lable at ScienceDirect

Journal of Archaeological Science 36 (2009) 504–512

Contents lists avai

Journal of Archaeological Science

journal homepage: ht tp: / /www.elsevier .com/locate/ jas

Reconstructing late Holocene palaeoenvironments in Bangladesh:phytolith analysis of archaeological soils from Somapura Mahavihara sitein the Paharpur area, Badalgacchi Upazila, Naogaon District, Bangladesh

A.K.M. Masud Alam a,b,*, Shucheng Xie a, Lynley A. Wallis c

a State Key Laboratory of Geological Processes and Mineral Resources, China University of Geosciences, Lu Mo Road 388, Wuhan 430074, PR Chinab Department of Natural History, Bangladesh National Museum, Shahbagh, Dhaka 1000, Bangladeshc Department of Archaeology, Flinders University, GPO Box 2100, Adelaide, SA 5001, Australia

a r t i c l e i n f o

Article history:Received 19 March 2008Received in revised form 29 August 2008Accepted 29 September 2008

Keywords:PhytolithsArchaeologyBangladeshPalaeoenvironmentPaharpurLate HoloceneEl Nino Southern Oscillation

* Corresponding author. State Key Laboratory of GeoResources, China University of Geosciences, Lu Mo RChina. Tel.: þ88 01924778532.

E-mail addresses: akm.masudalam@gmail.com163.com (S. Xie), lynley.wallis@flinders.edu.au (L.A. W

0305-4403/$ – see front matter � 2008 Elsevier Ltd.doi:10.1016/j.jas.2008.09.038

a b s t r a c t

Palaeoenvironmental reconstruction based on phytolith analysis of late Holocene-aged soils in andaround the Somapura Mahavihara archaeological monastery site in the Paharpur area, BadalgacchiUpazila of Naogaon District in northwestern Bangladesh was undertaken. Results indicate five climatezones marked by alternatively cool and temperate events. The phytolith assemblages of Zones 1, 3 and 5are clearly characterized by a higher proportion of Pooideae phytoliths. The climate indices for thesezones are all greater than 50, indicating cooler climatic conditions. In contrast, the phytolith assemblagesof Zones 2 and 4 are typified by a higher proportion of Panicoideae phytoliths; their climate indices are42 and 44, respectively, indicating the presence of a warmer, more temperate climate. In general, thedominance of grasses over broad- leaved trees throughout the sequences suggests that generally cool totemperate and dry conditions with some cyclical variability persisted around Paharpur and thesurrounding region throughout the Pala Dynasty from AD 730 to 1080. The question of whether theseshifts in vegetation might be linked with El Nino Southern Oscillation events, and whether these mightin turn have cultural responses, is raised.

� 2008 Elsevier Ltd. All rights reserved.

1. Introduction

Palaeoenvironmental reconstruction is a fundamental compo-nent of any archaeological research project. Since the publication ofa seminal paper by Rovner (1971), palaeoenvironmental recon-structions using phytolith analysis have made increasing contribu-tions to our understanding of both archaeology and palaeoecology.Phytoliths (also known as plant opals or opaline silica) are siliceousmicrofossils that form through the process of silica precipitation inand/or between the cells of living plant tissues. They occur in many,though not all, plant families but are especially abundant anddiverse in the grass family (Poaceae), assuming shapes diagnostic ofthe grasses in which they are found (Twiss et al., 1969; Blackman,1971; Clifford and Watson, 1977; Brown, 1984; Piperno and Pearsall,1998; Piperno, 2006). Owing to their differing morphologies, insome cases grass phytoliths can be indicative of either the C3 or C4photosynthetic pathway and can therefore be used to help establish

logical Processes and Mineraload 388, Wuhan 430074, PR

(A.K.M.M. Alam), xiecug@allis).

All rights reserved.

palaeoclimatic conditions (Twiss et al., 1969; Tieszen et al., 1979;Livingstone and Clayton, 1980; Twiss, 1992, 2001; Wu et al., 1992);C3 grasses are dominant in cool regions with high latitudes orelevations, while C4 grasses tend to occur in semi-arid to arid, sunnyand warm regions.

Beyond their morphological variability, the other strength ofphytoliths lies in their durability; being inorganic in nature they arenot broken down by bacteria such as are many other microfossils.They are stable across a wide pH range (3–9) and preserve well inboth wet and dry, as well as alternating wet and dry conditions(Piperno, 2006). Hence, unlike organic plant remains, they do notrely on exceptional conditions for survival and are thereforewidespread in sediments and soils.

As a consequence of their abundance, durability and diagnosticmorphologies, phytoliths have been increasingly used to recon-struct aspects of Late Quaternary palaeoenvironments in numeroussediment types, including loess (Lu and Wang, 1991; Lu et al., 1996;Madella, 1997; Blinnkov et al., 2002); lake muds (Carter, 2002;Thorne, 2004); sand dunes (Horrocks et al., 2000; Boyd, 2005);tephra sequences (Sase et al., 1987; Parr, 1999; Lentfer et al., 2001),and coastal plain sequences in addition to other sediment types(Fredlund and Tieszen, 1997; Carter and Lian, 2000; Prebble andShulmeister, 2002; Lu et al., 2002; Abrantes, 2003; Piperno and

A.K.M.M. Alam et al. / Journal of Archaeological Science 36 (2009) 504–512 505

Jones, 2003; Charles and Isabel, 2005). However, their reliability aspalaeoenvironmental indicators in soils has been less well studied(but see for example Kelly et al., 1991; Piperno and Becker, 1996;Alexandre et al., 1999; Prebble et al., 2002; Stromberg, 2004;Stromberg et al., 2007). Similarly, numerous studies have beenreported on phytoliths extracted from archaeological sedimentsand artefact surfaces (e.g. Bowdery, 1998; Barboni et al., 1999; Graveand Kealhofer, 1999; Kealhofer et al., 1999; Mercader et al., 2000;Piperno et al., 2000; Wallis, 2000, 2001; Vrydaghs et al., 2001;Ishida et al., 2003) but the analysis of phytoliths recovered fromarchaeological soils is less common (Rovner, 1971; Pearsall, 1978;Lewis, 1981; Delhon et al., 2003; Sullivan and Kealhofer, 2004).

Similarly, owing to a lack of locally based researchers, ourexisting knowledge about the Holocene climatic history of Ban-gladesh is limited (but for exceptions see Ahmed, 1993; Choudhury,1994); hence there is ‘‘an urgent need for further study to betterunderstand and forecast El Nino teleconnections’’ in the region(Glantz, 2000: 23).

