Professional Documents
Culture Documents
1979 The South American Herpetofauna
1979 The South American Herpetofauna
1979 The South American Herpetofauna
Duellman
Editor
Library of the
Museum of
Comparative Zoology
THE SOUTH AMERICAN HERPETOFAUNA:
ITS ORIGIN, EVOLUTION, AND DISPERSAL
THE SOUTH AMERICAN
HERPETOFAUNA: ITS ORIGIN,
EVOLUTION, AND DISPERSAL
WILLIAM E. DUELLMAN
EDITOR
and
MONOGRAPH
OF THE
MUSEUM OF NATURAL HISTORY,
THE UNIVERSITY OF KANSAS
NUMRER 7
1979
MONOGRAPH OF THE MUSEUM OF NATURAL HISTORY,
THE UNIVERSITY OF KANSAS
Number 7, pages 1^85, 172 figures in text
Issued December 28, 1979
© 1979 by The Museum of Natural History, The University of Kansas, Lawrence, Kansas.
All rights reserved. No part of this book may be reproduced in any form or by any means without permission
in writing from the publisher.
JUN 51985
HARVARD
UNIVERSITY PRINTED BY
UNIVERSITY OF KANSAS PRINTING SERVICE
LAWRENCE, KANSAS, USA
Dedicated
to the memories of
three herpetologists who
contributed so much to our knowledge of
the South American herpetofauna:
11. Origin and Distribution of the Herpetofauna of the Dry Lowland Regions of
Northern South America.
Carlos Rivero-Blanco and James R. Dixon 281
15. The HeqDetofauna of the Andes: Patterns of Distribution, Origin, Differentiation and
Present Communities.
William E. Duellman 371
16. Refugia, Refuges and Minimum Critical Size: Problems in the Conservation of the
Neotropical Herpetofauna.
Thomas E. Lovejoy 461
United States; most of these were reported on In this brief review, each family is dis-
amphibians and reptiles broadens to include glossa and Oedipina) also occur in South
studies on the ecology, life history, and be- America. There, Oedipina (2 species) occurs
havior, the need still remains for descriptive only in the Choco, whereas Bolitoglossa (24
morphology and systematics. Ever increasing species) inhabits the Choco, Amazonia, and
human disturbance of natural environments, the northern Andes. Plethodontid salaman-
especially the rainforests, eliminates forever ders entered South America from Central
many components of the biota before they be- America after the closure of the Panamanian
come known to science. Portal (Wake, 1966).
1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA
Table 1:1. — Taxonomic Composition of the South Table —
1:3. Comparison of Numbers of Species of
American Herpetofauna. Tetrapod Vertebrates in South America with World
(
° = endemic to South America) Fauna.
Genera Species
Family Total Endemic Total Endemic
Amphibia
Plethodontidae 2 24
Pipidae 1
Leptodactylidae 41 37
Bufonidae 7 4
Brachycephalidae" .. 2 2
Rhinodermatidae° .. 1 1
Dendrobatidae — 3
Pseudidae" 2 2
Hylidae 22 15
Centrolenidae 2 1
Ranidae 1
Microhylidae 16 13
Rhinatrematidae° .... 2 2
Typhlonectidae" .... 4 4
Caeciliidae 9 6
Reptilia
Pelomedusidae 1
Chelidae 7
Kinostemidae 1
Chelydridae 1
Emydidae 2
Testudinidae 1
Gekkonidae 16 7
Iguanidae 27 20
Teiidae 38 28
Scincidae 1
Anguidae 2 1
Amphisbaenidae 6 5
Anomalepidae 4 1
Leptotyphlopidae _ 1
Typhlopidae 1
Boidae 5 2
Aniliidae 1 1
Tropidophiidae 2
Colubridae 77 39
Micruridae 2 1
Viperidae 3
Crocodylidae 4
MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
cene (Baez and Gasparini, this volume). Al- America are indicative of immigration of eleu-
though the family is unquestionably of Gond- therodactylines into Central America prior to
wanan origin, the relationships of the lepto- the establishment of the isthmian link in the
dactylids are not clear. Lynch (1971) recog- late Pliocene (Savage, 1973; Lynch,
1976).
nized the South American and Australian frogs The Ceratophryinae are widespread in Cha-
plus the South African Heleophryne as one coan, Amazonian, and Guianan lowlands. The
family, the Leptodactylidae, but he (1973) Elosiinae are restricted to the Brasilian
separated the Old World genera into the Shield. The Leptodactylinae
are widespread
Myobatrachidae. This arrangement was fol- in tropical and subtropical lowlands, with a
lowed generally by Savage (1973), Duellman primitive genus (Pleurodema) also inhabiting
(1975), Heyer (1975), and Heyer and Liem Patagonia, austral forests, and the Andes
( 1976), but
not by Tyler (this volume). Hey- (Duellman and Veloso, 1977). Physalaemus,
er ( 1975 ) suggested that leptodactylids might Pleurodema, and Leptodactylus have entered
have evolved from leiopelmatids; this idea was Central America, and the latter also is in the
elaborated upon by Lynch (1978). West Indies.
Within South America, the primitive Tel- —
Bufonidae. The earliest fossil bufonids
matobiinae are primarily distributed in tem- are from the Paleocene of Brasil (Estes and
—
perate regions the tribe Telmatobiini in Pat- Reig, 1973), followed by the Oligocene Neo-
agonia, austral forests, and the high Andes. procoela, which is a member of the Eurasian
More advanced telmatobiines are in temper- Bufo calamita group, according to Tihen
ate and tropical regions
—Odontophrynini in ( 1962)
and Baez and Gasparini (this volume)
the Chaco, southeastern Brasil, and nonfor- but referred to the telmatobiine leptodactylids
ested regions in eastern Brasil, Grypiscini on by Lynch ( 1971 ) By the Miocene, Bufo was
.
the Brasilian Shield, Eleutherodactylini most present in South America, North America,
diverse in northwestern South America but Europe, and Africa (Tihen, 1972). The ab-
also occurring on the Brasilian and Guianan sence of bufonids from the Australo-Papuan
shields and in Amazonia, and also speciose in Region (except for the introduced Bufo ma-
Middle America and the West Indies. The rinus), combined with the fossil history of the
diversity of eleutherodactyline genera and the group, strongly suggests a western Gondwana-
differentiation of Eleiitherodactyhis in Middle land origin of the family ( Blair, 1972; Savage,
Table 1:4. — Postulated Geographic Origins of Families of Amphibians and Reptiles Inhabiting South America.
(NA = North America; SA = South America; f= Extinct in South America)
Tropidophiidae (SA)
Micruridae (SA)
1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA
Table 1:5. — Distribution of Herpetofaunal Family Groups in Major Eco-physiographic Regions
in South America.
g 2
•c 13
o o
Family group
c/j
c
J3
c c 3 s
3
c
o O < O 03 O u
Amphibia
Plethodonndae + + + +
Pipidae
+ + + +
Ceratophryinae + + + + + + +
Telmatobiinae + + + + + + + + + + +
Elosiinae + +
Leptodactylinae . + + + + + + + + + + +
Bufonidae + + + + + + + + + + + +
B rachycephalidae +
Rhinodermatidae +
Dendrobatidae + + + + + + + +
Pseudidae + + + +
Phyllomedusinae + + + + + + +
Hemiphractinae + +
Amphignathodontinae + + + + +
Hyhnae + + + + + + +
Centrolenidae + + + + +
Ranidae + + +
Microhylidae + + + + + + +
Rhinatrematidae + + + +
Typhlonectidae + + + +
Caeciliidae + + + + + +
Reptilia
Pelomedusidae - + + +
Chelidae + + + + +
Kinosternidae + + + + + + +
Chelydridae +
Emydidae + + + +
Testudinidae + + + + + + + +
Gekkoninae + + + + + + + +
Sphaerodactylinae + + + +
I guanines + + + + +
Basiliscines + +
Anolines + + + + + +
Tropidurines + + + + + + + +
Teiidae + + + + + + + +
Scincidae + + + + + + +
Anguidae + + + + +
Amphisbaenidae + + + + + + + +
Anomalepidae .._ + + + + + +
Leptotyphlopidae + + + + + + +
Typhlopidae + + + + +
Boidae + + + + + + +
Aniliidae + +
Tropidophiidae + +
Xenodontinae + + + + + + + +
Colubrinae + + + + + +
Micruridae + + + + + + +
Crotalinae + + + + + + +
Crocodylidae + + + + + + +
MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA
1973; Laurent, this volume). Bufo and six and thus arc American origin. Spe-
of South
other genera occur in South America, and cies of all three genera occur in lower Central
Bufo and seven other genera occur in Africa America, presumably having arrived there
(principally tropical western Africa), but after the closure of the Panamanian Portal in
Bufo and five other genera inhabit southeast- the late Pliocene.
ern Asia and adjacent islands. One genus Pseudidae. —An autochthonous South
(Crepidophryne) is endemic Central
to American family containing two genera and
America, and Bufo is widespread in the Hol- four species (Gallardo, 1961) and widely dis-
arctic Region. Differing views have been ex- tributed in tropical and subtropical cis-An-
pressed on the dispersal of Bufo ( Blair, 1972; dean lowlands, these aquatic frogs have been
Savage, 1973; Duellman, this volume; Lau- considered as relatives of the leptodactylids
rent, this volume). (Savage and Carvalho, 1953) or hylids
Bufo occurs throughout South America, (Lynch, 1973).
but only members of the Bufo spinulosus HyUdae. Although— Estes and Reig
group are present in Patagonia, the austral (
1973 ) mentioned the existence of Paleocene
forests, and the high Andes (Cei, 1968, 1972). hylid material from Brasil, these specimens
Rhamphophryne and Atelopus are primarily have not yet been described. The mid-Mio-
northern Andean; Dendrophryniscus is in cene Australobat melius from Australia has
Amazonia and the Brasilian Shield, Melano- been referred to the Hylidae by Tyler ( 1974 ) ;
phryniscus in the Chaco and adjacent areas, a presumed hylid is known from the Oligo-
and Oreophrynella in the Guiana Highlands cene of North America (Holman, 1968), and
(Trueb, 1971; McDiarmid, 1971). Hyla is known from the Miocene of Europe
—
Brachycephalidae. Unknown in the fossil (Noble, 1928). By far the greatest diversity
record, the two small frogs comprising this of hylids South America ( 22 genera, 313
is in
family are restricted to humid coastal low- species), as compared with Middle America
lands of southeastern Brasil (Izecksohn, (
15 genera, 129 species; Duellman, 1970). Six
1971). Although superficially resembling a Hyla, plus two endemic genera (Osteopdus
specialized bufonid, brachycephalids lack and Calyptahyla) inhabit the West Indies
Bidder's Organs (McDiarmid, 1971), an ( Trueb and Tyler, 1974 )
. The Holarctic hylid
uniquely derived character in the Bufonidae fauna depauperate, but in the Australo-
is
(Lynch, 1973). The phylogenetic position of Papuan Region 118 species are known in the
this endemic South American family is not genera Litoria and Nyctimystes (Duellman,
clear, but presumably it arose from a lepto- 1977; Tyler and Da vies, 1978), and nine more
dactylid-primitive bufonid stock. if Cyclorana is included in the
family (Tyler,
—
Rhinodermatidae. Known from two spe- et al., 1978; Tyler, this volume). Like the
cies restricted to austral forests (Formas, et leptodactylids, the Australian hylids are of
al., 1975), Rhinoderma
considered to be
is questionable relationship with the South
most closely related to the bufonids by Lynch American hylids. Savage (1973) resurrected
( 1971, 1973 ) and must be considered as of the family name Pelodryadidae for the Aus-
temperate South American origin. tralo-Papuan "hylids" and considered them to
—
Dendrobatidae. Lacking a fossil record be derived independently from the Neotropi-
but composed of three Recent genera, the cal hylids. Tyler (this volume) emphasized
dendrobatids are especially speciose in the the lack of evidence for such an arrangement.
northern Andes, Choco, and western Ama- In Australia, Cyclorana seems to be intermedi-
zonia, but also occur in eastern Amazonia and ate between the Australian "leptodactylids"
on the Guianan and Brasilian shields. Lynch and "hylids" and may prove to establish a
( 1971 ) demonstrated that the dendrobatids phylogenetic link between the two families on
are derived from the elosiine leptodactylids that continent. No such intermediates are
Fig. 1:1. Major eco-physiographic regions of South America. Temperate regions: Austral forests (AF),
Patagonia (PAT). Amazonia (AM), Choco (CH), Atlantic coast (AC). Tropical
Tropical evergreen forests:
and subtropical nonforests: Caribbean coastal desert (CD), Llanos (LL), Savannas (black), Caatinga (CA),
Cerrados (CE), Gran Chaco (GC), Pampas (PA), Monte (MO), Espinal (ES), Matorral (MA), Atacama
Desert (AD). Mountains (stippled): Andes (A), Guiana Highlands (G), Brasilian Highlands (B).
Regiones ecofisiograficas mayores de Sudamerica.
8 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
continent, except one species of Phrynohyas compelling arguments based on diverse mor-
and several species groups of Hyla that enter phological, developmental, and biogeographic
Central America. Two species of the Middle evidence against the Starrett and Savage
American Smilisca enter South America. model.
All South American microhylids belong to
Assuming that at least the Neotropical
hylids arose in South America, some stocks the subfamily Microhylinae, which is shared
must have entered Central America by waif with North America, tropical southeastern
dispersal prior to the closure of the Panaman- Asia, and the Malayan Archipelago. It is most
ian Portal in the late Pliocene. These stocks logical biogeographically and phylogenetically
were the ancestors of the several genera and that the Neotropical microhylids evolved in
isolation in South America and that a stock
species groups of Hyla endemic to Middle
America. After the closure of the portal sev- that subsequently gave rise to the North
eral groups dispersed northward into Central American Ga.strophryne and Hypopachus
America (PhyUomedusa, Hemiphractus, Gas- managed to enter Central America from the
trotheca, Phrynohyas, Hyla albomarginata, south during mid-Cenozoic times; microhy-
H. boans, H. bogotensis, H. leucophyllata, and known from the Miocene
lines are of Florida
H. rubra groups) and representatives of two (Holman, 1967). The 16 genera of South
Middle American genera (Agalychnis and American microhylids occur throughout the
Smilisca) dispersed into South America. tropical and subtropical lowlands with the
Centrolenidae. —
No fossils are known. greatest diversity in the southern part of their
Two genera and 46 species inhabit cloud for- range, particularly on the Brasilian Shield.
ests in the Andes, whereas a few species occur Three genera (3 species) entered Central
on the Guianan and Brasilian shields and in America after the connection of the conti-
nents.
Amazonia, the Choco, and Central America
(Duellman, 1977). Obviously of South Amer- The monotypic Geobatrachus in the Sierra
ican origin with Late Cenozoic dispersal into Nevada de Santa Marta in northern Colombia
Central America, the relationships of the cen- tentatively was referred to the Microhylidae
trolenids usually are thought to be with the by Lynch (1971), but Duellman (1975)
showed that this small frog has a combination
hylids, but no convincing evidence
is avail-
any family as presently defined. Geobatra- South America (Eocene to Pleistocene). The
chus has not been included in the numerical seven genera endemic to South America are
account of the microhylids. widely distributed in the cis-Andean tropics,
—
Rhinatrematidae. Lacking fossils and en- mostly in Amazonia.
demic to South America, these primitive cae- Kinosternidae. —
Known as far back as the
cilians are a sister group of the ichthyophiids Oligocene in North America, the great ma-
(Nussbaum, 1977), a family restricted to In- jority of kinosteraids (4 genera, 19 species)
dia, tropical southeastern Asia, Malayan Arch- occur in North America and northern Central
ipelago, and the Philippines. Roth South America. Only three species of Kinosternon
American genera occur on the Guianan Shield occur in lower Central America; one of these
and one (Epicrionops) has four species on also is widespread in cis-Andean South Amer-
the forested slopes of the northern Andes. ica, and two vicariant species occur in the
—
Typhlonectidae. The specialized aquatic Choco. The kinosternids obviously are a post-
caecilians are autochthonous to South Amer- portal entrant into South America from the
ica, where they are distributed discontinuous- north.
ly in the Caribbean and Amazonian lowlands —
Chelydridae. The snapping turtles have
and in the Parana Rasin. an extensive fossil record throughout the
Caeciliidae. —The presence of a single fos- Cenozoic in North America, where two genera
sil from the Paleocene of Rrasil (Estes and are extant. One species of Chelydra inhabits
Wake, 1972 ) possibly referable to this family,
,
lower Central America, and one occurs in the
signifies a long history of caecilians in South Choco in South America. Chelydra obviously
America, where nine genera and 46 species is a recent
immigrant into South America.
now occur in the humid lowland tropics and —
Emydidae. This family has an extensive
forested slopes of the northern Andes. Four fossil record in the Holarctic Region and to-
genera inhabit Middle America; the endemism day is distributed mainly in North America
there in Dermophis and Gymnopis indicates and the Oriental Region. The genus Chry-
that a caecilian stock entered Central America semys (=Pseudemys) is speciose in North
prior to the establishment of the isthmian America and in the West Indies, occurs in
link,whereas the other two (Caecilia and Central America (same species in northern
Oscaecilia) are both widespread in South South America), and is represented by anoth-
America, and evidently dispersed into Central er species in the Parana Rasin. Two species
America after the closure of the Panamanian of the Central American Rhinoclemys occur
Portal. The presence of caeciliids in tropical in the Choco, and Rhinoclemys is known from
Africa (6 genera), India (3 genera), and the the Pleistocene of Ecuador. Possibly an early
Seychelles Islands (3 genera), as well as in Chrysemys stock waifed to South America,
South America, is indicative of a widespread but Rhinoclemys and Chrysemys scripta cer-
Gondwanan distribution prior to the Late tainly entered the continent from the north
Cretaceous. subsequent to the establishment of the isth-
Pelomednsidae. —The classical, present mian link.
Gondwanaland distribution pattern of pelo- Meiolaniidac. —This extinct family known
medusid turtles is complicated by their occur- from South America (Late Cretaceous to
rence in Cretaceous deposits in Europe and early Eocene) and Australia (Miocene to
North America and in the Eocene of Asia. Pleistocene) seems to antedate the testudi-
However, the family has an extensive fossil nids. Their fossil record suggests a Gondwa-
record beginning in the Cretaceous in both nan history similar to that of the chelids.
South America and Africa (Wood, 1970). The —
Testudinidae. This family is cosmopoli-
single South American genus, Podocnemis, is tan, except in the Australo-Papuan Region,
widely distributed in the cis-Andean tropical and possibly had its initial radiation in Lau-
lowlands. rasia (Cracraft, 1974). The genus Geoche-
—
Chelidae. Considered to be a derivative lone occurs today in South America, Africa,
of the Pelomedusidae (Gaffney, 1975, 1977), India, southeastern Asia, and on the Gala-
the chelids are known are fossils only from pagos Islands; it is known from the late Oligo-
Australia (early Tertiary to Pleistocene) and cene through the Pleistocene in South Amer-
10 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
cur in Africa, southern Asia, and North Amer- South America and has two species in Pana-
ica. The diplodactylines are dominant in ma. Sphaerodactylus is most speciose in the
Australia and have dispersed onto islands in West Indies but with a few species in Central
the southwest Pacific. Sphaerodactylines are America, two of which extend into South
restricted to tropical America. The gekko- America.
nines are pantropical, being most diverse in Iguanidae.
—This large and diverse family
the Indian, Oriental, and Ethiopian regions is first known
in the fossil record from the
but also with many representatives in Aus- Upper Cretaceous of Brasil (Estes and Price,
tralia and on Pacific islands. Of the 11 gen- 1973) and is diverse in the late Paleocene of
era of gekkonines in South America, only Brasil (Estes, 1970). The earliest North
eight are endemic to the American tropics, if American fossils are from the Eocene. Evi-
the Old World species presently assigned to dently an early iguanid stock reached North
Phyllodactyhis are not considered to be con- America prior to the separation of the conti-
generic, as suggested by Dixon and Anderson nents or waifed between the two. Informally,
(1973). Except for the speciose Phyllodac- the iguanids are divided into five major groups
tyhis in dry habitats in western and northern (Etheridge, 1964, 1967). The sceloporines dif-
South America (also Middle America, West ferentiated in North America and were paral-
Indies, and Galapagos Islands), most of the leled by an extensive radiation of tropidurines
endemic genera are represented by only one in temperate South America. The iguanines
or two species and all live in eastern South and basiliscines are primarily Middle Amer-
America, save for the monotypic Thccadacty- ican with genera of the former endemic to the
lus in western Amazonia, the Choco, Guianan Galapagos Islands (Amblyrhynchus and
Shield, Central America, and Lesser Antilles. Conolophus) and the West Indies (Cyclura);
The only genus with more than two species is one species of Iguana and three of Basiliscus
Homonota (8 species) in temperate cis-An-
dean areas; the related Garthia with two spe- 'Smith et al. (1977) recognized the South American
Lygodactylus as the sole representatives of the
genus
cies occurs in the southern Atacaman Region.
Vanzoia, but Kluge (pers. coram.) informed me that
The other gekkonines in South America the South American species are perfectly good exam-
( Gymnodactylus
— 3 species, Hemidactylus — ples of Lygodactylus.
1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 11
have entered northern South America. The 75 additional species in Africa, Madagascar,
anolines are widely distributed in the tropics southern Asia, and the Pacific islands, plus an
of South America, Middle America, and the additional two species in Central America and
West Indies.Obviously waif dispersal be- the Antilles. Tihen (1964) suggested an Eur-
tween South America and the emerging An- asian origin of the North American skinks
tilles and Central America during the Tertiary (Eumeces and Scincella), but an African ori-
accounted for some of the patterns of distribu- gin of the South American Mabmja seems to
tion, many of which have been masked by be reasonable.
more recent dispersal after the closure of the
—
Anguidae. Presently distributed primar-
Panamanian Portal. It certainly seems safe to ily in North America, western Eurasia, and
assume that the origin of tropidurines and southeastern Asia, the anguids have an ex-
some anolines was in South America. tensive fossil history dating from the Late
The presence of two genera of iguanids on Cretaceous in North America (Meszoely,
Madagascar has been interpreted as evidence 1970). Two genera occur in South America.
for the former occurrence of the family in Diploglossus is speciose in Central America
Africa with subsequent extinction there (per- and the West Indies, and one of the South
haps owing to competition with agamids and American species is shared with Central
chamaeleontids). The similarity of caudal America. The endemic South American Ophi-
structure of the Madagascaran iguanids to the odes (4 species) in the south-central part of
tropidurines (Etheridge, 1967) and the pres- the continent apparently evolved from an
ence of iguanids in the Cretaceous of South anguid stock that entered South America
America do not contradict that hypothesis. from the north in the Cenozoic.
The iguanine Brachylophus, endemic to Amphisbaenidae. —Numerous fossil am-
islands in the southwest Pacific, apparently is phisbaenians are known from the Paleocene
an example of long-distance rafting via the to Miocene in North America, Eocene to Plio-
Trans-Pacific Current (Cogger, 1974). cene of Europe, and Oligocene of Mongolia.
Teiidae. —
Although presumed teiids are With the exception of Bipes and the Pale-
known from the Late Cretaceous in North arctic Blanus, the amphisbaenids (sensu Ber-
America, those do not seem to be ancestral to man, 1973) are Neotropical and African 10 —
living North American teiids, whereas late genera in Africa (one ranging into Europe)
Paleocene teiids of South American resemble and six in South America (plus one endemic
extant Neotropical genera (Estes, 1970). All to Cuba). The generic and specific differen-
living teiid genera occur in South America, tiation in Africa (10 genera, 52 species) and
where they are widespread throughout the South America (6 genera, 45 species), and the
continent, except for Patagonia and the aus- possible presence of Amphisbaena in both
tral forests.Five genera have representatives Africa and South America (Gans, 1967), plus
in the West and 10 genera extend into
Indies, 10 species endemic to the West Indies and
Central America. With the exception of Cne- one South American species extending into
midophorns, which is widespread and spe- Central America, are suggestive of an African-
ciose in North America, all other teiids prob- South American amphisbaenid interchange.
ably arrived in Central America after the However, the place of origin of the amphis-
formation of the isthmian link. baenids remains problematic.
still
In South America the family is widespread in North America and the Eocene of Europe. A
trans-Andean and eis-Andean tropical low- Late Cretaceous snake, Dinilysia from Pata-
lands. gonia, is considered to be related to aniliids
—
Leptotyphlopidae. The single genus in ( Rage, 1977 and hue aniliids were reported
) ,
this family containing about 64 species is from the Eocene of Brasil by Baez and Gas-
widespread in tropical and subtropical South parini (this volume), who support Cracraft's
America, Middle America, southwestern ( 1974 ) contention that aniliids are of Laura-
United States, Africa, and southwestern Asia. sian origin. Nonetheless, entry into South
No fossils are known. On the basis of present America possibly was by way of Africa. The
distribution it is reasonable to suggest that one living South American aniliid is wide-
the leptotyphlopids had a western Gondwa- spread in cis-Andean tropical lowlands.
naland origin and subsequently spread north- —
Tropidophiidae. Structurally, the tropi-
ward into Central and North America and dophiids are intermediate between the boids
independently into Asia. and colubroid snakes. Tropidophis has three
—
Typhlopidae. These fossorial snakes are widely dispersed species in South America
known from two genera ( Typhlina, 33 species (northern Andes and southeastern Brasil) and
in Australia, New Guinea, Solomon and Fiji 12 species in the West Indies. Trachyboa and
islands) and Typhlops (about 114 species Ungaliophis occur in the Choco and Central
throughout tropical and subtropical parts of America. The tropidophiids are considered to
the world, except Australia). Only three spe- be of South American origin with subsequent
cies occur in South America; another five are northward dispersal.
in Central America, and 16 occur in the West
—
Colubridae. The poor fossil record and
Indies. Thus, with respect to the total differ- taxonomic chaos of the colubrids (sensu lato)
entiation of the family, the Neotropics are permit only the most general comments to be
poor in typhlopids. The only fossils (Eocene- made about this immense and important fam-
Miocene of Europe) are of no help in inter- ily. My use of subfamilial designations fol-
preting the paleobiogeography of the group. lows that of Dowling (1975) but eliminates
In the absence of any evidence for the occur- some apparent misapplications not especially
rence of typhlopids in North America, a west- germane to the Neotropical colubrids. A brief
ern Gondwanaland origin for the Neotropical summary of the colubrid snakes follows.
stocks might be suggested. 1. Xenodontinae: 93 genera, about 570
Boidae. —
Represented by an extensive, species. Sixty genera occur in South America;
world-wide fossil record from the Upper Cre- 22 of these are shared with Central America.
taceous through the Eocene (only Pleistocene Seven genera are endemic to the West Indies;
in Australia ) the boids seem to have been the
,
13 are restricted to North America (north of
dominant snakes throughout the world in the the Isthmus of Tehuantepec in southern Mex-
Early Cenozoic. Evidently they had dispersed ico), and 13 are Middle American. Included
widely before the breakup of Pangaea. The in this group of rear-fanged genera are fossor-
Early Cenozoic South American boid Madtso- ial, terrestrial, and arboreal snakes.
aquatic,
ia also is known from the Late Cretaceous of Most arboreal xenodontines are nocturnal, and
Madagascar, and a related boid, Wonambi, is few (Alsophis, Dromicus, Leimadophis, Ly-
known from the Pleistocene of Australia (Ty- gophis) are diurnal racer-like snakes. Most
ler, this volume). In some respects the dis- xenodontines feed on frogs, lizards, or other
tribution of these fossils parallels that of liv- snakes.
ing boines
—eight
genera in the Neotropics, Lycodontinae: 79 genera, about 285
2.
two in Madagascar, one in New Guinea and species. These rear-fanged snakes are distrib-
islands in the southwest Pacific. Presently the uted primarily in the African and Oriental
family is widespread in tropical South Amer- regions and peripherally in northern Australia
ica and especially diverse in Amazonia. and the Palearctic Region.
Aniliidae. —These
problematic fossorial 3. 74 genera, about 440 spe-
Colubrinae:
snakes have a long history in northern conti- cies.Widespread in the Holarctic Region,
nents, dating from the Middle Cretaceous in some genera occur in the Ethiopian and Ori-
1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 13
ental regions, and two genera reach Australia. vaded Central America, and some of
New World; them may have waifed to South Amer-
Thirty-two genera occur in the
South America, but only ica before the closure of the Pana-
of these, 12 are in
one of those (Drymoluber) does not occur in manian Portal in the late Pliocene.
Central America. Most of the colubrines in 5. With the closure of the Panamanian
South America Dendrophidion, Portal in the late Pliocene there was
(Chironius,
Dn/marchon, Drymobius, Drymoluber, Lep- an interchange of South American
Masticophis, Mastigodryas, Spilotes) xenodontines northwards and Middle
tophis,
are diurnal racer-like snakes that are terres- American colubrines and xenodontines
trial or arboreal. southward.
4. Natricinae: 34 genera, about 170 spe- 6. Natricines never have extended farther
cies (excluding from the Natricinae the Old south than Central America.
World snakes more appropriately referred to 7. The West Indian colubrid fauna is com-
Acrochordidae and Homalopsinae). The na- posed of xenodontines derived either
tricines are widely distributed in the Holand- from Central or South America (some
ric and Oriental regions, with a few represen- widespread colubrine species are recent
tatives in Africa and one in northern Aus- immigrants into the Lesser Antilles ) .
(Holman, 1976), when they first ap- America (Burger, 1971). The earliest North
pear in South America (Baez
and American fossil crotalines are of Miocene age
Gasparini, this volume). (Holman, 1977). The presence of the pre-
2. Xenodontines evolved in South Amer- sumably primitive crotaline Lachesis muta in
ica, where they are the dominant snakes northwestern South America and lower Cen-
today and the only colubrids in the tral America suggests that the crotaline vipers
southern part of the continent. entered South America after the establishment
3. Some xenodontine snakes were present of the isthmian link. On the other hand,
in Middle America in the Cenozoic; the presence of many species of Bothrops
these differentiated, and some of them throughout South America as far south as
Patagonia, as well as many different species
dispersed as far as northeastern North
America. in Middle America, is suggestive of an earlier
4. Colubrine stocks in North America in- dispersal into South America. Bothrops is
14 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
cies endemic to the llanos). The other two reptiles. Africa has one endemic family of
lizards (Cordylidae), one of caecilians (Sco-
genera (Melanosuchus and Paleosnchus) are
in western Amazonia. lecomorphidae), and one of frogs that has
dispersed to Madagascar and the Seychelles
Islands ( Hyperoliidae Australia has no en-
)
.
eluded the differentiation of genera and spe- Table 1:6. —Taxonomic Diversity of Anurans and
cies on each continent and the numbers of
Lizards on Gondwanan Continents.
sive in South America than in Australia, and Antilles, existed between nuclear Central
paramos stand in marked contrast to the m.y.b.p.) South America had direct land con-
nections with Africa (Grant, 1971; Reyment
pampas, llanos, caatinga, and Atacama Desert,
which harbor distinctly different bi-
in turn and Tait, 1972; Larson and Ladd, 1973) and
otas than do the lowland and montane rain- with Australia via Antarctica until the Eocene
forests. or Oligocene («50 m.y.b.p.) (McGowran,
1973; Veevers and McElhinny, 1976).
the mid-Cretaceous. An island arc, the proto- persing northward via the island route. Pre-
1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 17
Table 1:7. —
Postulated Herpetofaunal Exchange Be- America but known from the Paleocene-Oli-
tween South America and Central America. gocene of North America.
Across Via
Reptilian dispersal via the islands appar-
Family Panamanian Isthmian
Link ently was much more extensive than that of
group Portal
the amphibians. Probable dispersers via the
N^SS ->~N N -> S S -> N
— — + proto-Antillean island arc are primitive igua-
Plethodonhdae
— — ~ nid lizards ( south to north ) and anguid lizards
Pipidae
Eleutherodactylini
— + +? + (north to south). Dispersal via the Middle to
— — — + Late Tertiary Central American Archipelago
Leptodactylinae
Bufonidae — — + + included testudinids, iguanines, crocodylids,
Dendrobatidae — — — +
— and perhaps some colubrines and crotalines
Phyllomedusinae + + +
Hemiphractinae
— — — + (
all north to south ) and gekkonines, anolines,
Amphignathodontinae
— — — + teiids, leptotyphlopids, typhlopids, xenodon-
Hylinae
— + + + tines, and perhaps sphaerodactylines and mi-
Centrolenidae — — — + crurids (all south to north). The iguanine
Ranidae — — + —
— — + dispersal is postulated for the migration of
Microhylidae +
— + + + an Ambhjrhynchus-Conolophus stock to the
Caeciliidae
Kinostemidae — — + — Galapagos Islands from the South American
Chelydridae
— — + — mainland, but possibly this stock waifed di-
+ — — —
Testudinidae
— — rectly from Central America ( Avery and Tan-
Gekkoninae + + ner, 1971). Colubrine southward dispersal
Sphaerodactylinae
— + + +
— + — — probably was late in the history of the archi-
Iguanidae (primitive)
Iguanines + — + — pelago, if indeed these snakes did enter South
Leptotyphlopidae
— + — — persed via the proto-Antillean island arc.
Typhlopidae
— + — —
— — — Overland dispersal after the establishment
Tropidophiidae +
— of the isthmian link involved more northward
Xenodontinae + + +
Colubrinae +? + +? than southward dispersal by amphibians, but
Micruridae — +? +? + it did permit entry into South America for the
persed only into lower Central America and South American herpeto-
to a discussion of the
—
have not speciated there Pipa (eastern Pan- fauna, it is germane to this essay to ascertain
ama); Pleurodema, Bufo typhonius, Elachis- the herpetofaunal relationships of the two
tocleis,Relictivomer, Caecilia, and Oscaecilia regions inasmuch as many genera and some
(central Panama); Hemiphractus and Gastro- species are common to the two. The history
theca (western Panama); Glossostoma (Costa of the Caribbean Plate and the tectonic move-
Rica); and Bufo haematiticus (Nicaragua). ments in the Antillean-Caribbean region have
The South American Phyllomedusa buckleiji not been resolved, but Rosen (1975) sum-
group has a species endemic to lower Central marized (and extended) the existing geologi-
America (Duellman, 1970). All three genera cal data and proposed a plausible vicariance
of South American dendrobatids occur in low- model of Caribbean biogeography.
er Central America; each has undergone some Excluding introduced taxa, the herpeto-
speciation (Savage, 1968). The Atelopus vari- fauna of the West Indies ( not including Trini-
us group has invaded Central America (to dad, Tobago, Bonaire, Aruba, and Curacao)
Costa Rica) and has undergone a bewilder- consists of 505 species in 57 genera ( Schwartz
ing diversification (Savage, 1972). and Thomas, 1975 ) 476 of the species and 18
;
Late Tertiary and Quaternary overland of the genera are endemic to the West Indies.
dispersal amongst reptiles also was extensive. Schwartz ( 1978 ) gave a brief description of
Southward dispersal brought chelydrid and the herpetogeography of the West Indies.
kinosternid turtles, basiliscine lizards, and Twenty-two genera of reptiles are primar-
possibly colubrine and crotaline snakes into ily mainland taxa having one or two species
South America for the first time, whereas si- extending into the West Indies. Fifteen of
these are South American taxa that extend
multaneously the first Mabuija, Amphisbaena,
and anomalepid and tropidophiid snakes into the Lesser Antilles —
Phyllodactylus (also
reached Central America. Mesoamerican Greater Antilles), Thecadactylus, Iguana, Ba-
groups originally derived from South Amer- chia, Cnemidophorus, Gymnophthabnus, Kcn-
ican stocks (beta anoles, Cnemidophorus, and tropyx, Mabuya (also Greater Antilles), Boa,
many xendodontine snakes) dispersed into CoraUus, Chironius, Clelia, Mastigodryas,
South America. Many South American taxa Pseudoboa, and Bothrops. Five are Central
(Thecadactylus, alpha anoles, Ameiva, micro-
American taxa that extend into the Greater
teiids, and xenodontine and micrurid snakes) Antilles —
Gonatodes, Tretanorhinus, and
moved northward. Ctenosaura, Boa, and Coniophanes only reach-
The herpetofauna of eastern Panama con- ing Isla San Andres and/or Isla Providencia.
Three are North American taxa that reach
tains many genera that are chiefly South
American —Pipa, Hemi-
—
Cuba Natrix fasciata, Kinosternon bauri, and
Rhamphophryne,
phractus, Gastrotheca, Elachistocleis, Caeci- Crocodylus acutus; the latter also has in-
vaded the Lesser Antilles from South Amer-
lia,Geochelone, Lepidoblepharis, Enyalioides,
ica, and there is an endemic species of Croco-
Echinosaura, Amphisbaena, CoraUus, Trachy-
boa, Atractus, Diaphorolepis, and Pseudoboa. dylus on Cuba. The geckos Tarentola and
The herpetofauna of the Chocoan lowlands Hemidactylus may have arrived by waifing
of northwestern South America contains many
from any one of many sources, although the
other species of Tarentola are circum-Medi-
species that are familiar to the herpetologist
terranean. Two of the Leptodactylus and one
working in lower Central America, whereas
each of Eleuthcrodactylus and Hyla are main-
the fauna in the Amazon Basin is greatly dif-
land species. The five west Indian Chrysemys
ferent (at least at the species level); com-
pare data given by Savage (1966) with those probably stem from an invasion from North
America.
presented by Lynch (
this volume ) and Dixon
Among the endemic or taxonomically rich
(this volume).
genera in the West Indies, the hylid genera
(Calyptaliyla, Osteopilas, and Hyla) were
Herpetofauna of the West Indies
studied by Trueb and Tyler (1974), who in-
Although the West Indies are peripheral ferred five invasions of the Greater Antilles
1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 19
by separate hylid stocks, probably from South from an early sphaerodactyline invasion; if so,
America by rafting; however, it is possible the few mainland species (Central America
that the stocks for some of these were on the and Choco) are the result of dispersal of
stocks back to the mainland.
proto-Antilles and drifted part of the way
to
their present positions. The same might be All of the colubrid snakes ( save the North
true for the monotypic Cuban eleutherodac- American Natrix fasciata) in the West Indies
tyline Sminthillus and some of
the West In- are xenodontines. (1970) demon-
Maglio
dian stocks of Eleutherodactylus. Lynch strated relationships the seven endemic
of
(1971) suggested that most of the West In- genera and Alsophis with diverse mainland
dian Eleutherodactylus, plus Sminthillus and xenodontines and concluded that four sepa-
the Mexican Syrrhophus and Tomodactylus rate xenodontine invasions of the West Indies
possibly represented one eleutherodactyline from either Central or South America were
lineage and that the Eleutherodactylus inop- necessary in the evolution of the West Indian
tatus group of Hispaniola and the mainland xenodontines. Presumably the tropidophiid
Eleutherodactylus formed another lineage. If stock that gave rise to the 12 species of Tropi-
these suppositions are correct, minimally two dophis in the Greater Antilles and the Ba-
eleutherodactyline invasions of the West In- hamas came from South America, perhaps via
dies are required. The nine Greater Antillean Central America.
Bufo seem to be related (Schwartz, 1972), Too little is known about the relationships
but their affinities with mainland taxa have of the Antillean Aristeltiger, amphisbaenids,
yet to be determined. Leptotyphlops, and Typhlops to speculate on
Amongst the lizards, the dominant genus their origins, except that it is unlikely that
is Anolis, represented by two groups of spe-
they invaded from North America.
cies (alpha and beta, fide Etheridge, 1959),
plustwo endemic genera (Chamaeleolis and
Chamaelinorops in Cuba and Hispaniola, re- CONTINENTAL PATTERNS
spectively). The alpha anoles inhabit the OF DISTRIBUTION
Greater and Lesser Antilles and are wide-
spread in South America, whereas the beta It is becoming increasingly evident that
anoles occur in Central and South America the patterns of climate and vegetation have
and the Greater Antilles. Williams ( 1969 and changed drastically in South America since
in Trueb and Tyler, 1974 ) required minimally the beginning of the Cretaceous. Axelrod
two invasions of the Greater Antilles by Anolis (
1972 ) argued convincingly that the interior
and two for the endemic genera. Cyclura is of the large African-American continent was
related to Ctenosaura of Middle America arid prior to the birth of the South Atlantic
(Avery and Tanner, 1971); the ancestral stock Ocean, which brought maritime and mesic
of Cyclura presumably arrived in the Greater climates to western Africa and eastern South
Antilles from Central America. This also America for the first time. Some elements of
probably is true for the ancestral xantusiid the arid-adapted west Gondwanan flora sur-
stock that gave rise to Cricosaura endemic to vived in South America (and Africa) (Sol-
Cuba (Savage, 1963) and that of Diploglossus brig, 1976; Sarmiento, 1976), while much of
represented by some Central American and the continentwas mesic. Subsequent to the
many West Indian species. Possibly an earlier Eocene, temperate South America gradually
or separate invasion was responsible for the became cooler and drier Axelrod and Bailey,
(
the developing Humboldt Current resulted in instead I present a general picture of the di-
desiccation of the land in the Late Tertiary verse distribution patterns and give examples
of each.
(Jeannel, 1967). Pleistocene climatic fluctua-
tion effected the entire continent with cool
and warm periods in the south (Baez and Temperate Herpetofaunas
Scillato Yane, this volume), humid and dry —
Austral Forests. The cool, moist forests of
periods in the lowland tropics (Haffer, this
southern Chile and adjacent Argentina repre-
volume ) and extensive glaciation in the Andes
sent an unique biotope in South America,
(Simpson, this volume). characterized by a highly endemic herpeto-
Although the fossil record of the herpeto- fauna composed mostly of primitive lepto-
fauna in South America is still fragmentary,
sufficient material exists, when dactylid frogs ( Alsodes, Batrachyla, Caudi-
especially
verbera, Eupsophus, Hylorina, Tehnatobufo)
placed with the better data from mammals
and the paleofloras, to give a faint impression mostly restricted to forests south of 37°S Lat
of past distributions, especially in the southern (Fig. 1:2A). The distributions of all of the
mapped by Formas
species are this volume ) .
with the uplift of the Andes and probably are Patagonian herpetofauna contains some an-
cient relicts ( telmatobiine leptodactylid frogs
not older than the Pliocene. Speciation in
and some tropidurine iguanid lizards) but
many lowland tropical groups (Haffer, 1974,
also many species of Liolaemus that have dif-
this volume) and Andean groups (Simpson,
ferentiated in the Pleistocene (Cei, this vol-
1975, this volume; Duellman, this volume)
seems to have occurred in the Pleistocene. ume). Some Patagonian groups have relatives
in the adjacent monte and the pampas, or in
Thus, we are faced with
contrasting pic-
tures —presumed recent speciation and appar- the austral forests, but the major latitudinal
expansion has been northward in the Andes,
ently rapid evolutionary rates in the lowland
best exemplified by Liolaemus (Fig. 1:2B).
tropics and in the Andes, as well as in some
temperate groups adapted to xeric conditions, Tropical and Subtropical Herpetofaunas
versus the survival of many old taxa in habitat
refugia in the austral forests and also in the Herein distinction is made between two
ancient Brasilian and Guianan highlands primary biotopes, as follow: 1) Forests
—
(Hoogmoed, this volume). tropical evergreen forests, including rainforest
Various contributors to this volume have and cloud forest, and 2) Nonforests the de- —
analyzed distributions within certain regions ciduous, scrub or thorn forests, and savannas,
(e.g., Patagonia) or biotopes (e.g., lowland grasslands and deserts. Although each of
tropical rainforests); here I attempt to provide these categories, especially the latter, contains
a broad synthesis of patterns in the entire diverse vegetation formations, they seem to
continent. Data for many groups and/or re- have reality with respect to major patterns of
gions are inadequate for a detailed analysis; distribution of the herpetofauna.
1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 21
Fig. 1:2. Distribution patterns of the South American herpetofauna: A. Batrachyla in austral forests (black);
Centrolcnclla principally in cloud forests but also entering lowland tropical rainforest (stippled). B. Liolac-
mus, an austral group entering adjacent subtropical areas and extending northward in the Andes (stippled);
Pleurodema brachijops, widespread in nonforests in northern South America (black). C. Hyla parviceps
group with vicariant species in lowland rainforests; three species are in the Amazon Basin and one each in the
other areas (stippled); Phyllopezus with disjunct populations in nonforested areas of the caatinga, cerrados, and
pampas (black) (after Vanzolini, 1974). D. Osteocephalus with species on Andean and Guianan slopes and
others in lowland rainforests. E. Tropidurus with species inhabiting diverse nonforested environments through-
out tropical South America and the Galapagos Islands. F. Pleurodema, a temperate South American genus with
vicariant species in the Andes and in nonforested environments to Panama.
Patrones de distribution de la herpetofauna sudamericana. A. Batrachyla en los bosques australes (negro);
Centrolenella principahncnte en bosques neblinos pero tambien entra las tierras bajas de la selva lluviosa tropical
(punteado). B. Liolaemus, un grupo austral entra los areas subtropicales adyacentes y se extende hacia el
norte en los Andes (punteado); Pleurodema brachyops diseminado en ambientes no forestales en el norte de Sud-
america (negro). C. Hyla parviceps con especies vicarias en las tierras bajas del bosque lluvioso, tres especies
en la Amazonia y una especie en cada una de los otros areas (punteado); Phyllopezus poblaciones disjuntas en
areas sin bosque de caatinga, cerrado y pampas (segun Vanzolini, 1974) (negro). D.
Osteocephalus con espe-
cies en las laderas andinas y guianense y otras especies en las tierras bajas del
bosque pluvial. E. Tropidurus
con especies habitando diversos ambientes no forestales atraves de Sudamerica tropical y las Islas Galapagos.
F. Pleurodema, un genero de la region templada de Sudamerica con especies vicarias en los Andes y en ambi-
entes no forestales hasta Panama.
22 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Forests. —
The vast Amazonian rainforests and east of the Amazon Basin are the dry
and the smaller areas of rainforest in the areas of northeastern Brasil (Caatinga), the
Choco and along the southeastern coast of interior savannas (cerrados), and the Gran
Brasil contain the richest herpetofaunas in Chaco. Subtropical, cis-Andean, nonforests
South America (Lynch, this volume; Dixon, include the pampas, espinal, and monte in
this volume). Especially diverse in these for- Argentina; west of the Andes are the matorral
ests are dendrobatid and hylid frogs, anoline and the Atacama Desert (Fig. 1:1).
and teiid lizards, and xenodontine colubrid Distributions of many species of plants
snakes. The herpetofaunas of the montane and animals indicate that various combina-
rainforests or cloud forests in the Andes, Gui- tions of these nonforest environments were
ana Highlands, and the Brasilian Highlands continuous with one another in the not-too-
are primarily altitudinal extensions of the distant past. Pleistocene climatic fluctuations
lowland groups (Hoogmoed, this volume; resulted in drier periods ( interglacials ) that
Duellman, this volume). However, in the allowed for expansion of the nonforests ( Haf-
montane forests, certain groups are either en- fer, 1974; Vanzolini, 1976). Gallardo (1969,
demic or far more diverse than in the low- 1971, this volume) emphasized the faunal re-
—
lands frogs of the families Centrolenidae lationships among the chaco, pampas and
(Fig. 1:2A), Dendrobatidae (Colostethus), monte, and Vanzolini (1968, 1974, 1976)
and Leptodactylidae (Eleutherodactylus) demonstrated distribution patterns in the cer-
and salamanders of the genus Bolitoglossa rados and caatinga. The herpetofaunal rela-
(Andes only). tionships among the coastal deserts, llanos
Distribution patterns are highly variable and savannas of northeastern South America
(see Duellman, 1978, Fig. 197, for examples are analyzed by Rivero-Blanco and Dixon
of Amazonian distributions). A few wide- (thisvolume) and Hoogmoed (this volume).
spread species, such as Boa constrictor, in- The trans-Andean area consists of a nar-
habit all of the lowland forests, but these row coastal strip and Andean slopes to 2000-
species usually also inhabit intervening non- 3000 m, 1-37°S Lat. The dry upper Maranon
forest areas. More commonly, vicariant spe-
Valley and associated valleys in the Huanca-
cies occur in the different areas of rainforest bamba Depression are separated from the
(Fig. 1:2C). Widespread and speciose gen- trans-Andean arid zone by passes at less than
era,such as Eleutherodactylus, Hyla, and An- 3000 m. The coastal deserts and matorral
olis, are found throughout the lowland and have a small, but largely endemic, herpeto-
montane forests, but usually there are distinct fauna including three endemic genera of liz-
combinations of species at different elevations, ards —Garthia (2 species), Callopistes (2)
as shown for Eleutherodactylus by Lynch and Dicrodon The dominant groups are
(3).
and Duellman (1979). These patterns are two genera of Tropidurus Dixon and
lizards, (
nonforested biotopes are more extensive, di- Ecuador. This current has been important in
verse and fragmented than the forests. In rafting stocks of Atacaman reptiles Tropi-
—
northern South America are the coastal des- durus, Alsophis (=Dromi-
erts, savannas, and the extensive llanos; south cus)
—toPkyllodactylus,
the Galapagos.
1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 23
regions (Fig. 1:2B), whereas others are re- nosis for the study of the South American
stricted to the coastal deserts or llanos. Some herpetofauna. South America has the richest
ceratophryine frogs, some snakes, and several herpetofauna of any continent, but the fauna
lizards have series of populations, subspecies is still poorly known taxonomically. Our
from the caatinga south-
or species distributed knowledge of systematic and ecological rela-
westward in the cerrados to the chaco or even tionships is even less. Human devastation of
into the monte or pampas (Fig. 1:2C). Such vast areas of forest that a few years ago were
open habitats obviously were continuous, or unexplored is eliminating forever important,
nearly the past so as to permit the dis-
so, in and in many cases unknown, aspects of the
persal of such nonforest taxa as Tropidurus, biota. Although biologists have had some in-
Cnemidophorus, and Bufo granulosus (Webb, fluence on the control of this exploitation,
1978 ) . Some of the taxa in the Atacama Des- there is little hope that we can preserve all
ert are not represented east of the Andes, that we may
wish to save. Thus, we are faced
except in the upper Mar anon Valley, but —
with two courses of action salvage collecting
Phyllodactylus also is diverse in northern and preservation of diverse natural preserves.
South America, Middle America, and the Random collecting of the biota, even in
West Indies. Tropidurus widely distributed
is
reasonably well known areas, commonly re-
in tropical nonforested environments on both sults in new information on distributions, tax-
sides of the Andes (Fig. 1:2E). At least one
onomy, or life histories. However, collecting
temperate group (Pleurodema) has dispersed efforts need to be intensified and planned to
northward in nonforests to northern South sample biota before they are destroyed. In the
America (Duellman and Veloso, 1977) (Fig. case of amphibians and reptiles, efforts must
1:2F). be made to obtain not only series of well-
preserved specimens but also tissues for karyo-
Montane Herpetofaunas
logical and biochemical studies, colored pho-
The
three major highland regions of South tographs, tape recordings of calls (of frogs),
America —
the ancient Guianan and Brasilian life history data, and extensive notes on habi-
highlands and the young Andes have little — tats and behavior. We
cannot necessarily ex-
in common herpetologically. With few excep- pect that a visit to the same region five or ten
tions, there are no isolated sister groups at years hence, or even next year, will permit
high elevations that do not have relatives at the collection of these data. The collection of
low elevations. Hylid frogs of the genera these kinds of materials and data must be
Cryptobatra chits (northern Andes) and Ste- encouraged and supported at every level inter-
fania (Guiana Highlands) and teiid lizards, nationally. The impending biological crisis
Euspondylus in the same regions, are primary has no national boundaries; responsible and
examples. Hylid frogs of the genera Gastro- effective collectors should be encouraged to
theca and Flectonotus occur in the Andes and make adequate collections throughout the
in the Brasilian Highlands, but some of these continent. Systematic collections under re-
species occur at low to moderate elevations, sponsible direction of trained biologists will
even though at present none lives in the inter- be one of the most important biological re-
vening lowlands. The herpetofaunas of the sources of the future; materials in these col-
highland regions seem to have been derived lections made available internationally to
independently from the adjacent lowlands. In qualified investigators will be the basis for not
the case of the Andes, the fauna is composed only systematic studies but much evolutionary
of a southern assemblage derived from Pata- synthesis.
gonia and a northern assemblage derived from The establishment of large natural reserves
the tropical lowlands (Duellman, this vol- in areas of high species richness and ende-
ume). mism can preserve large segments of the her-
24 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
petofauna. However, such reserves should not pielago proveo las veces de filtro entre Norte
be established strictly for conservation pur- y Sur America, el cual fue cruzado en ambas
poses. It is in these reserves that biologists direcciones por algunos grupos. El mayor in-
can undertake long-range studies of communi- tercambio entre las dos faunas se llevo a cabo
ties, population dynamics, behavior, and life una vez fue establecida la conexion entre los
history. The resulting kinds of information dos continentes. Seis familias de anfibios y 15
complement those obtained from salvage col- de reptiles son compartidas por Norte y Sur
lecting and contribute substantially to our America. Ademas cuatro familias de anfibios
totalunderstanding of the biota. y cuatro familias de reptiles sudamericanas
Actions of these kinds are necessary now; tambien se encuentran en Centro America.
a few years hence will be too late for some of Gran parte de la herpetofauna Antillana se
the regions and their herpetofaunas. Without compone de grupos neotropicales, algunos de
such actions the papers assembled in this vol- los cuales invadieron las especialmente
islas,
ume will not be the preliminary assessments las islas menores de las Antillas
provenientes
as intended, but instead they might be the de Sudamerica; otros grupos invadieron las
last word on the South American herpeto- islas desde Centro America.
reptiles en 203 generos y 22 familias (exclu- ono respondiendo a cambios de las condi-
los
yendo los taxa marinos). De los 318 generos ciones climaticas durante el Cenozoico; apar-
y 2,210 especies, 201 generos y 2,017 especies entemente muchas de las especies que existen
son endemicas de este continente. Entre las actualmente evolucionaron en el Quaternario.
familias de reptiles, no encontramos ninguna Las herpetofaunas de las regiones templadas
endemica en sudamerica; en cambio existen incluyen aquellas encontradas en los bosques
cinco familias endemicas de anfibios. australes y las estepas patagonicas. Los
Durante 4.5-50 millones de aiios la fauna bosques estan restringidos a zonas aisladas,
sudamericana evoluciono aislada del resto de mientras que las estepas se han dispersado
los continentes formaban Gondwanalandia, hasta el monte subtropical adyacente y hacia
sos y agregaciones asociadas con ambientes Burcer, W. L. 1971. Genera of pitvipers (Serpen-
sin bosqucs grandes
—pampas, monte, espinal, tes: Crotalidae). PhD Dissert. Univ. Kansas,
186 p.
chaco, matorral, cerrado, caatinga, sabanas, Cei, J. M. 1968. Remarks on the geographical dis-
llanos y desiertos. Cada una de estas regiones tribution and phyletic trends of South American
tiene su fauna caracteristica.Las herpeto- toads. Pearce-Sellards Ser. Texas Mem. Mus.
faunas de las Andes, en
tierras altas en los (13): 1-20.
Guiana y en Rrasil tienen poco en coniun; Cei, J. M. 1972. Bufo of South America, pp. 82-92
in Blair, W. F. (ed. ). Evolution in the genus
aparentemente cada una se derivo independi- Bufo. Univ. Texas Press, Austin, 459 p.
entemente de las faunas encontradas en las Cogger, H. G. 1974. Voyage of the banded iguana.
tierras bajas adyacentes. Australian Nat. Hist. 18:144-149.
Con de adquirir material para estu-
el fin Cogger, H. G. 1975. Reptiles and amphibians of
Australia. Reed, Sydney, 584 p.
dios en el futuro y conservar agregaciones de
Cracraft, J. 1973. Continental drift, paleoclimatol-
la herpetofauna sudamericana antes de
que la ogy, and evolution and biogeography of birds. J.
destrucion causada por el hombre elimine aun Zool. 179:455-545.
mas la biota, es necesario hacer extensas y Cracraft, J. 1974. Continental drift and vertebrate
bien documentadas colecciones de la herpeto- distribution. Ann. Rev. Ecol.
Syst. 5:215-261.
Dietz, R. S., Holden, J. C. 1970. Reconstruction of
fauna a lo largo del continente asi como
Pangaea: Breakup and dispersion of continents,
establecer grandes reservas naturales en areas Permian to present. J. Geophys. Res. 75:4939-
ricas en especies y endemismos donde estudios 4956.
biologicos de la fauna puedan ser llevados a Dixon, J. R., Anderson, S. C. 1973. A new species
cabo en el futuro. and genus of gecko (Sauria: Gekkonidae) from
Iran and Iraq. Bull. South. California Acad. Sci.
72:155-160.
Dlxon, R. Huey, R. B. 1970. Systematics of the
LITERATURE CITED lizards
J. ;
of the
gekkonid genus Phyllodactylus of
mainland South America. Nat. Hist. Mus. Los
Anderson, S., Jones, J. K., Jr. 1967. Recent mam- Angeles Cty. Contrib. Sci. (192):l-78.
mals of the world: A synopsis of families. Ronald Dlxon, J. R., Wright, J. W. 1975. A review of the
Press, New York, 453 p. lizards of die iguanid genus Troj)idurus in Peru.
Auffenberg, W. 1971. A new fossil tortoise, with Ibid. (271): 1-39.
remarks on the origin of South American testu-
Dowling, H. G. 1975. A provisional classification of
dines. Copeia 1971 ( 1 ) : 106-117. snakes. 1974 Yearbook Herpetology. Herpetol.
Avery, D. F., Tanner, W. W. 1971. Evolution of Infor. Search Syst., New York, pp. 167-170.
the iguanine lizards Sauria, Iguanidae ) as deter-
(
Duellman, W. E. 1966. The Central American
mined by osteological and myological characters.
herpetofauna: An ecological perspective. Copeia
Brigham Young Univ. Sci. Bull. Biol. Ser. 12(3): 1966(4):700-719.
1-79.
Duellman, W. E. 1970. The hylid frogs of Middle
Axelrod, D. I. 1967.
Drought, diastrophism, and America. Univ. Kansas Mus. Nat. Hist. Monogr.
quantum evolution.
Evolution 21:201-209.
(l):l-753.
Axelrod, D. I. 1972. Edaphic aridity as a factor in
Duellman, W. E. 1972. South American frogs of
angiosperm evolution. Amer. Nat. 106:311-320.
the Hyla rostrata group (Amphibia, Anura, Hy-
Axelrod, D. I., Bailey, H. P. 1969. Paleotempera-
ture analysis of Tertiary floras. Paleogeogr. Paleo- lidae). Zool. Meded. Rijks Mus. Nat. Hist. Leiden
47:177-192.
climatol. Paleoecol. 6:163-195.
Berman, D. S. 1973. Spathorhynchus fossorium, a Duellman, W. E. 1975. On the classification of
middle Eocene amphisbaenian frogs. Univ. Kansas Mus. Nat. Hist. Occas. Pap.
Reptilia ) from (
Bons, J., Pasteur, G. 1977. Solution histologique a Duellman, W. E., Crump, M. L. 1974. Speciation
un probleme de taxonomie herpetologique interes- in frogs of the Hyla parviceps group in the upper
sant les rapports paleobiologiques de I'Amerique Amazon Basin. Univ. Kansas Mus. Nat. Hist.
du Sud el de l'Afrique. Acad. Sci. Paris, Ser. D Occas. Pap. (23): 1-10.
284:2547-2550. Duellman, W. E., Veloso M., A. 1977. The phy-
Brodkorb, P. 1971. Origin and evolution of birds, logeny of Pleurodema (Anura: Leptodactylidae ) :
pp. 20-55 in Farner, D. S., King, J. R. (eds.). A biogeographic model. Ibid. (64): 1-46.
Avian biology, I. Academic Press, New York, Dunn, E. R. 1931. The herpetological fauna of the
586 p. Americas. Copeia 1931 (3): 106-119.
26 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Estes, R. 1970. Origin of the Recent North Amer- Greer, A. E. 1974. The generic relationships of the
ican lower vertebrate fauna: An inquiry into the scincid lizard genus Leiolopisma and its relatives.
fossil record. Forma Functio 3:139-163. Australian J. Zool. Suppl. Ser. (31): 1-67.
Estes, R. 1975. Fossil Xenopus from the Paleocene Haffer, 1969. Speciation in Amazonian forest
J.
Upper Cretaceous of Rrasil. Science 180:744- Neotropical Region: A zoogeographic and eco-
751. logical review, pp. 311-^131 in Keast, A., Erx,
F. C., Glass, B. (eds. ). Evolution, mammals,
Estes, R. Reig, O. A. 1973. The early fossil record
of frogs: A review of the evidence, pp. 11-63 in
and southern continents. State Univ. New York
Press, Albany, 543 p.
Vial, J. L. (ed.). Evolutionary biology of the
anurans: Contemporary research on major prob- Heyer, W. R. 1973. Systematics of the marmoratus
lems. Univ. Missouri Press, Columbia, 470 p. group of the frog genus Leptodactylus. Nat. Hist.
Mus. Los Angeles Cty. Contr. Sci. (251): 1-50.
Estes, R., Wake, M. 1972. The first fossil record of
Heyer, W. R. 1975. A preliminary analysis of the
caecilian amphibians. Nature 239:228-231.
intergeneric relationships of the frog family Lep-
Etheridge, R. 1959. The relationships of the anoles
todactylidae. Smithsonian Contrib. Zool. (199):
(Reptilia: Sauna: Iguanidae): An interpretation
1-55.
based on skeletal morphology. PhD Dissert. Univ.
Heyer, W. R., Analysis of the
Liem, D. S. 1976.
Michigan, 236 p.
intergeneric Australian frog
relationships of the
Etheridge, R. 1964. The skeletal morphology and
family Myobatrachidae. Ibid. (233): 1-29.
systematic relationships of sceloporine lizards.
Holdex, J. C, Dietz, R. S. 1972. Galapagos core,
Copeia 1964(4):610-631. NazCoPac and
triple junction Carnegie/Cocos
Etheridge, R. 1966. The systematic relationships of ridges. Nature 235:266-269.
West Indian and South American lizards referred Holmax, A. 1967. Additional Miocene anurans
J.
to the iguanid genus Leiocephalus. Ibid. 1966 from Florida. Florida Acad. Sci. 30:121-
Q. J.
(1):79-91. 140.
Etheridge, R. 1967. Lizard caudal vertebrae. Ibid. Holmax, J. A. 1968. Lower Oligocene amphibians
1967(4) :699-721. from Saskatchewan. Ibid. 31:273-289.
Formas, R., Pugin, E., Jorquera, R. 1975. La iden- Holman, J. A. Snakes from the Rosebud For-
1976.
bdad del batracio chileno Heminectes rufus Phi- mation (Middle Miocene) of South Dakota. Her-
lippi, 1902. Physis 34:147-157. petologica 32:41-48.
Gaffnev, E. S. 1975. A phylogeny and classifica- Holmax, J. A. 1977. Upper Miocene snakes (Rep-
tion of higher categories of turtles.
the Bull.
tilia,Serpentes) from southeastern Nebraska. J.
Amer. Mus. Nat. Hist. 155:387-436. Herpetol. 11:323-335.
Gaffney, E. S. 1977. The side-necked turtle family
Izecksohx, E. 1971. Novo genero e nova especie de
Chelidae: A theory of relationships using shared Brachycephalidae do Estado do Rio de Janeiro,
derived characters. Amer. Mus. Novit. (2620):1- Brasil (Amphibia, Anura). Bol. Mus. Nac. Rio de
28.
Janeiro, Zool. (280): 1-12.
Gallardo, J. M. 1961. On the species of Pseudidae. Jaxzex, D. H. 1976. The depression of reptile bio-
Bull. Mus. Comp. Zool. Harvard Univ. 125:108- mass by large herbivores. Amer. Nat. 110:371-
134. 400.
Gallardo, J. M.
1969. Las especies de saurios Jeannel, R. 1967. Biogeographie de l'Amerique
(Reptilia) de la Provincia de Santa Fe, Argen- australe, pp. 401-460 in Delamere Deboutte-
tina, y consideraciones sobre su ecologia y zoo- ville, C, Rapoport, E. (eds.). Biologie de
geografia. Neotropica 15:73-81. l'Amerique australe, 3. C. N. R. S. Groupe Fran-
Gallardo, J. M. 1971. Composicion faunistica de cais Argiles C. R. Reun. Etud., Paris, 834 p.
los saurios de laProvincia de La Pampa, Repvib- Keast, A. 1972. Continental drift and the evolution
licaArgentina. Ibid. 17:44—48. of the biota on southern continents, pp. 23-87 in
Gans, C. 1967. A check list of Recent amphisbae- Keast, A., Erk, F. C, Glass, B. (eds.). Evolu-
nians ( Amphisbaenia, Reptilia). Bull. Amer. Mus. tion, mammals, and southern continents. State
Nat. Hist. 135:61-105. Univ. New York Press, Albany, 543 p.
Grant, N. K. South Atlantic, Benue Trough,
1971. Kluge, A. G. 1967. Higher taxonomic categories of
and Gulf of Guinea Cretaceous triple junction. gekkonid lizards and their evolution. Bull. Amer.
Geol. Soc. Amer. Bull. 82:229.5-2298. Mus. Nat. Hist. (135): 1-60.
Greer, A. E. 1970. A subfamilial classification of Larson, R. L., Ladd, J. W. 1973. Evidence for the
scincid lizards. Bull. Mus. Comp. Zool. Harvard opening of the South Atlantic in the Early Creta-
Univ. 139:151-183. ceous. Nature 246:209-212.
1979 DUELLMAN: SOUTH AMERICAN HERPETOFAUNA 27
Lynch, J. D. 1971. Evolutionary relationships, oste- Noble, G. K. 1928. Two new fossil Amphibia of
ology,and zoogeography of leptodactyloid frogs. zoogeographic importance from the Miocene of
Univ. Kansas Mus. Nat. Hist. Misc. Publ. (53): Europe. Amer. Mus. Novit. (303): 1-13.
1-238. Nussbaum, R. A. 1977. Rhinatrematidae: A new
Lynch, J. D. 1973. The transition from archaic to family of caecilians (Amphibia: Gymnophiona).
advanced frogs, pp. 133-182 in Vial, J. L. (ed.). Occas. Pap. Mus. Zool. Univ. Michigan (682):
Evolutionary biology of the anurans: Contem- 1-30.
porary research on major problems. Univ. Mis- Patterson, B., Pascual, R. 1968. The fossil mam-
souri Press, Columbia, 470 p. mal fauna of South America. Q. Rev. Biol. 43:
Lynch, J. D. 1976. The species groups of the South 409-451.
American frogs of the genus Eleutlicrodactijlus Peters, J. A., Donoso-Barros, R. 1970. Catalogue
( Leptodactylidae). Univ. Kansas Mus. Nat. Hist. of the Neotropical Squamata: Part II. Lizards
Occas. Pap. (61): 1-24. and amphisbaenians. U.S. Natl. Mus. Bull.
Lynch, J. D. 1978. A re-assessment of the telmato- (297): 1-293.
biine frogs of Patagonia. Ibid. (72): 1-57. Pianka, E. R. 1971. Lizard species diversity in the
Lynch, J. D., Duellman, W. E. 1979. The Elcu- Kalahari Desert. Ecology 52:1024-1029.
therodactylus of the Amazonian slopes of the Pianka, E. R., Huey, R. B. 1971. Bird species di-
Ecuadorian Andes (Anura: Leptodactylidae). versity in the Kalahari and the Australian deserts.
Univ. Kansas Mus. Nat. Hist. Misc. Publ. (in Koedoe 14:123-130.
press ) .
biomes in the late Pleistocene. Proc. Zool. Soc. anurans: Contemporary research on major prob-
London 141:395-421. lems. Univ. Missouri Press, Columbia, 470 p.
Moreau, R. E. 1969. Climatic changes and distribu- Savage, J. M. 1974. The isthmian linkand the evo-
tion of forest vertebrates in West Africa. J. Zool. lution of Neotropical mammals. Nat. Hist. Mus.
158:39-61. Los Angeles Cty. Contrib. Sci. (260):1-51.
28 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Savage, J. M, Carvalho, A. L. de. 1953. The fam- Tyler, M. Davies, M. 1978. Species groups with-
J.,
ily position of
Neotropical frogs currently referred in the
Australopapuan hylid frog genus Litoria
to the genus Pseudis. Zoologica 38:193-200. Tschudi. Australian I. Zool Suppl. 63:1—17.
Savitzky, A. H. 1978. The origin of the New World Tyler, M. J., Davies, M., King, M. 1978. The Aus-
proteroglyphous snakes and its bearing on the tralian frog Chiroleptes dahlii Boulenger: Its sys-
study of venom delivery in snakes. PhD Dissert. tematic position, morphology, chromosomes and
Univ. Kansas, 387 p. distribution. Trans. Roy. Soc. South Australia
Schwartz, A. 1972. The native toads (Anura: Bu- 102:17-23.
fonidae) of Hispaniola. J. Herpetol. 6:217-231. Vanzolini, P. E. 1968. Geography of the South
Schwartz, A. 1978. Some aspects of the herpeto- American Gekkonidae (Sauria). Arq. Zool. (Sao
geography of the West Indies, pp. 31-51 in Gill, Paulo) 17:85-112.
F. B. (ed.). Zoogeography in the Caribbean. Vanzolini, P. E. 1974. Ecological and geographical
Acad. Nat. Sci. Philadelphia Spec. Publ. (13): distribution of lizards in Pemambuco, northeastern
1-128. Brasil (Sauria). Pap. Avulsos Zool. (Sao Paulo)
Schwartz, A., Thomas, R. 1975. A check-list of 28:61-90.
West Indian amphibians and reptiles. Carnegie Vanzolini, P. E. 1976. On the lizards of a cerrado-
Mus. Nat. Hist. Spec. Publ. (1):1-216. caatinga contact: Evolutionary and zoogeographi-
Sill, W. D. 1968. The zoogeography of the croco- cal implications (Sauria). Ibid. 29:111-119.
dilians. Copeia 1968( 1 ):76-88. Vanzolini, P. E., Williams, E. E. 1970. South
Simpson, B. B. 1975. Pleistocene changes in the American anoles: The geographic differentiation
flora of the high tropical Andes. Paleobiology and evolution of the Anolis chrysolepis species
1:273-294. group (Sauria, Iguanidae). Arq. Zool. (Sao
Smith, H. M., Martin, R. L., Swain, T. A. 1977. A Paulo) 19:1-298.
new genus and two new species of South Amer- Veevers, ]. J., McElhinny, M. W. 1976. The sep-
ican geckos (Reptilia: Lacertilia). Pap. Avulsos aration of Australia from other continents. Earth
Zool. (Sao Paulo) 30:195-213. Sci. Rev. 12:139-159.
Sokol, O. M. 1975. The phylogeny of anuran lar- Vuilleumier, F. 1968. Origin of frogs of Patagonian
vae: A new look. Copeia 1975( 1 ) :l-23. forests. Nature 219:87-89.
Solbric, O. T. 1976. The origin and floristic affini- Wake, D. B. 1966. Comparative osteology and evo-
ties of the South American desert and semi-desert lution of the lungless salamanders, family Pletho-
regions, pp. 7—49 in Goodall, D. W. (ed. ). Evo- dontidae. Mem. South. California Acad. Sci. (4):
lution of desert biota. Univ. Texas Press, Austin, 1-111.
250 p. Webb, S. D. 1978. A history of savanna vertebrates
Starrett, P. H. 1973. Evolutionary patterns in lar- in the New World. Part II: South America and
val morphology, pp. 251-271 in Vial, J. L. (ed.). the great interchange. Ann. Rev. Ecol. Syst. 9:
Evolutionary biology of the anurans: Contem- 393-436.
porary research on major problems. Univ. Mis- Wermuth, H. Liste der rezenten Amphibien
1965.
souri Press, Columbia, 470 p.
und Gekkonidae, Pygopodidae, Xantu-
Reptilien:
Tihen, J. A. 1962. A review of New World fossil siidae. Das Tierreich (80) 1-246. :
bufonids. Amer. Midi. Nat. 68:1-50. Wermuth, H. 1967. Liste der rezenten Amphibien
Tihen, J. A. 1964. Tertiary changes in the herpeto- und Reptilien: Agamidae. Ibid. (86):1-127.
faunas of temperate North America. Sencken- Wermuth, H. 1968. Liste der rezenten Amphibien
bergiana. Biol. 45:265-279. und Reptilien:Cordylidae. Ibid. (87): 1-30.
Tihen, J. A. 1972. The fossil record, pp. 8-13 in Williams, E. E. 1969. The ecology of colonization
Blair, W. F. (ed.). Evolution in the genus Bufo. as seen in the zoogeography of anoline lizards on
Univ. Texas Press, Austin, 459 p. small islands. Q. Rev. Biol. 44:345-389.
Trueb, L. 1971. Phylogenetic relationships of cer- Wolfe, J. A. 1971. Tertiary climatic fluctuations
tain Neotropical toads with the description of a and methods of analysis of Tertiary floras. Paleo-
new genus (Anura: Bufonidae). Nat. Hist. Mus. geogr. Paleoclimatol. Paleoecol. 9:27-57.
Los Angeles Cty. Contrib. Sci. (216): 1^40. Wood, R. C. 1971. The fossil Pelomedusidae (Testu-
Trueb, L. 1974. Systematic relationships of Neo- dines, Pleurodira) of Africa. PhD Dissert. Har-
tropical horned frogs, genus Hcmiphraetus (Anu- vard Univ., 345 p.
ra: Hylidae). Univ. Kansas Mus. Nat. Hist. Occas. Wood, R. C, Patterson, B. 1973. A fossil triony-
Pap. (29): 1-60. chid turtle from South America. Breviora (415):
Trueb, L., Tyler, M. J. 1974. Systematics and evo- 1-10.
lution of the Greater Antillean hylid frogs. Ibid. Zweifel, R. G. 1972. Results of the Archbold Ex-
(24): 1-60. peditions No. 97. A revision of the subfamily
Tyler, M. J. 1974. First fossil frogs from Australia. Asterophryinae, family Microhylidae. Bull. Amer.
Nature 248:711-712. Mus. Nat. Hist. 148:411-546.
2. The South American Herpetofauna:
An Evaluation of the Fossil Record
Ana Maria Baez
Departamento de Geologia
Facultad de Ciencias Exactas y Naturales
Univcrsidad de Buenos Aires
Buenos Aires, Argentina
Zulma B. de Gasparini
The presence
of fossil remains of amphib- in figures 2:4-7. A checklist of the material
ians and reptiles related to living taxa in recorded there and the corresponding litera-
South America has been documented since ture references appear in Appendix 2:1.
the last century. Nevertheless, examination Inspection of the fossil record reveals that
of this literature reveals that in many cases many families comprising the present South
the namesare only mentioned and the ma- American heqoetofauna have a considerable
terialhas not been studied; in many other antiquity in that continent. Their presence,
cases it is evident that a revision is badly and even that of some recent genera, extends
needed. back to the late Mesozoic and early Tertiary
Partial reviews concerning several South (Figs. 2:1-3). Thus, it is essential to con-
American countries have been carried out sider the past changes in the geographical
(Argentina: Pascual, 1970; Pascual and Odre- position of South America and its connections
man Rivas, 1971; Gasparini and Baez, 1975; with other continents, especially from the
Brasil: Paula Couto, 1970; Colombia: Hoff- Middle Mesozoic onward, for those changes
stetter, 1970a; Ecuador: Hoffstetter, 1970b; must have affected distribution patterns.
Peru: 1970c; Uruguay: Mones,
Hoffstetter, According to recent paleomagnetic data
1972; 1975). Also, Baez and Gasparini (1977) (Vilas and Valencio, 1978a), South America
critically examined the Cenozoic record of was part of Gondwanaland, which may have
amphibians and reptiles in that continent and existed up to the Late Jurassic. It was still
analyzed distributional shifts, relating them to joined to Africa during the Early Cretaceous
the Cenozoic environmental changes that oc- (Reyment and Tait, 1972; Douglas, Moullade
curred as a result of different geological and Nairn, 1973; Vilas and Valencio, 1978),
events. while faunal relations with Antarctica-Aus-
In this paper, the available paleontological tralia seem to have been possible up to the
data are summarized in an attempt to evalu- Early Tertiary (Raven and Axelrod, 1975).
ate the information that the fossil record can There is no clear evidence that a direct land
provide about the historical development of connection between North and South America
these groups in South America. The taxo- existed from the Jurassic to the end of the
nomic assignment and the geographic and Tertiary (Malfait and Dinkelman, 1972;
stratigraphic references are sometimes doubt- McKenna, 1973). Nevertheless, a discontin-
ful;questionable referrals are not considered. uous pathway along a chain of volcanic is-
Some recent finds are currently under study, lands could have been established at different
and identification of genera and species is not times (Haffer, 1970; Malfait and Dinkelman,
yet available; thus, only an analysis at the 1972 ) There is no agreement concerning the
.
family level is possible at present. The main time of final reconnection between both
areas that have yielded fossil amphibians and Americas. Some evidence indicates that the
reptiles referable to modern groups are shown Isthmus of Panama was established during
29
30 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
FAMILIAL
GROUPS
1979 BAEZ & GASPARINI: FOSSIL RECORD 31
FAMILIAL
GROUPS
32 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
FAMILIAL
GROUPS
1979 BAEZ & GASPARINI: FOSSIL RECORD 33
(Sebecidae) and eusuchian (Alligatoridae, ments and the continuous uplift of that area
Crocodylidae, Gavialidae) crocodilians, throughout the late Cenozoic made the pres-
which is explained by the presence there of ervation of remains unlikely.
an extensive lowland with local swamps and Numerous crocodilian fossils, all referable
lakes (Van Houten and Travis, 1968). All of to the suborder Eusuchia (Alligatoridae, Gav-
the families recorded from those horizons also ialidae, Nettosuchidae and probably Croco-
are represented in older deposits of more dylidae) are recorded in the upper Pliocene
southern latitudes, except gavialid crocodil- of northern Venezuela (Fig. 2:7). The
gigan-
ians, whose affinities are still uncertain. exhibited by many members of these
tic size
Oligocene localities in central taxa and the appearance of the extant alli-
Patagonia
(Fig. 2:6) have yielded remains of the extant gatorid, Melanosuchus, is noteworthy. The
lepodactylid genera Eupsophus and Caudi- possible presence of trionychid turtles there
verbera, whose presence there was made pos- should be noted. This group is not repre-
sible by the prevalence of more mesic condi- sented in the present fauna of South America,
tions at that time. The oldest known testu- and may have reached the continent from
dinid turtle in South America, a species of the the north.
genus Geochelone related to the living G. In Pliocene times the Rio Parana consti-
chilemis (Auffenberg, 1971), is recorded in tuted, as it does today, an important pathway
the upper Oligocene of Patagonia. for the southward migration of elements of
the subtropical biota. This explains the rec-
Very rich, late Miocene faunas are known
from the upper Magdalena River Valley, Co- ord of alligatorid and gavialid crocodilians in
lombia, where the general environmental con- Pliocene deposits near the city of Parana,
ditions of Oligocene times
did not change Argentina (Fig. 2:7). Among them, Caiman
latirostris and possibly Rhamphostomopsis
significantly, although important tectonic
events took place in the Miocene were already represented Miocene
in the late
(Irving,
faunas of Colombia.
1971). Most of the families present there
have previous records in the continent, the Numerous remains of amphibians and rep-
tiles have been
majority being represented by extant genera reported from horizons of late
or species: Bufo marinus, Podocnemis Pliocene age in Monte Hermoso, Argentina
expan-
sa, Tupinambis cf. T. teguixin, Dracaena, Eu- (Fig. 2:7). The presence there of the living
teiidgenus Callopistes is especially significant
nectes, Caiman. Also a turtle, Geochelone
hesterna, closely related to the living G. car- considering its present range, which is on the
bonaria and G. denticulata arid Pacific lowlands of Peru and Chile.
(Auffenberg,
1971) was recorded. The appearance of two
The relatively few Pleistocene sites that
families, Colubridae and Nettosuchidae, in
have yielded amphibians or reptiles are im-
the fossil record is noteworthy. The latter are portant because they document the past pres-
endemic eusuchian crocodilians seemingly re- ence of families that currently inhabit South
stricted to tropical regions from Miocene to America, and that have not been recorded in
Plio-Pleistocene times. older deposits. Among these families are emy-
The southernmost
records of Cenozoic am-
dids, viperids, and amphisbaenids. The Pleis-
tocene records also suggest that significant
phibians or reptiles on the continent come
from early to middle Miocene environmental changes occurred at that time,
deposits of
southern Patagonia (Fig. 2:6). because the distribution of some taxa is incon-
Iguanids,
sistent with the
teiids and other lizards, turtles and snakes present distribution of their
have been found there, but all of this ma- habitats. For example, the Pleistocene faun-
ulesfrom Talara, northwestern Peru (Lemon
terial has yet to be restudied. The
post-Mio- and Churcher, 1961; Hoffstetter, 1970c) and
cene history of the Patagonian
herpetofauna from the Santa Elena Peninsula, southwestern
is not documented in the fossil record. The Ecuador (see Appendix 2:1), indicate that
increasing aridity that developed there as a more mesic environments prevailed in those
consequence of the Andean tectonic move- areas in the near past.
34 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
JO»
1979 BAEZ & GASPARINI: FOSSIL RECORD 35
J0°
A
36 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
ao°
A
19(79 BAEZ & GASPARINI: FOSSIL RECORD 37
^0°
A
38 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
The fossil record of modern amphibian early presumed records outside of South
in South America largely restricted
is America have been discarded or questioned
groups
to anurans. The only known caecilian fossil (Hecht, 1963; Estes, 1964, 1969; Reig, 1968;
is from the late Paleocene of Brasil; this rec- Lynch, 1971; Savage, 1973). Available evi-
ord implies that caecilians are old components dence indicates that the Leptodactylidae, in
No its restricted sense (Lynch, 1973) may have
of the South American herpetofauna.
differentiated in South America from a stock
salamander fossils are known from South
America. associated with temperate forests (Lynch,
1971; Savage, 1973; Heyer, 1975). The pro-
The considerable diversity of anuran
posed relationships to leiopelmatids (Heyer,
groups already present by the early Tertiary
1975) are significant in that the latter inhab-
in South America is noteworthy. Further-
ited Patagonia by Jurassic times (Estes and
more, their phylogenetic relationships and the
Reig, 1973). Leptodactylids must have dis-
fact that some of them were represented by
taxa similar to extant South American genera persed northward early in their history. Wide-
or species indicate that they have had a long spread occurrence of subhumid environments
in middle latitudes in the Late Jurassic and
that
history there. However, many families
the anuran fauna are un- Early Cretaceous might have been influential
comprise present
known as fossils.
in the development of xeric-adapted types of
leptodactylids (Baez and Gasparini, 1977).
The aquatic frogs of the family Pipidae
Telmatobiine leptodactylids, which represent
now South America and Africa; they
live in
an ancient radiation, are recorded in the Ter-
constitute an ancient, independent lineage.
tiary of Patagonia. The extant genus Caudi-
The earliest pipids are from the Early Creta- verbera appears in early Eocene, early and
ceous of Israel (Nevo, 1968), but their origin late Oligocene, and late Miocene deposits. By
probably extends back into the Jurassic. Un- the early Oligocene two other genera were
questionable pipids occur in the Cretaceous —
present the living Eupsophus and the ex-
and Tertiary of both Africa and South Amer- tinctNeoprocoela; the assignment of Neopro-
ica; strongly suggests that pipids were
this
coela to this family has been discussed by
components of the Gondwanan fauna. Fur- Tihen (1962, 1972) and Lynch (1971). These
thermore, remains referable to the African records furnish evidence of their former wider
genus Xenopus have been recorded from both distribution east of the Andes, and therefore
continents (Ahl, 1926; Vergnaud-Grazzini, reflect the mesic climate that prevailed in
1966; Broin et al., 1974; Estes, 1975a,b; Baez, those now arid regions (Gasparini and Baez,
1976). The fossil record indicates that differ-
1975). Ceratophryine leptodactylids, which
ent phyletic lines of pipids have existed in
some authors give familial status, are first
South America. The oldest known example recorded in the late Miocene of northern
there is Saltenia ibanezi from the Late Cre-
Patagonia, but they must have originated
taceous of northwestern Argentina. Members much The basic adaptation of this
earlier.
of the extant genus Xenopus have been re- an arid environment Heyer, 1975 )
group to (
corded in the late Paleocene of Brasil and Ar- could account for the paucity of paleontologi-
gentina. No fossil taxa directly ancestral to cal evidence for its early evolution, because
the living Neotropical Pipinae are known. few fossil-bearing horizons representing those
The Leptodactylidae are another ancient environments are known, particularly in the
component of the South American fauna. The Late Mesozoic and Early Tertiary. Remains
oldest known remains unquestionably refer- referable to the extant genus Ceratophrys
able to the family come from late Paleocene (specific allocation undetermined) are known
deposits in Brasil; undescribed taxa close to from the late Pliocene and middle Pleistocene
living genera and perhaps even an extant of Argentina (Baez and Gasparini, 1977).
genus are present there (Estes and Reig, The living C. ornata probably is represented
1973). This group probably is represented in in the late Pliocene of Argentina (Reig, 1958)
1979 BAEZ & GASPARINI: FOSSIL RECORD 39
and late Pleistocene of Bolivia, and C. aurita tion to the biogeographical history of hylids.
is known from the latest Pleistocene in Brasil Frogs are ancient members of South Amer-
(Lynch, 1971). Records of other leptodac- ican fauna. Pipids, leptodactylids, bufonids
tylids are limited to the living genus Lepto- and hylids are old components and are the
dachjlus from Pleistocene deposits. only frog families represented in the South
The earliest known record of the cosmo- American fossil record. Other groups such as
politan bufonid toads (absent from Australia rhinodermatids and dendrobatids, differenti-
and Madagascar) is from the late Paleocene ated on that continent probably during the
of Brasil. Although this material has not yet Tertiary.On the other hand, ranids are late
been described, the presence of members of immigrants from the north.
living species groups of Bufo has been recog-
nized (Estes and Reig, 1973). This supports
the proposal that bufonids are ancient com- REPTILES
ponents of the South American fauna and that
they could have had a southern origin Pelomedusid turtles presently inhabit
(Lynch, 1971; Reig, 1972; Savage, 1973). No South America, Africa and Madagascar, but
unquestionable bufonids are known in North they were more widely distributed in Late
America prior to the early Miocene (Tihen, Mesozoic and Early Tertiary times (Romer,
1972). Moreover, assignment of specimens of 1966; Gaffney and Zangerl, 1968; Wood, 1970,
Early Tertiary age from Europe to the Bufo- 1976b; Jimenez Fuentes, 1971, 1975; Gaffney,
nidae is highly doubtful; remains referable to 1975). The earliest-known pelomedusids are
the genus Bufo are unknown there before from Early Cretaceous deposits in Africa
the middle Miocene (Tihen, 1972). Different (Broin et al., 1974), and remains of Late
evidence suggests that South America is the Cretaceous age have been reported from
most likely area of origin of Bufo (Blair,
South America, North America, Africa, and
1972). Although sparse, the fossil record in- Europe. The greater proximity of continents
dicates that considerable diversification with- at that time (Smith, Briden and
Drewry,
in that genus took place in South America at 1973), coupled with the presumed marine
least since the Early
Tertiary. Members of
habits of some of these turtles (Wood, 1972,
the marinus group of Bufo have been reported 1976b) could have favored such wide distri-
as far back as the bution. The majority of the South American
late Miocene, although an
earlier differentiation of that group seems fossil pelomedusids that have been described
Brasil, but the material is still undescribed Podocnemis, but there is no general agree-
(Estes and Reig, 1973). This testifies to the
ment concerning the validity of species
early presence of hylid frogs in South Amer- (Wood and Gamero, 1971; Baez and Gaspari-
ica, and is consistent with the high degree of ni, 1977). By the late Miocene the extant P.
differentiation that they attained there
during expanse was already in existence. Remains
the time of isolation, as well as their referable to the genus Podocnemis are cited
possible
origin on that continent (Savage, 1973). The outside of South America, from Africa and
now widespread genus Hula has been re- western Europe. Pelomedusids referable to
corded from the early Oligocene of Canada other genera have been described from the
(Holman, 196S) and the early Miocene of 1
Dr. F. de Broin (pers. coram.) considers that Po-
Florida in the United States (Tihen, 1964).
docnemis elegans may belong to the extinct pelome-
The paleontological data add little informa- dusid genus Roxochehjs, related to Podocnemis.
40 MONOGRAPH MUSEUM OF NATURAL HISTORY NO.
Cenozoic of South America. A species of the the north during the Oligocene or even
genus Taphrosphys occurs in the Eocene of earlier. It noteworthy that even though
is
coastal Peru. That genus, which includes ma- chelonians are known since the Late Creta-
rine forms, is also recorded from the Creta- ceous, no testudinids have been reported be-
ceous of North America and Paleocene of fore the late Oligocene, whereas their pres-
Africa (Wood, 1975). The gigantic Pliocene ence is documented frequently since the late
Stupendemys geographicus exhibits many un- Miocene. Assuming an entrance from the
usual characteristics, and its affinities are still north, this partially could result from the fact
not clear (Wood, 1976b). that most known early Tertiary reptile-bear-
The freshwater turtles of the family Cheli- ing sites are in the southern part of the conti-
dae occur today in South America, Australia nent. However, an arrival earlier than late
and NewGuinea. The geographical location Eocene times seems improbable. Geochelone
group indicates that
of fossils assigned to this gringorum is
closely related, and may be an-
they could have had an essentially similar cestral, to the living G. chilensis, which be-
former range, because the only known un- came adapted to drier conditions (Auffenberg,
questionable records come from the early Eo- 1971). Increasingly xeric environments devel-
cene of southern Argentina, middle Tertiary oped in western mid-latitudes east of the
of Australia, and Oligocene or Miocene of Andes since the Pleistocene (Baez and Scil-
Tasmania (Warren, 1969). That evidence lato Yane, thisvolume). The related Pliocene
suggests an area of dispersal that included G. gallardoi has been recorded in areas where
also Antarcticaand that the past relations a marked dry season presumably existed.
between South American and Antarctica-Aus- Geochelone hesterna from the late Miocene of
tralia (McGowran, 1973; Dalziel et al., 1973; Colombia is thought to be ancestral to the
Sclater and Fisher, 1974) could account for extant species G. carhonaria and G. denticu-
the present range. A
species of the extant lata that now live in the northern part of the
South American genus Hydro medusa is the continent (Auffenberg, 1971).
oldest member ( Eocene ) of the family re- Although emydid turtles now occur in
corded so far (Wood and Moody, 1976). South America, their presence there during
Two extinct species of the now monotvpie the Tertiary is still uncertain. The existence
genus Chelus have been described C. co- — of remains referable to the Emydidae among
lombianus from the late Miocene of Colombia the material collected from late Miocene de-
and C. leuisi from the Pliocene of Venezuela posits in Colombia was mentioned by Medem
(Wood, 1976a). Neither of these seem to (1966, 196S), but the record has not been
have been directly ancestral to the living substantiated. The extant genus Geocmyda is
species, thereby indicating that different line- known from the late Pleistocene of Ecuador.
ages evolved within the genus (Wood, No fossil records of the Chelydridae and
1976a). Kinosternidae are known in South America.
Testudinid turtles are represented in South Trionychids, old members of the North Amer-
America by the genus Geochelone, with a ican faunas, are not present today in South
world-wide distribution in the Early Tertiary. America. Their presence in northern Vene-
Fossils having similarities with an extant Asi- zuela in the late Pliocene could have resulted
atic specieshave been recorded from Eocene from waif dispersal, but their colonization
deposits in North America and Africa, and was not successful (Wood and Patterson,
also from the early Oligocene of North Amer- 1973).
ica, Asia and Western Europe (Auffenberg, In the Late Cretaceous and Early Tertiary,
1974 ) The
earliest representative of the genus
. a peculiar group of land turtles of uncertain
Geochelone in South America is G. gringorum affinities, the Meiolaniidae, was also part of
from the late
Oligocene of Patagonia (Simp- the herpetofauna of southern South America.
son, 1942). Auffenberg (1971) suggested that They also have been recorded in Australia,
the ancestors of the South American species, where they survived until Pleistocene times.
comprising the distinctive subgenus Cltclo- In
summary, pelomedusids are ancient
noides, probably entered that continent from components of the ehelonian faunas of South
1979 BAEZ & GASPARINI: FOSSIL RECORD 41
America, being continuously represented since and resembling primitive living South Amer-
the Late Mesozoic. Podocnemis, the only liv- ican taxa have been recorded from the late
ing genus in South America, is known since Paleocene of Brasil (Estes, 1970). Their ab-
the Late Cretaceous; it has undergone con- sence in the Cretaceous is not surprising be-
siderable speciation. Chelids also are old cause continental Middle and Late Mesozoic
members of the fauna, even though no occur- faunas are still poorly known in South Amer-
rence prior to the early Eocene is known. No ica. Practically all later teiid fossils (Mio-
land turtles are known from the Late Meso- cene-Pleistocene) have been referred to the
zoic and Early Tertiary besides meiolaniids. extant South American genera Tupinambis,
Their ecological role was assumed later by Dracaena, Callopistes, and Dicrodon, al-
the testiudinid Geochelone, which entered the though differences in distributions are evident
continent probably in Eocene-Oligocene when compared with their present ranges
times. (Baez and Gasparini, 1977).
The record of the lizard family
fossil Very few fossil records of the other lizard
Iguanidae South America is fragmentary, a
in families now inhabiting South America are
situation that contrasts with its significance known from that continent. However, gek-
in the present Neotropical herpetofauna. Al- konids occur in the upper Paleocene of Brasil.
though the former presence of this group in Although this is the earliest known record of
Asia and Europe has been discarded (Hoff- the family, the origin of gekkonids could ex-
stetter, 1962; Estes, 1970), fossils referable to tend back to the Late Mesozoic, upon consid-
this family have been recorded in North eration of their relationships with some Juras-
America, but none antedates the early Eocene sic European and Asiatic taxa (Hoffstetter,
(Estes and Price, 1973). The earliest record 1964; Kluge, 1967). Living South American
of an iguanid is from the Upper Cretaceous representatives might have been derived from
of Brasil, and the family is represented by at different sources; the strong African affinities
least five species in the late Paleocene fauna of sphaerodactylines (Kluge, 1967) could be
from Itaborai, Brasil, thereby indicating an interpreted as resulting from the past connec-
early radiation in South America (Estes, tions of those continents (Estes, 1970). The
1970). Presently-known data are consistent presence of scincids in South America was
with suggested Gondwanan origin of
the extended back to the Paleocene ( Estes, 1976 ) .
iguanids (Cracraft, 1973). The group could These lizards are seemingly
already repre-
have entered North America from the south sented in the Late Cretaceous North Amer-
by waif dispersal, becoming quite diversified ican faunas (Estes, 1976).
by Miocene times ( Robinson and Van Deven- Even though the fossil record of lizards in
By the Miocene, iguanids attained
der, 1973). South America is extremely poor, it is evident
a wide distribution in South America, but that iguanids and teiids, both well repre-
little is known of the taxa present at that time sented there today, were characteristic com-
(Baez and Gasparini, 1977). Extant repre- ponents of the Cenozoic herpetofaunas of
sentatives of the family are recorded from late that continent, where they underwent consid-
Pleistocene deposits: Iguana in coastal Ecua- erable diversification. Those groups, and also
dor ( Hoffstetter, 1970b) and Leiosaurus bellii gekkonids, seem to comprise an ancient faunal
in central-western Argentina (Van Devender, element. The presence of anguids in South
1977). America has not yet been documented by the
The evolution of teiids took place primar- fossilrecord. They are present in North
ily in South America, where they are most America since the Late Cretaceous, and seem
diverse and widely distributed today. In the to have had an essentially northern dispersal
Late Cretaceous teiids were present in North (Hoffstetter, 1962; Estes, 1964; Meszoely,
America (Estes, 1964, 1969), although these 1970).
early records do not seem to be of direct sig- Snake remains have been recorded in
nificance in the establishment of the Recent South America from the Late Cretaceous
teiids there (Estes, 1970). On the other through the Pleistocene. With a few excep-
hand, fossil remains referable to the Teiidae tions, no good descriptions are available, and
42 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
subfamily Madtsoiinae, which were recorded America are from the late Pleistocene. True
from the late Paleocene and early Eocene of colubrids appear in the late Eocene of Europe
Patagonia. Species of that genus also are (Rage, 1974). In North America their pres-
known from the Late Cretaceous of Mada- ence has been documented since the early
gascar ( Hoffstetter, 1961) and of Niger
Miocene (Holman, 1976b); they become
(Broin et al., 1974), a distribution that sug- dominant elements of snake faunas by the late
gests their derivation from a Gondwanan
Miocene (Holman, 1976a). Thus, an entry in
stock. Remains assigned to the Boidae have South America from the north during the
been cited from the late Paleocene of Brasil Miocene was suggested (Hoffstetter, 1967b;
and late Pleistocene of Bolivia. The living Hoffstetter and Rage, 1977). However, it is
South American genus Eunectes occurs in noteworthy that the relationship of some Mio-
late Miocene deposits of Colombia. Boids cene colubrids from the United States to living
Central or South American forms could indi-
probably were also represented in the early-
middle Miocene of southern Patagonia and cate that they were derived from a more
the Pliocene of Parana, Argentina, but those southern source (Tihen, 1964).
records have not been substantiated (Gas- Fossil viperids, which were referred to the
parini and Baez, 1975; Baez and Gasparini, Crotalinae, are known in South America only
1977). from late Pleistocene deposits in Bolivia. They
Aniliids, now restricted to the Oriental could be late immigrants from the north
and Neotropical regions, were represented in (Reig, 1962; Baez and Gasparini, 1977), al-
the Early Tertiary faunas of South America. though South America should
their arrival in
A snake of controversial phylogenetic position, have preceded that record. In the early Mio-
Dinilysia patagonica, was recorded from the cene, typical viperids appear in Europe
Late Cretaceous of central-western Argentina (Hoffstetter, 1962). Viperid remains, tenta-
(Smith Woodward, 1901; Huene, 1929). Its tively assigned to the Crotalinae, were re-
affinities to modern aniliids were pointed out ported from the middle Miocene of North
by Estes, Frazzeta and Williams (1970), but America ( Holman, 1976c ) All evidence indi-
.
it was retained in the monotypic family Dini- cates, as in the case of colubrids, that the early
lysiidae. However, Dinilysia was considered history of this group is still largely unknown.
to be closer to boids, although not ancestral to The record of snakes in South Amer-
fossil
group also occurs in Eocene deposits of Eu- taxa assigned to that group, Necrosuchus ion-
rope (Hoffstetter, 1962; Rage, 1974). Rein- en-sis Simpson, 1937, inhabited southern Ar-
terpretation of relationships of Dinilysia and gentina. The affinities of this crocodilian have
the known fossil record of aniliids seems to be not been clearly established. Crocodylids ap-
more consistent with the postulated Laurasian pear in the fossil record in Cretaceous de-
1979 BAEZ & GASPARINI: FOSSIL RECORD 43
posits of comparable age in North America, habited South America in the Tertiary. It is
Africa and Asia (Sill, 1968). Thus, it is still controversial the South American Ion-
if
doubtful if the ancestors of Necrosuchus came girostrine crocodilians are true gavialids or
from North America, at this time is is impos- constitute a different lineage ( Langston, 1965;
sible to determine their area of origin. From Gasparini, 1968; Sill, 1968, 1970; Hoffstetter,
the Oligocene onward, crocodylids are re- 1970a; Hecht and Malone, 1972; Baez and
corded north of the Amazonian Basin, with a Gasparini, 1977). The oldest gavialids are
latitudinal distribution similar to the present known from Eocene deposits in Egypt (Hecht
one. No fossil Crocodylus is known from and Malone, 1972). The Asiatic and South
South America, except for a doubtful record American forms could have originated from
from the Pliocene of Maranhao, coastal Brasil an African stock; their dispersal would have
(Maury, 1923). Available evidence suggests been facilitated by their capacity to swim in
that the living representatives of that genus marine waters. In South America the first
are late immigrants from the north. The rela- gavialids appear in deposits of late Oligocene-
tionships of the peculiar crocodylid, Charac- early Miocene age from the northern part of
tosuchus fieldsi Langston, 1965, from the late the continent; they became extinct by the
Miocene of Colombia are still unknown. Plio-Pleistocene. The referral of South Amer-
Alligatorid crocodilians are distributed ican forms to the genus Gavialis is doubtful,
more widely than crocodylids in South Amer- because they are different from the Asiatic
ica. The earliest known representatives in members of that genus.
that continent occur in the late Paleocene of The extinct Nettosuchidae is an endemic
Brasil, and they probably also are present in group from mesic tropical environments of the
deposits of that age from southern Argentina. northern and central part of the continent
In both cases, the material has not been from the late Miocene to the earliest Pleisto-
studied yet. The validity of the Paleocene cene.
"Notocaiman stromeri" from Patagonia, pro- Crocodilians of the more primitive and
posed as an ancestor of Caiman, has been dis- extinct suborder Mesosuchia are known in
carded (Gasparini and Baez, 1975). The South America during a large part of the Ter-
presence of alligatorids in North America ex- tiary. These forms belong
to the family Se-
tends back to the Cretaceous, and a center of becidae Sebecosuchia, Gasparini, 1972) and
(
origin there was suggested by Sill (1968). are known from the Paleocene through the
Nevertheless, available data are inconclusive. Miocene. They were peculiar, mostly terres-
The Eocene Eocaiman cavernensis is the old- and inhabitants of the tropical
trial crocodiles,
est member of the family described from forests (Langston, 1965; Molnar, 1969, 1977;
South America. Langston ( 1965 ) pointed out Neill, 1971; Gasparini, 1972). Sebecids seem
its modern aspect and suggested that it could tohave had a long history on the continent,
have been ancestral to Caiman, Melanosuch- where they probably originated from bauro-
us, and perhaps also to the peculiar Balanero- suchids.
dus. Kalin 1955 ) did not rule out the pos-
(
The great diversity of Cenozoic crocodil-
sibility that it could be referred to the extant ian faunas in South America is noteworthy.
genus Caiman. This indicates that taxa close- Species referable to five families (Sebecidae,
ly related to living South American represen- Crocodylidae, Alligatoridae, Gavialidae, Net-
tatives of this family were already present on tosuchidae) have lived there, but only two of
that continent by the Early Tertiary. In the those groups (Crocodylidae, Alligatoridae)
late Miocene of Colombia an extant genus are present now. Available evidence suggests
and probably also a species Caiman latiro-
(
that the differentiation of sebecids and netto-
known (Gasparini and Baez, 1975).
stris) are suchids occurred in South America. The most
The recent Caiman yacare is known from the conspicuous groups were the Sebecidae and
Pliocene of Argentina and Melanosuchus from Alligatoridae, the latter being represented
the Pliocene of Venezuela. throughout the Cenozoic and the most im-
Today, gavialids are restricted to the portant crocodilian family there today. On the
Ganges Basin, India, but they may have in- other hand, the fossil record of crocodylids is
44 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
change of the biota in the region of the pres- La mention de anfibios y reptiles ceno-
ent upper Magdalena Valley, Colombia, in zoicos en America del Sur es relativamente
relation to the uplift of the Eastern Cordillera frecuente, sin embargo, muchas de las asigna-
and increasing aridity ( Howe, 1974 ) is an- ciones taxonomicas como las referencias geo-
other example. Many components of the rich graficas y cronoestratigraficas son dudosas.
late Miocene La Venta fauna, living on the Ello, conjuntamente con el hecho de que la
broad flood plains that prevailed there are mayoria de los hallazgos son aislados, sin
absent from that now semiarid area (Fields, formar parte de asociaciones representativas,
1959). hace que las conclusiones resulten aiin tenta-
Many fundamental aspects in the history
tivas. En general los datos disponibles per-
of the South American herpetofauna have yet miten un analisis a nivel familiar.
to be elucidated. The amount of paleontolog- El examen del registro senala la notable
ical information that has accumulated in re- antigiiedad, en America del Sur, de muchas
cent years enables us to expect that future de la familias de anfibios y reptiles que viven
studies will clarify these problems. en ese continente. Todas las familias regis-
tradas en el Terciario temprano son inte-
grantes de la herpetofauna actual sudameri-
ACKNOWLEDGMENTS cana, excepto las tortugas meiolanidas y los
cocodrilos sebecidos, ambos e.xtinguidos. Es
For their valuable comments and provision evidente que esas familias tienen distinto
of data, we are indebted to Drs. France de
abolengo, habiendose integrado o diferenci-
Broin, Richard Estes, J. Alan Holman, Bryan ado in situ alocronicamente, desde fines del
Patterson and Roger Wood. Special thanks Cretacico al menos. De acuerdo a las eviden-
are extended to Dr. William E. Duellman, cias paleontologicas la antigiiedad en Ameri-
who kindly revised and corrected the manu- ca del Sur de los pipidos, iguanidos, pelo-
script.
medusidos, meiolanidos, muy probablemente
los quelidos, y posiblemente los leptodacti-
cien a partir del Oligoceno se conocen regis- Auffenberg, W. 1974. Checklist of fossil land tor-
tros de los grupos considerados en el extremo toises (Testudinidae). Bull. Florida State Mus.
que alii se
registra muy es la distinta de Baez, A. M. 1977. Sobre Teracopluys Ameghino,
nomina nuda (Anura, Leptodactylidae de la
que habita ese area en nuestros dias. )
Burmeister, H. 1883. Monografia de los terrenos Estes, R. 1970. Origin of the Recent North Amer-
marinos terciarios del Parana. An. Mus. Puhl. ican lower vertebrate fauna: An inquiry into the
Buenos Aires 3:45-94. fossil record. Forma Functio 3:139-163.
Burmeister, H. 1885. Examen eritico de los niamf- Estes, R. 1975a. Xenopus from the Palaeocene of
feros y reptiles fosiles denominados por don Brazil and its zoogeographic importance. Nature
Augusto Bravard. Ibid. 3:95-174. 254:48-50.
Casamiquela, R. 1958. Un anuro gigante del Mio- Estes, R. 1975b. Fossil Xenopus from the Palaeo-
ceno de Patagonia. Rev. Asoc. Geol. Argentina cene of South America and the zoogeography of
13:171-183. pipid frogs. Herpetologica 31:263-278.
Casamiquela, R. 1960. Datos preliminares sobre un Estes, R. 1976. Middle Paleocene lower vertebrates
pipoideo fosil de la Patagonia. Actas y Trabajos I from the Tongue Formation, southeastern Mon-
Congr. Sudamer. Zool., La Plata 1959, 4:17-21. tana. J. Paleontol. 50:500-520.
Casamiquela, R. 1963. Sobre un par de anuros del Estes, R., Frazzetta, T., Williams, E. 1970. Stud-
Mioceno de Rio Negro (Patagonia) Wawclia ger- ies on the fossil snake Dinilysia patagonica Wood-
lioldi n. gen. et. sp. ( Ceratophrydidae y Gigan- )
ward. Part I. Cranial morphology-. Bull. Mus.
tohatrachus parodii Leptodactylidae
( Ameghini- ) . Comp. Zool. Harvard Univ. 140:25-74.
ana 3:141-162. Estes, R., Price, Ll. 1973. Iguanid lizard from the
Casamiquela, R. 1967. Sobre un nuevo Bufo fosil upper Cretaceous of Brazil. Science 180:748-751.
de la Provincia de Buenos Aires ( Argentina ) . Estes, R., Reig, O. 1973. The early fossil record of
Ibid. 5:161-168. frogs: A review of the evidence, pp. 11-63 in
Cattoi, N., Freibebc, M. 1958. Una nueva especie Vial, J. L. (ed. ). Evolutionary biology of the
de "Podocncmis" del Cretaceo argentine Physis anurans: Contemporary research on mayor prob-
21 (60): 58-66. lems. Univ. Missouri Press, Columbia, 470 p.
Cei, J. 1972. Bufo of South America, pp. 82-90 in Estes, R., Wake, M. 1972. The first fossil record of
Blair, W. F. (ed.). Evolution in the genus Bufo. caecilian amphibians.Nature 239:228-231.
Univ. Texas Press, Austin, 459 p. Estes, R., Wassersug, R. 1963. A Miocene toad
Cham, J. 1976. Relaciones de un nuevo Teiidae from Colombia, South America. Breviora (193):
(Lacertilia) fosil del Plioceno superior, Callopistes 1-13.
bicuspidatus n. sp. Inst. Miguel Lillo, Univ. Nac. Fields, R. Geology of the La Venta badlands,
1959.
Tucuman, Publ. Especial, pp. 133—153. Colombia, South America. Univ. Calif. Publ.
1973. Vertebrate evolution and bio- Geol. Sci. 23:403-444.
Cracraft, J.
geography in the Old World 373-393 Fox, R. 1975. Fossil snakes from the Upper Milk-
tropics, pp.
in River Formation (Upper Cretaceous), Alberta.
Tarlinc, D., Runcorn, S. (eds.).
Implica-
tions of continental drift to the earth sciences, Canad. J. Earth Sci. 12:1557-1563.
Academic New Gaffney, E. 1975. A revision of the side-necked
Vol. 1. Press, York, 622 p.
turtle Taphrosplujs sulcatus (Leidy) from the
Dalziel, I., Kligfield, R., Lowrie, N., Opdyke.
Cretaceous of New Jersey. Amer. Mus. Novit
N. D. 1973. Palaeomagnetic data from the
southernmost Andes and the Antarctandes. Ibid.: (2571): 1-24.
87-101. Gaffney, E., Zangerl, R. 1968. A revision of the
chelonian genus Bothremys ( Pleurodira: Pelome-
Dawson, M., West, R., Lancston, W., Hutchison, dusidae). Fieldiana Geol. 16:193-239.
1976. Paleogene terrestrial vertebrates: north-
J.
Gallardo, J. M. 1962. A proposito de Bufo varie-
ernmost occurrence, Ellesmere Island, Canada.
gatus (Giinther), sapo del bosque humedo An-
Science 192:781-782.
tartandico, y las otras especies de Bufo neotropi-
Douglas, R., Moullade, M., Nairn, A. 1973. cales. Physis 23 (64):93-102.
Causes and consequences of drift in the South
Gasparini, Z. B. de. 1968. Nuevos restos de Rham-
Atlantic, pp. 513-534 in Tarling, D., Runcorn,
plwstomopsis neogaeus (Burm.) Rusconi 1933
S. (eds.). Implications of continental drift to the ( Reptilia, Crocodilia) del "Mesopotamiense" (Pli-
earth sciences, Vol. 1. Academic Press, New oceno medio-superior) de Argentina. Ameghini-
York, 622 p. ana 5:299-311.
Emiliant, C, Gaertner, S., Lidz, B. 1972. Neo-
Gasparini, Z. B. de. 1972. Los Sebecosuchia (Croco-
gene sedimentation on the Blake Plateau and the dilia) del territorio argentino. Consideraciones
emergence of the Central American Isthmus. sobre su "status" taxonomico. Ibid. 9:23-34.
Palaeogeogr., Palaeoclimatol., Palaeoecol., 11:1-
10.
Gasparini, Z. B. de. 1973. Revision de los Crocodilia
(Reptilia) fosiles del territorio argentino. Su
Estes, R. 1961. Miocene lizards from Colombia, evolucion, sus relaciones filogeneticas, su clasifica-
South America. Breviora ( 143) 1—11. :
cion y sus implicancias estratigrahcas. Tesis Doc-
Estes, R. 1964. Fossil vertebrates from the late toral, Univ. Nac. La Plata, 169 p.
Cretaceous Lance Formation eastern Wyoming.
Gasparini, Z. B. de, Baez, A. M. 1975. Aportes al
Univ. Calif. Publ. Geol. Sci. 49:1-180. conocimiento de la herpetofauna terciaria de la
Estes, R. 1969. Relationships of two Cretaceous Argentina. Actas I Congr. Argent Paleontol.
lizards (Sauria, Teiidae). Breviora (317): 1-8. Bioestr., Tucuman 1974, 2:377-415.
