A sediment-based record of Lateglacial and Holocene environmental changes from Guangfulin, Yangtze delta, eastern China

The Holocene 17,8 (2007) pp. 1221–1231

A sediment-based record of Lateglacial and Holocene environmental changes from Guangfulin, Yangtze delta, eastern China Freea Itzstein-Davey,1 Pia Atahan,2* John Dodson,3 David Taylor1 and Hongbo Zheng4 ( 1Museum Building, Department of Geography, School of Natural Sciences, University of Dublin, Trinity College, Dublin 2, Ireland; 2M004, School of Earth and Geographical Sciences, The University of Western Australia, 35 Stirling Highway, Crawley WA 6009, Australia; 3Institute for Environmental Research, Australian Nuclear Science and Technology Organisation, Menai NSW 2234, Australia; 4 School of Ocean and Earth Sciences, Tongji University, 1239 Siping Road, Shanghai 200092, China) Received 10 April 2007; revised manuscript accepted 26 August 2007

Abstract: Multiproxies of past environmental conditions, comprising 53 sediment samples analysed for their lithostratigraphic properties (mainly their charcoal, phytoliths and pollen contents) from an AMS 14C-dated sequence of sediments accumulating at Guangfulin, Yangtze delta, are presented. The oldest sediments recovered date to the Lateglacial when a mosaic of mixed (conifer-deciduous) temperate forest and wetland vegetation characterized the study area. The Lateglacial–Holocene transition and much of the early Holocene record to c. 7400 yr BP appears to be missing from the sequence. The earliest evidence possibly representing human activities in the study area (the remains of cereals and indicators of forest) date to c. 7000 yr BP. A large increase in macrocharcoal remains c. 4700 yr BP is a more certain indication of human activities close to the study site, and may indicate the first occupation of what is now the location of a major archaeological excavation at Guangfulin. Technological changes during the Eastern Zhou Dynasty (770–221 BC) may be responsible for an increased abundance of rice (Oryza sp.), and possibly also foxtail or Chinese millet (Setaria italica), detected in the Guangfulin record after c. 2400 yr BP. An abrupt sedimentary change at c. 4000 yr BP may represent a short-lived episode of catchment instability. Aside from this, the sediment record from Guangfulin contains no evidence of dramatic environmental changes that could have led to a major decline in agricultural productivity c. 4000 yr BP, as has been suggested for the lower Yangtze by some researchers, who associate this with the cultural transition from Liangzhu to Maqiao. The findings do, however, add weight to the argument that developments in rice-based agriculture on the Yangtze delta varied both spatially and temporally. Key words: Charcoal, Neolithic, phytoliths, pollen, rice, sponge spicules, Yangtze delta, Guangfulin, Lateglacial, Holocene, China.

Introduction The Holocene evolution of the Yangtze delta in eastern China has long been a focus for sediment-based research (eg, Chen and Stanley, 1995; Chen et al., 1997). An extensive coring programme through the Quaternary sediments of the region shows sequences up to 300 m thick in places, with the Holocene represented variably from a few metres up to 30 m, and has begun to reveal a very *Author for correspondence (e-mail: [email protected])

dynamic late Quaternary history. During the last ice age the palaeo-Yangtze estuary was located some hundreds of kilometres east of the present coast (Elvin and Su, 1998). Rising sea-level during the Lateglacial brought the coastline closer to its present location, although it was still over 100 km farther east by c. 12 000 yr BP (Chen et al., 2000; Liu et al., 2000) and what is now the delta was part of a wide coastal plain drained by a network of incised river channels (Stanley and Chen, 1996). According to Stanley and Chen (1996), the delta developed as a half saucershaped feature during the early to mid Holocene, with a zone of

© 2007 SAGE Publications

10.1177/0959683607085128 Downloaded from hol.sagepub.com at NATIONAL UNIV SINGAPORE on February 20, 2015

1222

The Holocene 17,8 (2007)

RUSSIA KHASAKSTAN

MONGOLIA

Ya ng tze

SOUTH KOREA

Beijling

CHINA

Study Area

Shanghai

INDIA

Ri

ve r

TAIWAN

MIANMAR

LAOS

31°30’N

PHILIPPINES

7

Lake

1

3

2

Taihu 9

Shanghai

8 12

Qingpu 5 Guangfulin 13 10

4 11

0 120˚E

10

20

30

40

50 km

Songjiang 14

31°N

13

121˚E

122˚E

Figure 1 Location of Guangfulin palaeoecological site and selected archaeological sites on the Yangtze delta. The dotted line indicates the approximate palaeoshoreline c. 6000 BP and the solid lines are chenier ridges laid down during delta progradation (Stanley et al., 1999; Hori et al., 2002; Chen et al., 2005). The Guangfulin archaeological site is shown as number five and the Qingpu site of an earlier study is also shown (ItzsteinDavey et al., 2007a). Known archaeological sites within approximately 25 km of the site are shown (see Stanley et al. (1999) and Chen et al. (2005) for additional information). Sites one and two were inhabited during the Majiabang period, sites one to six were inhabited during the Songze period, sites one, four, and seven to eleven were inhabited during the Liangzhu period and sites 10–14 were inhabited during the Maqiao. The archaeological site at Hemudu is located just to the south of the area covered by the figure

relatively high levels of deposition along its seaward margins bounding a lower-lying central area, influenced by inundation and slowly infilling with sediment. Allied investigations of archaeology, palaeoecology and palaeoclimatology indicate that postglacial warming led to a midHolocene climatic optimum, when temperatures were 2–4°C warmer than present (Wang and Gong, 2000; Yi et al., 2003; Chen et al., 2005). At this time edaphic factors probably limited extensive forest development on the delta. Published pollen records indicate that vegetation in low-lying parts of the delta was largely characterized by wetland (including halophytic) communities; on the better drained topographic highs and inland areas, the midHolocene climatic optimum appears to have been associated with the replacement of temperate deciduous-conifer forests by subtropical taxa (eg, Liu et al., 1992; Chen and Chen, 1996; Chen et al., 1997; Yi et al., 2003). This replacement is said to have been reversed by the onset of cooler and drier climates from c. 4000 cal. yr BP (Sun and Chen, 1991; Yi et al., 2003). The archaeological record indicates that the Yangtze delta was settled by humans from c. 7000 cal. yr BP (Yu et al., 2000). This is a relatively late date for China and, if accurate, could be a reflection of the challenges posed by highly dynamic local environmental conditions. Human settlements, once established, have given rise to one of the world’s highest concentrations of prehistoric sites (Chang, 1986; Wu, 1988). These fall into four distinct cultural phases, termed the Majiabang, Songze, Liangzhu and Maqiao. Majiabang (c. 7500–5900 cal. yr BP) occupation sites are located close to areas of surface water centred in the Lake Taihu region and have yielded abundant evidence of the use of a wide range of aquatic plants, including the two subspecies of Asian domesticated rice (Oryza sativa indica (Xia rice) and O. sativa japonica (Keng rice)). According to Zou et al. (1998), Majiabang agriculture had a three-stage development, with the later developmental

