558649 HOL0010.1177/0959683614558649The HoloceneKahn et al. research-article2014 Research paper Mid- to late Holocene landscape change and anthropogenic transformations on Mo‘orea, Society Islands: A multi-proxy approach The Holocene 2015, Vol. 25(2) 333–347 © The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/0959683614558649 hol.sagepub.com Jennifer G Kahn,1 Cordelia Nickelsen,2 Janelle Stevenson,3 Nick Porch,4 Emilie Dotte-Sarout,5 Carl C Christensen,6 Lauren May,7 J Stephen Athens8 and Patrick V Kirch9 Abstract Archaeology’s ability to generate long-term datasets of natural and human landscape change positions the discipline as an inter-disciplinary bridge between the social and natural sciences. Using a multi-proxy approach combining archaeological data with palaeoenvironmental indicators embedded in coastal sediments, we outline millennial timescales of lowland landscape evolution in the Society Islands. Geomorphic and cultural histories for four coastal zones on Mo‘orea are reconstructed based on stratigraphic records, sedimentology, pollen analysis, and radiocarbon determinations from mid- to late Holocene contexts. Prehuman records of the island’s flora and fauna are described utilizing landsnail, insect, and botanical data, providing a palaeo-backdrop for later anthropogenic change. Several environmental processes, including sea level change, island subsidence, and anthropogenic alterations, leading to changes in sedimentary budget have operated on Mo‘orea coastlines from c. 4600 to 200 BP. We document significant transformation of littoral and lowland zones which obscured earlier human activities and created significant changes in vegetation and other biota. Beginning as early as 440 BP (1416–1490 cal. AD), a major phase of sedimentary deposition commenced which can only be attributed to anthropogenic effects. At several sites, between 1.8 and 3.0 m of terrigenous sediments accumulated within a span of two to three centuries due to active slope erosion and deposition on the coastal flats. This phase correlates with the period of major inland expansion of Polynesian occupation and intensive agriculture on the island, indicated by the presence of charcoal throughout the sediments, including wood charcoal from several economically important tree species. Keywords charcoal, Eastern Polynesia, geomorphology, human-induced landscape change, landsnails, multi-proxy, plant macrofossils, pollen, Society Islands, subfossil insects Received 19 May 2014; revised manuscript accepted 9 September 2014 Introduction Environmental archaeology plays a key role in understanding long-term human-ecological impacts. Archaeology’s increasing ability to provide long-term datasets of natural and human landscape change has positioned the discipline as an inter-disciplinary bridge between the social and natural sciences (Kirch, 2005; Sandweiss and Kelley, 2012; Van der Leeuw and Redman, 2002). By utilizing multiple lines of proxy indicators of past environmental conditions and social processes embedded within archaeological sediments, archaeological studies have revealed decadal and millennial timescales of landscape system behavior (Cremaschi et al., 2014; Dearing, 2008; Lubos et al., 2013). Extending temporal timescales through palaeoenvironmental analyses provides a strong context within which to analyze long-term trends and temporal scales of climatic, landscape, and social change and to parse out causality, whether natural or anthropogenic (Dearing et al., 2012; Kelley, 2014; Kirch et al., 2004; Rosen and Rivera-Collazo, 2012; Van der Leeuw, 1998; Varien et al., 2007; Wilkinson et al., 2007). Furthermore, there is increased recognition that understanding environmental and climatic change in the present can benefit from analyses of past human–environment interactions (Fisher, 2005; Fisher and Feinman, 2005; Ford and Nigh, 2009). Here, we present an array of multi-proxy data for Mo‘orea Island in the Society Islands to outline its mid- to late Holocene environmental history over a period that spans prehuman and post-colonization phases. (Polynesian arrival is estimated to be no later than c. 950 BP.) Limited prior palaeoenvironmental and archaeological research on Mo‘orea suggests that stratified evidence for early Polynesian land use would be found within recent 1Department 2Oceanic of Anthropology, College of William & Mary, USA Archaeology Laboratory, University of California, Berkeley, USA 3SCHL, College of Asia and the Pacific, Australian National University, Australia 4School of Life and Environmental Sciences, Deakin University, Australia 5School of Social Sciences, University of Western Australia, Australia 6Department of Natural Sciences, Bishop Museum, USA 7Cultural Surveys Hawai’i, USA 8International Archaeological Research Institute, Inc., USA 9Departments of Anthropology and Integrative Biology, University of California, Berkeley, USA Corresponding author: Jennifer G Kahn, Department of Anthropology, College of William & Mary, Williamsburg, VA 23185, USA. Email: jgkahn01@wm.edu 334 The Holocene 25(2) (b) (a) Reef C B Pihaena Point D 342 Mo’orea 50 m tre am N Va iom aS 341 0 3 0 1 Kilometers Kilometers 150 (d) 50 ef (c) Re ’Opunohu Bay m 50 m m ea Str hu no pu ’O 50 343 0 15 0 250 m 0.5 0 0 25 Kilometers 346 0 15 0.5 Kilometers Figure 1. (a) Map of Mo’orea, with context of main research sites and transects, (b) close-up of site context and transects for ScMo-341, 342, (c) close-up of site context and transect for ScMo-343, (d) ScMo-346 trench. coastal alluvial and colluvial deposits (Green et al., 1967; Hamilton and Kahn, 2007; Kahn, 2012; Lepofsky, 1995; Lepofsky et al., 1992, 1996; Pirazzoli et al., 1985). Lepofsky et al. (1992) demonstrated significant effects of Polynesian land use in the ‘Opunohu Valley, including forest clearance and erosion deriving from intensive inland cultivation. In 2011, we sampled subsurface depositional contexts at four localities around the lowlands of Mo‘orea. Our overarching goal was to integrate diverse lines of proxy data to provide evidence for geomorphological transformations of the island’s littoral and lowland zones from c. 4600 to 200 BP, including changes in vegetation and other biota, which we outline in a four-phase sequence. The early palaeo-sequence combines geomorphological data with analyses of insects, landsnails, and botanical specimens to reconstruct the mid- to late Holocene environment encountered by colonizing Polynesian populations c. 1000 BP. The post 1000 BP portion of our sequence focuses on anthropogenic contributions to sedimentation as well as biotic changes because of animal and plant translocation and human landscape use, most notably, slash and burn agriculture. Comparison of palaeo-period and Polynesian period data indicate a history of over 4000 years of landscape change, both natural and anthropogenic, with the anthropogenic (Polynesian settlement) period associated with greatly accelerated patterns of change. Mo‘orea Island: Environmental setting Mo‘orea, along with Tahiti, forms part of the Windward group of the Society Islands, a linear, age-progressive, volcanic chain (Figure 1). The island’s shield volcano emerged between 1.51 and 1.72 Ma; volcanic activity ceased c. 1.36 Ma (Maury et al., 2000). Subsequent faulting, caldera collapse, and subaerial erosion led to the dissected topography evident today. The two largest amphitheatre-headed valleys, ‘Opunohu and Paopao, have permanent streams emptying into deep bays formed by drowned valleys which were down-cut during late Pleistocene periods of lower sea level. An extensive system of fringing and barrier reefs surrounds Mo‘orea, with deep passes at the heads of ‘Opunohu and Paopao (Cook’s) Bays. The formation of coastal lowlands with calcareous sediments derived from these reefs is dependent upon interactions between eustatic sea level fluctuations, local tectonics, and changing local sediment budgets including the effects of human land use. Across the central Pacific, a variety of geomorphological indicators and radiometric dating supports an inferred peak high stand of c. +1 m between 4 and 6000 years ago. Model predictions for the Windward Society Islands put the mid-Holocene hydroisostatic emergence at 2.3–2.7 m, while observed palaeo-shoreline indicators suggest an actual high stand of 0.5 m (Dickinson, 2001: Table 1). Dickinson projects a thermal subsidence rate of 0.026–0.029 mm/yr or less (Dickinson, 2001: Section 5.2.3 Table 2) for the archipelago, a rate that would affect coastal archaeological sites. Thermal isostatic subsidence after the mid-Holocene peak would have increased the quantity of calcareous, reefderived sediments available to be transported across the reef flats, increasing the sediment budget for the island’s littoral zone. Relative sea level in the Society Islands may also have been influenced by subsidence. Subsidence of Tahiti and Mo‘orea is dominated by isostatic point loading at the active hotspot volcano 335 Kahn et al. Table 1. AMS radiocarbon dates. Beta-# Site Unit Layer Depth (cmbs) Material Conventional Age δ C13 337215 341 TP1 V 254 Cocos nucifera endocarp 3210 ± 30 BP −24.4 308005 341 TP1 VI 270 Pandanus drupe 2430 ± 30 BP −25.8 308007 341 TP1 VII 300 Cocos nucifera endocarp 4640 ± 30 BP −25.7 308006 341 TP1 VII 309 Pandanus drupe 4650 ± 30 BP −27.1 321016 342 TP1 II 47–52 Hibiscus tiliaceus wood 60 ± 30 BP −27.0 321018 342 TP2 IV 275 Pandanus drupe 3740 ± 30 BP −28.7 321017 308008 342 342 TP2 TP1 V V 230–250 270 Cocos nucifera endocarp Pandanus drupe 310 ± 30 BP 3030 ± 30 BP −23.0 −23.9 309302 342 TP1 V 272 Aleurites moluccana endocarp 440 ± 30 BP −22.4 308009 308010 342 343 TP1 