48 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Haffer, J. 1970. Geologic-climatic history and zoo- Holman, J. 1976b. A boid and a colubrid snake
geographic significance of the Uraba Region in from the Harrison Formation ( Lower Miocene,
northwestern Colombia. Caldasia 10:603-636. Arikareean) of Sioux County, Nebraska. Ibid.
Hecht, M. 1963. A reevaluation of the early history 32:387-389.
of the frogs. Part II. Syst. Zool. 12:20-35. Holman, J. 1976c. Snakes of the Split Rock Forma-
Hecht, M., Malone, B. 1972. On the early history tion ( Middle Miocene ) Central Wyoming. Ibid.
,
McKenna, M. 1973. Sweepstakes, filters, corridors, Paula Couto, C. de. 1970. Evolucao de comuni-
Noah's and beached Viking funeral ships in
arks, dades, modificacoes faunisticas e integracoes bio-
palaeogeography, pp. 295-308 in Tarlinc, D., cenoticas dos vertebrados cenozoicos do Brasil.
Runcorn, S. (ed.). Implication of continental Actas IV Congr. Latinoamer. Zool., Caracas 1968,
drift to the earth sciences. Vol. 1. Academic 2:907-930.
Press, New York, 622 p. Petkiella, B., Archangelsky, S. 1975. Vegetation
Medem, F. 1966. al conocimiento
Contribuciones y ambiente en el Paleoceno de Chubut. Actas I
sobre la y distribution geografica de
ecologia Congr. Argent. Paleontol. Bioestr., Tucuman,
Phrynaps (Batrachemys) dahli; (Testudinata, 1974, 2:257-270.
Pleurodira, Chelyidae). Caldasia 9:467-489. Price, Ll. 1953. Os quelonios de Formacao Bauru,
Medem, F. 1968. Desarrollo de la herpetologia en Cretaceo terrestre do Brasil meridional. Dept.
Colombia. Rev. Acad. Colombiana Cien. Exactas, Nac. Prod. Min., Div. Geol. Min., Rio de Janeiro
Fis. Nat. 13:149-200. 147:1-34.
Medina, C. 1976. Crocodilians from the Late Terti- Price, Ll. 1954. Um quelonio pleurodiro no Cal-
ary of Northwestern Venezuela: Melanosuchus cario da Serie Apodi, Cretaceo do Estado do Rio
fisheri sp. nov. Breviora (438):1-14. Grande do Norte. Dept. Nac. Prod. Min., Div.
Meszoely, C. 1970. North America fossil anguid Geol. Min., Notas Prelim. Estud, Rio de Janeiro
lizards. Bull. Mus. Comp. Zool. Harvard Univ. 85:1-12.
139:87-149. Price, Ll. 1964. Sobre o cranio de um grande croco-
Molnar, 1969. Jaw musculature and jaw mechan-
R. dilideo extinto do Alto rio Jurua, Estado de Acre.
ics ofthe Eocene crocodilian Sebecus icaeorliinus. An. Acad. Brasil. Cien. 36:59-66.
M.A. Thesis. Univ. Texas, Austin, 89 p. Rage, J.-C. 1974. Les serpents des Phosphorites du
Molnar, R. 1977. Crocodile with laterally com- Quercy, Palaeovert, 6:274-303.
pressed snout: First find in Australia. Science, Rage, J.-C. 1975. Un serpent du Paleocene du
197:62-64. Niger. Etude preliminaire sur l'origine des cae-
Mones, A. 1972. Lista de los vertebrados fosiles del nophidiens Reptilia, Serpentes
( ) . C. R. Acad.
Uruguay, I. Chondrichthyes, Osteichthyes, Rep- Sci. Paris, Ser. D
281:515-518.
ulia, Aves. Comun. Paleontol. Mus. Hist. Nat. Rage, J.-C. 1977. La position phyletique de Dinily-
Montevideo 1:23-36. sia patagonica, serpent du Cretace superieur.
Ibid. 284:1765-1768.
Mones, A. 1975. Notas paleontologieas uruguayas,
III. Vertebrados fosiles nuevos o poco conocidos Raven, P., Axelrod, D. 1975. History of the flora
( Chondrichthyes, Osteichthyes, Amphibia, Mam-
and fauna of Latin America. Amer. Sci. 63:420-
429.
malia). Ameghiniana 12:343—349.
1921. Bracliygnalliosuchus braziliensis, a Reic, O. 1958. Notas para una actualization del
Mook, C.
new fossil crocodilian from Brazil. Bull. Amer. conocimiento de la fauna de la Formation Chapad-
Mus. Nat. Hist. 44:43-50. malal. Acta Geol. Lilloana, Tucuman 2:255-283.
Mook, C. 1941. A new fossil crocodilian from Co- Reic, O. 1959. Primeros datos descriptivos sobre los
lombia. Proc. U.S. Natl. Mus. 91:55-58. anuros del eocretacico de la provincia de Salta,
Royo Y Gomez, 1945-46. Los vertebrados del Smith, A., Briden, J., Drewry, G. 1973. Phanero-
J.
Terciario continental colombiano. Rev. Acad. Co- zoic world maps, pp. 1-42 in Hughes, N. (ed.).
lombiana Cien. Exactas Fis. Nat. 6:496-512. Organisms and continents through time. Spec.
Rusconi, C. 1933. Observaciones criticas sobre rep- Pap. Palaeontol. ( Paleo. Assoc. London) 12:1-
tilesterciarios de Parana ( Alligatoridae) Rev. . 334.
Univ. Nac. Cordoba 20:57-106. Smith Woodward, A. 1901. On some extinct rep-
Rusconi, C. 1935. Observaciones sobre los gaviales tilesfrom Patagonia of the genera Miolania, Dini-
fosiles argentinos. An. Soc. Cient. Argentina 119: hjsia and Genyodcctes. Proc. Zool. Soc. London,
203-214. 1901:169-184.
Rusconi, C. 1946. Ave y reptil oligoceno de Men- Staesche, K. von. 1937. Podocnemis brasiliensis
doza. Bol. Paleontol., Buenos Aires 21:1-3. n. sp. aus der Oberen Kreide Brasiliens. Neuens
Savage, 1973. The geographic distribution of
J. Jahrb. Mineral. Geol. Palaeontol. Abh. Abt. A,
frogs: Patterns and predictions, pp. 351—445 in 77:291-309. Reprod. by Minist. Agricult., Div.
Vial, J. L. (ed.). Evolutionary biology of the Geol. Min., Brasil, Bol. 114 (1944).
anurans: Contemporary research on major prob- Stirton, R. 1953. Vertebrate paleontology and con-
lems.Univ. Missouri Press, Columbia, 470 p. tinental stratigraphy in Colombia. Geol. Soc.
Savage, 1974. The isthmian link and the evolution
J.
Amer. Bull. 64:603-622.
of Neotropical mammals. Nat. Hist. Mus. Los Suarez, J. 1969. Urn quelonio da formacao Bauru.
Angeles Cty. Contrib. Sci. (260): 1-51. Dept. Geograf., Fac. Filos. Cien. Letras, Pres.
Schaeffer, B. 1949. Anurans from the Early Ter- Prudente 2:35-54.
tiary of Patagonia. Bull. Amer. Mus. Nat. Hist. Tihen, J. 1962. New world fossil bufonids. Amer.
93:41-68. Midi. Nat. 68:1-50.
Sclater, J., Fisher, R. 1974. Evolution of the East Tihen, J. 1964. Tertiary changes in the herpeto-
Central Indian Ocean, with emphasis on the tec- faunas of temperate North America. Senckenber-
tonic setting of the Ninetyeast Ridge. Geol. Soc. giana Biol. 45:265-279.
Amer. Bull. 85:683-702. Tihen, ]. 1972. The fossil record, pp. 8-13 in Blair,
Segura, P. 1944. Estudio de la primera especie de W. F. (ed. ). Evolution in the genus Bufo. Univ.
tortuga fosilde Costa Rica con algunas generali- Texas Press, Austin, 457 p.
dades sobre el orden Testudines. Esc. Farm. Van Denvender, T. 1977. Observations on the Ar-
Guatemala, 6:9-29. (Not seen) gentine iguanid lizard Leiosaurus bellii Dumeril
Sige, B. 1968. Dents du micromammiferes et frag- and Bibron Reptilia, Lacertilia, Iguanidae). J.
(
Notostylops Beds of Patagonia. Bull. Amer. Mus. Vilas, J., Valencio, D. 1978. Paleomagnetism of
Nat. Hist. 67:1-22. the South American and African rocks and the
Simpson, G. 1933b. A new crocodilian from the age of the South Atlantic. Rev. Brasil. Geocien-
Notostylops Beds of Patagonia. Amer. Mus. Novit. cias 18:3-10.
Simpson, G. 1937b. An ancient eusuchian crocodile Warren, J. 1969. Chelid turtles from the mid-Ter-
from Patagonia. Ibid. 965:1-20. tiary of Tasmania. J. Paleontol. 43:179-182.
Simpson, G. 1938. Crossochelys, Eocene horned
Wieland, G. 1923. A new Parana Pleurodira. Amer.
turtle from Patagonia. Bull. Amer. Mus. Nat.
J. Sci. 5:1-14.
Hist. 74:221-254.
Simpson, G. 1942. A Miocene tortoise from Pata-
Williams, E. 1950. Testudo cubensis and the evo-
lution of Western Hemisphere tortoises. Bull.
gonia. Amer. Mus. Novit. (1209): 1-6.
Amer. Mus. Nat. Hist. 95:1-36.
Simpson, G., Minopiuo, L., Patterson, B. 1962.
The mammalian fauna of the Oivisadero Largo Williams, E. 1956. Podocnemis bassleri, a new
Formation, Mendoza, Argentina. Bull. Mus. species of pelomedusid turtle from the late Terti-
Comp. Zool. Harvard Univ. 127:239-293. ary of Peru. Amer. Mus. Novit. (1782): 1-10.
1979 BAEZ & GASPARINI: FOSSIL RECORD 51
Wood, R. 1970. A review of the fossil Pelomedusi- Wood, R. 1976h. Stupendemys geographicus, the
dae Testudines,
( Pleurodira ) of Asia. Breviora world's largest turtle. Ibid. (436):1-31.
(357): 1-23. Wood, R., Gamero, M. de. 1971. Podocnemis vene-
Wood, R. 1971. The fossil Pelomedusidae (Testu- zueletisis, a new fossil pelomedusid Testudines,
(
dines, Pleurodira) of Africa. PhD Dissert. Har- Pleurodira ) from the Pliocene of Venezuela and
vard Univ., 345 p. a review of the history of Podocnemis in South
Wood, R. 1972. A fossil pelomedusid turtle from America. Ibid. (376)1-23.
Puerto Rico. Breviora (392): 1-13. Wood, R., Moody, R. 1976. Unique arrangement of
Wood, R. 1975. Redescription of "Bantuchelys" con- carapace bones in the South American chelid
turtle Hydromcdusa maximiliani (Mikan). J.
golensis, a fossil pelomedusid turtle from the
Zool. 59:69-78.
Paleocene of Africa. Rev. Zool. Bot. Africaines
Wood, R., Pattersox, B. 1973. A fossil trionychid
89:127-144.
turtle from South America. Breviora (415):1-10.
Wood, R. 1976a. Two new species of Chelus (Tes- Zancerl, R. 1947. Redescription of Taphrosphys
tudines, Pleurodira) from the Late Tertiary of olssoni, a fossil turtle from Peru. Fieldiana Geol.
northern South America. Breviora (435): 1-26. 10:29-40.
APPENDIX
Appendix — The
fossil amphibians and reptiles recorded from the areas shown in
2:1.
figures 4-7 and their cor-
responding bibliographic references are listed below; the numbers correspond to those on the maps (Figs. 4-7).
The areas have been designated by conspicuous geographic names. Taxonomic entities not recognized in re-
cent studies are excluded.
CRETACEOUS
1. Alemania, Provincia de Salta, Argentina (Late Cretaceous).
anura: Pipidae: Saltcnia ibanezi Reig, 1959 (Reig, 1959; Parodi Bustos et al., 1960; Baez, 1975).
2. Laguna Umayo, Departamento de Puno, Peru (Late Cretaceous).
anura: Leptodactylidae ? (Sige, 1968).
crocodilia (Sige, 1968).
3. Mossoro, Estado do Rio Grande do Norte, Brasil ( Late Cretaceous ) .
testudines: Pelomedusidae:
Roxochelys vilavilcmis Broin, 1971.
?
5. Sao Jose do Rio Preto, Estado de Sao Paulo Brasil (Late Cretaceous).
testudines: Pelomedusidae: Podocnemis brasilieniis Staesche, 1937 2 (Price, 1953; Arid and Vizotto, 1966;
Broin, 1971); Roxochelys wanderleyi Price, 1953; Podocnemis elegans Suarez, 1969. 2 Pieropolis, Estado
de Minas Gerais, Brasil (Late Cretaceous).
squamata: Sauria: Iguanidae: Prist iguana brasilicnsis Estes and Price, 1973.
6. Northwest of Lago Colhue-Huapi, Provincia del Chubut, Argentina (Late Cretaceous ?).
testudines: Meiolaniidae: Niolamia patagonica Ameghino, 1899 (Smith Woodward, 1901; Simpson
1938).
PALEOCE NE-EOCE NE
1. Golfo de San Jorge, Provincia del Chubut, Argentina (early Paleocene).
testudines: (Gasparini and Baez, 1975; Baez and Gasparini, 1977).
crocodilia: Crocodylidae: Necrosuchus ionensis Simpson, 1937 (Gasparini and Baez, 1975).
2. Itaborai, Estado de Rio de Janeiro, Brasil (late Paleocene).
anura: Pipidae: Xenopus romeri Estes, 1975 (Estes, 1975a, b); Leptodactylidae (Estes, 1970); Bufoni-
dae (Estes, 1970); Hylidae (Estes, 1970).
gymnophiona: Caeciliidae: Apodops pricei Estes and Wake, 1972.
testudines: Pelomedusidae: Podocnemis
(Paula Couto, 1970).sp.
squamata: Sauria: Iguanidae (Estes, 1970); Teiidae (Estes, 1970; Paula Couto, 1970); Gekkonidae
(Estes, 1970). Serpentes: Boidae (Estes, 1970; Paula Couto, 1970); Aniliidae (Estes, 1970; Hoffstetter
and Rage, 1977).
crocodilia: Sebecidae: Sebecus sp. (Paula Couto, 1970); Alligatoridae (Paula Couto, 1970).
J
F. de Broin ( pers. comm. ) considers the assignment of P. brasilicnsis and P. elegans to the genus Podocnemis
to be questionable.
52 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
OLIGOCE NE-MIOCE NE
3
1. Campo Waldo, Departamento de Santander, Colombia (Oligocene).
crocodilia: Sebecidae: Sebecus sp. (Langston, 1965); Crocodylidae ( Langston, 1965).
testudines (Stirton, 1953).
2. Scarrit Pocket, Provincia del Chubut, Argentina (early Oligocene).
anura: Leptodactylidae: Caudiverbera caudiverbcra (Linnaeus) (Schaeffer, 1949; Lynch, 1971); Eu-
psophus sp. (Schaeffer, 1949); Neoprocoela edentatus (Schaeffer, 1949).
3. Tremembe, Estado de Sao Paulo, Brasil (early Oligocene).
testudines: Chelidae (Wood and Patterson, 1973).
4. South of Lago Colhue-Huapi, Provincia del Chubut, Argentina (late Oligocene).
anura: Leptodactylidae: Caudivcrbera sp. (Schaeffer, 1949; Baez, 1977).
5. Chaparral, Departamento de Tolima, Colombia (late Oligocene-early Miocene).
crocodilia: Alligatoridae: Balancrodus logimus Langston, 1965; Gavialidae (Langston, 1965).
6. Gaiman, Provincia del Chubut, Argentina (late Oligocene).
testudines: Testudinidae: Geochelonc gringorum (Simpson, 1942) (Williams, 1950; Auffenberg, 1971;
de la Fuente, pers. comm.).
7. Southern Provincia de Santa Cruz, Argentina (early -middle Miocene).
squamata: Sauria: Iguanidae (Ameghino, 1899; Gasparini and Baez, 1975; Baez and Gasparini, 1977);
Teiidae: Diasemosaurus occidentalis Ameghino, 1893 (Gasparini and Baez, 1975); Serpentes (Ameghino,
1899).
8. Coyaima, Departamento de Tolima, Colombia (late Miocene).
testudines: Chelidae: Chelus colombianus Wood, 1976a.
squamata: Sauria: Teiidae: cf. Tupinambis (Estes, 1961).
crocodilia: Sebecidae: Sebecus sp. (Langston, 1965); Alligatoridae (Langston, 1965); Crocodylidae
(Langston, 1965); Gavialidae: ? Gavialis colombianus Langston, 1965.
9. Carmen de Apicala, Departamento de Tolima, Colombia (late Miocene).
testudines: Pelomedusidae (Royo y Gomez, 1945-1946; Stirton (1953); Chelidae: Chelus colombianus
Wood, 1976a.
crocodilia: Alligatoridae: Eocaiman sp. (Langston, 1965); Caiman neivensis Mook, 1941 (Langston,
1965).
Tertiary amphibian and reptile bearing deposits of Colombia are assigned chronologically according to Van
3
The
Houten and Travis (1968) and Irving (1971); in an earlier paper the authors (1977) followed Stirton (1953).
1979 BAEZ & GASPARINI: FOSSIL RECORD 53
Chelus colombianus Wood, 1976a; Testudinidae: Geochelone (Chelonoides) hesterna Auffenberg, 1971; Em-
ydidae (Medem, 1968).
squamata: Sauria: Iguanidae (Estes, 1961); Teiidae: Tupinamhis cf. T. tequixin (Estes, 1961); Dra-
caena colombiana Estes, 1961; Serpentes: Aniliidae: Colombophis portai Hoffstetter and Rage, 1977; Boi-
dae: Eunectes stirtoni Hoffstetter and Rage, 1977; Colubridae ( Hoffstetter, 1967b; Hoffstetter and Rage,
1977).
crocodilia: Sebecus huilensis Langston, 1965; Scbecus sp. (Langston, 1965); Alligatoridae:
Sebecidae:
Eocaiman sp. (Langston, 1965); Caiman neivensis Mook, 1941 (Langston, 1965); Caiman cf. C. latirostris
( Daudin, 1802) (Langston, 1965; Baez and Gasparini, 1977); Crocodylidae: Charactosuchus fieldsi Lang-
ston, 1965; Nettosuchidae: Mourasuchus atopus Langston, 1965 (Langston, 1966); Gavialidae: cf. Rham-
phostomopsis (Langston, 1965).
11. North of Lago Buenos Aires, Provincia de Santa Cruz, Argentina (late Miocene).
anura: Leptodactylidae: Caudiverbera caudiverbera Linnaeus Casamiquela, 1958; Lynch, 1971).
(
12. Barranca de los Loros, Provincia de Rio Negro, Argentina (late Miocene).
anura: Leptodactylidae: Caudiverbera caudiverbera Linnaeus (Casamiquela, 1963; Lynch, 1971).
13. Ingeniero Jacobacci, Provincia de Rio Negro, Argentina (late Miocene).
anura: Leptodactylidae: Wawelia gerholdi Casamiquela, 1963.
PLIOCENE-PLEISTOCENE
1. Urumaco, Estado de Falcon, Venezuela (middle Pliocene).
testudines: Pelomedusidae: Stupetidemys geograpliicus Wood, 1976b; Chelidae: Chelus lewisi Wood,
1976a; Trionychidae (Wood and Patterson, 1973); Testudinidae (Wood and Patterson, 1973).
crocodilia: Alligatoridae: Melanosuchus fisheri Medina, 1976; Crocodylidae ?: Gryposuchus sp. (Pat-
terson, pers. comm.); Gavialidae: Ikanogavialis gameroi Sill, 1970; Nettosuchia: Mourasuchus amazon-
ensis Price, 1964 (Patterson, pers. comm.).
2. Valle de Santa Maria, Provincia de Catamarca, Argentina ( middle Pliocene ) .
squamata: Sauna: Teiidae (Ambrosetti, 1890; Ga;parini and Baez, 1975; Baez and Gasparini, 1977); Ser-
pentes: Boidae (Bravard, 1858; Burmeister, 1883, 1885).
crocodilia: Alligatoridae: Caiman latirostris (Daudin, 1802) (Gasparini and Baez, 1975); C. australis
(Burmeister, 1885); cf. C. jacare (Daudin, 1802) ( Gasparini and Baez, 1975); C. sp. (Gasparini and Baez,
1975; Baez and Gaspirini, 1977); Gavialidae: Rhamphostomopsis neogaeus (Burmeister, 1885) (Rusconi,
1933, 1935; Gasparini, 1968).
4. Monte Hermoso, Provincia de Buenos Aires, Argentina (late Pliocene).
anura: Leptodactylidae: Ceratophrys prisca Ameghino, 1899 (Rovereto, 1914); Bufonidae (Gasparini
and Baez, 1975).
testudines: Testudinidae: Geochelone gallardoi (Rovereto, 1914) (Auffenberg, 1974).
squamata: Sauria: Teiidae: Tupinamhis sp. (Rovereto, 1914); Callopistes bicuspidatus Chani, 1976.
5. Rio Quequen Salado, Provincia de Buenos Aires, Argentina (late Pliocene).
anura: Bufonidae: Bufo pisanoi Casamiquela, 1967.
6. Chapadmalal, Provincia de Buenos Aires, Argentina (late Pliocene).
anura: Bufonidae: Bufo pisanoi Casamiquela, 1967; Leptodactylidae: Ceratophrys sp. Reig, 1958).
(
Investigator Titular
Fundacion Miguel Lillo
Miguel Lillo 205
4000 Tucumdn, Argentina
At the height of the Matthewsian theory much poorer than it is now. Presently, some
of continental biogeography ( Matthew, 1915; credence can be given to negative evidence,
Darlington, 1957), the very title of this paper namely, the lack of fossils, in North America
would have been almost preposterous, at least and Europe of groups that elsewhere have
in the influential herpetological centers of extensive fossil records (e.g., turtles and
North America dominated by Noble (1931), crocodilians).
Dunn (1923, 1931), and Schmidt (1946). The New evidence definitely points to the
dissident voices of Jeannel (1942), Du Toit union of Africa and South America into a sin-
(1937) and others were stifled as inconse- gle continent from at least the Late Carbonif-
quential. Africa and South America were sup- erous until the Cretaceous. Then a graben
posed to have had no relationships whatever formed a narrow gulf north and south of a
for a very long time. Indeed, the differences residual bridge between northeastern Brasil
between the Ethiopian and Neotropical her- and Nigeria. During the last half of the Tu-
petofaunas are striking and seem to support ranian, the continents were split with the birth
the idea of independent histories. Most domi- of the South Atlantic Ocean, which initially
nant groups are represented in Africa and was quite narrow. This scenario was de-
South America by pairs of adaptive equiva- scribed by Reyment ( 1975 ) and is supported
lents or vicarious groups in Simpson's (1965) by convincing geological evidence paleo-
—
sense (Table 3:1). magnetism ( Creer, 1973 ) fit of the continents
,
South America have been explained by ex- (Heirtzler, 1968; Francheteau, 1973), and
tinctions in the Holarctic Region. In some plate tectonics with the fitting of cratons and
cases the fossil record seems to bear out this rocks ( Hurley, 1968; Dietz and Holden, 1970;
explanation (e.g., turtles of the family Pelo- Douglas et al., 1973). The stratigraphic con-
medusidae). When such paleontological evi- cordances are especially striking (Reyment
dence was lacking, it often was implied; Dunn and Tait, 1972), as well as the structure of
(
1931 ) emphasized that it would not be sane the coastal basins with their immense amounts
reasoning to postulate a trans-Atlantic land of fresh water sediments (Martin, 1968),
bridge just because no fossils were known which suggest a graben phase like that of the
from the northern continents. When Dunn African Rift Valleys. The salt deposits in An-
made his statement the fossil record was gola, Gabon and Brasil are reminiscent of
—
Table 3:1. Family Groups of Amphibians and Rep- later phases like that of Lake Turkana or the
tiles that have Trans-Atlantic Counterparts.
Red Sea (Reyment, 1975).
Paleontological data are still more con-
South America Africa
vincing (Baez and Gasparini, this volume).
Leptodactylidae Ranidae (sensu Liem, 1970) The amphi-Atlantic distribution of the meso-
Hylidae Hyperoliidae (sensu Liem, 1970) saurians of the early Permian (Romer, 1966)
Iguanidae Agamidae + Chamaeleontidae
is strong evidence for a Gondwanan land mass
Teiidae Lacertidae
Boinae Pythonini ( Colbert, 1973 ) . The early Triassic of Argen-
Crotalinae Viperinae tina has revealed a Cynognathus fauna almost
identical to that of the upper Beaufort beds
Investigator Principal del CONICET. of South Africa (Bonaparte, 1967). Two
55
56 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
groups of mammals
provide evidence that the SUMMARY OF DISTRIBUTION
Atlantic was not the wide ocean of today in PATTERNS
the Eocene and Oligocene. Primates and cav-
iomorph rodents, conspicuously lacking in the The first step in a biogeographic analysis
early South American mammalian fauna, nev- must be a summary of distribution patterns
ertheless entered the continent long before or "generalized tracts" (Croizat, 1964). These
the Late Cenozoic invasion via Panama. patterns are listed below.
Classically, it was assumed that they entered 1. Gondwanan or West Gondwanan
South America by island-hopping from North —
groups. Geotry petes- Apodops, Pipi-
America (Wood, 1950; Simpson, 1965; Patter- dae, Pelomedusidae, Amphisbaenidae,
son and Pascual, 1968). However, another Typhlopidae, Lcptotyphlopidae.
hypothesis assuming an African origin and a 2. South American groups that invaded
trans-Atlantic migration has been maintained
Africa before the separation of the con-
by Lavocat (1974) and Hoffstetter (1972, tinents. —Bufonidae, Iguanidae (
ex-
1977). Krommelbein (1971) showed that the tinct in Africa but surviving in Mada-
fossil,fresh water ostracods from the coastal
gascar).
basins of Reconcavo, Sergipe, Brasil, and of 3. Presumed Indian groups that invaded
Gabon are almost identical. Two different am- South America via Africa before the
monite faunas (one in the Potiguar Basin in
separation of the continents.
— Microhy-
northern Brasil and another in the Sergipe- lidae.
Alagoas Basin) existed until the lower Turan- 4. African groups that invaded South
ian. Then the African-Brasilian isthmus America after the separation of the
disappeared, and the two faunas mingled continents. —
Gavialidae (?), Gekkoni-
(Beurlen, 1961; Reyment, 1958, 1970). These nae, Scincidae, Amphisbaenidae (?),
data provide evidence for the birth of the Colubrinae (
? )
.
Atlantic Ocean and the separation of Africa 5. Holarctic groups tliat invaded Africa
and South America in the middle Turanian, and South America from the north. —
Testudinidae, Crocodylidae, Colubri-
about 90-95 m.y.b.p.
nae (?).
This establishes a firm basis on which to 6. African groups that recently invaded
proceed to determine which groups existed in South America by a northern route. —
both continents before their separation and Ranidae.
which emigrated from one to the other before 7. Neotropical groups absent from Afri-
or after their separation. The Hennigian sys-
ca. —
Rhinatrematidae. Dermophiinae, 1
tematists and biogeographers Caeciliidae, Typhlonectidae, Bolito-
Brundin, 1966;
(
3
Croizat, 1964; Nelson, 1973; Croizat et al., glossini, Leptodactylidae,-' Hylidae,
Centrolenidae, Pseudidae, Dendrobati-
1974) have insisted on the necessity of vi-
dae, Sphacrodactylinae, Iguanidae,
cariance events in determining generalized
Teiidae, Anguidae, Boini, Xenodonti-
tracts reconstructing biogeographic his-
in
nae, Micrurinae, Crotalinae.
tories. Their conclusions are in general agree-
-
ance at the expense of geocenters and dis- The African Heleophryninae are considered to be
members of the Myobatrachidae by Lynch (1973).
persal. (Appendix 3:1). Their approach is
3
The Hylidae, as well as other groups, like Disco-
applicable to the study of relationships be- glossidae, Pelobatidae, Salamandridae, and Angui-
tween the African and South American faunas. dae, are present in northern Africa. That part of
There was an old African-Brasilian (Inabre- Africa belongs to the Palaearctica Region and is not
considered to be relevant here, although the past
sian, fide Jeannel, 1942) fauna, and there also existence of some of its fauna in the Ethiopian Re-
are vicariant groups. gion cannot be ruled out entirely.
1979 LAURENT: AFRICA AND SOUTH AMERICA 57
8. African and Old World groups absent Geotrypetes has as many chromosomes as
—
from South America. Scolecomorphi- the ichthyophiids, which also include the In-
dae, Herpelinae, Heleophryninae, Hy- dian Uraeotyphlus (Nussbaum, pers. comm. ).
peroliidae, Agamidae, Chamaeleonti- Also, caecilians, like other lissamphibians,
dae, Lacertidae, Cordylidae, Varanidae, demonstrate a negative correlation between
Pythonini, Lycodontinae, Dasypeltinae, chromosome number and the number of de-
Elapidae, Viperinae. rived character states
(Laurent, press). in
9. Groups that apparently once lived in Thus, Apodops might belong to an African-
Africa but are now extinct there. Igua-
— Brasilian group of primitive caecilians, from
nidae. which the African and Neotropical caecilians
10. that icere
descended. By this reasoning, the other cae-
Groups presumably pan-
Gonduanan in the Jurassic. —Leiopel- South America
cilians are actually different in
pipids, which apparently are an early special- rocco (Vergnaud-Grazzini, 1966)]. Xenopus
ized derivative from Jurassic frogs. The pipids presently is speciose in Africa, where the spe-
have a typical west Gondwanan distribution cialized Hijmenochirus and Pseudhymenochi-
and are now common to Africa and South rus occur in forests. Xenopus is extinct in
America. They are known from both conti- South America, but perhaps through a Sal-
nents by a number of Mesozoic and Cenozoic teniaAike stock it gave rise to the modern
fossils, beginning with the Early Cretaceous South American pipids, now restricted to for-
of Israel (Nevo, 1968). It has been argued ested regions of northern and eastern South
that the pipids might have lived in the north- America (Estes, 1975b).
ern continents and subsequently migrated in- Other archaeobatrachians have lived in
dependently into South America and Africa. South America Baez and Gasparini, this vol-
(
However, the family is conspicuously absent ume). Vieraella (Lower Jurassic) and Noto-
in the fossil record in the north; it is replaced batrachus (Upper Jurassic) were placed in
there by the ecologically similar palaeobatra- the Leiopelmatidae by Estes and Reig
chids, now extinct, but apparently common (
1973 ) who suggested that the leiopelmatids
,
from the Jurassic to the Pliocene. 5 Estes radiated during the Jurassic in Gondwana-
(1975a) reported a fossil of Xenopus from land and that Ascaphus is a remnant of a
the Paleocene of Brasil and referred the northward migration at the end of the Meso-
Eocene Shelania from Patagonia to the same zoic. Leiopelma survived in New Zealand,
genus. Therefore, there is little reason to where it is the only frog. Discovery of fossil
doubt the Inabresian origin of the family. leiopelmatids other parts of Gondwana-
in
Estes (1975b) synthesized the phylogeny of land, especially Africa, is expected.
question, because the Australian region is de- sternal Noble (1931) sup-
pectoral girdle.
void of bufonids. The last was suggested bv ported Boulenger. Orton (1957) emphasized
Rlair (1972b). Laurent (1972, 1975) did not the apparent primitiveness of the microhylid
reject dispersal through North America but tadpoles, which are similar in many respects
proposed that bufonids also dispersed through to those of pipids. Orton believed that it was
Africa. Laurent was influenced by Estes' unlikely that such an adaptive complex of
( 1970, pers. coram. ) insistence that no bufo- features as the larval mouth in most anurans
nid entered Nortli America before the Mio- would be lost; therefore, she thought that the
cene. There is further evidence in favor of an microhylids were related to the pipids and
African dispersal route. 1) The South Amer- represented an early radiation among frogs.
ican bufonid radiation is the largest (Trueb, This contention was resisted by several herpe-
1971; McDiarmid, 1971; Cei, 1972) and the tologists beginning with Griffiths (1963).
African is next with seven genera and about The traditionalists include Griffiths and Car-
50 species (Tihen, 1960; Tandy and Keith, valho (1965), Tihen (1965), and Kluge and
1972), followed by Eurasia with only six gen- Farris (1969). The "Ortonists" include
era and about 40 species and finally North Hecht (1963), Inger (1967), and Starrett
America with one genus and some 20 species. ( 1973 ) ,
who based her opinion on a detailed
2) Few bufonids have retained an omoster- study of tadpoles. Savage (1973) enthusi-
num, a plesiomorphic character for the fam- astically based a new zoogeographic scheme
ily. These include the Bufo haematiticus on the apparent strengths of Starrett's conclu-
group in northern South America, the African sions, and even included Australia in the
genus Nectophrynoides, and possibly Wemer- original Gondwanan realm of the family ( see
ia in west Africa. 7 3) Relationships between Tyler, this volume, for contrary zoogeographic
the Neotropical and African bufonids is sup- arguments Lynch ( 1973 ) showed that such
)
.
ported by the high degree of genetic compat- an evolutionary scheme of the microhylids re-
ibility between the South American Bufo quired the independent acquisition of no less
arenarum and the African B. regularis (Blair, than 13 characters present in the ranids. In a
1972a), by the striking similarity of peculiar detailed study of tadpole structure, Sokol
species like the African B. superciliaris and (
1975 ) demonstrated that the microhylids
the Neotropical B. blombergi (Blair, 1972b), had the larval papillae, denticles, and
lost
and by serological affinities (Cei, 1977). Such horny beaks.- Therefore, the microhylids are
evidence induced Laurent (in press) to em- related to the ranoids.
phasize the African route rather than the
Microhylids have a semirelictual distribu-
North American one. A dispersal following intermediate between the scattered pat-
tion,
the described route need not exclude a Mio- terns of some old families, such as the Disco-
cene invasion of North America from South
glossidae and Pelobatidae, and the compact
America, but the African dispersal was much
patterns of the more modern, still radiating
earlier (±90 m.y.b.p. versus ±25 m.y.b.p.) groups, like the Ranidae and Bufonidae.
and therefore much more important to the Therefore, the microhylids must be older than
evolutionary biogeography of the family. the ranids. This idea is supported by the pres-
The systematic position of the Microhyli- ence of 28 chromosomes in the microhylid
dae is the most controversial matter in the Kaloula; the number is not the result of sec-
taxonomy of frogs. Boulenger (1882) con- ondary fusions and therefore is a
plesiomor-
sidered them (as the Engystomatidae) to be phic feature similar to the karyotype of Dk-
related to the Ranidae, because of their firmi- coglossus. The presence of a variety of micro-
hylids in South America also supports the
'Andersson (1903) mentioned the presence of a
vestigial omosternum in his description of Steno-
'
antiquity of the family; the ranids barely America. The genera living in temperate
enter South America. Asia (Microhijla) and North America {Gas-
trophryne) have reduced pectoral girdles.
Assuming that the microhylids evolved, Carvalho (1954), Nelson (1966), Nelson and
like the ranoids, in the eastern part ofGond-
Cuellar ( 1968 ) questioned the affinities be-
wanaland, their radiation was around the
tween the Asiatic and American microhylines.
Indian Ocean ( tropical Asia, East Indies, and
There is a possible trans-Gondwanan path-
Madagascar). The most primitive subfamilies
and Dyscophinae) live in way from India to Madagascar to Africa to
( Scaphiophryninae
South America that is marked by a series of
Madagascar and tropical Asia. The eastern
genera having complete pectoral girdles
—
groups in New Guinea ( Sphenophryninae and
Kalophrynus (tropical Asia), Mclanohatra-
Asterophryinae) are not relevant here, but
—
the other subfamilies are Rrevicipinae, Hop-
chus (India), Scaphiophryninae and Dysco-
9 phus (Madagascar), Brevicipinae and Parho-
lophryninae and Phrynomerinae in Africa,
plophryne (Africa) and Otophryne and
Cophylinae in Madagascar, and Microhylinae Dermatonotus (South America). If such a
in Asia and America.
dispersal took place, it might have occurred
We can dismiss as improbable an Holarc- before or slightly after the birth of the At-
tic origin of the family, for the primitive sub- lantic Ocean.
families live in the tropics. Thus, only Mada- Chromosome numbers support this hypoth-
gascar and the Gondwanan part of Asia are esis ( Morescalchi, 1973; Bogart and Nelson,
likely centers of origin of the microhylids. 1976; Bogart et al., 1976). In Asia, Kaloula
Both are possible, for during its northward has 28 chromosomes, which equals the rela-
drift India apparently was connected at times
tively primitive number of Discoglossus; other
with Madagascar by garlands of islands (Mc- Asian genera ( Uperodon, Ramanella, Micro-
Kenzie and Sclater, 1973), some of which hijla ) are known to have 26 chromosomes. In
continue to exist as the Seychelles, Amirante, Africa there are 26 in Phrynomerus and 24 in
Mascarene, Maldive and Laccadive Islands Breoiceps. In America there are 26 chromo-
(Appendix 3:2). Later, contacts were with somes in the primitive genera Otophryne and
Malaysia and Indochina through the Nicobar Glossostoma, 24 in Chiasmocleis and 22 in
and Andaman islands, allowing the Indian seven genera, including the widespread Gas-
fauna to invade eastern Asia, Indonesia, and trophryne and Elachistocleis.
even the East Indies. The African invasion Considering now only those families that
likely passed through the Mozambique Chan- are present on one side of the Atlantic, we see
nel and its islands (e.g., Comores). The Af- some striking parallelisms. South American
rican groups of microhylids are now highly leptodactylids, rhinodermatids and dendro-
differentiated. batids are paralleled by the terrestrial ranids
The American microhylines are a problem. and hyperoliids in Africa; the hylids and cen-
According to the Matthewsian theory, the trolenids in South America are paralleled by
microhylines originated in tropical Asia, in- the arboreal ranids (Chiromantis) and hyper-
vaded the eastern Palaearctic Region, and oliids in Africa. Moreover, some peculiar
passed into North America by the Bering isth- adaptations in one continent have counter-
mus and into South America by island-hop-
—
parts in the other the aquatic South Amer-
ping well before the Pliocene. There are seri- ican pseudids (coexisting with pipids) versus
ous objections to this hypothesis. In Asia, as African pipids; the rheophilous South Ameri-
well as in America, the primitive genera hav- can telmatobiines versus African heleophry-
ing a complete pectoral girdle are in the trop- nines; atelopine bufonids and brachyecpha-
ics; these are Kalophrynus, Chaperina, Me- lids in South America versus Didynamipus in
lanobatrachus and Gastrophrynoides in Asia Africa. In some cases the resemblances are
and Otophryiie and Dermatonotus in South striking. For example, compare Physalacmus
biligonigerus in South America with Tomop-
Mclanohatrachus included in the
terna delalandii (and congeners) in Africa;
is Microhylinae
(Savage, 1973; Laurent, in press). compare the South American Leptodactylus
1979 LAURENT: AFRICA AND SOUTH AMERICA 61
fuscus with African species of Ptycltadena, son, 1942), Miocene for Emydidae (Medem,
and the Hyla leucophyllata group in South 1968), a short Pliocene apparition for the Tri-
America with species of Afrixalus. On the onychidae in Venezuela (Wood and Patter-
other hand, some adaptations are unique to son, 1973). Only the Testudinidae and Tri-
one continent. Africa has nothing like the onychidae became established in Africa,
marsupial tree frogs (Amphignathodontinae); where they are known since the Miocene
South America has no frog emulating the sex- (Romer, 1966).
ual dichromatism of the tribe Hyperoliini (for
other examples, see Laurent, 1973).
Crocodilians. —The living crocodilians do
not show significant similarities between
South America and Africa. The Crocodylidae
Reptiles is present in both continents but relatively
Chelonians. —The Pelomedusidae is a clas- unimportant in South America; the Alligator-
sical case of an amphi-Atlantic distribution, idae, absent from Africa, radiated impressive-
which has been explained by Matthewsians ly in South America. However, before the
as an Holarctic origin and southward migra- formation of the Atlantic Ocean in the Creta-
tions. Others have explained the distribution ceous, the crocodilian fauna, composed ex-
by the fragmentation of a primitively Gond- clusively of mesosuchians, was much the same
wanan range. The alternatives are not clear, in South America and Africa. Thus, the Afri-
for there are northern fossils. Some very old can Libycosuchidae were small, blunt-snouted
turtles from Germany generally classified in crocodiles, very similar to the South American
other suborders are really pleurodires (de Notosuchidae (Sill, 1968; Buffetaut, 1976)."
Broin, pers. comm.). These are the Triassic
The gigantic, long-snouted pholidosaurid ge-
Proterochersis ( Proganochelydia ) and the Ju- nus Sarcosuchus was common to Brasil and
rassic Platychelys ( Amphichelydia). 10 west Africa (Buffetaut, pers. comm.). After
Quite
an array of other northern genera are known the severance of the last remnants of a bridge,
from Upper Cretaceous to Oligocene beds be- the faunas gradually became different. Al-
longing to the shores of the young Atlantic though the mesosuchians were dominant over
Ocean. According to de Broin (pers. comm.), the eusuchians until well into the Cenozoic in
the marine coastal Bothremydidae is a sister the scattered Gondwanan continents, they
family of the Pelomedusidae. The
oldest pelo- became subordinate them in Laurasia ( Buf-
to
medusid fossil is Platycheloides from the Low- fetaut, pers. comm.). The Dyrosauridae (Up-
er Cretaceous of Africa. Thus, the
family was
per Cretaceous and Early Cenozoic mesosuch-
in existence before the birth of the Atlantic ians), extremely long-snouted, gavial-like
Ocean. Later, the exclusively African pelo- creatures, although essentially African, also
medusines (not known before Oligocene) lived at the end of the Cretaceous on the west
were separated from the Podocneminae, side of the still trench-like Atlantic, but this
which flourished in South America, Africa, can be ascribed to their littoral habits (Buf-
still narrow ocean in the Eocene seems to be Vanzolini, 1968). This dispersal was easier
the only explanation of their distribution, for when the Atlantic Ocean was narrower. Other
the family is first known in the Eocene ( Hecht gekkonine stocks (e.g., Briba and Bogertia)
and Malone, 1972) of Africa. may have entered South America in the Late
Lizards. —Only the Gekkonidae and Igua- Cretaceous or Early Cenozoic. Possibly some
nidae are involved in the Afro-American sep- (e.g., Homonota) immigrated into South
aration. Other families of lizards seem to be America before the separation of the conti-
parallel radiations after the
—
separation Aga-
nents. Bons and Pasteur (1977) suggested an
midae and Chamaeleontidae similar to the early immigration for the two Neotropical
Iguanidae; Lacertidae, Cordylidae and Scin- species assigned to the African genus Lygo-
cidae emulating the Teiidae and some terres- dactylus.
trial iguanids; the Varanidae copied by large Estes and Price (
1973 ) believed that the
teiids, like Tiipinaml)is. Also there is the eco- Iguanidae originated in South America when
it was still united to Africa and invaded Af-
logical similarity between the numerous "mi-
croteiids" and the lygosomine skinks (Lau- rica, where they became extinct, and Mada-
rent, 1973). gascar, where they survived. Alternatively,
they could have originated in Africa, where
Although some archaic lizards were pres- the related agamids and chamaeleontids sup-
ent in the Triassic, these apparently have little
in common with Jurassic ones, which belong planted them. If the Iguanidae and Teiidae
have a common ancestor, the presence of
to the modern infraorders (Robinson, 1967;
teiids in North America in the Cretaceous,
Hoffstetter, 1955, 1967). Some of the modern
were contrasting to the absence of iguanids there
families present in the Cretaceous.
Thus, when Africa and South America drifted (Estes, 1970), is puzzling and cannot be ex-
plained with our present data.
apart, several recent families were already in
The past existence of iguanids in Africa is
existence. The iguanid, Pristiguana (Estes
and from the Cretaceous of Bra- hardly questionable, for they are still living in
Price, 1973 ) ,
silhas some characters of the Teiidae. Chrom- Madagascar. Is their extinction in Africa the
result of competition with agamids? Not
osome morphology is similar in iguanids and
teiids (Gorman, 1970). likely, because the African agamids are not
diverse and therefore unlikely to out-compete
The Gekkonidae may be the oldest family the diversified iguanids. Also, it is unlikely
of modern lizards. It existed in Brasil in the
that the chameleons out-competed the igua-
Paleocene (Estes, 1970). The Jurassic Ardeo-
nids, except for possibly some arboreal types,
sauridae presumably is ancestral to the gek-
for the chameleons are a highly specialized
konids and so similar to them (Hoffstetter,
group of lizards. Furthermore, the agamids
1964) that in a cladistic system they could be are probably relatively recent immigrants into
included in the gekkonids. Therefore, the
Africa from Eurasia. Rafting of iguanids and
presence of gekkonids in western Gondwana- teiids from South America to Africa is not
land before the formation of the Atlantic
possible now, because of the direction of the
graben is realistic.
ocean currents, but 50-80 m.y.b.p. such an
The relict and disjunct distribution of the event was more likely. In the reverse direc-
Eublepharinae is best explained by a northern
tion, the feasibility of a successful crossing has
origin. On the other hand, the Sphaerodac- been proved by Ilemidactylus (Kluge, 1969),
tylinae is likely to be a strictly Neotropical but no cases are documented for agamids,
derivative of a Gondwanan stock. Most of chamaeleontids, lacertids, or cordylids.
the South American Gekkoninae seem to have If, as indicated by Estes and Price (1973),
come from Africa by waif dispersal after the iguanids and teiids are respectively the roots
formation of the Atlantic Ocean. This is fairly of the Iguania and Scincomorpha radiations,
certain and recent for the species common to an eastern invasion of "eoteiids" is suggested.
both continents, such as Ilemidactylus hrooki Estes ( pers. comm. ) sees the teiids as an
and //. mabouia (Kluge, 1969) and hardly essentially Neotropical radiation with a lacer-
less obvious for Turentohx (Kluge, 1967; toid derivation through northern Gondwana-
1979 LAURENT: AFRICA AND SOUTH AMERICA 63
land. The Scincidae have relationships be- lineage (Rage, 1977), using the Bering path-
tween the southern Atlantic continents, but way, entered Africa in the lower Miocene.
the few South American species of Mahmja Rage ( pcrs.
comm. ) contemplates three pos-
probably came from Africa long after the sep- sible origins for the Pythonini —
Africa or Aus-
aration of the continents. tralasia, both of which he considers to be
Amphisbaenians.-
—The worm-lizards have doubtful, and Asia, a choice also favored by
a typical western Gondwanaland distribution. Underwood (1976, pers. comm.). Thus, boids
They are an old group, and their existence in supposedly migrated from North America to
the Inabresian continents is likely. Fossils are Asia via the Bering land bridge. A fourth
known from North America (Eocene to Pleis- possibility is the Indian raft (see Appendix
tocene) and belong to several extinct genera, 3:2), which may explain why boids flourish
as well as to the extant Rhincura and Lcpo- in Australasia and how they later came to
sternon, now
surviving only in South America Africa from Asia, presumably in Miocene
(Romer, 1966). The extinct Omoiotyphlops times. 13
is from the Eocene-Pliocene of Europe ( Rom- The least understood of all is the
groups
er, 1966; Hoffstetter, Therefore the
1962). vast array of higher snakes, the caenophidians
range of the amphisbaenids underwent a con- or Colubroidea. Until recently, the fossil rec-
traction similar to that of many tropical ord of caenophidians was exclusively Holarc-
groups that lived in Europe and North Amer- tic and only back to the Miocene. New data
12
ica. show that caenophidians existed in Europe in
Snakes. —Among
the primitive scoleco- the lower Eocene, and the Colubridae (sensu
phidians, the Leptotyphlopidae has about the lato) is the middle Oligocene. 14
known from
same range as the amphisbaenians and pre- Rage (1975) described Nigerophis from the
sumably the same history. The Typhlopidae Paleocene of Africa; this genus seems to be
has a pantropical range, which can be deemed intermediate between the caenophidians and
pan-Gondwanan until there is
contrary evi- the Palaeophidae. Such a systematic position
dence.
suggests an aquatic origin of modern snakes
The Boidae also is considered to be a and makes their paleogeographic history even
Gondwanan group; nonetheless, the present more difficult to interpret. Rabb and Marx
distribution is the result of complex migra- (
1973 ) suggested that the group perhaps had
Primitive fossil genera have been found
tions. a tropicopolitan distribution before the Gond-
in southern continents —
Laparrcntophis in the wanan fragmentation, but Rage (1976) dis-
Lower Cretaceous of northern Africa, Dini-
agreed. The scarcity and primitiveness of the
Jysia in Patagonia (Upper Cretaceous), and few colubroids from the early Cenozoic sup-
Madtsoia in South America, Africa, and Mad-
port Rage, rather than Rabb and Marx.
agascar (Upper Cretaceous to Paleocene). The problem is compounded by the taxo-
Presently, the surviving boids in South nomic uncertainty that prevails within the
America and Africa are not closely related.
Colubridae (sensu lato). Most attempts to
The Neotropical tribe Boini may be de-
clarify the systematica have resulted in 1)
scended in situ from archaic South American
recognition of small groups that can be sep-
boids, but the African Erycinae and Pythonini
arated from the bulk of the genera, or 2 ) par-
probably came from elsewhere. Hoffstetter
tition of larger groups that are highly contro-
and Rage (1972) believed that the Erycinae,
versial(Dunn, 1928; Bogert, 1940; Bourgeois,
which may have originated in North America
from a South American boine stock, was pres- 1968; Underwood, 1967; Dowling, 1975; Smith
ent in North America in the Paleocene or et al., 1977). Dowling's (1975) classification
earlier and in Europe in the Eocene. Another
is beautifully simple, resulting in geograph- lanticOcean. The reverse migration has never
ically discrete groups, namely New World been advocated, except in the beginning of
Xenodontinae ( following Dunn's scheme ) and the Atlantic era, when the sea was very nar-
Old World Lycodontinae. This concept is in row. Now the ocean currents are favorable
agreement with the probable belated birth for westward rafting in the tropics. If this
and radiation of the family as suggested by situation prevailed for a long time, it may
the fossil record. Thus, parallel radiations in- explain why the Neotropical fauna is so obvi-
to terrestrial, aquaticand fossorial groups took ously richer than the Ethiopian fauna. South
place in South America and Africa. America might have received a sizable fau-
On the other hand, Underwood's 1967 ) nistic contribution from Africa without giving
(
revolutionary classification, which was severe- anything in exchange, at least for the last 50
million years.
ly criticized by others (namely, Dowling,
1975; Hoffstetter, 196S; Smith et al., 1977), Savitzky (1978) provided evidence that
implied trans-Atlantic relationships. The simi- the micrurines are a derivative of Neotropical
larities between the fossorial African Cala- rear-fanged colubrids, such as Elapomorplms,
melaps and the South American Atractus, the rather than relatives of the Old World cobras.
aquatic Limnophis (or Hydraethiops) with This removes a zoogeographic problem. Thus,
Helicops, or the terrestrial Lycophidion and Africa contributed no venomous snakes to
Oxyrhopus simply may be the result of con- South America, for the Crotalinae are absent
in Africa. The Viperidae appears in the fossil
vergence. Of course, this is the most likely
hypothesis, but rafting should not be dis- record in the lower Miocene in northern con-
missed off-handedly. tinents. Its general range suggests an open
Three adaptive types of arboreal colubrids radiation without insular or peninsular traps.
Pre-Atlantic Gondwanan
— groups A.
Fie. 3:1. distribu-
tions present South America and Africa
in
before the Last pre-Atlantic
separation. B. faunistic
exchanges — groups migrated from one continent
that
to the other just before the separation or perhaps
soon afterwards. C. Post-Atlantic distributions and
later one-way dispersals.
A. Distribuciones gondwanense preatldnticas—
grupos prescntes en Sudamerica y Africa largo tiempo
antes de la division. B. Ultimo intercambio faunistico
preatldntico —grupos que migraron de un continente
al otro, justo antes dc la division o, tal vez, pronto
despues de clla. C. Distribuciones postatldnticas y
despues dispersion "waif" unidireccional.
MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
graficas de Matthew (1915) y Darlington Tarentola se hizo a fines del Cenozoico. Por
(1957). Los grupos dominantes son complet- consiguiente se puede suponer que tales mi-
amente distintos aunque en general ecologica- graciones ocurrieron durante todo el Ceno-
1979 LAURENT: AFRICA AND SOUTH AMERICA 67
zoico, con frecuencia decreciente, por supues- Bonaparte, J. F. 1967. New vertebrate evidence for
to, a medida que los continentes se alejaban.
southern transatlantic connection during the
a
lower and the middle Triassic. Palaeontol. 10:
La direction de los corrientes favorece clara-
554-63.
mente las travesias de Este a Oeste, de ma- Boss, J., Pasteur, G. 1977. Solution histologique a
nera que America del Sur ceso temprano de un probleme de taxinomie herpetologique interes-
sant les rapports paleobiologiques de l'Amerique
enriquecer la fauna africana, mientras que al
de Sud et de l'Afrique. C. R. Acad. Sci. Paris,
contrario Africa mando probablemente emi-
D
Ser. 284:2547-50.
sarios bastantenumerosos a Sudamerica, no
Boulenger, G. A. 1882. Catalogue of the Batrachia
solamente salamanquesas, sino tambien escin- Salientia S. Ecaudata in the collection of the
cidos del genero Mabuija y, tal vez varios gru- British Museum, London, 503 p.
Bourgeois, M. 1968. Contribution a la morphologie
pos de culebras arboricolas. Es interesante
comparee du crane des ophidiens de l'Afrique
notar que los anfibios no tuvieron exito en Centrale. Publ. Univ. Off. Congo Lubumbashi
este tipo de aventura, porque, en general no 18:1-293.
aguantan bien el agua salada y, tal vez la Bhoin, F. de, Buffetaut, E., Koeniguer, J. C, Race,
J. C, Russell, D., Taquet, P., Vercnaud-Graz-
competencia de los Anfibios sudamericanos no
zini, C, Wenz, S. 1975. La faune de vertebres
permitia lo que se llamo el acceso ecologico continentaux du gisement d'ln Beceten (Senonien
en su continente. du Niger). C. R. Acad. Sci. Paris, Ser. D 279:
469-472.
Brundin, L. 1966. Transantarctic relationships and
their significance. K. Sven. Vetenskapsakad.
LITERATURE CITED Handl. (4)ll(l):l-472.
Buffetaut, E. 1976. Sur la repartition geographi-
Ahl, E. 1926. Anura: Aglossa, Xenopodidae, pp. que hors d'Afrique des Dyrosauridae, Crocodiliens
141-142 in Kaiser, B. (ed. ). Die Diamanten- Mesosuchiens du Cretace terminal et du Paleo-
wiiste Siidwest Afrikas 2. D. Reimer, Berlin. gene. C. R. Acad. Sci. Paris, Ser. D283:487-90.
Amiet, J. 1976.
L. Observations anatomiques et Carroll, R. L., Currie, P. J. 1975. Microsaurs as
possible apodan ancestors. J. Zool. 57:229-247.
biologiques sur le genre Werneria Poche, 1903
(Amphibiens, Anoures, Fam. Bufonidae) Rev. .
Carvalho, A. L. de. 1954. A preliminary synopsis
Zool. Africaine 90:33-45.
of the genera of American microhylid frogs.
Occas. Pap. Mus. Zool. Univ. Mich. (555): 1-19.
Andersson, L. C. 1903. Neue Batrachier aus Kam-
arun von den Herren Dr. Y. Sjostedt und Dr. S. Casamiquela, R. 1961. Un pipoideo fosil de Pata-
gonia. Rev. Mus. La Plata, Secc. Paleontol. 4:71-
Junger gesammelt. Verh. Zool. Bot. Ges. Wien
123.
53:141-145.
Cei, J. M. 1972. Bufo
of South America, pp. 82-90
Beurlen, K. 1961. Die palaogeographische Entwick- in Blair, W. F.
Evolution in the genus
(ed.).
Iung des siidatlantischen Ozeans. Nova Acta Leo-
Bufo. Univ. Texas Press, Austin, 459 p.
poldina 24:1-36.
Cei, J. M. 1977. Serological relationships of the
Blair, W. F. 1972a. Evidence from hybridization,
Patagonian toad Bufo variegatus (Giinther). Bull.
pp. 196-232 in Blair, W. F. (ed.). Evolution in Serol. Mus. 52:2.
the genus Bufo. Univ. Texas Press, Austin, 459 p.
Colbert, E. H. 1973. Continental drift and the
Blair, W. F. 1972b. Summary, pp. 329-343 in distribution of fossil
reptiles, pp. 395—412 in
Blair, W. F. (ed.). Evolution in the genus Bufo. Tarling, D. H., Runcorn, S. K. (eds.). Implica-
Ibid. tions of continental drift in the earth sciences,
Blommers-Schlosser, R. M. A. 1975. Observations Vol. Academic Press, New York, 622 p.
1.
on the larval development of some Malagasy Creer, K. M. 1973. On the arrangement of the
frogs, with notes on their ecology and biology landmasses and the configuration of the geomag-
(Anura: Dyscophinae, Scaphiophryninae and Co- netic field during the phanerozoic, pp. 47-76 in
phylinae). Beaufortia 24:7-25. Tarling, D. H., Runcorn, S. K. (eds.). Implica-
Bogart, J. Nelson-, C. E. 1976. Evolutionary
P., tions of continental drift in the earth sciences.
implications from karyotypic analysis of frogs of Ibid.
the families Microhylidae and Rhinophrynidae. Croizat, L. 1964. Space, time, form: The biological
Herpetologica 32:199-208. synthesis. Published by author, Caracas, 881 p.
Bogart, P., Pyburn, W. F., Nelson, C. E.
J. 1976. Croizat, L., Nelson, G., Rosen, D. E. 1974. Cen-
The karyotype of Otophrync robusta (Anura: ters of origin and related concepts. Syst. Zool.
Microhylidae). Ibid. 32:208-210. 23:265-287.
Bocert, C. M. 1940. Herpetological results of the Darlincton, P. J., Jr. 1957. Zoogeography: the
Vernay Angola Expedition, with notes on African geographical distribution of animals. John Wiley
reptiles in other collections. Part I. Snakes, in- and Sons, New York, 675 p.
cluding an arrangement of African Colubridae. Dietz, R. S., Holden, J. C. 1970. The break-up of
Bull. Amer. Mus. Nat. Hist. 77:1-107.
Pangaea. Sci. Amer. 223:30-41.
68 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Douglas, R. G., Moullade, M., Nairn, A. E. M. of the gavialid crocodilians. Herpetologica 28:
1973. Causes and consequences of drift in South 281-284.
Atlantic, pp. 517-537 in Tarlinc, D. H., Run- Heirtzler, J. R. 1968. Sea-floor spreading. Sci.
the upper Cretaceous of Brasil. Science 180: hension et histoire de la sous-famille. Ann. Paleon-
748-751. tol. Vertebr. 58:81-124.
Estes, R., Reig, O. 1973. The early fossil record of Huene, F. von. 1956. Palaontologie und Phylogenie
frogs: A review of the evidence, pp. 11-63 in der niederen Tetrapoden. Fischer, Jena, 761 p.
Vial, J. L. (ed.). Evolutionary biology of the Hurley, P. M. 1968. The confirmation of continen-
anurans: Contemporary research on major prob- tal drift. Amer. 218:52-64.
Sci.
lems. Univ. Missouri Press, Columbia, 470 p. Inger, R. F. 1967. The development of a phylogeny
Estes, R., Wake, M. 1972. The first fossil record of of frogs. Evolution 21:369-383.
caecilian amphibians. Nature 239:228-231. Jeannel, R. 1942. La genese des faunes terrestres.
Francheteau, 1973. Plate tectonic model of the
J. Presses Univ. France, Paris, 514 p.
opening of the Atlantic ocean south of the Azores, Kluge, A. G. 1967. Higher taxonomic categories of
pp. 197-202 in Tarlinc, D. H., Runcorn, S. K. gekkonid lizards and their evolution. Bull. Amer.
(eds.). Implications of continental drift in the Mus. Nat. Hist. 135:1-59.
earth sciences. Vol. 1. Academic Press, New Kluge, A. G. 1969. The evolution and geographical
York, 622 p. origin of the New World Hemidactylus nxabouia-
Fryer, O, Iles, T. D. 1972. The cichlid fishes of brooki complex (Gekkonidae, Sauria). Misc.
the great lakes of Africa. T. F. H. Publ., Neptune Publ. Mus. Zool. Univ. Michigan (138): 1-78.
City, N.J., 641 p. Kluge, A. G., Farris, S. J. 1969. Quantitative phy-
GafFx\ey, E. S. 1975. A phytogeny and classifica- letics and the evolution of the anurans. Syst. Zool.
tion of the
higher categories of turtles. Bull. 18:1-32.
Amer. Mus. Nat. Hist. 155:391-436. Khommelbein, K. 1971. Non-marine Cretaceous os-
Gorman, G. C. 1970. Chromosomes and the sys- tracodes and their importance for the hypothesis
tematics of the family Teiidae (Sauria, Reptilia). of Gondwanaland. Proc. 2nd. IUGS Gondwana
Copeia 1970(2):230-245. Symp. South Africa, pp. 617-619.
Griffiths, I. 1963. The phylogeny of the Salientia. Laurent, R. F. 1972. La distribution des amphib-
Biol. Rev. 38:241-292. ious et les translations continentales. 17th. Inter.
Griffiths, I., Carvalho, A. L. de. 1965. On the Cong. Zool., Monaco, 1:1-16.
validity of employing larval characters as major R. F. 1973. A
Laurent, parallel survey of equatorial
phyletic indices in Amphibia Salientia. Rev. Bra- amphibians and reptiles in Africa and South
sileira Biol. 25:113-121.
America, pp. 259-266 in Mecgers, B. J., Ayensu,
Hecht, M. 1963. A reevaluation of the early history E. S., Duckworth, W. D. (eds.). Tropical forest
of the frogs. Part II. Syst. Zool. 12:20-35. ecosystems in Africa and South America. Smith-
Hecht, M., Malone, B. 1972. On the early history sonian Inst. Press, Washington, D.C., 350 p.
1979 LAURENT: AFRICA AND SOUTH AMERICA 69
Laurent, R. F. 1975. La distribution des amphib- Nelson, C. E., Cuellar, H. S. 1968. Anatomical
iens et les translations continentales. Mem. Mus. comparison of tadpoles of the genera Hypopachus
Natl. Hist. Nat. Paris Ser. A Zool. 88:176-191. and Gastrophryne ( Microhylidae). Copeia 1968
Laurent, R. F. In press. Systematique et distribu- (2):423-424.
tion geographique (des Amphibiens ) in Grasse, .
Nelson, G. 1973. Comments on Leon Croizat's bio-
P. P. (ed.). Traite de Zoologie 14, fasc. 2 Am- geography. Syst. Zool. 22:312-320.
phibiens. Nevo, E. 1968. Pipid frogs from the Early Cretace-
Laughton, A. Sclater, J. G., McKenzie, D. P.
S., ous of Israel and pipid evolution. Bull. Mus.
1973. The and evolution of the Indian
structure Comp. Zool. Harvard Univ. 136:255-318.
ocean, pp. 203-212 in Tarling, D. H., Runcorn, Noble, G. K. 1931. The biology of the Amphibia.
S. K. (eds.). Implications of continental drift in McGraw-Hill Book Co., New York, 577 p.
the earth sciences. Vol. 1. Academic Press, New Nopsca, F. 1928. The genera of Reptiles. Paleobiol.
York, 622 p. 1:163-188.
Lavocat, R. 1974. What is an hystricomorph ro- Nussbaum, R. A. 1977. Rhinatremidae: A new fam-
dent? Symp. Zool. Soc. London 34:7-20. ily of Caecilians (Amphibia: Gymnophiona ) .
Liem, K. F., Marx, H., Rabb, G. B. 1971. The Occas. Pap. Mus. Zool. Univ. Mich. (682): 1-30.
viperid snake Azemiops: Its comparative cephalic Orton, G. L. 1957. The bearing of larval evolution
anatomy and phylogenetic position in relation to on some problems in frog classification. Syst. Zool.
Viperinae and Crotalinae. Fieldiana Zool. 59: 6:79-86.
65-126. Patterson, B., Pascual, R. 1968. The fossil mam-
Liem, S. S. 1970. The morphology, systematics and mal fauna of South America. Q. Rev. Biol. 43:
evolution of the Old World tree frogs (Rhaco- 409-451.
phoridae and Hyperoliidae ) Fieldiana, Zool. 57:
.
Rabb, G. B., Marx, H. 1973. Major ecological and
1-145. geographic patterns in the evolution of colubroid
Lynch, J. D. 1973. The transition from archaic to snakes. Evolution 27:69-83.
advanced frogs, pp. 133-182 in Vial, J. L. (ed.). Rage, J.-C. 1974. Les Serpents des phosphorites du
Evolutionary biology of the anurans: Contempo- Quercy. Palaeovert. 6:274-303.
rary research on major problems. Univ. Missouri Rage, J.-C. 1975. Un serpent du Paleocene du
Press, Columbia, 470 p. Niger. Etude preliminaire sur l'origine des cae-
Martin, H. 1968. A
review of the evidence
critical nophidiens (Reptilia, Serpentes). C. R. Acad.
for a former direct connection of South America Sci. Paris, Ser. D 281:515-18.
with Africa, pp. 25-53 in Fittkau, E. J., Illtes, Rage, J.-C. 1976. Paleontologie, phylogenie et paleo-
J., Klinge, H., Schwabe, G. H., Sioli, H. (eds.). biogeographie des serpents. PhD Dissert. Univ.
Biogeography and ecology in South America. Paris.
W. Junk, The Hague, 946 p. Rage, J.-C. 1977. L'origine des Colubroi'des et des
Matthew, W. D. 1915. Climate and evolution. Acrochordoides (Reptilia, Serpentes). C. R.
Ann. New York Acad. Sci. 24:171-318. Acad. Sci. Paris, Ser D 286:595-597.
McDiarmid, R. W. 1971. Comparative morphology Reyment, R. A. 1958. Factors in the distribution of
and evolution of frogs of the Neotropical genera fossilcephalopods. Stockholm Contrib. Geol. 1:
Atelopus, Dendrophryniscus, Melanophnjniscus 91-184.
and Oreophrynella. Nat. Hist. Mus. Los Angeles Reyment, R. A. 1970. Vertically inbedded cepha-
Cty. Sci. Bull. (12;: 1-66. lopods shells. Some factors in the distribution of
McKenna, M. C. 1973. Sweepstakes, filters, corri- fossil cephalopods. Palaeogeogr. Palaeoclimatol.
dors, Noah's arks and beached Viking funeral Palaeoecol. 7:103-111.
ships in palaeogeography, pp. 295-308 in Tarl- Reyment, R. A. 1975. Paleontologie evolutive et
inc, D. H., Runcorn, S. K. (eds.). Implications nouvelle tectonique. Mem. Mus. Natl. Hist. Nat.
of continental drift in the earth sciences. Vol. 1. Paris Ser. A Zool. 88:8-18.
Academic Press, New York, 622 p. Reyment, R. A., Tait, E. A. 1972. Biostratigraph-
McKenzie, D. P., Sclater, J. G. 1973. The evolu- ical dating of the early history of the South
tion of the Indian Ocean. Sci. Amer. 228:62-72. Atlantic Ocean. Phil. Trans. Roy. Soc. London,
Medem, F. 1968. Desarrollo de la herpetologia en B. Biol. Sci. 264:55-95.
Colombia. Rev. Acad. Colombiana Cien. Exactas Robinson, 1967. The evolution of the Lacertilia.
P.
Fis. Nat. 13:149-200. Colloq. Int. C. N. R. S., Paris 163:395-407.
Mook, C. C. 1934. The evolution and classification Romer, A. S. 1956. Osteology of the reptiles. Univ.
of the Crocodilia. J. Geol. 42:295-304. Chicago 772
Press, Chicago, p.
Morescalchi, A. 1973. Amphibia, pp. 233-348 in Romer, A. S. 1966.
Vertebrate Paleontology. 3rd
Chiarelli, A. B., Cat-anna, E. (eds.). Cyto- ed. Univ. Chicago Press, Chicago, 468 p.
taxonomy and vertebrate evolution. Academic Savage, J. M. 1973. The geographic distribution of
Press, NewYork, 783 p. frogs: Patterns and predictions, pp. 351-445 in
Nelson, C. E. 1966. The evolution of frogs of the Vial, J. L. (ed.). Evolutionary biology of the
family xMicrohylidae. PhD Dissert. Univ. Texas anurans: Contemporary research on major prob-
Diss. Abst. 27, 3719 B. lems. Univ. Missouri Press, 470 p.
MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
70
Amer. Zool. 5:309-318. when adequate data are lacking. It might be said
genocenters, as Sokal and Crovello ( 1970 )
Trueb, L. 1971. Phylogenetic relationships of cer- for did
tain Neotropical toads with the description of a for the biological species concept, that the concept is
new genus ( Anura:Bufonidae). Nat. Hist. Mus. not operational. However, this does not preclude the
Los Angeles Cty. Contr. Sci. (216):l-40. existence of centers of origin, even if we are unable
Underwood, G. 1967. A contribution to the classi- to discover their location, exactly as the non-opera-
fication of snakes. Brit. Mus. (Nat. Hist.), Lon- tionality of the biological species concept
does not
don, 179 p. eliminate the fact that the cessation of gene exchange
Underwood, G. 1976. A systematic analysis of boid between two populations is such a momentous event
snakes, pp. 151-175 in Bellairs, A., Cox, C. B. in evolution that inconceivable to ignore
it is it, even
(eds.). Morphology and Biology of Reptiles. if we are unable to pinpoint its occurrence.
Linnean Soc. London Symp. Series 3, 290 p.
Vanzolini, P. E. 1968. Geography of the South
American Gekkonidae (Sauria). Arq. Zool. (Sao Appendix 3:2. —The Indian raft.
Paulo) 17:85-111.
It is now generally believed that India drifted
Vergnaud-Grazzini, C. 1966. Les amphibiens du
Miocene de Beni-Mellal. Notes Serv. Geol. Maroc away from Antarctica, Madagascar and Africa some-
27:43-69. time at the end of the Mesozoic and travelled north-
1972. wards through the India Ocean to collide with Laura-
Vergnaud-Grazzini, C. Hoffstetter, R.
sia in the Miocene. Tire dating of the separation is
Presence de Palaeobatrachidae (Anura) dans des
gisements tertiaires francais. Caracterisation, dis- still
—
doubtful about 100 m.y.b.p. from Antarctica
tribution et affinites de la famille. Palaeovert. and maybe the Paleocene (60 m.y.b.p) from Mada-
5:157-177. gascar. Little attention has been given to the impact
Wake, D. 1966. Comparative osteology and evolu- of the Indian fauna on the evolutionary zoogeography
tion of the lungless salamanders, family Pletho- in the Tertiary. This is an unfortunate omission, not
dontidae. Mem. South. Calif. Acad. Sci. (4):1-
justified by lack of evidence. The Eocene lndo-
111.
batrachus seems to belong to the Myobatrachidae, and
Wood, A. E. Porcupines, paleogeography and
1950.
the primitive snake family Uropeltidae survives in
parallelism. Evolution 4:87-98.
southern India and Sri Lanka. The rationale for my
Wood, R., Patterson, B. 1973. A fossil trionychid
as follows:
turtle from South America. Breviora (415): 1-10. hypothesis is
1979 LAURENT: AFRICA AND SOUTH AMERICA 71
1. India broke from Madagascar and Africa dur- finally along the Nicobar and the Andaman
ing the late Cretaceous or early Tertiary. islands.
2. As a large island, it rafted away from Mada- 7. The fate of much of the Indian fauna must
gascar northwards along the Mascarene Ridge, have been extinction.
leaving behind the Seychelles, Amirante and 8. Some elements escaped early to Madagascar
finally the Laccadive and Maldive islands and Africa and proved successful in their ex-
( Laughton et al., 1973; McKenzie and Sclater, pansion (e.g., Microhylidae).
1973). 9. Other elements escaped later in northern,
3. The Indian fauna evolved in isolation for about northeastern and eastern directions (e.g., other
50 million years; evolution was enhanced by Microhylidae and perhaps Agamidae, Varani-
the changing climates. dae, Pythonini, Elapidae).
4. Some faunistic exchanges remained with Mada-
gascar and Africa through the intervening Four groups of reptiles (Agamidae, Varanidae,
islands, like the Seychelles, and possibly others Pythonini, Elapidae) have patterns of distribution
that have since disappeared. that can be explained by an Indian differentiation.
5. These small islands provided opportunities for Each has a strong Indo-Malaysian component, another
genetic drift and quantum and tachytelic evo- strong Australasian component, and a weak African
lution favoring major adaptive shifts (e.g., component, as if there had been a late invasion of
Microhylidae, Savage, 1973). Africa from Asia.
6. Perhaps other exchanges took place in front of Subsequent to writing this account, I have been
and/or on the eastern side when the Indian informed by R. Hoffstetter that recent data show that
Noah's Ark (McKenna, 1973) drew near Lau- the Indian Subcontinent collided with Laurasia not
rasia, gliding along the Ninety-east Ridge and later than the Eocene.
4. Herpetofaunal Relationships of South America
With Australia
Michael J. Tyler
Department of Zoology
University of Adelaide
Adelaide, South Australia 5001
Australia
In reviewing the extent of South American modern families could just as well have orig-
herpetofaunal relationships with Australia, inated there as on the adjacent landmasses.
while simultaneously considering South Amer- Thus any realistic concept of intercontinental
ican-African relationships (Laurent, this vol- exchange avoids reference to "journeys" along
ume), it is helpful to recognize that vast dif- "routes," and only visualizes the expansion
ferences have existed in the opportunity for and retraction of populations. Cartoons in
faunal exchange. South America and Africa an otherwise serious paper by Rich ( 1975 ) on
may be regarded as lovers who experienced the origins of the Australian nonpasserine avi-
and exploited a large zone of contact and had fauna, illustrate the errors to which some in-
considerable opportunity for interchange and vestigational philosophies may have suc-
exchange across it. In contrast, the South cumbed.
American-Australian relationship suffered The study of intercontinental herpeto-
from being in the form of an arranged engage- faunal relationships faces problems of varia-
ment of longer duration. The couple never so tion of systematic interpretation of taxa, and
much as touched one another at any time. these materially influence the degree of faunal
The only contact was via a related intermedi- similarity. For example, if the numerically
ary named Aunt Arctica, whose presence be- dominant Australian terrestrial and arboreal
tween them effectively prevented a compar- frogs are regarded as members of the Lepto-
able degree of intimacy, and who is now dactylidae and Hylidae, respectively, all anu-
outwardly cool and distinctly secretive about ran families found in Australia are shared
revealing what took place between them. with South America. Superficially at least,
The benefit of employing such an analogy the anuran relationship appears likely to
lies inemphasizing the fact that Australia and prove a close one. However, if the names
South America have always been physically Myobatrachidae and Pelodryadidae are em-
separated. This separation always has been ployed for these same groups, it is difficult to
extensive, because the intervening Antarctica avoid a bias towards a quite different inter-
is a vast continent with a surface area of
pretation. In fact it would appear that, for
1,165,500,000 km 2 , comparable in size to the purposes of intercontinental comparisons,
South America north of the Tropic of Capri- there is a mystique surrounding a family
corn, and considerably greater than Australia name that does not extend to other nomen-
(7,700,000 km ). A North to South traverse
2
clature.
of Antarctica involves a distance of approxi- Over the past few years there have been
mately 4,000 km. substantial contributions to the study of plate
When Antarctica was an integral com- tectonics, continental drift, palaeoclimate and
ponent of Gondwanaland, the herpetofaunal the past flora and fauna of Australia. Many of
elements shared at any one time by South these papers are highly relevant to the inter-
America and Australia also would have oc- pretation of evolutionary opportunities and
curred on Antarctica. Certainly a topography, the nature of the diversification of the herpe-
climate, and vegetation equable to the main- tofauna. Here I have attempted to bring to-
tenance of reptiles and amphibians had to gether the most recent literature as a general
exist on Antarctica, and at least some of the background before examining the evidence to
73
74 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
lAl»0t
the onset of the northwards drift of Australia continent whose palaeoclimatic conditions
away from east Antarctica, ranging from a have been misinterpreted. Axelrod (1960)
low of 43 m.y.b.p. (Jardine and McKenzie, envisaged deterioration throughout the Ceno-
1972) to a high of 180-100 m.y.b.p. (Fooden, zoic leading to arid to semiarid conditions by
1972; Savage, 1973). However, it is now the Miocene. There is now evidence that cen-
placed at 55-52 m.y.b.p., with most authors tral Australia bore large, permanent lakes in
favoring 53 m.y.b.p. ( McGowran, 1973; Sclat- the Miocene. Lungfishes, teleosts, turtles,
er et al., 1974; Coleman and Packham, 1976; crocodiles, lizards, and frogs shared the site
Veevers and McElhinny, 1976). McGowran's with a vast diversity of marsupials and birds.
studies of the Antarctic-Australian suture led The surrounding vegetation was dense, rang-
him to suggest that, despite the onset of drift, ing from rainforests to extensive areas of
there was no substantial barrier to the pas- grassland. Gallery forests extended along the
sage of land animals prior to the early Eocene watercourses, and the presence of Nothofagus
(49 m.y.b.p.). and Podocarpus are interpreted to be evi-
Climatic, floral and faunal changes. Be- — dence of high rainfall. There was southern
cause Nothofagus forests now occur in some communication with the sea at some stage
temperate areas, such as southeast Australia (Callen and Tedford, 1976). The records of
including Tasmania, and in New Zealand, it crocodiles and turtles at former freshwater
has been possible to deduce that Nothofagus sites in central Australia are particularly nu-
is associated classically with temperate cli- merous (see also Newsome and Rochow,
matic conditions (Axelrod, 1975). Thus, with 1964). However, these represent the most
evidence of Nothofagus occurring in the Eo- conspicuous and most readily recognized rep-
cene at several localities in southern Australia, tile fossils; the search for smaller material has
the inference might be drawn that, at the only just begun.
time of the separation of Australia from East At some stage, the area between central
Antarctica, the southern Australian fauna was Australia and the north coast also was moist.
probably cool-temperate. Certainly this as- This is demonstrated by the cabbage palms,
sumption would be valid for N. fusca and N. Livistona mariae, now restricted to a colony
menziesi, which now exist in Australia, New of 3,000 at Finke River in central Australia.
Zealand, Chile and Argentina. However, the Their nearest relatives lie 1,000 km away in the
important species is N. brassi, which now northwest of the continent ( Latz, 1975 ) Pa- .
exists in New Guinea and New Caledonia and laeontological and geomorphological evidence
clearly is a subtropical species. Formerly, its demonstrate that central Australia provided
distribution was far more extensive, being numerous niches for mesic animals until the
known in Australia and New Zealand from end of the Pleistocene Wopfner and Twidale,
(
the Early Cretaceous to the mid-Pliocene, in 1967; Mabbut, 1967; Twidale, 1972). Cer-
West Antarctica from the early Palaeocene to tainly deserts have featured in Australia for a
the mid-Eocene, and from Chile and Argen- long period and have had an essential role in
tina from the Early Cretaceous to the late lizard speciation (Pianka, 1972). It has been
Oligocene (Schlinger, 1974). suggested that in the Quaternary much of the
Further evidence of the southern Austral- now moist extreme southwest of the continent
ian climate being subtropical has been estab- was arid (Glassford and Killigrew, 1976).
lished by Lange (1976) from his study of However the extent of Australia affected by
microfossil epiphyllous germlings, and by aridity appears to have been exaggerated.
Christophel and Blackburn (1978) from their
The minimal morphologicaldifferentiation
of the central Australian hylid
assessment of the Eocene South Australian frog fauna is
Maslin Bay flora. The geomorphological evi- wholly consistent with aridity being a late
Pleistocene feature. Thus the species Litoria
dence of widespread subtropical conditions
caerulea and L. rubella are relicts of a much
are summarized by Bowler (1976). richer fauna, surviving because of tolerance
Central Australia is another portion of the of adults or larvae to high
temperature or
76 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
possibly for some other reason. But for the ea, and of the area to the east and west is
contrary evidence of Glassford and Killigrew extremely complex. Similarly only the broadest
(
1976 ) it is possible that Australia has not
,
of principles of the nature of the collision of
been any more arid than it is today and, to the plates has yet been established. The fol-
judge from the nature and abundance of lowing contributions provide a brief spectrum
vegetation cover now stabilizing sand dunes of opinions and are a source of many other
in some areas, a trend towards climatic ame- references Thompson ( 1967 ) Falvey and
:
,
lioration has already begun. However, the Taylor (1974), Coleman (1975), Denham
Pleistocene record is of minimal relevance to (1975), Mackenzie (1975), Taylor
(1975),
this review. Tilbury (1975) and Coleman and Packham
(1976). Only recently attempts have been
made to reconstruct the nature of the plate
THE NATURE OF THE collision. Mackenzie (1975) suggested that
AUSTRALIAN-ORIENTAL COLLISION mountain ranges now on the north coast of
New Guinea represent an arc of islands that
Modern New Guinea is composed of three
persisted through to the Miocene, and became
distinct and roughly longitudinally arranged accreted during the collision of the plate
portions. The southern portion and the inter- margins (Fig. 4:2). Coleman and Packham
vening Arafura Sea originally represented the (1976:204) favored this concept: "For the
leading edge of the Australian continental moment, we accept the likelihood that north
plate. The central cordillera is predominantly coastalNew Guinea is a piece of crust, prob-
a much younger feature; uplift commenced in an arc segment, in collision with Aus-
ably
the Miocene. Finally there is a row of iso- tralia-New Guinea." New Britain, to the east
lated mountain ranges on the north coast, each of New
Guinea, therefore represents an island
of which is composed of older volcanic rocks. of the same arc, but which did not come di-
The history of the evolution of New Guin- rectly into contact with New Guinea.
^---^ TORRICELLI
1979 TYLER: SOUTH AMERICA AND AUSTRALIA 77
nan source, reinforced following the drifting ognize the existence of the additional pre-
collision Oriental unit.
of Africa, followed by a later North American
infusion. For Australia there was similarly
an initial Gondwanan source followed by an ORIENTAL ELEMENTS
Oriental one.
A basic problem is to determine for each
Ranidae
continent which taxa are of Gondwanan an-
cestry. The recent literature
includes several The overall distribution of the Ranidae in
assessments (Keast, 1971, 1973; Cracraft, the Australian Region is wholly consistent
1973, 1974, 1973). For Aus-
1975; Savage, with the concept of entry from the adjacent
between the Gondwanan
tralia the distinction Oriental Region to the west. What is less
element and the more recent Oriental one of two
satisfactorily explained is the existence
should be distinguishable on the basis of endemic species of Platymantis in Fiji far to
morphological divergence and by the nature the east, whereas none occurs in Australia.
of geographic distribution. This is because the In terms of diversity and abundance of ranid
Australian-Oriental collision occurred in the species, New Guinea is equally anomalous,
to
mid-Miocene, so that animals in Australia of the extent that this component of its fauna is
Oriental origin should have distinct affinities depauperate when compared with those of
and comparable geographic distributions with smaller islands to the west and to the east.
animals in the Oriental Region. Conversely, Thus there are 20 ranids in the Philippine
such relationships should be lacking among Islands, 10 on New Guinea, but 24 on the
the Gondwanan component and there should Solomon Islands. Those anomalies are high-
be minimal geographic distribution outside lighted by the study of the genus Platymantis,
the Australian continent. Thus I propose to including species previously referred to Cor-
establish the constituents of the Gondwanan nufer (Fig. 4:3). Viewing such a distribution
element of the Australian fauna primarily by pattern has led to the assumption that the
a process of identifying, and so eliminating, distribution of Platymantis in New Guinea is
the Oriental element. relictual (Zweifel, 1969).
On an historic and biogeographic basis In support of a concept that Platymantis
the Oriental element of the Australian herpe- was formerly far more widely distributed in
tofauna will fit into one of two categories, as New Guinea than it is today, there is evidence
follows: 1) Animals that occurred within the of close phylogenetic relationships existing
northern chain prior to the mid-Miocene col- between species that are geographically iso-
lision with Australia. All of these would have lated from one another. An example is P.
entered that area from the west. They could batantae of Batanta adjacent to the Vogelkop
range from the Philippine Islands to Fiji. To Peninsula of Irian Jaya (West New Guinea),
be recognizable as pre-collision components, which Zweifel ( 1969 ) considered most closely
they should be more abundant on islands east related to P. giUiardi and P. mimicus of New
of New Guinea than in New Guinea itself. Britain about 2000 km distant.
Platymantis
2) Animals that have dispersed from west to punctata of northern New
Guinea and P.
east following the accretion of the chain myersi of Bougainville, Solomon Islands, with
within northern New Guinea. Such animals which it has affinities, are separated by a simi-
are likely to show a progressive west to east lar distance.Inger (1954:355) evidently drew
reduction in diversity and to be poorly repre- comparable conclusions when he suggested
sented on the islands east of New Guinea. that the closest relations of P. meyeri of the
It is worth contemplating that some of the Philippine Islands ". are not with other
. .
78 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
B
O
R
N
E
O
1979 TYLER: SOUTH AMERICA AND AUSTRALIA 79
Islands (Fig. 4:4). The direction of second- Indonesia and the Philippines, New Guinea
demonstrable and northern Australia. The Australopapuan
ary radiations is a reflection of
affinities of the extant species. unit is the most prolific, with 13 genera and
pliylogenetic
New Britain evidently was a major source of 102 species ( Zweifel, 1972; Menzies and Tyler,
colonizing species, leading to the occurrence 1977), compared with 16 genera but only 32
of P. batantae on Batanta and P. punctata on species in South America (Walker, 1973;
Walker and Duellman, 1974; Nelson, 1975).
Waigeo Island off the Vogelkop Peninsula,
P. cheesmanac on the Cyclops Mountains, Parker (1934) was the last contributor to
and possibly P. papuensis on the Finisterre treat this family in its entirety. Subsequent
Mountains (subsequently extending through- contributions have tended to examine single
out northern New Guinea). geographic components, and the overall phy-
When the Australian continental mass col- logenetic relationships of the diverse genera
lided with the Oriental island chain, it there- remain obscure. Cracraft ( 1973 ) and Bogart
fore acquired four species (or stocks) of and Nelson ( 1976 ) outlined the principal
Platymantis (Fig. 4:4). Three of them (ba- issues, of which contention has centered on
tantae, cheesmanae and punctata) have re- interpretation of the origin of the family,
and
mained almost entirely within the original its relationship to the Ranidae.
particularly
confines of the islands on the north coast and In reality, the wide variety of opinions that
have not spread appreciably in New Guinea. have been offered on the origin of this family
The fourth (papuensis) has extended as far reflects the extreme morphological complexity
as the south coast in the extreme west of of the constituent members and the absence
New Guinea. This species ranges to New of a satisfactory, modern synthesis. This is
Britain and the Solomon Islands, but it is demonstrated particularly well by the varying
known to inhabit the intertidal zone and is subfamilial classifications that have been pro-
well suited to dispersal by land and by sea
posed.
(Tyler, 1976a). Nevertheless, with the
time
Within the context of South American-
scale available, its dispersal in New Guinea
Australian faunal studies, the contributions of
remains modest (Fig. 4:5). Perhaps this im-
com- Savage (1973) must be considered in detail.
plies the existence of an Australopapuan
Savage's conclusions differed quite strikingly
and hence an ecological, rather than a
petitor from those of Parker (1934). Whereas the
physical, barrier to dispersal. latter recognized two subfamilies occurring
The source of the stock that gave rise to within and confined to the Australopapuan
the two endemic species on Fiji is uncertain.
area (Asterophryinae and Sphenophryninae),
The intervening and florally rich New Hebri-
Savage recognized only one, to which the
des lacks Platymantis or any other endemic
name Asterophryinae was applied. Savage
species of frogs. The Australian hylid Litoria
(1973:355) considered that the only distinc-
aurea has been introduced there recently, pos- tion between the Asterophryinae and the
sibly from New Caledonia, where it was in-
Sphenophryninae was that ". the former
. .
p
1979 TYLER: SOUTH AMERICA AND AUSTRALIA 81
Fig. 4:5. Distribution of Platymantis papuensis on New Guinea and adjacent islands. Populations on islands
east of the mainland may represent an undescribed species (R. G. Zweifel, pers. comm. ).
Distribution dc Platymantis papuensis en Nucva Guinea y islas adyacentes. Las poblaciones de las islas al
este del contincnte pudieran ser una especie no descrita (R. G. Zweifel, pers. com.).
phryne in particular. Zweifel ( 1971 ) con- musculature (Tyler, 1974a) provide addition-
cluded that, despite certain equivocal fea- al data supporting such a recognition of two
Tongue Subcircular, entirely adherent, often with Oval, half-free behind, lacking
a median furrow and posterior pouch median furrow and posterior pouch
p
H
I
L
I
: 13
N
E
S
(JJ PALAU
BATANTA/WAIGEU
NW. NEW GUINEA
18
JEW BRITAIN
NEW GUINEA 6
*T\ SOLOMON
\A ISLANDS
\ 1
FIJI
AUSTRALIA [O]
Fig. 4:6. Distribution and numbers of species of fi ana in Australia and the adjacent Oriental Region and
Pacific area.
Distribution y numcros de especies de Rana en Australia y la adyaccnte Region Oriental y areas del Pacifico.
Unquestionably, the former union provided a tion can only be appreciated when the con-
but it involves resurrecting and modifying view can be supported on historical, biogeo-
concepts refuted by Savage. graphic, and morphological grounds. Pro-
Few authors have contemplated the pos- vided with the evidence of the nature of the
sibility of microhylids being widely distrib-
collision of the Australian continental plate
uted in Gondwanaland in the Cretaceous. with the pre-existing chain of Oriental islands,
Admittedly, the concept of the entry of the an Oriental origin seems highly likely for the
family into Australia from Indonesia ante- Australopapuan stock. Thus, Cophixalus,
dated acceptance of continental drift and sea- Oreophyrne, and Sphenophryne were prob-
floor spreading. However such an entry has ably established within the chain at the time
been supported by many authors, of which of the collision. The absence of these genera
Laurent (1975) is the most recent. Savage in the Solomon Islands and islands farther
(1973) proposed the interesting hypothesis south, and the presence in New Britain of
that the Microhylidae was present in the trop- only a single species each of Oreophryne and
ical portions of each of the southern land Sphenophryne (Tyler, 1967) indicate that
masses prior to the fragmentation of Gond- microhylids passed eastwards after the colo-
wanaland. He further put forward an in- nization of the same areas by ranids. It fol-
genious account of a turbulent climatic his- lows that the ancestry of the Papuan micro-
tory for Australia, so as to account for the hylid fauna must be far less complicated than
family's present abundance in New Guinea an examination of the diverse modern genera
(13 genera, 95 species) and almost total ab- would indicate. The important criterion for
sence from Australia (2 genera, 7 species). their success appears to have been the in-
Savage's hypothesis demands considerable herent ability to colonize the New Guinean
mobility for the Australopapuan populations. montane environments that evolved during
In particular there is the need for extinction the rapid elevation immediately after the col-
of the Australian component, followed by di- lision.
part of southern Australia [therefore exclud- reserves,and of decreased demands upon the
ing geographic areas affected by the aridity use of the larval digestive apparatus.
that Savage (1973) believed to have caused
There are several Oriental microhylids
their demise].
that exhibit trends towards delayed emer-
Whereas Savage visualized the Australo- gence. Inger (1966) provided ecological
papuan microhylids as a Gondwanan element notes of species in which enlarged and un-
and thus a group whose origins involved a pigmented ova have been found. In the genus
direct ancestry to South American
frogs, I Kalophrynus the eggs of some species are pig-
subscribe to the more orthodox opinion of an mented, whereas in others such as K. pleuro-
Oriental ancestry for the Papuan stock. This stigma they are unpigmented, and the larvae
84 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
C5
Fig. 4:7. The Australian and adjacent Oriental and Pacific areas showing the distribution of the Microhyli-
dae (broken line) and the range of the most widely distributed Australopapuan genus, Oreophryne (con-
tinuous line ) .
Distribution de Microhylidae (linea entrccortada) y el rango del mas extendido de todos los generos austral-
opapua, Oreophryne (linea continua) en Australia y las areas Oriental y Pacifico adtjacentes.
North America
South America Africa
Madagascar Madagascar
Asia New Guinea
New Guinea
Australia
Greer considered the Scincinae to be the gascar as well. Eumeces has a remarkably
most primitive group within a distribution disjunct distribution, with isolates ranging
that is
predominantly Laurasian but also oc- from North Africa to India, China to Vietnam,
cupies the entire African continent and Mada- and Middle to North America. Disjunctions
1979 TYLER: SOUTH AMERICA AND AUSTRALIA 87
The carettochelyid turtles have an unusual The colubrids occur mainly in the northern
distribution pattern, but, as yet, there is no and eastern portions of the Australian con-
evidence of their former presence in South tinent. Some of the genera, such as Stegono-
America. Known from
the Eocene of North tus, range through Australia and New Guinea
America and the late Miocene to Present of into the Oriental Region. The phylogenetic
New Guinea (Glaessner, 1942), the sole liv- affinities of Stegonotus appear to be with the
ing representative
Carettochelys insculpta Oriental Dinodon (McDowell, 1972). An
was reported from rivers of northern Aus-
first Oriental route of entry for the colubrids as
tralia by Cogger (1970), and since has been a whole seems to be unquestionable.
shown by Schodde, Mason, and Wolfe (1972)
to be distributed quite widely in the Northern Acrochordidae and Uropeltidae
Territory. This species evidently has a high
tolerance to salt water. Carettochelyids rep- Two acrochordids occur in New Guinea;
resent a relict family, but it is not necessarily in Australia they are restricted to the extreme
one of any great antiquity within the region. north of the continent. These aquatic species
They probably represent the first trionychoid are either both referred to Acrochordus, or
invasion that was followed by Pelochelys bi- one to that genus and the other to Chersydrus.
broni, which did not extend beyond New Each species is distributed extensively in In-
Guinea. Both species probably entered the donesia and farther west, and they represent
Australian Region in the Miocene. A triony- a recent Oriental invasion. The uropeltid
chid has been found in the middle Pliocene genus Cylindrophis can be included within
of Venezuela (Wood and Patterson, 1973). the fauna of the Australian Region because it
reaches the Am
Islands between Australia
and New Guinea
(McDowell, 1975). In
Elapidae
other respects, it is an exclusively Oriental
Within Australia and New Guinea the genus, and it has certainly entered the Aus-
elapid fauna is exceptionally diverse. Authors tralian Region very recently.
vary in the number of species and genera that
they recognize, but Cogger (1975) recog-
Typhlopidae
nized 26 genera and 61 species in Australia.
Such numbers and diversity would seem to Typhlopids occur on almost all continents.
require a great evolutionary time span. This Typhlops has a range almost equivalent to the
seems to conflict with the existence of endem- entire family, extending throughout Asia to
ism within the Solomon Islands (Salomone- New Guinea. The Australian species now are
laps and Loveridgelaps) and even as far as referred to the genus Typhlina, which resem-
Fiji (Ogmodon) (McDowell, 1970). More- bles Typhlops in external features, but differs
over, elapids extend in a slightly disjunct arc substantially in the nature of the male geni-
talia (Guibe, 1948; Robb, 1966). Cogger
through to Africa; an Oriental origin seems
likely for many of them. Whether this applies (1975) listed 22 species of Typhlina in Aus-
to all components of the Australian tralia, and McDowell (1974) listed 11 from
elapid
fauna will have to await completion of the New Guinea and the Solomon Islands. Ty-
splendid work commenced by McDowell phlina extends as far south as Fiji. The nature
of the distribution pattern indicates an Orien-
(1967, 1970).
The fossil record currently con-
tal origin for the Australian component, but
elapid is
ered history. The distribution of Python in an the Boidae, Australia may have been colo-
almost continuous arc from Africa to Aus-
—
nized by the family twice an initial entry
tralia provides further evidence of origin out- antedating the drift of Australia from Antarc-
side Australia. There has been a fairly pro- tica, and a second entry, presumably in the
nounced successful radiation within Australia, Miocene. If so, when Python first appeared
attributed by Storr (1964) to the absence of in the north of the continent, Wonambi or its
the Felidae. McDowell (1975) recognized ancestors inhabited at least the southeastern
three groups among the Australian Pythoni- part.
nae, distinguished by the presence or absence
The only other fossil record of Australian
of labial scale pits and prehensile adaptations boids is the report by Archer and Wade
to the structure of the tail. The first of these (
1976 ) of three vertebrae of a very large
contains Liasis, and the second Python, More- species in the lower Pliocene Allingham For-
lia and Chondropython. McDowell consid- mation in north Queensland. These authors
ered the latter two genera to be only weakly do not associate it with a modern species, but
defined and maintained that a good case noted (op. cit: 385) that it is ". . . mor-
could be made for referring them all to phologically very similar to modern species of
Morelia."
Python. Such is the state of herpetological
exploration in Australia that a giant new
species of Python wasdiscovered in the GONDWANAN ELEMENTS
Northern Territory in 1975 (Gow, 1977).
However, the third group containing Aspi- As demonstrated here, the Gondvvanan
clites was defined more satisfactorily by pos- elements are predominantly anuran. The fos-
sil record is only just being assembled, and it
sessing several features not exhibited by other
genera. The Boinae is represented in New is best dealt with here rather than within the
A rich Pleistocene fauna, including 166 Tyler (1971) studied a suite of characters
frog of extant species, has
ilia been taken at associated with the superficial mandibular
two cave sites at Naracoorte in the lower musculature, and the vocal sac that it con-
southeast of South Australia (Tyler, 1977). tains, in representatives of numerous families
The fossils include Litoria ewingi, Limno- and including almost all known hylid genera.
dijnastes cf. dumerili, L. tasmaniensis, Rani- He noted that anatomical divergence occurred
della signifera, and Geocrinia cf. laevis, all of principally in association with taxonomic
which occur in that area today. Unfortunate- units recognized as genera. He demonstrated
ly, no known frog fossils of
as yet there are that the Australopapuan Hyla constituted a
an age predating the Australopapuan-Orien- single morph distinguishable from species
tal collision. However, it is noteworthy that from other parts of the world. Consequently
Limnodijnastes and Litoria were established he proposed the resurrection of Litoria Tschu-
in the Miocene at Lake Palankarinna. di to accommodate the Australopapuan spe-
cies and further considered Litoria and Nyc-
of these and other frogs, Duellman ( 1975 ) examination of the diversity of struc-
perficial
offered a number of solutions to controversial ture render the casual observer critical of
may
issues. Among his actions, he continued to Australopapuan hylid systematics. Neverthe-
include Australopapuan species in the Hy- less the magnitude of diversity need be no
lidae. indication of polyphyly. Tyler ( 1971 ) pro-
In order to crystalize the current contro- posed the concept of monophyletic origin on
versy and to aid the transcontinental study of the basis of his studies of superficial mandibu-
the Hylidae, it is necessary to restate or clarify lar musculature and vocal sac structure. Sub-
the following issues: 1) the phylogenetic re- sequent karyotypic data assembled by Steph-
lationships of the Indonesian hylids ( those on enson and Stephenson (1970), Woodruff
the periphery of the Australian population ( 1972 ) Morescalchi and Ingram ( 1974 ) and
, ,
and geographically closest to the Oriental Menzies and Tippett (1976) have in no way
species); 2) whether the Australopapuan caused this concept to change. All of the
species genuinely constitute a monophyletic hylids karyotyped to date have 2n = 26, ex-
group; 3) the phylogenetic relationships of cept L. infrafrenata (2n
=
24), and in that
the Australian species; and 4) the phylo- instance a model for derivation from 2n = 26
genetic relationships of the Australian and has been proposed (Menzies and Tippett,
South American hylid faunal units. 1976).
The Indonesian hylid fauna. —The north- Phylogenetic relationships of Australian
western geographic limit of Litoria occurs in species.
— Tyler and Davies ( 1978a ) examined
the Indonesian islands of Timor and the Less- the morphology, osteology, myology, distri-
er Sunda Islands of Sumba, Savu and Alor. bution, and biology of 92 of the 94 species of
This latter assemblage represents the eastern Litoria currently recognized. They found that
end of an archipelago forming an intimate these species can be associated in no less than
link to the Malaysian Peninsula far to the 37 species groups. Insofar as all geographic
northwest. Therefore, the Litoria fauna of areas occupied by Hyla are concerned, this
the Timor-Lesser Sunda group is of impor- total of groups not exceptionally high. In
is
tance to any contemplation of entry of hylids reality the number of species per species
into Australia from the northwest. group is
remarkably similar in several geo-
The only species (L. everetti) occurring graphic areas (Table 4:3). Although more
in the relevant area is a member of the L.
concerned with the initial step of establishing
peroni group represented elsewhere in the phonetic groupings, Tyler and Davies demon-
strated that Australian hylids
Australopapuan area by five described species occupy an in-
—peroni, amboinensis, rothii and
everetti,
credible gamut of niches.
darlingtoni. The total geographic range of Tyler (1970, 1972a) suggested that there
this species group is
exceptionally extensive existsa close phylogenetic relationship be-
(Tyler and Da vies, 1978a). It appears that tween Australian hylids and leptodactylids.
the group evolved in Australia or New Guinea The core of this suggestion related to the lep-
and is now radiating in several directions and todactylid genus Cyclorana. That genus as
extending its range. Thus, it is confirmed that then constituted comprised a group of squat-
the phylogenetic affinities of L. everetti are bodied and also some elongate species. Lynch
with other Australopapuan species and not ( 1971 ) regarded them an integral component
with the southernmost Oriental hylid (Hyla of the Australian leptodactylid fauna.
Tyler
chinensis). (1971, 1972a) studied superficial mandibular
Monophyletic or polyphyletic origins. In- — musculature in all Australian leptodactylids
sofar as the Australopapuan fauna is con- and his conclusions differed from those of
cerned, the issue is whether the frogs referred Lynch only in his appraisal of Cyclorana, in
to the Hylidae are a monophyletic group. which he noted distinct hylid affinities. Tyler
Ecologically, at least, they are incredibly di- (1970) suggested that the similarities of the
verse, filling a spectrum of niches occupied on Australian hylids and leptodactylids implied
other continents by different families. A su- the existence of a single common ancestor.
92 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Table 4:3. — Relationship Between Number of Species and Number of Species Groups of Selected Hylid Genera
on Different Continents.
Geographic area
1979 TYLER: SOUTH AMERICA AND AUSTRALIA 93
grounds for an intrinsic division within the ance: "Despite the divergent nature of Acris
current concept of the Hylidae. Therefore, I with respect to other hylids, and the super-
favor recognition of a subfamily to accommo- ficial similarity of Acris to ranids, the inescap-
date Litoria, Nyctimystes and (as indicated able fact remains that Acris has procoelous
in later discussion) Cyclorana. Because all vertebrae, an arciferal pectoral girdle, inter-
family group names are of equal status for calary cartilages and claw-shaped terminal
the purposes of nomenclatural priority, Pelo- —
phalanges a combination of characters that
dryadinae (derived from Pelodryadidae seemingly inextricably ally the genus with
Giinther, 1858) takes priority over Nyctimy- the hylids."
stinae Laurent, 1975. It is equally reasonable to suggest that
It is worth noting that when Maxson and organismal evolution is unlikely to be con-
Wilson (1975) implemented Savage's concept strained along any linear path of morpho-
of the Pelodryadidae, because of results of logical divergence as assessed by human ob-
estimated mean albumin-immunological dis- servers. Hence, systematists have a quandry
tances between continental populations, they that is own making, and while the
of their
overestimated the continental divergence Hylidae remains defined as it is now, the
time. Their immunological distance of 100 Pelodryadinae remains an integral component
units equates with 60 million years, so that of it.
million years isolation between the popula- species and the absence of intercalary struc-
tions. Their calculation of an immunological tures it formerly has been accommodated in
distance of approximately 129 units between the Leptodactylidae (Parker, 1940; Lynch,
the relevant Australian and South American 1971). More evidence has demon-
recent
populations can be interpreted in two ways, strated between Cyclorana and
similarities
but may well be excessive. The physical pelodryadine hylids in myology (Tyler,
separation of Australia from Antarctica is now 1972a), adrenal catecholamines (Robinson
established at 52-55 m.y.b.p. This total com- and Tyler, 1973), in larval structure and biol-
pares with 77 m.y.b.p. calculated by immu- ogy (Watson and Martin, 1973), and in
nological techniques. If the latter is the cranial osteology (Fig. 4:10). A closer exam-
period of isolation of the stocks, ecological or ination of the moqjhology of Cyclorana spe-
physical barriers on the Antarctic land mass cies resulted in the discovery of intercalary
are called for to explain the separation of structures in C. inermis, C. alboguttatus, and
populations prior to rifting. C. dahlii, and led to these species being re-
Wallace, Maxson, and Wilson (1971) ferred to the hylid genus Litoria by Straughan
found greater immunological distances exist- (1969), Tyler (1974b) and Tyler, Davies and
ing between South American and the adjacent
1
King (1978) respectively.
North American species, than between North In consequence of these actions and of the
American and the geographically distant sin- resurrection of one species and the descrip-
gle Australian species examined. Maxson tion of five new species (Tyler and Martin,
(1978) demonstrated a high degree of 1975, 1977), Cyclorana now is composed ex-
compatability between North American and clusively of robust fossorial frogs lacking in-
European Hyla. Maxson and Wilson (1971) tercalary structures but retaining a closer af-
noted that where discrepancies exist between finity to hylid than to leptodactylid frogs.
organismal resemblance and albumin resem- Awareness of this presumably led Heyer and
blance, it is to be attributed to differential Liem (1976) to omit Cyclorana from their
rates of organismal evolution. As an example
they cited Acris, which exhibits albumin and 1
The customary term "intercalary cartilages" is not
haemoglobin affinities to North American used because these structures are bony in 46 of 71
Hyla, and yet is strikingly different from such Australopapuan hylid species studied (Tyler and
Davies, 1978a). Ossification bears no correlation
species in anatomy, gross structure, biology, with finger length, habits or geographic distribution.
and ecology. Duellman (1970:647) accepted
However, all large or moderately large arboreal
such evidence with considerably less toler- species retain a cartilaginous state.
94 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
5mm 5mm
Litoria raniformis Litoria alboguttata
Fig. 4:10. Skulls of certain species of Cyclorana and Litoria.
Crdneos de ciertas especies de Cyclorana ij Litoria.
phylogenetic analysis of the Australopapuan group may prove to be the sister group of
leptodactylidae ( Myobatrachidae ) .
Cyclorana and merit elevation to distinct ge-
Because all hylids exhibit axillary amplex- neric identity.
us and Australian leptodactylids (except Mix-
ophyes) inguinal, it follows that the embrace
of Cyclorana should be of relevance in de- Leptodactylidae (Myobatrachidae)
and the Myobatrachinae. Lynch (1971) pro- Myobatrachus and Pseudophryne. More re-
vided a splendid historical account of the cently, another new and as yet undescribed
classification of the Leptodactylidae, and in genus (Tyler et al., 1979) was discovered in
so doing, placed Parker's contribution in an the northern portion of Western Australia and
historical perspective. Following the publica- the Northern Territory. This form produces a
tion of Parker's monograph, A. R. Main and foam nest, has tadpoles with suctorial mouths
his colleagues undertook the first detailed and elongate tails, and the adult exhibits enor-
Certainly Neobatrachus was resurrected from scribed three more new genera to accommo-
the synonymy of Heleioporus by Main, Lee, date —
known species Paracrinia for Crinia
and Littlejohn ( 1958 ) but in over 25 years haswelli; Australocrinia to accommodate two
,
which the majority of species of one family tylid genera and approaches some primitive
could be distinguished from those of the leptodactylids from other geographical areas
other. I draw attention to this fact not in an ( Heleophryne, the Ceratophryinae, many Tel-
all with 2n = 26).
attempt to score a point, but solely to high- matobiinae,
light the fact that the case for considering
the On gross morphological grounds, the re-
Myobatrachidae a separate family still needs semblance between some Australian and
to be substantiated. South American genera is striking. If the
Duellman (1975) was able to accommo- South American Ratrachyla should be found
date Rheobatrachas within the Myobatrachi- tomorrow in the cool, temperate forests of the
dae, arguing that neither the suite of primi- southern section of the Australian Great Di-
tive morphological character states, nor the viding Range, it would be compared with
bizarre reproductive mode should exclude it. Kyarranus and Philoria, found to be highly
Lynch (
1971 ) employed a graphic technique similar, and attract little comment.
South America is curious indeed. Morescalchi family (the Gekkonidae) containing each of
(
1973 ) and Morescalchi and Ingram 1974 ( )
the other four units as subfamilies. Some of
noted that Cyclorana alboguttatus (now Li- Kluge's concepts of the origins, dispersal, and
toria albo guttata) is karyologically less dif- evolutionary relationship of these subfamilies
ferentiated than many Australian leptodac- have been variously criticized by Maderson
1979 TYLER: SOUTH AMERICA AND AUSTRALIA 97
have demon-
(1972), Moffat (1973),
and Russell (1976). karyotypic studies in Australia
strated within well-established gekkonine and
interpretation was made
1967 with
Kluge's ( )
of numer-
the assumption that continental drift was diplodactyline species the existence
ous species, none of
which has yet
neither tenable as an hypothesis, nor germane biological
to the resolution of the study. This led to
been accorded formal taxonomic status ( King,
1973, 1975, 1977; King and Rofe, 1976).
For
conclusions that must be reexamined within
recognized five dis-
the context of continental drift as an accept- example, King (1977)
cretely distributed and chromosomally
dis-
able hypothesis. For example, there is his
tinct populations within what is now termed
assumption that the Australian Diplodactyli- new
nae evolved from a primitive southeast Asian Diplodactylus vittatus. When all of these
chelid fauna was provided by Warren ( 1969 ) tal origins. The extant species of Crocodylus
who reported Emydura sp. macquari from
aff. are clearly of Oriental origin or derived from
siltstones in Tasmania. Reported to be of an Oriental stock, and are confined to the
Oligocene-Miocene age they are now con-
—
north of the Region C. porosus of south-
sidered somewhat older and of Early Tertiary eastern Asia, New Guinea and Australia, C.
of New Guinea and C. johnsoni
age (J. W. Warren, pers. coram.). Chelids are novaeguineae
no longer extant in Tasmania, and E. mac- of Australia.
quari is now confined to the Murray-Darling The fossil fauna is substantial both in
drainage system of the southeastern Australian quantity and diversity, and a number of
mainland. With the discovery of freshwater highly significant finds has been reported re-
turtle remains from Tasmania and also from cently. Molnar ( 1977 ) described from Chil-
various Miocene to Pleistocene deposits on the lagoe in North Queensland an incomplete
mainland (Callen and Tedford, 1976; Archer skull, with a high and laterally compressed
and Wade, 1976), it is likely that previously snout and probably xiphodont dentition. The
chelids have occupied almost all of the Aus- subsequent discoveiy of Pleistocene Palor-
tralian continent and New Guinea as well. chestes cf. P. azael at the same site was in-
ney's ( 1975 ) study of the phytogeny and to the Australian Region. Kluge (1974) rec-
classification of turtles almost exclusively re- ognized eight genera and 30 species of pygo-
lied upon cranial characters, and, in common podids. Kluge's (1976) analysis of phylo-
with most Australian chelid remains, the head genetic relationships within the family led to
and neck of Chelycarapookus remain un- the recognition of only six genera. Following
known. the work of Underwood (1957), there has
been general acceptance that the pygopodid-
gekkonid relationship is extremely close, and
Crocodiles
it follows that the most
likely origin for the
The crocodiles include two quite distinct pygopodids is within Australia directly from
a gekkonid stock.
components differing in their ancestry and arc
probably of separate Gondwanan and Orien- Many pygopodid species are restricted
1979 TYLER: SOUTH AMERICA AND AUSTRALIA 99
westward dispersal of animals by rafting from Myers (1949), and Gorham (1968).