This paper presents the results from a study of phytolithassemblages recovered from the late Holocene-aged archaeologicalsoils of the world heritage listed Somapura Mahavihara archaeo-logical monastery site located in Badalgacchi Upazila of NaogaonDistrict, northwestern Bangladesh. In addition, the palae-oenvironmental data thus obtained is compared with local claymineralogical data (Alam et al., 2008) and global palaeo-tempera-ture curves proposed by Warrick and Ahmad (1996), and thequestion of whether the El Nino Southern Oscillation (ENSO) hasaffected the patterning observed raised.

2. Study site location

The Somapura Mahavihara (meaning ‘‘large monastery’’)archaeological site is located in northwestern Bangladesh, and isarguably the most important and the largest known Buddhistmonastery south of the Himalayas. The site is situated in thevillage of Paharpur, Badalgacchi Upazila of Naogaon District,between latitudes 25�1.790–25�1.950N and longitudes 88�58.500–88�58.700E (Fig. 1). Paharpur is situated in the monsoon regionwith a summer dominant rainfall, lying just north of the Tropic ofCancer. The climate of the area is generally warm and humid,though based on rainfall, humidity, temperature and wind pres-sure four seasons are recognized: (a) pre-monsoon, (b) monsoon,(c) post-monsoon and (d) winter. The maximum daily tempera-ture ranges from 37 �C to 39 �C and the minimum from 7 �C to10 �C. Rainfall is very light from November to February, increasessomewhat in March and April and continues uniformly at about250 mm/month during the monsoon months June to September.In May and October the rainfall average decreases to approxi-mately 130 mm/month. The total average rainfall for the year is5080 mm.

The Paharpur region is well-drained by numerous small,entrenched meandering streams and rivers. Approximately 4.5 kmto the west of Somapura, the Old Jamuna River flows from north tosouth, maintaining a meandering course, forming many oxbowlakes. The traces of relict watercourses are detectable from the airand mostly run in a north–south direction, although an east–westarm appears to have extended laterally just south of the monastery(Chowdhury, 2003).

The Somapura Mahavihara site is currently surrounded by richalluvial farmland with small open fields that are intensively culti-vated, though small pockets of natural vegetation persist in someareas. Artificial cultivation platforms and homestead areas are usedfor growing vegetables and bananas, fruit trees and Betel leaf (pan).Mixed forest, scrub and replanted Gajari (Assam Sal) and Sal(Shorea robusta) occupy patchy areas of the higher ridges of theBarind terraces.

3. Archaeological background

The archaeological ruins of the Somapura Mahavihara monas-tery were discovered by Buchanan Hamilton who visited the sitebetween 1807 and 1812 whilst surveying for the East IndiaCompany (Sanday et al., 1983) and the first excavations were con-ducted at the site later in the 19th century by Sir Alexander Cun-ningham (1879). The ruins were placed on the Bangladesh List ofProtected Monuments in 1919, and inscribed to the World HeritageList in 1985.

The site is extensive, covering approximately 10 km2 (10 ha). Itoccupies a quadrangular court measuring ca. 274 m on each side,with high enclosure walls ca. 5 m thick and between 3.6 and 4.5 mhigh. This gigantic establishment is dominated by a central shrinewith 177 surrounding monastic cells and is conspicuous by its loftyheight and unusual architectural design (Alam, 2003; see Fig. 2).The site lies surrounded by a slightly elevated terrace, meaning thearea remained flood-free throughout historic times, a feature whichit is thought may have been an important factor in allowing the PalaEmpire to flourish.

The chronology of the occupation sequence at Sompura Maha-vihara has largely been established on the basis of excavationsundertaken from 1927–28 and 1933–34 (see Dikshit, 1938). Withthe support of UNESCO, the Department of Archaeology of theMinistry of Cultural Affairs, Bangladesh subsequently completedfurther excavations from 1981–82, 1984–85, 1988–89, and 1990–91,which have further contributed to our understanding of thesequence of events at the site. Such excavations have revealed thatSomapura Mahavihara was occupied predominantly by the PalaDynasty from the middle of the 8th century through to the early12th century. Dharmapala, the second Pala monarch who ruledfrom about AD 770–810, founded Somapura Mahavihara in thesecond half of the 8th century. Towards the end of the 9th centurythe Palas were defeated by the Gurjara-Pratiharas, but roughlya century later the Pala Dynasty flourished once more. In the 11thcentury, the Pala again suffered a setback and were devastated bywar, though towards the end of the century prosperity returnedonce more. In the 12th century the Senas, who were followers ofHinduism, replaced the Palas, and from this time onwards thehuman use of Somapura Mahavihara gradually declined before themonastery was permanently abandoned (Sanday et al., 1983).

The site sits on the Pleistocene–Holocene aged Barind ClayResiduum (Alam et al., 1990), which is the largest Quaternaryphysiographic unit of the Bengal Basin and comprises slightlyelevated landform terraces within the alluvium. The underlyingsedimentsdknown as the Pleistocene–Holocene Madhupur Clay/Barind Claydare considered to be marine–estuarine deposits(Brammer, 1996). The Barind Clay Residuum is a strongly iron-stained, clay-rich unit of weathered alluvial sediment that iscompact and resistant to erosion. In the study area it consists of twozones: an upper zone comprising a grey to yellowish brownmottled clayey silt/silty clay, with a lower zone of red andbrownish-yellow mottled clayey silt; the amount of sand increasesgradually with depth. Both zones feature abundant pottery frag-ments, brick chips and plant roots, and are described as archaeo-logical soils belonging to the Holocene epoch.

4. Materials and methods

4.1. Sampling

The archaeological soils of Somapura were sampled in May 2006and later in August of the same year. For the collection of samplestwo trenches (labelled Trench 1 and Trench 2), were excavated fromthe surface to the first occupation level of the monastery. Trench 1is located in the vicinity of the northern outer wall of the monastery

Fig. 1. Location map. (A) General map of Bangladesh (after Khan, 1991). (B) More detailed view of the Paharpur study area showing the location of Trenches 1 and 2 from which soilsamples were collected (based on an ASTER FCC image).

A.K.M.M. Alam et al. / Journal of Archaeological Science 36 (2009) 504–512506

(near the main entrance of the monastery), has a maximum depthof 1.50 m below surface, and the lower and upper zone of theBarind Clay Residuum is clearly present. Trench 2 is located in thevicinity of the eastern inner wall of the monastery and reacheda maximum depth below surface of 1.38 m (see Fig. 3), and revealsthe upper zone of the Barind Clay Residuum. Samples for phytolithanalysis were mostly collected from Trench 1 at regular intervals of5 cm, while charcoal samples for radiocarbon dating were collectedfrom Trenches 1 and 2 at depths of 67 cm and 138 cm belowsurface, and 10 cm and 95 cm depth, respectively.