stages associated with irrigation in response to climatically driven reductions in the height of the water-table and the extension of paddy fields beyond the area that flooded naturally. The archaeological record suggests that the succeeding phase, the Songze (c. 5900–5200 cal. yr BP), developed from the Majiabang but was characterized by social stratification and peaks in ceramic complexity and decorative sophistication (Chang, 1986; Shao 2005). Environmental and climatic conditions during the mid Holocene, in tandem with technological developments, may have facilitated increased cultural complexity and social stratification by allowing agricultural surpluses that could be used to support non-agricultural tiers in society. A trend towards increased social stratification together with improved technologies, including those relating to agriculture, continued into the late Neolithic (Chalcolithic) Liangzhu culture (c. 5200–4200 cal. yr BP) (Chang, 1986; Shao, 2005). Artefacts from Liangzhu occupation sites provide evidence of advanced agriculture (Cao et al., 2006), combining the cultivation of a range of plants (including Keng and Xia rice, melon, peach and water-caltrop) with domesticated animals, such as water buffalo and sheep, a pronounced social structure (involving elaborate funerals, human sacrifice and the iconography of power) and the ritual use of high quality worked jades (Chang, 1986). In addition to distinct cultural and technological changes, the Neolithic in the lower Yangtze was also characterized by alterations in the distribution of settlement as a result, presumably, of a complex and dynamic combination of environmental factors, including relative sea level, climate and topography (Stanley and Chen, 1996; Stanley et al., 1999; Yu et al., 2000). A significant shift is claimed to have occurred around 4000 cal. yr BP, marked by reduced per capita food production, urbanization and incessant warfare between numerous clan-based states (Chen, 1997; Jin and Liu, 2002). This shift, associated with the abrupt termination of the Neolithic Liangzhu and the beginning of the Bronze Age

Downloaded from hol.sagepub.com at NATIONAL UNIV SINGAPORE on February 20, 2015

Freea Itzstein-Davey et al.: Lateglacial and Holocene environmental changes, Yangtze delta

Maqiao (c. 3900–3200 cal. yr BP) societies, may have been in part driven by the challenges posed to food production by climatic change (Chen et al., 2005; Tao et al., 2006; Wang et al., 2006). The research presented in this paper, based on a sequence of sediments from Guangfulin within close proximity to an on-going archaeological excavation, provides a reconstruction of environmental changes on the Yangtze delta from c. 12 400 yr BP. The reconstruction is discussed within the context of available archaeological and palaeoecological evidence, including recently published data from a similar study at Qingpu (Atahan et al., 2007; Itzstein-Davey et al., 2007a, b).

Location, archaeological history and previous research Guangfulin is located about 40 km to the southwest of Shanghai (Figure 1). Agriculture, notably the cultivation of rice, vegetables and fruit, is the predominant land use around the sample site. The sample site at Guangfulin is set within an area rich in archaeological sites, although most have not been excavated (Zhang, 2001), and the available evidence suggests the area was first settled by humans by at least c. 5000 cal. yr BP (Chen, 2002; Zhang et al., 2003; Li et al., 2006). An archaeological excavation at Guangfulin covering 546 m2 (Cheng et al., 2006), has yielded artefacts from the Neolithic (Liangzhu) to the Eastern Zhou Dynasty (770–221 BC) (Zhang et al., 2003). Some of the pottery unearthed, dating to c. 4300 cal. yr BP and of a style not previously recovered from the lower Yangtze, is of particular interest as its presence at Guangfulin appears to indicate strong cultural and/or trade links with northern China (Cheng et al., 2006). Results of preliminary analyses of pollen and phytolith remains in material from the archaeological excavation are available. The published pollen record covers, on the basis of four radiocarbon dates, the last c. 5300 cal. yr BP (Chen, 2002; Li et al., 2006). The assemblages are dominated by Poaceae, with a small arboreal component. The phytolith remains attest to the local presence of rice since the early–mid Liangzhu Period (Zhang et al., 2002, 2003), although dating control, in the form of a single radiocarbon date, is weak.

Methodology This paper is based on sediment samples obtained from the exposed face of a trench (31°3.870′ N, 121°11.500′ E, 6 m above mean sea level) excavated to a depth of 191 cm and approximately 200 m south of the main archaeological excavation at Guangfulin. Sediment samples were collected as 2 cm thick slices from the freshly cleaned exposed face. Sediment slices were collected every 5 cm between 191 and 144 cm depth, and contiguously from 140 to 40 cm. The uppermost 40 cm of the profile was not sampled, because of disturbance. A total of 53 samples were analysed for their lithostratigraphic (magnetic susceptibility, particle size and stratigraphy) and charcoal (macro- and micro-), phytoliths and pollen contents. Sponge spicules were also recorded. Pottery fragments were scattered throughout the profile exposed in the trench. Eight samples from the exposed profile were submitted for AMS 14C dating. Five of these were pollen residues: residue preparation used 10% NaOH to remove humic colloids; 15% HCl to remove carbonates; 40% HF to remove silicates and sieving through a 5 µm mesh to remove the fine fraction. The three other samples for AMS 14C dating comprised macrofossils and were pre-treated using 30% HCl and 10% NaOH. All samples were sent to The Institute of Geological and Nuclear Sciences, New Zealand, for dating. Uncalibrated radiocarbon dates are presented here as ‘yr BP’; calibrated dates are assigned ‘cal. yr BP’.

