A record of Holocene environmental and ecological changes from Wildwood Lake, Long Island, New York
W. Wyatt Oswald1,2, David R. Foster2, Elaine D. Doughty2, and Dana MacDonald3 1Emerson College, Boston, MA 02116, USA 2Harvard Forest, Harvard University, Petersham, MA 01366, USA 3Department of Geological Sciences, Brown University, Providence, RI 02912, USA Corresponding author: W. Wyatt Oswald Emerson College Boston, MA 02116, USA Phone: 617-824-3502 Fax: 617-824-8735 Email: w_wyatt_oswald@emerson.edu For resubmission to Journal of Quaternary Science; December 11, 2009 RUNNING HEAD: Holocene lake-sediment record from Long Island, New York KEYWORDS: charcoal; fire; paleoecology; pitch pine; pollen
1

1 ABSTRACT 2 Analyses of pollen, charcoal, and organic content in a lake-sediment core from Wildwood 3 Lake, Long Island, New York, provide insights into the ecological and environmental history of 4 this region. The early-Holocene interval of the record (~9800-8800 cal. a BP) indicates the 5 presence of Pinus rigida-Quercus ilicifolia woodlands with high fire activity. A layer of sandy 6 sediment dating to 9200 cal. a BP may reflect a brief period of reduced water depth, consistent 7 with widespread evidence for cold, dry conditions at that time. Two other sandy layers, 8 bracketed by 14C dates, represent a sedimentary hiatus from ~8800 to 4500 cal. a BP. This 9 discontinuity may reflect the removal of some sediment during brief periods of reduced water 10 depth at 5300 and 4600 cal. a BP. In the upper portion of the record (<4500 cal. a BP), subtle 11 changes at ~3000 cal. a BP indicate declining prevalence of Quercus-Fagus-Carya forests and 12 increasing abundance of Pinus rigida, perhaps due to reduced summer precipitation. Elevated 13 percentages of herbaceous taxa in the uppermost sediments represent European agricultural 14 activities. However, unlike charcoal records from southern New England, fire activity does not 15 increase dramatically with European settlement. These findings indicate that present-day Pinus 16 rigida-Quercus ilicifolia woodlands on eastern Long Island are not a legacy of recent, 17 anthropogenic disturbances. 18 19 INTRODUCTION 20 Numerous paleoecological and paleoenvironmental records provide a broad view of the 21 post-glacial vegetation and climate history of the northeastern United States (“the Northeast”; 22 e.g., Bernabo and Webb, 1977; Gaudreau and Webb, 1985; Williams et al., 2001; Huang et al., 23 2002; Shuman et al., 2002a; 2002b; 2004; 2005; 2006). Some parts of the region, however, have
2

1 received less recent study and thus aspects of their environmental history are not well 2 understood. Long Island, New York, is one area where additional information about past 3 ecological and climatic changes is needed, in particular to improve our understanding of the 4 following: (1) the long-term history of Pinus rigida (pitch pine) woodlands in central and 5 eastern Long Island; (2) the distribution of Fagus grandifolia (American beech) along the 6 Northeast coast during the middle Holocene; and (3) the timing and magnitude of century-scale 7 variations in climate in this region during the Holocene. 8 There are various unresolved questions regarding the origin and history of forests 9 dominated by fire-dependent Pinus rigida and Quercus ilicifolia (scrub oak) on Long Island 10 (e.g., Backman, 1984; Kurczewski and Boyle, 2000; Jordan et al., 2003), which provide habitat 11 for a range of rare and endangered species (CPBJPPC, 1995). It is uncertain whether the Pinus 12 rigida woodlands of eastern Long Island have been maintained for millennia by frequent fire, or 13 if they have expanded more recently in response to European settlement, forest clearance, and 14 burning (e.g., Kurczewski and Boyle, 2000). Pinus rigida communities are threatened by 15 development and fire suppression, and their conservation, management, and restoration should 16 be guided by information about their past (e.g., CPBJPPC, 1995; Jordan et al., 2003; Parshall et 17 al., 2003). 18 Recent paleoecological studies on Cape Cod, Massachusetts, and adjacent islands (Fig. 1) 19 have documented the widespread occurrence of an unusual coastal vegetation type during the 20 middle Holocene (Foster et al., 2006). Beginning with the abrupt collapse of Quercus 21 populations ~5500 calibrated 14C years before present (cal. a BP), a portion of the coast of 22 southern New England was dominated by Fagus grandifolia. The geographic extent of coastal 23 Fagus forests, which persisted until ~3000 cal. a BP, has not been determined for other parts of
3

1 the Northeast coast, including Long Island. Several pollen records published some four decades 2 ago (Sirkin, 1967; 1971) suggest that the post-glacial vegetation history of Long Island 3 resembles that of much of southern New England, which can be explained in terms of orbital4 scale changes in climate (e.g., Davis, 1969; Suter, 1985; Gaudreau, 1986; Shuman et al., 2001; 5 2004; Oswald et al., 2007). Late-glacial forests featured cold-tolerant Picea (spruce) and Pinus 6 banksiana (jack pine), warmer and drier conditions during the early Holocene resulted in 7 dominance by Pinus strobus (white pine), and wetter conditions after ~8000 cal. a BP promoted 8 the expansion of various temperate forest taxa, including Quercus. There is no indication of a 9 middle-Holocene interval dominated by Fagus grandifolia on Long Island (Sirkin, 1967; 1971), 10 but the coarse sampling resolution and very limited chronological control restrict the utility of 11 the existing Long Island records. 12 Recent studies in southeastern Massachusetts, located northeast of Long Island (Fig. 1), 13 show that lake-sediment records from the Northeast feature not only the orbital-scale changes in 14 climate responsible for the broad sequence of vegetation shifts (e.g., Shuman et al., 2001; 2004), 15 but also century-scale intervals of reduced water depth evidenced by sand layers in near-shore 16 cores (e.g., Newby et al., 2009; Shuman et al., 2009). Relatively brief periods of dry conditions 17 associated with changing ocean conditions appear to have occurred at ~9200, 8300, 5300, 4600, 18 3800, 2900, 2200, and 1500 cal. a BP and may have served as important drivers for some aspects 19 of the vegetation history of the region (Shuman et al., 2004; 2009; Newby et al., 2009). 20 In this paper we discuss a detailed lake-sediment record from Wildwood Lake, located in 21 eastern Long Island (Fig. 1) near the Central Pine Barrens, the largest area of Pinus rigida22 Quercus ilicifolia woodlands on Long Island (CPBJPPC, 1995; Jordan et al., 2003). Preliminary 23 data from this record were reported elsewhere (Foster et al., 2006). Our analyses of charcoal,
4

