Marine Geology: Toru Tamura, Yoshiki Saito, Mark D. Bateman, V. Lap Nguyen, T.K. Oanh Ta, Dan Matsumoto

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 14

Marine Geology 326–328 (2012) 140–153

Contents lists available at SciVerse ScienceDirect

Marine Geology
journal homepage: www.elsevier.com/locate/margeo

Luminescence dating of beach ridges for characterizing multi-decadal to centennial


deltaic shoreline changes during Late Holocene, Mekong River delta
Toru Tamura a, b,⁎, Yoshiki Saito a, Mark D. Bateman b, V. Lap Nguyen c, T.K. Oanh Ta c, Dan Matsumoto a
a
Geological Survey of Japan, AIST, Central 7, 1-1-1 Higashi, Tsukuba, Ibaraki 305‐8567, Japan
b
Sheffield Centre for International Drylands Research, University of Sheffield, Winter Street, Sheffield S10 2TN, UK
c
HCMC Institute of Resources Geography, VAST, 1 Mac Dinh Chi St., 1 Dist., Ho Chi Minh City, Vietnam

a r t i c l e i n f o a b s t r a c t

Article history: Empirical understanding of decadal- to centennial-scale deltaic shoreline changes in the past is essential for
Received 12 March 2012 understanding the fate of coasts in the coming decades and centuries. We tested the effectiveness of quartz
Received in revised form 14 August 2012 optically-stimulated luminescence (OSL) dating of beach ridges to constrain shoreline changes of the Mekong
Accepted 14 August 2012
River delta over the last 3500 years. Forty-seven OSL ages have been analyzed from the Tra Vinh delta plain, cen-
Available online 25 August 2012
tral Mekong River delta. The Tra Vinh beach ridges are recurved and branching, showing the hierarchy of shore-
Communicated by J.T. Wells line changes, which include discontinuous shifts c. 5 km seawards caused by the emergence and elongation of a
delta-front bar/island, and subsequent downdrift accretion of spits. The spit accretion is interrupted by changes
Keywords: of shoreline orientation. Each of the discontinuous shifts resulted in a cluster of beach ridges, which is referred to
delta as ridgeset. The beach ridge sediments were found to have excellent luminescence properties resulting in low
Mekong River age uncertainties of c. 5%. The OSL chronology agrees well with shoreline changes over recent decades and
monsoon with radiocarbon ages of tidal flat sediment underlying the beach ridges. The chronology clearly illustrates the
OSL dating coastal progradation from 3500 years (relative to AD 2010) onwards. Two main types of beach ridge are present:
sediment discharge
recurved and trunk ridges. OSL ages of recurved ridges are consistently younger both downdrift and seawards,
Vietnam
documenting decadal- to centennial-scale shoreline migration especially over the last 1500 years. Trunk ridges
in contrast have less systematic ages because a trunk ridge is formed where the sediment is likely to have
been reworked by waves. Ages of updrift trunk ridges characteristically show relatively young ages in each
ridgeset, representing the extensive erosion and resedimentation of the updrift coast. Major changes in shoreline
orientation occurred around 400–500 years, suggesting strengthening of the northeasterly winter monsoon at
the beginning of the Little Ice Age. The rate of progradation of the Tra Vinh delta plain is described in two
ways: frequency of the discontinuous shift, and growth rate of the delta plain area. Both suggest the coast has
expanded regularly over the last 1000 years. A decrease in sand supply to the coast in the last few decades
due to river dam construction and fluvial sand dredging is inferred, possibly affecting the behavior of the modern
and future shorelines, which can be compared with the less human-influenced, centennial- to decadal-scale past
shoreline changes reconstructed in this study.
© 2012 Elsevier B.V. All rights reserved.

1. Introduction Better understanding and prediction of coastal changes in the coming


decades and centuries necessitate a long-term perspective on continuous
Knowledge of past coastal changes is essential to a better under- shoreline histories with a comparable time resolution. Unlike European
standing and to prediction of the present dynamics of coasts within coasts, where observations have been recorded for the last several hun-
the context of long-term fluctuations. Changes to the depositional dred years (e.g., Ballarini et al., 2003), knowledge of historical changes is
coasts occur in relation to many factors, including sea level, sediment generally limited along much of the world's coast. Southeast Asian river
supply, climate, and human activity. Past changes related to fluctua- deltas are typical regions with limited historical observations, but have
tions in these factors, if known precisely and accurately, are expected clearly been subject to significant coastal changes as a result of the
to suggest a range of phenomena that potentially occur in the future high sediment discharge. Although changes to the deltaic shoreline
coasts. have critical impacts for the now densely populated deltaic lowlands,
they are affected by complex interplays of waves, tides, and river dis-
charge (Galloway, 1975), along with additional factors, including climate
variations and basement configuration. These interplays are not easily
⁎ Corresponding author. Fax: +81 29 861 3747. represented by numerical prediction. The millennial-scale shoreline
E-mail address: toru.tamura@aist.go.jp (T. Tamura). changes of large Asian deltas during the Holocene have been relatively

0025-3227/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.margeo.2012.08.004
T. Tamura et al. / Marine Geology 326–328 (2012) 140–153 141

well documented on the basis of their stratigraphy, coupled with radio- The South China Sea coast of the Mekong River delta is mesotidal,
carbon dating (e.g., Allison et al., 2003; Storms et al., 2005; Ta et al., and is exposed to moderate- to low-energy waves. Along this coast,
2005; Tanabe et al., 2006; Tamura et al., 2009). In contrast, centennial- the tide is semi-diurnal with an upper mesotidal range (Gagliano
to decadal-scale shoreline changes of these deltas are generally limited and McIntire, 1968); in Vung Tau (Fig. 1A), the mean tidal range is
both in the historical or geological past, which prevents comprehensive 2.6 m, and the spring and neap tidal ranges are 3.3 and 1.9 m, res-
assessment of the present coasts characterized by increasing human pectively (Tamura et al., 2010). Waves and tides along the coast of
activities. the Mekong River delta have resulted in the formation of bar-trough
Beach ridges are the most evident marker of past sandy and/or (ridge and runnel) beaches. A subtidal delta-front platform, which
gravelly shorelines during the Holocene (e.g., Woods and Leahy, occurs from the low tide level to −4 m (relative to mean sea level),
1986; Mason, 1993; Tamura, 2012). The most significant problem in extends from just below the beach to 5–10 km offshore, beyond
using beach ridges as a past shoreline indicator was the scarcity of which occurs the northeast-striking delta slope (Fig. 1B). The delta
datable material (e.g., Isla and Bujalesky, 2000), except on plain of Tra Vinh occupies the central South China Sea coast of the
carbonate-rich coasts. However, reliable direct dating of beach ridges delta. The 40-km-long shoreline of Tra Vinh extends from northeast
has been enabled since optically-stimulated luminescence (OSL) dat- to southwest, and is slightly convex to the southeast. It is studded
ing has been greatly improved by a single aliquot regenerative-dose with arcuate beach ridges and inter-ridge swales from the shoreline
(SAR) protocol (Murray and Wintle, 2000). Over the last few decades, to 50 km inland. Beach ridges are generally less than 5 m high
many case studies have been practiced using quartz SAR OSL dating relative to the mean sea level (Fig. 2C), and are composed of well-
for beach-ridge chronology (e.g., Murray-Wallace et al., 2002; sorted fine sand, whereas inter-ridge swales are characterized by a
Ballarini et al., 2003; Argyilan et al., 2005; Nielsen et al., 2006; muddy substrate.
Roberts and Plater, 2007; Brooke et al., 2008a, 2008b; López and Clear differences exist in the river discharge and coastal wave
Rink, 2008; Nott et al., 2009; Forsyth et al., 2010; Rink and López, conditions between the wet summer and dry winter monsoons
2010; Reimann et al., 2011; Tamura, 2012). While the majority of (Gagliano and McIntire, 1968; Mekong River Commission, 2008;
these attempts have been applied to beach ridges on coastal barriers Tamura et al., 2010). The maximum discharge rate in the wet season
and strand plains, detailed dating of deltaic beach ridge complexes has is about ten times higher than the average rate in the dry season.
never been undertaken despite the potential to establish high- Waves associated with the summer monsoon are weaker and southerly
resolution characterisation of rapidly prograding deltaic shorelines. to south-westerly, whereas the winter monsoon enhances relatively
In this paper, we attempt high-resolution quartz OSL dating of strong waves from the north‐northeast. Spring and autumn are transi-
beach ridges in the Mekong River delta. The Mekong River delta, an tional periods, and their wave directions are bimodal. The dominance
excellent example of a mixed tide- and wave-dominated delta of the winter monsoon leads to a south-westward net sediment drift
(Galloway, 1975), has prograded more than 200 km seawards since (Tamura et al., 2010). An annual cycle of sediment characteristics occurs
c. 8000 years (Tamura et al., 2007, 2009). Tamura et al. (2012), along the updrift portion of the Tra Vinh coast as mud and very fine sand
based on 31 OSL ages of beach ridges on several delta plains of the are deposited on the beach during the wet season and then reworked
Mekong delta coupled with sediment drill cores and their radio- and transported south-westwards during the following dry season
carbon dates (Ta et al., 2005), clarified the multi-centennial to (Tamura et al., 2010). Waves on the Tra Vinh coast are not energetic.
millennial-scale, three dimensional evolution of the delta. In the Tamura et al. (2010) reported 0.5–1.0 m high breakers observed during
Tra Vinh delta plain, central Mekong delta, the beach shoreline was the dry season of 2005–2008. High offshore waves exceeding 4 m occur
initiated c. 3500 years (Tamura et al., 2012) and has prograded, only in the dry season, but such high waves are likely attenuated over
resulting in a huge delta plain with beach ridges extending up to the wide and shallow delta-front platform (Tamura et al., 2010).
50 km inland from the present shoreline. Here we present 32 new Sediment drill cores and seismic surveys of the delta plain and
OSL age determination from the Tra Vinh delta plain which when its coastal sea, coupled with radiocarbon dating, have revealed the
combined with the 15 ages reported by Tamura et al. (2012) are Holocene sedimentary evolution of the Mekong River delta
used to demonstrate the effectiveness of OSL dating for constraining (Nguyen et al., 2000; Ta et al., 2002, 2005; Tanabe et al., 2003;
multi-decadal to centennial-scale shoreline changes during the Late Tamura et al., 2007, 2009; Proske et al., 2010, 2011; Tjallingii et
Holocene. We further discuss causes of reconstructed changes, and al., 2010; Xue et al., 2010). Early Holocene transgressive deposits
examine implications for the modern shoreline. exist in the incised valley that extends from the modern shelf to
southern Cambodia (Ta et al., 2005; Tamura et al., 2009; Tjallingii
et al., 2010). The maximum flooding of the sea occurred
2. Study area c. 8000 years after a period of rapid sea-level rise, and was recorded
in the brackish-water tidal flat sediment occurring between fresh-
2.1. The Mekong River and delta water facies (Tamura et al., 2009). The Mekong delta then has
prograded seawards to result in the extensive delta plain during
The Mekong River rises in Tibet and flows 4600 km from its source the relative sea-level highstand (Fig. 3). Ta et al. (2002, 2005) com-
(National Astronomical Observatory, 1999; Hori, 2000) through six piled many radiocarbon-dated sediment cores, clarifying the
countries before draining into the South China Sea (Fig. 1A). Its an- sediment architecture and temporal evolution of the regressive
nual water discharge and sediment discharge are 470 km 3 and delta succession. Beach ridges generally occur in the lower delta
1.6 × 10 8 t, respectively tenth and ninth largest among the world's plain, rather than the upper delta plain, although several ridges
rivers (Milliman and Syvitski, 1992). The huge sediment supply to occur at Cai Lay (Fig. 1B), which is characterised by a shallower
the coast has resulted in an extensive delta plain with an area of pre-Holocene basement (Nguyen et al., 2000; Proske et al., 2010,
5.0 × 10 4 km 2 (Hori, 2000). The delta apex lies just north of Phnom 2011). The beach ridge distribution suggests that the sandy beaches
Penh, Cambodia, more than 200 km inland from the present river characterize only the later progradation of the central Mekong River
mouth (Tamura et al., 2007, 2009), from which the delta fans out to delta. In the Tra Vinh delta plain, the beach shoreline initiated
the southeast (Fig. 1A). South of Phnom Penh, the Mekong River c. 3500 years by aggradation on basement shoals (Tamura et al.,
bifurcates, forming two main channels called the Bassac and Mekong 2012). The delta plain propagated laterally later, evolving from river
rivers, respectively. The Mekong River then further branches into sev- mouth bars. While the Tra Vinh delta plain has prograded consistently
eral distributaries. The southernmost distributary and the Bassac over the last 3500 years, the majority of other delta plains in Ben Tre
River bound the Tra Vinh delta plain. and Tien Giang have prograded since 1500 years (Fig. 1B).
142 T. Tamura et al. / Marine Geology 326–328 (2012) 140–153