South America in the direction of Australia. —
Neic Zealand. The most well known com-
The west coast of South America is swept ponent of the New Zealand herpetofauna is
by the north flowing Humboldt Current which the rhynchocephalian reptile, the Tuatara
meets the transpacific southern equatorial cur- Sphenodon punctatus. This relic was prob-
rent at mid-latitudes. The latter current
ably quite widely distributed elsewhere in the
travels westward and eventually disperses Mesozoic, as evidenced by the extensive fos-
around all of the islands of the Pacific south sil record of
rhynchocephalians at that time.
of the Equator. The remaining terrestrial reptiles are 35 spe-
Thirty years ago Thor Heyerdahl's raft cies of lizards (almost all are endemic) of the
Kon-Tiki demonstrated the transport potential families Gekkonidae and Scincidae. The
of the southern equatorial current by travel- geckos include three endemic genera that are
ling from Peru to the Tuamotu Archipelago in unique among gekkonids in being ovovivipar-
French Polynesia south of the Marquesas ous, whereas the endemic skinks are members
Islands. Had his vessel been driven two or of genera widely distributed outside New
three degrees northwards, he woidd have Zealand.
passed between the Marquesas and the Tua- In their recent survey of New Zealand
motu Archipelago, and continued much far- vertebrates, Bull and Whittaker (1975) ac-
ther west on a longer journey terminating in
cepted Kluge's (1967) interpretations of the
Western Samoa, Fiji, or Tonga. This longer origin of the gekkonids, resulting in their
journey is probably the route of the ancestral statement (op. cit: 239): "The geckos form
stocks of the iguanids of Fiji and Tonga — part of the Malayo-Pacific element of the New
Brcichylophus fasciatus and B. brevicephalus, Zealand fauna and probably entered New
respectively. South American derivates are Zealand in the Miocene when the climate was
not evident elsewhere, and the remaining ele- warmer and the land more extensive than
ments of the herpetofauna of the islands in now." They further visualized fairly extensive
the west and southwest Pacific area are repre- oceanic dispersal by rafting from New Cale-
100 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
donia or directly from Australia. A totally dif- fected in a millenium or the subfossils repre-
ferent interpretation would result by consider- sent extinct species; in either case it seems to
ing the connection between New Zealand and be unprofitable to speculate about the char-
the Lord Howe Rise and the exceptionally acteristics of their immediate postdrift an-
close proximity of the Lord Howe Rise to cestors.
Australia in Early Cenozoic (Griffiths and Within the context of discussion of south-
Varne, 1972). west Pacific biological origins, the concepts of
New Zealand Nur and Ben-Avraham (1977) on a lost Pa-
If geckos really arrived in
no earlier than the Miocene, they have been cificcontinent (formerly lying close to the
east coast of Australia) must be considered.
evolving rapidly ever since along a unique
Nevertheless the present study of the anurans
path. Therefore, acceptance of an Early Ceno-
zoic entry avoids any concept of an explosive has not required such a landmass to explain
their origins.
radiation, and provides an adequate time span
for speciation in situ along novel lines. Kluge's
(1967) hypothesis of gekkonid evolution and CONCLUSIONS
dispersal did not accommodate continental
drift. Accordingly, it is not surprising that his
When the amphibian and reptile families
interpretations of the source of the faunal now found in Australia are examined, one by
ancestors is at variance with one incorporat-
one, to determine whether their affinities lie
ing this phenomenon. with South American or with Oriental stocks,
The endemic frog fauna of New Zealand it
rapidly becomes apparent that Oriental
is even more bizarre than the reptile fauna sources predominate, and that Australian-
and is represented by three species of Leio- South American links are few indeed. At the
pelma. Whether Leiopelma and Ascaphus of commencement of drifting in the Eocene, the
North America should be placed in a single Australian herpetofauna included the follow-
family ( Ascaphidae), or whether Leiopelma ing families shared with South America.
should constitute the Leiopelmatidae remains Gekkonidae (Diplodactylinae). The na- —
a matter of debate. However, there is no ture of the extensive radiation within Aus-
argument to the concept that these genera tralia well support the concept that the
may
represent relics of a fauna that was widely Gekkonidae was the only lizard family pres-
distributed. Estes and Reig (
1973 ) referred ent in Australia.
Vieraella and Notobatrachus of the Early and Boidae (P Madtsoinae). —
The presence of
Late Jurassic of Patagonia to the Ascaphidae. this family hinges upon the Pleistocene Wo-
Until recently, Leiopelma hamiltoni was nambi whose phylogenetic affinities are with
known only from a small heap of stones oc- snakes outside the Oriental Region. Extant
cupying one quarter of a hectare on the South genera certainly arrived in the Miocene.
Island. Now is known from an additional
it Chelidae. A — Gondwanan component
on the small Maud Island off
fifteen hectares probably of considerable antiquity. Within
the coast (Rull and Whitaker, 1975). Leio- Australia, fossils extending to the Early Ter-
pelma archeyi and L. hochstetteri are dis- tiary represent modern species. The origin of
tributed somewhat more widely on the North the family is uncertain, but the Lower Cre-
Island, but Bull and Whitaker (op. cit: 235) taceous Chehjcarapookus arcuatus, for which
stated that in the recent past probably (". . . Warren (1975) erected the Chelycarapooki-
within the last 1,000 years. Leiopelma
. .
.") dae, exhibits postcranial features that render
was far more widespread, being known from it a potential chelid ancestor.
—
Leptodactylidae. Irrespective of the final typed by Mrs. J. Russell-Price. I am also
assessment of the familial disposition of the indebted to Richard G. Zweifel for permis-
Australian genera and species, the South sion to reproduce the map of New Guinea
American affinities of the Australian members upon which figure 4:5 was prepared.
are indisputable.
These six families represent the elements
of the Australian herpetofauna destined to
RESUMEN
persist through to the Holocene. It follows
that by modern standards the total herpe- Las oportunidades para un intercambio
tofauna was incredibly depauperate, and herpetofaunistico gondwanico entre Sud
lacked many significant elements. Pygopodids America y Australia fueron muchos menores
evolved in Australia probably during the que aquellos entre Sud America y Africa,
period of isolation. debido a la expansion de la Antartica. Al
The comienzo de la separation de Australia de
serious deficiencies were remedied
from the Miocene onwards Sud America (53 m.a.a.p.) el continental su-
and
as Australia
the eastern outliers of the Oriental Region
reno Australiano era subtropical. Condiciones
humedas con una diversidad de anfibios y
approached one another on their collision
course. The commencement of colonization reptiles occupaban el centra de Australia hasta
el Pleistoceno superior cuando un incremento
by families such as the Scincidae probably
antedated the mid-Miocene. Others such as de la aridez elimino muchos sitios acuaticos.
the Typhlopidae, Carettochelyidae, Microhy- De todos los posibles modos de establecer
lidae, Varanidae, and Crocodylidae probably cuales segmentos de la herpetofauna australi-
were acquired at the time of the collision, ana eran compartidas con Sud America como
while the Ranidae entered then and again in resultado de un intercambio gondwanico, se
a subsequent wave. In broad terms, the Ori- ha seleccionado determinar que familias que
ental influence upon Australia probably was estan hoy presentes en Australia y Nueva
more significant than the North American in- Guinea, fueron adquiridas cuando Australia
fluence upon South America, because South coludio con la Region Oriental en el Mioceno
America had a more diverse and better estab- medio. Un analisis de tal evento demuestra
lished herpetofauna before the time of con- que un grupo de islas adheridas a lo que es
tact. la actual costa norte de Nueva Guinea. Por
ende, si estas islas estaban poblados con an-
fibios y reptiles algunos, al menos, debieron
ACKNOWLEDGMENTS haberse integrado a la heipetofauna australo-
papua a traves de esta via.
For the invitation to participate in the Se propone un modelo que involucra la
symposium "The South American Herpeto- dispersion de algunos generos a traves de las
fauna," and for the funding making this pos- Islas Orientales desde las Filipinas a Fiji en
sible, I am greatly indebted to The University el Mioceno (el
genera de batracios ranidos
of Kansas, and particularly to the convenor, La colision de tierras arego
Platy mantis).
William E. Duellman. este rana a la fauna de Nueva Guinea y los
Many of the ideas and assessment pro- actuales extremos de diversidad al Oeste y al
posed paper arose from or were stimu-
in this Este de Nueva Guinea reflejan la edad de
lated by discussions or correspondence with especiacion. En contraste otros generos arri-
many of my colleagues. For this great help I baron desde el Oeste despues de la colision,
gratefully acknowledge the contributions of resultando en una progresiva reduction del
Bob Lange, John M. Legler, Brian McGow- numero de especies desde el Oeste al Este.
ran, Allen E. Greer, Rowley Twidale and Entonces, los patrones de distribution geo-
George R. Zug. grafica de los elementos Orientales en Aus-
Margaret Davies prepared figures 4:1 and tralia varian de acuerdo con la fecha de en-
4:10,whereas figures 4:3-4:7 and 4:9 are the trada relativa a la colision de Australia. Sin
work of Debra Bennett. The manuscript was embargo, el topico mas importante es las rela-
102 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
cion de los anfibios y reptiles de Australia que Bai, A. R. K. 1956. Analysis of chromosomes in two
genera of Microhylidae (Amphibia: Anura).
aquellos de la Region Oriental adyacente. Proc. Natl. Inst. Sci. India, Part B 22:1-5.
Bajo esta criterio, las Ranidae, Microhylidae, Barbour, T. 1923. The frogs of the Fiji Islands.
Varanidae, ? Scincidae, Agamidae, Caretto- Proc. Acad. Nat. Sci. Philadelphia 75:111-115.
Colubri- Beaufort, L. F. de. 1951. Zoogeography of the
chelyidae, Trionychidae, Elapidae,
land and inland waters. Sidgwick and Jackson,
dae, Acrochordidae, Uropeltidae, Typhlopi-
London, 208 p.
dae, y Boidae entraron a la region geografica Becak, M. L., Demaro, L., Becak, W. 1970. Poly-
australiana despues del Mioceno medio. Los ploidy and mechanisms of karyotypic diversifica-
Boidae son linicos, por tener un genuino com- tion in Amphibia. Cytogenetics 9:225-238.
toceno se relaciona con Madtsoia), y otro numerical approach. Australian ]. Zool. 21:119-
Oriental. 149.
la Blommers, R. 1971. Karyotype de Anodontohyla
Correspondientemente, hcrpetofauna
montana Angel (Anura, Microhylidae) de massif
gondwanica de Australia no poseia la gran de TAndringitra (Madagascar). Terre Mai. 10:
mayoria de los elementos de esa fauna. Un 261-265.
buen ejemplo esta dado por la presencia de Blommers-Schlosser, R. M. A. 1976. Chromosomal
losHylidae y Leptodactylidae ( y por la man- analysis of twelve species of Microhylidae (Anu-
tencion del uso de estos nombres familiares ra) from Madagascar. Genetica 46:199-210.
Bocart, J. P., Nelson, C. E. 1976. Evolutionary
para los animales modernos de Australia). implications from karyotypic analysis of the fami-
Aparte de estos, existian solo los Cheliidae y lies Microhylidae and Rhinophrynidae. Herpeto-
los geckos diplodactylinos, de los cuals los logica 32:199-208.
Bogart, J. P., Pyburn, W., Nelson, C. E. 1976. The
Pygopodidae probablemente evolucionaron
karyotype of Otophnjne robusta (Anura: Micro-
como los linicos representantes australianos.
hylidae). Ibid. 32:208-210.
Una deriva pasiva permitio a los largartos Bowler, J. M. 1976. Aridity in Australia: Age, ori-
largas distancias desde la gins and expression in aeolian landforms and sedi-
iguanidos viajar
ments: Earth-Sci. Rev. 12:279-310.
costa oeste de Sud America a Fiji y a Tonga.
Brown, W. C, Alcala, A. C. 1970. The zoogeog-
A pasar de que la direccion de las corrientes raphy of the herpetofauna of the Philippine
transpacificas (Este a Oeste) favorece Is dis- Islands, a fringing archipelago. Proc. California
persion desde Sud America a Australia mas Acad. Sci. 38(6) 10.5-130.
:
Nat., Paris Sec. C 88:72-86. Cocger, H. G. 1974. Voyage of the banded iguana.
W. 1975. The presence of Australian Nat. Hist. 18:144-149.
Bagnara, J. F., Ferris,
phyllomedusine melanosomes and pigments in Cogger, H. G. 1975. Reptiles and amphibians of
Australian hylids. Copeia 1975(3) :592-595. Australia. Reed, Sydney, 584 p.
Bagnara, J. T. 1976. Color change, pp. 1-52 in Cole, C. J., Zweifel, R. G. 1971. Chromosomes of
Lofts, B. (ed.). Physiology of the Amphibia. a New Guinean microhylid frog, Cophixalus ri-
Vol. Ill, Academic Press, New York, 597 p. IHirius Zweifel. Herpetol. Rev. 3:15-16.
1979 TYLER: SOUTH AMERICA AND AUSTRALIA 103
Coleman, P. J. 1975. The Solomons as a non-arc. Chelidae: A theory of relationships using shared
Bull. Australian Soc. Explor. Geophys., 6(2-3): derived characters. Amer. Mus. Novit. ( 2620 ) :
60-61. 1-28.
Coleman, P. J., Packham, G. H. 1976. The Melane- Claessner, M. F. 1942. The occurrence of the
sian borderlands and India-Pacific plates' bound- New Guinea turtle (Carettochchjs) in the Mio-
ary. Earth-Sci. Rev. 12:197-233. cene of Papua. Rec. Australian Mus. 21:106-109.
Corben, C. J., Ingram, G. J., Tyler, M. J. 1974. Glassford, D. K., Killigrew, L. P. 1976. Evidence
Gastric brooding: Unique form of parental care for Quaternary westward extension of the Aus-
in an Australian frog. Science 186:946-947. tralian desert into south-western Australia. Search
Cracraft, J. 1973. Vertebrate evolution and bio- 7:394-396.
geography in the Old World tropics: Implications Gorham, S. W. 1965. Fiji frogs (with synopses of
of continental drift and palaeoclimatology, pp. the genera Cornufcr and Plat y mantis). Zool.
373-393 in Tarling, D. H., Runcorn, S. K. Beitr. 11:381-435.
(eds. ). Implications of continental drift to the Gorham, S. W. 1968. Fiji frogs. Life history data
earth sciences. Vol. 1. Academic Press, London, from field work. Ibid. 11:427-446.
622 p. Gow, G. F. 1977. A new species of Pytlion from
Cracraft, J. 1974. Continental drift and vertebrate Arnhem Land. Aust. Zool. 19:133-139.
Ann. Rev. Ecol. Syst. 5:215-261.
distribution. Greer, A. E. 1970. A subfamilial classification of
Cracraft, J. 1975. Mesozoic dispersal of terrestrial scincid lizards. Bull. Mus. Comp. Zool. Harvard
faunas around the southern end of the world. Univ. 139:151-184.
Mem. Mus. Natl. Hist. Nat. Paris, Ser A 88:29- Griffiths, Barne, R. 1972. Evolution of the
J. R.,
54. Tasman Macquarie Ridge and Alpine Fault.
Sea,
Darlington, P. J. 1957. Zoogeography: The geo- Nature (Phys. Sci.) 235:83-86.
graphical distribution of animals. Wiley, New Guibe, J. 1948. Contribution a l'etude de l'appareil
York, 675 p. genital des tvphlopides (Ophidiens). Bull. Soc.
Denham, D. 1975. Distribution of underthrust litho- Zool. France 73:224-228.
spheric slabs and focal mechanisms Papua New — Guttman, 1973. Biochemical techniques and
S. I.
Guinea and Solomon Islands Region. Bull. Aus- problems anuran evolution, pp. 183-203 in
in
tralian Soc. Explor. Geophys. 6(2-3) 78-79. : Vial, J. L. (ed. ). Evolutionary biology of the
De Vis, C. W. 1894. The lesser chelonians of the anurans: Contemporary research on major prob-
Nototherian drifts. Proc. Roy. Soc. Queensland lems. Univ. Missouri Press, Columbia, 470 p.
10:123-127. Hecht, M. K. 1975. The morphology and relation-
De Vis, C. W. 1897. The extinct fresh-water turtles ships of the largest known terrestrial lizard, Mega-
of Queensland. Ann. Queensland Mus. 3:3-7. lania prisca Owen, from the Pleistocene of Aus-
Dixon, J. R., Anderson, S. C. 1973. A new species tralia. Proc. Roy. Soc. Victoria (new ser.) 87(2):
Mus. Nat. Hist. 155:387-436. earth sciences. Vol. 1. Academic Press, London,
Gaffney, E. S. 1977. The side-necked turtle family 622 p.
104 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
King, M. 1973. Chromosomes of two Australian MacKenzie, D. E. 1975. Volcanic and plate tectonic
lizards of the families Scincidae and Gekkonidae. evolution of central Papua and New Cuinea. Bull.
Cytologia 38:20.5-210. Australian Soc. Explor. Geophys. 6(2-3) :65-67.
Kinc, M. 1975. Chromosomal studies on Australian Maderson, P. F. A. 1972. The structure and evolu-
lizards. Ph.D. Dissert. Univ. Adelaide, 150 p. tion of holocrine glands in sphaerodaetyline and
King, M. 1977. Chromosomal and morphometric eublepharine gekkonid lizards. Copeia 1973 (3):
variation in the Gekko Diplodactylus vittatus 559-571.
Gray. Australian J. Zool. 25:43-57. Main, A. R., Lee, A. K, Littlejohn, M. J. 1958.
King, M., King, D. 1975. Chromosomal variation in Evolution in three genera of Australian frogs.
the lizard genus Varanus (Reptilia). Australian Evolution 12:224-233.
J. Biol. Sci. 28:89-108. Maxson, L. R. 1976. The phylogenetic status of
King, M., Rofe, R. 1976. Karyotypic variation in phyllomedusine frogs ( Hylidae ) as evidenced
the Australian Cecko Phyllodactylus marmoratus from immunological studies of their serum albu-
(Gray) (Gekkonidae: Reptilia). Chromosoma mins. Experientia 32:1149-1150.
54:75-87. Maxson, L. R., Wilson, A. C. 1975. Albumin evo-
Kluge, A. G. 1967. Higher taxonomic categories of lution and organismal evolution in tree frogs
gekkonid lizards and their evolution. Bull. Amer. (Hylidae). Syst. Zool. 24:1-15.
Mus. Nat. Hist. 135:1-60. McDowell, S. B. 1967. Aspidomorphus, a genus of
Kluge, A. G. 1974. A taxonomic revision of the New Guinea snakes of the family Elapidae, with
Pvgopodidae. Misc. Publ. Mus. Zool.
lizard family notes on related genera. J. Zool. 151:497-543.
Univ. Michigan (147): 1-221. McDowell, S. B. 1970. On the status and relation-
Kluge, A. G. 1976. Phylogenetic relationships in the ships of the Solomon Island elapid snakes. Ibid.
lizard family Pygopodidae: An evaluation of 161:145-190.
theory, methods and data. Ibid. (152): 1-72. McDowell, S. B. 1972. The species of Stegoiwtus
Lance, R. T. 1976. Fossil epiphyllous "Germlings," (Serpentes, Colubridae) Papua New Guinea.
in
their living equivalents and their palaeohabitat Zool. Meded. Rijks Mus. Nat. Hist. Leiden 47:
indicator value. Neues Jahrb. Geol. Palaeontol. 6-26.
Abh. 151:142-165. McDowell, S. B. 1974. A catalogue of the snakes
Latz, P. K. 1975. Notes on the relict palm Livistona of New Guinea and the Solomons, with special
mariae F. Muell. in central Australia. Trans. Roy. reference to those in the Bernice P. Bishop Mu-
Soc. South Australia 99:189-195. seum. Part 1, Scolecophidia. J. Herpetol. 8:1-57.
Laurent, R. F. 1975. La distribution des amphib- McDowell, S. B. 1975. A catalogue of the snakes
iens et les translations continentales. Mem. Mus. of New Guinea and the Solomons, with special
Natl. Hist. Nat., Paris 88:176-191. reference to those in the Bernice P. Bishop Mu-
Leon, P. E. 1970. Report of the chromosome num- seum. Part II, Anilioidea and Pvthoninae. Ibid.
bers of some Costa Rican anurans. Rev. Biol. 9:1-79.
Trop. 17:119-124. McDowell, S. B., Bocert, C. M. 1954. The sys-
Liem, D. S. 1973. A new genus of frog of the family tematic position of Lanthanotus and the affinities
Leptodactylidae from SE. Queensland, Australia. of the anguimorphan lizards. Bull. Amer. Mus.
Mem. Queensland Mus. 16:459-470. Nat. Hist. 105:1-142.
Longman, N. A. 1929. Palaeontological notes. McGowran, B. 1973. Rifting and drift of Australia
Specimens from a well at Brigalow. Ibid. 9:247- and the migration of mammals. Science 180:
252. 759-761.
Lydekker, R. L. 1889. Catalogue of fossil amphib- Menzies, J. I., Tippett, J. 1976. Chromosome num-
ians and reptiles in the British Museum Natural (
bers of Papuanhylid frogs and the karyotype of
History) Part 3. British Museum, London, 254 p. Litoria infrafrenata (Amphibia, Anura, Hvlidae).
ology and zoogeography of leptodactyloid frogs. Menzies, J. I., Tyler, M. J. 1977. The systematica
Univ. Kansas Mus. Nat. Hist. Misc. Publ. (53): and adaptations of some Papuan microhylid frogs
1-238. which live underground. J. Zool. 183:431-464.
Lynch, J. D. 1973. The transition from archaic to Mittleman, M. B. 1952. A generic synopsis of the
advanced frogs, pp. 133-182 in Vial, J. L. (ed.). lizards of the subfamily lygosominae. Smithson-
Contem- ian Misc. Collect. 117:1-35.
Evolutionary biology of the anurans:
porary research on major problems. Univ. Mis- Moffatt, L. 1973. The concept of primitiveness
souri Press, Columbia, 470 p. and its bearing on the phylogenetic classification
Mabbutt, J. A. 1967. Denudation chronology in of the Gekkota. Proc. Linn. Soc. New South
Central Australia. Structure, climate and land- Wales 97:275-301.
form inheritance in the Alice Springs area, pp. Molnar, R. E. 1977. Crocodile with laterally com-
144-181 in Jennings, J. N., Mabbutt, J. A. pressed snout: First find in Australia. Science
(eds.). Landform studies from Australia and 197:62-64.
New Guinea. Australian Natl. Univ. Press, Can- Molnar, R. E. 1978. Age of the Chillagoe crocodile.
berra, 434 p. Search 9(4) 156-158.
:
1979 TYLER: SOUTH AMERICA AND AUSTRALIA 105
Morescalchi, A. 1968a. Initial cytotaxonomic data Schllnger, E. I. 1974. Continental drift, Notho-
on certain families of amphibious Anura Diplas- ( fagus, and some ecologically associated insects.
iocoela, after Noble, T. C. ). Experientia 24:280- Ann. Rev. Entomol. 19:323-343.
283. Schodde, R., Mason, I., Wolfe, T. O. 1972. Fur-
Morescalchi, A. 1968b. Hypothesis on the phytog- ther records of the pitted-shelled turtle (Caretto-
eny of the Salientia, based on karyological data. chclys insculpta) from Australia. Trans. Roy. Soc.
Ibid. 24:964-966. South Australia 96:115-117.
Morescalchi, A. 1973. Amphibia, pp. 233-248 in Sclater, J. G., Von Der Borch, C, Veevers, J. J.,
Chiarelli, A. B., Capanna, E. (eds.). Cytotax- Hekinian, R., Thompson, R. W., Pimm, A. C.,
onomy and vertebrate evolution. Academic Press, McGowran, B., Gartner, S., Johnson, D. A.
London, 783 p. 1974. Regional synthesis of the deep sea drilling
Morescalchi, A., Ingram, G. J. 1974. New chro- results from leg 22 in the eastern Indian Ocean,
mosome numbers in Australian Leptodactylidae pp. 815-831 in Von Der Borch, C, Sclater,
(Amphibia, Salientia). Experientia 30:1134— J. G., Veevers, J. J. Initial reports of the Deep
1135. Sea Drilling Project 22, 890 p.
Natarjan, R. 1953. A note on the chromosomes of Smith, M. J. 1975. The vertebrae of four Australian
Cacopus systoma. Proc. 40th Indian Sci. Congress, elapid snakes. Trans. Roy. Soc. South Australia
Lucknow pp. 180-181. 99:71-84.
Nelson', C. E. 1975. Another new miniature 4-toed Smith, M. J. 1976. Small fossil vertebrates from
South American microhylid frog (Genus: Syn- Victoria Cave, Naracoorte, South Australia. IV
cope). J. Herpetol. 9:81-84. Reptiles. Ibid. 100:39-51.
Newsome, A. E., Rochow, K. A. 1964. Vertebrate Stephenson, E. M., Stephenson, N. G. 1970. Kar-
fossils from Tertiary sediments in central Aus- yotypes of two Australian hylid frogs. Chromo-
tralia. Australian J. Sci. 26:252. soma 30:38-50.
Nur, A., Ben-Avraham, Z. 1977. Lost Pacifica con- Stirton, R. A., Tedford, R. H., Miller, A. H. 1961.
tinent. Nature 270:41-45. Cenozoic stratigraphy and vertebrate paleontology
Parker, H. W. 1929. Two fossil frogs from the of the Tirari Desert, South Australia. Geogr. Rev.
Lower Miocene of Europe. Ann. Mag. Nat. Hist. 62:40-70.
(10) 4 (21):270-281. Storr, G. M. 1964. Some aspects of the geography
Parker, H. W. 1934. A monograph of the frogs of of Australian reptiles. Senckenbergiana Biol. 45:
the family Microhylidae. British Museum (Na- 577-589.
tural History), London, 208 p. Strauchan, I. R. 1969. Hyla inermis (Peters), a
Parker, H. W. 1940. The Australasian frogs of the species hitherto erroneously referred to the lepto-
family Leptodactylidae. Novit. Zool. 42:1-106. dactylid genus Cyclorana (Anura, Hylidae/Lep-
Pianka, E. R. 1972. Zoogeography and speciation of todactylidae). Zool. Meded. Rijks Mus. Nat. Hist.
Australian desert lizards: An ecological perspec- Leiden 43:207-212.
tive. Copeia 1972 (1):127-144. Straughan, I. R., Lee, A. K. 1966. A new genus
Rabello, M. N. 1970. Chromosomal studies in Bra- and species of leptodactylid frog from Queens-
zilian anurans. Caryologia 23:45-59. land. Proc. Roy. Soc. Queensland 77:6.3-66.
Rhodin, A. G. J., Mittermeier, R. A. 1977. Neural Taylor, L. W. H. 1975. Deposition and tectonic-
bones in chelid turtles from Australia and New patterns in the western Coral Sea. Bull. Australian
Guinea. Copeia 1977 (2):370-372. Soc. Explor. Geophys. 6:33-35.
Rich, P. V. 1975. Antarctic dispersal routes, wan- Tilbury, L. A. 1975. Lineations in the Bismarck
dering continents, and the origin of Australia's Sea. Ibid. 6:72.
non-passeriform avifauna. Mem. Natl. Mus. Vic- Twidale, C. R. 1972. Evolution of sand dunes in
toria (36):63-125. the Simpson Desert, Central Australia. Trans.
Robb, J. 1966. The structure and possible function Instit. British Geogr.
(56):77-109.
of the cloacal pouches of male Australian typhlo- Tyler, M. J. Microhylid frogs of New Britain.
1967.
pids. Australian J. Zool. 14:27-30. Trans. Roy. Soc. South Australia 91:187-190.
Robinson, R. L., Tyler, M. J. 1972. The catecho- Tyler, M. J. 1970. Patterns of distribution and the
lamine content of the adrenal glands of frogs as origins of the Papuan hylid frog fauna. Search
an index of phylogenetic relationships. Comp. 1:246-247.
Gen. Pharmacol. 3:167-170. Tyler, M. J. 1971. The phylogenetic significance of
Russell, A. P. 1976. Some comments concerning vocal sac structure in hylid frogs. Univ. Kansas
interrelationships amongst gekkonine geckos, pp. Mus. Nat. Hist. Occas. Pap. ( 19) :319-360.
217-244 in Bellaxrs, A. d'A., Cox, B. (eds.). Tyler, M. ]. 1972a. Superficial mandibular muscu-
Morphology and biology of reptiles. Linn. Soc. lature, vocal sacs and the phylogeny of Australo-
Symp. (Ser. 3), 290 p. Papuan leptodactylid frogs. Rec. South Australian
Savage, J. M. 1973. The geographic distribution of Mus. 16:1-20.
frogs: Patterns and predictions, pp. 351-445 in Tyler, M. J. 1972b. A new genus for the Australian
Vial, J. L. (ed.). Evolutionary biology of the leptodactylid frog Crinia darlingtoni. Zool. Me-
anurans: Contemporary research on major prob- ded. Rijks. Mus. Nat. Hist. Leiden 47:193-201.
lems. Univ. Missouri Press, Columbia, 470 p.
106 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Tyler, M. J. 1974a. Superficial mandibular mus- Underwood, G. L. 1954. On the classification and
culature and vocal sac structure in the Anura. evolution of geckos. Proc. Zool. Soc. London 124:
M.Sc. Thesis, Univ. Adelaide, 134 p. 469-492.
Tyler, M. J. 1974b. First frog fossils from Aus- Underwood, G. L. 1957. On lizards of the family
tralia. Nature 248 (5450) :711-712. Pygopodidae. A contribution to the morphology
Tyler, M. J. 1974c. The systematic position and and phvlogeny of the Squamata. J. Morphol.
geographic distribution of the Australian frog 100:207-268.
Chirolcptes alboguttatus Gunther. Proc. Roy. Soc. Veevers, J. J., McElhinny, M. W. 1976. The sep-
aration of Australia from other continents. Earth-
Queensland 85:27-32.
Sci. Rev. 12:139-159.
Tyler, M. J. 1976a. Frogs. William Collins, Syd-
ney, 236 p. Walker, C. F. 1973. A new genus and species of
Tyler, M. J. 1976b. Comparative osteology of the microhylid frog from Ecuador. Univ. Kansas Mus.
pelvic girdle of Australian frogs and description Nat. Hist. Occas. Pap. (20): 1-7.
of a new fossil genus. Trans. Roy. Soc. South Walker, C. F., Duellman, W. E. 1974. Descrip-
Australia 101:3-14. tion of a new species of microhylid frog, Cliiasmo-
Tyler, M. J. 1976c. A new genus and two new spe- cZeis, from Ecuador. Ibid. (26): 1-6.
Tyler, M. J. 1977. Pleistocene frogs from caves at evolutionary clock hypothesis. Proc. Natl. Acad.
Naracoorte, South Australia. Trans. Roy. Soc. Sci. USA 68:3127-3129.
South Australia 101:85-89. Warren, J. W. 1969. Chelid turtles from the mid-
Tyler, M. J. 1979. The introduction and current tertiary of Tasmania. J. Palentol. 43:179-182.
distribution in the New Hebrides of the Aus- Warren, J. W. 1975. A fossil chelonian of probably
tralian hylid frog Litoria aurea. Copeia. (In Lower Cretaceous age from Victoria, Australia.
press ) Mem. Natl. Mus. Victoria 29:23-28.
Tyler, M. J., Davies, M. 1978a. Species groups Watson, G. F., Martin, A. A. 1973. Life history,
within the Australopapuan hylid frog genus Li- larval morphology and relationships of Australian
toria Tschudi. Australian J. Zool. Suppl. 63:1—47. Trans. Roy. Soc. South Aus-
leptodactylid frogs.
Tyler, M. J., Davies, M. 1978b. Phylogenetic rela- tralia 97:33-45.
tionships of Australian hyline and Neotropical Patterson, B. 1973. A fossil trionychid
Wood, R.,
phyllomedusine frogs of the family Hylidae. Her- turtlefrom South America. Breviora 415:1-10.
petologica 34:219-224.
Woodruff, D. S. 1972. Australian anuran chromo-
Tyler, M. J., Davies, M., King, M. 1978. The Aus-
some numbers. Herpetol. Rev. 4:208.
tralian frog Chiroleptes dahlii Boulenger: Its sys-
Wopfner, H., Twidale, C. R. 1967. Geomorpho-
tematic position, morphology, chromosomes and
distribution. Trans. Roy. Soc. South Australia logical history of the Lake Eyre Basin, pp. 119-
102:17-23.
143 in Jennings, J. N., Mabrutt, J. A. (eds.).
Landform studies from Australia and New Guinea.
Tyler, M. J., Martin, A. A. 1975. Australian lepto-
Australian National Univ. Press, Canberra, 434 p.
dactylid frogs of the Cyclorana australis complex.
Ibid. 99:93-99. Zweifel, R. G. 1969. Frogs of the genus Platyman-
tis ( Ranidae ) in New Guinea, with the descrip-
Tyler, M.
J., Martin, A. A. 1977. Taxonomic
studies of some Australian leptodactylid frogs of tion of a new species. Amer. Mus. Novit. (2374):
the genus Cyclorana Steindacfmer. Rec. South 1-19.
Australian Mus. 17:261-276. Zweifel, R. G. 1971. Results of the Archbold Ex-
Tyler, M. J., Rorerts, J. D. 1973. Noteworthy peditions. No. 96. Relationships and distribution
range extensions for some South Australian frogs. of tliomsoni, a microhylid frog of
Genyophryne
South Australian Nat. 99:143-147. New Guinea. Ibid. (2469): 1-13.
Tyler, M. J., Martin, A. A., Davies, M. 1979. Biol- Zweifel, R. G. 1972. Results of the Archbold Ex-
ogy and systematics of a new limnodynastine genus peditions. No. 97. A revision of the frogs of the
(Anura: Leptodactylidae) from northwestern subfamily Asterophryinae Familv Microhvlidae.
Australia. Australian J. Zool. 27 ( 1 ) : 135-150. Bull. Amer. Mus. Nat. Hist. 148:411-546.
5. Quaternary Biogeography of Tropical Lowland
South America
Jiirgen Haffer
Tommesweg # 60
4300 Essen-1
West Germany
107
108 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fig. 5:1. Distribution of humid tropical lowland forest and location of rainfall centers in Middle and
South America. Explanations: Shaded = humid forest, often semideciduous around savanna regions. Hatched
vertically
= areas receiving over 2500 mm of rain per year. Solid = Andean Cordilleras and Middle American
mountains of more than 2000 m elevation. Heavy dashed lines delimit the dry transverse zone of lower Ama-
zonia characterized by numerous isolated savanna enclaves. Letters designate areas of paleoecological research.
(See text for details.)
Distribution de selva humeda y position de centros de Uuvia en las tierras bajas tropicales de Centro y Sud
America. Explicaciones: Matizado = selva humeda; frecuentemente selva semidecidua alrededor de regiones
de savanas. Rayado vertical = areas recibiendo mas que 2500 mm de Uuvia anuales. Negro = Cordilleras
andinas y montafias de Centro America con alturas mayores que 2000 m. Lineas rayadas anchas delimitan la
zona seca transversal de Amazonia baja caracterizada por numerosas savanas aisladas. Las letras indican las
areas de investigaciones paleoecologicas. (Ver el texto para detalles.)
110 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
(Rondonia. Brasil)
(Llanos Onerrfoles .Gotomtoo ) iRupsKn. Guyana)
I0 2O3O4O5O6O7O809O KDO%
0-H00
21601
23401
wt
i^ Eem of tie tropical foresf
.'. ,
[
Maunna I
50 100%^°
G/amneae Cyperoceoe
,
B
Fig. 5:2. Pollen diagrams from central and northern South America (after van der Hammen, 1974). Ex-
B. Scheme of the succession of
planations: A. Pollen diagram from Capoeira and Kati ia, Rondonia, Brasil.
in the
vegetation types (as reflected in the pollen diagram) during an interglacial-glacial-interglacial cycle
present coastal plain of Guyana-Surinam. C. Pollen diagrams of the late Pleistocene and
Holocene from
areas at present covered with savanna vegetation ( Llan os Orientales, Colombia; Rupununi, Guyana). Four-
digit numbers indicate years before present.
Diagramas palinologicos de la parte central y septe ntrional de Sudamerica (segun van dcr Hammen, 1974).
Explicaciones: A. Diagrama paiinologico de Capoeira y Katira, Rondonia, Brasil. B. Esquema de sucesion
de
la vegetation (segun diagramas palinologicos) durante un ciclo interglacial-glacial-interglatial en la planitie
costanera de Guyana-Surinam. C. Diagramas palinolo gicos del Pleistoceno tardio y Holoceno de regiones hoy
cubiertas por savanas (Llanos orientates de Colombia; Rupununi, Guyana).
ested during dry climatic periods forming iso- icalSouth America (see earlier reviews by
lated "forest refugia." Sombroek, 1966; Haffer, 1969, 1974; Vanzolini
During Quaternary periods of raised sea and Williams, 1970; Fittkau, 1974; van der
level ( interglacials ) extensive areas of the , Hammen, 1974; Gamer, 1974, 1975; Raven
central and upper Amazonian lowlands were and Axelrod, 1974; Brown, 1977a,b).
converted into shallow inland lakes. Sioli Palynological data (Figs. 5:2-3). The
—
(1957) and (1976) discussed certain Irion pollen contents of bore hole samples from
aspects of the history of river sedimentation Katira Creek, 120 km southeast of Porto
and erosion as related to sea level fluctuations. Velho (Rondonia, upper Rio Madeira, Brasil;
The evidence for the above generalizations O of Fig. 5:1), indicate that in this rainforest
about Pleistocene climatic-vegetational fluc- region the forest was temporarily replaced by
tuations in tropical lowland South America open savanna vegetation, presumably during
is derived from palynological, geomorpho- a late Quaternary period of drought ( van der
logical, geological, and climatological studies Hammen, 1972, 1974; Absy and van der Ham-
conducted in widely scattered parts of trop- men, 1976 ) Pollen profiles from lake deposits
.
1979 HAFFER: QUATERNARY OF TROPICAL LOWLANDS 111
Displacement of Fluctuations of
vegetation zones lake level
5000-
10,000-
15,000
20,000'
25,000?-
30,000?-
112 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Andes (L and P of Fig. 5:1), during the naus region) that were deposited during the
Quaternary has been interpreted by Garner last glacialstage and in postglacial time. The
(1959, 1966, 1968) on the basis of a sequence authors did not identify indications of peri-
of alternating humid and dry periods. The odic influence of dry savanna climate in these
Caroni region in southern Venezuela was sediment cores. Possibly, fairly humid condi-
unforested repeatedly when severe planation tions and broad gallery forests persisted along
and aggradation occurred (Fig. 5:4). Forests the Amazon River and its tributaries during
have returned to this area comparatively re- (late) Pleistocene dry climatic phases. Irion
cently. Lateritic crusts, stone lines, dissected and Absy (1978) further felt that the surface
gravel deposits, settled dunes, or the morphol- land forms in central Amazonia do not favor
ogy of the landscape of parts of central and an interpretation of repeated climatic reversals
northern Brasil indicated to Barbosa (1958), in that area during the Quaternary. Hope-
Ab'Saber (1957, 1967), Cole (1960) and fully,sedimentological and palynological ma-
Hammer (1963:299, 305) that the present terial will soon become available from the
humid climate was preceded by a more arid terra firme regions between large Amazonian
climatic phase. Sarma (1974) obtained cor- rivers. These regions are more critical for
responding results in coastal Ecuador. testing the theory of Pleistocene vegetational
An interpretive map of South America by fluctuations in Amazonia than the broad river
Ab'Saber (1977) depicts the distribution of valleys where probably more or less extensive
major vegetation types during the dry climatic gallery forests remained during dry climatic
phase of the last glacial period (18,000- periods.
13,000 y.b.p. ) as reconstructed by the author —
Climatological data. Regional variation
from primarily geomorphological data (e.g., of annual rainfall in central South America is
pediments, fluvial terraces, stone lines, paleo- mainly determined by the surface relief which
sols, iron crusts) gathered over 20 years. The had originated at least in Late Tertiary time,
lowland tropical forests are shown to be re- although some uplifting of the Andes and the
stricted to a number of isolated areas that Guianan and Brasilian shields probably did
agree in size and location to refugia proposed occur later during the Quaternary. Precipi-
by several biogeographers as discussed below. tation is increased near the Atlantic coast, on
Geological data. —
The occurrence of exposed slopes of the Guianan mountains, and
arkosic sands (30-60% felspars) of Late Wis- near the Andes. By contrast, the total rain-
consin age in the shelf sediments off north- fall decreases on the leeward side of moun-
eastern South America led Damuth & Fair- tains and in a zone across lower Amazonia.
bridge (1970) to postulate that the climate Even though climatic patterns during Pleisto-
of tropical lowland South America during the cene glacial and interglacial periods probably
last glacial phase was much drier than
today.
differed considerably from the present situa-
Milliman et al. (1975) and Irion (1976) chal- tion, appears reasonable to assume that,
it
lenged this conclusion and suggested the during arid phases, the forests in Amazonia
possibility of an Andean or central Ama- disappeared first from areas that presently
zonian origin of the arkosic sands, which receive the least amount of rain per year and,
would diminish their paleoclimatic signifi- conversely, survived longest in areas of heavy
cance. Further data and interpretations of rainfall. Cool dry periods occurred 20,000-
Quaternary continental shelf deposits in trop- 13,000 y.b.p., 11,000-9,500 y.b.p., and 3,500-
be of interest. Studies of
ical latitudes will 2,800 y.b.p. (van Geel and van der Hammen,
sea-bottom cores taken in the Caribbean Sea 1973; Vanzolini and Williams, 1970; Heine,
and of foraminiferal assemblages and eolian 1974 ) The last two periods probably did not
.
biogenic detritus indicate aridity during cause so much forest fragmentation as the
temperature minima (Bonatti and Gartner, earlier which coincided with the last
one,
1973; Prell, 1973; Parmenter and Folger, glacial period. Zonneveld (1968) discussed
1974). the climatic changes in northern South Amer-
Irion and Absy (1978) studied lake sedi- ica.
at a number of localities in several South the origin of the large diversity of life in the
American countries, especially in Argentina, Neotropics and mainly refer to the question
southeastern Brasil, and Ecuador, indicate the of which processes led to the separation of
existence of a peculiar Neotropical mammal small populations in the vast tropical low-
fauna in forest and nonforest habitats during lands. Spatial isolation was known since the
the Tertiary and Quaternary periods (Patter- studies of Charles Darwin, Moritz Wagner
son and Pascual, 1968; Paula Couto, 1975). and Ernst Mayr ( 1942, 1963 )
as the prerequi-
1
This fauna had evolved in geographic isola- site of
geographic (allopatric) speciation.
tion during the Mesozoic and Tertiary from Agents causing the isolation of daughter pop-
an ancestral mammal fauna of western Gond- ulations from a previously more widespread
wanaland (Fittkau, 1969; Fooden, 1972) and parent population in tropical South America
came in direct contact with the North Amer- are thought to be 1 ) paleogeographic changes
ican mammal fauna in the Pliocene, when the in the distribution of land and sea, 2) the
Central American isthmus was closed ( review development the system of broad Ama-
of
bird remains are known from northwestern under the changing climate of the past. These
Peru and western Ecuador (Campbell, 1976, three theories are discussed in some detail
below.
1979). The famous fossil deposits in the lime-
stone caverns at Lagoa Santa, Minas Gerais —
Constancy theory. No major speciation
(Brasil) may be only subrecent in age. or subspeciation occurred in the tropical low-
Despite the outstanding general scientific lands during the Quaternary, when equable
significance of the South American fossil environmental conditions were fairly constant
faunas, their contribution toward elucidating and continued unchanged from the Tertiary
biogeographical events in the tropical low- into Recent times (or, at least, climatic
lands during the Quaternary is still limited. changes were of minor importance). Extant
The notable extinction of mammals in South Neotropical species and subspecies, as well as
America during the late Pleistocene, mainly their distribution patterns, are thought to
of the larger, more specialized herbivores and have originated during the Tertiary when
their predators, has been attributed to cli- populations were repeatedly isolated and re-
matic-vegetational changes, to hunting pres- connected owing to paleogeographic changes
sure by prehistoric man, or to both (Martin in the distribution of land and sea caused by
and Wright, 1967; Axelrod, 1967). Hersh- epeirogenic and orogenic events (e.g., uplift
kovitz (1969:27) stated that ".postglacial
. . of the shield areas and the Andes, respec-
decrease in humidity, aggravated by drastic tively, or foundering of forelands off the
seasonal fluctuations, might well have been present coast lines ) Tertiary species presum-
.
the more important factor in the extermina- ably survived relatively unchanged until to-
tion of largerMiddle and South American day, although some impoverishment of the
mammals." However, the problem remains fauna may have occurred. This theory was
unsolved, and the discussion of late Pleisto- generally accepted during early parts of this
cene extinctions continues (MacNeish, 1976; century and, among recent authors, by Croizat
Grayson, 1977; Webb, 1978). (1958, 1976) for the Neotropical fauna, Ems-
ley (1965) for butterflies, and Hershkovitz
(1969) for certain mammals. This interpreta-
QUATERNARY BIOGEOGRAPHY OF tion may also be called the "Tertiary island
TROPICAL LOWLAND SOUTH AMERICA
theory." According to this theory, differences
between the forest faunas to the west and east allied species of animals [Wallace, 1853;
of the Andes presumably originated from the Bates 1864 (for various groups of animals);
uplift of the Andes, which split a pre-existing, Sclater and Salvin, 1867; Hellmayr, 1910,
widespread lowland fauna into a trans- 1912; Snethlage, 1913; Mayr, 1942:228, and
Andean and a cis-Andean (Amazonian) por- Sick, 1967 (for birds); Hershkovitz, 1968,
tion (Chapman, 1917:89; Croizat, 1976:766). 1969 ( for monkeys ) ] However, many of these
.
It appears certain that the isolation of species circumvent large streams or, in the
to case of subspecies, hybridize in the head-
large populations owing Tertiary paleo-
and also water regions where the narrowing rivers
geographic climatic-vegetational
cease to be barriers to dispersal ( Haffer, 1969,
changes did play a major role in establishing
basic distributional and evolutionary patterns 1974, 1978). The distribution of numerous
of the Neotropical fauna and flora at higher other species is unaffected by Amazonian
taxonomic levels of families and genera. How- rivers. The river theory also leaves unex-
ever, the dataon Pleistocene fossils and on plained howa forest of the size, or nearly the
environmental fluctuations, reviewed above, size, of the present Amazon forest could have
indicate conspicuous changes in the composi- existed without major rivers. The develop-
tion and distribution of the Neotropical biotas ment of the great forest and of its large rivers
during the Quaternary period, probably lead- probably was one interrelated process. The
forest cannot be envisioned without the rivers
ing to extensive evolutionary changes at the
and vice versa.
species and subspecies levels in many groups
after the close of the Tertiary. This, of course, flooding of the Amazon Valley
Partial
does not exclude the possibility of members of during interglacial periods and shifting river
groups in the present fauna, espe-
less plastic courses in this region indeed may have iso-
cially among invertebrates, having survived lated some local animal populations on tem-
relatively unchanged since Tertiary times. porary islands, where they may have differen-
Because the Andes arose in the form of strings tiated as subspecies or even species (Willis,
of growing islands from a marine geosynclinal 1969:393). The same applies to islands on
basin, there was no pre-Andean, continuous the continental shelves of South America,
and widespread lowland fauna occupying which were connected with the mainland
what was later to become the Andes and their during glacial periods of lowered world sea
forelands. Moreover, the separation of trans- level and isolated during interglacial periods
and cis-Andean lowlands and their faunas was (Miiller, 1970; Vanzolini, 1973; Reboucas-
not complete; thus frequent exchanges of Spieker, 1974).
these faunas through the Caribbean lowlands —
Refuge theory. Forest and nonforest
of northern Colombia north of the Andes were biomes alternately broke into isolated blocks
possible during the Quaternary. and expanded and coalesced under the chang-
River theory. —This theory is an attempt ing humid to arid climatic conditions of the
to explain the differentiation in the rich Ama- Quaternary period. According to this theory,
zonian biota. Uniform and widespread popu- plant and animal populations isolated in the
lations areassumed to have been divided into restricted forest and nonforest refugia during
two more subpopulations and effectively
or adverse climatic phases either became extinct,
isolated by the developing network of large survived without much change or, more often,
Amazonian rivers during the Early Quater- differentiated to the level of subspecies or
nary period. The isolated subpopulations de- species before they came into secondary con-
viated to the level of subspecies and species tact with previously conspecific populations
on opposite river banks. This theory has been of other refugia during a following favorable
invoked on different occasions to explain cer- expansive phase. If a population reached
tain situations but has never been proposed sexual and ecological isolation
neigh- from its
formally or quantitatively tested. Field nat- boring allies, it could disperse widely in the
uralists observed repeatedly that the Amazon now continuous habitat before its extensive
or one of its tributaries separate for some dis- range was fragmented during the next adverse
tance the ranges of numerous subspecies of climatic phase. Rivers in Amazonia probably
116 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
were not a causal factor of speciation ( except ports of neospecies should be disregarded."
perhaps in a few cases) but merely limited Similarly, Ballmann ( 1977 ) concluded that
or modified the dispersal of populations from most of the extant avian species originated
the forest refugia. during the Pleistocene. Vuilleumier (1975:
This theory proposes that the intensive 480) summarized the current consensus stat-
earlier processes of faunal differentiation dur- ing that ". the distribution patterns of taxa
. .
(Vanzolini, 1970, 1973; Vanzolini and Wil- taken place in the tropics. . . ."
liams, 1970), and vertebrates in general An analysis of forest and nonforest refugia
2
(Miiller, 1972, 1973 ). Subsequently, it has could be based on criteria derived from the
been corroborated by the results of studies on following evidence. 1) Direct evidence from
Amazonian Drosophila flies (Spassky et al., fossil records and from studies in various
1971;Winge, 1973), butterflies (Brown et al., other geoscience fields, such as palynology,
1974; Brown, 1976, 1977a,b), and trees pedology, and geomorphology. As shown
(Prance, 1973). The theory is essentially the above, the sparse data presently available do
refuge model developed to explain faunal dif- permit several general conclusions to be
ferentiation during the Pleistocene in northern drawn regarding vegetational fluctuations in
latitudes (Stresemann, 1919; Reinig, 1937, tropical lowland South America. However,
1938a,b). This model was applied later to it is still inadequate to map in detail the
Australia (Keast, 1959, 1961), Africa (Car- changing distribution of forest and nonforest
casson, 1964; Moreau, 1966; Kingdon, 1971) vegetation through time or to delineate accu-
and finally to the Neotropics. rately forest refugia of varying sizes that must
As to the
chronology of the speciation have existed during arid climatic phases.
process, recent authors agree that species may 2) Indirect evidence based on an historic in-
evolve under favorable conditions in geo- terpretation of biogeographical data concern-
logically short periods of time [i.e., during the ing the present flora and fauna. Because the
time span of Pleistocene climatic phases or forest and nonforest refugia are geological-
even shorter periods (Selander, 1971; Van- paleontological rather than biological phe-
zolini,1970, 1973)]. Brodkorb (1971:46 ff.) nomena, conclusions based on such indirect
stated in a paleornithological review: "The evidence alone should be formulated cau-
Quaternary was a period of almost constant tiously. 3 ) Combination of direct and indirect
change. The living species of birds arose at evidence derived from the available paleo-
various times during the Pleistocene epoch, ecological and fossil data, as well as from
.... Postglacial evolution has been mainly at extant faunas and floras. Given the present
the subspecific level. The evidence con-
. . .
state of knowledge this method appears to be
tinues to amass that the most promising today. It is reviewed in
all living species of birds
arose during the Quaternary the following section in some detail, for the
scarcely . . .
more than 13 [living species of birds] have results so far obtained readily explain many
been reported, tenuously, from the Tertiary. biogeographic phenomena observed in the
It was formerly held that neospecies Neotropical Region.
origi-
nated in the Pliocene or even earlier, but The authors who advocated the Tertiary
little evidence exists to support this theory. . . . island theory, the Quaternary river theory,
Pending critical re-evaluation, all Pliocene re- and the Quaternary refuge theory determined,
as a first step, general patterns of vicariance
!
Several earlier authors briefly suggested in general in the continental biotas (i.e., the patterns of
terms climatic-vegetational fluctuations as a cause
allopatric differentiation of species and sub-
of Pleistocene faunal differentiation without dis-
cussing details (Fox, 1949; Turner, 1965; Miiller, species, including phenomena of secondary
1968); see also the review by Vuilleumier (1971). contact between vicariant populations). The
1979 HAFFER: QUATERNARY OF TROPICAL LOWLANDS 117
events, and .
subsequent dispersal modify-
. . species and/or subspecies with similar eco-
ing earlier vicariant patterns." The under- logic preferences; area of high endemism. A
lying causes of allopatric differentiation
in the purely descriptive term. Synonyms are core
3
lowlands during the Quaternary area of distribution, core area of endemism.
Neotropical
Mapped by superimposing the ranges of spe-
probably were processes of alternating frag-
mentation and coalescence of vegetation cies in thegroup considered and drawing con-
zones [i.e., "small-scale" vicariance events as tour lines of equal numbers of species in the
more or less constant during adverse climatic and partly fused in the past. Such suture
and plant populations. This is an interpretive existence of major ecological barriers, al-
nonbiological term referring to climatological, though the barriers themselves have long
since disappeared. In several cases horizontal
pedological, palynological and other geosci-
ence phenomena. The term "refuge" is some- shifts of the contact zones may have occurred
what inappropriate as animals and plants did after contact was established. For this rea-
not actively seek shelter in the refuge. Rather, son, conclusions about the presumed location
remnant populations remained hi the remnant of former barrier zones should be based on
habitats of varying size where they did or did several or many contact zones so as to mini-
not survive under fairly constant ecological mize potential errors involved. (Steps 1 and
conditions. 2 are independent of each other and may be
reversed.
A Method of Historic-Biogeographical
)
'^\ °fl •
o'.
U
/ Vol
Vi
gV \
o
s/
0'. \ -
1. 2. ^~-;-c J?
/
AW
/ \ \»o,
}X I
l|
\
120 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
expression and the location of separating hy- muchas veces se juntan en determinadas rcgiones.
brid belts more or less coincide in the differ- El numero de jormas difcrenciadas de las especies
superpuestas estd indicado (n
= 11 6 12). El nivel
ent species (Fig. 5:6). Contour lines illustrate
promedio de hibridizacion en las poblaciones de
varying hybrid levels of populations around especies combinadas estd mapcado por tineas de con-
the centers. Analyses also may be based on lomo (0 = "puro"). Una barrera para el flujo
the population structure of a single species, genetico (un gran rio 6 una montana) estd indicada
and distribution maps of several individual esquematicamente por una faja ncgra. ha matjoria
de las poblaciones en ambos lados de la barrera son
characters may be prepared (e.g., Vanzolini
"puras"; un flujo genetico restringuido ocurrc sola-
and Williams, 1970, for the lizard Anolis chrij- mente entre unas pocas poblaciones separadas.
solepis). The character maps may then be
superimposed and contoured in a similar
center probably was a center of dispersal in
manner those of subspecies. Coinciding
as the past, we proceed to postulate the approxi-
areas of uniform characters with low variabil- mate location, size and shape of the corre-
ity often cluster in core areas that are sep-
sponding refuge area of forest or nonforest
arated by zones of high character variability
vegetation within the central portion of the
coinciding with hybrid belts between sub- center, taking into consideration all available
species ranges. geoscience data from the region, however de-
Step 4: of the refngia related to
Mapping data may be. Medium elevation
ficient these
—
the dispersal centers. Ideally, this can be areas with high rainfall near the base of
accomplished using geoscience data exclu- mountains or plateau regions are prime candi-
sively. Having established on the basis of dates for forest refugia whose postulated ex-
zoogeographical data that a given distribution tent during the maximum of the arid climatic
1979 HAFFER: QUATERNARY OF TROPICAL LOWLANDS 121
period remains highly speculative. The model probably permitted a direct connection of the
assumes that the species and subspecies char- Amazonian fauna and flora with those of the
acterizing a given center were confined to the Atlantic forest in eastern Brasil across the
postulated refuge prior to dispersal and prior central Brasilian Plateau and along the coastal
to establishing secondary contact with other region of northeastern Brasil, where scattered
forms spreading from distant centers. Obvi- forests are still preserved on isolated moun-
ously, the former existence and changing size tains (Muller, 1973; Haffer, 1974; Brown,
of the refugia ultimately can be traced only 1976, 1977b). 3) The cis- Andean and trans-
through detailed palynological, pedological, Andean forest biotas probably were connected
and geomorphological studies rather than repeatedly both north of the Andes in the
through zoogeographical analyses. After all, Caribbean lowlands of northern Colombia
the refugia are geological-paleoclimatological, and through the north Peruvian Andes from
not biological, phenomena. However, as long the upper Maranon Valley via the low Por-
as only scattered geoscience data are avail- culla Pass so as to reach the forested Pacific
able, zoogeographical analyses as outlined lowlands of Colombia from the south (Chap-
above will help our understanding of biotic man, 1926; Haffer, 1967, 1975; Muller, 1973).
differentiation in lowland tropical South 4) The relations between the Middle and
America during the Quaternary. In general, South American nonforest bird faunas
the locations of distribution centers of the are less pronounced than those of the forest
Amazonian forest biota correlate well with faunas of these two areas (Mayr, 1964).
the locations of forest refugia tentatively de- 5) There are numerous conspecific popula-
rived from rainfall, relief and geoscience data tions inhabiting the nonforest regions of
alone (p. 10S), thus strengthening an histor- northern South America and central Brasil,
ical interpretation of the biogeographical core respectively, being separated by the entire
areas (Simpson and Haffer, 1978; Brown and width of the Amazon forest. Because their
Ab'Saber, 1979). dispersal under present climatic-vegetational
Endler (1977) suggested that parapatric conditions is
highly unlikely, a rather recent
speciation occurs frequently in nature and
direct communication of nonforest faunas
that the hybrid zones in Amazonia might ac- across Amazonia during one or more arid cli-
be zones of primary intergradation matic periods seems probable. This theory
tually
caused by strong environmental gradients. also explains the close of the
relationship
However, the geoscience data reviewed above fauna and flora of the Amazonian savanna
enclaves with those of the nonforest regions
(but not discussed by Endler) favor the inter-
to the north and south of Amazonia ( Hueck,
pretation of allopatric rather than parapatric
1966; Haffer, 1969; Muller, 1973).
speciation for the many hybridizing and non-
hybridizing populations of plants and animals The authors
of the studies reviewed below
in contact. The Amazonian forest refugia areaware of the tentative nature of the results
are geological-paleoclimatical, not biological obtained and of the suggestions made. More-
phenomena. over, the nascent theory of ecological refugia
during the Quaternary, like any theory,
cannot be proven but only disproven by the
Biogeographical Studies on the results of additional studies on the paleo-
Forest Fauna and Flora
climatology, vegetational history, and zooge-
Some of the major regional conclusions re- ography of South America. I wish to empha-
the size with Meggers (1977) that the efforts
garding Quaternary differentiation and by
dispersal of Neotropical biotas are summar- biologists and archeologists at present are not
ized, as follows: 1) Extensive speciation took more than a search for correlations and pat-
place in many groups of Neotropical animals terns useful as guides for investigation. The
and plants that were repeatedly isolated in strikingly similar results as to the basic pat-
refugia and later expanded their ranges dur- terns of differentiation in various unrelated
ing favorable expansive phases. 2) During groups of Neotropical animals tend to sup-
humid climatic periods, extensive forests port the refuge theory and justify continued
122 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
for more details). A number of other species hastante localizadas forman los centros de distribu-
tion (A-F; abajo) cada uno de cllos formado pot
have an irregular, spotty distribution or are It) a 50 especies. Adaptado de Haffer, 1974, Fig.
known from only a single locality. The 9.13 y Haffer, 1978. Otros grupos de especies en-
mapped contact zones in Amazonia (Fig. 5:7) demicas caracterizan la selva atldntica de Brasil meri-
are mostly located between distributional core dional y la sclva transandina de Sudamerica norocci-
dental.
areas which probably functioned as centers of
dispersal. The complementarity of contact upper and lower Amazonian centers are in
zones and distribution centers is less well de- contact within the area of the "weak" Imeri
veloped in central Amazonia, where a num- and Rondonia centers (characterized by com-
ber of birds spreading from the species-rich paratively few species).
1979 HAFFER: QUATERNARY OF TROPICAL LOWLANDS 123
Based on geosciencc data, Quaternary for- broad centers of dispersal but did not sug-
est refugia in Amazonia probably were lo- fest specific refugia. A comparison shows that
cated near the windward base of hilly ranges the boundaries of Miiller's centers rarely co-
and mountains, such as the northern margin incide with those of avifaunal core areas but
of the Brasilian Plateau, the eastern base of fall in the peripheral gradient of decreasing
the Andes, and the northern base of the moun- species numbers or coincide with river bar-
tains in the interior Guianas. These areas riers. Miiller (1973) did not state the criteria
correlate well with the location of the zoo- he used to delimit the various broad centers,
geographically defined distribution centers which on his maps are frequently separated
which, therefore, are assumed to indicate the only by narrow corridors.
approximate location of Quaternary forest In recent years, several herpetologists ac-
refugia. Additional refugia for the avifauna cepted the notion of a Pleistocene origin of
(Fig. 5:8) based on similar geological and numerous species and subspecies of Neo-
zoogeographical criteria have been proposed tropical frogs, lizards and snakes in forest
for the Atlantic forest of southeastern Brasil Duellman (1972, 1978), Heyer
refugia:
(Miiller, 1973; Jackson, 1978) and for the (1973), Duellman and Crump (1974), and
trans-Andean forest regionin western Colom- Silverstone (1975, 1976) for certain Amazon-
bia and in Middle America (Haffer, 1967, ian and trans-Andean frogs, Gallardo ( 1965,
1969, 1974, 1975). A recent study of a com- 1972 ) and Lynch ( this volume ) for the South
plex Neotropical avian genus on the basis of American amphibian fauna generally, Dixon
the refuge concept is Fitzpatrick's ( 1976 ) (this volume) for Amazonian reptiles, Ech-
analysis of the Todirostrum flycatcher group; ternacht (1973) for Middle American lizards
also see the recent review by Dorst ( 1976 ) .
(Ameiva), Hoogmoed (1973, this volume)
—
Lizards and frogs. Analysing the popula- for the herpetofauna of the Guianas, Jackson
tion structure and character variation in the (1978) for two genera of eastern Brasilian
Amazonian Anolis chrysolepis group, Vanzo- iguanid lizards, and C. W. Myers (1973,
lini and Williams (1970) and Vanzolini 1974) for two genera of snakes. Dixon (this
(1970, 1973) arrived at conclusions regarding volume) and Lynch (this volume) mapped
the history of faunal differentiation during clusters of endemic reptile and amphibian
the Quaternary that are strikingly similar to species in upper and lower Amazonia. These
those reached by Haffer (1969, 1974). Sev- centers of distribution more or less coincide
eral extensive areas of uniform character ex- with similar centers established for the forest
pression (core areas) in Anolis chrysolepis avifauna. At this stage, a more quantitative
are separated by regions where complex char- treatment of available distributional data
acter variation suggests hybridization or in- would be desirable, such as the derivation of
trogression along zones of secondary contact. herpetological distribution centers by means
Vanzolini and Williams (1970) interpreted of contoured diversity maps of localized spe-
this situation as being the result of secondary cies clusters and a comparison of the location
intergradation of populations that had differ- of secondary contact zones with that of dis-
entiated in geographic isolation, and they as- persal centers and postulated refugia.
sumed changes in the distribution of forest in Insects. —The distribution and population
Amazonia during the course of climatic fluctu- structure of two groups of Neotropical insects
ations. These authors reconstructed several are relatively well known and permit initial
forest refugia around the periphery of Ama-
biogeographic analyses 1)
—Certain Dro-
zonia (Fig. 5:8), most of which coincide flies that have been collected and
sophila
closely with those proposed by Haffer ( 1969,
analysed extensively in the course of genetic
1974) for birds. Additional forest refugia and 2) butterflies of the
investigations, genus
probably determined the differentiation of Heliconius. Spassky et al. (1971) summarized
Neotropical forest reptiles and will be iden- the distribution of several semispecies and
when other species are studied in detail.
tified
closely related species of Drosophila. These
Miiller (1972, 1973) using mostly herpetolog- authors and Winge (1973) concluded that
ical and ornithological data described several the various forms may have originated in for-
124 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fig. 5:8. Location of presumed Quaternary forest refugia (hatched) in tropical lowland South America.
Explanations: 1. Reconstruction based on the distribution patterns of Neotropical birds ( Haffer, 1967, 1969,
1974). 2. Reconstruction based on the population structure of Amazonian lizards (Anolis chnjsolcpis group,
Vanzolini and Williams, 1970, Vanzolini, 1970; wide hatches indicate core areas). 3. Reconstruction based
on the distribution patterns of subspecies and species of Heliconius butterflies ( Brown et al., 1974, Brown
1977a,b). 4. Reconstruction based on the distribution patterns of four families of Amazonian trees (Prance,
1973).
Position de presuntos refugios cuaternarios de selva humeda (rayado) en las Metros hajas tropicales de
Sudamerica. Explicaciones: 1. Reconstruction basada en los patrones de distribution de aves neotropicales
(Haffer, 1967, 1969, 1974). 2. Reconstruction basada en la estructura de poblaciones de lagaitos amazonicos
(grupo de Anolis chrysolepis, Vanzolini y Williams, 1970, Vanzolini, 1970; Hneas verticales indican areas nu-
clearcs). 3. Reconstruction basada en la distribution de subespicies y espeties <le mariposas del gSnero Heli-
conius (Brown et al, 1974, Brown, 1977a,b). 4. Reconstruction basada en la distribution de cuatro jamilias de
drboles amazonicos (Prance, 1973).
1979 HAFFER: QUATERNARY OF TROPICAL LOWLANDS 125
est refugia as proposed for the evolution of postulating the location and shape of forest
birds and reptiles. The colorful Heliconius refugia in the central portion of the centers
butterflies inhabit the entire humid Neotropi- of evolution. The majority of the 40 forest
cal Region. They are famous for their com- refugia in Middle and South America recon-
plicated Muellerian mimicry rings compris- structed from data on Heliconius species (Fig.
ing many races that are strikingly different 5:8) coincide with refugia postulated for
phenotypically. Rich material of these but- other groups. In several instances discrepan-
terflies in private and public collections per- cies exist, as expected. Brown (pers.
may be
mits fairly detailed analyses. On the basis of coram. ) currently is preparing more detailed
genetic and distributional studies of several contour maps defining core areas on the basis
species of Heliconius, Turner (1965, 1971) of superimposed species maps, which illus-
felt that climatic-vegetational fluctuations of trate by contour lines the hybrid levels of
the Pleistocene might explain the origin of local butterfly populations (Brown and
secondary range overlap of related forms and Ab'Saber, 1979).
repeated speciation, which led to the complex Plants. — D. R. Simpson (1972) supported
pattern of diverse, related and mimicking, the refuge theory, in particular for the forest
species in this genus. Additional detailed
region of eastern Peru, on evidence of genera
biogeographical studies of the genus Heli- of Rubiaceae. Parallels between the distribu-
conius and other groups involved in mimicry tion of certain Neotropical palms and birds
rings by Brown et al. (1974), Brown (1976, also exist (Moore, 1973). On the basis of a
1977a,b), Turner (1976a,b, 1977) and Shep- phytogeographic study of several genera of
pard et al. ( 1978 ) have greatly extended this Amazon forest trees in the families Chryso-
initial analysis (Fig. 5:9), see also the work
balanaceae, Dichapetalaceae, and Caryocara-
by Blandin (1977) and Descimon (1977). ceae. Prance ( 1973 also supported the theory
)
Conclusions are based on a revision of 162 of plant speciation in Quaternary forest re-
species of Heliconiini and Ithomiinae, family fugia of tropical South America (see also
Nymphalidae, representing a total of 905 Forere, 1976). In each of the tree genera
geographically differentiated forms. Brown studied, there are several areas of repeated
(unpubl.) analysed the distribution and dif- endemism coinciding with the refuges pro-
ferentiation of other groups of Neotropical
posed (Figs. 5:8, 5:10). Hopefully, a quan-
butterflies (Morpho, Phycoides, Callicorini, titative derivation of the postulated refugia
Anaea, Parities) and obtained results that are on the basis of contoured plant species diver-
similar to those on the species of Heliconius.
sity maps and, if available, the location of
A number of very different subspecies occupy contact zones will be published when phyto-
forest areas of varying size and are connected
geographical data on additional groups of
by zones of hybridization of varying widths. Amazon plants become available.
The method of analysis was, as follows. By
The distribution patterns of two other
comparing the location of hybrid zones and
groups of Amazon —members of
trees the
Leguminosae and Memecyleae— do not
the ranges of differentiated forms (subspe-
clear-
cies) in the various species, these entomolo-
ly support the interpretation of speciation in
gists derived centers of endemism (based on restricted refugia; yet they do not negate the
8 to 50 local differentiated forms of the
species possibility (Langenheim et al, 1973; Mor-
analysed). However, Brown (1977a,b) did ley, 1975 ). The ranges of most of these plants
not explain details of delimination of the cen- are quite extensive and, as pointed out by
ters of endemism, i.e., which criteria he used the authors cited above, expansion from re-
to determine the widely may have obscured refuge boundaries.
varying width of the fugia
areas separating the centers. However, four species of Mouriri occupy re-
Considering geo-
morphological, palynological, and climatolog- stricted ranges in northeastern Amazonia; this
ical data, Brown et al. (1974) and Brown correlates with refugia proposed by Prance
(1976, 1977a,b) continued their analysis by (1973). Nevertheless, Morley (1975) felt that
126 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fig. 5:9. Distribution patterns of certain butterflies (A,B), lizards (C), and frogs (D) in
tropical lowland South America. Explanations: A. Distribution of geographical races of
Heliconius crato shown by different patterns. Hybrid zones are omitted. Overlap of shading
indicates variable populations (after Brown et al., 1974). B. Distribution of geographical races
of Hypothyris ninonia shown by different symbols; distribution in Amazonia is probably con-
tinuous (after Brown, 1977b). C. Anolis chrysolepis, females; summary of geographic differ-
entiation, number of scales across snout (after Vanzolini and Williams, 1970, map 6, p. 166).
D. Distribution of several poison-arrow (or poison-dart) frogs of the genus Dendrobates :
Fig. 5:10. Distribution of certain Neotropical monkeys (A), plants (B), birds (C), and
of the Ge-Pano-Carib ancestral language stock among Indian tribes (D). Explanations:
A. Distribution of tamarins, genus Saguinus: Saguinus sp. (1), S. inustus (2), S. midas (3),
S. leucopus (4), S. o. oedipus (5a), S. o. geoffroyi (5b) (after Hershkovitz, 1969). B. Distri-
bution of various species of Caryocar: C. amygdaliferum (1), C. nuciferum (2), C. gracile
(3), C. amygdaliforme (4), C. pallidum (5), C. dentatum (6), C. eduZe (7) (after Prance,
1973). C. Distribution of the Green Jacamars, Galbida galbula superspecies: G. ruficauda
melanogenia (1), also widespread in Middle America G. r. ruficauda (2), G. r. rufoviridis (3),
G. galbula (4), G. cyanescens (5), G. tombacea (6), G. pastazae (7) (after Hafter, 1974).
D. Distribution of the Ge-Pano-Carib linguistic stock (shaded) and suggested dispersal routes
of the nonforest fauna from central Brasil into and across Amazonia (after Meggers, 1977).
Distribution de ciertos memos neotropicales (A), plant as (B), aves (C) y del grupo lin-
guistico ancestral Ge-Pano-Carib cntre tribus de indios (D). Explicacioncs: A. Distribution
de monos del genero Saguinus: Saguinus sp. ( 1), S. inustus (2), S. midas (3), S. leocopus (4),
S. o. oedipus (5a), S. o. geoffroyi (5b) (segun Hershkovitz, 1969). B. Distribution de varias
the distribution of the species he studied can oedipus. A critical appraisal of the differen-
be largely accounted for on the basis of tiation patterns in these and other groups of
present conditions ( "although with some diffi- Neotropical mammals in the light of the
culty"). Clearly, numerous additional groups refuge theoiy is needed. Hershkovitz (1969,
of Neotropical plants must be studied in 1972) and Moynihan (1976) did not attach
order to gather the basic data needed for a major importance to the refuge theoiy (but
critical appraisal of the refuge theory from a see the remarks by Avila-Pires, 1974:178 and
botanical point of view. Kinzey and Gentry, 1979). In the case of
—
Mammals. The patterns of distribution African forest mammals, Kingdon (1971:67)
of several groups of Neotropical forest mam- stated: ". with confidence that past cli-
. .
mals and their differentiation at the subspe- matic changes, repeatedly isolating and then
cies and species level ( Hershkovitz, 196S, reuniting the forest areas of Africa have been
1969, 1972, 1978; Avila-Pires, 1974) appear to a major mechanism in the speciation of forest
be similar to patterns observed in birds and mammals. Monkeys, duikers, squirrels and
insects. Therefore, suggest for these exam-
I other rodents show morphological differences
ples of mammals an historic interpretation of that frequently coincide with the forest
differentiation on the basis of Quaternary refuges."
climatic-vegetational fluctuations. This would Prehistoric man. —
In an interesting appli-
seem to offer a fairly straight-forward expla- cation of the biogeographical model of di-
nation of speciation in cases like that of the versification to cultural distributions in
trop-
tamarin monkeys (Saguinus) (Fig. 5:10). ical lowland South America, Meggers (1977:
Several species of tamarins coexist in upper 292) stated that ". . . some of the characteris-
Amazonia. The only lower Amazonian species tics observed by biologists can be discerned,
(S. midas) may have originated in the Guiana among them heterogeneity, disjunct distribu-
refuge and later established secondary con- tions, and widespread
dispersals." Prehistoric
tact with the upper Amazonian species in the man probably arrived in South America dur-
upper Rio Negro region (S. "inustus") and ing the late Pleistocene (MacNeish, 1976)
also crossed the lower Rio Amazonas south- and witnessed
several climatic reversals.
ward. Another example is the several species Therefore, certain cultural adaptations and
of Capuchin monkeys ( Cebus albifrons super- dispersal patterns may be explained on the
species). The trans-Andean Cebus capucin- basis of vegetational fluctuations. The
pres-
us, the upper Amazonian C. albifrons and the ent distribution of the Ge-Pano-Carib ances-
Guianan C. griseus divide the Neotropical tral language stock
(Fig. 5:10) correlates
forests among themselves without geograph- closely with the routes postulated for the
ical overlap in the manner of many avian intrusion of nonforest fauna into and across
superspecies Kinzey and Gentry (1979) re- Amazonia from open habitats in central Rrasil
cently interpreted the distributional history (Meggers, 1977).
of the Amazonian titi
monkeys Callicebus Fish. —
No detailed studies of the pattern
moloch and C. torquatus on the basis of of geographic differentiation in the
aquatic
Pleistocene forest refugia. vertebrate and invertebrate faunas of tropical
A number of contact zones lowland South America are available. How-
between re-
lated mammals are clustered in northwestern ever, it appears that the strong climatic-vege-
Colombia near the junction of the Panaman- tationalchanges and sea level fluctuations of
ian isthmus with the South American main- the Quaternary had numerous effects on the
land and could have been established when aquatic fauna of Neotropical rivers, as indi-
sea level fell in postglacial time and/ or cated by the comment of Lowe-McConnell
when
forests reoccupied this region after a post- (1975:261).
glacial or late Pleistocene period of drought— In the rivers of South America the
adaptive
the tapirs Tapirus bairdii/T. terrestris, the radiations of the characoids and various catfish
spider monkeys Ateles fusciceps/A. geoffroyi, groups are as remarkable as those of the cich-
lids in the African Great Lakes. Differentia-
the howler monkeys Alouatta villosa/A. seni-
tion has proceeded much further in these
culus, the tamarins Saguinus o. geoffroyi/S. o. groups, certain families or subfamilies now
1979 HAFFER: QUATERNARY OF TROPICAL LOWLANDS 129
having particular feeding habits (approaching phases of the Quaternary permitting the ex-
the condition in marine fishes). The semi-
isolated conditions in tributary streams, varzea,
pansion of nonforest faunas appears to be
oxbow and marginal lakes, would appear to more realistic. In this connection, detailed
offer ideal conditions for allopatric specia- studies of the biota presently inhabiting iso-
tion. Oscillations in river levels, due to factors lated nonforest enclaves (savanna islands)
ranging in scale from sudden local downpours Amazon
of rain to long-term climatic cycles, and geo-
within the forest region would be
morphological changes leading to river cap-
particularly interesting ( e.g.. Brown and Ben-
tures, give abundant opportunities for species son, 1977 ) Short ( 1975 ) reviewed the chaco
.
Fig. 5:11. Distribution of certain nonforest birds in tropical lowland South America (after Short, 1975).
Explanations: A. Ranges of the wren Campylorhynchus griseus (stippled) and the white woodpecker Mcla-
nerpes candidus (solid). Note isolated populations on Amazonian savanna enclaves; C. griseus has another
isolated population in southwestern Mexico. B. Disjunct ranges of Picazuro Pigeon (Cohtmba picazuro; 1, 2)
and its allospecies C. corensis (3). C. Ranges of Fluvicola pica (hatched) and F. nengeta (solid); southern
race of F. pica is partly migratory as indicated by arrows and dashed line. D. Range of Red-crested Finch
(Coryphospingus cucullatus, stippled) and its allospecies C. pileatus (hatched). E. Narrow range disjunc-
1979 HAFFER: QUATERNARY OF TROPICAL LOWLANDS 131
least two effects through which genetic- formation; other causes such as founder effect
and adaptation to the physical environment do
morphological differentiation was initiated. not have to act in order that races should be
1) Interruption of gene-flow permitted local formed."
adaptations and chance deviations to develop
in short periods of time (Mayr, 1942, 1963), Rate of evolution. —
The present consensus
the amount of differentiation depending upon is few vertebrate species date back to
that
the degree of geographic isolation, the size of the Pliocene, most neospecies having devel-
the refuge population, and the "plasticity" of oped since the beginning of the Pleistocene
the various systematic groups. 2) If gene- 1.8 of 2 million years ago (see p. 117). In-
flow normally is unimportant in maintaining creased knowledge on the absolute age and
genetic cohesion within species such as Ama- the regionally varying effect of climatic cy-
zonian butterflies (Ehrlich & Gilbert, 1973; cles willpermit speculations on the rate of
Turner, 1976a,b), geographic isolation may Quaternary speciation in particular groups of
cause evolution through other effects. Turner animals. Models will be constructed that
(
1977 ) suggested disorderly extinction of assign periods of species or subspecies differ-
species in the refugia as probably responsible entiation to particular phases of forest or
for rapid deviation of refuge populations in savanna fragmentation. Thus Vanzolini and
mimetic butterflies. The species composition Ab'Saber (1968) related the dispersal and
of the floras and faunas in the refugia prob- differentiation of two species of Liolaemus,
ably came to differ through disorderly ex- southeastern Brasilian lizards, to dated cli-
tinction, even though the distribution of spe- matic-vegetational shifts. However, Miiller
cies in the original continuous forest may have and Steiniger (1977) contested their conclu-
been comparatively uniform. Turner (1977: sions on the basis of studies of the systematics
110-115) concluded: of these lizards. Turner (1976) discussed
"The on the ecosystem
effects of extinction fairly recent speciation and subspeciation in
are to be far Even if no certain Heliconius butterflies in relation to
likely reaching.
butterfly species became extinct in any refuge, late and postglacial arid climatic phases.
tions, zones of hybridization, and other zones of avian interaction involving Chacoan birds; black lines indicate
location of zones involving 5 to 12 situations and dashed lines indicate location of zones involving 3 to 4 situa-
tions.
Distribution de ciertas aves savanofilas en las tieiras bajas tropicalcs de Sudamerica (segun Short, 1975).
Explicacioncs: A. Areas de distribution del cucarache ro Campylorhynchus griseus (puntcado) y del carpintero
bianco Melanerpes candidus (negro). Notese las poblaci ones de estas cspecies aisladas en las savanas de la selva
amazonica; C. griseus tambien se encuentra en el sudo este de Mexico. B. Areas disyunctas de las palomas
Columba picazuro (1,2) tj de su alloespecie C. corensis (3). C. Areas de distribution de Fluvicola pica (raya-
das) y de F. nengeta (negras); la raza meridional de F. pica cs parcialmente migratoria lo </i«- esta indicado por
flcchas y una linca raijada. D. Areas de distribution di Coryphospingus cucullatus (punteadas) y de su allo-
especie C. pileatus (rayadas). E. Disyuncioncs cstrccli cs, zonas de hibridizacion y otras zonas de interaction
de aves de la region chaqucna. Lineas continuas ind; can la position de zonas que se refieren a 5—12 parejas
aviarias; lineas entrecortadas de zonas que se reficren a 3—1 parejas aviarias.
132 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
7. Study nonforest faunas which have Terciario tardio sobrevivieron en los refugios
been somewhat neglected in recent pleistocenicos sin apreciable cambio evolu-
years compared to forest faunas and cionario.
analyze the relationship of forest and Un metodopropuesto para el analisis his-
nonforest faunas (e.g., Sick, 1966; Van-
torico-biogeografico y la reconstruccion de
zolini, 1974). antiguos refugios reeomienda, primero, ma-
8. Study the processes of differentiation pear la posicion de zonas de contacto secun-
that operated in the ecological refuge dario de especies parapatricas y de subespe-
islands and evaluate the significance cies que hibridizan. Estas zonas muchas veces
and, if possible, the amount of fauna] se encuentran mas o menos supcrpuestas en
extinction during adverse climatic-vege- ciertas regiones. Estas 'zonas de sutura' in-
tational periods. dican la posicion de antiguas barreras, aiin
cuando las barreras ecologicas mismas de-
subespecie o especie antes de establecer con- naron tambien la evolution de los animales
tacto secundario con poblaciones de otros re- de ambientes no-forestales sudamericanos. Se
fugios durante peiiodos expansivos bajo un necesita pruebas directas adicionales para de-
clima favorable. Numerosas especies prob- terminar la posicion de los refugios de selva
ablemente se extinguieron durante las fluc- y savana durante el cuaternario basadas en
tuaciones pleistocenicas de la vegetacion. datos geomorfologicos, pedologicos y palinol-
Posiblemente solo unas pocas especies del ogicos.
134 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
pansao dos climas secos na America do Sul, por fugios quaternarios, e conservacao de patrimonios
ocasiao dos periodos glaciais quaternaries. Paleo- geneticos na regiao neotropical: Padroes de dif-
climas 3:1-19. erenciacao em Ithomiinae (Lepidoptera: Nym-
Absy, M. L., van der Hammen, T. 1976.
Some
phalidae). Acta Amazonica 7:75-137.
southern
palaeoecologieal data from Bondonia, Brown, K. S., Jr., Ab'Saber, A. N. 1979. Ice-age
part of the Amazon Basin. Acta Amazonica 6:
forest refugia and evolution in the Neotropics:
293-299. Correlation of paleoclimatological, geomorpholog-
Avila-Pires, F. D. de. 1974. Caracterizacao zoo- ical and pedological data with modern biological
geografica da Provincia Amazonica. II
A familia — endemism. Inst. Geogr. (Sao Paulo), Paleoclimas
Callithricidae e a zoogeografia Amazonica. Anais 5:30 p.
Acad. Brasil Cienc. 46:159-181. Brown, K. S., Jr., Benson, W. W. 1977. Evolution
extinctions of
Axelrod, D. I. 1967. Quaternary in modern Amazonian non-forest islands: Hcli-
mammals. Univ. California Publ. Sci. 72: conius hermathena. Biotropica 9:95-117.
large
1-42. Brown, K. S., Jr., Sheppard, P. M., Turner, J. B. G.
Ballmann, P. 1977. Neuere Erkenntnisse iiber die 1974. Quaternary refugia in tropical America:
zeitliche der rezenten Vogelartcn.
Entstehung Evidence from race formation in Hcliconius but-
Mitt. Bayer. Staatssamml. Palaeontol. Hist. Geol. terflies. Proc. Boy. Scot. London B, 187:369-378.
17:169-175. Campbell, K. E., Jr. 1976. The late Pleistocene
Barbosa, A. 1958. Geomorfologia do Territorio do avifauna of La Carolina, southwestern Ecuador.
Bio Branco. Not. Geomorfol. ( Sao Paulo ) 1(1): Smithsonian Contrib. Paleobiol. 27:155-168.
16-18. Campbell, K. E., Jr. 1979. The Pleistocene avi-
Bates, H. W. 1864. The Naturalist on the Biver fauna of the Talara tar seeps, northwestern Peru.
Amazons. John Murray, London, 406 p. Royal Ontario Mus. Life Sci. Contrib. (in press).
Behle, H. W. 1963. Avifaunistie analysis of the Carcasson, B. H. 1964. A preliminary survey of the
Great Basin region of North America. Proc. XIII zoogeography of African butterflies. East African
Int. Ornith. Congr. pp. 1168-1181. Wildl. J. 2:122-157.
Bicarella, J. J. Subsidies para o estudo das
1965. Chapman, F. M. 1917. The distribution of bird-life
variacoes de nivel oceanico no quaternario bra- in Colombia. A contribution to a biological sur-
sileiro. An. Acad. Brasil. 37(suppl.) :263-278. vey of South America. Bull. Amer. Mus. Nat.
Bigarella, J. Variagoes climaticas no qua-
J. 1971. Hist. 36:1-729.
ternario superiordo Brasil e sua dataeao radio- Chapman, F. M. 1926. The distribution of bird-life
metrica pelo metodo de carbono 14. Inst. Geogr. in Ecuador, a contribution to the study of the
(Sao Paulo), Paleoclimas 1:1-22. origin of Andean bird life. Ibid. 55:1-784.
Bigarella, J. J., Andrade, G. O. de. 1965. Contri-
Cole, M. M. 1960. Cerrado, caatinga and pantanal:
bution to the study of the Brazilian Quaternary.
The distribution and origin of the savanna vege-
Geol. Soc. Amer., Spec. Pap. 84:433-451.
tation of Brasil. Geogr. J. 126:166-179.
Blandin, P. 1977. La distribution geographique des
Brassolinae ( Lepidoptera —
Nymphalidae). Faits
Croizat, L. 1958. Panbiogeography, 1 (The New
World). Published by author, Caracas, Venezue-
et problemes, pp. 161-218Descimon, H. (ed.).
in
la, 1058 p.
Biogeographie et Evolution en Amerique tropi-
cale. Publ. Lab. Zool. Ecole Normale Sup., 9: Croizat, L. 1976. Biogeografia analitica y sintetica
1-344. ( "Panbiogeografia"
de las Americas. Biblioteca
)
Caribbean climate
1973.
Acad. Cien. Fis., Mat. Nat., Caracas, 15-16:1-
Bonatti, E., Gartner, S.
890.
during Pleistocene ice ages. Nature 244:563-565.
Bremer, H. 1973. Der Formungsmeehanismus im Croizat, Nelson, G., Rosen, D. E. 1974. Cen-
L.,
ters and related concepts. Syst. Zool.
of origin
tropischen Begenwald Amazoniens. Z. Geomor-
23:265-287.
phol., Suppl. 17:195-222.
Brodkorb, P. 1971. Origin and evolution of birds, Damuth, J. E., Fairbridge, B. W. 1970. Equatorial
Atlantic deep-sea arkosic sands and ice-age aridity
pp. 19-55 in Farner, D. S., King, J. B. (eds.).
Avian Biology. Vol. 1. Academic Press, New in tropical South America. Geol. Soc. Amer. Bull.
Descimon, H. 1977. Biogeographie, mimetisme et Franzle, O. 1976. Die Schwankungen des pleisto-
speciation dans le genre Agrias Doubleday ( Lep., ziinen Hygroklimas in Sudost-Brasilien und Sii-
Nymphalidae, Charaxinae ) , pp. 307-344 in Des- dost-Afrika. Biogeographica 7:143-162.
cimon, H. (ed.). Biogeographie et Evolution en Callardo, J. M. 1965. The species Bufo granulosus
Amerique Tropicale. Publ. Labor. Zool. Ecole Spix (Salientia: Bufonidae) and its geographic
Norm. Sup. 9:1-344. variation. Bull. Mus. Comp. Zool. Harvard Univ.
Diamond, J. M. 1976. Relaxation and differential 134:107-138.
extinction on land-bridge islands: Applications Gallardo, J. M. 1972. Origen de las faunas sud-
to natural preserves. Proc. 16th Int. Omith. americanas de amfibios. Comun. Mus. Argentino
Congr. pp. 616-628. Cienc. Nat. "Bernardino Rivadavia" Inst. Nac.
Dorst, J. 1976. Historical factors influencing the Invest. Cien. Nat. Zool. 4:17-32.
richness and diversity of the South American Garner, H. F. 1959. Stratigraphic-sedimentary sig-
avifauna. Proc. 16th Int. Ornith. Congr. pp. 17- nificance of contemporary climate and relief in
35. four regions of the Andes mountains. Geol. Soc.
Duellman, \V. E. 1972. South American frogs of Amer. Bull. 70:1327-1368.
the Hijla rostrata group (Amphibia, Anura, Hy- Garner, H. F. 1966. Derangement of the Rio Ca-
lidae). Zool. Meded. Rijks Mus. Nat. Hist. Leiden roni, Venezuela. Rev. Geomorphol. Dynamique
47 (14): 177-192. 2:53-80.
Duellman, W. E. 1978. The biology of an equa- Garner, H. F. 1968. Tropical weathering and re-
torialherpetofauna in Amazonian Ecuador. Univ. lief, in Fairbridge, R. W. (ed.). Encyclopedia
Kansas Mus. Nat. Hist. Misc. Publ. (65):l-352.' of Geomorphology. Vol. 3. Reinhold Book Corp.,
Duellman, W. E., Crump, M. L. 1974. Speciation New York, 1295 p.
in frogs of the Hijla parviceps group in the upper Garner, H. F. 1974. The origin of landscapes. Ox-
Amazon Basin. Unix-. Kansas Mus. Nat. Hist. ford Univ. Press, London, 734 p.
Occas. Pap. (23): 1-40. Garner, H. F. 1975. Rain forests, deserts and evo-
Echternacht, A. C. 1971. Middle American lizards lution. An. Acad. Brasil. Cienc. 47:127-133
of the genus Ameiva (Teiidae) with emphasis on ( Suplemento ) .
Hellmayr, C. E. 1912. Zoologische Ergebnisse Kingdom, J. 1971. East African mammals. An atlas
einer Reise in das Miindungsgebiet des Amazonas. of evolution in Africa, Vol. 1. Academic Press,
Vbgel. Abh. Bayer. Akad. Wiss. Math.-Phys.
II London, 446 p.
26(2): 1-142.
Kl. Klammer, G. Zur Morphologie der Triim-
1963.
Hershkovitz, P. 1968. Metachromism or the prin- mererze des Eisenquarzitmassivs von Urucum in
ciple of evolutionary change in mammalian tegu- Mato Grosso, Brasilien. Z. Geomorphol. N.F. 7:
mentary colors. Evolution 22:556-575. 289-307.
Hershkovitz, P. 1969. The Recent mammals of the Klammer, G. 1971. tiber plio-pleistozane Terrassen
Neotropical Region: A zoogeographic and eco- im unteren Amazonasgebiet. Ibid. 15:62-106.
logical review. Q. Rev. Biol. 44:1-70. Klammer, G. 1975. Beobachtungen an Hangen im
Hershkovitz, P. 1972. The Recent mammals of the tropischen Regenwald des unteren Amazonas.
Neotropical region: A zoogeographic and eco- Ibid. 19:273-286.
logical review, pp. 311—432 in Keast, A., Erk, Lancenheim, J. H., Lee, Y.-T., Martin, S. S. 1973.
F. C, Glass, B. (eds.). Evolution, mammals, and An evolutionary and ecological perspective of
southern continents. State Univ. New York Press, Amazonian Hylaea species of Hyincnaea Legu- (
related to Lygopliis boursieri and a new species Reinic, W. F. 1938b. On the history of Holarctic
floral and faunal habitats in the Quaternary
(Colubridae). Amer. Mus. Novit. (2522): 1-37.
period. Res. Prog. 4:67-72.
Myers, C. W. 1974. The systematics of Rhadinaca
W. F. 1950. Chorologische Voraussetzungen
Reinic,
(Colubridae), a genus of New World snakes. fur die Analyse von Formenkreisen, pp. 346-378
Bull. Amer. Mus. Nat. Hist. 153:1-262.
in Jordan's, A. von, Peus, F. (eds.). Syllego-
Myers, C. W., Daly, J. W. 1976. Preliminary eval- mena Biologica, Festschrift Otto Kleinschmidt.
uation of skin toxins and vocalizations in taxo- Ziemsen, Wittenberg, 472 p.
nomic and evolutionary studies of poison-dart Remington, C. L. 1968. Suture-zones of hybrid
frogs ( Dendrobatidae ) . Ibid. 157:173-262. interaction between recently joined biotas, pp.
Parmenter, C, Folcer, D. W. 1974. Eolian bio- 321—428 in Dobzhansky, T. (ed. ). Evolutionary'
genic detritus in deep sea sediments: A possible Biology. Vol. 2. Appleton-Centurv-Crofts, New
index of equatorial ice age aridity. Science 185: York, 458 p.
695-698. Rosen, D. E. 1978. Vicariant patterns and historical
Patterson, B., Pascual, R. 1968. The fossil mam- explanation in biogeography. Syst. Zool. 27:
mal fauna of South America. Q. Rev. Biol. 43: 159-188.
409-451. Sarma, A. V. N. 1974. Holocene paleoecology of
Paula Couto, C. 1975. Mamiferos fosseis do
de. south coastal Ecuador. Proc. Amer. Philos. Soc.
Quaternario do sudeste brasileiro. Int. Symp. on 118 (1):93-134.
the Quaternary, southern Brazil (July 1975), Savage, J. M. 1974. The Isthmian link and the evo-
Spec. Contrib., Bol. Paranaense de Geociencias lution of Neotropical mammals. Nat. Hist. Mus.
33:89-132. Los Angeles Cty. Contr. Sci. (260):1-51.
Platnick, N. J., Nelson, G. 1978. A method of Schwerdtfecer, W. (ed.). 1976. Climates of Cen-
analysis for historical biogeography. Syst. Zool. tral and South America. World survey of clima-
27:1-16. tology. Vol. 12. Elsevier, Amsterdam, 532 p.
138 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Sclateb, P. L., Salvin, O. 1867. List of birds col- Tricart, J. 1974. Existence de periodes seches au
lected by Mr. Wallace on the lower Amazons and Quaternaire en Amazonie et dans les regions voi-
Rio Negro. Proc. Zool. Soc. London 1867:566- sines. Rev. Geomorph. Dynamique 4:145-158.
596. Tricart, J. 1975. Influence des oscillations climati-
Selaxdeb, R. K. 1971. Systematics and speciation ques recentes sur le modele en Amazonie Ori-
in birds, pp. 57-147 in Farner, D. S., King, J. R. entale ( Region de Santarem ) d'apres les images
(eds.). Avian Biology. Vol. 1. Academic Press, radar lateral. Z. Geomorphol. N.F. 19:140-163.
New York, 586 p. Turner, J. R. G. 1965. Evolution of complex poly-
Sheppard, P. M., Turner, J. R. G., Brown, K. S., Jr., morphism and mimicry in distasteful South Amer-
Benson, W. W. 1978. Genetics and the evolu- ican butterflies. Proc. XII Intern. Congr. Entomol.
tion of Miillerian mimicry in Heliconius butter- London (1964). 267.
flies. Phil. Trans. Roy. Soc. London (in press). Turner, J. R. G. 1971. Studies of Miillerian mimi-
Short, L. L. 1975. A zoogeographic analysis of the cry and its evolution in Bumet moths and heli-
South American Chaco avifauna. Bull. Amer. conid butterflies, pp. 224-260 in Creed, R. (ed.).
Mus. Nat. Hist. 154:163-352. Ecological genetics and evolution. Blackwell
Sick, H. 1965. A fauna do cerrado. Arq. Zool. (Sao Scient. Publ. Co., Oxford, 391 p.
Paulo) 12:71-93. Turner, J. R. G. 1976a. Muellerian mimicry: Classi-
Sick, H. 1966. As aves do cerrado como fauna cal "beanbag" evolution and the role of ecological
arboricola. An. Acad. Brasil. Cienc. 38:355-363. islands in adaptive race formation, pp. 185-218
Sick, H. 1967. Rios e enchentes na Amazonia como /ii Karlin, S., Nevo, E. (eds.). Population ge-
obstaeulo para a avifauna, pp. 495-520 in Lemt, netics and ecology. Academic Press, New York,
H. ( ed. ) Atas do Simposio sobre a
. Biota 832 p.
Amazonica. Vol. 5 ( Zool. ) Conselho Nacional
.
Turner, J. R. G. 1976b. Adaptive radiation and
de Pesquisas, Rio de Janeiro, 603 p. convergence in subdivisions of the butterfly genus
Silverstone, P. A. 1975. A revision of the poison- Heliconius Lepidoptera:
( Nymphalidae). 1.
Vanzolini, P. E. 1974. Ecological and geographical Region postulated by several authors during the last
distribution of lizards in Pernambuco, northeast- ten years ( Haffer, Vanzolini, Miiller, Prance, Brown,
ern Brazil (Sauria). Pap. Avul. Zool. (Sao Paulo) Sheppard, Turner; see compilation and table in
28:61-90. Brown, 1976:228-235; 1977a,b). These refugia prob-
Vanzolini, P. E., Ab'Sabeb, A. N. 1968. Divergence ably were viable during one or another adverse
rate in South American lizards of the genus Lio- climatic period or even repeatedly during the Qua-
laemus (Sauria, Iguanidae). Ibid. 21:205-208. ternary. Their locations are illustrated in figure 5:8.
Vanzolini, P. E., Williams, E. E. 1970. South
American anoles: The geographic differentiation Middle American Refugia
and evolution of the Anolis chrysolepis species 1. Guatemala refuge: Caribbean slope and foreland.
group (Sauria, Iguanidae). Arq. Zool. (Sao 2. Caribbean Costa Rica: Caribbean slope and fore-
Paulo) 19:1-298. land.
Vuilleumier, B. S. 1971. Pleistocene changes in 3. Pacific of southeastern
Chiriqui refuge: slopes
the fauna and flora of South America. Science
Costa Rica and southwestern Panama.
173:771-780. 4. Darien refuge: Mountains of eastern Panama.
Vuilleumier, F. 1975. Zoogeography, pp. 421-496
in Farner, D. S., King, J. R. (eds.). Avian biol-
Peri-Andean Refugia
ogy. Vol. 5. Academic Press, NewYork, 586 p.
Wallace, A. R. 1853. A Narrative of travels on the 5. Nechi refuge: Northern base of Colombian Cor-
Amazon and Rio Negro. Reeve and dilleras.
Co., London
Dover reprint (1972) of Ed. 2, 1889, 363 p.
6. Choco refuge: Western slope and foreland of
D. 1978. A Cordillera Occidental of Colombia.
Webb, S. history of savanna vertebrates
in the New World. Part II: South America and 7. Chimborazo refuge: Western base and foreland
the great interchange. Ann. Rev. Ecol. 9:
of Cordillera Occidental in Ecuador (about 1°
Syst.
393-426. north to 2° south).
8. Cauca refuge: Upper Rio Cauca Valley in south-
Wetterberc, G. Jorge-Padua, M. T., Castro,
B.,
C. S. de. 1976. Uma analise de prioridades em central Colombia.
conservacao da natureza na Amazonia. Seria
9. Magdalena refuge: lower eastern slopes of Cen-
Tecnica, 8. Ministerio de Agricultura, Inst. Brasil. Cordillera of Colombia.
tral
Venezuela.
Forest Remnants
29. Guiana refuge: Southern Surinam and southeast-
ern Guyana, possibly also comprising forests Remnant forests exist under present climatic con-
northwest of Manaus, northern Brasil. ditions on isolated mountains in generally dry north-
30. Oyapock refuge: Eastern, forested part of Tumuc- eastern Brasil (e.g., Serra de Ibiapaba, Serrania de
Humac Mountains on the border of Cayenne Baturite) and in central Brasil forming "minirefugia."
( French Guiana ) and Amapa, Brasil.
Islands
Amazon Basin Refugia
A. Sierra Nevada de Santa Marta: Montane island
31. Belem refuge: Lowlands southeast of Belem to in northern Colombia. Endemic elements inhabit
western and southern Maranhao, Brasil. moderate to high elevations. Tropical species of
32. Tapajos refuge: Low hills in southwestern Para, forest along seaward and southeastern sides are
Brasil from Itaituba on the Rio Tapajos south to mostly undifferentiated, related to Nechi forms.
the Serra do Cachimbo. B. Tobago:Island near Trinidad.
33. Rondonia refuge: Hill region between Rio Jipa- C. Eastern Marajo: Scrubby forest at the eastern tip
rana and upper Rio Roosevelt, Rondonia, Brasil. of Marajo Island in the mouth of the Amazon
34. Madeira refuge: Lowlands between Rios Madei- harbors a number of differentiated forms of but-
ra and lower Purvis, north of Humaita, Brasil. terflies.
35. Tefe refuge: Lowlands between the lower Punis D. Tapani: Isolated ridges in the Amazon opposite
and Jurua rivers, south of Tefe ( =Ega), Brasil. the mouth of the Rio Tapajos ( Santarem ).
6. Late Cenozoic Environmental Changes in
Temperate Argentina
Ana Maria Baez
Departamento de Geologia
Facultad de Ciencias Exactas ij
Naturales
Universidad de Buenos Aires
Buenos Aires, Argentina
Present temperate South America lies potamia. At the western border of the plains
mainly south of the Tropic of Capricorn are the easternmost slopes of the Sub-Andean
(
23°27'S ) According to the Koppen climatic
.
Ranges ( in the north ) and the faulted blocks
classification, based on both temperature and of the Pampean Ranges ( in the south ) West .
rainfall, temperate climates are found in a of these the Sud-Andean Region, which ex-
is
large area including most of Paraguay, parts tends west to the foot of the Andes. The most
of Bolivia, Brasil and Argentina, and Uru- southern region is Patagonia, with its northern
guay, as well as in a region along the Pacific limit coinciding approximately with the Rio
coast south of 38°S (Eidt, 1969). Neverthe- Colorado.
less, we also have taken into consideration In this paper we discuss the different lines
Patagonia and a strip of land lying in the of evidence concerning the environmental
rain-shadow east of the Andes, where arid changes that occurred in those regions since
climates now prevail (Eidt, 1969). Their late Miocene times. It is pertinent to point
inclusion is justified because the geological out here that particularly in this part of South
events that provoked the development of America neither the majority of the litho-
increasing xeric conditions in Patagonia and stratigrapliical units of the Cenozoic nor the
the sub-Andean area also delineated the phys- chronostratigraphical ones (especially Stages)
ical features and the basic climatic distribu- have been established formally. The latter
tion that presently exists in the other regions are the base upon which the corresponding
here considered. chronological intervals (Ages), or geochron-
From the climatic point of view an east- ological units, are differentiated according to
west change from the more humid eastern the American Code of Stratigraphical Nomen-
regions to the dry regions close to the Andes clature. On the other hand, the faunistic
overlaps a latitudinal change between the content, fundamentally mammals, of the non-
warm north and the cool temperate south. marine deposits has been comparatively well
Thus, within the extensive area under con- studied. The results of these studies have
sideration, several major regions can be rec- permitted the establishment of a sequence
ognized besides the Andean one. Those in of Provincial or Land-Mammal Ages, based
the Argentinian territory, to which we will on the characteristics of the mammal assem-
refer principally, are Mesopotamia, Chaco- blages. Tentatively, these Ages have been
Pampean Plain, Sud-Andean Region, and referred to the world-wide accepted Epochs
Patagonia (Fig. 6:1). (Simpson, 1940; Pascual et al., 1965). The
Mesopotamia is distinctly defined geo- stratigraphy and chronology of the continen-
graphically by the Rio Parana and the Rio tal Cenozoic, not only in the region consid-
Uruguay. Extensive lowlands, the Chaco- ered here, but in South America as a whole,
Pampean Plains, extend to the west of Meso- are based mainly on the sequence of mamma-
141
142 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
lian faunas. The Provincial Ages assigned to mont areas of the northern Patagonian Andes,
the Pliocene and Pleistocene Epochs, as well as as in northern Patagonia.
well A pro-
as the principal lithostratigraphical units re- nounced uplift at that time is indicated by
ferred to them are shown in figures 6:2-3. No the abundance of psephitic material coming
reliable radiometric dates of the latter ages from the neighboring Andes (Pascual and
are available. Odreman Rivas, 1971 ) This is consistent with
.
The chronology used in this paper is the other evidence based on radiometric studies
one usually applied by South American pale- indicating an increase in Andean tectonics
Thus, the Miocene-Pliocene from the late Miocene onwards (Rutland,
ontologists.
10-12 Guest and Crastley, 1965 ) As a consequence
m.y.b.p. However,
.
boundary is placed
recent studies in the European type areas in- of those movements, the Patagonian Andes
dicate that this boundary should be placed reached sufficient elevations so as to act as a
about 5.5 m.y.b.p. This formal change im- barrier to the moisture-laden Pacific winds.
This had a profound ecological effect in the
plies the reference of the Chasicoan and
Huayquerian Ages to the Miocene. development of increasing xeric conditions in
Patagonia and the subsequent improverish-
ment of the biota. Many savanna-woodland
GENERAL CONDITIONS DURING adapted animals retreated northwards, while
THE MIOCENE others, perhaps more eurythermal ones, sur-
vived in the more humid western area or
An understanding of the environmental west of the Andes Gasparini and Baez, 1975;
(
changes that occurred during the late Ceno- Baez and Gasparini, 1977). Paleobotanical
zoic in temperate South America must take data also show this pattern (Menendez, 1971:
into consideration the conditions existing in 364). Possibly there was a marked lowering
geologically earlier times. To be more pre- of mean temperatures, especially in the south-
cise, the late Cenozoic changes are the cul- ernmost region. Berggren and Van Couvering
mination of processes that began in the late (1974) suggested that an extensive ice sheet
Miocene. was present in western Antarctica from 7 to
10 m.y.b.p. (early Pliocene according to the
Early and middle Miocene (Santacruzian
Age) sedimentary rocks are exposed in south- chronology used in this paper). Thus, tem-
ern Patagonia, where they extend to the aus- peratures might have been significantly cooler
tral end of the continent. A milder and more than in earlier times.
equable climate at that time should have per- A progressive climatic deterioration in
mitted the presence of a biota in latitudes up Patagonia, at least since the Miocene, is indi-
to 50°S, which has extant relatives living cated by palynological data from deep-sea
mainly in subtropical forested areas. That is cores from the southwestern Atlantic Ocean
the case, for instance, of ceboid monkeys, (Groot et al., 1967). The Miocene assem-
echimyid rodents, myrmeeophagid and mega- blages have a high percentage of arboreal
therioid edentates; among the latter are the pollen and the appearance of a few xero-
Megalonychidae and Nothrotheriinae which phy tic elements ( such as Ephedra ) which are
,
are closely related to living tree sloths (Pas- absent in Eocene assemblages and abundant
cual, 1970; Pascual and Odreman Rivas, in the Pleistocene ones.
1971) Nevertheless, since the early Tertiary The "Paranense Sea" constituted a con-
in Patagonia the herbivorous mammals show
spicuous geographic feature in late Miocene
increasing dental adaptations indicating the times. This Atlantic transgression extended
spread of grasslands (Pascual and Odreman over most of the Chaco-Pampean region and
and Pascual, 1972).
Rivas, 1971; Patterson western Mesopotamia, from northeastern
In the late Miocene (Friasian Age) sedi- Patagonia to southern Paraguay (Camacho,
mentation predominandy occurred in pied- 1967).
144 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
CHAPADMALAL
FM.
MONTEHERMOSAN
MONTE
HERMOSO
FM.
Id
MESOPO- ARAUCA-
TERTIARY PLIOCENE a HUAYQUERIAN RIONEGRENSE EPECUEN TAMIENSE' NENSE"
o FM.
ARROYO
CHASICOAN
chasico'fm.
Fig. 6:2. Pliocene South American Provincial Ages and main referable mammal-bearing lithostratigraphical
units in the regions under study.
Edadcs Provinciates pliocenicas sudamcricanas tj principals unidades litoestratigrdficas mamalijeras re-
fcribles en /as regiones estudiadas.
PERIOD
146 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Nevertheless, a few elements of the mega- gonia and the advance of a postulated "great
fauna survived into Holocene times. For Somuncura glacier" extending to the sea dur-
example the ground-sloth Mylodon is known ing the youngest glaciations also have been
from deposits younger than 5394 ± 55 y.b.p. suggested by Auer (1956, 1958, 1960, 1974).
in southern Chile (Saxon, 1976). Such schemes have been mostly discarded
for recent investigations have demonstrated
Glaciations that much of the evidence of glacial action
upon which those hypothesis were based are
Very often the concept of the Quaternary the result of different geomorphological proc-
has been equated with the Ice Age and the
esses (Polanski, 1963, 1965; Methol, 1967; Fi-
base of the Pleistocene considered synchro-
nous with the onset of glaciation. But, in fact, dalgo and Riggi, 1970; Fidalgo, 1973). There
climatic deterioration to accelerate
is now general agreement that east of the
began
about 10 million years ago (early Pliocene) Patagonian Andes glaciation extended for a
relatively short distance, reaching the present
(Berggren and Van Couvering, 1974). Dif-
Atlantic coast only south of about 52°S (An-
ferent lines of evidence indicate the occur-
and Fidalgo,
tevs, 1929; Caldenius, 1932; Flint
rence of glaciation in polar and mountainous
areas of both hemispheres at least since Plio- 1963, 1968). On
the other hand, westward
cene times (Curry, 1966; Hays and Opdyke, flowing glaciers reached the Pacific Ocean as
far north as 43°S (Mercer, 1976). The glaci-
1967; Fleck et al., 1972; Mercer and Sander,
ated areas progressively decreased north-
1975), but this does not imply the synchro-
nous presence of lowland ice sheets in tem- wards, where they were confined mainly to
mountainous areas even during full glacial
perate regions.
times (Groeber, 1936; Frenguelli, 1957b).
The four classic Pleistocene glaciations of
Generally, three or four glaciations have
Europe (Giinz, Mindel, Riss, Wiirm) tradi- been recognized on the eastern side of the
tionally have been correlated to those of
Patagonian Andes, although the interpretation
North America ( Nebraskan, Kansan, Illinoian,
of those glacial events has differed. They
Wisconsin), but the problem is complex and have been considered either as minor episodes
has been oversimplified. Recent evidence has
roughly equivalent to the Wurm-Wisconsin
led to a complete revision of that scheme,
and perhaps also to an older event (Calden-
and accurate intercontinental correlation of
ius, 1932; Flint and Fidalgo, 1963, 1968) or
glacial events is still under debate (Berggren as different major glaciations ( Feruglio, 1949;
and Van Couvering, 1974). Widely differing
Auer, 1956, 1960, 1970). Moraine belts in the
opinions have been expressed concerning the Chilean lake district have been attributed to
number of major glacial episodes as well as
three major glaciations (Laugenie, 1971; Mer-
the extension of glaciated areas in southern
cer, 1972, 1976); the last was named the
South America.
Llanquihue Glaciation by Heusser (1974).
Four postulated major glaciations (Valli- Recently, glacial deposits in southwestern
manca, Colorado, Diamante, Atuel) on the Patagonia (about 50°S) have been dated be-
Argentinian side of the Andes have been cor- tween 1.2 and probably 1.0 m.y.b.p.; thus,
related with the classic four glaciations in the they correspond to an early Pleistocene glaci-
Northern Hemisphere by Groeber ( 1952, ation ( Mercer and Sander, 1975 ) The avail-
.
