NORTHEAST REGIONAL CONTEXT
As this overview demonstrates, the Northeast region is unique in Paleoindian studies for its dynamic marine borders—the northwest Atlantic Ocean transgressed overland north of Long Island and then regressed to leave a large emergent seaboard plain, and the inland Champlain/Goldthwait Seas filled the St. Lawrence Lowland and wrapped around to western Vermont. Although sites and hunting grounds have since been inundated, the potential for Paleoindians to take marine game and other resources is unproven but likely. The Younger Dryas climate reversal was very strongly expressed in the Northeast, where the temperate vegetation pulled back from its earlier northern limits, thereby reducing the biotic diversity on which people had no doubt relied farther west. With the possible exception of eastern Beringia, no other inhabited area of North America confronted people with such challenges and opportunities. River drainages were adapting to newly deglaciated landforms; large expanses of glacial lake floors lay as poorly drained swamps over many square miles, and rivers running to the changing sea level cut deeply into shore sediments. What is today New England was semi-isolated from land areas to the west, almost a peninsula pendant from Vermont, with the deep Hudson trough carrying massive drainage. Maine east of the Penobscot was apparently uninhabited, but the Maritime Provinces of Canada saw some use by Paleoindians. Mammoths were apparently extinct in the region before the Younger Dryas, while mastodonts died out just before or soon after people appeared. The only "Big Game" were cervids—elk, caribou, moose, white tail deer. The herds of caribou modeled for Ontario do not appear to have been present southeasterly at this time. North of the marine incursion lay the waning Laurentide Ice Sheet that affected climate in both summer and winter, partly offsetting the increased solar insolation of the time. Strongly seasonal climate, and the rarity of caves, has made organic remains extremely scarce. Consequently, the status of claims for very early sites cannot be evaluated. Research on Paleoindians in the Northeast is today colored by the extreme patchiness of environmental variables including knappable tool stones, by the private landholdings that characterize the area along with its early industrialization and intensive agriculture, and by the near invisibility of Native Americans until the last third of the twentieth century when consultation became possible and necessary. In addition, comparison with other areas has been delayed because of contrastive cultural histories north and south in the region. Especially in Late Paleoindian and Early Archaic times, these differences confounded comparisons.
Information about Paleoindian sites in the northeastern part of the United States has accumulated at an unprecedented rate in the last 15 years, apparently as a benefit of mandated Cultural Resource surveys in the several states. Most of the new discoveries are sites in the small class, significantly increasing knowledge of site distributions, single-component inventories, and density of footprints of phases of Paleoindian use of the landscapes. The region has long been notable for the presence of a few sites larger than the norm in any other region of the continent, all, however, lacking the megafaunal remains that characterize Paleoindian research in the public mind.
For this study, the Northeast encompasses the six New England states, New York, Pennsylvania, New Jersey, and Delaware. Of these ten, only Delaware lacks deposits of the Wisconsinan glacier, but all were affected by glacial outwash downstream, adjustments of biotic ranges, and coastal regression and transgression during the Pleistocene.
Paleoindian research has been lively in these states since the 1950s when very large sites were discovered. Both Shoop in Pennsylvania and Bull Brook in Massachusetts are multi-acre sites with hundreds of fluted points and fragments that recall Clovis in the West, where sites of such size and structure are unknown with Clovis points. The Folsom and Clovis excitement was slow to arrive in the Northeast because northeastern sites are never found with the spectacular megafaunal remains conspicuous in the West. Most Paleoindian sites have been damaged by surficial disturbance, and often by active collecting or digging for private ownership. This latter activity has increased alarmingly in recent years, fueled by rising market prices commanded by exquisite and ancient artifacts.