Fig. 1. A) Map of the Mekong River delta showing locations of sediment drill cores reported by previous studies. Cores KS, PK, and PSG are from Tamura et al. (2009), core GGM from
Proske et al. (2010), cores DT1 and BT2 from Ta et al. (2002, 2005), and cores 69‐2, 70‐2, 72‐2, and 72‐3 from Tjallingii et al. (2010). B) Geomorphology of the Mekong River delta
and bathymetry of the coastal sea relative to mean sea level (Nguyen et al., 2000; Ta et al., 2005) showing beach-ridge OSL ages (Tamura et al., 2012). The delta-front platform
extends from the shoreline to 4-m-deep isobath, offshore of which occurs the delta-front slope. The delta-front slope grades offshore into the prodelta and shelf at depths of
18–20 m. Sediment cores TA and GGN are from Proske et al. (2011), and other cores are from Ta et al. (2002, 2005). OSL and radiocarbon ages are expressed relative to AD 2010.

2.2. Recent changes to the Tra Vinh coast An elongate bar is recognized on the delta-front platform about
5 km offshore from the north-eastern Tra Vinh coast in a QuickBird
Recent changes to the Tra Vinh coast were reported by Tamura et satellite image taken on 14 April 2006, when the tide level was
al. (2010), who analyzed historical maps and satellite images and 0.6 m below mean tide level (Fig. 2A). The exposed bar was 1 km
conducted a series of field surveys to clarify changes of the beach wide, and extended 3 km roughly parallel to the shoreline. In anoth-
profile (Fig. 2). The Tra Vinh shoreline shows asymmetric changes er, unpublished QuickBird image taken two months previously at
between its north-eastern and south-western parts. The north- high tide, the bar was not exposed but its position was marked by
eastern Tra Vinh coast is divided into two parts by a creek (Fig. 2A). breaking waves. Thus, the bar emerged only at low tide in AD 2006
South of the creek, the shoreline prograded seawards up to 600 m (4 years). Surf was also seen slightly seaward of the bar in a Landsat
between AD 1936 and 2001 (i.e., 74 and 9 years before AD 2010, image taken in AD 1989, suggesting that the bar has been present at
respectively) to form a strand plain with beach ridges sub-parallel least since AD 1989.
to the present shoreline and trending from the north to south
(Fig. 2C). This strand plain narrows southwards. In contrast, the 3. Methods
coast north of the creek retreated by up to 400 m from AD 1936 to
2001. After AD 2001, coastal erosion apparently propagated south- The chronology of beach ridges developed on the Tra Vinh delta
wards across the creek, causing the shoreline over 2 km south of plain was determined in order to decipher past shoreline changes that
the creek to retreat. In the southwest, shoreline changes are caused they record. Beach ridges are formed within or adjacent to the beach,
mainly by the south-westward accretion of a recurved, 2–3-km-long and are subsequently isolated from active nearshore process due to
spit (Fig. 2B). The spit accreted 15 km downdrift between AD 1936 further beach progradation, at which point they are preserved as relict
and 2001, which increased the area of the coastal plain by 20 km 2. elongate mounds. Geometry and depositional age of beach ridges thus
After AD 1981 (29 years), the accretion rate decreased to 50–70% generally indicate past shoreline position and shape (e.g., Woods and
of that before AD 1981. The spit accretion resulted in a series of Leahy, 1986; Mason, 1993; Otvos, 2000; Tamura, 2012). We sketched
recurved beach ridges. Meanwhile, the shoreline was partly eroded the axis of each beach ridge on a Landsat image taken in 1989 (Fig. 4),
between AD 1981 and 1989 (21 years), but by AD 2001, the eroded as distinguished by the denser vegetation than in the surrounding
portion had recovered. marsh of the inter-ridge swale. Landform classifications adopted by
T. Tamura et al. / Marine Geology 326–328 (2012) 140–153 143

A Age (yr before AD 2010)


1000 8000 6000 4000 2000 0

+5

Elevation (m)
-10

-20

-30

Fig. 3. Relative sea-level curve of the Mekong River delta during the Holocene (Ta et al.,
2002; Hanebuth et al., 2011).

International Drylands Research luminescence laboratory. Sample


B preparation for the OSL dating was entirely performed under controlled
red light that does not affect the quartz OSL signal. Approximately 2 cm
of sediment was discarded from tube ends as it may have been exposed
to sunlight during sampling. This sediment was used for establishing
moisture content and sample dose rates. Quartz grains 125–180 μm in
diameter were cleaned and extracted from the bulk of sample following
the method of Bateman and Catt (1996). A monolayer of quartz was
mounted on a disc 9.8 mm in diameter to form an aliquot. Aliquots
were measured using TL-DA-20 automated Risø TL/OSL reader equipped
with blue light emitting diodes for stimulation (e.g., Bøtter-Jensen et al.,
C 2002) and a 90Sr/90Y beta source for laboratory irradiation. Measurement
+5
of OSL was filtered with a Hoya U-340 filter.
+4
The single aliquot regenerative-dose (SAR) protocol was used to de-
+3
Elevation (m)

termine the palaeodose (De) using the OSL response to a 5.6 Gy test
+2 1-7
1-6 HTL dose to monitor and correct for sensitivity changes (Murray and
+1
Wintle, 2000). OSL measurements were made at 125 °C with a stimula-
0 MTL
tion time of 80 s. Dose recovery preheat plateau tests (using a four re-
-1 generation point SAR protocol and a range of preheat temperatures of
-2 LTL
between 160 °C and 260 °C in 20 °C increments) were carried out for
-3 sample Shfd10087 to experimentally determine the most appropriate
600 400 200 0 100
preheat to apply within the SAR protocol, and 260 °C preheat was cho-
Distance from the shore (m)
sen (see Supplementary information for further details). A preheat time
Fig. 2. A) QuickBird image taken on 14 April 2006 when the tide level was 0.6 m below of 10 s was used in all cases. A 160 °C cutheat for the OSL response to
mean tide level showing shoreline changes in the updrift Tra Vinh coast since AD 1936 the test dose was used for all samples. For the data used in the final
(74 years before AD 2010), compiled using historical maps and satellite images. An age calculation, five regeneration points were measured including
elongate offshore bar is recognized near the upper right corner. B) Landsat image of AD
0 Gy and a replicate of the first regeneration point which was used to
1989 showing shoreline changes in the downdrift Tra Vinh coast. Locations of images
are shown in Fig. 1B. C) Cross section across the beach ridge plain in the updrift Tra
check whether the sensitivity correction procedure, through calculation
Vinh coast, measured in November, 2006. Location of the section is shown in Fig. 2A. of a recycling ratio, was performing adequately (Supplementary infor-
mation). Rangefinder tests and IR tests (Stokes, 1992) were performed
for each sample to determine regeneration points and to see if there is a
Gagliano and McIntire (1968) and by Nguyen et al. (2000) were also significant feldspar contamination. Resultant OSL shine down curves
considered in the identification of beach ridges. showed that the sand had a high OSL sensitivity so that even relatively
Quartz optically-stimulated luminescence (OSL) dating was young sample had sufficiently bright natural OSL signals (Supplementa-
performed on quartz sand sampled from beach ridges. Forty-seven ry information). Intervals of 0–1.6 s and 64–80 s after the start of stim-
samples were collected from boreholes deeper than 60 cm that ulation were used for initial OSL signal integration and background
were drilled with a sand auger at 46 sites (Table 1) in April and subtraction, respectively.
November 2010. Light-tight stainless steel tubes 30 cm long and A minimum of 20 replicates were measured for all samples. Data
6.5 cm in diameter were hammered into the bottom of boreholes from aliquots were rejected if recycling values fell beyond 1.0 ± 0.1.
to sample sand unexposed to light for luminescence dating above Statistical outliers were also removed using a criterion that removes
or just around the ground-water table. We avoided areas of human aliquots with De that falls outside the 25th and 75th percentiles
disturbance and aeolian dunes (often found on beach ridge crests) from the mean De value (Boulter et al., 2010). In this way, it was
so that the OSL ages reflect the timing of sediment deposition at hoped to exclude any poorly behaved aliquots or those with a signif-
the shoreline. All samples were measured at the Sheffield Centre for icant poorly/less-well bleached component. If retained, these aliquots
144 T. Tamura et al. / Marine Geology 326–328 (2012) 140–153