1 pollen, and organic content in the Wildwood record provide additional evidence for millennial2 to centennial-scale variations in climate and ecosystems in the Northeast (e.g., Newby et al., 3 2009; Shuman et al., 2009) and contribute to an improved understanding of the origin and 4 dynamics of Pinus rigida forests on Long Island. 5 6 STUDY AREA 7 Wildwood Lake (40.892°N, 72.673°W) is a 26-ha lake located on eastern Long Island, 8 New York; a single outlet to the north flows into the Peconic River. The lake features two basins 9 separated by a shallow shoal: the northern basin is ~3-m in depth; the southern basin is ~15-m in 10 depth at its deepest point, which is near the southernmost lakeshore (Fig. 2). Much of the 11 lakeshore currently is occupied by residential and commercial development. The lake sits on the 12 Ronkonkoma Moraine (e.g., Fuller, 1914; Sirkin, 1971), which is bordered to the north and south 13 by sandy outwash plains (Fig. 1). The dominant soils of the area are Plymouth loamy sand and 14 Carver and Plymouth sands, both of which have relatively low moisture-holding capacity 15 (USDA, 1972; 1975; Kurczewski and Boyle, 2000). Historical maps and witness-tree data 16 (Kurczewski and Boyle, 2000; Cogbill et al., 2002) suggest that outwash plains featured Pinus 17 rigida, Quercus ilicifolia, and some tree oaks (e.g., Q. alba, Q. coccinea, and Q. velutina) at the 18 time of European settlement in the mid-seventeenth century; moraines appear to have been 19 dominated by tree oaks (e.g., Q. rubra, Q. prinus, and Q. velutina), with lesser amounts of Pinus 20 rigida, Castanea dentata (American chestnut), Carya glabra (Pignut hickory), and Fagus 21 grandifolia. The pollen source area (Sugita, 1994) for this relatively large lake likely includes 22 areas of both moraine and outwash. A charcoal record from nearby Deep Pond (Fig. 1) indicates 23 that fires burned periodically during the late Holocene, increasing in frequency and intensity
5

1 after European settlement (Backman, 1984). Historical records show that large fires burned 2 across Long Island in the late-eighteenth and nineteenth centuries (Kurczewski and Boyle, 2000). 3 4 METHODS 5 We recovered a sediment core from Wildwood Lake in July 2001. The core was 6 collected near the center of the southern basin of the lake, ~100 m northwest of its deepest point; 7 water depth at the coring location was 14 m (Fig. 2). Coring ceased at 765-cm sediment depth 8 due to mechanical problems. The uppermost sediments (165 cm) were obtained with a 7-cm9 diameter plastic tube fitted with a piston. Lower sediments were collected in 1-m drive lengths 10 with a 5-cm-diameter modified Livingstone piston sampler. The surface core was extruded 11 vertically in 1-cm intervals; lower cores were extruded horizontally and wrapped in plastic and 12 aluminum foil. 13 Sediment organic content was estimated by measuring percent weight loss-on-ignition 14 (LOI); 1-cm3 samples were dried at 90°C and ashed at 550°C (Bengtsson and Enell, 1986). 15 Macroscopic charcoal content of the sediment was determined by soaking 1-cm3 samples in 16 KOH, sieving through a 180-μm screen, and counting charcoal pieces >180-μm at 20X 17 magnification (Long et al., 1998). Samples of 1 cm3 from 116 levels were prepared for pollen 18 analysis following standard procedures (Faegri and Iversen, 1989), and tablets containing 19 Lycopodium spores were added to the samples to estimate concentrations and influx (Stockmarr, 20 1971). Pollen residues were analyzed at 400X magnification until a minimum of 500 pollen 21 grains and spores of upland taxa was identified for each level. Pollen percentages are expressed 22 relative to the sum of upland tree, shrub, and herb pollen. Only ~30% of Pinus pollen grains 23 were identifiable to the subgenus level, but ~90% of the identified grains were Pinus subgenus
6