The information available on northeastern Paleoindian sites is mainly published, since Paleoindian information reliably commands interest. The extant literature is large and varied in quality, ranging from antiquarian notices of finds of single specimens through a few detailed analyses of collections, summary site reports, proposed mobility models, and regional syntheses. Surveys undertaken for Section 106 reviews reveal many such sites, which are rarely broadly excavated. Published notices are likely to follow, with few details given, but the unpublished and somewhat more detailed reports are difficult to access. Few northeastern Paleoindian sites have been nominated to the National Register of Historic Places, probably because they are typically heavily damaged. For instance, the Bull Brook site, the nation's largest, is not listed, having been essentially destroyed by commercial gravel extraction and the collecting fever that accompanied it (Byers 1956). Recent large-scale summaries and interpretations include those by Anderson (1990a), Curran (1999), Custer (1996), Dincauze (1993a,b), Doyle et al. (1985), Ellis and Deller (1997), Ellis et al. (1998), Gramly and Funk (1990), Petersen (1995), and Spiess et al. (1998).
The Northeast enjoys a relatively large radiocarbon data set for Paleoindian sites. While not an unmitigated advantage to scholarship (see Chronological Considerations for the Northeast), it provides at least adequate correlation to the paleogeography and paleoclimatology of the period 11,500 to 13,000 years ago. Fortunately, northeastern archeologists have the benefit of a stylistic seriation developed in Michigan and Ontario that appears accurate as a sequence, supported by geological ages of the site situations north of the Great Lakes (Ellis and Deller 1997). This typology needs to be refined in the United States to accommodate complications not yet encountered in Ontario, which are presented by sites in the Maritime Provinces and northern New England (Table 5). Understanding Paleoindian sites and distributions requires familiarity with the diverse landforms, relief, and glacially deranged river patterns in the region. In addition, marine geomorphology is central to understanding the availability of landforms to Paleoindian occupants of the region. The section, Geography and Environmental Conditions for the Northeast, reviews the dynamics of landform and coastal evolution in the context of extreme climatic change and the changing flora and fauna adapting to the physical environment.
Not every state has a State Plan, so there is no official index or inventory of Paleoindian Property Types for the region. Paleoindian categories in State Plans exist in Pennsylvania (Lantz 1985), New Jersey (Marshall 1982; Grumet 1990), and Massachusetts (published regional summaries). In the Property Types section, a purely heuristic set of distinctions is presented to highlight as much of the essential site diversity as possible at this stage of research, to structure the discussion that follows, and to aid in the evaluation of significance for the sites and site clusters that are the topics of this report. Few detailed interpretative studies have been made of Paleoindian sites to support consideration of such anthropological topics as population densities, site demographics, site-specific and culture-specific activities, or degree of logistic mobility.
The section Resource Distributions in the Northeast builds on the descriptive data in the first four sections to work toward a predictive model for Paleoindian site distributions. The ambiguity in the term "resources" here emerges as a strength, permitting discussion of site locations in terms of the economic resources sought by the Paleoindians themselves. The priorities that correlate with Paleoindian sites in several distinct ecological zones display subregional differences in economic ecology and in population distributions and densities as now known. Much more careful, fine-grained analysis will be needed before we can begin to discuss confidently such topics as colonization strategies, seasonal resource extraction, responses to changes in prey numbers and kinds, or social structures in lightly populated, diverse, dynamic landscapes. So far we have recovered little evidence for more elusive categories of human activities, such as belief systems and ideologies.
In the last century, investigation into initial human colonization of the Northeast has been blighted or enlivened (according to one's taste) by notorious misconceptions, false claims, and outright frauds. These distractions have slowed the increase of knowledge and contributed to a second-tier literature that tenaciously holds onto some of them. I have chosen to acknowledge a selection of these phenomena in Table 6, and to ignore the rest. Uncertainty remains about the integrity and even identification of some sites, specifically the DEDIC/Sugarloaf and Hannamann sites in Massachusetts, which cannot be considered candidates for protection until the mists are cleared.