A sands, respectively. Samples collected in October were water-saturated


but should be unsaturated during dry seasons as they were collected
just around the water table in the rainy season. Their average moisture
was thus assumed to be the mean of 7% and its measured value. Shallow
unsaturated samples taken in April, the end of dry season, have mois-
tures lower than 20%. It is assumed that these samples were saturated
during wet seasons, and their annual moisture was calculated from the
mean of 20% and its measured value. For deep samples collected close
to the water table in April, the measured values were used. Past changes
of moisture content are not predictable, so an absolute error of 5% was
applied to the assumed annual value to account for them. Cosmic dose
rate was estimated based on Prescott and Hutton (1994).
To obtain an age for a sample, the final De value was divided by the
calculated environmental dose rate for each sample. All results are
shown in Table 2. All ages are expressed relative to AD 2010, when
the sampling was carried out.

4. Results

4.1. Beach ridge architecture


10 km
The beach ridges exhibit the architectural hierarchy, which is
characterized by a ridgeset, sub-ridgeset, and individual beach ridge
B (Figs. 4, 5). Ridgesets are defined as consisting of a trunk, which orig-
inates at or near the bank of the distributary that bounds the north-
eastern end of the Tra Vinh delta plain, and have branches that
originate from the trunk. The branches form a series of recurved beach
ridges, generally 1–2 km long, and have an arcuate shape convex to-
wards the south. Recurved ridges accrete downdrift (south-westwards).
Some trunks further bifurcate to define sub-ridgesets. We identified
eleven ridgesets (numbered 1–11 with increasing distance from the
shoreline; Fig. 5). Each ridgeset is associated with up to five
sub-ridgesets, each of which was assigned a character a–e from the sea
to land in addition to the ridgeset number (e.g., the sub-ridgeset b in
Ridgeset 5 is Sub-ridgeset 5b). The width of beach ridge is generally
1–2 km, while the muddy inter-ridge swales are c. 5 km wide on
average.
A typical beach ridge in general gradually recurves away from the
trunk. However, in some places, the trunk and branch ridges contact
at a higher angle (e.g., the updrift part of Sub-ridgeset 5c; Fig. 5).
Updrift trunk ridges exhibit seaward accretion of subparallel ridges
(e.g., Ridgeset 8 and Sub-ridgeset 1b). In places, there are breaks of
trunk ridge (e.g., Sub-ridgeset 3a). Several ridges occur subparallel
10 km to the distributary near the updrift end (e.g., Ridgeset 2).
Ridgesets show variations in internal and external shapes. Recurved
Fig. 4. A) Landsat image of the Tra Vinh delta plain, and B) line drawing of beach ridges ridges are more pronounced in ridgesets 1–6, and in Ridgeset 9 that has
based on the image. The delta-front bar is submerged during the high tide, when the well-defined recurved ridges in its updrift part. Ridgesets 3, 4, and 7 are
Landsat image was taken, and its occurrence is shaded to be distinguished from the
partially located in the updrift part, and their shapes are more promi-
supratidal land and island.
nent seawards. Ridgeset 2 appears to have enclosed the downdrift lee
of ridgesets 3 and 4. The area and width defined by the ridgeset and
would lead to an age under- or over-estimation. All samples showed a its landward swale are also variable as summarized in Table 3.
unimodal De distribution (see Supplementary information for further
details). Over dispersion (OD) values were generally low and thus are 4.2. OSL ages
considered to have been well bleached prior to burial. As such, aver-
age De values for each sample were derived using the Central Age Forty-seven OSL ages were determined for beach ridges at 46 sites
Model (Galbraith et al., 1999). in the Tra Vinh delta plain (Tables 1, 2). Each sample site was
Both naturally-occurring radioisotope contribution and cosmic assigned a number within an individual ridgeset as reflecting the in-
radiation were considered for determining environmental dose rate. ferred order of deposition (Fig. 5). The depositional order was deter-
Concentrations of potassium, uranium, thorium, and rubidium were mined assuming that a seaward sub-ridgeset is younger and that
quantified by inductively-coupled plasma mass spectrometry (ICP-MS), within a sub-ridgeset the beach ridge grows from the updrift to
and were converted into dose rate based on data from Adamiec and downdrift parts. Three types of beach ridges were recognized. Twenty
Aitken (1998) and Marsh et al. (2002). Past moisture content should and twenty-six OSL ages were obtained for recurved and trunk ridges,
have drastically changed seasonally, as suggested by measurement of respectively. An OSL age was determined for the ridge sub-parallel to
samples taken in April and October, which are ends of dry and wet sea- the distributary at the updrift end of Ridgeset 2 (site 2–5).
sons, respectively. According to measured values for samples, moisture OSL ages of ridgesets are consistently older with increasing distance
contents were assumed to be 7% and 20% for unsaturated and saturated from the shoreline, clearly illustrating the age structure of a prograded
T. Tamura et al. / Marine Geology 326–328 (2012) 140–153 145

Table 1
OSL samples of the beach ridges in the Tra Vinh delta plain: site number and location, distance to the present shore, depth, contents of radionuclides, and annual average moisture.

Site no. Lab no. Latitude Longitude Distance to Depth K U Th Rb Moisture


(degree N) (degree E) shore (km) (m) (%) (ppm) (ppm) (ppm) (%)

Ridgeset 1
1-1 Shfd 10196⁎ 9.5464 106.4367 1.3 0.8 1.4 1.23 6 64.9 15
1-2 Shfd 10198 9.5858 106.5139 1.9 1.0 1.3 1.38 6.2 59 14
1-3 Shfd 10197⁎ 9.6188 106.5323 2.5 1.0 1.2 5.88 28.1 59.1 13
1-4 Shfd 10195 9.6288 106.5607 0.4 0.7 1 2.22 11.1 49.2 14
1-5 Shfd 10194 9.6443 106.5666 0.5 0.8 1.1 5.21 23.2 58.2 14
1-6 Shfd 10193 9.6746 106.5791 0.0 0.8 1.1 1.52 6.3 50.2 13
1-7 Shfd 10192⁎ 9.6746 106.5757 0.4 1.4 1 1.67 6.2 48.5 13
Ridgeset 2
2-1 Shfd 10090⁎ 9.5997 106.3864 9.0 1.3 1.1 1.29 6 57.8 18
2-2 Shfd 10141 9.6098 106.4425 8.0 1.2 1.4 1.77 9.3 74.2 20
2-3 Shfd 10140 9.6386 106.5024 6.8 1.9 1.1 1.26 6 56.9 19
2-4 Shfd 10089 9.6580 106.5056 7.6 1.9 1.1 2.14 10.9 58.5 21
2-5 Shfd 10136 9.8099 106.5052 10.6 1.3 1.1 1.27 5.7 55.4 19
Ridgeset 3
3-1 Shfd 10139⁎ 9.6630 106.4194 14.8 1.5 1.5 1.24 6 72.8 22
3-2 Shfd 10161 9.6793 106.4868 10.8 1.7 1.1 1.73 8.5 55.6 17
3-3 Shfd 10138 9.6907 106.4948 10.1 1.7 1.2 1.29 6.3 58.3 13
3-4 Shfd 10087 9.7317 106.5217 8.4 1.1 1.2 1.42 6.4 58.1 13
3-4 Shfd 10088 9.7317 106.5217 8.4 1.9 1.1 1.52 7.2 53.9 20
3-5 Shfd 10142 9.6744 106.4111 16.4 1.2 1.6 1.48 6.6 80.8 14
3-6 Shfd 10091 9.7076 106.4909 11.3 1.8 1.2 1.69 7.8 59.2 15
3-7 Shfd 10144 9.7144 106.4863 12.3 0.9 1.2 1.95 10.4 57.4 13
3-8 Shfd 10137 9.7835 106.5048 10.0 1.9 1.1 1.77 7.7 56.6 12
Ridgeset 4
4-1 Shfd 10143 9.6311 106.3064 14.7 0.9 1.3 1.42 6.5 61.4 20
4-2 Shfd 10145⁎ 9.7505 106.4870 13.5 0.8 1.2 2.82 12 60.2 12
Ridgeset 5
5-1 Shfd 10148⁎ 9.6694 106.3509 19.1 0.9 1.3 1.67 7.8 74.8 14
5-2 Shfd 10146 9.7186 106.4277 18.3 1.2 1.3 1.77 7.9 61.8 12
5-3 Shfd 10094⁎ 9.6767 106.2912 21.8 0.8 1.5 1.48 8.2 72 14
5-4 Shfd 10093 9.7536 106.4174 21.4 1.7 1.2 1.56 6.9 56.3 18
5-5 Shfd 10147 9.7842 106.4506 18.7 0.9 1.4 1.17 5.6 68.5 11
5-6 Shfd 10092⁎ 9.7825 106.4397 20.3 1.7 1.1 1.56 7.1 53.4 17
Ridgeset 6
6-1 Shfd 10153 9.7001 106.2164 27.4 1.9 1.4 1.44 6.6 77.6 22
6-2 Shfd 10151 9.7250 106.3153 27.4 1.6 1.5 1.3 6.4 75.9 22
6-3 Shfd 10149 9.7709 106.3755 26.5 1.3 1.4 1.3 6.4 67.7 13
6-4 Shfd 10150 9.8042 106.4078 24.5 1.3 1.2 1.71 6.4 66.4 22
6-5 Shfd 10152⁎ 9.8462 106.4096 27.3 1.6 1.5 1.22 5.9 75.5 16
Ridgeset 7
7-1 Shfd 10162 9.9223 106.3622 31.0 1.1 1.2 2.59 11.5 63.1 19
Ridgeset 8
8-1 Shfd 10097 9.7272 106.1913 31.4 1.4 1.4 1.76 8.8 75 17
8-2 Shfd 10096 9.7685 106.3228 31.1 1.6 1.4 1.5 6.8 67.9 14
8-3 Shfd 10163 9.8751 106.3483 33.7 1.9 1.3 1.22 5.6 66.1 21
8-4 Shfd 10095⁎ 9.8740 106.3407 34.7 1.4 1.5 2.02 9.7 74.9 18
Ridgeset 9
9-1 Shfd 10155 9.7950 106.1693 42.4 1.3 1.4 1.74 8.6 70.6 21
9-2 Shfd 10154 9.9080 106.3237 37.7 1.5 1.5 1.31 5.9 81.3 14
9-3 Shfd 10211 9.8747 106.2200 47.1 1.2 1.2 2.48 11.9 60.8 13
9-4 Shfd 10160⁎ 9.9179 106.3218 38.0 1.1 1.5 2.02 9.2 85.9 20
Ridgeset 10
10-1 Shfd 10199⁎ 9.9045 106.2099 50.1 0.8 1.3 2.33 9.6 62.1 12
Ridgeset 11
11-1 Shfd 10157⁎ 9.9034 106.0733 59.2 1.5 1.3 1.44 7.2 61.6 13
11-2 Shfd 10156 9.9619 106.1891 55.3 0.8 1.4 2.31 10.1 81.2 16
11-3 Shfd 10099⁎ 10.0753 106.1899 57.6 1.4 1.8 1.34 6.2 89.9 25