TP2 V IIIa 275 125 Cocos nucifera endocarp Aleurites moluccana endocarp 3120 ± 30 BP 250 ± 30 BP −25.7 −22.6 308011 343 TP2 IV 128–130 Pandanus drupe 4610 ± 30 BP −28.2 312349 346 BH1 Peat deposit 190–200 Persicaria sp. seeds 140 ± 30 BP −27.8 308012 348 TP1 II 78 Aleurites moluccana endocarp 120 ± 30 BP −22.9 321534 349 TP2 II 65 Unidentified nutshell 126.2 ± 0.5 pMC −25.6 of Mehetia (Dickinson, 2001). Excavations along the northern coast of Mo‘orea at Papetoa’i (Figure 1) exposed submerged archaeological deposits 0.25–.30 m below sea level which were dated to the 13th–15th centuries (Green et al., 1967: 182). An additional factor contributing to late Holocene sediment budgets on Mo‘orea would have been the nature of the island’s vegetation. Today, native forest dominated by endemic and indigenous taxa is confined to the higher ridges and peaks. The valley interiors and lower slopes are covered in a secondary growth mixture of Polynesian and European-introduced species, while the coastal vegetation is dominated by economic and ornamental plants (e.g. Cocos nucifera, Artocarpus altilis, and Hibiscus rosa-sinensis) along with widespread Pacific littoral taxa (e.g. Cordia subcordata, Calophyllum inophyllum, and Heliotropium foertherianum). In contrast, during the pre-contact period, particularly during the 17th–18th centuries, at the height of Polynesian occupation and land use, much of the island’s interior was under intensive cultivation, as demonstrated by archaeological and palaeobotanical studies of the ‘Opunohu Valley (Kahn et al., 2014; Lepofsky, 2003; Lepofsky et al., 1996). One objective of our study was to obtain evidence for changes in littoral zone and lowland vegetation from the period prior to human settlement up to European contact in the late 18th century. Calibrated Age 1594–1588 BC (0.9%) 1532–1418 BC (94.5%) 730–691 BC (7.4%) 660–650 BC (1.4%) 544–398 BC (86.6%) 3516–3396 BC (77.9%) 3386–3358 BC (17.5%) 3517–3396 BC (80.9%) 3386–3362 BC (14.5%) AD 1692–1728 (23.1%) AD 1811–1920 (72.3%) 2274–2256 BC (3.3%) 2209–2035 BC (92.1%) AD 1485–1650 (95.4%) 1396–1195 BC (94.9%) 1139–1134 BC (0.5%) AD 1416–1490 (94.0%) AD 1602–1610 (1.4%) 1450–1291 BC (95.4%) AD 1521–1575 (14.6%) AD 1585–1590 (0.4%) AD 1626–1679 (55.2%) AD 1764–1800 (21.3%) 1938– … (4.0%) 3510–3426 BC (57.9%) 3382–3339 BC (36.8%) 3204–3196 BC (0.7%) AD 1669–1780 (43.1%) AD 1798–1891 (36.8%) AD 1908–1944 (15.5%) AD 1678–1764 (32.6%) AD 1800–1940 (62.8%) AD 1686–1698 (11.4%) AD 1726–1733 (4.1%) AD 1806–1815 (8.4%) AD 1834–1879 (52.8%) AD 1916–1929 (18.7%) The timing of Polynesian arrival in the Society Islands has been a matter of debate (Allen, 2014; Anderson and Sinoto, 2002; Kahn, 2012; Kirch, 1986; Mulrooney et al., 2011; Wilmshurst et al., 2011). A recent synthesis (Lepofsky and Kahn, 2011) posits initial settlement of the Society archipelago c. AD 1050–1150 BP, while strict chronometric analyses have proposed a later settlement date c. 830–925 BP (Wilmshurst et al., 2011). In either scenario, initial settlement is likely to have been restricted to the coastal lowlands. Inland expansion intensified after 700 BP, when agricultural terrace complexes were constructed (Lepofsky, 1994). By 550 BP, there is substantial archaeological evidence for well-developed communities in interior valleys such as the ‘Opunohu (Green, 1996; Kahn, 2005a, 2007; Kahn and Kirch, 2013). Peak population, marked by chiefly control of surplus production and elaborate temples, occurred after 350 BP (Kahn, 2011, 2013; Kahn and Kirch, 2011; Sharp et al., 2010). The ‘Opunohu Valley has the most robust dataset in the Society archipelago relating to human land use and settlement patterns (Emory, 1933, Green et al., 1967; Kahn, 2005b, 2006, 2011; Kahn and Kirch, 2011, 2013; Lepofsky, 1994; Lepofsky et al., 1992, 1996; Lepofsky and Kahn, 2011; Sharp et al., 2010). Based on stratigraphic and palaeoethnobotanical evidence, Lepofsky et al. (1996) argued that vegetation clearance and burning associated 336 with shifting cultivation in the ‘Opunohu uplands led to massive erosion and deposition of sediments on the valley floor. Direct evidence for initial anthropogenic use of the ‘Opunohu Valley alluvial flat (charcoal from a hearth) was dated to 750–950 BP. Parkes (1997) analyzed sediment cores from Lake Temae and dated changes thought to represent initial colonization to 960– 1260 BP. Our project is analyzing and dating newly taken cores from Lake Temae which also suggest initial anthropogenic impact on sedimentation and vegetation after 1250 BP. Additional evidence for early Polynesian settlement comes from excavations at the GS-1 site near Pihaena point opposite the pass into Paopao Bay (Kahn, 2012). A sample of Hibiscus tiliaceus wood in the lowest cultural deposit at GS-1 was AMS dated to 870–950 BP. Materials and methods Coastal transects, coring, and test excavations Four localities (sites ScMo-341, ScMo-342, ScMo-343, and ScMo-346) were selected for transect coring and test excavations, three along the northern shore on the headlands between ‘Opunohu Bay and Paopao (Cook’s) Bay, and one in the alluvial bottomlands of the ‘Opunohu Valley (Figure 1). These locales represent two contrastive depositional environments, the first across from barrier reefs and major passes where there was likely to be major input of calcareous sediments of marine origin, the second at the base of the island’s largest valley where the main sedimentary input was inland terrigenous sources. Transects were laid out at ScMo-341, ScMo-342, and ScMo343 from the base of the colluvial slopes to the lagoon shore. Coring with a bucket auger proceeded at 10- to 20-m intervals along each transect. Sediments were screened through 1/8 and 1/16 inch mesh; all artifacts and faunal materials were quantified on standardized forms. Bulk sediment samples were retained for extraction of landsnails and for archival purposes. Based on the augering data, we developed predictions of areas with the densest subsurface cultural deposits, which were then opened up for detailed investigation by excavation of 1 m × 1 m test pits or by larger block excavations and trenches via backhoe excavation. These test pits and backhoe trenches allowed us to view broader stratigraphic profiles and to conduct additional pollen and sediment sampling. Stratigraphy and radiocarbon dating After cleaning, stratigraphic profiles of each excavation unit were drawn to scale (Figure 2). Observations were made of lithology, color, and texture for each stratum. Systematic sediment samples were collected for laboratory analysis. A total of 15 samples of wood charcoal and macrobotanical remains were collected from the archaeological deposits, identified to species by Emilie Dotte-Sarout, and submitted to Beta Analytic for AMS dating (Table 1). Where possible, fast-maturing reproductive parts, such as fruits, nuts, and seeds, were chosen for dating, in order to eliminate the ‘old wood’ problem of inbuilt age (Allen and Huebert, 2014; Rieth and Athens, 2013). Calibration of 14C ages was carried out with OxCal program 4.2 (Bronk Ramsey, 2009). Wood charcoal identification Wood charcoal fragments from screened deposits and subsurface features at ScMo-341 and ScMo-342 were collected for taxonomic identification. Anatomical features were described after Wheeler et al. (1989), using Dotte-Sarout’s (2010) Pacific Wood Anatomy Database and reference collection. Sampling and interpretations refer to anthracology principles (Théry-Parisot et al., 2010). Some assemblages were of limited size: ScMo-341 TP1 The Holocene 25(2) samples were relatively rich with 101 and 112 identifiable fragments for Layers II and III; ScMo-342 TP1 samples yielded only 39 and 32 identifiable fragments for Layers III and IV, with lower layers presenting less than 10 fragments each. Sediment analyses Sediment columns with 10-cm sampling intervals were taken at ScMo-341, ScMo-342, and ScMo-343; samples did not cross natural stratigraphic boundaries. Sediment samples were weighed and subsamples removed for microfossil analysis. Munsell color was recorded (dry) and pH was determined. Organic content was measured by the loss-on-ignition procedure (Dean, 1974), after heating the samples in a muffle furnace at 560°C for 1 h. Grain size distributions were determined by wet sieving samples through nested geological screens with mesh sizes ranging from −4Φ to 3Φ (Wentworth scale), and weighing the dried screen contents with a digital scale. Microscope slides were prepared from the 0Φ (1 mm, very coarse sand) fraction and examined under a stereo microscope for lithology. Pollen and plant macrofossils Nine samples from ScMo-341 and ScMo-342 were processed for pollen using standard techniques (Bennett and Willis, 2001). Approximately 23,000 exotic Lycopodium spores were added to each sample to calculate pollen and charcoal concentrations. Only black, opaque angular particles greater than 10 µm were counted as charcoal. Anaerobically preserved plant macrofossils from ScMo-341, ScMo-342, and ScMo-343 recovered by wet screening in the field were identified using botanical reference materials in the Herbarium of the Bishop Museum (Honolulu). Insect and arachnid remains Waterlogged, organic, clay-rich sediments at the base of ScMo341 (TP1), ScMo-342 (TP2), and ScMo-346 were sampled after removing 10 cm from excavation walls. Field processing involved manual disaggregation of sediments in water and density separation of organic materials. Repeated washing with water decantation of the lighter waterborne fraction of the sediment resulted in the separation of organics which were bagged separately; remaining clastic materials were