1223

The magnetic susceptibility of 10 g of sediment, oven dried at 38°C, was determined using a Bartington MS2 meter with a sensitivity of 0.1 × 10−8 SI. Samples for particle size analysis were pre-treated with 15% H2O2, 12% HCl and 5% Calgon solution. Measurement was conducted on a Beckman Coulter LS 230 counter, with a measuring range of 0.04–2000 µm. Processing of samples for pollen and microcharcoal analyses followed standard preparation techniques as outlined in Moore et al. (1991). A sediment sample of 1 cm3 was treated with 10% NaOH to remove humic acids, 15% HCl to remove calcareous minerals and 40% HF to remove silicates and acetolysis to remove unwanted organic matter. Owing to low pollen concentrations towards the base of the trench, an additional 2 cm3 was processed for samples at 149, 154 and 184 cm depth and 3 cm3 for samples at 159, 169, 179 and 189 cm. Lycopodium spores were added prior to processing to estimate pollen and microcharcoal concentrations. Pollen residues were mounted in silicon oil and scanned under an Olympus Nikon microscope at 400× magnification. Pollen counts are expressed as percentages of the total pollen sum, which was 300 in all samples except two. Samples from 164 and 174 cm in the profile had pollen sums of 241 and 236 grains, respectively. Microcharcoal (5–150 µm) was quantified using the point-count method (Clark, 1982) and presented as concentrations (cm2/cm3). Samples of 1 cm3 were prepared for macrocharcoal (>150 µm) through gentle disaggregation in 5% Calgon solution and sieving through a 150 µm mesh. All charcoal particles larger than 150 µm were counted under a stereomicroscope at 20× magnification. Pollen identification was made with reference to Wang et al. (1995). The majority of pollen grains was assigned to family level and, where possible, to genus. Poaceae pollen was divided into two size categories (< 40 µm and > 40 µm). Poaceae grains larger than 40 µm have previously been interpreted to represent cereals (Dickson, 1988) and pollen from cultivated rice (Oryza sativa) has been reported to be present in this size category (see Wang et al., 1995; Chatuvedi et al., 1998). It is acknowledged that the use of a size threshold to distinguish Poaceae pollen produced by cereals is not entirely satisfactory (see Maloney, 1990; Tweddle, et al., 2005). In the current context, however, size criteria are used in conjunction with other forms of evidence, notably other pollen types, charcoal and phytolith remains, to reconstruct past agricultural activities, including the cultivation of crops. Quercus pollen was also separated into two size categories as suggested by Chang and Wang (1986): grains with the longest axis > 30 µm were classified as ‘Quercus (deciduous comp.)’, and grains smaller than < 30 µm as ‘Quercus (evergreen comp.)’. It is acknowledged, however, that this method does not conclusively distinguish evergreen from deciduous species (see Liu et al., 2007). Difficulties were encountered in distinguishing Moraceae from Urticaceae pollen, and thus the two pollen types have been combined in the current work. Morus is likely to have been a common component of vegetation locally because of its role in silk production. Based on this and on published information (eg, Punt and Malotaux, 1984) and available type material, it is possible that Morus was the source of Moraceae/Urticaceae-type pollen found in sediments from Guangfulin. Phytoliths were extracted from 1 g crushed air-dried sediment samples using HCl to remove carbonates, agitation followed by settling to separate clays and firing in a muffle furnace to remove organics. The phytolith fraction was finally separated from the remaining inorganics using density (heavy liquid, sodium polytungstate) separation. Phytolith counts were conducted under light microscopy using a Meiji Techno Co. Ltd. ML5000 series laboratory microscope at 400× magnification until a total of 400 single-celled phytoliths were noted. If ten or more multicell morphotypes were encountered in the 400 single-celled count, then a count of multicelled morphotypes was continued until 50 multicelled types were recorded. Phytolith results are presented as percentages of the total sum.

Downloaded from hol.sagepub.com at NATIONAL UNIV SINGAPORE on February 20, 2015

1224

The Holocene 17,8 (2007)

Figure 2

Stratigraphy, mean particle size, magnetic susceptibility and pollen concentration for the trench at Guangfulin

Phytolith morphotypes were identified according to Wang and Lu (1993) following nomenclature of ICPN Working Group et al. (2005). Other key references utilized include Piperno (1988), Bozarth (1992), Rosen (1992), Runge (1999) and Lu et al. (2006). A list of phytolith types identified in this research, and their botanical affinities, is provided in ItzsteinDavey et al. (2007a). In particular, rice (Oryza sp.) phytoliths

were studied in detail. Rice species can be identified by four single-cell phytolith morphotypes: cuneiform (fan-shaped) bulliforms; bilobate (dumbbell) short cells, ‘bumpy’ long cells and double-peaked glume phytoliths (Piperno, 2006). Multicell phytoliths originating in the husks of Oryza were also identified (Jiang, 1995; Lu et al., 1997; Itzstein-Davey et al., 2007a, b). Unfortunately the total number of Oryza-type phytoliths

Table 1 AMS 14C dates for sediment samples from the trench at Guangfulin Calibrated dates are determined from the calibration curve IntCal4 (Reimer et al., 2004) using the program OxCal v 4.0.1 (Bronk Ramsey, 2001) Depth (cm)

Laboratory code

Age (14C yr BP)

62–64 70–72 88–90 124–126 138–140 154–156 174–176 179–181

NZA 26016 NZA 26011 NZA 26017 NZA 26012 NZA 26013 NZA 26014 NZA 26264 NZA 26015

945 ± 30 2057 ± 30 2453 ± 30 6209 ± 30 6375 ± 30 12218 ± 45 5517 ± 55 12366 ± 55

Calibrated date (95.4% prob.) 1025–AD 1158 170 BC–AD 16 753 BC–411 BC 5295 BC–5056 BC 5468 BC–5306 BC 12256 BC–12000 BC 4461 BC–4259 BC 12796 BC–12133 BC AD

Downloaded from hol.sagepub.com at NATIONAL UNIV SINGAPORE on February 20, 2015

Dated material wood fragment pollen residue charcoal fragment pollen residue pollen residue pollen residue macroscopic charcoal fragments pollen residue

Freea Itzstein-Davey et al.: Lateglacial and Holocene environmental changes, Yangtze delta

1225

encountered was too small to permit a statistically meaningful separation into likely-wild and likely-domesticated rice morphotypes (see Itzstein-Davey et al., 2007b for an example of an application of this separation technique to material from the lower Yangtze). C2 version 1.4.2 software (Juggins, 2003) was used to construct all stratigraphic diagrams. Zones were determined from variations in the remains of pollen, phytoliths and charcoal using CONISS software in TGView (Grimm, 1987, 1992).