1 Pinus. We therefore assume that most of the Pinus pollen in the Wildwood record is derived 2 from Pinus rigida. 3 Fossil-pollen assemblages were compared with modern-pollen spectra from eight sites 4 (Figs. 1 and 3) representing the three primary types of upland forest for Long Island and the 5 coastal area of southern New England (Motzkin et al., 2002). Three sites from Cape Cod are 6 dominated by Pinus rigida (Eagle, Fresh, and Round Ponds; Parshall et al., 2003), two Cape Cod 7 sites are dominated by Quercus species (Deep and Sandy Hill Ponds; Parshall et al., 2003), and 8 three sites from Naushon Island are dominated by Fagus grandifolia (Blaney’s Pond, Duck 9 Pond, and Mary’s Lake; W. Oswald and D. Foster, unpublished). Relationships between fossil 10 and modern assemblages were described by the squared-chord distance index of dissimilarity 11 (SCD; Overpeck et al., 1985). Six taxa were included in this analysis: Pinus, Quercus, Fagus, 12 Carya, Betula, and Tsuga. The majority of identified Pinus grains in these modern samples were 13 Pinus subgenus Pinus. Previous studies have shown that SCD values <0.1 indicate a strong 14 degree of similarity in the composition of the source vegetation (e.g., Overpeck et al., 1985). 15 Chronological control is provided primarily by accelerator mass spectrometry 14C 16 analysis of bulk sediment samples, each spanning 1 cm core depth (Table 1). 14C dates were 17 converted to calibrated years before present (cal. a BP) with CALIB 5.0 (Reimer et al., 2004). 18 The chronology of the upper sediments is based on 210Pb analysis (Table 2; Binford, 1990) and 19 pollen stratigraphy: the rise in agricultural taxa at 130 cm (Fig. 4B, 5A) was assigned an age of 20 300 cal. a BP (1650 AD; Kurczewski and Boyle, 2000). 21 22 RESULTS 23 Lithology and chronology
7

1 The Wildwood core is lacustrine sediment (LOI ~30%), with the exception of three sandy 2 layers (LOI <25%): 650-646 cm, 592-580 cm, and 562-554 cm (Fig. 4A, 4C). Organic content 3 also declines above 130 cm during the interval of European settlement. The five 14C ages below 4 599 cm (Table 1) indicate that the lower interval of the record was deposited quickly (~160 cm 5 in ~1000 years) and a straight line fit to the median calibrated 14C ages dates the sandy layer at 6 ~650 cm to 9200 cal. a BP (Fig. 4B). A preliminary chronology for this record presented in a 7 regional summary (Foster et al., 2006) assumed that this feature represented the 8200 cal. a BP 8 cold event (e.g., Alley et al., 1997; Alley and Ágústsdóttir, 2005), which appears as an interval of 9 reduced organic content in records from elsewhere in eastern North America (e.g, Spooner et al., 10 2002; Kurek et al., 2004). However, additional 14C dating and emerging recognition of a 11 climatic event at ~9200 cal. a BP at many sites worldwide (e.g., Fleitmann et al., 2008), 12 including southeastern Massachusetts (Newby et al., 2009; J. Hou and Y. Huang, unpublished), 13 support the updated age-depth curve presented here (Fig. 4B). The dates bracketing the sandy 14 layers at ~590 and ~560 cm suggest that these features represent a sedimentary hiatus from 15 ~8800 to ~4500 cal. a BP. The age model for the upper interval of the core (~4500 cal. a BP to 16 the present) involves a straight line fit to the European-settlement horizon and the median 17 calibrated ages of the six 14C dates above 571 cm, and second-order polynomial fit to the 18 European-settlement horizon and the sixteen 210Pb dates (Tables 1 and 2, Fig. 4B). 19 20 Pollen and charcoal records 21 The lower interval of the Wildwood core (~9800-8800 cal. a BP; Fig. 5C) features high 22 percentages of Quercus (~45%) and Pinus pollen (~35%), with lower percentages of Tsuga 23 (hemlock), Betula (birch), Myrica-Comptonia (bayberry-sweetfern), Poaceae (grass), and
8

1 Ambrosia (ragweed). These fossil assemblages are similar to modern spectra from the Quercus2 (SCD ~0.08) and Pinus-dominated sites (SCD ~0.1; Fig 6E) and dissimilar to those from Fagus 3 grandifolia-dominated sites (SCD ~0.2). The highest value for Pinus pollen (>50%) falls within 4 the sandy layer at ~650 cm that dates to 9200 cal. a BP. Charcoal influx values fluctuate 5 between ~40 and 80 pieces cm-2 yr-1, and Pediastrum influx values vary between ~100 and 2900 6 algal cell nets cm-2 yr-1 (Fig. 6D). 7 The upper interval of the record (~4500 cal. a BP to the present; Fig. 5B) features 8 gradually increasing Pinus pollen percentages (~20 to 40%) and gradually decreasing Quercus 9 values (~50 to 30%). Several taxa have higher pollen percentages in the upper interval than in 10 the early-Holocene part of the record, including Betula, Acer (maple), Fagus, Carya, and 11 Castanea. Tsuga pollen-percentage values increase from ~0 to 5% between ~4000 and 2000 cal. 12 a BP; Fagus and Carya pollen percentages decrease slightly during that interval. Poaceae and 13 Ambrosia pollen percentages are lower between ~4500 cal. a BP and European settlement than 14 they were during ~9800-8800 cal. a BP. The fossil assemblages from ~4500 to 2000 cal. a BP 15 are similar to all three modern-forest types (SCD ~0.06-0.1), but the similarity to the Fagus 16 grandifolia-dominated sites declines (SCD ~0.1-0.2) and the similarity to modern Pinus samples 17 increases (SCD ~0.05) after ~2000 cal. a BP (Fig. 6E). Charcoal influx values for the middle 18 and late Holocene are ~5-10 pieces cm-2 yr-1. Sediment organic content is high between ~4400 19 and 3000 cal. a BP, peaking at 3800 cal. a BP, then peaks again at ~2000 cal. a BP (Fig. 6B). 20 Influx of Pediastrum algal cell nets declines between ~4000 and 3000 cal. a BP (Fig. 6D). 21 The changes at 130 cm (300 cal. a BP) represent the era of European settlement and 22 forest clearance (Fig. 5A). A sharp drop in organic content is presumably attributable to 23 increased erosion of mineral material into the lake (Fig. 6B), and a rise in Pediastrum influx may
9

