal of
Petrology, 25,956-983.
PIGRAM C. J. & DAVIES H. L. 1987. Terranes and the
accretion history of the New Guinea orogen. BMR
Journal of Australian Geology and Geophysics, 10,
193-211.
- -& Symonds, P .A. 1991. A review of the timing of
the major tectonic events in the New Guinea
Orogen. Journal of Southeast Asian Sciences, 6,
307-318.
ROGERSON, R. J., HILYARD, D. B., FINLAYSON, E. J.,
HOLLAND, D. J., NION, S. Z. S. ET AL. 1987. The
geology and mineral resources of the Sepik
Headwaters region, Papua New Guinea. Geological
Survey of Papua New Guinea, Memoir, 12.
STRUCKMEYER,n. I. M., YEUNG, M. & PIGRAM, C. J.
1993. Mesozoic to Cainozoic plate tectonic and
palaeogeographic evolution of the New Guinea
Region. In: CARMAN, G. J • CARMAN, Z. (eds)
Petroleum Exploration in PNG: Proceedings of the
First PNG Petroleum Convention, Port Moresby.
PNG Chamber of Mines and Petroleum, Port
Moresby, 261-290.
WILSON, A. C., BARRETT, R., HOWE, R. & LEU, L. K.
1993. Occurrence and character of outcropping
limestones in the Sepik Basin: implications for
hydrocarbon exploration. In: CARMAN, G. J &
CARMAN, Z. (eds) Petroleum Exploration in PNG:
Proceedings of the First PNG Petroleum
Convention, Port Moresby. PNG Chamber of Mines
and Petroleum, Port Moresby, 111-124.
 Gondwana origin of the Baoshan and Tengchong terranes of
west Yunnan

HELMUT WOPFNER
Geologisches Institut, Universitdt zu Kgln Ziilpicherstrasse 49, D-50674 K61n, Germany

Abstract: Upper Palaeozoic glaciomarine deposits combined with the presence of cold-water
faunas and Glossopterisindicate a Gondwana provenance of the Baoshan and Tengchong Blocks
in western Yunnan. The glacial origin of these Permo-Carboniferous deposits is verified by a
three-fold subdivision of the basal sequence comprising, in ascending order: diamictites with
faceted and striated clasts, laminites with dropstones, and a black anaerobic de-glaciation facies.
The succeeding carbonates of the upper Permian and Lower Triassic originated during a post-
glacial phase of a much warmer climate. Although the two blocks are adjacent to each other their
Upper Palaeozoic sequences show significant differences. The glacigene successions of the
Baoshan Block are comparatively thin and are overlain by thick basalts and red beds. On the
Tengchong Block the glacigene marine deposits exceed 1000 m and are followed by thick reefal
limestones of Lower Permian age. An active volcanic rift setting is suggested for the Baoshan
Block and a proximal passive margin environment at a somewhat lower latitude for the
Tengchong Block. The succession of the latter is comparable with Permo-Carboniferous sections
in NW Australia, especially that of the Bonaparte Gulf basin, and it is assigned to the Sibumasu
tectono-stratigraphic unit. The Baoshan sequence shows similarities to sequences of Tibet and
NE India and is therefore assigned to the Tibetan realm. Both terranes separated from Gondwana
in the late Early Permian. Docking commenced in the Late Triassic concomitant with the closure
of the Changning-Menglian Belt. Lateral displacements in the course of the Himalayan orogeny
moved the Tengchong Block north, bringing it into juxtaposition with the western margin of the
Baoshan Block. This tectonic contact is now the Nujiang Line.

In southwest China the east-west tectonic trends of
the Himalaya turn sharply into a north-south
direction. This direction is maintained essentially
throughout Burma, Laos and Thailand, up to the
southern tip of the Malay Peninsula. In western
Yunnan the latitude-parallel structural elements
control the paths of the three great rivers, the
Yangtze, Mekong (Lancangjiang) and Salween
(Nujiang). South of latitude 26°N some southeast-
trending tectonic elements branch off the
north-south alignment. This is reflected by the
course of the Red River (Yuanjiang) and the
middle and lower reaches of the Mekong. Both
north-south and southeast-striking tectonic fabrics
resulted from the collision of the Indian plate with
Laurasia in Tertiary times, thereby masking a
number of structural features of earlier tectonic
events, especially the original boundary between
detached G o n d w a n a fragments and Laurasia
(Cathaysia).
Within the area of the Three River region of SW
China and the adjoining Golden Triangle in Burma
and Laos, a number of sutures have been suggested
which supposedly form the boundary between
Gondwana and Laurasia. These concepts, which
have been summarized by Jin Xiaochi (1994),
were developed either from palaeontological
and/or palaeobotanical evidence or from field
observations, as for instance the presence of rocks
like serpentinites or ophiolites, or diagnostic facies
associations such as banded cherts. An important,
but hitherto not unequivocally accepted piece of
evidence to determine the position of the boundary
is the presence of Upper Palaeozoic deposits of
presumed glaciomarine origin, generally referred to
as pebbly mudstones, as found in two structural
units in western Yunnan. Recent investigations in
that region carried out by the author and by Jin
Xiaochi (1994) have substantially enhanced the
probability that these pebbly mudstones are indeed
of glacial origin and hence originated from part of
Gondwana (Wopfner & Jin 1993; Jin Xiaochi
1994).
Regional overview
Southwestern Yunnan consists of six major
tectonic provinces which are shown in Fig. 1. These
comprise the Yangtze Platform which, in middle to
late Palaeozoic times was clearly part of Cathaysia,
although its Upper Proterozoic and Cambrian
successions show certain affinities to Gondwana.
To the west and southwest of the Yangtze Platform
are the L a n p i n g - S i m a o Fold System and the

From Hall, R. & Blundell, D. (eds), 1996, TectonicEvolution of SoutheastAsia, 539

Geological Society Special Publication No. 106, pp. 539-547.
540
g
BURMA
o 50 100
I I I
kilometres
~8= loo,
Fig. 1. Schematic map of the major tectonic provinces of southwestern Yunnan.
Changning-Menglian Belt. Both comprise folded
sedimentary and some volcanic rocks ranging in
age from Silurian to Lower Triassic. Within the
Lanping-Simao Fold Belt these older rocks are
covered by a thick succession of little-deformed red
beds of Lower Jurassic to Eocene age. In contrast,
the Palaeozoic succession of the Changning-
Menglian Belt is fairly well exposed, including
thick successions of banded cherts and basic
volcanics (Liu Benpei et al. 1991). The Ailaoshan
Belt which is wedged between the Yangtze
Platform and the Lanping-Simao Fold System
consists mainly of migmatites and other high grade
metamorphic rocks and some Palaeozoic and
Mesozoic cover. Mylonites attest to intense sinistral
shearing in mid-Tertiary times (Tapponnier et al.
1990). The westernmost parts of Yunnan are
formed by the Baoshan Block and the Tengchong
Block. These two blocks host Upper Palaeozoic
marine successions with diamictites and pebbly
mudstones which are interpreted to be of glacigenic
origin.
The Baoshan Block is built up substantially by
sedimentary and some volcanic rocks, ranging in
age from (?) Upper Proterozoic to Cretaceous. An
angular unconformity exists between Visean
carbonates of the Pumenqian Formation and older
rocks and the Upper Carboniferous/Permian
H. WOPFNER
26"
gtze Platform
/ S o u t h China
Fold System
\ VIETNAM
22-
,} L A O S '"
/ ~ , .,,I,-,.L.i.~ 'lo21 104
J
Dingjiazhai Formation, the basal unit of the
Pangaea sequence (Fig. 2). Its base contains
diamictites and other coarse-grained deposits of
presumed glaciomarine origin. The Dingjiazhai
Formation is succeeded by basalts, red beds and
thick carbonates which extend into the Ladinian.
The presence of bauxites above the Ladinian
carbonates indicates at least a temporary interval of
subaerial weathering. Mildly deformed middle
Jurassic red beds unconformably overlie the
Pangaea sequence in some places. These red beds
represent the molasse stage of the Indosinian
movements.
The Tengchong Block adjoins the Baoshan
Block in the west and is separated from it by the
Nujiang tectonic line. By far the largest portion of
the block is built up by the Gaoligongshan
Metamorphics which also form the mountain range
of the Gaoligongshan, immediately west of the
valley of the Nujiang. The metamorphic rocks,
which form a large synform, consist mainly of
gneisses, granulites, mylonites and some marbles.
They are of Middle Proterozoic age. The sedi-
mentary cover comprises some Lower Devonian
siliciclastic rocks and dolomites succeeded by the
Pangaea sequence, commencing with the Upper
Carboniferous-Permian Menghong Group. This
sequence comprises the thick sandstone-siltstone
 BAOSHAN AND TENGCHONG TERRANES, W Y U N N A N 541

TENGCHONG BONAPARTEGULF BASIN

NW AUSTRALIA
' ~ Yanzipo
Limestone ---~ TENGCHONG
T RI A S S I C TERRANE
intrusive
3 granite
Damuchang ~ ~ Lote
Formation ~ Permian Dodongchong
- " ~ Formotion
BAOSHAN I ~ ~ ~ j
i.'.'.:.'i" 'J K--~yting Formotion
Artinskion
andesite Siguaping Tr--~ocheryShole ~ block
2- ~ ] bauxite 2 ., ,..i ii I Formation ---- __--__
Assetion/Sokm. ~ lutite
Illl:lll'I II
L~I I~!~!~-.-~'~)!~ o

~~ ~ Kongshuhe
./ // ¢ Hewanjie ~ ~,~J Assetion Formotion
~ ~.
~\// Formation

i.It~.i.~~.i,~i[ ~~ ShazipoFm. [ ~ I Luogengdi

Formation '~.'-.'.'- i Stephonion
1- - - - - ~ YongdeFm. 1
V V v V~
v W ~ Woniusi ~ ~ l ~
Formation Bangdu NamurionOml
-,~Eorly Zishi
Formotion
v v v v v v v? Formation
Dingjiazhai
Fomation
0 0
no base No bose
PumenqianFm. exposed
( Visean)
Fig. 3. The Pangaea sequence of the northern
Tengchong Block and that of the distal Bonaparte Gulf
Fig. 2. Lithostratigraphy of the Upper Palaeozoic basin in northwestern Australia.
successions (Pangaea Sequence) of the Baoshan and
Tengchong Blocks.

as Gondwana-derived terranes. The question is, just

unit of the Bangdu Formation at the base,
succeeded by glaciomarine deposits. Within the
southern Tengchong Block the latter are termed, in
ascending order, Luogengdi, Siguangping and
Damuchang Formations (Fig. 2). In the northern
part, near the Burmese border, the basal unit termed
the Zishi Formation consists entirely of thick sand-
stones, and the glaciomarine succession is referred
to as the Kongshuhe Formation. The succession of
the northern Tengchong Block is shown in Fig. 3.
The glaciomarine beds are overlain by thick
coralline limestones of Permian age. The whole
region has been intruded by Indosinian granites,
some of which are tin-bearing. They form large
stocks and make up about 15% of the outcrop area.
The youngest rocks are Miocene or Holocene
basaltic lava flows, pyroclastic deposits and
perfectly preserved volcanic cones (Bureau of
Geology Yunnan 1990).
Glacial deposits or mud flows?
how strong is the evidence for a glacial origin of
these rocks and can it withstand scrutiny?
In a marine environment it is much more difficult
to prove sediment transport by glaciers than on
land, where lodgement tills, and glaciated pave-
ments with multi-directional striations and pressure
gouges can provide unequivocal proof for the
action of glaciers. The glacial load shed into a
marine realm is much more subject to reworking
by currents, slurry transport and other mass-
movements which will obscure the original glacial
signature. Thus single indicators on their own may
not be conclusive. Only a combination of features
viewed in context of their interrelationship will be
able to provide convincing evidence.
A glacigene sequence is characterized by a
general three-fold subdivision reflecting, in
ascending order, the final glacial high-stand, the
periglacial stage and finally the deglaciation stage
(Wopfner & Diekmann 1992; Diekmann 1993).
The rock units which correspond to these stages
are:

The Pangaea sequence of both the Baoshan and the
Tengchong Blocks contains within its basal rock
units diamictites, pebbly mudstones and other
ill-sorted, rudaceous deposits which have been
interpreted to be of glaciomarine origin. A glacial
character of these rocks would identify both blocks
(1) diamictites and other rudaceous deposits both
in autochthonous position and as slumped and
slurried re-deposited masses (Dowdeswell &
Scourse 1990);
(2) graded laminates with dropstones and
inter-calated, usually sandy, turbidites and
542 I-I. WOPVN~g

dark-coloured siltstones, containing a typical
cold-water assemblage of bryozoa and
mollusca (bry-mol facies), as observed not
only in many Lower Permian glaciomarine
successions but also in Pleistocene/Holocene
arctic and subarctic environments (Henrich
et al. 1992);
(3) black, pyritic lutites with high organic contents,
carbonate concretions and sulphates (Wopfner
1994).
Well-developed diamictites are present in the
lower parts of the Dingjiazhai Formation as well
as in the Kongshuhe Formation and its southern
equivalents. They are generally massive diamictites
with angular to subangular clasts, supported in a
dark grey to dark greenish-grey lutite matrix. The
composition of the latter reflects the lithology of
the source area scraped by the glaciers. Despite its
fine-grained nature it contains few primary clay
minerals. Within the Baoshan Block, where the
glaciers traversed Lower Carboniferous limestones,
the matrix is calcareous, whereas in the diamictites
of the Tengchong Block, which were derived
mainly from metamorphic rocks, the matrix
resembles a ground-up arkose (Jin 1994).
The clast sizes vary from an average of about
20 cm in the lower Kongshuhe Formation to about
6-10 cm in the Dingjiazhai Formation (Fig. 4a).
Generally there is a dominance of trapezoid or
pentagonal-shaped, subangular to subrounded
clasts, some facetted or soled due to grinding
during intra- or sub-glacial transport. Finer grained
diamictites with an average clast size of 1-1.5 cm
and chaotic texture are interpreted as distal slump
masses. They contain sporadic clasts up to 20 cm,

j. : >,

.¢ ....
+ ; •

7
7
Fig. 4. Features indicating a glacial origin of the Dingjiazhai Formation (Baoshan Block) and Kongshuhe Formation
(Tengchong Block): (a) massive diamictite from section along Burma Road, about 20 km SW of Baoshan, western
Baoshan Block; (b) slumped mass of fine-grained diamictite with faceted and faintly striated rain-out clasts,
Qingshuigou Section, central Baoshan Block; (c) graded laminates with cluster of drop stones, Liuku Section,
northern Baoshan Block; (d) siderite concretions in black, pyritic lutites, Kongshuhe Section, northern Tengchong
Block. Scales, including those on hammer, in centimetres; diameter of coin is 21 ram.
 BAOSHAN AND TENGCHONG TERRANES, W YUNNAN
some of them well faceted and with faint striations.
They represent rain-out clasts, material which was
ice-rafted into more distal basin positions and
released upon melting, which accounts for the
preservation of some of the glacial signatures
(Fig. 4b).
The heavy mineral fraction of the diamictites and
associated deposits contains large amounts of
garnet and other unstable minerals, making up
50-75% of the heavy mineral spectrum in some
diamictites of the Kongshuhe Formation (Fig. 5).
Garnet is resistant to mechanical abrasion but
susceptible to chemical attack, thus its dominance
within the heavy mineral spectrum is a typical
feature of glacial deposits (Hamilton & Krinsley
1967; Gravenor 1979; Diekmann 1993).
Other features indicative of a glacial or
periglacial origin are rhythmically bedded dark-
grey siltstones with lonestones, measuring up to
25 cm, and graded laminates with dropstones as
depicted in Fig. 4c. These laminates consist of
regular couplets and may represent seasonal
deposits. Locally, the periglacial siltstones contain
an abundance of bryozoa and brachiopoda, a fossil
assemblage regarded as typical of a cold-water
marine environment (cf. Shi & Archbold 1993).
The black lutite facies forms the third member of
a glacigene sequence. It reflects depositional and
climatic conditions during the final deglaciation.
KSH4
D Garnet = 75.0
[] Tourmaline = 11.0
[] Zircon = 5.0
[] Opaque = 8.0
• Others = 1.0
KSH
[] Garnet = 49.0
[] Tourmaline = 7.9
[] Zircon = 6.3
[] Opaque = 36.4
• Epidote = 0.5
Fig. 5. Heavy-mineral spectrum of the basal diamictite
(KSH 4) and a diamictite from the lower third (KHS) of
the type section of the Kongshuhe Formation, showing
very high garnet concentrations (from Jin Xiaochi 1994).
543
The rocks consist of black, kaolinitic and pyritic,
generally massive or faintly bedded shales or
siltstones. They are anaerobic sediments which
characteristically show intermediate to high
contents of organic carbon, generally with a
pronounced participation of algal kerogen.
Commonly, large carbonate concretions are present
in the lower parts of this facies (Fig. 4d). In the case
of Yunnan, the concretions consist of siderite (Jin
1994). Gypsum and baryte may also be present.
This facies has a very wide distribution. Besides
Yunnan it has been recognized in South and East
Africa, Oman, northern India and northwestern
Australia (Wopfner 1994).
Table 1 shows a compilation of the most
common features typical of glacial deposits. From
the evidence presented here the author is very
confident that the successions of both the
Dingjiazhai Formation and of the Kongshuhe
Formation and equivalents represent glacial,
periglacial and postglacial deposits. During the
Late Carboniferous/Early Permian, the Baoshan
and Tengchong Blocks must have been in a position
where they could be traversed by glaciers of the
Gondwana glaciation. Thus they are terranes which
were separated from Gondwana after the glaciation.
Morphotectonics
Although the Baoshan and Tengchong Blocks are
in direct contact with each other today, their
Permo-Carboniferous successions are quite
different, which is apparent from a comparison of
the two stratigraphic columns of Fig. 2.
The glacial beds of the Baoshan Block are
relatively thin and, especially the periglacial units,
are quite fossiliferous. The glacigene succession is
overlain by thick basalts of the Woniusi Formation,
Table 1, Features of glacial deposits which have been
observed within the basal Pangaea Sequence of Baoshan
and Tengchong Blocks in western Yunnan are indicated
by crosses
Baoshan Tengchong
diamictites with source matrix × x
glaciated pavements 0 0
pentagonal clasts × ×
striated clasts × 0
faceted clasts x x
graded laminates x x
dropstones x 0
slurried diamictite x 0
slump masses with rain-out clasts x x
high content of garnet x x
bry/mol biofacies x x
deglaciation lutites x x
544 H. WOPFNER
followed by red beds of the Yongde Formation and
Artinskian/Kungurian carbonates of the Shazipo
Formation. Geochemical data indicate that the
basalts of the Woniusi Formation are of intra-
cratonic origin (Liu Benpei, pers. comm. 1992).
The glacigene succession of the Tengchong
Block (Fig. 3) on the other hand, may exceed
1000 m, indicating a much higher sediment
accumulation rate. The black lutites of the
deglaciation phase exhibit a well-developed,
fossiliferous transition zone to the overlying Lower
Permian coralline carbonates. Volcanic rocks are
completely absent and there is no trace of red beds,
suggesting a more distal environment than that of
the Baoshan Block.
From the evidence available so far, the author
suggests that the Pangaea sequence of the Baoshan
Block originated in a volcanic rift setting which,
during the Late Permian and Early Triassic, was
transformed by crustal foundering to a restricted
carbonate platform, indicated by the Lower Triassic
dolomites and bituminous limestones of the
Hewanjie Formation (Fig. 2). For the succession of
the Tengchong Block a passive margin setting is
suggested, probably distal shelf to upper slope,
which subsequently developed into a carbonate
platform.
Origin of Yunnan terranes
In view of the environmental interpretation of the
Pangaea sequences of the Baoshan and Tengchong
Blocks, their original position within Gondwana
may be deduced by identifying comparable
morphotectonic settings with similar lithostrati-
graphy at the margin of known Gondwana
fragments.
Permian volcanic rift structures have been
identified by Gaetani & Garzanti (1991 ) in Kashmir
and Zanskar, at the northwestern margin of the
Indian Gondwana fragment, and a similar morpho-
tectonic situation was also suggested by Acharyya
(1987). Figure 6 shows a comparison between the
succession of the Baoshan Block with that of
Kashmir. There, a comparatively thin glacigene
sequence of Asselian age rests unconformably on
the mid-Carboniferous Fenestella Shale. The
unconformity reflects disturbances of the
Hercynian orogenic phase, also recognized in the
Lahul-Spiti district (Kanwar & Ahluwalia 1979).
The Agglomeratic Slate which conformably
overlies the glacigene sequence contains an
Artinskian faunal assemblage. It is followed by the
Nishatbag Formation with Gangamopteris and
Glossopteris. The thick basaltic to andesitic rocks
of the Panjal Traps which form the major part of the
Pangaea sequence of Kashmir, are of Kazanian to
Lower Tatarian age (Murgabian-Midian; Stampfli
KASHMIR-
NW-INDIA
Daonelta BAOSHAN
ShaLe TERRANE
-Khunamuh
Formation _ _ ~
ZewanFro. Ar~
ShazipoFm.
VvVvV
v vv Panjar YonggdeFro.
Trops
VvVvVv~
500~ Woniusi
Formation
Aggtomera[ s°°l ilv~-~v~
Sokm.~
Beds Assel v v v
Assel ~'~'~ Dingjiazhai
Fenestetta om -:' Formation
vis. ~
i!ikPumenq ian
Shore
Formation
Fig. 6. Comparison between the sedimentary sequence
of the Baoshan Block with the Pangaea succession of
Kashmir (after Gaetani & Garzanti 1991).
et al. 1991). They are thus younger than the basalts
of the Woniusi Formation and apparently reflect the
separation stage of the Gondwana terranes from the
supercontinent. The basalts of the Woniusi
Formation, on the other hand, were associated with
the inception of rifting.
Gondwana successions in NE India and Bhutan,
as for instance in the Rangit Valley tectonic
window in Sikkim (Sinha Roy 1973) show many
similarities with the Baoshan succession. The
three-fold subdivision of diamictites-laminites-
black lutites is quite clearly developed, and at least
the upper part of the succession was deposited in a
marine environment (Singh 1987). Amygdaloidal
basalts of up to 3 m thickness occur together with
the 'pebble-slate' (Sinha Roy 1973). The glacial
sequence is overlain by sandstones with inter-
bedded coal seams, but in comparison with the coal
fields of peninsular India, seam thickness is much
reduced, indicating a position closer to the shores of
the Tethys. Palaeontological evidence presented by
Singh (1987) and Singh & Archbold (1993) shows
that the postglacial marine beds of the eastern
Himalaya are of Lower Permian age and that they
correlate with similar postglacial beds in Tibet
(Lhasa etc.) and these in turn may be equivalents of
the postglacial siltstones of the Baoshan Block.
This correlation is strengthened by the fact that the
glacigene sequence of the Lhasa region as well as
that of the Baoshan Block are overlain by thick
basalt accumulations (Jin 1994). There are also
many similarities between the Lower Permian
brachiopod faunas of the eastern Himalayas
(Arunachal Pradesh etc.) with those of Western
Australia (Singh & Archbold 1993). It is suggested,
 BAOSHAN AND TENGCHONG TERRANES, W YUNNAN 545

therefore, that the Baoshan Block occupied a both sections are capped by an anaerobic, black
position adjacent to northeastern India, which is lutite deglaciation facies. New palynological data
consistent with the available evidence. obtained by Yang Wei-ping (1994) from the
The Pangaea sequence of the Tengchong Block black lutites of the Kongshuhe section demonstrate
lacks volcanic rocks and is characterized by equivalence with the Treachery Shale of the
great thicknesses of the glacigene succession. Bonaparte Gulf section. According to Foster
Furthermore, the latter is preceded by a thick sand- (1986) the Treachery Shale belongs to the
stone or siltstone unit (Fig. 2). Such thick glacio- Granulatisporites confluens Oppel Zone (Foster &
marine successions occur in distal parts of Permian Waterhouse 1988) of Upper Asselian to Lower
basins of western and northwestern Australia, as for Sakmarian age. The black lutites of the Kongshuhe
instance in the Carnarvon, Canning and Bonaparte Formation thus fit in perfectly with the Early
Gulf basins. The Pangaea sequence of these basins Permian deglaciation event, as suggested by
is floored in places by Lower Palaeozoic rocks, and Wopfner (1994). The succession of the Tengchong
in others by Proterozoic metasediments or higher- Block also compares well with Permo/
grade metamorphic rocks. Carboniferous sequences in Thailand, western
Figure 3 shows a comparison between the Malaya and Sumatra, thus connecting it with the
Pangaea sequence of the northern Tengchong Block Sibumasu Block in the sense of Metcalfe (1993).
and that of the Bonaparte Gulf basin in north- The wide distribution of tin-bearing granites of
western Australia. In both cases the diamictites are Indosinian age adds another similarity.
underlain by Upper Carboniferous sandstones
which apparently reflect a periglacial facies during
Conclusions
the early stage of glacier advance. The glacio-
marine beds are unusually thick and contain both The evidence presented in this paper suggests that
rain-out tills and mass-transported material, and the Baoshan Block was the easternmost part of the

~5
o
J
P
-..7
:.'"
"......."
.