1954). He contended that during the first two able data permit reliable intercontinental
assumed glacial events, which were ascribed comparison of only the most recent glacial
to the "Eoquaternary," extensive ice sheets fluctuations.
covered the Andean region and Patagonia. In
northern Patagonia glaciers were supposed to
Environmental Changes from Pliocene
have reached the Atlantic Ocean south of Rio to Holocene Times
Colorado. The other two glaciations were re-
ferred to the "Neoquaternary" and were re- It is attempt a detailed ac-
premature to
stricted mainly to the Andean region. The count of the and environmental
climatic
existence of local centers of glaciation in Pata- changes in temperate South America from
1979 BAEZ & SCILLATO YANE: CENOZOIC OF ARGENTINA 147
Pliocene to Holocene times and their in- also could indicate a woody environment
fluence on biogeographic patterns. Many (Scillato Yane, Uliana and Pascual, 1976).
important factors and much of the evidence According to sedimentological and geomor-
for the evaluation of past climates have been phological data provided by Andreis (1965)
studied inadequately, and interpretations fre- and Volkheimer ( 1971 ) by the late Pliocene
,
quently are contradictory. Therefore, only an the climate at those latitudes had changed to
account of major events and general trends is drier and colder conditions.
sary in order to reconstruct the climatic fluc- of more humid conditions than at the present.
tuations of such a short interval of the earth's Paleosoils from the tablelands of central and
history. northern Patagonia are being studied by J. A.
In many parts of the world, mammalian Ferrer (pers. comm.).
faunas have furnished interesting clues to Prevalence of dry conditions in eastern
possible late Cenozoic climatic fluctuations, Patagonia during Quaternary times is sug-
and the inferences based on them have been gested by palynological analysis of sea cores
supported by other kinds of evidence. Thus, from the Argentine Basin in the southwestern
a basic chronological framework of the cli- Atlantic Ocean (Groot and Groot, 1964,
matic record has been established. It is note- 1966). A two types of
cyclic alternation of
worthy that in southern South America, Pleis- pollen zones occurs with depth. Those indica-
tocene mammals have only been recorded tive of more pronounced arid environments
quite far from the glaciated area. No taxa (high percentage of Chenopodiaceae pollen)
indicative of extreme conditions have been were considered to represent "glacial stages."
recognized. Although in Pleistocene times Furthermore, certain features suggest eustatic
glaciations were mainly restricted to the lowering of sea level, an inference also sup-
mountainous areas, temperature and moisture ported by the study of diatoms (Groot et al.,
changes during glacial and interglacial inter- 1967). More humidity could account for the
vals probably caused important shifting of relative increase of the arboreal pollen in the
the biotas of the nonmountainous regions. In samples attributed to the "interglacial stages."
order to provide a better understanding of the This agrees with the climatic interpretation
Late Cenozoic environmental changes, each of the palynology of interglacial peat beds in
region is considered separately. northwestern Patagonia and Tierra del Fuego
—
Patagonia. By the late Miocene, the de- (Auer, 1958, 1970).
velopment of increasingly xeric conditions Humid glacial ages and relatively warm
had begun in this region. Nevertheless, in and dry interglacial ages in the Andean region
northern Patagonia, especially in the area of of Argentina and central and southern Chile
influence of the pre-Colorado and pre-Negro were proposed in accordance with a pollen
rivers, a warmer and more humid climate diagram and the fluctuations of the carbonate
than that of today seems to have persisted content of a deep-sea core from the southeast-
during early and middle Pliocene. This is sug- ern Pacific Ocean (Groot and Groot, 1966).
gested by the record of a megalonychid eden- However, palynological spectra of strati-
tate of the subfamily Orthotheriinae, which graphic sections in the southern Chilean lake
148 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
district reflect more complex climatic changes (Bossi, 1969; Ramos, 1970; Caminos, 1972;
since the last interglacial, as well as drier Gordillo and Lencinas, 1972).
conditions during the Llanquihue Glaciation The regions along the eastern side of the
(Heusser, 1974). Andes, then lower and wanner than today,
Evidently, eustatic descent of sea level constituted an important pathway for migra-
caused the emersion of much of the present tion between the subtropical regions and the
continental shelf; thus, eastern Patagonia sup- more temperate southern Pampean Plain.
porting a steppe vegetation became larger. Moreover, similar ecological conditions to the
The probable connection of the Malvinas present phytogeographic Chacoan Province
Islands to the mainland could explain the (sensu Cabrera, 1971), with open xerophytic
presence there of terrestrial species with rela- woodland, apparently existed in the south-
tives on the continent, for example the canid, western part of Provincia de Buenos Aires.
Dusycion australis (Auer, 1958:223-224). The Upper Pliocene elements recorded from there,
sub-Antarctic forests, which now extend along such as cariamid birds (Tonni, 1974) myrme-
the eastern slopes of the Andes as far north cophagid anteaters ( Kraglievich, 1934), and
as about 36° (Fernandez, 1976), could have the anuran Bufo paracnemis (Gasparini and
descended to lower altitudes in glacial times Baez, 1975) now are absent in that area.
(Vuilleumier, 1971). The Sub-Andean Region was affected by
Sub-Andean Region. —The rich mamma- important tectonic events during, and subse-
lian faunas recorded in the Pliocene of this quent to, the Plio-Pleistocene. These had con-
region have a marked subtropical character, siderable physiographic and climatic conse-
which extended over a great part of the Sea" (Pascual and Odreman Rivas, 1971,
Chaco-Pampean Plain during the late Mio- 1973). Pliocene and Pleistocene stratigraphic
cene (Frenguelli, 1920; Rossi de Garcia, sequences and their paleontological content
1966). Within the "Mesopotamiense" there is are best known in the Provincia de Buenos
an evident alternation of sediments corre- Aires; therefore, our analysis will deal prin-
sponding to different environments but most cipally with that portion of the Chaco-Pam-
are fluvio-deltaic sediments. pean Plain.
Fossil remains are indicative of humid The available studies of successive faunas
subtropical conditions (Pascual and Odre- and sediments of Pliocene and Pleistocene age
man Rivas, 1971; Gasparini and Baez, 1975; have not yet furnished evidence of drastic
Baez and Gasparini, 1977). Presumably, the changes. In general there was an overall
gallery forests along the river favored the cooling trend from the more warm temperate
southward extension of the tropical biota conditions of the early Pliocene to the more
(
Pascual and Odreman Rivas, 1971 A south- )
. moderate present ones with colder periods
ward extension of tropical flora (Cabrera, probably occurring during the Pleistocene.
1971) and fauna (Ringuelet, 1955, 1961) is Concerning precipitation, at present the re-
evident today, but the southward penetration gion has a permanently humid climate, with
probably was greater in Pliocene times, when a change to drier conditions and some pre-
the climate was warmer. There is an abun- dominantly dry months in the western part.
dance of fossil tree trunks, turtles, large croco- This regime seems to have prevailed during
diles, megalonychid and pampatheriine eden- most of the Pliocene and Pleistocene, but an
tates, and dinomyid rodents in the Pliocene increase of xeric conditions, especially during
deposits. There is an absence of chinchillid certain intervals, is evident (see below).
and octodontid rodents and hegetotheriid Since this region came into existence, the
and mesotheriid notoungulates, which are dominant vegetation type has been grass
characteristic of open environments. steppe. However, the climatic fluctuations
The composition of these assemblages is just mentioned would have favored greater
quite different from those of about the same eastward expansion of xerophytic forest,
age in the Sub-Andean and Pampean regions. which presently borders the Pampean Plain
These differences reveal that these areas had to the west and north.
distinct ecological and biogeographical fea-
Echimyid rodents, now living in tropical
tures. and subtropical regions, are recorded from
Pleistocene deposits are widely distributed this area in the Pliocene and lower Pleisto-
in Mesopotamia, but in most cases neither cene. Also in early Pliocene times (Chasi-
their stratigraphy nor their paleontological coan Age) megalonychid and nothrotheriine
content has been studied adequately; there- tree sloths were abundant there, but they are
fore, their
chronological assignment is still inconspicuous in later ages (Scillato Yane,
uncertain. Even
so, the presence of certain 1977, 1979). These groups of rodents and
taxa is quite interesting from the paleoen- edentates are found frequently in Miocene
vironmental point of view. The record of the deposits of southern Patagonia; the north-
caviid rodent Dolichotis in the middle Pleis- ward shift of their southern limit of distribu-
tocene of northwestern Mesopotamia (Alva- tion reflects the climatic deterioration in the
rez, 1974) indicates that at certain times the south. The occurrence of large dasypodids,
climate was drier than at present; drier con- closely related to the tropical and subtropical
ditions also are indicated by sedimentological living Enphractus, in the Pliocene and of
data (Alvarez, 1974). At present Dolichotis procyonids in the middle and upper Pliocene
isrestricted to the Patagonian, Sub-Andean and lower Pleistocene is also significant. Ac-
and southwestern Pampean regions. cording to these and other faunal evidence,
—
Chaco-Pampean Plain. T his region it can be postulated that throughout the Plio-
started to acquire its present features after cene and early Pleistocene ( Uquian Age ) the
the regression of the late Miocene "Paranense climate was wanner than today followed by
150 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
a gradual cooling. This agrees with the more inary palynological analysis of some samples
pronounced deterioration of climatic condi- from the Argentine shelf off the mouth of the
tions in Patagonia after early Pleistocene time, Rio de la Plata also suggests more pronounced
as indicated by palynological data (Groot aridity during the glacial stages (Groot et al.,
ronment similar to that of the present Cha- mates can be explained by the changes in the
coan phytogeographic province, with open distribution of ocean currents and wind sys-
xerophytic woodlands, has been inferred there tems over the South American continent dur-
from the composition of the fossil fauna ing glacial and interglacial stages (Damuth
(Tonni, 1974; Baez and Gasparini, 1977). and Fairbridge, 1970). During glacial phases
The stratigraphical units of early Pleisto- the Chaco-Pampean Plain was under the in-
cene to Recent age, especially in the coastal fluence of the dry southwesterly winds that
area and in the northeastern part of the acted as the depositional agent for the loess.
Provincia de Buenos Aires exhibit an alterna- Even though the fossil record is rich, the
tion of subaqueous (fluvial, lacustrine, palu- available paleozoological data do not provide
dal )and aeolian deposits ( Kraglievich, 1952, clear support to the environmental fluctua-
1953; Frenguelli, 1957a). These alternations tions during the Pleistocene and Holocene, for
of continental deposits of different origins much of the material is not accurately as-
have been associated with fluctuating climatic signed to stratigraphic levels. Furthermore,
conditions, but there is no agreement as to formational units have a long time range,
what climatic condition, especially tempera- during which climatic fluctuations took place.
ture, corresponds to each of the lithological This probably explains the records of taxa
changes. The existence of alternating humid that are indicative of opposing climatic con-
and arid cycles during the late Pleistocene ditions from a single locality and formation.
and Holocene also is indicated by geomorpho- Nevertheless, significant distributional shifts
logical data (Tricart, 1973). in the biota can be attributed to Pleistocene
marine and brackish-water
Intercalated environmental changes. Drier conditions in
occur discontinuously in different areas
levels the Pampean Plain are indicated by the occur-
and have been correlated with the above rence of characteristic elements of the present
mentioned continental subaqueous deposits Sub-Andean and Patagonian regions, such as
which probably indicate humid conditions the mustelid Lijncodon and the caviids Do-
(Frenguelli, 1957a; Tricart, 1973). If the lichotis and Microcavki, in Pleistocene de-
marine levels are interpreted as resulting from posits. Dolichotis also occurs in sediments of
eustatic sea level rise, the correlation of the the same age in southern Uruguay (Calca-
humid phases with the interglacial stages fol- terra, 1972). On the other hand, more equa-
lows (Tricart, 1973). Accordingly, cold and ble climatic conditions would have allowed
arid or semiarid climates seem to have alter- procyonids and the dasypodid Euphractus to
nated with warm and humid ones. Prelim- live on the Pampean Plain.
1979 BAEZ & SCILLATO YANfi: CENOZOIC OF ARGENTINA 151
matic fluctuations within both the xeric and fluctuations in southeastern Brasil has been
humid phases 1964; Ab'Saber,
provided mainly by sedimentological, geo-
(Bigarella,
morphological, paleontological and phytogeo-
1977). These cyclic changes are also docu- studies. An alternation between
graphical
mented by phytogeographic data, although warm and humid and relatively cold and dry
the chronology is far from clear (Klein,
periods, with minor fluctuations within those
1975). phases, has been proposed.
152 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
y Prof. Guillermo del Corro (Museo Argen- graficos) han suministrado information que
tino de Ciencias Naturales, Buenos Aires) for permite esbozar un relato de la evolution
their valuable comments. Special thanks are ambiental en las regiones consideradas.
extended to Dr. Victor Ramos (Servicio Na- Hasta el Mioceno medio a tardio las con-
tional Minero Geologico, Buenos Aires), who diciones prevalecientes en Patagonia fueron
mucho mas calidas y humedas que las hoy
critically read the manuscript and provided
useful information. aliivigentes; asi lo indica el registro de nu-
merosos vertebrados cuyos parientes actuales
viven en areas intertropicales, parcialmente
RESUMEN boscosas o selvaticas. A partir del Mioceno
tardio se produjo un notable desecamiento
La actual zona templada de America del (motivado por la circunstancia de que los
Sur incluye (de acuerdo con la clasificacion Andes australes alcanzaron ya una altura
climatica de Koppen): Uruguay, la mayor suficiente como para actua como efectiva ba-
parte de Paraguay, partes de Bolivia, Brasil rrera a los vientos del Pacifico ) y una
humedos
y Argentina, asi como una portion territorial mas progresiva atemperacion (en parte coin-
a lo largo de la costa pacifica al sur de los cidente con un partial englazamiento de
38°S. Para los fines del presente estudio, se Antartida occidental). Estos cambios deter-
han considerado tambien la region subandina minaron la retraction hacia el norte y oeste
y Patagonia, donde actualmente prevalecen de muchos elementos de la biota.
climas aridos. Esto se debe a que su historia La consideracion en particular del Ceno-
paleoambiental se halla intimamente ligada a zoico tardio nos remite, primeramente, al
la de los aludidos ambitos actualmente tem- problema de los limites entre las Epocas im-
plados. plicadas (Plioceno, Pleistoceno y Holoceno).
La mayor parte de evidencias disponi-
las El limite plio-pleistocenico no ha sido aun
bles indicadoras de cambios ambientales, claramente determinado en Sudamerica;
especialmente durante el Terciario tardio, se tentativamente se lo ubica hacia el Uquiense
refieren al territorio argentino, en el
que pue- temprano (Blanquense tardio de USA). Tal
den considerarse cuatro regiones principales: incertidumbre responde tanto a la ausencia
1) Patagonia, 2) Area Subandina, 3) Meso- de un criterio unanime a nivel mundial en
potamia, y 4) Llanura Chaco-Pampeana. esta materia, como a las dificultades implicitas
Una adecuada comprension de los cam- en correlacionar las Edades locales sudameri-
bios climatico-ambientales acaecidos en Sud- canas con las de otros continentes. El limite
america templada durante el Cenozoico tardio pleisto-holocenico coincide aproximadamente
requiere inevitablemente, consideration
la con el fin de la Edad Lujanense, serialado por
de los tiempos geologicos anteriores. Los datos una extincion masiva de la megafauna.
disponibles indican un pronunciado deterioro El estudio de los eventos glaciales cuya —
de las condiciones que, aun hacia el Mioceno importancia fundamental en la evolution
medio a tardio, eran mucho mas benignas y climatica del Cenozoico tardio es universal-
uniformes que las actuates. —
mente reconocida se halla aun en Sudameri-
Cabe que en esta parte de Ameri-
seiialar ca en una etapa preliminar. No existe acuerdo
ca del Sur lamayoria de las unidades lito- en cuanto al numero y extension de las gla-
estratigraficas y cronoestratignificas del Ceno- eiaciones, pero puede desde ya desecharse el
zoico continental no han sido formalmente esquema simplista que establecia una exacta
establecidas. De hecho, ha sido el reconoci- correspondencia con las cuatro clasicamente
1979 BAEZ & SCILLATO YANE: CENOZOIC OF ARGENTINA 153
reconocidas en el Pleistoceno del Hemisferio tir su influencia tanto en esta region como en
Norte. En el SO de Patagonia fueron reciente- territorio uruguayo.
mente reconocidos varios cpisodios glaciales Llanura Chaco-Pampeana. —Casi toda la
que datan del Plioceno tardio y Pleistoceno documentation disponible se refiere a la
temprano. Mucho mejor documentada esta mitad austral de esta area (Pampasia). Los
la glaciacion suprapleistocenica,
groseramente estudios de las sucesivas "faunas" y sedi-
correlacionada con
Wiirm-Winsconsin del mentitas pliocenicas y pleistocenicas no han
Hemisferio Norte; pero debe remarcarse que revelado la ocurrencia de cambios drasticos.
afecto especialmente a la region andina, y No obstante, a partir del Plioceno temprano
que solo al sur de los 52°S los glaciares al- tuvo lugar una progresiva atemperacion.
canzaron la actual costa atlantica. Desde el Pleistoceno medio hasta el Holoceno
La evolucion ambiental durante el Ceno- es factible reconocer varias fluctuaciones cli-
zoic tardio en las cuatro areas del territorio maticas: lapsos relativamente calidos y hii-
argentino mencionadas precedentemente pue- medos (con frecuente depositation fluvio-
de sintetizarse asi: lacustre) alternaron con otros relativamente
Patagonia.
—Con exception dc la region
frios y secos (en los que predomino
amplia-
mas septentrional, el clima fue frio y semi- mente la sedimentation de "limos loessoides,"
arido a arido durante la mayor parte del o aun de verdaderos loess). Un analisis pre-
Plioceno. Estas condiciones parecen haberse liminar del contenido paleomastozoologico de
acentuado durante el Cuaternario, si bien las respectivas unidades
estratigraficas avala,
ciertas evidencias sugieren algunos lapsos mas en general, esta interpretation. Es muy im-
hiimedos. Los datos disponibles no indican portante seiialar que los lapsos presumible-
un extenso englazamiento. mente calidos y hiimedos parecen coincidir
Area Subandina. —Durante el Plioceno el
con ingresiones marinas, en tanto que los
numero- frios y secos corresponden a regresiones.
registro paleomastozoologico incluye
sos taxa tipicamente intertropicales, in- Estos ascensos y descensos del nivel marino
que
dican condiciones parcialmente boscosas y podrian ser de origen eustatico.
calidas, posiblemente algo mas humedas que
En lo que ataiie a otras areas fuera del
las actuales; de tal modo, esta area
(continua
territorio argentino, no existe information
entonccs con la chaquefia hacia el norte) suficiente sobre el Terciario tardio. En cam-
obro corao una "via de conexion" del SO de bio, hay claras evidencias de fluctuaciones
la llanura climaticas ciclicas especialmente para el Cua-
pampeana con ambitos francamente
ternario del sudeste de Brasil. De acuerdo a
subtropicales. No obstante, como consecuen-
cia de los Movimientos Andinos, existian ya datos de diversa indole, lapsos calidos y
cuencas intermontanas en la parte septentrio- hiimedos alternaron con otros relativamente
nal donde pudieron prevalecer condiciones frios y secos.
mas Los eventos tectonicos del Plio-
aridas.
ceno tardio-Pleistoceno temprano acentuaron
LITERATURE CITED
notablemente el bloqueo a los vientos hii-
medos del NE, con la consecuente desecacion
Ab'Sabeb, A. N. 1977. Espacos ocupados pelo ex-
progresiva del clima en toda esta region. pansao dos climas secos na America do Sul, por
—
Mesopotamia. Durante el Plioceno las ocasiao dos periodos glaciales quatemarios. Paleo-
condiciones fueron aiin mas calidas y hume- climas 3:1-19.
das que las hoy vigentes. El area influenciada Alvarez, B. B. 1974. Los mamiferos fosiles del
Cuaternario de arroyo Toropi, Corrientes (Argen-
por el rio pre-Parana obro ya entonces como tina). Ameghiniana 11(3) :295-311.
una importantisima via de conexion" entre la Amechino, F. 1902. L'age des formations sedimen-
llanura bonaerense y ambitos biogeograficos tairesde Patagonie. An. Soc. Cient. Argentina
mas septentrionales. En el transcurso del 54(6):283-342.
Cuaternario siguieron prevaleciendo condi- Amechino, F. 1906. Les formations sedimentaires
du Cretace superieur et du Tertiaire de Pata-
ciones calidas y humedas, pero muy probable-
gonie. An. Mus. Nac. Hist. Nat. Buenos Aires
mente algunos lapsos mas secos hicieron sen- (3a)8: 1-568.
154 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Andreis, R. 1965. Petrografia y paleocorrientes de Catamarca, La Rioja y San Juan, pp. 41-79 in
la Formacion Rio Negro Tramo Genera] Conesa-
( Leanza, A. L. (ed.). Geologia Regional Argen-
Boca del Rio Negro). Rev. Mus. La Plata, Secc. tina. Acad. Nac. Ciena Cordoba, 869 p.
Geol. 5(36):245-310. Curry, R. P. 1966. Glaciation about 3,000,000 years
Antevs, E. 1929. Maps of the Pleistocene glacia- ago in the Sierra Nevada. Science 154:770-771.
tions. Geol. Soc. America Bull. 40:631-720. Damuth, J. E., Fairbridge, R. W. 1970. Equatorial
Auer, V. 1956. The Pleistocene of Fuego-Patagonia. Atlantic deep-sea arkosic sands and Ice-Age arid-
Part I: The Ice and Interglacial Ages. Ann. ity in tropical South America. Geol. Soc. Amer.
Acad. Sci. Fennicae, Ser. A3 45:1-226. Bull. 81:189-206.
Auer, V. 1958. The Pleistocene of Fuego-Patagonia. Eidt, R. C.1969. The climatology of South America,
Part II: The history of the Flora and Vegetation. pp. 54-81 in Fittkay, E. J., Illies, J., Klinge,
Ibid., 50:1-239. H., Schwabe, G., Sioli, H. (eds.). Biogeography
Auer, V. 1960. The Quaternary history of Fuego- and ecology in South America. Vol. I, Junk, The
Patagonia. Proc. Roy. Soc. London, Ser. B152: Hague, 447 p.
507-516. Fernandez, J. 1976. El limite boreal de los bosques
Auer, V. 1970. The Pleistocene of Fuego-Patagonia. andino-patagonicos. Bol. Soc. Argentina Bot. 17:
Part V: Quaternary Problems of Southern South 307-314.
America. Ann. Acad. Sci. Fennicae Ser. A3 100: Feruglio, E. 1949. Description geologica de la
1-194. Patagonia. Dir. Grab Yac. Petrol. Fiscales, Buenos
Auer, V. 1974. The isorhythmicity subsequent to Aires 2:1-349.
the Fuego-Patagonia and Fenoscandian ocean Fidalgo, F. 1973. Consideraciones sobre los bajos
level transgressions and regressions of the latest situados al Norte de la provincia de Santa Cruz.
glaciation. Ibid. 115:1-88. Act. V Congr. Geol. Argentino 5:123-137.
Baez, A. M., Gasparini, Z. B. de. 1977. Origenes Fidalgo, F., Riggi, J. C. 1970. Consideraciones
y evolucion de los anfibios y reptiles del Ceno- geomorficas y sedimentologicas sobre los Rodados
zoico de America del Sur. Acta Geol. Lilloana Patagonicos. Rev. Asoc. Geol. Argentina 25:430-
14:140-232. 443.
Berggren, W. A., Van Cowering, J. A. 1974. The Fleck, R. J., Mercer, J. H., Nairn, A. E. M., Peter-
Late Neogene. Palaeogeogr., Palaeoclimatol., Pa- son, D. N. 1972. Chronology of Late Pliocene
laeoecol. 16:1-216. and Early Pleistocene glacial and magnetic events
Bicarjella, J. J. 1964. Variacoes climaticas no Qua- in Southern Argentina. Earth Planet. Sci. Lett.,
ternario e suas implicacoes no revestimento floris- 16:1.5-22.
tico do Parana. Bol. Paranaense Geogr. Curitiba: Flint, R. F. 1965. The
Plio-Pleistocene boundary,
10-15. pp. 497-533 International Studies on the
in
Bodenbender, G. 1924. El calchaqueno y los Estra- Quaternary. Geol. Soc. Amer. Spec. Pap. 84.
tos de la Puna de Penck. Bol. Acad. Nac. Cien. Flint, R. F., Fidalgo, F. 1963. Geologia glacial de
Cordoba 26:405-468. la zona de borde entre los paralelos 39° 10' y
BONDESIO, P., LAZA, J. H., SCLLLATO YaNE, G., TON- 41°20' de latitud sur en la Cordillera de los Andes
ni, E., Vucetich, M. G. 1979. Estado actual del (Argentina). Bol. Dir. Nac. Geol. Min. (Buenos
conocimiento de los vertcbrados de la Formacion Aires) 93:1-35.
Arroyo Chasico (Edad Chasiquense, Plioceno Flint, R. F., Fidalgo, F. 1968. Drift glacial al Este
Temprano) de Provincia de Buenos Aires, Re-
la de los Andes entre Bariloche y Esquel. Min. Ec.
publica Argentina. Act. II Congr. Argentino y Trab., Secret. Est. Energ. y Min., Inst. Nac.
Paleontol. y Bioestr. I Congr. Latinoamer. Pa- Geol. Min., Argentina, Bol. 119:1-19.
leontol. (in press).
Frenguelli, J. 1920. Contribution al conocimiento
Bossi, G. E. 1969. Geologia y Estratigrafia del sec- de la geologia de Entre Rios. Bol. Acad. Nac.
tor sur del valle de Choromoro. Parte I: Tecto. Ciena Cordoba 24:55-256.
Acta Geol. Lilloana 10:17-64.
Frenguelli, J. 1937. Investigaciones geologicas en
Cabrera, A. L. 1971. Fitogeografia de la Republica la zona saltena del Valle de Santa Maria. Inst.
Argentina. Bol. Soc. Argentina Bot. 14(1-2): Mus. Univ. Nac. La Plata. Obra Cincuentenario
1-42. 2:215-572.
Calcaterba, A. 1972. Dos roedores fosiles nuevos Frenguelli, J. 1957a. Neozoico, pp. 1-115 in
para Uruguay y confirmation de otro. Comun. Geografia de la Republica Argentina. Soc. Arg.
Paleontol. Mus. Hist. Nat. Montevideo 1:11-21. Est. Geogr. GAEA, 2, 218 p.
Caldenius, C. 1932. Las glaciaciones cuaternarias Frenguelli, J. 1957b. El glaciarismo cuaternario,
en la Patagonia y Tierra del Fuego. Publ. (Bue- pp. 117-218 iit Geografia de la Republica Argen-
nos Aires) Dir. Nac. Minas Geol. 95:1-150. tina. Ibid. 2, 218 p.
Camacho, H. H. 1967. Las transgresiones del Cre- Gasparini, Z. B. de, Baez, A. M. 1975. Aportes al
tacico Superior y Terciario de la Argentina. Rev. conocimiento de la herpetofauna terciaria de la
Asoc. Geol. Argentina 22:253-280. Argentina. Act. I Congr. Argentino Paleontol.
Caminos, R. 1972. Sierras Pampeanas de Tucuman, Bioestr. 2:377-115.
1979 BAEZ & SCILLATO YANE: CENOZOIC OF ARGENTINA 155
Patterson-, B., Pascual, R. 1972. The Fossil Mam- filogeneticas y zoogeograficas sobre los Euphrac-
mal Fauna of South America, pp. 247-309 in tini. Act. I Congr. Argentino Paleontol. y Bioestr.
Keast, A., Erk, F., Glass, B. (eds.). Evolution, 2:449-461.
Mammals and Southern Continents. State Univ. Scillato Yane, G. J. 1977. Nuevo Megalonychidae
NewYork Press, Albany, 543 p. (Edentata, Tardigrada) de Edad Chasiquense
Paulo Couto, C. de. 1975. Mamiferos fosseis do (Plioceno temprano) del sur de la Provincia de
Quaternario do Sudeste Brasileiro, pp. 89-132 in Buenos Aires (Argentina). Su importancia filo-
Bigarella, J. J., Becker, R. D. (eds.). Internat. genetica, biostratigrafica y paleobiogeografiea.
Symp. Quat. Bol. Paranaense Geocien., Curitiba Rev. Asoc. Cienc. Nat. Litoral (Santo Tome,
33, 370 p. Santa Fe, Argentina) 8:45-54.
Polanski, J. 1963. Estratigrafia, Neotectonica y Scillato Yane, G. J. 1979. Nuevo Nothrotheriinae
Geomorfologia del Pleistoceno Pedemontano entre ( Edentata, Tartigrada ) de Edad Chasiquense
los rios Diamante y Mendoza (Provincia de ( Plioceno temprano ) del sur de la Provincia de
Mendoza). Rev. Asoc. Geol. Argentina 17:127- Buenos Aires, Republica Argentina. Su impor-
349. tancia bioestratigrafica, filogenetica y paleobio-
Polanski, J. 1965. The maximum glaciation in the geografica. Act. VII Congr. Geol. Argentino
Argentine Cordillera, pp. 453-472 in Wright, ( in press )
.
H. E., Jr., Frey, D. (eds.). International Studies Scillato Yane, G. J., Uliana, M. A., Pascual, R.
on the Quaternary. Geol. Soc. America. Spec. 1976. Un Megalonychidae (Edentata, Pilosa) del
Pap. 84. Plioceno de la Provincia de Rio Negro (Argen-
Ramos, V. 1970. Estratigrafia y estructura del Ter- tina). Su importancia bioestratigrafica y paleo-
ciario en la Sierra de los Colorados (Provincia de biogeografica. Act. VI Congr. Geol. Argentino
La Rioja, Republica Argentina). Rev. Asoc. Geol. 1:579-591.
Argentina 25:358-382. Simpson, G. G. 1940. Review of the Mammal-bear-
Reig, O. A. 1952. Sobre la presencia de mustelidos ing Tertiary of South America. Proc. Amer. Phil.
mefitinos en la Formacion Chapadmalal. Rev. Soc. 83:649-709.
Mus. Municipal de Cienc. Nat. Traditional Mar Solbric, O. T. The origin and floristic affini-
1976.
del Plata 1:45-51. ties of American temperate desert and
the South
Ringuelet, R. 1955. Panorama zoogeografico de la semidesert regions, pp. 65-99 in Goodall, D. W.
provincia de Buenos Aires. Notas Mus. La Plata ( ed. ) . Evolution of the desert biota. Univ. Texas
18 (Zool. 156):1-15. 250
Press, Austin, p.
Ringuelet, R. 1961. Rasgos fundamentales de la Tonni, E. 1974. Un nuevo cariamido (Aves, Grui-
zoogeografia de la Argentina. Physis 22:151-170. formes) del Plioceno Superior de la Provincia de
Rossi de Garcia, E. 1966. Contribution al conoci-
miento de los Ostracodos de la Argentina. I. — Buenos Aires. Ameghiniana 11:366-372.
Tricart, J. L. F. 1973. Geomorfologia de la Pampa
Formacion Entre Rios de Victoria, Provincia de
deprimida. Base para los estudios edafologicos y
Entre Rios. Rev. Asoc. Geol. Argentina 21:194-
agronomicos. Seer. Est. Agric. y Ganad. de la
208.
Nation, Inst. Nac. Tecnol. Agrop., Coleccion
Rovereto, C. 1914. Los estratos araucanos y sus
Cient., 12, Imprenta "CONI," Buenos Aires,
fosiles.An. Mus. Nac. Hist. Nat. Buenos Aires
202 p.
25:1-250.
Volkheimer, W. 1971. Aspectos paleoclimaticos
Rutland, R. W. Guest, J. E., Crastley, R. L.
R.,
del Terciario argentino. Rev. Mus. Arg. Cienc.
1965. and Andean uplift. Nature
Isotopic ages
Nat. "B. Rivadavia," Paleontol. 1:243-262.
208 (5001): 677-678.
Vuilleumier, B. S. 1971. Pleistocene changes in
Saxon, E. C. 1976. La prehistoria de Fuego-Pata-
the fauna and flora of South America. Science
gonia: colonizacion de un habitat marginal. An.
173:771-780.
Inst. Patagonia 7:63-73.
Scillato Yane, G. J. 1975. Nuevo genero de Dasy- Webb, S. D. 1976. Mammalian faunal dynamics of
podidae (Edentata, Xenarthra) del Plioceno de the great American interchange. Paleobiology
Catamarca (Argentina). Algunas consideraciones 2:220-234.
7. Quaternary Biogeography of the
High Montane Regions
of South America
1
Beryl B. Simpson
Department of Botany
Smithsonian Institution
Washington. D.C. 20560
USA
From 11°N to 55°S, the South American Simpson, 1975), the "Andes" actually consist
Andes span 66 degrees of latitude, a distance of a complex array of mountains that can be
of over S000 km (Fig. 7:1). At their widest partitioned in severalways for purposes of
point, they are about 500 km broad. Such an discussion. meaningful to discriminate
It is
enormous land mass, extending across one units of the Andes by partitioning the Cor-
hemisphere and projecting into the other, dillera into geomorphological units as done
must necessarily exhibit great diversity. Else- by Harrington (1956), Gansser (1973) and
where in South America, significantly high Simpson (1975) (Fig. 7:2). Although in this
mountains are found only in southeastern treatment, mountains often are referred to by
Brasil and along the southern edge of the the country in which they occur, the use of the
Guiana Shield (Fig. 7:1), but the area of country is merely for convenience as to lo-
these mountains above 2000 m elevation is cality; the units
being discussed are the geo-
very small. In several previous papers (Vuil- morphological units or their parts in the
leumier, 1971; Simpson, 1973, 1975), I dis- regions mentioned. A different and somewhat
cussed Pleistocene events in the Andes, but more detailed subdivision of the Cordillera
within the last few years, several studies have into "tectonic segments" was proposed by
added to the knowledge of Andean Tertiary Sillitoe (1974), Gansser (1973) and others.
and Quaternary history. In this review of the The boundaries of the major tectonic seg-
Quaternary history of the high montane re- ments correspond with the boundaries of the
gions of South America, I first review the geomorphological units (Fig. 7:2).
current information about the formation of
Perhaps the single most important aspect
the Andean Cordillera and the other montane of the geological history of all the Andean
regions. A background knowledge of the units is their recency of uplift (see Simpson,
historicalgeology of the regions is necessary 1975, Fig. 2). Although there is some dis-
because their differing ages have an im- crepancy in the exact timing of final uplift,
portant bearing on the composition and di- the last major upheaval of almost all units
versity of the floras and faunas that were was end of the Tertiary or within the
at the
affected by Quaternary climatic changes. The final uplift of the Sierra
Quaternary.
After outlining the historical geology, I briefly Nevada de Santa Marta, the most northerly
describe the modern climate and vegetation
part of the Cordillera, has been dated toward
of these regions and then turn to a discussion the end of the Pleistocene (Gansser, 1955),
of Pleistocene biogeographical events.
although the use of late moraines as evi-
dence can be questioned. Recent studies (van
GEOLOGICAL HISTORY OF THE der Hammen, 1974; van der Hammen et al.,
MONTANE REGIONS 1973) have continued to document that the
high elevation zones (above 2000 m) of the
The Andes eastern Andes of Colombia were produced
during the middle and later part of the
As mentioned by numerous authors (e.g., Pliocene. Garner (1975) questioned the evi-
1
Present address: of Botany, University
dence for late uplift of the northern moun-
Department
of Texas, Austin, Texas 78712, USA. tain ranges. The absence of traces of glacia-
157
158 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
tions before the last part of the Pleistocene, elusive evidence, because earlier moraines
used as evidence for late uplift (Herd and can be easily eliminated. Few studies ex-
Naeser, 1974), can not be considered as con- tending to the beginning of the Pleistocene
800 Kilometers
1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 159
have been made, but careful examination of Cordilleras extending through Ecuador have
one important long pollen core (Fig. 7:3) received little attention in recent years, but
clearly shows that plants characteristic of a summary of previous studies in this region
high elevations were absent in the Colombian (Saner, 1971) indicated three stages of final
Andes until the beginning of the Pleistocene uplift, the last of which occurred during the
(van der Hammen et al., 1973). Limited Pleistocene. This last uplift was postulated
studies have been made in the Cordillera to have accounted for all elevations above
Occidental of Colombia, but the available 2000 m (Sauer, 1965, diagram 7). The youth-
evidence seems to suggest that they were ini- fulness of the high volcanic peaks in Ecuador
tially uplifted later than the ranges to the is attested to, in part, by still active Volcan
east (Shagam, 1975) and that they reached Cotopaxi.
their present elevations after the end of the Earlier investigations of both the Eastern
Tertiary (Burgl, 1961; Haffer, 1970a; Herd and Western Cordillera through Peru and
and Naeser, 1974). In Venezuela, geological Bolivia (Steinmann, 1930; Dollfus, 1959/1960;
investigations of the Merida Andes (Shagam, Ahlfeld, 1970) indicated Pliocene-Pleistocene
1975) show several early orogenies with the final uplifts of about 1000 to 3000 m. More
last initiated in the Oligocene and continuing recently the orogenic history of the Andes
to the present. Significantly high elevations has been interpreted in relation to plate tec-
were achieved only relatively late in the tonics (James, 1971, 1973; Gansser, 1973).
Pleistocene (Schubert, 1974a). James, in his interpretation of the uplift of
The portions of the eastern and western the Bolivian Andes (
1971 ) , envisioned an ini-
Fig. 7:1. Distribution of high elevation areas in South America. Areas above 2000 m
are shaded on
the western side of the continent. For the southeastern Brasilian Highlands and the Guiana Tablelands, areas
above 1000 are indicated. Inserts give climate diagrams showing mean monthly temperature and precipitation
regimes at several high elevation stations. The changes in annual dispersion of rainfall and yearly patterns of
temperature from north to south along the Andes as well as the aridity of the western part of the central
Andes are evident. In the inset diagrams, mean monthly values of both temperature and precipitation are con-
nected. Areas shaded with vertical bars indicate times of excess precipitation; stippled areas are times of mois-
ture deficit. The left hand, vertical axis is calibrated in units of 10°C; the right hand vertical axis is calibrated
in units of 20 mm of precipitation. The horizontal axis indicates the month of the year ( reversed in different
hemispheres). Shaded areas along the base of the diagram show the months from which freezing temperatures
have been recorded. For each of the stations with an inset, the following are given: name, latitude, longitude,
altitude, number of years from which data were recorded, mean annual temperature and mean annual precipi-
tation for this period. A. Bogota, Colombia, 4°38' N X 74 "05' W, 2556 m, 94 years, 13.°2C, 940.9 mm.
B. Riobamba, Ecuador 00°22'S X 78°34'W, 3058 m, 8 years, 11.5°C, 1361 mm. C. Cajamarca, Peru, 07°08'S
X 78°28'W, 2621 m, 9 years, 14°C, 716 mm. D. Cuzco, Peru, 13°33'S X 71°59'W, 3312 m, 17 years, 12.5°C,
750 mm. E. Arequipa, Peru, 16°19'S X 71°33'W, 2525 m, 37 years, 13.8°C, 104 mm. F. El Alto (airport,
La Paz), Bolivia, 16°30'S X 68°12'W, 4105 m, 28 years, 7.5°C, 564 mm. G. Oruro, Bolivia, 17°58'S
X 67°07'W, 3708 m, 10 years, 7.5°C, 282 mm. //. La Quiaca, Argentina, 22°06'S X 65°36'W, 3459 m,
8 years, 9.5°C, 322 mm. I. Cristo Redentor, Argentina, 32°50'S X 70°5'W, 3829 m, 27 years, -1.7°C, 354
mm. /. Itatiaia, Brasil, 22°20'S X 44°43'W, 2200 m, 20 years, 12.9°C, 2417 mm. Data for Figure insets A-I
from Schwerdtfeger 1976) and inset J from Brade (1956). Climate data for a complete year are not available
(
CARNEGIE
RIDGE
SANTA MARTA
EASTERN COROILLERA
1 WESTERN CORDILLERA
]
CENTRAL COROILLERA
CORDILLERA OCCIDENTAL
CORDILLERA ORIENTAL
PRINCIPAL COROILLERA
1 PATAGONIAN CORDILLERA
tialCretaceous collision between the Nazca (Turner, 1972; Yrigoyen, 1972). Paralleling
and the South American plates (Fig. 7:2) the Principal Cordillera from about latitude
leading to the original outline of most of 14°S to about 44°S is the Coastal Cordillera
the Cordillera. He dated the final "crunch" (Fig. 7:2), which apparently was uplifted
and major upheaval as occurring at the end somewhat than most of the Andes
earlier
of the Tertiary. During this last period, the (Okada, 1971). Recent evidence (Cobbing
Altiplano was postulated to have been raised et al., Dalmayrac et al., 1977) indicates
1977;
to its present elevation of 3000 and 4500 m. that two small sections of this cordillera, one
Although James' (1971) study was directed in Peru and one in Chile, are of Precambrian
particularly at the Rolivian/Peruvian Andes, age. Cobbing et al. (1977) suggested that
his conclusions probably extend to the ma- the Arequipa Massif in Peru is a rifted por-
jority of the Andean Cordillera, because the tion of the old South American shield areas.
contact between the Nazca and South Amer- Plio-Pleistocene movements were primarily
ican plates extends from Ecuador to southern involved in altering preexisting structure
Chile. Gansser (1973) also discussed the (Dessanti, 1972).
Andean orogenies in terms of plate tectonics, Dott et al. (1977) indicated that the up-
but he included the areas of the Andes where liftof the Patagonian Cordillera of southern
the South American Plate is in contact with Chile was later than 77 to 81 million years
other plates (Fig. 7:2) and stressed the re- ago and that most orogenic movements were
cent evidence that initial movements caused
completed by the Miocene.
by contact of the plates were very early,
probably Mesozoic, in some areas. The Mountains of Southeastern Brasil
In northwestern Argentina, the Pampean
Ranges (Fig. 7:2) were raised (Simpson and In southeastern Brasil, the Serra do Mar,
Vervoorst, 1977) after final Quarternary up- Serra da Mantiqueira, and associated ranges
lift of the Principal Cordillera to the west were formed by the arching and fracturing
Fig. 7:2. Schematic drawing of the major geomorphological units of the Andean Cordillera (redrawn
from Simpson, 1975) and the principal shields of the South American continent (from Putzer, 1968). The
tectonic segments are drawn from Sillitoe (1974) and Gansser (1973). Numbers to the left of the dotted
lines refer to boundaries of tectonic segments ( transverse faults that segment the descending lithosphere ) as
numbered in Sillitoe (1974). Missing numbers were assigned to segments omitted here. Numbers included
the following: 1. Amotape Zone, the northern limit of the Central Andes. 2. Huancabamba deflection, locality
of the change in direction of the Andes from NW to NNE. 4. Pisco or Abancay deflection, a proposed divi-
sion of the Central Andes; a sharp step marks the beginning of the Coastal Cordillera. 5. Northern limit of the
Altiplano-Puno block. 6. A change in strike of the Andes from N to NW
and the narrowing of the Eastern
Cordillera (the Ichilo Fault or the Arica Elbow line). 10. The southern edge of the Puno block. 13. Boundary
of the Norte Chico and Central Chilean regions, the northern limit of the Central Valley and the southern
edge of the Precordillera and Sierra de Cordoba in Argentina. 16. Southern limit of the Cordillera Principal.
In Southeastern Brasil letters refer to tectonic belts (from de Almeida, 1966). Dashed line is the Tropic of
Capricorn.
Dibujo esquemdtico de las unidades gcomorfologicas maijores de la Cordillera de los Andes (redibujado de
Simpson, 1975) y de los principales escudos del continente Sudamericano (de Putzer, 1968). Los segmentos
tectonicos estdn dibujados de Sillitoe (1974) y Ganscr (1973). Los numeros indican las fronteras de los seg-
mentos tectonicos: 1. Zona de Amotape, el limitc nortc de los Andes Centrales. 2. Desviacion de Huancabam-
ba, localidad del cambio de direccion de los Andes de NO a NNE. 4. Desviacion de Pisco o Abancay, una
division de la Cordillera Central propuesta; un marcado paso sehala el cc-mienzo de la Cordillera de la Costa.
5. Limite norte del bloque Altiplano-Puno. 6. Cambio direccional de los Andes de N a NO
y el cstrechamiento
de la Cordillera Oriental (la falla de Ichilo o la linea de Codo de Arica). 10. Mdrgen sureno del bloque del
Puno. 13. Separacion del Norte Chico y Region Central en Chile, limite norte del Valle Central y borde sur
de la Precordillera y la Sierra de Cordoba en Argentina. 16. Limite sur de la Cordillera Principal. En el su-
destc del Brasil, las letras indican cordones tectonicos (de de Almeida, 1966). La linea entrecortada senala al
Tropico de Capricornio.
162 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
ELEMENTS
OUERCUS
PLEISTOCENE
PLIOCENE
ALNUS
500 000
Years Ago
1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 163
what in age, all of these basement rocks are ing from the Guiana Shield was disputed.
Most South American geologists tended to
essentially Precambrian in age, and the Bra-
favor an interpretation of very ancient initial
silian platform must have consolidated fol-
uplift (e.g., Oliveira and Leonardos, 1943)
lowing an Precambrian orogenic cycle, pos-
but some geologists such as Gansser (1954)
sibly about 1,800 million years ago. More
restricted orogenies apparently occurred be-
concluded that they were first raised more
tween 900 and 1,300 million years ago. A final recently, perhaps in the Cretaceous. Part of
the difficulty in interpreting the history of
Precambrian orogeny that affected 40 per
these mountains was the lack of fossils or
cent of the shield area (de Almeida et al.,
geological stratigraphic sequences that could
1973) has been dated at about 450 to 700
million years ago.
be correlated with those of known ages else-
where on the continent. Since 1960, rock
Because most of the Brasilian Shield above
1200 m is covered with Cretaceous sediments, samples from a portion of the highlands in
Guyana have been Rb/Sr dated as between
uplift to appreciable elevations must have oc-
1,500 to 2,000 million years of age (Snelling,
curred primarily during the Cenozoic. De
1963). In southern Venezuela, Rb/Sr dates
Freitas (1951) concluded that three periods
from several localities indicate rock ages of
of epeirogenic movements account for the
m
or higher. The first of these
1,700 to 2,100 million years (Hurley and
uplift to 1200
Rand, 1973). The moderninterpretation of
probably occurred at the end of the Creta- the geological history of the Guiana High-
ceous and the last two in the Tertiary. The lands (Lexico Estrat. Venezuela, 1970) is of
third may have extended into the Quaternary. a Precambrian basement about 3,000 to 3,100
It was during these last three phases of uplift million years old that was initially uplifted
that the Serras do Mar, Mantiqueira, Espin- about 2,000 to 2,100 million years ago. Sub-
haco and Borborema wer fully formed. sequent uplifts apparently occurred in the
Fig. 7:3. Pollen core diagrams (redrawn from van der Hammen, 1974, Figs. 3 and 5) from the high plain
of Bogota, Colombia. The base of the right hand core has been dated as older than 4 million years (Pliocene).
The uppermost ( separated ) part of the core on the left is from Fuquene, Colombia. High elevation ( subpara-
mo) elements first appear in the uppermost Pliocene. Appreciable amounts of pollen (indicating species that
are abundant) of paramo plants are found only at the beginning of the Pleistocene. In the diagrams, percent-
age of the total pollen recovered from a piece of a core that is represented by plants of a given vegetation type
is recorded from left to right. Zero percent (absence) is on the left, and 100 percent would fill the box. The
diagram also shows the times of first appearance in this area of the North American tree genera, Alnus and
Quercus which migrated into South America after the closure of the Panama Portal. During the Pleistocene,
interglacials are indicated by increases in the percentage of the total pollen that belongs to Andean forest
genera. In glacial times, pollen of paramo plants dominate. Increases in the amounts of Polylepis indicate coo]
but quite moist conditions. The dotted arrow shows the corresponding portions of two cores taken from the
high plain. On the right core, letters are designations of disjunct core segments.
Diagramas de cortes palinologicos (redibujados de van der Hammen, 1974, Figs. 3 y 5) de la Meseta de
Bogota, Colombia. La base del corte de la derecha tiene una cdad mayor que 4 millones de aiios (Plioceno).
La parte aha del corte de la izquierda es de Fequene, Colombia. Elementos de altura (subpdramo) aparecen
recien en el Plioceno tardio. Considerables cantidades de polen (indicadores de las especies mas abundantes)
de plantas del paramo se encuentran solo al comienzo del Plcistoccno. En los diagramas, el porcentaje del
total de polen cstd registrado en la base de izquierda a derecha; cero por cicnto (ausencia) estaria a la izquierda,
micntras que un 100 por cicnto llcnaria el diagrama. El diagrama tambien mucstra los tiempos de la primera
aparicion de los generos de drbolcs norteamcricanos Alnus y Quercus los cuales migraron a Sudamerica luego
de establccida la comunicacion en el 1st mo de Panama. Durante el Plcistoceno los inter glaciales estun indicados
por incrementos en el porcentaje del total de polen perteneciente a los generos forcstales andinos. En tiempos
glaciales, el polen de plantas del paramo domina. Los incrementos en la cantidad dc Polylepis indica condi-
ciones frias pero humedas. La flecha entrecortada muestra las porciones correspondientcs de los cortes tornados
en la sabana. En el corte de la derecha, las letras representan seementos de cortes disjuntos.
164 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Mesozoic, the Paleocene, and again later in ent across the continent. Moreover, because
the Tertiary. The recent history has been one of elevations, they have added com-
their
of intense erosion and minor eustatic changes. plexities due to orographic effects. A brief
description of the major climatic regimes pro-
vides a background against which Pleistocene
Summary
climatic changes can be assessed. This out-
This cursory analysis of the geological line consists of climate diagrams (Fig. 7:1)
history of the high mountains of South Amer- and brief descriptions of the major tempera-
ica has pointed out that there are two very ture and precipitation regimes. A much more
distinct sets of montane areas. All of the complete description of South American cli-
Andean Cordillera, extending the length of mates and references to local meterological
the western edge of the continent is very data can be found in Schwerdtf eger ( 1976 ) .
young; in fact, it was raised above sea level Trewartha ( 1961 ) still provides one of the
only after the end of the Cretaceous, and most lucid accounts of the more complex
elevations above 2000 m
were achieved only climates.
within the last 2 to 5 million years. The
faunas and floras of these high elevations
Tropical Andes
could not have migrated into high elevation
habitats nor begun to differentiate until the Most of the Colombian Andes and the
later Pliocene or Pleistocene. Quaternary cli-
Merida Range of Venezuela lies within the
matic changes did not, for the most part, Equatorial Trough (zone over which the
modify a preexisting biota but were part of Intertropical Convergence Zone travels dur-
the actual story of the initiation and de- ing the course of a year) and receives rain-
fall during two periods each
velopment of the high Andean flora and year. The time
of the heaviest precipitation is October to
fauna. For the mountains of southeastern
Brasil and the Guiana Highlands, which are November; a second maximum occurs in April
and May (Fig. 7:1A). Throughout this area,
composed of Precambrian/ Cambrian rocks
and which were progressively uplifted precipitation is greatest at intermediate ele-
vations (500-1500 m) and decreases with
throughout the Mesozoic and Cenozoic, the
Pleistocene was merely the latest stage of a elevation. Both the Caribbean and the Pa-
long developmental history. Prior to the cific slopes receive about 1500 to 2000 mm
was a well-developed and of rainfall annually. The very high precipita-
Pleistocene, there
established biota on which the climatic tion ( up to 14,000 mm per year )
on the Pacific
coastal lowlands decreases inland and with
changes of the Quaternary had the combined
effects of augmentation, decimation and mod-
elevation. The Cordillera Central of Colom-
ification.
bia has slightly higher amounts of annual
rainfall (up 2600
to mm
per year) than the
eastern or western Cordilleras. The major
MODERN CLIMATE inter-Andean valleys of Colombia ( Cauca and
Magdelena) are dry at low elevations, but
The Andean Cordillera, flanking the west- their slopes receive as much as 1000 to 3000
ern coast of South America, is strongly in- mm of rain (respectively) a year (Johnson,
fluenced by its proximity to the Pacific Ocean. 1976). The depth of these valleys and the
The highlands of southeastern Brasil, 10° to aridity of their floors have acted as barriers
20° longitude farther east, come under the to east-west migrations of montane forest and
as one goes south along the Andes, there is stantly during the year, with 24 to 28 days
a change in the dispersion pattern in the of each month receiving some form of pre-
rainfall, a reduction in the total precipitation, cipitation (Miller, 1976).
and an increasing difference in the amount On the eastern side of the Andes south
of moisture that falls on the eastern and of the tropic (actually south of 29°S), the
western slopes. The discrepancy between the Andean slopes become dry. The moisture-
two sides of the Andes reaches its maximum laden winds of the tropical Atlantic reach
across the southern Altiplano (Fig. 7:1E,F). the eastern
only to about 29°S. In addition,
On the western slopes of southern Peru and slopes are the rain shadow flank of the Andes
northern Chile, the climate is extremely arid and are too far inland to receive any oceanic
and produces a barren desert. At the same influences. As in the case of the western
latitude on the eastern slopes of southern rainfall be-
slopes at about the same latitude,
Bolivia, enough precipitation falls to support comes bimodal
briefly and then shifts to a
upper montane cloud forest. The severe arid- winter rainfall near
pattern of predominantly
is caused by a com-
ity of the western slopes the Rio Colorado and Rio Negro (37°S)
bination of several factors— 1) the presence (Prohaska, 1976). Precipitation increases
of a cold upwelling along the coast, 2) the southward. South of 37°S, the Andean slopes
flow of a cold oceanic current parallel to intercept sufficient moisture to support
de-
the coast, 3 ) the production of a rainshadow ciduous beech forest.
by the Eastern Cordillera, and 4) the strong From 27 °S to the southern tip of the
Pacific anticyclone that lies off the coast of
continent, the variation in the annual march
some years, the Pacific anti-
temperature becomes more and more pro-
Peru. During of
cyclone, the force of the Humboldt
current
nounced, and yearly effects overshadow any
and the amount of upwelling are lessened.
diurnal temperature differences. In southern
At such times, the Intertropical Convergence areas of the continent, the growing season is
Zone extends farther south than in normal restricted to a few months of the year.
in relatively heavy
years, thereby resulting
rains on the coast and coastal mountains. The
Eastern Tropical Mountains
sporadic occurrence of this phenomenon,
known as "El Nino," has led some researchers Along the coast of southeastern Brasil, 70
to postulate that a similar climatic pattern to 80 percent of the rain fallsbetween No-
was prevalent during glacial periods (see vember and April with July and August being
section on Theories of Glacial Climatology). the driest part of the year. Total precipitation
166 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 167
is
quite high (2000 mm) and is thought to
— 6°C have been recorded. Annual rainfall
be attributable to the buildup of the sub- at this elevation exceeds 2400 mm (Brade,
tropical anticyclone over the "hump" of Bra- 1956; Fig. 7:1J). Highlands farther inland
sil
(Trewartha, 1961). At elevations above from the coast are influenced primarily by
2200 m on Itatiaia, mean monthly tempera- the flow of unstable equatorial air that an-
ture differences during a year are between nually moves across central Brasil. Here,
15°C and 27°C. Minimum temperatures of rainfall is lower than along the coast, and
Fig. 7:4. Barriers to population expansions of high elevation taxa during either glacial or interglacial
periods. 1. Lowland area of northeastern Colombia separating Sierra Nevada from the Eastern Cordillera
of Colombia. 2. Low arid area separating the Eastern Cordillera of Colombia and the Merida Andes of Vene-
zuela. 3. Rio Magdalena Valley separating the Cordillera Oriental from the Cordillera Central of Colombia.
4. Rio Cauca Valley separating the Cordilleras Occidental and Central of Colombia. 5. Northern Peru low area.
In all of the cases of barriers 1-5, east-west or north-south dispersal of high elevation elements is now pre-
vented by areas of unsuitable low elevation habitat. During glacial periods, the effects of these barriers would
have been less severe. 6. Upper Maranon Valley, an arid valley that was a glacial and interglacial barrier
preventing east-west exchange of elements, probably more effective in integlacial than in glacial times.
7. Eastern side of Lake Titicaca where glacial ice flowed into the lake and prevented north-south biotic ex-
change. This area is not a barrier at the present time. 8. Clacial barriers for arid elements formed by lakes
and bogs across the surface of the Altiplano during cold periods. 9. Zone of continuous aridity across the Cor-
dillera which has served as a barrier to north-south migration during glacial and interglacial times. 10. Rio
Bio-Bio. A modern and glacial barrier to north-south dispersal because of the climatic change at this lati-
tude, the presence of the river itself and the glacial ice which followed the course of the river. Asterisks indi-
cate habitats isolated on mountain peaks during interglacial periods (paramos in the north and alpine habitats
in the south). During glacial times, these habitats expanded, facilitating exchange. The letter A refers to the
area of lowland rainforest and llanos separating the Andes from the Guiana Highlands and preventing east-west
colonization. In glacial times, stepping stones of subtropical habitat may have been present in this region. Let-
ter B indicates thorn scrub ( monte/chaco ) separating the Andes from the highlands of southeastern Brasil. A
modern barrier, presumably of narrower extent in glacial times. Letter C designates the Amazon lowlands which
separate the Guiana and the southeastern Brasil highlands. An effective filter zone in both glacial and inter-
glacial times but with stepping stones possible in glacial periods formed on the low tablelands of central Brasil.
The triangles indicate arid barriers between relic woodlands on the western slopes of the Peruvian Andes. In
humid periods, presumably during glacial times, these arid areas received sufficient moisture to allow continu-
ous forest growth.
Ban-eras para la expansion de las poblacioncs de biota de altura durante los periodos glaciates o intcrglacialcs.
1. Area de tierras bajas del noreste de Colombia separando la Sierra Nevada de la Cordillera Oriental de Co-
lombia. Area drida ij baja separando la Cordillera Oriental de Colombia y los Andes de Merida de Vene-
2.
zuela. Valle del Rio Magdalena separando la Cordillera Oriental de la Cordillera Central de Colombia.
3.
4. Valle del Rio Cauca separando las cordilteras Occidental y Central de Colombia. 5. Areas bajas del norte
del Peru. En todos los casos de las barreras 1—5, la dispersion este-oeste o nortc-sur de los clcmentos de alturas
estd prevenido por areas de poca clcvacion. Durante los periodos glaciates, los efectos de estas barreras hab-
rian sido menos severos. 6. Valle del Alto Maranon, un valle drido en cual durante la epoca glacial c intergla-
cial fue barrcra cual evito el intercambio de etementos de este-oeste; probablemente la nuis effectiva de las
la
dos fue la epoca 7. Lado oriental del Lago Titicaca donde el hiclo glacial flujo y previno el inter-
interglacial.
cambio biotico del norte a sur. Esta area no es una barrcra al tiempo presente. 8. Barreras glaciates para
etementos dridos fonnados por lagos y pantanos en el Altiplano durante periodos frios. 9. Zona de aridez
continua a lo largo de la cordillcra que ha servido de barrcra a la migracion norte-sur durante los tiempos
glaciates e intcrglacialcs. 10. Rio Bio-Bio, una barrcra moderna y glacial para la dispersion norte-sur a causa
del cambio climdtico a esta latitude, la presencia del rio mismo y del hiclo glacial que siguio el curso del rio.
Los astcriscos indican ambicntes aislados sobrc los topes de montanas durante los periodos intcrglacialcs (paramos
en el norte y ambientes alpuws en el sur). Durante el tiempo glacial estos ambientes se expandieron facilitando
el intercambio. La letra A reficre a las areas de sclva pluvial de tierras bajas
y de llanos separando los Andes
de las Alturas Guayanesas y previmiendo la colonizacion este-oeste. En tiempos glaciates, rutas de migracioncs
cntrccortadas de ambicntes subtropieales pucden habcr estado presente en esta region. La letra B indica un
matorral espinoso (monte/chaco) separando los Andes de las alturas del sudeste del Brasil. Una barrcra mod-
erna, probablemente de extension menor en tiempos glaciates. La letra C designa las tierras bajas amazonicas
que separan la y las alturas del sudeste brasileno. Un filtro efjectivo en los tiempos glaciates e intcr-
Guayana
glacialcs pero con rutas de migracioncs entrecortadas posiblcmente en periodos glaciates formadas en las mesctas
bajas del Brasil central. Los triangidos indican las barreras dridas cntre selvas relictuales sobrc las ladcras oc-
cidentates de los Andes peruanos. En periodos humedos, probablemente durante los periodos glaciates, estas
areas humedas recibidn suficicnte humedad como para permitir el crecimiento forestal continuo.
168 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
On
the summits of the Guiana Highlands,
rainfall is higher than on the surrounding
lowlands. The eastern group of these table-
lands receives about 2000 to 2500 per mm
year and the western portion about 3300 to
F MAMJJA SOND 3500 mm. At elevations above 1000 m, rain
falls almost continuously throughout the year
with slight reductions in March and Septem-
ORURO, BOLIVIA 17°S, 67°W (3703 M) ber (Maguire, 1970; Snow, 1976). Because
this area is so far removed from any oceanic
influences, it is probable that the moisture
has been derived, in part, from reevaporation
or from evapotransporation of precipitation
that previously fell as rain over the Amazon
Basin (Snow, 1976). Temperatures are cool
throughout the year, often dipping to 5°C
at night (Maguire, 1970).
are often changes in species of characteristic an upper limit at a somewhat lower elevation,
genera at key biogeographical points. In usually only to 3600 m, and can be consid-
each case, a description of the physiognomy ered the true "ceja." The dominant Podo-
of the vegetation is given so as to present a carpus is P. nubigenus and the
predominant
visual aspect of the habitats. More detailed Weimnannia, W. fagaroides. Oreopanax, Al-
treatments can be found in ( 1966 ) Hueck ,
nus (Juglandaceae) and Clusia (Guttiferae)
Hueck and 1972 ) and Cabrera and
Seibert (
are also very common. South of Santa Cruz,
Willink (1973). Recause so much detail is Bolivia, and continuing into northern Argen-
lost in maps of large scale, I have not at- tina (16°S-27°S) there is a noticeable differ-
tempted to include a vegetation map. Hueck ence in the vegetation. At lower elevations,
and Seibert (1972) produced the most com- Podocarpus (P. parleteoreii) Eugenia (Myr- ,
of 2400 to 3800 m and locally ascends to sae), Geranium (Geraniaceae), and Hyperi-
4200 m. The species of Weinmannia that cum Hypericaceae ) are familiar to temper-
(
dominate in Venezuela are W. jahnii and ate botanists. In wet paramos, the ground
W. microphylla, along with Podocarpus olei- is often spongy with accumulated sphagnum
folius, P. montanus, and P. rospiglossii. Spe- and plant debris, and the landscape is usually
cies of Oreopanax ( Araliaceae ) Ilex (Aqui- , shrouded in mist, if not drizzle. Analyses of
foliaceae), and Brunellia (Rrunelliaceae) are the vegetation have shown that the dominant
commonly mixed with other characteristic elements have not been derived by point by
elements. From northern Ecuador to central point vertical differentiation of the upper
Bolivia (1°N-16°S), the montane forest has montane flora (Simpson, 1975). Rather, the
170 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
differentiation of the flora has primarily been shrubs have many xerophytic characteristics.
horizontal within the paramo zone, although Tightly compacted mat plants, such as Azor-
several of the characteristic genera were orig- ella (Umbelliferae), Pycnophyllum (Caryo-
inally derived from Neotropical stocks that phyllaceae) and Werneria (Compositae), are
grew at lower elevations. Many genera and common. Along the eastern side of Altiplano,
species are restricted to one or two paramos the vegetation is often termed "wet puna"
(e.g., Espeletia, Dipholostephium, and Lori- (Troll, 1959), because it is more lush than
caria) or have undergone immense radiations that in the west. In the western portions,
within this regions. Other elements belong cacti, often ground-hugging, and various
to genera common in high elevations through- Bromeliaceae (Puya, TiUandsia) are locally
out the Andes (Senecio, Baccharis, Festuca, abundant. Trees (Polylepis) are found in
Mutisia, Polylepis) or to those which migrated humid microsites up to elevations of 5200 m,
from North or Central America after the clos- but individuals are widely spaced, small, and
ing of the Panamanian portal (Lapinus). very gnarled with reduced leaves.
The southwestern altiplano of Bolivia and
The High Elevations of Ecuador and Peru adjacent regions of northern Argentina and
and Easternmost Bolivia Chile, are exceedingly dry (see Modern
Climate ) even grasses are sparse. The domi-
;
There has been argument for years as to nant plants are scattered Compositae (Para-
whether the area of Ecuador and Peru with
stephia,LepidophijUum and Nordophyllum) .
low, humid, herbaceous vegetation dominated Shrubs (1-1.5 m tall) of these genera, spaced
by grasses and Compositae should or should up to 6 m apart, grow with small twisted
not be called paramo, but several authors shrubs of Adesmia horrida (Leguminosae).
(Cuatrecasas, consider the paramo
1968), Cacti are sometimes found, and in southern
zone to extend from Costa Rica to Peru. The Peru, bromeliads are locally dominant. In
most conspicuous difference between the intermontane valleys at lower elevations
areas south of northern Ecuador and those
(3400 m-1000 m) in southern Bolivia and
to the north is the absence of Espeletia to
Argentina, columnar cacti (Trichocereus
the south. Weberbauer
1945 ) tended to call
(
T. and scrubby trees
terscheckii, pascana)
humid high elevation areas in Peru "jalca." and shrubs Prosopis ferox, Zuccagnia, Legu-
(
However, in Ecuador the term "paramo" minosae; Gochnatia and Baccharis; Composi-
often is used as the common name for supra-
tae) are common. This semiarid scrub vege-
forest grasslands. Many of the same grass tation becomes dominant (with different spe-
genera found in Colombia are dominant there,
cies of Prosopis) the eastern
along slopes of
and the typical high elevation Ranuncula-
the Argentine Andes south of the ceja forest
ceae, Gentianaceae, Geraniaceae, legumes
and Compositae are common. (27°S) until the latitude (36°S) where the
deciduous Nothofagus forest begins to appear.
Beginning in southwestern Ecuador and
increasing in breadth across Bolivia to Argen-
tina (4 or 5°S-27°S) the supraforest vegeta- The West Slopes of the Andes of Colombia
tion becomes (sporadically) drier and usual-
is called "puna." As in the case of the Because of peculiar climatic events (Tre-
ly
wartha, 1961), the westernmost slopes of the
paramo, grasses and members of the Com-
Colombia Andes (extending into northern
positae dominate the vegetation. However,
the most abundant grasses here are Poa (P. Ecuador) are covered by extremely humid
montane forest, which extends narrowly in-
humilis), Stipa (S. ichu) and Festuca. The
puna extends from elevations of about 3400 m
land and to about 1500 m
elevation. Domi-
The Western Slopes of the Andes from carpus again becomes a dominant element (P.
Ecuador to Central Chile nubigenus) as well as Drimys winteri. Coni-
fers such as Fitzroya cupressoides, and Arau-
Along the west coast from south-central caria become conspicuous. Along with these
Ecuador to northern Chile (2°S-21°S), the
stately trees are laurels, including Laurclia
vegetation becomes increasingly xerophytic and Mijrceugenella (Myrtaceae). The forest
and sparse. In southern Ecuador into Peru becomes diminished in species and lower in
to about the latitude of Lima (12°S). the stature as the Andes descend and the climate
western slopes are covered with thorn scrub becomes harsher from 40°S to the tip of
and cacti. Legumes such as Prosopis and Tierra del Fuego. The beeches, Nothofagus
Acacia, Bromeliaceae (Puya), Cactaceae
pumilio and N. antarctica, although reduced
(Trichoccreus), and Compositae such as in size, continue to Tierra del Fuego.
Proustia, and Diplostephium, are
Franscria
dominant Weberbauer, 1945; Tosi, 1960).
(
The High Southern Andes
This scrub extends to, and merges with, the
dry puna at elevations of about 3000 m. In In the southernmost Andes (26°S to about
southern Peru, the vegetation becomes more 51°S) the area above tree line and below the
and more sparse until, west of Arequipa, only level of the glaciers is covered by humid
scattered bromeliads and a few cacti are visi- meadows similar in aspect to those of the
ble. In northern Chile, vegetation of any Alps. As would be expected, rosette herbs
prominence is virtually absent. When the are abundant, especially members of the
infrequent rains occur, ephemerals appear, Compositae (Perezia, Leucheria and Nassau-
but such periods of flowering can be decades via). Genera of families typical of the north
apart. temperate zone are common Ranunculus —
Once the latitude of about 21° to 27°S is (Ranunculaceae), Cardamine ( Crucif erae ) ,
called the evergreen rainforest or the Val- ma (Bignoniaceae), Geonoma, and other
divian Forest. In addition to the evergreen palms (Euterpe, Cocus). In the more south-
species of Nothofagus I>etuloides and Podo- ern sector, Eugenia (Myrtaceae), Roupala
172 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
speaking, to the Cerrado vegetation ("campos and to the resolution of some old problems.
rupestres" )
. As one might expect from the discussion
of the climate and vegetation of the different
The Guiana Highlands mountainous areas throughout the continent,
their Pleistocene histories differed consider-
The Guiana Highlands are the least bio-
ably depending on elevation, latitude, longi-
logically explored of any of the South Amer- tude and exposure. In the most general terms,
ican montane regions. Ry 1970, over 2,000 we can say that within the Pleistocene, and
plant species had been described from the
during times of world wide sea-level lowering
few explored tepuis. Maguire (1970) esti-
(glacial periods), cooler conditions existed
mated that at least 4,000 taxa eventually with or without, an absolute increase in
would be found. In contrast to the other moisture. In times of high, world-wide sea
regions described, endemism here
is extreme-
was a warming
ly high
—
perhaps as much as half of the flora.
levels
of
(interglacials),
climate sometimes
there
accompanied by in-
Certain families such as the Rapateaceae are creased precipitation.
almost confined to this area and have many
genera with species restricted to the various The Northern Andes
mountain tops. Among the other families with
especially large numbers of endemic genera As a result of 20 years of careful palyno-
and species are the Melastomaceae and the and geological work of van der Ham-
logical
Myrtaceae. men and his colleagues, a firmly dated and
Many primitive and endemic Compositae fairly complete sequence of changes over the
also are found in this region. Most of these, last 4 million years is available for the East-
Stenopadus, Glossarion, Neblina, Chimantea, ern Andes of Colombia (van der Hammen,
Quelchia, Achropogon and Duidea belong to 1974,and references therein). The sequences
the tribe Mutisieae and are considered by shown in a representative pollen core (Fig.
some (Carlquist, 1957, 1974) to be among 7:3) indicate several things. First, high ele-
the most primitive members the family. of vation, cool-climate-adapted taxa appear only
The relationships of the flora of these high- at the very end of the Pliocene. The earliest
lands are obscure, because of the high per- of these elements includes Myrica sp. (Myri-
centage endemic genera and families.
of caceae). As the plain around Bogota, from
There are few taxa closely related to groups which these cores were made, became pro-
found in the high Andes and only slightly gressively uplifted (van der Hammen, 1974;
more show with elements
floristic similarities van der Hammen et al., 1973), more and more
in thesurrounding lowlands and mountains of plant taxa characteristic of high elevations
southeastern Rrasil (Brade, 1956; Maguire, appeared in the fossil record [e.g., Lycopodi-
1970; Carlquist, 1974). urn (Lycopodiaceae); Gunnera (Haloragida-
1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 173
Andes is lower than that of the Colombian 1967). Mercer and Palacios (1977) dated
Andes. It is possible that there were fewer times of ice advance during the last glaciation
opportunities for speeiation in situ (fewer in the Cordillera Vilcanota. The major ad-
times of effective glaciation) than in the vance ended between 28,000 and 14,000 years
Colombian Andes. b.p. but minor advances occurred at about
Sauer's (
1971 ) summary of the research 11,500 and between 600 and 300 years b.p.
on Ecuadorian Quaternary history indicated There is still much dispute about the
three glacial periods, the last of which re- amounts of precipitation received in Peru
duced snow line 1500 to 2000 m. According during glacial periods. Nevertheless, with
to his reconstructions (Sauer, 1971, Fig. 7:8), the exception of Nogami (1972), most authors
the Ecuadorian Andes were being continu- have come to the conclusion that there was
ously uplifted throughout the Pleistocene, and an increase in total moisture received on the
the first glacial advance had only minor ef- western slopes of the Peruvian and adjacent
fects because of the lack of appreciable land Chilean Andes at some time during the vari-
above 2000 m. By the time of the last glacia- ous glacial cycles. One or more periods of
tion, the mountains had reached their present precipitation increase were postulated by
elevation. Despite the differing intensities Garner (1959) and Dollfus (1976), and ac-
feltfrom the three glaciations in Ecuador, cepted by Gansser (1973). Mercer and Pala-
Sauer ( 1971 ) correlated them with the Min- cios ( 1977 were unable to come to con-
)
del, Riss and Wiirm glaciations ofEurope. clusions about ice age wetness. Possible
During the last glaciation at least, southward moisture increases on the western montane
and northward (and probably to some extent by Simpson (1975), would
flanks, as outlined
across the central valley) migrations of high have allowed southward expansion of upper
elevation taxa occurred in cooler and/ or more montane forest down the Pacific facing slopes
humid Because the high elevation
periods. (Fig. 7:4). The present occurrence of relic
areas of Ecuador are now in the form of rela- patches of upper montane forest in sheltered,
tively isolated peaks (Fig. 7:4), many taxa humid canyons along the western Andes
exhibit differentiation from mountain to prompted Koepcke (1961) to postulate that
mountain. The restriction of taxa to moun- glacial periods had been accompanied by in-
tains of certain regions canbe seen in Poly- creases in precipitation in this region. An-
lepis (Simpson, in press), numerous other thropological finds now indicate that rivers
plant genera such as Niphogeton and Arraca- flowing from the Andes to the Pacific carried
cia (Umbelliferae), Llerasia and Mutisia much more water at earlier times in the
(Compositae) (see Simpson, 1975), and in Pleistocene than at present and were able to
some avian groups such as the Atlapetes support human settlements (Gansser, 1973).
schistaceous superspecies (Paynter, 1972). In eastern Peru, which normally receives
In Peru, where more stratigraphic work a plentiful supply of rain, the depression of
has been done, researchers ( Hastenrath, 1967; snow line and the increases in moisture would
Clapperton, 1972; Dollfus, 1976; and Nogami, have been relatively less pronounced. It is
1972), have interpreted levels of moraines as doubtful that even the "arid" phases experi-
phases of the last or last two world-wide enced by the Eastern Cordillera as envisioned
glacial advances (see below), although Stein- by Garner ( 1959, Fig.
8 ) would have altered
mann (1930), postulated that there were at the habitat sufficiently so as to have created
least three (one very ancient) glaciations in a north-south migratory corridor for arid or
Peru. In any event, moraines show that snow semi-arid elements. However, during periods
line was lowered about 1000 to 1500 m in the of increased precipitation on the highest por-
east and 500 to 1000 m in the west (the drier tions of the Cordillera, it is likely that there
slope of the Andes) several times during was an exchange of mesophytic taxa from
the Quaternary. In a north-south direction east to west across the nudos of the Peruvian
across Peru, the snow line depression was Andes (Koepcke, 1961, Fig. 2). These nudos
fairly constant, despite the present decreasing are east-west extensions of the Cordilleras
annual precipitation southward (Hastenrath, Oriental and Occidental that form "bridges"
1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 175
across low inter-Andean regions. The most patchy nature of suitable microhabitats and
pronounced nudos occur slightly north of drastic cutting by man. In the southern part
Lima (the Nudo de Pasco connecting the of the Altiplano near Tarija, fossil remains of
Cordillera Blanca and Negra with the Cordil- Megatherium, Scelidontherium, Macrauchen-
lera Oriental) and at about 15°S (Nudo de ia, and Equus (Ahlfeld and Branisva, 1960)
Vilcanota, connecting the Cordilleras Occiden- indicate the presence of grasslands in glacial
tal, Oriental and Auzangate). A common times. This region is now covered by desert
distribution pattern for plant taxa that indi- scrub vegetation.
cates such a migratory route is the presence In addition to humidity changes, studies
of populations in mesophytic habitats on the of moraines on the mountains to the east of
western slopes near Lima and then, disjunctly, Lake Titicaca indicate at least four significant
across the Andes on the eastern slopes in or advances of ice, some of which extended into
around the Urubamba Valley (Simpson, the lake itself. Unfortunately, no correlations
1975). of these advances have been made with ad-
Data from the entire puna surface is pri- vances elsewhere and it is uncertain whether
marily geological, although several of the they represent stages in a late glaciation or
lakes on the Altiplano eventually may prove independent major glacial advances.
to be sources of continuous pollen
reliable Northwestern Argentina and the northern
cores. The work of Kessler ( 1963 )
careful part of Chile differed in their Pleistocene his-
and the earlier work of Troll (1927), provide tory just as they now differ in their climate
a basis for interpretations about the changing and vegetation. In this area of northern
regimes and fluctuating snow lines of the Argentina, both the vegetation of the puna
Quaternary. Their findings indicate that dur- and the ceja forests reach their southernmost
ing "climax" phases of the last glaciation ( and limit. As in the case of the entire upper mon-
perhaps earlier) lake systems covered much tane forest from Colombia to this latitude,
of the Altiplano (Simpson, 1975, Fig. 20) and Pleistocene climatic changes probably had
that snow line was lowered 700 m below its little disruptive effect on the
continuity of
present level on mountains flanking the plain. the band of forest. A more significant change
Vegetational changes on the nonflooded por- that occurred during the Quaternary would
tions of the Altiplano during such periods have been the southward spread of many
may have been similar to those recorded near North American forest taxa that entered
Lago Junin in central Peru (cited in Dollfus, South America after the closing of the Pana-
1976), namely, the production of bogs, per- manian Portal. Important genera such as
haps seasonal bogs, dominated by members Alnus, Jiiglans, and Quercus merged with
of the Cyperaceae and Juncaceae. Under native tropical and subtropical taxa such as
such conditions, elements of the arid-scrub Drimys and Podocarpus as they spread south.
associations would have been restricted to Alders and walnuts are now among the most
higher, west-facing slopes of the Andes. The characteristic trees of the upper forests in
distributional ranges of xerophytic members northern Argentina. Although the addition of
of genera such as Lepidophyllum, Mutisia, these trees would not have altered the physi-
Perezia, Baccharis, Senecio (all Compositae) ognomy of the forest, the fact that they be-
and other genera, such as Cremolobus (Cru- came dominants and increased the amount of
ciferae), would have been severely frag- deciduousness would have had a pronounced
mented (see maps of present distributions effect on the insects and other animals for-
of some of these taxa in Simpson, 1975; Fig. merly associated with the montane forests.
7:4). In contrast to these taxa, elements such In the inter-Andean valleys of northern
as species of Pohjlepis, which are now re- Argentina, conditions were quite different.
stricted to river valleys and/ or cloud belt An analysis of many of the internally drained
areas, would have expanded their ranges. basins, such as the Bolson de Pipanaco
However, in the case of Pohjlepis, the exceed- (27°S), indicates that at least once in the
ingly restricted modern distribution pattern Pleistocene, presumably during glacial peri-
is a result of both the present localized, ods, some were covered by lakes (Simpson
176 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
and Verhoorst, 1977). In addition, many of Farther to the south in Chile between 27°
the higher peaks of the region such as the and 30°S, there is a poor record of Pleistocene
Nevados del Aconquija and the Sierra de conditions, although there are remnants of al-
Ambato were glaciated, with permanent snow luvial terraces in Copiapo (Paskoff, 1977).
lines reaching 500 m lower than any modern Still farther to the south
(30°-33°S), evidence
vestiges of ice. The combination of the for- of pronounced Quaternary events begin to
mation of these lakes and the increase in appear. This area, as indicated above, now is
permanent snow indicates that cold periods subject to pronounced changes resulting from
at the latitude of the tropic were accompanied latitudinal movements of the nearby atmos-
by increases in precipitation. If conditions pheric pressure system. During the Pleisto-
were appreciably wetter as well as colder, the cene, these movements must have been al-
monte vegetation now found in these valleys tered in some fashion. All of the areas above
would have been eliminated or, at least, 4000 m in this area were glaciated with ice
greatly restricted. The arid-scrub monte asso- extending to 2100 m elevation at 30°S, 2800
ciations presumably persisted farther south m at 33°S, 1700 m at 33°30'S (Maipo Valley),
on the east side of the Andes and/or in parts and to 1200 m at 34°S in the lower Central
of the modern Chaco. The depauperate na- Valley (Paskoff, 1977). This substantial in-
ture of the faunas now found in these isolated crease in the amount of ice formation indi-
northern inter-Andean valleys has been at- cates both a significant glacial depression in
tributed, in part, to the slow recolonization ambient temperature and an increase in total
of the isolated pockets, following Quaternary moisture received during the year. Terraces
periods of decimation (Mares, 1976). on the Pacific shore caused by increased pre-
On the western side of the Andes be- cipitation have been shown to correspond in
tween latitudes 18° and 27°S there is evidence time to periods of eustatic sea-level changes
of slight glaciation in the Chilean Provinces dated as glacial. The combined effect of
of Tarapaca and Antofagasta. Most of the cooler and wetter conditions in the region
evidence of Pleistocene glaciations has been would have led to an environment amenable
eroded away by modern weathering processes. to growth of southern beech forests
(Simp-
However, those moraines that remain indicate son, 1973; Fig. 7:6). Remnants of the Notho-
that on the Payachata Volcanos (18°10'S), fagus forest, which was pushed to the north
glaciations reached to 4500 m. Ice is now and capable of surviving in this area during
found only at elevations over 5500 m. On glacial periods, can now be seen in small
Volcan Sajama as well, ice lobes may have areas such as Fray Jorge (30°30'S), where
reached as low as 4500 m and on Yaricova woodlands with Drimys, Myrceugenia, Gun-
20°S down to 4000 m (Paskoff, 1977). Near nera, and Escallonia, still persist. Recently
the Salar de Huasco on the western edge of there has been discussion as to whether or
the Altiplano, Tricart (cited in Paskoff, 1977) not these woodlands are Pleistocene or pre-
found evidence of two, an older and a Quaternary relicts (Kummerow, et al., 1961).
younger, glaciations, both of which left mo- Geological data (Rirot, 1970) and paleon-
raines at about 4200 m. Although most gla- tological evidence ( Hoffstetter and Paskoff,
cial advances in this area remain undated, 1966) support an hypothesis of a Pleistocene
Tricart correlated these with times of high age for these woodlands. Consequently, the
shorelines (presumably equivalent to lower last time(s) forests reached these areas would
sea levels) and hence, glacial periods on a have been during cold, wet phases of the
world-wide scale. Pleistocene.
Lakes were formed during cold periods In the centralmost portion of Chile at
in Tarapaca near the Peruvian border, as about the latitude of Santiago (33°S) and
indicated by cold-water-adapted diatoms southward (to 39°S), evidence of glaciation
found fossilized in dry lake basins. The Cen- is abundant. Alpine glaciers extended down
tral Valley in the northern part of Anto- to elevations of 2800 m at Portillo and to
fagasta also was covered by a lake (Paskoff, 1600 m at Guardia Vieja (Paskoff, 1977). In
1977). the valley of the Rio Bio-Bio, a low area of
1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 177
'.
'
l r . i
u.X
ul
*-»
i 5
O Q IDDI
178 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
the Cordillera (Simpson, 1973), an ice tongue occurred at 57,000 and 43,000 years b.p. Paly-
apparently reached as low as 1100 to 1200 m nological studies from cores showed that the
(Fig. 7:6). Although so postulated by Briig- coldest period was during the earliest ad-
gen ( 1950 ) there has been no recent evidence
, vance and that grasses and Compositae
in this part of Chile to support an hypothesis become dominant during all of the glacial
of an ancient glaciation. Paskoff ( 1977 ) ac- stages. Several taxa now present, including
cepted the evidence for only two, relatively some species of Nothofagus and Podocarpus,
recent, major advances in this region. Fossil disappeared from the island. However, dur-
remains of cool-habitat-adapted vertebrates ing the middle advance, other taxa, such as
(Mastodon, Mylodon, Equus) have been as- Poclocarpus andinus, now found only in the
sociated with the last of these advances Andes 1200 to 1800 m, were
at elevations of
(Casamiquela, 1969-1970, in Paskoff, 1977). present. Other high latitude herbaceous spe-
Traces of these large vertebrates disappeared cies, such as Lycopodium fuegiana and Dra-
after 11,000 years b.p. pctes muscosa (Thymelaeaceae), were found
In the southern lake region there firm in fossil associations with
is
Podocarpus andi-
documentation for four glaciations. Between nus. Thus, it is possible that parts of Chiloe
latitudes 39°-40°S, ice reached out from the did harbor in glacial periods plant (and ani-
Andes on several occasions westward to the mal) taxa that are normally characteristic of
base of the Coastal Cordillera. The last two higher elevations and latitudes.
of these advances have been named the Rio On the eastern side of the Andes, a wealth
Negro and El Salto, respectively. The latter of new data has been gathered and sum-
corresponds to the Llanquihue Glaciation of marized by Mercer (1976) about Pleistocene
Heusser (1974). The earlier advances have conditions of the lake region between 39°
neither been named nor dated. and 50°S. With the exception of the Sabana
South of 40°, the Pleistocene geology is de Bogota in Colombia, this area now has
simplified.Ice from the Andes flowed west- the most complete and well-dated Quaternary
ward and downward into the Pacific, virtu- sequence of any region of South America.
ally eliminating all of the biota on the western Mercer and his colleagues have dated the old-
est glacial advance with geological remains
slopes (Fig. 7:6). However, the flora and
fauna of the southern forest and bog associa- (49°28'S) at 3.5 million years. The glacial
tions persisted, because the climatic condi- advance which left the most extensive re-
tions allowed it to expand toward the Equator mains occurred 1.2 million years ago. Sub-
north of the sheet of ice (Figs. 7:6-7; see also sequent to this, several smaller advances, cul-
Vuilleumier, 1971, Fig. 2, for a more detailed minating at about 13,000 years b.p., took
map of the extent of glacial ice). place. The most extensive of this last series
For biogeographers interested in Quater- was about 56,000 years b.p. In his analysis,
Mercer (1976) concluded that the Patagonian
nary biotic changes, one area in southern
Chile has long remained controversial the — gravel, hypothesized by Auer ( 1970 ) to have
Island of Chiloe. Many authors have postu- been deposited by a piedmont glacier, is, in
fact, a composite mixture of glacial outwash
lated that the islandwas free of glacial ice
and served as a refugium for southern ele- deposited over a time span dating from the
ments at times of glacial maxima. Others mid-Pliocene.
have claimed that the island was covered by Correlation of the Argentine glacial se-
ice and could have served no such function. quences with those of the Northern Hemi-
Recent geological and palynological investi- sphere is still uncertain, but three advances
gations by Heusser and Flint (1977) settled during the Recent period (4,600-4,200 years
the debate. The northern part of the island b.p., 2,700-2,000 years b.p., and ca. 1250 AD)
was, indeed, covered by glacial ice during coincide with advances of glacial ice else-
parts of three independent advances. The
where in the world (Mercer, 1976).
first advance has not been dated, but the last Mercer's work does not include palyno-
two have been radiocarbon dated as having logical investigations but some conclusions
1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 179
STRONG
GLACIAL
Fig. 7:7. Climatology of a maximum glacial (A) and interglacial (B) cycle following Fairbridge (1972)
showing the proposed ocean current, pressure systems and wind direction changes. Continental outlines and
the extent of glacial ice were drawn from Vuilleumier (1971). It is assumed in this model that during glacial
periods, the intensities of the Atlantic and Pacific anticyclones (high pressure systems) and the force of the
Humboldt Current were increased. The Amazon Basin was drier than at present because the influx of humid
air from the northeast was prevented. Dotted arrows are January wind directions, solid lines are wind direc-
tions in July.
CUmatologia de un pcriodo glacial maxima (A) y de un interglacial (B) siguicndo a Fairbridge (1972).
Se muestran las corrientes oceanicas propuestas, los sistemas de )ircsion, y los cambios de direction de los
vientos. Los limitcs continentales y la extension de los hielos glaciates fueron tornados de Vuilleumier (1971).
Se asume en cste modelo que durante los periodos glaciates, las intensidades de los anticiclones (sistemas de alto
prcsion) del Atlantico y del Pacifico, y la fuerza de la corriente de Humboldt se incrementaron. La cuenca
amazonica fue mas scca que a lo presente debido a la auscncia de los vientos humedos dcsde el noreste. Las
flcchas entrecortado.s indiean direction del viento en Enero; las flcchas continuas, la direction de los vientos en
Julio.
about the effects of the long series of glacial glacial periods (Fig. 7:6). Forest elements,
advances on the biota of southeastern South in contrast, would have been shifted in distri-
America can be drawn. First, the biota obvi- bution, but their ranges would not have been
ously has been subjected to periods of intense subjected to fragmentation either during gla-
cold interspersed with periods as warm as, or cial or interglacial periods.
warmer than, the present, since the middle The investigations of Mercer (1976) also
of the Pliocene. Such fluctuations, as previ- confirm the glacial limit at the base of the
ously outlined by Simpson (1973), would Andes on the eastern flanks as first established
have caused the high elevation biota of the by Caldenius ( 1932 ) This limit of glacial ice
.
region to migrate northward and/or down- provides a guide to the eastern limit of the
ward several times. For elements in habitats southern beech forests during glacial periods.
above tree line, a succession of downward The tableland of Patagonia would have been
and outward migrations followed by retreat neither a solid sheet of ice nor a broad ex-
upward would have led to rapid differentia- panse of forest. Rather, it probably was a
tion of populations that were reduced in size boggy moorland periodically subjected to
and restricted in distribution durinff inter- flooding and deposition of rubble during gla-
180 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
cial periods. For further discussion of Qua- len,1970) argued that the cooling on this iso-
Patagonia, see Baez and lated mountain could not have been sufficient
ternary climates in
Yane to have causedice formations at this low
Scillato (this volume).
elevation. Consequently, the presence or
absence of an alpine glacier on Itatiaia during
Eastern Highlands
Pleistocene glacial periods still remains un-
In Eastern South America, there is poor certain.
documentation for events during the Qua- There is no doubt, however, that the high-
of est parts of Itatiaia, the rest of the Serra da
ternary and some debate as to the effects
various climatic changes. In the Serra do Mantiqueira, and other highlands in south-
Mar, Bigarella and Mousinho (1966) de- eastern Brasil were subjected to periods of
scribed a series of terraces that they ascribed cold and (at least) increased aridity relative
to changes in humidity during the Pleisto- to that of the present. In terms of the biota,
cene. Whenan arid climatic cycle occurred, the periods of cold would have led to a de-
valleys were filled with debris (from short- pression of vegetation zones accompanied by
lived stream flow) and sudden but sporadic an increase in the superficial extent of the
showers produced slope retreat at the base vegetation zones of the upper elevations. This
of mountains leading to the formation of pedi- increase in superficial distribution would have
ments. In succeeding humid phases these promoted migration to, and exchange of ele-
pediments were partially degraded (because ments with, neighboring mountains (Fig.
of continuous washing by rainfall), and the 7:4). Several authors (Brade, 1956; Smith,
debris in the valleys were partially swept and 1962; Miiller, 1968; Klein, 1975) have pointed
scoured away. A series of such climatic out the relationships of components of the
changes led to the superposition of eroded biota of the southeastern highlands with the
pediments and stream-bed patterns that these biotas of the Andes and the lowland south
authors correlated with times of low sea level temperate regions. Many genera, such as
or glacial periods. However, as Beurlen Araucaria, Poclocarpus, which show connec-
(1970) has cautioned, beach levels per se are tions with the Andes, are remnants of ancient,
not necessarily an accurate indication of gla- widely distributed temperate forests (i.e.,
cial advances. Other corroborative evidence Early Tertiary), but many of the herbaceous
of temporal correlations is needed. This evi- genera, such as Azara (Flacourtiaceae),
dence is provided, in part, by pollen se- Boopis (Calyceraceae), Escallonia (Escallon-
quences found elsewhere in South America iaceae), which are now found disjunctly in
that indicate dry phases were coincident (in the highlands of southeastern Brasil may have
tropical areas) with glacial advances in the spread to this region during cool periods of
Andes (van der Hammen, 1974). the Pleistocene. Other plant genera that are
At higher elevations, in the Serra da Man- more widely distributed in high elevation
tiqueira, there are geomorphological remains and/or temperate regions of South America
that have led some authors ( cited in Beurlen, and also occur disjunctly in the highlands of
to the formation of a perma- southeastern Brasil (e.g., Jamesonia, Gyno-
1970) postulate
nent glacier on Itatiaia during glacial times. grama, Polypodiaceae; Anemone, Clematis,
Such formations include the presence of Ranunculus, Ranunculaceae; Berberis, Ber-
and Labiatae) also
huge scattered boulders and U-shaped valleys beridaceae; Lepicliina,
may have reached these mountains during
(Ebert, 1960). However, there are no actual
moraines. It is possible, in opposition to the cold periods.
In the Guiana no
conclusion of Ebert, that the boulders and Highlands, virtually
the U-shaped valleys were products of mud Pleistocene geological studies have been con-
flows and solifluction rather than glacial ice. ducted. All of the presumed reconstructions
of Quaternary climatology have been made
Moreover, the presumed elevation of the
Itatiaia glaciation was 2100 to 2500 m, an from inferences derived from data from other
elevation lower than any Andean glaciation areas. Such data from the Andes to the west,
at similar latitudes. Odeman (cited in Beur- the southeastern Brasilian Highlands to the
1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 181
south,and the Caribbean to the north (Ro- remotely related to the flora of either the
nattiand Gartner, 1973) leave little choice Andes or the surrounding lowlands.
but to assume that the Pantepui tablelands
were cooler in times of glacial advance else-
where in the tropics of South America and THEORIES OF GLACIAL
that vegetation zones of the upper elevations CLIMATOLOGY
were lowered similarly to those elsewhere.
Refore extensive studies in the Southern
As in the case of the paramos and the campos
of the Brasilian highlands, a lowering of the Hemisphere had been made, many authors
upper elevation vegetation zones on these (e.g., Budel, 1951) assumed that Pleistocene
tablelands would have slightly increased their climatic changes in both hemispheres involved
areal distribution and produced patches of simply an equatorial shift and latitudinal
subtropical forest on low mountains that are compression of the earth's climatic belts. The
now covered by tropical forest. These changes net effect of such shifts would have been to
in vegetation patterns would have fostered push high latitude climatic zones toward the
exchange among the tepuis and increased Equator and to reduce the latitudinal extent
of the mid-latitude semi-arid zones. This
exchange of some elements with the Andes
model of Quaternary climatology was based
(Mayr and Phelps, 1967; Haffer, 1970b).
Cook 1974 ) in a study of the origin of the
,
on European data and incorporated the ap-
(
avifauna of these highlands, pointed out that parent synchrony of cold and wet periods. It
there is little correlation between the areal is now clear that in
many regions portions of
extent of the modern upper
elevation zones, or glacial cycles (often the coldest part) were
the complexity of the vegetation of the vari- arid rather than wet (Hammond, 1976). In
ous tepuis, and the number of breeding bird South America, investigations have shown
taxa found on each. This lack of correlation that the Quaternary climatic picture was par-
that there is not an equilibrium as ticularly confusing, because the continent
implies
would be expected following the predictions spans two hemispheres and is dominated by
of the model of island biogeography (Mac- exceedingly high mountain ranges. A further
Arthur and Wilson, 1967). Cook interpreted impediment to an elucidation of the Pleisto-
the lack of an equilibrium as an indication of cene climates across the entire continent has
been the lack of precise temporal correlations
the recent arrival of the avifauna. Mayr and
1967 ) also pointed out that the ma- for what appear to be actual discrepancies in
Phelps (
an original, ancient avifauna. Nevertheless, it Some authors (e.g., Nogami, 1972) have
is interesting that there is little floristic con- suggested that there was no major change at
nection between the tepuis and the Andes, any time in the Pleistocene in the atmospheric
despite the evidence from avian relationships. circulation pattern over western South Amer-
In contrast to the relatively low level of en- ica. On the basis of observations that the
demicity in the avifauna (10-30%), over 50 snow line in the Andes during glacial ad-
percent of the flora is estimated to be en- vances was parallel to that now found, No-
demic. Moreover, the relationships of the gami ( 1972 ) concluded that the "apparent"
flora of the Guiana Highlands, albeit distant, increase in moisture was only a reduction in
are primarily with the highlands of south- evaporation caused by lower temperatures.
eastern Brasil or even Africa rather than with However, most authors disagree, favoring in-
the Andes. In view of the presumed recency stead changes in the force and/or location
of the avifauna and its successful migration of the various high pressure systems that af-
eastward during the Pleistocene, it is puzzling fect the South American climate. Fairbridge
that the flora has remained so distinct and (
1972 ) divided, for purposes of glacial clima-
182 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
tology, cycles of climatic change (considered would have become comparatively warm, and
to be periods of 100,000 years) into four sub- glaciers would have retreated in Antarctica
stages
—thekataglacial with a cold, wet cli- and on the Eastern Cordillera. Garner (1959)
mate; the peniglacial with a very cold, very proposed that at this time there was increased
dry climate; the anaglacial or initial inter- precipitation on the Western Cordillera and
glacial phase with a cool, diy climate changing the expansion of glaciers there (Fig. 7:8).
to mild and wet conditions; and the true Garner's hypothesis, while taking into ac-
interglacial with conditions similar to those count the complexities of the Andes and the
of the present. As this complex scheme indi- apparent nonsynchrony of glacial advances
cates, times of glacial advance would involve on the two sides of the Peruvian Andes, ap-
both arid and humid periods; Fairbridge pears to conflict with data from the Brasilian
(1972) attributed the older associations of lowlands (Haffer, 1974) that there was arid-
glacial periods and "pluvials" to interpreta- ity, rather than increased humidity, during
tions made on data from anaglacial or kata- full glacial phases. The Eastern Cordillera
glacial phases of the cycle. Fairbridge further is under the same climatic influences that
proposed that the aridity of the full glacial affect the Amazon Basin. Nevertheless, it is
phase in the low latitudes of southern conti- possible that the height of the Andes was
nents was caused by ocean-surface cooling sufficient to cause increased precipitation at
and the buildup of semi-stable high pressure high elevations when the air from the south-
areas near tropical continents (Fig. 7:7). In east that picked up moisture and caused
South America, semi-stable highs would have aridity in the lowlands reached them (see
been located off the north coast of the also Fairbridge, 1972).
Guianas, in the Atlantic Ocean off the tip of In the most recent discussion of Andean
Brasil, and off the coast of Peru (Fig. 7:7). Pleistocene climatology, Dollf us ( 1976 ) stated
Although the model proposed by Fair- that the climatic conditions on the two sides
bridge ( 1972 ) explains the climates that have of the Equator differed during glacial maxima.
been proposed for eastern tropical South North of the Equator, cold periods were asso-
America, does not account well for the com-
it ciated with dry conditions during peniglacial
plexities exhibited in the western tropical phases, as proposed by Fairbridge (1972);
Andes nor for conditions found at high lati- south of the Equator, peniglacial phases were
tudes. Garner (1959) argued that Pleistocene cold and wet. Dollfus postulated that during
glacial advances on the eastern and western major glacial advances, the equatorial high
slopes of the Peruvian Andes had been out of pressure system off western South America
phase (Fig. 7:8) and suggested that times of "weakened" and the onshore winds were re-
expansion of glacial ice on the Eastern Cordil- duced in force. The strength of the Humboldt
lera were synchronous with those in Antarc- Current was correspondingly reduced, and
tica (i.e., in phase with glacial advances in the northern Andes of Peru received increased
most areas of the world). He hypothesized moisture in the form of showers. This hypoth-
that during such periods, the southern oceans esis conflicts with that of Garner ( 1959 )
.
and the Humboldt Current (originating in One point emphasized by Dollfus is that local
southern waters ) would have become increas- conditions would have altered major climatic
ingly cold as the glacial cycle progressed, patterns and led to anomalies during Pleisto-
causing a general cooling of the atmosphere cene glacial periods. The aridity of the Alti-
in South America. The resulting general tem- plano, which he dated as glacial, would have
perature depression would have led to the been a result of such local phenomena. He
accumulation of ice (reduced melting) in based this conclusion on the fact that, under
areas where moisture was plentiful, such as present conditions, the Altiplano is driest
over the Eastern Cordillera. He further pro- when the Amazon Basin receives abnormally
posed that the ^Vestern Cordillera would have high amounts of rainfall. It is evident that
experienced increased aridity. During the in- Dollfus' hypothesis conflicts with the findings
terglacial cycles in the Antarctic region, the Amazon Basin
that during glacial periods the
southern oceans and the Humboldt Current was dry, not wet, and with the studies of
1979 SIMPSON: QUATERNARY OF MONTANE REGIONS 183
CORRESPONDING
LEVELS l ALTIPLANO
PEDIPLAIN
THROUGH
FLOWING
ARROYO RIVER
PRESENT
ARID
HUMIDITY
SLIGHT GLACIAL
MINOR
HUMIDITY SUBARID
ARIDITY
HUMID
GLACIAL
MAJOR
HUMIDITY
SEVERE
ARIDITY
SEVERE
ARIDITY HUMID
SEVERE GLACIAL
MAJOR
HUMIDITY SEVERE
ARIDITY
SEVERE
ARIDITY
MAJOR
HUMIDITY
HUMIDITY
SEVERE ARIDITY
Fig. 7:8. Geomorphological sequences indicating Pleistocene climatic changes on the eastern and western
slopes of the Peruvian Andes after Garner (1956). The eastern sequence is from the Urubamba Valley,
Cordillera Oriental, Peru and the western sequence from southern Peru in the Cordillera Occidental west of
Arequipa. Shading indicates aggredation accumulated during times of aridity. In humid periods, strongly
flowing rivers and streams incise the rubble previously deposited in stream beds. Numbers are assigned to
level of the same presumed age. Non-synchrony of arid and humid phases on the two Cordilleras is indicated
by the divergent geomorphological processes at the same level.
Secuencias geomorfologicas indicando los cambios climdticos del Pleistoceno en las vertientes oriental y
occidental de los Andes peruanos de acuerdo a Garner (1956). La secuencia oriental corresponde al Valle de
Urubamba, Cordillera Oriental, Peru, y la sceuencia oeste corresponde al sur del Peru, en la Cordillera Occi-
dental al oeste de Arequipa. El sombreado indica deposiciones acuniuiadas durante los tiempos de aridez. En
periodos humedos, fuertes flujos fluviales y de arroyos carcomieron la coma de piedra previamente depositada
en la eorriente. Los numeros corresponden a niveles de una misma presupuesta (clad. La ausencia de sinconi-
sidad de las fuses humedos y dridas en las dos Cordilleras estd indicada por los procesos geomorfologicos diver-
gentes al misma nivel.
184 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Kessler (1963) that showed increased pre- un levantemiento final e importante al tiempo
cipitation and lake formation across the del Plio-pleistoceno. Por el contrario, las Al-
Altiplano during glacial periods. turas Brasilenas del Sudeste y las Alturas
Obviously there is still disagreement about Guayanesas han sido masas de tierras conti-
the time of Pleistocene humid phases during nentales desde tiempos pre-Cambricos. Ele-
or after glacial cycles throughout South vaciones hasta 1000 m
o mas probablemente
America. Nevertheless, the various theories ocurrieron desde el Cretaceo. Sin embargo,
about Pleistocene climatology show the im- como una consecuencia de las diferencias
portance of taking local factors into account temporales para guarecer plantas y animales,
and the dangers of overgeneralization from las alturas del Este tienen una biota mas an-
data derived from a limited area. As empha- tigua y endemica que la de los altos Andes.
sized by Beurlen (1970), more palynological Esto es porque todos aquellos plantas y ani-
data and absolute datings are necessary be- males que viven por sobre 2000-3000 m de
fore a truly meaningful and comprehensive elevacion en los Andes han sufrido inmigra-
picture of South American Quaternary clima- cion, colonization, y diferenciaeion solo desde
tology can be drawn. el termino del Terciario. Los eventos cli-
maticos y geologicos del Cuaternario fueron,
entonces, una parte integral de la formation
ACKNOWLEDGMENTS de esta biota. En las alturas del Este, los
elementos del Cuaternario constituyeron mer-
Thanks are due J. Haffer and J. L. Neff amente una etapa mas de una larga historia
for their helpful suggestions on the manu- evolutiva.
script and to A. Tangerini and S. Yankowski Dada la extension latitudinale de la Cor-
for the excution of the drawings.
dillera de los Andes, clima y vegetation difier-
en grandamente de norte a sur. Del mismo
RESUMEN modo, los efectos de los eventos climaticos
del Cuaternario fueron diferentes de acuerdo
Las regiones altas (sobre 2000 m) de a la position latitudinal. En general, las tem-
Sudamerica ocurren principalmente en la peraturas disminuyeron a traves de la Cor-
parte oeste del continente donde la Cordil- dillera en forma sincronica con las expan-
lera de los Andes corre paralela a la costa siones mundiales de los hielos glaciales (de-
por mas de 8000 km. Otras regiones altas con nominados aqui "periodos glaciales"). A lo
elevaciones por sobre los 1000 m
se encuen- largo de los Andes, la linea de las nieves tam-
tran en el sudeste del Brasil y a lo largo de la bien bajo. El algunas areas, como en los
frontera venezolana-guayanesa-brasilera. Los Andes del norte, porciones de los ciclos gla-
areas por encima de los 2000 m
en estas dos parecen haber sido mas secos que lo
ciales
regiones, sin embargo, son bastante pequenas. que son hoy en dia. En otros areas,
ellos
Las regiones montanosas del sudeste brasileiio como en el Altiplano y en las regiones tem-
son conocidas colcctivamente como las Al- pladas del sur, las condiciones glaciales
turas Brasilenas del Sudeste, y aquellos del parecen haber sido mas humedas que al
norte comoAlturas Guayanesas. presente. El numero de ciclos de expansion
Geologicamente, el sistema andino comple- de los hielos glaciales a lo largo de la Cor-
to es muy joven, teniendo casi todas sus dillera varia desde un episodic tardio en
por-
ciones bajo el nivel del mar hasta al final del algunos recientemente elevados picos en el
Cretaceo. Aiin cuando hay diferencias en extreme norte, masta al menos cuatro en la
Guayanesas, los periodos glaciales fueron sin Bigahella, J. J., Moushino, M. R. 1966. Slope de-
duda oportunidades en que los ambientes de velopment in southeastern and southern Brazil.
Z. Geomorphol. 10:150-160.
altura se expandieron y, consecuentemente,
Bihot, P. 1970. Esquisse comparative des differents
durante las cuales las migraciones dentro de types de modele en roches cristallines dans la
los sistemas montaiiosos se facilitaron. En el zone temperee chaude de l'Amerique du Sud.
caso de Rev. Geogr. Phys. Geol. Dynam. 4:299-312.
las montarias del sudeste brasileno,
Bonatti, E., Gartner, S., Jr. 1973. Caribbean
una inmigracion desde los altos Andes y/o
climate during the Pleistocene ice ages. Science
desde los areas temperados de altas latitudes 244:563-565.
tambien ocurrio durante los ciclos glaciales. Brade, A. C. 1956. A flora do Parque Nacional de
Aiin cuando la migration de aves desde los Itatiaia. Bol. Parque Nac. Itatiaia, Brazil. Rio
de Janeiro. 5:1-92.
Andes a las Alturas Guayanesas pudiera haber
Bruggen, J. 1950. Fundamentos de la Geologia de
sido similarmente promovida durante los pe- Chile. Inst. Geogr. Militar, Santiago, Chile, 374 p.
riodos frios, un de la flora de esta
analisis Budel, J. 1951. Die Klimazonen des Eiszeitalters.
region demuestra que pocas, si
algunas, Eiszeitalter 1:16-26.
plantas colonizaron hacia el este durante el Burcl, H. 1961. Historia geologica de Colombia.
Rev. Acad. Colombiana Cien. Exactas Fis. Nat.
Pleistoceno.
11:137-191.
Varias teorias han sido elaboradas para Cabrera, A. L., Willink, A. 1973. Biogeografla de
America Latina. General Secretariat of the Or-
explicar los diversos regimenes de humedad/
ganization of the American States, Washington,
temperatura durante las diversas fases del D.C. Monografia 13. Ser. Biologia, 120 p.
Pleistoceno en diferentes partes del conti- Caldenius, C. C. 1932. Las glaciaciones Cuater-
186 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
der Westseite der peruanischen Anden. Bonn. Entwicklung Sudamerikas, pp. 1-24 in Fittkau,
Geogr. Abhandl. 29:1-320. E.J., Illies, J., Klinge, H., Schabe, G. H, Soli,
Kummebow, J., Matte, V., Schlegel. F. 1961. H. (eds.). Biogeography and Ecology in South
Zum problem der Nebelwalder and der zentral- America. Junk. The Hague. Vol. 1. 445 p.
chilenischen Kiiste. Ber. Dsch. Bot. Ges. 74:135- Saver, W. 1965. Geologia del Ecuador. Min. Edu-
145. cation, Quito, Ecuador, 384 p.
Labouriau, M. S. L., Schubert, C. 1977. Glacial Sauer, W. 1971. Geologie von Ecuador. Gebriider
and palynological studies in die Venezuelan Andes Borntraeger, Berlin, 316 p. [Vol. 11 of Beitrage
and late Quaternary instability. I\' Symp. Int. zur regionalen Geologie der Erde. H. I. Martini
Ecol. Tropical, Panama, 7-11 March, 1977. p. (ed.)].
103 [Abstract]. Schubert, C. 1974a. Late Pleistocene Merida gla-
Lexico Estratigraftco de Venezuela (ed. 2). ciation, Venezuelan Andes. Boreas 3:147-152.
1970. Bol. Geol.. Publ. Especial 4. Minist. Minas Schubert, C. 1974b. Late Pleistocene glaciation of
Hidrocarburos., Dir. Geol., 758 p. Paramo de la Culata. north-central Andes. Geol.
MacArthur, R. H., Wilson-, E. O. 1967. Theory of Rundsch. 63:516-538.
Island Biogeography. Princeton Univ. Press, Schwerdtfeger, W. (ed.). 1976. World Survey of
Princeton, N.J., 203 p. Climatology. Vol. 12. Climates of Central and
Macuire, B. 1970. On the flora of the Guayana South America. Elsevier, Amsterdam, 532 p.
Highland. Biotropica 2:85-100. Shagam, R. 1975. The northern termination of the
Mares, M. A. 1976. Convergent evolution of desert Andes, pp. 325—420 in Nairn, A. E. M.. Stehli,
rodents: Multivariate analysis and zoogeographic F. G. (eds.). The Ocean Basins and Margins,
implications. Paleobiology 2:39—63. Vol. 3. The Gulf of Mexico and the Caribbean.
Mavd, Phelps, W. H.. Jr. 1967. The origin of
E., Plenum Press, N.Y., 583 p.
the bird fauna of the south Venezuelan highlands. Sillitoe, R. H. 1974. Tectonic segmentation of the
Bull. Amer. Mus. Nat. Hist. 136:269-328. Andes: Implications for magnetism and metallog-
Mercer, J. H. 1976. Glacial history of southernmost eny. Nature 250:542-545.
South America. Quat. Res. (N.Y.) 6:125-166. Simpson, B. B. 1973. Contrasting modes of evolu-
Mercer, J. H., Palacios M., O. 1977. Radiocarbon tion in two groups nf Perezia (
Mutisieae; Com-
dating of the last glaciation in Peru. Geology positae) of southern South America. Taxon 22:
5:600-604. 525-536.
Miller. A. 1976. The climate of Chile, pp. 113- Simpson, B. B. 1975. Pleistocene changes in the
146 in Sciiwedtfecer, \V. (ed.). Climates of flora of the high tropical Andes. Paleobiology
Central and South America. Elsevier, N.Y., 532 p. 1:273-294.
[Vol. 12 of World Survey of Climatology, H. E. Simpson, B. B. In press. A revision of the genus
Landsberg, ed. in Chief]. Polijkpis (Rosaceae: Sanguisorbeae). Smithsoni-
Muller, P. 1968. Die Herpetofauna der Insel von an Contrib. Bot.
Sao Sebastiao (Brasilien). Saarbrucker Zeitung Simpson, B. B., Vervoorst, F. 1977. Physiographic
Verlag, 68 p. settings of the study sites, pp. 16-25 in Orians,
Nogami. M. 1972. The snowline and climate during G. II., Solbric. O. T. (eds.). Convergent Evo-
the last glacial period in the Andes mountains. lution in Warm Deserts. Dowden, Hutchinson
Quat. Res. (Tokyo) 11:71-81. and Ross, Stroudsburg, Pa., 333 p.