1 indicate changes in the aquatic ecosystem in response to disturbance in the watershed (Fig. 6D; 2 e.g., Jankovska and Komarek, 2000). Rumex (sorrel), Poaceae, and Ambrosia increase 3 dramatically, while Carya, Quercus, and especially Pinus exhibit declining abundances at the 4 beginning of this interval. Decreasing percentages of the herbaceous taxa and increasing Pinus 5 pollen percentages (~20 to 50%) in levels dating to ~1950 AD reflect the decline of agricultural 6 activity and reforestation. Charcoal influx values increase from ~5 to 15 pieces cm-2 yr-1 at ~100 7 cal. a BP (1850 AD), then reach a peak of ~20-35 pieces cm-2 yr-1 in levels dating to ~1920-1970 8 AD (Fig. 5A). 9 10 DISCUSSION 11 The Wildwood Lake record provides new insights into past environmental and ecological 12 changes on Long Island, New York, including century-scale variations in climate and the long13 term history of vegetation and fire. 14 15 Early-Holocene vegetation and fire 16 This sediment core from Wildwood Lake contains only a ~1000 year interval of the early 17 Holocene, but nonetheless does contribute to our understanding of Long Island ecosystems 18 between ~9800 and 8800 cal. a BP. The prevalence of Pinus and Quercus, paucity of mesic19 forest taxa such as Fagus grandifolia and Carya, relatively high abundance of Ambrosia, and 20 very high charcoal influx values suggest open Pinus-Quercus woodlands with frequent fire on 21 the Ronkonkoma moraine and adjacent areas of outwash. Similar pollen assemblages in early22 Holocene samples from other sites on Long Island (Sirkin, 1967; 1971) indicate that this type of 23 vegetation was widespread. High fire activity and open forest structure, as evidenced by
10

1 abundant Ambrosia (Faison et al., 2006), are consistent with evidence suggesting that regional 2 climate was relatively dry during this interval (Fig. 6A; Shuman et al., 2001; 2004; 2009; Newby 3 et al., 2009). 4 5 Evidence for century-scale climatic events 6 The decline in organic content that is dated securely to 9200 cal. a BP likely represents an 7 interval of particularly dry climate and reduced water levels at Wildwood Lake. A similar 8 feature occurs in the sediments of New Long Pond in southeastern Massachusetts (Newby et al., 9 2009) and geochemical data from Blood Pond in south-central Massachusetts indicate a ~4°C 10 temperature decline at that time (J. Hou and Y. Huang, unpublished). Emerging evidence 11 suggests that a widespread climatic event took place at ~9200 cal. a BP, presumably caused by a 12 weakening of Atlantic meridional overturning circulation (e.g., Fleitmann et al., 2008). The peak 13 in Pinus pollen abundance at 9200 cal. a BP in the Wildwood Lake record may represent a 14 change in vegetation in response to cold, dry conditions. This finding is consistent with other 15 evidence for century-scale ecological changes associated with early-Holocene variations in 16 climate (e.g., Shuman et al., 2004). 17 The organic-content and 14C data demonstrate that the Wildwood core features a >4000 18 year hiatus between ~8800 and ~4500 cal. a BP. One interpretation of this finding is that climate 19 was dry and water levels were low throughout this entire period. Webb et al. (1993) offered a 20 similar interpretation based on lake-sediment data from other sites in southern New England. 21 More-recent reconstructions of lake-level changes for sites across the Northeast (e.g., Lavoie and 22 Richard 2000; Newby et al., 2000; Dieffenbacher-Krall and Nurse 2005; Shuman et al. 2005), 23 including Crooked Pond in southeastern Massachusetts (Fig. 6A; Shuman et al., 2001), however,
11

1 indicate that the period from ~8000 to 5500 cal. a BP was relatively wet, suggesting that the 2 sedimentary hiatus may not simply represent a continuous interval of dry conditions. Instead, we 3 hypothesize that the pronounced dry events recorded at New Long Pond at ~5300 and 4600 cal. a 4 BP (Shuman et al., 2009) occurred across the region and also caused major water-level 5 reductions at Wildwood Lake. If the portion of the lake bottom cored for this study were 6 substantially shallower than at present, perhaps by >10 m, older sediments could have been re7 deposited to the deeper part of the lake, resulting in the loss of that interval of the record from 8 ~8800 cal. a BP to the time when deeper water levels once again reached this site. The sandy 9 layers at ~590 and ~560 cm presumably were re-deposited from the exposed lakeshore and 10 shoal; the 14C dates indicate that those sediments were mixed. Our data suggest that the 11 accumulation of lake sediment resumed ~4500 cal. a BP, a date consistent with the lake-level 12 reconstructions for Crooked Pond (Shuman et al., 2001) and New Long Pond (Shuman et al., 13 2009). 14 15 Middle- and late-Holocene vegetation and fire 16 The similarity of middle-Holocene pollen spectra to modern samples from three different 17 types of vegetation (Pinus, Quercus, and Fagus) may be attributable to the large pollen source 18 area of Wildwood Lake. Presumably, the lake receives some pollen from the local vegetation on 19 the Ronkonkoma moraine, as well as some pollen from vegetation on the nearby outwash plains. 20 The similarity of middle-Holocene pollen assemblages to modern Pinus samples suggests that 21 areas of sandy soil were dominated by Pinus rigida, as seems to have been the case during the 22 early Holocene. The forests of the moraine itself, on the other hand, appear to have featured 23 Quercus with some Carya and Fagus grandifolia. The abundance of Fagus, however, is
12