\F

z'
'i};
. "..: . .: :" ...

... '"~..y""

,...
? '/
(....; .... .... ,....'"'.".~-
""-.
'. ~..
.. • ¢t

•.' ............ .13:'

( .............. :: ~..
..... ~ (". ),...,, .:~:.:, --'.:..........~
"".. !

% ('K- q~

Fig. 7. Reconstruction of Gondwana showing occurrences of Permo-Carboniferousglacial deposits (dashed outlines)

and the limit of glaciation in Africa and Arabia after Wopfner (1991b). Latitudes are drawn relative to an Early
Permian South Pole (star) similar to that given by Levellet al. (1988). The fragments of Gondwana to the north of
India and northwest of Australia are schematic. The position of the glacial deposits, indicated by triangles, is well
within the latitude of active glaciation.
546 H. WOPFNER
Lhasa Block. It originated in a volcanic rift setting
which extended from Kashmir to a position north-
east of Arunachal Pradesh. The Tengchong Block
was part of the Sibumasu Block and its provenance
was a non-volcanic rift off the northwest coast of
Australia. The two rift systems connected at a triple
junction with the Westralian trough (WA trough in
Fig. 7). This triple junction was situated northwest
of the present northwest coast of the Australian
continent (Veevers 1988). The inception of rift
movements along the western and northwestern
coast of Australia during the Late Palaeozoic
glaciation is well documented by pronounced
differences in thickness of the glacial deposits
across active faults (Veevers 1988; Wopfner 1991a)
and comprehensive geophysical data (O'Brien
1993).
The volcanic rift along the Tethyan margin is
well established in Kashmir and Zanskar (Acharyya
1987; Gaetani & Garzanti 1991) where substantial
parts of the rift system remained attached to the
Indian plate. Further east, most of the rift basin was
detached and drifted off as the Lhasa Block and the
Baoshan Block. The difference may be explained if
removal was effected by detachment in the case of
Kashmir/Zanskar but caused by delamination and
subcrustal shear in the case of Lhasa and Baoshan
Blocks (cf. O'Brien 1993).
The timing of initial separation of the terranes
from Gondwana may be deduced from the ages
when deposition of terrigenous detritus ceased. In
the case of the Tengchong Block and presumably
also of the remaining Sibumasu Block this stage
References
ACHARYYA, S. K. 1987. Limits of Greater India Plate
during Gondwana time. The Palaeobotanist, 36,
290-301.
BUREAU OF GEOLOGY AND MINERAL RESOURCES OF
YUNNAN PROVINCE. 1990. Regional Geology of
Yunnan Province. Geological Memoirs, Series 1, 21,
Geological Publishing House, Beijing (in Chinese
with English summary).
CHEN HAIHONG. 1992. Cenozoic rotations of East Asia
and a possible dynamic genesis. Memoirs
Lithospheric Research, 1992, 52-54, Beijing.
DIEKMANN, B. 1993. Pal~ioklima und glazigene Karoo-
Sedimente des sp~ten PaI~iozoikums in SW-
Tansania. SonderverOffentlichungen Geologisches
Institut UniversitiT"t KOln, 90.
DOWDESWEEE, J. A. & SCOURSE, J. D. 1990. On the
description and modelling of glacimarine sediments
and sedimentation. In: DOWDESWELL J. A. &
SCOURSE J. D. (eds) Glacimarine Environments,
Processes and Sediments. Geological Society,
London, Special Publications, 53, 1-13.
FOSTER, C. B. 1986. A review of Earl3, Carboniferous-
was reached not later than Late Sakmarian although
there are still some disparities in the dating of the
limestones succeeding the glacigene sequence. In
the case of the Baoshan Block terrigenous intake
was interrupted after deposition of the red beds
of the Yongde Formation in about Artinskian
times. This corresponds closely with the models
suggested by Acharyya (1987), Gaetani & Garzanti
(1991) and Liu Benpei et al. (1991), whereas Liu
Guanghua & Einsele (1994) suggest a somewhat
younger date. Neo-Tethys was formed between the
new margin of Gondwana and the departing
fragments.
In Late Triassic to Early Jurassic times the
drifting terranes and micro-continents collided with
elements of Cathaysia, thereby closing an oceanic
realm generally referred to as Palaeo-Tethys
(Acharyya 1987). In Yunnan, the result of this
collision was the Changning-Menglian Fold Belt
(Liu Benpei et al. 1991). At this time, structural
elements of Yunnan were still more or less latitude-
parallel, as indicated by structural analyses and
palaeomagnetic data (Chen Haihong 1992) and the
Baoshan and Tengchong Blocks were probably
not in contact with each other. The present juxta-
position of the two blocks is due to dextral strike-
slip movements along the Nujiang dislocation zone
during the Himalayan orogeny in mid-Tertiary
times.
The author records his gratitude to Deutsche
Forschungsgemeinschaft (DFG) and the National Science
Foundation of China (NSFC) for financial support of this
project.
Early Permian selected samples from Kulshill Nos.
1 & 2, Leseur No. 1, Lacrosse No. 1 Pelican Island
No. 1 and Berkley No. 1 of the Bonaparte Basin.
Australian Geological Survey Organisation, Open
File Report.
-- & WATERHOUSE, J. B. 1988. The Granulatisporites
confluens Oppel-zone and Early Permian marine
faunas from the Grant Formation on the Barbwire
Terrace, Canning Basin, Western Australia.
Australian Journal of Earth Sciences, 35, 135-157.
GAETANI, M. & GARZANTI, E. 1991. Multicyclic history of
the northern India continental margin (northwestern
Himalaya). AAPG Bulletin, 75, 1427-1446.
GRAVENOR,C. P. 1979. The nature of the Late Paleozoic
glaciation in Gondwana as determined from an
analysis of garnets and other heavy minerals.
Canadian Journal of Earth Sciences, 16,
1137-1153.
HAMILTON, W. & KRINSLEY,D. 1967. Upper Paleozoic
deposits of South Africa and Southern Australia.
Geological Society of America Bulletin, 78,
783-800.
 BAOSHAN AND TENGCHONG TERRANES, W YUNNAN
HENRICH, R., HARTMANN, M., REITNER, J., SCH.~.FER, P.,
FREIWALD, A., ET AL. 1992. Facies belts and
communities of the Arctic Vesterisbanken
Seamount (Central Greenland Sea). Facies, 27,
71-104, Erlangen.
JIN XIAOCHI. 1994. Sedimentary and palaeogeographic
significance of Permo-Carboniferous sequences in
western Yunnan, China. SonderverOffentlichungen
Geologisches Institut Universitiit K61n, 99.
KANWAR, S. S. & AHLUWALIA,A. D. 1979. Lithostrati-
graphy of Upper Palaeozoic Tethyan Sequence in
Chandra Valley near Bara Lacha La, District Lahul
and Spiti, Himachal Pradesh, India. In: GUPTAV. J.
(ed.) Upper Palaeozoics of the Himalayas.
Hindustan Publishing Corp., Delhi, 147-153.
LEVELL, B. K., BRAAKMAN,J. H. & RUTTEN, K. W. 1988.
Oil-bearing sediments of Gondwana glaciation in
Oman. AAPG Bulletin, 72, 775-796.
L ~ BENPEI, FENG, QINGLAI & FANG, NIANQIAO. 1991,
Tectonic evolution of the Paleo-Tethys in
Changning-Menglian Belt and adjacent regions,
westem Yunnan. Journal China Univ. Geosciences,
2 (1), 18-28.
L ~ GUANGHUa& E~SELE, G. 1994. Sedimentary history
of the Tethyan basin in the Tibetan Himalayas.
Geologische Rundschau, 83, 32-61.
METCALFE, I. 1993. Southeast Asian terranes,
Gondwanaland origins and evolution. In: FINI)LA¥,
R. H., UNRUG, R., BANKS, R. M. & VEEVERS,J. J.
(eds) Gondwana Eight. Balkema, Rotterdam,
181-200.
O'BRIEN, G. W. 1993. Some ideas on the rifting history of
the Timor Sea from the integration of deep crustal
seismic and other data. Petroleum Exploration
Society of Australia Journal, 21, 95-113.
SHI, G. R. & ARCrmOLD,N. W. 1993. Permian brachiopod
faunal sequence of the Shah Thai terrane, biostratig-
raphy, palaeobiogeographical affinities and plate
tectonic implications. Abstracts of 3rd International
Symposium IGCP 321: Gondwana Dispersion and
Asian Accretion, Kuala Lumpur, 1993, 46.
SINGH, T. 1987. Permian biogeography of the Indian
supercontinent with special reference to the marine
fauna. 7th Gondwana Symposium. American
Geophysical Union, Geophysical Monograph, 41,
239-249.
-- & ARCHBOLD, N. W. 1993. Brachiopoda from the
547
Early Permian of the Eastern Himalaya. Alcheringa,
17, 55-75.
SrNHA ROY, S. 1973. Gondwana pebble-slate in the Rangit
Valley tectonic window, Darjeeling Himalayas and
its significance. Journal of the Geological Society of
India, 14, 31-39.
STAMPFLI, G., MARCOUX, J. & BAUD, A. 1991. Tethyan
margins in space and time. Palaeogeography,
Palaeoclimatology, Palaeoecology, 87, 373-409.
TAPPOr,mmR, P., LACASSrN,R., LELOUP,P. H., SCHA_RER,U.,
ZHONG DALAI ET AL 1990. The Ailao Shan/Red
River metamorphic belt, Tertiary left-lateral shear
between Indochina and South China. Nature, 343,
431-437.
VEEVERS, J. J. 1988. Morphotectonics of Australia's
northwestern margin - a review. In: PURCELL,V. G.
& PURCELL, R. R. (eds) The Northwest Shelf
Australia. Proceedings of Petroleum Exploration
Society of Australia Symposium, Perth, 19-27.
WOPFNER, H. 1991a. Permo-Triassic sedimentary basins
in Australia and East Africa and their relationship to
Gondwanic stress pattern. In: ULBRICH H.
ROCHA-CAMPOS A. C. (eds) Gondwana Seven.
Proceedings of the 7th International Gondwana
Symposium, Sao Paolo, 1988, 133-146.
- - 1991b. Extent and timing of the Late Palaeozoic
glaciation in Africa. Sonderverb'ffentlichungen
Geologisches Institut Universitiit KiJln, 82,
447-453.
- -1994. Early Permian deglaciation sequences reflect
end of "ice house" conditions in northern
Gondwana. Permophiles, 24, 22-23.
- -& DJ.EKMANN, B. 1992. Neue Ergebnisse aus der
spatpal~iozischen Abfolge in der Karoo Tansanias.
Zentralblatt fiir Geologie und Paliiontologie, Teil 1,
1991, 2689-2701, Stuttgart.
- - & JIN, XIAOCHI. 1993. Baoshan and Tengchong
Blocks of Western Yunnan (China) in the Late
Palaeozoic mosaic of the eastem Tethys. Abstracts
of the 3rd International Symposium IGCP 321-
Gondwana Dispersion and Asian Accretion, Kuala
Lumpur, 1993, 47-49.
YANG, WEI-PING. 1994. Lower Permian palynological
studies of the Tengchong Block in western Yunnan,
China. Abstracts of the International Symposium on
Permian Stratigraphy, Environment & Resources,
Guiyang, China, 1994, 44.
 Early Mesozoic orogeny in Fujian, southeast China
Z U Y I Z H O U 1, Q I U Y U A N L A O 1, H U A N J I A N G C H E N l, SIJIANG D I N G 2
& ZHONGTING LIAO 1
1Department of Marine Geology & Geophysics, Tongji University, Shanghai 200092, China
2Hainan Geological Survey, 24 Nanhang Road, Haikou, Hainan 570005, China

Abstract: The Mesozoic sedimentary successions of west Fujian evolved from Lower Triassic
deep marine fine turbidites to Upper Triassic and Jurassic coal-bearing molasse deposits, and
represent a typical foreland basin sedimentary sequence. This foreland basin was generated by an
early Mesozoic continental collision to the south, between the South China block and the South
China Sea block. This collision led to the formation of Permo-Triassic S-type granites, and a
first stage of thrusting in a NNW-SSE direction. Post-collisional convergence resulted in the
accumulation of thick molasse deposits. Two main unconformities, within the molasse and
separating the molasse from the underlying flysch, necessitate a re-evaluation of the traditional
concepts of 'Indosinian' and 'Yanshanian' movements in SE China. The tectonic and
stratigraphic evolution of the west Fujian foreland basin, together with the structural features in
the basin, show that the Early Mesozoic orogeny in the region was a mild and continuous process.