The Northeast Paleoindian area contrasts least with the Midwest and adjacent Canada, and most strongly with the Southeast. Although Québec has no fluted point sites, Atlantic Canada and southern Ontario were contiguous territories for northeastern makers of fluted points and the lanceolate successors. The Great Lakes lowland and the St. Lawrence seas and lakes were major travel and transport routes into the Northeast; lying as they did just south of the Laurentide Ice Sheet they formed an influential limit on Paleoindian colonization of eastern North America. Canadian sites, geological dating, and tool typologies dominate Northeastern and Midwestern archaeology, but have been little noted or appreciated in the Southeast. The Debert site study was particularly influential. People moving into the uniquely dynamic environment of the northern lakes had to learn quickly how best to use the rapidly changing, often unpredictable ecology. They had to abandon skills and assumptions they relied on to the west and south. Initial recognition of these constraints on Paleoindians (e.g., Dincauze1981c) was based on awareness of ecological patchiness. The dismissal of the constaints set northeastern Late Pleistocene studies into dead end debates for nearly two decades; the Northeast became viewed as marginal, a judgment unearned by Paleoindian behavior. Once overextension of the "Clovis" terminology is abandoned, comparison of northeastern sites and cultural behavior is likely to be increasingly made to the Midwest, as differences in material culture and economies are seen more clearly to the south. It may eventually be possible to deal fully with the apparent delay in occupation of the glacial border areas of the continent. The Southeast takes prominence in comparisons in later Paleoindian times, when lanceolate bifaces were moved northeastward from both the northern Prairies and from Dalton and related cultures, and most definitively in the early Holocene as Early Archaic lifeways extended northward. The resulting Cis-Appalachian Archaic cultural sphere apparently interrupted cultural influences from the west for several millennia.
The northeastern United States offers five distinct but not independent routes to chronologies for Late Glacial and Early Holocene archeological complexes: 1) stratigraphy; 2) Great Lakes geomorphology; 3) radiocarbon assays; 4) stylistic seriation; and 5) the Younger Dryas climatic subchron with its direct ties to ice layer, varves, and U-Th dating elsewhere. Other possible methods such as faunal and pollen zone associations have so far failed for lack of preservation in the regional sites. Late-glacial placement in general is well established on the basis of a few scraps of cold-adapted faunal remains (Spiess et al. 1985), association with late-glacial landscape features and vegetation, and stylistic comparisons on a continental scale.
Radiocarbon ages for northeastern fluted point sites are all younger than Clovis ages in the western Plains and Southwest (Haynes et al. 1984). The difficulties of radiocarbon dating in the northeastern late Pleistocene go well beyond the problems of finding culturally-affiliated organic materials in sites of that age (Bonnichsen and Will 1999; McWeeney 1995). Limitations inherent in the radiocarbon method itself, deriving from its dependence upon variable atmospheric states, have made reliable and accurate ages very elusive in the Younger Dryas period, with its plateaus of indeterminacy at critical time spans (Bartlein et al. 1995; Fiedel 1999; Grafenstein et al. 1999).
Northeastern radiocarbon ages cluster into one group in the early Younger Dryas climatic interval and a second toward the end of the Younger Dryas. The clusters associate sites sharing distinct fluted point styles (Table 5), in the correct order. The Initial Holocene period, in contrast, groups many dissimilar cultural manifestations, some of which are sequential and others coeval.
Data so far support ages for the initial settlement of the Northeast sometime later than 13,500 years ago. There is a long and active history of older claims, not all of which are dismissible. The recent spate of new candidates for ages older than 14,000 years in North and South America assures that the debate remains healthy.
Stratified sequences containing fluted points of different styles are not reported in the Northeast. The longest stratigraphical sequences in the region are from the western and eastern edges of Pennsylvania. The oldest is at the Meadowcroft Rockshelter in the southwest. A second long stratigraphical sequence, at Shawnee-Minisink, has Paleoindian in a basal aeolian deposit, followed by alluvial deposits with Early and later Archaic cultural remains (Dent 1985, 1999). The Arc site in western New York lies on a weathered Late Glacial diamicton sealed under an early Holocene peat (Tankersley et al. 1997). The nearby Hiscock site is a Late Glacial organic deposit with stray isolated fluted points in it. These sites together imply a position for earliest Paleoindian artifacts coeval with the withdrawal of the Wisconsinan ice from the United States. More detail is not now available.