Results Stratigraphy and sediment analysis Several stratigraphic layers were visible in the exposed trench wall (Figure 2). The lowermost layer, 191–130 cm, was composed of homogenous light grey clay with numerous plant fragments. Mean particle size and magnetic susceptibility values were generally greater in this layer than in overlying sediment. Pollen concentrations were low (less than 10 000 grains/cm3) but increased from 140 cm. Between 130 and 105 cm a darker grey silty clay horizon was present, containing green mottles and some plant fragments. Pollen concentrations were greater than the underlying layer. At 105–103 cm a rounded stone layer with fine gravel fragments was noted, while between 103 and 88 cm the sediment was a dark grey silty clay, rich in pottery and containing abundant plant and charcoal fragments. A peak in magnetic susceptibility at 96 cm is evident and coincides with relatively large mean sediment particle sizes (Figure 2). An abrupt transition is noted at 88 cm into a dark grey, silty clay containing plant and, occasionally, pottery fragments and large stones. Pollen concentrations were very high in this layer, with values exceeding 600 000 grains/cm3, while mean particle size and magnetic susceptibility values were generally low compared with underlying layers. A diffuse boundary, from 50 to 40 cm, separated the uppermost, clearly disturbed, cultivation layer from underlying sediments.

AMS 14C dating AMS 14C dating reveals that the age of sediment samples from Guangfulin trench ranges from c. 12 400 to c. 400 yr BP (Table 1 and Figure 2). An age reversal occurs in the lower sediments (the sample at 174 cm is younger than three overlying dates and has been rejected as a result). Deltaic sediments can be difficult to date by radiocarbon techniques (Stanley and Chen, 2000; Stanley and Hait, 2000; Itzstein-Davey et al., 2007a, b) and the anomalously young date from Guangfulin could result from contamination resulting from the incorporation of younger carbon. The age–depth curve based on the remaining seven AMS 14C dates correlates well with the chronology from the archaeological excavation at Guangfulin (Chen, 2002; Li et al., 2006) and indicates two main accumulation phases, separated by either a period of slow sediment accumulation or, perhaps most likely given the sedimentary setting, a hiatus that may account for the Lateglacial–Holocene transition and much of the early Holocene to c. 7400 yr BP. Inferred ages are calculated for the current research, based on two least square regression lines. These indicate a rate of sedimentation of 1.7 mm/yr below 154 cm and 0.13 mm/yr above 138 cm.

Charcoal, phytoliths and pollen analyses Microfossil records for Guangfulin are presented in Figures 3–6. The preservation of phytoliths varied throughout the profile, being poorest at the base. Phytolith numbers were also low, presumably as a result of poor preservation, around 120 cm. Low concentrations of pollen occurred towards the base of the trench section. The microfossil data have been divided into four zones (GFL 1–4). A Pearson’s correlation calculation indicates a significant

Figure 3

Pollen and charcoal record for the trench at Guangfulin

Downloaded from hol.sagepub.com at NATIONAL UNIV SINGAPORE on February 20, 2015

1226

The Holocene 17,8 (2007)

C3

Oryza sp.

C4

Lo

ng Lo ce ng ll e Tr ce chi ap ll r na ez od te ifo s rm Bu Sq lif ua o re Bu rm lif cu o n Bu rm e lif cu ifor or n m m ei ro for sm H ai un m l r de lg pe cel d e ak l e Tr d ap ep ez ide ifo rm rm is R R on ec de ta ng l le G lo bu Bi lar lo ba ech te in at e Sa dd C le ro s D s ic o Bu t tra m ch R py eat ic lo e e n s do bu g c ub lifo ell le rm -p ea typ ke e d R ic e hu s

In fe rr D ed ep ag Lo th e ng (c (ye ce m) ars ll BP sm ) oo th

k

Poaceae C3/C4

54 64

1400

GFL4

74 2400

84 94

3400

GFL3

104

4400

114 5400

GFL2

124

6400

134

7400

144

11400

154 164

GFL1

174 184

12400

20

0

Figure 4

40

60

0

0

0

20

0

0

0

20 0

0

0

20

0

20 0

0

20

0

0

0

0

0

0

Single-celled phytolith record of Poaceae types for the trench at Guangfulin

co m b t Bi rac e ty lo p ba hea e” te te of s As Bu te ra lif ce or ae C m ro ss Lo ng un cell id s en mo tif ot ie d h un g id ra en ss tif 1 ie d un gr id as en s tif 2 ie d D h ia us to k m s Sp on ge sp ic Sp ul on es ge -c sp om ic pl ul et es e -b ro ke n

ey

ic

on

D

“H

ot

pe ty ria ta Se

D

In

fe

rre

d ep age ( O th ( ry cm yea za ) rs BP sp . )

outside count

54 64

1400

GFL4

74 2400

84

3400 4400

94

GFL3

104 114

5400

124

6400

134

7400

144

11400

154

GFL2

GFL1

164 174 12400

184 0

Figure 5

20

40 0

20

0

0

0

20 0

0

0

0

20 0

20 0

20 0

20 0

20 0

25 50 75 100 125

Multicell phytolith record for the trench at Guangfulin

positive correlation between levels of micro- and macrocharcoal in the same level (r = 0.857, r2 = 0.734, p = 0.0).

GFL 1 (191–140 cm, c. 12 400–7000 yr BP) This zone includes three AMS 14C dates: 12 366 ± 55 yr BP (179–181 cm); 12 218 ± 45 yr BP (154–156 cm) and what appears to be the erroneous date of 5517 ± 55 yr BP (174–176 cm) (Table 1).

The upper boundary for the zone coincides with a fourth date: 6375 ± 30 yr BP (138–140 cm). GFL 1 is characterized by pollen from arboreal taxa (Figure 3), in particular Pinus and Quercus, by phytoliths mainly assignable to C3 grasses (a few C4-type grass phytoliths were also recorded) (Figure 4) and by sponge spicules (Figure 5). Pollen from Betula, Castanopsis/Castanea, Juglans, Salix and Tsuga were also present. The vast majority of Quercus

Downloaded from hol.sagepub.com at NATIONAL UNIV SINGAPORE on February 20, 2015

Freea Itzstein-Davey et al.: Lateglacial and Holocene environmental changes, Yangtze delta

Figure 6

1227

Simplified combined pollen, charcoal and phytolith record for the trench at Guangfulin

CULTURAL RECORD

GUANGFULIN PALAEOENVIRONMENTAL RECORD

Han (2156 – 1730 BP)

GFL 4: “Human-dominated” Grasses & cereal-type pollen dominant. Low, small-scale burning. Rice & both C4/C3 phytoliths.

1500 2000 2500

cal. yr BP

3000

Qin (2171 – 2157 BP) Eastern Zhou (2720 – 2171 BP) Western Zhou (2977 – 2721 BP) Shang (3650 – 2977 BP)

3500

Maqiao (3900 – 3200 cal. yr BP)

4000

Cultural Interruption

4500

Liangzhu (5200 – 4200 cal. yr BP)

5000 5500

Songze (5900 – 5200 cal. yr BP)

6000 6500

GFL 3: Rice agriculture Reduction in arboreal pollen, increase in grasses. High burning. Increase in C4 & reduction in C3 phytoliths, increase in rice. Peak magnetic susceptibility & PSA.

