The Quaternary of Virginia - a Symposium Volume

Home , Bootherium, Environment of Virginia, Hemphillian, Irvingtonian, Mylohyus, Natural Trap Cave

VIRGINIA DIVISION OF MINERAL RESOURCES PUBLICATION 75 THE QUATERNARY OF VIRGINIA - A SYMPOSIUM VOLUME

Edited by J.N. McDonald and S.O. Bird

COMMONWEALTH OF VIRGINIA

DEPARTMENT OF MINES, MINERALS AND ENERGY DIVISION OF MINERAL RESOURCES Robert C. Milici, Commissioner of Mineral Resources and State Geologist

CHARLOTTESVI LLE, VI RGIN IA 1 986 VIRGINIA DIVISION OF MINERAL RESOURCES PUBLICATION 75 THE QUATERNARY OF VIRGINIA - A SYMPOSIUM VOLUME

Edited by J.N. McDonald and S.O. Bird

COMMONWEALTH OF VIRGINIA

DEPARTMENT OF MINES, MINERALS AND ENERGY DIVISION OF MINERAL RESOURCES Robert C. Milici, Commissioner of Mineral Resources and State Geologist

CHARLOTTESVILLE, VIRGINIA 1 986 FRONT COVER: Quaternary travertine-marl deposits at Beaverdam Falls, Alleghany County. The falls are approximately 25 f.eet high (Photograph by David A. Hubbard, Jr.). VIRGINIA DIVISION OF MINERAL RESOURCES PUBLICATION 75

THE QUATERNARY OF VIRGINIA _ A SYMPOSIUM VOLUME

Edited by J.N. McDonald and S.O. Bird

COMMONWEALTH OF VIRGINIA

DEPARTMENT OF MINES, MINERALS AND ENERGY DIVISION OF MINERAL RESOURCES Robert C. Milici, Commissioner of Mineral Resources and State Geologist

CHARLOTTESVI LLE, VI RGINIA 1 986 DEPARTMENT OF MINES, MINERALS AND ENERGY

RICHMOND. VIRGINIA

O. GENE DISHNER. Director

Commonwealth of Virginia, copyright 1986 FOREWARD River and some of its tributaries on selected headwater parts of the drainage basin by Kite, The Virginia Division of Mineral Resources Bell and Armstrong, and on travertine and marl celebrated the 150th anniversary of its founding deposits known from the state by Hubbard, Gi- in 1984 and 1985. One of several events sponsored annini and Lorah. Holman and Fay, in separate by the Division in commemoration of this anni- papers, present evidence that Pleistocene and versary was a Symposium on the Quaternary of modern assemblages of reptiles and amphibians Virginia, held in Charlottesville on September (herptiles) from middle Appalachian localities 26-28, 1984. This Symposium was designed to are harmonious, unlike those of mammals; herp- meet several important objeetives: to summarize tiles appear to have tolerated the extreme the state of knowledge on the Quaternary of ehanges in climate more easil5r than mammals. Virginia and its environs; to call attention to Changes in the composition of plant communities research underway on the Quaternary of Virginia in the Appalachians over the last 23,000 years and to identify problems, priorities, and direc- are described by Delcourt and Delcourt. Rem- tions for future work; to assemble scientists, nants of several Arctic and boreal plant species students and layman with interest and expertise which evidently became established in Virginia in the Quaternary of Virginia and its environs during glacial stages exist today in isolated areas in order to promote and facilitate the exchange near the summits of some of the State's higher of ideas and information on the subject; and to mountains as shown by Woodward and Ruska. promote new interest in the study of the Qua- Two recent studies of Pleistocene vertebrate fau- ternary of the region. The program provided a nas are included: one from a fissure-fill in Penn- comprehensive overview of the status of knowl- sylvania (Fonda and Czebieniak), the other from edge of the Quaternary of Virginia. Knowledge a fluvial (?) deposit in northern Virginia (Ott and about the megafauna of the region was empha- Weems). Eshelman and Grady give detailed des- sized in commemoration of Thomas Jefferson's criptions of Quaternary vertebrate localities in interest in this subject area. Virginia, ineluding lists of bird and mammal Eight of the thirteen invited papers and ten species found at these sites. McCary summarizes of the eleven poster papers presented at the some of his lifelong work on Paleo-Indians in Symposium are included in this volume. The Virginia by describing the distribution of Clovis keynote address presented by Silvio Bedini en- projectile points from around the State and notes titled "Thomas Jefferson and American Paleon- that nowhere in the State have ancient human tology" was published in 1985. The meeting closed activities been found in association with Pleis- with a field trip to Saltville led by Jerry McDo- tocene animal remains. Papers by White and nald (1984). Barfield summarize aspeets of recent archeolog- Contained in this volume are papers dealing ical investigations at one of the Saltville sites with many of the major topical areas of Quater- (aaSm5f). Rose uses archeologieal data and his- nary studies for Virginia and environs. A general torical accounts by early European observers to summary of geomorphic processes and an eval- document the occurrence and distribution of uation of evidence for Pleistocene overprints on large mammals in Virginia during the late pre- them are given by Conners, who finds a great historic and early historic period. McDonald need for detailed work on Pleistoeene geology in gives a historical sketeh of the development of Virginia. Three examples of geomorphic studies Quaternary vertebrate paleontology and zooar- ineluded in the present volume are those sum- cheology within Virginia, lists modern work un- marizing differences measured in fans on either derway on these topics, reviews some of the major side of the Blue Ridge by Kochel and Simmons features of ongoing field work at Saltville, and and on alluvial deposits lying above the Potomac recommends more emphasis on vertebrate paleontology and zooarcheology at Virginia univer- References sities and museums. A novel idea of Pleistocene extinctions, dwarfism, and deformity resulting Bedini, S., 1985, ThomasJefferson and American from inferred climate-induced decreases in blood paleontology: Virginia Division of Mineral supply to the mammalian uterus is presented by Resources Publieation 61,26 P. Mclean. Boaz recounts some of the early ideas Hoffer, F. B., 19??, Bibliography of Virginia on evolution of J. L. Cabell. geology and mineral resources-1960-1969: The editors here make readers aware of two Virginia Division of Mineral Resources Pub- important bibliographic works on the geology of lication 1, 68 p. Virginia, including those dealing with Quater- McDonald, J. N., 1984, The Saltville, Virginia nary geology, that mightotherwise be overlooked. locality: A summary of research and field trip One is the matchless product of a supreme effort guide: Charlottesville, Virginia Division of by J. K. Roberts (1942) and the other is a fine Mineral Resources 45 P. work eovering the decade from 1960 to 1969 by Roberts, J. K., 1942, Annotated geologieal bib- Boyd Hoffer (19??). We have compiled a new liographyof Virginia: Riehmond, Dietz Press, bibliography on the Quaternary of Virginia for 726p. those years to the present not included in the earlier cited works. This reference work was obtained from the GeoRef file of the Dialog Sys- tem in California. It is on open-file at the Virginia Division of Mineral Resourees and can be obtained by writing the Division Librarian.

J. N. McDonald S. O. Bird, editors THE QUATERNARY OF VIRGINIA CONTENTS Invited Papers Page

Quaternary geomorphic proeesses in Virginia, by John A. Conners 1

Late Quaternary vegetational change in the central Atlantie states, by Hazel R. Delcourt and Paul A. Delcourt Zg

The known herpetofauna of the late Quanternary of Virginia poses

Quaternary vertebrate localities of Virginia and their avian and mammalian fauna, by Ralph Eshelman and Frederick Grady 4g Early man in Virginia, by Ben C. McCary 7l

Late prehistoric and protohistoric large mammal zoogeography of on the status of Quaternary vertebrate paleontologT and zooarcheology in Virginia, by Jerry N. McDonald ...... gg

Embryogenesis dysfunction in the Pleistocene/Holoeene transition mammalian extinction, dwarfing, and skeletal abnormality, by Dewey M. Mclean ...... rOb

Abstracts of Poster Papers

Preliminary observations on floodplain, terrace, and fan deposits in the upper Potomac River basin of virginia and west virginia, by J. steven Kite,

Quaternary alluvial fans in central Virginia, by R. Craig Kochel and David W. Simmons ...... 123

Quaternary travertine - marl deposits of virginia, by David A. Hubbard, Jr., William F. Giannini, and Michelle M. Lorah ...... 126 wisconsinan herpetofaunas of the central Appalachians, by Leslie p. Fay ...... 126

Late Pleistocene fauna of the Hollidaysburg fissure, by shirley s. Fonda

The Hot Run site: A new Pleistocene vertebrate locality in northern Virginia, by James R. Ott and Robert E. Weems ...... l}g

A Pleistocene legacy: Artic and boreal elements in the eontemporary biota of Virginia, by Susan L. Woodward and William F. Ruska, Jr...... :...... 181

Projectile points from a multieomponent site (44Sm51) in the Saltville Valley,

Tools anddebitage: Cultural debris from a multicomponent site (44Sm51) in the Saltville Valley, Virginia, by Eugene Barfield ...... 1BB

Virginia's first evolutionist: James Lawrence Cabell, by N.T. Boaz ...... l8b QUATERNARY GEOMORPHIC PROCESSES IN VIRGINIA

John A. Connersr

CONTENTS

?age Introduction ...... ,..,

Geomorphic processes and climatic change: problems and overview ...... 6

Processes of eavern and karst

Suggestions for the Quaternary

ILLUSTRATIONS Page Figure

l: H"?;1,H1'#:"*;il'""1t#ffi'Hf,,;;;re;;, N",tn c"";ii;; : 3 3. view of northern slopes of Rich Mountain, showing foot-slope deposits g 4. Partof a seree slope near Sherando Lake, Augusta County . .... l0 5. Large dislocated blocks of resistant metabasalt along the Greenstone Trail . . . 10 6. Large isolated boulder along Skyline Drive ...... 11 7. Terrace-like features along east side of Big Walker Creek valley ...... 1g 8. Meadow Creek Falls near New Castle .. . 1b

TABLES

INTRODUCTION ternary of the United States, for example, devotes less than two of its 900 pages to "Quaternary With few exceptions (notably the pioneering processes in the Atlantic Coastal Plain and Ap- work of John T. Hack), relatively little geomor- palachian Highlands" (Judson, 1965). The geo- phic research has been done in Virginia in recent morphic literature on the inland areas of Vir- decades and only a small part of that has focused ginia, the focus of this paper, is especially sparse. directly on the influences of Quaternary climatic As will be noted. a handful of recent studies have fluctuations on denudational and depositional advanced our knowledge of some aspects of Qua- processes. The indispensable volume, The Qua- ternary geomorphology in Virginia. In addition, research is under way focusing on various alluvial and colluvial deposits, periglacial phenomena, surface accumulations of calcium carbonate (marl, tufa, travertine), and a variety of paleon- 11901 Downey Street, Radford, Virginia 24141 tological and archeological sites. VIRGINIA DIVISION OF MINERAL RESOURCES

The purpose of this paper is to summarize the present day. A major meehanism in estab- current knowledge of geomorphic events and lishing the southeasterly streamflow has been the processes whieh appear to be associated with capture of higher-elevation, northwest-flowing Quaternary climatic changes in Virginia. Unfor- streams by the steeper gradient streams flowing tunately, little is known about the specific denu- directly southeast to the Atlantic. The New River dational processes influencing landforms and as- will fall victim to an impending piracy by trib- sociated deposits, and our understanding of the utaries of the Roanoke River in the not-too-distant complex responses of those processes to the many future, geologically speaking (Figure 1). During climatic fluctuations of the recent geological past all this time, widening of the Atlantic Ocean, is even more deeply shrouded in mystery. Con- uplift of the Appalachian highlands, and the sequently, recognizing and accurately interpret- denudation and sculpturing of the Virginia land- ing the evidence can be very difficult. As a result, scape have been progressing more-or-less contin- much of what follows will of necessity be spec- uously, albeit at varying rates. ulative in nature. (5) Beginning roughly 55 million years ago, a long cooling trend called the "Cenozoic Decline" HISTORICAL PERSPECTIVE began. This trend culminated in the series of dramatic climatic fluctuations which have char- To help place our topic in geological perspee- aeterized Earth history during the past 2 million tive, a brief summary of the evolution of the years or so. Table 1 provides some details of the Virginia landscape is offered. For more details, current "Ice Age". In Virginia, the focus of Qua- the reader may consult Judson (1975) and Hack ternary studies has been on the most recent 50,000 (1e82). years or so, especially the Late Wisconsinan gla- (1) During the late Precambrian and early cial stade and the ensuing Holocene interglacial. Paleozoic eras, low-gradient streams flowed southeast from the Canadian Shield to the ancestral Atlantic Ocean. (2) During the mid-to-late Ordovician (460-435 million yrs B.P.), this proto-Atlantic Ocean apparently began closing as the ancestral North American continental plate approached the Eu- ropean and African plates, initiating a monumen- tal series of orogenies which were to continue for the next 220 million years. Because this mountain-building activity was centered to the east of today's Appalachian highlands, streamflow across Virginia was eventually reversed towards the west and northwest. The New River's north- westerly flow is apparently an inheritance of these ancient upheavals. (3) During the Triassic period, a massive thermal convection cell arose within the mantle producing tensional forces in the lithosphere and the gradual rifting apart of North America from Europe and Africa. During the uparching, o-ro rn ..,r.E"Oi"':'';,t' stretching and faulting associated with early I ..'.;'';.i.i;'*Hil'';';l;'""'."--- -r"J":1::':ti':!r1':irr:'i'--""'r rifting, crustal thinning occurred. As the new .....,' 4"" Atlantic Ocean widened and the eastern edge of - ,;:.-' errerrrc ocrrr North America withdrew from the center of .: f spreading (the Mid-Atlantic Ridge), the eastern margin of the continent subsided and drainage across Virginia began reversing towards the Figure 1. Map showing approximate relation of southeast once again Virginia to the Laurentide iee sheet, sea level ( ) The southeasterly drainage reversals prob- (roughly 395 ft. (120 m) below present mean sea ably began some 170 millionyears ago in northern level), and the northern edge of the Polar Frontal Virginia (Potomac River) and have continued to Zone during Pleistocene glacial maxima. PUBLICATION 75

Table l. Outline of Cenozoic glacial chronology.

Years B.P. Euents

70Mu-65M Earliest evidence of possible glaciation. Preliminary studies of marine microfossils in Transantarctic Mountains believed to have been carried upslope by glaciers, show glaciation cycles at roughly 70-65M, 50-40M, 30-20M, and 7-3M yr B.P.b

55M Long cooling trend, the "Cenozoic Decline", begins. Apparently related to drift of continents into polar regions. Baffin Bay begins to open allowing eold waters of the Arctic Ocean to enter the Atlantic.

38M Rapid worldwide temperature decline and formation of near-freezing Antarctic bottom waters.

28M-27M Good evidence of small, widespread glaciers in Antarctica.

14M-10M Antarctic Ice Sheet buildup following opening of Drake Passage and full development of the Antaretic South Ring stream.

10M-8M Antarctic Ice Sheet expands to about half present volume and becomes permanent. Evidence for earliest Cenozoic glaciation in North America.

3.5M Late Pliocene uplift, including closing of Isthmus of Panama which strengthened Gulf Stream, helped trigger onset of "ice age" climates and circulations. Modern oceanic circulation established by 3.2M yr B.P. Glaciers in Alaska (3.6M yr B.P.) and California (3M yr B.P.)

3.0-2.5M Pronounced increase in fluctuations between warm and cold climatic episodes. Northward expansion of Antarctic polar front and growth of northern ice sheets.c

2.4M Ice sheets about 213 aslarge as Wisconsin sheets maghaveexisted in North America.

1.8M-1.6M Beginning of Pleistocene Epoch, based on cold-water, marine fossils.

1.6M-125,000 Because of much conflicting data, no detailed ehronology is attempted here. Canada and the northern United States experienced multiple glaciations interspersed with warm interglacial episodes. Good evidence for at least 17 major fluctuations exists in Europe for the last 1.7M yr, based on loess deposits and paleosols. Deep-sea cores indicate at least 21 cyeles of glaeiation in the last 2.3M yr, based mainly on oxygen- isotope data.

125,000-100,000 Last major interglacial (Sangamon).

85,000 Earliest definitive evidence of Wisconsinan glaciers in North America.

75,000 Start of rapid Wisconsinan ice growth in North Ameriea.

70,000-45,000 Mid-Wisconsinan or "Port Talbot" interstade: North American ice sheets retreated to core areas.

25,000 Rapid readvance begins.

23,000 Marked cooling and pond formation at several sites in eastern U. S. between latitudes 34oand 37oN. 4 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 1. (cont.)

Years B.P. Eaents 20,000-19,000 Maximum extent of Wisconsinan ice sheets: 5.8M mi2 (15M kmz) of land in North America glacier-covered; sea level 395 feet (120 m) lower-Bering Shelf was 620 mi. (1000 km)-wide land mass connecting North America and Asia: firnline averaged about 3280 feet (1000 m) lower than at present in mid-latitudes and temperatures were roughly 3'C colder near oceans and 17"C eolder in some mid-continent areas.

18,000-6000 Iee sheets generally in retreat but with many fluctuations, including major readvances of Laurentide sheet at 13.5, 11.8, and 8.2 thousand yr B.P.

16,500 Climatic amelioration beginning in Tennessee.

15,000-14,000 Undeniable evidence of human oceupation of North America.

14,000 Rapid rise in sea level begins and continues until about 10,000 yr B.P. Firnline rising: Yellowstone Plateau, San Juan Mountains, Mauna Kea, other areas deglaciated. Pluvial lakes of western U. S. receding.

13,500-8000 Glacial Lake Agassiz in existence.

13,000 24,000 years of drought ending in Florida.

12,500-12,000 Cordilleran Glacial Complex retreats to north of U. S.-Canadian border. Last catastrophic megaflood of Lake Missoula. Last permafrost in Pennsylvania. Warm climate with summer drought on Gulf Coastal Plain.

11,000 Accelerated ablation of ice sheets. Prairie replacing spruce forests in arid parts of Great Plains.

11,000-10,000 Extinction of many species of large vertebrates.

10,400 Lake Algonquin (Georgian Bay-Lake Superior area) began draining northeast into Champlain Sea and Atlantic. Prior to this, most meltwater drained south to Gulf of Mexico.

10,000 "Official" end of Pleistocene and start of Holocene. No break in ongoing climatic adjustments. Deglaciation eontinues. Vast ice sheet still exists over eastern Canada and sea level is some 66 feet (20 m) below present level. Radiation curves for tilt and precession reinforce each other causing average 7% greater solar radiation than at present and encouraging inereased rainfall, slightly warmer temperatures in northern land areas-possibly influenced also by early Holocene expansions of pluvial lakes.

7500 Last readvance of Laurentide iee sheet.

7500-5500 Altithermal interval of exr-eptionally warm, dry climate in many areas: Lake Bon- neville was reduced to a playa; 7000 years ago the prairie/forest border in U. S. midwest was 62 mi. (100 km) east of present location.

6500 Last remnant of Laurentide ice sheet melts on Quebec-Labrador Plateau.

6000 Change from zonal to meridional weather patterns bringing more rain to eastern and central U. S.: slightly cooler; more bogs; pines replace deciduous trees in many hreas; flooding more frequent. PUBLICATION 75

Table 1. (cont.)