South China is geologically a complicated region.
It is traditionally considered as being composed of
two main tectonic elements, the so-called 'Yangtze
para-plafform' and 'the South China Caledonian
orogen' (e.g. Ren et al. 1980; Guo et aI. 1983; Ai et
al. 1985; Wang 1986). The deformed Upper
Palaeozoic and Mesozoic cover of south China is
supposedly typical of a para-platform (Ren et al.
1980). Intense Mesozoic igneous activity in SE
China is related to a postulated Mesozoic sub-
duction of a Pacific plate under east China and the
adjacent regions (e.g. Guo et al. 1983). These
assumptions, however, have been challenged as
oversimplifications because they failed to explain
some new geological data which have emerged
during the past years (Hsti et al. 1987). In order to
explain these new data, Hsti et al. (1987, 1990)
postulated that the Palaeozoic and Mesozoic rocks
of south China are components of a Mesozoic
orogen called the Huanan (south China) Alps and
that the deformation was caused by continental
collisions.
Much of the geological terminology that
dominated Chinese geological literature before the
1980s was also used, naturally enough, for the
geology of SE China. Thus, on one hand there are
erroneous extrapolations of European features such
as 'Ca/edonian' or 'Hercynian' movements and
fold belts, while on the other hand, new orogenic
events like 'Indosinian' and 'Yanshanian' and more
were invented, based mainly on the practice of
relating unconformities to orogenic movements.
These terms often hid ignorance of regional
tectonics and other geological puzzles. They have
all caused great confusion when attempting to
distinguish one orogenic event from another in
regions where there are several unconformities.
The SE Chinese province of Fujian consists of at
least two major tectonic entities, i.e. the east Fujian
block and the west Fujian block, with boundaries
which are still controversial (Fig. 1). Widely
varying views exist as to the tectonic nature and
evolutionary patterns of the region (Guo et al.
1983; Ai et al. 1985; Wang 1986; Ren 1986; Zhou
1989; Bian et al. 1993). New discoveries have
convinced some that it is a collage of several
continental blocks (Guo et al. 1984; Zhou 1989;
Bian et al. 1993). Complex interactions among the
blocks and present Eurasia and Pacific plates and
their precursors should therefore be recorded by
structural, petrological and geophysical character-
istics of the region, although intense magmatic
activity in the Cretaceous has rendered these
characteristics obscure.
In this paper, more emphasis will be directed
towards the geological characteristics of the west
Fujian foreland basin. From these characteristics
and other available geophysical and regional
geological data from this region and the northern
shelf of the South China Sea, the Early Mesozoic
orogeny and its regional significance will be
discussed.
West Fujian foreland basin
General setting
It was traditionally believed that the NNE trending
'Zhen'he-Da'pu fault' (Fig. 1) is the boundary
between two different geological units, i.e. the west

From Hall, R. & Blundell, D. (eds), 1996, Tectonic Evolution of Southeast Asia, 549
Geological Society Special Publication No. 106, pp. 549-556.
550 ZUYI ZHOU ET AL.

Yangtze Block Shangha~l

/ I Fuzhou
/ SouthChina Fujian~ I#~ I
/ Block I J~ h'~iwan
/ Nan'ning / N aI,~
n ' a....
o ~I''/( II
.J'Hong Kong
Reiqiong-~ / J
Strait ~
South China
an Sea Block

Synclines " ~ EastFujianTectonicZone is./.]

[--//
Anticlines " ~ CretaceousVolcanicRocks ~
Permo-Triassic Granites
Palaeozoic-Mesozoic Rocks~ .... !
Profiles I ~ Pre-CambrianRocks ~f~
Fig. 4 II
Fu'an-Nan]inFault F-N Fault
Zhen'he-Da'pu Fault Z-D Fault

Fig. 1, Tectonic map of Fujian, SE China and its regional location. Dashed line marks the boundary of the west
Fujian foreland basin.

Fujian block and the east Fujian block. This belief
was based largely on the restriction to the west
Fujian block of the present surface distribution of
pre-Sinian metamorphic rocks and Upper
Palaeozoic strata. However, in the past decade,
similar strata have been discovered in the area to
the east of the 'Zhen'he-Da'pu fault' and to the
west of the 'Fu'an-Nan'jin fault' which is also
NNE trending. Furthermore, the Bouguer anomaly
map of the region also shows evidence of the
'Fu'an-Nan'jin fault' (Fig. 2), but not of the
'Zhen'he-Da'pu fault'. It has therefore been
suggested by Chen et al. (1982) that the boundary
between the west and east Fujian should be placed
along the 'Fu'an-Nan'jin fault'. Neither of the
two faults, however, can be traced northwards
into Zhejian province to form a conspicuous long
fracture in SE China using geological or geo-
physical evidence. In addition, due to the cover of
thick Upper Mesozoic igneous rocks, the existence
of 'Fu'an-Nan'jin fault' remains geologically
unconvincing. Nevertheless, evidence of the large
geological contrast between coastal east Fujian and
west Fujian can be easily identified, with coastal
east Fujian having been discussed fairly recently
by many authors (Zhou & Lao 1990; Li et al. 1993;
Lu et al. 1993). While many geologists in recent
years (including those who identified the so-
called 'Zhen'he-Da'pu fault' and its northern
equivalent in Zhejian - the 'Zhen'hai-Li'shui
fault') have discarded the concepts surrounding
the definition of these faults ( Bian, pers. comm.
1989; Zhou 1989), it is still quite common in the
English literature to invoke this NNE trending
fault in the southeast corner of south China as a
plate margin (e.g. Hsti et al. 1990).
The northern, eastern and southern boundaries of
the west Fujian foreland basin as shown in Fig. 1
are much more clearly defined than the two NNE
trending faults in Fujian and the basin continues
southwestwards into northeastern Guangdong. The
region that is covered by the basin is usually called
the Yongmei depression and is believed to be either
a 'Hercynian fold belt' (Bian & Gao 1982; Wang
1986) or a 'Hercynian-Indosinian fold belt' (Wang,
E. K. et al. 1982). Zhou (1992a, 1992b) indicated
that the region had evolved from a Late Palaeozoic
passive continental margin to an Early Mesozoic
foreland basin, which was related to an Early
Mesozoic collision and the subsequent compres-
sion between the SE China and the South China Sea
continental blocks.
Stratigraphy
Upper D e v o n i a n - U p p e r Permian. The Upper
Devonian succession in west Fujian consists
mainly of a series of slightly metamorphosed quartz
 EARLY MESOZOIC OROGENY IN FUJIAN, SE CHINA 551

I
118"E
.... . . . . . .

;> 38.8o\

• +~o

'-.•.

• "___._.L. . . . . ~ . .-35.."

"", .

~ . .'"

~ /:/~
37.60
// •

• ~@ ."~' 24ON •

Fig. 2. Bouguer anomaly contours (dashed lines, in mgal) after processing to remove the effect of the upper 15 km
of the crust, and isopachs (solid lines, in kilometres) of the crust of Fujian and environs (original data from Chen
et al. 1983).

sandstones and sandy conglomerates, which siltstones and radiolarian-bearing siliceous

unconformably overlie pre-Devonian meta-
morphosed rocks. The Lower Carboniferous
sandstones have been interpreted as fluvial and
mudflow deposits by Li (1989). The Middle
Carboniferous deposits in some parts of the basin
contain small amounts of volcaniclastic rocks
whose S i t 2 contents are bimodal, i.e. either <52%
or >65%. The REE patterns of the basalts inter-
bedded with these volcaniclastic rocks are similar
to those formed in a continental rift setting (Wang,
D. P. et al. 1982). As the thickness of the Upper
Devonian to Upper Carboniferous sedimentary
rocks gradually decreases northwards, the sediment
grain size increases. The lower part of the Permian
sequence in the basin consists of coalbeds inter-
bedded with sandstones and sandy siltstones, while
the upper part comprises siltstones, calcareous
sedimentary rocks.
Lower Triassic. The lower Triassic Xikou
Formation consists of sandstones, calcipelites,
mud-siltstones and siltstones. Deep-water turbidite
sedimentary structures are well developed in the
Xikou Formation, among which are Bouma
sequences, convolute laminations, ball-and-pillow
structures, scour marks and flute marks. The
palaeocurrent direction determined from flute casts
and climbing-ripple beddings is around 300 ° in the
central area of the basin, i.e. Da'tian and Yong'an.
In the same area, contourites and carbonate gravity
flow deposits were identified (Yang et al. 1993).
The basin was asymmetric with water depth
decreasing towards the northwestern part of the
basin. Syndepositional slumps and brecciation of
552 ZtWI ZHOU ET AL.

strata are also well developed. The grain size

frequency diagram and C - M diagram of 35 Age & Rock Tecton-
samples indicate suspension and saltation trans- Units Lithology stratigraphy
portation (Zhou 1992a). The rate at which 156 Ma
. V . . V . . V , ,
sediments of Xikou Formation were accumulated -..V, . . V . . . . . .

(0.136->0.460 m per 100 years) exceeds that of the ChanglinFm.:ill iiiiiii:

Palaeozoic deposits (0.01-0.17 m per 100 years) in J3c
the region and is within the range of that of the ....?.....?....?.
flysch deposits in other foreland basins in the world 163 Ma =~"~°~~ :
(0.15-1.927 m per 100 years, Schwab 1986). The Zhangping ~
Fm
major element compositions of the Xikou dZz :-.~.--~.:-~.
Formation rocks exhibit low Fe203 + MgO 187Ma .............. Molasse
(3-5%), TiO 2 (0.6-0.95%), A1203/SiO 2 (0.2-0.3%) ".~::~-:.~: ~:.~: ~.:' Deposits
and high K20/Na20. Discriminant function .'~.'.' .~.': .~. in
.............. Foreland
analysis (Zhou 1992a) of sandstones from the Lishan Fm . . . . . . . . . . . . . . Basin
Xikou Formation using 11 major element oxides as J1 ..............
variables and adopting the values of unstandardized
. . . . . . . . . . . . . .
discriminant coefficients for sandstone suites of 208 Ma ..............
eastern Australia (Bhatia 1983) suggest that the
provenance of the Xikou Forrmation was an active Wenbingshan . . . . . . . . . . . . . .
Fm ..............
continental margin (Zhou 1992a). T3w ~ . . c,>.. ~ . .
Another prominent feature of the Xikou
~ : ~ . ~ : ~ . ~ : .~
Formation is its well-developed structures. While 230 Ma
some of these structures are of syndepositional ~ m n Fm
T2n " - : , . = . .x .~x :' Flysch '
origin, the majority were caused by the post- Deposits
240 Ma
depositional compression evident from the closely . . . . . . = ....... [11

spaced mesoscopic scale isoclinal folds, chevron ..... • ..... Foreland

Xikou Fm ...r, . . . . . . ~ .... Basin •
folds, duplexes and associated cleavages (Zhou Tlx ' ' '
1992b; Yang et al. 1993). 245 Ma
p - ; j ~
Middle Triassic-Upper Jurassic. The Middle 250 Ma .............. Shelf -
Triassic calcareous sandstones were deposited in .............. Deposits •
r - - '...T - - - ' . , "--:.- , I
small-scale shallow marine depressions. They are C - : , . - • . = • .'
overlain unconformably by the thick Upper Triassic ., ? , - 1 -
360 Ma I 1 l
conglomerates, pebble-bearing sandstones, sandy ..~' " ~ ' - C~-- -

conglomerates and coal-beating sandstones. The Tianwadong . . . . . . . . . . . . . . Fluvial '