discarded. Organic fractions were sieved through nested sieves from 2 to 0.25 mm. Coarser fractions (>0.5 mm) were sorted in water under a binocular microscope. Plant macrofossils, insect remains, and other remains (e.g. termite ‘coprolites’) were sorted from the sample. Fine fractions (0.25– 0.5 mm) were processed using kerosene flotation with the previously sorted coarser material in order to recover any remaining insect material. Insect and other arthropod remains were identified by comparison with Porch’s (2008) reference material. Landsnails Nonmarine mollusk shells were encountered in certain excavation units, particularly at ScMo-343. Specimens were recovered from sediment samples by flotation or by examination of sieve fractions of 1.00 mm mesh size or larger. Identifications were made by Christensen using reference collections in the Malacology Department at the Bishop Museum. Results ScMo-341 and ScMo-342,Vaioma Valley region Our first two transects are situated on the northwestern headland of Paopao Bay, on a broad coastal plain at the mouth of the 337 Kahn et al. Figure 2. Stratigraphic profiles of sections at (a) ScMo-341, (b) ScMo-342, and (c) ScMo-343. Vaioma Valley (Figure 1). Here, the coastal plain consists primarily of an alluvial fan emanating from the Vaioma Valley, which we call the Pihaena fan. ScMo-341. This site is located on the grounds of the Richard Gump Research Station, where Kahn (2012) had previously excavated stratified deposits pre-dating 600 BP. The ScMo-341 transect ran from near the base of a steep ridge (140 m) across the gently sloping coastal plain to the shoreline, a total distance of 100 m. Three 1 m × 1 m test pits were excavated along this transect. TP1 was the most inland, c. 50 m inland of the circle-island road and had the deepest stratification, as seen in the profile of the unit’s western face (Figure 2a). Seven depositional units were encountered: 338 Layer I (0–50 cm). Dark brown (10 YR 4/3) massive deposit of sandy clay loam, well sorted with occasional angular to subangular volcanic clasts and occasional charcoal flecks. Layer II (50–65 cm). Dark brown (10 YR 4/3) deposit of fluvially transported sand and gravel incorporating larger subrounded to rounded volcanic clasts (0.5–11 cm size range). This most likely represents a flood event from the nearby intermittent stream. Infrequent animal bone was recovered which was too fragmented to identify to species. Layer III (65–145 cm). Dark grayish brown (10 YR 4/3-2) deposit of massive, structureless, well-sorted, highly compact silty clay. Occasional subangular volcanic clasts and charcoal flecks. Layer IV (145–150 cm). Dark grayish brown (10 YR4/3-2) lens of silt and sand with rounded and subrounded volcanic clasts (1–5 cm size range), representing a single fluvial depositional event. Layer V (150–273 cm). Dark brown (7.5 YR 4/2) massive deposit of silty clay largely devoid of volcanic clasts. Layer VI (273–282 cm). Dark gray (N4/) gleyed, very compact, sticky clay, entirely lacking volcanic clasts. During excavation, a thin (c. 1 cm thick) lens of compressed peat was noted and sampled at the upper limit of this deposit. Infrequent animal bone was recovered, but it was too fragmentary to identify to species. Layer VII (282–322 cm). Light gray (10 YR 5-6/2) calcareous marine sand incorporating marine gastropods, becoming increasingly coarse-grained with depth and including water-rolled pieces of Acropora branch coral. Water table appears at c. 294 cm below surface. The lower deposit is currently waterlogged and includes anaerobically preserved Pandanus drupes (sections of the female fruit) and other plant macrofossils. High frequencies of historic and modern artifacts (>40 years old), including metal, ceramics, tiles, and glass, were recovered from the upper 50 cm of sediments across the coastal flat (see Kahn, 2012). Layer I represents modern and historic fill. Layers II–V represent alternating episodes of low- and high-energy fluvial deposition of terrigenous sediments whose source was the adjacent volcanic ridge and the small, intermittent stream at its base. These deposits had infrequent animal bone and basalt artifacts (including flakes, a bifacial tool, and a polished adze flake) and moderate amounts of charcoal; each decreased with increasing depth. Layer VI is an organically enriched layer of gleyed clay and peat with frequent shell. Layer VII, incorporating coral, marine fauna, and anaerobically preserved plant remains, is permanently waterlogged. Results of laboratory analysis of the sedimentary sequence in TP1 are presented in Figure 3. The sediments are dominated by silt- to clay-sized fractions, except for the basal Layer VII which consists largely of medium- to coarse-grained sands. There is a shift from calcareous to terrigenous sediments at 260 cm, indicative of a major change in the source of sediment input. While silt- and clay-sized materials dominate the upper terrigenous sediments, there are four distinct pulses when larger pebble and gravel-sized grains are evident. These most likely represent short-term high-energy erosional events, possibly associated with tropical cyclones. Unlike at site ScMo-342 (see below), microscopic charcoal was not observed in the ScMo341 sediments. Four radiocarbon dates were obtained on samples from Layers V, VI, and VII (Table 1). Two samples from Layer VII (Beta308006 and Beta-308007) on preserved coconut endocarp and a Pandanus drupe have nearly identical ages of 4650 and 4640 ± 30 BP. A Pandanus drupe from overlying Layer VI returned an age The Holocene 25(2) of 2430 ± 30 BP, while a coconut endocarp from Layer V dated to 3210 ± 30 BP. This stratigraphic inversion of the dates from Layers VI and V suggests that there may have been mixing of materials during deposition of the Layer V fluvial event. In aggregate, dates from the base of TP1 indicate that the waterlogged deposits below 2.5 m depth substantially pre-date Polynesian occupation of the island. Earlier work at the same site, but at its southern extent, dated the upper part of the fluvial deposit (presumably the same as our LII), to between c. 250 and 100 BP (Kahn, 2012). A cultural deposit underlying this fluvial deposit (and apparently not extending to the northern portion of the site) was dated c. 750–950 BP. Botanical samples from the fluvially transported sediments (LII–LV) were dominated by charred fragments of Aleurites moluccana and Cocos nucifera nuts. Both have varied subsistence and economic uses. In total, 22 taxa were recorded from the wood charcoal assemblages of LII–LV. Hibiscus tiliaceus, an indigenous species known to thrive in anthropogenically disturbed environments (Florence, 2004), was recovered at moderately high frequencies (1/4–1/6 of the assemblages). Other Polynesian introductions found with low frequency include Cordyline fruticosa, Artocarpus altilis, and an Anacardiaceae identified as cf. Spondias dulcis. Indigenous species represented at low frequency include Barringtonia asiatica, Thespesia populnea, several Rubiaceae and Cunoniaceae, Myrtaceae cf. Metrosideros, Euphorbiaceae cf. Phyllanthus, and Mimosaceae cf. Serianthus myriadena (Butaud et al., 2008; Florence, 1997, 2004). The wood charcoal data demonstrate that Layers III–V date to the Polynesian period. The mix of endemic and introduced or cultivated species suggests an anthropogenic landscape dominated by tree crops and secondary species, albeit with remnants of native forest. Plant macrofossils from Layers VI and VII exhibited less diversity. All of the represented species were indigenous taxa including Pandanus, Cocos, Barringtonia asiatica, Cordia subcordata, and Hibiscus tiliaceus. ScMo-342. This transect is situated 500 m west of ScMo-341, near the western edge of an alluvial fan that emanates from the mouth of Vaioma Valley. The N-S oriented transect ran 300 m from the shore across the gently sloping coastal flat to the intermittent Vaioma Stream channel (Figure 1). The stream channel parallels the transect about 40 m to the west, and provided the main source of sediment input to the coastal flat. Two 1 m × 1 m test units were excavated, in addition to a 4 m × 3 m block excavation (Figure 4). Stratigraphy was similar to that encountered at ScMo-341 (Figure 2b). The stratigraphic sequence at TP1 is approximately mid-way along the coastal flat 150 m from the shoreline and 170 m away from the inland stream. Layer I (0–38 cm). Dark reddish gray (5 YR 3-4/2). Blocky ped structure; upper 15 cm is the A horizon. Occasional scattered volcanic clasts and scattered charcoal flecks. Contact with II diffuse. Layer II (38–62 cm). Reddish brown (5 YR 4/3), highly compact, sandy clay loam. Considerable quantity of rounded to subrounded volcanic clasts and charcoal flecking. In the southwestern part of the unit, large basalt cobbles forming a pavement were associated with dense charcoal concentrations in a hearth, indicating the presence of a former occupation and possibly a cookhouse. Layer III (62–184 cm). Dark brown (7.5 YR 4/4) massive, structureless, silty-clay loam with occasional charcoal flecking, the lowest 15 cm becoming strongly mottled (10YR 6/6 brownish yellow). 