Alluvial sequences in the valleys of rivers draining into the Atlantic typically begin at or subsequent to Paleoindian time (e.g., Crissel 1998; McNett 1985). Early Holocene sites buried in alluvium enclose unfluted lanceolate points with radiocarbon dates implying ages younger than 11,500 years ago (Maymon and Bolian 1992; Sanger et al. 1992). Most extant fluted point sites lie on Late Glacial topographic surfaces, few of which are sealed by later deposits. Sites with point styles IV to VI are situated near the surface, typically in plowzones (e.g., Plenge, Reagen, Potts, Nicholas). The implication is that Late Glacial sediment accumulation on northeastern upland surfaces ended before the close of the period of fluted points, and inland river aggradation intensified subsequent to it.
There are a number of Late-Glacial geomorphological landscape features dated by radiocarbon. These offer some potential to assignrelative dates to Paleoindian sites if they can be shown to be earlier, contemporary with, or later than those features. During the Younger Dryas, more than 13,000 years ago, the Laurentian ice margins stood near the north shore of the maritime Gulf of St. Lawrence. That precluded the possibility of fresh-water ice-margin lakes in Paleoindian time southeast of the Great Lakes. Several sites in central and northern New England are in or near sand dunes. Also, there has been research in Connecticut on the basic chronology of southern New England dune building, related to shifts in late-glacial climates and wind regimes. This work (Thorson and Schile 1995) supports the view that although there may have been some reactivation during the Younger Dryas episode, dune-building in that area had been essentially completed 13,000 years ago, prior to the arrival of humans. Farther north, dune-building has not been securely dated; the living surface at the Hedden site in Maine had charcoal aged 12,300-12,600 years ago underlying what appeared to be naturally deposited dune sands.
This very traditional geological method of dating by landform association is not without its problems when applied to Paleoindian chronology. The differing styles of fluted points in the Great Lakes and Midwest have been reasonably well associated with a series of strandlines of Late Glacial and ice-margin lakes in the eastern Great Lakes area and these geomorphological features have extensive suites of radiometric dates. In turn these Great Lakes dates have been used to approximate ages for the Northeast sites yielding similar fluted point styles but where radiocarbon ages exist for New England sites, the Great Lakes ages are older by substantial amounts. It is possible that these indirectly estimated geomorphological ages for the Great Lakes Paleoindian sites which yield the type specimens of fluted points will be modified by future research. It seems equally likely that those more westerly Great Lakes sites from which the fluted point styles come are indeed older than the sites yielding similar fluted point styles in the northeastern United States.
Many of the radiocarbon ages for northeastern Paleoindian sites have large error ranges or other uncertainties (Bonnichsen and Will 1999; Curran 1996; Fiedel 1999; Levine 1990; Stuiver and Reimer 1993). Published radiocarbon ages, when *13C corrected for fractionation and calibrated to the Stuiver et al. (1998) curves, show older than uncorrected ages. The age and the rates of change implied by uncorrected and uncalibrated years are erroneous and misleadingly short, making time seem to move more quickly than it did in fact (Taylor et al. 1996). Also, computer bugs can affect calibrations under specified conditions (email alert from Belfast lab, 6/26/99): the literature is not clean. All this implies that precision in calendrical time in the Late Quaternary is still elusive. The choice made here is to use ages calibrated by the Struiver et al. 1998 standard, leaving age ranges unspecified in expressions of "years ago."
In the Northeast, there are additional complications: northeastern acidic soils have been churned by freeze-thaw cycling, burrowing animals of many sizes, and tree roots, for all the years since Paleoindian occupation of the area. Consequently, there is a real contamination problem that first surfaced with efforts to date Bull Brook (Byers 1959; see Bonnichsen and Will  for a disheartening catalog).