Lower to Middle Jurassic quartz conglomerates, Fm >.-o.-c..
D3t .............. Deposits ,
coarse and fine sandstones and siltstones inter- ~..C3,..6>..
•,, ,~.,. ~,.. ~,m i
bedded with marine deposits are in turn overlain 376 Ma --,.. ~.. r-,- -
unconformably by thick tuffaceous sandy cong-
lomerates, sandstones and siltstones of the Upper Tuffaceous Sandstone
Jurassic. The accumulation rates of these sediments
Coal-bearing sandstone
range between 0.078 m and 0.490 m per 100 years,
close to those of the molasse deposits in other Conglomerate
foreland basins (0.10-0.40m per 100 years, Fine Sandstone
Schwab 1986). Calcpelite ~:~
In summary, the Mesozoic sedimentary succes-
sions of west Fujian evolved from Early Triassic Mudstone
flysch deposits to coal-bearing Upper Triassic to Sandstone
Upper Jurassic molasse deposits, forming a typical
Mud-Siltstone
foreland basin sedimentary sequence (Fig. 3). This
foreland basin trended ENE and increased in size Siliciclastic C a r b o n a t e
during its later evolution. In the Early Triassic, the Carbonate [ ~
foredeep was in Da'tian, central Fujian, where Sandy Conglomerate
deep-water turbidites were deposited. Sub-
sequently, as the size of the basin increased, the Fig. 3. Generalized Upper Palaeozoic-Mesozoic
water depth decreased. This process culminated in stratigraphic column of west Fujian.
Late Triassic and Late Jurassic time when the
 EARLY MESOZOIC OROGENY IN FUJIAN, SE CHINA 553

extremely thick (maximum thickness of 3400 m in Permian

Da'tian) molasse sequence was rapidly deposited SouthChina
SouthChinaBlock SeaBlock
and the width of the basin reached a maximum.

Foreland fold-and-thrust belt

Widely developed folds and low-angle thrusts in
west Fujian have been recognized by regional Early Triassic
geological surveys and drill cores that penetrated
into both the sedimentary cover and the basement.
The foreland fold and thrust belt of west Fujian is
characterized by multistage and multilevel
thrusting of the cover and basement decoupling.
The most prominent folds were developed in the
Lower Permian coal-bearing sediments, and the
folding in these rocks is intense and complex. This
Late Jurassic
coal-bearing Lower Permian sequence also acted as
the detachment zone which enabled the develop-
ment of imbricate thrusts (Fig. 4) and folds in the
overlying sedimentary cover. Other detachment
zones in the basin were in the Lower and Upper
Carboniferous strata.
Low angle thrusts in the basin usually occur in ~ Continental
groups withvery similar dips (15-20 °) and strikes Crust Granites
(ENE-WSW). These occur not only as cover 1 Oceanic ~ Attenuated
Crust Crust
thrusts related to different layers of detachment, but
also as steeper thrusts in the basement of the basin.
Fig. 5. Schematic cross-section to show the evolution of
Kinematic studies of the thrusts reveal two main
west Fujian foreland basin. Symbols are the same as
stages of thrusting. The earlier thrusting episode those in Fig. 3.
has been related to Early Mesozoic north-south
convergence between two continental blocks (Fig.
5) and was overprinted by Early Cretaceous thrust-
ing from west to east (e.g. Tao 1987). The north- SSE to NNW, with the root of the thrust belt, there-
west limbs of the anticlines are shorter and steeper, fore, in the south of the basin.
and at some localities these anticlines are even
overturned. Microfabric studies show that the Permian-Triassic granites
direction of the earlier stage of thrusting was from
Permian to Triassic S-type granites in the west
Fujian foreland basin generally cut across the
fractures and the anticlinal axes. They are alkaline
I granites with high 87sr]S6Sr ratios and low Fe3+/
(FEZ++ Fe 3+) ratios. In addition, they are also
LREE enriched (Wang & Liu 1986). A discriminant
diagram of Rb against Y + Nb (Fig. 6) suggests a
syncollisional and volcanic arc origin. Most of
these granites postdate the closure of the previous
ocean basin and may be related to the partial
II -~SE melting of the partly subducted oceanic crust or to
voluminous sediments accumulated between the
two continents before the final closure of the
intervening ocean basin.