339 n( %) pH cro Sh Ma ell rin s( %) e Lo ss on I gn itio Mi Or ga Pa nic rti cle s( %) Ca lc Se areo dim us en t (% ) Te rri Se geno dim u en s t (% ) Cla y( >4 ) Sil t Ve ry an d Co ars e De pt h( cm Str ) ati gra ph y 14 CD ate sB .P. Pe bb les (-4 ) Sa nd (0 ) Kahn et al. 0- 50 - 100 - 150 - 200 - 250 2430 +/- 30 4640 +/- 30 4650 +/- 30 300 0 20 40 0 20 0 20 40 60 80 0 20 40 60 80 100 0 20 40 60 80 0 5 10 0 5 10 0 5 10 0 5 10 Figure 3. Sedimentary profile, ScMo-341. Figure 4. Backhoe trench block excavations at ScMo-342 with Layer VII (peat) outlined in white and Layer VIII (calcareous sand) at the basal limit; the latter was rich in macrobotanical remains. Layer IV (184–200 cm). Dark gray, gleyed (N4/), very compact, structureless, well-sorted clay, containing some fibrous plant material. Layer V (200–240 cm). Gray (5 YR 5/1) calcareous marine sand, fine to medium grained, with fragments of marine shell. Anaerobically preserved plant remains are abundant. Analysis of sediment samples from the column taken from the face of TP1 illustrate a shift from calcareous to terrigenous sediments in the basal part of the section. Layer IV in TP1 is remarkable for its acidic pH (mean value 3.15) and corresponding high degree of organic matter including anaerobically preserved organic particles (wood, fiber) in the 0Φ size fraction. The upper terrigenous sediments, mostly dominated by silt and clay-sized particles, show two pulses with larger sized clasts, as at ScMo-341, presumably marking high-energy erosional events. Charcoal particles were present in appreciable quantities in the 0Φ size fraction samples from Layer IV and from higher up in the section (above 100 cm depth). As at ScMo-341, the fluvial deposits in Layer III at ScMo-342 had infrequent faunal remains (including a fish dentary) as well as charcoal. The underlying Layer IV and V deposits appear to be somewhat mixed, with evidence for infrequent bird bone and faunal materials (including rat bone) which likely derived from Layer III in addition to frequent organic materials and marine fauna. The basal Layer V deposit had frequent anaerobically preserved plant macrofossils, including Pandanus tectorius drupes, coconut endocarps, Terminalia sp., and Cordia subcordata seeds (Figure 5). TP2 at ScMo-342 was opened farther inland, 220 m from the shoreline and 100 m to the point where the intermittent stream debouches from Vaioma Valley. The overall depositional sequence in TP2 was similar to that of TP1, but with some important differences: Layer I (0–40 cm). Dark reddish brown (5 YR 3-4/2) silty-clay loam with a few scattered subrounded volcanic clasts (2–3 cm size range). Layer II (40–75 cm). Dark reddish brown (5 YR 3/2) sandy clay loam with occasional cobbles to 20 cm size indicative of a higher energy fluvial environment. Layer III (75–139 cm). Dark brown (7.5 YR 3-4/2) massive, structureless deposit of silty-clay loam with scattered charcoal flecks. Layer IV (139–145 cm). Dark reddish brown (5 YR 3/3) lens of gravel and small cobbles in a matrix of sandy clay loam; high frequency of subrounded volcanic clasts (1–5 cm size range) representing a higher energy fluvial event. Layer V (145–264 cm). Dark brown (7.5 YR 3-4/2) massive, structureless deposit of silty-clay loam with occasional subrounded volcanic clasts (1–5 cm size range). Layer VI (264–291 cm). Very dark gray (5 YR 3/1) with mot- 340 The Holocene 25(2) Figure 5. Macrobotanicals identified in the ScMo-342 site excavations, including (a) Pandanus tectorius drupes, (b) Cordia subcordata seeds, (c) Terminalia sp. seeds, and (d) Cocos endocarp. tlings of reddish brown (5 YR 4/4); densely compacted gravel and cobbles; clasts well rounded with the majority in the 5– 10 cm size range, but occasional cobbles up to 35 cm diameter. This represents a major, high-energy fluvial event, presumably a flood from the Vaioma Stream. Water table was hit during this layer at c. 264 cm below surface. Layer VII (291–296 cm). Black (2.5 YR N2.5/), compact deposit of peat with anaerobically preserved organic matter. Layer VIII (296–332 cm). Very dark gray (5 YR 3/1) medium- to coarse-grained calcareous sand with marine shells and course organic material. Analysis of the sediment column from TP2 is presented in Figure 6. The sequence is similar to that at TP1, although in this case terrigenous sediments are present throughout with the basal Layer VIII having an admixture of calcareous and terrigenous grains. This dominance of terrigenous grains even in the basal deposits likely reflects the location of TP2 closer to the mouth of Vaioma Stream. Layers VI and VIII again are strongly acidic (pH 2.82–3.02) with significant quantities of anaerobically preserved plant materials. As in TP1, charcoal first appears just above the peat deposit (in the bottom of Layer V), and may reflect initial human land clearance with burning in the vicinity. Charcoal is then relatively abundant again near the top of the depositional sequence. Periodic high-energy fluvial events are represented by three major pulses of large-sized clasts, again presumably associated with major flood events from the Vaioma Stream, possibly induced by cyclones. Six AMS radiocarbon dates were obtained for ScMo-342, four from TP1 and two from TP2 (Table 1). Three of these samples from anaerobically preserved plant macrofossils (coconut endocarp and Pandanus drupes) from the basal, waterlogged sediments (Layer V in TP1 and Layer VII in TP2) yielded ages of 3030 ± 30, 3120 ± 30, and 3740 ± 30 BP (Beta-308008, Beta-308009, Beta321018). These dates are consistent with those from the basal deposits at site ScMo-341, indicating that the marine sands and peat deposits that underlie the Pihaena alluvial fan are pre-Polynesian in age, dating to between 4600 and 2400 BP. The remaining three dates from ScMo-342 are from higher stratigraphic contexts associated with Polynesian land use. Sample Beta-309302 with an age of 440 ± 30 BP, although recorded as coming from a depth of 220–240 cm, Layer V in TP1, is the Polynesian-introduced candlenut (Aleurites moluccana) and most likely became mixed into the top of Layer IV in an initial highenergy fluvial event after Polynesian colonization of the island. This date suggests that the thick terrigenous deposits making up the Pihaena alluvial fan did not begin to accumulate until after c. cal. 460–534 BP. A second sample (Beta-321017) from Layer V in TP2 dates to 310 ± 30 BP, also confirming that the alluvium from the Vaioma Stream accumulated in this area after c. cal. 300–465 BP. A third sample (Beta-321016) from the dense charcoal associated with a probable cookhouse feature in Layer II of TP1 dated to 60 ± 30 BP, with a most likely calendar age of cal. 222–258 BP. Plant macrofossils and wood charcoal. In total, 15 taxa were found in the wood charcoal assemblage from the Layer II hearth and oven rake out. The assemblage is dominated by Hibiscus tiliaceus (55% of all identifiable fragments) and Syzgium malaccense (13%). A suite of other Polynesian introductions are present in low frequencies (<5%), including Inocarpus fagifer, Artocarpus altilis, and Cordyline (stem). Also recovered in low frequencies are indigenous species, including several Rubiaceae, a Euphorbiaceae, Cordia cf. subcordata, and Thespesia populnea. Cocos nucifera is represented by two stem fragments and numerous nut fragments. These data illustrate the presence of a mosaic-type vegetation (arboricultural lands, disturbed areas, remnant native woodlands) where Polynesians were collecting firewood during the late pre-contact and early historic periods. The Layer III fluvial deposits have a similar makeup but have a higher taxonomic diversity (23 taxa). In addition to those listed for Layer II, Polynesian introductions include Aleurites moluccana, possibly the tiare and noni trees (Gardenia cf. taitensis, Morinda cf. citrifolia), and tuber remains tentatively identified as Colocasia. Indigenous species are more frequent, with Calophyllum inophyllum, Ficus sp., Fagraea berteroana, a Myrtaceae, an Apocynaceae, and two Rubiaceae occurring in addition to taxa found above. The Layer III wood charcoal assemblage represents both burning of the inland native forest as well as transformation to a cultivated landscape. 341 Mi pH cro Sh M a ell rin s( %) e Lo ss o Ign n i ti o n( %) lca Se reou di m s en t (% ) Ch arc oa l (% Or ) ga Pa nic r ti cle s( %) Ca Te rri g Se e n o di m u s en t (% ) (>4 ) Ve ry Sa Coar nd se (0) S il ta nd C la y De pt h( cm Str ) ati gra ph y 14 CD at e sB .P. Pe bb les (-4 ) Kahn et al. 0 - 50 - 100 - 150 - 200 - 250 - 3740 +/- 30 300 - 310 +/- 30 320 0 10 20 30 40 50 60 0 10 0 10 20 30 40 50 60 0 20 40 60 80 100 0 20 40 60 0 2.5 0 5 10 15 0 2.5 0 5 10 0 5 10 Figure 6. Sedimentary profile, ScMo-342. Significant numbers of anaerobically preserved macrobotanical remains were recovered from the waterlogged basal layers at ScMo-341 and ScMo-342 (Figure 4). The most abundant remains were of Pandanus tectorius, an indigenous species that is widely distributed across the central and eastern Pacific (Figure 4a). A total of 578 drupes were recovered – evidence that Pandanus tectorius was a major component of the coastal forests of Mo‘orea in the mid-Holocene. Also present were endocarps of two other indigenous trees commonly found in the littoral zones of Polynesian islands: Cordia subcordata and Terminalia sp. (Figure 4b and c). More surprising and of considerable note was the occurrence of numerous examples of well-preserved endocarps of a small, presumably wild (i.e. non-domesticated) form of coconut, Cocos nucifera (Figure 4d). The small overall mean length and width of the endocarps, 8.5 cm × 7.0 cm (sd. 0.6 length, 0.7 width) and their oblong shape conforms to the size patterns described for Lepofsky et al. (1992). Two samples were directly dated by 14C to 3210 ± 30 and 3120 ± 30 BP (Beta-337215 and Beta-308009 respectively), confirming that wild coconut was present in the coastal zone of Mo‘orea in the pre-Polynesian era. Pollen assemblages. Nine samples were taken from deposits at ScMo-341 and ScMo-342 for pollen analysis. In all, 25 different pollen types were