Years B.P. Euents

Neoglaciation: numerous readvances of alpine glaciers with 2 or 3 major advances, each with substages, represented in many mountain ranges. A prelude to the next full glacial? 100-0 Meltbacks of many small alpine galciers; sea level rises about 0.5 ft. By the mid- 1980's, evidence of small increases in mean global temperature and of significant increases in CO, and other "greenhouse" gases in the atmosphere due to human activities leads to growing concern over human impacts on climate. (Principal sources: Bowen, 1978; Eyles, 1983; Wright, 1983) a M = million b Science News,1983, vol. 123, p. 6 " Science News,1986, vol. L29,p.374

Very little evidence has been uneovered in Vir- the Susquehanna River would be carrying glacial ginia which can be related confidently to earlier meltwaters and sediments to the Atlantic Ocean Quaternary events. whose shoreline lay some 60 miles (97 km) eastward of its present-day position. THE QUATERNARY Strong katabatic winds, composed of air that PALEOENVIRONMENT was frigid, dry, and heavy, would slide down over the vast Laurentide ice sheet and sweep south As an aid in visualizing the environment during and southeast across Virginia towards the Polar a full glacial, imagine driving northeast from Frontal Zone which lay some 180 miles (290 km) Winchester in northern Virginia along Inter- south of the Virginia-North Carolina border state-S1 and, in less than four hours driving time near latitude 34oN (Figure 1). Along the Polar (about 180 miles), encountering a great mass of Frontal Zone, these cold and dry air masses met ice blocking the highway. Imagine traveling due with warm, moist air masses from the Gulf of west from Winchester along Route 50 and encoun- Mexico to form a steep north-south temperature tering that same wall of ice after 7 or 8 hours gradient. As long as an ice sheet persisted in driving time (350 highway miles). Such would northeastern North America a fairly strong east- have been the situation 20,000 years ago. Climates west temperature gradient also existed, and at- as warm as the present Holocene interglacial mospheric circulation across the United States probably are characteristic of only about 10 per- tended to be strongly zonal (from west to east) cent of a typical glacial-interglacial cycle. The with little seasonal variation in basic circulation duration of such a cycle, over the last 500,000 patterns (Knox, 1983). Because of the conver- years at least, has been about 100,000 years, and gence of westerlies and tropieal air masses, pre- appears to be closely related to the eccentricity cipitation was abundant along the broad Polar of the Earth's orbit about the Sun (Imbrie and Frontal Zone itself and warm-temperate forests Imbrie, 1979). similar to those of present-day Virginia existed The great ice mass noted above was part of the there during glacials (Delcourt and Delcourt, Laurentide ice sheet. It rose steeply from its 1984). In Florida, south of this zone, dry, windy margin to a maximum elevation roughly twice conditions and scrub brush prevailed (Watts, that of Virginia's highest peak. It covered 4 19S3). North of the Polar Front, boreal forest with million square miles (10 million kmz) of land- abundant jack pine and spruce dominated the a high, almost unbroken, glacial desert stretching eastern United States. North of this forest, a northward to well beyond the Arctic Circle. Be- fringe of tundra over 93 miles (150 km) wide in cause so much water was locked up in glaciers, places (Watts, 1979) bordered much of the ice mean sea level was about 395 feet (120 m) lower sheet itself. Patches of treeless tundra also ex- than at present (Bloom, 1983a). Thus a journey tended southward along the high Appalachian eastward from Winchester 20,000 years ago Mountains to the Smoky Mountains (Wright, would have found a broad valley where Chesa- 1981). Although minor pockets of permafrost peake Bay now exists. In the center of the valley, probably existed at higher elevations south of VIRGINIA DIVISION OF MINERAL RESOURCES

Pennsylvania, it is very unlikely that sufficient sites; (c) improving knowledge of the paleoenvir- precipitation occurred to allow formation of small onment; (d) more accurate correlation and dating glaciers such as those which Berkland and Ray- of Quaternary events and materials; and (e) in- mond (1973) suggested were present on Grand- telligent land use (e.g.; locating safe waste-dis- father Mountain, North Carolina. Nival processes posal sites, stable building sites, good construc- associated with snowbanks may well have been tion materials and dependable water supplies- operative, but proof of even this is lacking. The see Eyles, 1983X lower limits of tundra in Virginia are not known Three major difficulties are encountered when but evidence of Late Wisconsinan tundra at2620 attempting to unravel paleo-climatic influences ft. (800 m) elevation in Maryland (Maxwell and on geomorphic processes and landforms. Davis, 1972) and at 3280 ft. (1000 m) in West (1) Singularity of landforms: Like fingerprints, Virginia (Watts, 1979) suggests it may have ex- no two landforms are completely identical. It may tended to similar levels at favorable locations in be possible to identify certain threshold condi- Virginia. tions at which an active feature (e.g., slope or Detailed temperature and precipitation values stream channel) changes from one equilibrium for glacial-age Virginia are still conjectural. Cur- state to another. However, the uniqueness or rent mean annual temperatures range from near "singularity" (Begin and Schumm, 1984) of land- 16oC in southeast counties to near 9oC in moun- forms means that critical parameters quantified tainous counties. Ice-age temperatures were at one location cannot be applied with certainty probably on the order of 10o-15o colder overall to another locality, even though the environmen- with very cold winters accounting for most of the tal settings may appear to be nearly identical. difference. Precipitation may have been some 20 (2) Meager understanding of contemporary to 50 percent less than today during full glacial geomorphic processes: Recognizing the processes conditions with a significantly higher percentage shaping today's landscapes is far more difficult of that falling as snow than does today (Deleourt, than it may at first appear. To use one common 1979; Barry, 1983). example, it may be apparent that a rock has been broken down into smaller fragments which now GEOMORPHIC PROCESSES AND CLIMATIC lie scattered along a slope. Problems ensue when CHANGE we attempt to ascertain exactly what processes caused the breakdown, quantify the contribution We will now pose some questions of major of each process involved, and determine when and importance to Quaternary geomorphologists, dis- for how long those processes were operative. cuss the difficulties, both general and specifie, Often we are reduced to vague references about which must be confronted when attempting to "weathering and erosional processes" as a result answer those questions, and then review the re- of our inability to delineate the impacts of sueh lations between specific processes and Quater- activities as hydration, hydrolysis, oxidation-re- nary ehanges of climate in Virginia. duction, frost action, simple solution, overland Among the major questions of concern are flow influences, or organic activity. these: (1) What were the nature and magnitude (3)Extrapolation to past conditions: The diffi- of the variations in paleoclimate? (2) When did eulties of attempting to apply our limited knowl- climatic changes oecur in specific loeations and edge of complex interacting processes to the what were the durations of the climatic regimes? uncertain conditions of the past should be man- (3) How do climatic fluctuations influence the ifest. The task is further complicated by the often various types of processes acting upon the land- profound impacts man has had on vegetative scape-what is their relative importance and cover (deforestation, farming, grazing), biogeo- what are the rates at which they progress? (4) chemieal cycles (acid rain, fertilizers, pesticides), What are the resulting impacts upon the land- streamflow (dams, diversions, accelerated runoff, scape, how much evidence of these impacts re- water table lowering), and the like. For example, mains, and how do we recognize that evidence the present mechanieal sediment load of streams and place it in a meaningful time-frame? (5) How in the eastern United States is probably 4 to 5 do we apply our knowledge of past and present times what it was prior to extensive human processes and their known or suspected results modification of the land (Meade, 1969). to such practical problems as (a) the geological In spite of such difficulties, Pleistocene stream interpretation of known Quaternary fossil and terraees, buried channels, talus accumulations, archeological sites; (b) the loeation of new and lake deposits, distinctive unconformities in sed- potentially rewarding fossil and archeological iments and other forms of evidence have been PUBLICATION ?5

correlated with specific paleoclimatic intervals Few attempts have been made to quantify pos- in many parts of the world, including numerous sible differences in denudation rates in the Ap- locations in the western, central and northeastern palachians between Quaternary and Tertiary Pe- United States. In Virginia, however, the imprints riods, or between glacial and interglaeial of climatic change upon the landscape appear to episodes. Denny (1956) estimated that plateau have been much more subtle, and most correla- tops in north-central Pennsylvania have been tions to past environments are tentative or un- lowered as much as 197 feet (60 m) during the proven. Aside from some late Quaternary sequen- Pleistocene (about 38 mm/1,000 yr). Hack (1965) ces in the coastal regions, very few deposits or estimated a long-term lowering of 226 feet (69 landforms of any kind in Virginia reveal a clear m)/million yr (69 mm/1000 yr) for Virginia's record of alternating Quaternary environments. Shenandoah Valley. Kane (1975) concluded that This may seem somewhat surprising in view of the current denudation rate of the 613 miz (1595- the numerous climatic fluctuations the area must kmz) Rappahannock River Basin of Virginia had have experienced during the last 1.5 million years decreased from 0.95 to 0.87 inches (24 to 22 mm)l or so. 1000 yr over a recent 20 yr period. The lowering A fundamental reason for this laek of clear rate in small basins is generally much higher than paleoclimatic evidence of a geomorphic nature in large ones-Kane found a 53.5 inches (1358 appears to be that the climatic changes influenc- mm)/1000 yr denudation rate for a small basin ing the Virginia terrain generally did not cross of 0.08 *iz (0.2 km2) near Reston, Virginia. De- major geomorphic thresholds as they did in cer- nudation estimates can be used to help explain tain other environments. An important factor in why deposits of alluvium or colluvium revealing the lack of threshold crossover was the buffering a long history of Quaternary sedimentation are effect of the forests, which maintained their hold rare in the Appalachian region (e.g., Mills, 1981, on most of the terrain during glacials as well as p.2I8-2r9). interglacials. Even though the forest type Although such denudational rates imply a cer- changed markedly, the influences on slope pro- tain gross uniformity of overall effect during the cesses and sediment yield over most of the region Quaternary, changes in the rates of certain types may not have changed sufficiently to trigger any of weathering, mass wasting, erosion and depo- major wisespread adjustments in basic erosional sition have undoubtedly taken place during the and depositional processes. As a consequence of Quaternary. Evidence for, and some speculation this lack of drastic change in surficial processes concerning, such changes will now be reviewed and the slow ongoing uplift of the Appalachian for the major surficial processes. highlands (Hack, 1982), the terrain as a whole continued to experienee denudation during both WEATHERING AND MASS WASTING glacials and interglacials. The eonstant wearing away of landforms has left behind little concrete In a colder climate, such as existed in Virginia evidence of climatic changes for the twentieth during glacials, chemical weathering should be century geomorphologist, and evidence of mid- inhibited because lower temperatures decrease to-early Pleistocene environments is virtually un- the rates of most chemical reactions. Mechanical known. weathering and mass wasting rates should in- A number of estimates of denudational rates crease during glaeials, mainly through the have been made. Young's (1969) survey of de- greater efficacy of frost action. The importance nudation rate estimates led him to conclude that of frost action at cold temperatures appears un- most areas of steep relief are being lowered from deniable, especially where permafrost exists 3.9 to 39 inches (100 to 1000 mm)/1000 yr. Hack (Washburn, 1980). Where permafrost is not pres- (1980) compiled data for the Appalachian region ent, and even where mean annual temperatures and coneluded that the long-term rate of denu- are above freezing, frost action may still be an dation was about 16 inches (40 mm)/1000 yr. important, even dominating, process, but it be- Houser (1980, p. 101-102) suggests that solutional comes more difficult to identify positively its processes on the carbonate rocks of the region effects and isolate them from similar effects of are responsible for about 5.4 inches (13.5 mm) non-cryogenic processes. of that 16 inch (40 mm)/f000 yr total, the rest Doubts do exist regarding the importance of being attributable to mechanical processes. It freeze-thaw action, however. White (1976) advo- should be noted that most of this estimated de- cates hydration shattering as the major process nudation occurred prior to the Pleistocene Epoeh. causing the splitting apart and fractionation of VIRGINIA DIVISION OF MINERAL RESOURCES

bedrock. Experiments and observations have yet weathering and mass wasting did progress at a to document that ice is capable of cracking solid more rapid rate during glacials is strongly sug- rock. Even in a cold environment, it is highly gested by the observation that periglacial-type probable that such unjustifiably neglected weath- features and associated deposits increase in fre- ering processes as hydration, hydrolysis, and quency from south to north alongthe Appalachian especially exudation (interstitial salt crystalliza- highlands and by the sheer abundance of collu- tion) play an important role in preparing the way vium in certain locations (Mills, 1981). for, and reinforcing, frost action. Much of the geomorphic evidence that has been The nature of the deposits and landforms pro- used to document the former existence of peri- duced will vary with such factors as bedrock glacial conditions consists of weathering and lithology and structure, the degree of seasonal mass wasting features such as stone stripes, stone temperature change, the amount and distribution polygons, block fields, and colluvial deposits of precipitation, and latitude. An understanding which show distinct vertical zonations that can of the Pleistocene periglacial environment of Vir- be related to past climatic intervals. All these ginia is complicated by the lack of present-day features are present in Virginia and adjacent analogues-i.e., a mid-latitude periglacial cli- areas but colluvium with informative vertical mate in an area of similar elevation, topography zonation is rare-most exposures of residuum and and geology. For example, the angle of the sun colluvium are notoriously unilluminating (Fig- is much higher in Virginia than in modern tundra ure 2). Many of the features once widely assumed areas such as those of Canada and Siberia. More to be of relic periglacial origin (Smith, 1949) have direct rays of sunlight combined with longer been recently shown to be quite active under the winter days and longer summer nights may ha-re present interglacial climate (Hack, 1965; Houser, had a significant impact upon freeze-thaw action 1980). A case in point are diamictons. by increasing the number of freeze-thaw cycles Diamictons, or diamicts, are nonsorted terri- during the year. Aiding this would be the cold genous deposits containing a wide range of par- dry air which would encourage many more clear ticle sizes (Flint, 1971, p. 154). The Appalachian sunny days during winter, encouraging melting, region contains extensive diamictons ranging in and clear cold nights during summer, encouraging freezing. The rate of diurnal temperature change may also have increased considerably during glacials in Virginia. Rapid freezing causes more efficient near-surface freezing, forming a seal which can significantly increase the forces exerted by the underlying water as it solidifies and expands. In the colder yet drier environment that prevailed during most of the Pleistocene in Virginia, snowbanks may have been quite thin in many places, reducing their insulating capacities and enhanc- ing nival processes (Embleton, 1979). In the "Ice Age" springtime, the combination of frequent freeze-thaw cycles and abundant water from thawing snow and ground-ice would facilitate the production of detritus and enhance downslope creep and flow of regolith. These ac- Figure 2. Regolith in borrow pit in Rockingham tivities would tend to produce a thicker regolith, County, North Carolina along truck route 87 near thus reducing slope gradients, increasing stream Eden (Leaksville) in the Piedmont Province, load and inhibiting stream incision. During in- about 3.? miles (6 km) south of the Virginia terglacials, there would be a reduction in the rate border. Cut shows saprolite (decomposed meta- of mass wasting and regolith production, but not morphic bedrock) overlain by a generally fine- a cessation of it. Increased precipitation would grained, poorly sorted material which seems to yield more runoff, which would winnow fines be mostly colluvium. Upper zone has a very thick from slope deposits, increase streamflows, en- deep-red soil profile suggesting all this material courage stream incision and lead to steeper slopes may be pre-Quaternary in age. (Photograph taken (Hack and Goodlett. 1960). That mechanical August 19, 1980) PUBLICATION 75 age from probable Tertiary to contemporary. Most are products of mass wasting. The steep slopes of the Appalachian region are among the most landslide-prone in the nation. During periods of high rainfall, hundreds of debris avalanches, along with mudflows, slumps, flash floods, rockfalls, reactivated earthflows and the like, have been shown to occur (see, for example, Hack and Goodlett, 1960). In August, 1969, Hur- ricane Camille yielded thousands of mass movements-one study identified 1534 debris "slides" in a 36 -12 (93 kmz) area of the Spring Creek drainage basin in eastern West Virginia (Rad- bruch-Hall and others, 1976). An extreme event such as Camille-28 inches (710 mm) of rainfall Figure 3. View of northern slopes of Rich Moun- in 8 hours-has an estimated recurrence interval tain, showing footslope deposits, some 6.2 miles of about 1000 years. It produced foot-slope de- (10 km) east of Tazewell, Tazewell County, Vir- posits averaging from 12 to 24 inches (30 to 60 ginia. Note boulders and blocks along swales at cm) in thickness, some containing blocks up to foot of mountain. Are these "boulder streams" 6 feet (2 m) in diameter (Williams and Guy, 1973). Pleistocene deposits with the fines being win- Kochel and others (1982) found evidence for re- nowed by modern fluvial action, or debris flows peated debris avalanching in central Virginia from rare catastrophic floods of Holocene age, every 700 to 4000 years. Individual events can or neither? (Photograph taken June 22, 1980) be recognized using such criteria as buried soils and textural, structural and mineralogical differences between the deposits. Many of the boulder streams which have been to south, becoming rare south of Asheville, North described "rn the Valley and Ridge southwest of Carolina. The erosional and winnowing action of Roanoke are also diamictons. These bouldery runoff from intense rains can explain the first accumulations occur in the bottoms of small val- two observations and the latter two hardly con- leys on the sides of steep ridges, especially where stitute proof of periglacial, as opposed to more a resistant rock such as the Silurian Tuscarora recent, origins. Nevertheless, the evidence is sug- Sandstone directly overlies a nonresistant unit gestive of a periglacial origin. (Figure 3). Interpretations of the deposits have Mills (1981) noted that the great quantity, ranged from alluvium to defunct rock glaciers. depth, and apparent age of apron-like diamictons Although some may have accumulated by geli- at Grandfather Mountain, North Carolina sug- fluction and boulder creep under periglacial con- gested accumulation under periglacial condi- ditions, most appear to be the product of flash tions. These deposits appear older near inter- floods and debris flows during intense or "cat- fluves and younger near drainageways. A follow- astrophic" rainfalls (Mills, 1978, 1980a, 1980b). up study (Mills, 1982) on such bouldery foot-slope Houser (1980) also concludes that virtually all the deposits seems to document a distinct age dif- colluvial-type deposits in the New River region ferential of at least 10,000 years, based on the near Blacksburg are of recent rather than pe- percentage of clay, degree of oxidation, and clast riglacial origin. weathering in the B-2 soil horizons. Such a con- However, other diamictons exist, especially at trast in age would be very unlikely had the high elevations, which cannot be readily ex- deposits formed by ongoing random events such plained by Holocene processes. Michalek (1968) as those generated by catastrophic rainfalls. studied many bouldery fan-like colluvial deposits Features composed of coarse rock fragments and concluded that most were of periglacial or- and lacking fine interstitial materials have re- igin. His evidence was that (1) the deposits seem ceived much attention as paleoclimatic indica- to be undergoing erosion at the present time; (2) tors. Many such features ociur as large patches the upper ends of many deposits are covered with of stony rubble covering slopes. Talus, scree, and block fields; (3) no fans were found below about felsenmeer are all basically unvegetated, or 2300 feet (700 m) elevation; ( ) the fans diminish sparsely vegetated, areas covered by large broken in number and degree of development from north fragments of resistant rock (Figures 4 and 5). PUBLICATION 75 age from probable Tertiary to contemporary. Most are products of mass wasting. The steep slopes of the Appalachian region are among the most landslide-prone in the nation. During periods of high rainfall, hundreds of debris avalanches, along with mudflows, slumps, flash floods. rockfalls. reactivated earthflows and the like, have been shown to occur (see, for example, Hack and Goodlett, 1960). In August, 1969, Hur- ricane Camille yielded thousands of mass movements-one study identified 1534 debris "slides" in a 36 mi2 (93 km2) area of the Spring Creek drainage basin in eastern West Virginia (Rad- bruch-Hall and others, 1976). An extreme event such as Camille-28 inches (710 mm) of rainfall Figure 3. View of northern slopes of Rich Moun- in 8 hours-has an estimated recurrence interval tain, showing footslope deposits, some 6.2 miles of about 1000 years. It produced foot-slope de- (10 km) east of Tazewell, Tazewell County, Vir- posits averaging from 12 to 24 inches (30 to 60 ginia. Note boulders and blocks along swales at cm) in thickness, some containing blocks up to foot of mountain. Are these "boulder streams" 6 feet (2 m) in diameter (Williams and Guy, 1973). Pleistocene deposits with the fines being win- Kochel and others (1982) found evidence for re- nowed by modern fluvial action, or debris flows peated debris avalanching in central Virginia from rare catastrophic floods of Holocene age, every 700 to 4000 years. Individual events can or neither? (Photograph taken June 22, 1980) be recognized using such criteria as buried soils and textural, structural and mineralogical differences between the deposits. Many of the boulder streams which have been to south, becoming rare south of Asheville, North described in the Valley and Ridge southwest of Carolina. The erosional and winnowing action of Roanoke are also diamictons. These bouldery runoff from intense rains can explain the first accumulations occur in the bottoms of small val- two observations and the latter two hardly con- leys on the sides of steep ridges, especially where stitute proof of periglacial, as opposed to more a resistant rock such as the Silurian Tuscarora recent, origins. Nevertheless, the evidence is sug- Sandstone directly overlies a nonresistant unit gestive of a periglacial origin. (Figure 3). Interpretations of the deposits have Mills (1981) noted that the great quantity, ranged from alluvium to defunct rock glaciers. depth, and apparent age of apron-like diamictons Although some may have accumulated by geli- at Grandfather Mountain, North Carolina sug- fluction and boulder creep under periglacial con- gested accumulation under periglacial condi- ditions, most appear to be the product of flash tions. These deposits appear older near inter- floods and debris flows during intense or "cat- fluves and younger near drainageways. A follow- astrophic" rainfalls (Mills, 1978, 1980a, 1980b). up study (Mills, 1982) on such bouldery foot-slope Houser (1980) also concludes that virtually all the deposits seems to document a distinct age dif- colluvial-type deposits in the New River region ferential of at least 10,000 years, based on the near Blacksburg are of recent rather than pe- percentage of clay, degree of oxidation, and clast riglacial origin. weathering in the B-2 soil horizons. Such a con- However, other diamictons exist, especially at trast in age would be very unlikely had the high elevations, which cannot be readily ex- deposits formed by ongoing random events such plained by Holocene processes. Michalek (1968) as those generated by catastrophic rainfalls. studied many bouldery fan-like colluvial deposits Features composed of coarse rock fragments and concluded that most were of periglacial or- and lacking fine interstitial materials have re- igin. His evidence was that (1) the deposits seem ceived much attention as paleoclimatic indica- to be undergoing erosion at the present time; (2) tors. Many such features occur as large patches the upper ends of many deposits are covered with of stony rubble covering slopes. Talus, scree, and block fields; (3) no fans were found below about felsenmeer are all basically unvegetated, or 2300 feet (700 m) elevation; (4) the fans diminish sparsely vegetated, areas covered by large broken in number and degree of development from north fragments of resistant rock (Figures 4 and 5). 10 VIRGINIA DIVISION OF MINERAL RESOURCES