Majiabang (7500 – 5900 cal. yr BP)

GFL 2: Forest clearance – Early rice agriculture Increased pollen concentration, decreased arboreal pollen, increase in grasses. Low burning. C4 & some C3 phytoliths, some rice.

7000 7500

GFL 1: “Nature-dominated” wide coastal plain Quercus, Pinus, some Chenopodiaceae. C3 phytoliths.

12,000

Figure 7 A timeline showing Neolithic cultural periods, Ancient China and Early Imperial China dynasties (approximate age) with key environmental changes at Guangfulin trench site. The boundaries separating zones in the palaeoenvironmental record are based on uncalibrated radiocarbon dates.

Downloaded from hol.sagepub.com at NATIONAL UNIV SINGAPORE on February 20, 2015

1228

The Holocene 17,8 (2007)

pollen on the basis of size appears to be from evergreen taxa. Chenopodiaceae, Cyperaceae, Poaceae and Typha are prominent among non-arboreal pollen types, while fern spores are also relatively abundant. Very low numbers of phytoliths from Oryza sp. were present in three samples from this zone (189, 169 and 164 cm). The zone appears to contain at least one break (hiatus) in sedimentation from c. 11 400 to 7400 yr BP. The upper boundary of the zone is marked by a rapid reduction in Pinus pollen and, to a lesser extent, Chenopodiaceae and Quercus, and an increase in Poaceae and Typha pollen. Micro- and macrocharcoal concentrations are very low, averaging, respectively, 0.8 cm2/cm3 and 8.5 grains/cm3.

Zone GFL 2 (140–110 cm, c. 7000–4700 yr BP) This zone contains one AMS 14C date, in addition to 6375 ± 30 yr BP (138–140 cm): 6209 ± 30 yr BP (124–126 cm). The zone is characterized by high percentages of pollen from herbs (Artemisia, Cyperaceae, Poaceae and Typha) and greatly reduced levels of Pinus when compared with Zone GFL 1. Reduced percentages of Juglans, Quercus (deciduous comp.) and Tsuga pollen were noted. Arboreal pollen present includes Carpinus, Castanopsis/Castanea-type, Quercus (evergreen comp.) and Salix. In this zone, percentages of most pteridophyte spores are greatly reduced compared with GFL 1. Phytoliths from Oryza sp. are present in low numbers, and a Setaria-type morphotype was recorded. The remains of sponge spicules are more abundant than in the zone below, and concentrations of macro- and microcharcoal are higher (concentrations average, respectively, 2.7 cm2/cm3 and 134.1 grains/cm3).

Zone GFL 3 (110–80 cm, c. 4700–2400 yr BP) This zone contains a single AMS 14C date (2453 ± 30 yr BP (88–90 cm)) and is distinguished primarily by large increases in phytoliths from Oryza, particularly multicelled forms (Figure 5). It is also characterized by very high micro- and macrocharcoal concentrations when compared with GFL 2, with microcharcoal peaking at 25.8 cm2/cm3 at 100 cm and macrocharcoal peaking at 9692 grains/cm3 at 102 cm. Pollen from Artemisia, Poaceae and Typha remains common, and there are two distinct peaks in Moraceae/Urticaceae-type pollen (at 82 and 86 cm depths).

Zone GFL 4 (80–54 cm, c. 2400–400 yr BP) Sediments comprising Zone GFL 4 (80–54 cm, c. 2400–400 yr BP) were dated directly with two AMS 14C dates: 2057 ± 30 yr BP (70–72 cm) and 945 ± 30 yr BP (62–64 cm) (Table 1). Poaceae pollen is more abundant (percentages average 46.8%), although the number of grains > 40 µm remain largely unchanged when compared with GFL 3. Multicelled husks of Oryza sp. remain common and several Setaria-type phytoliths were recorded. Pollen from Typha is abundant, with Artemisia and Cyperaceae pollen also present but in lower abundances when compared with GFL 3. Pollen from the aquatic herb Myriophyllum is more abundant compared with GFL 3. Carpinus and Quercus, the latter most likely from evergreen forms, were commonly encountered pollen types. Micro- and macrocharcoal concentrations are substantially lower than in GFL 3.

Discussion According to the data presented here and summarized in Figures 6 and 7, conditions at Guangfulin during the Lateglacial were characterized by lower temperatures than present and more frequent inundation. Both forest and open, herbaceous vegetation occurred, with forests, including temperate elements such as Betula, Castanopsis/ Castanea, Pinus, Salix and Tsuga, presumably restricted to better

drained sites and higher altitudes in the wider catchment. The occurrence of pollen from Chenopodiaceae, Cyperaceae, Poaceae and Typha, together with the relative abundance of phytoliths from C3 (temperate) grasses, and abundant sponge spicules, indicate the presence of halophytic and other forms of open wetland habitat close to the sample site. Low charcoal abundances indicate that fires were infrequent events. Environments were thus ‘nature-dominated’ (in the sense of Messerli et al., 2000), with human activity possibly restricted to occasional visits by hunter-gatherers. Unfortunately the Lateglacial–Holocene transition and much of the early Holocene (c. 11 400–7400 yr BP) does not appear to be recorded in the profile of sediments exposed in the trench at Guangfulin. Vegetation is markedly different by c. 7000 yr BP: pollen from forest taxa, notably conifers, evergreen forms of Quercus and several temperate deciduous taxa, are far lower in abundance than earlier in the record, while pollen from more open types of vegetation (eg, Poaceae) is much more common. The contribution of C4 plants is also increased after this date. A major increase in Typha pollen, concomitant with a decline in Chenopodiaceae pollen, may indicate the expansion of freshwater habitats at the expense of more saline conditions. The occurrence of Oryza sp. phytoliths from c. 7000 yr BP and large sized (> 40 µm) pollen from Poaceae may represent the onset of incipient, rice-based agriculture in the study area during the Majiabang period. Burning linked to agriculture, including rice-based agriculture in particular (Cao et al., 2006), some distance from the study site would also explain the moderate increases in microcharcoal by c. 7000 yr BP. Major increases in both micro- and macrocharcoal from c. 4700 yr BP suggest an important shift in burning regime, and the occurrence of vegetation fires close to the sample site. The peak in magnetic susceptibility and particle sizes at 96 cm, dated c. 4000 yr BP, may represent a major erosive event, linked to catchment instability. However, the precise source of eroded material accumulating at the sample site is difficult to ascertain because of the nature of sedimentary environments associated with deltaic areas. Furthermore, an autochthonous origin for material producing the peak of magnetic susceptibility cannot be discounted. If the peak in magnetic susceptibility does indeed represent catchment instability in the form of deposition of eroded material, then the instability could be linked to events associated with the transition between the Liangzhu and Maqiao that are claimed by some to have led to widespread reductions in agricultural productivity and to have been driven by climate change (Chen, 1997; Stanley et al., 1999; Jin and Liu, 2002; Chen et al., 2005; Tao et al., 2006; Wang et al., 2006). This seems a little unlikely, however, as an increased abundance of rice phytoliths, together with other indicators of settled agriculture – such as Setaria-type phytoliths and large-sized Poaceae pollen – suggest a greater importance of agriculture locally, rather than its demise. Perhaps a more robust explanation is that the peaks in magnetic susceptibility and particle sizes, c. 4000 yr BP, result from erosion due to local agricultural activity (the construction of earthworks, for example) that had no longterm negative impact on agricultural productivity, or that the peaks are due to the incorporation in the sediment of small fragments of fired pottery (see Dearing, 1999). An increased abundance of Moraceae/Urticaceae-type pollen from c. 3000 yr BP may represent the cultivation of Morus, and the early production of silk close to the study site. Environments continued to be human-influenced post-c. 2400 yr BP. Rice continued to feature, although abundances of Oryza sp. phytoliths were low. A reduced abundance of Moraceae/Urticaceaetype pollen may also reflect a decline in the importance of sericulture. A change in fire regime, indicated by reduced levels of charcoal, could reflect changes in farming techniques, such as reduced burning of stubble (Cao et al., 2006). Alternatively,