Okada, A. 1971. On the neotectonics of the Atacama Smith, L. B. 1962. Origins of the flora of southern
fault —
zone region. Preliminary notes on late Brazil. Contrib. U.S. Natl. Herb. 35:21.5-249.
Cenozoic faulting and geomorphic development Snelling, N. J. 1963. Age of the Roraima Forma-
of the Coast Range of northern Chile. Bull. Dep. tion. British Guiana. Nature 198:1079-1080.
Geogr. Univ. Tokyo. 3:47-65. Snow, J. W. 1976. The climate of northern South
Oliveira, A. I. de., Leonardos, O. H. 1943. Geol- America, pp. 295-^404 in Schwerdtfegfr, W.
ogia do Brazil. Min. Agricultura, Rio de Janeiro, (ed.). Climates of Central and South America.
813 p. Elsevier, Amsterdam, 532 p. [Vol. 12 of World
Paskoff, R. P. 1977. Quaternary of Chile: The Survey of Climatology. H. E. Landsberg, ed. in
state of research. Quat. Res. -(N.Y.) 8:2-31. Chief].
Paynter, R. A., Jr. 1972. Biology and evolution of Steinmann, G. 1930. Geologia del Peru. Carl Win-
the Atlapctes schistaceus species group (Aves: ters, Heidelberg, 448 p.
Emberizinae). Bull. Mus. Comp. Zool. Harvard Tosi, J. A., Jr. 1960. Zonas de vida natural en el
Univ. 143:297-320. Peru. Inst. Interam. Cienc. Agric. OEA (Organ.
Prohaska, F. 1976. The climate of Argentina, Para- Estados Unidos) Cent. Trop. Invest. Ensenanza
guay and Uruguay, pp. 13-112 in Schwerdtfe- Grad. Inf. Zona Andina. Bol. Tech. 5:1-271.
cer, W. (ed.). Climates of Central and South Trewartha, G. T. 1961. The Earth's Problem Cli-
America. Elsevier, N.Y., 532 p. [Vol. 12 of World mates. Univ. Wisconsin Press, Madison, 334 p.
Survey of Climatology, H. E. Landsberg, ed. in Troll, K. 1927. Yon Titikakasee zum Pooposee und
Chief]. zum Solar von Coipasa. Petennanns Mitt. 73:
Putzer, H. 1968. Uberbliek uber die geologische 218-222.
188 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Troll, K. 1959. Die tropischen Giberge. Bonn. Pliocene and lower Quaternary of the Colombian
Geogr. Abhandl. 25:1-93. Eastern Cordillera and the early evolution of its
Troll, K. 1965. Seasonal climates of the earth, pp. high-Andean biota. Rev. Palaeobot. Palynol. 16:
19-25 in Rodenwaldt, E., Jusatz, H. J. (eds.). 1-122.
1965 World Maps of Climatology. Springer- Ver- Vuilleumier, B. S. 1971. Pleistocene changes in the
lag, New
York, 2nd ed., 28 p. fauna and flora of South America. Science 173:
Turner, J. C. M. 1972. Cordillera Oriental, pp. 117- 771-780.
142 in Leanza, A. F. (ed.). Geologia Regional
Weberbauer, A. 1945. EI Mundo Vegetal de los
Argentina. Acad. Nac. Cien., Cordoba, 869 p.
Andes Peruanos. Minist. Agric, Lima, Peru,
van- der Hammen, T. 1974. The Pleistocene changes
of vegetation and climate in tropical South Amer- 766 p.
ica. J. Biogeogr. 1:3-26. Yrigoyen, M. R. 1972. Cordillera Principal, pp.
van der Hammen, T., Werner, J. H., Dommelen, 345-364 in Leanza, A. F. (ed.). Geologia Re-
H. van. 1973. Palynological record of the up- gional Argentina. Acad. Nac. Cien., Cordoba,
heaval of the northern Andes: a study of the 869 p.
8. The Amphibians of the Lowland Tropical Forests
John D. Lynch
School of Life Sciences
University of Nebraska
Lincoln, Nebraska 68508
USA
Prior to the separation of South America estimates of approximate altitudinal distribu-
from Gondwanaland and through much of tions for the many amphibians closely asso-
the Cretaceous, most of the continent sup- ciated with the major mountain systems in
ported a lowland tropical flora with a temper- South America (Andes, Guiana Shield, Bra-
ate flora to the extreme south (Savage, 1973). silian Shield,and coastal ranges). Published
During the Cenozoic, tropical lowland floras data are especially scant for the diverse fauna
were restricted because of the emergence of of southeastern Brasil.
the Andean ranges and the expansion of arid Several previous biogeographic essays
and semiarid regions in western South Amer- (e.g., Parker, 1935; Lutz, 1972; Lescure, 1975)
ica (Axelrod, 1960). The Andes separated on the Amazonian herpetofauna have dealt
lowland forests of extreme northwestern South with members of both forested and nonfor-
America from those of central South America, ested assemblies in preliminary fashions. The
and the developing arid zones separated those current effort suffers from some of the same
of central South America from those of the limitations imposed on previous efforts, name-
southeast. ly,incomplete distributional data (but prob-
As used here, "lowland tropical forests" re- ably not so incomplete as suggested by Heyer,
fers to those forested environments from sea 1976) and indefinite phylogenetic constructs.
level to 1000 meters elevation. Amphibians The South American rainforests exist as
distributed within the savannas occurring in four or five disjunct elements (Fig. 8:1). The
forests are not considered forest elements. most southern, the Austral Forests, is tem-
—
Thus, the tree frogs Hyla crepitans, H. par- perate and very distinct faunistically, sharing
tialis, and H. raniceps
—
are ignored here, al- only one genus (Bafo) and no species with
though each occasionally invades one or more the other forests. The northern tropical for-
of the major South American rainforests. On ests are tenuously connected, chiefly by gal-
the other hand, the marine toad, Bnfo mari- lery forests and forest islands distributed
nus, is relatively common in nonforested en- through the nonforest areas separating them.
vironments in Central America and northern The Trans-Andean Forests are isolated from
South America and also is distributed through the vast Central Cis-Andean Forests by the
the northern forests penetrating even those northern Andes but are connected weakly via
aseasonal forests with very high annual pre- the Northern Forests of Colombia and Vene-
cipitation (Choco and Napo regions). zuela; the latter are not entirely discrete (ten-
Although I have not always segregated the uous connections through the forests fringing
evergreen forests from those that are season- the Merida Andes and Venezuelan llanos).
ally dry, I have pointed out the subdivision The largest of the forests ( approximately 80$
within each major forest. (The subdivisions of all tropical South American forests) is
usually are correlated with seasonal or asea- the Central Cis-Andean forest ( Amazonian or
sonal forests.)Those forests that have pro- Hijlaea) drained by the Rio Amazonas and
nounced and prolonged dry seasons are not its tributaries. The central forests become
included here. As a general rule, such forests drier southeastwardly and grade into the cer-
are inhabited by nearly the same suite of rado on the northwestern face of the Bra-
species found in adjacent nonforest environ- A less marked dry belt ( Reinke's
silian Shield.
ments (cerrado, llanos). The greatest diffi- Corridor) described by numerous, isolated
is
culty (and least confidence) was to extract and generally small savannas extending from
IV)
190 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
NORTHERN
1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 191
Caecilians
Caccilia subnigricans + +
Caccilia tcntaculata + +
Siphonops annulatus — + +
Anurans
Adenomera hylacdactyla + + +
Bufo marinus _ + + +
Bufo typhonius — + + + +
Centrolcnclla ftcischmanni _. + + +
Hyla albomarginata + + +
Hyla boans + +
Hyla egleri _ _ + +
Hyla gcographica
+ +
Hyla Icucophyllata ....
+ +
Hyla luteocellata + +
Hyla minuta + +
Hyla rubra + + +
Hyla senicula _.. + +
Leptodactylus bolivianus .... + +
Lcptodactylus pcutadactylus + + +
Leptodactylus poccilochilus + +
Lcptodactylus wagneri + + +
Phyllobatcs pictus +
Phyllomedusa trinitatus .. + +
Phrynohyas venulosa + + + +
Physalaemus pustulosus + +
Rana palmipes + + +
Relictivomer pearsei + +
these suggestions beyond these examples (1955), Cochran and Goin (1970), and Ri-
in part because participated in
I have vero (1961), treating southeastern Brasil, Co-
the recognition that "Eleutherodactylus con- lombia, and Venezuela, respectively. Lescure's
spicillatus (or Hijlodes gollmeri)" consisted (
1976 ) account of the frogs of French Guiana
of many species of frogs distributed through ismuch more useful. Lutz (1973) and the
the forested habitats of lowland and montane numerous and scattered works of Werner
South America (Lynch, 1975; Lynch and Bokermann are especially useful for Brasil.
Hoogmoed, 1977; 1961; Lynch and
Rivero, Taylor (1968) remains the most valuable
Myers, in prep.). I predict that within a dec- source for caecilian data, and Wake and
ade, we will find that there exist very few Lynch 1976, and references therein ) pro-
(
"widely distributed" species of forest amphib- vided a recent summary for salamander dis-
ians; concomitantly, distinctions among differ- tributions. The largest group of amphibians
ent forest faunas will become greater. (frogs and toads) is summarized the most
poorly. Duellman's (1977) account partially
fills a void for centrolenid, hylid, and
THE SOUTH AMERICAN AMPHIBIAN pseudid
data, but does not provide easy reference to
FAMILIES
primary distributional data.
Thirteen amphibian families occur in
South American lowland forests (Table 8:2)
PLETHODONTIDAE
but only one (Brachycephalidae) is endemic. (Lungless Salamanders)
The thirteen families differ in their relative The family is primarily North American
apparencies in tropical lowland forests and with one tribe radiating in the Neotropics
in their global distributions. (Wake, 1966). Bolitoglossa is well repre-
Primary distributional data are available sented in montane environments in Colombia
in the taxonomically dated works of Cochran and Venezuela (Brame and Wake, 1963,
192 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Table 8:2.- -Amphibian Complements of the Tropical Lowland Forests of South America.
( Number of endemic species given in parenthesis. )
Typhlonectidae _.
Brachycephalidae
Bufonidae
Centrolenidae
Dendrobatidae _..
Hylidae
Leptodactylidae
Microhylidae .
Pipidae
Ranidae
1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 193
Vhyllobates usually occur at elevations below in South America and lower Central America
1000 m(Savage, 1968; Silverstone, 1975, (Carvalho, 1954). The South American mi-
1976). crohylids are primarily lowland forest animals
but Glossostoma aequatorialc is Andean and
HYLIDAE Dermatonotus and Elachistocleis are non-
(Treetoads or Treefrogs) forest genera (Dunn, 1949).
Amphignathodontines, hylines, and phyllo-
medusines occur in a wide range of environ- PIPIDAE
ments along both altitudinal and moisture di- (Tongueless Frogs)
mensions, although only amphignathodontines The pipinae is restricted to South America
could be viewed as successfully venturing and adjacent Panama (Savage, 1973). Pipa
onto the high Andes (Duellman, 1977). parva is a nonforest
species but the other five
Hemiphractines are lowland forest animals re- pipines are lowland forest animals (Dunn,
stricted to northwestern South America and 1948, Izecksohn, 1976b).
Panama (Trueb, 1974). Hylines are widely
distributed outside of the Neotropics (Aus-
PSEUDIDAE
(Pseudid Frogs)
tralo-Papuan, Holarctic, Oriental regions);
the other three subfamilies are Neotropical. The aquatic pseudids fringe the rain-
Some primary distributional data are avail- forests of tropical South America (Cochran
able in Duellman ( 1971a, 1971b, 1972, 1973, and Goin, 1970; Gallardo, 1961; Savage,
1974), Duellman and Crump (1974), and 1973) but do not invade the wet forests. Be-
Trueb and Duellman (1971). cause their distribution is generally nonforest,
I have not included them elsewhere in this
LEPTODACTYLIDAE account.
(Leptodactylid Toads)
The family has invaded
RANIDAE
nearly every physi-
(True Frogs)
ographic division of South America except
the extremely xeric coast of Chile and Peru This Holarctic, Ethiopian, and Oriental
(Lynch, 1971). Bufonids have only slightly family barely penetrates Neotropical regions
narrower distributions (the two families are (Savage, 1973). A single species (Rana
the most wide-ranging amphibian families in palmipes) is distributed widely over the
South America). Ceratophryines and lepto- northern half of lowland South America.
dactylines are primarily lowland frogs ranging
through xeric and mesic environments. Elo- The two dominant families of amphibians
siines and telmatobiines primarily are found in South America (Hylidae and Leptodacty-
in mesic (low and high elevations) environ- lidae) also are dominant in the lowland tropi-
ments (Odontophrynus is a nonforest low- cal forests (Table 8:3). In each of the four
land genus). Primary distributional data for forests, these two families have the greatest
leptodactylid frogs are scattered widely. numbers of species, accounting for 48 percent
Lynch (1971,and references therein) pro- (Trans-Andean), 54 percent (Northern), 61
vided summaries of distributions for all gen- percent (Central Cis-Andean), and 84 per-
era. Other primary distributional data are cent ( Atlantic ) of the areal faunas. These fam-
available in Heyer (1970, 1973). Data for ilies are the only South American
amphibian
eleutherodactyline frogs represent summaries families having more than one family-group
of unpublished data accumulated from mu- (subfamilies, tribes) represented in the Neo-
seum study by the author. tropics. The two families account for 61 per-
cent of the lowland tropical forest fauna of
MICROHYLIDAE South America (327 of 530 species).
(Narrow-mouthed Frogs) The family Hylidae currently is divided
The family pantropical but invades
is into four subfamilies (Duellman, 1970, 1977),
north temperate regions (Parker, 1934). All three of which are Neotropical and essen-
South American genera are endemic or occur South American
tially ( Amphignathodonti-
194 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Table 8:3. — Hylid and Leptodactylid Complements of the Tropical Lowland Forests of South America.
(Numbers in parentheses are endemics.)
Family-group T-Andean Northern Central Atlantic
Hylidae
Amphignathodontinae 3(3) 2(2) 6(6) 7(7)
Hemiphractinae KD 4(4)
Hylinae 18(15) 7(2) 66(56) 53(45)
Phyllomedusinae 6(6) 2(1) 8(7) 9(9)
Leptodactylidae
Batrachylini 3(3)
Ceratophryinae KD KD KD
Eleutherodactylini 24(24) 4(4) 33(33) 15(15)
Elosiinae 16(16)
Grypiscini 13(13)
Leptodactylinae 8(3) 5(0) 20(16) 14(11)
Odontophrynini 5(5)
nae, Hemiphractinae, and Phyllomedusinae); ics at the family level are the Brachycephali-
the fourth subfamily, Hylinae, occurs in the dae, Elosiinae, and Grypiscini, because they
Australo-Papuan, Holarctic, Ethiopian (north- are not found outside of the Atlantic Forests.
ern), and Oriental regions as well. Hemi- Although the four forests may be separated
phractines occur only in Trans-Andean and and linked in various combinations on the
Central Cis- Andean Forests (Trueb, 1974). basis of family-group occurrences, generic
The family Leptodactylidae was consid- and specific distributions provide more pow-
ered to occur in the Nearctic, Neotropical, erful tools to describe and differentiate am-
South America (Savage, 1973). The sub- both forest and nonforest habitats.
family Telmatobiinae is divided into six Seventeen genera (3 caecilians and 14
tribes ( Batrachylini, Calyptocephalellini, anurans) occur only in Central Cis-Andean
Eleutherodactylini, Grypiscini, Odontophryn- Forests. All three caecilian genera (Brasilo-
ini, andTelmatobiini) by Lynch (1971, 1978). typhlus, Potomotyphlus, and Rhinatrema)
Two (Calyptocephalellini and Telma-
tribes apparently are restricted to lowland forest
tobiini) do not enter lowland tropical forests. habitats. Of the 14 anurans, the hylids —Allo-
The Batrachylini occur in the temperate rain- phnjne and Nyctimantis, the leptodactylids —
forests of Chile and Argentina and the trop- Dischidodactylus, Hydrolactare, Lithodytes,
ical forests of southeastern Brasil (Atlantic and Vanzolinus, and the microhylids Cteno- —
Forests). Grypiscines and odontophrynines phryne, Hamptophryne, Otophryne, Syncope,
occur in Atlantic Forests but not in the other and Synapturanus, are endemic forest genera.
tropical forests, whereas eleutherodactylines Pipa also occurs as a nonforest element in
occur in all four tropical forests. northwestern South America and Panama;
Thus, at the family-group level, the At- Edalorhina (Leptodactylidae) and Stefania
lantic Forests exhibit
pronounced endemicity (Hylidae) have upland forest species.
[Brachycephalidae, Batrachylini (shared with Twenty-one genera occur only in Atlantic
Grypiscini, and
Austral Forests), Elosiinae, forests. The Mimosiphonops, and
caecilian
Odontophrynini (but shared with nonforest the anurans Arcovomer, Brachyccplialus,
environments)]. The most significant endem- Crossodactylodes, Crossodactylas, Cycloram-
1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 195
phus, Dasypops, Fritziana, Hcmipipa, Hyo- the Magdalena forests, although a poison-dart
phryne, Macrogenioglottus, Mcgaclosia, Mij- frog, Dendrobates trunctatus, is restricted to
ersieUa, PsyUophryne, Scyihrophrys, Stereo- the pair of forests (Silverstone, 1975) and
cyclops, and Thoropa are lowland forest only weakly differentiated from the Chocoan
endemics. Hylodes and Zachaenus each have D. auratus. Hyla phlehodes (Hylidae), Oedi-
one upland species, Proceratophrys has at pina complex (Plethodontidae), and Smilisca
least one cerrado species, and Chthonerpeton sila (Hylidae) occur in the Nechi forests but
(if actually represented in forested environ- not in other Trans-Andean Forests in Colom-
ments; see Taylor, 1968) is primarily a non- bia but are not endemic to the Nechi forests
forest genus of the Rio Parana drainage. (Brame, 1968; Duellman, 1977). Eighty-eight
(70.4?) of the 126 Trans- Andean Forest am-
phibians are distributed in the Chocoan
THE TRANS-ANDEAN FORESTS Forests of South America and Central Amer-
ica but not in the Nechi or Magdalenan for-
The Trans-Andean occupy the
Forests ests.Fifty-six species (44.8% of the fauna)
lowlands west of the Andes in Colombia and are endemic to the South American Choco,
Ecuador (Choco lowlands) and extend into and only eight of these range the length
Central America. Also, I have included the of Choco. The northern half of the
the
seasonally dry forests in the area of the Rio Choco and southern half (divided at approx-
Sinii and lower rios Cauca and Magdalena,
imately Buenaventura) contain 24 endemic
and the evergreen forests of the Rio Magda-
species each (Fig. 8:3). The most diverse
lena. Haffer (1969) and Muller (1973, 1974) Trans-Andean Forest fauna is that found in
divided these Trans-Andean Forests into Cho- the Rio San Juan Drainage where the north-
coan and Nechi (Rio Sinu and Rio Cauca) ern and southern Chocoan faunas overlap.
units, and Muller (1973) separated the rain- Some of the relatively widespread species
forests of the Rio Magdalena as a distinct (Table 8:1) also occur in the Rio San Juan
distributional center. Those divisions were
Drainage, thereby heightening the diversity
accomplished largely on the basis of bird there, but these do not occur appreciably
distributions and are not supported by am- south of the drainage.
phibian distributions. The relative poverty of the Nechi and
One hundred and twenty-six amphibians Magdalena forests, as well as those in south-
occur in the Trans-Andean Forests (Appen- western Ecuador is marked. The most obvi-
dix 8:1), including most of the salamanders ous factor that potentially reduces diversity
known from lowland forests in South Amer- in these areas the seasonality of rainfall,
ica. Endemism is pronounced —
88 percent of
is
Fie. 8:3. Endemism among Trans-Andean Forest amphibians. A. The distributions of 88 of the 126
Trans-Andean species are encompassed by the hatched area. B. 56 species are restricted to the South Amer-
ican Choco. C. Numbers of endemics in the northern and southern South American Choco (hatched ) Nechi
,
phractus jcisciatus occurs at lower elevations by Miiller (1973) (Santa Marta, Catatumbo,
on the Andean slopes and appears to range and Venezuelan coastal forests centers), who
throughout the forests where lower montane considered the centers not well differentiated.
habitats exist. Bufo haematiticus, Caecilia He separated them on the basis of vertebrates
thompsoni, Eleiitherodactylus fitzingeri, E. other than amphibians.
longirostris, and Leptodactyhis pentadactylus This is the smallest and least diverse of
link the Choco and the wet Magdalena For- the lowland forest units and has the lowest
ests hut do not occur in the drier Nechi amount of endemism. Only 39 species of
Forests. amphibians are found in the Northern Forests
(Appendix 8:2). Endemism is low (56.4?)
THE NORTHERN FORESTS compared to the other three forests, but com-
parable to that of the strictly Chocoan ele-
The Northern Forests include those at the ments of the Trans-Andean Forests (44.8$).
base of the Sierra Nevada de Santa Marta, Some of the endemism (Eleiitherodactylus
the forest around the Maracaibo basin, those bicumulus, E. rozei, E. terraebolivaris, Cen-
in the Falcon District of Venezuela, those trolenella sp., Phyllomedusa medinae, Gastro-
associated with the coastal range of Vene- theca williamsoni, Flcctonotus pygmaeus)
zuela, and those along the northwestern edge may nothing more than montane slope
reflect
of the Venezuelan llanos. The Northern species invading the lowland forests. If that
is the case for these 9 species, the apparent
Forests are bisected by the Sierra de Perija
and the dry Peninsula de Guaira. The assem- endemism of the northern forests is mark-
blage was treated as three separate centers edly reduced to the point of being inconse-
1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 197
the two well-differentiated northern faunas linked by the following amphibians (asterisks
indicate number of other forests where spe-
(Trans-Andean and Central Cis-Andean).
cies occurs )
:
^v. /» C ^
GUYANA
PARA
NAPO
UCAYALI
MADEIRA-TAPAJOS
Fig. 8:6. Concentration of amphibian endemics within Central Cis-Andean Forests. Guyana, Madeira-
Tapajos, Napo-Ucayali, and Para subunits are labeled. The coarse-stippled zone (3 endemics) is supra-Amazon.
The remaining subunits include the Bolivia (4 endemics) and Venezuelan Guyana (14 endemics) subunits.
Concentraciones de anfibios endemicos dentro do las selvas cisandinas centrales. Guayana, Madeira-Tapajos,
Napo-Ucayali, y Para est an indicadas. El achurado grueso indica la zona supra-Amazonica (3 endemismos).
Las rest antes subunidadcs forestalcs incluyen a Bolivia (4 endemismos) y a la Guyana oenezolana (14 ende-
mismos) subunidades.
lantic forests are obviously unique in having tylus mystacinus ranges through the forests as
batrachyline, elosiine, grypiscine, and odonto- well as widely into nonforested environments
phrynine leptodactylid frogs. Eleven families (e.g., the Chaco; W. R. Heyer, pers. comm.).
of amphibians and 183 species are known Hyla faber is the Atlantic forest replacement
from the Atlantic forests. Of the 183 species, for Hyla Loans ( Lutz, 1973 ) .
168 (91.8%) are endemic to those forests. The frogs Gastrotheca fissipcs, Hyla leu-
Relatively few amphibians link all of the cophyllata, H. rubra, Leptodactylus mystace-
Atlantic Forests (Bufo crucifer, Hyla faber, us, L. troglodytes, L. Phyllodytes
icagneri,
H. geographica, Leptodactylus mystacinus, L. luteolus, and Proceratophrys arecristiceps
pentadactylus, and Siphonops annulatus ap- broadly distributed through the northern por-
pear to be distributed throughout the forests ) . tion of the Atlantic forests, and Hyla fusco-
These widespread species include two (Hyla marginata, H. senicula, and Siphonops paul-
geographica and Siphonops annulatus) also ensis are widely distributed through the south-
distributed over the whole of the Central Cis- ern portions of the Atlantic Forests. West and
Andean Forests and one (Leptodactylus pen- south of Cabo Frio, one encounters a series
dactylus) widely distributed in Central Cis- of frogs distributed in the coastal band of
Andean and Trans-Andean forests. Leptodac- the southern Atlantic Forests ( Crossodactylus
200 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fig. 8:7. Napo-Ucayali distributions. A. Eleutherodactylus acuminatus. B. E. ockendeni. Both are eleu-
therodactyline leptodactylids.
Distribueiones en el Napo-Ucayali. A. Eleutherodactylus acuminatus. B. E. ockendeni. Amhos son lep-
todactylidos cleutherodactylinos.
dispar, Cycloramphus asper, Flectonotus fis- dactyla, Bufo typhonius, Hyla alboirwrgi-
silis, Gastrotheca viridis, Hyla catharinae, H. nata, H. H. rubra, Leptodactylus wag-
egleri,
hayii, Leptodactylus sp. N., Phrynohyas imi- neri, Phrynohyas venulosa, and Rana pabn-
tatrix, Phyllomedusa fimbriate!, Procerato- ipes) and from the west (Hyla geographica,
phrys appendiculata, and P. boiei). H. leucophyllata, H. senicula, Leptodactylus
Almost a dozen species of frogs link the pentadactylus, Phyllobates pictus, and Phryn-
northern and southern (chiefly coastal) por- ohyas venulosa).
tions of the Atlantic Forests. These are Bufo Some of the distinctiveness of the Atlantic
ictericus, B.
typhonius, Eleutherodactylus Forests is reduced if one views dendrobatids
binotatus, Hyla albomarginata, Hyla microps, (abundant in Central Cis- Andean Forests)
Macrogenioglottus alipioi, Osteocephalus and elosiine leptodactylids (prominent in At-
langsdorffii,Phrynohyas mesophaea, P. venu- lantic Forests) as vicariants. Such a view is
losa (from the west in the southern Atlantic consistent with current notions of the rela-
Forests and from the north in the northern tionships of these two groups of frogs ( Lynch,
Atlantic Forests), Thoropa miliaria, and Tra- 1971). Adding these to the other sets of vi-
chycephalus nigromaculatus. cariants of these two major forests (Table
The nonendemic species can be divided 8:4) tends to lessen the impression of nearly
into —
two groups those invading the Atlantic- complete distinction of the two forests.
Forests from the north (Adenomera hylae- Muller (1973) divided the Atlantic forests
1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 201
Fig. 8:8. Guianan and Napo-Ucayali distributions. A. Circles are Syncope antenori, squares represent both
Syncope carvalhoi and Euparkerella myrmecoides ( Microhylidae and eleutherodactyline leptodactylids, respec-
tively), and triangles, Eleutherodactylus chiastonotus. B. Squares, Eleutherodactylus nigrovittatus; circles, E.
zeuctotylus.
Distribuciones en Guyana y Napo-Ucayali. A. Los circulos representan a Syncope antenori, los cuadrados
representan a Syncope carvalhoi y Euparkerella myrmecoides (Microhylidae y Icptodactylido eleutherodactylino ,
into three subcenters (one north of Ilheus, one Bahia subcenters requires combining the Ba-
between Ilheus and Cabo Frio, and the third hia and Espirito Santo units recognized here
encompassing what I term southern the Per- — as the Bahia Unit. By so doing, the new
nambuco, Bahia, and Paulista subcenters, re- Bahia Unit is second in endemicity only to
spectively). I divided the Atlantic Forests the Sao Paulo Unit ( 11 Bahia + 13 Espirito
into seven units and compiled species lists for Santo + 2 Bahia-Espirito Santo endemics =
each. Of the 183 species used, approximately 26 endemics). Endemicity is lowest in those
one-third are endemic to the coastal forests three subunits invaded by nonendemic taxa
between Cabo Frio and Parana (Fig. 8:9). (Pernambuco, Minas Gerias, and Parana
The other six areas exhibit low endemism, units). Some of the endemics scored for the
and all five coastal units are richer than the Sao Paulo Subunit may not occur below 1,000
two inland areas ( in endemics, as well as all m (
Ccntrolenella albotunica, C. divaricans,
amphibians). Recognizing Pernambuco and C. (labia, C. eurygnatha, and C. vanzolinii);
Fig. 8:10. Parapatric distribution patterns. A. Bufo guttatus group (3 species): Chocoan B. blombergi,
Napoan B. glaberrimus, and Guianan B. guttatus. B. Leptodactylus fuscus group (5 species): Amazonian L.
sp. "A," Guianan L. longirostris, Pernambucan L. mystaceus, Paulistan L. sp. "N," and Chocoan L. vcntrimac-
ulatus. C. Eleutherodactylus fitzingeri group (7 species): Chocoan E. achatinus, Guianan E. chiastonotus, Na-
poan E. conspicillatus (open hatching), Madeiran E. fenestrates (fine hatching), Ucayalian E. peruvianus
(coarse stipple), Venezuelan E. terraebolivaris and supra-Amazonian E. vilarsi (fine stipple). D. Hatching
(Eleutherodactylus sulcatus group, 2 species): Napo- Ucayalian E. sulcatus; Venezuelan E. maussi. Stipple
(Dendrobates tinctorius group, 4 species): Chocoan D. auratus, Nechian D. truncatus (coarse stipple), Guia-
nan D. tinctorius, and Paran D. galactonotus. Sources: A.) Cei (1972); B.) W. R. Heyer (pers. comm.);
C.) Lynch (in prep.); D.) Lynch (1975), Silverstone (1975).
Patrones dc distribucion parapdtricas. A. Grupo de Bufo guttatus (3 especies): B. blombergi del Choco, B.
glaberrimus del Napo, B. guttatus guy ones. B. Grupo de Leptodactylus fuscus (5 especies): L. sp. "A" del
Amazonas, L. longirostris de Guyana, L. mystaceus dc Pemamhuco, L. sp. "N" Paulistano, y L. ventrimacu-
latus del Choco. C. Grupo Eleutherodactylus fitzingeri (7 especies): E. achatinus del Choco, E. chiastonotus de
Guyana, E. conspicillatus dc Napo (lincado amplio), E. fenestratus de Madeira (lineado fino), E. peruvianus del
Ucayali (achurado grueso), E. terraebolivaris de Venezuela y E. vilarsi del Alto Amazonas (achurado fino).
D. Lineado (Grupo de Eleutherodactylus sulcatus, 2 especies): Napo-Ucayali E. sulcatus; Venezuela, E.
maussi. Achurado (Grupo de Dendrobates tinctorius, 4 especies): Choco, D. auratus; Ncchi, D. truncatus
(achurado grueso): Guyana, D. tinctorius; y Para, D. galactonotus. Fuentes: A.) Cci (1972); B.) W. R. Heyer
(pers. comm.); C.) Lynch (en prep.); D) Lynch (1975), Silverstone (1975).
204 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Table 8:5. — Distribution of Reproductive Modes Among Tropical Lowland Forest Amphibians.
Reproductive Mode
8 9 10
Plethodontidae X
B rachy cephalidae
Bufonidae X
Centrolenidae _ X
Dendrobatidae __ X
Amphignathodontinae X
Hemiphractinae X
Hylinae X
Phyllomedusinae
Batrachylini
Ceratophryinae
Eleutherodactylini
Elosiinae X 1
Grypiscini
Leptodactylinae
Odontophynini _ X
Microhylidae __ X
Pipidae _.
Ranidae
1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 205
/FOREST MODES
Fig. 8:11. Reproductive modes among rainforest anurans. Mode 9 (Pipa with direct development) and
Mode 10 amphignathodontine and hemiphractine hylids, direct development) not shown. Modes 5 and 7
(
involve foam nests; eggs are laid on vegetation in Mode 4 (after Crump, 1974).
Modos reproductivos de los anuros de la pluviselva. A los Modo 9 (Pipa con desarrollo directo) y 10 (hyli-
dos amphignathodontinos y hemiphractinos, desarrollo directo) no se los muestra. Modos 5 y 7 tienen nidos de
espuma; los huevos son puestos en la vegetation en el Modo 4 (de Crump, 1974).
ingly greater risks (and concomitantly, apply more to the adult than the embryos.
greater dependency) are taken by those am- Species with Mode 6 have terrestrial eggs
phibians depositing their eggs out of water which, upon hatching, yield tadpoles that
(Modes 4-8, 10). The larvae of amphibians pass some time on the parent's back but
having Modes 5 and 7 presumably secure eventually are transported to bodies of water
some protection from the foam. Those spe- for "normal'' development. Terrestrial eggs
cies employing Mode 4 (eggs on vegetation and embryos resulting in miniature replicas
over water, larvae in water) have a more of the adult produced without submersion in
obvious association with wet environments water characterize Modes 7 and 8.
(and therefore, more frequently forests) but If the asserted sequence is correct, we
also occur in nonforested habitats (llanos, should expect to discover different patterns of
scrub )
. distribution (
and patterns of endemism if
13/8
1/1
16/15)
Fig. 8:12. Species densities of am phibians having various reproductive modes in South
and endemicities
American rainforest units. The first number
the rami ber of species of the group in the area; the number fol-
is
lowing the solidus is the number of endemic species of the group in the area. A. Mode 1, Hylinae. B. Mode
5, Leptodactylidae (except Adenomera). C. Mode 4, Centrolenidae, Hylinae, Phyllomedusinae. D. Mode 6,
Dendrobatidae ( the numbers in parentheses in southeast ern Brasil refer to Elosiinae). E. Mode 8, Eleuthero-
dactylini ( the numbers in southeastern Brasil refer to Grypiscini).
Densidad dc especies y endemismos de varios atrftb ios con diversos modos reproductivos en las selvas lluvi-
osas de Sud America. El primer numero es el numero i/c especies del grupo en el area; el numero siguiente a
la linca diagonal representa el numero de especies ende micas del grupo en el area. A. Modo 1, Hylinae. B.
Modo 5, Leptodactylidae (exccpto Adenomera). C. M
odo 4, Centrolenidae, Hylinae, Phyllomedusinae. D.
Modo 6, Dendrobatidae (los numeros cntre parentesis en el sudestc del Brasil refieren a Elosiinae). E. Modo
8, Eleuthcrodactylini (los numeros en parentesis en el sudestc del Brasil refieren a Grypiscini).
Table — Frequencies
8:6. of Frogs and Salamanders
of Tropical Lowland Forests Exhibiting 10 Repro-
ductive Modes Compared to Frequencies for Non-
endemic Species.
208 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
lids and centrolenid frogs (and possibly cae- Middle America prior to the Neogene uplift
ciliidand hyline stocks ) do not fall into either of thePanamanian Isthmus, but most of the
pattern. Amphignathodontines and centrolen- radiation of the group occurred in the West
ids are recent invaders of Central America Indies and in northern South America.
(Savage, 1973) but are usually found in up- The development of biogeographic con-
lifted areas (discontinuous distributions). structs (historical without
biogeography)
Phyllomcdusines are much better represented some defensible phylogenetic constructsis not
in Central America (Agalychnis and Pachy- science (see Rosen, 1975, for philosophy and
medusa) and are essentially replaced in methodology). Unfortunately, the sorts of
South America by Phyllomedusa; the replace-
phylogenetic input required to generate ro-
ment suggestive of geographic separation
is bust biogeographic constructs are not avail-
of the two forest stocks during much of the able for most lowland Amphibia. In spite of
Tertiary. the impressive strides made in the past decade
Some
of these amphibian groups were iso- in tabulating systematic data for some of the
lated and radiating in Central America during larger amphibian groups in South America,
much of the Tertiary; Savage (1973) cited our knowledge of the relationships within
bufonids, hylids (hylines and phyllomedu- those groups remain regretfully nascent. This
sines), microhylids, and eleutherodactyline is less a critique of
previous work than a
leptodactylids. Savage and Wake (1972) statement of what remains to be done for
identified the Middle American caeciliids iso- most groups of tropical amphibians and rep-
lated in Central America during the same tiles. I do not believe that we have the
time period. Most of these isolates are recog- systematic data necessary to generate formu-
nizable in that the species groups or genera lations of how the tropical lowland forest
are essentially endemic to Central America amphibian groups developed in time and
and extend only into the Choco in South space.
America. Savage ( 1973 ) suggested that Eleu-
therodactijlus and some derived genera under-
went an Eocene-Miocene radiation in Middle ACKNOWLEDGMENTS
America and subsequent to the uplift of the
Panamanian Portal radiated into South Amer- Most of the distributional data presented
here represents summaries of unpublished
ica. My
study of Eleutherodactylus does not
confirm such a conviction. Ancestral relicts data collected by colleagues. My thanks go
to William E. Duellman, M.
(Hylactophryne and Ischnocnema) persist at J. Fouquette, Jr.,
Eleutherodactylus (Mexico to Bolivia and for serving as sources of stimulation and criti-
eastern Brasil), as well as South American cism; special thanks are extended to William
groups in lower Central America (hiporcatus E. Duellman, Linda Trueb, and Charles F.
and unistrigatus groups; Lynch, 1976). Some Walker. I defer to identify the many others
eleutherodactyline frogs certainly invaded to avoid failing to recall some.
1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 209
Patagonia. Los Andes del norte han contri- America. Univ. Kansas Mus. Nat. Hist. Monogr.
buido a una fragmentation de ciertos grupos (l):l-753.
Duellman, W. E. 1971a. A taxonomic review of
(ranas dendrobatidads y eleutherodactylinas, South American hylid frogs, genus Phrynohyas.
y cecilios rhinatrematidos )
.
Univ. Kansas Mus. Nat. Hist. Occas. Pap. (4):
Los cecilios rhinatrematidos
y typhlonec- 1-21.
tidos y los anuros
amphignathodontinos, Duellman, W. E. 1971b. The identities of some
Ecuadorian hylid frogs. Herpetologica 27:212-
brachycephalidos, centralenidos, dendroba- 227.
tidos, la mayoria de los eleutherodactylinos, Duellman, W. E. 1972. South American frogs of
elosiinos, grypiscinos, hemiphractinos, y pipi- the Hyla rostrata group Amphibia, Anura, Hy-
(
dos se originaron y diferenciaron al parecer lidae). Zool Meded. Rijks Mus. Nat. Hist. Leiden
47:177-192.
en las selvas tropicales de tierras bajas de
Duellman, W. E. 1973. Frogs of the Hyla geo-
Sud America. Ciertos otros grnpos (cecilios
graphica group. Copeia 1973 (3):515-533.
caeciliidos, algnnos bufonidos, hylinos, micro- Duellman, W. E. 1974. A reassessment of the taxo-
hylidos, y phyllomedusinos, y posiblemente nomic status of some Neotropical hylid frogs.
Univ. Kansas Mus. Nat. Hist. Occas. Pap. (27):
algunos eleutherodactylinos) se originaron en
1-27.
estas selvas pero irradiation en Centro Amer-
Duellman, W. E. 1977. Liste der Rezenten Am-
ica tambien. Las salamandras plethodontidos
phibien und Reptilien. Hylidae, Centrolenidae,
y los anuros ranidos son invasores recientes Pseudidae. Das Tierreich 95:1-225.
desde las selvas de Centro America, y los Duellman, W. E., Crump, M. L. 1974. Speciation
in frogs of the Hyla parviceps group in the upper
anuros batrachylinos, la mayoria de los bufo-
Amazon Basin. Univ. Kansas Mus. Nat. Hist.
nidos, ceratophryinos, leptodactylinos, y Occas. Pap. (23): 1-40.
odontophryinos son invasores de ambientes Dunn, E. R. 1948. American frogs of the family
surefios temperados y subtropicales. Pipidae. Amer. Mus. Novit. (1384): 1-13.
Dunn, E. R. 1949. Notes on South American frogs
of the family Microhylidae. Ibid. (1419):1-21.
Estes, R„ Reig, O. A. 1973. The early fossil record
LITERATURE CITED of frogs: A review of the evidence, pp. 11-63
in Vial, J. L. (ed.). Evolutionary biology of the
Bufo. Univ. Texas Press, Austin, 459 p. Heyer, W. R. 1973. Systematics of the marmoratus
Cochran, D. M. 1955. Frogs of southeastern Brazil. group of the frog genus Leptodactylus (Amphibia,
Bull. U.S. Natl. Mus. (206): 1-423. Leptodactylidae). Ibid. (251): 1-50.
Cochran, D. M., Goin, C. J. 1970. Frogs of Co- Heyer, W. R. 1975. A preliminary analysis of the
lombia. Ibid. (288): 1-655. intergeneric relationships of the frog family Lep-
Crump, M. L. 1974. Reproductive strategies in a todactvlidae. Smithsonian Contrib. Zool. (199):
'
Heyer, W. R., Liem, D. S. 1976. Analysis of the Lynch, J. D., Myers, C. W. [in preparation]. The
intergeneric relationships of the Australian frog frogs of the fitzingeri group of Eleutherodactylus
family Mvobatrachidae. Smithsonian Contrib. in the Chocoan lowlands.
Zool. (233): 1-29. McDiarmid, R. W. 1971. Comparative morphology
Hueck, K. 1966. Die Walder Siidamerikas. Fischer and evolution of frogs of the Neotropical genera
Verlag, Stuttgart, 422 p. Atelopus, Dendrophryniscus, Melanophryniscus,
Izecksohx, E. 1971. Novo genero e nova especie de and Oreophrvnella. Nat. Hist. Mus. Los Angeles
Brachycephalidae do Estado do Rio de Janeiro, Cty. Sci. Bull. (12): 1-66.
Brasil. Bol. Mus. Nac. Rio de Janeiro Zool. (280): Moore, J. A. 1944. Geographic variation in Rana
1-12. pipiens Schreber of eastern North America. Bull.
Izecksohx, E. 1976a. O status sistematieo de Phryn- Amer. Mus. Nat. Hist. 82:345-370.
iscus proboscideus Boulenger ( Amphibia, Anura, Moore, J. A. 1975. Rana pipiens — the changing par-
Bufonidae). Rev. Bras. Biol. 36:341-345. adigm. Amer. Zool. 15:837-849.
Izecksohx, E. 1976b. Una nova especie de Pipa, do Muller, P. 1973.
Dispersal centres of terrestrial
estado do Amazonas, Brasil (Amphibia, Anura, vertebrates the Neotropical Realm.
in Biogeo-
Pipidae). Ibid. 36:507-510. graphica, 2, Junk, The Hague, 244 p.
Lescure, J. 1973. Notes biogeographiques sur quel- Muller, P. 1974. Aspects of zoogeography. Junk.
ques Amphibiens du bassin superior du Maroni. The Hague, 208 p.
C. R. Seances Soc. Biogeogr. (439):58-63. Noble, G. K. 1931. The biology of the Amphibia.
Lescure, J. 1975. Biogeographie et ecologie des McGraw-Hill Book Co., New York, 577 p.
Amphibiens de Guyane Francaise. Ibid. (440): Nussbaum, R. A. 1977. Rhinatrematidae: A new
68-82. family of caecilians (Amphibia: Gymnophiona).
Lescure, J. 1976. Contribution a l'etude des Am- Occas. Pap. Mus. Zool. Univ. Michigan (682):
phibiens de Guyane francaise. VI. Liste pre- 1-30.
liminaire des Anoures. Bull. Mus. Natl. Hist. Pace, A. E. 1974. Systematic and biological studies
Nat. Zool. Paris (265): 475-525. of the leopard frogs (Rana pipiens complex) of
Lutz, A. 1929. Taxonomia e biologia do genero the United States. Misc. Publ. Mus. Zool. Univ.
Cyclorhamphus. Mem. Inst. Oswaldo Cruz 22(1): Michigan (148): 1-140.
1-25. Parker, H. W. 1934. A monograph of the frogs of
Lutz, A. 1931. Observacoes sobre batrachios brasi- the family Microhylidae. Brit. Mus. (Nat. Hist.),
leiros. Taxonomia e biologia des elosiinas. Ibid. London, 208 p.
24(4): 195-222. Parker, H. W. 1935. The frogs, lizards, and snakes
of British Guiana. Proc. Zool. Soc. London 1935
Lutz, B. 1972. Geographical and ecological notes
on Cisandine to Platine frogs. J. Herpetol. 6: (3):505-530.
83-100. Peters, J. A. 1973. The frog genus Atelopus in
Ecuador (Anura: Bufonidae). Smithsonian Con-
Lutz, B. 1973. Brazilian species of Hyla. Univ.
trib. Zool. (145): 1-49.
Texas Press, Austin, 265 p.
Pianka, E. R. 1969. Habitat specificity, speciation,
Lynch, J. D. 1971. Evolutionary relationships, oste-
and zoogeography and species diversity in Australian desert lizards.
ology, of leptodactyloid frogs.
Univ. Kansas Mus. Nat. Hist. Misc. Publ. (53): Ecology 50:498-502.
1-238. Richards, P. W. 1952. The tropical rain forest.
Cambridge Univ. Press., Cambridge, 450 p.
Lynch, J. D. 1973. The transition from archaic to Rivero, A. 1961. Salientia of Venezuela. Bull.
J.
advanced frogs, pp. 133-182 in Vial, J. L. (ed.).
Mus. Comp. Zool. Harvard Univ. 126:1-207.
Evolutionary biology of the anurans: Contempo- Rosen, D. E. 1975. A vicariance model of Caribbean
rary research on major problems. Univ. Missouri
biogeography. Syst. Zool. 24:431—464.
Press,Columbia, 470 p.
Salthe, S. N., Duellman, W. E. 1973. Quantitative
Lynch, J. D. 1975. The identity of the frog Eleu- constraints associated with reproductive modes in
therodactylus conspicillatus (Gunther), with de- anurans, pp. 229-249 in Vial, J. L. (ed.). Evo-
scriptions of two related species from northwest- lutionary biology in the anurans: Contemporary
ern South America. Nat. Hist. Mus. Los Angeles research on major problems. Univ. Missouri Press,
Cty. Contrib. Sci. (272): 1-19. Columbia, 470 p.
Lynch, J. D. 1976. The species groups of the South Savage, J. M. 1968. The dendrobatid frogs of Cen-
American frogs of the genus Eleutherodactylus tral America. Copeia 1968(4) 745-776. :
APPENDICES
Appendlx 8:1. —Amphibians of the Trans-Andean Forests.
Nonendemic taxa (i.e., found inone or more other tropical lowland forests) are noted by an asterisk and letters
denoting the lowland forests (T = Trans- Andean, N = Northern, C = Central Cis-Andean, A = Atlantic).
Salamanders Barchyolos pulcher Dendrobates truncatus
° NC
Oedipina complex Leptodactylus bolivianus Dendrobates viridis
Oedipina parvipes Leptodactylus melanonotus Phyllobates anihonyi
SAC
Bolitoglossa biseriata Leptodactylus pentadactylus Phyllobates aurotaenia
"" bicolor
Bolitoglossa chica Leptodactylus poccilochilus Phyllobates
Bolitoglossa medemi Leptodactylus ventrimaculatus Phyllobates boulengeri
° NCA
Bolitoglossa phalarosoma Leptodactylus wagneri Phyllobates espinosai
*"
Bolitoglossa silverstonei Physalaemus pustulosus Agalychnis calcarifer
Bolitoglossa sima Eleutherodactylus sp. "A" Agalychnis callidryas (expected)
Bolitoglossa taijlori ( expected ) Eleutherodactylus acliatinus Agalychnis litodryas
Eleutherodactylus anomalus Agalychnis spurrelli
Caecilians
Eleutherodactylus areolatus Phyllomcdusa sp.
Epicrionops marmoratus Eleutherodactylus biporcatus Phyllomedusa venusta
Epicrionops parkeri Eleutherodactylus bufoniformis Hemiphractus fasciatus
Caecilia antioquiaensis Eleutherodacttjlus caryopliyllaceus Gastrotheca angustifrons
Caecilia caribea Eleutherodactylus caprifer Gastrotheca cornuta
Caecilia guntheri Eleutherodactylus cruentus Gastrotheca nicefori
* NC
Caecilia leucocephala Eleutherodactylus diastema Hyla boons
Caecilia nigricans Eleutherodactylus fitzingeri Hyla boulengeri
Caecilia perdita Eleutherodactylus gaigeae Hyla ebraccata
Caecilia subnigricans "" Eleutherodactylus gularis Hyla gryllata
Caecilia tentaculata ° c Eleutherodactylus latidiscus Hyla miliaria
Caecilia tliompsoni Eleutherodactylus longirostris Hyla pellucens
Dennopliis parviceps Eleutherodactylus mow Hyla picturata
Oscaecilia och rocephala Eleutherodactylus ornatissimus Hyla phlebodes
Oscaecilia polyzona Eleutherodactylus raniformis Hyla quinquefasciata
Parvicaccilia nicefori Eleutherodactylus ridcns Hyla rosenbcrgi
° NCA
Parvicaecilia pricei Eleutherodactylus roseus Hyla rubra
Typhlonectes natans Eleutherodactylus subsigillatus Hyla ruhracyla
Eleutherodactylus tacniatus Hyla subocularis
Anurans
Eleutherodactylus walkeri Hyla sugillata
• NCA
Atelopus balios Eleutherodactylus sp. "Z" Phrynohyas venulosa
Atclopus elegans Colostethus chocoensis Smilisca phacota
Atelopus glyphus Colostethus imbricolus Smilisca sila
Atelopus longibrachius Colostethus latinasus Trachyccphalus jordani
Atelopus longirostris Colostcthtis nubicola Centrolenclla flcischmanni * N0
Atelopus mindoensis Colostethus pratti Centrolenella ilex
Atelopus spurrelli Colostethus talamancae Centrolenclla ocetlifera
Bufo blombergi Dendrobates auratus Centrolenella prosoblcpon
Bufo coniferus Dendrobates fulguritus Centrolenella spinosa
Bufo huematiticus Dendrobates histrionicus Glossostoma aterrim um
1979 LYNCH: AMPHIBIANS OF LOWLAND TROPICAL FORESTS 213
James R. Dixon
Department of Wildlife and Fisheries Sciences
Texas AirM University
College Station, Texas 77843
USA
The task of summarizing the distribution, Dixon and Soini, 1975, 1977; Rand and
natural history and evolution of approximate- Humphrey, 1968) have investigated these
ly half of the known species of reptiles of topics.
South America is a nearly futile exercise be- Vanzolini (
1967 ) discussed the problems
cause the taxonomy and distribution of many of application of ecological principles to the
important groups are poorly known. Never- Amazonian biota without having a firm under-
theless, there are some redeeming values in standing of the taxonomy and evolution of
any exercise —
in this case, a summary of the its constituent groups. Only through detailed
state of our knowledge of the distribution and taxonomic studies of various lowland tropical
ecology of the South American reptilian rainforest genera and species of reptiles (Uz-
fauna. zell, 1966; Oliver, 1948; Duellman, 1958; My-
As a matter of convenience throughout ers, 1974; Savage, 1960; Kluge, 1969; Peters,
this paper, I refer to rainforest reptiles, rather 1960; Gans, 1971; Dixon, 1973, 1974a,b; Roze,
than "tropical lowland rainforest reptiles." 1967; Ruibal, 1952) and geographic and/ or
Rainforest reptiles include those genera and ecological information ( Rand and Humphrey,
species that occur below 1000 elevationm 1968; Crump, 1971; Reebe, 1944a-c, 1945;
and within the environs of rainforest (i.e., Vanzolini, 1968, 1972, 1974, 1976; Vanzolini
yearly temperature varying between 23° and and Williams, 1970; Test, et al., 1966; Duell-
28°C; more or less continuous rainfall be- man, 1978) will the reptilian rainforest fauna
tween 150 and 600 cm annually). In the con- ever be fully understood.
text of this paper, any reptile that occurs The following information obtained for
within the conditions mentioned above quali- forest associations and their included reptile
fies as a rainforest species, although it may is only a rough estimate and, without
species
not be one ecologically. There are about 14 doubt, will be subject to much future revision.
genera of savanna relicts within the confines We are raising questions that cannot be
of the rainforest environment, an additional answered firmly without further studies on
22 genera are aquatic, and about 30 genera fossil histories, taxonomy, and ecology of par-
are fossorial (Table 9:1). Several genera are ticular reptile groups. Time is short, for the
wide ranging, occupying many vegetation being altered by man
rainforests are rapidly
zones, including rainforests. Therefore, the (Richards, 1973), and data concerning the
task of assigning a genus and/ or its attendant reptile fauna soon may be impossible to
species as rainforest endemics, frequently is obtain.
difficult without some basic knowledge of Without detailed natural history observa-
their ecology. tions on the majority of the species, little can
There is a limited understanding of the be accomplished concerning the relationships
evolution and distribution of rainforest rep- among rainforest reptiles. For example, the
tiles in South America because only a few
taxonomy and distribution of Anolis is so
individuals (Vanzolini, 1972, 1974, 1976; Van- poorly known (Williams, 1976) that the large
zoliniand Reboucas-Spieker, 1973; Duellman, number of described South American taxa
1978; Fitch, 1970; Myers, 1974; Crump, 1971; (80) is enough to confuse and mask the rela-
217
218 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Table 9:1. —
Reptilian Genera that Contain: A) Species that Are Savanna Relicts; B) Species that Are
Aquatic or Semi-aquatic; C) Species that Are Semi-fossorial within Lowland Rainforest Situations.
Rrasilian shields, the Orinoco (Llanos) Basin rainforests were continuous geographically
of Venezuela, or the Parana Basin of Para- throughout much of the Early and Middle
guay, is indicated by the presence of brack- Cenozoic, subjected to disruption during the
ish and saltwater fossil deposits found in Miocene and Pliocene orogenies (Emiliani,
eastern Peru and Ecuador (Haffer, 1974). 1956; Dorf, 1959; Tanner, 1968), and severely
It has been suggested that during the Late restricted to relatively small refugia by wind,
Cenozoic the upper Amazon Basin was a temperature, and rainfall patterns during and
huge inland sea, in which sediments were following the Pleistocene (Flint, 1971; Sioli,
constantly being deposited through torrential 1975; Simpson, 1975; Raven and Axelrod,
rains upon the Andes. These runoffs created 1975; Emiliani, 1972; Haffer, 1974; Vuilleu-
vast flood plains that were crossed by numer- mier, 1971; Damuth and Fairbridge, 1970).
ous meandering streams that now form the Most modern reptilian genera with wide
major river systems of the Amazon Basin. distributions that also occur in rainforests
lack fossil records in South America (Amphis-
Although the narrow straits between the
Guiana and Brasilian shields may have been baena, Ameiva, Iguana, Mabuya, Clelia, Epi-
blocked at times (with subsequent flow crates,Erythrolamprus, Lachesis, Leptodeira,
through the Orinoco or Parana basins) an Leptophis, Mastigodryas, Spilotes, Tantilla,
almost continuous connection with the At- Tripanurgos) However, their current distri-
.
lantic Ocean seems likely. Haffer's (1974) butions imply that they may have evolved
review of other studies indicates that Marajo concurrently with those widespread genera
Island, at the mouth of the Amazon, is under- with recorded fossil histories in the Miocene
lain by over 4000 m of fossiliferous marine and Pliocene (Dracaena, Tupinambis, Teius,
and brackish, Cretaceous and Tertiary sedi- Diplolaemus and Crotalus). Some of these
ments. The evidence suggests that continen- genera are also xeric adapted; thus, it is im-
tal fluvial sediments were transported both possible to postulate their origins without
from east and west to the sub-Andean basin fossils or other evolutionary evidence. Estes
of the upper Amazon. Permanent, easterly and Price (1973) identified Pristiguana from
directed flow of sediments probably followed theUpper Cretaceous in Brasil as the earliest
the final major uplift of the Andes in late known iguanid and suggested a Gondwanan
Pliocene and early Pleistocene. origin of the family based on morphological,
Raven and Axelrod (1975) showed that paleontological, and zoogeographical consid-
erations.
the angiosperm flora of South America had
a definite relationship to Africa through the
Eocene, but as the plates ( continents ) drifted The Fossil Evidence
apart, each continent developed a unique The fossil history of rainforest reptiles is
flora. The process was enhanced
by the Mio- relatively good for those groups with large
cene orogenies in eastern Africa and western
skeletal features (e.g., turtles and crocodil-
South America, changes in wind and tempera-
ians) but extremely poor for amphisbaenians,
ture patterns associated with those orogenies,
lizards, and snakes. The fossil record of am-
and the influence of angiosperms from tropi-
phibians and reptiles in South America has
cal Asia thatimmigrated to the approaching been documented by Baez and Gasparini
African continent via sweepstakes dispersal.
(this volume). Only a few comments espe-
Much of the diversity of the tropical flora of
cially pertinent to rainforest groups are in-
Africa was diminished by the orogenies of the
cluded here.
eastern plateau of Africa and associated wind,
The presence of relatively modern gavials
temperature, and hydrologic pattern shifts,
(Langston, 1965) in the Rio Magdalena of
but South America apparently maintained
Colombia during the upper Oligocene and in
several centers of floral refugia (Haffer, 1974; the Pliocene in Argentina suggests that the
Vanzolini,1973; Midler, 1973; Vuilleumier, lowland tropical vegetation was extensive
1971) and thus retained a greater diversity throughout much of South America for long
of angiosperms. Historically, it seems that the periods of time. This is further enhanced by
220 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
the probable replacement of gavials as pisciv- pendemys, Trionyx and possibly Taphros-
orous carnivores by longirostrine crocodiles, phys) and fourteen genera of crocodilians
first in northern South America and later in (Gavialis, Sebecus, Charactosuchus, Rham-
the south. Langston (1965) suggested that phostomopsis, Nettosuchus, Eocaiman, Bra-
were more adaptive
longirostrine crocodiles chygnathosuchus, Gryposuchus, Proalligator,
and were probably replacing the gavials ( tak- Leptorrhamphus, Purassaunis, Ilchunaia, No-
ing over the fish-eating niche) throughout the tocaiman, Necrosuchus) The majority of the
.
Cenozoic Era. Langston pointed out that the species of those genera that survived the
significance of the record is that longirostrine climatic,edaphic, and hydrologic shifts of
crocodiles were assuming this role, not alli- Miocene/ Pliocene eras (Paleosuchus, Mela-
gatorids. However, the fossil record indicates nosuchus, Caiman, Crocodyhis, Podocnemis,
that the alligatorids were evolving at the Chelus, Phrynops, Kinosternon, Rhinoclem-
same time that longirostrine crocodiles pre- mys, Chrysemys, Geochelone) are those that
sumably were competing with gavials. Cli- exist today in Neotropical rainforests (Wood,
matic changes brought about by Mio-PIiocene 1976; Wood and Gamero, 1971; Wood and
orogenies may have severely limited the habi- Patterson, Romer, 1956; Auffenberg,
1973;
tatand tolerances of longirostrine crocodiles 1974; Langston, 1965; Medina, 1976).
and they in turn, were competively excluded The presence of the lizard genera Dra-
by alligatorids. The presence of modern caena, Tupinambis, and cf. Polychrus (Estes,
crocodiles occupying parapatric geographic 1961) in the Miocene deposits of the upper
ranges in South America suggests that they Rio Magdalena of Colombia strongly suggests
(C. acutus, C. intermedins) are relicts of that many contemporary lizard genera were
some ancient stock or populations that were widespread rainforest occupants. Although
excluded in earlier competition, were (or few fossil South American lizard and snake
became) salt tolerant, survived in island re- genera are currently known, many modern
fugia, and are recent invaders. Cis-Andean and Trans-Andean genera were
The fossil history of South American tur- probably well differentiated prior to the
tles is even more confusing, perhaps owing orogeny of the Andes (Estes, 1961).
to the conservative nature of the morphologies Although we have little evidence that
of the groups and the widespread distribution tropical environments have remained, in part,
of the pleurodires in South America and the from the Mesozoic to modern times, the fossil
rest of the world. Many early fossils of pleu- evidence of tropical reptiles in South America
rodires were misidentified and, only recently, suggests persistent tropical environments in
have workers (Auffenberg, 1974; Wood, 1972, most areas of South America prior to the up-
1976; Wood and Gamero, 1971) attempted to lift of the Andes (Harrington, 1962).
unravel the myriad of names assigned to fos-
sils discovered in the past two centuries. The Historical Relationships of Currently
presence of Podocnemis and Chelus in the Disjunct Rainforests
Pliocene in Estado Falcon, Venezuela, Chelus
The distribution of rainforest reptiles is
in the Miocene in the upper Rio Magdalena
dependent ultimately upon the history of en-
Valley, Colombia, Podocnemis in the Late vironmental factors that have affected the
Cretaceous in Brasil and in the Eocene/ Plio-
habitat and the reptilian gene pool. The
cene in Peru, suggests that turtles and croco-
increased vagility of one group compared
dilians were occupying similar habitats dur-
with another, the inherent, fixed attributes of
ing comparable periods of their evolution in
a particular species versus those with highly
South America.
variable attributes, the physical and ecolog-
The orogenies of the late Miocene and ical barriers between different and/or equal
throughout much of the Pliocene disrupted habitat areas through time, and diffuse com-
the general distributions of many turtle and
petition from other organisms within the
crocodile species and apparently caused the niche all and/or
hypervolume play major
demise of four genera of turtles in South minor roles in formulating the distribution of
America (Schwoeboemys, Parahijdraspis, Stu- various groups of reptiles.
1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 221
Amazonian rainforests, and finally, the Cho- (1970), Vuilleumier (1971), Simpson (1975)
coan rainforests were separated from the and Raven and Axelrod (1975) have garnered
Amazonian by the closure of the
forests evidence concerning the Quaternary history,
Huancabama Based on fossil and
Deflection. especially dealing with arid/ humid
that
present distributions of some reptiles, there ( glacial/ interglacial) phases from the Pleisto-
are indications that a lowland rainforest cor- cene to 2500 years ago. Most data place the
ridor may have existed between the Chocoan end of the Wisconsin glaciation at 10,000 to
and Amazonian forests through the Huan- 15,000 y.b.p., with two short arid cycles 4,000
cabama Deflection during the last major in- and 2,500 y.b.p. These may have caused the
terglacial period. The fossil record reveals present hiatus between the Amazon and At-
several genera that are geologically and genet- lantic forests. Three to four glacial periods
ically older than the Andes and that occur are postulated for the Andes and most other
now, or occurred in the past, on both sides regions, and there are five terrace levels along
of the mountains. The ranges of several spe- the Rio Amazonas resulting from cyclic ero-
cies of snakes
belonging to three genera sion and aggradation (Haffer, 1974, this vol-
I assume there was at least one pre-
(Chironius, Leimadophis, Leptophis) seem to ume).
have been connected recently, probably glacial period of long duration prior to the
through the Huancabama Deflection. Pleistocene, that was followed by four inter-
The water gaps between Central Amer- glacial periods of various lengths of time after
ica and South America may have been nar- each of the four major glacral periods of the
row enough to allow some interchange of Pleistocene. Because the Amazonian forest
faunas between the two continents during the is relatively young—800,000 to 1,800,000 years
Cenozoic ( Hershkovitz, 1966; Savage, 1974). and has been affected variously by flooding
Land connections (not necessarily continu- and/ or cool arid cycles, I assume that the
ous) between those regions beginning in the five, recognized terrace levels are the result
Pliocene (Haffer, 1974) certainly allowed of preglacialand interglacial flooding. If we
faunal interchange on a larger scale. Rain- assume that each arid glacial ) phase created
(
forest faunas probably were exchanged when broad savanna regions and restricted the rain-
222 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Table 9:2. —
A) North and Central American and/or Endemic Reptilian Genera that Contain Lowland Trop-
ical Rainforest Species that Reach Their Southern Distributions in the Northern and Chocoan forests (
" =
Endemic to the Choco Forest). B) South American Reptilian Genera that Contain Lowland Tropical Rain-
forest Species that Reach Their Northern Limits in the Northern and Choco forests ( f Indicates a Savanna
Relict).
forest to relatively small refugia (Fig. 9:1), ic variation,suggesting that a large number
the union of the restricted Atlantic and Ama- of species may have reached the Atlantic
zonian rainforests may have occurred an forest during the last Amazon-Atlantic con-
suggests that the "Diagonal of Open Forma- ber of times this genus extended its distribu-
tions" (Vanzolini, 1963, 1974) has been pres- tion into the Atlantic forest during interglacial
ent from Late Tertiary to present with only periods with alternate glacial periods of spe-
minor alterations. Therefore, a belt of Ama- ciation in both the Amazonian and Atlantic
zonian forest may have reached the Atlantic forests.
forestas indicated above, or via the Rio In the case of Cliironius, there seems to
Madeira-Rio Parana Drainage around the be a progressive intrusion of recently-differ-
west and south sides of the Brasilian Shield. entiated Amazonian stocks resulting from the
A filter corridor through the Brasilian Shield repeated contraction of the forest into refugia
along gallery forests also may have been pos- during at least four glacial periods. Each
sible, but these routes probably were restric- interglacial expansion of the Amazonian forest
tive. Many Amazonian reptiles have reached that reached the Atlantic forest seemed to
only the northern part of the Atlantic forest carry with it the more primitive members of
(e.g., Anolis punctatus, A. ortoni, A. fusco- each evolving gene pool. Without fossil evi-
auratus, Kcntropyx calcaratus, Bothrops bi- dence, the model is purely hypothetical, but
lineatus, Chironius carinatus, Coleodactylus Wiest ( pers. comm. ) constructed an inde-
meridionalis, Polychnis marmoratus, Tripan- pendent model, and we differ only on minor
urgos compressus, Mastigochyas boddaerti, points. If we assume that the primitive stock
Dipsas indica), thereby suggesting that the of the gene pool is peripheral to the de-
former route is the most plausible. Most veloping core and the set of characters defin-
Amazonian forest species that have reached ing the primitive conditions for Chironius are
the Atlantic forest show little or no geograph- correct, the following sequence of events
Distribucion propuesta para el Plcistoceno dc los tipos de vegetation Suramericana durante los pcriodos
de maxima glacial. D= Desicrto; GI = Hielo Glacial; GL = Pastizales; = Monte; NE = Bosque de
M
Nothofagus; P = Paramo (Alpino); PS = Matorral semi-desertico dc Patagonia; S = Savana; TF/S =
Matonal TM = montano; TR = Bosque tropical lluvioso; TS-E/D
dc bosque espinoso; =
Bosque tropical
Bosque tropical semi-siempreverde/detiduo. La zona punteada y con Uncus diagonal cs en la cuenca ama-
zonica puede habcr sido un bosque lu'imcdo tropical.
1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 223
Fie. 9:1. Proposed Pleistocene distribution of South American vegetation types during maximum glacial periods.
D = Desert;GI = Glacial Ice; GL = Grassland; M = Monte; NF = Nothofagus Forest; P = Paramo
(Alpine); PS = Patagonian (Semi-desert Scrub); S = Savanna; TF/S = Thorn Forest/Scrub; TM = Trop-
ical montane; TR = Tropical rainforest; TS-E/D = Tropical semi-evergreen/deciduous Forest. Dashed and
diagonal lined area in Amazon Basin may have been an additional isolated tropical rainforest.
224 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
should represent the various invasions of 5. The second interglacial period allowed
Chironius stock (and possibly other reptiles) a reunion of the lowland rainforests.
into the Atlantic forest. Chironius scurrulus invaded the ex-
1. Froto-Chironius stock invaded the At- panded Atlantic forest for the first time,
lantic during a pre-Pleistocene
forest while C. multiventris and exoletus did
period when the two lowland tropical so for the second time. Chironius fo-
rainforests were connected along a veatus had developed in the Atlantic
coastal corridor, from the mouth of the forest as an ecological equivalent of
Amazon eastward and southward along C. multiventris, and the two species
the northeastern part of Brasil. formed competitive populations. Chir-
2. During the first major Pleistocene gla- onius bicarinatus did not differentiate.
cial period, the forests were discontin- Previously allopatric populations of C.
uous refugia with C. multiventris and exoletus interbred along zones of con-
C. exoletus as developing taxa in Ama- tact.
zonian refugia and C. bicarinatus in 6. The third glacial period reduced the
Atlantic refugia. Two additional spe- forests to discrete refugia. Chironius
cies, C. quadricarinatus and flavolinea- exoletus, multiventris and scurrulus re-
tus, were speciating in gallery forests mained undifferentiated populations in
within savannas (the "Cerrado-Caatin- Amazonian refugia. Chironius carina-
ga Diagonal" of Vanzolini, 1963, 1974) tus became a developing biological spe-
between the two rainforests. cies from an isolated C. exoletus stock
3. The major Pleistocene interglacial
first in the Guianan Forest Refugium. Chi-
period allowed expansion of the forest ronius bicarinatus and foveatus re-
refugia and eventual reconnection of mained relatively undifferentiated in
the Atlantic and Amazonian forests. Atlantic forest refugia, while the com-
Chironius quadricarinatus and C. fla- peting Atlantic populations of C. multi-
volineatus evolved into full, biological ventris became extinct. The semi-iso-
species and extended their
ranges lated Atlantic population of C. exoletus
within the diagonal of open formations differentiated subspecifically (C. e.
in Brasil where they have remained pyrrhopogon), and an isolated Atlantic
to the present time. In Amazonia, C. population of C. scurrulus diverged into
multiventris and exoletus had devel- C. s. laevicollis.
oped into biological species and now 7. During the third interglacial period, a
invaded the Atlantic forest, coming into reunion of the two forests allowed for
contact with C. bicarinatus, a species the reinvasion of C. exoletus, multiven-
with a different diet and probably not tris, scurrulus and an initial invasion of
a competitor. C. carinatus into the Atlantic Forest.
4. The second, major glacial period again Chironius bicarinatus and foveatus re-
reduced the forests to discrete refugia. mained intact, although the latter again
The C. exoletus stock remained in the became competitive with the invading
Atlantic refugia with some contact with C. multiventris. The once isolated pop-
C. exoletus populations in Amazonia ulations of C. exoletus-scurrulus formed
along gallery forests, while the Atlantic intergrading complexes.
population of C. multiventris became 8. The fourth glacial period reduced the
isolated and formed a vicariant bio- forests into discrete refugia once more.
logical species (C. foveatus). Chiron- Chironius exoletus, scurrulus, multi-
ius bicarinatus remained undifferenti- ventris, and carinatus remained undif-
ated in Atlantic refugia while C. multi- ferentiated in Amazonian forest refugia,
ventris did likewise in the Amazonia while an isolated population of Ama-
refugia. Chironius scurrulus developed zonian C. scurrulus developed into a
into a biological species in one of the biological species (C. fuscus). The
several Amazonian forest refuges. isolated Atlantic Forest population of
1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 225
briefly invaded the Atlantic forest dur- reptiles in theWestern Hemisphere that are endemic
to, or shared with, each of the two major continents
ing a connection of the two forests and the Central American corridor.
somewhere between 10,000 and 4,500
El porcentaje de los 335 generos de reptiles en el
and the latter species invaded
y.b.p., Hemisferio Occidental que son endemicos para o
for the first time. Atlantic populations compartidos con cada uno de los continentcs mayores
of C. carinatus and fuscus currently are y con el corredor centroamcricano.
throughout the Atlantic Forest. era and 49.5 percent of the species. It also
represents 44.5 percent of all genera and
17.8 percent of the species in the Western
COMPOSITION OF THE
HERPETOFAUNA Hemisphere (Fig. 9:2).
Of the 149 reptile genera, 21 (14.1%) are
shared between the Choco and the Amazon;
Faunal Similarity
17 (10.7%) between the Atlantic and the Ama-
At the rate taxonomists are describing new zon; 3 (2.7%) between the Choco and the
taxa of reptiles from the Neotropics ( 10-30 Atlantic; 38 (25.5%) among all major Neo-
species per year), it will be a long time be- tropical rainforests (Fig. 9:3, Table 9:3). Of
fore we obtain sufficient knowledge to allow the 550 species, 23 (4.3%) are shared between
226 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
GENERA
SPECIES
the Choco and the Amazon; 43 (8.0%) be- cies, 284 (51.4%) are snakes, 235 (43.0%) are
tween the Atlantic and Amazon; 1 (0.02%) lizards [includes 30 (5.5%) amphisbaenids],
between the Choco and the Atlantic; 18 24 (4.4%) turtles and 6 (1.1%) crocodilians.
(3.1%) all major Neotropical forests
among Of 95 lizard and 94 snake genera in South
(Fig. 9:4,Appendix 9:1). The smaller trans- America, 63 (66.3%) and 72 (76.6%), respec-
Andean forests (Magdalena, Sinu, Nechi, tively, occur in rainforests.
Santa Marta, Maracaibo, and coastal Vene- I agree with Burt (1958) and Udvardy
zuela) show little endemism ( tian 4%) ( 1969 )
that any formula chosen to obtain
and those closest to the Choco are essentially indices for similarities or differences between
identical in their reptile faunas to the latter. faunal elements depends upon the accuracy
Each of the more isolated of these forests of the basic data set, the relative size and/or
(Santa Marta, Maracaibo, coastal Venezuela) equivalent nature of each of the faunal areas
contains one lowland rainforest endemic spe- and elements, and the taxonomic category
cies. The coastal Venezuelan Forest shows chosen for the comparison (Fig. 9:5A).
Amazon Forest, whereas
closer affinities to the Perhaps one of the best ways to compare
the Maracaibo and Santa Marta forests con- faunal elements is to obtain subsets of data
tain a number of species that are broadly from equivalent areas within any given, major
distributed in both Central and South Amer- faunal region. The investigator thus gains an
ica (see Appendix 9:1). understanding of the variation within and be-
Of
the 149 genera of rainforest reptiles, tween geographic subsets of data. It seems
72 (48.3%) are snakes, 63 (42.3%) are lizards obvious that an increase in the number of
[includes 5 (3.4%) amphisbaenids], 10 (6.7%) species of any particular group from a given
turtles and 4 (2.7%) crocodilians; of the spe- area will decrease the faunal similarity index
1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 227
Table 9:3. —Genera Shared among the Three Major Tropical Lowland Rainforests.
Amazon/ Atlantic Amazon/Choco Amazon/Choco/Atlantic
Apostolepis Alopoglossus Ameiva Mabuya
Cercosaura Bach ia Amphisbaena Mastigodryas
Coleodactylus Boa Anolis Micrurus
Colobosaura Cnemidophorus Atractus Oxyrhopus
Eny alius Corallus Bothrops Polyclinic
Hemidactylus Dendrophidion Caiman Pscudoboa
Kentropyx Drymarchon Chironius Pscustes
Lcpostcmon Drymobius Clelia Rhadinaca
Palcosuchus Enyalioidcs Dipsas Spilotcs
Pantodactylus Gonatodes Epicrates Siphlophis
Philodryas Gymnophthalmus Erythrolamprus T ant ilia
Plirynops Imantodes Geoclielone(?) Tripanurgos
Platemys Lepidoblepharis Hclicops Tropidophis( ?)
Podocnemis Morunasaurus Iguana Tupinamhis
Thamnodynastcs Ninia Kinostemon Typhlops
Tropidurus Oxybelis Lachesis Xenodon
Tupinamhis Pliocercus Leimadophis
Xenopholis Prionodactylus Leposoma
Choco/Atlantic Ptychoglossus Lcptodcira
Diploglossus Rhinobothryum Leptophis
Liotyphlops Synophis Leptotyphlops
Lygophis Thccadactylus Liophis
Although the Amazonian and Atlantic for- The variation in size of independent rain-
ests tend to show a closer relationship to each forests does not seem to dictate the number
other than to the Choco Forest, there are 18 of silvicolous lizards. The predictors of di-
species and 38 genera that are common to all versity are the same for each forest —
1) low-
three forests. When the common species are land tropical rainforest; 2) latitudes; 3) cli-
removed from consideration and attention is mate; 4) structural heterogeneity; and 5) de-
focused upon the probable origin of the spe- gree of trophic specialization. The only in-
cies and/or genera, a more meaningful rela- dependent variable is total land area occupied
tionship can be observed between various by the rainforests. Assuming that the pre-
South American ecophysiographic regions dictors for each species in each forest are
Atlantic (77)
CenlroUmin (28)
Amazon (318)
Atlantic (150)
Amazon (166)
Amazon (531
Amazon 11121
Atlantic 1141
Amazon (41)
Amazon 126)
Atlantic 1111
1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 229
be relatively constant, providing that the age, recourse for quality information on faunal
evolution, and environmental conditions with- relationships (Pianka, 1974). For example,
in the forests are similar. My
knowledge of the numbers of lizards inhabiting the forest
silvicolous lizards suggests that the Choco floor of the Choco, Amazon, and Atlantic, and,
contains about 38 species, Amazon 44, Congo in parentheses, those that are true leaf litter
39, and Thailand 40. Thus, the predicted inhabitants of primary rainforest are 22(18)
similarities in numbers of species between in the Choco, 43(35) in the Amazon, and
forests, regardless of size, apparently are 19(18) in the Atlantic forest. These figures
realistic. Therefore, if we have knowledge of indicate that forest size may be important
certain evolutionary and ecological parameters for terrestrial lizards; almost twice as many
concerning the habitat and its attendant spe- species of terrestrial lizards occur in the
cies, we should be able to predict the number Amazon than in the other two forests. Among
of species that may occupy that habitat re- several possible explanations for the influence
gardless of their generic or familial status or of forest size on the number of Amazonian
size of the forest. Iguanids and teiids basi- lizards are the following. 1) Various seg-
cally fill the silvicolous-food-generalist role in ments of the Amazonian forest may have had
the Choco and Amazon forests, gekkonids and different evolutionary histories owing to cli-
chamaelionids in the Congo, and gekkonids matic alterations during the Quaternary,
and agamids in Thailand. whereas the Choco and Atlantic forests may
The constraints placed upon a terrestrial have remained as single units (Fig. 9:1).
lizard are much more difficult to discern than 2 Physical attributes of the forest floor ( sand
)
those of a silvicolous arboreal one. Compari- hills, lowland swamps, different soil proper-
son of arboreal, eurytrophic lizards is pos- ties,slope exposure, depth of leaf litter) may
sible between rainforests of similar latitudes, have affected speciation over time. 3) A
temperatures, and general physiognomy. longer ecotone between savanna and forest
However, comparison of terrestrial lizards is affected heterogeneity. I favor the first al-
more difficult owing to the patchiness of the ternative because the distribution of reptiles
forest floor. Adequate knowledge of the coincides with postulated Quaternary refugia
habits of the animals in question is the only (Figs. 9:8-9). Why such refugia would pro-
Fig. 9:5.(A) The three tropical lowland rainforests of South America compared for reptilian faunal similar-
ities based on genera and on species. Simpson's formula (C/N,) X 100 is used because it indicates the
percentage of common taxa in the smaller of the two faunas, regardless of the size of the larger one. How-
ever, the larger fauna must also be compared to obtain similar data in the opposite direction. The solid lines
of the triangle represent the percent of the common taxa in the larger of the two faunas (direction of arrow
is from larger to smaller fauna), whereas the inner dashed line shows percentage of common taxa in the
smaller of the two faunas (direction of arrow from smaller to larger fauna). Numbers in parentheses following
names of forests represent number of taxa within that forest. (B) A Simpson formula (C/Ni) X 100 com-
parison of species of lizards and snakes from three local rainforest sites within the Iquitos region, Peru. When
faunas of equal size are compared (number of snakes between Centra Union and Mishana), then Duellman's
(1965) modified formula (2C/Ni + N = ) X 100 is used because the resulting Faunal Resemblance Factor
compares Simpson's formula in both directions with a resultant average of the two percentages. (C, D, and
E) Simpson's Formula (C/N,) X 100 comparing various reptilian species components in the three tropical
lowland rainforests of South America.
(A) Comparacion de los tres bosques tropicales de tierra baja de Sur America, en base a similaridad de la
fauna reptiliana en cuanto a generos y especies. Se uso la formula de Simpson (C/Ni) X 100 porque indica el
porcentajc de taxones coryiunes in la mas pequena de las dos faunas, sin importar el tamano de la mayor. Sin
embargo, la fauna mayor debe de ser tambien comparada para obtener datos similares en direccion opuesta.
Las lineas solidas del tridngido representan el porecntaje de laxones comunes en la mas grande de las dos
faunas (la direccion de la flecha va de la mayor a la menor fauna), mientras que la linea quebrada interior
muestra el porcentaje de taxones comunes en la mas pequena de las dos faunas (la direccion de la flecha
va de la menor a la mayor fauna). (B) Comparacion de las especies de la lagartos y serpientes de tres locali-
dades de bosque amazonico dentro de la region de Iquitos, Peru, usando la formula de taxa (como las de
Ccntro Union y Mishana), entonces, usamos la formula modifieado por Duellman (1965) (2C/N, + N : ) X 100
porque el factor de similaridad resultante compara en dmbas direcciones con una resultante de la media de
los dos porcentajes. (C, D, and E ) Comparacion de varias especies componentes de tres bosques tropicales de
tierra baja en Sur America usando la formula de Simpson.
230 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fie. 9:6. Percentage contribution of nonendemic reptilian species to a particular rainforest. Number
within a circle represents approximate percent of endemism of reptilian species for a particular forest.
Contribution porcentual de especies no endemicas de reptiles a un basque tropical particular. El numero
dentro del circuit) representa el porcentaje aproximado de endemismo de especies de reptiles para un bosque
en particular.
1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 231
75
232 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fie. 9:8. Generalized ranges of rainforest lizards that show restricted distributions. In addition, there
are 18 species of lizards that are distributed throughout the Amazonian forest.
Distribuciones generalizadas de saurios de bosque </nc muestran distribuciones restringidas. Hay IS espe-
cies adicionales distribuidas a lo largo del valle amazonico.
1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 233
Fig. 9:9. Generalized ranges of rainforest snakes that show restricted distributions. In addition, there are
46 species of snakes that are distributed throughout the Amazon Basin.
Distribuciones generalizadas de serpientes de bosque que muestran distribuciones restringidas. Hay 46
especies adicionalcs distribiiidas a lo largo del voile amazonico.
234 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
por una o mas especies en esa formacion. Choco estan intimamente asociados a la fauna
Las pluviselvas tropicalcs siempreverdes Centroamericana, mientras que la fauna de
de America del Sur han sufrido procesos de ex- reptiles del Atlantico muestra menor ende-
pansion y contraccion durante los periodos mismo, menor numero de especies, y una
glaciales e interglaciales, del Pleistoceno, y a mayor afinidad con la de la pluviselva Ama-
menudo, se han conectado entre si mediante zonica.
corredores selvaticos de variables dimen- Mientras que la teoria de refugios selva-
siones, permitiendo asi repetidas migraciones ticos durante maximas glaciales puede ex-
de reptiles amazonicos hacia las selvas de la plicar la ocurrencia de endemismos en varios
costa Atlantica del Rrasil y en menor grado segmentos de grandes porciones selvaticas, el
hacia las selvas del Choco en Colombia. Las alto numero de formas minadoras o semi-
migraciones en sentido opuesto tambien pare- minadoras endemicas puede haber evolucion-
cen haber ocurrido, pero en menor intensidad. ado a traves de cambios edaficos y climaticos
Las selvas del Atlantico y del Amazonas han independientes de la formacion de refugios.
estado en contacto con mas frecuencia, mien- Ciertamente, lo inverso parece ocurrir hoy dia
tras que la orogenesis andina ha obstaculizado en el caso de la fauna relictual de savanas
y restringido la formacion de corredores con- aisladas dentro de la pluviselva. La ausencia
tinuos que favorecieran migraciones entre las de refugios acuaticos de importancia durante
selvas del Choco y el Amazonas. las maximas glaciales puede ser la razon de
Antes de que la orogenesis andina ocu- la existencia de pocas especies endemicas de
siderables de 2000 y mas metros. Beebe, W. 1944b. Field notes on the lizards of Kar-
tabo, British Guiana, and Caripito, Venezuela.
Las estrategias de vida de los reptiles Part 2. Iguanidae. Ibid. 29:195-216.
modernos de la pluviselva tropical son pri- Beebe, W. 1944c. Field notes on the snakes of Kar-
mariamente terrestres (43%), seguidos por los tabo, British Guiana, and Caripito, Venezuela.
arboricolas (26%), semi-minadores y mina- Ibid. 31:11-52.
dores (21%) y acuaticos (10%). Muchas de Beebe, W. 1945. Field notes on the lizards of Kar-
tabo, British Guiana, and Caripito, Venezuela.
las formas terrestres ocupan zonas selvaticas
Part 3. Teiidae, Amphisbaenidae and Scincidae.
abiertas, semi-abiertas o modificadas. Algunas Ibid. 30:7-32.
de las ultimas son formas relictuales de las Berry, E. W. 1938. Tertiary flora from the Bio
formaciones de savanas de los periodos gla- Argentina.
Pichileufu, Geol. Soc. Amer. Spec.
ciales. Algunas son antropicas Pap. (12): 1-149.
y el resto esta
formada por generalistas euritipicos. Solo Burt, W. H. 1958. The history and affinities of the
recent land mammals of western North America,
unas pocas especies y generos provenientes pp. 131-154 in Hubbs, C. L. (ed.). Zoogeog-
de los Andes y que han evolucionado aparen- raphy. Publ. 51, Amer. Assoc. Adv. Sci., Wash-
temente a lo largo del pie de monte han sido ington, D.C. 509 p.
Crump, M. L. 1971.
Quantitative analysis of the
capaces de invadir con exito las selvas de
ecological distribution of a tropical herpetofauna.
tierra baja. De alii que muchas de las
espe- Univ. Kansas Mus. Nat. Hist. Occas. Pap. (3):
cies de reptiles que estan asociadas fisica- 1-62.
mente a situaciones selvaticas (46%) no son Damuth, J. E., Fairbridge, B. W. 1970. Equatorial
asociados ecologicos. Atlantic deep-sea arkosic sands and ice-age aridity
in tropical South America. Geol. Soc. Amer. Bull.
Como puede esperarse los reptiles del 81:189-206.
236 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Diamond, J. M. 1973. Distributional ecology of Hershkovitz, P. 1966. Mice, land bridges and Latin
New Guinea birds. Science 179:767. American faunal interchange, pp. 725-751 in
Dixon, 1973. A systematic review of the teiid
J. R. Wenzel, R. L., Tipton, V. J. (eds. ). Ectopara-
lizards,genus Bachia, with remarks on Hctero- sites of Panama. Field Museum of Natural His-
dactylus and Anotosaura. Univ. Kansas Mus. Nat. tory, Chicago, 861 p.
Hist. Misc. Publ. (57): 1-47. Hueck, K., Seibert, P. 1972. Vegetationskarte von
Dixon, J. R. 1974a. Systematic review of the lizard Sudamerika. Gustav. Fischer Verlag, Stuttgart,
genus Anotosaura (Teiidae). Herpetologica 30: 71 p.
13-18. James, D. E. 1971. Plate tectonic model for the evo-
Dixon, R. 1974b. Systematic review of the micro-
J. lution of the central Andes. Geol. Soc. Amer.
teiid genus Iphisa. Ibid. 30:133-139. Bull. 82:3325-3346.
Dixon, J. R., Soini, P. 1975. The reptiles of the up- Jenks, W. F. 1956. Peru. Geol. Soc. Amer. Mem.
per Amazon Basin, Iquitos Region, Peru. I. Liz- (65):213-247.
ards and amphisbaenians. Milwaukee Publ. Mus. Kluge, A. G. 1969. The evolution and geographical
Contrib. Biol. Geol. (4): 1-58. origin of the new world Hemidactylus mabouia-
Dixon, J. R., Soini, P. 1977. The reptiles of the up- brookii complex (Gekkonidae, Sauria). Misc.
per Amazon Basin, Iquitos Region, Peru. II. Publ. Mus. Zool. Univ. Michigan (138): 1-78.
Crocodilians, turtles and snakes. Ibid. (12):1-91. Lancston, W., Jr. 1965. Fossil crocodilians from
Dorf, E. 1959. Climatic changes of the past and Colombia and the Cenozoic history of the Croco-
present. Contrib. Mus. Paleontol. Univ. Michigan dilia in South America. Univ. Calif. Publ. Geol.
(13): 181-210. Sci. (52): 1-157.
Duellman, W. E. 1958. A monographic study of Lohmann, H. H. 1970. Outline of tectonic history
the colubrid snake genus Leptodeira. Bull. Amer. of Bolivian Andes. Bull. Amer. Assoc. Pet. Geol.
Mus. Nat. Hist. 114:1-152. 54:735-757.
Duellman, W. E. 1965. A biogeographic account MacArthur, R.
H., MacArthur, J. W., Preer, J.
of the herpetofauna of Michoacan, Mexico. Univ. 1962. On species diversity, II. Prediction of bird
Kansas Mus. Nat. Hist. Misc. Publ. 15:627-709. census from habitat measurements. Amer. Nat.
Duellman, W. E. 1978. The biology of an equa- 96:167-174.
herpetofauna in Amazonian Ecuador. Univ.
torial Medina, C. J. 1976. Crocodilians from Late Tertiary
Kansas Mus. Nat. Hist. Misc. Publ. (65): 1-352. of northwestern Venezuela: Melanosuclius fisheri,
Emiliani, C. 1956. Oligocene and Miocene tempera- Sp. Nov. Breviora (438): 1-14.
tures of the equatorial and subtropic Atlantic Muller, P. 1973. The dispersal centres of terres-
trial vertebrates in the neotropical realm. Bio-
Ocean.Geology 64:281-288.
geographica 2, W. Junk, The Hague, 244 p.
Emiliani, C. 1972. Quaternary paleotemperatures
and the duration of the high temperature inter- Myers, C. M. The
systematics of Rhadinaca
1974.
Roze, J. A. 1967. A
check list of the new world Vanzolini, P. E. 1967. Problems and programs in
venomous coral snakes
( Elapidae ) with descrip- ,
Amazonian zoology. Atas do Simposio sobre a
tions of new forms. Anier. Mus. Novit. (2287): Biota Amazonica. 5 (Zool.) 85-95.
:
APPENDIX
Appendix 9:1. —
Distribution of reptilian genera and numbers of species associated with tropical rainforests
of South America. An asterisk indicates the occurrence of a savanna relict within the forest area. A question
mark indicates possible occurrence but distributional records are too general to establish exact habitat. The
Amazonian forest reptilian fauna is divided into distributional tracts —
western Amazon (western third), Eastern
Amazon (eastern third), Middle Amazon (middle third), and Widespread Amazon (those species that range
throughout the Amazonian forest). The number in parentheses following the generic name is the number
of species associated with rainforests, although some may not be ecological associates.
Melanosuchus 1 —
Paleosuchus (2)
( )
—
Turtles
Chelydra (1)
Chains (1)
Geochelone (2)
Hydromcdusa (2) ...
Kinosternan (3)
Phnjnops (4)
Platemys (3)
Pseudemys (1)
Podocnemis (5) —
Rhinoclemys (2)
Amphisbaenians
Amphisbaena (30) .
Aulura ( 1)
Bronia ( 1 )
Leposternon (5)
Mesobaena ( 1 )
Lizards
Alopoglossus (6) 4
Amapasaurus ( 1 )
—
Ameiva (5)
Anadia 2 ) (
Anisolepis ( 1 )
Anolis (33-46) ..
Arthrosaura (2)
Bachia (13)
Basiliscus ( 3 )
Cnemidophorus ( 1 )
Corytlwphanes ( 1 ) .
Cercosaura (1)
Colobodactylus (4)
Colobosaura ( 1 )
Crocodihirus ( 1 )
Diploglossus (4)
Dracaena ( 1 )
Ecpleopus ( 1 )
Enyalius (8) - _.
Emjalioides (7)
Echinosaura ( 1 )
Euspondylus (2)
Gymnodactylus ( 1 ) .
Conatodes (7)
Hemidactylus (2) ...
Hetewdactylus ( 1 )
Iguana ( 1 )
Iphisa (
1 )
Kenlropyx (6)
1979 DIXON: REPTILES OF LOWLAND TROPICAL FORESTS 239
Lepidoblepharis (7) „
Leposoma (8)
Liolaemus ( 1 ) _
Mahwja ( 7 )
Morunasaurus ( 1 )
Neusticurus (5)
Ophiodes (1)
Oph ryoessoides ( 4) .
Pantodactylus (2)
Plica (2)
Placosoma (2)
Pohjchrus (4) 1
Prionodactylus (4) __. 3
Pseudogonatodes (2)
Piijchoglossus ( 4 ) 2
Sphaerodactylus (2) .
Sfoiocercu.s ( 2 ) 2
Stcnolepis ( 1 )
Strobilurus ( 1 )
Thecadactylus ( 1 ) ...
Trctioscincus(2)
Tropidurus (4)
Tupinambis (2)
Uraccntron (3)
Uranoscodon ( 1 )
Urostrophus ( 1 )
Snakes
Anilius ( 1 )
Anomalepis ( 1 )
Apostolepis ( 5 )
Atractus (25) 10
Boa (1) _
Bothrops (21) 5
Chironius (12) 1
Cfc/ia (3) 1
Corallus (3)
Crotalus (1) .
Dendrophidion (2) .
Di;>mk (18)
Diaphorolepis ( 2 ) ...
Drepanoides ( 1 ) __
Drymarchon ( 1 )
Drymohius ( 2 )
Drymoluber ( 1 )
Epicratcs ( 1 )
Elapomorphus (3) .
Emmochliophis (1) .
Enulius (2)
Erythrolamprus (4)
Eunectes (3)
Geophis ( 1 )
Helicops (11)
Helminthophis ( 1 ) .
Hydrodynastes ( 1 ) .
Hydrops ( 2 )
Imantodes (3)
Lachesis ( 1 )
Lampropeltis ( 1 ) _.
Leimadophis (8)
Leptodeira ( 2)
240 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Leptophis (4) 2 .. 1 1 3 1
Leptotyphlops (12) _ 5 7 2 3 2
Liotyphlops (5) 1 4
Liophis (9) - 3 2 3
Lygophis (2) 1
Lystrophis (
1 )
1
Mastigodryas ( 3 )
2
Micrurus (23) 3
Ninia (2)
Nothopsis ( 1 )
Oxybelis ( 4 )
Oxyrhopus (5)
Philodryas (
4)
Phimophis (1)
Pliocercus (
1 )
Pseudoboa (3)
Pseudoeryx ( 1 )
Pseustes (4) 1
Ptychophis ( 1 )
1?
Rhadinaea (10) 5
Rhinobothyrum (2) _
Saplienophis ( 1 )
Scaphiodontophis ( 1 )
Sibon (1)
Sibynomorphus ( 1 ) _
Siphlophis (4)
Sordellina ( 1 )
Spilotes (1)
Stenorrliina ( 1 )
Sy nophis (2)
Tantilla (4)
Thamnodynastes (3)
Trachyboa (1)
Trctanorhinus ( 1 ) ....
Tripanurgos ( 1 )
Tropidodryas (2)
Tropidophis (3)
Typhlophis ( 1 )
Typhlops ( 4 )
Ungaliophis ( 1 )
Xenoboa 1 ( )
Xenodon (6)
Xenopholis ( 1 )
10. The Herpetofauna of the Guianan Region
Marinus S. Hoogmoed
Rijksmuscum van Natuurlijkc Historic
Postbus 9517
2300 RA Leiden, The Netherlands
Although this paper deals with a highland Guyana, several of which even at the present
fauna, it not limited to the reptiles and
is day remain unvisited. The most renowned of
amphibians that occur at elevations of more these tepuis is Roraima with an altitude of
than 1000 m. One of the main reasons for 2810 m, discovered in 1838 by Robert Schom-
this is that "highlands" above 1000 in the m burgk and climbed for the first time in 1884
Guiana area are few and occupy only a very by Im Thurn and Perkins.
small part of the total area of the Guiana The first zoological collection ever made
Shield. Another reason is that our knowledge near any tepui was assembled there in
of the herpetofauna at higher elevations in 1842 by Richard Schomburgk. Other collec-
Guiana is still very fragmentary. These facts tionswere made at the foot in the 1880's.
prompted me to deal with the herpetofauna The first herpetological specimens from the
of the entire Guiana Shield. summit were secured by Quelch and McCon-
Thecoast of the Guianas was discovered nell in 1894. In 1898 they made a second
in 1499 by Alonso de Ojeda and Amerigo expedition to the summit plateau. The ma-
Vespucci. After the discovery of the so-called terial of these expeditions contained several
"Spanish Main" or "Wild Coast," numerous new species. Boulenger (1895, 1900) studied
expeditions tried to explore the interior in them and described the OreophryneUa
frogs
search of the fabulous El Dorado. Most quelchii, O. macconneUi, Otophryne robusta
famous of these adventurers was Sir Walter and Hylodes marmoratus, and the lizards
Raleigh, who undertook several expeditions Neusticurus rndis and EuspondyJas leucostic-
into the interior of Guiana. Zoologically, tus. Of these species, only the first and the
these expeditions were of no importance last came from the summit of the mountain;
whatsoever. From about the beginning of the other species were collected at the base.
the 18th Century, zoological specimens, main- The next zoological expedition, on a much
ly from Surinam, started to reach Europe, larger scale and under the auspices of the
and an important percentage of the species American Museum of Natural History, visited
of reptiles and amphibians described by Lin- Roraima in 1927-28, spending two weeks on
naeus in 1758 originated from the Guianas. the summit and about two months at the
One of the first scientific explorers of the base (Tate, 1928, 1930a,b, 1932, 1939; Chap-
interior was Von Humboldt, who in 1801 man, 1931). Among the material collected
visited the Rio Orinoco and the Cassiquiare were several reptiles and amphibians. In
Canal (Gleason, 1931). In 1835 Sir Robert 1971, 1973 and 1974 Roraima was visited
Schomburgk started his explorations in Guy- again, this time by parties with herpetologists
ana and adjacent countries in order to settle as members (Warren, 1973). Their collec-
the frontiers. During these explorations, zoo- tions contained many novelties.
logical collections were made that supplied In 1928 Mount Duida, at the western
a wealth of new data. Since Schomburgk's end was explored zoo-
of the series of tepuis,
travels, an increasing number of scientific logically. The expedition was the first that
expeditions penetrated into the interior of succeeded in climbing the mountain and
lowland Guiana and would lead us too far
it
spent three months at the summit. Among
astray to try to deal with them here in any the herpetological material collected were the
detail. I make an exception for the expedi- types of the teiid lizards Pantodactylus tyleri
tions exploring the tepuis in Venezuela and and Arthrosaura tatei, of the hylid frog Stef-
241
242 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
ania goini, and of the leptodactylid frog with the biogeographical and geographical
Elosia duidensis. In 19.37-38 Auyantepui was data (Fig. 10:1). The Serra Acarai and the
zoologically explored. From 1938 until the Tumuc Humac Mountains, forming the divide
present, considerable exploration took place between the French authors' Guiana and
in the tepui region, mostly ornithological and Amazonia apparently do form a geographical
botanical, but as a by-product many herpe- barrier for a number of endemic species
tological specimens were collected, some of (mainly frogs), but this is too small a pro-
which at the time being new to science were portion of the entire fauna to justify the
described by Roze
( 1958a,b; Auyantepui, definition of the Guiana area as they do. Far
Chimantatepui), Rivero (1961, 1965, 1966, more species are spread on both sides of
1967a,b, 1968a-d, 1970, 1971; Duida, Mara- the divide and occur both in the Orinoco and
huaca, Chimantatepui, La Escalera region Amazon Basin (Haffer, 1974; Miiller, 1973).
and other tepuis) and Lancini (1968; Cerro In a discussion of the Guiana herpetofauna
Jaua). The most recent biological explora- I think it is better to take into consideration
tion of some large tepuis were the expeditions biogeographical data of the majority of the
to Cerro Jaua and Sarisarinama in 1974 and (herpeto) fauna being studied, rather than
to Cerro Yapacana in 1978. These were some rely only on those of some endemic frogs.
of the rare expeditions in which herpetologists In that way the biogeographical definition of
participated (Nott, 1975; Orejas-Miranda and Guiana, as accepted here, agrees closely with
Quesada, 1976). The herpetological results the geographical, geological and climatolog-
of these expeditions have not yet been pub- ical data. However, there are good grounds
lished. for considering the Guiana of the French
authors as a subregion of the Guiana as here
defined.
DELIMITATION AND DESCRIPTION Geologically, this area is a unit known as
OF GUIANA the Guiana Shield (Gansser, 1954; Fittkau,
1974), of which small parts are situated west
Guiana is the area bordered by the Rio of the area as here delimited (Fig. 10:2).
Orinoco, the Cassiquiare Canal (connecting Along the edges, notably in the north, the
the Orinoco and Amazon drainages), and the east and the south, there are belts of alluvial
Rio Negro in the west, by the Rio Amazonas deposits; however, the core is made up of
in the south and by the Atlantic Ocean in the pre-Cambrian metamorphic and
igneous
north and the east. The area comprises three rocks. Together with the Brasilian Shield, it
political units in their entirety, namely, Guy- can be considered as part of the geological
ana, Surinam and French Guiana. Of Vene- foundation of South America. Since Paleozoic
zuela it comprises the Estado Bolivar and the times these shields have not been submerged.
Territorio Federal Amazonas, known under During the Mesozoic both shields were con-
the common denomer Guayana. Of Brasil nected, for the Amazon was not yet present.
it comprises the Territorio do Amapa, the During the Late Cretaceous the area was
Territorio de Roraima and those parts of slightly uplifted and the first signs of the
the states of Para and Amazonas that are present Amazon Basin became visible. In
situated north of the Rio Amazonas and Rio the Tertiary there was a further uplift ( Haffer,
Negro. Recently Lescure ( 1977 ) and Des- 1974). The higher, central parts of the Guiana
camps et
al. (1978) defined Guiana as the Shield are covered with sandstone remnants
area bordered in the west by the Rio Barama of the Roraima Formation. Deposition of this
(Venezuela) and in the southeast by the sandstone took place in Proterozoic time,
Rio Araguari (Brasil). The southern border 1600-1800 m.y.b.p., as stream and delta de-
would be formed by the watershed between posits laid down in continental to epiconti-
riversemptying directly into the Atlantic nental environments (Priem et al., 1973).
Ocean and rivers belonging to the Amazonian After uplift, this formation covered the Gui-
drainage. In my opinion, this definition of ana Shield as an extensive sandstone plateau
Guiana is artificial and not in accordance or tableland, on which the early Guianan flora
1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 243
Fig. 10:1. Map of Guiana, showing the borders of the area as here defined (heavy broken line) and as
defined by Descamps et al. (1978) and by Lescure (1977) (heavy dotted line). Presumed forest refugia
are gray and indicated by numbers: 1 = Imeri Refuge, 2 = Guiana Refuge, 3 = Imataca Refuge, 4 = Tepui
Refuges. The line of fine dots (in this and the follow ing maps of Guiana) represents the 200 m contour line.
Mapa de la Guayana, mostrando los limites del terri torio definido aqui (linea cntrecortada gruesa) y el de-
finido por Descamps et al. (1978) y por Lescure (1977) (linea gruesa punteada). Supuestos refugios forestales
en gris,indicados con numeros: 1 = Refugio de Imeri, 2 = Refugio de Guayana, 3 = Refugio de Imataca,
4 = Refugios de Tepuyes. La linea punteada f\na (en este y los siguientes mapas) representa la linea de con-
torno de 200 m.
developed (Maguire, 1970). During the Cre- of Brasil and Guyana, with many isolated
taceous and Tertiary uplift of the area, ero- remnants in the western part of the Estado
sion shaped the present-day table mountains Bolivar and in the Territorio Federal Ama-
or tepuis. These mountains consist of layered, zonas and two outlying remnants in eastern
unfossiliferous, pink sandstones, with dolerite Guyana and in central Surinam (Bisschops,
dikes and sills, reaching a maximum thickness 1969; Priem et al., 1973) (Fig. 10:2). Some
of about 2400 m in Auyantepui in southeastern geologists (e.g., Priem et al., 1973:1677) are
Venezuela and decreasing to 700 m in the of the opinion that "It is impossible to decide
Tafelberg in central Surinam (Haffer, 1974). whether occurrences represent erosional rem-
At present, the Roraima Formation covers an nants of a once-continuous cover or sediments
area of about 450,000 km 2 and is spread over deposited in a number of isolated basins."
a total area of 1,200,000 km 2 in Venezuela, However, most geologists and biologists re-
Guyana, Brasil and Surinam (Priem et al., gard the present-day sandstone mountains as
1973). The
greater part of the Roraima sand- remnants of a once-continuous sandstone
stone concentrated in the Gran Sabana
is cover. Also, one could imagine a combination
region of Venezuela and the adjacent parts of the possibilities, in which the western
244 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fig. 10:2. Map of Guiana showing the extent of the Guiana Shield (gray area) and of the Roraima
sandstone formation (black). The white areas in the sandstone represent gabbro (after Bisschops, 1969;
Gansser, 1954; Priem et al., 1973).
Mapa de la Guayana mostrando la extension del Escudo Guayani (gris) y de la formation Roraima
(negro). Las areas blancas en el gres de Roraima representan gabbro (segiin Bisschops, 1969; Gansser, 1954;
Priem et al, 1973).
Roraima Formation once formed a continu- ing in conjunction with the Andean uplift
ous cover and the two outlying areas in De- toward the end of the Tertiary and at the
merara and Surinam could have been de- beginning of the Pleistocene" (Haffer, 1974).
posited in isolated basins. However, decisions The Guiana Shield consists of an elevated
on this subject should be reached by geolo- portion in the west, rising from sea level to
gists, although perhaps biologists may con- well over 1000 m
in relatively extensive areas.
tribute to the solution. In this paper I adhere This portion bears the sandstone tepuis of
to the view that the tepuis are remnants of which the highest attain a height of 2810 m
a once-continuous formation. It seems useful (Mount Roraima) and 3014 m (Serra de
to state that the arch of sandstone tepuis Neblina). These tepuis are mostly flattopped,
in southern Colombia, west of the Rio Orin- with perpendicular cliffs several hundred
oco, and ending quite close to the Andes in meters high separating the plateau summits
the Sierra de Macarena, is not of the same from the talus formed by the accumulation of
as the Roraima Formation (Lescure, 1977). erosional products at the base of the cliffs
This sandstone is much younger and probably (Figs. 10:3-4).
represents a deposition of erosional products The western part is separated from an
of the Roraima Formation (Haffer, 1974; eastern elevated part by a depression formed
Paba-Silva and Van der Hammen, 1960). by the river systems of the Rio Branco and
These tepuis probably arose by "Block fault- the Essequibo River, which may be connected
1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 245
Fig. 10:3. The Tafelberg in Surinam, the easternmost remnant of the Roraima formation. Note the flat top,
the steep, bare upper reaches of the flanks and the sloping talus covered with forest.
La Montana Tafelberg en Surinam, el residuo mas oriental de la formaeion Roraima. Observase la cumbre
aplanada, las flancos escarpados y rasos en su parte superior y el tabid inclinado y cubierto de selva.
Fig. 10:4. View of several tepuis south of El Manteco, Estado Bolivar, Venezuela.
Vista de algunos tepuyes al sur de El Manteco, Estado Bolivar, Venezuela.
246 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fig. 10:5. Map of Guiana with the contour lines of 200, 1000 and 1500 m (after Mayr and Phelps, 1967).
Mapa de la Guayana con lineas de contorno de 200, 1000 y 1500 m (segi'm Mayr y Phelps, 1967).
during the rainy season when large areas are important role. In the western and north-
inundated. The eastern part is much lower western portion of the shield there are sa-
than the western one, reaching a maximum vannas more or less continuous with the llanos
height of 1280 m
(Julianatop) in central of central Venezuela. In the three Guianas
Surinam. From the divide between the Ama- there is a band of coastal savannas on white
zon Basin and rivers flowing north to the sand, reaching from Georgetown in the west
Atlantic Ocean (nowhere over 1000 m high), to Cayenne in the east. East of Cayenne and
the country gradually slopes down to sea in Amapa the white sand is absent and some
level. Thus, although the topography of east- extensive swamps in that region are dry
ern Guiana may be rather rugged, with many savannas in the dry season. In Amapa this
mountain ranges and valleys separating them, coastal belt is bordered on the west by a belt
the area hardly ever exceeds 1000 m. In of cerradrj— savanna with isolated trees. Iso-
western Guiana, the topography is more or lated, extensive savanna complexes of the cer-
less the same, but on a higher level, with the rado type are present (Hills, 1969) in south-
consequence that more extensive areas are western Guyana (Rupununi), in southeastern
over 1000 m. However, superimposed on the Venezuela (Gran Sabana) and on the border
Guiana Shield in this region are sandstone between Surinam and Brasil (Sipaliwini/Paru
tepuis that may reach elevations of almost savannas). Smaller, isolated savannas occur
3000 m (Fig. 10:5). The Guiana Highlands in Surinam and in Venezuela both on the
are also known as Pantepui. Roraima sandstone and on other substrates
The greater part of the areacovered by
is (Fig. 10:6). On the higher points, starting at
tropical rainforest, but savannas also play an about 800 to 1000 m, cloud forest occurs with
1979 HOOGMOED: HERPETOFAUNA OF GUI AN AN REGION 247
Fig. 10:6. Map of Guiana showing the distribution of forest and savannas. Forested areas white, savannas
gray, inundated savannas hatched. The zone with lower rainfall (cf. Fig. 10:7) has been indicated with heavy
broken lines (after Hills, 1969; Muller, 1973; Oldenburger et al., 1973; Prance, 1973; Romariz, 1974 and
personal field data).
Mapa de la Guayana mostrando la distribution de selva y sabana. Selva en bianco, cerrado en gris, campo
rayado. La zona menos lluviosa (cf. Fig. 10:7) se ha indicado con una linea cntrecortada gruesa (segun Hills,
1969; Muller, 1973; Oldenburger et al., 1973; Prance, 1973; Romariz, 1974 y observacioncs personales).
thick layers of mosses covering the trees, is characterized by two dry and two rainy
shrubs and the ground. This is especially so seasons per year. Their duration and the
on the talus of many of the tepuis. The pla- period of the year in which they fall are some-
teau summits of the smaller tepuis have only what variable, and at higher elevations the
a shallow layer of soil, which is insufficient to distinction between dry and rainy seasons
support forest; thus, the vegetation is low, may be hardly evident, but in general this
often savannalike. The plateau summits of the division holds true for the greater part of the
larger (more extensive) tepuis is more diver- area. Within the area a wide zone with dis-
sified,and in some places a sufficiently deep tinctly lower rainfall extends northwest-south-
layer of soil has accumulated to support mod- east connecting the llanos of Venezuela
erately high forest; however, in other places with the caatinga and cerrado region of cen-
there is only sparse vegetation (Chapman, tral Brasil (Figs. 10:6-7). Within this zone,
1931; Gleason, 1931; Maguire, 1945, 1955, which roughly covers the extreme southwest-
1970; Mayr and Phelps, 1967; Tate, 1928, ern part of Surinam, southern Guyana, south-
1930a,b, 1932, 1938a,b, 1939; Tate and Hitch- eastern Venezuela and the Guianan part of
cock, 1930). Para, the annual rainfall is 2000 mm
or less.
The climate of the region under discussion To the northeast and to the southwest the
248 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
70 60 50
:
:
S^™s:SS^S '00 0- (1000 mountains above 2000
: 15 00
Rainfall (mm) distribution in northern South America (after Prance, 1973; Reinke, 1962).
Fig. 10:7.