gently sloping (usually less than 15o) or even flat areas, and the blocks are typically larger in mean size than talus or scree, often a meter or more in diameter. If the rocks form a linear body extending down a hillside, "block stream" or some similar term is often used. If blocks are rounded, "boulder field" or "boulder stream" are appropriate terms. As with many groups of landforms, these patches of broken rocks appear to be part of a suite of gradational features and deciding on the most appropriate term for some sites can be difficult. Figure 4. Part of a scree slope near Sherando Generally, the gently-sloping, very coarse- Lake, Blue Ridge Mountains, Augusta County. grained felsenmeer-type features appear to be Quartzite blocks lie on steep southeast-facing periglacial in origin and inactive in today's cli- slopes at about 2000 feet (610 m) elevation. Ev- mate. They are very well developed in Pennsyl- idence indicates this slope is currently quite vania, where several are popular tourist attrac- active. (Photograph taken June 12, 1984) tions, and become less frequent, less distinct, and more poorly developed southward along the Ap- palachians. Among the better known felsenmeer in Pennsylvania are the Devil's Race Course, Devil's Potato Patch, Blue Rocks Block Field, and the Hickory Run Boulder Field (Geyer and Bolles, 1979). The latter may be the largest such feature in the eastern United States, measuring 400 x 1800 feet (122 x 549 m) in areal extent. The Blue Rocks Block Field, which is at least 59 feet (18 m) deep in places and contains blocks of Tuscar- ora Sandstone up to 39 feet (12 m) long, was interpreted as a product of creep and gelifluction during the Wisconsinan glacial by Potter and Moss (1968) who noted its close resemblance to gelifluction sheets and terraces in Alaska. More recently, Psilovikos and van Houten (1982) came to the same basic conclusion after studying the Figure 5. Large dislocated blocks of resistant Ringing Rocks field along the Delaware River metabasalt along the Greenstone Trail, Blue in eastern Pennsylvania. Felsenmeer require the Ridge Parkway, Augusta County. This location presence of massive, resistant bedrock with quite is just a few miles from the scree of Figure 4 widely spaced, distinct joints. Where the resistant and several hundred feet higher in elevation. rock strata are sufficiently thick, and well-de- Large block to left is about 20 feet (6 m) high veloped, widely spaced joints are present, enor- with a large cleft behind it. Unlike many screes, mous blocks the size of boxcars or small buildings such large-scale blocks in the poorly developed (Figure 6) may become detached and migrate felsenmeer or rock city-type areas of Virginia downslope by a process called cambering, which appear to be inactive periglacial relics. (Photo- is especially effective in a periglacial environ- graph taken June 13, 1984) ment (Eyles, 1983, p. I22). Huge blocks, especially if isolated, may also result from the unearthing of large corestones from deeply Slopes covered by talus and scree are typically weathered saprolite (Hupp, 1983). In the resistant steep, usually over 25o, and the rock fragments bedrock itself, deeply weathered joints and tend to be of moderate size, often averaging slightly displaced blocks may form the deep nar- around 8 inches (20 cm) in mean diameter. Talus row passageways of "rock cities." Similar fea- is derived directly from a cliff at the head of the tures include periglacial tors, castellated rocks, deposit while scree lacks an immediately adjacent boulder caves, and assorted fissures and clefts cliff. Felsenmeer, or block fields (bloekfields) are (Figure 5). Geyer and Bolles (1979) provided 10 VIRGINIA DIVISION OF MINERAL RESOURCES

gently sloping (usually less than 15o) or even flat areas, and the blocks are typically larger in mean size than talus or scree, often a meter or more in diameter. If the rocks form a linear body extending down a hillside, "block stream" or some similar term is often used. If blocks are rounded, "boulder field" or "boulder stream" are appropriate terms. As with many groups of landforms, these patches of broken rocks appear to be part of a suite of gradational features and deciding on the most appropriate term for some sites can be difficult. Figure 4. Paft of a scree slope near Sherando Generally, the gently-sloping, very coarse- Lake, Blue Ridge Mountains, Augusta County. grained felsenmeer-type features appear to be Quartzite blocks lie on steep southeast-facing periglacial in origin and inactive in today's cli- slopes at about 2000 feet (610 m) elevation. Ev- mate. They are very well developed in Pennsyl- idence indicates this slope is currently quite vania, where several are popular tourist attrac- active. (Photograph taken June 12, 1984) tions, and become less frequent, less distinct, and more poorly developed southward along the Ap- palachians. Among the better known felsenmeer in Pennsylvania are the Devil's Race Course, Devil's Potato Patch, Blue Rocks Block Field, and the Hickory Run Boulder Field (Geyer and Bolles, 1979). The latter may be the largest such feature in the eastern United States, measuring 400 x 1800 feet (I22x 549 m) in areal extent. The Blue Rocks Block Field, which is at least 59 feet (18 m) deep in places and contains blocks of Tuscar- ora Sandstone up to 39 feet (12 m) long, was interpreted as a product of creep and gelifluction during the Wisconsinan glacial by Potter and Moss (1968) who noted its close resemblance to gelifluction sheets and terraces in Alaska. More recently, Psilovikos and van Houten (1982) came to the same basic conclusion after studying the Figure 5. Large dislocated blocks of resistant Ringing Rocks field along the Delaware River metabasalt along the Greenstone Trail, Blue in eastern Pennsylvania. Felsenmeer require the Ridge Parkway, Augusta County. This location presence of massive, resistant bedrock with quite is just a few miles from the scree of Figure 4 widely spaced, distinct joints. Where the resistant and several hundred feet higher in elevation. rock strata are sufficiently thick, and well-de- Large block to left is about 20 feet (6 m) high veloped, widely spaced joints are present, enor- with a large cleft behind it. Unlike many screes, mous blocks the size of boxcars or small buildings such large-scale blocks in the poorly developed (Figure 6) may become detached and migrate felsenmeer or rock city-type areas of Virginia downslope by a process called cambering, which appear to be inactive periglacial relics. (Photo- is especially effective in a periglacial environ- graph taken June 13, 1984) ment (Eyles, 1983, p. I22). Huge blocks, especially if isolated, may also result from the unearthing of large corestones from deeply Slopes covered by talus and scree are typically weathered saprolite (Hupp, 1983). In the resistant steep, usually over 25o, and the rock fragments bedrock itself, deeply weathered joints and tend to be of moderate size, often averaging slightly displaced blocks may form the deep nar- around 8 inches (20 cm) in mean diameter. Talus row passageways of "rock cities." Similar fea- is derived directly from a cliff at the head of the tures include periglacial tors, castellated rocks, deposit while scree lacks an immediately adjacent boulder caves, and assorted fissures and clefts cliff. Felsenmeer, or block fields (blockfields) are (Figure 5). Geyer and Bolles (1979) provided PUBLICATION 75 11

forming bedrock by forests which are older near the outcrop than lower on the slope (Hupp, 1988). In addition to resistant upslope bedrock with suitable joint or fracture patterns, the development of large "stone patch" phenomena is apparently favored by the presence of a nonresistant rock underlying the slopes below the resistant formations. In Virginia the nonresistant bedrock is often a shale or other clay-rich material such as the Upper Ordovician Martinsburg Forma- tion. The impermeable nature of such rock limits downward infiltration of water, leading readily to saturated conditions in the overlying colluvium. The pore pressure which develops within Figure 6. Large isolated boulder along Skyline the unconsolidated regolith greatly decreases its Drive (near Mile 79) in Shenandoah National cohesion. In steeply-sloping treeless scree and Park. No cliffs are nearby and slopes in the forest talus areas, gravitational attraction can over- above the boulder are gentle. A case of periglacial come the weakened cohesive forces during peri- boulder creep, unearthing of a corestone, or both ods of heavy or extended rainfall, causing slope . , or something completely different? (pho- failures, especially at the base of the slopes where tograph taken June 13, 1984) saturation and pore pressures are greatest. Re- cent movements of scree on northern Massanutten Mountain have been shown to coincide with pe- descriptions and photographs of many such fea- riods of heavy rainfall and high streamflow tures, with periglacial action frequently cited as (Hupp, 1983). The numerous "blow-out pits" critical to their origin. Most of these phenomena (landslide headwall hollows or "cirques") which reveal a heavy lichen cover, exhibit deeply pitted have been observed on scree slopes following and weathered rock surfaces, and show no sign heavy downpours attest to the importance of this of present activity, attesting to their relic nature. process. Basal lubrication and sliding along the In contrast, scree slopes, which are especially clay-rich underbase, interstitial water flow (pip- common along steep slopes in the northern Blue ing), and freeze-thaw may also play a roll in Ridge and Valley and Ridge provinces of Virginia modern as well as glacial-age activity along such (Figure 4), frequently show such evidence of slopes. It is generally believed that formation of current activity as (1) tilted trees and scree- talus and scree was more effective during glacials inundated shrubs at the base of the slope; (2) and that many presently active screes are merely abundant fresh rock surfaces and other evidence remnants of the more extensive block-covered of recent turnover; (3) less than75 percent lichen slopes of a colder climate. cover; (4) depressions or hollows on the scree Freeze-thaw activity may encourage movement surface; (5) broad tumescences or bulges along along rock-strewn slopes in several ways. Early lower scree slopes; (6) recent rock-slide tracks and season snowfall packs in around blocks. Melting, scars on slopes below the scree; (Z) damming of penetration and refreezing of the water occurs, scree behind large logs or other blockages on the replacing air spaces with ice which, unlike air, slope; and (8) frequent scars or deformations on is a good conductor of heat. The result is accel- trees located on or adjacent to the scree slopes erated heat loss from the block field or scree, (Hack, 1965; Hupp, 1983). encouraging ever deeper freezing. Water seeping Where scree slopes and felsenmeer-type fea- downslope along the surface between the imper- tures are found in the same area, the scree gen- meable underlying material and the colluvium erally occupies steeper slopes at lower elevations. will freeze beneath the barren rubble causing It appears that many scree slopes are currently additional expansion and pressure. In the spring- undergoing downward migration by a series of time, the upper part of the ice surrounding the complex mass movements as indicated by slide blocks will melt and be replaced by stagnant air tracks and scars near the bottom of the scree which helps to insulate the lower ice and retard slope. More stable slopes are evident at and its melting (Hack and Goodlett, 1960). Even to- beyond the upper part of the slope. Scree is day, ice has been found in Virginia scree as late separated from upslope outcrops of the scree- as July, and in some deep crevasses or block caves PUBLICATION 75 11

forming bedrock by forests which are older near the outcrop than lower on the slope (Hupp, 1g83). In addition to resistant upslope bedrock with suitable joint or fracture patterns, the development of large "stone patch" phenomena is apparently favored by the presence of a nonresistant rock underlying the slopes below the resistant formations. In Virginia the nonresistant bedrock is often a shale or other clay-rich material such as the Upper Ordovician Martinsburg Forma- tion. The impermeable nature of such rock limits downward infiltration of water, leading readily to saturated conditions in the overlying colluvium. The pore pressure which develops within Figure 6. Large isolated boulder along Skyline the unconsolidated regolith greatly decreases its Drive (near Mile 79) in Shenandoah National cohesion. In steeply-sloping treeless scree and Park. No cliffs are nearby and slopes in the forest talus areas, gravitational attraction can over- above the boulder are gentle. A case of periglacial come the weakened cohesive forces during peri- boulder creep, unearthing of a corestone, or both ods of heavy or extended rainfall, causing slope . , or something completely different? (pho- failures, especially at the base of the slopes where tograph taken June 13, 1984) saturation and pore pressures are greatest. Re- cent movements of scree on northern Massanutten Mountain have been shown to coincide with pe- descriptions and photographs of many such fea- riods of heavy rainfall and high streamflow tures, with periglacial action frequently cited as (Hupp, 1983). The numerous "blow-out pits" critical to their origin. Most of these phenomena (landslide headwall hollows or "cirques") which reveal a heavy lichen cover, exhibit deeply pitted have been observed on scree slopes following and weathered rock surfaces, and show no sign heavy downpours attest to the importance of this of present activity, attesting to their relic nature. process. Basal lubrication and sliding along the In contrast, scree slopes, which are especially clay-rich underbase, interstitial water flow (pip- common along steep slopes in the northern Blue ing), and freeze-thaw may also play a roll in Ridge and Valley and Ridge provinces of Virginia modern as well as glacial-age activity along such (Figure 4), frequently show such evidence of slopes. It is generally believed that formation of current activity as (1) tilted trees and scree- talus and scree was more effective during glacials inundated shrubs at the base of the slope; (2) and that many presently active screes are merely abundant fresh rock surfaces and other evidence remnants of the more extensive block-covered of recent turnover; (3) less than Zb percent lichen slopes of a colder climate. cover; (4) depressions or hollows on the scree Freeze-thaw activity may encourage movement surface; (5) broad tumescences or bulges along along rock-strewn slopes in several ways. Early lower scree slopes; (6) recent rock-slide tracks and season snowfall packs in around blocks. Melting, scars on slopes below the scree; (Z) damming of penetration and refreezing of the water occurs, scree behind large logs or other blockages on the replacing air spaces with ice which, unlike air, slope; and (8) frequent scars or deformations on is a good conductor of heat. The result is accel- trees located on or adjacent to the scree slopes erated heat loss from the block field or scree, (Hack, 1965; Hupp, 1983). encouraging ever deeper freezing. Water seeping Where scree slopes and felsenmeer-type fea- downslope along the surface between the imper- tures are found in the same area, the scree gen- meable underlying material and the colluvium erally occupies steeper slopes at lower elevations. will freeze beneath the barren rubble causing It appears that many scree slopes are currently additional expansion and pressure. In the spring- undergoing downward migration by a series of time, the upper part of the ice surrounding the complex mass movements as indicated by slide blocks will melt and be replaced by stagnant air tracks and scars near the bottom of the scree which helps to insulate the lower ice and retard slope. More stable slopes are evident at and its melting (Hack and Goodlett, 1960). Even to- beyond the upper part of the slope. Scree is day, ice has been found in Virginia scree as late separated from upslope outcrops of the scree- as July, and in some deep crevasses or block caves L2 VIRGINIA DIVISION OF MINERAL RESOURCES of New York and New England, it may persist Virginia include sorted stone polygons, nets, all summer. Thus it is quite probable that patches stripes, and steps. The larger of these forms of permafrost existed in Virginia during glacials. (roughly a meter or more in width or diameter) Part of a scree slope cemented and rendered are generally regarded as periglacial relics impermeable by such persistent ice forms a bar- (Clark, 1968). Some of the smaller features may rier to downslope subsurface water seepage, caus- be currently active. Table 2 lists a number of such ing locally high pore pressures and encouraging features and indicates the probability of finding failure (e.g., a rock avalanche) at that place on them in Virginia. Clark (1984) reports that stud- the scree slope. Even without development of such ies on such features are nearing completion at frozen barriers, the ground saturation of slopes Whitetop Mountain and Interior, Virginia and and the frequent refreezing during the perigla- other locations. It is hoped that radiocarbon and cial regimes would enhance both the production thermoluminescence dating of critical soil horiz- of more blocks and downslope movements. ons will begin to yield a chronology which can Other periglacial-related features occurring in be applied to such phenomena.

Table 2. Occurrence of relic periglacial features in Virginia.

Feature O ccurr ence in Virgi,nia

Felsenmeer (blockf ields) yes Rock cities yes; small compared to those further north. Large cambered blocks yes Scree and talus slopes yes; considerable ongoing activity. Sorted and stone polygons, nets yes; small forms may be active at and stripes high elevations. Tors yes; other origins possible; some active. Rock glaciers improbable. Cryoplanation terraces probable. Thermokarst depressions improbable. Pingos highly improbable. Palsas (palsen) improbable; easily destroyed - preservation unlikely. Ice-wedge polygons and casts highly improbable. Soil-wedge polygons improbable. Periglacial involutions yes; origins may be questionable. (frost-heaved soils) Gelifluction fans, lobes, aprons, yes; difficult to distinguish from terraces Holocene mass wasting features. Periglacial soil mounds probable; origins of such features remains debatable. Ice-rafted clasts in alluvium. yes.