Early Mesozoic continental collision in

SE China
Fig. 4. Structural profiles to show the thrusting Studies of the geology of SE Asia and of south
activity in west Fujian. See Fig. 1 (I and II ) for the China in the last decade or so have culminated
locations of the profiles. in the recognition of several continental blocks
554
Rb
ppm S y n - C O L ? /
1000
100
~ ',~ WPG
10
Y+Nb ppm
I'O 16o ic;oo
Fig. 6. Discriminant diagram of Rb against Y+Nb
for Permo-Triassic granites from west Fujian. The
determination of tectonic settings is based on Pearce
et al. (1984). ORG = ocean ridge granites;
Syn-COLG = syn-collisional granites; VAG = volcanic
arc granites; WPG = within plate granites. The original
data are from Fujian Bureau of Geology and Mineral
Resources (1985).
that were assembled mainly during the Late
Palaeozoic-Early Mesozoic (e.g. Seng0r 1990).
Most of the blocks originated from Gondwana. The
Middle-Late Triassic collision between the South
China block and the Indochina block along the
Black River (or Song Da) suture and the Red River
(or Song Ma) suture is well documented, which is
in marked contrast to the little attention that has
been paid to the possible suture zones further east
in the northern shelf of the South China Sea and the
south China margin. According to Zhu (1987),
there existed a pre-Cretaceous active margin along
the northern shelf of the South China Sea, which
passes through the southern part of Hainan Island
and may be connected to the Red River suture. The
similarity in Cambrian sedimentary successions,
ore deposits, trilobites and brachiopods between
Hainan Island and Australia, together with the
discovery of Upper Palaeozoic glacio-marine
,-/ ) ~"
~,/ ~ t
--
/ .
Fig. 7. Postulation of the positions of Early Mesozoic suture zones on south China margin. Lines 1, 2, 3 are suture
zones proposed by HsiJ et al. (1990), Zhen (1989) and Wang (1986), respectively.The area marked 4 is the possible
location of the Early Triassic suture between the South China and South China Sea blocks suggested in this paper.
zuYi ZHOU E T AL.
deposits of similar origin to those in Tibet, west
Yunnan and SE Asia (Yu 1989) suggest that Hainan
Island is a fragment that rifted from Gondwana at
or after the end of the Palaeozoic. The location of
the suture zone in Hainan Island and its vicinity
remains a problem to be solved. Wang (1986) pro-
posed that the suture zone extends along the
Reiqiong Strait that separates Hainan Island from
mainland China. Others (Zhen 1989) have argued
that it is located in northern Hainan (Fig. 7).
Further east, Hsti et al. (1990) postulated the
existence of a Mesozoic (Triassic or Jurassic)
collision between the Huanan and the Donnanya
Blocks. However, it is worth mentioning here that
the suture zones to the south and to the east of
Fujian in SE China are of different origins and ages
(e.g., Zhou 1992b).
The deep-water flysch deposits of the Lower
Triassic Xikou Formation, the intense deformation
of these deposits, the formation of S-type granites
and the first episode of thrusting in west Fujian
fold-and-thrust belt have all been interpreted above
to be genetically related to a continent-continent
collision which took place to the south of the west
Fujian foreland basin in the Early Triassic. This
collision marked the beginning of the Early
Mesozoic orogeny in the region. However, defor-
mation did not stop as collision ceased. Continuing
post-collisional convergence made the orogeny a
long process and led to the uplift of colliding
continental blocks, resulting in the deposition of
thick molasse in the adjacent west Fujian foreland
basin (Figs 3 & 5). The difficulty in identifying the
suture zone has been due mainly to the massive
magmatic and deformation activity related to the
subduction process to the east of Fujian in the
Taiwan Strait in the Cretaceous (Zhou e t al. 1992b),
and to submergence of the area beneath the South
China Sea. However, geophysical data from the
region record some evidence of the existence of an
ENE trending tectonic zone. The Bouguer anomaly
trend in east Fujian swings from NE to ENE at
//Pearl River /"
h Basin /'
II _-- . . . . . . //
........ ,oo
 EARLY MESOZOIC OROGENY IN FUJIAN, SE CHINA
the southern border of the west Fujian foreland
basin. Folding and thrusting have resulted in the
thickening of the crust in the basin (Fig. 2). Along
the coast from Nan'ao to Hong Kong is an ENE
tending tectonic zone characterized by a relatively
low aeromagnetic anomaly and gravity gradient,
reflecting proposed ultrabasic igneous rocks at a
depth of 8-12 km (unpublished report of No. 909
Aeromagnetic Survey Brigade, Ministry of
Geology and Mineral Resources of China). In the
Pearl River Mouth Basin, three ENE trending
aeromagnetic lows are also attributed to deep basic
intrusive rocks (Guong et al. 1989). In order to
unravel the basement structures beneath the
Tertiary basins of the northern South China Sea
shelf, an integrated geological-geophysical
approach has been used to interpret several N-S
trending profiles for which there are gravitational,
magnetic, seismic and borehole data supplemented
by onshore data. Forward and inversion iterative
modelling was applied to the profiles. These data
enabled the postulation of Palaeozoic metamorphic
rocks and Upper Mesozoic sedimentary and meta-
morphic rocks in the basement. Fault mapping
revealed an ENE fault system cross-cut by younger
NW trending faults. All these phenomena are
consistent with a collisional event. A schematic
evolution for west Fujian is proposed in Fig. 5,
though at this juncture we can only conjecture as to
the position of the suture zone (Fig. 7).
Discussion
According to Seng/3r (1985), the Cimmerian
Continent rifted from the northern margin of
Gondwana during the latest Palaeozoic to the
earliest Mesozoic. This continent disintegrated
as it moved through the Tethyan domain.
Palaeomagnetic and palaeontological data suggest a
possible Gondwana origin for the Yangtze, south
China and Hainan Island blocks. The evidence of
Early Mesozoic continental collision in SE China
suggests that, before the Cimmerian Continent
united with Eurasia during the Late Triassic-
Middle Jurassic (Sengtir 1985), some of the conti-
nental blocks (e.g. South China block and South
China Sea block in this paper) had already collided
with each other during their northward drifting. The
collisions of continental blocks that constitute the
Cimmerides therefore took place diachronously.
References
At, C. X., CHEN, B. W. & HUANG,H. 1985. Preliminary
study of iron ore-control tectonics of south Fujian
and east Guangdong. Bulletin of the Institute of
Geology of Chinese Academy of Geological
Sciences, 13, 53-66 (in Chinese)
BHATIA, M. R. 1983. Plate tectonics and geochemical
555
The so-called 'Indosinian' movement in Indochina
and south China is the regional geological phenom-
enon that resulted from this diachronous collision
process.
As mentioned above, the Early Mesozoic
orogeny in SE China was not a short event, because
N-S post-collisional convergence continued until
the end of the Jurassic. The two regional uncon-
formities within the syn-orogenic molasse and
separating the molasse and the underlying flysch
(Fig. 3), reflect two main stages of tectonic
relaxation (or tectonic quiescent periods of Blair &
Bilodeau 1988) in the collision zone. The coarse
sediment above the unconformities actually
represents flexural rebound of the thrust belt as the
uplifted area was eroded, and the finer sediment
above the two coarse units might represent renewed
uplifts that were related to the continuing N-S
compression. The traditional practice of relating
these two unconformities to, respectively,
'Indosinian' and 'Yanshanian' orogenies should
therefore be reassessed.
The sedimentary sequence and evolution of the
west Fujian foreland basin are comparable with
those of typical foreland basins (e.g. the
Appalachian foreland basin, Tankard 1986).
Thrusting and drcollement took place along ramps
at different levels, and the detached pieces could
not displace freely on a perfect 'sole' of the
southern Appalachian type (Zhu 1989). Finally,
the Lower Triassic deep marine fine turbidites
in the foredeep of the west Fujian foreland basin
conformably overlies the Upper Permian deposits
(Fig. 3), which shows that the transition from
passive continental margin to active continental
margin was gradual. The fact that neither mud nor
shale was deposited over the fine turbidites also
reveals that the compression was a slow and
continuous process. In addition, the lack of
evidence of metamorphism shows that the
continental collision was a mild collision.
This paper was prepared at the University of Wales,
Aberystwyth in early 1994, while the senior author was
a recipient of a Royal Society Visiting Fellowship. R. J.
Whittington, W. R. Fitches, T. Horscroft, J. Charvet,
M. Allen and B. C. Burchfiel are thanked for their con-
structive comments and contributions to the manuscript.
Robert Hall is thanked for improving the figures and
manuscript. This research was supported by the National
Science Foundation of China (project NSFC49202034).
composition of sandstones. Journal of Geology, 91,
611-627.
BIAN,X. Z. & GAO, T. J. 1982. Discussion on the types of
iron ore deposits and their exploration in 'Yongmei
depression'. Geology of Fujian, 1 (2), 53-67 (in
Chinese).
556 zuYI ZHOU ET AL.
--, Cue, Z. X. & ZHOU, W. D. 1993. Framework of
Palaeozoic-Mesozoic tectonic evolution of Fujian
Province. Geology of Fujian, 12 (4), 280-291 (in
Chinese).
BLAIR, T. C. & BILODEAU, W. L. 1988. Development of
tectonic cyclothems in rift, pull-apart, and foreland
basins: sedimentary response to episodic tectonism.
Geology, 16, 517-520.
CHEN, Y. A. & WANG, P. Z. 1982. A study on the regional
geophysical and geochemical characteristics of
Fujian. Geology of Fujian, 1 (2): 69-82 (in
Chinese).
FUJIAN BUREAU OF GEOLOGY AND MINERAL RESOURCES.
1985. Memoirs of the Geology of Fujian. Geology
Press, Beijing, (in Chinese).
Guo, L. Z., SHI, Y. S. & MA, R. S. 1983. Formations and
evolution of Meso-Cenozoic active continental
margin and island arcs of west Pacific. Acta
Geologica, 57 (1): 11-18 (in Chinese).
--, YE, S. F. & Lu, H. E 1984.
Tectonostratigraphic terranes of Southeast China.
Journal of Nanjing University (Natural Sciences
Edition), 20 (4), 732-739 (in Chinese).
GUONG, Z. S., JIN, Q. J., WAND, S. S. & MENG, J. M. 1989.
Geology, tectonics and evolution of the Pearl River
Mouth Basin. In: ZHU X. (ed.) Chinese Sedimentary
Basins. Elsevier, Amsterdam, 181-196.
Hst3, K. J., SUN, S. & LI, J. L. 1987. Huanan Alps, not
South China Platform. Scientia Sinica, B series,
1107-1115.
, LI, J. L., CHEN, H. H., WANG, Q. C., SUN, S. &
SENGOR,A. i . C. 1990. Tectonics of South China:
Key to understanding the West Pacific Geology.
Tectonophysics, 183, 9-39.
LI, J. L., HE, H. Q., YANG, M. E, WANG,Z. M. & L1N, D.
Y. 1993. Structural features of the Mesozoic
ophiolitic melange in coastal Fujian. In: LI, J. L.
(ed.) Lithosphere structure and geological evolution
of SE China. Metallurgy Industry Press, Beijing,
199-206 (in Chinese).
LI, Y. W. 1989. Sedimentary and structural analysis of
Lower Palaeozoic Yongmei depression. M Sc thesis,
Tongji University (in Chinese).
Lu, H. F., J1A, D., Guo, L. Z., SHI, Y. S., ZHANG, Q. L. &
WANG, Z. H. 1993. The formation and colliding
history of Fujian-Taiwan microcontinent. In: LI, J.
L. (ed.) Lithosphere structure and geological
evolution of SE China. Metallurgy Industry Press,
Beijing, 12-26 (in Chinese).
PEARCE, J. A., HARRIS, N. B. W. & TINDLE, A. G. 1984.
Trace element discrimination diagrams for the tec-
tonic interpretation of granitic rocks. Journal of
Petrology, 25, 956-983.
REN, J. S. 1986. Some problems about the geotectonics of
South China. Scientia Bulletin, 49-51 (in Chinese).
, JIANG, C. E, ZHANG, Z. K. & Q1N, D. Y. 1980.
Geotectonics of China and its evolution. Science
Press, Beijing (in Chinese).
SCHWAB,E L. 1986. Sedimentary 'signatures' of foreland
basin assemblages: real or counterfeit? In: ALLEN,P.
A. & HOMEWOOD, P. (eds) Foreland Basins.
Blackwell Scientific Publications, 359-410.
SENGOR, A. i . C. 1985. The story of Tethys: how many
wives did Okeanos have? Episodes, 8 (1): 3-12.
- - 1990. Plate tectonics and orogenic research after 25
years: a Tethyan perspective. Earth Science Review,
27, 1-201.
TANKARD, A. J. 1986. On the depositional response
to thrusting and lithospheric flexure: examples
from the Appalachian and Rocky Mountain basins.
In: ALLEN, P.A. & HOMEWOOD, P. (eds) Foreland
Basins. Blackwell Scientific Publications,
369-392.
TAO, J. H. 1987. Characteristics and mechanics of thrust
tectonics in southwest Fujian. Geology of Fujian, 7
(3), 196-204 (in Chinese).
WANG, D. P., LIu, Z. J. • WANG, D. Q. 1982.
Palaeogeography and structures of Makeng-type
iron ore deposits, Fujian. Journal of Changchun
College of Geology, 3, 43-58 (in Chinese).
WANG, D. Z. & LIU, C. S. 1986. Distribution regularities
and genetic series of granites of Hercynian-
Indosinian cycle in Southeast China. Acta
Petrologica Sinica, 2 (4), 1-13 (in Chinese).
WANG, E. K., ZHANG, J. Z. t~ OUYANG, Z. 1982. The
discovery of fusulina in metamorphic rocks in
Datian, Fujian and its significance. Journal of
Nanjing University (Natural Science Edition), 18
(2), 25-32 (in Chinese).
WANG, H. Z. 1986. Geotectonic development of China. In:
ZUNYI YANG, YUQI CHEN & HONGZHENWANG (eds)
The Geology of China. Clarendon Press, Oxford,
235-276.
YANG, M. E, LI, J. L., HAO, J., Hou, Q. L. & HE, H. Q.
1993. The discovery of Lower Triassic contourites
and carbonate gravity flow deposits in southwest
Fujian. In: El, J. L. (ed.) Lithosphere structure and
geological evolution of SE China. Metallurgy
Industry Press, Beijing, 213-218 (in Chinese).
Yu, Z. Y. 1989. The determination of the early Permian
glaciomarine deposit in Hainan Island and its
tectonic significance. Journal of Nanjing University
(Natural Science Edition), 25 (1), 108-119 (in
Chinese).
ZHEN, J. Z. 1989. Tectonics of west Hainan Island. MSc
thesis, Tongji University (in Chinese).
ZHOU, Z. Y. 1989. The study on the basement of SE
China: a review. Geology of Fujian, 8 (1), 46-53
(in Chinese).
- - 1992a. The depositional environment and tectonic
setting of Xikou Formation in western Fujian.
Experimental Petroleum Geology, 14 (2), 135-142
(in Chinese).
-- 1992b. Eastern Fujian tectonic zone. In: Lru, G. D.
(ed.) Geological-Geophysical Features of China
Seas and Adjacent Regions. Science Press, Beijing,
320-327 (in Chinese).
- - & LAO, Q. Y. 1990. Mesozoic evolution of eastern
Fujian and its adjacent areas. In: WILEY, T. J. &
HOWELL, D. J. (eds) Terrane Analysis of China and
the Pacific Rim. Circum-Pacific Council for Energy
and Mineral Resources, Earth Science Series, 13,
185-187.
Znu, X. 1987. On the evolution of Chinese continental
margins and basins. Marine Geology and
Quaternary Geology, 7 (3), ll5-119 (in Chinese).
1989. Remarks on Chinese Meso-Cenozoic
sedimentary basins. In: ZHU X. (ed.) Chinese
Sedimentary basins. Elsevier, Amsterdam, 1-5.
 Index

References in italics are to Tables or Figures

accretionary prisms and complexes Bacan 445

Banda arc 53, 57 crustal isotopic signature 448-9
Borneo 252, 253 K/Ar dating 502, 503, 505
Songpan Ganzi 99 plate setting 483
Sumatra 23-4 Sibela continental suite 483-4
Timor trough 81-2 Sibela ophiolite 486-7
Visayas 515-16 tectonic evolution 490-3
Aceh sliver microplate 22-3 Tertiary stratigraphy 487-90, 501
Adang fault 382-3 Bacan Formation 487, 491
Aibi-Xingxing suture 99 Bakit Mersing line 248, 252
Aifam Group 468 Balangbaru Formation 359-61,366, 367
Ailaoshan belt 540 Ban Ang Formation 236
Ailaoshan suture 99, 106 Ban Thalat Formation 236
Aileu Formation 57 Banda arc 11, 47, 64, 451
Air Bangis granite 331,333 backarc region 56, 68-70
Aitape Basin 527 forearc structure 68
Ala Shan terrane 99 indentor model 70-2
Altyn Tagh fault 124, 125 orogen evolution 57-8
Amasing Formation 489 reflection profiles
Ambon 445 methods 48-9
crustal isotopic signature 447-8 results 49-56
amphibolite, Darvel Bay 265 Banda arc (east) see Aru trough
Andaman Sea 160 Banda orogen 194-5, 197, 451
Anggi granite 468 Banda Ridges 445
apatite fission track analysis crustal isotopic signature 449
Papua New Guinea 528-31 Banda Sea
Thailand 244-5 Miocene history 462
western Sulawesi 404-6, 424 Miocene setting 144-5
4°mr/39Arages, Kaibobo complex 457-9 tectonic blocks 140, 141
arc-continent collision 525 tectonic setting 175-7
Arip Volcanics 253 Bangdu Formation 541
Aru trough 86-7 Banggai granite 474
deformation front 92-4 Banggai Island stratigraphy 474
gravity data 89-92 Banggai-Sula continental fragment 160, 466
seismic reflection data 88-9 Banggai-Sula Spur 445
Asem-Asem Basin 252 Banggong suture 99, 107
Ashmore platform 62, 63 Bantimala melange complex 354-5
asphalt 431 lithology 355
ATLAS model 154 outcrop distribution 356-61
Aurora volcanic ridge 518 structure 355-6
Australian plate tectonic evolution 361-2
collision with Philippine Sea plate 137 Banyak thrust 25
continental shelf characters Baoshan block 101,540
crustal thickness 63--6 diamictite 541-4
gravity survey 66-8 terrane motion 544-6
marginal slope features 69-81 Barail Formation 134
reflection profile 49-51 Barisan Mts 321-2
structure 63 Barisan orogen 189-91,197
isotope characteristics 446 Barito Basin 252
plate motion 32 Batu fault 21
Ayu trough 168 Bawang Dacite 253