4.2. Phytolith extraction techniques

Fifteen soil samples were processed using phytolith extractionprocedures as described by Wang and Lu, 1993 at the Key

Fig. 2. Photograph showing the archaeological ruins of the Somapur

Laboratory of Biogeology and Environmental Geology in ChinaUniversity of Geosciences. Samples were air dried at roomtemperature and then crushed with a mortar and pestle andsubsequently sieved through a 100 mm mesh (0.154 mm) to removeroot fragments and the gravel fraction. Approximately 10 g of eachsoil sample was then placed into sterile beakers. Organic matterwas removed using 30% hydrogen peroxide, followed by a 10%hydrochloric acid wash to remove any carbonates present. Thesamples were then rinsed four times with distilled water in order toremove clay particles and to return the suspension to a neutral pH(pH 7).

Residual samples were transferred to test tubes and 5 ml of zinciodide heavy liquid (adjusted to a specific gravity of 2.35 to ensurephytoliths would ‘‘float’’) was added. The samples were thencentrifuged to separate the phytoliths from the heavier mineral

a Buddhist monastery, Paharpur, Naogaon District, Bangladesh.

Fig. 3. Lithological description of Trench 1 and 2 sections, respectively.

Table 1Climate index value calculated from relative percentages of warm and cool groupphytoliths (after Twiss, 1992, 2001)

Index value Temperature

20 Warm30 Temperate40 Temperate50 Cool60 Very cool70 Very cool

A.K.M.M. Alam et al. / Journal of Archaeological Science 36 (2009) 504–512 507

fraction. The suspended phytoliths were poured into clean testtubes and rinsed three times with distilled water. This was followedby a 5% acetic acid wash to remove any remaining carbonates. Thephytolith fractions were then rinsed again with distilled water andcentrifuged, before being rinsed with absolute alcohol and trans-ferred to vials for long-term storage. Phytoliths were mounted forviewing on slides using glycerin, covered with cover slips andsealed with paraffin. The studies were then examined using a Zeissstereomicroscope at 400� magnification.

4.3. Phytolith identification and quantification

No modern phytolith reference database was available for thestudy region and it was beyond the scope of this project to developone. In the absence of a regional phytolith database, the archaeo-logical phytoliths were identified, described and classifiedfollowing Twiss’ (Twiss et al., 1969; Twiss, 1992, 2001) classificationscheme supplemented by published keys including Mulholland andRapp (1992), Pearsall and Elizabeth (1992), Fredlund and Tieszen(1994), Kealhofer and Piperno (1998), Pearsall (2000), Lu and Liu,2003, Wallis (2003) and Piperno (2006). Phytolith descriptions arebased on the new phytolith code of terminology (IPCN) (Madellaet al., 2005). The number of phytoliths counted ranged from 445 upto 595 per slide in order to identify at least 250 short-cell mor-photypes per sample. Phytoliths with unrecognizable shapes werecounted and listed in the unidentified category. Sponge spiculesand microcharcoal were also recorded and are included in thegraphs as percentages of the overall assemblage.

Short-cell phytoliths were separated into two main categories:those coming from the Pooideae (predominantly C3 grasses) versusthose derived from the Panicoideae (predominantly C4 grasses)sub-families. The Pooideae-types are characterized by trapeziformsinutes and trapeziform polylobates (Mulholland and Rapp, 1992;Twiss, 1992; Wang and Lu, 1993; Fredlund and Tieszen, 1994;Pearsall, 2000; Lu and Liu, 2003; Piperno, 2006). It should be notedthat these phytolith morphotypes have also been described fromthe intercostal zones of the epidermis of Chloridoideae and Pan-icoideae grasses, but are generally produced therein only in smallamounts and hence are considered relatively reliable indicators ofthe Pooideae (Brown, 1984; Mulholland, 1989; Barboni et al., 1999).

The Panicoideae-type subfamily is indicated by the presence ofdistinctive bilobate and polylobate morphotypes (Brown, 1984;Mulholland and Rapp, 1992; Twiss, 1992; Pearsall, 2000; Piperno,2006).

Other grass morphotypes identified in the phytolith assem-blages included those from the epidermis long-cells, hair cellmesophyll and bulliform cells, including the forms elongatesmooth, elongate echinate, scutiform hair cell (point-shaped type),cuneiform (fan-shaped) bulliforms and parallelepipedal (square/rectangular shaped) bulliforms (Twiss et al., 1969; Pearsall andElizabeth, 1992; Twiss, 1992, 2001; Wang and Lu, 1993; Lu and Liu,2003; Piperno, 2006; Barboni et al., 2007).

4.4. Climate index

In Twiss’ (Twiss et al., 1969; Twiss, 1992, 2001) classification ofgrass phytoliths, the division and classification of Festucoid (Pooid),Panicoid and Chloridoid type phytoliths generally correspond withdifferent climates. For example, Panicoid-type phytoliths aremainly produced by C4 grasses that predominantly grow in warmand humid areas. C4 grasses also produce Chloridoid-type phyto-liths, but these grasses are typical of warm semiarid or highlyseasonal regions well beyond the limits of the study area. Pooid-type phytoliths are produced dominantly by C3 grasses thatflourish in cool seasons and climates in high latitudes or elevations.Hence, changes in the relative abundance of Panicoid versus Pooidphytolith types throughout a section can be used to infer pastclimatic change (cf. Gu et al., 2007). For the purposes of this anal-ysis, after comprehensively considering the different photosyn-thetic pathways in plants phytolith morphotypes were divided intotwo groups: The cool group includes Pooid-type morphologies (i.e.trapeziform polylobates, trapeziform sinutes), elongates (bothsmooth and echinate) and point-shaped (scutiform hair cell) phy-toliths; in contrast, the warm group morphologies include Pani-coid-type shapes (i.e. bilobates and polylobates), cuneiformbulliforms and parallelepipedal bulliform. The ratio of cool to warmgroup phytoliths is used here as a climate index, calculated as:

Climate=temperature index ðIcÞ

¼ Cool group phytolithsðCool group phytolithsþwarm group phytolithsÞ � 100

Subsequently, high values are suggestive of a cooler climate(allowing more C3 grasses with an optimum temperature range ofbetween 18 and 23 �C to influence the phytolith assemblage), whileresulting low values are indicative of a warmer climate (allowingmore C4 grasses with an optimum temperature range of between32 and 35 �C to influence the phytolith assemblage) (see Table 1).