Principal sources: Eyles, 1983; Pewe, 1983;Washburn, 1980. PUBLICATION 75 13

FLUVIAL PROCESSES effects in Virginia where the mean annual rainfall is about 47 inches (1200 mm) per year. Dis- The impacts of climatic change upon fluvial tinct terraces and alluvial deposits which can be processes are highly complex. Even within a attributed to climatic change are very rare in single region, agiven change mayproduce stream the central and southern Appalachians and pro- erosion in one drainage basin while inducing cesses currently operative can explain most such deposition in another. Factors influencing stream features. Many stream terraces in the Valley and response to a given shift in clirnate include vege- Ridge province appear to be a product of stream tation, relief, steepness and orientation of slopes, piracy in which a stream forced to erode through size and shape of the drainage basin, nature of resistant bedrock is eventually intersected or bedrock and regolith, topographic texture, and captured by another which has incised more geomorphic setting. The last named factor may deeply by exploiting a zone of nonresistant rock. include such processes as active uplift, subsi- The result is a lowering of base level for the dence, volcanic or thermal phenomena, loess de- captured stream, often followed by accelerated position, sea-level changes in coastal regions, and downcutting and the development of straths and the like. Basically, anything that alters the load- alluvial terraces along the valley (Figure 7). discharge ratio of a stream is likely to trigger Hack (1965) describes another, more subtle some response by the stream (Flint, 1971, p. 304- type of piracy in which streams flowing from 313). rugged mountain terrain onto broad valleys in It is commonly held that in most mid-latitude Virginia deposit fans and aprons at the base of continental areas a colder climate will induce the mountains. Streams which flow only upon the stream aggradation by increasing stream load- less resistant rocks of the valley lack both the a result of larger quantities of clastic sediment high load of the mountain streams and their foot- being supplied to streams by glaciers, wind, and/ slope deposits. Eventually such a stream, or a or increased rates of mechanical weathering and tributary thereof, will capture a higher-level mass wasting. Many areas do indeed contain mountain stream. This stream now takes on the distinctive terraces and alluvial deposits, often load of the mountain stream, becomes aggrada- of regional extent and clearly contemporaneous tional, and rvill itself eventually be pirated in a with cold glacial episodes. More directly, climatic changes alter the distribution and magnitude of runoff, causing profound changes in stream behavior, especially in more arid regions. For example, as mean annual precipitation increases from zero, rapid increases in runoff, flood potential and sediment yield result (Knox, 1983). When rainfalls totaling from 12-20 inches (300-500 mm) per year are reached, sediment yield is at a maximum and the desert begins to change to grassland. As precipitation increases beyond this major threshold, increased vegetation density helps stabilize slopes and retard runoff, leading to a very pronounced reduction in flashflooding and sediment yield. At 31- 35 inches (800 to 900 mm) of rainfall per year, forests begin to take over from the grassland. Forests are slightly more effective in dampening Figure 7. Terrace-like features (stratigraphic flood effects and sediment yield than grasslands, benches, straths, alluvial terraces?) along east perhaps because they are less severely impacted side of Big Walker Creek Valley southwest of the by fires and overgrazing. As rainfall increases State Route 42152 junction, Bland County, Vir- beyond about 35 inches (900 mm) per year, forest ginia. Are these and similar features found cover persists and rivers are no longer as respon- around the state related to bedrock control, base sive to such phenomena as heavy rainfalls and level variations controlled by stream piracies, climatic changes as they are in more arid areas. changes in the rate of downcutting in water gaps, This scenario helps explain why Quaternary or changes in stream regimen.caused by climatic climatic changes did not yield more pronounced variations? (Photograph taken June 8, 1984) PUBLICATION 75 13

FLUVIAL PROCESSES effects in Virginia where the mean annual rainfall is about 47 inches (1200 mm) per year. Dis- The impacts of climatic change upon fluvial tinct terraces and alluvial deposits which can be processes are highly complex. Even within a attributed to climatic change are very rare in single region, a given change may produce stream the central and southern Appalachians and pro- erosion in one drainage basin while inducing cesses currently operative can explain most such deposition in another. Factors influencing stream features. Many stream terraces in the Valley and response to a given shift in clirnate include vege- Ridge province appear to be a product of stream tation, relief, steepness and orientation of slopes, piracy in which a stream forced to erode through size and shape of the drainage basin, nature of resistant bedrock is eventually intersected or bedrock and regolith, topographic texture, and captured by another which has incised more geomorphic setting. The last named factor may deeply by exploiting a zone of nonresistant rock. include such processes as active uplift, subsi- The result is a lowering of base level for the dence, volcanic or thermal phenomena, loess de- captured stream, often followed by accelerated position, sea-level changes in coastal regions, and downcutting and the development of straths and the like. Basically, anything that alters the load- alluvial terraces along the valley (Figure 7)" discharge ratio of a stream is likely to trigger Hack (1965) describes another, more subtle some response by the stream (Flint, 1971, p. 304- type of piracy in which streams flowing from 313). rugged mountain terrain onto broad valleys in It is commonly held that in most mid-latitude Virginia deposit fans and aprons at the base of continental areas a colder climate will induce the mountains. Streams which flow only upon the stream aggradation by increasing stream load- less resistant rocks of the valley lack both the a result of larger quantities of clastic sediment high load of the mountain streams and their foot- being supplied to streams by glaciers, wind, and/ slope deposits. Eventually such a stream, or a or increased rates of mechanical weathering and tributary thereof, will capture a higher-level mass wasting. Many areas do indeed contain mountain stream. This stream now takes on the distinctive terraces and alluvial deposits, often load of the mountain stream, becomes aggrada- of regional extent and clearly contemporaneous tional, and will itself eventually be pirated in a with cold glacial episodes. More directly, climatic changes alter the distribution and magnitude of runoff, causing profound changes in stream behavior, especially in more arid regions. For example, as mean annual precipitation increases from zero, rapid increases in runoff, flood potential and sediment yield result (Knox, 1983). When rainfalls totaling from 12-20 inches (300-500 mm) per year are reached, sediment yield is at a maximum and the desert begins to change to grassland. As precipitation increases beyond this major threshold, increased vegetation density helps stabilize slopes and retard runoff, leading to a very pronounced reduction in flashflooding and sediment yield. At 81- 35 inches (800 to 900 mm) of rainfall per year, forests begin to take over from the grassland. Forests are slightly more effective in dampening Figure 7. Terrace-like features (stratigraphic flood effects and sediment yield than grasslands, benches, straths, alluvial terraces?) along east perhaps because they are less severely impacted side of Big Walker Creek Valley southwest of the by fires and overgrazing. As rainfall increases State Route 42152 junction, Bland County, Vir- beyond about 35 inches (900 mm) per year, forest ginia. Are these and similar features found cover persists and rivers are no longer as respon- around the state related to bedrock control, base sive to such phenomena as heavy rainfalls and level variations controlled by stream piracies, climatic changes as they are in more arid areas. changes in the rate of downcutting in water gaps, This scenario helps explain why Quaternary or changes in stream regimen.caused by climatic climatic changes did not yield more pronounced variations? (Photograph taken June 8, 1984) l4 VIRGINIA DIVISION OF MINERAL RESOURCES

similar fashion. The overall result is a continuing deposition as heavy rainfalls destabilized rubble- denudation of the region in a very complex, covered slopes which then delivered excessive dynamic fashion which can leave behind numer- amounts of sediment to streams. These related ous terrace remnants and alluvial deposits of events may help explain the extensive alluviation widely varying ages. As Hack notes, although which apparently took place along the Little climatic change probably altered somewhat the Tennessee River during the later stages of gla- rates of fluvial activity, no fundamental change cial-interglacial and stadial-interstadial transi- in basic procedures occurred and clear evidence tion periods (Delcourt, 1980). Following such a of various climatic regimes is diffieult to recog- transitional phase, the more frequent occurren- nize in the alluvial sequences. ces of large river discharges would.tend to lead The influence of catastrophic rainfalls is also to more rapid downcutting than occurred during important to an understanding of fluvial pro- the eolder periods. Generally, records indicate cesses. As noted above, many boulder streams that wetter climates tend to induce stream in- found along small first and second order stream cision and drier elimates stream aggradation valleys are apparently formed by debris flows and (Knox, 1983). flash floods (Mills, 1980b). Such deposits may be At higher elevations in Virginia, lengthy pe- gradational between colluvium and alluvium. riods of subfreezing temperatures and extensive Kochel and Johnson (1984) conclude that many areas of frozen ground were probably common humid-temperate "alluvial" fans in central Vir- during glacial winters. Where the snowpack is ginia are actually debris avalanche deposits built shallow, vegetation sparse, and temperatures suf- up during infrequent storms occurring about ficiently cold, the regolith may be invaded by once every 3000 to 6000 years. Heavy runoff tends dense impermeable ice ("concrete frost") which to redistribute the fan deposits by eroding ma- may persist through part or all of the summer terial at the fan head and depositing it on the (Dunne and Leopold, 1978, p. 167). Under such lower slopes. conditions, wintertime discharge along mountain Major floods along larger streams may dras- streams may cease. As in many cold regions today, tically alter the valley floor and stream channel much of the annual streamflow may have been (Hack and Goodlett, 1960), stripping away vege- discharged during the spring melting period. tation, piling up deposits which may later form Runoff would be rapid over the frozen ground terraces, and even destabilizing the stream, al- and springtime floods, some triggered by tering its behavior for years to come. In some breakup of ice jams along rivers, would be com- stream systems, such events, with recurrence mon. The magnitude of these floods and their intervals of several hundred years, may have impaets would in all probability be far less than more geomorphic impact than all lesser events those from a Hurricane Camille, however, and combined. Differentiating the effects of such epi- clear-cut evidence of such past processes is rare. sodic events from possible climatic effects is a One indication of such events may be the anom- major challenge that needs to be met if alluvial alously large, striation-bearing clasts found in sequences are to be accurately interpreted. Pleistocene deposits along the lower reaches of One important influence of a cold climate on many major Appalachian rivers such as the fluvial systems in Virginia would probably be a James and Rappahannoek (Wentworth, L928; reduction, perhaps the near-elimination, of in- Flint, 1971, p. 312-313). The clasts are very sim- tense and catastrophic rainfall events. Because ilar to those found in today's Arctie streams where the Polar Frontal Zone was well to the south of river ice action is vigorous. Such clasts seem to Virginia during glacial maxima, little warm occur only in sediments deposited during cold moist air could mix with the cold dry westerlies climatic intervals. and their size and abundance flowing across the state. Such conditions should decrease from the Susquehanna River south to help stabilize streams and reduce their erosion the Roanoke River. and sediment transportation rates. At the same In summary, the responses of fluvial systems time, production and migration of loose rubble in the Virginia highlands to climatie changes are over most of the terrain was probably greater complex and poorly understood. Evidence of fluv- than during warm periods. As the shift from ial processes during eold glacial intervals is often colder to warmer climate occurred and warm subtle and difficult to distinguish from the effects moist air masses began following the Polar Front of catastrophic Holocene rainfall events and non- as it moved northward, rainfall would increase. climatie influences such as stream piracy. It is The initial effeet might have been increased probable that aggradation tended to dominate PUBLICATION 75 15

along most streams during, and especially at the Springs, seeps or streams heading in carbonate close of, colder and drier episodes and degrada- rocks frequently precipitate calcium carbonate tion tended to prevail during warmer, wetter at favorable locations. The resulting deposits can intervals. sometimes grow to impressive size, extending for miles along streambeds, building outward to PROCESSES OF CAVERN AND KARST form waterfalls, or accumulating as mounds and DEVELOPMENT fans (Figure 8). These deposits suggest extensive subsurface circulation of waters and dissolution The utility of karst terrain in providing clues of carbonate bedrock. Deposition is favored where to climatic changes is somewhat controversial the water stream spreads out, allowing more COo (Jennings, 1981) but, in general, appears to be loss, more evaporation, and more algae and moss limited. Wells and Gutierrez (1981) report that growth (which takes COo from the water), all of changes in base level may be reflected as changes which tend to cause CaeO, precipitation. Tufa in slope along longitudinal profiles of karst deposits may therefore record periods of depo- streams, but dating and relating the previous sition, stagnation, or erosion which can be related base levels to specific climatic intervals is very to climatic variations. Tufa, as well as travertine difficult in most areas. formations in caves, may also yield valuable cli- Straths and terraces along major rivers often matic information from entrapped pollen and represent former extended base levels. In nearby other fossil material, fluid inclusions, isotopic soluble rocks, large integrated cavern systems variations in minerals, and embedded clastic sloping gently towards the base level may have materials. existed. The cave systems developed slightly Many tufa and travertine deposits, as well as below and roughly parallel to the water table clastic cave sediments, have been dated using which was graded to local river levels. The vast radiocarbon, thorium-uranium, and remnant Flint Ridge Mammoth Cave system in Kentucky, plaeomagnetism techniques (Latham, 1981). Re- which is some 310 miles (500 km) in length and sults are very encouraging, with 230Th/zgaU dating still being extended, shows seven primary levels of travertine and tufa providing the most abun- which are correlated with terrace remnants and horizontally extended surface landforms controlled by base levels of the Ohio River during late Tertiary and Pleistocene time. The higher the cave levels, the older they appear to be. When the number of levels is compared with available ages for travertine and sediments, there appears to be about half as many cave levels as glacial- interglacial cycles. Hess and Harmon (1981) suggest a complex series of responses involving excavation of cave passages and cave sediment during interglacials when the Ohio River was downcutting, thus lowering base levels and increasing the hydraulic gradient in the caves. This excavation phase would be followed by partial infilling of the lower cave levels by sediment during glacials when base level was higher because of alluviation along the Ohio River. It Figure 8. Meadow Creek Falls near New Castle, appears that two interglacials are required for Craig County, Virginia. This creek drains the base level to get low enough for complete removal eastern part of Sinking Creek Valley, a large of the sediment in the existing cave and the breached anticline underlain by limestones. The initiation of a new cycle of cave development at falls are on Silurian sandstones, but the presence a lower level. of carbonate rocks upstream is evident along the Increased rainfall in an area should favor in- creek-the upper part of the falls seen here flows creased solution ofcarbonate rock and more rapid over a large tufa mound some 30 feet (10 m) high. accumulation of residual debris on the surface. Most tufa buildups are not so obvious, but they However, recognition of such changes in the terra may contain valuable clues to the paleoenviron- rossa or other karst regolith is very difficult. ment. (Photograph taken August 26,7982) PUBLICATION 75 15

along most streams during, and especially at the Springs, seeps or streams heading in carbonate close of, colder and drier episodes and degrada- rocks frequently precipitate calcium carbonate tion tended to prevail during warmer, wetter at favorable locations. The resulting deposits can intervals. sometimes grow to impressive size, extending for miles along streambeds, building outward to PROCESSES OF CAVERN AND KARST form waterfalls, or accumulating as mounds and DEVELOPMENT fans (Figure 8). These deposits suggest extensive subsurface circulation of waters and dissolution The utility of karst terrain in providing clues of carbonate bedrock. Deposition is favored where to climatic changes is somewhat controversial the water stream spreads out, allowing more COo (Jennings, 1981) but, in general, appears to be loss, more evaporation, and more algae and mosi limited. Wells and Gutierrez (1981) report that growth (which takes CO, from the water), all of changes in base level may be reflected as changes which tend to cause CaCO" precipitation. Tufa in slope along longitudinal profiles of karst deposits may therefore recbrd periods of depo- streams, but dating and relating the previous sition, stagnation, or erosion which can be related base levels to specific climatic intervals is very to climatic variations. Tufa, as well as travertine difficult in most areas. formations in caves, may also yield valuable cli- Straths and terraces along major rivers often matic information from entrapped pollen and represent former extended base levels. In nearby other fossil material, fluid inclusions, isotopic soluble rocks, large integrated cavern systems variations in minerals, and embedded clastic sloping gently towards the base level may have materials. existed. The cave systems developed slightly Many tufa and travertine deposits, as well as below and roughly parallel to the water table clastic cave sediments, have been dated using which was graded to local river levels. The vast radiocarbon, thorium-uranium, and remnant Flint Ridge Mammoth Cave system in Kentucky, plaeomagnetism techniques (Latham, 1981). Re- which is some 310 miles (500 km) in length and sults are very encouraging, with 230Th/234U dating still being extended, shows seven primary levels of travertine and tufa providing the most abun- which are correlated with terrace remnants and horizontally extended surface landforms controlled by base levels of the Ohio River during late Tertiary and Pleistocene time. The higher the cave levels, the older they appear to be. When the number of levels is compared with available ages for travertine and sediments, there appears to be about half as many cave levels as glacial- interglacial cycles. Hess and Harmon (1981)suggest a complex series of responses involving excavation of cave passages and cave sediment during interglacials when the Ohio River was downcutting, thus lowering base levels and increasing the hydraulic gradient in the caves. This excavation phase would be followed by partial infilling of the lower cave levels by sediment during glacials when base level was higher because of alluviation along the Ohio River. It Figure 8. Meadow Creek Falls near New Castle, appears that two interglacials are required for Craig County, Virginia. This creek drains the base level to get low enough for complet-e removal eastern part of Sinking Creek Valley, a large of the sediment in the existing cave and the breached anticline underlain by limestones. The initiation of a new cycle of cave developmenr ar falls are on Silurian sandstones, but the presence a lower level. of carbonate rocks upstream is evident along the Increased rainfall in an area should favor in- creek-the upper part of the falls seen here flows creased solution of carbonate rock and more rapid over a large tufa mound some 30 feet (10 m) high. accumulation of residual debris on the surface. Most tufa buildups are not so obvious, but they However, recognition of such changes in the terra may contain valuable clues to the paleoenviron- rossa or other karst regolith is very difficult. ment. (Photograph taken August 26,1982) 16 VIRGINIA DIVISION OF MINERAL RESOURCES

dant and reliable data, most of it for the last because of reduced precipitation as rain, in- 150,000 years (Hening and others, 1983). The creased ground ice which limits infiltration, and dates clearly show variations in calcite growth greatly decreased biotic activity. rates through time. The data may then be corre- In some caves, suites or generations of speleo- lated with chronologies obtained from terrigen- thems can be distinguished (Hennig and others, ous sediments, deep-sea cores and other sources. 1983). Their differing physical characteristics Flint (1971, p. 301-302) noted that caves had are probably related to differing elimatic re- been under-exploited as a source of information gimes. Abrupt changes in the nature of cave on paleoclimates. While still true, the situation decorations, such as abundant caleite crystals or is improving, in some cases with rather surpris- globularites below a certain level, shelf-like pro- ing results. For example, field observations on jections of travertine, or abandoned rimstone limestones in New York and elsewhere indicate deposits usually indicate former water levels in that some caves have formed since the retreat the cave which may reflect climatic conditions. of the Wisconsinan ice sheet (White, 1979). This The travertine bands in speleothems yield in- is supported by model calculations showing that formation for dating, as noted above, and also a fracture about 25 micrometers wide can evolve for paleotemperature determination. Most spe- into a subterranean flow path or conduit of about leothems, especially stalagmites and flowstone, 5 mm diameter in some 3000 years. Concentrated, contain 0.05 to 0.5 percent trapped water, repre- turbulent groundwater flow along such a conduit senting the water which formed the travertine can then enlarge it to a full-scale cave passage in the past. Oxygen and hydrogen isotope ratios of about 3 feet (1 m) diameter in less than 10,000 for the entrapped water and coexisting caleite years. This indicates that cave networks at given allows temperatures at the time of deposition to levels can indeed be formed in the time span be successfully calculated in many cases (Flint, represented by favorable climatic regimes of the 1976; Yonge, 1981). Travertine may be richer in Quaternary, and fits in nicely with the theory of iron oxides during warmer, moister periods. Hess and Harmon (1981) described above. Zones of dirty travertine may correlate to drier, Speleothem formation is generally favored by dustier cave environments when speleothem increases in rainfall and vegetation cover. The growth was minimal. Such dirty flowstone may vegetation is important because water must ab- possess a stronger remnant paleomagnetism, aid- sorb considerable carbon dioxide to be capable ing dating efforts (Latham, 1981). of dissolving significant amounts of the limestone Considerable controversy has surrounded the or dolomite it moves through, and because the interpretation of speleothems composed of min- major source of the carbon dioxide is organic erals other than calcite, especially aragonite. activity in soils. When water seeping through the Many have contended that aragonite tends to ground enters the air in a vadose cave, carbon crystallize only in higher-temperature caves dioxide is lost and calcite is deposited as trav- (above 13-17"C) (Flint, 1971, p. 303-304; Sweet- ertine. Should the water entering the cave still ing, 1973; p. 188). Moore (1956) believed caves be capable of corrosion, however, existing spe- in the western United States reflected this, with leothems may experience dissolution instead of more calcite speleothems at high latitudes and precipitation of calcite. more aragonite at lower latitudes. However, ad- In the arid southwestern United States, in- ditional studies indicate that aragonite deposition creased rainfall and vegetation cover favored may be complicated by numerous faetors other speleothem growth during glacials. In most lo- than temperature (Hill, 1976, p. 89), and the value calities around the world, however, it is the of using it as a paleoenvironment indicator re- interglacials which clearly correspond to periods mains to be proven. of increased speleothem growth (Gascoyne, 1981; Large quantities of clastic debris ranging from Glazek and Harmon, 1981; Hennig and others, clays to huge breakdown blocks are common. in 1983). During glacials, colder temperatures re- caves. Surficial material washed into caves may duce vegetation growth rates along with bacterial contain clues to the paleoenvironment. For ex- and auto-oxidative processes in the soil, thus ample, Flint (1971, p. 303) notes that rounding limiting the carbon dioxide available to infiltrat- and calcite coatings on sand grains may be related ing water. In some areas, including Virginia, to precipitation. Thick, massive clay fills in caves decreased precipitation would have an additional may contain thin bands of charcoal washed in inhibitins influence. During very cold periods, following forest fires on the surface. Wood, shells, speleothem growth may come to a virtual halt bones and other informative and datable material PUBLICATION 75 t7