OSL age reported by Tamura et al. (2012).

shoreline (Tables 1, 2; Fig. 6). These ages are comparable with recent c. 30 m. Sites 1–2, 1–4, and 1–5 are located landward of the AD 1936
shoreline changes (Fig. 2; Tamura et al., 2010) and with reported radio- shoreline, and their ages are concordantly older than 74 years. Radio-
carbon ages of sediment cores (Fig. 7; Ta et al., 2002; Tanabe et al., carbon ages of molluscan shells in subtidal to intertidal flat facies of
2003). Shoreline positions since AD 1936 (74 years before AD 2010) sediment drill cores TC1 and TV1, from ridgesets 9 and 3, are 2957–
provides an independent age evidence for Ridgeset 1. Sites 1–1 and 2810 cal years BP and 760–678 cal years BP (equivalent to 3017–
1–7 are located between shorelines in AD 1936 and 1981 (29 years), 2870 years and 820–738 years relative to 2010 AD), respectively
and their OSL ages are accordingly within a range of 29–74 years. Site (Figs. 5, 7). Tidal flat facies lie below beach ridges, and thus should be
1–6 is suggested to be younger than AD 1989 (21 years), but it has an older than the beach ridge occurring above or seaward of it. The upper-
age of 33 ± 4 years, and thus is slightly overestimated. This may be most part of the tidal flat facies in core TV1 is older than OSL ages of
due to the uncertainty in the past shoreline position as the Landsat Ridgeset 3 (Table 2), and nearly identical to or slightly older than
image used for showing AD 1989 shoreline has a spatial resolution of Ridgeset 4. OSL ages of Ridgeset 9 are also younger than the age of
146 T. Tamura et al. / Marine Geology 326–328 (2012) 140–153

Table 2
Type of dated beach ridge, dose rate, numbers of aliquots measured and aliquots used for De determination, palaeodose, over dispersion, and OSL age. OSL ages with standard error
exceeding 10 years are rounded to the nearest decade.

Site no. Lab no. Ridge type Dose rate Aliquots measured Aliquots used De OD Age
(mGy/a) (mGy) (%) (years)

Ridgeset 1
1-1 Shfd 10196⁎ Recurved 1.944 ± 0.099 40 27 75 ± 10 0 39 ± 5
1-2 Shfd 10198 Recurved 1.916 ± 0.095 20 18 299 ± 7 6 156 ± 9
1-3 Shfd 10197⁎ Trunk 4.150 ± 0.196 20 20 740 ± 9 3 178 ± 9
1-4 Shfd 10195 Recurved 2.155 ± 0.098 20 17 339 ± 9 20 157 ± 8
1-5 Shfd 10194 Recurved 3.616 ± 0.169 20 18 522 ± 8 3 144 ± 7
1-6 Shfd 10193 Trunk 1.813 ± 0.087 40 34 60 ± 7 23 33 ± 4
1-7 Shfd 10192⁎ Trunk 1.745 ± 0.082 32 28 120 ± 7 25 69 ± 5
Ridgeset 2
2-1 Shfd 10090⁎ Recurved 1.630 ± 0.103 20 20 477 ± 9 3 290 ± 20
2-2 Shfd 10141 Recurved 2.115 ± 0.102 20 18 790 ± 9 12 370 ± 20
2-3 Shfd 10140 Trunk 1.594 ± 0.078 20 16 505 ± 8 0 320 ± 20
2-4 Shfd 10089 Recurved 1.996 ± 0.107 20 19 792 ± 13 9 400 ± 20
2-5 Shfd 10136 Along channel 1.594 ± 0.077 20 18 505 ± 8 6 320 ± 20
Ridgeset 3
3-1 Shfd 10139⁎ Recurved 1.844 ± 0.096 20 18 962 ± 12 7 520 ± 30
3-2 Shfd 10161 Trunk 1.880 ± 0.089 20 20 1065 ± 14 5 570 ± 30
3-3 Shfd 10138 Recurved 1.826 ± 0.091 20 18 1110 ± 11 4 610 ± 30
3-4 Shfd 10087 Trunk 1.889 ± 0.093 20 17 409 ± 9 6 220 ± 10
3-4 Shfd 10088 Trunk 1.689 ± 0.081 20 17 340 ± 11 48 200 ± 10
3-5 Shfd 10142 Recurved 2.197 ± 0.114 20 15 1366 ± 11 11 620 ± 30
3-6 Shfd 10091 Trunk 1.951 ± 0.112 20 19 1476 ± 15 7 760 ± 40
3-7 Shfd 10144 Recurved 2.231 ± 0.105 20 20 1672 ± 13 3 750 ± 40
3-8 Shfd 10137 Trunk 1.953 ± 0.093 20 18 1153 ± 11 6 590 ± 30
Ridgeset 4
4-1 Shfd 10143 Recurved 1.813 ± 0.089 20 18 1526 ± 14 7 840 ± 40
4-2 Shfd 10145⁎ Trunk 2.563 ± 0.119 20 17 1960 ± 18 6 770 ± 40
Ridgeset 5
5-1 Shfd 10148⁎ Recurved 2.089 ± 0.101 20 18 2019 ± 22 7 970 ± 50
5-2 Shfd 10146 Trunk 2.146 ± 0.104 20 17 2141 ± 20 6 1000 ± 50
5-3 Shfd 10094⁎ Recurved 2.245 ± 0.112 20 17 2491 ± 20 5 1110 ± 60
5-4 Shfd 10093 Recurved 1.810 ± 0.100 20 19 2525 ± 22 4 1400 ± 80
5-5 Shfd 10147 Trunk 1.985 ± 0.102 20 19 1877 ± 23 15 950 ± 50
5-6 Shfd 10092⁎ Recurved 1.759 ± 0.101 20 15 2796 ± 20 4 1590 ± 90
Ridgeset 6
6-1 Shfd 10153 Recurved 1.818 ± 0.092 20 18 2969 ± 34 19 1630 ± 80
6-2 Shfd 10151 Recurved 1.866 ± 0.096 20 16 3393 ± 39 7 1820 ± 100
6-3 Shfd 10149 Trunk 2.024 ± 0.103 20 18 3628 ± 21 3 1790 ± 90
6-4 Shfd 10150 Trunk 1.729 ± 0.083 20 19 3039 ± 30 5 1760 ± 90
6-5 Shfd 10152⁎ Trunk 1.972 ± 0.103 20 18 3187 ± 32 5 1620 ± 90
Ridgeset 7
7-1 Shfd 10162 Trunk 2.255 ± 0.105 20 17 4611 ± 45 6 2050 ± 100
Ridgeset 8
8-1 Shfd 10097 Recurved 2.138 ± 0.104 20 16 4338 ± 26 5 2030 ± 100
8-2 Shfd 10096 Trunk 2.063 ± 0.104 20 17 3857 ± 36 5 1870 ± 100
8-3 Shfd 10163 Trunk 1.683 ± 0.085 20 15 3847 ± 39 6 2290 ± 120
8-4 Shfd 10095⁎ Trunk 2.295 ± 0.112 20 19 4776 ± 34 8 2080 ± 100
Ridgeset 9
9-1 Shfd 10155 Trunk 2.026 ± 0.099 20 18 4732 ± 67 14 2340 ± 120
9-2 Shfd 10154 Trunk 2.043 ± 0.106 20 16 4304 ± 55 5 2110 ± 110
9-3 Shfd 10211 Recurved 2.451 ± 0.114 20 19 4861 ± 72 7 1980 ± 100
9-4 Shfd 10160⁎ Trunk 2.208 ± 0.108 20 17 5121 ± 50 6 2320 ± 120
Ridgeset 10
10-1 Shfd 10199⁎ Trunk 2.381 ± 0.113 20 18 6627 ± 86 7 2780 ± 140
Ridgeset 11
11-1 Shfd 10157⁎ Trunk 2.004 ± 0.099 20 20 6670 ± 74 5 3330 ± 170
11-2 Shfd 10156 Trunk 2.377 ± 0.113 20 19 6740 ± 77 5 2840 ± 140
11-3 Shfd 10099⁎ Trunk 2.016 ± 0.107 20 18 7193 ± 66 6 3570 ± 190