1 substantially lower for this interval than in records from Cape Cod and nearby islands, where its 2 pollen percentages reach 30-40% during ~5500-3000 cal. a BP (Foster et al., 2006). Climate 3 appears to have become cooler and wetter between ~4000 and 2000 cal. a BP, as evidenced by 4 rising water levels at Crooked Pond (Shuman et al., 2001), for example. This change may be 5 manifest in the Wildwood Lake record by the declines in Pediastrum influx and organic content 6 at that same time. Several tree taxa exhibit subtle changes in abundance between ~4000 and 7 2000 cal. a BP. Most importantly, Quercus, Carya, and Fagus become somewhat less common, 8 whereas Pinus increases in abundance. We interpret this gradual change as a decline in the 9 prevalence of Quercus-dominated hardwood forests and the coincident expansion of some Pinus 10 rigida on the Ronkonkoma moraine. The declining abundance of Fagus in response to cooler, 11 wetter conditions is counterintuitive given its prevalence in northern parts of New England (e.g., 12 Thompson et al., 1999; Cogbill et al., 2002). Shuman and Donnelly (2006) hypothesize that the 13 seasonality of precipitation may have shifted between the middle and late Holocene, so that even 14 though annual precipitation increased after ~3000 cal. a BP, the summer months were drier. 15 Reduced summer rainfall might explain the decline of Fagus and rise of drought-tolerant Pinus 16 rigida in the Wildwood record. Relatively stable charcoal influx values after ~4000 cal. a BP 17 suggest that fire activity was not altered by these changes in climate and vegetation. This finding 18 contrasts with other fire-history data from eastern North America, which indicate an increase in 19 burning during since ~3000 cal. a BP (e.g., Carcaillet and Richard, 2000; Power et al., 2008). 20 21 Recent changes in fire and Pinus rigida abundance 22 Fire activity, as evidenced by charcoal influx, does not appear to increase at the time of 23 European settlement. This stands in contrast to fire-history data from Deep Pond, located ~15
13

1 km northwest of Wildwood (Fig. 1; Backman, 1984), and elsewhere in southern New England 2 (Parshall and Foster, 2002), which show a rise in burning. Historical accounts suggest that large 3 fires burned across eastern Long Island in the late-eighteenth and nineteenth centuries (e.g., 4 Kurczewski and Boyle, 2000), but the Wildwood charcoal record indicates that those fires were 5 typical for the last few millennia. Indeed, they were probably smaller and less intense than fires 6 of the early Holocene; fire activity on Long Island was very high ~9800-8800 cal. a BP when 7 climate was relatively dry. Historical records suggest that fire frequency and intensity increased 8 when the Long Island Railroad reached eastern Long Island in the mid-nineteenth century 9 (Kurczewski and Boyle, 2000). This change may be reflected in the Wildwood charcoal record 10 by the elevated influx values from ~100-0 cal. a BP (~1850 to 1950 AD). 11 The increase in Pinus and Quercus pollen percentages in the uppermost sediments of 12 Wildwood Lake reflect the establishment of Pinus rigida-Quercus ilicifolia woodlands following 13 the abandonment of agriculture over the last century. Pinus abundance at present appears to be 14 slightly higher than it has been over the last millennium, perhaps indicating that Pinus rigida 15 benefited modestly from the disturbances associate with European settlement, deforestation, and 16 agriculture (Kurczewski and Boyle, 2000). This interpretation is consistent with the ability of 17 Pinus rigida to establish under early-successional conditions (e.g., Motzkin et al., 1996; 18 Windisch, 1999). However, Pinus rigida does not appear to be dramatically more abundant than 19 it was prior to European settlement, suggesting that this vegetation type is not necessarily a 20 legacy of recent anthropogenic disturbances. 21 These analyses of the sedimentary record from Wildwood Lake improve our 22 understanding of Holocene environmental and ecological changes on Long Island, New York, 23 but some questions deserve additional study. Detailed reconstructions of lake-level history (e.g.,
14

1 Shuman et al., 2001; Newby et al., 2009) for this and other sites are needed to test our 2 hypotheses about past changes in climate and water depth, and the analysis of pollen and 3 charcoal data from other sites on Long Island would provide further insights into past spatial 4 patterns of vegetation and fire. 5 6 ACKNOWLEDGEMENTS 7 We thank Brian Hall, Tim Parshall, Glenn Motzkin, and Sarah Truebe for field and 8 laboratory assistance. The manuscript was greatly improved by suggestions from Konrad 9 Gajewski, Marilyn Jordan, and two anonymous reviewers. This project was funded by the 10 National Science Foundation, the A. W. Mellon Foundation, the Nature Conservancy, the 11 National Ocean Sciences Accelerator Mass Spectrometry Facility, and a Huret Faculty 12 Development Award. 13 14
15