Downloaded from hol.sagepub.com at NATIONAL UNIV SINGAPORE on February 20, 2015

Freea Itzstein-Davey et al.: Lateglacial and Holocene environmental changes, Yangtze delta

reduced burning could reflect either a more general decline in human activity, brought about by increased inundation (the remains of sponge spicules increased in abundance from c. 2400 yr BP, as did pollen from Myriophyllum and Typha), or the shortage of combustible material owing to the almost complete eradication of forests. The results presented above are in broad agreement with the less finely resolved findings of Zhang et al. (2003) and Li et al. (2006), who reported an expansion of rice agriculture at Guangfulin, based on pollen and phytolith remains, during the early–mid Liangzhu Period. These previous records, however, based on sediment taken from the archaeological excavation, found no earlier indication of rice agriculture. In contrast, pollen and phytolith evidence presented here indicate minor amounts of rice agriculture may have occurred in the area from as early as c. 7000 yr BP. Furthermore, Li et al. (2006) also suggested a slight cooling and drying of climate during the late Liangzhu, based on increased Artemisia and Chenopodiaceae and decreased Quercus pollen. There is, however, no indication in the more finely resolved and better dated pollen data presented here of such a change. An abrupt event may have impacted the study site c. 4000 yr BP, however, marked in the sediment record analysed as a sudden and short-lived pulse of material with an increased magnetic susceptibility and particle size. If this abrupt change in sediments represents an erosive event, then the driver or drivers of this event have yet to be identified, as it does not appear to be reflected in the charcoal, phytoliths or pollen data from the trench at Guangfulin. Agriculture, including rice, appears to have greatly expanded during the Liangzhu period at Guangfulin, with the changes possibly reflecting technological advancements and increased population levels. Prior to this agricultural expansion, inhabitants of the lower Yangtze area could have practiced cultivation while continuing to rely in part on hunting and gathering (Lu, 2006). Rice appears to have been a component of food production at Guangfulin before its occurrence at Qingpu (Itzstein-Davey et al., 2007a, b) and may have been introduced around the same time as at Hemudu, where it dates to possibly as early as c. 7000 cal. yr BP (Huang et al., 2002; Fuller et al., 2007) (both Qingpu and Hemudu are located on the Yangtze delta). Single-cell phytoliths from Oryza sp. are recorded in much lower abundances at Guangfulin when compared with Qingpu (Itzstein-Davey et al., 2007a, b). However, Oryza husks are much more abundant, particularly from c. 5000 yr BP, and it is possible that the sample site at Guangfulin, given the evidence of occupation close by, was associated with the processing of rice, once gathered from the fields, prior to storage and consumption. Dating of rice remains from Guangfulin, when viewed in the context of other evidence for rice-based agriculture, appears to add further support to the notion that rice production on the Yangtze delta was patchy and strongly influenced by local environmental conditions, notably topography and frequency of inundation (Itzstein-Davey et al., 2007a, b).

Conclusion The multiproxy record from Guangfulin trench provides new insight into human activity and environmental dynamics on the Yangtze delta from c. 12 400 yr BP. A mosaic of forest and wetlands dominated the region during the Lateglacial. Reduced forest cover by c. 7000 yr BP coincides with the first appearances of possible agricultural indicators, particularly the remains of Oryza sp., and a change in fire regime, while less frequent inundation may have made the area more attractive for human settlement and agriculture. The burning of vegetation seems to have become more prominent from c. 4700 yr BP, and may have approached the study site. Reduced burning after c. 2400 yr BP could reflect

1229

changes in farming techniques, such as reduced burning of stubble (Cao et al., 2006), associated with the Eastern Zhou Dynasty. Alternatively, reduced burning could reflect either a more general decline in human activity in the study area, brought about by increased inundation, or deforestation and the consequent shortage of combustible material. Results of the research described here add to the growing body of evidence on the interactions between humans and their environment during prehistoric times in a part of the world associated with both very high population levels and rapidly changing environmental conditions. In particular, results from the research add further weight to the argument that rice-based agriculture on the Yangtze delta has varied both spatially and temporally in accordance with changing human cultural and environmental conditions.

Acknowledgements This research would not have been possible without the support of the Australian Research Council, the Irish Research Council for Science, Engineering and Technology and the Ministry of Education of China, and this support is gratefully acknowledged. Thanks are also due to Jungan Qin for fieldwork assistance and extensive support in China and to Jungan Qin and Rui Liu for translating substantial sections of Chen Jie’s PhD thesis. Xiaqiang Li is thanked for his assistance with pollen identification. We are grateful to Arlene Rosen and Houyuan Lu for help with phytolith identifications. Thanks also to Bill Wilson and Lorraine Wilson for laboratory support, and to two anonymous reviewers for their comments on an earlier draft of this paper.