may be preserved in cavern sediment, or even In Virginia, the continental shelf was not actually encased in travertine. In New York, cave sed- exposed to the current sub-sea 395-foot-depth iments from glacial periods tend to be lower in contour because the massive Laurentide ice sheet organic content and higher in soluble components to the north had caused isostatic crustal depres- than those from interglacials (Mylroie, 1979). sion. Hence net submergence of coastal plain Rockfall debris beneath near-surface domes, features off the Virginia coast may be only 195 breccia zones in cave sediments, or coarser zones - 262 feet (60-80 m) rather than 395 feet. Between of talus near entrance pits may reflect increased 16,000 and 10,000 yrs B.P., crustal rebound frost action and correlate with colder periods. caused by the retreat of the Laurentide ice sheet Reworked eolian sediment may be found in caves may have caused emergence (regression) along and might correspond to a warm, dry interval. the Virginia coast even as rising sea levels were Variations in cave passageways include inner inundating coastal landscapes along the south channels cut into a cave floor, horizontal galleries Atlantic coast (Bloom, 1983a, p.2I9; Clark, 1981). not related to bedrock variations, potholes, and Mastodons and mammoths roamed the broad knickpoints. Such features often result from windswept coastal plain during glacials. River changes in water levels, solution rates, water valleys, lagoons, estuaries, barrier islands, and temperature and load, and the like, and some have other features undoubtedly existed over extensive been linked to Pleistocene climate changes areas of the present continental shelf, but the (Sweeting, 1973, p. 155-15?). reworking of deposits by wave aetivity during Unfortunately, little concrete information of ensuing transgressions has extensively altered the above type is available from Virginia. But such features. The cold Labrador Current from in view of the information reviewed above and the North Atlantic dominated the coastal waters the abundance of caves in the state, the potential of Virginia during glacials creating a climate of cave studies for yielding valuable paleoenvi- more like present-day Nova Scotia than Virginia ronmental information should not be underesti- Beach, with sea temperatures some 6-8'C eolder mated. in August than at present (Imbrie and others 1983). COASTAL AND EOLIAN PROCESSES Wind is an important agent in coastal areas today, especially on barrier islands, which total A detailed examination of the intensively stud- some 69,000 acres (28,000 heetares) in Virginia. ied coastal regions of Virginia and adjaeent areas Data on eolain effects during the Pleistocene in is beyond the scope of this paper and no attempt Virginia are limited. Small sand dune areas were will be made to review the many recent contri- apparently present on parts of the Virginia butions and controversies relating to the Quater- Coastal Plain during the dry, windy glacials, and nary geology of this region.. Rather, a very brief thin loess deposits along with ventifacts are scat- overview of major coastal zone events of impor- tered about. Among the beach ridges, spits, flats tance to surficial processes is offered to complete and other abandoned coastal zone features of our perspective of Quaternary geomorphology in Virginia are the shallow, circular-to-elliptical Virginia. depressions called "Carolina Bays", most of which Fluctuating sea levels have produeed complex are now believed to be Pleistocene relics of eolian sequences of transgressive and regressive de- origin (Pettry and others, 1979). posits and associated landforms in the coastal lowlands of Virginia. Quaternary sediments and SUGGESTIONS FOR THE QUATERNARY various relic terraces or "flats", scarps, strand- FOSSIL HUNTER lines, bars and related features in the coastal zone provide evidence of past sea levels (Bloom, 1983a, Because much of the interest in Quaternary 1983b; Farrell, 1979; Johnson, and others 1982; geology in inland Virginia has centered about Pilkey and Evans, 1981). Superimposed upon the vertebrate fossils, the location of potentially pro- many sea-level responses to the growth and ab- ductive Quaternary fossil sites merits special lation of glaciers were slow tectonic uplift, sea- attention. Even if a promising site or deposit ward tilting, and glacio-isostatic influences. contains no vertebrates, it is likely to yield other Sea levels were quite close to those of the useful information about the paleoenvironment, present during past interglacials, but when ice such as spores, pollen, invertebrate fossils, buried sheets were at their maximum extent, sea level soils, and valuable stratigraphic and sedimento- was some 395 feet (120 m) lower than at present. logical data. Some general suggestions follow. 18 VIRGINIA DIVISION OF MINERAL RESOURCES

(1) Alluvium and colluvium: Although most found, their lacustrine sediments provide an ideal subaerial deposits of this type lack fossils, rapid site for fossil preservation. The well-known Salt- burial or special conditions may allow preserva- ville Valley (Figure 10) is an excellent example tion of organisms, including some megafauna. of special conditions giving rise to a classic fossil The stratigraphy of any thick depositional se- locality. Saltville lies along a typical strike valley quence, if exposed, may provide valuable infor- eroded upon the nonresistant Maccrady Forma- mation about past processes, even if no fossils are tion of Mississippian age. Evidence indicates that present. Unfortunately, as noted previously, con- a sizable river flowed through the Saltville sec- tinuing denudation has tended to limit major tion of the valley until about 14,500 yr B.P. when accumulations which contain sequences of past piracy of the old "Saltville River" occurred glacial and interglacial deposits. The base level (McDonald and Bartlett, 1983; Stocking and oth- for most Wisconsinan alluvium was probably very ers 1979). Shallow lakes have existed at Saltville close to current floodplain levels and distinguish- almost continuously since then. The origin of the ing Holocene from Pleistocene deposits may be lake depressions is uncertain. Three possibilities difficult. The best preservation of old alluvium are (a) subsidence caused by dissolution of e- tends to be on limestone or dolomite. Because of vaporite beds in the underlying Maccrady For- their solubility, most drainage is internal (Figure mation; (b) mass wasting and colluvium build- 9), greatly reducing surface runoff and erosion up in the water gap immediately north of the of surficial materials (Hack, 1965; Houser, 1980). valley; (c) alluvial deposition, perhaps during a Where carbonate bedrock lies close to the base flood, along the North Fork of the Holston River of mountainous areas, thick colluvial and alluvial which occupies the valley immediately north of deposits are likely to be preserved, but vertebrate the water (now a wind) gap. The presence of the fossil finds in such deposits are rare. In eastern lake, plus both fresh and salt water springs, Virginia, low-lying marshy environments along attracted many late Quaternary mammals-and coasts and river valleys may have entrapped and later, humans-to the Saltville Valley. preserved numerous animals during glacials, but (3) Caves and fissures: Most residuum and sap- these locations are now beneath the Atlantic rolite are unproductive and uninformative. How- Ocean. ever, where deep fissures have been weathered into (2) Depressions: Natural lake basins in ungla- bedrock (usually limestone or dolomite), filled with ciated areas such as Virginia are rare, but where residuum and exposed in river canyons, quarries

Figure 9. Small, flat-floored karst depression (polje) near Dublin, Pulaski County. A spring- Figure 10. View north across Saltville Valley, fed stream flows into a swallow hole near its Smyth County, Virginia's most famous verte- center and exposes several meters of fill. Old brate fossil locality. Stream piracy and creation topographic maps show this area as a lake. Such of a shallow lake made this an ideal area for the a location might yield useful information about preservation of late Quaternary fossils and pre- late Quaternary climates and life forms. (Pho- historic human artifacts. (Photograph taken June tograph taken April 10, 1980) 25. 1984) 18 VIRGINIA DIVISION OF MINERAL RESOURCES

(1) Alluvium and colluvium: Although most found, their lacustrine sediments provide an ideal sulcaerial deposits of this type lack fossils, rapid site for fossil preservation. The well-known Salt- burial or special conditions may allow preserva- ville Valley (Figure 10) is an excellent example tion of organisms, including some megafauna. of special conditions giving rise to a classic fossil The stratigraphy of any thick depositional se- locality. Saltville lies along a typical strike valley quence, if exposed, may provide valuable infor- eroded upon the nonresistant Maccrady Forma- mzltion about past processes, even if no fossils are tion of Mississippian age. Evidence indicates that pr(3sent. Unfortunately, as noted previously, con- a sizable river flowed through the Saltville sec- tinuing denudation has tended to limit major tion of the valley until about 14,500 yr B.P. when accumulations which contain sequences of past piracy of the old "Saltville River" occurred glacial and interglacial deposits. The base level (McDonald and Bartlett, 1983; Stocking and oth- for most Wisconsinan alluvium was probably very ers 1979). Shallow lakes have existed at Saltville clc,se to current floodplain levels and distinguish- almost continuously since then. The origin of the ing Holocene from Pleistocene deposits may be lake depressions is uncertain. Three possibilities difficult. The best preservation of old alluvium are (a) subsidence caused by dissolution of e- tends to be on limestone or dolomite. Because of vaporite beds in the underlying Maccrady For- thr:ir solubility, most drainage is internal (Figure mation; (b) mass wasting and colluvium build- 9), greatly reducing surface runoff and erosion up in the water gap immediately north of the of surficial materials (Hack, 1965; Houser, 1980). valley; (c) alluvial deposition, perhaps during a Where carbonate bedrock lies close to the base flood, along the North Fork of the Holston River of mountainous areas, thick colluvial and alluvial which occupies the valley immediately north of deposits are likely to be preserved, but vertebrate the water (now a wind) gap. The presence of the fosrsil finds in such deposits are rare. In eastern lake, plus both fresh and salt water springs, Virginia, low-lying marshy environments along attracted many late Quaternary mammals-and coasts and river valleys may have entrapped and later, humans-to the Saltville Valley. preserved numerous animals during glacials, but (3) Caves and fissures: Most residuum and sap- these locations are now beneath the Atlantic rolite are unproductive and uninformative. How- Ocean. ever, where deep fissures have been weathered into (2) Depressions: Natural lake basins in ungla- bedrock (usually limestone or dolomite), filled with ciated areas such as Virginia are rare, but where residuum and exposed in river canyons, quarries

Figure 9. Small, flat-floored karst depression (polje) near Dublin, Pulaski County. A spring- Figure 10. View north across Saltville Valley, fed stream flows into a swallow hole near its Smyth County, Virginia's most famous verte- center and exposes several meters of fill. Old brate fossil locality. Stream piracy and creation topographic maps show this area as a lake. Such of a shallow lake made this an ideal area for the a llocation might yield useful information about preservation of late Quaternary fossils and pre- la1;e Quaternary climates and life forms. (Pho- historic human artifacts. (Photograph taken June tograph taken April 10, 1980) 25. 1984) PUBLICATION 75 19

and so forth, they provide a likely place for fossil fluence slope development and relate to erosional finds. Similar conditions exist among the deep and mass wasting processes is needed. Studies clefts of some large felsenmeer or rock city areas. of recent variations in fluvial processes and drain- Karst shafts (Figure 11) and collapsed or filled- age development such as stream captures, strath in cave entrances, as well as the caves themselves, and terrace formation, and lithologic and struc- represent other places worth investigating. Tufa tural influences should also prove productive. and other mineral deposits are common near Caves are in special need of serious scientific springs and along certain rivers in Virginia (Fig- study. ure 8; Hubbard and others, this volume) and may As understanding of modern surficial processes yield useful data. and landforms expands, we will be better able to assess the impacts of past climatic changes, SUMMARY recognize evidence of such changes, and apply our knowledge to a plethora of related fields of Although the Quaternary paleoenvironments interest from Quaternary fossil sites to modern did not alter the terrain of Virginia as drastically land use planning. as they did in many regions, their imprints are clearly present in some areas, especially at high elevations and in coastal areas. It is also apparent that currently active processes are capable of REFERENCES producing or maintaining many features once thought to be of relic periglacial origin. The Barry, R. G., 1983, Late-Pleistocene climatology, Quaternary climatic changes influenced the rate, i,n Wright, H. E., Jr., ed., Late-Quaternary areas of operation, and relative effectiveness of environments of the United States, vol. 1: surficial processes but did not alter the basic Minneapolis, Univ. Minnesota Press p. 390- denudational pattern that has prevailed for tens 407. of millions of years in Virginia. Begin, Z. B. and Schumm, S. A., 1984, Grada- With the possible exception of the coastal zone, tional thresholds and landform singularity: modern geomorphological research in Virginia Significance for Quaternary studies: Quat. is still in its infancy. More precise knowledge of Res., vol. 21, p.267-274. specific weathering processes and how they in- Berkland, J. O. and Raymond, L. A., 1973, Pleis- tocene glaciation in the Blue Ridge Moun- tains, North Carolina: Science, vol. 181, p. 651-653. Bloom, A. L., 1983a, Sea level and coastal morphology of the United States through the Late Wisconsin glacial maximum, iz Wright, H. E., Jr., ed., Late-Quaternary environments of the United States, vol. 1: Minneapolis, Univ. Minnesota Press, p. 2L5-229. 1983b, Sea level and coastal changes, 'in Wright, H. E., Jr., ed., Late-Quaternary En- -, vironments of the United States, vol. 2: Min- neapolis, Univ. Minnesota Press, p.42-51. Bowen, D. Q., 1978, Quaternary geology: Oxford, U. K., Pergamon Press, 221p. Clark, G. M., 1968, Sorted patterned ground: New Appalachian localities south of the glacial Figure 11. Pig Hole, Giles County. This 1l9-foot- border: Science, vol. 161, p. 355-356. deep karst shaft in Middle Ordovician limestone L984, Periglacial features in the high Ap- provides access to some 2460 feet (750 m) of cave palachians south of the glacial border: Cur- passages. Pits, crevasses, and caves have good -, rent Res. Univ. Tenn., vol. 1, p. 85-87. entrapment potential and commonly contain hu- Clark, J. A., 1981, Comment (on "Late Wisconsin man artifacts, fossils, and geomorphic evidence and Holocene tectonic stability of the United of climatic change. (Photograph taken August 19, States mid-Atlantic coastal region"): Geology, 1981) vol. 9, p. 438. PUBLICATION 75 19

and so forth, they provide a likely place for fossil fluence slope development and relate to erosional finds. Similar conditions exist among the deep and mass wasting processes is needed. Studies clefts of some large felsenmeer or rock city areas. of recent variations in fluvial processes and drain- Karst shafts (Figure 11) and collapsed or filled- age development such as stream captures, strath in cave entrances, as well as the caves themselves, and terrace formation, and lithologic and struc- represent other places worth investigating. Tufa tural influences should also prove productive. and other mineral deposits are common near Caves are in special need of serious scientific springs and along certain rivers in Virginia (Fig- study. ure 8; Hubbard and others, this volume) and may As understanding of modern surficial processes yield useful data. and landforms expands, we will be better able to assess the impacts of past climatic changes, SUMMARY recognize evidence of such changes, and apply our knowledge to a plethora of related fields of Although the Quaternary paleoenvironments interest from Quaternary fossil sites to modern did not alter the terrain of Virginia as drastically land use planning. as they did in many regions, their imprints are clearly present in some areas, especially at high elevations and in coastal areas. It is also apparent that currently active processes are capable of REFERENCES producing or maintaining many features once thought to be of relic periglacial origin. The Barry, R. G., 1983, Late-Pleistocene climatology, Quaternary climatic changes influenced the rate, in Wright, H. E., Jr., ed., Late-Quaternary areas of operation, and relative effectiveness of environments of the United States, vol. 1: surficial processes but did not alter the basic Minneapolis, Univ. Minnesota Press p. 390- denudational pattern that has prevailed for tens 407. of millions of years in Virginia. Begin, Z. B. and Schumm, S. A., 1984, Grada- With the possible exception of the coastal zone, tional thresholds and landform singularity: modern geomorphological research in Virginia Significance for Quaternary studies: Quat. is still in its infancy. More precise knowledge of Res., vol. 21, p. 267 -27 4. specific weathering processes and how they in- Berkland, J. O. and Raymond, L. A., 1973, Pleis- tocene glaciation in the Blue Ridge Moun- tains, North Carolina: Science, vol. 181, p. 651-653. Bloom, A. L., 1983a, Sea level and coastal morphology of the United States through the Late Wisconsin glacial maximum, in Wright, H. E., Jr., ed., Late-Quaternary environments of the United States, vol. 1: Minneapolis, Univ. Minnesota Press, p. 215-229. 1983b, Sea level and coastal changes, in Wright, H. E., Jr., ed., Late-Quaternary En- -, vironments of the United States, vol. 2: Min- neapolis, Univ. Minnesota Press, p.42-51. Bowen, D. Q., 1978, Quaternary geology: Oxford, U. K., Pergamon Press, 22I p. Clark, G. M., 1968, Sorted patterned ground: New Appalachian localities south of the glacial Figure 11. Pig Hole, Giles County. This 1l9-foot- border: Science, vol. 161, p.355-356. deep karst shaft in Middle Ordovician limestone 1984, Periglacial features in the high Ap- provides access to some 2460 feet (750 m) of cave palachians south of the glacial border: Cur- passages. Pits, crevasses, and caves have good -, rent Res. Univ. Tenn., vol. 1, p. 85-87. entrapment potential and commonly contain hu- Clark, J. A., 1981, Comment (on "Late Wisconsin man artifacts, fossils, and geomorphic evidence and Holocene tectonic stability of the United of climatic change. (Photograph taken August 19, States mid-Atlantic coastal region"): Geology, 1981) vol. 9. p. 438. 20 VIRGINIA DIVISION OF MINERAL RESOURCES

Delcourt, H. R., 1979, Late Quaternary vegeta- 1980, Rock control and tectonism-their tional history of the eastern highland rim and importance in shapingthe Appalachian High- adjacent Cumberland Plateau of Tennessee: -, lands: U. S. Geol. Survey Prof. Paper 1126- Ecol. Monogr., vol. 49,p.255-280. B, p. B1-817. Delcourt, H. R. and Delcourt, P. A., 1984, Ice age !982, Physiographic division and differen- haven for hardwoods: Nat. Hist., vol. 93, no. tial uplift in the Piedmont and Blue Ridge: 9, p.22-28. -, U. S. Geol. Survey Prof. Paper 1265, 48 p. Delcourt, P. A., 1980, Quaternary alluvial ter- Hack, J. T., and Goodlett, J. C., 1960, Geomor- races of the Little Tennessee River Valley, phology and forest ecology of a mountain east Tennessee, im Chapman, J., ed., The 1979 region in the central Appalachians: U. S. archaeological and geological investigations Geol. Survey Prof. PaPer 347,66P- in the Tellico Reservoir: Univ. Tenn. Dept. Hennig, G. J., Grun, R., and Brunnaeker, K., Anthropology Rept. Investig. 29, p. 110-121. 1983, Speleothems, travertines, and paleoc- Denny, C. S., 1956, Surficial geology and geomor- limates: Quat. Res., vol. 20, P. L-29- phology of Potter County, Pennsylvania: U. Hess, J. W. and Harmon, R. S., 1981, Geochro- S. Geol. Survey Bull. 51-C, p. 95-104. nology of speleothems from the Flint Ridge Dunne, T., and Leopold, L. 8., 1978, Water in Mammoth Cave system, Kentucky, U.S.A., iza environmental planning: San Francisco, Beck, B. F., ed., Eighth International Con- Freeman and Co., 818 p. gress of Speleology Proceedings, vol. 2: Amer- Embleton, C., 1979, Nival processes, in Em- icus, Ga., Georgia Southwestern College. p. bleton, C. and Thornes, J., eds., Process in 433-436. geomorphology: New York, Halsted Press, p. Hill, C. A., 1976, Cave minerals: Austin, Tx., 307-324. Speleo Press, 137 p. Eyles, N., 1983, Glacial geology: An introduction Houser, 8., 1980, Erosional history of the New for engineers and earth scientists: Oxford, River, southern Appalachians, Virginia: U.K., Pergamon Press,409 p. Ph.D. dissertation, Virginia Polytechnical In- Farrell, K. M., 1979, Stratigraphy and geomor- stitute and State Univ., Blacksburg, c.225 phology of late Pleistocene deposits of the p. northern neck between the Corrotoman River Hubbard, D. A., Jr., Giannini, W. F., and Lorah, and the Chesapeake Bay, Virginia (abs.): Va. M. M., 1985, Quaternary travertine-marl dep- Jour. Sci, vol.30, no.2,p.79. osits of Virignia: this volume. Flint, R. F., 1971, Glacial and Quaternary geol- Hupp, C. R., 1983, Geo-botanical evidence of late ogy: New York, Wiley and Sons, 892 p. Quaternary mass wasting in block field areas 1976, Physical evidence of Quaternary cli- of Virginia: Earth Surface Processes and matic change: Quat. Res., vol. 6, p. 519-528. Landforms, vol. 8, p. 439-450. -,Gascoyne, M., 1981, A climate record of the Imbrie, J., and Imbrie, K. P., 1979, Ice ages: Yorkshire Dales for the last 300,000 years, solving the mystery: Hillside, N. J., Enslow inBeck, B. F., ed., Eighth International Con- Publ.,224p. gress of Speleology Proceedings, vol. 1: Amer- Imbrie, J., Mclntyre, A., and Moore, T. C., Jr., icus, Ga., Georgia Southwestern College, p. 1983, The ocean around North America at the 96-98. last glacial maximum, in H.8., Wright, Jr., Geyer, A. R. and Bolles, W.H., 1979, Outstanding H. E., ed., Late-Quaternary environments of scenic geological features of Pennsylvania: the United States, vol. 1: Minneapolis, Univ. Harrisburg, Pa. Geol. Survey, 508 p. Minnesota Press, p. 230-236. Glazek, J. and Harmon, R. S., 1981, Radiometric Jennings, J. N., 1981, Morphoelimatic control- dating of Polish cave speleothems: Current A tale of piss and wind or a case of the baby results, in Beck, B. F., ed., Eighth Interna- out with the bathwaler?, in Beck, B. F., ed., tional Congress of Speleology Proceedings, Eighth International Congress of Speleology vol. 2: Americus, Ga., Georgia Southwestern Proceedings, vol. 1: Americus, Ga., Georgia College, p.424-427. Southwestern College, p. 367-368. Hack, J. T., 1965, Geomorphology of the Shenan- Johnson, G. H., Peebles, P. C., and Berquist, C. doah Valley, Virginia and West Virginia, and R., 1982, Pleistocene depositional patterns of origin of the residual ore deposits: U. S. Geol. the southeastern Virginia and northeastern Survey Prof. Paper 484, 78 p. North Carolina coastal plain (abs.): Va. Jour. PUBLICATION 75 2l