OSL age reported by Tamura et al. (2012).

tidal flat in core TC1. These consistencies with other age evidence indi- those of adjoining ridges except for ages from the site 3–4 (Fig. 6).
cate the OSL age estimates are robust overall. Isochronous lines in a OSL ages highly overlap in ridgesets 7–9, which exhibit large uncer-
shore-normal cross section (Fig. 7) define the progradation of the delta- tainties relative to OSL ages in younger ridgesets. OSL age uncertainties
ic clinoform that extends from the beach ridge to deeper deltaic sedi- increase in proportion to the absolute age. There is in contrast a more
ment facies (Tamura et al., 2012). The isochronous lines suggest that than 1000 years difference between ridgesets 9 and 11. The only
the progradation of the beach shoreline and delta front initiated sample from Ridgeset 10 has an age slightly younger than those of
c. 3500 years, and that the clinoform steepened towards the present Ridgeset 11. The range of the ridgeset age varies, ranging from
coast (Fig. 7). 110 years (Ridgeset 2) to 600 years (Ridgeset 5). Ridgeset 5 has ages
There are variations in the range of OSL ages of an individual of 950–1590 years, and apart from this, ridgesets 1–4 and 6–8 show
ridgeset. Ridgesets 1–6 do not have any OSL ages overlapping with average recurrence interval of c. 200–250 years as 4 and 3 ridgesets
T. Tamura et al. / Marine Geology 326–328 (2012) 140–153 147

Fig. 5. Definition of the hierarchy of beach ridges on the Tra Vinh delta plain, and locations of sampling sites for OSL dating. Even-numbered ridgesets are shaded to enable distinc-
tion of adjacent ridgesets.

formed in 0–1000 and 1600–2300 years, respectively. In contrast, The internal age structure of a ridgeset is systematic when consid-
ridgesets 9–11 formed in 1600 years between 3570 and 1980 years, ering only recurved ridges. Ages of relatively younger recurved
and their average recurrence interval is 500–600 years. ridges, from sites 6–2 to 1–1, consistently agree with the inferred
order of deposition (Fig. 6). Ages of trunk ridges in contrast are
more scattered. The updrift trunk generally exhibits much younger
ages than the rest of ridgeset (e.g., 1–6, 1–7, 3–4, 3–8, 5–5, 6–5).
Table 3 Ridgesets 7–11 have few recurved ridge ages, and ages of their
Area of the ridgeset and its landward inter-ridge swale, cumulative delta plain area
trunk ridges are highly scattered and apparently not systematic.
after Ridgeset 11, and representative OSL age of the past shoreline marked by each
ridgeset.
5. Discussion
Ridgeset Area Cumulative delta Shoreline age
(km2) plain area (km2) (year before
AD 2010) 5.1. Shoreline change characterization
1 238.3 1815.7 –
2 137.7 1577.4 200 ± 10 Integration of the architecture and OSL ages of beach ridges suggests
3 106.5 1439.7 520 ± 30 the hierarchical shoreline changes of the Tra Vinh coast during the last
4 94.2 1333.2 770 ± 40 3500 years (Fig. 8). Ridgesets generally indicate cyclic discontinuous
5 245.2 1239 950 ± 50
shifts of shoreline 5 km seawards on average. Each of the cyclic shifts
6 77.9 993.8 1620 ± 90
7 62.5 915.9 – is followed by the downdrift (southwestward) elongation and accretion
8 280.7 853.4 1870 ± 100 of the recurved spit or island, resulting in trunk and recurved beach
9 330.2 572.7 2110 ± 110 ridges that comprise an individual ridgeset. The downdrift accretion of
10 242.5 242.5 – the barrier is interrupted with changes in shoreline orientation that
11 – 0 2840 ± 140
are suggested by the bifurcations of trunk ridges. These hierarchical
148 T. Tamura et al. / Marine Geology 326–328 (2012) 140–153

4000
OSL age (yr before AD 2010)

3000

2000

1000

40 30 20 10 1
Inferred order of deposition
Fig. 6. Plot of OSL age obtained from the Tra Vinh beach ridge against inferred order of deposition. The deposition order is determined assuming the seaward sub-ridgeset is youn-
ger, and that the individual sub-ridgeset is younger downdrift. Two different ranges of OSL ages for an individual ridgeset are indicated with or without considering 1-σ age
uncertainties.

shoreline changes have intrinsic time scales: the discontinuous shift in the modern shoreline (Figs. 2, 8). In ridgesets 1–6, recurved ridges
occurred per several centuries; several changes in shoreline orientation have ages systematically younger both downdrift and seawards,
occur between the discontinuous shift, indicating their average interval while ages of trunk ridges are scattered. The modern updrift coast
is several decades to a few centuries; and the downdrift spit/island ac- modestly prograded seawards to result in ridges sub-parallel to the
cretion occurs on annual- to decadal-scale. shoreline (Fig. 2A). This indicates the updrift trunk ridge could be
The changes to the Tra Vinh coast over the last decades are re- younger than downdrift recurved ridges. It is also suggested that
ferred for interpreting the internal age structure of the ridgesets as the updrift trunk ridge is likely reworked as seen in the recent coastal
each ridgeset shows a downdrift accretion, similarly to that observed erosion (Fig. 2A). Relatively young trunk ridges are present at sites

+5

-10
Elevation (m)

-20

-30

-40

-50

Fig. 7. Cross section along a shore-perpendicular transect of the Tra Vinh delta plain (Tamura et al., 2012; modified from Ta et al., 2005). Transect location is shown in Figs. 1B, 5. OSL
and calibrated radiocarbon ages are expressed relative to AD 2010, not to AD 1950.
T. Tamura et al. / Marine Geology 326–328 (2012) 140–153 149

A B C

Fig. 8. Schematic illustration showing the discontinuous shift and change in shoreline orientation inferred from hierarchical beach ridge structure in the Tra Vinh delta plain.

6–5, 5–5, 4–2, 3–8, 3–4, 1–6, and 1–7. Ages of these trunk ridges are form the core of the next discontinuous shoreline shift (Fig. 8). Such
generally close to the initiation of seaward ridgeset or sub-ridgeset. an island formed on the delta front is among those defined as the deltaic
The trunk ridge at site 3–4 is continuous to Ridgeset 2, and its age is barrier island (Stutz and Pilkey, 2002), of which formation is common
intermediate between ages of ridgesets 1 and 2. This suggests that in a variety of wave-influenced deltas (e.g., Rodriguez et al., 2000;
the extensive shoreline reworking occurred at site 3–4 in relation to Bhattacharya and Giosan, 2003; van Maren, 2005). Stutz and Pilkey
the formation of Ridgeset 2 and/or just before the shoreline shift (2002) analyzed the shapes and dimensions of deltaic barrier islands
from ridgesets 2 to 1. A similarly young age is present in site 5–5, in the world, and classified them into four prograding types and a trans-
where the trunk age is c. 600 years younger than the neighbouring gressive type, based on the length of island and width of inlet. Each of
recurved ridge at site 5–6, being as young as the termination of the prograding types is characterized by a degree of relative contribu-
Ridgeset 5. Recurved ridges are more likely protected from the shore- tions of wave, tide, and fluvial processes. The coast of the Mekong
line erosion due to their shape and position, and thus show the sys- River delta is characterized by a low wave influence, large tide range,
tematic age structure. The young trunk ridge in contrast tends to and gentle slope of shoreface (Stutz and Pilkey, 2002). The term deltaic
reflect the younger limit of the ridgeset age. Causes of age structure barrier island as defined by Stutz and Pilkey (2002), however, also in-
irregularities in ridgesets 7–11 are uncertain, and may include larger cludes the barrier shorelines even attached to the mainland. Here, for
uncertainties in age estimation. the clarification, an island that is developed on the delta front and iso-
No abrupt change in shoreline orientation as inferred from the lated from the land is referred to as a delta-front island, and the island
bifurcations of trunk ridges have been observed. Ridgeset 1 exhibits after attached to the mainland is called a spit. Although the formative
two series of recurved ridges that comprise downdrift portions of process of the delta-front island in the Mekong delta cannot be
sub-ridgesets 1a and 1b, respectively. According to the shoreline discussed effectively due to the lack of adequate data, such as detailed
position in AD 1936 and OSL ages from sites 1–2, 1–4, and 1–5 (Fig. 2A, bathymetry of the delta front and wave conditions, processes from
B), a change in shoreline orientation occurred sometime between 70 other deltas have been documented. In the study of the Red River
and 150 years to separate these sub-ridgesets. Ridgeset 2, although delta, based on the process-based numerical modelling (Delft-3D) and
defined as a ridgeset, may just represent a relatively large-scale change historical observation, van Maren (2005) showed that the delta-front
in shoreline orientation. The change in shoreline orientation from island evolves from a bar on the delta front, of which sediment is initial-
ridgesets 3 to 2 is larger than others, coupled with the prominent ly transported offshore in the river plume and later reworked onshore
shape of the Ridgeset 3 shoreline, resulting in a long distance of shoreline by waves. As the bar evolves to become a subaerially-exposed island,
migration in the downdrift part. This change in orientation was accom- the longshore transport becomes more important, enhancing the elon-
panied by the reworking of the updrift coast as revealed by the young gation of the island (van Maren, 2005). The island then is attached to
OSL age at site 3–4. A similar extensive reworking of the updrift shoreline the mainland, defining a new shoreline. Similar processes were ob-
occurred also in relation to the orientation change revealed by the served in other wave-influenced deltas (e.g., Rodriguez et al., 2000;
transition from sub-ridgesets 5c to 5b. Giosan et al., 2005).
The actual duration of the ridgeset formation is not constrained Factors controlling the interval of the discontinuous shoreline
strictly, as it is not guaranteed that the youngest and oldest parts of shift have not been understood well although they are critical for
the individual ridgeset were exactly sampled. The range of OSL ages predicting the behavior of the actively prograding coast. Intervals
obtained, however, probably estimates an approximate duration of of the discontinuous shoreline shift are variable, ranging from a
the ridgeset, as ridgesets 1–6 show neither a remarkable overlap few decades in the Brazos River delta (Rodriguez et al., 2000),
nor gap in OSL age between neighbouring ridgesets. The ages at site through c. 100 years in the Red River delta (van Maren, 2005), to
3–4 seem to be an exception, but the deposition at site 3–4 is related several centuries on the Tra Vinh coast (Fig. 6). The interval in prin-
to formation of Ridgeset 2, and not to the building stage of Ridgeset 3. ciple is equivalent to the distance between the delta-front island and
Thus, the formative durations of ridgesets 2 and 3 are estimated shoreline divided by the rate of net coastal progradation. The
200–400 and 520–760 years, respectively. progradation rate is affected by the rate of sediment discharge, and
by the length of the shoreline to which the sediment is distributed.
5.2. Causes of shoreline changes The distance between the delta-front island and shoreline is roughly
equivalent to the width of the delta-front platform that reflects dis-
5.2.1. Shoreline progradation charge of muddy sediment (Stutz and Pilkey, 2002). Thus, the inter-
Cyclic discontinuous shoreline shifts c. 5 km seawards are the prin- val of the delta-front island is strongly correlated with the quantity
cipal mode in which the Tra Vinh coast appears to prograde. An and quality of sediment discharge onto the coast. In the Tra Vinh
elongate bar is present on the modern delta-front platform (Fig. 2A), delta plain, however, the width of delta-front platform appears to
and has the potential to grow into a sub-aerially exposed island and have changed (Fig. 7; Tamura et al., 2012), which may be the result
then, once attached to the land, into a spit. Such a spit could then of the millennial-scale evolution. The initial delta progradation
150 T. Tamura et al. / Marine Geology 326–328 (2012) 140–153