5. Results

5.1. Dating

Four charcoal samples were analyzed for 14C in the Quaternaryage dating laboratory of Peaking University, Beijing. The 14C age of

A.K.M.M. Alam et al. / Journal of Archaeological Science 36 (2009) 504–512508

each charcoal sample was converted into a calibrated calendar ageby OxCal software version 3.10 (Bronk-Ramsey, 2005) using cali-bration curve of IntlCal04 (Reimer et al., 2004). The results ofthe OxCal calibration of the measured 14C age are summarized inTable 2. The radiocarbon dates thus obtained fit well with theknown occupation history of the site based on written records, witharchaeological sediment accumulation appearing to commenceduring the middle of the 8th century AD coincident with the firstconstruction of the site.

5.2. Phytolith assemblages

Phytoliths were plentiful and well preserved in all 15 samples, aswere other siliceous microfossils such as sponge spicules fromaquatic organisms. While phytoliths are known to be transportedvia wind and water vectors, it is assumed in this case that theassemblages are predominantly the result of in situ decay andtherefore are representative of local vegetation and by proxy localclimatic conditions. It is assumed the sponge spicules have beendeposited through the periodic flooding of the landscape.

The major plant groups represented in the samples are Poaceaeand broad-leaved trees. Poaceae phytoliths include short cells,long cells, bulliform cells, hair cells, Pooid-type morphotypes(trapeziform polylobates and trapeziform sinutes) dominate theassemblages, followed in much lower relative frequency by thePanicoid-types (bilobate and polylobate). Broad-leaved trees areindicated by two characteristic types, platelets polygon (polyhedralplate-like) and globular granulate (spherical rugose), abundant indeciduous trees and subtropical-tropical regions, respectively(Wang and Lu, 1993; Kealhofer and Piperno, 1998; Piperno, 2006).

The phytolith assemblages can be divided into five major zoneson the basis of the dominant types and climate index. These fivezones and the relative percentages of phytolith assemblages arepresented in Figs. 4 and 5 and described below.

Zone 1 (150–95 cm below surface): Charcoal samples from thiszone yielded ages of AD 730 and 850 at 138 and 95 cm belowsurface, respectively. Poaceae phytoliths dominate the samplesfrom this zone (69%), whereas broad-leaved tree phytoliths repre-sent only 8%. Among the Poaceae phytoliths, bulliform cells (29%)and Pooideae-types (23%) are the most common. Scutiform haircelland long cell phytoliths have percentage values close to 8%, whilePanicoideae-type phytoliths (reflecting warm and humid climaticconditions) are absent in the lower part of this zone, but present invery low quantities in the upper part. In contrast microcharcoalaccounted for approximately 22% of the overall assemblage, andsponge spicules about 0.2%. The climate index value is relativelyhigh throughout this zone, averaging ca. 57, reflecting the presenceof a cool climate.

Zone 2 (95–70 cm below surface): This zone is also dominatedby Poaceae phytoliths with a slight decrease to about 64%, while theless dominant broad-leaved tree types increase in percentagevalues to about 11%. There are slight increases in microcharcoal andsponge spicule quantities relative to Zone 1. Within the Poaceaemorphotypes, bulliform cells increased (to ca. 36%) while Pooi-deae-types decreased (to about 14%). Long cell phytoliths showedvalues similar to those of Zone 1, but the relative abundance of

Table 2Radiocarbon dates from Trenches 1 and 2, Paharpur archaeological site, Bangladesh

Lab. No. Trench Depth belowsurface (cm)

Matdate

BA 06800 1 67 ChaBA 06801 1 138 ChaBA 06802 2 10 ChaBA 06803 2 95 Cha

scutiform haircell decreased to 6%, while Panicoideae-type phyto-liths slightly increases to about 0.4%. The climate index decreased toabout 44 and 41 at 87.5 cm and 77.5 cm depth below surface,respectively, reflecting the existence of a more temperate climate.The climate index value of 41 at 77 cm depth is the lowestencountered in the whole section.

Zone 3 (70–45 cm below surface): Poaceae phytolithsincreased (to 66%) and broad- leaved trees decreased (to 9%) incomparison to quantities in Zone 2. Microcharcoal fragmentsslightly decreased in relative abundance. Among the Poaceaephytolith types, bulliform cells decreased and Pooideae-typesincreased to about 30% and 20%, respectively. Scutiform haircellsshow a slight reduction in value (7%), while long cell and Pan-icoideae-type phytoliths showed values similar to those of Zone 1.Sponge spicule frequency is little lower (about 0.8%) than thatwitnessed in Zone 2. The climate index value increased to about 53,reflecting the presence of a cool climate.

Zone 4 (45–35 cm below surface): This zone is also dominatedby Poaceae phytoliths (63%) with a low content of broad-leavedtrees (10%). Microcharcoal increases in relative abundance. Bulli-form cells increase and Pooideae-type phytoliths decrease to ca.34% and 13%, respectively. Scutiform hair cell abundance decreasesto about 5% whereas long cell and Panicoideae-type phytolithsincrease in relative abundances to approximately 10% and 0.3%.Sponge spicule frequency is much higher (3%) in this zone. Thetemperature index value decreases to about 44, reflectinga temperate climate.

Zone 5 (35–5 cm below surface): Charcoal samples from thiszone yielded ages of AD 915 and 1080 at 67 and 10 cm depth belowsurface, respectively. Poaceae phytoliths increase slightly in abun-dance to about 64%, while broad-leaved tree phyotliths decrease toabout 8%. Microcharcoal decreased in relative abundance. Bulliformcells decreased to about 31% while Pooideae-type, scutiform haircell and long cell phytoliths all slightly increased to about 14%, 6%and 13%, respectively. The relative abundances of Panicoideae-typephytoliths decreased to about 0.25% of the total assemblage.Sponge spicule frequency is relatively lower (about 2%) than thatobserved in Zone 4. The climate index increases to about 51,reflecting the presence of a cool climate.