Sci. vol.33, no.3, p. 15. 1980a, Observations on boulder streams in Judson, S. 1965, Quaternary processes in the the Valley and Ridge Province, Giles County, Atlantic Coastal Plain and Appalachian -, Virginia (abs.): Geol. Soc. America Abstraets Highlands, in Wright, H. 8., Jr., and Frey, with Programs, vol. 12 p.201-202. D. G., eds., The Quaternary of the United 1980b, The influence of boulder deposits on States: Princeton, N. J., Princeton Univ. the retreat of mountain slopes in the Valley Press, p. 133-135. -, and Ridge Province, Giles County, Virginia Evolution of Appalachian topography, (abs.): Geol. Soc. America Abstracts with iz Melhorn, W. N., and Flemal, R. C., eds., Programs, vol. 12, p. 484-485. -,t975,Theories of landform development: Bingham- 1981, Some observations on slope deposits ton, N. Y., Publ. Geomorph, p.29-44. in the vicinity of Grandfather Mountain, Kane, W. F., 1975, Rates of denudation for se- -, North Carolina: Southeastern Geol., vol.22, lected Yirginia basins (abs.): Va. Jour. Sci., p.209-222. vol.26, no.2, p.85. L982, Long term episodic deposition on Knox, J. C., 1983, Responses of river systems to mountain foot-slopes in the Blue Ridge Prov- Holocene climates, inWright, H.8., Jr., ed., -, ince of North Carolina: Evidence from rel- Late-Quaternary environments of the United ative-age dating: Southeastern Geol., vol. 23, States, vol. 2: Minneapolis, Univ. Minnesota p. L23-t28. Press, p.26-4L. Moore, G. W., 1956, Aragonite speleothems as Kochel, R. C., and Johnson, R. A., 1984, Geomor- indicators of paleo-temperature: American phology and sedimentology of humid-temper- Jour. Sci., vol.254, p.746-753. ate alluvial fans, central Virginia, iz Koster, Mylroie, J. E., 1979, Glaciation and karst geo- E. H., and Steel, R. J., eds., Sedimentology morphology in Schoharie County, New York of gravels and conglomerates: Canadian Soc. (abs.): Natl. Spel. Soc. Bull., vol. 41, p. 112- Pet. Geol. Memoir 10, p. 109-122. 113. Kochel, R. C., Johnson, R. A., and Valastro, S., Pettry, D. E., Scott, J. H., Jr., and Bliley, D. J., Jr., 1982, Repeated episodes of Holocene de- 1979, Distribution and nature of Carolina bris avalanching in central Virginia (abs.): Bays on the eastern shore of Virginia: Va. Geol. Soc. America Abstracts with Programs, Jour. Sci., vol.30, no. 1, p.3-9. vol. 14, p.31. Pewe, T. L., 1983, The periglacial environment Latham, A. G., 1981, Magnetostratigraphy from in North America during Wisconsin time, in speleothems: Establishment and applica- Wright, H. 8., Jr., Late-Quaternary environ- tions, inBeek, B. F., ed., Eighth International ments of the United States, vol. 1, Minnea- Congress of Speleology Proceedings, vol. 1: polis, Univ. Minnesota Press, p. 157-189. Americus, Ga., Georgia Southwestern Col- Pilkey, O. H., and Evans, M., 1981, Rising seas, lege, p. 358-361. shiftin g sho r es, in J ackson, T. C., and Reische, Maxwell, J. A., and Davis, M. 8., 1972, Pollen D., eds., Coast alert: San Francisco, Friends evidence of Pleistocene and Holocene vegeta- of the Earth, p. 13-47. tion on the Allegheny Plateau, Maryland: Potter, N., Jr., and Moss, J. H., 1968, Origin of Quat. Res., vol. 2, p. 506-530. the Blue Rocks block field and adjacent de- McDonald, J. N., and Bartlett, C. S., Jr., 1983, posits, Berks County, Pennsylvania: Geol. Soc. An assoeiated musk ox skeleton from Salt- America Bull., vol. 79, p.255-262. ville, Virginia: Jour. Vert. Paleontol, vol. 2, Psilovikos, A., and Van Houten, F. 8., 1982, p.453-470. Ringing Rocks barren rock field, east-central Meade, R. H., 1969, Errors in using modern Pennsylvania: Sed. Geol., vol.32, p.233-243. streamload data to estimate natural rates of Radbruch-Hall, D. H., and others, 1976, Prelim- denudation: Geol. Soc. America Bull., vol.80, inary landslide overview map of the conter- p. L265-L274. minous United States: U. S. Geol. Survey Map Michalek, D. D., 1968, Fanlike features and re- MF-771, scale 1:7,500,000, with text. lated periglacial phenomena of the southern Smith, H. T. U., L949, Pleistocene climatic Blue Ridge: Ph.D. dissertation, Univ. North ehanges in nonglaciated areas: Eolian phen- Carolina, 198 p. omena, frost action, and stream terracing: Mills, H. H., 1978, Some observations on "alluv- Geol. Soc. America Bull., vol.60, p. 1485-1516. ial" fans in the southern Appalachians (abs.): Stocking, J., Thompson, J., and Bartlett, C. S., Tenn. Tech. Jour., vol. 13, p.72. Jr., L979, Preliminary reports on the Saltville 22 VIRGINIA DIVISION OF MINERAL RESOURCES

Pleistocene bone beds: Regional geology and ern states: Geol. Soc. America Bull., vol. 39, geomorphology (abs.): Va. Jour. Sci., vol. 30, p. 941-953. no. 2, p. 81. White, S. E., 1976, Is frost action really only Sweeting, M. M., 1973, Karst landforms: New hydration shattering? - A review: Arctic York, Columbia Univ. Press, 362 p. Alpine Res., vol. 8, no. 1, p. 1-6. Washburn, A. L., 1980, Permafrost features as White, W.8., 1979, Water balance, mass balance, evidence of climatic change: Earth-Sci. Rev., and time scales for cave system development: vol. 15, p.327-402. Natl. Spel. Soc. Bull., vol. 41, p. 115. Watts, W. A., 1979, Late Quaternary vegetation Williams, G. P., and Guy, H. P., 1973, Erosional of central Appalachia and the New Jersey and depositional aspects of Hurricane Ca- Coastal Plain: Ecol. Monogr., vol. 49, p. 427- mille in Virginia, 1969: U. S. Geol. Survey 469. Prof. Paper 804, 80 p. 1983, Vegetational history of the eastern Wright, H. E., Jr., 1981, Vegetation east of the United States 25,000 to 10,000 years ago, in Rocky Mountains 18,000 years ago: Quat. -, Wright, H. E., Jr., ed., Late-Quaternary en- Res., vol. 15, p. 113-125. vironments of the United States, vol. 1: Min- , €d., 1983, Late-Quaternary environments of neapolis, Univ. Minnesota Press, p.294-310. the United States (Two volumes): Minneapo- Wells, S. G., and Gutierrez, A. A., 1981, Geomor- lis, Univ. Minnesota Press, 407 p. and277 p. phic adjustments of fluvial systems to ground- Yonge, C. J., 198L, Fluid inclusions in speleo- water hydrology in semiarid and humid karst, thems as paleoclimate indicators, in Beck, B. inBeck, B. F., ed., Eighth International Con- F., ed., Eighth International Congress of Spe- gress of Speleology Proceedings, vol. 1: Amer- leology Proeeedings, vol. 1: Americus, Ga., icus, Ga., Georgia Southwestern College, p. Georgia Southwestern College, p. 301-304. 2t6. Young, A., 1969, Present rate of land erosion: Wentworth, C. K., 1928, Striated cobbles in south- Nature, vol.224, p. 851-852. PUBLICATION 75 23

LATE QUATERNARY VEGETATIONAL CHANGE IN THE

CENTRAL ATLANTIC STATES

}Jazel R. Delcourtr PauI A. Delcourtz

CONTENTS Page

Abstract 23 Introduction ...... 24 Vegetation of the Late wisconsin full glacial interval (29,000 to 16,b00 yr B.p.) 24 vegetation of the Late \{isconsin late glacial interval (16,b00 to 12,b00 yr B.p) 27 Vegetation of Holocene (12,b00 the interval yr B.p. to the present) ...... 29 Late Quaternary gradients in plant and animal communities ...... 31 Directions for future research 34 Acknowledgments 34 References 34 ILLUSTRATIONS

Figure 1. Map of late Quaternary sites discussed in text ...... 24 2. summary diagram of the arboreal pollen and plant-macrofossil sequenee from the Saltville Valley, Smyth County, Virginia 28 TABLE

Major late Quaternary palynological sites in the central Atlantic states 25

ABSTRACT munities indicates that vegetational disequilibrium and disharmonious faunal assemblages occurred We review the changes in vegetation, climate, from 13,000 to 10,000 yr B.P. During that time, and geomorphic regimes in the central Atlantic climatic amelioration and dynamic landscape states region (36o to 41"N, 74" tn 84"W) over the changes occurred with the transition from colluvial past 23,000 years sinee the onset of the last full to alluvial geomorphic regimes. The Valley and glacial interval. Examination of gradients in late Ridge Physiographie Province served as a major Pleistocene and Holocene plant and animal com- migrational corridor for plant species at the transition from late Pleistocene to Holocene conditions. Holocene vegetation has shifted from cool-temperate to warm-temperate forests, reflecting rProgram climatic for Quaternary Studies ofthe Southeast- warming, continued northward migrations of ern United States, Graduate Program in Ecology warm-temperate plant taxa, and, in the Atlantic and Department of Botany, University of Tennes- Coastal Plain, the rise in sea level. Future research see, Knoxville, Tennessee 3?996. priorities include establishing a network of radiocarbon-dated late pollen plant- zProgram Quaternary and for Quaternary Studies ofthe Southeast- macrofossil sites in the Allegheny Plateau, Valley ern United States, Graduate Program in Ecology and Ridge, Blue Ridge, Piedmont, and Atlantic and Department of Geological Scienees, University Coastal Plain physiographic pt'ovinces of Virginia of Tennessee, Knoxville, Tennessee 9?g96. and sumounding states. 24 VIRGINIA DIVISION OF MINERAL RESOURCES

INTRODUCTION

Studies of late Quaternary vegetational history in the unglaciated southeastern United States were pioneered in Virginia, with one of the earliest studies thatof Lewis and Cocke (1929) inthe Dismal Swamp (Figure 1). Despite the early emphasis upon understanding the late Quaternary history of vegetation and climate along the Atlantie seaboard, relatively few radiocarbon-dated, late Quaternary palynological sites have been studied within the state of Virginia (Figure 1; Table). In this paper, we review the available literature g4ow 74"W concerning late Pleistocene and Holocene plant- fossil studies in Virginia and the surrounding Figure 1. Map of late Quaternary sites discussed region of the central Atlantic states. We restrict in text. Letter codes for the palynological sites are our review to the time interval within which ra- referenced in the Table. Four faunal sites include: diocarbon analysis allows vegetational reconstruc- NP4, New Paris No. 4, PA (Guilday and others, tions to be made based upon an absolute time scale, 1964); CC, Clark's Cave, VA (Guilday and others, with the greatest emphasis upon biotic changes 1977); SA, Saltville Valley, VA (Ray and others, since the onset of the last full glacial period, from f967); and BB, Baker Bluff Cave, TN (Guilday and 23,000 yr B.P. to the present. We summarize vege- others, 1978). The dashed line corresponds to the tational ehanges for each of the five physiographic full glacial limit of the Laurentide Ice Sheet. Phy- provinces represented in the state of Virginia, siographic regions discussed in the text are num- including the Allegheny Plateau, the Valley and bered as follows: (1) Allegheny Plateau; (2) Valley Ridge, the Blue Ridge, the Piedmont, and the and Ridge; (3) Blue Ridge; (4) Piedmont; and (5) Atlantic Coastal Plain. In order to characterize Atlantic Coastal Plain. changes in vegetation and climate within the context of these broad physiographic divisions, we drawupon late Quaternary palynological sites from areas surrounding Virginia and located between Buckle's Bog, Maryland (Maxwell and Davis, 36o and 41o N latitude and between 74" and 84o L972), is located at2670 feet (814 m) elevation along W longitude (Figure 1). We discuss (1) the rela- a broad ridge on the Allegheny Plateau (Figure tionship of late Pleistocene vegetational and faunal 1, Table). The site is an alluvial glade, with 8.5 gradients to habitat availability and migrational feet (258 cm) of clay and peat representing the time corridors; (2) biotic responses to climatic amelio- interval from 19,000 yr B.P. to the present. During ration during the late glaeial and early Holocene the full glacial interval, the vegetation of the Al- intervals; and (3) patterns and processes in Hol- legheny Plateau was alpine tundra dominated by ocene vegetational development. Finally, we indi- sedges (Cyperaceae), grasses (Gramineae), and cate important future directions for research in other herbs. Low pollen-influx values (generally Quaternary vegetational history of Virginia and its less than 2000 grains'em-2'yr-1) indicate that trees surrounding region. were absent at elevations above approximately 1480 feet (450 m) (Maxwell and Davis, 1972). VEGETATION OF THE LATE WISCONSIN On the central Atlantic Coastal Plain, full glacial FULL GLACIAL INTERVAL pollen spectra are available from three areas. Near (23,ooo To 16,500 YR B.P.) Trenton, New Jersey (Figure 1) a pollen spectrum dating 16,?00 yr B.P. (Sirkin and others, 1970) is No pollen or plant-maerofossil sites yet studied dominated by pine (Pimn), with grass, bfteh(Bet- within the state of Virginia date from the Late ula), and spruce (Picea ), indieating taiga vegeta- Wisconsin full glacial interval, 23,000 to 16,500 yr tion (an open boreal woodland). Peat lenses buried B.P. We can infer full glacial vegetation for por- by sand dune sheets on the Delmarva Peninsula tions of Virginia from palynological evidence from of Maryland (Figure 1) date from 23,300 to 17'800 several sites nearby in Maryland, Delaware, and yr B.P. (Sirkin and others, 1977; Denny and others, North Carolina that have radiocarbon dates from 19?9). These peat lenses contain a pollen a"ssem- this time interval. blage consisting of pine, spruce, northern shrubs, PUBLICATION ?5 25

.29= z tra EE:E € +$? e ; E : e. g ; Et H ?? es;sseec? ? Ensse fl gg g3#sEaEsF S SeFEa atr Fl z -irng-4 trl>cdX,r4lcd8B.= H r€Ee t J.q>6d! J XP 6 = $ .gsii F X

t *4 - t-bog bo Fl= r *i ! FJ <1 0 . o gdg9 6g !He*H ! $ Ee F ** -o X =z o.Ei5 I qn * 'E:;e€ oFA&=6 $E 3 ar sE n r i CN ts 0r> n -:: E =9 jrE E 'o3tr6 d +) ^ c€ EE SEES E fr fi}3 +)a = c') +t +) gH I a EE s .Ec s s-e i r'1 :: S Eb S S A €! 2 ! t E, e!? e o Ecis +E+c €cis so t 33 +) F EE EEEE sESE $EEE tr ri 4) 'E.o.o co de 6O i c, trtr- 66

b0 .= s R roi[o6a@C\OIo s tH$8e eOO0r(0 >. i2$3s s n c€

! 8 Eq € a cqoorr/a cB mmM .'J tr (l) +) 6t sB _d oX +) c! E Fx tr r'rd:Fj .Fo r d d d 3 €e AsE {E?.g ?F"" H; cq d t F e -t 2= sF ! * E" * E E E E< H AF E €s s 5 E n 26 VIRGINIA DIVISION OF MINERAL RESOURCES

s e g zg za F. s s; sgii- ;s $g sF ? €*5e s E-€a6.aa Q = +trE e.d;e e Fl F.Egg 'Fs .eE HE'F H Hs Fb He 3 h= h aB gio Eeo-8535 5 F F= F E F F= fiEdF F A 6 F U) FJ .dE z € E EE z .Fbo € eE € bes E 'E€#e € .E rrl H H € FE H FE X E E # Fl = = a 'J 5 i >; t ;Ht F t:at n