Distribution dc la lluvia (mm) que cae anualmente en la parte norte de la America del Sur (segun Prance,
1973; Reinke, 1962).
annual rainfall increases, reaching maxima of restricted to refuges, mostly along its periph-
over 3000 mm
in northeastern French Guiana ery (Brown et al., 1974; Haffer, 1969, 1974,
and coastal Amapa and of some 2500 to 3000 this volume; Vanzolini, 1970a). Inversely,
mm in the upper Orinoco region (Reinke, during the wet climatic phases the rainforest
1962; Prance, 1973). Mean annual tempera- spread again from the refugia and the sa-
tures are between 24° and 27°C in the low- vanna vegetation and fauna retreated into
lands and decrease with increasing altitude. refuges. Of importance in this connection are
During the last few years it has become the Guiana, Tepui and Imeri forest refuges
increasingly clear that Pleistocene and Holo- of Haffer (1969, 1974); the Guiana (forest),
cene climatic changes had a profound influ- Pantepui (montane forest) and Roraima (sa-
ence on the vegetation of northern South vanna) centers of Midler (1973); and the
America, especially Amazonia and adjacent
in Guiana, Imataca and Imeri refuges of Prance
regions. It is presumed that during dry cli- (
1973 ) all of which are situated within the
,
matic phases the rainforest disappeared from limits of the area considered here (Fig. 10:1).
large stretches of the Amazon Basin and was The aforementioned belt with a lower pre-
1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 249
cipitation played an important role in the mentary knowledge of these groups. Thor-
distribution of plants and animals during the ough herpetological exploration of the tepui
different climatic periods. At present, the region, starting with the now easily accessible
savanna complexes of interior Guiana are situ- La Escalera region in eastern
Venezuela,
ated in this belt (Fig. 10:6). However, during probably will provide us with many interest-
dry climatic phases probably much larger ing finds. Because our present knowledge is
areas of it were covered with savanna, thereby so fragmentary, it is often difficult to decide
providing a dispersal route for savanna in- whether a certain species is really restricted
habitants, either to the north or to the south, to one tepui or not. The available data permit
and at the same time forming a barrier to some zoogeographical conclusions, but those
east-west dispersal of forest inhabitants. On regarding the so-called endemics certainly
the other hand, both areas with higher rain- have to be drawn with much reserve.
fall, adjacent to this dry belt, are thought to Presently a total of 408 species of reptiles
be the areas where forest refuges were situ- and amphibians is known to occur in the
—
ated during arid phases to the northeast the Guiana region (Table 10:1, Appendix 10:1).
Guiana refuge, to the southwest the Imeri Seventy-six species are represented by 108
refuge. During wet climatic phases, the forest subspecies, which raises the number
of spe-
spread from these refuges and invaded the cies-group taxa for the region to 440. The
savanna belt, fragmenting it into several iso- herpetofauna of Guiana can be allocated to
lated savanna complexes, as is the case today
eight groups, which in turn can be partly
(Fig. 10:6). The montane forests covering subdivided.
the slopes of the tepuis in southern Venezuela 1. Endemic Guiana region:
in
can be regarded as isolated occurrences of A. Highland (over 1000 m)— 18 am-
rainforest on places with favorable climatic
phibians, 9 reptiles.
conditions (high elevation, high rainfall) gen- B. Lowland (below 1000 m) —74 am-
erally having unfavorable climatic, and pos- phibians, 50 reptiles.
sibly edaphic, conditions (Gran Sabana area). 2. Amazonian:
These forests,which are different from the A. Periferal along western and northern
tropical lowland rainforests, probably were —
margin of basin 10 amphibians, 19
only connected with the lowland forests dur- reptiles.
ing very wet climatic phases. Although Mid- B. With disjunct populations in upper
ler's concept of the Guiana center is much
Amazonia and near the mouth of the
wider ( and based on several different groups ) Amazon —2 amphibians, 1 reptile.
than Haffer's, Prance's and others' Guiana C. Species of Amazon Basin occurring
refuge, I think we can synonymize the two on southern edge of Guiana and
without problems; the same is true for the
along eastern margin, where they
Pantepui center and the Tepui refuge. There may reach French Guiana 3 am- —
is no parallel in Midler's concepts of Haffer's
phibians, 11 reptiles.
and Prance's Imeri refuge. The Imataca ref- D. Widespread Amazonian, occurring
uge, which was postulated by Prance ( 1973 ) throughout greater part of Guiana
—
for plants only substantiated further by
is
39 amphibians, 62 reptiles.
data from butterflies (Brown et al., 1974)
3. Widespread species (distribution ex-
(Fig. 10:8).
tending from Mexico or Central Amer-
ica over entire cis-Andean tropical
HERPETOFAUNA South America): 12 amphibians, 35
reptiles.
Although since 1894 quite a substantial 4. Species reaching their eastern distribu-
number of reptiles and amphibians has been tion limit on the Guiana Shield, from
collected from the sandstone tepuis, only a Central America, northwestern South
small part of it was collected by herpetolo- America or upper Amazonia: 11 am-
gists. This partly explains our scant and frag- phibians, 17 reptiles.
250 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
haffer,i969 ,
1974 vanzolini,1970
Fig. 10:8. Location of forest refugia during arid periods according to several authors; numbers as in
Fig. 10:1.
Situation de los refugios de selva durante los periodos secos segtin algunos autores; los numeros como
en la Fig. 10:1.
reaching Surinam or even Venezuela: group, it has not been established beyond
8 amphibians, 13 reptiles. doubt that they occur in the Guiana region.
6. Cosmopolitan species: amphibians, 6 Thus, there remain five important groups,
reptiles. totaling 177 amphibians and 217 reptiles, that
7. Species imported from the Caribbean reflect the complicated history of the Guiana
region: 1 amphibian, 4 reptiles. herpetofauna and that are dealt with in detail
8. Species with limited or uncertain dis- later.
tributions that may occur in the region: Considerable differences exist between the
amphibians, 3 reptiles. percentages of reptiles and amphibians in five
The three groups in the tabulation
last
different groups and subgroups (Fig. 10:9,
above are of no importance in the following Table 10:2). These groups are highland (1A)
considerations. The five cosmopolitan species and lowland (IB) endemics, disjunct (2B)
of sea turtles and one species of cosmopolitan and widespread ( 2D ) Amazonian, and wide-
1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 251
—
Table 10:2. Composition of the Guianan Herpetofauna.
The columns contain the totalnumber of species having a certain distribution and the percentage this group
forms of the total number of species in the Guianan Region.
1979 HOOGMOED: HERPETOFAUNA OF GUI AN AN REGION 253
cies have restricted distributions, usually con- by assuming that the genus Stefania arose
sisting only of the summit or talus slopes of from hylid stock in the Guiana Highlands,
one or a few adjacent tepuis (Figs. 10:10-11). probably prior to the Oligocene. Initially, the
As stated before, this either reflects our frag- stock split into two groups, which during the
mentary knowledge of the herpetofauna of most recent uplift of the area in Mio-Pliocene
the Guiana Highlands, or these distributions times became isolated on several tepuis and
are real and the comparable habitat on other since differentiated into the several species
tepuis is occupied by a related species. How- now composing the two species groups. The
ever, this has only been documented (and occurrence of Stefania evansi in lowland areas
poorly so) for the endemic frog genus may be regarded, as in Otophryne, as being
Stefania. secondarily, induced by the Pleistocene cli-
The bufonid genus Oreophrijnella from matic changes, which lowered the general
Roraima and Auyantepui is considered to be temperature of the area by about 3°C (Van
a specialized derivative from the general ate- der Hammen, 1974).
lopodid stock and to have evolved in isola- Species showing a slight degree of Andean
tion since the Early Tertiary or the Cretaceous relationships aremembers of the frog genera
(McDiarmid, 1971). The same is true for the Centrolenella and Eleutherodactylus, and of
microhylid frog Otophryne (not an altitudinal the colubrid snake genus Atractus; all three
endemic), composed only by O. robusta with genera probably evolved in or near the Andes,
—
two subspecies one restricted to high eleva- either in the foothills or in the lowlands, and
tions on Chimantatepui, the other occurring subsequently spread to the east. However,
in the greater part of interior Guiana at eleva- the endemic altitudinal species belonging to
tions of 200 to 1666 m. Like Oreophrijnella, these genera have no direct relations with
Otophryne also shows a combination of prim- Andean species and probably are altitudinal
itive, derived and unique characters. This is forms derived from lowland species. The
most easily explained by assuming that these matter is slightly different for the species
frogs were subject to a long evolution in of Euspondylus, a genus of teiid lizards of
isolation on the sandstone formation, prob- Andean origin, members of which live at
ably since the Cretaceous or Early Tertiary; medium to high altitudes in the Andes from
the invasion of tropical lowland Guiana by Peru to Venezuela; two species reached the
Otophryne may be considered as secondary. higher of the Guiana Shield, pos-
altitudes
According to Lynch (pers. coram.), the lepto- sibly during a time of Pleistocene climatic
dactylid frog Hylodes duidensis belongs to an depression. The altitudinal endemics of the
undescribed genus of the tribe Eleutherodac- tree frog genus Hyla all apparently are re-
Its relations are not clear, but it
tylini. may lated to lowland species groups.
have developed on the Guiana Shield as a Riolama, a monotypic, endemic teiid lizard
highland derivative of the eleutherodaetyline genus restricted to the summit of Mount
stock. Stefania is an endemic, egg-brooding Roraima, is known only from the type speci-
hylid frog genus clearly related to the north- men. Presumably it is related to Leposoma
ern Andean Cryptohatrachus. According to and its relatives, but its history is not clear.
Rivero (1970), these frogs can be divided It may have evolved from lowland microteiids
into the Stefania goini group, with two spe- by isolation on a sandstone tableland prior
cies, and the Stefania evansi group with five to the Oligocene, aswas probably the case
species (and three undescribed ones). One The colubrid snakes Liophis and
in Stefania.
member of the goini group occurs on Mount Thamnodynastes occur in lowland Amazonia
Duida in the west, the other on Chimantatepui and Guiana, but they seem to be of southern
in the east. One member of the evansi group Brasilian origin and to have evolved into sev-
occurs on Cerro Marahuaca and the others eral altitudinal species in Guiana. The near-
on the eastern part of the Roraima Formation. est relative of the iguanid lizard, Tropidurus
The distribution of the members of these bogerti, is T. torquatus hispidus (R. Ether-
species groups can be explained most easily idge, pers. comm.), a member of a species or
254 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
species complex, which may be of southeast- fauna can be explained with the aid of the
ern Brasilian origin. The few altitudinal en- following theories.
demic subspecies have evolved by isolation
all
1. The Mountain Bridge Theory as pre-
at higher altitudes from lowland relatives of
sented by several authors (Todd and
different origins.
Carriker, 1922; Haas, 1957) apparently
Attempts to explain the origin of the fauna is useless, because there is no geological
of Pantcpui have been based on the distribu- evidence for a connection of southern
tion of birds (Chapman, 1931; Haffer, 1974; Venezuela with the Andes. As has been
Mayr and Phelps, 1967), mammals (Tate, pointed out, the sandstone mountains
1939), frogs (Rivero, 1965) and snails (Haas, (Sierra de Macarena) in southern Co-
1957). Because of different dispersal abilities lombia are not the remnants of such a
and different geological ages of the groups
bridge.
concerned, these studies came to different con- 2. The Plateau Theory, starting from the
clusions.For instance, birds supposedly were that "a more extensive
assumption
able to reach Guiana from the Andes by tableland probably did exist on the
(simply stated) flying from one mountain Guayana shieldduring the Mesozoic
with suitable climate to the next. This pos- and Tertiary, prior to an intensive ero-
sibility doesn't exist for the other groups. The sional dissection" (Haffer, 1974:163) is
distribution of the endemic Guianan herpeto- useful to explain the presence of several
1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 255
orogenic movements that shaped it into merly (Chapman, 1931; Tate, 1939).
a mountain range before erosion graded However, it may have facilitated the
it into a plateau, which in turn was dispersal of certain organisms, because
uplifted and eroded into its present all life zones on mountains shifted to
shape, serves well to explain the distri- lower altitudes, thus creating suitable
bution of the genus Stefania ( Rivero, habitats for subtropical organisms in
1970). places where they were formerly ab-
The Modified Cool Climate Theory de- sent. These still widely-separated, sub-
parts from the assumption that during tropical habitats could have been of
the glacial periods of the Pleistocene, importance for birds. The distribution
the lowlands between the Andes and of amphibians and reptiles apparently
Guiana, and within the Amazonian related to Andean taxa could not have
basin had a cooler climate. This indeed gone only via those "stepping stones"
was true, the temperature of the low- but most likely through the lowlands
lands having been about 3°C lower at times of cooler temperatures.
256 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
level to a maximum of 2400 m. In a number There are five ( monotypic lowland en-
)
bates steyermarki (poison-arrow frog), Hyla pasaurtis is restricted to a small area in east-
ern Guiana and Mesobaena to western Gui-
ginesi, H. benitezi, H. kanaima, H. lemai, H.
ana. The ranges of the first four genera and
sibleszi, Stefania marahuaquensis, S. woodlcyi
(hylid frogs) and Eiiparkerella sp. "A" (lep-
of many endemic species coincide with that
of the postulated Guiana Forest Refuge Haf-
todactylid frog), have narrower elevational
(
cene) dry or wet climatic phases occurring Depression, 32 (43%) only occur east of the
in northern South America.
depression, and 24 (32%) only occur west of
A few of the lowland endemics occurring it. Of the 50 endemic species of lowland rep-
in western Guiana, mainly around the head- tiles (Table 10:2,
Appendix 10:1) IS (36%)
waters of the Rio Orinoco, seem to strengthen occur on both sides of the
Essequibo-Rio
view of an Imeri forest refuge. These
Haffer's Branco Depression, 17 (34%) only occur east
endemics include one monotypic genus of the depression, and 15
(30%) only occur
(Mesobaena, amphisbaenian), eight species west of it. The picture changes
distinctly
[Aparasphenodon venezolanus (tree frog), when the altitudinal endemics also are con-
Dendrobates leucomelas and D. steyermarki sidered. In that case the number of
amphib-
ians restricted to the western
(poison-arrow frogs), Atractus insipidus, Hel- part of Guiana
icops hogei, Liophis canaima (all colubrid becomes 42 and the corresponding number
snakes), PhyUodactylus dixoni of reptiles 24. The
(gekkonid percentages change ac-
lizard), Crocodijlus intermedins (crocodile)] for amphibians
cordingly, respectively 18
and three subspecies [Hydrops triangularis (22%), 32 (39%) and 42 (51%); for reptiles
venezuelensis, Leptophis ahaetulla copei respectively 18 (31%), 17 (29%) and 24 (41%).
(both colubrid snakes), Micrurus surinamen- Among the widespread endemics the propor-
sis nattereri
(elapid snake)]. A similar situa- tion of reptiles is
considerably higher than
tion is known in birds, with one endemic that of amphibians; in both the western and
genus and nine endemic species (Haffer, eastern endemics the proportion of
amphib-
1974). Assuming a similar divergence rate for ians is higher than that of
reptiles, reflecting
the organisms involved, this seems to the greater mobility of reptiles. When
point to only
at least three arid phases the lowland endemics are considered, the
during which the per-
forest fauna was isolated in this Imeri forest centage of amphibian species restricted to the
refuge. east is distinctly higher than that of
species
Different patterns of distribution exist in restricted to the west, in reptiles it is
only
Guiana. The endemic species are not evenly slightly This probably reflects the
higher.
distributed throughout the area. As has been greater importance of the Guiana Refuge for
noted in the section on altitudinal distribu- amphibians, as compared to the importance
of the Imeri Refuge. For reptiles, both
endemics are restricted to
tion, all altitudinal
the western part of the Guiana Shield, the refuges apparently were equally important.
area west of the Essequibo-Rio Rranco De- Why the Guiana Refuge was more important
for amphibians than for reptiles remains a
pression. The ranges of most of the species
that supposedly originated on the matter of conjecture. However, possibly it
higher parts
of the sandstone area do not extend far be- results from the greater dependence of am-
yond; only a few reach the Essequibo River phibians on water and moist habitats. Thus,
in the east. Exceptions, like the isolation in different refuges was more severe
microhylid
for amphibians than for reptiles;
frog, Otophryne robusta, and the teiid lizard, reptiles re-
Neusticurus rudis, extend their ranges beyond stricted to different forest refuges
probably
the The Essequibo-Rio came into contact earlier than the amphibians,
Essequibo River.
Rranco Depression seems to have been a bar- thus diminishing the possibilities of
having
rier to the eastward distribution of a number attained reproductive incompatibility. Maybe
it was
of species, mainly
Pantepui species. On the simply a matter of size, the Guiana
other hand, it was a barrier to the westward Refuge having been larger (and therefore
distribution of a number of species. The ef- possibly harboring more than the
species)
fect of this barrierevident from the ranges
is Imeri Refuge. Perhaps both factors played a
of lowland endemics
(Fig. 10:11). Of the role.
74 endemic species of lowland amphibians The Branco Depression
Essequibo-Rio
(Table 10:2, Appendix 10:1) 18 (24%) occur also served as a route for lowland Amazonian
on both sides of the Essequibo-Rio Branco species invading the northern part of Guiana.
258 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
120 100 80 60 40
Distribution of species belonging to the groups 2-5 (Table 10:2, Appendix 10:1).
Fig. 10:12.
Distribution de especies pertenecientas a los grupos 2-5 (Tabla 10:2, Apendicc 10:1).
1 = Pseudopaludicola pusilla (group 4). 2 = Htjla geographica (group 2d). 3 = Lcpidoblepharis festae
(group 2b). 4 = Crocodilurus lacertinus (group 2c).
120 100 80 60 40
Distribution of species belong to the groups 2-5 (Table 10:2, Appendix 10:1).
Fig. 10:13.
Distribution de espccies pertenecientas a los grupos 2—5 (Tabla 10:2, Apendice 10:1).
1 = Lysapsus limellus (group 5). 2 = Lysapsus limellus laevis (group 5). 3 = Hyla senicula melan-
argyrea (group 5). 4 = Leptodactylus rhodomystax (group 2a). 5 = Phrynohyas venulosa (group 3).
Surinam no such connections occur between with areas of high rainfall (over 2500 mm)
the Corantijn or Marowijne river systems and (Fig. 10:7) and may have been caused by the
the Amazon Basin; this explains the absence most recent arid phase, which apparently
of these two species in that country. ended 2000 years ago and caused a separation
Other species of the Amazon Valley appar- of the upper and lower Amazonian forests
ently succeeded in reaching eastern French (and the animals living in them) (Haffer,
Guiana but did not penetrate farther west. 1974). A number of these species are dis-
The distribution of a few species with dis- tributed in an arciform area from Bolivia
junct populations in upper Amazonia and along the eastern foot of the Andes to the
near the mouth of the Rio Amazonas (Table Guianas. This arc can be termed the Ama-
10:2, Appendix 10:1, Fig. 10:12) is correlated zonian Arc. Lescure ( 1977 ) called the north-
260 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
ern part of this arc ( Serra do Navio to Loreto, Paru area. Fluctuation of the size of the
forests, thereby at times forming a barrier
Peru) the Roraima Arc, because he believed
that all sandstone in this region belonged to between the inland and coastal savannas, was
the Roraima Formation. As pointed out be- responsible for the isolation of the inland sa-
fore, this is not the case. However, the exist- vannas; lizards living there are distinctly dif-
ence of an arciform distribution pattern in sev- ferent from the populations of the same spe-
eral reptiles and amphibians seems to be real cies in savannas farther north (Hoogmoed,
Most of the species in this group appar- Most of the species from southeastern or
ently had their origin in Amazonian South central Brasil reaching Guiana do not extend
America; from there they dispersed into farther west than French Guiana or Surinam
southern Central America (Fig. 10:13); a few (Fig. 10:13); only seven [Hyla x-signata (tree
are of Central American origin and dispersed frog), Leptodactyhis fuscus (leptodactylid
into South America. One example of this last frog), Pseudis paradoxus (pseudid frog),
subgroup is the teiid lizard Cnemidophorus 1.
Phrynops geoffroanus (chelid turtle), Liophis
lemniscatus, occurring from Honduras to the miliaris (colubrid snake), Coleodactylus mer-
mouth of the Amazon. This species occurs idionalis (gekkonid lizard), and Tropidurus
only along the coast in Guiana. The fact
that
torquatus (iguanid lizard)] reach Venezuela.
this species is still extending its range along The majority, if not all, of these species are
the lower Amazon (Vanzolini, 1970b) and inhabitants of savannas or open swamps, and
that it does not occur in the far interior of the their distributions are closely associated with
Guianas indicate that it is a recent immigrant those habitats. Apparently these species are
from the northwest. The presence of forests recent immigrants from the southeast that
in southern Surinam and French Guiana, either used the savanna corridor (central
was a barrier to the dispersal of C. I. lem- and northeastern Brasil to southeastern Vene-
niscatus ( a savanna inhabitant ) into the large zuela) during the last arid phase (about
inland, edaphic savannas in the Sipaliwini/
2000-3500 years ago), when the greater part
1979 HOOGMOED: HERPETOFAUNA OF GUI AN AN REGION 261
Table 10:5. —
Comparison of Rainforest Snake Faunas
ANALYSIS OF GEOGRAPHIC of Different Regions in Northern South America.
DISTRIBUTION Common
Species in
ana. For forest-inhabiting frogs there is a open formations are found in the Iquitos area;
distinct barrier within Guiana, formed by the all of these are either associated with open
Essequibo-Rio Bianco Depression and also, aquatic or edge situations. Real savanna spe-
the forest anuran fauna is distinctly separated cies are absent, because no suitable habitat
from that to the southeast. However, these is available in the
region (Dixon and Soini,
conclusions are based only on data from four 1977). When comparing these snakes (Table
areas (three of which chosen with a certain 10:6) with the open formation species of
bias) and therefore should be treated with Guiana, it is clear that the resemblance be-
much reserve, although they do confirm the tween Iquitos and the three parts of Guiana is
picture that emerged from a first study of small. Within Guiana there is a lower
degree
distribution maps. These findings for the of resemblance between the snake faunas of
anuran faunas can be most easily explained western Guiana and the Brasilian part of
by assuming that the Essequibo-Rio Branco Guiana, but this is caused by the presence in
Depression not only served as a connection western Guiana of several species reaching
(and dispersal route for aquatic species) their eastern distribution limits there and in
between Guiana and Amazonia, but also was the Brasilian part of Guiana of species reach-
the area that retained its savanna vegetation ing their northern distribution limits there,
longest, as still indicated by the presence of and of species that are known only from the
large savanna areas in the border region of Amazon Basin. However, the data for snakes
Guyana and Brasil and in the coastal area again are based only on four areas.
near the mouth of the Essequibo River. Thus, The lizards and amphisbaenians ("sauri-
this area formed an efficient barrier to the ans") appeared to offer the best possibilities
dispersal of forest frogs; at the same time, it for a faunal analysis, because there were sev-
formed a dispersal route for savanna frogs. eral places from which representative samples
This situation apparently lasted until fairly seemed to be present (Tables 10:7-8). How-
recently, untilunder the influence of an in- ever, upon closer examination, it soon turned
creasingly wet climate the forests in the out that the data were not very reliable. This
Guiana and Imeri refuges started to expand holds true for the forest lizards and amphis-
and met in the Essequibo-Rio Branco De- baenians of Lely Mountains, Paramaribo, and
pression. This explains why many forest spe- Alto Maraca (Amapa). When compared with
cies have their eastern or western distribution the entire region of which they are part, re-
limits at the Essequibo-Rio Branco Depres-
spectively eastern Guiana (twice) and the
sion. Several species that apparently were Brasilian part of Guiana, they show resem-
associated with one of the refuges in the re- blance factors of only 0.69, 0.59 and 0.64, re-
gion succeeded in crossing the depression, but spectively. These are hardly more, or even
this could have taken place only lower, than their respective resemblance fac-
recently
when the savanna vegetation was substituted tors with the lizard fauna of Belem (0.62,
by forest. 0.68 and 0.57). For Cayenne the situation
Comparison of the snake and lizard faunas seems to be better; when compared with east-
of several localities gives quite a different ern Guiana it shows a resemblance factor of
picture.The data for forest-inhabiting snakes 0.73, but here we should keep in mind that
(Table 10:5) show that there is a great re- the total number of species reported from this
semblance between different parts of the locality also contains old records that pos-
Guiana Region and that the resemblance with sibly refer to specimens that were shipped
the forest snake fauna of Iquitos is fairly from Cayenne but actually did not occur
good, but distinctly lower than within Guiana. there. From the remaining data it is clear that
The data suggest a gradual transition within there is a diminishing resemblance westward
Guiana from west to east and also from Iqui- between the rainforest lizard faunas of Belem
tos to Guiana,but owing to lack of data from and Guianan areas. Within Guiana the re-
intermediate localities this last hypothesis semblance is high, and nowhere is a clear
cannot be proved. Only five snakes inhabiting break apparent. The resemblance factor be-
264 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
tween Belem and Iquitos is slightly higher distribution of a number of eastern forest
than that between Iquitos and the different species to the west and of western forest spe-
parts of Guiana. Resemblance between Iqui- cies to the east. No such function is present
tos and western Guiana is smaller than that for savanna inhabitants that, with the excep-
with the Brasilian part of Guiana. This ap- tion of local endemics, are spread throughout
parent reversal of expected resemblances is the area. There is no separation to the south-
the result of the influence of Amazonian spe- east savanna-inhabiting species, which
for
cies in the Brasilian part of Guiana and in consequently show a great resemblance with
eastern Guiana. Apparently there is a fairly the savanna fauna of northeastern Brasil.
well-developed barrier southwest of Guiana, The data do not present any evidence for
separating the lizard faunas of Guiana and the recognition of a Guiana Region as de-
upper Amazonia. Again, our knowledge of fined by Lescure (1977); distinct breaks be-
areas intermediate between Iquitos and Gui- tween eastern Guiana and the Brasilian part
ana is poor, and the conclusions must be of Guiana are nowhere evident.
regarded as preliminary. When comparing
the savanna lizard faunas of different regions
(Table 10:8) we get a very different picture. CONCLUSIONS
There is no resemblance with Iquitos, where
this category of lizards is
completely absent. The herpetofauna of Guiana, as it is
The agreement between Belem and the dif- known at present, is a composite fauna with
ferent Guianan areas diminishes westward, a complex history. A number of endemic spe-
and there seems to be a break between west- cies belong to old genera (endemic or with
ern Guiana on the one hand and eastern disjunct, relict distributions) that apparently
Guiana and the Brasilian part of Guiana on inhabited certain parts of the area since the
the other. Upon closer examination, this ap- Cretaceous. Other endemics probably origi-
parent break is caused completely by the nated in the region during periods of isola-
presence in western Guiana of a number of tion in forest or savanna refuges, which are
altitudinal endemics and of western species assumed to have existed during arid and wet
just reaching their eastern limits in Guiana. phases in the Pleistocene-Holocene, respec-
When thesespecies (forming 605? of the tively. The most important
forest refuge was
savanna lizard fauna) are excluded, there are the Guiana Refuge on the northern slopes of
no breaks for the remaining general savanna the Tumuc-Humac and Acarai mountains. A
lizards within Guiana, neither with Belem. lessimportant role was played by the Imeri
The only break is between Iquitos and Gui- Refuge in the region of Serra Imeri and Serra
ana, and this can be completely explained da Neblina. The species restricted to higher
by the absence of suitable savanna habitat in altitudes survived the arid phases in disjunct
upper Amazonia. As for savanna-inhabiting forests on the higher slopes of the tepuis, col-
frogs, the Essequibo-Rio Branco Depression lectively known as Tepui Refuges. During
formed no barrier to that part of the savanna arid phases of the Pleistocene and Holocene,
inhabitants that had been in the area rela- the species isolated in the refuges underwent
tively long. For a number of local savanna differentiation and, depending on the time of
endemics it seems to act as such, mainly be- their arrival in the area and also on their rates
cause those endemics did not have the chance of evolution, they differentiated into endemic
to expand their area of distribution. genera, species, or subspecies.
On the of the data presented in
basis Although some species show relationships
Tables 10:.3-8 can be concluded that for
it to Andeanspecies, these are not direct and
forest inhabitants Guiana seems to be a real only indicate that both Andean and Guianan
herpetogeographic entity, well separated from species evolved from the same or related low-
surrounding areas to the southwest and the land species. The Guianan species of Atrac-
southeast. Within the area, the Essequibo- tus (colubrid snakes), Eleutherodactylus
Rio Branco Depression forms a barrier for the ( leptodactylid frogs) and Centrolenella
1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 265
de los tepuyes y de una manera suficiente tan tion de Riolama no queda muy clara. Los
solo su aspecto herpetologico. Dado este es- generos endemicos residuales (AUophnjne,
tado de cosas, por fuerza hemos de limitarnos Rliinatrema, Pekocephalus, Amapasaurus,
a interpretar los datos obtenidos de la herpe-
Mesobaena) son endemicos de llanura y su
tofauna de la altiplanicie de la Guayana. historial esta
probablemente asociado con el
Actualmente se conocen en la Guayana de de Guayana y de Imeri. Una
los refugios
178 especies de anfibios y 230 reptiles, totali- especie endemica de Euparkerella apunta
zando asi las 408 especies heipetologicas. No tener alguna relation con las montanas del
se conoce las salamandras. En dicho numero sudeste brasileno. La "Hylodes" duidensis
se comprenden: cinco tortugas marinas, un que hasta hace poco se creia estaba emparen-
chacon (Gekkonidae) cosmopolita, cinco tada con formas del sudeste brasileno, resulta
especies importadas de las Antillas y tres cuya ser bastante diferente y mas bien
representa
presencia en la Guayana es de origen dudoso. una derivation de los eleutherodactylini de
El residuo puede clasificarse en cinco grandes llanura. La mayor parte de los endemicos de
grupos
—
endemicos (38%), amazonicos (37%), altura son derivaciones subtropicales de
pari-
de vasta extension (12%), llegando el limite entes de la llanura tropical. Los endemicos
oriental de su extension hasta el escudo gua- de altura tienen una extension limitada a uno
yani (7%) y a la Guayana del Brasil central y o varios tepuyes. El origen de la mayor
parte
sudeste (5%). Si se consideran estos datos de los endemicos de llanura se explica prob-
aisladamente, referidos a los reptiles y a los ablemente por la formation en el pasado de
anfibios por separado, constatamos diferencias refugios forestales a traves de los cambios
importantes. De 52 porciento
los anfibios, el climaticos del periodo cuatemario. De esos
son endemicos, el 30 porciento son amazonicos supuestos refugios en region es el de la
la
representar una radiation reciente y la posi- extension de las vegetaciones abiertas del ul-
1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 267
caracteristica en formas de llanura se debe Gleason, H. A. 1931. Botanical results of the Tyler-
Duida Expedition. Bull. Torrey Bot. Club 58:
probablemente, en la mayoria de los casos a 277-506.
la competition ecologia u a otras condiciones Haas, F. 1957. Zur Tiergeographie von Amazonien
particulares del medio ambiente.
und dem Guayana Schild. Mitt. Naturforsch. Ges.
Bern 14:59-64.
Haffer, J. 1969. Speciation in Amazonian forest
birds. Science 165:131-137.
LITERATURE CITED
Haffer, J. 1974. Avian speciation in tropical South
America. With a systematic survey of the tou-
Bisschops, J. H. 1969. The Roraima Formation in cans Rhamphastidae ) and jacamars ( Galbuli-
(
Surinam. Verh. K. Ned. Geol. Mijnbouwkd. dae). Publ. Nuttal Ornithol. Club, Cambridge
Genoot. Geol. Ser. 27:109-117. (14): 1-390.
Boulencer, G. A. 1895. Description of a new ba- Hills, T. L. 1969. The savanna landscapes of the
trachian ( Oreophryne Quelchii ) discovered by
Amazon Basin. Savanna Res. Ser. Dep. Geogr.
Messrs. J. J. Quelch and F. McConnell on the McGill Univ. Montreal, (14): 1-38.
summit of Mt. Roraima. Ann. Mag. Nat. Hist.
Hoocmoed, M. S. 1973. Notes on the herpetofauna
(6)15:521-522. of Surinam IV. The lizards and amphisbaenians
Boulencer, G. A. 1900. Reptiles: 53-54, Batraehi- of Surinam. Biogeographica 4:1-419.
ans:55-56, in
Lankster, E. R., 1900. Report
Lancini, A. R. 1968. El genero Euspondijhts (Sau-
on a collection made by Messrs. F. V. McConnell
and J. J. Quelch at Mount Roraima in British ria, Teiidae) en Venezuela. Publ. Occas. Mus.
Cienc. Nat. Zool. (12): 1-8.
Guiana. Trans. Linn. Soc. London Zool. (2)8(2):
51-76. Lescure, J. 1975. Biogeographie et ecologie des
Brown, K.
S., Sheppard, P. M., Turner, J. R. G.
Amphibiens de Guyane Francaise. C. R. Seances
1974. Soc. Biogeogr. 440:68-82.
Quaternary refugia in tropical America:
Evidence from race formation in Heliconius but- Lescure, J. 1977. Diversite des origines biogeo-
terflies. Proc. Roy. Soc. London, B, 187:369-378. graphiques chez les Amphibiens de la region
Chapman, F. M. 1931. The upper zonal bird-life of guyanaise. Publ. Lab. Zool. ficole Normale Super.
Mts. Roraima and Duida. Bull. Amer. Mus. Nat. 9:53-65.
Hist. 63:1-135. Maguire, B. 1945. Notes on the geology and geog-
Crump, M. L. 1971.
Quantitative analysis of the raphy of Tafelberg, Suriname. Geogr. Rev. 35:
ecological distribution of a tropical herpetofauna. 563-579.
Univ. Kansas Mus. Nat. Hist. Occas. Pap. (3): Maguire, B. Cerro de la Neblina, Amazonas,
1955.
1-62. Venezuela. A newly
discovered sandstone moun-
Cunha, O. R. da. 1961. II. Lacertilios da Ama- tain. Ibid. 45:27-51.
zonia. Os lagartos da Amazonia brasileira, com Macuire, B. 1970. On the flora of the Guayana
referenda aos representados na colecao
especial highland. Biotropica 2:85-100.
do Museo Goeldi. Bol. Mus. Paraense Emilio Mayr, E., Phelps, W. H., Jr. 1967. The origin of
Goeldi Nova Ser. Zool. (39): 1-189. the bird fauna of the South Venezuelan High-
Descamps, M., Gasc, J. P., Lescure, J., Sastre, C. lands. Bull. Amer. Mus. Nat. Hist. 136:269-328.
1978. Etude des ecosystemes quyanais. II. Don- McDiarmid, R. W. 1971. Comparative morphology
nees biogeographiques sur la partie orientale des and evolution of frogs of the Neotropical genera
Guyanes. C. R. Seances Soc. Biogeogr. 467: Atelopus, Dendrophryniscus, Melanophryniscus,
55-82. and OreophryneUa. Nat. Hist. Mus. Los Angeles
Dixon, J. R„ Soini, P. 1975. The reptiles of the Cty. Sci. Bull. (12): 1-66.
upper Amazon Basin, Iquitos Region, Peru. I. Muller, P. 1973. The dispersal centres of terrestrial
Lizards and amphisbaenians. Contrib. Biol. Geol. vertebrates in the Neotropical realm. A study in
Milwaukee Publ. Mus. (4):l-58. the evolution of the Neotropical biota and its
Dkon, J. R., Soini, P. 1977. The reptiles of the native landscapes. Biogeographica 2:1-244.
upper Amazon Basin, Iquitos Region, Peru. II. Nott, D. 1975. Into the lost world. A descent into
Crocodilians, Turtles and snakes. Ibid. (12): prehistoric time. Prentice-Hall, Englewood Cliffs,
1-91. 186
N.J., p.
268 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Rivero, J. A. 1965 (1964). The distribution of Tate, G. H. H. 1930a. Notes on the Mount Roraima
Venezuelan frogs. V. The Venezuelan Guayana. region. Geogr. Rev. 20:53-68.
Caribb. J. 4:411-420.
Sci. Tate, G. H. H. 1930b. Through Brazil to the sum-
mit of Mount Roraima. Natl. Geogr. Mag. 58:
Rivero, J. A. 1966. Notes on the genus Cnjpto-
584-605.
batrachus (Amphibia, Salientia) with the de-
Tate, G. H. H. 1932. Life zones at Mount Roraima.
scription of a new race and four new species of a
new genus of hylid frogs. Ibid. 6:137-149. Ecology 13:235-257.
Tate, G. H. H. 1938a. Auyantepui. Notes on the
Rivero, J. A. 1967a. Anfibios colleccionados por la
Phelps Venezuelan expedition. Geogr. Rev. 28:
expedicion Franco- Venezolana al Alto Orinoco 452^174.
1951-1952. Ibid. 7:145-154.
Tate, G. H. H. 1938b. A new "Lost World." Nat.
Rivero, J. A. 1967b. A new race of Otophryne ro- Hist. 42:107-120, 153.
busta Boulunger (Amphibia, Salientia) from the
Tate, G. H. H. 1939. The mammals of the Guiana
Chimanta-Tepui of Venezuela. Ibid. 7:155-158. Region. Bull. Amer. Mus. Nat. Hist. 76:151-229.
Rivero, J. A. 1968a. A new species of Elosia (Am- Tate, G. H. H., Hitchcock, C. B. 1930. The Cerro
phibia, Salientia) from Mt. Duida, Venezuela. Duida Region of Venezuela. Geogr. Rev. 20:
Amer. Mus. Novit. (2334): 1-9. 31-52.
Rivero, J. A. 1968b. Los centrolenidos de Venezuela Todd, W. E. C, Cahriker, M. A., Jr. 1922. The
(Amphibia, Salientia). Mem. Soc. Cienc. Nat. birds of the Santa Marta region of Colombia: A
La Salle 28:301-334. study in altitudinal distribution. Ann. Carnegie
Rivero, J. A. 1968c. A new species of Eleuthero- Mus. 14:3-611.
dactylus (Amphibia, Salientia) from the Guyana van der Hammen, T. 1974. The Pleistocene
region, Edo. Bolivar, Venezuela. Breviora (306): changes of vegetation and climate in tropical
1-11. South America. J. Biogeogr. 1:3-26.
Rivero, J. A. 1968d. A new species of Htjla (Am- Vanzolini, P. E. 1970a. Zoologia systematica, geo-
phibia, Salientia) from the Venezuelan Guyana. grafia e a origem das especies. Univ. Sao Paulo,
Ibid. (307): 1-5. Inst. Geografia Tese Monogr. 3:1-56.
Rivero, J. 1970. On the origin, endemism and
A. Vanzolini, P. E. 1970b. Unisexual Cnemidophorus
distribution of the genus Stefania Rivero (Am- lemniscatus in the Amazonas valley: a prelim-
phibia, Salientia) with a description of a new inary note (Sauria, Teiidae). Pap. Avul. Zool.
species from southeastern Venezuela. Bol. Soc. (Sao Paulo) 23:63-68.
Venezolana Cienc. Nat. 28:456-481. Warren, A. N. 1973. Roraima. Report of the 1971
Rivero, J. A. Notas sobre los anfibios de
1971. British expedition to Mount Roraima in Guyana,
Venezuela. I. Sobre los hilidos de la guayana South America. Private publication, Oxford,
venezolana. Caribb. J. Sci. 11:181-193. 152 p.
1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 269
a
„_, o
| E 8 o a
tDD-OajQjGCQj3>-i.CGajn)3(uajajaQj!DrtrtftaicD
- £ "'
_i<:j<_i<j<
tS » II cl -1
B
t v.
sji _ 5
bg §
-3a 8" P5P5
II
11 II
ceo
< pa
So «;
. c N C-3
2 .2 £ ;tfp;p:p:
•43 ft c ^ C CPL,
: : irt :
irttf :
ipa ;
g °is s
aa c
N C 'S
rt ro cs
:<J :P3 IX : A !
« tf ;<&<&&& \ :
pa : :
pfi : :
ps «
H >
-J o
N COo
_o « :< : :
iX P5 ipap; :pwB5ffi :X P3
S P 3
i i
^
CJCO e
C 4j S § §.2
*1 '<! : : :X : : :
X PS :
« PS :
PS PS « : :
X PS : :
PS :
PS
3
!.| d
E "3 a.
Iq s 1. ~CH 3
"K c r N
3
a ii II m Ssg» :W :
5"
- a
Q'g § 5W 1*8 |
z«s d at 3 < :PS : : i
PS ! i
PS i
PS :PSPSPS : i
PS PS PS i
PS PS
fl 3
Oh - C .
o~ a
Sill « §*J
-C ft
ft
5 £ c
vi
<D
-"-
3 PjjPSpSPS 22 x ;
^
'
A 1
°5 (a i
PS PS PS PS X c/i ;
:| WW^PSPS^
(3
II
g
ca C a
C ^ Im 9
$«PSPSPS bS PS ipSpSPS ipSpSpSpS PS PS PS PS PS PS PS PS
o CO - *3 : :
:
:
O >*
wo
*°^
O 3
-<
_-T3 ni
&in §to •
3 (U O
-^ O
CO
3 .2 5 ^
o o O)
to
o <m o <N CO
in oo oo h
OOOOOOOOOt^coO^-,
„ nOONHHfflH«»)MOSnOO OOOOOCO CO
OS)
H O rt ''nn«iO'»-
I <J (1) J J J J I
P -t<COC-)^HCOCOCMCOCDT|t-ilo^gt-t-COt~CO-H§ ^t-^^
I I
ft
1
<-»-(
° Q.
O0 100 rt ioo-lci»o4coooci>cic!>4.|
HIOHO <M{N
dooo<icJ>2:
lO COCO ^ en 05
CO^ Scici -— Tl
D
II c/o
w CO CO ^H
i
»—I
o
£o> «
is a . CO.
S.S5
a s
o .a
In 2 >
•3-3
oXS£ c
3 ^a
tS O 0.
2' K o 1
Bu a-
w tS ||
P. K c 2
< <J5 00 N <->
270 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
O XI
£ J
-3
c o 'ps
co o- c/} PS X I I I OS la- lt« lS i«
PS
« I
: i
:
I
n S C "
2 * « eu PC OS
9<
°<S '3 jr 'PS
co co co PS PS PS I :oS I
;0S I IBS PS I I to PS
g IPS I
o
S 3-w
c
« a
2 '"
3 'IX
PS X PS I
"2 PS
PS
I eu :
os PS I
X :
PS g IPS IPS
S§3.2 iW ;cucopSpSPSJ5«PS I eu ;
p- X co |W ; pS :
K cu
ps
< u
u a 4)
CO
« c r S :
W co I co i
W I i
PS : co I : co co co I
!
W PS
£« S g 0=
jS ° c
CO "C -- w co co co PS PS PS co PS I"- IPS i
PS PS :PS ;
etf a) 3 PS °g
m
£ «
^ aj
co co co PS IPS leuQtfW PS :PSpS •
co co PS
co
li
•—
3
i
g PS J? ipScogWpSPS
a E § co
~ .3 iW cococoosgpS icopi^oS IPS IPSPS ICOPSPSPSPS ;
^p^S tf ^*^ ! :
:p=
PS
WO
100 OOO
^tiooooooio o ooookmoo -J O _ O O
oconoiooim^ coo
C7>
o
'S9
<McoooomcooiM 00 o t
-.(M (_,-to-rooo-ro ln ooc<i<MO<Mco<M <^o oot^
'3° ooooooooo
Tf m
OS on
CO
2 o o^oooooo
-r*r^co in
oociooooo
ao COO
CO O iei-^oS^i
N, ^ —i
i—
i co
** -N
w in
2 *-w
<cj a a
C
"a "2
'
ago
Cj "C
u 5 1-^
-
&§l
JiCS cj 60 <->
=» ^ a
O ; «• ^ <4» C — .**
*3
5 :=> -i-
at=S2» p^ <£» ^J
"2; •*-
a. a.
3 O
~ iC tj .^ 1)
5i< ?2
CO&O&dc^SoCO^Co
^ r^l ^^^Siii^^
rn^Ti — ~J
«^> *^> *ii .1 £] •£! (31 •-) ~) *-'
C>3 «* p- ni
c
—o 2
S3
iX IPS : 1
W :
'PS
: X : 00 :
: 00
5 «
ft rt
IPS PSX 1
PS :oo w 'PS PS"
B
o 2<
o
— :
K PS i
CC ico W W H
PS
'5 3
w
8 c E
Pi :
PS PS co «W W Hm IX
S 'S o • PS
0) *3 2 3
ft
i ? "j
i
w I
00 CO 00 ! co 00 : i : 00
O a-
>
i o c
ted
3 CO X v.PS CO UJ !
W PS PS
W !
icoSoopS ; co : :
ico : : ico
pq
ftO
3
Ho 0) 3
x£?i«^ : 00 W ;
3
& 6 g
B ii =3
^ '3 pScoPSpSco^pScoooH X :gbJ co co ™
,W co PS PS co co 00 00 PS X 00 pS 00
o : 1 1
:
o
bo
O
a> © ~* ©
CD
©el OO OlO h
O o CO o CO
CD © O © tn
OO 00
lOOCt
O
O O O CD
.ot-oiiooo-r-^o — 00-H o r-,-i-"i<iooo Of-clOoOOOOTOT.-sr—
=> n.
©© ooo <
J.JI.2.II
o © o — o omooo J. J I 1
00 G> ^H 0<M(N
W CO t^ CT>
<i>
5SS
~r 'Sj
*->'-*-«!- ^ CY, 1- . __
a .g « o
~ ^ cj
% s S 8 Cu c S 2 '3 S
^j \u O O a c ^ a -S
<-=> -a -a -a -o 5, ,53 ^ 60 &fi
MM Csc
53 _53 ja ja a a
272 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
c
^ .2
S £ CD CD 0)0)
11
N C
d cd .
05 its : : :
W I
?3
—
CD
c :«
G G Pi
PSpSoS :P5
WWgcC I
J-. o — -
N C »
J a
ft ra c3
c
eSeSoS i :
WH^fS :BS<
1-,
C
oSeSpS ICC ;
< W
ctj b .2 *<"
(S OS PS !
Sgw ;
!
!(S
E "« <u
, ,
1
w :p3
s a c/3
l
<3J
8
I" ! rt i« ;
!(«S9u i
;
«
fl « 3
OlX« , w OS PS PS : WW" 3
«PS w PS^ccc^^'itS^
,
PS' 5 "i
£o
c c
o StntfM !«WW| SPSgpSXPSgW^p- ipScfl co PS PS PS
M r?
MOO
oo >o COOOOOOWO (M o
ooo OO o
ooSoo<
— o
'
O -F
C5 <N
,
3| <N — *
— -^
<M
d --ocinionoioof,
MO 'MOO a> in
O O O oo in
, O"
aa
I I
"?
c:
oooodoooo«^^2^^ Soo
'
oo'o.o oo o«
i
—
i . ^H f—I
in °*
'
w CO 0>
I ! I
•s
S
Z< «i .i=-S
fc -Si
cu •«. o o S to c^ .'o fc
= "
'3
<u
-2
3 -= S £ a 5 '
~:s?.<«b =
ft
CO
W b "a ;«
Co CO Co
SuCsooaooii
Co
fa r^r* r* *s 5^ V o « a
« o r, «
a s S 3 ? 5
•Si cu
"9
Cy
"53
Cy
"9"a-~
Cy Cy
ft ft o a a a a a a a o o
CO CO l
O CJ CJ Cj o o o o o cy
a a o o o p a
i
-S -s
J)
s>
Co
i»
^
s
co
3; ?
i:
L?
i^
l?^^,^,^=^-~
i:^ ™ « ™ ^?"^: -Ji -Jt>i i- 1- §
•*--•*-.
ijcjdi^^^^^^jj *t- •*- s =
a; a; a; S; a; a; a; a;* as a;
1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 273
274 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
e
uo ao
6 .S
C
8.1 §1
O
oj rt oj
J-2 g
aft
s
-3
aPi
a
o
1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 275
o '.C
iX 'PC
I :X<*> UN
'OCK '
,
ips :<! i
:<;«: <CS : i
W
<
s as pstd Ed
PC PC* W pCggpSPC ict < PCCS ieCeC
XX :
:X : : W X -aw cCpS ;« SSpS Ed
PS PC PC
S3
-a C/5
a
a XX :X W W :X
.Ed Ed
psps << ;<<
:
PC PS 'S3 i
pc 1
a
o)
"a "5
a C C o Ed
:M6C PC< < c/3
S « «> c S PS
IPC
PC
PC
,
W
OS
>
c c
~
.2
-F .3 iX : i
igo-pSgpC l|eoK
:
g PS tf i
PC :
<< :<< < ^ < < PS IW
M ftO
5 £ IX IX I^XCSPC :
:g :pCg
I
;g
I
PS
g g K PS :
PC PC -< :
<< I
< :<« ig
C M
It c
& i « X OS :X l« X g w CC g PC X« <
o
I : I
PC i
«j |
:
gg PS PC I I I
$ I
i^ .
jg PC PC IX
wo
oo
oo
ooo
-J"
3 !
w
276 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
c
<~ .2
O B
H
^3
. en oo
< «
2QJ
S n C
« «
a
9 i.s
u
.a
K-rt i =
JCPS£
0) a) a>
a
o
CO 3
"3
a ftO
|
I CO - —
-a
a
ft
cd
&
o
W
1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 277
e
^ .2
sua J a ftflSESSE <>>
<<<<<<< E'v^ C n U G U fl O Q'O u &rt s fl s E a cs rt
H .9 St
i-J < t/3 Ui
PS 'PS
PS :PS IPS
w
PS PS :g PS
a H S »
Q.
rt rt « PSW ipsxgpspsps
.
W B
g-a pq
iff! rt
*s
; :ps PS
'
w :pS IPS :
IPS
"*
PS
o 00
'a a
:pSW
PS
PS PS PS IPSX Wps
PS
:03 :
PS 'PS
' '
W
"3 oo
§ psw XPSpSPSX WpSpSPS IPS
W , PL, ,
^
« ,
.
w icnps W
w
,
Bio-
,
'
•
ps PS PS p;
< O
-2
aa
C «o CO (75
! W X ! : W : i : oo
^ '« w.
^wg
: : : : ;
S PS PS
3
^ o c CD
W W
$ 3
:
« PS :
IX Wpsps :wpS PS :psg :
:& :
PS : :
,g ! cn PS
PS
aj PS
=5
£ g 00
a; C
IPS PS XPSpSPSXWpSpScnoopS BSPSPSH
V, .2 :pS£
PS PS :^S| :
|«WgpH^| gX
aa
(3
wo
CD
O oO
cd in (N
0)OMO .OOl50 n lN o o o OO
CD
>
y ^ ob cDooait^cocoincsi§"j
'
in
9
ooo ^o 0210 o
NMO w
Aeii^ciiiScici
CO OM" CO . oooo
oo oooo in °
—I <N
OP ci.
w (N CM ^H
in
I I ! I I
o
I ! !
! !
§2 3
2 o
go. . ==
-^
co irf a t3
a ffl-s o ^
a ~-o
IT
<=> 3 £-2 u' 5 -5 £ 2 -S 5 2
p =0
a !•* :- 2 CD
§ s- ?
5 «o
<u c G 2
**•
ci
:
^ ^
-a IsS. ^ b^
co co a ^ o x 2 co 2~ <»
5 g
— -S a. w Z a -a .a
*^ -^! "? '^l "S "J^ "^ c\,
'"a 5 »5> =o co Co oo
o :u -Q
CO
cocoCoCoCocoCoCoCo _ ^) Cii Qj O 1- Cj U,
o o o a a.
ah "a "c oa oa 2o oa
to a go
-c-a-c"a Cl
o o o o o
i.
is ?
^ P
nS 6a ou -=»
c c e c c
u u Q; (u &) a > C9 ^j
j- j- j- j-
2 2 °
•
%) : •S "o "o 5 o o o
x >< xxx >-j : «5 «* as «pa JUOOUOU ^-5 <:<^
278 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
B
„_, O
o B
CCc8^C2CCJftfl)(BU^C5)|v«caj)ci)DO,c:-vJ _Q CD CD CD C CD G
JJJJ <C J W W
CD CD CD CD CD CD CD
E-i co
-3
en C/5 en
:^PS oSW w iW .
w
c OS os
m OS
c '5 C/3
w
.2 fc
wg^w ' Wen pS OS OS w
CPh os os OS
PS
JJ N C
ww ^ososw
C/3
w w
os ps
wos OS '03 '* ;
S PS
5 os os OS
Pi' '
CO
gg r|eSPSW W OS i
w
OS os
os
w
iX !
W : Wen IX ;cn PS :2cn
S '3 o os OS
< O
en
A^ < CO
W
en
CD CD
:
W :
OS
:
Wen
OS
us>
WW ji n
iWa^jgu
en en
W < H CO n «i<
« 2 OS OS I^PSOSW icnpS :
pS oS OS K^OS < :
:WpS igcnXpSOS^X
ftO OS
03
w
i 5
—
a W p; WenpSpSDS PS§enenen^ XXW as en
W tn pS OS X £j X
PL.a^'OSW'I'npSXS
i :
CO • ;
CD 3 £j
PS
Pi
-.
a rt
en
r,
o 3g
-* :|nW w .«„;
^ .wS^
K os : W cS PS PS PS !
S-«1
Ch
CD
60 £6 osg
o
oo
e>)0 ot^roo-roo ooo
in o o m o oi o 3oooo
mioocdoo©
noiOnHoejHHHO-OrtO in co
OJr-^COcJJCO^Ol-Hin (M^fflSHOeDCJ-i«ng n ricJ
Cv.
oo ooooo
w
.« to
CJ 2
1979 HOOGMOED: HERPETOFAUNA OF GUIANAN REGION 279
^O o
'43
S a £ S k,K. d C C
<<<<< ??JDJJJ
<d oj j) jj OJ q;
c c o
WW
«BS
« « S"> w
i W
C/2
W W [
K
:
EG
o
c -o
BJK
c c
< W
'
a
r *PS
C/3 c« C/i
! § £ c
: W
:
OJ
>
m r i °2 OWlfi
KB
a a
rt :t-no
—I <N
PQ
C/3 C/5
:X X <m
" —
i
_ °°
rt in
'
— £0
a m m in
& aSofi ww (fi CS
-H r-f CI
wo
o
° o
°oT
in -* :
s° s
o
o i
5 in
00
:
^£ ~ o
o
=
bo o *
S3 S § sy
'Soli
c
a 2-=
a a a a OSS"*
~ Co
w 3 * G ~ u S a a
-5 « a a a
o 2 S S £
III It
ex, o, a; E-< E-i
11.
Origin and Distribution of the Herpetofauna of the Dry
Lowland Regions of Northern South America
The lowland open formations of north- provided us with a list of amphibian and
rep-
eastern South America include the llanos of tilespecies for the El Manteco region of the
the Orinoco, and the Caribbean arid and southeastern llanos of Venezuela.
semi-arid lowlands (Fig. 11:1). The dry low-
lands of the Guianas, are discussed by Hoog-
moed (this volume). The area of lowland
VEGETATION
open formations is large; the llanos of Vene-
zuela alone occupy more than 180,000 km 2 The llanos are extremely diverse and con-
(Ramia, 1967), and the Caribbean arid and tain many habitat types different forest asso-
:
semi-arid lowlands cover some 50,000 km2 ciations, various types of savannas, and the
(Sarmiento, 1976). ever-changing aquatic along with
habitats
These areas are geologically recent.The subtle topographic features. This diversity
llanos of Colombia and Venezuela emerged makes it difficult to understand the llanos
during the Pleistocene and developed with without collating the geologic, topographic,
the contribution of Andean sediments. The floristic, faunistic,edaphic, pyric, climatic and
coastal desert areas were highly influenced by human intervention factors. In order to better
sea level oscillations during the same period; understand such diversity and the character-
most of it is sedimentary in origin. ( Direction istics of the reptilian and
amphibian faunas
de Geologia, MMH, 1969). Although the one must refer to Beard's ( 1953 and Ramia's
)
topography of these areas varies, the domi- (1967) comprehensive works on the Neo-
nant flat-land physiography is
conspicuous tropical savannas.
throughout. The herpetofauna of the llanos of the
Climatic conditions vary considerably, but Orinoco and the Caribbean arid and semi-
some general trends are noted. Annual rain- arid lowlands encounters a wide range of
fall maxima vary from 1000 to 2000 mm, but environments in the form of swift gradients.
the dry periods are usually six months or The rainfall patterns vary from almost six
longer (Ewell and Madriz, 1968). Mean an- months of rain in the llanos to almost no rain
nual temperatures for the area are above 24°C in the coastal deserts. Vegetation provides
(Ewell and Madriz, 1968). abundant and (in some places) continuous
Data pertaining to amphibian and reptile shade in the llanos, but vegetation is sparse
distributionswere obtained from Cochran in the drier areas.
and Coin (1970), Dunn (1957), Medem The following description of plant com-
(1968) and Ruthven (1922) for Colombia; munities based on Ramia (1967), Sarmi-
is
Donoso-Barros (1968), Roze (1966), Rivero ento (1976), and our own familiarity with
(1961), and Staton and Dixon (1977) for the area. These habitat types fall within more
Venezuela. We utilized a number of publica- generalized climatic types or life zones that
tions denoting the ranges of individual spe- have been proposed for by the area by Ewell
cies, but they are too numerous to mention and Madriz (1968) and Walter and Medina
here. Gorzula and Hoogmoed (pers. comm.)
(1971).
281
282 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fig. 11:1. The open savannas and hot, dry lowland regions of northern South America (after Hueck, 1966).
Stippled areas are forests and mountains; heavily dotted area is part of Panama excluded from the study. Num-
bers indicate locality zones in Appendix 11:1.
Sabanas abiertas y regiones de tierra haja seca y ca liente del Norte de Sur America. Las Areas punteadas
son bosques y montanas, y la zona con puntos gruesos es parte de Panama, que no se incluyc in el estudio.
Los numeros indican zonas de localidad in el ancxo 11:1.
Fig. 11:2. Gallery forest of a tributary of the Rio Apure, Estado Apure, Venezuela, during the late wet
season.
Bosque de galena de un tributario del Rio Apure, Estado Apure, Venezuela, en plena epoca de lluvias.
Fig. 11.3. Gallery forest of the Rio Apure, Estado Apure, Venezuela,
during the late wet season.
Bosque de galena del Rio Apure, Estado Apure, Venezuela, a finales de la epoca de lluvias.
284 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fig. 11:4. Pure stand of Palma Llanera, Copernicia tectorum, in the Estero de Camaguan, Estado Guarico,
Venezuela, during the dry season.
Palmar de Copernicia tectorum en el Estero de Camagudn, Estado Guarico, Venezuela, durante la esta-
cion seca.
nate these sandy savannas (Fig. 11:6). In of water level during the year increase plant
some areas there are no trees, whereas in diversity in these habitats. The Rancos are
others fire resistant trees (Curatella, Bow- the higher ground, formed by the accumula-
dichia, Byrsonima and Roupala) are common. tion of sediment along the streams and never
These well-drained grasslands are rarely covered by water in the rainy season. Some
flooded and are burned annually so as to bancos are pure savanna, whereas others are
provide fresh food for cattle. partly or completely forested (Fig. 11:4).
—
Paspalum savannas. During the wet sea- Rajios are covered with up to 20 cm of water
son these savannas normally are inundated, during the wet season. Esteros are the lowest
and apparently only one grass (Paspalum ground and may be covered by 1 m of water
fasciculatum) is capable of living permanent- during the rainy season. The plant species
ly in the area (Fig. 11:7). Normally, it attains composition in these savannas is constantly
a height of 2 m
and forms dense, pure stands. changing throughout the year. These changes
These savannas are more common in the wet- probably result from the rather short periods
test areas of the western llanos of Venezuela. when ideal conditions are available; a dy-
Savannas of bancos, bajios and esteros. — namic dry-wet cyclic succession occurs during
This type of savanna is more diverse in spe- a few months.
cies of grasses than the former two savannas. Medanales. —The driest sites in the llanos,
It seems as though subtle topographic and lacking both trees and grass, are sand dunes
structural variations of the soil and fluctuation in close association with Trachipogon savan-
1979 RIVERO-BLANCO & DIXON: HERPETOFAUNA OF LLANOS 285
nas and located mainly south of the Rio Tropical Thorn Woodland
Apure. Ramia (1967) described them as dy- Sarmiento (1976) distinguished this vege-
namic systems of sand dunes.
habitats. — The
tation type as lacking a continuous canopy
Aquatic rivers, canos, la-
and having large spaces of bare soil between
goons and temporary aquatic situations, such
sparse trees and shrubs (Fig. 11:9). The
as esteros and bajios, form an important habi-
most common trees are those belonging to the
many amphibians and reptiles of eco-
tat for
—
nomic importance turtles of the genus Po- genera Capparis, Prosopis, Cercidium and
Pithecolobium. The Cactaceae is represented
docnemis and crocodilians. This habitat is
by MammaUaria. Melocactus, Opuntia and
unique in the llanos and, although intrinsi- Lemaireocerus. The arid climate of this zone
cally different from the terrestrial habitats, it
has an irregular precipitation pattern; how-
shares the dry-wet effects of the climate and
ever, there is a tendency for rains to increase
oscillates in size
(some disappear in the dry towards the end of the year (Walter and
season). This results in brumation of many
Medina, 1971).
fish, reptiles, and amphibians.
Dry formations common in certain inter-
Andean valleys in Venezuela (usually around
Very Dry Tropical Forest
1000 m or above) are typical rain shadows
This type of forest occurs from sea level to (Fig. 11:10). Ewell and Madriz (1968) clas-
600 m where the mean annual temperatures sified the vegetation and climate of this life
286 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
i fi
Fig. 11.6. Trachipogon savannas in Estado Anzoategui, Venezuela, with scattered Chaparro, Curatella sp.
Sabanas de Trachipogon del estado Anzoategui, Venezuela, con algunos Chaparros Curatella sp.
Fig. 11:7. Paspalum savannas of the lower Rio Apure, Estado Apure, Venezuela, with pure stands of
Paspalum fasciculatum.
Sabanas de Paspalum fasciculatum del Bajo Rio Apure, Estado Apure, Venezuela.
1979 RIVERO-BLANCO & DIXON: HERPETOFAUNA OF LLANOS 287
Fig. 11:8. Very dry tropical forest in early Janu- Fig. 11:9. Tropical thorn woodland community
ary, near the Laguna Unare, Estado Anzoategui, in Estado Lara, Venezuela.
Venezuela. Comunidad de matorral espinoso tropical en el
Bosque tnuy seco tropical a principios de enero, Estado Lara, Venezuela.
cerca de la Laguna de Unare, Estado Anzoategui,
Venezuela.
Desert Scrub and Desert Andes, provided the necessary physical traits
to alter the climate from wet tropical to dry
The climatic regime of the desert scrub tropical, semiarid and/or arid desert regimes.
does not differ appreciably from that of the The climatic and elevational shifts eliminated
tropical thorn woodland. Floristically, it is an those less adaptive tropical species of the
impoverished thorn woodland equivalent to Magdalena and Maracaibo basins, while the
the deserts of arid temperate areas (Sarmi- Orinoco Embayment remained unchanged for
ento, 1976). The most arid areas are salt or a longer period of time. The latter region
sand deserts in the western Caribbean coast slowly filled with Andean and fluvial sedi-
of Venezuela and Colombia (Fig. 11:11). ments from the north and west from Cre-
Salt deserts are common along the coast with taceous to Recent times and is underlain by
low shrubby Chenopodiaceae, while sand some 10,000 meters of deposits (Direccion
dunes are inhabited by Acacia, Opuntia and de Geologia, MMH, 1970). The extensive
several geophytes (Sarmiento, 1976). Colombian and Venezuelan llanos are prob-
288 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
#
'^:/.
"<-
Fig. 11:10. Premontane thorn woodland community of a rain shadow desert in the vicinity of Lagunillas,
Estado Merida, Venezuela (above 1000 m, mean temperature less than 24°C).
Comunidad de matorral espinoso premontano en el vallc inter-andino dc Lagunillas, Estado Merida, Vene-
zuela (Sobre 1000 m.s.n.m. y con una tempcratura media anual por debajo de 24°C).
Fig. 11:11.Typical Medanos (sand dunes) in the desert scrub community of Estado Falcon, Venezuela.
Medanos (Dunas de Arena) en la comunidad de Maleza Desertica en el Estado Falcon, Venezuela.
ably less than one million years of age and tained by edaphic and climatic factors.
formerly may have been more patchy than Studies of the cactus flora (Sarmiento,
they are today. There is evidence that some 1976) along the coastal zones of northeastern
areas of the llanos are maintained as a fire South America suggest that endemism is low
climax vegetation, while other parts are main- and that phytogenetically, these zones are
1979 RIVERO-BLANCO & DIXON: HERPETOFAUNA OF LLANOS 289
more closely allied with similar areas in Cen- northern limit of Leptodactylus bolivianus is
tral America and the West Indies than that Costa Rica and L. pentadactylus Honduras,
of northeastern Brasil. Sarmiento (1976) sug- and both range southward to Bolivia. Lepto-
gested that the origin of the flora is hetero- dactylus fragilis and L. poecilochdus reach
geneous, with many plant species being Mexico and Costa Rica, respectively, but their
contributed from nearby and less dry plant southern limits border the Orinoco Basin.
formations. In addition, the composition of Leptodactylus fuscus, macrosternum and wag-
the vegetation is strongly influenced by cosmo- neri reach their southern limits in Brasil and
politan or western hemisphere weeds which Peru, but their northern limits are the dry
have greater numbers of species both north zones of northeastern South America, and L.
and south of the middle latitudes. The num- rugosus is restricted to the Guiana region.
bers of relictual, endemic taxa are relatively Thus, only one of the above species may be
few, and are comprised of species typical of regarded as endemic to the dry zones of
certain edaphic conditions or common forms northern South America. The four endemic
of communities, such as tropical deciduous species of amphibians are those that belong
and dry tropical forests. to rather widespread genera that occur both
Although there is a series of arid vegeta- north and south of the zone —
e.g., Hyla min-
Andes of Venezuela,
tion islands along the uscula, H. wandae, Physalaemus enesefae, and
Colombia and Peru, and many of these are Pipa parva; one crocodilian, Crocodylus inter-
only 200 to 300 km apart, the plant species medins; one turtle, Podocnemis vogli; eleven
show relationship to those of the south-
little lizards, Arwlis annectens, A. onca, Bachia bi-
ern South American deserts. Most of the flora color, B. guianensis, B. talpa, Qonatodes vit-
is more closely related to the arid floras of tatus, Hemidactylus palaichthus, Lepidoble-
Mexico, Central America and the western pharis sanctaemartae, Ophryoessoides eryth-
Caribbean. Furthermore, if we examine the rogaster, Phyllodactylus dixoni, Tretioscincus
plants and animals of the small, relictual, bijasciatus; nine snakes Bothrops lansbergi,
open formations within the present rainforests Crotalus vegrandis, Helicops danieli, H. sca-
and those of the pantanal of western Brasil, Leptotyphlops dimidiatus, Micrurus cir-
laris,
we are left with the impression that both the cinalis, M. dissoleucas, Phimophis guianensis,
flora and fauna once were connected across and Typhlops lehneri.
the Amazon Basin, probably during Pleisto- Based upon present distribution patterns,
cene glacial maxima. a number of species of amphibians and rep-
Herpetofaunal affinities of the northern tiles may have evolved within the dry zones
lowland dry zones of South America are at of northeastern South America and later
variance with Sarmiento's (1976) conclusions spread north and south into similar habitats
concerning the flora. Some of the amphibians when favorable climates and corridors were
and reptiles occurring in the dry forests in available. Among these are the following:
northern South America are more common in Bufo gramdosus, Hyla crepitans, Hyla rostra-
lowland rainforests. At least 14 of 36 species ta, Leptodactylus bolivianus, Pseudis paradox-
of amphibians and 44 of 108 reptiles that oc- us, Geochelone carbonaria, Ameiva ameiva,
cur in northern dry forests also inhabit Ama- A. bifrontata, Cnemidophorus lemniscatus,
zon forests. However, many of these species Gymnophthalmus speciosus, Tropidurus tor-
are fossorial, aquatic, or show wide tolerances quatus, Crotalus durissus, Mastigodryas pleei,
for ecological situations; thus it is difficult to Micrurus isozonus, Pseudoboa neuwiedi, Tan-
assign these species to a particular vegetation tilla scmicincta, and Thamnodynastes strigilis.
zone.
Many wide ranging species occur in the
Some species and genera have affinities to northeastern South American dry zones, but
Central America and to the dry zones of Bra- some {Bufo marinus, Hyla microcephala, H.
sil, whereas others are widely distributed rubra, Caiman crocodylus, Phrynolujas venu-
north and south of the dry zones of north- losa, Chrysemys scripta, Kinosternon scorpi-
eastern South America. For example, the oides, Iguana iguana, Eunectes murinus, Heli-
290 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
cops angulatus, Leptodeira annulata) are than a less diverse one, and that more species
present because of the abundance of surface of amphibians and reptiles are associated
water six to eight months of the year or be- with plant communities that are structurally
cause of prey associated with abundant heterogeneous. The structurally simple desert
moisture. scrub community of coastal Venezuela ap-
Other species (Ameiva ameiva, Thecadac- pears to contain only one reptile (Cnemido-
tylus rapicaiidus, Tupinambis teguixin, Boa phorus lemniscatus) and no amphibians. As
constrictor, Clelia clelia, Drymarchon corais, many as 48 species of amphibians and reptiles
Epicrates cenchria, Imantodes cenchoa, hep- occur in a structurally heterogeneous dry for-
tophis ahaetulla, Mastigodryas boddaerti, M.
est zone near the center of the llanos, and 75
bifossatus, Oxybelis aeneus, Spilotes pullatus, species in a moderate heterogeneous dry
Sibon nebulata, Tantilla melanocephala) are forest zone in the Yuruari Savanna (Appen-
dix 11:1).
adaptively plastic and survive in many harsh
environments. Amphibians more abundant where
are
In summary, herpetofauna of the lowland water or rainfall more constant; the same is
is
Mata, Bancos,
Deciduous Bajios &
&Gallery Trachipogon Paspalum Esteros
Species Forests Palmares Morichales Savanna Medanales Savanna Savannas
_____
Anurans
Bufo granulosus
— + — — — + +
Bufo guttatus + — — — — — +
Bufo marinus + + + + + + +
+ _
_____
Dendrobates leucomelas
Elachistocleis ovalis — + + + — + +
Hyla crepitans + + + — — + +
+ + + — — + +
-)-______
Ihjla microcephala
Hyla rostrata -j_
_
Hyla rubra + + + — — + +
Hyla wandae
Leptodactylus bolivianus + + + — — + +
Leptodactylus fuscus _ + + + — — + +
Leptodactylus macrosternon .... + + + — — + +
Leptodactylus fragdis _ .._ + + + — — + +
Leptodactylus wagneri _. + + + — + +
Physalaemus enesefae + + — — — — +
Physalaemus pustulosis _ + + — — — + +
Phrynohyas venulosa . + + + — — — +
Phyllotnedusa hypocondrialis „_+ —
—
— —
—
—
—
—
—
—
—
Pipa pipa ._ + -—
Mata, Bancos,
Deciduous Bajios &
&Gallery Trachipogon Paspalum Esteros
Species
Amphisbaenians
Amphisbaena alba
Amphisbacna fuliginosa
Lizards
Anolis auratus
Cnemidophorus lemniscatus _..
+_____
Forests
-|-
—
+
Palmares
+
+
Morichales
—
+
Savanna
+
+
_
Medanales
—
+
Savanna
+
+
Savannas
+
+
+
+
Gymnophthalmus spcciosus — — + — — — + +
+ — + — — — +
____
Hemidactylus pdlaichthus
Iguana iguana + + + — — — +
Kentropyx striatus + — + — — + +
Phyllodactylus dixoni
— — — + — —
Phyllodactylus vcntralis
_ +
Tropidurus torquatus + + + — — — +
Tupinambis tcguixin + + + + + + +
Snakes
Boa constrictor _ + + + + + + +
Chironius carinatus + + + — — — +
Corallus cnhydris + — + — — --
-)-_____
+
+
_(-_____
+
+
+
+
—
—
—
—
+
+
+
+
Leimadophis typhlus
Leptophis ahaetulla
Lygophis lincatus
Masticophis mentovarius
Mastigodryas bifossatus
.)-_____
+
—
—
—
—
—
—
—
—
—
+
+
—
—
—
—
—
+
+
+
Micrurus isozonus + + + + + + +
Oxybclis aencus
Oxyrhopus petola
Phimophis guianensis
Pscudoboa neuwiedii
Spilotcs pullatus
_)______
+
+
+
+
+
—
+
+
+
—
+
+
—
—
+
—
—
—
—
+
—
—
—
+
+
+
Thamnodynastes strigilis + + + — — + +
Comparison of lizard faunas of the low- caibo Basin; Aineiva ameiva subsp., Cnemi-
land dry forests of northern South America dophorus lemniscatus subsp., Phyllodactylus
and the upper Amazon rainforest in Peru dixoni, Crotalus vegrandis, Crocodylus inter-
(Dixon and Soini, 1975), reveals 34 and 38 medins, Podocnemis vogli to the llanos; Lepi-
species, respectively. Duellman (1978) listed doblepharis sanctaemartae, Ophryoessoides
29 species of lizards from an Amazonian site erythrogaster, Helicops danieli to the Santa
(Santa Cecilia, Ecuador); 26 of these occur Marta Region; Anolis onca and Tretioscincus
at Iquitos, Peru. Owing to several physical bifasciatus to the dry coastal areas of Vene-
and/ or climatic barriers between various zuela.
vegetation zones in the llanos and dry north- The llanos and northern dry forests con-
ern forests, local endemism is greater than in tain 56 species of snakes, whereas the upper
the upper Amazon Basin, thereby resulting Amazon region has 92 species (88 at Iquitos,
in a large total number of lizard species in 53 at Santa Cecilia). Santa Cecilia and the
the former region. As examples Pipa parua, — dry zones in northern South America share
Anolis annectens are endemic to the Mara- almost 30 percent of their species of snakes.
292 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Santa Cecilia shares 92 percent of its snake ing us the opportunity to present this paper
fauna with Iquitos, whereas Apure, Vene- on such short notice.
zuela (llanos site), shares 43 percent of its
snakes with Iquitos. At any one 10,000 km
2
= RESUMEN
200,000 km )
2
site in the llanos region (
total
only 14-17 species of snakes are present, a En presente contribution hemos tratado
la
much lower number than at given sites in the de examinar las relaciones zoogeograficas de
upper Amazon Basin. Thus the structure of la herpetofauna de las zonas bajas y secas de
the land area, complexity of vegetation com- Sur America Nor-Occidental. En area abar-
munities, and climate, seem to be limiting ca concretamente los Llanos del Orinoco y
factors for some, but not all, species. Nearly las zonas aridas y semi-aridas de las costas del
equal numbers of species of lizards occur
in
Caribe, mas de 300,000 km de
e incluye
2
vanna, and William Pyburn for information en la fauna y especialmente en los anfibios
resistir la
on anuran distribution Colombian
in the que estan pobremente dotados para
llanos. We thank William Duellman for giv- sequia prolongada.
1979 RIVERO-BLANCO & DIXON: HERPETOFAUNA OF LLANOS 293
Con el objeto de poner en claro las pre- que unas pocas son endemicas a la region.
ferencias ecologicas de las especies de anfibios Segun nuestra lista de especies y su distri-
y reptiles de la zona en estudio, hemos tratado bucion, y la diversidad de habitats accesibles
de hacer una clasificacion de habitats que esta en la zona en estudio se puede notar que el
enmarcada dentro de tipos climaticos gen- numero de especies es mayor en aquellos
erales reconocidos para la region (Ewell y habitats de mayor diversidad fisica como los
Madriz, 1968). Dentro de cada uno de los bosques, o de mayor estabilidad microclima-
tipos climaticos presentamos una lista de los tica, o donde las condiciones son mas pre-
habitats accesibles a la herpetofauna basan- decibles de ano a ario.
donos en experiencias personales y en cono- Los ciclos estacionales de gran influencia
cidas clasificaciones hechas por otros autores ecologica como el del agua en la region 11a-
(Ramia, 1967; Sarmiento, 1976). nera tienen influencia muy especial en la com-
La lista de habitats es la siguiente: I. — position faunistica, en el sentido de propor-
Bosque Seco Tropical: A) Bosques decidu- cionar un dinamismo unico que presenta un
ous, Matas y Bosques de galeria. (Figs. 11: reto para aquellas especies que no se pueden
2-11:3); B) Palmares (Fig. 11:4); C) Mori- adaptar rapidamente a situaciones extremas
chales (Fig. 11:5); D) Sabanas de Trachi- de carestia o abundancia de agua. Muchas
pogon (Fig. 11:6); E) Sabanas de Paspalum especies estan activas solo durante la parte
(Fig. 11:7); F) Sabanas de Bancos, Bajios favorable del ano y esto hace que sea dificil
y Esteros (Fig. 11:4); G) Medanales; y de observar la totalidad de las especies en un
H) Rios, cafios y lagunas. determinado periodo del ciclo de lluvias.
II. —Bosque muy Seco Tropical. Bosque En ambientes mas secos como los de las
muy seco (Fig. 11:8); Monte seco siempre- regiones xerofilas del litoral, las condiciones
verde; Monte deciduo. climaticas provocan, entre otros fenomenos,
III.— Monte Espinoso Tropical (Fig. 11: una sequia prolongada, que solo es soportada
9). por un numero escaso de reptiles y por muy
IV. — Maleza Desertica y Desierto (Fig. pocos anfibios.
11:11). La distribucion de algunos grupos, como
Presentamos una lista de anfibios y rep- largartos y serpientes, esta sujeta a diferencias
tiles en la que se seiialan las preferencias inherentes a dichos organismos en relacion a
ecologicas de las especies en relacion con los su position trofica y muy especificamente a
habitats terrestres que hemos identificado en sus requerimientos como consumidores, asi
el Bosque Seco Tropical (Tabla 11:1). El como tambien a sus requerimientos especi-
Apendice 11:1 presenta la distribucion cono- ficos en cuanto a la estructura del habitat.
cida de las especies de anfibios y reptiles con Las serpientes paracen no diferir de los
respecto a las grandes zonas consideradas en lagartos en cuanto a su adaptabilidad hacia el
el estudio. medio fisico y estructural del habitat. En
Sarmiento (1976), sugiere que el origen cambio las serpientes parecen consumi-
ser
de la flora de las regiones secas de la costa del dores terciarios que pueden explotar de ma-
Caribe es heterogeneo. Algunas especies o nera mas especializada el alimento disponible,
grupos son cosmopolitas, otros endemicos, mientras que los lagartos son mas generali-
otros vienen del norte o del sur de America y zados en sus dietas y por ello mas propensos
la mayoria ha inmigrado a la zona en dife- a competir por el recurso disponible. Debido
rente epoca y por diferentes. a dichas caracteristicas, creemos poder ex-
Los Reptiles y Anfibios de esa region pa- plicar la existencia de casi el doble de especies
recen confirmar las conclusiones de Sarmi- de serpientes en la zona estudiada.
ento en lo que respecta al origen de la flora.
La herpetofauna muestra afinidades con gru-
LITERATURE CITED
pos de origen centro-americano y tambien con
las zonas mas secas de Brasil.
Algunas espe- Arnold, S. J. 1972. Species densities of predators
cies son decididamente cosmopolitas mientras and their prey. Amer. Nat. 106:220-236.
294 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Beard, J. S. 1953. The savanna vegetation of north- patchy environment. Proc. Natl. Acad. Sci. 51:
ern tropical America. Ecol. Monogr. 23:149-215. 1207-1210.
Cochran, D., Goin, C. J. 1970. Frogs of Colombia. Medem, F. 1968. El desarrollo de la herpetologta in
Bull. U.S. Natl. Mus. 288:1-655. Colombia. Rev. Acad. Colombiana Cien. Exactas
Direccion de Geologia, MMH, Venezuela. 1970. Fis. Nat. 13:149-199.
Lexico Estratigrafico de Venezuela. Bol. Geol. Ramia, M. 1967. Tipos de sabanas en los llanos de
Publ. Especial 4, 757 p. Venezuela. Bol. Soc. Venezolana Cien. Nat. 112:
Dixon, J. R„ Soini, P. 1975. The reptiles of the 264-287.
upper Amazon basin, Iquitos region, Peru I. Liz- Rivero, J. A. 1961. Salientia of Venezuela. Bull.
ards and Amphisbaenians. Milwaukee Publ. Mus. Mus. Com. Zool. Harvard Univ. (126): 1-207.
Contr. Biol. Geol. (4):l-58. Roze, J. A. 1966. La taxonomia y zoogeografia de
Dixon, J. R., Soim, P. 1977. The reptiles of the los ofidios de Venezuela. Univ. Central Vene-
upper Amazon basin, Iquitos region, Peru II. zuela, Caracas, 362 p.
Crocodilians, turtles and snakes. Ibid. ( 12 ) 1— 91. : Ruthven, A. G. 1922. The amphibians and reptiles
Donoso-Barros, R. 1968. The lizards of Venezuela of the Sierra Nevada de Santa Marta, Colombia.
(Check list and key). Caribb. J. Sci. 8:105-122. Misc. Publ. Mus. Zool. Univ. Michigan (8): 1-69.
Duellman, W. E. 1978. The biology of an equa- Sarmiento, G. 1976. Evolution of arid vegetation in
torial herpetofauna in Amazonian Ecuador. Univ. tropical America, pp. 65-99 in Goodall, D. W.
Kansas Mus. Nat. Hist. Misc. Publ. 65:1-352. (ed.). Evolution of Desert Biota. Univ. Texas
Dunn, E. R. 1957. Contributions to the herpetology Press, Austin, 250 p.
Colombia 1943-1946. Privately printed, 296 p.
of Staton, M. A., Dixon, J. R. 1977. The herpeto-
Ewell, J., Madriz, A. 1968. Zonas de Vida de fauna of the central llanos of Venezuela: note-
Venezuela, Memoria explicativa sobre el Mapa worth records, a tentative checklist and ecological
Ecologico. MAC, Dir. Invest., Caracas, 265 p. notes. J. Herpetol. 11:17-24.
Holdridge, L. R. 1967. Life zone ecology tropical Veillon, J. P. 1963. Relation de ciertas caracter-
science center. San Jose, Costa Rica, 206 p. isticas de la masa forestal de unos bosques de
Hueck, K. 1966. Die Walder Sudamerikas. Gustav las zoneas bajas de Venezuela con el factor cli-
Fischer Verlag, Stuttgart, 422 p. matico: Humedad pluvial. Acta Cient. Venezo-
Hueck, K., Seirert, P. 1972. Vegetationskarte von lana 14:30-41.
Sudamerika. Gustav. Fischer Verlag, Stuttgart, Walter, H., Medina, E. 1971. Caracterizacion cli-
APPENDIX
Appendix 11:1. —Species
amphibians and reptiles (150) recorded from savannas and dry forests of northern
of
South America, exclusive of Ecuador and the northern Atacama Desert. Holdridge's (1967)
vegetation zone
system was utilized in determining vegetative communities. Numbers (1-5) of locality zones correspond to
locations in figure 11:1.
o ft
o
c
ft H T3 o
o
° g a
C/3 8.
Species
S o «
V 0,0 bo £
£ E
c S
o -c
o
MO <s fci
CS
fcCJ
a
Anurans
B1//0 granulosus + + + + + + +
Bufo guttatus + +
Bufo marinus — + + + + + + +
Ceratophnjs calcarata + + +
Dcndrohatcs Icucomelas .. + +
Elachistoclcis ovalis + +
llyla albomarginata +
Hyla blairi + +
Ht/Ia crepitans + + + +
Hyla egleri + +
Hyla geographica + +
Hyla kennedyi + +
Hyla leucophyllata + +
Hyla lutcocellata + +
Hyla microcephala + + + +
Hyla minuscula + +
Hyla multifasciata + +
Hyla rostrata + +
Hyla rubra + +
Hyla wandae .. + +
Hyla x-signata + +
Leptodactylus bolivianus + + + +
Leptodactylus fragilis + + + +
Leptodactylus fuscus + + + +
Leptodactylus macrosternum + + +
Leptodactylus pentadactylus + + +
Leptodactylus poecilochilus _ + +
Leptodactylus rugosus + +
Leptodactylus wagneri + + +
Phrynohyas venulosa + + + +
Physalaemus cnescfae + + +
Physalaemus pustolosus + + + +
Phyllomcdusa lujpocondrialis + +
Pipa parva + +
Pipa pipa + +
Pleurodema brachyops + + + +
Pscudis paradoxus .. + + + +
Pscudopaludicola pusilla + + +
Pseudopaludicola sagittifcr + +
Rana palmipes +
Rclictivomer pearsei + +
Sphaenorhynchus eurhostus + +
Crocodilians
Caiman crocodilus + + + + +
Crocodylus acutus + + +
296 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
p.
o
I-
H _, H
H-o
It 0) ° £
CO
t/1
Species
.5 o
a o O
u 4) Q £Q
t£ " 6
CO 2
JH 3!
U.U
Crocodylus intermedins +
Paleosuchus palpebrosus +
Turtles
Chclus fimbriatus + +
Chrysemys scripta + + +
Geochelone carbonaria + -t-
Kinostcrnon scorpioides + +
Phry nops geoffroanus +
Podocnemis erytherocephala ... +
Podocnemis expansa +
Podocnemis nnifilis +
Podocnemis vogli +
Bhinoclemys punctularia +
Amphisbaenians
Amphisbaena alba +
Amphisbaena fuliginosa _ +
Lizards
Ameiva ameiva _ +
Ameiva bifrontata +
Anolis annectens +
Anolis auratus +
Anolis gaigei +
Anolis onca _ +
Bachia bicolor +
Bachia guianensis +
Bachia heteropa + +
Bachia talpa +
Basiliscus basiliscus +
Cercosaura ocellata +
Cnemidophorus lemniscatus + + + +
Gonatodes albogularis + + +
Gonatodes vittatus _ + + +
Gymnophthalmus speciosus + +
Gymnophthalmus underwoodi . +
Hemidactylus brooki .... +
Hemidactylus palaichthus +
Iguana iguana + + + +
Kentropyx borkianus +
Kentropyx striatus +
Lcpidoblcpharis sanctaemartae +
Leposoma rugiceps +
Mabuya mabouya +
Ophryoessoides crythrogaster ... +
Pliyllodactylus dixoni +
Phyllodactylus ventralis ... + + +
Polychrus marmoratus +
Pseudo gonatodes lunulatus _ +
Sphaerodactylus molei +
Thecadactylus rapicaudus + + +
1979 RIVERO-BLANCO & DIXON: HERPETOFAUNA OF LLANOS 297
Species
298 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Species
12. Composition, Distribution y Origen de la
Herpetofauna
Chaqueiia
Jose M. Gallardo
Museo Argentino de Ciencias Naturales
Avenida Angel Gallardo 470
Buenos Aires, Argentina
fauna chaqueiia se tienen a traves de los re- Chaco. Cubre los departamentos de Boque-
lators y libros de los sacerdotes misioneros de ron y Olimpo de Paraguay, al oeste del Rio
la epoca colonial (Siglos XVII-XVIII). Los Paraguay. En la Argentina abarca toda la Pro-
Padres Ruiz de Montoya, Lozano, Paucke, vincia de Santiago del Estero y penetra en los
Juarez, Sanchez Labrador (vease referencias valles del este de Jujuy, el este de Catamarca
en Gallardo, 1961b) tienen descripciones y a y de La Rioja y el norte de San Luis (Fig.
veces grabados sobre diversas especies cha- 12:1). Presenta areas ecotonales o de transi-
queiias de ofidios, saurios, cocodrilos, y tor- tion con otras faunas penetrando en el Para-
tugas, con informaciones por observation guay hacia el este. En Brasil algunos elemen-
directa y otras referencias de los indigenas tos faunisticos del Chaco llegan al Mato
del "Gran Chaco Gualamba." Algunos natur- Grosso, al sudeste y al nordeste a favor del
alistas viajeros como d'Orbigny (1847) tam- Cerrado y la Caatinga; Uegando aun hasta
bien se refieren a algunos de los anfibios y las Guayanas, y algunos elementos alcanzan
el este de Salta hasta la mitad de Formosa, o practicamente el doble, entre 1000 a 1460
299
300 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Map showing the area occupied by the Chacoan herpetofauna and the transition zone with the Litoral-
Mesopotamian fauna. Capital cities and principal rivers are shown.
mm, y en el area subandina son menos de que chaqueno, con bosques y arboles aislados
la mitad que en la chaquena (160-325 mm). (Acacia, Prosopis, ScJtinopsis, Aspklospcrma)
De modo que existe un escalon pluviometrico en pastizales de gramineas (Setaria, Digi-
coincidente en general con el topografico. taria, Trichloris) ;
a veces se trata de palmares
En cuanto a las temperaturas anuales hay ( de matorrales espi-
Tritrinax, Copernicia ) o
diferencia entre el area ocupada por la fauna nosos ultimos, en areas
(en especial, estos
de transition y la ocupada por la litoral-meso- degradadas), otras son areas de vegetation
potamica, de la mitad de Santa Fe hacia el halofila o falta la vegetation casi totalmente
sur. Asi en la primavera hay cinco meses de en detenninados espacios. Corresponde basi-
posibles temperaturas bajo cero (entre mayo camente a la Provincia Chaquena del Dominio
y septiembre inclusive), con temperaturas Chaquerio (Cabrera, 1971), la
herpetofauna
minimas de hasta —5° y — 6°C. En la mitad chaquena tambien puede extenderse por las
sur de Santa Fe hay seis o siete meses de provincias del Monte y del Espinal. No
posibles temperaturas bajo cero (de abril o existe,en cambio, en ambientes puramente
mayo a octubre inclusive), lo que amplia herbaceos como los del sur de Santa Fe y de
considerablemente el esquema anterior, ade- Cordoba, y la Provincia de Buenos Aires, que
mas minimas pueden alcanzar los — 10°C.
las corresponden a la llamada "Pampa humeda."
Esto se halla ligado a un cambio faunistico, Es alii donde vemos que existe una com-
que coincide con la desaparacion hacia el sur ponente climatica y fitogeografica que no
de la fauna herpetologica chaquena. favorece el avance de la herpetofauna cha-
En la region considerada domina el par- quena hacia el sur.
1979 GALLARDO: HERPETOFAUNA CHAQUENA 301
Tabla 12:1. — Lista de los Anfibios y Reptiles Chaquenos y Sus Distribuciones en Regiones Ultracbaqueiias.
( + = presente;
— = ausente; X = representado por otra subespeeie;
* =
basicamente endemica de Chaco)
Litoral- Region
Especies Mesopotamica Pampas Patagonia Subandina
Batracios
Ceratophrys pierotti"
Ceratophrys sp.°
Lepidobatrachus asper"
Lepidobatrachus llanensis" ._
Lepidobatrachus salinicola" ...
Leptodactylus anceps"
Leptodactylus bujonius"
Leptodactylus chaquensis" ...
Leptodactylus gualambcnsis"
Leptodactylus laticeps" .
Odontophrynus americanus ._ + +
Physalaemus albonotatus"
Physalaemus biligonigerus + +
Pleurodema borellii"
Pleurodema guayapae +
Pleurodema tucumana"
Bufo arenarum X X +
Bufo granulosus ~ X X
Bufo paracnemis _
Melanophryniscus stelzneri ... X X
Pseudis paradoxus X X
Hyla acuminata''
Hyla fuscovaria"
Hyla raniceps
Hyla x-signata
Phrynohyas venulosa —
Phyllomedusa hypocondrialis
Phyllomedusa sauvagii"
Elachistocleis bicolor
Dermatonotus mulleri"
Saurios
Homonota horrida"
Phyllopezus pollicaris"
Leiosaurus paronae"
Liolaemus chacoensis"
Ophryoessoides caducus"
Pristidactylus vautieri"
Proctotretus doellojuradoi" ...
Tropidurus spinulosus
Tropidurus sp.°
Cnemidophorus Icachi"
Gymnophthalmus rubricauda"
Kentropyx lagartija"
Kcntropyx viridistriga"
Teius teyou
Tupinambis rufescens
Ophiodes intermedins +
Mabuya frenata +
Anfisbenios
Amphisbacna camura"
Anops kingii + + +
Lcpostcrnon microccphalum . + +
Ofidios
Leptotyphlops albipuncta" ...
Leptotypldops unguirostris +
Leptotyphlops weyrauchi"
Constrictor constrictor + + +
Epicrates cenchria _ +
1979 GALLARDO: HERPETOFAUNA CHAQUENA 303
Litoral- Region
E species Mesopotamica Pampas Patagonia Subandina
Eunectes notaeus 1-
Clclia clclia +
CIclia occipitolutea .
Elapomorpltus tricolor
Leimadophis sagittifer
Ltjsirophisdorhignyi _
Lystrophis semicinctus
Oxyrhopus rhombifer
Philodryas aestivus _
Philodryas baroni
Philodryas patagonicnsis ...
Philodryas psammophideus
Phimophis vittatus ...
Pseudotomodon trigonatus .
Quelonios
Kinosternon scorpioides
Gcochelone petersi"
Cocodrilos
Caiman latirostris
304 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fig. 12:2. Esquema de la utilization del habitat de algunos saurios y anfisbenios en la zona norte de
Cordoba y sur de Santiago de Estero, Argentina.
Sketch of the habitat utilization by some lizards and amphisbaenians in the area north of Cordoba and south
of Santiago del Estero, Argentina.
1. Tropidurus spinulosus, Tropidurus sp., 2. Homonota horrida, 3. Leiosaurus paronae, 4. Pristidactylus
vautieri, Ophiodes intermedins, 6. Amphisbaena camura, Anops kingii, 7. Tupinambis rufescens, Teius
5.
Tabla 12:2.- -Especies y Subespecies Vicariantes o Cercanas Entre si en las Faunas Chaquefia y Litoral-
Mesopotamia.
Chaqi Litoral-Mesopotamia
Ceratophrys sp. Ceratophrys ornata
Leptodactylus anceps Leptodactylus prognathus
Leptodactylus bufonius Lep todactylus m y statin us
Leptodactylus chaquensis Leptodactylus ocellatus
Leptodactylus gualambcnsis Leptodactylus gracilis
Bufo arcnarum chaguar Bufo arcnarum platensis
Bufo granulosus major Bufo granulosus fernandczac
Pseudis paradoxus occidcntalis Pseudis paradoxus platensis
Liolaemus chacoensis Liolacm us wiegmannii
Cnemidophorus leachi Cnemidophorus lacertoides
Teius teyou cyanogaster Teius teyou teyou
Tupinambis rufescens Tupinambis tequixin
1979 GALLARDO: HERPETOFAUNA CHAQUElVA 305
anteriores (Gallardo, 1965b, 1969b) se de- Aires (Monte Hermoso), de acuerdo a los
scriben 14 subespecies para Bufo granulosus trabajos paleontologicos de Tonni (1974) y
(Fig. 12:3), correspondientes basicamente a de Chani (1977), pues segun dichos autores
los refugios que otros autores han asignado esto se deberia a que habria habido fluctua-
para diversas especies de animales (Haffer, ciones paleoecologicas de climas secos y
1969, 1974). Sin embargo, al mismo tiempo climas humedos entre el Plioceno superior y
esas subespecies se extienden por los sistemas la actualidad. Vease Baez y Scillato Yane
hidrograficos correspondientes; a la fauna (este volumen) para una exposition com-
chaquena corresponde Bufo granulosus major. pleta de los paleoclimas de la region.
En otro trabajo sobre Pseudis paradoxus En general se nota actualmente un em-
(Gallardo, 1961c) tambien se encuentra una pobrecimiento faunistico hacia el sur, con re-
coincidencia de este tipo, correspondiendo a specto a la fauna de anfibios del sudeste de
la fauna chaquena, Pseudis paradoxus occi- Brasil y la Provincia de Misiones, Argentina,
dentalis. en la fauna litoral-mesopotamica. Es asi como
Los ofidios son de mas amplia distribution en la Provincia de Buenos Aires de 22 especies
306 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fernandez, J. 1976. Hallazgo de peces pulmonados ciones sobre Ios saurios de dicha provincia argen-
fosiles en la puna Ann. Soc. Cient. Argen-
jujeria. tina y del oeste de Buenos Aires. Neotropica 12:
tina 201:1.3-18. 13-29.
Gallardo, J. M. 1951. Sobre un Teiidae (Reptilia, Gallardo, J. M. 1966b. Las especies argentinas del
Sauna poco conocido para la fauna argentina.
) genero Ophiodes Wagler. Rev. Mus. Argent.
Com. Mus. Argent. Cienc. Nat. Bernardino Riva- Cienc. Nat. Bernardino Rivadavia Inst. Nac. In-
davia Nac. Invest. Cienc. Nat. Zool. 2:1-8.
Inst. vest. Cienc. Nat. Zool. 9:123-146.
Gallardo, J. M. 1953. El escuerzo como animal de Gallardo, J. M. 1966c. Zoogeografia de los anfibios
terrario. Ichthys 1:75-79. chaquenos. Physis 26:67-81.
Gallardo, J. M. 1957. Las subespecies argentinas Gallardo, M.
1968a. Relaciones zoogeograficas
J.
de Bufo granulosus Spix. Rev. Mus. Argent. de la fauna batracologica del oeste de la Provin-
Cienc. Nat. Bernardino Rivadavia Inst. Nac. cia de Santa Fe (Argentina). Com. Mus. Argent.
Invest. Cienc. Nat. Zool. 3:337-374. Cienc. Nat. Bernardino Rivadavia Inst. Nac. In-
Gallardo, J. M. 1959. Sobre un Iguamdae del vest. Cienc. Nat. Ecol. 1:1-13.
noroeste argentino, Leiocephalus caducus ( Cope ) .
Gallardo, J. M. 1968b. Las especies argentinas del
Acta. Zool. Lilloana 17:485-497. genero Mabuya Fitzinger Scincidae, Sauria).
(
Gallardo, J. M. 1961a. Estudio zoogeografico del Rev. Mus. Argent. Cienc. Nats. Bernardino Riva-
genero Lciosaurus (Reptilia, Sauria). Physis 22: davia Inst. Nac. Invest. Cienc. Nat. Zool. 9:177-
113-118. 196.
Gallardo, J. M. 1961b. Panorama zoologico argen- Gallardo, J. M. 1968c. Sobre la validez de algunas
tino: Batracios y Reptiles. Ibid. 22:171-180. especies argentinas de Pleurodcma ( Anura, Lepto-
Gallardo, J. M. 1961c. On the species of Pseudidae dactylidae). Physis 28:135-144.
(Amphibia, Anura). Bull. Mus. Comp. Zool. Gallardo, J. M.
1969a. Especies de saurios (Rep-
Harvard Univ. 125:111-134. tilia) de Provincia de Santa Fe, Argentina y
la
The word Patagonia is derived from the longed erosion. Scattered through the region
term "Patagones," meaning big-legged men, are extensive areas of extrusive basaltic rocks.
applied to the tall Tehuelche Indians of The open landscape is dissected by transverse
southernmost South America by Ferdinand rivers descending from the snowy Andean
Magellan in 1520. Subsequently, this pic- cordillera; drainage is poor near the Atlantic
turesque name came to be applied to a con- coast. Patagonia is subjected to severe sea-
spicuous continental region and to its biota. sonal drought with about five cold winter
Biologically, Patagonia can be defined as months and a cool dry summer, infrequently
that region east of the Andes and extending
interrupted by irregular rains and floods.
southward to the Straits of Magellan and
eastward to the Atlantic Ocean. The northern
boundary is not so clear cut. Elements of the HISTORY OF THE PATAGONIAN BIOTA
Pampean biota penetrate southward along the
coast between the Rio Colorado and the Rio In contrast to the present, almost uniform
Negro (Fig. 13:1). Also, in the west Pata- steppe associations in Rio Negro, Chubut,
gonian landscapes and biota enter the vol- and Santa Cruz provinces, during Oligocene
canic regions of southern Mendoza, almost and Miocene times tropical and subtropical
reaching the Rio Atuel Basin. The Pata- vegetation occurred along with xerophytic
gonian region has a wide ecotonal zone with woodlands with luxuriant mesophytic gallery
the Chacoan region (Gallardo, this volume). forests. A
comparison of the rich Miocene
The monte vegetation (Morello, 1958) with Leufu, Rio Negro (Berry, 1938)
flora of Pichi
its several formations containing numerous with analagous associations from Mirhoja,
subtropical elements extends south to the Chubut; Valcheta, Rio Negro; and Rio Chalia,
Peninsula de Valdes; the monte enters the Santa Cruz, shows a mixture of mesic tropical
Rio Chubut drainage and extends westward elements (Ficns, Fagara, Nectandra, Tabe-
to the Rio Neuquen, Rio Agrio, and Rio buia, Mijristica, Sterculia, tree ferns, Erythro-
Limay
valleys. South of the Rio Negro, the monte xylon, Oreopanax, Maytenus), including
associations exist in a system of saline low- climbers (Buettneria, Banisteria, Bignonia,
lands (bajos) and reach irregular spurs of Cissus, Paullinia, Sapindus, Strychnos), to-
the Meseta de Somuncura, a typical Patagon- gether with nontropical genera (Araucaria,
ian environment (Cei, 1969a,b; Ruiz Leal, Azora, Berberis, Ginkgo, Laurelia, Emboth-
1972). Nevertheless, there is a general, some- rium, Fitzroya, Libocedrus, Podocarpus, Lo-
times remarkable, agreement between the matia, Peumus, Myrceugenia, Drimys). Most
phytogeographic boundaries of the Monte- of the latter are characteristic components of
Pampean and the Patagonian regions and the the present temperate Valdivian forest. Nev-
distribution patterns of their herpetofaunas. ertheless, xeric areas in the Middle Tertiary
Herein I emphasize the biota of the Cis- of Patagonia are suggested by certain paleo-
Andean steppe to the near exclusion of the floras containing Schinopsis, Schinus, and Cu-
Trans-Andean austral forest ecosystems pania. The former is a significant genus of
treated by Formas (this volume). trees in the subtropical Chacoan region.
Patagonia is a region of sedimentary rocks Nothofagus forests were widespread in
and soils, mostly tablelands subjected to pro- Patagonia in the Eocene and Oligocene, but
309
310 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
&0,
/Canquel
Meseta
SOUTHERN TEMPERATE
V PHYTOGEOGRAPHICAL REGIONS
Greaf~Central ^V
1979 CEI: PATAGONIAN HERPETOFAUNA 311
400 KM
TERTIARY PATAGONIAN
FLORAS AND HERPETOFAUNA
Miocene Patagonian
Tropical and Valdivian
Floral Remains
Oligocene Eupsophus
and Neoprocoela
[Scarritt Pocket)
Miocene Ceratophrynid
Wawelia
Oligo-Miocene
I Patagonian Caudiverbe ra
Fig. 13:2. Paleontological records of the lower Tertiary Patagonian flora and of leptodactylid frogs.
Hallazgos paleontologicos de flora patagdnica del Terciario inferior y de anuros leptodactylidos.
312 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
these associations decreased and retreated history of tropical elements in Patagonia. The
southward and westward in the Middle and snake Dinilysia patagonica from the
fossil
Late Tertiary. Chusquea bamboo groves oc- Upper Cretaceous of Neuquen is related to
cur in Cenozoic deposits at Laguna Hunco, boids and aniliids that are widespread in trop-
Chubut. Further evidence of paleoclimatic ical South America. Furthermore, boid snakes
conditions in Patagonia is derived from the (Madtsoia), crocodilians (Necrosuchus, Se-
extensive paleo-mammal faunas ( Raez and becus, Eocaiman), and meiolaniid and pelo-
Scillato Yane, this volume) and more limited medusid turtles from Paleocene-Eocene de-
paleo-herpetofaunas (Raez and Gasparini, posits in Chubut are indicative of tropical
this volume). environments (Gasparini and Raez, 1975).
Primitive leptodactylid, ceratophrynine, Iguanid (Erichosaurus debilis) and teiid (Di-
and bufonid frogs have been recorded by asemosaurus occidentalis) lizards lived in
Schaeffer (1949), Chaffee (1952), and Casa- southern Santa Cruz in the Miocene.
Northern or Ancient
Patagonian Major
Animal Habitat
500 KM
Shield
Shield
Francisco
sin
tan
CONTINENTAL
GEOTECTONIC UNITS
Deseado
"x Massif
1979 CEI: PATAGONIAN HERPETOFAUNA 315
NORTHERN OR
ANCIENT MAJOR ANIMAL
HABITAT
SOUTHERN OR
SANTA CRUZ MAJOR
ANIMAL HABITAT
800 KM
Fig. 13:5. Location of the southern massifs and their relation to the major Patagonian herpetofaunal
regions.
Ubication de los macizos aust rales y su relation con las regiones herpctofaunisticas fundamentales
patagonicas.
316 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Tertiary age along the Andean front; their Herbs Spiny or Sclerotic Plants
small lakes are close to the last eastern patches Stipa Chuquiraga
Festuca Nassauvia
of Nothofagus and Araucaria forests. Along Poa Ephedra
the base of the Andes there is an Austral- Senecio filaginoides Styllingia
Patagonian ecotonal zone; however, it is Grindelia chilocnsis Verbena
much narrower Verbena ligustrina Pantacantlia
than the Monte-Patagonian
Acaena caesiritosa Adesmia
ecotone extending from the Rio Colorado to Shrubs Austrocactus
Valcheta, onto the northern spurs of the Me- Mulinum spinosum Hydrophyllic Plants
seta de Somuncura, and southward to near Colliguaja integerrima Juncus
Bcrbcris cuncata Carex
Bahia Camarones on the Atlantic coast (Ca-
Lijcium tenuispinosurn Ranunculus
brera, 1951). Anarthrophijllum rigidum Hi/psela
Floristically, theancient Patagonian re- Anarthrophyllum Plagioboth rys
gion is characterized by a predominantly desideratum Acaena macrostemon
Trcvoa spinifer Caltha
steppe vegetation with scattered green bushes Prosopis patagonica Cortaderia
(Mulinum spinosum), several grasses, low Larrea ameghinoi AzoreUa
herbaceous plants, some spiny plants, and a Halophyllic Plants
variety of low bushes; additional kinds of Atriplex
Frankenia
plants are present in saline environments and Spartina
riparian situations (Table 13:1). The most
characteristic element of the monte forma-
tions the creosote bush (Larrea), which is
is rugged southernmost Andean cordillera. The
Andes are commonly bordered by sharp-
represented by five sympatric species in the
ecotonal zone at Valcheta (Rio Negro). In edged basaltic mesetas having elevations of
1000 to 1500 m.
typical Patagonian landscapes only the low
Larrea ameghinoi is present. Phytogeographically, the Santa Cruz
The cool Patagonian steppes dominated by Faunal Region agrees with Cabrera's ( 1951 )
Mulinum and Stipa, with scattered creeping, —
Patagonian districts Patagonico Subandino,
cushionlike plants, exist in western Neuquen, Patagonico Central, and Golfo de San Jorge.
southwestern Rio Negro, and in most of Chu- The Patagonico Subandino includes the ba-
In these areas the steppes are saltic mesetas (e.g., Meseta Vizcachas, Me-
but, provinces.
seta Asador, Meseta de la Muerte, and the
commonly associated with basaltic landscapes
Meseta de Lago del Sello) and the southern
resulting from the rampant Cenozoic volcanic
The steppes are discontinuous in humid lowlands. 1
In these areas open steppe
activity.
northern Neuquen and Rio Negro provinces,
associations of Festuca monticola, Bromus
where they occur mostly at elevations of more macranthus, Hordeum cornosum, Agropyron
than 900 to 1000 m. magellanicum, Poa sp., and Dcschampsia sp.
The southern or Santa Cruz Faunal Re- predominate, but some shrubs (Bcrberis
cuncata, Nassauvia aculeata, or Mulinum spi-
gion extends from about 45°S to the Straits
of Magellan (Fig. 13:7). This region encom- nosum) are present.
passes some distinct physiographic areas. The Phytogeographic differences between the
arid valley of the Rio Deseado borders the Sub-Andean district and the central and San
northern limits of the large Altiplanicie Cen- Jorge districts are evident by the monotonous
tral, a dead volcanic landscape with scattered grasslands of Stipa humilis in the latter. The
clay basins and petrified early Cenozoic trees grasses are interrupted by the broad circular
(Auracarites). South of the great plateau the
bushes of the blackish "mata negra" (Verbena
drainage basins of the rios Chalia, Santa Cruz, 1
Coyle, and Gallegos provide more moist low- Although the Sub-Andean District is considered to
be a single physiographic unit (Fig. 13:7), for pur-
lands extending to the Straits of Magellan.
poses of herpetofaunal analysis, I distinguish the
These rivers drain the glacial valleys of the Humid Southern Lowlands.
1979 CEI: PATAGONIAN HERPETOFAUNA 317
—
Table 13:2. Comparison of the Herpetofaunas in Ten Districts in Patagonia.
( Numbers of species is a given district are in boldface; numbers of species in common to two districts are in
Roman, and the italics are Faunal Resemblance Factors [N, + N2 /2C (Duellman, 1966)].
318 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
400 KM
I
HERPETOFAUNAL AND
PHYSIOGRAPHIC DISTRICTS
Ancient Patagonian
Steppe Climax
Patagonian
Phytogeographical Range
Monte
Fig. 13:6. Herpetofaunal and physiographic districts of the ancient Patagonian region.
Distritos herpetofaunisticos ij fisiogrdficos de hi region patagdnica antigua.
1979 CEI: PATAGONIAN HERPETOFAUNA 319
Fig. 13:7. Herpetofaunal and physiographic districts of the southern Patagonian region.
Distritos herpetofaunisticos y fisiograficos de la region sur-patagonica.
320 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
PERCENTAGE HERPETOLOGICAL
ELEMENTS IN FAUNAS
z
<
E
Z
<
1979 CEI: PATAGONIAN HERPETOFAUNA 321
Fie. 13:9. Southern limits of distribution of taxa in the monte formation in northern Patagonia. Areas of
sympatry of Pleurodema bufonina and P. thaul are indicated.
Limites mcridionales de distribution de taxa en la formation del monte en el norte patagonico. Se indi-
ca las areas de simpatria de Pleurodema bufonina y P. thaul.
neuquensis from the thermal brooks on the Luan. In other ecotonal and subandean areas
sandy Meseta Lonco Luan, 1500 m elevation of Neuquen, Pleurodema thaul, Liolaemus
in Neuquen (Cei, 1976), they belong to a chilensis, and L. buergeri are found. The lat-
transitional herpetofauna of the austral-Pata- ter occurs sympatrically with the typical Pata-
gonian ecotone. Other transitional species are gonian lizards Liolaemus elongatus and L.
the frog Pleurodema thaul and two lizards, kriegi. Likewise, the characteristic lizards of
Liolaemus tenuis and Diplolaemus leopardin- rocky Patagonian communities, Phymaturus
us, all characteristic inhabitants of Araucaria palluma and P. patagonicus zapalensis occur
forests (Cei, 1970a, 1974b). On the xeric on the basaltic plateaus. Phymaturus palluma
Meseta de Lonco Luan, dead patches of extends northward in the Andes at elevations
Nothofagus and Chusquea exist near the bor- of 3000 to 3500 mto La Rioja and San Juan,
der of the Araucaria forests. Liolaemus lineo- and P. patagonicus has distinct populations
maculatus occurs on the Meseta de Lonco (P. p. payuniae and P. p. nevadoi) to the
322 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Tentative Boundary
of So. Chacoan Province or Monte
*1 Tentative Distribution of
. Liolaemus darwlni
500 KM
Fig. 13:10. Distribution of Liolaemus darwinii, a characteristic species of the monte formation. The south-
ern edge of this formation is the southern limits of distribution of Geochelone and of colubrid snakes in north-
ern Patagonia.
Distribucion de Liolaemus darwinii, especie caracter islica dc la formation del monte. El limite meridional
de esta formacion es el limite meridional de Geochelone y de los ofidios colubridos en la Patagonia septentrional.
1979 CEI: PATAGONIAN HERPETOFAUNA 323
casuhatiensis
B Tentative Oistributio
of Di plolaemus
1 pata q onicus
2 indigtinplug
3 somuncurensis
4 2apalensis
5 pa yuniae
6 nevadoi
Tentative Distribution ot
Leiosaurus belli!
500 KM
A nitoi
A solitarius
A reverberii
Alsodes sp
Somuncuria somuncurensis
north in southern Mendoza Province (Cei demic subspecies of lizards. The endemic
and Castro, 1973; Cei and Roig, 1975) (Fig. Liolaemus ruizleali inhabits rocky summits of
13:12). the meseta at 1200 to 1700 m. The two en-
Theisolated Meseta de Somuncura (150 demic, telmatobiine leptodactylid frogs have
X 80 km), with elevations to 1700 m, has unique morphological and ecological traits.
some peculiar habitats. Between 800 and The monotypic Somuncuria somuncurensis
1700 m, above the Monte-Patagonian ecotone, lives in streams at 800 to 1000 m issuing from
11 species amphibians and reptiles are
of thermal springs on the northeastern slopes
known. Six of the species are
widespread in of the meseta. Inhabiting the same streams
—
Patagonia Pleurodema bufonina, Homonota having a temperature of about 18°C is the
darwinii, Diplolaemus darwinii, Liolaemus endemic characid fish Gymnochacinus bergi.
bibronii, L. bendengeri, and L. rothi. Three Conversely, Atelognathus reverberii, a nearly
other species and two subspecies are endemic fossorial frog, inhabits the arid plateau at
— Phymaturus patagonicus somuncurensis and elevations of more than 1000 m
and breeds in
Liolaemus elongatus petrophilus are the en- small temporary pools.