References Atahan, P., Grice, K. and Dodson, J. 2007: Agriculture and environmental change at Qingpu, Yangtze delta region: a biomarker, stable isotope and palynological approach. The Holocene 17, 507–15. Bozarth, S.R. 1992: Classification of opal phytoliths formed in selected dicotyledons native to the Great Plains. In Rapp, G.J. and Mulholland, S.C., editors, Phytolith systematics. Plenum Press, 193–214. Bronk Ramsey, C. 2001: Development of the radiocarbon program OxCal. Radiocarbon 43, 355–63. Cao, Z., Ding, J., Hu, Z., Knicker, H., Kögel-Knabner, I., Yang, L., Yin, R., Lin, X. and Dong, Y. 2006: Ancient paddy soils from the Neolithic age in China’s Yangtze River Delta. Naturwissenschaften 93, 232–36. Chang, K.-C. 1986: The archaeology of ancient China. Yale University Press. Chang, K.-T. and Wang, P.-L. 1986: Pollen morphology of Quercus L. in China. Acta Phytotaxonomica Sinica 24, 362–69 (in Chinese). Chatuvedi, M., Datta, K. and Nair, P.K.K. 1998: Pollen morphology of Oryza (Poaceae). Grana 37, 79–86. Chen, J. 2002: Culture and paleoenvironment during the Neolithic of the Yangtze Delta. PhD thesis, East China Normal University (in Chinese). Chen, S. 1997: Developmental history of paddy rice and arid agriculture in ancient Chinese civilization. Agricultural Archaeology 86, 51–53. Chen, Z. and Chen, M. 1996: Preliminary study of Quaternary stratigraphy and palynology in the Yangtze delta, eastern China. Journal of the Geological Society of China 39, 59–72. Chen, Z. and Stanley, D.J. 1995: Quaternary subsidence and river channel migration in the Yangtze Delta plain, eastern China. Journal of Coastal Research 11, 927–45. Chen, Z., Chen, Z. and Zhang, W. 1997: Quaternary stratigraphy and trace-element indices of the Yangtze Delta, eastern China, with special reference to marine transgressions. Quaternary Research 47, 181–91. Chen, Z., Song, B., Wang, Z. and Cai, Y. 2000: Late Quaternary evolution of the sub-aqueous Yangtze Delta, China: sedimentation,

Downloaded from hol.sagepub.com at NATIONAL UNIV SINGAPORE on February 20, 2015

1230

The Holocene 17,8 (2007)

stratigraphy, palynology, and deformation. Marine Geology 162, 423–41. Chen, Z., Wang, Z., Schneiderman, J., Tao, J. and Cai, Y. 2005: Holocene climate fluctuations in the Yangtze delta of eastern China and the Neolithic response. The Holocene 15, 915–24. Cheng, H.S., Zhang, Z.Q., Song, J., Gao, M.H., Zhu, D., Lin, J.W. and Feng, S.L. 2006: PIXE study on ancient pottery from Chinese Shanghai area. Nuclear Instruments and Methods in Physics Research Section B: Beam Interactions with Materials and Atoms 249, 601–603. Clark, R.L. 1982: Point count estimation of charcoal in pollen preparations and thin sections of sediments. Pollen et Spores 24, 523–35. Dearing, J.A. 1999: Holocene environmental change from magnetic proxies in lake sediments. In Maher, B.A. and Thompson, R., editors, Quaternary climates, environments and magnetism. Cambridge University Press, 231–78. Dickson, C. 1988: Distinguishing cereal from wild grass pollen: some limitations. Circaea 5, 67–71. Elvin, M. and Su, N. 1998: Action at a distance: the influence of the Yellow River on Hanzhou Bay since A.D. 1000. In Elvin, M. and Liu, T.U.-J., editors, Sediments of time: environment and society in Chinese history. Cambridge University Press, 411–46. Fuller, D.Q., Harvey, E. and Qin, L. 2007: Presumed domestication? Evidence for wild rice cultivation and domestication in the 5th millennium BC of the Lower Yangtze region. Antiquity 81, 316–31. Grimm, E.C. 1987: CONISS: A Fortran 77 program for stratigraphically constrained cluster analysis by the method of incremental sum of squares. Computers and Geosciences 13, 13–15. ––— 1992: Tilia software. Illinois State Museum. Hori, K., Saito, Y., Zhao, Q. and Wang, P. 2002: Architecture and evolution of the tide-dominated Changjiang (Yangtze) River delta, China. Sedimentary Geology 146, 249–64. Huang, W., Wang, L., Zhao, X., Xiong, J. and Yao, X. 2002: Gems of the Hemudu culture. Museum of the Hemudu site. Cultural Relics Publishing House. ICPN Working Group, Madella, M., Alexandre, A. and Ball, T. 2005: International code for phytolith nomenclature 1.0. Annals of Botany 96, 253–60. Itzstein-Davey, F., Atahan, P., Dodson, J.R., Taylor, D. and Zheng, H. 2007a: Environmental and cultural changes during the terminal Neolithic: Qingpu, Yangtze delta, eastern China. The Holocene in press. Itzstein-Davey, F., Taylor, D., Dodson, J., Zheng, H. and Atahan, P. 2007b: Wild and domesticated forms of rice (Oryza sp.) in early agriculture at Qingpu, lower Yangtze, China: evidence from phytoliths. Journal of Archaeological Science in press. Jiang, Q. 1995: Searching for evidence of early rice agriculture at prehistoric sites in China through phytolith analysis: an example from central China. Review of Palaeobotany and Palynology 89, 481–85. Jin, G.Y. and Liu, D.S. 2002: Mid-Holocene climate change in North China, and the effect on cultural development. Chinese Science Bulletin 47, 408–13. Juggins, S. 2003: User guide: C2, software for ecological and palaeoecological data analysis and visualization. University of Newcastle, 1–69. Li, C.-H., Chen, J. and Wang, W.-M. 2006: Pollen records from Guangfulin relics, Songjiang district of Shanghai. Acta Micropalaeontologica Sinica 23, 175–81 (in Chinese). Liu, K.-B., Sun, S. and Jiang, X. 1992: Environmental change in the Yangtze River delta since 12,000 years BP. Quaternary Research 38, 32–45. Liu, Y.-S., Zetter, R., Ferguson, D.K. and Mohr, B.A.R. 2007: Discriminating fossil evergreen and deciduous Quercus pollen: a case study from the Miocene of east China. Review of Palaeobotany and Palynology 145, 289–303. Liu, Z.-X., Berne, S., Saito, Y., Lericolais, G. and Marsset, T. 2000: Quaternary seismic stratigraphy and paleoenvironments on the continental shelf of the East China Sea. Journal of Asian Earth Sciences 18, 441–52. Lu, H., Wu, N. and Lu, B. 1997: Recognition of rice phytoliths. In Pinilla, A., Juan-Tresserras, J. and Machado, M.J., editors, Estado actual