recorded. The pollen concentration is extremely low in the basal samples from ScMo-342 TP2 and from ScMo341 TP1. The assemblages across all three sites are dominated by fern spores (Figure 7). The three pollen samples from waterlogged Layer VII at ScMo-341, while dominated by fern spores, reveal a decrease in the frequency of Pandanus, and a corresponding increase in sedges (Cyperaceae) and grasses in the upper two samples. This suggests changes in the coastal vegetation from a closed moist lowland forest to a more diverse lowland forest environment, possibly incorporating open sedge/grassland. There is also over a 10-fold increase in pollen concentration from the lowest to uppermost sample, with the uppermost sample being the most diverse in pollen content and where Casuarina and palm pollen appear for the first time. Charcoal is absent from all of the ScMo-341 samples. In TP1 at ScMo-342, the pollen assemblage from lower Layer V is composed of fern spores, Pandanus pollen, and Ficus pollen. Ficus decreases in the central sample from Layer V, with the inclusion of small amounts of Rutaceae pollen (allocated to the other category) and the first appearance of charcoal, possibly derived from anthropogenic burning. The uppermost sample from Layer V is more diverse palynologically. Grass pollen appears for the first time, Ficus expands, and Hibiscus pollen constitutes the ‘other’ category. While the pollen concentration is constant across all three layers, the charcoal concentration is much higher in the uppermost sample from Layer V, suggestive of human presence in the adjacent landscape. In TP2 at ScMo-342, samples were analyzed from the peat deposit (Layer VII, 372–384 cm) and from the basal Layer VIII (at 384–390 and 390–401 cm). TP2 differs from TP1 in several respects. First, there is almost a 10-fold increase in pollen concentration from lowermost sample to uppermost sample. Second, charcoal values are much lower overall in TP2 compared with TP1 and decrease up the sequence. While ferns, Pandanus, and Ficus continue to dominate, the ‘other’ category is composed of two types in 390–401 cm, three Types in 384–390 cm, and eleven types in 372–384 cm. Glochidion is present in all three levels. Overall, the sequences show the development of closed lowland forest to a more open and diverse forest at site ScMo-342. 342 The Holocene 25(2) 341 TP1 Layer V 342 TP1 Layer IV 342 TP2 Layer VII Ferns Poaceae Pandanus Ficus Cyperaceae Other Dam/Crumpled 13% Dam/Crumpled 3% Dam/Crumpled 7% 341 TP1 Layer VI 342 TP1 Layer V (upper) 342 TP2 Layer VIII (upper) Dam/Crumpled 9% Dam/Crumpled 3% Dam/Crumpled 9% 341 TP1 Layer VII 342 TP1 Layer V (lower) 342 TP2 Layer VIII (lower) Dam/Crumpled 13% Dam/Crumpled 0% Dam/Crumpled 0% Figure 7. Pie charts representing pollen concentrations at ScMo-341 and ScMo-342. Insect and arachnid remains. Insect subfossils are diverse and moderately rich in ScMo-342 TP2 and less abundant and less well preserved in ScMo-341 TP1; in the discussion below, we focus on the ScMo-342 assemblage. Both insect assemblages are dominated by a range of taxa that are considered part of the indigenous fauna of the Society Islands and Eastern Polynesia generally. However, many of these taxa are today absent from lowland contexts or high islands or absent from low islands altogether. They are commonly restricted to mid–high altitude forests where human impact is less pervasive. Evidence that the ScMo-342 assemblage was deposited in a shallow wetland habitat includes the presence of the tachyine ground beetle Paratachys sexguttatus, a species widespread in wetland habitats in Eastern Polynesia today (Figure 8). Likewise, the occurrence of the water beetle genera Paracymus, Enochrus, and Copelatus, the latter a new record for French Polynesia, signals a wetland habitat. The presence of a black fly, Simulium (Inselellium) sp., suggests occasional flowing water contribution to the fauna (Craig and Porch, 2013). The beetle fauna is otherwise dominated by saproxylic (dead wood) taxa like cossonine weevils (many species in multiple genera), cryptorhynchine weevils (Ampagia and several unidentified genera), the eastern Polynesian endemic genus Proterhinus (multiple species), and zopherids (Zopheridae). These are well represented with several Pycnomerus and Bitoma species, and a new species each of Antilissus and Epistanus (first records of these genera for the Society Islands). Both the Zopheridae and cossonine weevils are far more diverse and abundant in the prehuman record than they are today in eastern Polynesian environments, implying significant extinction or massive under-collection. Other saproxylic taxa include several genera of osoriine staphylinids, Dorcatomiella (Ptinidae), and Aprostomis, a genus which is widespread in the Pacific and known from fern fronds and similar microhabitats (Lawrence et al., 1999). Phytophagous taxa include the weevils Miocalles and probably Rhycogonus, present today at moderate to high elevations, a tiny jewel beetle (cf. Maoraxia), and bark beetles (Scolytinae) represented by at least several species of Xyleborus and Hypothenemus. A range of other taxa common in forest habitats are present, including the species of the beetle families Ciidae, Nitidulidae, Ptiliidae, and Histeridae. Taxa other than beetles include numerous genera of oribatid mites and several spiders including a thomisid, most likely Misumenops melloleitaoi, a species that is today limited to medium to high altitudes in Tahiti and Mo‘orea (Garb, 2006). Ants are infrequent but include several specimens of Strumigenys and an unidentified ponerine genus. As an assemblage, the ScMo-342 fauna imply an intact native rainforest with abundant decaying wood. A closed forest canopy is implied by the insect assemblage and plant macrofossil assemblages associated with the insect samples. ScMo-343,Vaipahu coastal flats ScMo-343 is situated near the eastern headland of ‘Opunohu Bay (Vaipahu Point; Figure 1). The coastal plain is approximately 170 m wide, terminating at the base of a steep ridge that ascends to an elevation of 250 m. Auger holes at 10-m intervals revealed exclusively calcareous sediments over most of the 170-m-long transect, with some input of terrigenous sediments in the last two auger holes closest to the volcanic slope. We excavated two 1-m test pits in this area, which today consists of swampy ground with dense Hibiscus tiliaceus vegetation (Figure 1). Four stratigraphic deposits were exposed in the TP1 and TP2 (Figure 2c) excavations: Layer I (0–6 cm). Dark reddish brown (5 YR 3-4/2) silty-clay loam with a few scattered subrounded volcanic clasts (2–3 cm size range). Layer II (9–96 cm). Dark brown (7.5 YR 3-4/2) massive, structureless deposit of silty-clay loam with scattered charcoal flecks and a few scattered subrounded volcanic clasts (2–3 cm size range). Diffuse basal boundary. Layer IIIA (96–c. 140 cm). Dark gray, gleyed (N4/), very compact, structureless, well-sorted clay, containing some fibrous plant material. Highly irregular but distinct basal boundary. Basal limit has a notably high concentration of landsnails. Layer IIIB (c. 140–190 cm). Lens of dark gray, gleyed (N4/), very compact, structureless, well-sorted clay, containing some 343 Kahn et al. Figure 8. Beetle (Coleoptera – (a)–(k), (m)–(o)), ant (Hymenoptera: Formicidae – (l)) and spider (Aranae: Thomisidae – (p)) subfossils from ScMo-342, TP2 excavation. (a) Bitoma sp.1 pronotum (Zopheridae); (b) Pycnomerus sp.2 pronotum (Zopheridae); (c) Proterhinus sp.1 (Belidae); (d) and (e), (i) and (k) Cossoninae genus indet. pronota (Curculionidae); (f) Nitidulidae genus indet. sp.1 elytron; (g) Ampagia sp.1 head (Curculionidae); (h) Antilissus sp. ‘Moorea’ pronotum (Zopheridae); (l) Strumigenys sp. head (Formicidae); (m) Miocalles sp.2 pronotum (Curculionidae); (n) Osoriinae gen. indet. head (Staphylinidae); (o) Staphylinidae genus indet. head (Staphylinidae); and (p) Misumenops cf. melloleitaoi carapace (Thomisidae). Scale (1) is 0.5 mm and applies to all images. fibrous plant material found within Layer IV. Highly irregular distinct basal boundaries. Layer IV (140–200 cm). Very dark gray (5 YR 3/1) mediumto coarse-grained calcareous sand and marine shells. Partially below the water table, with anaerobic preservation of Cocos nucifera endocarp and Canarium seeds. High marine shell and coral content (including boulder-sized Porites heads). Results of sediment analysis of the TP2 column are summarized in Figure 9. There is a major shift in the depositional regime at 1 m below surface. Below this depth the sediments are dominated by calcareous sands, containing significant quantities of micro-marine mollusks. This represents a period when the main source of sediment would have been the adjacent reef flat, with only minor input of terrigenous sediments from the steep ridge. Above 1 m in the profile, terrigenous sediments dominate (although some calcareous grains continue to be present), indicating that the main source of sediment input was now the volcanic ridge. This shift is accompanied by an increase in organic content and an increase in the silt- to clay-sized fractions. A single high-energy depositional event is suggested by the sample from 70–80 cm, which has a high frequency of larger sized clasts. Of particular note is the increase in organic particles (anaerobically preserved pieces of wood and fiber) present in the samples higher than 120 cm depth. This is suggestive of a transition from an open beach depositional environment to a swampy back-beach environment. Radiocarbon dates. Two samples from TP2 at site ScMo-343 were submitted for AMS radiocarbon dating. An anaerobically preserved