€€€€ ;6ce €E gT 3 F E gT ?sg€ E E 3E i E ir rE !F f6rE*s t tE €; € gF E sars .E E :EE H HE"T s: Ea s n"sE € i EES € €frEE EE H H s s ss $ € €r $ ; ; ;.,# frfr E* E * z € H i€ € E i Ea*.,*n E-sE

q E isi iEsi€i i ?EisiE 3siE F 5E,s a5E5Ea fi A.eSEAA SrErf;u

,; k OC\

h?. H S SX I R

frt >-A () H H 3> H H E E # Ei3

F, d * V ? 3 i * d 3 F g; ,f € fr : t.g* : ; E E F Fq ; 3 E € t s g t t i i s s s sg E fr # E E$ E E PUBLICATION 75 27

and upland herbs indicative of a cold, dry taiga and shrubs of crowberry (Empetrum ef.. nigrurn) environment in which ephemeral bogs and thaw and shrub birch(Betula glanduloso) persisted until lakes formed and were subsequently filled by drift- approximately 12,500 yr 8.P., when an open, boreal ing sand. spruce woodland established (Watts, 1979). To the Farther south on the Coastal Plain of northeast- south, along the unglaciated portion of the Valley ern North Carolina, Roekyhock Bay (Figure l, and Ridge, at Crider's Pond, Pennsylvania (Figure Table) reeords vegetational changes dating from 1; 950 feet,290 m, elevation), late glacial vegetation 27,700 yr B.P. to the present (Whitehead, 1981). was charaeterized by a boreal forest dominated by Duringthe full glacial interval, pollen assemblages spruee, with minor amounts of jack pine but no at Rockyhock Bay were dominated by ?5 percent evidence of tundra herbs (Watts, 1979). Between to 90 percent jack pine (northern Diploxylon Pirnr^s), 13,000 and 12,500 yr B.P., balsam fft (Abies bal- with spruce and a mixture of herbs that indicate somea), tree birch, alder, and willow (Solzr) es- an unstable landscape with patches of taiga vege- tablished locally within the boreal forest. tation interspersed with the exposed surfaces of In the Valley and Ridge Province of Virginia, active sand dunes (Whitehead, 1981). several sites provide evidence of late glacial vege- Although no full glacial pollen sites are known tation. Hack Pond, a karst pond located at 1480 from the Valley and Ridge, Blue Ridge, or Pied- feet (451 m) elevation at 38o N latitude in the mont provinces in the region, we infer that alpine Shenandoah Valley of northern Virginia (Figure tundra was restricted to highest elevations of the L), dates from greater than 13,000 yr B.P. to the Allegheny Plateau and the Blue Ridge Mountains, present (Craig, 1969). During the late glacial in- with a relatively open boreal forest extending east- terval, the vegetation was a elosed spruce-pine ward at lower elevations to the Coastal Plain (Del- forest. After 12,700 yr B.P., white pine (Pinus court and Deleourt, 198f). stuoAus), b i rch, alder, and oak ( 8a er ws) establ i shed within the forests (Craig, 1969). VEGETATION OF THE LATE WISCONSIN The oldest radiocarbon-dated, continuous record LATE GLACIAL INTERVAL of late Quaternary sediment accumulation and (16,500 TO 12,500 YR B.P.) vegetational change available from the Valley and Ridge Provinee of Virginia is represented by the On the Allegheny Plateau of Maryland, alpine section from the Saltville Valley (Figures 1, 2; tundra persisted throughout the late glacial inter- Table). Alluvial glades at 1722 feet (525 m) ele- val and was replaced byopen sprueewoodland after vation within the Saltville Valley date from 15,500 12,700 yr B.P. (Maxwell and Davis, 1972). Farther to 2000 yr B.P. During the late glacial interval, south, at Cranberry Glades, an alluvial glade lo- from 15,500 to 12,500 yr B.P., the vegetation was cated at 3375 feet (1029 m) elevation on the Al- a species-rich mosaic of boreal coniferous and hard- legheny Plateau of West Virginia (Figure l), alpine woods forests. Sandstone ridges in the uplands were tundra persisted until 12,200 yr B.P. (Watts, 1g?g). oecupied by jack pine, aspen (Popu,lus balsomifera Six sites in the Valley and Ridge Province date and P. tremulnides type), tree birch, and oak. Ash from the late glacial interval (Figure 1). The Tan- (Franinus). maples (Acer pensglnanimm and A. nersville, Pennsylvania, site (Figure 1) is a kettle- sa,ccharam), beech (Fag,Ls grarLd,ifolia), hickory bloek depression formed during the retreat of the (Carya), elm (Ulmas), and cherry (Pruntus) oecu- Laurentide Ice Sheet. From about 14,000 yr B.P. pied mesic slopes. Within the poorly-drained alluv- to 13,300 yr B.P., the newly deglaciated landscape ial bottomlands, dominant trees included red was poorly drained, with extensive wetlands dom- spruee (Pdceantbms) and balsam fir with tamarack inated by sedges. At 13,300 yr B.P., trees of spruee, (Larin) (Figure 2). Open glades were locally dom- aspen (Puyulus ef.. tremuloides), eedar (Juniperus inated by boreal shrubs, grasses, and sedges, with type), and shrubs of green alder (Alrws erispa) a rich assemblage of herbs indicative of both boreal invaded and stabilized the glacial moraines (Watts, woodlands and marshes. 1979). Between 13,300 and 12,500 yr B.P., the Late glacial vegetation on the Piedmont is re- vegetation changed in composition from taiga to presented only at Marsh, Pennsylvania (Figure 1). closed boreal forest dominated by spruce, pine, fir The Marsh site, a small alluvial swale located (Abi,es), and birch. Similarly, at Longswamp, Penn- within 53 miles (85 km) of the ice margin at 41o sylvania(Figure l), inthe periglacial zone adjacent N latitude, dates to approximately 13,600 yr B.P. to the southernmost limit of glaeial advance, late (Martin, 1958). The late glacial pollen assemblage glacial grass-dominated tundra with arctic herbs from this site was dominated by nonarboreal pollen such as Dryas infegri.folia and Sagi,na ef. nnd,osa types, including primarily sedges and grasses, with 28 VIRGINIA DIVISION OF MINERAL RESOURCES

6i $$ '6 ! a F €E $ *$ $e * iI €*

.9 €>oi agFc') .^ 5 q.thct q) =(n >c€6!a oo +)o>a d3 +s9q) trtr ...-{FE +t qrB 2a 6E 35 9= =Ea-a 8E .Fto9 fi C€ HE F€ EcB EgFE. E+) Fts 6EaCh EX ';i5 ! fie -c) 6o Pq) c!3 * 9F' "r*g:"";e "- tr $"i rrro: .l z E" cB(,H€ kar t- frl E g\ts.E *t E j L *::g"HBI E:r:l E;: €E s; ;;; h5 z *6tsE rr-l E [] Ea ao g gl 63 /eo.O f g b s qs aio ir'6e oor.le $ B $ $ij 2d .$3 f&{ F PUBLICATION 75 29

substantial representation of pollen'of northern vegetation occurred during the late glaeial period. Diploxylon pines. Minor amounts of spruce, fir, To the north, tundra occurred near the ice margin birch, and willow were also present in the pollen at Tannersville Bog and Longswamp. Coniferous assemblage. The vegetation growing in the vicinity taiga occurred farther south at Crider's Pond, and of the Marsh site during the late glacial period has closed, species-poor boreal forest dominated by been interpreted as on the ecotone between taiga spruce oceurred at Hack Pond in the Shenandoah and tundra, the latter represented by herbs such Valley of Virginia. In southwestern Virginia, as Polem,oruium, Sarifraga cf.. opposit:ifolia, and closed, species-rich boreal forest of spruee, fir, and Polygonum bktortoides (Martin, 1958). pine with numerous cool-temperate species of hard- In a pollen spectrum dated 13,200 yr B.P. from wood trees and shrubs charaetefized the Saltville the Coastal Plain at Trenton, New Jersey (Figure Valley throughout the late glacial interval. 1), co-dominance of pine, spruce, and birch with The late glacial vegetation of the Piedmont Prov- grass reflected late glaeial taiga persisting approx- ince remains largely unknown, as only the Marsh imately 30 miles (50 km) south of the glacial limit record is available, identifying the position of the (Sirkin and others, 1970). In the central Delmarva tundra-taiga ecotone. Peninsula of Maryland and Delaware (Figure 1), The eentral Atlantic Coastal Plain, from central peat lenses interbedded within the Parsonsburg New Jersey to as far south as northeastern North Sand have been radiocarbon-dated between 16,300 Carolina, was characterized by species-poor taiga and 13,400 yr B.P. (Sirkin and others, 1977; Denny throughout the late glacial interval. and others, 1979). The pollen spectra doeument taiga dominated by pine, spruce, and birch, with VEGETATION OF THE HOLOCENE herbs in the grass, sedge, and aster families. INTERVAL A pollen spectrum from sediments at the mouth (12,500 YR B.P. TO THE PRESENT) of Chesapeake Bay, Virginia (Figure 1), dates from t5,280 yr B.P. (Harrison and others, 1965). The ' For convenience, we distinguish three major time pollen assemblage is interpreted as a relatively intervals of the Holocene: (1) the early Holoeene open, eoniferous boreal forest dominated by pine, interval, from 12,500 to 8500 yr B.P.; (2) the middle with spruce and nonarboreal pollen important Holocene interval, from 8500 to 4000 yr B.P.; and (Harrison and others, 1965). (3) the late Holocene interval, from 4000 yr B.P. On the Coastal Plain of North Carolina, approx- to the present (Delcourt and Delcourt, 1985). imately 310 miles (500 km) south of the glaeial After 12,500 yr 8.P., open spruce woodland was margin,late glacial pollen assemblages were char- replaced by elosed boreal forest on the Allegheny acterized by relatively low pollen influx of 2000 Plateau of Maryland. At Buckle's Bog, the early to 5300 grains.cm-2.yr-1, persistence of high per- Holocene boreal forest was composed of spruce, centages of jack pine pollen, and continuous rep- pine, fir, ash, hornbeam (CarpimtslOstrya type), resentation of fir pollen (\{hitehead, 1g8l). Spruce and oak (Maxwell and Davis, L972). By 10,500 yr pollen percentages increased after 16,000 yr B.P. B.P., white pine, birch, and hemlock (?szga,) es- and remained relatively high throughout the late tablished in the vicinity of Buckle's Bog within a glacial interval, and tamarack was present. The mixed conifer-northern hardwoods forest. vegetation of the central Atlantic Coastal Plain thus During the middle Holoeene, white pine was remained in pine-spruce taiga during the late gradually replaeed by hemlock and oak. Hemlock glacial interval (Whitehead, 1981). deereased in abundance approximately 3000 yr During the late glacial interval, a tundra zone 8.P., and plant succession following the hemlock generally occurred within 30 to 53 miles (50 to 85 decline resulted in late Holocene forest communi- km) of the ice margin (Martin, 1958; \[atts, 19?g; ties dominated by oak, chestnut (Costanea d,entata), Pewe, 1983), and alpine tundra extended at high beech, and hickory (Maxwell and Davis, L972). elevations southward along the Allegheny Plateau At Cranberry Glades on the Allegheny Plateau at least as far south as Cranberry Glades, West of West Virginia, mixed conifer-northern hard- Virginia (Watts, 1979). Alpine tundra may also woods forests were predominant from 12,200 to have persisted through the late glaeial interval at 4900 yr B.P. Early Holocene forests at Cranberry higher elevations of the Blue Ridge Provinee into Glades were dominated by hemlock, with birch, the southern Appalachian Mountains (Delcourt and oak, and hornbeam. After the collapse of hemlock Delcourt, 1981). populations at 4900 yr B.P., upland forests were In the Valley and Ridge Province, a pronounced dominated by oak, birch, and hiekory, with beech gradient in both physiognomy and composition of and chesinut. Within the last 4500 years, spruce 30 VIRGINIA DIVISION OF MINERAL RESOURCES

populations have reexpanded near the site (Watts, ing boreal conifers after 9000 yr B.P. (Watts, f 979). 1979). In the Shenandoah Valley of Virginia, the se- At Tannersville Bog, within the Valley and Ridge quence ofvegetational changes during the Holocene of Pennsylvania, closed boreal forest of pine, spruce, was similar to that occurring in southern Penn- and fir persisted from 12,500 to 11,500 yr B.P. sylvania. At Hack Pond, boreal coniferous forest (Watts, 1979). With the arrival of white pine and dominated by jack pine, white pine, and spruce paper birch (Betula papyrifera), an early-succes- persisted from 12,700 to 9000 yr B.P. (Craig, 1969). sional, mixed conifer-northern hardwoods forest From 9000 yr B.P. to the present, temperate forests replaced the boreal forest after 11,500 yr B.P. At have been dominated byoak, chestnut, and hickory. about 9800 yr B.P., oak and hemlock populations Hemlock was present locally until about 4500 yr expanded as white pine and paper birch dimin- B.P. White pine and southern Diploxylon pine ished. Hemlock declined at 4600 yr B.P., resulting (Pinus rtgida) expanded into upland sites during in a ehange in forest composition to dominance by the past 4500 years (Craig, 1969). pine, birch, oak, and chestnut. Within the last 4000 Potts Mountain Pond, at 3576 feet (1090 m) years, hemlock, spruce, and tamarack populations elevation in the Valley and Ridge of west-central increased locally (Watts, 1979). Although the com- Virginia (Figure 1), has a radiocarbon-dated pollen position has shifted in terms of relative dominance sequence that extends from 11,100 yr B.P. to the of tree taxa, the mixed conifer-northern hardwoods present (Watts, 1979). At Potts Mountain Pond, forest type persisted in the vicinity of Tannersville early Holocene mixed coniferous-northern hard- Bog throughout the Holocene. woods forest of oak, white pine, spruee, fir, and In contrast to Tannersville Bog, early Holocene birch persisted until about 9000 yr B.P. (Watts, vegetation at longswamp, Pennsylvania, was bo- 1979). Boreal conifers became locally extinct after real forest dominated by jaek pine, with some fir, 9000 yr B.P. and were replaced by oak, chestnut, birch (Betula populifolia), and oak (Watts, 1979). black gum, and hickory. White pine reexpanded Jack pine dominated from 12,000 to 6000 yr B.P., loeally in the late Holocene (Watts, 1979). after which it was replaced by forests of oak and In the Saltville Valley of southwestern Virginia, chestnut. The patterns of Holocene vegetation ob- from 12,500 to about 10,000 yr 8.P., a mixed served at Longswamp are similar to those recorded conifer-northern hardwoods forest was dominated farther to the west at Panther Run Pond (Figure by oak, spruee, white pine, and jack pine, with 1). At Panther Run Pond, closed boreal forest with hemlock, fir, birch, hornbeam, and maples (Figure spruce, fir, and jack pine dominated from 12,600 2). From 10,000 to 4700 yr 8.P., hemlock dominated until about 6400 yr B.P. (Watts, 1979). In the late the forests. Hemlock populations collapsed 4700 Holocene, boreal speeies were replaced by oak, years ago, and were replaced by oak, ash, and chestnut, hemlock, white pine, and black gum chestnut (Figure 2). (Nyssa Wlnatica) (Watts, 1979). At Bear Meadows Shady Valley Bog, located in northeastern Ten- Bog, located near Panther Run in central Penn- nessee (Figure 1), dates from approximately 10,000 sylvania (Figure 1), early Holocene boreal forest yr B.P. to the present (Barclay, 1957). At Shady of pine, spruce, fir, and tamarack was replaced in Valley Bog, hemlock was the forest dominant from the middle Holocene by oak, chestnut, hemlock, 10,000 yr B.P. to the middle Holocene interval. The beech, and black gum (Kovar, 1965). decline in hemloek populations at Shady Valley At Crider's Pond in south-central Pennsylvania, probably occurred at the same time as in Saltville closed boreal forest of jack pine, spruce, fir, and Valley, between 5000 and 4500 yr B.P. Late Hol- birch persisted until 9000 yr B.P. (Watts, 1979). ocene forests of Shady Valley were predominantly After 9000 yr B.P., oak forest with hiekory and oak, hickory, and chestnut. During the late Hol- hemloek replaced the pine forest. Hemloek popu- ocene interval, there was a minor reexpansion of lations were reduced at about4500 yr B.P., followed spruce and fir locally within Shady Valley (Bar- by minor reexpansion in Diploxylon pine in the late elay, 1957). Holocene.interval. Big Pond, to the southwest of No Holoeene pollen or plant-macrofossil records Crider's Pond (Figure 1), contains a Holocene rec- are yet available for the Blue Ridge Province of ord dating from the last 11,000 years (Watts, 1979). Virginia or its imrnediate vicinity. On the Pied- At Big Pond, Holocene vegetational changes were mont, the Marsh record from Pennsylvania (Mar- similar to Crider's Pond in timing and sequenee tin, 1958) indicates that boreal forests were re- of replaeementof boreal forest species bytemperate placed in the early Holocene interval by mixed forests, with oak, chestnut, and black gum replac- eonifer-northern hardwoods forest of pine, hem- PUBLICATION ?5 31

lock, birch, and oak. Deciduous forest of oak and levels approximately 6000 years ago. Although the chestnut established in the middle and late Hol- oldest radioearbon date available is 8900 yr B.P., ocene intervals. peat apparently began to accumulate as early as On the central Atlantic Coastal plain, Holocene 12,000 yr B.P. (Whitehead, 1972; Whitehead and vegetational changes were influenced by changes Oaks, 1979). The pollen record from the Dismal in climate and sea level. Two sites in north-cential Swamp shows four intervals of major vegetational pond New Jersey, Helmetta Bog and Szabo (Figure change. From 12,000 to about 10,600 yr B.P., open 1), record vegetational changes over the past 11,000 boreal forests of northern Diploxylon pine occurred years (Watts, 1979). In the early Holocene interval. with minor populations of spruce, birch, and oak. from 11,400 to 9000 yr B.P., the vegetation ai Between 10,600 and 8200 yr B.P., the vegetation Helmetta and Szabo ponds was boreal forest of changed to a mixed conifer-northern hardwoods s_p{uce and pine, with shrub birch (Betula glan- forest as beech and hemlock migrated into the area dulosa). After 9000 yr B.P., spruce populations and as birch and oak expanded their populations. became locally extinct, and both white pine and Between 8200 and 3500 yr 8.P., the vegetation was southern Diploxylon pine (P'irws rigid,a) estab- dominated by oak, with minor populations of south- lished. Oak increased between g000 and 8b00 yr ern Diploxylon pine, bald cypress (Tarod,ium d,is- B.P. yr After 8500 8.P., the forests were dominated ti,chum), h i ckory, an d sweetg um (Liquidamb ar sty - by oak, with hickory, black gum, and southern raciflua). At about 3500 yr B.P., a minor rise in Diploxylon pine (Watts, 1g?g). In the pine Barrens sea level resulted in a rise of the regional ground- region of southern New Jersey (Figure 1), southern water table and an expansion in the size of Dismal Diploxylon pines and oaks have eodominated for Swamp. Swamp forests of bald cypress, Atlantic the last 10,500 years (Florer, lgTZ). Traces of pollen white cedar (Chamaecyparis thy oides), tupelogum of cool-temperate and boreal species such as spruce, (Nysso aquatica), and red maple (Aeer rubru,m) hemlock, and birch persisted until about 10,000 yr expanded over the area of Dismal Swamp. The B.P. (Florer, Lg72). At Trenton, New Jersey (Fig- upland vegetation of the Coastal Plain of Virginia ure 1), a pollen spectrum dated 10,800 yr B.p. was was dominated by oak and southern Diploxylon dominated by pine, with minor amounts of spruce pine during the late Holocene interval (Whitehead and hemlock; however, spectra dating beiween and Oaks, 1979). 5100 and 2800 yr B.P. reflect middle to late Hol- At Rockyhock Bay, on the outer Coastal Plain ocene oak-pine forests on the Coastal plain (Sirkin of northeastern North Carolina, jack pine-domi- and others, 1970). Farther to the south, on the nated taiga persisted until about 10,000 yr B.P. central Delmarva Peninsula of Maryland, pollen (Whitehead, 1981). After 10,000 yr B.P., jack pine spectra dating 9000 and 8900 yr B.P. contain pollen and spruce were replaced by more dense forests assemblages codominated by oak, pine, and birch, of southern Diploxylon pine, white pine, oak, hick- with traces of spruce, hemlock, black gum, and ory, hornbeam, birch, beech, and hemlock. This hickory. closed, mixed forest persisted until about 5000 yr At the entranee to the modern Chesapeake Bay B.P., when it was replaced by a warm-temperate (Figure 1), radiocarbon dates are availrable from assemblage including southern Diploxylon pine, sediments at 11,600 and 10,800 yr B.p. (Harrison oak, bald cypress, tupelogum, red maple, and mag- and others, 1965). Pollen spectra associated with nolia (Magnolia). these dates indicate that 11,600 years ago, a mixed conifer-northern hardwoods forest dominated by pine, spruce, fir, hemlock, birch, willow, and oal LATE QUATERNARY GRADIENTS IN occupied the outer Coastal Plain at gZ" N latitude. PLANT AND ANIMAL COMMUNITIES After 10,300 yr B.P., forest composition shifted to a temperate assemblage dominated by oak and The central Atlantic states region has been char- southern Diploxylon pine. acterized as an area of both vegetational disequi- The development of Holocene vegetation on the librium and disharmonious faunal assemblages outer Coastal Plain of Virginia has been studied during the late Pleistoeene and early Holocene extensively at the Dismal Swamp (Figure 1). The (Watts, 1979; Delcourt and Delcourt, 1g8B; Lun- swamp developed at the mouth of a drowned river delius and others, 1983; Graham and Lundelius, valley as sea level rose during the postglacial 1984). Were the latitudinal gradients in composi- interval. The rate of peat accumulation in this tion of flora and fauna comparable during the late freshwater swamp and marsh complex is tied to Pleistocene and early Holoeene intervals? Can we the rise in sea level and its stabilization at modern interpret the past plant and animal communities 32 VIRGINIA DIVISION OF MINERAL RESOURCES in terms of the changing late Quaternary climate 1983) maintained alpine tundra and inhibited es- and geomorphic environments? tablishment of tree populations throughout the late North-to-south gradients in distribution and glacial interval. The periglacial, alpine-tundra en- composition of vegetation during the late Pleisto- vironment would have been an effective barrier to cene in the eentral Atlantic states were constrained migrations of boreal and temperate plant species with respect to physiography, geomorphic regime, both east and west of the Valley and Ridge Prov- and climate. During the late Pleistocene, severe ince. periglacial conditions occurred adjacent to the Within the central Atlantic states, both geomor- southern margin of the Laurentide Ice Sheet across phic processes and vegetational patterns and pro- all physiographic provinces represented in the cen- cesses were locked into a fundamentally glacial tral Atlantic states (P6w6, 1983). Discontinuous elimatic regime until approximately 12,500 yr B.P. permafrost extended along the Allegheny Plateau At lower elevations, two environmental regimes as far south as West Virginia and along the Blue were dominant throughout the late Pleistocene: (1) Ridge Province as far south as North Carolina and unglaciated landscapes of the Valley and Ridge Tennessee (Delcourt and Delcourt, 1981; Delcourt Province were shaped by colluvial processes; and and Delcourt, 1985). At lower elevations within the (2) land surfaces within the Atlantic Coastal Plain Valley and Ridge Province, under climatic condi- were modified by eolian processes. tions that were less severely cold than those prev- The high peaks of the Blue Ridge Mountains alent at higher elevations, colluvial processes served as an orographic barrier that screened moved regolith and rock debris downslope. This moisture from eastward-moving storms. The mois- regime of aceelerated mass wastage resulted in the ture and unstable regolith were transported down- rapid filling of poorly-drained depressions with slope, westward into the Valley and Ridge Province mineral sediment. In the northern portion of the and eastward onto the Piedmont. The landscapes Valley and Ridge Province, in central Pennsylva- of the Valley and Ridge were modified by fteeze- nia, the late Pleistocene landscape was eharacter- thaw processes, where localized disturbances such ized by extensive patches of open ground, with a as debris avalanches, solifluetion lobes, and the flow mosaic of tundra and taiga communities. Geomor- of block streams created open patches of exposed phic disturbances would have continually "reset" sediments within an otherwise closed boreal forest plant succession, resulting in a predominance of (see Conners, this volume). These geomorphic ep- early-successional or pioneer communities of herbs isodes of mass-wastage increased the area main- and shrubs. Farther to the south, from central to tained in pioneer herb and shrub communities, as southwestern Virginia, the landscape was predom- well as in early-successional forest. inantly closed boreal forest, with more-open Within the Appalachian Mountains, the change patches of muskegs developed within poorly- from glaeial to interglacial climatic conditions drained valley bottoms. The increase in richness occurred between 13,000 and 12,000 yr B.P.; this of plant species in boreal forests from Hack Pond corresponded with the transition from a colluvial to the Saltville Valley may reflect the areal de- to a fluvial geomorphic regime. As a result, during velopment of mid- and late-successional forests on the Holocene, the overland flow of water by sheet- more stable landscapes. The frequency and inten- wash was less effective in erosion and transport sity of geomorphie disturbances were greater at of mineral debris. As the geomorphic threshold 38o N latitude than they would have been at 36o from colluvial to alluvial processes was passed, N during the late Pleistocene. trees eolonized the former colluvial surfaees and In eastern North America, elimatic amelioration closed forests were able to establish on stabilized began at latitudes of 34" to'36" N as early as 16,500 upland slopes. The extensive root systems of the yr B.P. (Deleourt and Deleourt, 1984). During the trees and thick organic litter accumulating beneath late-glacial interval, populations of temperate tree the forest canopy would have further inhibited species increased along the southern edge of the erosion of the soil. During the Holocene' concen- boreal forest at sites located at low elevations in tration of alluvial activity within well-defined western and middle Tennessee, Georgia, and South stream networks maintained treeless openings Carolina (Delcourt and Delcourt, 1984). In the within the vegetation only within alluvial glades Allegheny Plateau and Blue Ridge physiographie (wet meadows) and bogs. Sediment that filled provinces, persistence of severely cold climate and ponds, glades, and bogs changed from inorganie frequent disturbance by active geomorphic pro- to organic-rich at the transition from late glaeial cesses such as freeze-thaw churning of soil (P6w6, to early Holocene eonditions (Maxwell and Davis, PUBLICATION 75 33