1 REFERENCES CITED 2 3 Alley RB, Mayewski PA, Sowers T, Stuiver M, Taylor KC, Clark PU. 1997. Holocene climatic 4 instability: a prominent, widespread event 8200 years ago. Geology 25: 483-486. 5 6 Alley RB, Ágústsdóttir AM. 2005. The 8k event: cause and consequences of a major Holocene 7 abrupt climate change. Quaternary Science Reviews 24: 1123-1149. 8 9 Backman AE. 1984. 1000-year record of fire-vegetation interactions in the northeastern United 10 States: a comparison between coastal and inland regions. M.Sc. thesis, University of 11 Massachusetts, Amherst. 12 13 Bengtsson L, Enell M. 1986. Chemical analysis. In Handbook of Holocene Palaeoecology and 14 Palaeohydrology, Berglund BE (ed). John Wiley and Sons: Chichester, UK; 423-454. 15 16 Bernabo JC, Webb T, III. 1977. Changing patterns in the Holocene pollen record of northeastern 17 North America: a mapped summary. Quaternary Research 8: 64-96. 18 19 Binford MW. 1990. Calculation and uncertainty analysis of 210Pb dates for PIRLA project cores. 20 Journal of Paleolimnology 3: 253-267. 21 22 Carcaillet C, Richard PJH. 2000. Holocene changes in seasonal precipitation highlighted by fire 23 incidence in eastern Canada. Climate Dynamics 16: 549-559. 24 25 Central Pine Barrens Joint Planning and Policy Commission (CPBJPPC). 1995. Central Pine 26 Barrens Comprehensive Land Use Plan, Vol. 1-2, http://www.pb.state.ny.us. 27 28 Cogbill CV, Burk J, Motzkin J. 2002. The forests of presettlement New England, USA: spatial 29 and compositional patterns based on town proprietor surveys. Journal of Biogeography 29: 30 1279-1304. 31 32 Davis MB. 1969. Climatic changes in southern Connecticut recorded by pollen deposition at 33 Rogers Lake. Ecology 50: 409-422. 34 35 Dieffenbacher-Krall AC, Nurse AM. 2005. Late-glacial and Holocene record of lake levels of 36 Mathews Pond and Whitehead Lake, northern Maine, USA. Journal of Paleolimnology 34: 28337 310. 38 39 Faegri K, Iversen J. 1989. Textbook of Pollen Analysis. Fourth edition. John Wiley and Sons, 40 Chichester, UK. 41 42 Faison EK, Foster DR, Oswald WW, Doughty ED, Hansen BCS. 2006. Early-Holocene 43 openlands in southern New England. Ecology 87: 2537-2547. 44 45 Fleitmann D, Mudelsee M, Burns SJ, Bradley RS, Kramers J, Matter A. 2008. Evidence for a 46 widespread climatic anomaly at around 9.2 ka before present. Paleoceanography 23: PA1102.
16

1 2 Foster DR, Oswald WW, Faison EK, Doughty ED, Hansen BCS. 2006. A climatic driver for 3 abrupt mid-Holocene vegetation dynamics and the hemlock decline in New England. Ecology 4 87: 2959-2966. 5 6 Fuller ML. 1914. The Geology of Long Island. United States Geological Survey Professional 7 Paper 82, Washington, DC. 8 9 Gaudreau DC. 1986. Late-Quaternary vegetational history of the Northeast: paleoecological 10 implications of topographic patterns in pollen distributions. Ph.D. thesis, Yale University, New 11 Haven. 12 13 Gaudreau DC, Webb T, III. 1985. Late-Quaternary pollen stratigraphy and isochrone maps for 14 the northeastern United States. In Pollen Records of Late Quaternary North American Sediments, 15 Bryant J, Holloway J (eds). American Association of Stratigraphic Palynologists: Dallas; 24716 280. 17 18 Huang Y, Shuman B, Wang Y, Webb T, III. 2002. Hydrogen isotope ratios of palmitic acid in 19 lacustrine sediments record late-Quaternary climate variations. Geology 30: 1103-1106. 20 21 Jankovska V, Komarek J. 2000. Indicative value of Pediastrum and other coccal green algae in 22 palaeoecology. Folia Geobotanica 35: 59–82. 23 24 Jordan MJ, Patterson WA, III, Windisch AG. 2003. Conceptual ecological models for the Long 25 Island pitch pine barrens: implications for managing rare plant communities. Forest Ecology and 26 Management 185: 151-168. 27 28 Kurek J, Cwynar LC, Spear RW. 2004. The 8200 cal. a BP cooling event in eastern North 29 America and the utility of midge analysis for Holocene temperature reconstructions. Quaternary 30 Science Reviews 23: 627-639. 31 32 Kurczewski FE, Boyle HF. 2000. Historical changes in the pine barrens of central Suffolk 33 County, New York. Northeastern Naturalist 7: 95-112. 34 35 Lavoie M, Richard PJH. 2000. Postglacial water-level changes of a small lake in southern 36 Quebec, Canada. The Holocene 10: 621-634. 37 38 Long CJ, Whitlock C, Bartlein PJ, Millspaugh SH. 1998. A 9000-year fire history from the 39 Oregon Coast Range, based on a high-resolution charcoal study. Canadian Journal of Forest 40 Research 28: 774-787. 41 42 Motzkin G, Eberhardt R, Hall B, Foster DR, Harrod JC, MacDonald D. 2002. Vegetation 43 variation across Cape Cod, Massachusetts: environmental and historical determinants. Journal of 44 Biogeography 29: 1439-1454. 45
17