de los estudios de fitolitos en suelos y plantas: the state-of-the-art phytoliths in soils and plants. Centro de Ciencas Medioambientales, 159–74. Lu, H.-Y., Wu, N.-Q., Yang, X.-D., Jiang, H., Liu, K.-B. and Liu, T.-S. 2006: Phytoliths as quantitative indicators for the reconstruction of past environmental conditions in China I: phytolith-based transfer functions. Quaternary Science Reviews 25, 945–59. Lu, T.D-L. 2006: The occurrence of cereal cultivation in China. Asian Perspectives 45, 129–58. Maloney, B.K. 1990: Grass pollen and the origins of rice agriculture in north Sumatra. Modern Quaternary Research in Southeast Asia 11, 135–60. Messerli, B., Grosjean, M., Hofer, T., Nunez, L. and Pfister, C. 2000: From nature-dominated to human-dominated environmental changes. Quaternary Science Reviews 19, 459–79. Moore, P.D., Webb, J.A. and Collinson, M.E. 1991: Pollen analysis. Blackwell Scientific Publications. Piperno, D.R. 1988: Phytolith analysis: an archaeological and geological perspective. Academic Press, Inc. ––— 2006: Phytoliths: a comprehensive guide for archaeologists and palaeoecologists. AltaMira Press. Punt, W. and Malotaux, M. 1984: Cannabaceae, Moraceae and Urticaceae. Review of Palaeobotany and Palynology 42, 23–44. Reimer, P.J., Baillie, M.G.L., Bard, E., Bayliss, A., Beck, J.W., Bertrand, C.J.H., Blackwell, P.G., Buck, C.E., Burr, G.S., Cutler, K.B., Damon, P.E., Edwards, R.L., Fairbanks, R.G., Freidrich, M., Guilderson, T.P., Hogg, A.G., Hughen, K.A., Dromer, B., McCormac, G., Manning, S., Ramsey, C.B., Reimer, R.W., Remmele, S., Southon, J.R., Stuiver, M., Talamo, S., Taylor, F.W., van der Plicht, J. and Weyhenmeyer, C.E. 2004: IntCal04 terrestrial radiocarbon age calibration, 0–26 cal kyr BP. Radiocarbon 46, 1029–58. Rosen, A.M. 1992: Preliminary identification of silica skeletons from near eastern archaeological sites: an anatomical approach. In Rapp, G.J. and Mulholland, S.C., editor, Phytolith systematics. Plenum Press, 125–32. Runge, F. 1999: The opal phytolith inventory of soils in central Africa – quantities, shapes, classification, and spectra. Review of Palaeobotany and Palynology 107, 23–53. Shao, W. 2005: The formation of civilization: the interaction sphere of the Longshan period. In Allen, S., editor, The formation of Chinese civilization: an archaeological perspective. Yale University and New World Press, 85–124. Stanley, D.J. and Chen, Z. 1996: Neolithic settlement distributions as a function of sea level-controlled topography in the Yangtze Delta, China. Geology 24, 1083–86. ––— 2000: Radiocarbon dates in China’s Holocene Yangtze delta: record of sediment storage and reworking, not timing of deposition. Journal of Coastal Research 16, 1126–32. Stanley, D.J. and Hait, A.K. 2000: Deltas, radiocarbon dating, and measurements of sediment storage and subsidence. Geology 28, 295–98. Stanley, D.J., Chen, Z. and Song, J. 1999: Inundation, sea-level rise and transition from Neolithic to Bronze Age cultures, Yangtze Delta, China. Geoarchaeology 14, 15–26. Sun, X. and Chen, Y. 1991: Palynological records of the last 11,000 years in China. Quaternary Science Reviews 10, 537–44. Tao, J., Chen, M.-T. and Xu, S. 2006: A Holocene environmental record from the southern Yangtze River delta, eastern China. Palaeogeography, Palaeoclimatology, Palaeoecology 230, 204–29. Tweddle, J.C., Edwards, K.J. and Fieller, N.R.J. 2005: Multivariate statistical and other approaches for the separation of cereal from wild Poaceae pollen using a large Holocene dataset. Vegetation History and Archaeobotany 14, 15–30. Wang, F., Chien, N., Zhang, Y. and Yang, H. 1995: Pollen flora of China (second edition). Science Press. Wang, S.W. and Gong, D.Y. 2000: Climate in China during the four special periods in the Holocene. Progress in Natural Science 10, 379–86. Wang, Y. and Lu, H. 1993: The study of phytolith and its application. China Ocean Press (in Chinese). Wang, Z.H., Chen, Z.Y. and Tao, J. 2006: Clay mineral analysis of sediments in the Changjiang delta plain and its application to the Late

Downloaded from hol.sagepub.com at NATIONAL UNIV SINGAPORE on February 20, 2015

Freea Itzstein-Davey et al.: Lateglacial and Holocene environmental changes, Yangtze delta

Quaternary variations of sea level and sediment provenance. Journal of Coastal Research 22, 683–91. Wu, J.M. 1988: The distribution and change of the circumstance of prehistoric sites in the Islet in Yangtze River. Culture of Southeast China 6, 16–36. Yi, S., Saito, Y., Zhao, Q. and Wang, P. 2003: Vegetation and climate changes in the Changjiang (Yangtze River) Delta, China, during the past 13,000 years inferred from pollen records. Quaternary Science Reviews 22, 1501–19. Yu, S., Zhu, C., Song, J. and Qu, W. 2000: Role of climate in the rise and fall of Neolithic cultures on the Yangtze delta. Boreas 29, 157–65. Zhang, Q. 2001: Ancient tombs tell tales. Shanghai Star. Retrieved 4 October 2007 from: http://app1.chinadaily.com.cn/star/2001/0222/ fo6-1.html.

1231

Zhang, Y., Song, J. and Lu, B. 2002: New discovery in archaeology and exploration of living environment of ancestors from Guangfulin relics in Shanghai. Journal of Tongji University 30, 1454–57 (in Chinese). Zhang, Y., Zhang, M. and Song, J. 2003: Development of ancestors’ cultivation revealed in phytolith assemblages from Guangfulin relics. Chinese Science Bulletin 43, 287–90. Zou, J., Tang, L. and Wang, C. 1998: On the origin of cultivated keng rice. Scientia Agricultural Sinica 31, 75–81.

Downloaded from hol.sagepub.com at NATIONAL UNIV SINGAPORE on February 20, 2015