6. Discussion and conclusions

A great strength of phytolith analysis is that it allows thereconstruction of past floral communities and interpretation of theenvironmental conditions and changes they represent, particularlyin respect to the different grass communities present. The majorityof the phytolith assemblages recovered from the Holocenearchaeological soils of Somapura are dominated by Poaceae phy-toliths with a generally small contribution from broad- leaved trees.Among the Poaceae, Pooideae-type and bulliform cell phytolithsdominate, followed in relative frequency by elongate and point-shaped (scutiform haircell) types, with much lower relativefrequencies of the Panicoid-types (bilobates and polylobates). Thephytolith assemblages from Zones 1, 3 and 5 are characterized bya higher proportion of cool group Poaceae types (36%), while warmgroup Poaceae type phytoliths represented only 30% of the

eriald

Radiocarbon age(14C yr BP)

Calibrated age (cal yr AD)(95.4% probability, 2s)

rcoal 1105 � 45 AD 915 (810–1020)rcoal 1295 � 40 AD 730 (650–810)rcoal 960 � 40 AD 1080 (990–1170)rcoal 1175 � 40 AD 850 (720–980)

Fig. 4. Relative abundances of major phytoliths versus depth.

A.K.M.M. Alam et al. / Journal of Archaeological Science 36 (2009) 504–512 509

assemblages in these zones. The diagram of phytoliths and climateindex shows that all of these three zones are constantly dominatedby Pooid, scutiform hair cell and long cell phytoliths and the climateindexes are all higher than 50 indicating cool climatic conditions(Fig. 5). In contrast, patterning in the phytolith assemblage relatedto Zones 2 and 4 are effectively reversed, typified by a highproportion of warm group type phytoliths (ca. 35%) while coolgroup type phytoliths comprised only approximately 28% of theassemblage. These two zones are dominated by bulliform cells withlow climate indices of 42 and 44, respectively, indicating thepresence of a generally cool, temperate climate. Furthermore, asmentioned earlier it is evident that Pooid-type phytoliths aresignificantly more abundant than the Panicoid-types in all samples(Fig. 5). Characteristically, the presence of Pooid grasses are indic-ative of cool climatic conditions, while Panicoid grasses are adaptedto warm and wet environments. Because Panicoid phytoliths occurin such low frequencies throughout the Somapura Mahaviharasequence, it suggests that there was not an especially warm or wetenvironment present at any time. Similarly, the abundance ofgrasses over broad-leaved trees is supportive of the idea thatgenerally consistently dry conditions prevailed throughout the PalaDynasty from AD 730 to 1080.

Thus, the result implies that generally cool to temperate and dryconditions persisted around Paharpur and the surrounding regionthroughout the Pala Dynasty from AD 730 to 1080 with minorfluctuations through warmer and cooler periods. This scenario isconsistent with the clay mineralogical data from the same soilsamples as reported by Alam et al. (2008). Alam et al. (2008) haveshown that the dominant clay mineral in the Somapura sequencesis illite, with smaller amounts of kaolinite and chlorite. These

minerals are interpreted as having formed under generally cool totemperate and dry climates.

While the late Holocene is known to have featured much lessclimatic variability than the preceding early Holocene and Pleis-tocene periods, it has experienced some significant shifts charac-terized by a trend towards increased aridity and heightenedvariability after ca. 3000 BP (e.g. Whetton et al., 1990; McGloneet al., 1992; Allan et al., 1996). The geographical features and loca-tion of Bangladesh mean it is particularly susceptible to the effectsof ENSO in the late Holocene. During El Nino events the sea-levelpressure gradients in the southeaster Pacific weaken, accompaniedby a decrease in the strength of trade winds, resulting in reductionsin rainfall to Bangladesh during El Nino years while La Nina yearsare associated with increased rainfall and floods (see Ahmed, 1993;Choudhury, 1994; Shaman and Tziperman, 2005). Also of climaticrelevance during the late Holocene period was the occurrence ofthe Little Ice Age (beginning somewhere between the 13th centuryand 16th centuries and finishing in the mid-19th century AD),when the global average temperature was probably about 0.5 �Ccooler than that prevailing at the beginning of the 20th century(Warrick and Ahmed, 1996; see Fig. 6). The effects of this climaticevent are best documented in Europe, though there is evidence thatit was in fact a global phenomenon (Grove, 1988).

Using Fig. 6 as a starting point it suggests that during the periodfrom the early 8th to the late 10th century, there was a slightlycooler climate than at present (up to 0.5 �C cooler). Then towardsthe end of the 11th century, average temperatures increased byapproximately 0.75 �C meaning it was up to 0.25 �C warmer than atpresent. This warming trend within a generally cool temperateregime continued, and in the early 12th century (AD 1100) the

Fig. 5. Relative abundances of grass subfamilies and results of climate index versus depth.

A.K.M.M. Alam et al. / Journal of Archaeological Science 36 (2009) 504–512510

temperature increased slightly more to ca. 0.5 �C warmer thanpresent. Interestingly, as far as can be ascertained from thearchaeological and radiocarbon evidence, there is no indication ofthe Pala Dynasty having persisted in the Paharpur region after thelate 11th and/or early 12th century.

As Choudhury (1994) has pointed out, extensive famines inBangladesh in the historical period have been shown to be stronglycorrelated with El Nino events (e.g. the 1770, 1943, 1974 and 1982famines), typically as a result of reduced rainfall causing crops tofail, with the situation exacerbated by severe flooding associatedwith the transition into La Nina years. Such events cause massivereductions in agricultural productivity and production, leading towidespread poverty, migration and social unrest. As alluded to inSection 3, the Sompura Mahaviharan site has experienced a seriesof changes in occupation with social upheaval and changes in theruling powers; likewise, the phytolith analysis documents a series

Fig. 6. Past changes in global temperature for the last

of changes in local vegetation patterns and by inference localclimatic conditions. To what degree these changes in the archaeo-logical records can be argued to be causatively linked with theclimatic variability in the late Holocene, particularly the effects ofthe ENSO, is unknown, but is an interesting possibility that isworthy of further study.

In summary, the phytolith assemblages of the late Holocene-aged archaeological soils from Somapura Mahavihara at Paharpurin northwestern Bangladesh are dominated by large quantities ofgrass (Poaceae) phytoliths with a small contribution from broad-leaved trees. The corresponding climate indices suggest generallycool to temperate and dry conditions persisted in this regionthroughout the Pala Dynasty from 730 to 1080 AD, with periods offluctuating warmer and cooler periods. The emerging climatic databased on the phytolith assemblages are consistent with the globalpalaeo-temperature curve during this period, and may be related to

1000 years. (Courtesy: Warrick and Ahmed, 1996).

A.K.M.M. Alam et al. / Journal of Archaeological Science 36 (2009) 504–512 511

the ENSO phenomena. Furthermore, the data presented here are inclose agreement with the clay mineralogical data from Paharpur.However, because of our limited understanding of phytolithproduction patterns in modern soils and vegetation of Bangladesh,the general conclusions drawn from this study must be consideredpreliminary. Nevertheless, this study is significant as it representsthe first phytolith research conducted in Bangladesh and serves asa starting point for further detailed investigations of the extensiveSomapura Buddhist monastery site.