557
558 INDEX

Belaga Formation 248, 250, 251,253 Celebes Sea 140, 147, 164, 171
Bentong-Raub line 204 Cenozoic tectonic setting for SE Asia 124-5
Bia Formation 475 Champa Formation 236
Biak Formation 475 Changning-Menglian Belt 540, 546
biostratigraphy, Pak Lay sediments 229-30 Changning-Menglian suture 99, 107-8
Bird's Head 139, 142-3, 146 China, South 539, 549
Pb isotope ratios 446 collision tectonics 525
stratigraphy 470, 471 Asia-India 129, 133-5
tectonic reconstruction 477 Australia-Philippine Sea plate 137
tectonic setting 165, 174 Fujian evidence 553-5
Bisa, K/Ar dating 503 Indochina 225, 230
Bislig Bay 517 Philippines 512-16
Black River suture see Song Da Sulawesi 171-4
Bliri volcanics 527 continental fragments 445
Bobang Formation 474 Cotabato fault 518
Bogal Limestone Formation 468 Cotabato trench 39-40
Bonaparte basin 62, 63 Cretaceous studies
Borneo (Kalimantan) 252 Khorat Plateau 237-8
Schwaner Mts 252-3 palaeomagnetism 207
tectonic setting 157-9, 168-70 Salawati stratigraphy 468
see also under Kalimantan Sumatra plutonism 324, 325, 330
Bouguer gravity measurements terrane distribution 114
Aru trough 89-92 Tomori stratigraphy 475
Australian continental shelf 66 Crocker Formation 256-8
Boyan Melange 252 crustal provenance analysis
Boyan suture 99, 110-11 Pb-Nd isotope analysis
Browse basin 62, 63 methods 446
Buan Formation 251 results 447-50
Bunga basalt 253 results discussed 450-1
Bunta Formation 474
Burma see Myanmar
Buru, tectonic setting 168 Daguma fault 518-19
Buton 160, 167, 431-3,445 DAMAR line 48, 49-56, 64, 76, 78-82
displaced terrane 466 Damuchang Formation 541
palaeomagnetism study Dangerous Grounds Block 308
methods 434-5 Daram Sandstone Formation 470
results 435-4 1 Darvel Bay Complex 263-4
results discussed 441-2 metamorphic geology 264-73
Buya Formation 474 origins 273-7
ophiolite emplacement 277
declination 435,436, 437, 439
Cagayan ridge 164 Devonian studies
Camba Formation 354-5, 367 Fujian province 550-1
Camba-Enrikang-Mamasa Complex 397 rifting 104
Cambro-Ordovician terrane distribution 111 Salawati stratigraphy 468
Canning basin 62, 63 terrane distribution 111
Carboniferous studies diamict, Yunnan 541-3
Fujian province 551 Dingjiazhai Formation 540, 541,542
rifting 104 displaced terranes 466
Salawati stratigraphy 468 Dongnanya block 100
terrane distribution 112 Dungun Graben 295-6
Tomori stratigraphy 474
Carcar Formation 520
Caroline plate 5-6, 141, 145-7, 466 East Java Sea 384
tectonic setting 168, 177 East Malaya terrane 99, 102
Cascades fault zone 519 East Natuna Basin 299
Cathaysia block 100 East New Guinea composite terrane 140, 141
Cathaysialand 113 East Papua composite terrane 140
 ~NDEX 559

East Talaud bank 40-2 Huon-Finisterre arc terrane 13

Elat Formation 87 hydrocarbon potential
Embaluh Group 251,253, 254 Lao PDR 246
Eocene studies Malay Basin 281,285
Asia-India collision 129 Penyu Basin 281
plate settings 380-2 Salawati Basin 479
Sulawesi tectonics 383 Thailand 244
Tomori Basin 479

Facet Limestone Group 468

Fafanlap Formation 470 inclination 435,436, 437, 439
Faumai Formation 470 indentor model, Banda arc 70-2
Finnisterre terrane 140, 141, 145, 148 India-Asia collision
Flores thrust 11 kinematic model 125-8
forearc environments, examples of see Sumatra forward modelling 135-7
margin model constraints 133-5
'Fu'an-Nan'jin fault 550 rotation evidence 131-3
Fujian province 549 Tertiary development 129-31
foreland basin 549-50 Indian Ocean, seafloor spreading 127
collision tectonics 553-5 Indoburman flysch 134, 135
fold and thrust belt 553 Indochina block/terrane 99, 102
granites 553 tectonic history 204-6
stratigraphy 550-3 tectonic setting 157
Triassic collisional tectonics 225, 230
Indonesia
Gabaldon basin 519 orogenies 188-96
Gamta Limestone Formation 468 tectonic development 186-8
Gaoligongshan metamorphics 540 tectonic history 204-6
Garba pluton 331 tectonic setting 186
garnet amphibolite, Darvel Bay 272-3 Indosinian orogeny 230, 549, 555
geochemistry Indus Yarlung Zangbo suture 99
Darvel Bay 267-8, 269 Irian Jaya see Bird's Head; Salawati Basin
Papua New Guinea 532-3 isothermal remanent magnetism measurement
Sumatra granitoids 325-8 methods 435
western Sulawesi 397-400, 425 results 435, 436, 437
geochronology s e e 4°mr/39Ar;K]mr; Rb/Sr; Izu trench 8
Sm/Nd; U/Pb Izu-Mariana trench 16
glacigenic sediments, Yunnan 541-3
Gondwana terrane rifting 103-5, 539, 544-5
gravity see Bouguer
Gujuti Formation 504 Jass Formation 468
Gunung Api 68 Java, tectonic setting 160
Java trench 7, 8
Java-Sumatra margin, collision evidence 133-4
Hainan Island 99, 103, 554 Javan Forearc 337
Halmahera 29, 32, 499-500 Jinshajiang suture 99, 106
K/Ar dating Jurassic studies
methods 500-2 Fujian province 552
results 502-6 palaeomagnetism 206-7
results discussed 506-8 rifting 104
stratigraphy 500 Salawati stratigraphy 468
tectonic setting 165,491 Sumatra plutonism 324, 325, 330
Hatapang granite 331 terrane distribution 114
Himalaya trench 8 Tomori stratigraphy 475
Himalayan front 9-11
hornblende gneiss, Darvel Bay 265
Huai Hin Lat Formation 206, 209, 236 K/Ar dating
Huanan Alps orogen 549 Bacan 502, 503, 505
560 INDEX

K/Ar dating (contd) Kuantan Formation 324

Bisa 503 Kulapis Formation 313, 316
Halmahera arc 506-8 Kun Lun fault 125
Khorat granites 244 Kunlun suture 99
Obi 503, 504 Kunlun terrane 99
Papua New Guinea 531 Kurogegawa terrane 99, 103
Pak Lay 229 Kutei Basin 384
Rajang Group 253
Sumatra granitoids 328-9, 333
western Sulawesi 400-3
Kai Islands 87-8 Labang Formation 312, 316
Kaibobo granite 456 Lamasi Complex ophiolite 397, 400
Kaibobo ultramafic complex Lampang Group 208, 212, 214
4°Ar/39Ardating 457-9 Lancangjiang suture 99, 107
closure temperature 459 Langi Formation 366
implications for Banda Sea 462 Lanping-Simao fold system 539, 540
implications for obduction 461-2 Lao PDR 539
P-T-t modelling 459-61 Khorat Plateau 233-5
petrography 457 hydrocarbon potential 244, 246
Kais Formation 471 inversion history 237-45
Kalaw red beds 208, 210-11,212-13, 216 palaeomagnetic studies 206-7
Kalimantan (Borneo) stratigraphy 384 stratigraphy 235-7
Rajang Group 251-2, 253, 254-6 tectonic development 245
see also under Borneo Pak Lay Fold Belt
Kalomba Formation 475 history of research 225-8
Kaputusan Formation 489-90, 505 recent studies 229-31
Kasiruta stratigraphy 487-8 summary history 231
Kayasa Formation 505-6 Lassi pluton 331
Kayoa 505 Laurasia-Gondwana marginal terranes 539, 544-5
Kazakhstan terrane 99 leaky transform fault 70
Kelabit Formation 256 Legaspi Lineament 517
Kelalan Formation 250, 256 Lelinta Shale Formation 468
Kembelangan Group 468 Lenggurur thrust belt 146
Kemum Formation 468 Lhasa terrane 99
Kemum terrane 143 Lianga fault 517
Kerabai volcanics 253 Ligu Formation 468
Keskain Formation 468 Long Bawang Formation 256
Ketapang batholith 252 Longmen Thrust 124, 125
Ketungau basin 254 Lubok Antu Melange 248
Khlong Marui fault 159 Luogengdi Formation 541
Khok Kruat Formation 207, 209 Luok Formation 475
Khorat Group 225, 230, 231,233 Lupar fault 254
Khorat Plateau 233-5 Lupar Formation 248, 250, 251,253
hydrocarbon potential 244, 246 Lupar Line 248, 252
inversion history 237-45 Lupar Line ophiolite 248
palaeomagnetic studies 206-7 Lurah Formation 256
stratigraphy 235-7 Luwuk Formation 475
tectonic development 245 Luzon, Miocene tectonics 515
Kimberley block 62, 63
Kintom Formation 475
Kisar Island 68
Klamogun Formation 471 Macolod Corridor 511,520-1
Klasafet Formation 472 Makassar Strait 394-7
Klasaman Formation 472 tectonic setting 162
Klondyke Formation 515 Malacca microplate 323, 324
Kluet Formation 324 Malawa Formation 354, 366
Kongshuhi Formation 541,542 Malay Basin 281,292-3
Kontum massif 206 basement 283-4
 INDEX 561

heat flow 282 Muna 434

hydrocarbons 281,285 Mussau trench 6
stratigraphy 283, 293 Myanmar (Burma)
structure 284-7, 293-7 collision evidence 134-5
Malay blocks, tectonic setting 159-60 palaeomagnetic studies
Malay Peninsula methods 212
palaeomagnetic studies 207-10 results 212-13
tectonic setting 539 stratigraphy 210-11
Maleta basin 62, 63 tectonic setting 539
Manday suture 99, 110
Manunggal batholith 331
Mariana trench 8, 15-16
Marinduque Basin 521 Nagan granodiorite 331
Matano Formation 475 Nalang Formation 236
Matindok Formation 475 Nam Con Son Basin 300
Mekong basin 282, 299-300 Nam Set Formation 236
Melanesian orogeny 195-6, 197 Nam Thom Formation 236
melanges Nam Xoi Formation 236
Java 353 Nam-Phong Formation 206, 209
Nias Island 342 Nambo Formation 474
Sulawesi 356 Nan-Uttaradit ophiolite zone 225
Melawi Basin 254 Nan-Uttaradit suture 99, 107
Menanga Formation 474 Nanaka Formation 474
Mengliong Group 540 Nasa Formation 236
Mentarang Formation 256 natural thermal remanent magnetism 435,436, 437
Mentawai fault 21, 24 143Nd/144Nd and provenance
Mentawai fault zone 338, 345-6, 346-50 methods of analysis 406-7, 425, 446
Mentawai sliver microplate 21-2 results 447-50
Meratus Mts 252 results discussed 450-1
Meratus suture 99, 110 Neogene studies
Mergui microplate 323, 324 Halmahera arc stratigraphy 500
metachert, Darvel Bay 271-2 orogenies 188-96
metamorphic studies Philippine tectonics 511-12
Darvel Bay 264-73 docking 512-16
Gaoligongshan 540 extension 520-1
Sulawesi 140-2 post docking 516-20
metatuff, Darvel Bay 270-1 Sundaland rotations 217-18
Miangas ridge 35 see also Miocene; Pliocene
Mindanao collision zone 511, 512-15 New Guinea
Mindoro-Panay collision zone 511 convergence evidence 145-6
mineralogy (heavy), Yunnan glacigenics 543 indentor 43
Miocene studies Pb isotope ratios 446
Asia-India collision 130-1 plate motion 12-13
Banda Sea 462 see also Papua New Guinea
Malay and Penyu Basins 283 New Guinea Limestone Group 470
Papua New Guinea 533,534-6 New Guinea Mobile Belt 525, 526
Sulawesi tectonics 384 New Guinea orogen 147
Sulu Sea Basin 311-16, 317 New Hebrides trench 8, 13
Misool Island, stratigraphy 468, 470, 472, 571 Nias basin 21
Molucca Sea 29, 147 Nias Island 338-41
bathymetric transect features 32-43 deformation studies 341-6
geodynamic framework 32 fault pattems 346
lithospheric boundaries 43-5 significance of structures 346-50
subduction 492 Nofanini Formation 474
tectonic setting 174-5 North China terrane 99
Molucca-Sorong fault 467 North Sula-Sorong fault 467
Moon volcanics 489 Northeast China terrane 99
Mukus assemblage 323, 324 Nusa Babi monzodiorite 488-9, 503
562 INDEX