Great Lakes Style Sequence
A stylistic sequence developed for fluted point sites in the eastern Great Lakes area offers a basis for finer resolution. Ontario archeologists refined a point-style seriation defined in Michigan (Ellis and Deller 1990, 1997; Roosa 1963), providing a technique for discussing relative age among sites with fluted points (Table 5). It does not contradict in any way the radiocarbon ages available, and it is congruent with the geological ages of late-glacial landforms in Ontario. While the ages assigned to Paleoindian landforms in Ontario are presently older than the dated sites in the Northeast, there is no reason to assume a measurable time lag (see above). In order to introduce regional terminology, and to reduce the need to discuss terminological equivalences in the several states, I use six style groups to discuss the relative ages of sites. The groups are not cleanly discrete. There are intermediates between all of them, and more than one may appear in a given site. These two observations support the inference that Paleoindian use of the region was essentially continuous following the initial colonization.
Younger Dryas Chronology
With the Younger Dryas climate reversal securely dated by calibrated radiocarbon supported by ice-core and varve ages, U-Th, oxygen-isotope, and coraline rings chronologies (Ridge et al. 1999), it is clear that fluted point-using Paleoindians date within the Younger Dryas period. No efforts at cross-dating from cultural sites to the absolute ages of Younger Dryas time have yet been successful. The absence of evidence for this dramatic climatic event in the Meadowcroft cultural, biological, and stratigraphic chronology is cause for some concern; it should be recognizable if the sediments are as old as claimed.
Paleoindian and Early Archaic sites in the Holocene
Style groups IV-VI fall into post-glacial time, in a period of rapid warming that must have stressed all living things at middle and high latitudes (Petersen et al. 2000). In the Northeast, this observation leads to terminological and chronological confusion, because the beginning of the Early Archaic Period is around 11,600 years ago, but unfluted "Paleoindian" bifaces and late tool kits are of the same age (Dincauze 1986; Funk 1991).
What is becoming clear is that the Early Archaic period is time-transgressive, being earlier in the south than in the north. In Delaware, southern Pennsylvania, New Jersey, and southeastern New York, and very rarely beyond there, the late Paleoindian/Early Archaic Hardaway-Dalton biface, rarely fluted, marks the transition (A.J. Anderson 1964; Carr 1998; Custer 1986). Northward, notably in Maine and eastern Canada, the unfluted lanceolate Group VI bifaces, assigned to Late Paleoindians, are apparently contemporary with the Early Archaic notched bifaces of the Kirk and Palmer suites, which reach into central New England (Doyle et al. 1985). Their dating is frequently controversial, but they are all dependably Holocene.
The Northeast region of the United States for this study includes the New England states, Delaware, New Jersey, New York, and Pennsylvania. Northeastern Paleoindian studies normally include the Canadian Maritime Provinces, Quebec, and Ontario, whose important early sites and studies necessarily are cited here. This Northeast is bounded easterly by the Atlantic Ocean, northerly by the international boundary, westerly to the limit of the Allegheny Plateau, and southerly by the Mason-Dixon line, beyond the limits of Pleistocene glaciation.
The Northeast is geologically diverse. Precambrian basement (craton) rocks of the Canadian Shield appear in the Adirondacks of New York. Complex folded structures of the mountain belts trending SSW to NNE directed glacial flow and river drainages west of Maine, while the structural trend of the coastal zone runs NW to SE. Paleozoic mid-continental seas deposited the fossiliferous rocks of the Allegheny Plateau and folded Appalachians. Continental collisions and related tectonic movements of Precambrian, Paleozoic, and Mesozoic ages, associated with extensive volcanism and opening and closing early versions of the Atlantic Ocean, built mountains now worn to roots: e.g., Appalachian Ridge-and-Valley uplands, Taconic, White, and Green Mountains. Igneous and sedimentary Paleozoic rocks accreted to eastern New England with fragments of continents colliding from the East. Mesozoic volcanic and detrital deposits lie east of the Appalachian spine from Newfoundland to New Jersey. Unconsolidated Tertiary sediments underlie the Coastal Plain from south of Manhattan and east at least to Martha's Vineyard (Hunt 1974; Skinner and Porter 1995).