Acknowledgments

This research was undertaken as part of a Ph.D. by the primaryauthor, supported by the Chinese Government ScholarshipProgram through the China Scholarship Council (CSC), The Peoples’Republic of China. The authors are deeply grateful for this support,as well as for additional financial support provided by NSFC (grantno. 40525008) for conducting fieldwork and some of the laboratorywork in Bangladesh. The authors are also indebted to the ForeignStudent Office, China University of Geosciences for their logisticsupport. The authors also gratefully acknowledge the cooperationand support from the respective authorities of Bangladesh NationalMuseum, Department of Archeology, Bangladesh Space, Researchand Remote Sensing Organization (SPARRSO) and Department ofGeology, University of Dhaka, Bangladesh. Thanks are due to Dr.Sifatul Quader Chowdhury, Professor of Geology Department,Dhaka University for providing help in the field and Dr. Gu Yan-sheng, Associate Professor, School of Environmental Studies, ChinaUniversity of Geosciences for his assistance with the phytolithanalysis. Last but by no means least, the authors are grateful tothe unknown reviewers and the Editors for their critical readingof the manuscript; as always any errors in interpretation remain theresponsibility of the authors.

References

Abrantes, F., 2003. A 340,000 year continental climate record from tropical Africa-news from opal phytoliths from the equatorial Atlantic. Earth and PlanetaryScience Letters 209, 165–179.

Ahmed, A.S.M.S., 1993. A climatological study of Bangladesh and possible correla-tion with El Nino/Southern Oscillation. Unpublished Master of Science disser-tation, University of Dhaka, Bangladesh.

Alam, M.S., 2003. Paharpur. National Encyclopedia of Bangladesh, Banglapedia.http://banglapedia.search.com.bd/HT/P_0019.htm accessed 15 July 2007.

Alam, M.D., Khurshid, H., Shahidul, A.K.M., Khan, M.R., Whitney, J.W., 1990.Geological Map of Bangladesh at 1:1000,000. Government of the PeoplesRepublic of Bangladesh, Geological Survey of Bangladesh.

Alam, A.K.M.M., Xie, S., Saha, D.K., Chowdhury, S.Q., 2008. Clay mineralogy ofarchaeological soil: An approach to paleoclimatic and environmental recon-struction of the archaeological sites of the Paharpur area, Badalgacci upazila,Naogaon District, Bangladesh. Environmental Geology 53, 1639–1650.

Alexandre, A., Meunier, J.-D., Mariotti, A., Soubies, F., 1999. Late Holocene phytolithand carbon-isotope record from a Latosol at Salitre, South-Central Brazil.Quaternary Research 51, 187–194.

Allan, R., Lindesay, J., Parker, D., 1996. The El Nino Southern Oscillation and ClimaticVariability. CSIRO Publishing, Collingwood.

Barboni, D., Bonnefille, R., Alexandre, A., Meunier, J.D., 1999. Phytoliths as paleo-environemental indicator at the Middle Awash hominid site, Ethiopia. Paleo-geography, Paleoclimatology, Paleoecology 152, 87–100.

Barboni, D., Bremond, L., Bonefille, R., 2007. Comparative study of modern phytolithassemblages from inter-tropical Africa. Palaeogeography, Palaeoclimatology,Palaeoecology 246, 457–470.

Blackman, E., 1971. Opaline silica in the range grasses of southern Alberta. CanadianJournal of Botany 49, 769–781.

Blinnkov, M., Busacca, A., Whitlock, C., 2002. Reconstruction of the late Pleistocenegrassland of the Columbia basin, Washington, USA, based on phytolith recordsin loess. Palaeogeography, Palaeoclimatology, Palaeoecology 177, 77–101.

Bowdery, D., 1998. Phytolith Analysis Applied to Pleistocene–Holocene Archeo-logical Sites in the Australian Arid Zone. Hadrian Books, Oxford.

Boyd, M., 2005. Phytoliths as paleoenvironmental indicators in a dune field on thenorthern Great Plains. Journal of Arid Environments 61 (3), 357–375.

Brammer, H., 1996. The Geography of the Soils of Bangladesh. University PressLimited, Dhaka 1000, Bangladesh.

Bronk-Ramsey, C., 2005. Improving the resolution of radiocarbon dating by statis-tical analysis. In: Levy, T.E., Higham, T.F.G. (Eds.), The Bible and RadiocarbonDating: Archaeology, Text and Science. Equinox, London, pp. 57–64.

Brown, D.A., 1984. Prospects and limits of a phytolith key for grasses in the centralUnited States. Journal of Archaeological Sciences 11, 345–368.

Carter, J.A., 2002. Phytolith analysis and paleoenvironmental reconstruction fromLake Poukawa Core, Hawkes Bay, New Zealand. Global and Planetary Change 33,257–267.

Carter, J.A., Lian, O.B., 2000. Paleoenvironmental reconstruction from the lastinterglacial using phytolith analysis, south eastern North Island, New Zealand.Journal of Quaternary Science 15 (7), 733–743.

Charles, D.T., Isabel, I.A., 2005. Paleoenvironment and plant cultivation on terracesat La Quemada, Zacatecas, Mexico: The pollen, phytolith and diatom evidence.Journal of Archaeological Science 32, 341–353.

Choudhury, A.M., 1994. Bangladesh floods, cyclones and ENSO. Unpublished paperpresented at the International Conference on Monsoon Variability and Predic-tion. ICTP, pp. 8–9.

Chowdhury, S.Q., 2003. Paharpur (Framework). National Encyclopedia of Bangla-desh, Banglapedia. http://banglapedia.search.com.bd/HT/P_0020.htm accessed15 July 2007.

Clifford, H.T., Watson, L., 1977. Identifying Grasses: Data, Methods and Illustrations.University of Queensland Press, St. Lucia.

Cunningham, A., 1879. Report of a tour in Bihar and Bengal from Patua to Sunar-gaon. Archaeological Survey of India 15, 117–120.

Dikshit, K.N., 1938. Excavation at Paharpur, Bengal. Memoir of the ArchaeologicalSurvey, Delhi.

Delhon, C., Alexandre, A., Berger, J.-F., Thie bault, S., Brochier, J.-L., Meunier, J.D.,2003. Phytolith assemblages as a promising tool for reconstructing Mediter-ranean Holocene vegetation. Quaternary Research 59, 48–60.