Pandanus drupe from Layer IV (Beta-308011) yielded an age of 4610 ± 30 BP. From the base of Layer II, a preserved endocarp of Aleurites moluccana or candlenut (Beta-308010) was dated to 250 ± 30 BP (highest probability intercept of cal. 271– 324 BP). The date on this Polynesian-introduced tree indicates that the major shift in depositional regime marked by an increase in terrigenous sediments occurred about 300 years ago, slightly later than at TP2 at site ScMo-342. Landsnail remains. Layer IIIA, which post-dates 250 BP, had abundant landsnails. Several of the recovered species inhabit aquatic (Clithon) or shoreline and strandline environments (Laemodonta monilifera, Melampus sp., probably also Assiminea parvula, although the last-named species may also occur in inland locations). Five terrestrial species that were identified are endemic to Mo‘orea (Libera dubiosa, Libera jacquinoti, Libera new species, Minidonta new species, Sinployea modicella), but three are native to Mo‘orea, while three also occur on other nearby island groups (Georissa striata, Taheitia pallida, Lamellidea micropleura). The status of several recovered taxa (Sturanya sp., Nesopupa sp. cf. tantilla, Hiona sp. cf. verticillata, Hiona sp., Sinployea sp.) is unclear because of uncertainties as to their specific identity, but they are undoubtedly native to Mo‘orea and perhaps occur on other nearby islands (Christensen et al., unpublished; Garrett, 1884). A number of commensal terrestrial species were present, introduced into the islands of Oceania either prehistorically (Gastrocopta pediculus, Pupisoma dioscoricola, Allopeas gracile) or since the advent of European commerce (Gastrocopta servilis, Leptinaria unilamellata, Opeas hannense, Paropeas achatinaceum, Subulina octona). Additional commensal species (Lamellidea oblonga, Lamellidea pusilla, perhaps Elasmias sp., and Tornatellides oblongus) were transported within Oceania prehistorically, but their point of origin within that region is unknown (Christensen and Weisler, 2013). One aquatic species, Melanoides tuberculata, is cryptogenic in status; although it is a global invasive, the duration of its presence in Remote Oceania is uncertain. The prehistoric introduction of alien commensal species signals the start of the Polynesian period of anthropogenic change. Ecological change is likewise documented with the gradual extinction of certain endemic landsnail species. Of the species endemic to Mo‘orea that are recorded here, Libera new species 344 %) pH Lo ss o I gn n itio n( Mi Or ga Pa nic rti cle s (% ) cro Sh Ma ell rin s( %) e ) Ca lca Se reou dim s en t (% Te rri g Se eno dim us en t (% ) Sil ta nd Cla y( Ve ry Sa Coar nd se (0) De pt h Str (cm) ati gra ph y 14 CD ate sB .P. Pe bb les (-4 ) >4 ) The Holocene 25(2) 0- 50 - 100 - 250 +/- 30 4610 +/- 30 150 - 180 0 20 40 0 20 0 40 60 80 0 20 40 0 20 40 0 5 10 15 0 5 10 0 5 10 0 5 10 Figure 9. Sedimentary profile, ScMo-343. and Minidonta new species are known only from the present archaeological excavations; they apparently became extinct prior to European contact (1767) or early in the historic era. Libera jacquinoti was collected in 1839 but has not been observed since – it may be presumed to have become extinct prior to the 1880s. Libera dubiosa was collected by Garrett (1884) and others in the late 1800s but has not been seen subsequently; it probably went extinct before 1934. ScMo-346, ‘Opunohu Valley floor This site is located on the alluvial floodplain of the ‘Opunohu Valley, approximately 60 m from the present stream channel, and 600 m inland of the modern shoreline where the stream enters into the bay. This site was investigated by excavating a 3-m-long backhoe trench to a depth of 2 m at which a massive deposit of river gravel and cobbles was encountered: Layer I (0–60 cm). Reddish brown to dark reddish brown (5 YR 3-4/4) massive, structureless deposit of silty-clay loam; some charcoal flecking noted throughout. Layer II (60–130 cm). Reddish brown to dark reddish brown (5 YR 3-4/4) deposit, very similar to Layer I but with higher clay content. Some oxidation mottling appears just above the contact with Layer III. Within Layer II, there is a distinct lens of coarse-grained volcanic sand between 75 and 80 cm, representing a higher energy flood event. Layer III (130–200 cm). Dark gray (5 Y 4/1), massive, structureless deposit of gleyed clay with very little silt content. Below 150 cm, this deposit incorporates substantial quantities of anaerobically preserved, fibrous plant material. The water table was reached at 190 cm. Analysis of the sediment column from site ScMo-346 shows the sediments to be dominated throughout by silt- to clay-sized particles (>−4Φ), although Layer II is slightly less well sorted. The sediments are entirely terrigenous in origin, with no calcareous input. Layer III is fairly acidic (pH 4.4–4.5), whereas Layers I and II have pH values in the range of 5.4–6.5. Loss-on-ignition values are slightly higher for Layer III (9.9–10%) than for Layer II (7–9%). Charcoal grains were present throughout the 0Φ fraction that was microscopically examined. Radiocarbon date. A single sample (Beta-312349) of anaerobically preserved Persicaria sp. seeds from 190–200 cm depth within Layer III was dated to 120 ± 30 BP. This date has five possible calibrated age ranges between 280 and 0 BP but most likely pre-dates European arrival. Persicaria glabra is regarded as a Polynesian introduction to the Society Islands (Florence, 2004, vol. 2). Plant and insect remains. Twelve litres of organic sandy silt with abundant coarse plant remains were sampled from the base of the –346 trench at a depth of 190–200 cm. Of the several thousand seeds recovered from the sample, most belong to several species of Cyperus sedges (Cyperaceae) with fewer of Fimbristylis, Eleusine indica, Solanum americanum, Persicaria, Amaranthaceae, Brassicaceae, Cucumis, and another unidentified Cucurbit. Many of these taxa are probable Polynesian introductions (Prebble, 2008). The insect fauna is dominated by taxa most likely introduced prehistorically, but includes, in smaller numbers, indigenous and/or endemic terrestrial and freshwater species like the endemic Insellielum black flies (Craig, 1987). The species present in the ScMo-346 samples are yet to be identified; however, it is worth noting that elsewhere in Polynesia, human impact has resulted in the extinction of lowland black fly species (Craig and Porch, 2013). Examples of the likely human introductions include the earwig Euborellia annulipes, the bug Geotomus pygmaeus, the beetles Saprosites pygmaeus and Litargus balteatus, and many ants. Although the calibrated range of the radiocarbon determination on Persicaria seeds could be as young as mid-20th century, we believe it is likely the sample dates to the late prehistoric or earliest post-contact periods. This is based on the absence of a range of key taxa that, elsewhere in eastern Polynesia, occur ubiquitously in post-contact samples, combined with the depth of the sample in an apparently stable 20th-century landscape. The ant fauna of this sample, for example, contains a mix of taxa that is characteristic of Polynesian aged sediments from elsewhere in eastern Polynesia (Porch, unpublished) and lacks ants like Pheidole fervens, Cardiocondya sp., and Strumigenys rogeri that are characteristic of the post-European contact period. The absence of the herb/shrub Ludwigia octivalvis is similarly indicative of Polynesian aged sediments. Discussion The stratigraphic and palaeoenvironmental investigation of four localities revealed consistent evidence for the geomorphological evolution of the littoral and lowland zones of Mo‘orea Island from c. 4600 to 200 BP, including changes in the vegetation and 345 Kahn et al. other biota of those zones. Here, we synthesize this information as a series of four successive stages. Stage 1 (4600 to ~3000 BP) This phase is evidenced by the basal deposits at sites ScMo-341, ScMo-342, and ScMo-343, all of which consist of calcareous sands and reef detritus (branch coral fingers, marine mollusks) derived from the offshore fringing and barrier reefs. This phase correlates with a widely evidenced mid-Holocene sea level high stand which on Mo‘orea has been estimated at between 0.5 and 1 m higher than the modern level. Because of the increase in reefderived sediment occasioned by this high stand, narrow coastal beaches are formed at the base of the steep volcanic slopes around the headlands between ‘Opunohu and Paopao Bays. The presence of larger clasts (coral cobbles, branch coral fingers) suggests a higher energy shoreline in the vicinity of ScMo-341, ScMo-342, and ScMo-343 than is the case today. Plant macro- and microfossils incorporated within these basal sediments provide an indication of Mo‘orean coastal vegetation at this time. Of significance is the presence of a wild form of coconut with a small endocarp volume. It was previously unknown whether coconut had naturally dispersed as far east as the Society Islands, although Parkes (1997: 197) had reported Cocos pollen from Atiu Island in the Southern Cook group dated to 8600–7600 BP and pre-Polynesian Cocos pollen from the Lake Temae core on Mo‘orea (p. 192). That Cocos had naturally dispersed as far as Mo‘orea by at least 4600 BP suggests that it will likely appear in other fossil assemblages throughout central Eastern Polynesia. This discovery casts doubt as to whether coconuts reported by Lepofsky et al. (1992) from basal deposits in the ‘Opunohu Valley are evidence of Polynesian occupation. In addition to the coconuts, plant assemblages at ScMo-341, ScMo-342, and ScMo-343 are dominated by Pandanus tectorius