appears to have been a transitional state from the low-gradient


aggradational delta before 3500 years, after which the clinoform
has steepened as the prograding delta became more pronounced.
The relatively gentle clinoform may be reflected by the longer recur-
rence interval and longer distance of the discontinuous shoreline
shift revealed by ridgesets 8–11.
In addition to the interval of the discontinuous shoreline shift,
quantifying the rate of progradation of the Tra Vinh delta plain is im-
portant as it is directly related to the sediment supply. However, the
quantification is not straightforward because of the complicated pat-
tern of shoreline changes. As the shoreline accretes downdrift as well
as seawards, it is uncertain as to whether the progradation rate along
a selected single transect appropriately represents the general growth
trend of the delta plain. The area defined by the ridgeset and its land-
ward inter-ridge swale, and the representative age of a ridgeset are
thus used to assess the growth rate of the delta plain. The representa-
tive age of the ridgeset is assumed here to be the youngest OSL age
among those determined, which is typically found in the updrift
part (Table 3). As the width of the Tra Vinh delta plain is almost con-
sistent, the average progradation rate is proportional to the rate of the
area increment. There are two phases of the rate of the Tra Vinh delta
plain progradation (Fig. 9). Before 1500 years, the delta plain
prograded linearly at a higher rate. After 1500 years, the rate of
progradation abruptly decreased and was consistent at least until
the formation of Ridgeset 2. While the ages of the inland ridgesets Fig. 10. Close-up of line drawing of the beach ridges (Fig. 5) with OSL ages showing al-
are assumed to be the last timing of their reworking, Ridgeset 1 is ternations of two different phases of the shoreline progradation. Anti-clockwise shifts
still active, and its abandonment should be later than AD 2010 in of shoreline orientation occurred at c. 1100–1000 and 500–400 years following phases
of updrift-centered progradation. OSL ages are expressed relative to AD 2010.
Fig. 9. North of the Tra Vinh delta plain, there are three delta plains
in Ben Tre and one in Tien Giang, which are bounded by distributaries
and share the sediment supply from the Mekong River with the Tra shoreline, which results in a more uniformly oriented shoreline, as
Vinh coast. Although it is not possible to quantify the progradation seen in the transition from ridgesets 3 to 2 (Fig. 10). The downdrift elon-
of these coasts in detail, the majority of these other delta plains gation of the spit from the prominent delta lobe associated with the
have formed since 1500 years (Fig. 1B; Tamura et al., 2012). The updrift shoreline retreat typically occurs in wave-influenced deltas of
focal shift of the river sediment supply towards the northeast thus ac- which river sediment discharge has decreased, including the Ebro
counts for the declined progradation of the Tra Vinh delta plain (Jiménez et al., 1997), Po (Simeoni et al., 2007), and Volga deltas
(Fig. 9), and may have been controlled by morphological changes at (Anthony and Blivi, 1999). In these deltas, spit elongation has occurred
the bifurcation of the Mekong River located c. 100 km inland from under conditions of sediment scarcity, with sediment redistributed
the present river mouth (Fig. 1B). from the delta lobe to feed downdrift spit growth. On the Tra Vinh
coast in contrast, the morphological prominence revealed by ridgesets
3 and 4 suggests that they evolved from a delta-front barrier island,
5.2.2. Changes in shoreline orientation and is different from a typical delta lobe at the river mouth. However,
The abrupt changes in shoreline orientation revealed by trunk ridge even in this case the prominence is still maintained by the river
bifurcations lead to a seaward shift in the downdrift portion of the sediment discharge (van Maren, 2005). The changes in shoreline
orientation of the Tra Vin coast thus are interpreted as reflecting fluctu-
2000
ations in the balance between river sediment discharge and longshore
transport. The trunk ridge bifurcation indicates only the increase in
relative importance of longshore transport. Relative decline of the
Cumulative delta plain area (km2)

longshore transport in contrast may be reflected by the accretion of


1500 the updrift sub-parallel ridge as seen in Sub-ridgeset 1b, or by the
discontinuous shifts revealed by ridgesets. During the period
200–1600 years, while ridgesets 2–5 formed, two major downdrift
elongations of the spit occurred at transitions from sub-ridgesets 5c to
1000 5b, and from ridgesets 3 to 2, and their timings are constrained
1000–1100 years by ages from sites 5–2 and 5–3, and 400–500 years
by ages from sites 3–1 and 2–4 (Fig. 10). Although the terminations of
these spit elongations are not well constrained, their growth appears
500 to have become terminated by the initiation of younger ridgesets.
Present shoreline
Their terminations are thus older than 840 and 178 years, respectively,
Past shoreline
according to ages from sites 4–1 and 1–3.
Monsoons critically affect the fluctuations in balance between the
0 river discharge and longshore transport in the Mekong River delta. The
sediment and water discharge of the Mekong River are from the precip-
3000 2500 2000 1500 1000 500 0 -500
itation in its catchment, and are thus controlled by the south-westerly
Age (yr before AD 2010)
summer monsoon. The south-westward longshore sediment transport
Fig. 9. Temporal changes of the Tra Vinh delta plain area revealed by ridgesets and is enhanced by the north-easterly winter monsoon. The Asian summer
their representative ages as summarized in Table 3. monsoon is known to have fluctuated during the Late Quaternary from
T. Tamura et al. / Marine Geology 326–328 (2012) 140–153 151