Fredlund, G.G., Tieszen, L.L., 1994. Modern phytolith assemblages from the NorthAmerican Great Plains. Journal of Biogeography 21, 321–335.

Fredlund, G.G., Tieszen, L.L., 1997. Calibrating grass phytolith assemblages inclimatic terms: application to late Pleistocene assemblages from Kansas andNebraska. Palaeogeography, Palaeoclimatology, Palaeoecology 136, 199–211.

Glantz, M.H., 2000. Lessons Learned from the 1997–98 El Nino: Once Burned, TwiceShy? Report prepared for the UNEP/NCAR/UNU/WMO/ISDR.

Grave, P., Kealhofer, L., 1999. Assessing bioturbation in archaeological sedimentsusing soil morphology and phytolith analysis. Journal of Archaeological Science26, 1239–1248.

Grove, J.M., 1988. The Little Age. Methuen, London.Gu, Y., Pearsall, D.M., Xie, S., Yu, J., 2007. Vegetation and fire history of

a Chinese site in southern tropical Xishuangbanna derived from phytolithand charcoal records from Holocene sediments. Journal of Biogeography 35(2), 325–341.

Horrocks, M., Deng, Y., Ogden, J., Sutton, D.G., 2000. A reconstruction of the historyof a Holocene sand dune on Great Barrier Island, northern New Zealand, usingpollen and phytolith analyses. Journal of Biogeography 27, 1269–1277.

Ishida, S., Parker, A.G., Kennet, D., Hodson, M.J., 2003. Phytolith analysis from thearchaeological site of Kush, Ras al-Khaimah, United Arab Emirates. QuaternaryResearch 59, 310–321.

Kealhofer, L., Piperno, D.R., 1998. Opal Phytoliths in Southeast Asian Flora. Smith-sonian Institute Press, Washington DC.

Kealhofer, L., Torrence, R., Fullagar, R., 1999. Integrating phytoliths with use wear/residue studies of stone tools. Journal of Archaeological Science 26, 527–546.

Kelly, E.F., Amundson, R.G., Marino, B.D., Deniro, M.J., 1991. Stable isotope ratios ofcarbon in phytoliths as a quantitative method of monitoring vegetation andclimate change. Quaternary Research 35, 222–233.

Khan, F.H., 1991. Geology of Bangladesh. The University Press Limited, Dhaka,Bangladesh.

Livingstone, D.A., Clayton, W.D., 1980. An altitudinal cline in tropical Africangrass floras and its paleoecological significance. Quaternary Research 13,392–402.

Lentfer, C.J., Boyd, W.E., Torrence, R., 2001. Phytolith research relating to thearchaeology of West New Britain, Papua New Guinea. In: Meunier, J.D., Colin, F.(Eds.), Phytoliths: Applications in Earth Sciences and Human History. A.A. Bal-kema Publishers, Lisse, pp. 213–224.

Lewis, R.O., 1981. Use of opal phytoliths in palaeoenvironmental reconstruction.Journal of Ethnobiology 1, 175–181.

Lu, H.Y., Liu, K.B., 2003. Morphological variations of lobate phytoliths from grassesin China and the southeastern USA. Diversity and Distributions 9, 73–87.

Lu, H.Y., Wang, Y.J., 1991. A study of phytoliths in loess profile and palae-oenvironmental evolution at Heimugou in Luochuan, Shaanxi Province sincelate Pleistocene. Quaternary Science 1, 72–84.

Lu, H.Y., Wu, N.Q., Liu, D.S., Han, J.M., Qin, X.G., Sun, X.J., Wang, Y.J., 1996. Seasonalclimatic variation recorded by phytolith assemblages from the Baoji loesssequence in central China over the last 150,000 a. Science in China (Series D) 39,629–639.

Lu, H.Y., Liu, Z.X., Wu, N.Q., Berne, S., Saito, Y., Liu, B.Z., Wang, L., 2002. Ricedomestication and climatic change: Phytolith evidence from East China. Boreas31, 378–385.

Madella, M., 1997. Phytoliths from a Central Asia loess-paleosol sequence andmodern soils: their taphonomical and palaeoecological implication. In:Pinilla, A. (Ed.), The State of the Art of Phytoliths in Plants and Soils. Mono-grafias del Centro de Ciencias Medambioentales, Madrid, pp. 49–58.

Madella, M., Alexandre, A., Ball, T., 2005. International code for phytolith nomen-clature 1.0. Annals of Botany 96, 253–260.

A.K.M.M. Alam et al. / Journal of Archaeological Science 36 (2009) 504–512512

McGlone, M.S., Kershaw, A.P., Markgraf, V., 1992. El Nino/Southern Oscillationclimatic variability in Australasian and South American palaeoenvironmentalrecords. In: Diazm, H.F., Markgraf, V. (Eds.), El Nino: Historical and Palae-oclimatic Aspects of the Southern Oscillation. Cambridge University Press,Cambridge, pp. 435–462.

Mercader, J., Runge, F., Vrydaghs, L., Doutrelepont, H., Ewango, C.E.N., Juan-Tresseras, J., 2000. Phytoliths from archaeological sites in the tropical forest ofIturi, Democratic Republic of Congo. Quaternary Research 54, 102–112.

Mulholland, S.C., 1989. Phytolith shape frequencies in North Dakota grasses: Acomparison to general patterns. Journal of Archaeological Science 16, 489–511.

Mulholland, S.C., Rapp Jr., G., 1992. A morphological classification of grass silica-bodies. In: Rapp Jr., G., Mulholland, S.C. (Eds.), Phytolith Systematics: EmergingIssues. Plenum Press, New York, pp. 65–90.

Parr, J.F., 1999. The spatial patterning of palaeo-forest disturbance at an archaeo-logical site on Garua Island, West New Britain, as indicated by fossil phytoliths.Unpublished B. Appl. Sc. (Hons.) dissertation, Southern Cross University,Lismore.

Pearsall, D.M., 1978. Phytolith analysis of archaeological soils: Evidence for maizecultivation in formative Equador. Science 199, 177–178.

Pearsall, D.M., 2000. Paleoethnobotany: A Handbook of Procedures, second ed.Academic Press, San Diego.

Pearsall, D.M., Elizabeth, H.D., 1992. Developing a phytolith classification system. In:Rapp Jr., G., Mulholland, S.C. (Eds.), Phytolith Systematics: Emerging Issues.Plenum Press, New York, pp. 37–64.