326 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Southern Patagonian Herpetofauna narum, and all of the genera of snakes are
more northern groups.
The distributions of the 11 lizards, one
Bufo spinidosus and Alsodes are primarily
snake, and one frog are examined with re- Andean groups. The former enters Patagonia
spect to five ecophysiographic areas (Fig. in many disjunct valleys; the species does not
13:7)—1) Altiplano Central, 2) Coastal Dis- occur on major basaltic mesetas. In Pata-
trict, 3) Humid Southern Lowlands, 4) Sub-
gonia, Alsodes occurs only in the volcanic
Andean Area, and 5) Meseta de Lago del
highlands adjacent to the Andes.
Sello. The distributions of the species and
Pleurodema is a primitive leptodactyline
subspecies of amphibians and reptiles in these that may have originated in, and dispersed
five areas are tabulated in Appendix 13:1,
from, the austral forests (Duellman and Ve-
and the distributional relationships between
loso,1977); the genus has dispersed north-
the areas are analyzed in Table 13:2. ward in nonforested habitats to the Carib-
Nine species occur in the Altiplano Cen- bean. Two species are peripheral in Pata-
tral; all but the frog Pleurodema bufonina —
gonia P. thaul in the austral forest Pata- —
also occur in the coastal district. Only six spe-
gonian steppe ecotone and P. nebulosa in the
cies occur in the humid southern lowlands; monte. Pleurodema bufonina is a widespread
the southernmost frog, Pleurodema bufonina, species endemic to Patagonian habitats. Al-
is rather rare there. The lizard Liolaemas though and bufonina are distinctive
P. thaul
magellanicus, the only herpetofaunal species in their morphology and behavior, popula-
on Tierra del Fuego, is common on the south- tions intermediate between the species exist
ern end of the mainland. Only four species in high valleys in western Neuquen.
(
all lizards )
occur in the sub-Andean region. Lynch (1978) provided an hypothesis for
The Meseta de Lago del Sello at 47°S is the evolution of lower telmatobiine frogs in
nearly 50 km in diameter. On the top of the Patagonia. The peculiar monotypic Somun-
plateau, low grasses (Festuca, Poa, and curia is endemic to the Meseta de Somun-
Stipa) are dominant with thorny plants or cura, whereas five species of Atelognathus
cushion plants (Benthamiella azorella. Ver- occur in isolated basaltic areas in Patagonia.
bena, Senecio, Nassauvia); lichens are abun- The only extra-Patagonian Atelognathus is
dant in rocky areas. Owing to the proximity A. grandisonae from Puerto Eden in extreme
of the great Patagonian Ice Field, cold winds southern Chile.
whip the plateau, even in summer. Only three Among the iguanid lizards, the monotypic
iguanid lizards have been found on the pla- Vilcunia has characters of both Proctotretus
teau ( 1200-1600 m ) the two species of Lio-
; and Liolaemus and is endemic to southern
laemus are widespread in southern Patagonia, Patagonia (Donoso Barros and Cei, 1971).
whereas the monotypic Vilcunia sylvanae is Diplolaemus and Phymaturus are fundamen-
endemic. No amphibians have been found tally austral genera (Figs. 13:11-12). Cten-
on the plateau; the widespread Pleurodema oblcpharis, Leiosaurus, Proctotretus, and Pris-
bufonina ascends the slopes to only 900 m. tidactijlus are widely distributed to the north
of Patagonia (Cei, 1973c,d). Two species of
Proctotretus are distributed in temperate
ORIGIN OF THE HERPETOFAUNA areas in southern Brasil, Uruguay, and cen-
tral Argentina; P. pectinatus occurs in the
The Patagonian herpetofauna is distinc- monte-Patagonian ecotone. Pristidactylus fas-
tive in the diversity of telmatobiine leptodac- ciatus is primarily Patagonian, but congeners
tylidfrogs and tropidurine iguanid lizards. occur in disjunct Andean areas (P. scapula-
Most other species inhabiting Patagonia are tus) and in isolated extra- Andean massifs of
members of groups that are mainly distrib- central Argentina —
Sierra Grande de Cordoba
uted to the north of Patagonia. Thus, Geo- (P. achalensis) and Sierra de la Ventana (P.
chelone, Cnemidophorus, Mabuya, Hornono- casuhatiensis) (Gallardo, 1964, 1968; Cei and
ta, Leptodactylus, Odontophrynus, Bufo are- Castro, 1975).
1979 CEI: PATAGONIAN HERPETOFAUNA 327
Three major points relative to the origin tered relicts, unique witnesses to some of the
and evolution of the Patagonian heipetofauna most ancient steps in the history of continental
need to be emphasized. vertebrates, whereas the Recent speciation of
1. There has been a radiation of Liolaemus attests to the continued evolu-
primitive
leptodactylid frogs, remnants of Gond- tionary activity in the region.
wanan elements. Vuilleumier (
196S )
and Lynch (1971, 1978) noted the aus-
Evolutionary Radiation of
tral center of radiation of telmatobiine
Patagonian Liolaemus
leptodactylids in Patagonia and the
austral forests; Formas (this volume) Lizards of the genus Liolaemus are wide-
discussed the biogeography and ecol- spread in temperate South America. Three
ogy of the telmatobiines of the austral species range into southern Rrasil, and several
forests —Abodes,
Batrachyla, Caudi- species occur in the Andes, two extending
verbera, Eupsophus, Hylorina, Insueto- northward to central Peru
(Duellman, this
phrynus, and Telmatobnfo, some of volume). Twenty-six taxa are Patagonian or
which also occur in Argentina (Cei, Andean-Patagonian.
1978). The limited
paleontological Four major evolutionary units can be rec-
evidence supports the austral center of ognized among the Patagonian Liolaemus as
radiation (Schaeffer, 1949; Chaffee, follows: 1) L. fitzingeri complex, 2) L. elon-
1952; Casamiquela, 1963; Estes and gatus-kriegi complex, 3) L. kingii-archeforus
Reig, 1973). complex, and 4) L. magellanicus-lineomacu-
2. An austral South American center of latus complex (Cei, 1971a, 1972a, 1973a,b
evolution and adaptive radiation of an 1974a, 1975a,b, 1975d,e Cei and Scolaro,
;
ancestral stock of iguanid lizards is evi- 1977; Scolaro and Cei, 1977). The lizards that
dent (Cei, 1973c,d, 1975c; Cei and are not members of these groups are pri-
Castro, 1975 ) Fourteen genera are aus-
.
marily peripheral to Patagonia and/ or are
tral in Argentina and Chile; seven of ecotonal elements. Some of these are mem-
these are Patagonian. There are some bers of transcordilleran —
groups Liolaemus
20 genera of iguanids in tropical South altissimus, chilensis, cyanogaster, lemniscatus,
America and another ten genera in the pictus, and tenuis; L. bibronii is related to
Sonoran region of North America. With the Chilean L. fuscus. Liolaemus boidengeri,
the exception of Anolis and Sceloporus, darwinii, and gracilis are members of the
no other iguanid genus displays such more northern monte fauna and are primarily
an impressive adaptive radiation as does peripheral in Patagonia (Fig. 13:10).
Liolaemus, the dominant lizards in any Liolaemus fitzingeri complex. This group—
Patagonian community. In most of is characterized by 1) patch of enlarged
these same communities there exist rep- scales on posterior surfaces of thighs, espe-
resentatives of the other well developed in males (Fig. 13:14);
iguanid genera —Diplolaemus,Patagonian
Leiosau-
cially
2) high number (52-82) of blunt, slightly
rus, Phymaturus, Pristidactylus, Procto- keeled scales around body; 3) high number
tretus, and Vilcunia. (7-11) of preanal pores; 4) stout body and
3. An impressive post-Pleistocene adaptive relatively long tail, 1.5 times length of body;
radiation has taken place in four groups 5) tendency to have black venters and dark
of Patagonian Liolaemus (see following humeral 6) predominate dorsal color
collars;
section ) .
patterns consisting of wide transverse dark
Thus, far from being a totally barren re- blotches, bordered posteriorly by white, but
gion biologically, irregularly colonized by ele- spotted erythristic, and melanistic variations
ments from neighboring biotas, Patagonia has not uncommon.
been, and apparently still is, a center of active Content: L. fitzingeri canqueli, L. fitzin-
speciation of several hcrpetofaunal elements. geri fitzingeri, L. fitzingeri melanops, L. rothi,
The old radiations are supported by the scat- ?L. ruizleali. Liolaemus f. fitzingeri is the
328 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
\UiMW
52-82
7-11
anal pores
TENTATIVE DISTRIBUTION OF
PATAGONIAN
LIOLAEMUS FITZINGERI COMPLEX
Tentative N. and S.
Liolaemus ruizleali
?r£^ Endemic to Somuncura Mts.
72-120
anal pores
500 KM
TENTATIVE DISTRIBUTION OF
PATAGONIAN LIOLAEMUS
SPECIES GROUPS
pr-vy.'j;
;•,;;:' elongatus krie qi Complex
-
maoellanicus - lineomaculalus
Complex
58-84
6-10
anal pores
40-70
LJK
anal pores
ques mesofiticos de galena durante el Oligo- En los depositos del Oligoceno y del Mio-
ceno y el Mioceno. Bosques de Nothofagus ceno telmatobidos primitivos, ceratofrinidos,
existian durante el Eoceno y el Oligoceno. y bufonidos son conocidos, asi como boideos
A mediados del Terciario los climas se vol- primitivos, cocodrilos, y tortugas meiolanidas
vieron mas secos dando lugar a la expansion y pelomedusidas estan representados en los
de la vegetation xerofitica. depositos del Cretaceo superior y del Ceno-
336 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
zoico inferior. Los fosiles existentes, la pre- Berry, E. W. 1938. Tertiary flora from the Rio Pichi
Leufu, Argentina. Geol. Soc. Amer. Spec. Pap.
serve distribucion y los patrones de especia-
(12):1-149.
cion indican que la Patagonia ha sido una Bogart, J. P. 1970. Systematic problems in the am-
region importante para la evolucion de la phibians family Leptodactylidae (Anura) as indi-
austral. cated by karyotypic analysis. Cytogenetics 9:
herpetofauna
369-383.
Dos
regiones faunisticas se reconocen en
la Patagonia —
1) la region del norte, o Pata-
Cabrera, A.
la
L.
Republiea Argentina.
1951. Territorios fitogeognificos de
Bol. Soc. Argentina Bot.
Herpetologica 26:18-23.
region del sur. Cei, J. M. 1971a. Herpetologia Patagonica I. Lio- —
El grado de endemismo es alto, especial- laemus del grupo magellanicus. Caracteristicas
mente en los altiplanos volcanicos, en las taxonomicas y geneticas. Physis 30:417—424.
Cei, J. M. 1971b. Herpetologia Patagonica -II. —
estribaciones de los Andes y en las mesetas
Notas sobre la distribucion geografica del genero
aisladas. Los generos monotipicos Somun- Diplolaemus. Ibid. 30:471-474.
curia y Vdcunia son endemicos de las mesetas Cei, J. M. 1972a. Herpetologia Patagonica III. Re- —
laciones de afinidad seroproteinicas y fileticas en
asi como las especies y/o subespecies de Al-
el genero Liolaemus. Ibid. 31:411-422.
sodes, Atelognathus, Liolaemus, y Phijma- Cei, J. M. 1972b. Herpetologia Patagonica V. Las —
turus. especies extracordilleranas alto-Patagonicas del
Cei, J. M. 1975b. Herpetologia Patagonica X. El — Duellman, W. E., Veloso M., A. 1977. Phylogeny
conjunto evolutivo de Liolaemus elongatus: ana- of Plcurodema (Anura: Leptodactylidae): A
lisis serologico. Ibid. 34C:203-208. biogeographic model. Univ. Kansas Mus. Nat.
Cei, J. M. 1975c. Herpetologia Patagonica XI. — Hist. Occas. Pap. (64): 1-46.
Diferenciacion serologica de Diplolaemus dar- Estes, R., Reig, O. A. 1973. The early fossil record
wirti y Diplolaemus hibroni en poblaciones alo- of frogs: A review of the evidence, pp. 11-63
simpatridas. Ibid. 34C:209-210. in Vial, J. L. (ed.). Evolutionary
biology of the
Cei, J. M. 1975d. Liolaemus melanops Bunneister anurans: Contemporary research on major prob-
and the subspecific status of the Liolaemus fitz- lems. Univ. Missouri Press, Columbia, 470 p.
ingeri group ( Sauria-Iguanidae). J. Herpetol. 9: Gallardo, J. M. 1962. A proposito de Bufo varie-
217-222. gatus (Gunther) sapo del bosque humedo Ant-
Cei, J. M. 1975e. Southern Patagonian lizards of artandico, y las otras especies de Bufo neotropi-
the Liolaemus kingi group. Herpetologica 31: cales. Physis 23:93-102.
109-116. Gallardo, J. M. 1964. Los generos Urostrophus D.
Cei, J. M. 1976. Remarks on some Neotropical am- & B. y Cupriguanus gen. n. (Sauria, Iguanidae)
phibians of the genus Abodes from southern y sus especies. Neotropica 10:125-136.
Argentina (Anura, Leptodactylidae ) Atti Soc. .
Gallardo, J. M. 1968. Dos nuevas especies de
Italia Sci. Nat. Mus. Civ. Stor. Nat. Milano 117: Iguanidae (Sauria) de Argentina. Ibid. 14:1-8.
la
79-84. Gallardo, J. M. 1970. A
proposito de los Telma-
Cei, J. M. 1977. Serological relationships of the tobiinae (Anura, Leptodactvlidae ) patagonicos.
Patagonian toad Bufo variegatus (Gunther). Ibid. 16:73-85.
Serol. Mus. Bull. 52:2. Gasparint, Z. B., Baez, A. M. 1975. Aportes al cono-
Cei, J. M. 1979. Amphibians of Argentina. Monit. cimiento de la herpetofauna terciaria de la Argen-
Zool. Italiano Monog. Zool. (in press). tina. Acta I Congr. Argentino Paleontol. Biostrat.
Cei, M., Castro, L. P. 1973. Taxonomic and sero-
J. 2:377-413.
logic researches on the Phymalurus patagonicus Harrington, H. J. 1962. Paleogeographic develop-
complex. J. Herpetol. 7:237-247. ment of South America. Bull. Amer. Assoc. Petrol.
Cei, J. M., Castro, L. P. 1975. A serological con- Geol. 46:1773-1814.
tribution to the taxonomic status of Cupriguanus,
Lynch, J. D. 1971. Evolutionary relationships, oste-
a South American genus of iguanid lizards. Serol. and zoogeography of leptodactyloid frogs.
ology,
Mus. Bull. 51:5-6. Univ. Kansas Mus. Nat. Hist. Misc. Publ. (53):
Cei, J. M„
Roic, V. G. 1966. Caracteres biocenoticos 1-238.
de lagunas basalticas del oeste de Neuquen.
las
1978. A re-assessment of the telma-
D.
Lynch, J.
Bol. Est. Geog. Univ. Nac. Cuyo 13:182-201.
tobiineleptodactylid frogs of Patagonia. Univ.
Cei, J. M., Roic, V. G. 1968. Telmatobiinos de las Kansas Mus. Nat. Hist. Occas. Pap. (72): 1-57.
lagunas basalticas de Neuquen (Anura, Lepto-
Morello, J. 1958. La provincia fitogeognifica del
dactylidae). Physis 27:265-284. monte. Opera Lilloana 2:5-115.
Cei, J. M., Roic, V. G. 1975. A new lizard from the
Sierra del Nevado Mountains, central Argentina.
Peters, J. A., Donoso-Barros, R. 1970. Catalogue
of the Neotropical Squamata II. Lizards and am-
J. Herpetol. 9:256.
phisbaenians. Bull. U.S. Natl. Mus. (297): 1-293.
Cei, J. M., Scolaro, J. A. 1977. Herpetologia Pata-
gonica —
XIII. La identidad de Liolaemus goet- Ruiz Leal, A. 1972. Los confines boreal y austral
schi y de laforma melanops del grupo Liolaemus de las provincias patagonicas y central respectiva-
en Rio Negro y Chubut. Physis 36C: mente. Bol. Soc. Argentina Bot. 13 (Supple-
fitzingeri,
225-226. ment) .89-118.
Chaffee, R. G. 1952. The Deseadan vertebrate Schaeffer, R. 1949. Anurans from the early Terti-
fauna of Scarritt Pocket, Patagonia. Bull. Amer. ary of Patagonia. Bull. Amer. Mus. Nat. Hist.
Mus. Nat. Hist. 98:509-562. 93:47-68.
Donoso-Barros, R. 1966. Reptiles de Chile. Ed. Scolaro, J. M.
A., Cei, 1977. Herpetologia Pata-
Univ. Chile, Santiago, 458 p. gonica —
J.
Los iguanidos del grupo Liolaevius
XII.
Donoso-Barros, R., Cei, J. M. 1971. New lizards fitzingeri en Chubut: datos serologicos y position
from Patagonian volcanic tablelands of Argentina. taxonomica. Physis 36C:219-223.
J. Herpetol. 5:89-95. Tihen, J. A. 1962. A review of the New World fos-
Duellman, W. E. 1966. The Central American sil bufonids. Amer. Midi. Nat. 68:1-50.
herpetofauna: An ecological perspective. Copeia Vuilleumier, F. 1968. Origin of frogs of Pata-
1966(4) :700-719. gonian forest. Nature 219:87-90.
338 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
APPENDIX
Appendix 13:1. — Distribution of species and subspecies of amphibians and reptiles in ten districts of the two
major faunal regions in Patagonia.
c -3
B
8
w o
a
-a J
q
2o a
Taxon G
a U 3
« X O
ft*
5o o C/3 -a
'S c
bo <
o5
o O 3
> O t/3
Anurans
Alsodcs gargola gargola ... +
Abodes gargola neuquensis
- - - + -
Atelognathus nitoi _
Atelognathus patagonicus
- - - + -
Atelognathus praebasalticus praebasalticus
- - - + -
- - - + -
Atelognathus praebasalticus agilis
- - - + -
Atelognathus praebasalticus dobeslawi
Atelognathus praebasalticus luisi
- - + -
Atelognathus reverberii
Atelognathus solitarius +
_
—
Leptodacttjlus ocellatus
Odontophnjnus occidentalis
Pleurodema bufonina
Pleurodema nebulosa
Pleurodema thaul
Somuncuria somuncuriensis
Bufo arenarum
Bufo spimdosus
Lizards
Homonota darwinii ...
Homonota horrida
Ctenoblepharis donosobarrosi
Diplolacmus bibronii
Diplolacmus darwinii .. _
Diplolacmus leopardinus .
Leiosaurus bellii _ —
Liolaemus archeforus archeforus
Liolaemus archeforus sarmientoi .
Liolaemus uustromendocinus
Liolaemus bibronii
Liolaemus boulengeri
Liolaemus buergeri ._
Liolaemus ceii
Liolaemus chilensis
Liolaem us darwinii
Liolaemus elongatus elongatus
Liolaemus elongatus petrophilus ...
W
c 0)
a
•3
Taxon o
60
c a
o o
14. La Herpetofauna de los Bosques Temperados
de Sudamerica
J. Ramon Formas
Instituto de Zoologia
Universidad Austral de Chile
Casilla 567
Valdivia, Chile
341
342 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Tadla 14:1. —Temperatura Media, Humedad Relativa y Precipitaciones de Diferentes Puntos del sur de Chile
(segun Hajek y Di Castri, 1975).
Temperatura Humedad Precipita-
Media Relativa
Lugar Ubicacion °C
Punta Lavapie 37°08'S-37°35'W 13.3
Contulmo 38°02'S-73°12'W
Valdivia 39°48'S-73° 14'W
Pto. Montt 41°28'S-72°57'W
Castro 42°29'S-73°48'W
Melinka 43°54'S-73°46'W
Pto. Aysen 45°24'S-72°42'W
San Pedro 47°43'S-74°55'W
Pto. Eden 49°08'S-74°25'W
San Isidro 53°47'S-70°58'W
Navarino 55° 10'S-67°30'W
1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 345
norpatagonica que se extiende hasta el para- fiordos de la region de los canales, se forma-
lelo 48. Alii destacan como arboles mas im- ron en gran medida por la accion del hielo
portantes Podocarpus salignus y Pilgeroden- glaciar (Briiggen, 1950; Fuenzalida, 1965a).
dron uviferum. AI oeste de estas formaciones Los bosques australes sudamericanos, con
boscosas hay pantanos, los cuales se ubican sus elementos tipicos (Nothofagus, Araucaria
especialmente en las islas. Aqui hay Perneth- y Laurelia), han existido desde el Terciario
ija (Ericaceae), Gleichnia ( Pteridofita ) pero , (Jeannelle, 1967; Cerceau-Larrival, 196S) y se
tambien Nothofagus betuloides y Nothofagus acepta que tuvieron una distribution gond-
antarctica. wanica (Couper, 1960). En el extremo sur
de Sudamerica alcanzaron un rango de exten-
sion mucho mas amplio que el que tienen hoy,
HISTORIA DEL AREA llegando hasta la actual Patagonia (Menen-
dez, 1969 ) Durante el Eoceno muchos de los
.
dillera de los Andes. En el Cretacico Superior Rio Turbio, Argentina, mezclados con ele-
se hicieron sentir, en el oeste del continente, mentos australes (Menendez, 1969). Aqui
los primeros movimientos orogenicos del 11a- las capas mas inferiores muestran elementos
mado ciclo Andino (Harrington, 1962) que tipicamente surenos (Nothofagus) los cuales
originaron las diversas partes de la Cordillera son reemplazados en los estratos superiores
de los Andes. Durante el Eoceno, se aprecia por elementos tropicales (Persea, Psidun).
un aceleramiento de los procesos orogeneticos La coexistencia de una con una
flora austral
que alcanzan gran desarrollo en el Mioceno; flora tropical se explica debido a que estos
seguidos en el Plioceno de movimientos que ultimos elementos ocupaban las partes bajas
dieron origen a la forma actual del macizo (valles) y los australes las partes superiores
andino (Harrington, 1962; Haffer, 1970). de montanas (Briiggen, 1950; Menendez,
las
deformacion y plegamiento en varias regiones actuales del bosque temperado ( Solbrig, 1976;
del pais, al final de las cuales el territorio Menendez, 1969).
adquirio las caracteristicas de una region La extincion de los bosques australes en
1965a). Durante el Mioceno hubo una gran al impedir la pasada de los vientos frios y
transgresion marina que cubrio extensas areas humedos del oeste. Durante el Paleoceno y
del sur de Chile (Cecioni, 1970). A fines del el Eoceno el macizo andino estaba poco le-
Plioceno o comienzos de la epoca siguiente, vantado y los vientos del Pacifico llegaban
Pleistoceno, el territorio chileno fue profunda- hasta los sectores orientales del extremo sud-
mente modificado por un fuerte tectonismo. americano. Durante el Oligoceno la barrera
Este trajo como consecuencia el levantami- de los Andes llego a ser un obstaculo para
ento de la Cordillera de los Andes, de la ellos, los cuales se vieron definitivamente fre-
Cordillera de la Costa y la formacion del nados en el Mioceno. Al no haber lluvias ni
Valle Central (Briiggen, 1950; Fuenzalida, humedad en el sector oriental del macizo
america. Durante el Paleoceno el clima del de la Costa hacen presumir que las pendientes
continente fue mas calido
que hoy (Solbrig, occidentales de este macizo costero no tu-
1976) y despues del Eoceno se aprecia un vieron influencia glaciar. Heusser (1966) e
lilies (1970) indican que la Cordillera de la
gradual enfriamiento y desecacion (Axelrod
y Bailey, 1969; Wolfe, 1971) el cual eulmina Costa permanecio fuera de la action de estas
en el Pleistoceno durante las etapas glaciares. masas de hielo.
Fuera de los cambios anteriorniente re-
que la transgresion marina
feridos, es posible
COMPOSICI6N DE LA HERPETOFAUNA
delMioceno y las glaciaciones del Pleistoceno
hayan afectado la distribucion de la heipeto-
fauna austral. La entrada miocenica del mar,
La herpetofauna de los bosques australes
40
I
34
»8
• • •
• 1I
• •
OOO" 2
III! 4
52 SW® 5
++
+++ +
c
o
7
i 8
I
9
ho
in
75 67 73 65
Fig. 14:3. Patrones de distribution latitudinal de los Renews de anfibios y reptiles. Las areas en bianco
(11) corresponden a hielo continental o regiones carentes de anfibios o reptiles. Los numeros indican la den-
sidad generica.
Latitudinal patterns of distribution of the genera of amphibians and reptiles. The white areas (11) repre-
sent ice-covered areas or areas free of amphibians or reptiles. The numbers indicate the generic density.
350 m. Ambas cspecies son simpatricas en gaster tiene una subespecie (L. c. bratt-
una amplia Liolaemus pictus presenta
area. stroemi) que se distribuye en la Isla Grande
tres subespecies (L. p. chiloensis, L. p. major de Chiloe (Donoso-Barros, 1966). Liolaemus
y L. p. talcanensis) que se distribuyen en el tenuis ocupa alturas que van desde el nivel
archipielago de Chiloe ( Donoso-Barros, 1966; del mar hasta los 1000 m y en la region co-
Urbina y Ziiniga, 1977). Liolaemus cijano- stera es reemplazada por una subespecie, la
350 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
75
1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 351
75
352 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
75 75 73
T^37
Batrachyla
taeniata
Batrachyla
leptopus
41
Batrachy la
antartandica
t>£
•
{
200 Km
75
354 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
75
yj I
- 37
- 37
200 km
200 km
Fig. 14:13. Patrones de distribution de Liolae-
mus chilensis, L.
cyanogaster cyanogaster tj L. c.
brattstroemi en el sur de Chile.
Fic. 14:12. Patrones de distribution de Pristi-
Distribution patterns of Liolaemus chilensis, L.
dactylus torquatus y Liolaemus monticola villaricen-
cyanogaster cyanogaster and L. c. brattstroemi in
sis en el sur de Chile.
southern Chile.
Distribution patterns of Pristidactylus torquatus
and Liolaemus monticola villaricensis in southern
Chile. de la cola es afilada. Rhinodcrma darwinii
muestra un patron de desarrollo que tiene
aumento en el conocimiento del desarrollo fuertes desviaciones del esquema corriente
embrionario y larvario de sapos y ranas del presentado por los anuros (Noble, 1931).
bosque. En la Tabla 14:5 aparece la duracion Como consecuencia de ello la larva muestra
del periodo embrionario y larvario de algunos reducciones en las aletas, forma del cuerpo y
batracios del sur de Chile. La mayoria de las estructura de la boca (Jorquera, et al., 1972).
especies de anuros presentan larvas del tipo Las especies del genero Rhinoderma ( dar-
generalizado de Orton ( 1953 ) sin embargo
;
iviniiy rufum) muestran patrones de desarro-
existen tambien otros tipos larvarios adapt- llo unicos entre los anuros ya que algunas de
ados a ambientes especiales. Telmatobufo aus- las etapas del desarrollo transcurren en la
tralis posee un renacuajo adaptado a vivir en bolsa gutural del macho. En la figura 14:17
aguas frias y torrentosas, su boca tiene forma se muestra comparativamente los ciclos de
de ventosa y por medio de ella se adhiere a las desarrollo de las especies de Rliinodcrma.
piedras (Formas, 1972). La larva de Caudi- Los huevos de R. darwinii son depositados
verbera caudiverbera vive en grandes cuerpos en terreno vegetal hurnedo; posteriormente el
de aguas abiertas y su cuerpo esta adaptado macho toma con la boca los jovenes embri-
a estos ambientes nectonicos. Su forma es ones y los introduce en la bolsa bucal dentro
redondeada, las aletas son altas y la punta de la cual se realiza el resto del desarrollo
1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 355
Tabla 14:3. — Meses del Aiio Durante los cuales los Machos de las Especies Listadas Cantan y Mucstran
Callosidades Sexuales ( O = canto, X = callosidades sexuales, A = ambos).
Especies
356 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
75
Liolaemus
37
pictus
Alsophis
Cham i ssoni s
37
41
Tactiynien
chi lensis
Liolaemus p
chi loensis Liolaemus 41
nun major
Lio laemus p.
tatcanens 200 km
Alturas ( m )
ro
01
S o
o 8 o o
o o
J
A. grandisonae
B. taeniata
B. leptopus
B. antar tandica
C. caudiverbera
P. thaul
R. darwinii
R. r uf urn
B. variegatus
B. chi lensis
B. rubropunc tat us
E. vittatus
E. roseus
E. vanzolinii
E. migueli
H. sylvatica
A. monticola
I. acarpicus
T. venu s tus
T. austra lis
Biotemperaturas (°C)
Fig. 14:16. Distribution altitudinal de los anuros del hosque austral.
Altitudinal distribution of anurans in the austral forest.
Caudiverbera aparece mas cerca de los escuer- verbera estarian muy relacionados, pero no
358 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
ca c3
c c
o o
b* (-
Oj OJ 0J
So » ft ft
fc c c c
CD *o -o -o
*~*
'o
ca ca
.5 > __^ > >
CO CD q>a;
t — t"- f w
OS CO _Q
-O-
-* —
-jfa^ro 4) OJ
<N >>
i ^h O O
» - ca ca
en
35 ca ca ca D OJ
3 3
s s s
i* u !-
_ o o o O O O O
O fa fa fa - _ fa
3a c 9 000 0)00 c
4) oj
"
"H 7tT
,
°i
•-
.0.0
»° i
l/j <M CN CM "2
S £
S <<
HNM^ciro
C
SlJJ-
J3
E
w so
ft"
6
OJ
o
Q ?
D
fa
3
Q
1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 359
Rhinoderma darwinii
20 di'as 34 di'as
Rhinoderma r u f u m
7 di'as 14 di'as A meses
Fig. 14:17. Modalidades de desarrollo de las dos especies del gencro Rhinoderma.
Modes of development in the two species of Rhinoderma.
region al sur del paralelo 41, como ha sido generos (Caudiverbera y Tcbnatobufo) pre-
indicado por Vuilleumier, es dificil entender sentan una diversidad especifica muy baja.
entonces la presencia de anuros tales como Hasta ahora se han descrito solamente dos
Hijlorina sylvatica, Batrachijla antartandica especies de Telmatobufo (venustus y austra-
(Atalah y Sielfeld, 1976) y Atelognathus lis) y dos de Caudiverbera; una viviente (C.
ejemplos de plantas que indican que el bos- generos (2 especies), la antigiiedad de Cau-
que austral tuvo una expansion mas nortena diverbera (Terciario Inferior) y la presencia
(30°S) que la que tienen actualmente. La de caracteres primitives (26 cromosomas y
presencia de Batrachijla taeniata, anuro tipico dilatacion del coxis en T. venustus) en las
del sur de Chile, en los bosques relictos de especies de ambos generos, nos hace suponer
la provincia de Valparaiso (Quintero) (Ca- que Telmatobufo y Caudiverbera son miem-
purro, 1958; Cei y Capurro, 195S) hace su- bros de una antigua radiation de leptodac-
poner tambien que algunas especies siguieron tilidos que alcanzo poca diversification en el
el avance hacia el norte del bosque austral extremo sur de Sudamerica.
durante los periodos glaciares. Schaeffer (1949) describio un Eupsophus
E nt re los mas probables habitantes de los sp. en el Oligoceno de Chubut (sur de Ar-
refugios occidentales de la Cordillera de la gentina). En esta especie los nasales estan
Costa, estan las especies del genero Telma- en aparente contacto medio, condition que
tobufo, las cuales muestran fuertes adapta- no se encuentra en las especies vivientes del
ciones a los arroyos de montaiia; sin embargo, genero. Rogart ( 1970 ) sehala que el fosil no
Caudiverbera no pudo alcanzar estos refugios es separable de Eupsophus roseus, forma vi-
costeros debido a que es un genero adaptado viente en el bosque temperado. Eupsophus
a ambientes Ienticos y es muy posible que ha sido tradicionalmcnte un genero muy con-
haya tenido que emigrar hacia el norte. Pos- fuso y mal definido, lo que ha traido como
teriormente al desaparecimiento de los gla- consecuencia falsas intcrpretaciones biogeo-
ciares, Caudiverbera reinvadio el Valle Cen- graficas. Despues de las revisiones de Cei
tral,pero no logio penetrar en la Isla de (1960, 1962b), Grandison (1961) y Lynch
Chiloe ya que la existencia del Canal de (1972, 1975) se puede apreciar una cierta
Chacao fue una barrera real para impedir la consistencia taxonomica para el genero. Eup-
dispersion hacia el sur de esta especie. sophus esta exclusivamente distribuido en el
1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 361
bosque temperado de Sudamerica. Lynch genero irradio desde el sur de Chile a la Pata-
(197S) ubica a Eupsophus, Alsodes, Atelog- gonia Argentina.
natkus, Batrachophrynus, Hylorina, Insueto- Si las especies de Eupsophus realizan sus
algunos de sus miembros poseen caracteres sur de Chile y en las regiones costeras de
cromosomicos (2n = 26) y etologicos primi- Chile Central (B. rufum) (Formas et al.,
tives (amplexus inguinal en Batrachyla lep- 1975).
topus ) . El genero Bufo esta representado en el
Pleurodema es el unico genero de la sub- bosque austral por tres especies: B. chilen-
familia Leptodactylinae presente en el bosque sis, B. rubropunctatus y B. variegatus, siendo
de Nothofagus del sur de Chile. La unica esta ultima un anuro de dificil ubicacion taxo-
especie existente aqui es Pleurodema thaul; nomica, ya que sus relaciones son oscuras y
sin embargo Lynch ( 1971 ) seriala que esta controvertidas. Cei y Espina ( 1957 ) lo con-
especie podria tener un ran go de superespe- sideran vinculado al grupo de Bufo spinu-
cie. Duellman y Veloso (1977) indican que losus; sin Silva (1959)
embargo Capurro y
al menos dentro de los que se llama Pleuro- al hacer estudio cromatografico de la secre-
el
dema thaul existen tres especies las cuales se cion de las parotidas hayan claras diferencias
pueden diferenciar por su amplexus y cario- entre Bufo spinulosus (=
chilensis) y Bufo
tipo. El
genero Pleurodema tiene
amplia variegatus. Gallardo
(1962b) indica que
distribucion en Sudamerica, especialmente en Bufo variegatus no tiene ninguna relation con
los ambientes temperados y semi-aridos. los Bufo sudamerioanos del grupo spinulosus
Heyer ( 1975 ) considerando la presencia o y que sus afinidades estarian con el grupo
ausencia de glandulas lumbares, insinua un calamita de la region Holoartica. Martin
probable origen difiletico para el genero. (1972) coloca a Bufo variegatus en un grupo
Lynch 1971 ) basandose en la anatomia y
( propio (grupo variegatus); sin embargo dice
configuracion del renacuajo supone que Pleu- que miebros del grupo spinu-
es similar a los
rodema es un miembro muy primitivo de la losus con los cuales
probablemente estaria
subfamilia Leptodactylinae y que el genero estrechamente relacionado. Cei ( 1977 ) de-
presentaria relaciones con Eupsophus. Ambos muestra, desdc el punto de vista inmunolog-
se diferencian claramente en la morfologia ico,que Bufo variegatus presenta mayores re-
esternal y en la biologia reproductiva. En laciones con el grupo calamita que con las
la distribucion cladistica de Heyer (1975) especies sudamericanas del grupo spinulosus.
basada en caracteres derivados, Pleurodema Ademas de de Bufo anterior-
las tres especies
esagrupado con Batrachyla, Batrachophrynus mente ha citado en el area Pro-
referidas, se (
mente se ban descrito dos especies (Bhino- de taxa existente en bosque temperado de
el
derma darwinii y B. rufum) esta distribuido Sudamerica, ya que muchas familias y gen-
en los bosques temperados de Nothofagus del eros tipicamente neotropicales estan ausentes
1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 363
del bosque austral. Vuilleumier (1968) senala En segundo grupo faunistico se in-
el
que no se debe olvidar que los anuros del cluye genero Pleurodema el cual tiene
al
bosque del sur de Chile pertenecen a una amplia distribucion en Sudamerica, especial-
zona temperada, las cuales contrastan en gen- mente en los ambicntes semiaridos. Pleuro-
eral con las areas tropicales que son muy ricas dema es considerado un grupo originado en
en especies. el extremo sur del continente y que posterior-
Vuilleumier ( 1968) considera a la fauna de mente emigre al resto de Sudamerica, en el
anuros del bosque temperado del sur de Chile cual adquirio gran diversification (14 espe-
como compuesta de cuatro elementos fau- cies) (Duellman y Veloso, 1977). Tambien
nisticas: 1) grupos de Ieptodaetilidos autoc- en esta agrupacion se incluye a Atelognathus
tonos que no se han diversificado (Caudiver- el cual segun Lynch (1978) irradio del sur
y que tienen mas especies fuera del bosque tral sea secundaria. Segun Cei ( 1971 ) Bufo
austral que dentro de el (Pleurodema, Bnfo). chilensis y B. rubropunctatus aparecen como
De acuerdo a la discusion sobre el origen un conjunto de antiguo abolengo filetico, se-
y las relaciones de los anuros del bosque aus- guramente anterior al levantamiento andino.
tral, se proponen aqui tresgrupos faunisticos Schaeffer (1949) describio Neoprocoela eden-
para explicar la historia de los anfibios de este tata en depositos del Oligoceno inferior de
bosque. Chubut (Argentina). Esta es una forma de
El primer elemento faunistico esta com- problematica ubicacion familiar ya que se ha
puesto de Ieptodaetilidos ( Caudiverbera, Tel- considerado dentro de la familia Leptodacty-
matobufo, Hylorina, Batrachyla, Eupsophus, lidae (Schaeffer, 1949; Lynch, 1971) y Bu-
Insuetophrynus y Abodes) y rhinodermatidos fonidae (Tihen, 1962). Si Neoprocoela es un
(Rhinoderma) originados en el extremo sur miembro del genero Bufo como lo indica Ti-
de Sudamerica. Todos los generos aqui indi- hen (1962) la presencia de los anuros de este
cados son endemicos del bosque austral, en genero en el bosque austral pareceria ser muy
algunos casos tienen una antiguedad que se antigua.
remonta al Terciario Inferior y otros poseen La batracofauna austral presenta once ge-
caracteristicas muy primitivas. Es posible que neros de los cuales siete son endemicos de
los integrantes de este grupo faunistico se la region, otros (Pleurodema y Atelognathus)
haya refugiado en los sectores occidentales se han originado aqui, pero se diversificaron
de la Cordillera de la Costa durante las gla- en Sudamerica y Bufo tiene un origen pos-
ciaciones del Pleistocene La inclusion de iblemente secundario. Por lo tanto se puede
Rhinoderma, genero de enigmaticas rela- decir que la batracofauna de los bosques aus-
ciones filogeneticas, se debe principalmente a trales de Sudamerica es mayoritariamente en-
la estrictez de su distribucion dentro del demica del area.
bosque y sus adaptaciones reproductivas a Entre los reptiles existentes en los biotopos
este biotopo (cuidado de las crias en la bolsa australes se encuentran saurios del genero
del macho). La mayoria de los generos de Liolaemus, el cual segun Peters y Donoso-
este grupo tiene muy pocas especies y Eup- Barros (1970) incluye mas de 54 especies.
sophus y Alsodes parecen ser los mas diversi- Liolaemus se distribuye en el cono sur del
ficados. Esta primera agrupacion faunistica continente sudamericano, ocupando regiones
contiene la mayor cantidad de generos del tales como Argentina, Bolivia, Brasil, Chile,
bosque austral, es endemico del area y rara Paraguay, Peru y Uruguay. Las especies de
vez sus especies sobrepasan sus limites. este taxon tienen una gran versatilidad eco-
364 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
costa, en los desicrtos, en losbosques tropi- de losAndes. Esta especie junto con L. lo-
eales, estepas sabanas, grandes alturas de la renzmiilleri, L. schrbderi, L. nigroviridis, L.
Cordillera de los Andes y en las selvas austra- leopardinus, L. altissimus y L. buergeri repre-
les. Donoso-Barros (
1966 ) plantea que el sentan un grupo alto andino ( Hellmich, 1950 )
genero Liolaemus tiene su origen posible- y la especie que ocupa los limites mas bajos
mente en la region de Bolivia. Hellmich de los Andes es L. monticola la cual se des-
(
1952 ) basado en la diversification de las membra hacia el sur en varias subespecies
especies de Liolaemus, tanto al este como al (monticola chillanensis y monticola villaricen-
oeste del macizo andino, postula que la Cor- sis). De estas formas solamente Liolaemus
dillera de los Andes ha sido un elemento monticola uillaricensis penetra al bosque
diferenciador de especies, a partir de un austral.
tronco preexistente antes del levantamiento genero de iguanidos
Pristidactijlus es otro
de los Andes. Si Liolaemus se origino de un representado en bosque por una sola espe-
el
tronco preandino, cerca de la region de Bo- cie, Pristidactijlus torquatus. Existen dos
livia, es posible que la presencia de las espe- especies mas en Chile (Pristidactijlus alvaroi
cies del genero existente en el sur de Chile y P. valeriae) (Donoso-Barros, 1974) las cu-
haya sido un fenomeno de emigration se- ales estan en los bosques relictos del Cerro
cundario. Parece ser que la invasion de Lio- El Roble (Provincia de Valparaiso). En Ar-
laemus en el bosque austral ha sido por la gentina existen cuatro especies mas: P. sca-
Cordillera de los Andes y por el Valle Central. pulatus, P. achalensis, P. casuhatiensis y P.
A pesar de que la Patagonia es rica en rep- fasciatus. Las especies argentinas de Pristi-
tiles,no existen especies compartidas entre dactijlus muestran una distribution discon-
los dos biotopos; sin embargo se observa una tinua (Cei, 1973). Este patron sugiere que
leve invasion de especies patagonicas, en el las especies del genero se diferenciaron de un
sur del bosque austral tales como: Liolaemus tronco primitivo posiblemente antes del lev-
fitzingeri y L. bihronii (Donoso-Barros, 1960), antamiento de los Andes.
el saurioPhymaturus patagonicus (Cei y Cas- Dos especies de serpientes llegan al bos-
tro, 1973) y el anuro Pleurodema bufonina que austral: Tachymenis chilensis y Ahophis
(Cei, 1962a). La ingresion de especies pata- cliamissonis. Esta ultima especie es ovipara
gonicos, en los limites del bosque temperado, y no penetra muy al sur llegando hasta Val-
ha sido tambien referida por Cekalovic divia (Donoso-Barros, 1966). Maglio (1970)
(1974). Las especies de Liolaemus que pare- ha revisado las culcbras del genero Alsophis,
cen haber ingresado por el Valle Central seri- el cual se distribuye especialmente en las islas
an Liolaemus tenuis, Liolaemus cijanogaster, del Caribe, noroeste de Sudamerica y extremo
Liolaemus pictus y Liolaemus chilensis. L. sur de America Central. Un pequeiio grupo
chilensis ha sido incluida por Hellmich ( 1952) de especies, A. angustilincatus del Peru y A.
en su Grupo III, en el cual tambien esta L. cliamissonis, se extienden hacia el sur por el
gravenhorstii. Todo este grupo de especies oeste de Sudamerica. Maglio (1970) incluye
estepa de Acacia caven. La especie mas su- genus e indica que Alsophis cliamissonis mu-
rena de este grupo (Liolaemus chilensis) es cstra mucha similitud con Alsophis canthcri-
la que penetra en el limite norte del bosque genus de Cuba. Segun este autor es posible
austral. Liolaemus pictus, L. cijanogaster y que las especies de las Indias Occidentales
L. tenuis ban sido incluidas por Hellmich hayan derivado de un tronco ubicado en el
(1952) en su Grupo IV. De estas tres espe- oeste de Sudamerica en el cual se encuentran
cies, Liolaemus cijanogaster y L. pictus estan hoy A. cliamissonis, A. tachijmenoides y A.
estrictamente restringidas bosque de No-
al angustilineatus. Estas tres especies represen-
thofagus, pero Liolaemus tenuis alcanza por tarian un centra de especiacion sobre las
el norte hasta el paralelo 32. Finalmente Lio- Galapagos y Alsophis cliamissonis, segun
laemus monticola mucstra una penetration Maglio, parece ser un relicto del tronco prim-
1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 365
itivo. La entrada de esta especie en el bosquc reptiles, represented by the genera Liolae-
austral seria por lo tanto secundaria, pues su mus. Pristidactylus, Alsophis, and Tachymen-
centro originario debe haber estado colocado is, there are no endemics at the generic level.
al sur del Peru o norte de Chile. In general, the anurans have a low diversity;
El genero Tachymenis posee seis especies Caudiverhera, Hylorina, and Insuetophrynus
are monotypic.
(T. affinis, T. peruviana, T. tarmensis, T. at-
tenuata y T. surinamensis) (Peters y Orejas- This small herpetofauna is mainly re-
stricted to southern Chile between 37 and
Miranda, 1970) las cuales estan en Peru, Bo-
livia, Chile y Surinam. Al no existir registros 55°S but also occurs in parts of adjacent Ar-
fosiles ni estar claras las relaciones de estas gentina. In southern Chile, three physio-
refuges for much of the herpetofauna that group of frogs probably originated in south-
dwells today in southern Chile. ern South America and dispersed northward
The highest concentration of the herpeto- in the austral forests.
fauna occurs in the northern part of the aus- The second faunal element contains Pleu-
tral forest (37°S), whereas there is a gradual rodema and Atelognathus, which apparently
decrease in numbers towards the south. Only originated in the austral forest and secondarily
Bufo reaches 53°S. It is possible that the dispersed widely in South America. The third
declining temperature gradient is the respon- element contains the toads of the genus Bufo,
sible factor for this distribution pattern. Alti- which probably secondarily entered the aus-
tudinally, most of the amphibians reach 1000 tral forests. The reptiles had a different his-
m; Bufo and Pleurodema exceed 2000 m. tory; of
all the groups seem to be secondary
in the austral forests.
Despite the denseness of the austral for-
ests, no amphibians there have arboreal adap-
tations. The majority of anurans are restricted
BIBLIOGRAFIA
to the forest floor or to lotic or lentic aquatic
environments. Occasionally Hylorina syl-
Arenas, J. N. 1972. Morfometria del Lago Rinihue
vatica and Batrachyla leptopus have been (Prov. de Valdivia, Chile). Mus. Nac. Hist. Nat.
observed on branches and trunks of trees. (Santiago de Chile) Publ. Occas. 14:3-15.
Many lizards have arboreal tendencies. Atalah, A., Sielfeld, W. 1976. Presencia de Ba-
Several anurans have adaptations in the trachyla antartandica Barrio en Magallanes. An.
Inst. Pat. 7:169-170.
embryonic and larval stages for the forest Aubouin, J., Bohello, A. V., Cecioni, G., Charrier,
floor. One of the most notable occurs in R., Chotix, P., Frutos, J., Thiele, R. y Vicente,
C. 1973. et struc-
Rhinoderma, in which the larvae develop in J. Esquisse paleogeographique
turale des Andes Meridionales. Rev. Geog. Phys.
the vocal sac of the males. Other reproductive
Geol. Dynam. 15:11-72.
adaptations are related to egg-deposition sites. Auer, V. 1960. The Quaternary history of Fuego
Batrachyla taeniata and Rhinoderma lay eggs Patagonia. Proc. Roy. Soc. London 152:507-516.
in wet soil, whereas Eupsophus lay eggs in Axelrod, D. I., Bailey, H. P. 1969. Paleotempera-
ture analysis of Tertiary floras. Paleogeogr. Paleo-
water holes. Larvae of Telmatobufo australis climatol. Paleocol. 6:163-195.
are adapted to swift streams, whereas those of Barrio, A. 1967a. Batrachyla antartandica n. sp.
Caudiverbera caudiverhera are nectonic. ( Anura, Leptodaetylidae Descripcion y estudio
) .
The most outstanding reproductive adap- comparativo con la especie genotipica, B. leptopus
Bell. Physis 27:101-109.
tation in the reptiles is viviparity, which is
Barrio, A. 1967b. Observaciones etoecologicas so-
present in some Liolaemus (cyanogaster and bre Hylorina sylvatica Bell (Anura, Leptodaetyli-
pictus) and in the snake Tachymenis chilensis. dae). Ibid. 27:153-157.
The history of the herpetofauna can be Barrio, A. 1970. Insuetophrynus acarpicus, un nue-
vo leptodactilido finisternio sudamericano (Am-
traced back to the Early Tertiary. Caudi-
phibia; Anura). Ibid. 30:331-341.
verbera and Eupsophus have been found in
Barrio, A. 1973. Una nueva especie de Tclmato-
strata in Patagonia, where fossil plants asso- bius (Anura, Leptodaetylidae) procedente del
ciated with austral forests also were found, dominio austral cordillerano argentino. Physis 32:
207-213.
thereby indicating that the austral forests and
Barrio, A., Rinaldi de Chieri, P., 1971. Contribu-
at leastsome of its inhabitants were much cion al esclarecimiento de la posicion taxonletica
more widespread than they are today. de algunos batracios patagonicos de la familia
Analyses of patterns of distribution of the Leptodaetylidae mediante el analisis cariotipico.
Ibid. 30:673-685.
different taxa, their phylogenetic relation-
Bogart, J. P. 1970. Systematic problems in the
ships, and their age (inferred from the fossil
amphibian family Leptodaetylidae (Anura) as
data) provide die basis for the recognition of indicated by karyotypic analysis. Cytogenetics
three herpetofauna! groups. The first, charac- 9:369-383.
terized by endemics, primitive characters, and Bogart, J. P. 1973. Evolution of anuran karyotypes,
pp. 337-339 in Vial, J. L. (ed.). Evolutionary
antiquity, includes Caudiverbera, Telmato- biology of the anurans: Contemporary research
bufo, Hylorina, Batrachyla, Insuetophrynus. on major problems. Univ. Missouri Press, Co-
Alsodes, Rhinoderma, and Eupsophus. This lumbia, 470 p.
1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 367
Rruggen, J. 1948. La expansion del bosque en el de lAmerique australe, pp. 111-197 in Dela-
sur de Chile en la epoca post-glacial. Rev. Uni- mere Deboutteville, C, Rapoport, E. (eds.).
vers. 33:105-114. Biologie de rAmerique australe, 4. C. N. R. S.
Brugge.n, J. 1950. Fundamentos de la geologia de Groupe Francais Argiles C. R. Reun. Etud., Paris,
Chile. Ceogr. Militar, Santiago, 374 p.
Inst. 685 p.
Bhum-Zohrilla, N., Saez, F. A. 1968. Chromosomes Couper, R. A. 1960. Southern Hemisphere Meso-
of Leptodactylidae (Amphibia, Anura). E.xperi- zoic and Tertiary Podocarpaceae and Fagaceae
entia 24:969.' and their puleogeographic significance. Proc. Rov.
Busse, K. 1971. Desarrollo de Batrachyla Icptopus Soc. London 152:491-500.
Bell eon observaciones sobre su ecologia y com- Darlington, P. J. 1965. Biogeography of the south-
portamiento (Amphibia; Leptodactylidae). In- ern end of the world. Harvard Press, Cambridge,
vest. Zool. Cbil. 15:5-63. 236 p.
Cabrera, J. 1977. Adaptacion y cultivo de Caudi- Di Castri, F. 1968. Esquisse Ecologique du Chili,
verbcra caudiverbera (Linneus) en medios arti- pp. 7-82 in Delamere Deboutteville, C, Rap-
ficiales. Cayana 5:111-112. oport, E. (eds.). Biologie de rAmerique aus-
Capurro, L. F. 1958. Nota sobre la ovipostura y trale, 4. C. N. R. S. Groupe Francais Argiles
desarrollo de Eupsophus taeniatus (Girard). In- C. R. Reun. Etud., Paris, 685 p.
vest. Zool. Chi). 4:20S. Donoso-Barros, R. 1960. Ecologia de los reptiles
Capurro, L. F., Silva, F. C. 1959. Valor taxono- chilenos. Invest. Zool. Chil. 6:65-72.
mico del estudio cromatografico del veneno de Donoso-Barros, R. 1966. Reptiles de Chile. Univ.
las parotidas de Bufo spinulosis y Bufo variega- Chile, Santiago, 458 p.
tus. Ibid. 5:189-197. Donoso-Barros, R. 1974. Nuevos reptiles y anfibios
Casamiquela, R. M. 1963. Sobre un par de anuros de Chile. Bob Soc. Biol. Concepcion 48:217-229.
del Mioceno de Rio Negro (Patagonia) Wawelia Duellman, W. E., Veloso, A. 1977. Phylogeny of
gerholdi n. gen. et sp. (Ceratophrydidae) y Plcurodema (Anura: Leptodactylidae): A bio-
Gigantobatrachus parodii (Leptodactylidae). geographic Model. Univ. Kansas Mus. Nat. Hist.
Ameghiniana 3:141-160. Occas. Pap. (64): 1-46.
Cecioni, G. 1970. Esquema de paleogeografia chi- ENDESA. 1972. Caudales medios mensuales retro-
lena. Ed. Univers., Santiago, 143 p. spectivos. Empresa Nacional de Electricidad,
Cei, J. M. 1960. A survey of the leptodactylid frogs, Santiago, 211 p.
genus Eupsophus in Chile. Breviora 118:1-13. Estes, R., Reig, O. A. 1973. The early fossil record
Cei, J. M. 1962a. Batracios de Chile. Univ. Chile, of frogs: A review of the evidence, pp. 11-63 in
Santiago, 128 p. Vial, J. L. (ed. ). Evolutionary Biology of the
CEr, J. M. 1962b. El genero Eupsophus en Chile. anurans: Contemporary research on major prob-
Invest. Zool. Chil. 8:7-42. lems. Univ. Missouri Press, Columbia, 470 p.
Cei, J. M. 1970. La posieion filetica de Telmatobii- Formas, J. R. 1972. A second species of Chilean
nae, su diseusion reciente y significado critico de frog genus Tehnatobufo (Anura: Leptodactyli-
algunos inmunotests. Act. Zool. Lilloana 27:181- dae). Ibid. 6:1-13.
192. Formas, J. R. 1975. Las larvas de las especies chi-
Cei, J. M. 1971. Analisis sero-inmunologico de difer- lenas pertenecientes al genero Eupsophus, grupo
entes niveles de especiacion en Bufo del grupo nodosus (Anura; Leptodactvlidae). Bob Soc.
spinulosis. Ibid. 28:91-105. Biol. Concepcion 49:231-237.'
Cei, J. M. 1973. Comentarios sobre algunos generos Formas, J. 1976.
R. Descriptions of Batrachyla
de Iguanidos: Diplolaemus, Lciosaurus, Apero- (Amphibia, Anura, Leptodactvlidae) tadpoles.
pristis y Cupriguanus. Physis 32:269-276. J. Herpetol. 10:221-225.
Cei, J. M. 1977. Serological relationships of the Formas, J. R., Espinoza, N. D. 1975. Karyological
patagonian toad Bufo variegatus (Giinther). Ser- relationships of Tehnatobufo ( Anura; Leptodacty-
olog. Mus. Bull. 52:2. lidae ). Herpetologica 31:429-432.
Cei, J. M., Capurro, L. F. 1958. Biologia y desa- Formas, J. Pugin, E.
R., 1971. Reproduction y
rrollo de Eupsophus taeniatus (Girard). Invest. desarrollo de:
Batrachyla untaiiandica Barrio
Zool. Chil. 4:150-182. (Anura, Leptodactylidae). Mus. Nac. Hist. Nat.
Cei, J. M., Castro, L. P. 1973. Taxonomic and (Santiago de Chile) Bob 32:201-213.
serological researches on the Phymaturus pata- Formas, J. R., Pugin, E. 1978. Tadpoles of Eupso-
gonicus complex. J. Herpetol. 7:237-247. phus roscus and Bufo variegatus in Southern
Cei, J. M., Espina, S. 1957. La vibration sexual pre- Chile. J. Herpetol. 12:243-246.
ventiva en poblaciones de Bufo spinulosis de Formas, J. R., Pugin, E., Jorquera, B. 1975. La
Chile. Invest. Zool. Chil. 4:62-65. identidad del batracio chileno Heminectcs rufus
Cekalovic, T. 1974. Divisiones biogeograficas de la Philippi, 1902. Physis 34:147-157.
XII Region chilena ( Magallanes). Bob Soc. Biol. Fuenzalida, H. 1965a. Orografia. In Geografia
Concepcion 48:297-314. Economica de Chile Corfo, Santiago: 7-33.
Cerceau-Larrival, T. 1968. Contribution palyno- Fuenzalida, H. 1965b. Hidrografia. In Geografia
logique et biogeographique a l'etude biologique Economica de Chile Corfo, Santiago: 153-199.
368 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fuenzalida, H. 1965c. Clima. In Geografia Eco- Jorquera, B., Pugin, E., Goicoechea, O. 1974.
.\6mica de Chile Corfo, Santiago: 99-152. Tabla de desarrollo normal de Rhinoderma dar-
Fuenzalida, H. 1965d. Biogeografia. In Geografia wini Conception )
( Bol. Soc. Biol. Concepcion
.
Jorquera, B., Izquierdo, L. 1964. Tabla de desa- Chairelli, A. B., Capanna, E. (eds.). Cytotax-
rollo normal de Calyptocei>lialella gatji ( Rana onomy and vertebrate evolution. Acad. Press.
chilena). Biologica 36:43-53. London, New
York, 783 p.
Jorquera, B., Pugin, E., Goicoechea, O. 1972. Muller, 1973. The dispersal centres of terrestrial
P.
Tabla de desarrollo normal de Rhinoderma dar- vertebrates in the Neotropical Realm. A. Junk,
wini. Arch. Med. Veter. 4:1-15. The Hague, 224 p.
1979 FORMAS: HERPETOFAUNA DE BOSQUES TEMPERADOS 369
Munoz-Cristi, J. 1973. Geologia de Chile. Ed. Schaeffer, B. 1949. Anurans from the early Terti-
Andres Bello, Santiago, 209 p. ary of Patagonia. Bull. Amer. Mus. Nat. Hist.
Noble, G. K. 1931. The biology of the Amphibia. 93:47-68.
McGraw-Hill Book Co., New York, 577 p. Schmidt, K. P. 1952. A new leptodactvlid frog from
Oberdobfer, E. 1960. Pflanzensoziologisehe Studien Chile. Fieldiana Zool. 31:11-15.
in Chile, ein Vergleieh mit Europa. Verlag J. Schneider, C. O. 1930. Observaciones sobre batra-
Kramer, Weinheim, 208 p. cios chilenos. Rev. Chil. Hist. Nat. 34:220-223.
Orton, G. L. 1953. The systematics of vertebrae Silva, F., Veloso, A., Solervicens, J., Ortiz, J. C.
larvae. Syst. Zool. 2:63-75. 1968. Investigaciones Zoologicas en el Parque
Pefaub, J. 1971. Nota sobre Tehnatobufo bullocki Nacional Vicente Perez Rosales y Zona de Pargua.
Schmidt (Anura, Leptodactylidae Mus. Nac.
) .
Mus. Nac. Hist. Nat. (Santiago de Chile) Bol.
Hist. Nat. (Santiago de Chile) Bol. 32:215-225.
148:1-12.
Peters, J. A., Donoso-Barros, R. 1970. Catalogue
Solbrig, O. T. 1976. The origin and floristic affini-
of the Neotropical Squamata: Part II. Lizards
ties of the South American temperate desert and
and amphisbaenians. Bull. U. S. Natl. Mus. 297:
semidesert regions, pp. 7—49 in Goodall, D. W.
1-293.
( ed. ) Evolution of desert biota. Univ. Texas
.
William E. Duellman
Museum of Natural History and
Department of Systematics and Ecology
The University of Kansas
Lawrence, Kansas 66045
USA
The Andes —the longest mountain chain in the puna vegetation. In the extreme south
the world —extend nearly km along the
-8,000 the austral forests extend nearly to snow line.
northern and western edges of South America. Permanent snow and glaciers exist on the
This young mountain chain contains many higher peaks throughout the Andes, and the
active volcanoes and innumerable dormant Cordillera Real extending for about 300 km
ones. More than a dozen peaks reach heights in Bolivia and the Cordillera Blanca about
of more than 6,000 m; only the Himalayas and 400 km in length in Peru, are nearly continu-
Pamirs in Asia have peaks that are higher. ous snow-covered ranges. The melting snow
Frequent earthquakes attest to continuing and glaciers provide water for countless An-
tectonic activity. Spanning 66° of latitude dean lakes, of which are trapped in
many
through the tropics and southern temperate glacial cirques and reach gigantic proportions
zone and reaching to within 1,300 km of the in Lago Titicaca (177 X 56 km, 3812 m), and
Antarctic Circle, the Andes are a major fac- myriads of small streams, some of which fed
tor in the formation of climates in western by heavy precipitation on the Andean slopes,
South America. Blocking both easterly and grow and merge to form the giant tributaries
westerly moisture-laden winds, the massive of the Rio Amazonas. For more general in-
mountain range creates immense rain-sha- formation on the Andes and excellent photo-
dows west of the Andes between 5° and 35°S graphs, the reader is referred to the works by
and east of the Andes between 28° and 38°S. Morrison ( 1974,
1975 ) a real appreciation
;
The eastern face of the Andes in the tropics for early exploration in the Andes can be
and the western face north of the Equator gained from Whymper (1892).
and south of 37°S receive abundant rainfall. The complex topography and variety of
At high elevations daily temperatures vary as environments resulting from tectonic events
much as 20°C; in many areas freezing tem- and climatic fluctuations in the Pleistocene
peratures are a nightly occurrence. and continuing to the present provide an
Thus, the climates and environments of array of habitats for a diverse Andean fauna
the Andes are highly diverse. At lower lati- that is far richer than one might expect.
tudes, the slopes receiving moisture-laden More than 700 species of amphibians and
winds are covered with lush tropical forests, reptiles are known to inhabit the Andes. The
which give way at higher elevations to an purposes of this paper are to 1) describe the
elfin forest of stunted trees heavily laden with distributional patterns of the Andean herpe-
thick growths of mosses. Above tree line a tofauna, 2 ) determine the origin of the fauna,
variety of composites, including frailejones 3) hypothesize geological and climatic
and cushion plants are dominant life forms changes that influenced the differentiation and
in the paramos. In drier areas, vegetation dispersal of the Andean herpetofauna, and
may be nearly absent on the slopes and pres- 4) examine the existing herpetofaunal com-
ent only in valleys where bunch grasses form munities in the Andes.
371
372 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
zuela, Cei's (1962) study of Chilean am- species that are designated solely by letters.
phibians, Rivero's (1963a) summary
of the The 727 species of amphibians and reptiles
distribution of Venezuelan Andean frogs, known to occur in the Andes were tabulated
Donoso-Barros' (1966) account of the reptiles for 1) altitudinal ranges, 2) major habitats
of Chile, Roze's (1966) summary of Venezue- occupied, and 3) physiographic regions in-
lan snakes, and Cochran and Goin's (1970) habited (Appendices 15:1-3). Only species
account of Colombian frogs have been useful occurring above 1,000 m
are included. Many
sources of information, as have been the more species primarily inhabiting lowlands, and
limited papers on the Loja Basin in Ecuador only peripherally inhabiting Andean slopes,
2. Serrania de —
Turumiquire. Isolated
Substantial distributional data are in-
highland mass in northeastern Vene-
corporated numerous systematic studies, as
in zuela.
follow: Brame and Wake (1963) on sala- 3. Cordillera —
de la Costa. The coastal
manders of the genus Bolitoglossa; Cei 1971,
ranges of northern Venezuela.
(
vian frogs; and Veloso and Trueb (1976) on 9. Nudo de Pasto.—The highland mass in
southern Colombia and extreme north-
frogs of the genus Telmatobius.
1979 DUELLMAN: HERPETO FAUNA OF ANDES 373
ern Ecuador from which the Colombian in Bolivia, southern Peru, and north-
and Ecuadorian cordilleras diverge. ern Argentina.
10. Cordillera Occidental in Ecuador. — 24. Andes in northern Argentina.
—The
The western Andean range. ranges east of the Altiplano to 27°S
11. Cordillera Oriental in Ecuador. —The Lat.
eastern Andean
range. 25. Andes in northern Chile. —The ranges
12. Inter-Andean Basins Ecuador. —The
in west of the Altiplano to 27°S Lat.
high valleys lying between the eastern 26. —
Andes in southern Ctule. The west-
and western ranges. ern slopes of the Andes south of 27°S
Lat.
D. Huancabamba Depression 27. Andes in southern —
Argentina. The
13. Huancabamba Depression. —The low eastern slopes of the Andes south of
ranges and basins in northern Peru 27°S Lat.
and southern Ecuador.
Many of the regions are definitive physio-
(Senecio, Acaena, etc.), and some low frogs of genus Tehnatobius (currently
the
spiny sclerophylls (Ephedra, Adesmia, being studied by Linda Trueb), lizards of the
etc.). genus Proctoporus (currently being studied
7. Puna. —The montane habitat above tree by Thomas H. Fritts), and snakes of the genus
line that is drier than paramo and ex- Atractus. Taxonomic problems still remain in
tends from southern Ecuador to north- the large iguanid lizard genus Liolaemus and
ern Argentina is called puna. The puna the hylid frog genus Gastrotheca. Based on
is dominated by bunch grasses (Fes- the recent rate of acquisition of new species
tuca, Poa, and especially Stipa); in of frogs of the genera Centrolenella, Colo-
many extensive areas, grasses are the stethus, and Eleutheroclactylus, many more
only evident vegetation (Fig. 15:7). species remain to be discovered in unexplored
Composites, such as Baccharis, Lepi- and poorly collected ranges and valleys. At
dophyllum, and Senecio, are wide- the present time the Ecuadorian Andes, Me-
spread, whereas low trees (Polylepis) rida Andes, Cordillera de la Costa in Vene-
and cushion plants (principally the zuela, and southern Chilean Andes probably
umbellifer Azorella) are local in their are the best-known regions, whereas the Cor-
distributions. In many areas the puna dillera Central in Peru, Cordillera Oriental
1979 DUELLMAN: HERPETOFAUNA OF ANDES 375
"r -•i
J^*wi<irr i^>i flF
Fig. 15:1. Arid slopes on the west face of the Andes, 6 km E Lo Valdes, 2250 m, Provincia de Santiago,
Chile.
le.
Ladcras dridas en el hdo oeste de los Andes, 6 km E Lo Valdes, 2250 tn, Provincia de Santiago, Chile.
Fie. 15:2. Cloud forest in Cordillera de la Costa, Rancho Grande, 1100 m, Estado de Aragua, Venezuela.
Selva de neblina en la Cordillera de la Costa, Rancho Grande, 1100 m, Estado de Aragua, Venezuela.
376 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fig. 15:3. Dry forest in an eastern valley, 15 km S Quiroga, 1750 m, Departamento de Cochabamba, Bolivia.
Selva seca en un valle del este, 15 km S Quiroga, 1750 m, Departamento de Cochabamba, Bolivia.
Fig. 15:4. Nothofagus forest atLago de Huechulaf quen, 900 in, Provincia de Neuquen, Argentina.
Selvas de Nothofagus en el Lago de Huechulafquen, 9(H) m, Provincia de Neuquen, Argentina.
1979 DUELLMAN: HERPETOFAUNA OF ANDES 377
Fig. 15:5. Paramo dominated by Espeletia in Paramo El Angel, 14 km SW Tulcan, 3340 m, Provincia dt
Carchi, Ecuador.
Paramo dominado por Espeletia en el Paramo El Angel, 14 km SO Tulcan, 3340 m, Provincia de Carchi
Ecuador.
Fig. 15:6. Patagonian scrub, with scattered Adesmia, on south slope of Paso El Choique, 1950 m, Provincia
de Mendoza, Argentina.
Matorral patagonico, con Adesmia dispersas, en la ladera sur del Paso El Choique, 1950 m, Provincia de
Mendoza, Argentina.
378 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
'*m%hto
Fig. 15:7. Bunch grass puna, Pampas de Ramoscruz, 31 km W Orcos, 4120 m, Departamento de Ayacucho,
Peru.
Puna con pasto en champas, Pampas de Ramoscruz, 31 km O Orcos, 4120 m, Departamento de Ayacucho,
Peru.
~<fr- ,
'
-
?*v
$8B&&
in northern Peru, Cordillera Oriental in Bo- Andes (Liddle, 1946); but the Cordillera de
livia, and the Andes of northern Argentina la Costa apparently was uplifted little or not
probably are the poorest-known regions. De- at all in the Quaternary. Although some
spite these limitations of the data and in- workers (e.g., Rivero, 1964) distinguish two
complete knowledge of the taxonomy, it is major physiographic regions of the Cordillera
reasonable to determine broad distributional de la Costa (Serrania del Litoral including
patterns and to interpret these patterns on the the Serrania de Paria, and Serrania del In-
basis of generalized phylogenetic relation- terior including the serranias incorporating
ships and historical events in the Andes. Fi- Cerro Turumiquire and Cerro Periquito), I
nally, an apology. I have written more about have recognized one major unit, the Cordil-
(and hopefully most accurately) the regions lera de la Costa, and two isolated small
(
northern Andes ) and organisms ( frogs ) with —
ranges the Serrania de Paria and Cerro Tu-
which I am most familiar. rumiquire. The Cordillera de la Costa is sep-
arated from the Merida Andes by the low
(±500 m), subhumid Depresion de Bar-
PATTERNS OF DISTRIBUTION quisimeto. The highest peak in the Cordillera
de la Costa is Pico Naiguata at 2765 m
Eachof the six major Andean units is
(Aguerrevere and Zuloaga, 1937); this and
discussed with respect to 1) historical geol- 13 other peaks more than 2000 are in the m
ogy, 2) physiography, 3) climate, and 4) littoral which rises
section of the cordillera,
herpetofauna. With the exception of the lat-
abruptly from the Caribbean coast (Fig. 15:
ter, the reader is referred to Simpson (this 11). The interior section of the Cordillera
volume) for more complete discussions. The de la Costa is broad and has few peaks ex-
distributional data on the herpetofauna are
ceeding 1000 m. Disjunct to the east is Cerro
summarized in Appendices 15:1-3.
Turumiquire (2630 m). The Depresion de
Unare separates the main part of the Cor-
Venezuelan Andes dillera de la Costa from the narrow ranges on
the peninsulas of Araya and Paria, where
The highest mountains in Venezuela are
only two peaks exceed heights of 800 m
the Merida, or Venezuelan, Andes. The com- The higher (>850 m)
(Davey, 1949).
plex ranges and intermontane valleys that northern (seaward) slopes of the cordillera
comprise these mountains have a length of support luxuriant cloud forest, which dis-
about 400 km and breadth of about 100 km
integrates into dry forest on most of the lee-
(Oppenheim, 1937). The highest elevation is
ward slopes. Cloud forest occurs at elevations
5002 m on Pico Bolivar; snow line is between above 1500 m
on Cerro Turumiquire and as
4600 and 4700 m (Jahn, 1934). The Merida low as 600 m
in the Serrania de Paria.
Andes are separated from the Cordillera Ori- In addition to the general references cited,
ental by elevations of less than 600 m in the distributional data for the Venezuelan high-
Depresion de San Cristobal, or Cucuta, at the lands were obtained from Donoso-Barros
Colombian-Venezuelan border. According to
(1968), Rivero ( 1963a,b, 1964, 1968, 1972,
Liddle (1946), the Merida Andes were dis-
1974), Rivero and Mayorga (1973), Test, Sex-
tinct from the main chain of the Andes until ton and Heatwole (1966), Williams (1974),
the end of the Eocene; the last orogeny began and Williams, et al. (1970). Published data
in the Oligocene and continues to the present
were supplemented by information provided
(Shagan, 1975). According to Schubert by Scott J. Maness and by material collected
(1974), significantly high elevations were at- by me in 1974.
tained only late in the Pleistocene. Paramo
Forty-eight amphibians and 32 reptiles oc-
habitats,which descend to 2400 m in some cur principally at elevations of more than
places, comprise only3500 sq km of area in 1000 m in the Venezuelan highlands. 1 Of
the Merida Andes (Jahn, 1931).
'Rivero ("1976" [1978]) named three additional
The Cordillera de la Costa had a similar
species of frogs (Colostethus) from the Merida
geological history of uplift as the Merida Andes.
1979 DUELLMAN: HERPETOFAUNA OF ANDES 381
Fig. 15:11. The Venezuelan Andes. A. Merida Andes; B. Cordillera de la Costa; C. Cerro Turumiquire;
D. Serrania de Paria.
Los Andes venezolanos. A. Los Andes de Merida. B. Cordillera de la Costa. C. Cerro de Turumiquire.
D. Serrania de Paria.
these, only six are confined to elevations of Two widespread frogs (Hyla labialis
dilleras.
more than 2500 m. All six occur in the Merida and CentroleneUa buckleyi) inhabit subpara-
Andes and include four frogs (Eleutherodac- mo and paramo from 2000 to 2700 m in the
tylus boconoensis, E. ginesi, E. lancinii, Atelo- Merida Andes and similar habitats at 2400 to
pus mucubajiensis) and two lizards (Anadia 3000 m (Hyla) and 2100 to 3400 m (Centro-
bitaeniata, A. brevirostris) .
Only three other leneUa) in Colombia. The other species
—
species exceed 2500 m — the
salamander, Bo- Gastrotheca nicefori ( 1575 )
m
Anolis nigro-,
Utoglossa orestes, and the frog, CentroleneUa punctatus (1200 m), Chironius monticola
buckleyi, in the Merida Andes, and the frog, (1000-1600 m), Leimadophis bimaculatus
Colostethus mandelorum, on Cerro Turumi- (1400-2500 m), and Micrurus mipartitus
quire. (800-2000 m) inhabit cloud forest on both
sides of the Depresion de San Cristobal.
Only four species are known from the
Serrania de Paria and seven from Cerro Tu- Other species in the Merida Andes have af-
finities with species in the Colombian Andes.
rumiquire, each having three and five en-
demic species, respectively. 2 Of the 46 species Atelopus oxyrhynchus and A. mucubajiensis
in the Cordillera de la Costa, 34 are endemic. are members of the Atelopus ignescens group,
Thus, in these three areas, specific endemism which is speciose in Colombia and Ecuador
is 71-75 percent. Only 18 (56%) of the 32 and occurs in the Sierra Nevada de Santa
species in the Merida Andes are endemic.
Marta (Rivero, 1963b). Hyla platydactyla is
a member of the Hyla bogotensis group con-
Despite the low (600 m) separation of the
Merida Andes from the Cordillera Oriental taining four species in the main Andean Cor-
in Colombia, seven species occur in both cor- dillera.One salamander Bolitoglossa savagei,
(
The major faunal affinities between the Cordillera Oriental in Colombia; only seven
Merida Andes and Cordillera de Costa are
la species are in common. The most speciose
among six snakes — Atractus badius, Chironius genera in the Venezuelan highlands are Eleu-
monticola, Dendrophidion percarinatus, Lam- therodactylus, Colostethus, Centrolenella,
propeltis triangulum, Leimadophis zweifeli, Anolis, Anadia, and Atractus, all of which are
and Micrurus mipartitus, all but two of which widespread and diverse in humid lowland
have lower distributional limits of less than and foothill habitats. Therefore, it seems most
1000 m. No frogs or lizards are shared by likely that the species of these genera that
the two Cordilleras. are endemic to the highlands were derived
Two species of frogs (Colostethus her- from lowland ancestral stocks. A minor per-
minae and Eleutherodactylus urichi) are centage of the fauna of the Merida Andes ap-
shared between the Cordillera de la Costa parently was derived from highland stocks in
and Cerro Turumiquire; the Eleutherodacty- the Cordillera Oriental of Colombia. In the
lus also occurs in the Serrania de Paria and latter group are Bolitoglossa orestes, B. sa-
on Trinidad. According to John D. Lynch vagei, Atelopus mucubajiensis, A. oxyrhyn-
(pers. comm.), many of the Eleutherodacty- chus, Gastrotheca nicefori, Hyla labialis, Hyla
lus in the Cordillera de la Costa have affinities platydactyla, Centrolenella bucklcyi, and An-
with West Indian species rather than with olis nigropunctatus. Certainly in contrast to
those in the Andes of Colombia and Ecuador. the Colombian and Ecuadorian ranges of the
This faunal relationship also appears in the Andes, the fauna of the Merida Andes is
hylid frog Flectonotus pygmaeus that inhabits depauperate; this suggests that the Depresion
the Cordillera de la Costa and Isla Tobago de San Cristobal has been an effective barrier
north of Trinidad. Two other species are in- to the dispersal of most highland groups.
cluded in the genus — F. fitzgeraldi on Trini- Furthermore, the recency of elevation of the
dad and F.
fissilis
in the highlands of south- Merida Andes, combined with late Pleisto-
eastern Brasil. cene glaciation of the small areas now sup-
Five inhabitants of cloud forest in the porting paramo, may be partly responsible
Venezuelan highlands (2 frogs Gastrotheca — for the few high montane species of amphib-
longs to a group having a species in lower of only about 16,000 sq km. The highest ele-
Central America (P. lemur), one on the east- vations, such as Pico Cristobal Colon at 5775
ern slopes of the Andes in Ecuador ( P. buck- m, are perpetually covered with snow. The
leyi) and an unnamed species on the Pacific
, sierra rises abruptly from the Caribbean
slopes in Colombia (Duellman, 1970). coastal plain and is narrowly separated by
Two monotypic genera of snakes ( Umbri- arid lowlands at elevations of less than 500 m
vaga mertensi and "Urotheca" williamsi) are from the northern part of the Cordillera Ori-
endemic to the Cordillera de la Costa. No ental of the Andes; the northern part of the
species are shared with the Guiana Highlands. cordillera along the Colombian-Venezuelan
Thus, each of the
regions four of the border the Siena de Perija, which attains
is
Venezuelan highlands has endemic species of elevations of 3750 m. The first uplift of the
amphibians and reptiles; those of the two Sierra Nevada de Santa Marta was in the
small highland areas (Cerro Turumiquire and Miocene, but the final uplift did not occur
Serrania de Paria) seem to have been derived until the end of the Pleistocene (Gansser,
from the Cordillera de la Costa, which shares 1955). According to Carriker (1922), cloud
few species with the Merida Andes (Fig. 15: forest occurs at elevations of 1385 to 2200
m,
12). Thirty-two species occur in the Merida
but have observed cloud forest on Cerro
I
Andes, as compared with 103 species in the San Lorenzo (= Cerro Kennedy) at 2700 m.
1979 DUELLMAN: HERPETOFAUNA OF ANDES 383
384 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
cc
46/34
03/36
Herpetofaunal comparisons of the Sierra Nevada de Santa Marta with other highlands. Numbers
Fig. 15:13.
in blocks are numberof species/number of endemic species; numbers of shared species are within arrows. CC
= Cordillera de la Costa; CCC = Cordillera Central de Colombia; COcC = Cordillera Occidental de Colombia;
COrC = Cordillera Oriental de Colombia; = Merida Andes; SM = Sierra Nevada de Santa Marta.
MA
Comparaciones faunisticas de la herpetofauna de la Sierra Nevada de Santa Marta con otras tierras de
alturas. Nitnieros dcntro de bloques representan numero de especies/ numero de especies endemicas; numero
de especies en comun estdn en las flechas.
de la Costa in Venezuela, some from the dered to the northwest by the Rio Patia, to
Merida Andes, some from the Cordillera Ori- the northeast by the headwaters of the Rio
ental, and some from the surrounding low- Caqueta, and to the south by the Rio Chota.
lands (Fig. 15:13). With the exception of Much of the nudo is above 3000 m
with two
the widespread Micrurus mipartitus, no spe- peaks, Volcan Chiles and Volcan Cumbal,
cies are shared with the Central American reaching 4760 m.
highlands. Much exploration remains to be The Andes north of the Nudo de Pasto
done in the Sierra Nevada de Santa Marta, a form three distinct ranges. The western range,
region that probably has many more species the Cordillera Occidental extends for about
than known at present. Also, the geographi- 650 km between the Pacific lowlands and the
cally important but biologically unexplored valley of the Rio Cauca. The southern border
Sierra de Perija may hold the key to under- is the Rio Patia; the dry upper valley of the
standing the fauna! relationships of the Sierra river at about 1200 m separates the Cordillera
Nevada de Santa Marta. Occidental from the Nudo de Pasto. The
Cordillera Occidental is narrow ( <50 km)
Northern Andes and lacks continuous high ridges; the two
The northern Andes are comprised of five highest peaks are Cerro Tamana (4200 m)
and Pico Frontino (4080 m).
major north-south ranges diverging from the
Nudo de Pasto and a series of high intermon- The Cordillera Central extends 750 km
tane basins in Ecuador; the entire northern north from the Nudo de Pasto; this range
Andes extend for about 1800 km from the about 100 km in width is bordered on the
Caribbean lowlands at 10°50'N to the Huan- west by the valley of the Rio Cauca and on
cabamba Depression at 4°30'S. Central to the east by the valley of the Rio Magdalena.
the physiography of the northern Andes is the Extensive areas are above 3000 m, and four
high massif of the Nudo de Pasto in southern peaks with permanent snow exceed 5000 m
Colombia and northern Ecuador (Fig. 15:
— Nevado del Huila (5760 m), Nevado del
14). The Nudo de Pasto encompasses a Quindio (5400 m), Nevado del Ruiz (5400
north-south extent of about 110 km and a m), and Nevado del Tolima (5215 m).
breadth of about 130 km. The nudo is bor- The Cordillera Oriental is narrowly sep-
1979 DUELLMAN: HERPETOFAUNA OF ANDES 385
io° -
Fig. 15:14. The northern Andes. A. Sierra Nevada de Santa Marta; B. Cordillera Occidental de
Colombia; C. Cordillera Central de Colombia; D. Cordillera Oriental de Colombia; E. Nudo de Pasto;
F. Cordillera Occidental de Ecuador; G. Cordillera Oriental de Ecuador; H. Huancabamba Depres-
sion. The inter-Andean Basins lie between the Cordillera Occidental and the Cordillera Oriental in
Ecuador.
Los Andes del norte. Las hoijas interandinas estdn entre la Cordillera Occidental y la Cordillera
Oriental en Ecuador.
386 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
The northern Andes have received con- based on the extensive collections in the Mu-
siderable geological study (Biirgl, 1961; Herd seum of Natural Histoiy at The University
and Naeser, 1974; Sauer, 1965, 1971; Shagam, of Kansas and in the National Museum of
of the western cordilleras. The absence of (40), snakes 73 (57). Seven genera (Am-
evidence of glaciation on some high peaks phignathodon, Centrolene, Osornophryne,
suggests that their final uplift occurred after Phenacosaurus, Pholidobohis, Saphenophis,
the last major glaciation. Nonetheless, most Synophis) are endemic to the northern Andes,
areas above 3700 m
were glaciated; climatic and one (Cryptobatrachus) is endemic save
depression was in the magnitude of 6-7°C for one species in the Sierra Nevada de Santa
with a downward shift of environments of Marta.
about 1000 to 1200 m
(van der Hammen, Of the 70 species having ranges extending
1974). beyond the limits of the northern Andes, 43
The extensive areas above tree line are are the adjacent lowlands.
also present in
humid and cool with annual precipitation of Some of these also occur in the central Andes,
1000 to 2000 mm and little seasonal fluctua-
especially species that inhabit rainforest and
tion temperature, but daily variation of
in cloud forest. Thus, 15 species are shared be-
10°C or more ( Cuatrecasas, 1968). These tween the Cordillera Oriental in Ecuador and
areas paramo have grasses (Festuca),
of the Cordillera Central in Peru, and 20 are
rosette herbs (Espeletia and Senecio), cushion shared between the former and the Cordillera
plants (Distichia), and low bushes (Baccha- Oriental in Peru. Fourteen species on the
ris). The western slopes of the Cordillera Pacific slopes of the Cordillera Occidental in
Occidental and the eastern slopes of the Cor- Colombia and Ecuador, and /or the northern
dillera Oriental from the Depresion de San
parts of the Colombian cordilleras are shared
Cristobal southward support luxuriant cloud with the highlands in lower Central America;
forests. These humid montane forests also
these include six frogs (Gastrotheca nicefori,
occur locally in the Cordillera Central, espe-
Hemiphractus fasciatus, Centrolenella fleisch-
cially in the northern part. Subparamo is manni, C. griffithsi, C. prosoblepon, C. vaJ-
common but localized in the eastern and erioi), five lizards (Anolis antonii, A. chloris,
western cordilleras. The inter-Andean basins Basiliscus galeritus, Polychrus gutturosus, Pri-
in Ecuador possibly were subparamo prior to uertebralis), and three snakes
onodactylus
human modification into cultivated fields and
(Dendrophidion percarinatus, Micrurus mi-
grazing of livestock. partitus, Bothrops schlegeli). With
the ex-
In addition to the general publications al- of Gastrotheca and Centrole-
ception nicefori
ready cited, I have drawn information from nella griffithsi, all of these species range well
the works on anurans by Duellman ( 1972, below 1000 m. As noted previously, four
1973), Duellman and Altig (1978), Duellman
species are shared with the Sierra Nevada de
and Simmons (1977), Lynch ( 1975a,b, 1976a,
1979), Lynch and Duellman (1973, 1979),
3
Not included in these figures or in Appendix 15:2
Myers and Daly (1976a,b). and Peters are Atelopus carauta from 1300 m
in the Cordillera
Occidental of Colombia and Hernandez
Ruiz
(1973). The works Myers (1973, 1974) and
(
Santa Marta, and seven are shared with the southernmost basin ( Loja ) is considered to be
Merida Andes. part of the Huancabamba Depression.
only those species that do not occur
If Only considering those 73 species that do
below 2500 m
are considered, 72 of the 73 not occur below 2500 m, a much different
species are endemic to the northern Andes. picture is evident (Fig. 15:17). Endemism in
The exception is the boid snake Tropidophis each region ranges from 20 percent in the
taczanowskyi, which inhabits the Cordillera Nudo de Pasto to 100 percent in the Cordil-
Oriental in Ecuador and the Cordillera Cen- lera Occidental in Colombia. Whereas the
tral in Peru. eastern cordilleras in Colombia and Ecuador
Within the northern Andes, only two spe- and the western cordilleras in Colombia and
cies, Centrolenella buckleyi (2100-3400 m Ecuador shared the greatest numbers of spe-
in subparamo and paramo) and Eleuthero- cies when the entire fauna was considered, in
1230-2800 m in cloud an analysis of only the high montane species,
dactylus w-nigrum (
forest and subparamo) occur in all seven re- they have no species in common.
The greatest species richness and highest
gions. Eleutherodactylus vertebralis (2340-
3500 m
in subparamo and paramo) occurs in percentage of endemism in the northern An-
the three Colombian cordilleras, the Nudo des is amongst the frogs (especially Colo-
de Pasto, and the Cordillera Occidental in stethus, Eleutherodactylus, and Centrole-
Ecuador. Four species of frogs (Eleuthero- nella), which form 60 percent of the entire
Andean herpetofauna, but 65 percent in the
dactylus buckleyi, E. unistrigatus, Atelopus
northern Andes and 71 percent in the Cor-
ignescens, and Gastrotheca riobambae) dis-
tributed mostly above 2500 occur in four m dillera Oriental in Ecuador. Although the
three snakes entire fauna in the equatorial cordilleras is
regions; (Chironius monticola,
Rhadinaea latehstriga, and Micrurus miparti- large, local communities are much smaller.
the northern Andes, 155 species occur in two ules in the cloud forest, subparamo, and par-
or three regions, and 250 are endemic to a amo, with generally decreasing numbers of
at higher altitudes. Equatorial tran-
given region, with the largest number of en- species
demics in the Cordillera Oriental in Ecuador sects in the Cordillera Occidental and Cor-
dillera Oriental reveal the presence of 62 and
(74) and the Cordillera Occidental in Ecua-
dor (65), but with the highest percent of 79 species, respectively (Fig. 15:18). On the
endemism in the Cordillera Central in Co- eastern slopes there is a dimunition of species
lombia (54%). at about 2000 m; this is especially evident
Table 15:1. —
390 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
COcC COrC
CCC
10 > < 24 >
66/29 103/36
78/42
33 45
COcE COrE
< 15 * 10-
22/3
137/65 164/74
-21-
Fig. 15:16. Herpetofaunal comparisons of the regions of the northern Andes. Numbers in
blocks are numbers of species/number of endemic species; numbers of shared species are
within arrows. CCC = Cordillera Central de Colombia; COcC = Cordillera Occidental de
Colombia; COcE = Cordillera Occidental de Ecuador; COrC = Cordillera Oriental de
Colombia; COrE = Cordillera Oriental de Ecuador; IAB = Inter-Andean Basins; NP = Nudo
de Pasto. One species is in common between CCC and IAB.
Comparaciones faunisticas de la herpetofauna de las regimes dc los Andes del norte.
Numcros dentro de bloques rcpresentan numero de cspecies /ni'imero de cspecies cndemicas;
el numero de cspecies en comun estdn en las
flechas. Una cspecies cs compartida por CCC
y IAB.
1979 DUELLMAN: HERPETOFAUNA OF ANDES 391
Fig. 15:17. Herpetofaunal comparisons of the regions of the northern Andes using only
species thatdo not occur below 2500 m. Numbers and abbreviations are same as in figure 16.
Comparaciones faunisticas de la herpetofauna de las regiones de los Andes del norte,
comprendiendo solo aquellas especies que no bajan de los 2500 m. Numeros y abreviaciones
igual que en la figura 16.
392 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
PACFC
1979 DUELLMAN: HERPETOFAUNA OF ANDES 393
3000-
2000-
1000
M
Fig. 15:20. Profile of the Huancabamba Depression at 5°15'S Lat.
Perfil de la Depresion de Huancabamba a 5°15'Lat. S.
ECUADOR
394 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Kilometers
16
82
Fig. 15:22. The central Andes. A. Cordillera Occidental North; B. Upper Maranon Valley; C. Cordillera
Central; D. Upper Huallaga Valley; E. Cordillera Oriental North; F. Cordillera Occidental South; G. Man-
taro-Apurimac Valley; H. Cordillera Oriental South; I. Cordillera Oriental de Bolivia; J. Altiplano.
Los Andes centrales.
1979 DUELLMAN: HERPETOFAUNA OF ANDES 395
SOUTHWEST NORTHEAST
6000
5000
4000
3000
2000
1000
M
Fig. 15:23. Profile of the Cordillera Oriental in southern Peru.
nota, and Cordillera de Carabaya) having steep slopes dissected by deep ravines are
many peaks with permanent snow, the high- known as the Yungas.
est of which is the Nevado Salcantaya (6271 Rising from the lowlands of the upper
m). Lower front ranges (Cadena de Pau- .Amazon Basin, several mountain ranges
cartambo, Cadena de Pantiacolla) do not ex- reaching above 2500 m
are isolated from the
ceed 4500 m; the deep valleys separating the main Andean cordillera. These ranges in Peru
various ranges give a relief of 2000 to 3000 m are, from north to south, Cerros de Otanahui,
to the cordillera (Fig. 15:23). The major Cordillera Azul, Cerro de la Sal, and Serrania
rivers dissecting the mountains are the Rio de Sira.
Apurimac, Rio Urubamba, and Rio Vilcanota. The major montane valleys separating the
In Bolivia the eastern cordillera consists of principal cordilleras are those of the Rio
a single range, the Cordillera Real. In the Maranon, and Rio Huallaga flowing north-
northwest and to the southeast there are two ward in northern Peru, and the Rio Mantaro
—
ranges the Cordillera Central separated by flowing southeastward in central Peru, and
the Rio Caipe from the outer range, the Cor- the Rio Apurimac flowing northwestward in
dillera Oriental. The highest peaks are in the southern Peru. The Mantaro and Apurimac
snow-covered Cordillera Real, where four converge to form the Rio Ene; thus, these two
peaks exceed 6000 m, and the highest is Cerro valleys are placed together for the purposes
Illimani (6460 m). The Cordillera Real drops of analysis.
precipitously into the Amazon Basin; the The initial uplift of the central Andes was
396 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
high eastern slopes. The eastern and northern one of these inhabits dry forest in the northern
slopes of the easternmost ranges of the eastern Andes. Eight species (including 2 of 23 noted
cordilleras support lush cloud forest, the above) occur in the central Andes and the
crests of these outer ranges have subparamo Huancabamba Depression. Only 12 species
and/or wet puna habitats. The western and are shared between the central Andes and the
southern slopes of the outer ranges are dry southern Andes (including the Altiplano);
with puna and low xerophytic vegetation. one of these also ranges into the cis-Andean
The high ridges of the interior ranges of the lowlands, two occur on the Pacific lowlands,
eastern cordilleras are drier than the outer and nine are restricted to high ( > 2500 ) m
ranges and have extensive areas of puna. The elevations in the central Andes and at least
deep valleys between the ranges are dry with in the northern parts of the southern Andes.
low sclerophytic vegetation at higher eleva- No species of amphibian or reptile occurs
tions and dry scrub forest at lower elevations. in all nine of the regions within the central
The high montane valleys are extensively Andes. The most widespread species are the
cultivated and also support puna, much of frogs Pleurodema marmorata, Telmatobius
which is grazed. marmoratus, Bufo spinulosus, the lizard Lio-
The diverse herptofauna of the central laemus multiformis, and the snake Tachy-
Andes has never been summarized. In addi- menis peruviana. These are the only species
tion to the general works cited previously, the that occur in eastern and western cordilleras
following works are pertinent to the system- and intermontane valleys. The major simi-
atics and distribution of amphibians and rep- larities inthe cordilleras are among the east-
the central Andes: Barbour and Noble
tiles of ern ranges, which individually share 9 to 14
(1920a) on southern Peruvian taxa, Dixon species, with three species occurring in all
and Huey (1970), Dixon and Wright (1975), four regions (Fig. 15:24). Only three species
Fritts (1974), and Uzzell (1969, 1970) on are shared between the northern and southern
lizards; Schmidt and Walker (1943) and parts of the Cordillera Occidental, and no
Walker (1945) on snakes; Duellman (1976, more than five species are common to any
1978a-c), Duellman and Fritts (1972), Duell- part of the Cordillera Occidental and any
man and Toft (1979), Gallardo (1961), Ma- range in the eastern cordilleras. The similari-
cedo (1960), Schmidt (1954), Silverstone ties of the eastern ranges are principally in
(1975, 1976), and Vellard (1951-1960) on those species that inhabit the cloud forests
frogs. between 1000 and 2000 m; the three species
The herpetofauna of the central Andes, as with distributions including all four regions
presently known, consists of 159 species (75, of the eastern cordilleras are forest dwellers
1979 DUELLMAN: HERPETOFAUNA OF ANDES 397
Fie. 15:24. Herpetofaunal comparisons of the Andean ranges in the central Andes. Numbers in blocks are
number of species/number of endemic species; numbers of shared species are within arrows. CC = Cordillera
Central; COcN = Cordillera Occidental North; COcS = Cordillera Occidental South; COrB = Cordillera Ori-
ental de Bolivia; COrN = Cordillera Oriental North; COrS = Cordillera Oriental South.
Comparaciones faunisticas de la herpetofauna de las Cordilleras andinas de los Andes centrales. Numeros
dentro de bloques representan numero de especies /numero de especies endemicas; numero de especies en comun
estdn en las flcchas.
—two lizards (Prionodactijlus argulus and P. The species richness in the southern part
manicatus) and one snake (Chironius monti- of the Cordillera Oriental reflects the
complex
cola). The faunal list from the Cordillera topography and diverse habitats in that region
Central is unrealistic; the area has not been (Fig. 15:25). Elevational changes in one
studied adequately. inter-Andean valley result in striking differ-
398 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Southern Andes
Fig. 15:26. The southern Andes. A. Northern completely separated from the principal cor-
B. C. Northern Argentina; D. dillera. In the south, the Andes descend only
Chile; Altiplano;
Southern Chile; E. Southern Argentina. to the Patagonian plateaus at 600 to 1300 m.
Los Andes del sur.
The Pacific slopes of the Andes in northern
The Andes of northern Argentina have ex- elevations. From north to south (about 28°S
tensive areas over 4500 m and some peaks Lat.) there is a change from cloud forest to
400 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
evergreen forest and deciduous broadleaf for- and Veloso (1977) reviewed Pleurodema, and
est. At about 26°S Lat. to 38°S Lat., the Lynch (1978b) summarized data on Alsodes.
eastern slopes of the Andes are arid with de- The iguanid lizards have been studied by Cei
ciduous forest existing in river valleys and (1971, 1973, 1974a,b), Cei and Castro
Patagonian scrub infiltrating the lower slopes (1973), and Donoso-Barros (1972).
to 2000 m ( Roig, 1972 ) the Patagonian scrub
;
The herpetofauna of the southern Andes
crosses the continental divide at some low consists of 64 species (30 frogs, 31 lizards, 3
passes. South of 38°S Lat. on the Pacific snakes). No genera are endemic to the
slopes and 36°S Lat. on the eastern slopes are southern Andes, but austral endemics such as
the austral forests characterized by a diversity Alsodes are shared with the lowland forests;
of Nothofagus and in places dominated by Diplolaemus and Phymaturus are shared with
Araucaria or Fitzroya (Formas, this volume). Patagonia, and Garthia is shared with the
Within tins area is the so-called lake region Pacific coastal deserts. Of the 64 species, 36
of the southern Andes where cold streams cas- are endemic. Of the 28 nonendemic species,
cade down from glaciers. 12 also occur in arid habitats east of the
The Andes and Argen-
in southern Chile Andes, principally in Patagonia (e.g., the
tina were uplifted nearly to their present iguanid lizards Diplolaemus leopardinus, Lio-
heights by the end of the Miocene (Dott, et laemus bibronii, L. elongatus). Six others oc-
al., 1977). At that time, the initial uplift of cur in lowland Nothofagus forests (e.g., frogs
the northern part of the Argentine, Chilean, such as Alsodes nodosus and Bufo variega-
and Bolivian Andes and the Altiplano took tus ) five occur on the arid Pacific lowlands
;
place (Petersen, 1958; Rutland, et al., 1965). (e.g., lizards such as PhyUodactylus gerrhopy-
The final, major orogeny of the Altiplano and gus and Tropidurus peruvianus) . The faunal
the principal Cordilleras in northern Argen- similaritiesbetween the central and southern
tina and Chile was completed by the end of Andes have already been discussed.
the Pliocene (Turner, 1972; Yrigoyen, 1972; Within the southern Andes, only two spe-
James, 1971, 1973), whereas at least some of cies, the toad Bufo spinulosus and the snake
the extra-cordilleran ranges in Argentina were Tachymenis peruviana, occur in all five re-
elevated later (Simpson and Vervoorst, gions. Despite the high Andean divides, the
1977), and the coastal cordillera of Chile was southern Andes of Chile and those in Argen-
uplifted earlier( Okada,
1971 ) The southern
. tina sharemore species (10) than any other
Andes were extensively glaciated during the two regions (Fig. 15:27). Most of the en-
Pleistocene, with large montane glaciers per- demics are frogs of the genera Alsodes and
sisting to the present; at the height of glacia-
Telmatobius restricted to separate stream
tion, all of the Andes south of 30°S Lat. were drainages and highland lizards of the genus
entirely glaciated (Patterson and Lanning, Liolaemus.
1967; Heusser, 1974). The distribution of amphibians and rep-
The herpetofauna of Chile has been re- tiles on either side of the Andes in relation
viewed thoroughly by Cei ( 1962 ) and to passes through the high cordillera suggests
Donoso-Barros (1966, 1970). Cei (1979) re- that available structural habitat (or perhaps
viewed the amphibians of Argentina, but the food) may limit their distributions instead of
reptiles have not been summarized. The only altitude and the associated climatic stresses.
significant paper on the Altiplano is Parker's For example, at 33°S Lat. the cordillera is
(1940). Important works dealing with frogs breached by Puerto Bermejo at 3883 m. In
are, genus Telmatobius Vellard
as follow: — the immediate vicinity of the pass two lizards,
(1946), Callardo (1962), Laurent (1970, Liolaemus altissimus and L. fxtzgeraldi, reach
1973, 1977), Veloso and Trueb (1976), and elevations of 3500 m and 2800 m, respectively,
Cei (1977); genus Gastrotheca Laurent
— and occur on both sides of the Andes (Fig.
(1967, 1969a,b, 1976); genus Bufo— Gallardo 15:28). In the same area four other species
(1967), Cei (1968, 1972). Barrio (1965) dis- of lizards occur only on the eastern slopes —
cussed the Hijla pulchella complex; Duellman Liolaemus elongatus (up to 2800 m), L. rui-
1979 DUELLMAN: HERPETO FAUNA OF ANDES 401
Fig. 15:27. Herpetofaunal comparisons of regions in the southern Andes. Numbers in blocks
are numbers of species/number of endemic species; numbers of shared species are within
arrows. AAN = Andes of Argentina north; AAS =Andes of Argentina south; ACN =
Andes
of Chile north; ACS = Andes of Chile south; ALT = Altiplano.
Comparaciones faunisticas de la herpetofauna de las regiones de los Andes del sur.
Numeros dentro de bloques representan numero de especies/ numero de especies endemicas;
numero de species en comun estdn en las flechas.
ball (2900 m), Phymaturus palluma (3500 other hand, farther south there is continuity
m), and Pristidachjlus scapulatus (2900 m). of habitat through much lower passes. For
Likewise, two species are restricted to the example, Patagonian scrub continues through
—
western slope Liolaemus leopardinus (3000 Puerto de Buta Malin (37°30'S, 1800 m) onto
m) and L. nigroviridis (3000 m). On the the western side of the cordillera to Laguna
402 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Table 15:2. — Faunal Resemblance of the Herpetofaunain Six Major Andean Regions.
Numbers of species in a given region are in boldface; numbers of species in common to two regions are in
Roman, and the faunal resemblance factors t2C/(N, + Ns) = FRF] are in italics.
Sierra
Venezuelan Nevada de Northern Huancabamba Central Southern
Andes Santa Marta Andes Depression Andes Andes
Venezuelan Andes . 80 0.040 0.028 0.000 0.008 0.000
Sierra Nevada de Santa Marta ... __ 2 21 0.018 0.0(H) 0.000 0.000
Northern Andes 7 4 415 0.035 0.008 0.000
Huancabamba Depression 8 43 0.079 0.000
Central Andes 1 23 8 159 0.116
Southern Andes . 13 64
1979 DUELLMAN: HERPETOFAUNA OF ANDES 403
Serroni'a de Pana
Cerro Turumiquire
Cordillera de la Costa
Merida Andes
Altiplano
00 01 0.2 0.3 04 05
SIMILARITY
Fig. 15:29. Cluster analysis of 727 species of amphibians and reptiles in 27
physiographic regions of the Andes; analysis is by the unweighted pair-group
method using arithmetic means.
Andlisis de agrupaeion de 727 especies de anfibios y reptiles en 27 regiones
fisiogrdficas de los Andes. El andlisis usa el metodo de grupos de parejas no
compensadas en sus promedios aritmeticos.
dispersal in the Andes (Fig. 15:31). The minor barriers are within major regions be- —
physiographic barriers in the northern part of tween the Merida Andes and the Cordillera
the Andes coincide with those demonstrated de la Costa, and between the Cordillera Occi-
for birds by F. Vuilleumier ( 1977 ) The ma- . dental in Colombia and the rest of the north-
jor physiographic barriers separate the major ern Andes. The major ecological barrier is
—
Andean regions Merida Andes, Sierra Ne- the drastic change from cloud forest to rela-
vada de Santa Marta, northern Andes and tively dry puna, which follows the upper
central Andes —
from one another, whereas reaches of the outer ranges of the eastern cor-
404 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Menda Andes
Cordillera Occidental, Colombia
Huancabamba Depression
Cordillera Oriental North, Peru
£ Cordillera Central, Peru
Altiplano
00 1 0.2 03 04 05 06 07 08 09 10
SIMILARITY
Fie. 15:30. Cluster analysis of 147 species of amphibians and reptiles occurring only above 2500 in 23 m
physiographic regions of the Andes; analysis is by the unweighted pair-group method using arithmetic means.
Andlisis de agrupacion de 147 especies de anfibios ij reptiles que habitan solamente por encima de los
2500 m en 23 regiones fisiogrdficas de los Andes. Metodo al igual que figura 29.
dillera in the central Andes. Comparatively It is obvious that species richness is high-
broad latitudinal transition zones exist be- est in those tropical regions supporting both
tween the cloud forest and the deciduous for- cloud forest and equable habitats above tree
ests on the eastern slopes in northern Argen- line —Cordillera Oriental in Ecuador ( 164
tina, and between the Nothofagus forests of species), Cordillera Occidental in Ecuador
the southern Andes and the arid slopes to the (137 species), Cordillera Oriental in Colom-
north. Faunal comparisons of these eight ma- bia (103 species). Between the Equator and
jor ecogeographic regions in the Andes reveal 24 °S Lat. there is a dramatic decline in spe-
that the regions separated by major geo- cies richness, most notable in amphibians;
graphical barriers or
ecological differences farther south there is little change (Fig.
have faunal resemblance factors of less than 15:32).
0.1,whereas those separated by minor physio- Endemism is as high as 76 percent in the
graphic barriers have factors greater than 0.1 Sierra Nevada de Santa Marta, whereas the
(Table 15:3). average percentage of endemic species in any
1979 DUELLMAN: HERPETOFAUNA OF ANDES 405
lines = major physiographic barriers; single solid lines por sobre los 1000 m estdn incluidas.
= minor physiographic barriers; broken lines = major
ecological barriers; shaded areas
= transition zones Mapa diagramdtico de los Andes mostrando ba-
between ecologically different regions. A. Cordillera rreras para la dispersion de la herpetofauna. Dos
de la Costa; B. Merida Andes; C. Sierra Nevada de lineas continual = barreras fisigrdficas mayores; una
Santa Marta; D. Cordillera Occidental de Colombia; linea continua = barrera fisiogrdfica menor; lineas
E. Northern Andes; F. Eastern slopes of central discontinuas = barreras ecologicas mayores; areas
Andes; G. Nonforested central and southern Andes; sombreadas = zonas de transition entre regiones
H. Forested southern Andes. diferentes ecologicamente.
406 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Table 15:3. — Faunal Resemblance of the Herpetofauna in Eight Major Eco-geographic Regions of the Andes.
Numbers of species in a given region are in boldface; numbers of species in common to two regions are in
Roman, and the faunal resemblance factors [2C/(N, +N-) = FRF] are in italics.
S V
O
2
=3
c
O O
u +^
c
13 O 9
13 o
C 3 C/3
fa
<u 13
6 o
I J3
3
o o o o
O O Z fa
Cordillera de la Costa .. 46 0.129 0.030 054 0.008 0.013 0.000 0.000
Merida Andes 5 32 0.075 082 0.034 0.014 0.000 0.000
Sierra Nevada de Santa Marta 1 2 21 023 0.019 0.000 0.000 0.000
Cordillera Occidental, Colombia
Northern Andes
Eastern Front, Central Andes
Nonforested Cent. & S. Andes
Forested Southern Andes
110
3
2
2
7 4
1 66
37
1
0.163
386
23
1
0.011
0.093
107
7
0.000
0.004
0.049
78
3
0.000
0.000
0.000
0.055
31
one of the 27 regions is 40.9 (Table 15:4). available for most regions of the Andes. As
The apparent absence of endemics in the Cor- already discussed in the account of the north-
dillera Central of Peru probably is indicative ern Andes, a distinct middle elevation fauna
of the poor sampling of the region, whereas occurs in cloud forest and another element on
the low values for the Nudo de Pasto, inter- the high ridges; species richness is much
Andean valleys in Ecuador, and the Marafion higher in the cloud forest, and endemism is
Valley in Peru are realistic in that most spe- likely to be high in both, certainly higher
cies in these regions are shared with adjacent than in the adjacent lowlands (see especially
regions. The presence of many species of a Table 15:4 and Figs. 15:18-19).
given genus in one region is indicative of
local speciation and endemism. This is espe-
cially true of frogs of the genera Eleuthero-
EXTRA-ANDEAN FAUNAL
dactylus, Colostethus, and Centrolenella in the
RELATIONSHIPS
northern Andes, where the species per genus
is 3.3 to 3.9 in the eastern cordilleras of Co- Once the Andean herpetofauna has been
lombia and Ecuador and the Cordillera Occi- defined and circumscribed, it is necessary to
dental in Ecuador; this high number is ap- ascertain the relationships of the faunas of
proached only by the Cordillera Oriental in the Andes before attempting to determine the
southern Peru (2.8), where there is also great origin of the Andean herpetofauna. Primarily
differentiation of Eleutherodactylus, and the this will be done at the generic level. Eighty-
Altiplano (2.7) where there are six species of eight genera of amphibians and reptiles occur
Liolaemus (Table 15:4). However, there is at elevations of more than 1000 in the m
no direct correlation between the total per- Andes. More than half (47) of these genera
centage of endemic species and the numbers are not primarily Andean in their distribution.
of species per genus in that region; for exam- Some are represented in the Andes by lowland
ple, the genus per species value is 1.6 in the species that ascend the highlands to eleva-
Sierra Nevada de Santa Marta, where 76 tions of more than 1000 (
m
13 genera ) and
percent of the species are endemic. others (23 genera) have species restricted to
Substantial data on altitudinal differences Andean slopes, although most of the species
in species richness and endemism are not occur elsewhere. The Andes are peripheral to
1979 DUELLMAN: HERPETOFAUNA OF ANDES 407
Table 15:4. — Genera and Species of Amphibians and Reptiles in Each of 27 Regions of the Andes.
% % Species
Northern Northern Endemic Species per
Region Genera Genera Species Species Species Endemic Genus
Serrania de Paria 3 100
Cerro Turnniiquire 5
Cordillera de la Costa .._ . 26
Merida Andes 16
Sierra Nevada de Santa Marta - 13
Cordillera Occidental, Colombia 27
Cordillera Central, Colombia .. — - 34
Cordillera Central, Peru 12
Cordillera Oriental, Colombia 30
Cordillera Oriental, Ecuador 42
Cordillera Occidental, Ecuador — . 41
Nudo de Pasto, Colombia __ 11
Cordillera Oriental North, Peru _ 30
Cordillera Oriental, Bolivia 21
Inter-Andean Basins, Ecuador 12
Huancabamba Depression 23
Cordillera Oriental South, Peru 28
Andes North, Argentina 9
Cordillera Occidental North, Peru _— _ 10
Mantaro-Apurimac Valleys, Peru 8
Upper Maranon Valley, Peru 4
Upper Huallaga Valley, Peru 5
Altiplano _ 6
Cordillera Occidental South, Peru 9
Andes North, Chile 10
Andes South, Chile 9
Andes South, Argentina 9
408 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Table 15:5. —Patterns of Andean Distributions of Genera that are Principally Extra- Andean.
Lowland Species Differentiated on Peripheral to Range
on Lower Slopes Low-moderate Slopes of Genus
Leptodactylu.t Caecilia Euparkerella
1
Phyllobatcs Epicrionops Ischnocnema
Gonatodes Dendrohates Flectoiwtus
Polychrus Osteocephalus Phyllodactylus
Tropidums PhyUomcdusa Basiliscus
Alopoglossus Glossostoma Diplolaemus
Cercosaura Garthia Phymaturus
Neusticurus Lepidoblepharis Tropidophis
Ptychoglossus Pseudogonatodes Lampropeltis
Liotyphlops Anolis Liophis
Clelia Enyalioides Sibon
Leptoph is Pristidactylus Sibynomorphus
Micntrus Leptotyphlops
Cliironius
Dcndropliidion
Dipsas
Lcimadophis
Mastigodryas
Oxyrhopus
Philodryas
Rhadinaea
T ant ilia
Bothrops
Most of the species of Pbyllobates recognized by Silverstone (1976) and Appendix 15:2 have been
1
listed in
19 genera are endemic to the Andes; the 93 petofauna, I have chosen to analyze the
species in these genera and 337 Andean spe- herpetofauna first by means of areas of con-
cies in eight other genera that are primarily gruence. Subsequent analysis of generalized
Andean account for about 60 percent of the tracks of those groups that are well known
Andes has been summarized by Simpson (this stantial evidence for the recognition of a
volume), who also synthesized the evidence temperate South American adaptive radiation
for Late Cenozoic and Quaternary climatic of tropidurine iguanid lizards. Although the
changes; these phenomena also have been taxonomy of the large iguanid genus Liolae-
treated in detail for tropical South America by mus is still in an alpha state, it is evident that
Haffer ( this volume ) and for temperate South there a major diversity of species groups in
is
America by Baez and Scillato Yane (this Patagonia and the adjacent Andean foothills
volume). Modern phylogenetic approaches and that by comparison with these regions,
using cladistics as the basis for vicariance or the Liolaemus diversity is less in the Austral
dispersal biogeography (Croizat, et al., 1974; forests,Atacama Desert, and the high Andes.