a range of geological proxies (e.g., Wang et al., 2005; Colin et al., 2010; shoreline is commonly caused by a range of reasons, including
Zhang et al., 2011), but its centennial- to decadal-scale fluctuations storm, changes in wave condition, relative sea-level changes, and
affecting regional precipitation of Southeast Asia are poorly understood. bathymetry changes due to the formation of delta-front bar/island
The strength of the winter monsoon in contrast may possibly be evalu- (e.g., Giosan et al., 2005; Simeoni et al., 2007). The inferred
ated based on fluctuations detected in East Asia. Chinese historical doc- decrease in sediment supply thus is not regarded as an exclusive
uments suggest that the winter monsoon in East Asia has fluctuated at cause of the present coastal erosion (Fig. 2A, B). At decadal- to
decadal- and centennial-scales (Wang and Zhang, 1992). An increase centennial-scales, the discontinuous shoreline shift is rather an
in the winter monsoon strength 400–500 years, corresponding to the important issue for the coastal environment, because if it occurs it
beginning of the Little Ice Age (Lamb, 1995), is suggested by frequent will protect the present shoreline from open sea waves and further
aeolian dust fall (Zhang, 1984) and lower winter temperature (Ge et erosion. The decrease in sediment discharge potentially prevents
al., 2003) in China, and activation of coastal dunes along the Japan Sea the formation of the delta-front bar (Giosan et al., 2005), and thus
(Tamura et al., 2011). Therefore, it is probable that the downdrift elon- further discontinuous shoreline shift. It is thus important to under-
gation of the Tra Vinh coast beach ridges was enhanced by the strength- stand the mechanism of the delta-front bar/island, which primarily
ened winter monsoon. The decrease in the river sediment discharge controls the shoreline change of the Mekong River delta coast.
might have been associated with this, although it is uncertain how the
Asian winter and summer monsoons are related. This hypothesis indi-
cates the possibility that the morphology of the Mekong River delta 6. Conclusions
shoreline is affected by regional to global climate changes at decadal-
to centennial-scales, but it should be tested based on further examina- Luminescence dating of beach ridges in the Tra Vinh delta plain, cen-
tions of other proxies and numerical modelling of shoreline geometry. tral Mekong River delta successfully characterized the centennial- to
multi-decadal scale shoreline changes over the last 3500 years. The
5.3. Implications for the modern shoreline beach ridge sediments have excellent luminescence properties, and
OSL ages agree well with independent age evidence. The Tra Vinh
The characterization of the past shoreline changes in time scales shoreline has exhibited hierarchical changes that are suggested by the
comparable to the historical observation is critical for understanding presence of ridgesets. Each ridgeset corresponds to a discontinuous
the modern and future shorelines, with allowance being made for the shoreline shift c. 5 km seawards caused by the formation of a delta-
effects of recent industrial activity and extreme climate change that front bar/island, of which the interval ranges from 200 to 600 years.
do not concern the past. The most remarkable change that potentially The discontinuous shift is followed by the downdrift spit accretion driv-
occurs to the present coast is the discontinuous shift of the shoreline en by the dominant winter monsoon. Due to relatively small un-
caused by the delta-front bar/island (Fig. 2A). Assuming the consistent certainties of OSL ages, within an individual beach ridge set, it is
interval of the ridgeset formation at 200–250 years, the next discontin- possible to discern downdrift ridges to be younger, which represents
uous shift is suggested to occur in the next few decades, as according to centennial to decadal-scale shoreline shifts. Spit accretion has occurred
the OSL age at site 1–3, at least 178 ±9 years have passed since the in conjunction with changes of shoreline orientation, reflecting the
Ridgeset 1 shoreline emerged. Extrapolating the consistent rate of balance between river sediment discharge and longshore drift, which
delta plain progradation (Fig. 9) suggests, in contrast, that the present was possibly controlled by the intensity of Asian monsoons. A charac-
shoreline will be abandoned 200–300 years later. Along the present teristic change in shoreline orientation occurred 400–500 years, the
shoreline (Fig. 2B), the downdrift spit accretion has been inactive beginning of the Little Ice Age, suggesting the strengthened winter
since AD 2001. This may suggest either that the sediment supply has de- monsoon and/or declined river sediment discharge. The progradation
creased or that more sand is being transported offshore to build the rate of the delta plain was roughly constant, but abruptly decreased at
delta-front bar. 1500 years. This suggests that the focal shift of the river sediment sup-
The Mekong River delta is expected to experience a decrease in ply to the northeastern part of the Mekong River delta occurred at
the bedload supply due to recent human activities that have not 1500 years, possibly caused by some morphological changes at distrib-
been quantitatively assessed, which is an obstacle to predict the utary bifurcations, as the majority of other delta plains have been
shoreline change from the long-term perspective. Several studies formed after 1500 years. The changes of the Mekong River delta shore-
have examined changes of water discharge and suspended sediment line are thus affected by external climate changes and internal fluctua-
load of the Mekong River over the recent decades (Lu and Siew, tions in sediment distribution. Over the last 1000 years, the Tra Vinh
2006; Kummu and Varis, 2007; Walling, 2008). These studies con- shoreline has regularly shifted in terms of progradation rate and inter-
cluded there has been no remarkable change of discharge or val of discontinuous shoreline shift. Although ongoing human activities
sediment load in the lower reaches of the Mekong River since do not allow for simple extrapolations of past regular conditions in the
AD1960, despite the commissioning of the first major dam at the prediction of future shoreline change, the reconstructed shoreline
Lancang Cascade in the upper Mekong River in AD1993. However, changes provide the baseline for assessing future coastal behavior with-
there are no reliable data available to assess changes in bedload out the influence of human activities.
(Kummu and Varis, 2007), which critically determines the coastal Supplementary data to this article can be found online at http://
sediment budget, and is the parameter more likely affected by the dx.doi.org/10.1016/j.margeo.2012.08.004.
newly constructed dam. In addition to the dam construction, in the
lower reach of the Mekong River, many ships actively operate in
fluvial sand dredging. Although the dredging is also not quantitative- Acknowledgments
ly assessed, it directly decreases the sand supply to the coast. A re-
cent decrease in sand supply is thus inferred, and may have We express thanks to Adam Dunajko and Rob Ashurst for their
enhanced the erosion of the updrift Tra Vinh shoreline that was rec- help in processing luminescence samples. This work was carried out
ognized recently (Fig. 2). Breaks in inland beach ridges (Fig. 4B) while T.T. worked in the University of Sheffield with the aid of JSPS
however, suggest that updrift erosion occurs even if the sediment Postdoctoral Fellowships for Research Abroad, and partly supported
supply is not significantly altered by human activities. Even the re- by NAFOSTED for projects 105.09-2010.05 and 105.09-2010.02, and
cent erosion started between AD1936 and 1981, before the “Doi Megadelta Project of the Ministry of Environment of Japan. The man-
Moi”, the economic reform of Vietnam, initiated in AD1986 and the uscript was significantly improved by Edward Anthony, two anony-
first Chinese dam got commissioned. Updrift erosion of a deltaic mous reviewers, and journal editor John Wells.
152 T. Tamura et al. / Marine Geology 326–328 (2012) 140–153