Piperno, D.R., 2006. Phytoliths: A comprehensive guide to archaeologists andpalaeoecologists. Altamira Press, Maryland.

Piperno, D.R., Becker, P., 1996. Vegetational history of a site in the central Amazonbasin derived from phytolith and charcoal records from natural soils. Quater-nary Research 45, 202–209.

Piperno, D.R., Jones, J.G., 2003. Paleoecological and archaeological implications ofa Late Pleistocene/Early Holocene record of vegetation and climate from thePacific coastal plain of Panama. Quaternary Research 59, 79–87.

Piperno, D.R., Pearsall, D.M., 1998. The Silica Bodies of Tropical American Grasses:Morphology, Taxonomy and Implications for Grass Systematics and FossilPhytolith Identification. Smithsonian Institution Press, Washington, DC.

Piperno, D.R., Ranere, A.J., Holst, I., Hansell, P., 2000. Phytoliths in Cucurbita andother neotropical Cucurbitaceae and their occurrence in early archaeologicalsites from the lowland American tropics. Journal of Archaeological Science 27,193–208.

Prebble, M., Shulmeister, J., 2002. An analysis of phytolith assemblages for thequantitative reconstruction of late Quaternary environments of the Lower TaieriPlain, otago, South Island, New Zealand: II. Paleoenvironmental reconstruction.Journal of Paleolimnology 27 (4), 415–427.

Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Bertrand, C., Blackwell, P.G.,Buck, C.E., Burr, G., Cutler, K.B., Damon, P.E., Edwards, R.L., Fairbanks, R.G.,Friedrich, M., Guilderson, T.P., Hughen, K.A., Kromer, B., McCormac, F.G.,Manning, S., Bronk Ramsey, C., Reimer, R.W., Remmele, S., Southon, J.R.,Stuiver, M., Talamo, S., Taylor, F.W., van der Plicht, J., Weyhenmeyer, C.E., 2004.IntCal04 terrestrial radiocarbon age calibration, 0–26 cal kyr BP. Radiocarbon 46(3), 1029–1058.

Rovner, I., 1971. Potential of opal phytoliths for use in paleoecological reconstruc-tion. Quaternary Research 1, 345–359.

Sanday, J., Frost, A., Smyth, J., Lohuizen, J.V., 1983. Technical Report-Master Plan forthe conservation and preservation of the ruins of the Buddhist Vihara at

Paharpur and Historic Mosque-City of Bagherhat. Serial No. FNR/CLT/CH/83/258(UNDP).

Sase, T., Hosono, M., Utsgawa, T., Aoki, K., 1987. Opal phytolith analysis of presentand buried volcanic ash soils in Te Ngae Road tephra section, Rotorua Basin,North Island, New Zealand. Quaternary Research 27, 153–163.

Shaman, J., Tziperman, E., 2005. The effect of ENSO on Tibetan Plateau snow depth:A stationary wave telconnection mechanism and implications for the SouthAsian monsoons. Journal of Climate 18 (12), 2067–2079.

Stromberg, C.A.E., 2004. Using phytolith assemblages to reconstruct the origin andspread of grass-dominated habitats in the great plains of North America duringthe late Eocene to early Miocene. Palaeogeography, Palaeoclimatology, Palae-oecology 207, 239–275.

Stromberg, C.A.E., Werdelin, L., Friis, E.M., Gerçek, S., 2007. The spread of grass-dominated habitats in Turkey and surrounding areas during the Cenozoic: Phy-tolith evidence. Palaeogeography, Palaeoclimatology, Palaeoecology 250, 18–49.

Sullivan, K.A., Kealhofer, L., 2004. Identifying activity areas in archaeological soilsfrom a colonial Virginia house lot using phytolith analysis and soil chemistry.Journal of Archaeological Science 31 (12), 1659–1673.

Thorne, V.C., 2004. Phytolith evidence for C4-dominated grassland since the earlyHolocene at Long Pocket, northeast Queensland, Australia. Quaternary Research61, 168–180.

Tieszen, L.L., Senyimba, M.M., Imbamba, S.K., Troughton, J.H., 1979. The distributionof C3 and C4 grasses and carbon isotope discrimination along an altitudinal andmoisture gradient in Kenya. Oecologia 37, 337–350.

Twiss, P.C., 1992. Predicted world distribution of C3 and C4 grass phytoliths. In:Rapp Jr., G., Mulholland, S.C. (Eds.), Phytolith Systematics: Emerging Issues.Plenum Press, New York, pp. 113–128.

Twiss, P.C., 2001. A cormudgeon’s view of grass phytolitholog. In: Meunier, J.D.,Colin, F. (Eds.), Phytoliths in Earth Sciences and Human History. A.A. Balkema,Rotterdam, pp. 7–25.

Twiss, P.C., Suess, E., Smith, R.M., 1969. Morphological classification of grass phy-toliths. Soil Science of America Proceedings 33, 109–115.

Vrydaghs, L., Doutrelpont, H., Beeckmanm, H., Haerinckm, E., 2001. Identificationsof a morphotype association of Phoenix dactylifera L. lignified tissues origin ated-Dur (1st AD), Umm al-Qaiwain (UAE). In: Meunier, J.D., Colin, F. (Eds.),Phytoliths in Earth Sciences and Human History. A.A. Balkema, Rotterdam, pp.239–250.

Wallis, L.A., 2000. Phytoliths, late Quaternary environment and archeology intropical semi-arid northwest Australia. Unpublished PhD dissertation, TheAustralian National University, Canberra.

Wallis, L.A., 2001. Environmental history of northwest Australia based on phytolithanalysis at Carpenter’s Gap1. Quaternary International 82-85, 03–117.

Wallis, L.A., 2003. An overview of phytolith production patterns in selected north-west Australian flora. Review of Palaeobotany and Palynology 125, 210–248.

Wang, Y.J., Lu, H.Y., 1993. Phytolith Study and its Application. China Ocean Press,Beijing.

Warrick, R.A., Ahmed, Q.K., 1996. The Implications of Climate and Sea-level Changefor Bangladesh. Kluwer Academic Publications, The Netherlands.

Whetton, P., Adamson, D., Williams, M., 1990. Rainfall and river flow in Africa,Australia and East Asia linked to El Nino-Southern Oscillation events. In:Bishop, P. (Ed.), Lessons for Human Survival: Nature’s record from the Quater-nary. Geological Society of Australia, pp. 71–82.

Wu, N.Q., Lu, H.Y., Nie, G.Z., 1992. The study of phytoliths in C3 and C4 grasses andits paleoecological significance. Quaternary Sciences 3, 241–251.