1 Motzkin G, Foster DR, Allen AE, Harrod JC, Boone RD. 1996. Controlling site to evaluate 2 history: vegetation patterns of a New England sand plain. Ecological Monographs 66: 345-365. 3 4 Newby PE, Killoran P, Waldorf M, Shuman B, Webb T, III, Webb RS. 2000. 14,000 years of 5 sediment, vegetation, and water level changes at Makepeace Cedar Swamp, southeastern 6 Massachusetts. Quaternary Research 53: 352-368. 7 8 Newby PE, Donnelly JP, Shuman BN, MacDonald D. 2009. Evidence of centennial-scale 9 drought from southeastern Massachusetts during the Pleistocene/Holocene transition. 10 Quaternary Science Reviews 28: 1675-1692. 11 12 Oswald WW, Faison EK, Foster DR, Doughty ED, Hall BR, Hansen BCS. 2007. Post-glacial 13 changes in spatial patterns of vegetation across southern New England. Journal of Biogeography 14 34: 900-913. 15 16 Overpeck JT, Webb T, III, Prentice IC. 1985. Quantitative interpretation of fossil pollen spectra: 17 dissimilarity coefficients and the method of modern analogs. Quaternary Research 23: 87-108. 18 19 Parshall T, Foster DR. 2002. Fire on the New England landscape: regional and temporal 20 variation, cultural and environmental controls. Journal of Biogeography 29: 1305-1317. 21 22 Parshall T, Foster DR, Faison E, MacDonald D, Hansen BCS. 2003. Long-term history of 23 vegetation and fire in pitch pine-oak forests on Cape Cod, Massachusetts. Ecology 84: 736-738. 24 25 Power MJ, Marlon J, Ortiz N, Bartlein PJ, Harrison SP, Mayle FE, Ballouche A, Bradshaw 26 RHW, Carcaillet C, Cordova C, Mooney S, Moreno PI, Prentice IC, Thonicke K, Tinner W, 27 Whitlock C, Zhang Y, Zhao Y, Ali AA, Anderson RS, Beer R, Behling H, Briles C, Brown KJ, 28 Brunelle A, Bush M, Camill P, Chu GQ, Clark J, Colombaroli D, Connor S, Daniau A-L, 29 Daniels M, Dodson J, Doughty E, Edwards ME, Finsinger W, Foster D, Frechette J, Gaillard M30 J, Gavin DG, Gobet E, Haberle S, Hallett DJ, Higuera P, Hope G, Horn S, Inoue J, Kaltenrieder 31 P, Kennedy L, Kong ZC, Larsen C, Long CJ, Lynch J, Lynch EA, McGlone M, Meeks S, 32 Mensing S, Meyer G, Minckley T, Mohr J, Nelson DM, New J, Newnham R, Noti R, Oswald W, 33 Pierce J, Richard PJH, Rowe C, Sanchez Goñi MF, Shuman BN, Takahara H, Toney J, Turney 34 C, Urrego-Sanchez DH, Umbanhowar C, Vandergoes M, Vanniere B, Vescovi E, Walsh M, 35 Wang X, Williams N, Wilmshurst J, Zhan JH. 2008. Changes in fire regimes since the Last 36 Glacial Maximum: an assessment based on a global synthesis and analysis of charcoal data. 37 Climate Dynamics 30: 887-907. 38 39 Reimer PJ, Baillie MGL, Bard E, Bayliss A, Beck JW, Bertrand CJH, Blackwell PG, Buck CE, 40 Burr GS, Cutler KB, Damon PE, Edwards RL, Fairbanks RG, Friedrich M, Guilderson TP, Hogg 41 AG, Hughen KA, Kromer B, McCormac FG, Manning SW, Ramsey CB, Reimer RW, Remmele 42 S, Southon JR, Stuiver M, Talamo S, Taylor FW, van der Plicht J, Weyhenmeyer CE. 2004. 43 IntCal04 Terrestrial radiocarbon age calibration, 26 - 0 ka BP. Radiocarbon 46: 1029-1058. 44
18

1 Shuman B, Bravo J, Kaye J, Lynch JA, Newby P, Webb T, III, 2001. Late-Quaternary water2 level variations and vegetation history at Crooked Pond, southeastern Massachusetts. Quaternary 3 Research 56: 401-410. 4 5 Shuman B, Bartlein P, Logar N, Newby P, Webb T, III. 2002a. Parallel climate and vegetation 6 responses to the early Holocene collapse of the Laurentide Ice Sheet. Quaternary Science 7 Reviews 21: 1793-1805. 8 9 Shuman B, Webb T, III, Bartlein P, Williams JW. 2002b. The anatomy of a climatic oscillation: 10 vegetation change in eastern North America during the Younger Dryas chronozone. Quaternary 11 Science Reviews 21: 1763-1916. 12 13 Shuman B, Newby P, Huang Y, Webb T, III. 2004. Evidence for the close climate control of 14 New England vegetation history. Ecology 85: 1297-1310. 15 16 Shuman B, Newby P, Donnelly JP, Tarbox A, Webb T, III. 2005. A record of late-Quaternary 17 moisture-balance change and vegetation response from the White Mountains, New Hampshire. 18 Annals of the Association of American Geographers 95: 237-248. 19 20 Shuman B, Donnelly JP. 2006. The influence of seasonal precipitation and temperature regimes 21 on lake levels in the northeastern United States during the Holocene. Quaternary Research 65: 22 44-56. 23 24 Shuman B, Huang Y, Newby P, Wang Y. 2006. Compound-specific isotopic analyses track 25 changes in the seasonality of precipitation regimes in the northeastern United States at ca. 8200 26 cal. a BP. Quaternary Science Reviews 25: 2992-3002. 27 28 Shuman BN, Newby P, Donnelly, JP. 2009. Abrupt climate change as an important agent of 29 ecological change in the Northeast U.S. throughout the past 15,000 years. Quaternary Science 30 Reviews 28: 1693-1709 31 32 Sirkin LA. 1967. Late-Pleistocene pollen stratigraphy of western Long Island and eastern Staten 33 Island, New York. In Quaternary Paleoecology, Cushing EJ, Wright HE, Jr. (eds). Yale 34 University Press: New Haven; 248-274. 35 36 Sirkin LA. 1971. Surficial glacial deposits and postglacial pollen stratigraphy in central Long 37 Island, New York. Pollen et Spores 13: 93-100. 38 39 Spooner I, Douglas MSV, Terrusi L. 2002. Multiproxy evidence of an early Holocene (8.2 kyr) 40 climate oscillation in central Nova Scotia, Canada. Journal of Quaternary Science 17: 639-645. 41 42 Stockmarr J. 1971. Tablets with spores used in absolute pollen analysis. Pollen et Spores 13: 43 615-621. 44 45 Sugita S. 1994. Pollen representation of vegetation in Quaternary sediments—theory and method 46 in patchy vegetation. Journal of Ecology 82: 881-897.
19