NW Borneo-Palawan Trough 308 Khorat Plateau 206-7

NW Sabah Platform 308 Malay Peninsula 207-10
Nyaan Volcanics 253 Myanmar 212-13
Nyalau Formation 251 relation to plate setting 154-6, 217-19
Pangaea 539, 544-5
Papua New Guinea
apatite fission track analysis
Obi
methods 528-30
displaced terrane 466
results 530
K/Ar dating 503, 504
results discussed 530-1
Tertiary stratigraphy 501
geochemistry 532-3
oblique subduction/convergence 3, 19
plate setting 525
Oligocene studies
stratigraphy and structure 525-7
Asia-India collision 129-20
tectonic summary 533-6
Malay and Penyu Basins 283
thermochronology 531-2
New Guinea orogen 147
Pasir Basin 252
Papua New Guinea 534
Pb isotope ratios
South China Sea extension 248
methods of analysis 446
Sulawesi tectonics 383-4
results 447-50
Ombilin granite 324, 329
results discussed 450-1
ophiolites
Pedawan Formation 250
Darvel Bay 263-4, 277
Penyu Basin 281,297-8
Lamasi Complex 397, 400
basement 283-4
Lupar Line 248
stratigraphy 283
Nan-Uttaradit 225
structure 287-8
Sibela 486-7
Permian studies
Opol Formation 513
Fujian province 551,553
Ordovician studies 111
Khorat Plateau 235
orogenic events
palaeomagnetism 206
Banda 57-8, 194-5, 197, 451
rifting 104
Baresan 189-91,197
Salawati stratigraphy 468
Huanan Alps 549
Sumatra magmatism 323-4, 329
Indonesia 188-96, 230, 549, 555
terrane distribution 112
Melanesian 195-6, 197
Tomori stratigraphy 475
New Guinea 147
petroleum see hydrocarbon potential
Pacific 225
Phalat Formation 236
Sabah 258
Phetchbun Fold belt 233, 234
Sarawak 256
Philippine arc 511
South China Caledonian 549
Philippine archipelago
Sulawesi 191-4, 197
Neogene tectonics 137-9, 164-5, 511-12
Sunda 188-9, 197
docking 512-16
Talaud 191,192, 197
extension 520-1
Yanshanian 549, 555
post docking 516-20
Philippine fault 15, 511
Philippine Sea plate 5, 32, 466, 491
Pacific orogeny 225 boundary evidence 147
Pacific Plate, Mesozoic subduction 549 coupling with Australia 137, 145-6
Pak Lay Fold Belt rotation estimates 138-9
history of research 225-8 tectonic setting 162
recent studies 229-31 Philippine trench 15, 32, 44
summary history 231 bathymetry 33-5
palaeogeography, terrane distribution 111-14 Phon Hong Group 235, 236, 244
palaeomagnetic studies Phu Phanang Formation 236
methods 212, 434-5 Phulekphey Formation 236
regional results Piring granodiorite 253
Asia rotation evidence 131-3 Piyabung Volcanics 253
Banda Sea 142 plate motions 4, 5-6
Buton 435-42 boundary convergence 3
 INDEX 563

boundary zone deformation 6-7 Saleh Island stratigraphy 487, 489

Plateau Limestone 208, 212 Salodik Formation 474, 475
Pliocene studies 536 Sampolakosa Formation 435,438, 439, 440
Poh Formation 475 Sandakan Formation 314, 316
Prince Alexander Mts 527 Sangihe arc 29, 32, 39
provenance analysis see under crustal Sangihe basin/trough 39, 40
Pujada ridge 35-9 Sat Khua Formation 207, 209
Pumenqian Formation 540 Sapulut Formation 256
Sarangani-Davao depression 39
Sarawak
Qaidam terrane 99 Ketungau Basin stratigraphy 254
Qamdo-Simao terrane 99 Rajang Group stratigraphy 248-51
Qiangtang terrane 99 Sarawak orogeny 256
Qinling-Dabie suture 99, 109-10 Saysomboun Formation 236
Quaternary studies Schwaner Mts 252-3
Bacan stratigraphy 490 SE Asia, tectonic setting in Cenozoic 155, 158,
Halmahera arc 506 161,163, 166, 169, 170, 172, 173, 176,
178
SE Asia (SEA) plate 5
Sebuku Formation 256
radiometric dating s e e 4°Ar/39Ar, K/Ar; Rb/Sr; seismic sections
Sm/Nd; U/Pb Sandakan basin 314, 315
Rajang Group 253 Vientiane Plain 239-243
Kalimantan 251-2 Selangkai Formation 251,253
Sarawak 248-51 Semitau ridge 254
Ratburi Limestone 208, 212, 214 Semitau terrane 99, 103
Raub-Bentong suture 99, 108-9 Sepauk tonalite 252, 253
Rb/Sr dating Sepik Basin 526
Sumatra granitoids 324, 328-9 Seram 445, 455-7
western Sulawesi 406-7, 425 crustal isotopic signature 448
Red River fault 124, 125, 157, 204 tectonic setting 167-8
Red River suture see Song Ma ultramafic complex study
rifting events 103-5 4°Ar/39Ar values 457-9
Ruta Formation 489 closure temperature 459
Ryukyu trench 8, 15 implications for Banda Sea 462
implications for obduction 461-2
P-T-t modelling 459-61
Sabah 307-8 petrography 457
Central basin 311-16 Seram trough 86
Darvel Bay Complex 263-4 Serantak Volcanics 253
metamorphic geology 264-73 Setul Limestone 208, 212, 214
origins 273-7 Shan boundary 99, 110
ophiolite emplacement 277 Shan Plateau
NW region palaeomagnetic studies
deep regional unconformity (DRU) 308-10 methods 212
structure 311 results 212-13
Rajang Group 254-6 stratigraphy 210-11
Sabah orogeny 258 Shan-Thai block 126, 132
Sagaing suture 99, 110 see also Sibumasu block
Sahul platform 62, 63 Shazipo Formation 541,544
Sainabouli Province see Pak Lay Fold Belt shear faults 3
Salawati Basin Sibela continental suite 483-4
correlation to Tomori Basin 475-6 Sibela ophiolite 486-7
hydrocarbon potential 479 Sibolga granite 324, 329
stratigraphy 467-72 Sibumasu block/terrane 99, 100-2
structure 467 tectonic history 204
tectonic significance 476-9 tectonostratigraphy 205
Saleh diorite 489, 503 Triassic collision studies 225, 230
564 INDEX

Sibuyan sea fault 519-20 Sulawesi Group 475

Siguangping Formation 541 Sulawesi orogeny 191-4, 197
Sikuleh granite 331 Sulu Sea 164
Silantek Formation 248, 254 basin tectonics 311-16, 317
Silurian studies 468 Sumatra
Simao block/terrane 99, 101 plutonism 323-5
Sintang intrusives 254 geochemistry 325-8
Sirga Formation 470 geochronology 328-9
slip partitioning 3, 6-16, 19 tectonic setting 160, 322-3
slip vectors 19 Sumatra fault 21, 24, 160
sliver microplates 19, 21-3 Sumatra fault system (SFS) 322
Sm/Nd isochron, western Sulawesi 407, 425 Sumatra margin
Snellius ridge 42-3 accretionary prism transfers 23-4
Solomon trench 13 plate setting 20
Song Da (Black River) suture 105-6, 204, 554 sliver plates 21-3
Song Ma (Red River) suture 99, 105-6, 204, 554 subduction rates 20
Songpan Ganzi accretionary complex 99 tectonic model 24-7
Sorol trough 6 Sumatra trench 8
Sorong fault 12-13,467 Sumatran Forearc 7, 337-8
Sorong fault system 465-6 see also Nias Island
Sorong-Sulabesi fault 467 Sumba 445
South China block/terrane 98-100 crustal isotopic signature 449-50
Cenozoic motion 124, 125 metamorphic belt 140-2
South China Caledonian orogen 549 Sunda orogeny 188-9, 197
South China Sea Sundaland 250, 281
models of tectonic evolution 247-8 collision evidence 133-5
spreading mechanism 133 deftned 291
tectonic setting 157 marginal features 353,354
see also Malay Basin; Penyu Basin northern shelf basin studies see Malay; Penyu;
South Sula fault 467 West Natuna; East Natuna; Mekong; Nam
South West Borneo terrane 99, 103 Con Son
Sturt block 62, 63 palaeomagnetic studies
Sukadana granite 252, 253 history of study 206-10
Sula Islands 474 methods 212
Sula platform 167 results 212-17
Sulabesi Island 467 results discussed 217-19
Sulan pluton 331 setting for sites 210-12
Sulawesi significance of results 219-22
Bantimala melange complex 354-5 tectonic setting 171,204
lithology 355 tectonic summary 300-5
outcrop distribution 356-61 see also Sumatra
structure 355-6 suturing events 105-11
tectonic evolution 361-2
collisional tectonics 171-4
displaced terrane 466
metamorphic belt 140-2 Tagoloan fault 514
tectonic setting 160-2 Taiwan collision zone 511
Tertiary magmatic evolution 392-4 Talaud orogeny 191,192, 197
distribution of igneous rocks 397 Talaud ridge 40
geochemistry 397-400, 425 Tamangil Formation 87
geochronology 400-11,424-5 Tanamu Formation 474
petrogenesis 411-13 Tang Ting fault 125
petrology 397 Tanimbar trough 86
thermal history 141-15 Tanjong Formation 313
Tertiary stratigraphy 365-9, 394 Tarim block/terrane 99, 124, 125
depositional environments 375-7 Tatau formation 250, 251
facies description 369-75 Tawali Besar, stratigraphy 487-8
Tertiary tectonics 383-4, 415-23 Tawali Formation 487-8, 491
 INDEX 565

tectonic overview of SE Asia 155 Fujian province 551-2, 553

Temburong Formation 256 palaeomagnetism 206
Tengchong Block 540 rifting 104
diamictite 541-4 Salawati stratigraphy 468
terrane motion 544-6 Sumatra plutonism 324, 329-30
terrane evolution terrane distribution 112
origins 98-103 Tomori stratigraphy 475
palaeogeography 111-14 Trusmadi Formation 256
Tertiary studies Tukang Besi platform 160, 167, 434
Bacan stratigraphy 487
Bacan tectonics 490-3
Salawati stratigraphy 471
Sumatra plutonism 331-3 U/Pb dating, western Sulawesi 410-11,425
Sunda Shelf kinematics 301 Ujung Kulon fracture zone 21
Tomori stratigraphy 475 Ulai intrusion 331
see also Eocene; Miocene; Oligocene; Neogene;
Pliocene
Tetambahu Formation 474 Verde Passage fault 519-20
Tethys Ocean, phases of 103-5,546 Visayas, Neogene accretion 515-16
Thai-Malay Peninsula 159-60 volcanic arc, role in Banda orogen 55-6, 57
Thailand Vulcan basin 62, 63
hydrocarbon potential 244
palaeomagnetic studies 207-10
rotation 217-18
tectonic setting 539 Waaf Formation 470
Thangon Formation 236 Waigeo, tectonic setting 491
thermochronology, Papua New Guinea 531-2 Walanae Depression 366
Three Pagodas fault 159 Wallace Line 394
thrust faults 3 Wang Chao fault 159
Tibet block, Cenozoic motion 124, t25 Weber basin 87
Timor Weda Group 503, 505
role in Banda orogen 57 Weduar Formation 87
structural model 47 Weryahan Formation 87
TIMOR line 48, 49-56, 64, 76, 78-82 West Andaman fault 22, 23
Timor trough 11, 51-3, 86 West Burma Block 99, 102
evolution 75-6 West Kutei basin 253
seismic reflection profile 81 West Malaysia see Malay Peninsula
Tipuma Formation 468 West Natuna Basin 282, 298-9
Tokala Formation 474 stratigraphy 283
Tomori Basin 472-3 West New Guinea composite terrane 140,
correlation to Salawati Basin 475-6 141
hydrocarbon potential 479 West Philippine basin 32
stratigraphy 470, 471,474-5 Wetar thrust 11, 61
structure 473-4 Wharton basin 19
tectonic significance 476-9 Woi Formation 504
Tomori Formation 475 Woniusi Formation 541,543-4, 544
Tonasa Limestone Formation 354, 365, 366 Woyla suture 99, 110
depositional environment 377-80 Woyla terranes 99, 103
facies 368-75
origins 375-7
stratigraphy 367-9 Xianggui block 100
tectonic environment 383-4 Xikou Formation 551-2, 554
Tondo Formation 435, 438, 439, 440 Xilao-He suture 99
Tonga trench 8, 13-15
Torricelli Intrusive Complex 527
Torricelli terrane 147
Triassic studies Yangtze block 100
collisional tectonics 225 Yangtze para-platform 549
566
Yangtze Platform 539, 540
Yanshanian orogenies 549, 555
Yap trench 6
Yefbie Shale 468
Yongde Formation 541, 544
Yunnan (SW) 539-41
INDEX
glacigenic sediments 541-3
origins 544-6
'Zhen'he-Da'pu fracture 549-50
Zishi Formation 541