The structural trends of eastern North America influenced travel of the earliest humans advancing from either the West or South. People entered from the West through the glaciated or unglaciated Allegheny Plateau, following tributaries of the Ohio River or the shores of ice-margin lakes. East of the Great Lakes rise the Adirondacks and the rugged mountain ranges of New England, with moderate to high relief. Rivers breach the Appalachian peaks in a few crucial places, dropping along the Fall Line to the Coastal Lowland (Hunt 1974). The ancient, rolling, unglaciated surface of the Piedmont begins the lower elevations toward the coast. The route from the South led along the Piedmont and Coastal Plain, or through the Great Valley of the Appalachians. During glacial time, the Coastal Lowland extended eastward to include the exposed inner Continental Shelf, but by the time of the Paleoindians the shoreline north from Massachusetts had moved west and was withdrawing again seaward (Belknap et al. 1987; Oldale 1985).
An unknown number of Quaternary glaciations gouged bedrock and sediments, enlarged valleys, and deposited sand and gravel over all of New England, most of New York, the northwestern and northeastern corners of Pennsylvania, and northern New Jersey. In the latter two states earlier tills extend south of the Wisconsinan limit. Major valleys were choked with outwash. Glacially deranged drainage patterns poleward of the moraines left the area initially with abundant surface water and randomly exposed bedrock. During early stages of glacial melt, large ice-margin lakes filled major river valleys. After 14,000 years ago, an inland sea filled the St. Lawrence-Champlain lowland south of the ice sheet (Dyke and Prest 1987a). During the early Paleoindian period seacoast withdrawal dominated the eastern coastal areas, supporting pedestrian traffic across Long Island Sound and beyond the present limits of Martha's Vineyard and Nantucket islands, delaying the formation of biologically rich estuaries (Oldale 1985, 1986).
Paleoindian sites are associated with beaches of the Late Glacial Champlain Sea in Vermont, New York, and Ontario (Loring 1980), and the succeeding Goldthwait Sea (Dumais 2000). Along the Atlantic shore, where the sea has transgressed inland, sites are rare but there are indications in occasional dredgings that Paleoindians were active on the exposed continental shelf (Glynn 1969). The Hudson and Baltimore canyons slashed the exposed coastal zone, limiting latitudinal movement on those plains. Central Maine was flooded by the late-glacial Atlantic Ocean for some distance inland. Gravel deltas formed into that body of water and subsequently drained became favored locations for Paleoindian sites. Sea level rise and isostatic adjustments through the early Holocene caused river incision upstream and aggradation in the lower reaches, burying early sites deeply in alluvium.
The Ohio River basin drains the Allegheny plateaus to the Mississippi and the Gulf of Mexico; the rest of the Northeast drains into the Gulf of Maine and the Mid-Atlantic bight. Prior to ca. 14,000 years ago, meltwater torrents carried ice and sediment down the Ohio to the Mississippi River and the Gulf of Mexico (Kennett and Shackleton 1975) and down the Susquehanna, Delaware, and Hudson rivers to the Atlantic. Northern rivers participated later. Paleoindians moving east across New York would have confronted a deep Hudson trench even after its iceberg period (Dineen 1996). Early Holocene rivers incised rapidly, leaving Paleoindian sites on high alluvial terraces.
Water levels and shores of the deglaciating eastern Great Lakes fluctuated dramatically; their special names distinguish them from modern lakes. Briefly after 14,000 years ago, Lakes Erie and Ontario were at low water stages (Karrow and Warner 1988) as isostatic rise decanted them to the northeast. In contrast to the situation in Ontario, proglacial lakes in New England and New York drained prior to the appearance of fluted-point users (Curran and Dincauze 1977; Ridge and Larsen 1990). Biologically rich, extensive, swampy, post-lake inland basins attracted Paleoindians and later peoples (Nicholas 1988, 1998).