1 2 Suter SM. 1985. Late-glacial and Holocene vegetation history in southeastern Massachusetts: a 3 14,000-year pollen record. Current Research in the Pleistocene 2: 87-89. 4 5 Thompson RS, Anderson KH, Bartlein PJ. 1999. Atlas of relations between climatic parameters 6 and distributions of important trees and shrubs in North America. U. S. Geological Survey 7 Professional Paper 1650, http://pubs.usgs.gov/pp/p1650-a. 8 9 United States Department of Agriculture (USDA). 1972. General Soil Map, Suffolk County, 10 New York. 11 12 United States Department of Agriculture (USDA). 1975. Soil Survey of Suffolk County, New 13 York. 14 15 Webb RS, Anderson KA, Webb T, III. 1993. Pollen response-surface estimates of Late16 Quaternary changes in the moisture balance of the northeastern United States. Quaternary 17 Research 40: 213-227. 18 19 Williams JW, Shuman BN, Webb T, III. 2001. Dissimilarity analyses of late-Quaternary 20 vegetation and climate in eastern North America. Ecology 82: 3346-3362. 21 22 Windisch AG. 1999. Fire Ecology of the New Jersey Pine Plains and Vicinity. Ph.D. thesis, 23 Rutgers University, New Brunswick.
20

Table 1: 14C results for the Wildwood Lake sediment core

Depth (cm) 207 264 417 549 569 571 599 617 654 697 762

14C lab code AA-58100 AA-58101 OS-47543 OS-72387 OS-72388 OS-56478 OS-72389 OS-47544 OS-56479 OS-56480 OS-47545

δ13C -27.98 -27.70 -25.48 -25.43 -26.14 -25.80 -23.81 -24.49 -25.25 -24.87 -24.28

14C date 1300 ± 40 1240 ± 40 3080 ± 40 4510 ± 40 4030 ± 40 3860 ± 35 7980 ± 60 8070 ± 60 8210 ± 45 8440 ± 40 8790 ± 60

Cal age range (2σ) 1143-1301 1068-1270 3171-3383 5039-5310 4417-4784 4156-4412 8645-9004 8720-9240 9021-9302 9333-9531
9563-10,151

Median cal age 1237 1179 3301 5163 4497 4290 8844 8987 9175 9474 9822

Table 2: 210Pb results for the Wildwood Lake sediment core

Depth (cm) 0 4 8 12 16 20 24 28 32 36 40 44 48 52 56 64 72

Age (relative to AD 1950) ± S.D. -50.0 ± 0.6 -46.9 ± 0.6 -42.6 ± 0.6 -38.4 ± 0.6 -33.5 ± 0.6 -29.3 ± 0.6 -24.5 ± 0.6 -18.8 ± 0.6 -10.5 ± 0.6 3.4 ± 0.7 11.0 ± 0.7 18.5 ± 0.8 26.2 ±0.9 33.9 ± 1.0 41.7 ± 1.2 66.5 ± 2.2 95.5 ± 10.6

21

FIGURE CAPTIONS
Figure 1: Map of Long Island, New York and Cape Cod, Massachusetts featuring surficial geology and showing location of Wildwood Lake, other fossil-pollen sites on Long Island (Sirkin, 1967; 1971; Backman, 1984), Cape Cod sites used for modern-analogue analysis (Parshall et al., 2003; W. Oswald and D. Foster, unpublished), and Crooked and New Long Ponds in southeastern Massachusetts (Shuman et al., 2001; 2009; Newby et al., 2009).
Figure 2: Aerial photo and bathymetric map of Wildwood Lake. Contour interval is 3 m; black circle indicates coring location.
Figure 3: Modern-pollen data from sites in southeastern New England: Eagle Pond, Fresh Pond, Round Pond, Deep Pond, Sandy Hill Pond (Parshall et al., 2003), Blaney’s Pond, Duck Pond, and Mary’s Lake (W. Oswald and D. Foster, unpublished).
Figure 4: (A) Lithology of the Wildwood Lake sediment core; gray shading indicates lacustrine sediment; black shading indicates sandy layers. (B) Age-depth model; open circles are 210Pb dates, triangle is European settlement; squares are calibrated 14C dates. (C) Organic content estimated by percent weight loss-on-ignition.
Figure 5: Pollen percentages diagram for selected taxa and charcoal influx values. Separate diagrams are shown for: (A) 1000 cal. a BP to present; (B) ~4500 cal. a BP to present; (C) ~9800-8800 cal. a BP. Dashed charcoal-influx lines in (A) and (B) represent 2X exaggeration. Estimated age for European settlement is 300 cal. a BP.
Figure 6: (A) Lake-level reconstructions for sites in southeastern Massachusetts; water-depth estimates for Crooked Pond (Shuman et al., 2001); triangles are dry events evidenced by sand layers in sediments of New Long Pond (Newby et al., 2009; Shuman et al., 2009). (B) Wildwood Lake organic content, estimated by percent weight loss-on-ignition. (C) Influx of macroscopic charcoal for Wildwood Lake. (D) Influx of Pediastrum algal cell nets in the Wildwood record. (E) Mean squared chord distance (SCD) values for comparison of Wildwood
22

Lake fossil pollen assemblages with modern pollen spectra from sites dominated by Fagus (n=3; W. Oswald and D. Foster, unpublished), Quercus (n=2; Parshall et al., 2003), and Pinus (n=3; Parshall et al., 2003). SCD values <0.1 indicate similar vegetation composition.
23