Rosen, 1975; Platnick and Nelson, 1978) are Cei (1974a, pers. comm.) considers Ctenoble-
ideal for analysis of faunal origins and sub- pharis to be derived from Liolaemus.
Steno-
sequent differentiation, but in most groups of cercus seems to be a highland sister group of
Andean amphibians and reptiles, the phylo- the lowland Ophryoessoides (Fritts, 1974);
genetic relationships have yet to be deter- these genera apparently are derived from a
mined in any such refined manner. Procto-
tropidurine stock, perhaps close to
Two current methods of determining tretus (Etheridge, pers. comm.). Taehymenis
centers of dispersal (and possibly of origin) is the only genus of snakes widespread in
involve generalized tracks (Croizat, et al., the cool temperate region and the high
Table 15:6. — Distribution of Andean Genera of Amphibians and Reptiles in Adjacent Extra-Andean Regions
and Neotropical Highlands.
(
* = Endemic to Andes; numbers of species are Andean/Total )
South-
Central Amazon Guiana eastern Austral Atacama
Genus Species America Choco Basin Highlands Brasil Patagonia Forests Desert
Caecilia 12/29 +
Epicrionops 4/9
Bolitoglossa 14/63
Alsodes 6/7
Amblyphrynus" 2/2
Batrachophrynus* 2/2
Eleutherodactijlus 153/s400
Euparkerella 1/2
Ischnocnema ... 1/3
Leptodactylus 2/32
Phrynopus" 14/14
Pleurodema 4/12
Telmatobius" 30/30
Geobatrachus" 1/1
Atelopus 27/39
Bufo 20/ == 190
Osornophryne* _ 2/2
BJiamphopIiryne 3/6
Colostethus 39/64
Dendrobates 4/20
Phyllobates _ 3/20
Amphignathodon* .... 1/1
Cryptobatrachus" 3/3
Flectonotus 1/2
Gastrotheca - 33/40
Hemiphractus . 4/5
Hyla 21/=275
Osteocephalus 2/6
Phyllomedusa 4/29
Centrolene" .... _ 1/1
Centrolenella 45/64
Glossostoma 1/2
Garthia ~- 1/2
Gonatodes _ 3/17
Lepidoblepharis 1/8
Phyllodactylus 3/60
Pseudogonatodes _ 1/5
Anolis _ 18/=200
Basiliscus 1/4
Ctenoblepharis _ 5/11
Diplolaemus 1/3
Enyalioides _ 2/8
Liolaemus ._ 24/52
Phenacosaurus" 3/3
Phytnaturus 1/2
Polychroides" _ 1/1
Polychrus .__ 2/5
Pristidactylus ._ .._ 1/4
Stenocercus . 26/29
Twpidurus 2/20
Alopoglossus 1/6
Anadia .. 7/11
Cercosaura 1/3
Euspondylus 5/9
Macropholidus' 1/1
Neusticurus 3/8
Opipeuter" 1/1
Pholidobolus* 5/5
1979 DUELLMAN: HERPETOFAUNA OF ANDES 411
cal history and the distribution of the toads Atelopus apparently is most closely related to
in question. Melanophryniscus in subtropical South Amer-
NorthernHcrpetofaunal Assemblage.
— ica (McDiarmid, 1971). Trueb (1971) of-
Thirty-two genera and species groups com- fered various alternative hypotheses for the
prise this assemblage. Some of these genera relationships of Rhamphophryne; the most
are disjunct from the major area of congru- plausible one geographically is that Rham-
ence in the Andes of southern Colombia and phophryne is related to, but not derived
Ecuador, where 22 of the groups occur. The from, Atelopus.Ruiz and Hernandez (1976)
disjunct groups are in northern Colombia demonstrated that Osornophryne is a spe-
(Cryptobatrachus, Phenacosaurus, Euspondy- cialized derivative of Atelopus.
therodactylus fitzingeri group ancestor; frogs 1970). Although Gastrotheca also occurs in
of that group are primarily lowland in distri- southeastern Brasil, the major center of differ-
entiation is in the northern Andes; Amphig-
bution, although some species (notably E.
nathodon seems to be merely a Gastrotheca
w-nigrum) inhabit the Andes. Lynch (1975b)
with mandibular teeth.
analyzed the broad-headed eleutherodactyline
frogs and showed that Amblyphrynus was Although Centrolenella is diverse in Cen-
most closely related to frogs of the Eleuthcro- tral America, southeastern and the
Brasil,
dactylus sulcatus group in the Amazon Basin; Guiana Highlands, it most speciose on the
is
he suggested that Amblyphrynus might be an Andean slopes. Members of all species groups
Andean eleutherodactyline prototype. Three of Centrolenella occur in the northern Andes,
bufonid genera occur in the northern Andes — as does the only other genus in the family,
vada de Santa Marta and the Merida Andes; Polychrus, although Gorman, et al. (1969)
414 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
found no karyological differences and con- and the Bufo typhonius group, both of which
sidered Polychroides to be indistinguishable have species widely distributed in the low-
from Poly citrus. The other lizards are micro- lands of Central America, the Choco, and
teiids; the relationships of these have been the Amazon Basin, but which have species
partially analyzed by Presch (
1978 ) ,
who con- endemic to the Andean slopes.
sidered Prionodactylus, Proctoporus, and Opi- In summary, the Andean herpetofauna
peuter tobelong one lineage, Euspondylus
to evolved from ancestral stocks from the low-
and Pholidobolus to a second, and Anadia to lands. These are readily assignable to two
a third that also includes the genera Ecpleo- major —
Andean assemblages northern and
pus and Placosoma, which are extra-Andean. southern. The dispersal of the groups is re-
Of
the remaining groups in the northern flected in their diminishing importance in the
herpetofaunal assemblage, the microteiid liz- faunas of regions more remote from their
ard genera Euspondylus and Prionodactylus, centers of dispersal (Table 15:4) and the
also occur in the lowlands and have related peripheral limits of their distributions (Figs.
(perhaps derived) genera in the highlands
— 15:34-35). It is evident that the central
Pholidobolus and Proctoporus, respectively. Andes in Peru are a major area of transition
The highly speciose dendrobatid genus Colo- between the northern and southern assem-
stethus has some species in the lowlands, blages.
many species on the Andean slopes, and a
few species in the high Andes; Edwards DISPERSAL AND DIFFERENTIATION
(
1974 ) concluded from an analysis of Colo-
stethus that most of the high Andean species By comparison with the other highland
were derived from groups on the lower slopes. regions in South America, the Andes are very
The most speciose group in the northern young. Although some uplift occurred prior
Andes is the Eleuthcrodactylus unistrigatus to the Miocene, the major tectonic events that
group. Lynch (1976b) identified the center resulted in the uplift of the Andes began only
of diversity of thegroup as the Andes of about 20 million years ago with major
Colombia and Ecuador, but numerous species orogenies occurring toward the end of the
occur in the upper Amazon Basin, on the Pliocene and through the Pleistocene all —
slopes of the Cordillera Oriental in Peru, in within the last 2.5-4.5 million years. Climatic
the Sierra Nevada de Santa Marta, and the fluctuation during the Pleistocene is a final
Merida Andes, and in the Cordillera de la factor in the complex and geologically recent
Costa. The greatest species richness at any series of events that have molded the Andes
one site is in the cloud forests and in the and their environments into the features that
upper Amazon Basin. Those species groups we see today. Thus, the early evolution and
in the high Andes were derived from diverse dispersal of the herpetofauna must be viewed
ancestral stocks at lower elevations (Lynch, with respect to the mid-late Cenozoic events,
pers. comm.). whereas much of the dispersal and differen-
Extra-Andean Groups. —Many species in- tiation is intricately associated with Quater-
Fig. 15:34. Northern limits of distribution of genera of the southern herpetofaunal assemblage.
Limites de distribution norte de los generos del con junto herpetologico surefio.
416 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Fig. 15:35. Southern limits of distribution of genera of the northern herpetofaunal assemblage. Additional
species of Atelopus and Eleutherodactylus have entered the eastern slopes of Peru and Bolivia from the
Amazonian lowlands.
Limites de distribution sur de los ueneros del conjunto herpetoldgico norteno. Espeties adicionales de
Atelopus y Eleutherodactylus han cut rado a las laderas estes de Peru y Bolivia desde las tierras hajas del
Amazonas.
1979 DUELLMAN: HERPETOFAUNA OF ANDES 417
occurred in the Pleistocene, when the Andean with predictions of numbers of species and
orogeny thrust the cordilleras to their present endemics with island ( = paramo ) size and
heights, and the cold Humboldt Current as- distance from the presumed source of immi-
sumed its present course. Prior to the present gration (see also Mauriello and Roskaski,
positioning of the Humboldt Current, more 1974), but he did not consider Pleistocene
precipitation would have fallen on the west- climatic effects to be important to the avi-
ern slopes of the Andes in Peru and northern fauna. Haffer (1970, 1974) and van der Ham-
Chile (Simpson, 1975a). Furthermore, cli- men ( 1972 ) emphasized that Pleistocene cli-
matic fluctuation during the Pleistocene pro- matic fluctuations have been a primary factor
vided at least intermittent, less xeric condi- in the expansion of ranges of the Andean
tions along the coast (Campbell, 1976, 1979). biota. Simpson ( 1975b ) performed an analy-
The lacustrine deposits in the Altiplano prob- sis of distribution of some groups of the
ably are of Pleistocene age, when the area paramo flora and found higher correlations
had an extensive lake system (Lohmann, between areas of glacial paramos, their dis-
1970). tances from source areas, and the number of
During the Middle and Late Tertiary, the species than with the modern paramo param-
northern Andes were being elevated. Differ- eters.
ent ranges had somewhat different histories Few systematic and evolutionary studies
with respect to uplift and volcanism, but the on Andean amphibians and reptiles have at-
final uplift of all of the tropical ranges oc-
tempted to explain patterns of speciation and
curred in the late Pliocene and Pleistocene distribution as the result of Pleistocene cli-
(Ahlfeld, 1970; Gansser, 1955; Sauer, 1971; matic changes (Duellman, 1972, 1974; Fritts,
van der Hammen, et al., 1973). Climatic 1974; Montanucci, 1973). It is constructive to
change was principally a cooling effect with examine selected groups of the herpetofauna
continued humid, but probably more equable, in light of Late Tertiary and Quaternary tec-
temperatures, with the exception of fluctua- tonic events and climatic changes.
tion of temperatures and precipitation corre- —
Northern Paramos. The proposed depres-
lated with glacial and interglacial stages (van sion of tree line by about 1000 m at glacial
der Hammen, et al., 1973). The structural de- maxima would have resulted in extensive
pressions between some of the principal cor- areas of paramo where now only relatively
dilleras (e.g., Cauca and Magdalena valleys small islands exist (Simpson, 1975b).
paramo
and Huancabamba Depression) became more F. Yuilleumier's (1970) and Simpson's
significant as the cordilleras were elevated ( 1975b )
results on avian and floral distribu-
(Simpson, 1975b). suggest that herpetofaunal
tions, respectively,
Early workers, such as Chapman ( 1917 ) patterns should show positive correlations be-
and Chardon (1938), attempting to explain tween the numbers of species and paramo-
the origin, speciation, and distribution of high island size and distance from areas of emigra-
Andean biotic components, were hampered by tion. One of the basic tenets of MacArthur
the then persistent ideas of the antiquity and and Wilson's ( 1967 ) theory of island biogeog-
immutability of the lowland forests and little raphy is that there is dispersal from a source
or no knowledge of the Pleistocene climatic area and that the number of immigrants that
changes in the Andes. B. S. Vuilleumier become established on a given island is pro-
418 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
portional to the size of the island and in- tion and speciation during times of climatic
versely proportional to its distance from the elevation, 4) dispersal and secondary contact
source area. An important consideration in resulting in sympatry at subsequent time of
this concept is the mode of dispersal —active climatic depression, and 5) present isolation
sects), passive aerial (some plants and some Similar patterns are evident in the micro-
insects), and passive waiting (everything teiid lizards of the genus Pholiclobolus, five
else). With a few exceptions, amphibians species of which inhabit the Andes from 1800
and reptiles apparently reach oceanic islands to 3960 m in Ecuador; with the exception of
by rafting. However, rafting is not a means of two cases of sympatry, their distributions are
dispersal between montane islands. There- confined to particular combinations of Cor-
fore, the presence of a species of frog or dilleras and/or inter- Andean valleys (Mon-
lizard on presently isolated paramo islands tanucci, 1973). Likewise, frogs of the Eleu-
cannot be explained by direct dispersal; in- therodactyhis curtipes complex show allopat-
stead they must be considered as relictual ric distributions in the high Cordilleras in
isolates of former more widespread distribu- southern Colombia and Ecuador, as do frogs
tions. Brown ( 1971 reached the same con-
)
of the genus Phrynopus in paramos and sub-
clusion with respect to North American mon- paramos from Colombia to Peru ( Lynch,
tane mammals. 1975a) and members of the Gastrotheca mar-
Climatic depression of as much as 1000 m sapiata group in Peru ( Duellman and Fritts,
would have united most of the present para- 1972 ) The two species of high montane toads
.
mos in the Cordillera Oriental in Colombia (Osornophryne) are now isolated in separate
into one extensive paramo; the same is true paramos in the Cordillera Central of Colom-
for those in the Cordillera Central of Co- bia and the Nudo de Pasto, but climatic de-
lombia, the Nudo de and the Cordilleras
Pasto, pression of 1000 m would have united all of
in Ecuador (Fig. 15:36). There are groups of these paramos (Table 15:8).
species that are endemic to the present para- Some phyletic lineages occur at high ele-
mos in the Cordillera Oriental of Colombia. vations in the Cordilleras in Colombia and
For example, the Eleutherodactylus elegans Ecuador as well as in the Merida Andes
complex of the E. unistrigatus group has three and/or the Sierra Nevada de Santa Marta.
species (affinis, bogotensis, elegans) in the Thus, the 11 species in the Atelopus ignescens
paramos in the vicinity of Bogota, one species group are distributed principally above 2S00
(lynchi) farther north in the Paramos de m in the cordilleras of Colombia and Ecuador
Guantiva, Rusia, and Vigajual, and another (8 species), above 2350 m in the Sierra Ne-
(nicefori) still farther north in the Paramo vada de Santa Marta (1 species), and above
de Almorzadero. Only two species occur in 2010 m in the Merida Andes (2 species). As-
sympatry (bogotensis and elegans in Paramo suming that this group dispersed from the
de Choachi). All of the species currently live major northern Andean ranges, climatic de-
at elevations above 2800 m. Thus, the distri- pression would have had to have been more
bution of this species complex can be ex- than 1000 m in order to allow them to disperse
plained by 1) origin from one stock in the into the Sierra Nevada de Santa Marta and
Cordillera Oriental, 2) widespread dispersal the Merida Andes or at the time of dispersal
during time of climatic depression, 3) isola- the frogs were able to tolerate warmer condi-
Table 15:8. —Genera and Species Groups Endemic, or Primarily Restricted, to Regions of Major Pleistocene
Paramo- Islands.
(Letters refer to locations in fig. 36.)
10'
-10'
o°-
tions than they do now. The same is true for pus, Rhamphophryne, Gastrotheca, Hemi-
—
two other montane frogs Hyla labialis (pres- phractus, Hyla, and Centrolenella have close
ent range: 2000-3000 m) and Centrolenella relatives on opposite sides of the Andes ( Fig.
buckleiji (2000-3400 m). 15:38). One member of the predominately
Other groups of amphibians and reptiles lowland Elciitherodactylus fitzingeri group
in paramos and subparamos have distribu- (E. iv-nigrum) occurs at elevations of 1230-
tions that also extend down into cloud forests. 2800 m
in cloud forests on both slopes of the
Their presence as isolates in various ranges three cordilleras in Colombia and on the
in the Andes can be explained by their for- Pacificand Amazonian slopes in Ecuador, and
merly having wider distributions that in- the snake Chironius monticola ranges through
cluded now uninhabitable lowlands. As the cloud forests from Venezuela to Bolivia (Fig.
climates in the intervening lowlands became 15:39). Some of these groups also occur in
less equable, the animals became isolated in the cloud forests in the highlands of Panama.
various montane habitats. Examples are frogs —
Each of the following groups Rhampho-
of the genus Cryptobatrachus, lizards of the phryne, Gastrotheca, Hemiphractus, and the
genus Anadia, and salamanders of the genus —
Hyla bogotensis group has one representa-
Bolitoglossa (Fig. 15:37). The enigmatic frog tive there and many in the Andean cloud
Geobatrachus in the Sierra Nevada de Santa forests.
Marta presumably is a Pleistocene relict of —
Central and Southern Andes. Although
some unknown lowland ancestor. most taxonomic groups of the herpetofauna in
—
Cloud Forests. The montane rainforests the southern Andes are restricted to either
or cloud forests are fragmented by low val- the Patagonian or the Pacific slopes, some
leys and by high cordilleras. Presumably at groups have related species on both sides —
times of depression of vegetation zones owing Alsodes, Telmatobius, and Liolacmus. In the
to climatic factors, cloud forest extended far south, some Patagonian species follow
across some of the lower areas now supporting present continuous Patagonian scrub through
lowland tropical rainforest or dry forest. Thus, low Andean passes; thus Patagonian species
dispersal routes would have been available such as Pleurodema bufonina, Diplolacmus
between the Sierra Nevada de Santa Marta, leopardinus, Liolacmus kriegi, and Phymatu-
the Merida Andes, the Cordillera de la Costa, rus palluma occur on the Pacific slopes of
and the Cordillera Oriental in Colombia, Andean passes. Farther north, mostly differ-
thereby permitting dispersal of some cloud ent species occupy similar ecological niches
forest groups, such as Bolitoglossa, Eleuthero- on the two slopes, but Liolacmus fitzingeri has
dactylus, Atelopus, and Centrolenella in these populations isolated on both slopes. This
areas. Likewise, dispersal between cloud for- suggests that in the not too distant past, cli-
ests in the Ecuadorian and Peruvian cor- matic conditions were elevated sufficiently
dilleras would have been possible across the that the lizards were able to disperse across
Huancabamba Depression, where cloud for- the Andean divide.
ests exists now only on isolated ridges. Cli- The existing, continuous high montane
matic fluctuation apparently was important in habitats presently allow dispersal for such
the dispersal and speciation of stream inhabi- widespread species as Pleurodema marmo-
tants such as Colostcthus and Centrolenella, rata, Liolacmus multiformis, Tachymenis
many species of which are isolated in head- peruviana and members of the Bufo spinulo-
waters of confluent drainages but do not occur sus group; all are present on the Altiplano
now in the areas of confluence. and in the cordilleras and high inter-Andean
Trans-Andean dispersal of cloud forest in- valleys northward at least to central Peru
habitants must have occurred at times of (Fig. 15:34). The extensive and more con-
climatic elevation or when the Andes were tinuous high montane regions in the central
lower, with continuous cloud forest from the Andes of Peru, Bolivia, and northern Argen-
eastern to the western slopes. Numerous spe- tina and Chile seem to have afforded few
cies of the genera Elciitherodactylus, Atelo- opportunities for isolation and differentiation
1979 DUELLMAN: HERPETOFAUNA OF ANDES 421
-10°
Bolitoglossa
422 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
10' -\
-10°
Fig. 15:38. Distribution of the two primarily Andean species of horned frogs, genus Hemiphractus.
Distribucion de dos especies primariamentc andinas de amtros comudos, genero Hemiphractus.
1979 DUELLMAN: HERPETOFAUNA OF ANDES 423
10'
-10'
o°-
Fig. 15:39. Distribution of the snake Chironius monticola, an inhabitant of Andean cloud forests; the range
extends southward into Bolivia.
Distribution de la culehra Chironius monticola, tin habitante de la selva de neblina; su rango se extiende
hacia el sut hasta Bolivia.
424 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
of high montane species during the Pleisto- the dispersal and speciation of inhabitants of
cene. The notable exception is in the eastern the Andean cloud forests. The greatest
front ranges of the Andes where paramo amount of isolation of high montane species
islands apparently persisted, expanded, and took place in the dissected northern Andes
contracted resulting in isolated populations and southward in the front ranges of the
and speciation in frogs (Phrynopus and Gas- Cordillera Oriental in Peru. Hypothesized
trotheca) and lizards (Proctoporus) . Pleistocene parameters of paramos commonly
Two genera are diverse in the central encompass the ranges of species or groups of
Andes and deserve individual consideration. closely related species now isolated on
Lizards of the genus Stenocercus primarily smaller paramo islands. Inability of high An-
live in xeric habitats below 3500 m. Fritts dean species to disperse across unsuitable
(1974) analyzed the evolutionary relation- habitats between Pleistocene paramos re-
ships of the species and showed that three sulted in major highland areas each having
phyletic lines evolved in different parts of the groups of species independently derived from
central Andes; he considered that the com- lowland stocks, or at least separate centers of
plexities of the Andean orogenies in the Plio- differentiation in these areas (Fig. 15:40).
Pleistocene and climatic fluctuation in the Thus, the results of F. Vuilleumier ( 1970) and
Pleistocene were the primary factors in spe- Simpson ( 1975b ) relating to avian and floral
ciation events in Stenocercus. The important montane island biogeography differ from
thing here is that isolation was in valleys, not those dealing with the herpetofauna. Wide-
on mountain tops. At present, 23 species of spread distributions of some high montane
the aquatic frogs, Telmatobius, are recognized species in the Altiplano and the central Andes
in the central Andes; two others occur in the are indicative of post-Pleistocene dispersal
southern Andes and five others in the Huan- through existing suitable habitats. Tectonic
cabamba Depression and Ecuador. On the
in and glacial events in the central Andes af-
Altiplano and in the Peruvian Andes most fected the dispersal and isolation of inhabi-
species occur at elevations in excess of tants of the presently xeric dry valleys and
3000 m, where they live in cold streams; in aquatic inhabitants of the high cordilleras,
many cases they are found most commonly whereas in the southern Andes, these events
in small rivulets near the crests of ridges at resulted in the isolation of populations on
more than 4500 m. Most species are allo- eastern and western Andean slopes.
patric. It seems likely that the dispersal and
subsequent isolation and speciation of these
was associated with climatic fluctuations
ANDEAN HERPETOFAUNAL
frogs
and actual
COMMUNITIES
glaciation of the high central
Andes in the Pleistocene. The systematic rela-
tionships of the species of Telmatobius pres-
Community studies of the Andean herpe-
tofauna are limited to a few papers. Pefaur
ently are too poorly known to assess fully the
and Duellman (1977) reported on community
historicalbiogeography of the group.
The other aquatic high Andean frog, Bat- structure of the herptofauna at 10 Andean
analyses primarily were concerned
sites; their
rachophrynus, represented by two species in
central Peru, apparently is an ancient relict
with niche breadth and niche overlap with
of a group having a widespread distribution respect to structural habitat and food. Test,
southward prior to the uplift of the Andes. Sexton, and Heatwole (1966) reported on the
In summary, the patterns of dispersal and reptiles from the Rancho Grande region in
the Cordillera de la Costa in Venezuela; some
differentiation of the high Andean herpeto-
inferences about the reptile community can
fauna are intimately associated with the ele-
vation of the Andes in the late Pliocene and be gained from their paper. Lynch ( 1976a,
Pleistocene, glacial events during the Pleisto- 1979) and Lynch and Duellman (1979) com-
cene, and climatic fluctuation during the mented on eleutherodactyline frog communi-
Pleistocene. Likewise, these events influenced ties in Ecuador.
1979 DUELLMAN: HERPETOFAUNA OF ANDES 425
Fig. 15:40. Distributional limits of three species groups of montane marsupial frogs, genus Gastrotheca.
Limites distribucionales de ires grupos de especies de los anuros marsupiales de montana, genero Gastro-
theca.
426 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
.
6. Paramo de las Hermosas, 7 km NE
of Tenerife, Valle, Colombia, 2850 m 1000 KM
(P).
7. Paramo de Purace, Cauca, Colombia,
3300 m (P). Fie. 15:41. Location of 40 herpetofaunal com-
8. Nudo de Pasto, 8 km NE of Pasto, Na- munities in the Andes. See text for sites designated
rino, Colombia, 3020 m (P).
by numbers.
Localidades para 40 comunidades herpetoldgicas
9. Nudo de Pasto, 12 km E of Pasto,
</<• losAndes. Vcr cl texto para los sitios designados
Narifio, Colombia, 3050 m (P). par numero.
1979 DUELLMAN: HERPETOFAUNA OF ANDES 427
2 3
4
2
5
2
2
3
Centrolenids
Gekkonids
Iguanids
Microteiids
Snakes
_.._
__
5
-
2
__
3
4
1
4
__
3
4
2 11
11 3
1
4
Total Species 19 25 19 19 30 24 26
—
Table 15:10. Comparison of the Herpetofaunas in Seven Communities in Andean Cloud Forest.
[
Numbers of species in a given community are in boldface; numbers of species in common to two communities
are in Roman, and the Coefficients of Community are in italics. )
1979 DUELLMAN: HERPETOFAUNA OF ANDES 429
» *
r T
T
ft
o T
r
r
Fig. 15:42. Three dimensional plot of the multivariate means of 28 species of frogs in a cloud forest com-
munity at theRio Azuela, Ecuador. Axis I is size, increasing from left to right; Axis II is height above ground,
increasing from bottom to top; Axis III is association with water, increasing from front to back. Solid sym-
bols are nocturnal species; open circles are diurnal ones.
Distribution tridimensional de los promedios multivariados de 28 especies de anuros en una eomunidad de
sclva dc neblina en el Rio Azuela, Ecuador. Eje I es tamano, incremcntdndosc de izquierda a derecha; Eje II
es altura sobre el suelo, incremcntdndose de abajo hacia arriba; Eje III es asociacion con ag.ua, incremcntdndosc
desde el frente hacia atrds. Simbolos llenos represcntan especies nocturnas; circulos claros represcntan especies
diu mas.
>
Genus « o <
3 A 3 W
o > U a* 2 5
Bolitoglossa __.. 1
Elcutherodactijlus ... 1
Phnjnopus 1
Tclmatobius -
Colostcthus 1
Atelopus ..
Osornophryne „ ..
Gastrothcca ..
Hula 1
Centrolenella ._
Stenocercus _
Anadia 1
Pholidobolus ..
Prionodactylus ._
Proctoporus
Total Species 6
no species in common with sites in the other est adult male). The ratio of diurnal to noc-
Cordilleras (Tahle 15:12). turnal species varies from 1:3 to 5:3; all of
In the relatively simple paramo communi- the reptiles and frogs of the genera Atelopus
ties,the differential utilization of resources and Colostethus are diurnal. Using the Par-
was measured with respect to 1) distance amo de Vigajual as an example, it can be
from water, 2) utilization of rock cover, 3) seen that the seven species in the community
diel activity, and 4) snout- vent length (larg- are distributed throughout the spectrum of
430 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
—
Table 15:12. Comparison of the Herpetofaunas in Twelve Communities in Northern Andean Paramos.
(Numbers of species in a given community are in boldface; numbers of species in common to two communi-
ties are in Roman, and the Coefficients of Community are in italics. )
>
XI
1979 DUELLMAN: HERPETOFAUNA OF ANDES 431
Table 15:13. — Herpetofaunal Composition of Twenty-one Communities in the Central and Southern Andes.
Genus
432
i— l
— — — —
i l l i l . l ! h N H lO rt O
co <* -* co co cm -**
O M M o « n N
d
CO CO CO CO 1(5 © >o
d O
co in
oddN
<j< co co -* r>-
.S~ ID CO
rH
CM y—| l-H
» s
io >o
O O O O
^H
CM ^H »-H
q ^ n
>o ic
i-i d d d d
NHHHNNIMB SN CO CO
O)
00
0")N
CO t^ CO
© © o o © ©
co co co co
M< CO CO CM CM CM CM
© © © d © d ©
CO CO CO CO CO CO cm © lO CM
CO CO CM CM CM CM
>-i © o o o o o o
MW^WC'lNNNNNlCtON
CC(CCO(N(N^"HM»H
©©'©©©©©©d
rt^USCMOOOt^t^CMOJOCM
CO CO CO CM CM CM CM CM i-h •— <
ooooooo'o'o'o
° « r-l i-H 1» Ot^oOOt^c^CMO"MCM
C0t--CoCOCM^H'"-<CMCMCMCM
o
w £•
.3.
ooo'oo'oooooo
'C
rt
S
3 CM»/5©©— !©©©l^t^
cmcmcocococm^h*^
ooo'o'o'doo
O o Irt O O Cw O O O CO CO CM o
\£cmcmcocdco^cococmcm
CM
CM
IE >
©do© © © © © d © d d
i
i
* CM
CM
oonipEuitfnjj
d
CO N
E '2
3 3
1979 DUELLMAN: HERPETOFAUNA OF ANDES 433
Fig. 15:44. Three dimensional plot of the multivariate means of seven species in a high Andean community
at Santa Rosa, Peru. Axes and symbols are the same as in Figure 43.
Distribution tridimensional de los promedios multivariados de side especies en una comunidad altoandina en
Santa Rosa, Peru. Ejes y simbolos son los mismos que en la figura 43.
Fig. 15:45. Three dimensional plot of the multivariate means of ten species in a Patagonian-Andean com-
munity at Laguna Blanca, Argentina. Axes and symbols are the same as in figure 43.
Distribution tridimensional de los promedios multivariados de diez especies en una comunidad patagonico-
andina en Laguna Blanca, Argentina. Ejes y simbolos son los mismos que en la figura 43.
434 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
populated, heavily cultivated areas have been normal human activity. Furthermore, some
in continual, or nearly continual, use for about of these isolated paramos already have been
8000 years (Engel, 1976). Although it is set aside as national parks, especially in Co-
impossible to state with any degree of cer- lombia and Venezuela. Until the last decade
tainty that human activities have not caused most of the extensive areas of cloud forest
the extirpation of any species of amphibian were subject to little human disturbance. The
or reptile, it seems as though most human steep Andean slopes mostly were accessible
disturbance in the high Andes may be bene- only on foot or with pack animals; settlements
ficialto most members of the herpetofauna. consisted only of small ranchos or tambos one
At least at the present time, man is a predator day's travel apart. However, in recent years,
only on the large frogs Tehnatobius culeus many roads are being constructed from the
and Batrachophryniis in Andean lakes. On highlands to the Amazonian and Chocoan
the other hand, some agricultural practices lowlands; following the roads are colonists
are extremely beneficial to amphibians and who quickly clear the forest. When the road
reptiles. Irrigation systems, mostly small from Papallacta to Lago Agrio, Ecuador, was
channels, provide sites for the development opened in October 1971 it was possible to
of tadpoles of Gastrotheca, Bufo, Pleurodenw, drive for many kilometers through pristine
Cetitrolenella, and Tehnatobius, as well as cloud forest. By April 1972 many of the areas
adults of the latter. The clearing of fields of of cloud forest already were gone. Certainly
rocks and placing these in piles or as fences if the
unique and diverse biota of the cloud
creates seemingly ideal habitat for some frogs forests is to be preserved, several large re-
and especially some lizards of the genera Lio- serves must be set aside in the very near fu-
laemus, Stenocercus, Pholidobolus, and Proc- ture. Although steps have been taken in Peru
toporus. Fence rows of planted Agave also with the Parque Manii and preliminary work
provide excellent shelter for Gastrotheca, in Ecuador with the Proyecto Cayambe-Coca,
Stenocercus, and Pholidobolus. Perhaps the these are insufficient, because so many cloud
most detrimental human disturbance in the forest inhabitants have restricted ranges.
found effects on the herpetofauna. The clear- am grateful to Stephen Ayala, Jose M. Cei,
many species in a given area, but the conse- Raymond F. Laurent, Bruce MacBryde, Jaime
quent deleterious effects of erosion, silting, E. Pefaur, and Alberto Veloso M. for logistic
and chemical changes influence streams below support and introduction to terrain new to
the destroyed areas. Although the clearing me. Much information utilized herein was
of cloud forests is detrimental or even devas- gathered by my field companions; for their
most some and lizards efforts I thank David C. Cannatella, Dana K.
tating to species, frogs
reach large populations in cafetales or in Duellman, Thomas H. Fritts, Stephan Halloy,
forest-edge habitats. Oscar Ochoa M., and especially John E.
1979 DUELLMAN: HERPETOFAUNA OF ANDES 435
Simmons and Linda Trueb. Field studies in habitan a elevaciones mayores que 1000 en m
the Andes were supported by grants from the los Andes fueron tabulados de acuerdo a sus
National Geographic Society (1304) and the distribuciones altitudinales y a los ambientes
National Science Foundation (
DEB 74-01998, ocupados en 27 regiones fisiograficas de los
DEB 76-09986). Andes (Apendices 15:1-3). Estas regiones
John D. Lynch and Jaime E. Pefaur have fueron agrupadas en seis categorias mayores
been constant sources of information and ideas para propositos de analisis de los patrones de
while I have been preparing this distribucion.
manuscript.
Furthermore, Pefaur has brought to my atten- Cada una de las cuatro regiones dentro de
tion much pertinent literature and translated los Andes venezolanos tiene especies endem-
the Spanish resume.Lynch generously shared icas de anfibios y/o reptiles; aquellos de las
with me knowledge of the Andes
his first-hand —
dos pequenas regiones de altura Cerro Tu-
and his unequalled knowledge of Eleuthero- rumiquire y la Serrania de Paria parecen —
dactylus. Linda Trueb and Stephen R. Ed- haber derivado de la Cordillera de la Costa,
wards attempted to answer my multitudinous la cual comparte pocas especies con los Andes
questions about Telmatobius and Colostethns, de Merida. De las 32 especies en los Andes
respectively. Jose M. Cei, Richard E. Ethe- de Merida, solo siete se encuentran el la Cor-
ridge, and Thomas H. Fritts provided insights dillera Oriental en Colombia. En la aislada
into problems related to the Andean iguanid Sierra Nevada de Santa Marta, 16 de las 21
lizards, and the latter tried to relieve my especies en la selva nublada y paramo son
frustrations in dealing with the Andean micro- endemicas. Cuatro de las especies no-endem-
teiid lizards. John A. Wiest, Jr., generously icas tambien ocurren en la Cordillera Orien-
Andes peruanos. Dentro de los Andes del rata, Telmatobius marmoratus, Bufo spinido-
norte, 250 especies con endemicas en una de sus), una de saurios (Liolaemus multifor-
las siete regiones fisiograficas, con el mayor mis), y una culebra (Tachymenis peruviana)
numero de endemismos en la Cordillera Ori- se distribuyen ampliamente y son las unicas
ental en Ecuador (74), y en la Cordillera Oc- especies que ocurren en las cordilleras del
cidental en Ecuador (65), pero con el mayor este y del oeste. La mas alta tasa de endemis-
porcentaje de endemismo (54%) en la Cordi- mo en los Andes centrales ocurre en la parte
llera Central en Colombia. La mayor diversi- sur de la Cordillera Oriental en Peru, donde
dad y el mas alto porcentaje de endemismo 41 de las 76 especies son endemicas.
en los Andes del norte ocurre en los sapos y Los Andes del sur incluyen las cordilleras
ranas especialmente Colostethus, Eleuthero-
(
en Argentina y Chile y el Altiplano. Alii la
dactylus, y Centrolenella), quienes conforman herpetofauna consiste en 64 especies consti-
al 65 por ciento de la herpetofauna del area. tuidas por 30 anuros, 31 saurios y tres cu-
La herpetofauna de las hoyas interandinas lebras. Ningun genero es endemico de los
parece haber derivado principalmente de las Andes del sur, pero ciertos endemicos aus-
cordilleras adyacentes. La mas alta similari- trales como Alsodes se comparten con las
dad faunistica dentro de los Andes del norte selvas de tierras bajas, como Diplolaemus y
ocurre entre las cordilleras del este en Colom- Phymaturus con la Patagonia, y como Garthia
bia y Ecuador, pero cuando las especies de con el desierto de la costa pacifica. De las 64
sobre 2500 m se consideran unicamente, estas especies, 36 son endemicas; de las restantes
regiones no tiene especies en comun. En esta 28, 12 son comunes con la Patagonia, seis con
misma linea de analisis todas las especies las selvas australes, cinco con las tierras bajas
altoandinas de la Cordillera Occidental en aridas del Pacifico, y 12 con los Andes cen-
Colombia son endemicas. trales. La mayoria de endemicos son anu-
los
La Depresion de Huancabamba es la ma- ros de los generos Alsodes y Telmatobius re-
trales, y nueve con el desierto de Atacama. algunas especies quedan agrupadas, en aque-
Las relaciones genericas de la herpetofauna llos casos, caracteristicas de comportamiento
andina se establecen primariamente con las o reproduction tienden a separar a aquellas
tierras adyacentes y no con las tierras de al- especies.
turas del Brasil o de las Guayanas. La actividad humana en los altos Andes
La herpetofauna andina se origino por in- aparentemente ha contribuido poco al detri-
vasiones separadas desde las tierras bajas.Un mento de la herpetofauna; por el contrario, la
conglomerado faunistico sureno esta compu- contraction de canales de irrigation y el apila-
esto de grupos derivados de antiguos inmi- miento de rocas y cercos ha provisto con im-
grantes a la region andina y por inniigrantes portantes ambientes para varios anfibios y
tardios de la Patagonia y selvas australes. El reptiles. El sobre pastoreo puede tener ef-
conglomerado faunistico norteno consiste en fectos negativos en los insectos que son im-
grupos derivados de antiguos y tardios inmi- portantes como alimento de la herpetofauna.
grantes de tierras bajas tropicales. Los mo- La exploration de las selvas nublados en las
delos de dispersion y especiacion de la her- laderas andinas ha tenido graves consecuen-
petofauna altoandina estan estrechamente cias sobre la herpetofauna; solo algunas espe-
asociados con los movimientos orogenicos del cies son capaces de sobrevivir en los claros
Contrib. Paleobiol. (27): 1-211. Dixon, J. R., Huey, R. B. 1970. Systematics of the
Campbell, K. E., Jr. 1979. The Pleistocene avi- lizards of the gekkonid genus Phyllodactylus of
fauna of the Talara Tar Seeps, northwestern Peru. mainland South America. Nat. Hist. Mus. Los
Roy. Ontario Mus. Life Sci. Contrib. (In press) Angeles Cty. Contrib. Sci. (192): 1-78.
Carriker, M. A., Jr. 1922. Description of the Santa Dixon, J. R., Wright, J. W. 1975. A review of the
Marta Region, pp. 7-29 in Ruthven, A. G. The lizards of the iguanid genus Tropidurus in Peru.
amphibians and reptiles of the Sierra Nevada de Ibid. (271): 1-39.
Santa Marta, Colombia. Misc. Publ. Mus. Zool. Dollfus, O. 1960. Presentation de la structure des
Univ. Michigan (8): 1-69. Andes centrales peruviennes. Trav. Inst. Francais
Cei, J. M. 1962. Batracios de Chile. Ed. Univ. Etud. Andines 7:53-64.
Chile, Santiago, 128 cviii p.+ Donoso-Barros, R. 1966. Reptiles de Chile. Ed.
Cei, J. M. 1968. Remarks on the geographical dis- Univ. Chile, Santiago, 458 p.
tribution and phyletic trends of South American Donoso-Barros, R. 1968. The lizards of Venezuela
toads. Pearce-Sellards Ser. Texas Mem. Mus. (checklist and key). Caribb. J. Sci. 8:10,5-122.
(13): 1-21. Donoso-Barros, R. 1970. Catalogo herpetologico
Cei, J. M. 1971. Consideraciones sobre las relaciones Chileno. Mus. Nac. Hist. Nat. Bol. ( Santiago de
taxonomicas de Phymaturtis patagonicus y Phy- Chile) 31:49-124.
maturus palluma. Acta Zool. Lilloana 28:37-46. Donoso-Barros, R. 1972. Contribucion al conoci-
Cei, J. M. 1972. Bufo of South America, pp. 82-92 miento del genero Ctenoblepharis Tschudi y
in Blair, W. F. ( ed. ) Evolution in the genus
. Phrynosaura Wemer (Sauria: Iguanidae). Bol.
Bufo. Univ. Texas Press, Austin, 459 p. Soc. Biol. Concepcion 44:129-134.
Cei, J. M. 1973. Comentarios sobre algunos generos Dott, R. H., Jr., Winn, R. D., Jr., Dewit, M. L.,
de iguanidos: Diplolaemus, Leiosaurus, Apero- Bruhn, R. L. 1977. Tectonic and sedimentary
pristis y Cupriguanus. Physis 32:269-276. significance of Cretaceous Tekerika Beds of Tierra
Cei, J. M. 1974a. Two new species of Ctenoble- del Fuego. Nature 266:620-622.
pharis Reptilia, Iguanidae ) from the arid en-
( Duellman, W. E. 1970. The hylid frogs of Middle
vironments of the Central Argentina ( Mendoza America. Univ. Kansas Mus. Nat. Hist. Monogr.
Province). J. Herpetol. 8:71-75. ( 1 ) : 1-753.
Cei, J. M. 1974b. Revision of the Patagonian igua- Duellman, W. 1972. A review of the Neotropi-
E.
nids of the Liolaemus clongatus complex. Ibid. cal the Hyla bogotensis group. Univ.
frogs of
8:219-229. Kansas Mus. Nat. Hist. Occas. Pap. (11): 1-31.
Cei, J. M. 1977. A new species of Tclmatobius (Am- Duellman, W. E. 1973. Descriptions of new hylid
phibia, Anura, Leptodactylidae) from the north- frogs from Colombia and Ecuador. Herpetol-
ern mountains of San Juan, Argentina. Ibid. 11: ogica 29:219-227.
359-361. Duellman, W. E. 1974. A systematic review of the
Cei, J. M. 1979. Amphibians of Argentina. Monit. marsupial (Hylidae: Gastrotheca) of the
frogs
Zool. Italiano
Monog. Zool. (In press) Andes of Ecuador. Univ. Kansas Mus. Nat. Hist.
Cei, M., Castro, L. P. 1973. Taxonomic and sero-
J. Occas. Pap. (22): 1-27.
logical researches on the Phymaturus patagonicus Duellman, W. E. 1976. Centrolenid frogs from
complex. J. Herpetol. 7:237-247. Peru. Ibid. (52): 1-11.
Cei, J. M., Erspameh, V., Roseghini, M. 1972. Bio- Duellman, W. E. 1977. Liste der rezenten Ani-
genic amines, pp. 233-243 in Blair, W. F. ( ed. ) .
phibien und Reptilien: Hvlidae, Centrolenidae,
Evolution in the genus Bufo. Univ. Texas Press, Pseudidae. Das Tierreich (95): 1-225.
Austin, 459 p. Duellman, W. E. 1978a. Two new species of Eleu-
Chapman, F. M. 1917. The distribution of bird life therodactylus (Anura: Leptodactylidae) from
in Colombia. Bull. Amer. Mus. Nat. Hist: 36: the Permian Andes. Trans. Kansas Acad. Sci.
1-730. Sl:65-71.
1979 DUELLMAN: HERPETOFAUNA OF ANDES
Duellman, W. E. 1978b. New species of leptodac- Gaxsseh, A. 1955. Ein Beitrage zur Geologie und
tylid frogs of the genus Eleutherodactylus from Petrographie der Sierra Nevada de Santa Marta
the Cosnipata Valley, Peru. Proc. Biol. Soc. ( Kolumbien, Siidamerika). Schweiz. Mineral.
Washington 91:418-130. Petrogr. Mitt. 35:209-279.
Duellmax, W. E. 1978c. Three new species of Gansser, A. 1973. Facts and theories on the Andes.
Eleutherodactylus from Amazonian Peru (Am- J.Geol. Soc. London 129:93-131.
phibia: Anura: Leptodactvlidae ) .
Herpetologica Gorham, S. W. 1966. Liste der rezenten Amphibien
34:264-270. und Reptilien. Ascaphidae, Leiopelmatidae, Pipi-
Duellman, W. E.,Altic, R. 1978. New species of dae, Discoglossidae, Pelobatidae, Leptodactylidae,
tree frogs ( family Hylidae ) from Colombia and Rhinophrynidae. Das Tierreich. (85): 1-222.
Ecuador. Ibid. 34:177-185. Gorman, G. C, Huey, R. B., Williams, E. E. 1969.
Duellmax, W. A taxonomic
E., Fritts, T. H. 1972. Cytotaxonomic studies on some unusual iguanid
review of the southern Andean marsupial frogs lizards assigned to the genera Chamaeleolis, Poly-
(Hylidae: Gastrotheca). Univ. Kansas Mus. Nat. chrus, Polychroidcs, and Phenacosaurus, with be-
Hist. Occas. Pap. (9): 1-37. havioral notes. Breviora (316): 1-17.
Duellmax, W. E., Simmons, J. E. 1977. A new Haffer, J. 1970. Enstehung und Ausbreitung nord-
species of Eleutherodactylus (Anura: Leptodac- Andiner Bergvogel. Zool. Jahrb. Abt. Syst. Oekol.
tvlidae) from the Cordillera Oriental of Colombia. Geogr. Tiere 97:301-337.
Proc. Biol. Soc. Washington 90:60-65. Haffer, J. 1974. Avian speciation in tropical Soudi
Duellmax, W. E., Toft, C. A. 1979. Anurans from America. Publ. Nuttall Omithol. Club (14):1-
Serrania Amazonian Peru: Taxonomy
de Sira, 390.
and biogeography. Herpetologica 35:60-70. Ham, C. K., Herrera, L. J., Jr. 1963. Role of sub-
Duellmax, W. E., Veloso M., A. 1977. Phylogeny andean fault system in tectonics of eastern Peru
of Pleurodcma (Anura: Leptodactylidae A )
:
and Ecuador, pp. 47-61 in Childs, O. E., Beebe,
B. W. (eds.). Backbone of the Americas. Mem.
biogeographic model. Univ. Kansas Mus. Nat.
Hist. Occas. Pap. (64): 1-46. Amer. Assoc. Pet. Geol. 2:1-320.
Edwards, S. R. 1974. A phenetic analysis of the
Harrington, H. J. 1956. Main morphostructural
genus Colostethus (Anura: Dendrobatidae ) .
regions of South America, pp. xii-xviii in Jenks,
Ph.D. Dissert. Univ. Kansas. 419 p. W. F. ( ed. ) Handbook of South American geol-
.
Kinzl, H. 1968. La glaciacion actual y pleistocenica Lynch, J. D. 1979. Leptodactylid frogs of the genus
en los Andes centrales, pp. 77-90 in Troll, C. Eleutherodactylus from the Andes of southern
(ed. ). mountainous regions
Geo-ecology of the Ecuador. Univ. Kansas Mus. Nat. Hist. Misc.
of the tropical Americas. Colloq. Geog. 9:1-223. Publ. (66): 1-62.
Laurent, R. F. 1967. Descubrimiento del genero Lynch, J. D., Duellman, W. E. 1973. A review of
Gastrotheca Fitzinger en Argentina. Acta Zool. the centrolenid frogs of Ecuador, with descrip-
Lilloana 22:353-354. tions of new species. Univ. Kansas Mus. Nat.
Laurent, R. F. 1969a. Estudio complementario de Hist. Occas. Pap. 10:1-66.
Gastrotheca christiani Laurent. Ibid. 25:123-136. Lynch, J. D., Duellman, W. E. 1979. The Eleu-
Laurent, R. F. 1969b. Una segunda especie del therodactylus of the Amazonian slopes of the
genero Gastrotheca Fitzinger en Argentina. Ibid. Ecuadorian Andes (Anura: Leptodactylidae).
25:143-150. Univ. Kansas Mus. Nat. Hist. Misc. Publ. ( In
Laurent, R. F. 1970. Dos especies nuevas argen- press )
tinas del genero Telmatobius (Amphibia, Lepto- MacArthur, R. A., Wilson, E. O. 1967. The theory
dactylidae). Ibid. 25:207-226. of island biogeography. Monog. Pop. Biol. 1,
Laurent, R. F. 1973. Nuevos datos sobre el genero Princeton Univ. Press, Princeton, New Jersey,
Telmatobius en el noroeste Argentino, con de- 204 p.
scripcion de una nueva especie de la Sierra del Macedo, H. de. 1960. Vergleichende Untersuchun-
Manchao. Ibid. 30:163-187. gen an Arten der Gattung Telmatobius Am- (
Laurent, R. F. 1976. Nuevas notas sobre el genero phibia: Anura). Z. Wis. Zool. Abt. A 163:355-
Gastrotheca Fitzinger. Ibid. 32:31-64. 396.
Laurent, R. F. 1977. Contribucion al conocimiento Mauhiello, D., Roskoski, J. P. 1974. A re-analysis
del genero Telmatobius Wiegmann (4a Nota). of Vuilleumier's data. Amer. Nat. 108:711-714.
Ibid. 32:189-206. McDiahmid, R. W. 1971. Comparative morphology
Lazell, J. D. 1969. The genus Phcnacosaurus ( Sau- and evolution of frogs of the Neotropical genera
ria: Iguanidae). Breviora 325:1-24. Atelopus, Dcndrophryniscus, Mclanophryniscus,
Liddle, R. A. 1946. The geology of Venezuela and and Oreofihrynella. Nat. Hist. Mus. Los Angeles
Trinidad, 2nd ed. New York, 890 p. Cty. Sci. Bull. (12): 1-66.
Lohmann, H. H. 1970. Outline of tectonic history Mercer, J. II., Palacios, M. O. 1977. Radiocarbon
of Bolivian Andes. Bull. Amer. Assoc. Pet. Geol. dating of the last glaciation in Peru. Geology 5:
54:735-757. 600-604.
Low, B. S. Evidence from parotoid-gland
1972. Montanucci, R. R. 1973. Systematica and evolution
secretions, pp. 244-264 in Blair, W. F. (ed. ). of the Andean lizard genus Pholidobolus (Sauria:
Evolution in the genus Bufo. Univ. Texas Press, Teiidae). Univ. Kansas Mus. Nat. Hist. Misc.
Austin, 459 p. Publ. (59): 1-52.
Lynch, J. D. 1971. Evolutionary relationships, oste- Morrison, T. 1974. Land above the clouds. Andre
ology,and zoogeography of leptodactyloid frogs. Deutsch, Ltd., London, 223 p.
Univ. Kansas Mus. Nat. Hist. Misc. Publ. (53): Morrison, T. 1975. The Andes. Time-Life Inter-
1-238. national, Amsterdam, 184 p.
Lynch, J. D. 1975a. A review of the Andean Muller, P. 1973. The dispersal centres of terrestrial
leptodactylid frog genus Phrynopus. Univ. Kansas vertebrates in the Neotropical Realm. Biogeo-
Mus. Nat. Hist. Occas. Pap. (35): 1-51. graphica 3, Junk, The Hague, 244 p.
Lynch, J. D. 1975b. A review of the broad-headed Myers, C. W. 1973. A new genus for Andean snakes
eleutherodactvline frogs of South America. Ibid. related to Lygophis boursieri and a new species
(38):l-46. (Colubridae). Amer. Mus. Novit. (2522): 1-37.
Lynch, J. D. 1975c. The identity of the frog Eleu- Myers, C. W. 1974. The systematics of Rhadinaca
therodactylus eonspicillatus (Giinther), with de- (Colubridae), a genus of New World snakes.
scriptions of two related species from northwest- Bull. Amer. Mus. Nat. Hist. 153:1-262.
ern South America (Amphibia: Leptodactylidae). Myers, C. W., Daly, J. W. 1976a. Preliminary eval-
Contrib. Sci. Los Angeles Cty. Mus. (272): 1-19. uation of skin toxins and vocalizations in taxo-
Lynch, J. D. 1976a. New species of frogs (Lepto- nomic and evolutionary studies of poison-dart
dactylidae: Eleutherodactylus) from the Pacific frogs (Dendrobatidae). Ibid. 157:173-262.
versant of Ecuador. Univ. Kansas Mus. Nat. Hist. Myers, C. W., Daly, J. W. 1976b. A new species of
Occas. Pap. (55): 1-33. poison frog Dendrobates
(
from Andean Ecua-
)
Lynch, J. D. 1976b. The species groups of the dor, including an analysis of its skin toxins. Univ.
South American frogs of the genus Eleutherodac- Kansas Mus. Nat. Hist. Occas. Pap. (59): 1-12.
tylus (Leptodactylidae). Ibid. (61): 1-24. Myers, C. W., Daly, J. W., Malkin, B. 1978. A
Lynch, J. D. 1978a. A new eleutherodactvline frog dangerously toxic new frog (Phyllobates) used by
from the Andes of northern Colombia ( Lepto- Embera Indians of western Colombia, with dis-
dactylidae). Copeia 1978(1):17-21. cussion of blowgun fabrication and dart poisoning.
Lynch, J. I). A re-assessment of the telma-
19781). Bull. Amer. Mus. Nat. Hist. 161:307-366.
tobiineleptodactylid frogs of Patagonia. Univ. Noule, G. K. 1921. Some new lizards from north-
Kansas Mus. Nat. Hist. Occas. Pap. (72): 1-57. western Peru. Ann. N.Y. Acad. Sci. 29:133-139.
1979 DUELLMAN: HERPETOFAUNA OF ANDES 441
Oftedal, O. T. 1974. A revision of the genus Ana- (Amphibia, Salientia). Soc. Cien. Nat. Salle
dia (Sauna, Teiidae). Arq. Zool. (Sao Paulo) Mem. 28:301-334.
25:203-265. Rivero, 1970. On the origin, endemism, and
J. A.
Okada, A. 1971. On
the neoteetonics of the Ata- distribution of the genus Stefania Rivero (Am-
cama fault zone region —
Preliminary notes on late phibia, Salientia) with a description of a new
Cenozoic faulting and geoniorphic development species from southeastern Venezuela. Soc. Vene-
of the coast range of northern Chile. Bull. Dep. zolana Cienc. Nat. Bol. 28:456-481.
Geog. Univ. Tokyo 3:47-65. Rivero, J. A. 1972. La ranita "mas elevada" de
Oppenheim, V. 1937. Contribution to the geology Venezuela. Natura (Venezuela), Sept. 1972:48-
of the Venezuelan Andes. Bol. Inf. Asoc. Vene- 49.
zolana Geol. Min. Pet. 1:25-45. Rivero, 1974. On Atelopus oxyrhynchus Bou-
J. A.
Parker, H. W. 1932. Some new or rare reptiles and lenger (Amphibia, Salientia), with the descrip-
amphibians from southern Ecuador. Ann. Mag. tion of a new race and a related new species from
Nat. Hist. (10)9:21-26. the Venezuelan paramos. Soc. Venezolana Cienc.
Parker, H. W. 1934. Reptiles and amphibians from Nat. Bol. 29:600-612.
southern Ecuador. Ibid. (10)14:264-273. Rivero, J. A. "1976" [1978]. Notas sobre" los anfibios
Parker, H. W. 1938. The vertical distribution of de Venezuela II. Sobre los Colostethus de los
some reptiles and amphibians in southern Ecua- Andes Venezolanos. Soc. Cien. Nat. Salle Mem.
dor. Ibid.(11)2:438-450. 35:327-344.
Parker, H. W. 1940. The Percy Sladen Trust Ex- Rivero, J. A., Mayorga, H. 1973. Un nuevo Eleu-
pedition to Lake Titicaca in 1937. Amphibia. therodactylus (Amphibia, Salientia) del Paramo
Trans. Linnean Soc. London 1(2) :203-216. de Guaramacal, Estado Trujillo, Venezuela. Ca-
Patterson, T. C, Lanninc, E. P. 1967. Los niedios ribb. J. Sci. 13:75-79.
ambientes glacial tardio y postglacial de Sud- Roig, F. A. 1972. Bosquejo fisionomico de la vege-
america. Bol. Soc. Geog. Lima 86:1-19. tacion de la Provincia de Mendoza. Bol. Soc.
Paull, D., Williams, E. E., Hall, W. P. 1976. Argentina Bot. 13 (supplement): 49-80.
Lizard karyotypes from the Galapagos Islands: Rosen, D. E. 1975. A vicariance model of Caribbean
Chromosomes in phylogeny and evolution. Bre- biogeography. Syst. Zool. 24:431—464.
viora (441): 1-31. Rossman, D. A., Thomas, R. 1979. A new dipsadine
Pefaur, J. E., Duellman, W. E. 1977. Community snake of the genus Sibynomorphus from Peru.
structure in high Andean herpetofaunas. Herpe- Louisiana St. Univ. Mus. Zool. Occas. Pap. (54):
tol. Rev. 8(3):Suppl. 6-7. (Abstract) 1-6.
Peters, J. A. 1973. The frog genus Atelopus in Roze, J. La taxonomia y zoogeografia de los
1966.
Ecuador. Smithsonian Contrib. Zool. ( 145 ) 1—19. : ofidios de Venezuela. Univ. Central Venezuela,
Peters, J. A., Donoso-Barros, R. 1970. Catalogue Caracas, 362 p.
of the Neotropical Squamata: Part II. Lizards Ruiz-C, P. M., Hernandez-C, J. I. 1976. Osorno-
and Amphisbaenians. U.S. Natl. Mus. Bull. phryne, genero nuevo de anfibios bufonidos de
(297): 1-293. Colombia y Ecuador. Caldasia 11(54) :93-148.
Peters, J. A., Orejas-Meranda, B. 1970. Catalogue Ruiz-C, P. M., Hernandez-C, J. I. 1978. Una
of the Neotropical Squamata. Part I. Snakes. nueva especie Colombiana de Atelopus ( Am-
Ibid. (297): 1-347. phibia: Bufonidae). Ibid. 57:181-197.
Petersen, U. 1958. Structure and uplift of the Ruthven, A. G. 1922. The amphibians and reptiles
Andes of Peru, Bolivia, Chile, and adjacent Ar- of the Sierra Nevada de Santa Marta, Colombia.
gentina. Bol. Soc. Geol. Peru 33:57-143. Misc. Publ. Mus. Zool. Univ. Michigan (8):l-69.
Presch, W. 1978. Evolutionary relationships of mi- Rutland, R. W. R., Guest, J. E., Grastv, R. L.
croteiid lizards. Joint Annual Meeting SSAR, 1965. Isotopic ages and Andean uplift. Nature
ASIH, HL, Arizona State Univ., Tempe, Arizona, 208:677-678.
31 May-7 June 1978. (Abstract of presented Sauer, W. 1965. Geologia del Ecuador. Min. Edu-
paper) cacion, Quito, 384 p.
Platnick, N. I., Nelson, G. 1978. A method of Sauer, W. 1971. Geologie von Ecuador. Gebruder
analysis for historical biogeography. Svst. Zool. Bortraeger, Berlin, 316 p.
27:1-16. Savage, J. M. 1958. The iguanid lizard genera Uro-
Rivero, J. A. 1961. Salientia of Venezuela. Bull. saurus and Uta with remarks on related genera.
Mus. Comp. Zool. Harvard Univ. (126): 1-207. Zoologica 43(2):41-54.
Rivero, J. A. 1963a. The distribution of Venezuelan Savage, J. M. 1960. A revision of the Ecuadorian
frogs. II. The Venezuelan Andes. Caribb. J. Sci. snakes of the colubrid genus Atractus. Misc. Publ.
3:87-102. Mus. Zool. Univ. Michigan (112): 1-86.
Rivero, J. A. 1963b. Five new species of Atelopus Schmidt, K. P. 1954. Notes on frogs of the genus
from Colombia, with notes on other forms from Telmatobius with descriptions of two new Peru-
Colombia and Ecuador. Ibid. 3:103-124. vian species. Fieldiana Zool. 34:277-287.
Rivero, J. A. 1964. The distribution of Venezuelan Schmidt, K. P., Walker, W. F., Jr. 1943. Three
frogs. IV. The coastal range. Ibid. 4:307-319. new snakes from the Peruvian Andes. Field Mus.
Rivero, J. A. 1968. Los centrolenidos de Venezuela Nat. Hist. Zool. Ser. 24:325-329.
442 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Schubert, C. 1974. Late Pleistocene Merida glacia- Uzzell, T. 1973. A revision of lizards of the genus
tion, Venezuelan Andes. Boreas 3:147-152. Prionodacttjlus, with a new genus for P. leuco-
Shacam, R. 1975. The northern termination of the stictus and notes on the genus Euspondylus ( Sau-
Andes, pp. 325—420 in Nairn, A. E. M., Stehli, ria, Teiidae). Ibid. (159): 1-67.
F. G. (eds.). The ocean basins and margins. Vol. van der Hammen, T. 1972. Historia de la vegeta-
3. The Gulf of Mexico and the Caribbean. Ple- cion y el medio ambiente del norte sudamericano.
num Press, New York, 706 p. Mem. Symp. I Congr. Lat. Amer. Bot., Mexico,
Silverstone, P. A. 1975a. Two new species of D.F., pp. 119-134.
Colostethus (Amphibia: Anura: Dendrobatidae ) van der Hammen, T. 1974. The Pleistocene changes
from Colombia. Mus. Nat. Hist. Los Angeles Cty. of vegetation and climate in tropical South Amer-
Contrib. Sci. (268): 1-10. ica. J. Biogeog. 1:3-26.
Silverstone, P. A. 1975b. A revision of the poison- van der Hammen, T., Werner, J. H., Dommelen,
arrow frogs of the genus Dendrobates Wagler. H. van. 1973. Palynological record of the up-
Mus. Nat. Hist. Los Angeles Cty. Sci. Bull. (21): heaval of the northern Andes: A study of the
1-55. Pliocene and Lower Quaternary of the Colom-
Silverstone, P. A. 1976. A revision of the poison- bian eastern Cordillera and the early evolution of
arrow frogs of the genus Pliyllobates Bibron in its high-Andean biota. Rev. Paleobot. Palynol.
Sagra (Family Dendrobatidae). Ibid. (27): 1-53. 16:1-122.
Simpson, B. B. 1975a. Glacial climates in the eastern Vellard, J. 1946. El genero Telmatobius en la
tropical South Pacific. Nature 253(5486) 34-36. :
Republica Argentina. Acta Zool. Lilloana 3:
Simpson, B. B. 1975b. Pleistocene changes in the 313-326.
flora of the high tropical Andes. Paleobiology Vellard, 1951. Estudios sobre batracios andinos.
1:273-294. I. — El J.
Whymper, E. 1892. Travels amongst the great An- Williams, E. E., Reic, O. A., Kiblisky, P., Rivero-
des of the Equator. 2nd ed. John Murray, Lon- Blanco, C. 1970. Anolis jacare Boulenger, a
"solitary" anole from the Andes of Venezuela.
don, 456 p.
Ibid. (353): 1-15.
Williams, E. E. 1974. South American Anolis:
Yricoyen, M. R. 1972. Cordillera principal, pp. 345-
Three new species related to Anolis nigrolineatus 364 in Leanza, A. F. (ed.). Geologia regional
and A. dissimilis. Breviora (422): 1-15. Argentina. Acad. Nac. Cien. Cordoba, 869 p.
APPENDICES
Appendices 1-3 include lists of the 727 species (462 amphibians and 265 reptiles) that
formed the data base for the analysis of geographical distributions. For each species, the
known altitudinal limits of distribution are given; the general habitat of each is coded as noted
in the captions to the tables. The species are arranged sequentially in 1) ordinal categories
(caecilians, salamanders, frogs, lizards, snakes), 2) phylogenetically by families, and 3) alpha-
—
Appendix 15:2. Altitudinal and geographic distribution of amphibians and reptiles in the Andes of Colombia
and Ecuador, and the eastern Cordilleras of Peru, Bolivia, and northern Argentina; all highland species in the
Huancabamba Depression are included. Habitats: C = cloud forest, D = dry forest; P = paramo or puna,
R = rainforest, S = subparamo, X = unknown. Species noted by ' are those diat are restricted to outlying
ranges east of the Andes
— Sierra de Macarena, Cordillera de Cutucii, Cordillera del Condor, Serrania de Sira.
Species noted by
" occur in the Cordillera Occidental in northern Peni.
a
6 c 2
5 -3
"3
3 E
>
be u
75
O w £ « <
Elevation TJ 5 3 £«
w "c o
Species
^ 7. ?. *„ flj CJ f
(D
446 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
5 -5 ^ -5
S
5
T3
s £
5d u s u w ft,
§
co
3
d _-
cj Is 2 "3
Species Elevation t3 3 3 Js
q~ CJ OJ Qj
1979 DUELLMAN: HERPETOFAUNA OF ANDES 447
? a 3 | |
I g e||| «
| a |§ 5 | 2 g 5 ^. |
u
3
<u "(a
h
Q -S a 2 -2 .2 "S
Species Elevation -c 3
J J S | «
"3
£
6
2
loo
^
o
2
ooo 3 u
2
If
2 2 g 2
.§
S
I
Z — a a
J52 k_o;
a s a g u
~o _*; _±! ,33
=3
*
soo
13 3 =00 1?Tooc-p
OUU ZOO U U U ~ "S 15 12 i? Til £
~
§
_
.
o
T3
22 .2 O
.S & 3 .g
13 d
w 3 g he
-§ -2 3 o
« w
« « U o w Oh
Species Elevation 3
'o c m
53
U o O o
>
z _QJ J1J
CL»
JL> Q
^ ^ ^ 3
T3 -d
O o o 3 o o o 3
O U c3 U O O U ^
Eleutherodactylus viridicans __ 2170-2680
Eleutherodactylus walked 220-1270 RC
Eleutherodactylus w-nigrum 1230-2800 CS CS CS S CS CS
Eleutherodactylus sp. "A" . __ 1850-2160 ~
Eleutherodactylus sp. "B" ._ 2100-2750 C
Eleutherodactylus sp. "C" 3200 s
Eleutherodactylus sp. "D" 3750 p
Eleutherodactylus sp. "E" . 2900-3000 s
Eleutherodactylus sp. "F" 3400 I'
2
o 13
-o
3 £o J=
32
U o
S 3 W
—f _r o u w
« O _o
a "3 oj rt ^
Species Elevation is -g c
o c
o
O 3
as
^-
z ,i 0)
jj; ja O) 0* 0; 1* at n
a o
-o
S o o 3 o o o 3
o o O Z O u u o O
Colostethus sp. "2"" 1100-2400
Colostethus anthracinus 2500-3500 SP
Colostethus sp. "3" _ 1410-1740 c
Colostethus sp. "4" 3700
Colostethus bolivianus 800-1200 RC
Colostethus sp. "5" 900-1190 RC
Colostethus elachyhistus .. 800-2800 S D
Colostethus sp. "6" 1800-2900 CS
Colostethus fraterdanieli 1900 C
Colostethus fuliginosus 480-1740 RC
Colostethus sp. "7" 1900-2070
Colostethus infraguttatus ... 340-1270 RC
Colostethus kingsburyi 1150-1410
Colostethus lehmanni 1900
Colostethus mertensi 2170-2680 C
Colostethus sp. "8" 1900-2350 C
Colostethus sp. "9" 2300-3500 SP
Colostethus palmatus 900-2500 RC
Colostethus sp. "10" .... 1500-1700 c
Colostethus sp. "11"" 1830 c
Colostethus ramosi 1240
Colostethus sp. "12" 1150-2790 S c
Colostethus sp. "13" 2100-3500 P s
Colostethus subpunctatus .. .. 2100-3300 SP
Colostethus sylvatica .. . 1580-2970 D
Colostethus taeniatus _ 1740-2970 cs cs
Colostethus sp. "14" ,>
1830 c
Colostethus vertebralis .. 2500-3200 P .... p
Colostethus whymperi .. 1460-2120 C
00
Colostethus sp. "IS" 1800
Colostethus sp. "A" .. 670-2130 DC
Dendrobates abditus 1650-1700
Dendrohates lehmanni 850-1200 C ....
22 o
o
E
W 3 g
o o w
§ 2 1 O £
o O W
"3
O i: .B x; « a «
Species Elevation
.a
O O O s
*- *" t» hM M _
Z 2, ID <0 V V JO jp
^3 O
O 3 o o o o 3
e U z O U O
Hy/a sp. "B" 1750
Hyla sp. "C" 2300
Osteocephalus pearsoni 300-1620
Osteocephalus vcrrucigerus ... 500-1840
Phyllomedusa bakea° ___. 1280
Phyllomedusa boliviano 500-2000
Phyllomcdusa perinesos 1410-1490
Centrolene geckoideum 1920-2150 c c
Centrolenella anomala 1740
Centrolenella antioquiensis 1890-2475
Centrolcnella audax 1490-1700
Centrolenella buckleyi .. ..... 2100-3400 SP SP SP
Centrolenella cochranae . 1100-1410
Centrolenella flavopunctata 720-1800
Centrolenella fleisehmanni 200-1410 RC RC
Centrolenella grandisonae 1460-2000 C C
Centrolenella griffithsi .. 1200-2170 C C
Centrolenella johnelsi 2500
Centrolenella mariae' 1550
Centrolenella medemi .. 1100-1490
Centrolenella megacheira 1490-1740
Centrolenella ocellata .. 1200-1820
Centrolenella ocellifera 700-2300 '. RC
Centrolenella pellucida 1740
Centrolenella perkticta .. 1410-1460
Centrolenella pipilata ..... 1700-1740
Centrolenella prosoblepon 200-1410
Centrolenella siren .. 1200-1950
Centrolenella spiculata 1000-1700
Centrolenella truebae 1700
Centrolenella valeroi .. 200-1100
Centrolenella sp. "A" 1410-1490
Centrolenella sp. "B" .. 2660
Centrolenella sp. "C" .. 1140-1540
Centrolenella sp. "D" 1140
Centrolenella sp. "E" 1540
Centrolenella sp. "F" .. 1960-2150
Centrolenella sp. "G" 1960-2010
Centrolenella sp. "H" .. 1950
Centrolenella sp. "1" 1230
Centrolenella sp. "J" 1230
Centrolenella sp. "K" . 1230
Centrolenella sp. "L" 1750
Centrolenella sp. "M" 1750
Centrolenella sp. "N" 1750
Centrolenella sp. "O" 1050
Centrolenellasp. "P" 1200-1850
Glossostoma aequatoriale 2500-3615
1979 DUELLMAN: HERPETOFAUNA OF ANDES 453
Species
454 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Species ion
1979 DUELLMAN: HERPETOFAUNA OF ANDES 455
2 w u -2 3-3
a 2 3
0)
Species Elevation
456 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
o
ti
a
§ s
Species Elevation 13
cd
cd
£ JJ
1979 DUELLMAN: HERPETOFAUNA OF ANDES 457
Appendix 15:3. — Altitudinal and geographic distribution of amphibians and reptiles in the Andes of Peru,
Bolivia, Argentina, and Chile. Habitats: A = arid, C = cloud forest, N = Nothofagus forest, P = puna,
S = Patagonian scrub, V = Valley vegetation.
•S -2
Species Elevation
458 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
-S
-g
en
5 £
53 o
o 2 o
Z o
'rt *rt C
1) ^z —
Species Elevation 3 2 < U
b "2
d
O 0000 '0
-a
>
•c
3
a
<
-
0)
*3 'B
o o o o 2
u U u O X
1979 DUELLMAN: HERPETO FAUNA OF ANDES 459
"fi -s
•E
PL,
-s
4o -g"
3
Z
Species Elevation -j £ «
V o O
=a
-3 ^ ^
o o
u O U
16. Refugia, Refuges and Minimum Critical Size: Problems
in the Conservation of the Neotropical Herpetofauna
Thomas E. Lovejoy
World Wildlife Fund
1601 Connecticut Avenue, N.W.
Washington, D.C. 20009
USA
461
462 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
of the primary forest has been cut) could Vanzolini's (1973) reptilian refugia num-
have an important drying effect on climate, 1 bered about Such differences raise inter-
six.
an effect that climatological science seems esting questions about the size of the grain
unprepared or unable to predict. If Brasil's of the environment for different kinds of
plan to protect five percent of the Amazon organisms, and says to the extent that non-
in conservation units is thus imperiled, it well overlap merely reflects the abundance of
could mean that man's environmental effect refugia of a particular taxon, the job is merely
would be yet greater than the magnitude of to create more refuges. But suppose a taxon
the Pleistocene climatic changes. with few but large refugia, such as reptiles,
The paleoecologists to locate
failure of has a refugium which does not overlap with a
palynological evidence of the whereabouts of refugium from a second taxon with numerous,
the deciduous forests of the northern United smaller refugia. Given continued belief in
States during Pleistocene glaciations suggests the refugia hypothesis, the lack of coincidence
that this forest may not be the highly co- of refugia most likely indicates our data base
evolved unit it was for so long thought to be. is
inadequate, or may suggest interesting dif-
As a consequence we must exercise care not ferences in speciation in the different groups.
to leave government and nonscientific con- The Brasilian government has taken the
servationists with the impression that because refugia approach in its recent document "Uma
we talk of our centers of endemism as refugia Analise de Prioridades em Conservacao da
they are indeed likely to be refugia in times of Natureza na Amazonia" (Wetterberg et al.,
future climatic change. It should be force- 1977). choosing priority areas from overlays
fully pointed out that refugia, at least those of refugia for heliconian butterflies, four plant
not associated with features of geological re- families, birds, and lizards; highest priority
lief, may well have been in different places in has been assigned to areas with the greatest
different periods, and that dry periods varied coincidence of refugia. The document has
in duration and perhaps even in temperature. been included in the national development
The deciduous forest of the northeast or the plan, and further was adopted by the six
Amazonian hyalea did not necessarily move (non-Guianan) Amazonian nations at their
around as Yet with respect to the
a unit. second technical meeting on conservation in
Amazon one wonders how so many
forests Brasilia in July 1977. The Brasilian govern-
one to one relationships and intricate systems ment is to be applauded for taking biogeog-
haps not so apparent. However, loss of spe- size of reptilecommunities has been carried
cies from too small an area obviously is inevi- out by Rruce Wilcox, who has kindly made
table if only because low density species will his data available. His studies are primarily
occur in numbers too low to reproduce or concerned with the lizard faunas of islands
to withstand stochastic fluctuations. Undoubt- in the Gulf of California (Wilcox, 1978).
edly there are other processes involved in These land-bridge islands can be dated as to
the decay process of an ecosystem with a age of isolation on the basis of the depth of
species number too large for the area left the water separating them from the mainland
intact to date they have not been
by man, but and a knowledge of rates of sea level rises at
explored. An
important question to test is the end of the Pleistocene. Consequently,
one advanced by Lovejoy and Oren (1979), these islands provide the first time sequence
namely, whether this decay process leads to a evidence of the decay process of super-
predictable species composition of the im- saturated faunas of any sort.
poverished community. Wilcox also was able to estimate immigra-
There has been sharp controversy as to tion and extinction rates for these lizard
whether more species can be preserved by a island communities and found the former so
series of small reserves rather than a large low as to be negligible in his calculations.
one Diamond, 1975; Si'mberloff and Ameiva ameiva and Mabuya mabuija not-
(e.g..
Abele, 1976; Terborgh, 1976). Arguments withstanding, this would seem to make sense
that concentrate on simple numbers of species because lizards probably do not have the
can be faulted for having tended to treat all dispersal facility of birds. That the extinction
more importantly rate is lower for lizards than birds may be
species as equal, but for
reflective of a slower pulse to the
ignoring that what we really want to preserve dynamics
are functioning ecosystems, not some kind of of the poikilothermic system, or that poor dis-
glorified zoos.
2
Ecosystem protection will re- perses may be good persisters (Oren, pers.
areas or at least ones greater than comm. )
.
quire large
the minimum critical size. If immigration rates are low and oppor-
tunities for dispersal to a conservation unit
Terborgh (1974) has estimated that for
an area of about few, large areas are in order to maintain the
tropical rain forest birds,
needed in order to original lizard biota. However, Wilcox (in
1,000 square kilometers is
preparation) suggests that the lower energy
keep extinctions down to one percent of the
demands of poikilotherms should make higher
original species complement per century.
biomass and densities possible, which, all
This size is probably in the right order of
other things being equal, would indicate that
magnitude, but obviously minimum critical
the area need not be as large as for a com-
size will vary considerably with the taxon
and ecosystem concerned. parable avian community.
I am not aware of data on minimum criti-
In designing reserves with ecosystem pres-
cal size for amphibian ecosystems, although I
ervation in mind, the minimum critical size
would guess it to be yet smaller and frequent-
to be taken is that which is the largest of
ly involving special features of the environ-
those of all the taxonomic components. Yet
ment. In any case, this leaves plenty for
there may be occasions when a reserve should
herpetologists to do, as long as our species
be designed primarily for its herpetofaunal
doesn't reduce the possible future directions
community. Further, there are probably in- in which biological knowledge can grow by
teresting things to be learned by consideration
doing away with Neotropical ecosystems in
of factors relating to minimum critical size in
the name of supposed progress.
different kinds of organisms.
Almost all the work on minimum critical
ACKNOWLEDGMENTS
!
This is not meant to denigrate in any fashion the
T. Prance, David C. Oren, Rruce A. Wilcox, Remington, C. L. 1968. Suture-zones of hybrid in-
teraction between recently joined biotas, pp. 321-
and Richard O. Bierregaard.
428 in Dobzhansky, T., Hecht, M. K., Steere,
W. C. (eds.). Evolutionary Biology. Vol. 2.
452
RESUMEN Salati,
p.
E., Marques, J., Molion, L.-C. B. 1978.
Origem e distribucao das chuvas na Amazonia.
Gran parte de la herpetofauna asi como Interciencia 3:200-205.
otras biotas sudamericanas solo sobreviviran Simberloff, D. S., Abele, L. G. 1976. Biogeog-
raphy theory and conservation practice. Science
en parques nacionales y reservas biologicas. 191:285-286.
Las bases biogeograficas para el disefio de Smith, N. G. 1969. The importance of being para-
dichas areas se nava en base a la teoria de los sitized. Nature 219:690-694.
Terborgh, J. 1974. Preservation of natural diver-
refugios. Las reservas designadas para rep- The problem of extinction prone species.
sity:
tilesy anfibios podran ser mas pequenas que Bioscience 24:715-722.
aquellas para aves. Terborgh, J. 1976. Island biogeography and con-
servation: Strategy and limitations. Science 193:
1028-1029.
LITERATURE CITED Vanzolini, P. E. 1973. Paleoclimates, relief, and
species multiplication in equatorial forests, pp.
255-258 in Meggers, B. J., Ayensu, E. S., Duck-
Brown, K. S., Jr. 1977. Centres de evolucao, re-
worth, E. D. (eds.). Tropical forest ecosystems
fugios quaternaries e conservacao de patrimonios in Africa and South America. A comparative re-
geneticos na regiao neotropical: padroes de difer- view. Smithsonian Institution Press, Washington,
enciaeao em Ithomiinae ( Lepidoptera: Numpha-
D.C., 350 p.
Iidae). Acta Amazoniea 7:75-137. 1976.
Villa Nova, N." A., Salati, E., Matsui, E.
Diamond, J. M. 1975. The island dilemma: lessons Estimativa de evapotranspiracao na bacia Ama-
for the design of natural reserves. Biol. Conser.
zoniea. Acta Amazoniea 6:215-228.
7:129-146.
Wetterberg, G. B., Jorge Padua, M. T., Soares de
Haffeh, 1969. Speciation in Amazonian forest
Castro, C, Vasconcelos, J. M. C. 1977. Uma
J.
birds. Science 165:131-137.
analise de prioridades em conservacao da natureza
Lovejoy, T. E., Oren, D. C. 1979. Minimum criti-
cal size of ecosystems, in Burgess, R. L., Sharpe,
na Amazonia. Projeto de Desenvolvimento e
D. M. (eds. ). Forest island dynamics in Man- Pesquisa Florestal. Instituto Brasileiro do Desen-
New volvimento Florestal, Ser. Teen. (8), PNUD/
dominated landscapes. Springer- Verlag.
York In press). FAO/IBDF/BRA-545. Brasilia, 62 p.
(
SUBJECT INDEX
Roman numerals refer to text, italics to maps and photographs.
herpetofauna 57 100
origin of herpetofauna 64 fossil record 88, 97
separation from South America 29, 55, 219 herpetofauna 73
species/area 15 history of continent 74
Altiplano 373, 378, 398, 399 origin of herpetofauna 77, 83, 91
climate 165 species/area 15
herpetofauna 400 Australopapuan fauna
tectonics 161, 400 distribution patterns 77, 80, 84
vegetation 170, 399 origins 90
Altitudinal distribution Avifauna
Central Andes
398 distribution patterns 122, 124, 129
Guiana Highlands 252 paramo islands 417
Northern Andes 388
Southern Andes 350, 401 Barriers 166
Andes 23, 158, 371 Andean 165, 402, 405
barriers to dispersal 165, 166, 382, 389, 402, Austral forests 360
405 Depresion de San Cristobal 382
Central Andes 373, 394 Essequibo-Rio Branco Depression 264
climate 158, 164, 387 Huancabamba Depression 389
description 23, 157, 341, 371, 380, 381, 385, inter-Andean valleys 164
398, 399 Marafion Valley 167
dispersal routes 412, 420 Beringia 17
displacement of vegetation zones 111, 172, Biogeographical theories 55, 114, 254, 417
177 Brasilian Highlands
distribution patterns 380, 387, 389, 396, 400, age 163
402, 414, 415-416, 421-423, 425, 443 climate 167
ecophvsiographic regions 402, 405 glaciation 180
endemism 381, 383, 387, 389, 404 Quaternary history 180
faunal similarities within 383, 388, 397, 401 tectonics 163
faunal similarities without 406 vegetation 171
geological history 157, 346, 387, 398
glacial climates 173 Caatinga 22, 299
glaciation 151, 172, 400 Center of evolution 1 17
habitats 373, 375-379 Centers of origin 70
herpetofaunal communities 426 Central America 16, 383, 407
human modifications 374, 434 connection with South America 16, 173
inter-Andean basins 373, 386 faunal exchange with South America 17, 221,
origin of herpetofauna 409 414
Quaternary history 416
172, radiation of amphibians 208
species richness 387, 397, 404 Central Andes 373, 394
vegetation 169, 373 herpetofauna 389
Areas of congruence 409 physiographv '
396
Arid Andean habitats 373, 375, 399 Cerrados 22, 299
Atacama Desert 22, 23, 417 Chaco 299, 300
Aunt Arctica 73 Chaco-Pampean Plain 141, 142, 149
Austral forests 20, 310, 341, 342, 374, 376 Chilean Archipelago 342
climate 344 Chilean Lake District 344
distribution patterns 347, 349-354, 356 Climate
endemism 363 Altiplano 165
MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
within Andes 383, 388, 397, 401 Human modifications in Andes 374, 37.9, 434
within tropical forests 190, 221, 229, 291 Humboldt Current 20, 22, 99, 182, 417
Fossil record
Africa 57, 61, 63 Immunological distance and continental divergence
amphisbaenians 62 93
Australia 88, 89, 97 Immunological evidence of relationships
birds 114 hylid frogs 92
caecilians 57 phyllomedusine frogs 8
crocodilians 61, 90 Inabresia 56, 58, 64
Europe 62 Insect distributions and speciation 123
frogs 89
58, Inter-Andean Basins 373, 386
Laurasia 61 herpetofauna 388
mammals 114, 146, 148, 151 physiography 386
salamanders 57 Interglacial phases
snakes 63, 89 climate 147, 150, 173, 179
South American amphibians 312, 356
South American reptiles 219, 312 Karyological evidence of relationships
South American taxa, list of 51 hylid frogs 91
turtles 61, 88, 97 leptodactylid frogs 358, 360, 362
Fossil sites 34, 34-38, 311 microhylid frogs 60, 85, 87
Galapagos Islands 17, 22, 364 Lago Titicaca 175, 371, 399
Geological interpretations Lost Pacific continent 100
Amazon Basin 113, 287 Lithostratigraphic units 144
Geomorphological interpretations 111, 183 Llanos 281, 282, 284, 286
Geomorphological units 160, 314-315
Glacial climatology, theory 181 Mammal ages 144
Glacial phases Mammal distributions and speciation
ages 113, 173, 347 Amazon Basin 128
climate 108, 113, 147, 150, 173, 179 Mandibular musculature, frogs 81, 84, 86, 91
correlation with northern 146, 174, 178 Mantaro-Apurimac Valley 373, 393
number 146, 173 Maraiion Valley 22, 373, 393
Glaciers Megafaunal extinction 114, 145
Andean 151, 172, 400 Merida Andes 372, 381
Brasilian Highlands 180 herpetofauna 382
Chilean 178, 343, 347, 359 physiography 380
Patagonian 146, 178 Mesopotamia 141, 142, 148, 299
Gondwanaland Montane rainforest 374
breakup of 29, 55, 75, 219 Monte 309, 310
distributions in 14, 56, 65
historical components of 16 Natural reserves 23, 434
Guiana Shield forest refugia 132
age 163 size 462
climate 168, 247 New Guinea 74
delimitation 242, 243-244, 246-248 connection with Australia 76
erosion 164 herpetofauna 81
geology 163, 242 New Zealand 74
herpetofauna 249 connection with Australia 100
Quaternary history 180 herpetofauna 99
vegetation 172, 247 Nonforests 22, 282
Northern Andes 372, 385
Habitats in Andes 373 herpetofauna 387
Habitat utilization physiography 384
amphibians 351 Nudo de Pasto 372, 384
lizards 304
reptiles 351 Oceanic dispersal 99
Herpetofauna, South American
geographic origins of families 4 Paleofloras
fossil record 29, 51 Maslin Bay 75
review of families 2
Patagonian 311
taxonomic composition 3 Tertiary-Chaco 19, 309, 414
Historical biogeographic analysis, method 118 Valdivian 311
Huallaga Valley 373, 393 West Gondwanan 19
Huancabamba Depression 22, 218, 221, 373, 389 Palynological interpretations 110, 163, 172
468 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
Valle Central de Chile 342 Vicariance model 18, 70, 117, 409
Vegetation
Andes 169, 374 West Indies
Austral forests 345 endemism 18
Brasilian Highlands 171 origin of herpetofauna 19
Chaco 300 origin of islands 18
Guiana Highlands 172, 246, 247 taxonomic diversity 18
Llanos 281 Western Pacific island faunas 99
northern lowlands 281 dispersal routes 80
Patagonia 316 distribution patterns 82
Pleistocene 223
vertical shifts in Andes 111, 172, 177
470 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
TAXONOMIC INDEX
All scientific names of families and lower taxonomic levels are indexed, except those trivial names appearing
in appendices. Roman numbers refer to text, italics refer to figures and boldface to appendices.
Acacia 171, 287, 300, 373, 374 amciva 289, 290, 291, 463
caven 341, 342, 364 bifrontata289
Acacioxtjlon 148 Amphibolurus 87
Acaena 173, 374 Amphichelydia 61
caespitosa316 Amphignathodon 387, 410, 411, 413, 450
macrostemon 316 Amphignathodontinae 5, 8, 17, 61, 194, 204, 205,
Acantholippia seriphioides 317 208
Achropogon 172 Amphisbaena 11, 18, 218, 219, 222, 227, 238, 279,
Acontinae 87 296, 339
Acris 93 alba 291
Acrochordidae 13, 88 camura 302, 304
Acrochordus 88 fuliginosa 291
Adenomera 190, 206, 213, 214, 215, 269 Amphisbaenidae 3, 4, 5, 11, 14, 17, 32, 54, 56, 63,
andreae 197, 201 65, 251
bokermanni 201 Anadia 238, 372, 382, 407, 410, 411, 414, 420, 429,
191, 197, 200, 201
hylaedactyla 444, 453
marmorata 201 bitacniata 381
Adesmia 171, 316, 373, 374, 377, 399 brevirostris 381
horrida 170 pulchella 383
Adiantum 345 Anaea 125
Aegla 352 AnarthrophyUum
Aextoxicum punctatus 345 desideratum 316
Afrixalns 61 rigidum 316
Agahjchnis 8, 208, 212, 213 Anemone 180
Agamidae 14, 15, 55, 57, 62, 65, 66, 71, 87 Anguidae 3,4,5,11.17,32,56,65
Agave 434 Aniliidae 3, 4, 5, 12, 32, 42, 51, 53, 251
Agropyron magellanicum 316 Anilius 42, 218, 239, 274
Alligatoridae 31, 32, 33, 43, 44, 51, 52, 53, 54, 65, Anisolcpis 238
220, 251 Anodontolujla 87
Allophryne 194, 214, 256, 270 Anolis 2, 19, 22, 218, 227, 238, 277, 278, 296, 327,
nithveni 254 382, 383, 407, 408, 410, 413, 444, 453
Alnus 163, 169, 173, 175 annectens 289, 291
jorullensis 169 antonii 387
Alopoglossus 222, 227, 238, 278, 408, 410, 453 auratus 291
Alouatta chloris 387
senicuhts 128 chrysolepis 123, 126
villosa 128 chrysolepis group 123, 124
Abodes 20, 325, 326, 327, 338, 341, 351, 361, 363, fuscoauratus 222
365, 400, 407, 409, 410, 411, 412, 420, 457 nigropunctatus 381, 382
gargola 361 onca 289, 291
gargola 320 ortoni 222
neuquensis 321 punctatus222
monticola 347, 348, 351, 355, 361 Anomalepidae 3, 4, 5, 11, 17, 18, 251
nodosits 361, 400 Anomalepis 11, 218, 222, 239
pehuenche 361 Anops kingii 302, 304
verrucosus 361 Aparasphcnodon 214, 215, 271
Alsophis 12, 19, 22, 341 vcnczolanus 254, 257
angustilineatus 364 Apodichelys htcianoi 51
cantherigenus 364 Apodops 56, 57
348, 350, 352, 356, 364
chamissonis /(nee! 51
tachymenoid.es 364 Aporophis 274
Amapasaurus 238, 256, 278 Apmtolepis 13, 218, 227, 239, 274
tetradactylus 255 Apuleia fcrra 171
Amblyphrynus 383, 410, 411, 413, 445 Araucarta 151, 171, 180, 309, 316, 321, 334, 345,
383
ingeri 316,374, 400, 11 1
Bufo 4, 7, 39, 53, 54, 58,189, 190, 192, 212, 213, Callopistes 22, 33, 41, 53
214, 215, 270, 295, 301, 338, 341, 351, 353, Callopsis 54
362, 363, 400, 407, 410, 412, 426, 431, 434, Calhiella 81, 82
443, 449, 457 Caltha 316
arenarum 59, 301, 302, 320, 326 Calyptahyla 7, 18
Cercosaura 227, 238, 278, 296, 408, 410, 453 leachi 302, 304
Chaetophractus vellerosus 148 lemniscatus 289, 290, 291
Chamaeleolis 19 lemniscatus 260
Chamaeleontidae 55, 57, 62, 65, 66 Cocus 171
Chamaelinorops 19 Coleodactylus 10, 227, 277
Changlosaurus 63 meridionalis 222, 260
Chaperina 60 Colliguaja 342
Characoidae 128 integerrima 316
Charactosuchus 53, 220 Colobodactylus 238
fieldsi 43 Colobosaura 227, 238
Chelidae 3, 4, 5, 9, 14, 31, 32, 44, 52, 53, 97, 100, Colombophis 53
251 portai 42
Chelodina 97 Colostethus 2, 22, 192, 212, 213, 214, 215, 269, 372,
Chelonia 273 374, 380, 382, 383, 388, 406, 407, 410, 411,
Cheloniidae 251 414, 420, 427, 429, 443, 444, 450
Chelonoides 40, 53 abditaurantius 387
Chelus 40, 52, 53, 218, 220, 238, 273, 296 fuliginosus group 197
colonibianus 40 lierminae 382
fimbriatus 258 mandelorum 381
lewisi 40 Colubridae 3, 4, 12, 14, 32, 33, 42, 53, 54, 63 88
Chelyearapookidae 98 228, 251
Chehjcarapookus 98 Colubrinae 5, 12, 13, 17, 18, 56
arcuatus 98, 100 Columbia
Chelydra 9, 218, 222, 238 corensis 130
Chelydridae 3, 4, 5, 9, 17, 18, 31 picazuro 130
Chersydrus 88 Coniophanes 18, 222
Chiasmocleis 60, 214, 215, 273 Conolophus 10, 17
Chimantea 172 Constrictor constrictor 302
Chiromantis 60 Copernicia 282, 300
Chhonius 13, 18, 64, 220, 222, 227, 239, 275, 297, tectorum 282, 284
408,411,444,455 Cophixalus 83, 84, 87
bicarinatus 224, 225 Cophylinae 60
carinatus 222, 225, 291 Corallus 18, 227, 239, 274, 297
cxoletus 224, 225 enhydris 291
pijrrhopogon 225 Cordicephalus 58
flavolineatus224 Cordylidae 14, 15, 57, 62, 66
foveatus 224, 225 Cornufer 78
/uscus 224, 225 Cortaderia 316
monticola 381, 382, 388, 397, 420, 423 modesta 172
multivcntris 224, 225 Cori/p/jospingus
quadricarinatus 224 cucullatus 130
scurruhis 224, 225 pileatus 130
laevicollis 224, 225 Corythophanes 222, 238
Chondropython 89 Cremolobus 175
Chrysemys 9, 18, 218, 220, 222, 296 Crepidophryne 7
scripts 289 Cricetinae 145
Chrysobalanceae 125 Cricosaura 19
Chtiwnerpeton 192, 194, 214 Crinia 95
Chuquiraga 170, 316, 342 haswelli 95
aurea 317 Crocodylidae 3, 4, 5, 14, 17, 31, 32, 33, 42, 43, 51,
avellanedae 317 53, 56, 61, 66, 100, 101, 251
Chusquea 312, 321, 345, 374 Crocodilurus 218, 238, 278
penifolia 172 lacertinus 258
guj'Za 345 Crocodylus 14, 18, 43, 98, 218, 220, 222, 238, 274,
Cichlidae 128 295, 296
Ctssus 309, 345 acutus 18, 220
C/e/m 18, 219, 227, 239, 275, 297, 408, 411, 455 intermedins 220, 257, 289, 291
clelia 290, 303 johnsoni 98
occipitolutea 303 novaeguinae 98
Clemantis 180 porosus 98
CZuria 169 Crossochclys 52
Cnemidophorus 11, 18, 23, 218, 227, 238, 278, 296, Crossodactylodes 194, 215
326, 339 Crossodactylus 194, 215
lacertoides 304 dispar 199, 200
474 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
388, 389, 406, 407, 410, 416, 420, 427, 428, Eocaiman 52, 53, 220, 312
429, 443, 445, 446, 447, 448 cavernensis 43
achatinus 203 Eoxenopoides 58
acuminatum 200 Eophractus 356
affinis 418 Ephedra 143, 316, 374
alfrcdi group 208 Epicrates 219, 227, 239, 274, 297
biporcatus group 208 cenchria 290, 291, 302
bicumulus 196 Epicrionops 9, 192, 212, 213, 273, 408, 410, 445
binotatus 200 Epilobium 171
binotatus group 201 Equu.s 175, 178
boconoensis 381 Erethizontidae 148
bogotensis 418 Erctmochphjs 273
bogotensis group 418 Erichosaurus debilis 312
buckleyi 388 Erycinae 63
cajamarccnsis 389 Erymnochelys 61
carmelitae 383 Erythrolamprtis 219, 227, 239, 275
chiastonotus 201, 203 Erythroxylon 309
conspicillatus 191,203 Escallonia 169, 172, 176, 180
curtipes complex 418 Espeletia 169, 170, 173, 374, 377, 387
discoidalis 208 Eublepharidae 96
elegans 418 Eublepharinae 62
elegans complex 418 Eucryphia cordifolia 345
fenestratus 203 Eugenia 169, 171
fitzingeri 196 Eulychnia 171
fttzingeri group 201, 203, 208, 383, 413, 420 Eumeces 11, 86, 87
ginesi 381 Euncctcs 33, 42, 53, 218, 239, 274, 297
inoptatus group 19 murinus 289
insignitus 383 notaeus 303
lacrimosus 198 Euparkerella 213, 215, 256, 269, 408, 410, 448
lancinii 381 myrmecoides 201
longirostris 196 Euphorbia 317
lynchi 418 Euphractus 148, 149, 150
maussi 197, 203 Eupsophus 20, 33, 38, 52, 3ii, 312, 327, 341, 350,
nicefori 418 351, 360, 361, 362, 363, 365, 412
nigrovittatus 201 migueli 348
ockendeni 200 roseus 312, 347, 348, 352, 355, 358, 360
peruvianus 203 vanzolinii 348
prolixodiscus 383 vertebralis 348
pulvinatus 255 uittafus 347, 348, 352, 355, 358
ranifor mis 195, 197 Euspondylus 23, 238, 253, 265. 278, 396, 407, 410,
rozci 196 411,413,414,444,453,454
sulcatus 203 leucostictus 241
sulcatus group 203, 413 Eusuchia 31, 33, 61
terraebolivaris 196, 203 171
Euterpe
unistrigatus 208, 388
unistrigatus group 383, 411, 414, 418
Fagara 309
uric/ii 382
Ferfuca 170, 316, 326, 374, 387, 399
vertebraK? 388
monticola 316
t'l'/ars! 203
Feylininae 87
w-nigrum 388, 420
Ficus 309
zeuctotylus 201
Fitzroya 309, 345, 374, 400
EZosi'a duidensis 242, 253, 254
cupressoides 171, 345
Elosiinae 4, 5, 193, 194, 204, 206, 207
Fleetonotus 23, 214, 215, 381, 408, 410, 413, 444
Elseija 97, 98
fissilis 200, 382
Embothrium 309 382
fitzgeraldi
Emmochliophis 218, 222, 239 pygmaeus 196, 382
Emydidae 3, 4, 5, 9, 31, 44, 53, 54, 61, 251 F/i/ui'co/a
Emijdura 97 nengeta 130
macquari 98 pica 130
Engystomatidae 59 Frankcnia 316
Emdius 218, 222, 239, 297 Franseria 171
Enyalioides 18, 22, 222, 227, 238, 408, 410, 453 Fritziana 195, 215, 413
Enyalius 227, 238 Fuchsia 345
476 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
faber 199, 201 56, 57, 62, 65, 99, 123, 228, 251
fasicata 198 Iguaninae 5, 17
fuscomarginata 199 Ikanogavialis 53
fuscovaria 302 Ilchunaia 52, 220
geographica 191, 197, 199, 200, 258 Ilex 169
ginest 254, 256 paraguariensis 171
/tayu 200 Imantodes 227, 239, 275, 297
kanaima 256 cenchoa 290, 291
/aiiafo 381, 382, 420 Indobatrachus 70
labialisgroup 411,413 Insuetophrymts 327, 341, 351, 361, 363, 365
hmciformis 198, 201 acarpicus 348, 355
larinopygkm group 407, 411, 413 Iphisa 238, 278
/cmai 254, 256 Ischnocnema 208, 213, 215, 408, 410, 448
leucophijllata 8, 191, 197, 199, 200
leucophyllata group 61 Jamesonia 180
licteocellata 191 Jtiglans 169
microcephala 17, 289, 290 australis 169
microps 200 Juncits 316
mintiscula 289
minuta 191, 197 Kalophrynus 60, 83
multifasciata 198, 201, 254 pleurostigma 83
omatissima 254 Kaloula 60, 87
pardalis 189 Kankanophryne 95
parviceps group 2i Kentropyx 18, 218, 227, 238, 278, 296
phlcbodes 195 calcaratus 222
platydactyla 381, 382 lagartija 302
proboscidea 255 striates 291
pulchella complex 400 riridistriga 302
raniceps 189, 302 Kinosternidae 3, 4, 5, 9, 17, 18, 31, 40, 251
rhodopepla 198 Kinosternon 9, 218, 220, 227, 238, 273, 296, 303
rodriguezi 255 fcauri 18
rostrata 260, 289, 290 scorpioides 289
rostrata group 197 Kyarranus 96
rubra 8, 190, 191, 195, 197, 200, 289
senicula 199, 200 Lacertidae 15, 55, 57, 62, 65, 66
mclanargyrea 259 Lachesis 219, 227, 239, 277
sibleszi 254, 256 mi/ fa 13
wandae 289, 290 Lagostomus maximus 301
x-signata 260, 302 Lampropeltis 222, 239, 408, 411, 444
Hylactophryne 208 triangulum 382
Hylidae 3, 4, 7, 14, 15,
30, 32, 39, 55, 56, 60, 65, 73, Laparrentophis 63
75, 89, 90, 91, 92, 93, 100, 192, 193, 194, Larrca 316
207, 251 ameghinoi 316
Hvlinae 193, 194, 204, 206, 207,
5, 17, 208 Laurclia 171, 309, 345, 346, 414
Hylodes 195, 215, 270 serrata 345
duidensis 253, 254 Leguminosae 125
gollmeri 191 Leimadophis 12, 218, 221, 227, 239, 275, 276, 297,
marmoratus 241 339,407, 408,411,444,455
Hyhphorbus 81, 84 bimaculatm 381
Hylorina 20, 327, 341, 361, 363, 365 melanotus 291
sylvatica 347, 351, 355, 358, 360 regina 291
Hymenochirus 58 sagittifer 303
Hyophryne 195, 215 h/pWus 291
Hypericum 169, 173 zweifeli 382
Hyperoliidae 14, 55, 57, 60, 65 Leiocephatus 19
Hyperoliini 61 Lciopehna 3, 58, 100
478 MONOGRAPH MUSEUM OF NATURAL HISTORY NO. 7
fuegiana 178 65, 71, 79, 83, 84, 86, 101, 192, 193, 204,
Micrurus 13, 218, 227, 240, 277, 297, 303, 339, 408, /usca 75
411, 444, 456 menziesi 75
circinalis 289 obliqua 171, 345
dissoleucas 289 procera 171
frontalis 303, 320 pumilio 171, 345
isozonus 289, 291 Nothopsis 218, 240
mipartitus 381, 382, 383, 384, 387, 388 Nothrotheriinae 143, 149
surinamensis nattereri 257 Notobatrachus 3, 44, 58, 100
Mimosiphonops 194, 214 Notocaiman 220
Mixophys 94 stromeri 43
Morelia 89 Notosuchidae 61
Morpho 125 Nyctimantis 194, 213
Morunasaurus 222, 227, 239 Nyctimystes 7, 90, 93
Mourasuchus 53 Nyctimystinae 90, 93
Mouriri 125
Muehlenbergia 173 Octodon degus 352
Mulinum 316, 374 Octodontidae 149
spinosum 316 Odatria 85
Sfus musculus 352 Odontophrynini 4, 194, 204, 207
Mi/fma 170, 174, 175 Odontophrynus 193, 301, 326, 338
Myersiella 195, 215 americanus 302
Mylodon 146, 178 Oedipina 2, 192, 194, 212
Mylodontidae 145 complex 195
Myobatrachidae 56, 65, 70, 73, 94, 95, 96 Ogmodon 88
Myobatrachinae 95, 96 Omoiotyphlops 63
Myobatrachus 95 Ophiodes 11,218,239
Myrceugenella 171 intermedins 302, 304
Myrceugenia 176, 309 Ophryoessoides 222, 239, 296, 409
Myrica 172 caducus 302
Myristica 309 erythrogaster 289, 291
Myrmecophagidae 143, 148 Opipeuter 396, 410, 411, 413, 414, 454
Myzodemdrum 345 Opuntia 285, 287
Oreopanax 169, 309
Nassauvia 171, 316, 326 Oreophrync 83, 84
aculcata 316 Oreophrynella 7, 253, 255, 270
Natricinae 13 jnacconnelli 241
Mrfra fasciata 18, 19 quelchii 241
NebZfrui 172 Oscaecilia 9, 18, 57, 192, 212, 213, 214, 273
Nectocaccilia 213, 214 Osornophryne 192, 387, 410, 411, 413, 426, 418,
Necrosuchus 43, 220, 312 429, 449
tonensts 42, 51 Osteocephalus 21, 22, 213, 215, 272, 408, 410, 452
Nectandra 309 langsdorffi 200, 201
Nectophrynoides 59 taurinus 197, 201
Neobatrachus 95 Osteopilus 7, 18
Neochelys 61 Otophrync 60, 87, 194, 214, 253, 255, 273
Neoprocoela 4, 38, 39, 52, 32 i, 312, 363 rofcusfa 241, 253, 254, 256, 257, 260
edentata 363 Oxybelis 64, 227, 240, 276, 297
Nettosuchidae 31, 33, 43, 53 aenci/s 290, 291
Nettosuchus 220 Oxyrhopus 64, 218, 227, 240, 276, 297, 408, 411, 456
Neusticurus 218, 234, 239, 278, 408, 410, 454 pefo/a 291
racenisi 256 rhombifer 303
rudfa 241, 256, 257
totei 256 Pachymedusa 208
Nigerophis 63 Palacolama 151
Ninia 218, 227, 240, 276 Palaeophidae 63
Niolamia patagonica 51 Palaeobatrachidae 58
Paleosuchus 14, 218, 220, 227, 238, 274, 296
Niphogeton 174
170 Palorchestes 98
Nordophyllum
a^aeZ 98
Nothofagus 75, 170,176, 178, 223, 309, 312, 316,
Pantacantha 316
321, 345, 346, 347, 352, 359, 360, 361, 362,
364, 365, 374, 376, 400, 404, 414 Pantodactylus 227, 239
antarctica 171,345,346 (i//en 241
betuloides 171, 346 Paracophyla 87
dombeyi 171, 345 Paracrinia 95
1979 INDEX 481
Plethodontidae 2, 4, 5, 17, 57, 191, 192, 204, 207 torquatus 348, 351, 354, 364
Pletlwdontohyla 87 valeriae 364
Pleurodema 4, 18, 20, 21, 23, 270, 295, 301, 326, vautieri 302, 304
338, 341, 362, 363, 372, 400, 407, 409, 410, Pristiguana 62, 219
411, 412, 415, 431, 434, 457 brasiliensis 51
302
horellii Proalligator 220
brachyops 21, 23, 260, 290 Proceratophrys 195, 214
bufonina 317, 320, 321, 325, 326, 364, 402, fcoiei' 200
420 appendicular 200
cinerea 431 cristiceps 199
guayapae 302 Proctoporus 374, 398, 411, 413, 414, 424, 429, 431
marmorata 396, 420 434, 444, 454
nebulosa 326 /aet-i's 388
thaul 321, 326, 347, 348, 350, 351, 352, 355, occulatus 388
358, 362 Proctotretus 323, 326, 327, 339, 409
tucumena 302 doellojuradoi 302, 304
Pleurodira 52 pectinatus 320
Pleurofora patagonica 317 Procyonidae 145
Plica 234, 239, 278 Proganochelydia 61
Pliocercus 227, 240 Prosopis 170, 171, 285, 300, 373, 374
Poa 170, 316, 326, 342, 374, 399 /crox 170
humilis 170 patagonica 316, 317
Podocarpus 75, 169, 171, 175, 178, 180, 309, 345, Proterochersis 61
373 Proustia 171
andinus 178 Pseudemydura 97
lambertii 171 Pseudemys 9, 238
montanus 169 Pseudhymenochirus 58
nubigenus 169, 171 Pseudidae 3, 4, 5, 7, 30, 56, 60, 193, 251
oleifolius 169 PseuA.s 272, 295
parleteoreii 169 paradoxus 260, 261, 289, 290, 302, 305
rospiglossii 169 occidentalis 304, 305
salignus 346 plutcnsis 304
Podocneminae 61, 65 Pseudoboa 18, 218, 222, 227, 240, 276, 297
Podocnemis 9, 31, 39, 41, 51, 53, 218, 220, 227, 238, neuwiedi 289, 291
273, 285, 296 Pseudoeryx 218, 240, 276
argentinetisis 52 Pseudogonatodes 10, 238, 277, 296, 383, 408, 453
bassleri 53 Pseudonaja 88
brasiliensis 51 nuchalis 88
elegans 39, 51 Pseudopaludicola 270, 295
expansa 33, 39 /wsiHa 25S, 260
uogfc 289, 291 Pscudophryne 95
Polychroides 389, 410, 411, 413, 414, 453 occidentalis 95
Polychrus 220, 227, 239, 278, 296, 408, 410, 413, Pseudotomodon 339
414 trigonatus 303
gutturosus 387 Pseusfcs 227, 240, 276, 298
marmoratus 222 Psidum 346
Polylepis 163, 170, 173, 174, 175, 374 Psyllophryne 195, 214
australis 169 Ptychadena 61
cocuyensis 169 Ptychoglossus 227, 239, 279, 408, 411, 454
sericea 169 Ptychophis 240
Primula 171 Purussaurus 53, 220
Potomotyphlus 194, 213, 273 Puya 170, 171, 173, 373
Prionodactylus 222, 227, 239, 279, 407, 411, 414, Pycnophyllum 170
429, 454 Pygopodidae 14, 98, 100
arguhis 397 Pi/f/ion 89
manicatus 397 Pythonini 55, 57, 63, 66, 71
vertebrate 387
Pristidactylus 326, 327, 341, 364, 408, 410, 431, 458 Quclchia 172
achalensis 323, 326, 364 Quercus 163, 173, 175
alvaroi 364
casuhatiensis 323, 326, 364 Ramanella 60, 87
fasciatus 323, 326, 364 Ranfl 79, 82, 213. 214, 215, 269, 295
pectinatus 326 palmipes 8, 58, 190, 191, 193, 197, 200
scapuhtus 323, 326, 364, 401 pipicns 190
1979 INDEX 483
Ranidae 3, 4, 5, 8, 14, 15, 17, 30, 39, 55, 56, 58, 59, Scolecomorphidae 14, 57
60, 65, 77, 78, 101, 192, 193, 204, 207, 251 Scythrophrys 195, 214
Ranidclla signifera 90 Sebecidae 31,32,33,43,51,52,53
Ranunculus 171, 180, 316 Sebecosuchia 43
Relictivomer IS, 190, 213, 214, 273, 295 Sebecus 51, 52, 53, 220, 312
pearsei 191, 195 icaeorhinus 98
Rliacophoridae 90 Senec/o 169, 170, 172, 175, 326, 374, 387
Rhadinaea 218, 227, 240, 276, 408, 411, 444, 456 filaginoides 316
lateristriga 388 Setaria 300
Rliamnophis 64 Sliclania 58
Rhamphophryne 7, 18, 192, 213, 214, 372, 407, 410, Sifcon 222, 240, 274, 298, 408, 411, 456
411,413,418,420,449 nebulata 290
Rhamphostomopsis 33, 53, 220 Sibynomorphus 240, 408, 411, 456
Rlicobatrachus 96 oneilli 389
silus 95 turgidis 303
Rhinatrema 192, 194, 213, 256, 273 Siphlophis 222, 227, 240, 276
Rhinatrematidae 3, 4, 5, 9, 56, 57, 192, 207, 251 Siphonops 190, 213, 214, 273
Rhineura 63 annulatus 191, 197, 199
Rhinobothryum 227, 240, 276, 298 paulensis 199
Rhinoclemys 9, 218, 220, 238, 273, 296 Smilisca 8, 194, 212
Rhinodcrma 7,341, 351, 354, 359, 362, 363, 365 phaeota 195
darwinii 347, 348, 352, 354, 355, 358, 362 195
stfa
Cfe^
A£R 1 5 1996
-
m 3 1 ZQOZ
WAY 2 2002-
Contributors:
Jose M. Cei
James R. Dixon
»
William E. Duellman
• J. Ramon Formas
• JoseM. Gallardo
• Zulma B. de Gasparini
•
Jiirgen Haffer
* Marinus S. Hoogmoed
•
Raymond F. Laurent
• Thomas E. Lovejoy
• John D. Lynch
• Carlos Rivero-Blanco
•
Beryl B. Simpson
• Michael J. Tyler