drupes and Cordia subcordata and Terminalia cf. samoense endocarps. Wood charcoal analysis adds Hibiscus tiliaceus and Barringtonia asiatica to the list of taxa, while the pollen samples from ScMo-341 add Ficus and Phyllanthus in the immediate coastal zone. Ficus tinctoria is a common indigenous species in central Polynesia, while two species of Phyllanthus (formerly Glochidion), Phyllanthus manono and P. taitensis, are known from the Society Islands, frequently occurring in lowland habitats (Butaud et al., 2008: 124). All of these taxa indicate the presence of a welldeveloped native coastal forest on Mo‘orea during Stage 1. wood charcoal data) hint at human presence on the landscape, but the exact timing of this is difficult to ascertain given the lack of good stratification. Beginning as early as 440 BP (cal. 460–534 BP) at ScMo-342, however, and soon thereafter at ScMo-341 and ScMo-343, a major new phase of sedimentary deposition commenced which can only be attributed to anthropogenic effects. This phase is especially evident at ScMo-341 and ScMo-342, where between 1.8 and 3.0 m of terrigenous sediments accumulated within a span of two to three centuries. The source of this sediment was the valley and ridge slopes immediately inland of the sites, especially the Vaioma Valley in the case of ScMo-341 and ScMo-342. This phase of active slope erosion and deposition on the coastal flats correlates with the period of major inland expansion of Polynesian occupation and intensive agriculture on the island, previously documented in particular for the ‘Opunohu Valley, but doubtless an island-wide phenomenon. The presence of scattered charcoal throughout the thick alluvial and colluvial sediments indicates the use of fire on the inland slopes. The widespread recovery of wood charcoal from numerous economically important tree species indicates the presence of anthropogenically transformed landscapes. The fact that these assemblages are largely dominated by successional forest species and Polynesian cultigens indicates both shifting cultivation and the creation of arboricultural gardens on the interior slopes of the island. The stratigraphic sequence at ScMo-346, in the alluvial floodplain of the ‘Opunohu Valley, relates to this stage, reinforcing the conclusions of Lepofsky et al. (1996) that intensive land use in the valley led to massive alluvial in-filling of what had formerly been a more extensive embayment and swampy zone. The basal deposit at ScMo-346, dating to the period immediately prior to European contact, evinces a swampy environment some 600 m inland of the modern shoreline, which then became buried under nearly 2 m of alluvium, before stabilizing sometime in the modern period. Stage 4 (300–150 BP) During this final period, the coastal flats became stabilized, possibly because of an increased emphasis on arboriculture and a decline in the use of shifting cultivation. Furthermore, after initial European contact at the end of the 18th century, massive population declines led to abandonment of inland settlements and significant secondary forest regrowth on the interior slopes, further decreasing erosion of terrigenous sediments. Stage 2 (~3000–2400 BP) This stage saw the draw-down of sea level from the mid-Holocene high stand to its present level, accompanied by the formation of low energy silty deposits in swampy depressions presumably behind low beach ridges. These back-beach swamps were traps where organic matter accumulated, forming the anaerobic, organic peat-like deposits found at ScMo-341, ScMo-342, and ScMo-343. The insect fauna is consistent with this interpretation and suggests a closed canopy coastal forest surrounding these swamps, continuing to be dominated by such trees as Pandanus, Ficus, and Phyllanthus, but also with Hibiscus tiliaceus. Stage 3 (550–300 BP) The coastal zones stabilized after about 2400 BP, without much active sediment accumulation other than organic materials in the back-beach swamps. As a result, the period from c. 2400 to 550 BP is attested primarily by a gap in deposition. A few scattered indications of human-introduced biota (microscopic pollen influxes at ScMo-341 and ScMo-342 and the Conclusion Multiple lines of evidence support a model of natural and anthropogenic change over a 4600 year sequence in the Society Islands. Changing sedimentary records and macrobotanical and insect remains record shifts in geomorphology and the environment in the pre-Polynesian period linked to sea level changes. Variation in pre-Polynesian to Polynesian period macrobotanical and pollen sequences illustrates the removal of local forests and management of anthropogenic landscapes associated with shifting cultivation and arboriculture. At the same time, sedimentary records demonstrate massive slope erosion and accumulation of the coastal plain, in effect, creating a new landscape for residential and agricultural activities in the last c. 500 years. Anthropogenic activities had far reaching consequences, including the introduction of Polynesian landsnails and insects and the eventual extinction of endemic species. The Mo‘orea record illustrates the range of significant environmental changes, both natural and anthropogenic, operating on Society Island coastlines for the last four millennia. These processes have significantly obscured traces of early Polynesian 346 settlement of the island. In order to better understand the early colonization sequence in the Windward Society Islands, it will be necessary to seek out buried coastal contexts or, alternatively, protected landscapes where early settlement deposits may be preserved. Acknowledgements Kahn and Kirch co-directed the fieldwork and drafted the manuscript. Laboratory analysis of sediment were supervised by Kirch and Nickelsen, analysis of pollen by Stevenson, of wood charcoal by Dotte, of faunal remains by Kahn, of macrobotanical remains by Kahn and May, and of insect remains by Porch (supported by an Australian Research Council DECRA Fellowship (DE130101453). We express our appreciation to the following individuals who assisted or facilitated our research: Chantal Tahiata, Ministèrè de la Culture, de l’Artisanat, et de la Famille; Teddy Tehei, Chef de Service, Service de la Culture et du Patrimoine; Priscille Frogier, Chef de Service, Délégation à la Recherche de la Polynésie Française; M. Van Bastolaer, le Maire de Mo‘orea-Maiao; and Neil Davies, Director, Gump Research Station. Bellona Mou and Tamara Maric of the Service de la Culture et du Patrimoine facilitated the permit process. Hinano Murphy helped with Tahitian translations, discussions with property owners, and import regulations. Diana Izdebski drafted the figures. We thank the following property owners for permission to work on their lands: the Wilder family, Mari Mari Kellum, Eric de Bovis, Aimata Teariki, Nathalie Terai Richmond, and Adolphe Frogier. Permission to work in ‘Opunohu Valley was granted by Louis Sandford, Director, Domaine d’Opunohu. Funding This project was funded by National Science Foundation collaborative grants BCS-1029765 to Kahn and BCS-1030049 to Kirch. References Allen MS (2014) Marquesan colonisation chronologies and post-colonisation interaction: Implications for Hawaiian origins and the ‘Marquesan Homeland’ hypothesis. Journal of Pacific Archaeology 5(2): 1–17. Allen MS and Huebert JM (2014) Short-lived plant materials, long-lived trees, and Polynesian 14C dating: Considerations for 14C sample selection and documentation. Radiocarbon 56(1): 257–276. Anderson A and Sinoto YH (2002) New radiocarbon ages of colonization sites in East Polynesia. Asian Perspectives 41: 242–257. Bronk Ramsey C (2009) Bayesian analysis of radiocarbon dates. Radiocarbon 51(1): 337–360. Butaud J-F, Gérard J and Guibal D (2008) Guides des arbres de Polynésie française: Bois et utilisations. Tahiti: Au Vent Des Iles. Christensen CC and Weisler MI (2013) Land snails from archaeological sites in the Marshall Islands, with remarks on prehistoric translocations in tropical oceania. Pacific Science 67: 81–104. Craig DA (1987) A taxonomic account of the Black Flies (Diptera: Simuliidae) of the Society Islands – Tahiti, Moorea and Raiatea. Quaestiones Entomologicae 23: 372–429. Craig DA and Porch N (2013) Subfossils of extinct and extant species of Simuliidae (Diptera) from Austral and Cook Islands (Polynesia): Anthrogenic extirpation of an aquatic insect? Zootaxa 3641(4): 448–462. Cremaschi M, Zerboni A, Mercuri AM et al. (2014) Takarkori rock shelter (SW Libya): An archive of Holocene climate and environmental changes in the central Sahara. Quaternary Science Reviews 101: 36–60. The Holocene 25(2) Dean WE Jr (1974) Determination of carbonate and organic matter in calcareous sediments and sedimentary rocks by loss on ignition: Comparison with other methods. Journal of Sedimentary Petrology 44: 242–248. Dearing JA (2008) Landscape change and resilience theory: A palaeoenvironmental assessment from Yunnan, SW China. The Holocene 18(1): 117–127. Dearing JA, Yang X, Dong X et al. (2012) Extending the timescale and range of ecosystem services through paleoenvironmental analyses, exemplified in the lower Yangtze Basin. Proceedings of the National Academy of Sciences of the United States of America 109(18): E1111–E1120. Dickinson WR (2001) Paleoshoreline record of relative Holocene sea levels on Pacific islands. Earth-Science Reviews 55(3): 191–234. Dotte-Sarout E (2010) ‘The Ancestor Wood’ (Trees, Forests and Precolonial Kanak Settlement on New Caledonia Grande Terre, 3 volumes including Wood Atlas). Unpublished PhD Thesis, Université Paris I Sorbonne, The Australian National University. Emory KP (1933) Stone Remains