References Murray-Wallace, C.V., Banerjee, D., Bourman, R.P., Olley, J.M., Brooke, B.P., 2002. Optically
stimulated luminescence dating of Holocene relict foredunes, Guichen Bay, South
Australia. Quaternary Science Reviews 21, 1077–1086.
Adamiec, G., Aitken, M., 1998. Dose-rate conversion factors: update. Ancient TL 16, 37–50. National Astronomical Observatory, 1999. Chronological Scientific Tables (Rika-nenpyo):
Allison, M.A., Khan, S.R., Goodbred Jr., S.L., Kuehl, S.A., 2003. Stratigraphic evolution of Maruzen, Tokyo . 1058 pp. (in Japanese).
the late Holocene Ganges–Brahmaputra lower delta plain. Sedimentary Geology Nguyen, V.L., Ta, T.K.O., Tateishi, M., 2000. Late Holocene depositional environments
155, 317–342. and coastal evolution of the Mekong River Delta, southern Vietnam. Journal of
Anthony, E.J., Blivi, A.B., 1999. Morphosedimentary evolution of a delta-sourced, drift- Asian Earth Sciences 18, 427–439.
aligned sand barrier–lagoon complex, western Bight of Benin. Marine Geology 158, Nielsen, A., Murray, A.S., Pejrup, M., Elberling, B., 2006. Optically stimulated lumines-
161–176. cence dating of a Holocene beach ridge plain in Northern Jutland, Denmark.
Argyilan, E.P., Forman, S.L., Johnston, J.W., Wilcox, D.A., 2005. Optically stimulated lu- Quaternary Geochronology 1, 305–312.
minescence dating of late Holocene raised strandplain sequences adjacent to Nott, J., Smithers, S., Walsh, K., Rhodes, E., 2009. Sand beach ridges record 6000 year
Lakes Michigan and Superior, Upper Peninsula, Michigan, USA. Quaternary Re- history of extreme tropical cyclone activity in northeastern Australia. Quaternary
search 63, 122–135. Science Reviews 28, 1511–1520.
Ballarini, M., Wallinga, J., Murray, A.S., van Heteren, S., Oost, A.P., Bos, A.J.J., van Eijk, Otvos, E.G., 2000. Beach Ridges — definitions and significance. Geomorphology 32, 83–108.
C.W.E., 2003. Optical dating of young coastal dunes on a decadal time scale. Qua- Prescott, J.R., Hutton, J.T., 1994. Cosmic ray contribution to dose rates for luminescence and
ternary Science Reviews 22, 1011–1017. ESR dating: large depths and long-term time variations. Radiation Measurements 23,
Bateman, M.D., Catt, J.A., 1996. An absolute chronology for the raised beach deposits at 497–500.
Sewerby, E. Yorkshire, UK. Journal of Quaternary Science 11, 389–395. Proske, U., Hanebuth, T.J.J., Behling, H., Nguyen, V.L., Ta, T.K.O., Diem, B.P., 2010. The
Bhattacharya, J.P., Giosan, L., 2003. Wave-influenced deltas: geomorphological implica- palaeoenvironmental development of the northeastern Vietnamese Mekong
tions for facies reconstruction. Sedimentology 50, 187–210. River Delta since the mid Holocene. The Holocene 20, 1257–1268.
Bøtter-Jensen, L., Bulur, E., Murray, A.S., Poolton, N.R.J., 2002. Enhancements in lumi- Proske, U., Hanebuth, T.J.J., Gröger, J., Diem, B.P., 2011. Late Holocene sedimentary and
nescence measurement techniques. Radiation Protection Dosimetry 101, 119–124. environmental development of the northern Mekong River Delta, Vietnam. Qua-
Boulter, C., Bateman, M.D., Frederick, C.D., 2010. Understanding geomorphic responses ternary International 230, 57–66.
to environmental change: a 19 000-year case study from semi-arid central Texas, Reimann, T., Tsukamoto, S., Harff, J., Osadczuk, K., Frechen, M., 2011. Reconstruction
USA. Journal of Quaternary Science 25, 889–902. of Holocene coastal foredune progradation using luminescence dating — an ex-
Brooke, B., Lee, R., Cox, M., Olley, J., Pietsch, T., 2008a. Rates of Shoreline Progradation ample from the Świna barrier (southern Baltic Sea, NW Poland). Geomorphol-
during the Last 1700 Years at Beachmere, Southeastern Queensland, Australia, ogy 132, 1–16.
Based on Optically Stimulated Luminescence Dating of Beach Ridges. Journal of Rink, W.J., López, G.I., 2010. OSL-based lateral progradation and aeolian sediment accu-
Coastal Research 24, 640–648. mulation rates for the Apalachicola Island Complex, North Gulf of Mexico, Florida.
Brooke, B., Ryan, D., Pietsch, T., Olley, J., Douglas, G., Packett, R., Radke, L., Flood, P., Geomorphology 123, 330–342.
2008b. Influence of climate fluctuations and changes in catchment land use on Roberts, H.M., Plater, A.J., 2007. Reconstruction of Holocene foreland progradation using
Late Holocene and modern beach-ridge sedimentation on a tropical macrotidal optically stimulated luminescence (OSL) dating: an example from Dungeness, UK.
coast: Keppel Bay, Queensland, Australia. Marine Geology 251, 195–208. The Holocene 17, 495–505.
Colin, C., Siani, G., Sicre, M.-A., Liu, Z., 2010. Impact of the East Asian monsoon rainfall Rodriguez, A.B., Hamilton, M.D., Anderson, J.B., 2000. Facies and evolution of the mod-
changes on the erosion of the Mekong River basin over the past 25,000 yr. Marine ern Brazos Delta, Texas; wave versus flood influence. Journal of Sedimentary Re-
Geology 271, 84–92. search 70, 283–295.
Forsyth, A.J., Nott, J., Bateman, M.D., 2010. Beach ridge plain evidence of a variable late- Simeoni, U., Fontolan, G., Tessari, U., Corbau, C., 2007. Domains of spit evolution in the
Holocene tropical cyclone climate, North Queensland, Australia. Palaeogeography, Goro area, Po Delta, Italy. Geomorphology 86, 332–348.
Palaeoclimatology, Palaeoecology 297, 707–716. Stokes, S., 1992. Optical dating of young (modern) sediments using quartz: results from a
Gagliano, S.M., McIntire, M.G., 1968. Report on the Mekong Delta. Technical Report, 57. selection of depositional environments. Quaternary Science Reviews 11, 153–159.
Coastal Studies Institute, Louisiana University. 143 pp. Storms, J.E.A., Hoogendoorn, R.M., Dam, R.A.C., Hoitink, A.J.F., Kroonenberg, S.B., 2005.
Galbraith, R.F., Roberts, R.G., Laslett, G.M., Yoshida, H., Olley, J.M., 1999. Optical dating Late-Holocene evolution of the Mahakam delta, East Kalimantan, Indonesia. Sedi-
of single and multiple grains of quartz from Jinmium rock shelter, northern mentary Geology 180, 149–166.
Australia: Part I, experimental design and statistical models. Archaeometry 41, Stutz, M., Pilkey, O.H., 2002. Global distribution of deltaic barrier island systems.
339–364. Journal of Coastal Research SI36, 694–707.
Galloway, W.E., 1975. Process framework for describing the morphologic and strati- Ta, T.K.O., Nguyen, V.L., Tateishi, M., Kobayashi, I., Saito, Y., 2002. Holocene delta evolu-
graphic evolution of deltaic depositional systems. In: Broussard, M.L. (Ed.), Deltas, tion and sediment discharge of the Mekong River, southern Vietnam. Quaternary
Models for Exploration: Houston, Texas, Houston Geological Society, pp. 87–98. Science Reviews 21, 1807–1819.
Ge, Q., Zheng, J., Fang, X., Man, Z., Zhang, X., Zhang, P., Wang, W., 2003. Winter half-year Ta, T.K.O., Nguyen, V.L., Tateishi, M., Kobayashi, I., Saito, Y., 2005. Holocene delta evolu-
temperature reconstruction for the middle and lower reaches of the Yellow River tion and depositional models of the Mekong River delta, southern Vietnam. In:
and Yangtze River, China, during the past 2000 years. The Holocene 13, 933–940. Giosan, L., Bhattacharya, J.P. (Eds.), River Deltas – Concepts, Models, and Examples:
Giosan, L., Donnelly, J.P., Vespremeanu, E., Bhattacharya, J.P., Olariu, C., Buonaiuto, F.S., SEPM Special Publication, 83, pp. 453–466.
2005. River delta morphodynamics: examples from the Danube delta. In: Giosan, Tamura, T., 2012. Beach ridges and prograded beach deposits as palaeoenvironment re-
L., Bhattacharya, J.P. (Eds.), River Deltas — Concepts, Models, and Examples: cords. Earth-Science Reviews 114, 279–297.
SEPM Special Publication, no. 83, pp. 393–411. Tamura, T., Saito, Y., Sieng, S., Ben, B., Kong, M., Choup, S., Tsukawaki, S., 2007. Depositional
Hanebuth, T.J.J., Voris, H.K., Yokoyama, Y., Saito, Y., Okuno, J., 2011. Formation and fate facies and radiocarbon ages of a drill core from the Mekong River lowland near Phnom
of sedimentary depocentres on Southeast Asia's Sunda Shelf over the past sea-level Penh, Cambodia: evidence for tidal sedimentation at the time of Holocene maximum
cycle and biogeographic implications. Earth-Science Reviews 104, 92–110. flooding. Journal of Asian Earth Sciences 29, 585–592.
Hori, H., 2000. The Mekong: Environment and Development. United Nations University Tamura, T., Saito, Y., Sieng, S., Ben, B., Kong, M., Sim, I., Choup, S., Akiba, F., 2009. Initiation
Press, Tokyo . 398 pp. of the Mekong River delta at 8 ka: evidence from the sedimentary succession in the
Isla, F.I., Bujalesky, G.G., 2000. Cannibalisation of Holocene gravel beach-ridge plains, Cambodian lowland. Quaternary Science Reviews 28, 327–344.
northern Tierra del Fuego, Argentina. Marine Geology 170, 105–122. Tamura, T., Horaguchi, K., Saito, Y., Nguyen, V.L., Tateishi, M., Ta, T.K.O., Nanayama, F.,
Jiménez, J.A., Sánchez-Arcilla, A., Valdemoro, H.I., Gracia, V., Nieto, F., 1997. Processes Watanabe, K., 2010. Monsoon-influenced variations in morphology and sediment
reshaping the Ebro delta. Marine Geology 144, 59–79. of a mesotidal beach on the Mekong River delta coast. Geomorphology 116, 11–23.
Kummu, M., Varis, O., 2007. Sediment-related impacts due to upstream reservoir trap- Tamura, T., Bateman, M.D., Kodama, Y., Saitoh, Y., Watanabe, K., Yamaguchi, N., Matsumoto,
ping, the Lower Mekong River. Geomorphology 85, 275–293. D., 2011. Building of shore-oblique transverse dune ridges revealed by ground-
Lamb, H.H., 1995. Climate, History and the Modern World, second ed. Routledge, London . penetrating radar and optical dating over the last 500 years on Tottori coast, Japan.
433 pp. Geomorphology 132, 153–166.
López, G.I., Rink, W.J., 2008. New quartz optical stimulated luminescence ages for beach Tamura, T., Saito, Y., Nguyen, V.L., Ta, T.K.O., Le, M.D., Bateman, M.D., Matsumoto, D.,
ridges on the St. Vincent Island Holocene strand plain, Florida, United States. Yamashita, S., 2012. Origin and evolution of interdistributary delta plains; insights
Journal of Coastal Research 24, 49–62. from Mekong River delta. Geology 40, 303–306.
Lu, X.X., Siew, R.Y., 2006. Water discharge and sediment flux changes over the past decades Tanabe, S., Ta, T.K.O., Nguyen, V.L., Tateishi, M., Kobayashi, I., Saito, Y., 2003. Delta evo-
in the Lower Mekong River: possible impacts of the Chinese dams. Hydrology and lution model inferred from the Holocene Mekong delta, southern Vietnam. In: Sidi,
Earth System Science 10, 181–195. F.H., Nummedal, D., Imbert, P., Darman, H., Posamentier, H.W. (Eds.), Tropical
Marsh, R.E., Prestwich, W.V., Rink, W.J., Brennan, B.J., 2002. Monte Carlo determina- Deltas of Southeast Asia — Sedimentology, Stratigraphy, and Petroleum Geology:
tions of the beta dose rate to tooth enamel. Radiation Measurements 35, 609–616. SEPM Special Publication, 76, pp. 175–188.
Mason, O.K., 1993. The geoarchaeology of beach ridges and cheniers: studies of coastal Tanabe, S., Saito, Y., Vu, Q.L., Hanebuth, T.J.J., Ngo, Q.L., 2006. Holocene evolution of the Song
evolution using archaeological data. Journal of Coastal Research 9, 126–146. Hong (Red River) delta system, northern Vietnam. Sedimentary Geology 187, 29–61.
Mekong River Commission, 2008. Annual Mekong Flood Report 2007. Mekong River Tjallingii, R., Stattegger, K., Wetzel, A., Phach, P.V., 2010. Infilling and flooding of the
Commission (MRC), Vientiane, Lao PDR. 73 pp. Mekong River incised valley during deglacial sea-level rise. Quaternary Science
Milliman, J.D., Syvitski, J.P.M., 1992. Geomorphic/tectonic control of sediment Reviews 29, 1432–1444.
discharge to the oceans: the importance of small mountain rivers. Journal of van Maren, D.S., 2005. Barrier formation on an actively prograding delta system: the
Geology 100, 525–544. Red River Delta, Vietnam. Marine Geology 224, 123–143.
Murray, A.S., Wintle, A.G., 2000. Luminescence dating of quartz using an improved Walling, D.E., 2008. The changing sediment load of the Mekong River. Ambio 37,
single-aliquot regenerative-dose protocol. Radiation Measurements 32, 57–73. 150–157.
T. Tamura et al. / Marine Geology 326–328 (2012) 140–153 153

Wang, P.K., Zhang, D., 1992. Recent studies of the reconstruction of East Asian monsoon Xue, Z., Liu, J.P., DeMaster, D., Nguyen, V.L., Ta, T.K.O., 2010. Late Holocene evolution of
climate in the past using historical literature of China. Journal of the Meteorological the Mekong subaqueous delta, southern Vietnam. Marine Geology 269, 46–60.
Society of Japan 70, 423–446. Zhang, D., 1984. Synoptic-climatic studies of dustfall in China since historical times.
Wang, P., Clemens, S., Beaufort, L., Braconnot, P., Ganssen, G., Jian, Z., Kershaw, P., Scientia Sinica B27, 825–836.
Sarnthein, M., 2005. Evolution and variability of the Asian monsoon system: state Zhang, J., Chen, F., Holmes, J.A., Li, H., Guo, X., Wang, J., Li, S., Lü, Y., Zhao, Y., Qiang, M.,
of the art and outstanding issues. Quaternary Science Reviews 24, 595–629. 2011. Holocene monsoon climate documented by oxygen and carbon isotopes
Woods, A.J.C., Leahy, L.P., 1986. The applicability of multiparameter, relative dating from lake sediments and peat bogs in China: a review and synthesis. Quaternary
techniques to beach ridge correlation in Baja California Norte, Mexico. Journal of Science Reviews 30, 1973–1987.
Coastal Research 2, 1–15.

You might also like