Although the evidence is soft, people could have entered the Northeast with warming climate prior to 13,000 years ago. The abrupt Younger Dryas climatic reversal, emphasizing the strong continental seasonality of the region, stunted vegetation and chilled the air (Cwynar and Levesque 1995; Grimm and Jacobson 1992; Peteet et al. 1993). With perihelion in summer (Kutzbach 1987), winters continued severe in the higher latitudes. I expect that fluted point makers did not linger in the immediate international boundary area when the climate reversed (Mandryk 1993; Mott et al. 1986; Stea and Mott 1989). Only the earliest fluted point styles (Groups I-II) are common there; Group III is notably absent.
Northeastern Holocene climatic regimes affect the condition and visibility of ancient sites (McWeeney 1994; Thorson and Schile 1995). Summer temperate climates promote biota, which churn sites. Mid-Holocene dryness spawned frequent forest fires that may account for the young radiocarbon samples in so many sites. Winter freeze restricts the activity of decomposers, preserving charcoal better than farther south, while churning sediments and destroying by cryoturbation most features and stratification.
Biota of the Fourteenth Millennium
By 14,000 years ago, the tundra zone south of the ice was reduced to a fringe north of the international boundary, except for high-altitude areas in the mountains (Dincauze 1988; Mayle et al. 1993). South of the tundra, a spruce-park woodland dominated northern New England (Gaudreau 1988), with hardwoods invading from the south during warm periods before and after the Younger Dryas. Hardwoods show thinly on northeastern pollen diagrams before 13,000 years ago, but macrofossils were established well north of the pollen zones (McWeeney 1994). New Jersey, Delaware, and southern Pennsylvania had mappable densities of oak by 14,000 years ago (Jacobson et al. 1987), accompanied by temperate pine species (McWeeney 1994). Mountain heights were retarded in reforestation, in terms of southern species taking residence, but they were not significantly characterized by tundra within the span of human presence.
Soils development paces revegetation. Late Glacial warming speeded that process until the temperature reversal of the Younger Dryas (Anderson et al. 1990; Morgan 1987; Overpeck et al. 1992; Prentice et al. 1991). Postglacial thin and acid soils restricted the potential mix of vegetation in what was necessarily a patchy mosaic of microclimates and adapted vegetation. Until loamy soils developed, newly exposed glacial and coastal sediments were porous, deficient in soil moisture. Evidence for forest fires is consistent with this expectation.
Northeastern Pleistocene fauna included mastodont, giant beaver, and stag elk (Cervalces) until some time after 14,000 years ago; none, however, are reported in Paleoindian sites (Guilday 1982; Kurten and Anderson 1980). Paleoindian artifacts do occur among Pleistocene paleontological deposits at the Hiscock site in New York (Laub et al. 1996). Caribou were in the region from unspecified beginnings until they left Maine in the twentieth century. Virginia deer followed the hardwood species north from glacial refugia and were apparently present with the earliest Paleoindians in Pennsylvania (Guilday and Parmalee 1982: 171) and possibly New York (undated at Hiscock; Steadman et al. 1986). The now well-dated Younger Dryas episode marks the boundary for Pleistocene species in the Northeast (Elias 1999); subsequent faunas are entirely modern. In fact, the Meadowcroft record includes only modern fauna from the beginning (Guilday and Parmalee 1982).
Small temperate forest mammals dominate as raptor prey at Meadowcroft (Guilday and Parmalee 1982); beaver occurred at Bull Brook (Speiss et al. 1985:148). Fish occur in Late Glacial sediments of the Champlain Sea (Cronin 1977), but rarely in alluvium. Shawnee-Minisink exceptionally produced fish of unknown species (McNett 1985). Anadromous species recolonized from the sea; freshwater fish returned as spat or adults dropped by birds and other predators. There is no reason to suspect tardiness in their postglacial northward spread, beyond delays required for drainages to become biologically active.
The North Atlantic avian flyway, drastically shortened by glaciation, extended north as soon as vegetation and other food species permitted; recolonization by beetles of newly exposed land near the Great Lakes (Morgan 1987) shows how quickly that took place. Condors at the Hiscock site in New York may have been neighbors of Paleoindians (Steadman et al. 1986). In Champlain Sea sediments of Quebec, an eider duck has been dated to the Younger Dryas time span (Harington and Occhietti 1980).