in the Society Islands: Bernice P. Bishop Museum Bulletin, vol. 119. Honolulu, HI: Bernice P. Bishop Museum. Fisher CT (2005) Demographic and landscape change in the Lake P´atzcuaro Basin, Mexico: Abandoning the garden. American Anthropologist 107(1): 87–95. Fisher CT and Feinman G (2005) Introduction to ‘Landscapes over Time’. American Anthropologist 107(1): 62–69. Florence J (1997) Flore de la Polynésie Française, vol. 1. Paris: IRD Editions, publications scientifiques du Museum. Florence J (2004) Flore de la Polynésie Française, vol. 2. Paris: IRD Editions, publications scientifiques du Museum. Ford A and Nigh R (2009) Origins of the Maya Forest garden: May resource management. Journal of Ethnobiology 29(2): 213–236. Garb J (2006) A new species of thomisid spider (Araneae, Thomisidae) from the Society Islands with a description of the male of Misumenops melloleitaoi. Journal of Arachnology 34: 357–367. Garrett A (1884) The terrestrial Mollusca inhabiting the Society Islands. Journal of the Academy of Natural Sciences of Philadelphia 2(9): 17–114, 2–3. Green RC (1996) Settlement patterns and complex society in the windward Society Islands: Retrospective commentary from the ’Opunohu Valley, Mo’orea. In: Julien M, Orliac M and Orliac C (eds) Mémoire de Pierre, Mémoire d’Homme: Tradition et Archéologie en Océanie. Paris: Publications de la Sorbonne, pp. 209–228. Green RC, Green K, Rappaport RA et al. (1967) Archeology on the island of Mo’orea, French Polynesia. Anthropological Papers of the American Museum of Natural History 51(2): 111–230. Hamilton B and Kahn J (2007) Pre-contact population in the ’Opunohu Valley, Mo’orea: An integrated archaeological and ethnohistorical approach. In: Kirch PV and Rallu J-L (eds) The Growth and Collapse of Pacific Island Societies: Archaeological and Demographic Perspectives. Honolulu, HI: University of Hawaii Press, pp. 129–159. Kahn JG (2005a) Household and community organization in the late prehistoric Society Islands (French Polynesia). PhD Dissertation, Anthropology, University of California; University Microfilms. Kahn JG (2005b) An archaeological survey of the upper Amehiti sector, ‘Opunohu Valley, Mo’orea, Society Islands. In: Marchesi H (ed.) Bilan de la recherche archéologique en Polynésie française 2003–2004. Punaauia: Service de la Culture et du Patrimoine, Dossier d’Archeologie polynésienne, vol. 4, pp. 33–40. Kahn et al. Kahn JG (2006) Society Islands (Central Eastern Polynesia) chronology: 11 radiocarbon dates for the late prehistoric expansion and proto-historic periods in the ’Opunohu Valley, Mo’orea. Radiocarbon 48(3): 409–419. Kahn JG (2007) Power and precedence in ancient house societies: A case study from the Society Island chiefdoms (French Polynesia). In: Beck R (ed.) The Durable House: House Society Models in Archaeology (Carbondale: Center for Archaeological Investigations Occasional Paper No. 35). Carbondale, IL: Southern Illinois University, pp. 198–223. Kahn JG (2011) Multi-phase construction sequences and aggregate site complexes of the prehistoric windward Society Islands (French Polynesia). Journal of Island and Coastal Archaeology 6: 24–50. Kahn JG (2012) Coastal occupation at the GS-1 Site, Cook’s Bay, Mo’orea, Society Islands. Journal of Pacific Archaeology 3(2): 1–10. Kahn JG (2013) Temple renovations, aggregate Marae, and ritual centers: The ScMo-15 Complex, Lower Amehiti District, ’Opunohu Valley, Mo’orea (Society Islands). Rapa Nui Journal 27(2): 33–49. Kahn JG and Kirch PV (2011) Monumentality and the materialization of ideology in Central Eastern Polynesia. Archaeology in Oceania 46: 93–104. Kahn JG and Kirch PV (2013) Residential landscapes and house societies of the late prehistoric Society Islands (French Polynesia). Journal of Pacific Archaeology 4(1): 50–72. Kahn JG, Horrocks M and Nieuwoudt M (2014) Agriculture, domestic production, and site function: Micro-fossil analyses and late prehistoric landscapes of the Society Islands. Economic Botany 68: 246–263. Kelley AR (2014) The archaeological record as a paleoclimate and paleoenvironmental archive. Paper presented at 2014 Geological Science Meeting, Vancouver, 19–22 October. Kirch PV (1986) Rethinking East Polynesian prehistory. Journal of the Polynesian Society 95: 9–40. Kirch PV (2005) Archaeology and global change: The Holocene record. Annual Review of Environmental Resources 30: 409–440. Kirch PV, Hartshorn A, Chadwick O et al. (2004) Environment, agriculture, and settlement patterns in a marginal Polynesian Landscape. Proceedings of the National Academy of Sciences of the United States of America 101: 9936–9941. Lawrence JF, Hastings AM, Dallawitz MJ et al. (1999) Beetles of the World, Ver. 1.0. Canberra: SIRO Division of Entomology. Lepofsky D (1994) Prehistoric agricultural intensification in Society Islands, French Polynesia. PhD Dissertation, Anthropology, University of California; University Microfilms. Lepofsky D (2003) The ethnobotany of cultivated plants of the Maohi of the Society Islands. Economic Botany 57(1): 73–92. Lepofsky D and Kahn JG (2011) Cultivating an ecological and social balance: Elite demands and commoner knowledge in ancient Ma’ohi agriculture, Society Islands. American Anthropologist 113(2): 319–335. Lepofsky D, Harries HC and Kellum M (1992) Early coconuts on Mo’orea Island, French Polynesia. Journal of the Polynesian Society 101: 299–308. Lepofsky D, Kirch PV and Lertzman K (1996) Stratigraphic and paleobotanical evidence for prehistoric human-induced environmental disturbance on Mo’orea, French Polynesia. Pacific Science (50)3: 253–273. Lubos CCM, Dreibrodt S, Robin V et al. (2013) Settlement and environmental history of a multilayered settlement mound in Niederröblingen (central Germany): A multi-proxy approach. Journal of Archaeological Science 40: 79–98. Maury R-C, Le Dez A and Guillou H (2000) Notice Explicative, Carte Ggéol. France (1/25 000), Feuille Moorea, Polynésie Française. Orléans: BRGM. 347 Mulrooney MA, Bickler SH, Allen MS et al. (2011) High-precision dating of colonization and settlement in East Polynesia. Proceedings of the National Academy of Sciences of the United States of America 108: E192–E194. Parkes A (1997) Environmental change and the impact of Polynesian colonization: Sedimentary records from central Polynesia. In: Kirch PV and Hunt TL (eds) Historical Ecology in the Pacific Islands: Prehistoric Environmental and Landscape Change. New Haven, CT: Yale University Press, pp. 166–199. Pirazzoli PA, Montaggioni LF and Faure G (1985) late Holocene sea-level changes in the Society Islands and in the Northwest Tuamotu atolls. In: Proceedings of the 5th International Reef Congress, Tahiti, 27 May–1 June. Porch N (2008) Invertebrates in landscape archaeology. In: David B and Thomas J (eds) Handbook of Landscape Archaeology (World Archaeological Congress, Research Handbook Series). Walnut Creek, CA: Left Coast Press, pp. 457–467. Prebble M (2008) No fruit on that beautiful shore: What plants were introduced to the subtropical Polynesian islands prior to European contact? In: Clark G, Leach F and O’Connor S (eds) Islands of Inquiry: Colonisation, Seafaring and the Archaeology of Maritime Landscapes (Terra Australis 29). Canberra: ANU E Press, pp. 227–251. Rieth T and Athens JS (2013) Suggested best practices for the application of radiocarbon dating to Hawaiian archaeology. Hawaiian Archaeology 13: 3–29. Rosen AM and Rivera-Collazo I (2012) Climate change, adaptive cycles, and the persistence of foraging economies during the late Pleistocene/Holocene transition in the Levant. Proceedings of the National Academy of Sciences 109(10): 3640–3645. Sandweiss DH and Kelley AR (2012) Archaeological Contributions to Climate Change Research: The Archaeological Record as a Paleoclimatic and Paleoenvironmental Archive. Annual Review of Anthropology 41: 371–391. Sharp WD, Kahn JG, Polita CM et al. (2010) Rapid evolution of ritual architecture in central Polynesia indicated by precise 230Th/U coral dating. Proceedings of the National Academy of Sciences of the United States of America 107(30): 13234–13239. Théry-Parisot I, Chabal L and Chrzavzez J (2010) Anthracology and taphonomy, from wood gathering to charcoal analysis. A review of the taphonomic processes modifying charcoal assemblages, in archaeological contexts. Palaeogeography, Palaeoclimatology, Palaeoecology 291: 142–153. Van der Leeuw S (1998) The Archaeomedes Project: Understanding the natural and anthropogenic causes of land degradation and desertification in the Mediterranean Basin. Luxembourg: Office for the Official Publications of the European Union. Van der Leeuw S and Redman CL (2002) Placing archaeology at the center of socio-natural studies. American Antiquity 67(4): 597–605. Varien MD, Ortman SA, Kohler TA et al. (2007) Historical ecology in the Mesa Verde region: Results from the village ecodynamics project. American Antiquity 72: 273–299. Wheeler E, Baas P and Gasson P (1989) IAWA list of microscopic features for hardwood identification. IAWA Journal 10: 219–332. Wilkinson TJ, McGuire G, Christiansen J et al. (2007) Modeling Settlement Systems in a Dynamic Environment: Case Studies from Mesopotamia. In: Kohler TA and van der Leeuw S (eds) The Model-Based Archaeology of Socionatural Systems. Santa Fe, NM: School of American Research, pp.166–199. Wilmshurst JM, Hunt TL, Lipo CP et al. (2011) High-precision radiocarbon dating shows recent and rapid initial human colonization of East Polynesia. Proceedings of the National Academy of Sciences of the United States of America 108(5): 1815–1820. Copyright of Holocene is the property of Sage Publications, Ltd. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.