1972; Watts, 1979), reflecting decreased colluvial consistent with the paleoecologic reconstruction of activity and increased rates of peat growth in these boreal forest and poorly-drained muskegs within environments. the Saltville Valley during the late glacial interval. On the Atlantic Coastal Plain, cold and drv The three remaining sites contain assemblages climatic conditions occurred during the late pleis- of small mammals that date from approximately tocene within the rain shadow of the Appalachian 12,000 to about 10,000 yr B.P. (Guilday, 1984; Mountains. Extensive eolian activity duringthe full Guilday and others, 1978). Guilday and others glacial glacial and late intervals resulted in the (1978) document a latitudinal gradient in compo- reworking of sand across the land surface. for sition of small mammals from these sites. Although example, the Parsonsburg Sand layer that mantles boreal, temperate, and grassland mammals were the central Delmarva Peninsula (Denny and others, reported at each site, their relative abundances 1979). The migration of sand dunes and sheets changed systematically with latitude. At New Paris generated would have gaps within the boreal forest No. 4, the assemblage of small mammals indicated and maintained open taiga across the landscape. a mixture of tundra and taiga environments (Guil- Driven by wind, water currents eroded shores of day and others, 1964). At Clark's Cave, the mic- ponds lakes and and thus created oriented Carolina rotines were characteristic of dense boreal forests Bays, i.e., elongate, shallow lakes with sand dunes (Guilday and others, L977). The faunal assemblage developing along their southeastern margins at Baker Bluff Cave was compatible with a boreal, (Whitehead, 1981). The combination of eolian dis- mixed coniferous-deciduous woodland. From north turbance and nutrient-poor quartzose substrates to south, a climatie gradient from boreal to tem- was a limiting factor for primary productivity perate eonditions was inferred (Guilday and others, within upland taiga distributed from central New 1978). The fossil assemblages contained species of Jersey to North Carolina. The central Ailantic mammals whose modern ranges do not overlap, and Coastal Plain remained in a glacial climatic regime therefore have been called disharmonious assemb- until about 10,500 yr B.P. beeause of the maritime lages. These assemblages lacked good modern ana- influence of the cold Labrador Current extending logs and have been interpreted as evidence for more along the Atlantic Coast as far south as Soutl equable climatic conditions during the late Pleis- Carolina (Delcourt and Delcourt, 1984). tocene, with reduced seasonal extremes (Lundelius After 12,500 yr 8.P., boreal taxa spread both and others, 1983; Graham and Lundelius, lg84). northward onto newly deglaciated terrain and to In part, the north-to-south gradient in geomorphic plateau higher elevations within the Allegheny and processes .also maintained the fine-grained char- the Blue Ridge Province. Temperate taxa expanded acter of environmental heterogeneity of the late northward along migration routes that followed the Quaternary Appalachian landscapes. Valley and Ridge corridor and the Atlantic Coastal The disharmonious assemblages of fossil mam- Plain. The combination of climatic amelioration. mals appear to reflect the response of mammalian fundamental changes in geomorphic regimer, communities to rnajor climatic amelioration during plant-species migrations "nd northward resulted in the late glacial and early Holocene intervals, as vegetational disequilibrium and dynamically well as the increased diversity of habitats that changing landseapes during the early HolocenL resulted from both widespread geomorphic and interval, particularly between 12,500 and 10.000 vegetational instability. The vegetational disequi- yr (Deleourt B.P. and Delcourt, 1g8B). librium (Delcourt and Delcourt, 1988) resulted Four sites with late Pleistocene faunal informa- from northward migrations of plant species ilong tion province within the Valley and Ridge extend- the Valley and Ridge corridor, as well as from ing from Pennsylvania to Tennessee provide some ehanges in disturbance regimes with the transition ofthe best-dated evidence available concerning late from colluvial to alluvial geomorphic processes. Pleistocene faunal gradients (Lundelius and others. With the shift 10,000 years ago to a landscape 1983; Guilday, 1984). From north to south, these modified predominantly by alluvial processes, the sites pennsylva- inelude New Paris No. 4, central vegetational mosaie changed from the Pleistocene nia; Clark's Cave, western Virginia; Saltville, pattern (an open patchwork oftundra, open glades, southwestern Virginia; and Baker Bluff Cave. and boreal forests) to a Holocene pattern of pre- northeastern Tennessee (Figure 1). dominantly closed, temperate deciduous forests. Pleistocene fauna from Saltville date between The regional vegetation continued to ehange and 13,600 yr (Ray _1q,q00 B.P. and others, 196?; throughout the past 10,000 years because of sub- McDonald, this volume). The species-rieh assem- sequent northward migrations of warm-temperate blage of proboseidians, musk oxen, and caribou is trees, the expansion of coastal swamps related to 34 VIRGINIA DIVISION OF MINERAL RESOURCES the rise of sea level to its modern position about Denny, C. S., Owens, J. P., Sirkin, L. A., and Rubin, 6000 years ago, and relatively minor Holocene M., 19?9, The Parsonsburg Sand in the Central climatic change. However, the magnitude of en- Delmarva Peninsula, Maryland and Delaware: vironmental changes through the middle and late U. S. Geol. Survey, Prof. Paper 1067-8, 16 p. Holocene intervals was not as great as that char- Delcourt, H. R., and Delcourt, P. A., 1985, Qua- aeterizing the transition from the late glacial to ternary palynolory and vegetational history of the early Holocene. the southeastern United States, p. 1-37 'inBry- ant, V. M., Jr., and Holloway, R. G., eds., Pollen DIRECTIONS FOR FUTURE RESEARCH Records of Late-Quaternary North American Sediments: Amer. Assoc. Strat. Palynol. Foun- From this summary of late Quaternary vegeta- dation. tional changes known from the central Atlantie Delcourt, P. A., and Deleourt, H. R., 1981, Vege- states, it is apparent that very little information tation maps for eastern North America:40'000 is available for several major physiographic re- yr B.P. to the present, iro Romans, R., ed., gions. No sites with full glacial sequenees of pollen Geobotany II. New York, Plenum Press, p. 123- or plant-macrofossil assemblages have yet been 166. studied in the Blue Ridge, Valley and Ridge, or < 1983, Late-Quaternary vegetational dy- Piedmont provinces of Virginia. Patterns of vege- namics and community stability reconsidered: tational ehange are yet poorly understood for the Quat. Res., vol. 19, P.265-27L. central portion of the Atlantic Coastal Plain, with 1984, Late-Quaternary paleo-climates only the pioneering studies of Dismal Swamp re- and biotic responses in eastern North America presenting the state of Virginia. Future research and the western North Atlantic Ocean: Palaeo- should eoncentrate upon establishing a network of g:eogr., Palaeoelimatol., Palaeoeeol., vol. 48, p. radiocarbon-dated sites with detailed paleoecologie 263-284. studies throughout this critical region. Florer, L. E., lg7?,Palynology of a postglacial bog in the New Jersey Pine Barrens: Bull. Torrey ACKNOWLEDGMENTS Bot. Club, vol. 99, P. 135-138. Graham, R. W., and Lundelius, E. L., Jr., 1984'" This research was sponsored by the Ecology Coevolutionary disequilibrium and Pleistocene Program of the National Scienee Foundation extinctions,'i'n Mafiin, P. S., and Klein, R. G., through grants BSR-83-00345 and BSR-83-06915 eds., Quaternary extinctions, a prehistoric rev- to the University of Tennessee. We wish to thank olution: Tucson, Univ. Arizona Press, p.223- Jerry McDonald and the citizens of Saltville for 249. permission and assistanee in accomplishing the Guilday, J. 8., 1,984, Pleistocene extinction and field work in connection with pollen and plant- environmental ehange, case study of the Ap- macrofossil studies of the Saltville Valley section. palachians, inMartin, P. S., and Klein, R. G., This paper represents Contribution No. 35 of the eds., Quaternary extinetions, a prehistoric rev- p. Program for Quaternary Studies of the Southeast- olution: Tucson, Univ. Arizona Press, 250- ern United States, University of Tennessee, Knox- 258. ville, Tennessee, and is a United States contribution Guilday, J. E., Hamilton, H. W., Anderson, E., and to IGCP Project 158 B. Parmalee, P. W., 1978, The Baker Bluff eave deposit, Tennessee, and the late Pleistocene faunal gradient: Bull. Carnegie Mus. Nat. Hist., no. 11, 66 p. REFERENCES Guilday, J. E., Martin, P. S., and McCrady, A. D., 1964, New Paris No. 4: a late Pleistocene cave Barclay, F. H., 195?, The natural vegetation of deposit in Bedford County, Pennsylvania: Nat. Johnson County, Tennessee, past and present: Spel. Soe. Bull., vol. 26,p.121-L94. Ph.D. Dissertation, Univ. Tennessee, Knoxville, Guilday, J. E., Parmalee, P. W., and Hamilton, H. I47 p. W, L977, The Clark's Cave bone deposit and Craig, A. J., 1969, Vegetational history of the the late Pleistocene paleoecologa of the central Shenandoah Valley, Virginia: Geol. Soc. Amer- Appalachian Mountains of Virginia: Bull. Car- ica Spee. Paper 123,p.283-296. negie Mus. Nat. Hist., no.2,87 P. PUBLICATION 75 35

Harrison, W., Malloy, R.J., Rusnak, G. A., and neapolis, Univ. Minnesota Press, p. 157-189. Terasmae, J., 1965, Possible late Pleistocene Ray, C. E., Cooper, B. N., and Benninghoff, W. S., uplift: Chesapeake Bay entrance: J. Geol., vol. 1967, Fossil mammals and pollen in a late 73,p.201-229. Pleistocene deposit at Saltville, Virginia: J. Kovar, A. J., 1965, Pollen analysis of the Bear Paleontol., vol. 41, p. 608-622. Meadows Bog of central Pennsylvania: Penn- Sirkin, L. A., Denny, C. S,, and Rubin, NI.,Ig77, sylvania Acad. Sci., vol. 38, p. 16-24. Late Pleistocene environment of the central Lewis,I. R., and Cocke, E. C., 1929, Pollen analysis Delmarva Peninsula: Geol. Soc. America Bull., of Dismal Swamp peat: J. Elisha Mitchell Sci. vol.88, p. 139-142. Soc., vol. 45, p. 37-58. Sirkin, L. A., Owens,J. P., Minard,J. P., and Rubin, Lundelius, E. L., Jr., Graham, R. W., Anderson, M., 1970, Palynology of some upper Quaternary E., Guilday, J., Holman, J. A., Steadman, D. peat samples from the New Jersey eoastal plain: W., and Webb, S. D., 1983, Terrestrial verte- U. S. Geol. Survey Prof. Paper 700-D, p.77- brate faunas, in Porter, S. C., ed., Late-Qua- 87. ternary environments of the United States, the Watts, W. A., 1979, Late Quaternary vegetation of late Pleistocene: Minneapolis, Univ. of Minne- central Appalachia and the New Jersey Coastal sota Press, p. 3ll-353. Plain: Ecol. Monogr., vol. 49,p.427'469. Martin, P. S., 1958, Taiga-tundra and the full- Whitehead, D. R., 1972, Development and environ- glacial period in Chester CounW, Pennsylvania: mental history of the Dismal Swamp: Ecol. Amer. J. Sci., vol.256, p.470-502. Monogr., vol.42, p. 301-315. Maxwell, J. A., and Davis, M. 8., 1972, Pollen 1,981, Late-Pleistocene vegetational evidence of Pleistoeene and Holoeene vegeta- changes in northeastern North Carolina: Ecol. tion on the Allegheny Plateau, Maryland: Quat. Monogr., vol. 51, p.451-471. Res., vol. 2, p. 506-530. Whitehead, D. R., and Oaks, R. Q., Jr., 1979, P6w6, T. L., 1983, The periglacial environment in - Developmental history of the Dismal Swamp, North Ameriea during Wisconsin time, iz Por- 'in Kirk, P. W., Jr., ed., The great Dismal ter, S. C., ed., Late-Quaternary environments Swamp: Charlottesville, Univ. Press of Vir- of the United States. the late Pleistocene: Min- ginia, p.25-43. 36 VIRGINIA DIVISION OF MINERAL RESOURCES

THE KNOWN HERPETOFAUNA OF THE LATE QUATERNARY

OF VIRGINIA POSES A DILEMMA

J. Alan Holmanl

CONTENTS

Page

Late Pleistocene herpetofaunal studies in Virginia ...... 36

TABLE

The late Pleistocene herpetofauna of Virginia 38

ABSTRACT of Virginia and to suggest directions for future paleoherpetologieal studies. Although the mammalian fauna of the late Although few reports on Virginia's late Pleis- Pleistocene of Virginia contains extinct speeies tocene amphibians and reptiles have been done, and speeies that indicate boreal climatic condi- implications from these reports pose a serious tions, the late Pleistocene herpetofauna appears dilemma. Although Virginia's late Pleistocene almost identical to what it is today. This dilemma mammalian fauna contains extinct forms and is discussed, and future Pleistocene herpetolog- forms that suggest boreal conditions, the herpe- ical studies are suggested. tofauna appears to be almost identical to what it is today. The implieations of this dilemma will INTRODUCTION be given after the review of previous studies that follows. Among the most compelling problems of the late Pleistocene are the rapid extinction of the LATE PLEISTOCENE HERPETOFAUNAL mammalian megafauna, the postulated rapid STUDIES IN VIRGINIA change in the climate, and ways in which the spread of humankind might be related to the two Workers heretofore involved in late Pleistocene previous events. Paleoherpetological studies re- herpetofaunal studies in Virginia are, in chro- flect directly or indireetly on all of these ques- nological order: Olaf A. Peterson, Neil D. Rich- tions. Thus, the purpose of the present paper is mond, Helen McGinnis, John E. Guilday, Leslie to examine and reflect upon the record of am- P. Fay, J. A. Holman, and Jerry N. McDonald. phibians and reptiles from the late Pleistoeene Compared to avian and mammalian remains, herpetological fossils are quite difficult to identify. Thus it is not surprising that a few misidentifications of herpetological remains, or ob- rMuseum, Michigan State University, East jects thought to be herpetological remains, were Lansing, Michigan 48824 made. These early misidentifications led to some PUBLICATION 75 37 confusion in previous interpretations of Pleisto- prey of raptors that frequented the feature. The cene faunas and events. location, geology, and description of the site are O. A. Peterson (1917) reported "a large croc- found in Guilday (1962). Herpetological remains odile tooth" associated with the remains of a detailed in Guilday's paper were identified by mastodon in a "fossil-bearing alluvial deposit" at Neil D. Riehmond. Richmond identified two sa- Saltville, Virginia. This report led to much con- lamanders, five anurans, two turtles, one lizard, fusion. Anyone trying to discuss the faunal as- and ten snakes from the site. Fay (1984) restudied sociation of the Saltville deposit had to either the Natural Chimneys herpetofauna and added postulate some complex reason to account for the three salamanders, six anurans, one lizard, and presence of the crocodile in a predominantly seven snakes to the list. Fay eliminated the south- boreal fauna, or to consider the fauna hopelessly ern extral i m ital forms M astic op his fl ag ellum and heterochronic. N. D. Richmond (1963) pointed out Crotalus ad,amanteus from the list in Guilday that the "croeodile tooth" was actually a fossil (1962). Noteworthy is the fact that the little invertebrate, most likely an orthoceran cepha- ground skink, Scincella lateralis, is known as a lopod that had eroded out of a lower Paleozoic fossil only from the Natural Chimneys site (Fay, limestone in the area, and had washed into the 1984). The Natural Chimneys herpetofauna is site as a part of the gravel in which the Pleistocene listed in the Table. mammals were found. Ray, Cooper and Ben- The Clark's Cave deposits, 7.4 miles (12 m) ninghoff (1967) demonstrated that the specimen southwest of Williamsville, Bath County, Vir- was an aberrant mandibular tusk of a mastodon, ginia, was excavated in the early 1970's by field M ammut americ anum (Kew). parties from the Carnegie Museum. The site was Later N. D. Richmond (in Guilday, 1962) iden- discovered in 1972. The site lies at an elevation tified a Coachwhip Snake ( M asticophis flagellum) of 1480 feet (448 m). The bones are thought to and an Eastern Diamondback Rattlesnake (Cro- have been derived from the prey of raptors. A talus ad,amanteus) based on vertebral remains quite detailed account of the regional setting and from the late Pleistocene Natural Chimneys site, history of the site is given in Guilday and others Augusta County, Virginia. These species are not (1977). Helen McGinnis identified the amphibi- found in Virginia today although they do range ans and John E. Guilday identified the reptiles north to near its southern border (Conant, 1975). reported in this paper. This herpetofauna in- L. P. Fay (1984) restudied the snake material cluded two salamanders, seven anurans, two li- from the Natural Chimneys site and suggested zards, and two snakes. Fay (1984) added seven that the fossil material assigned to Masticophis salamanders, three anurans, one turtle, and flagellum should be reassigned to Coluber con- twenty snakes to the list. The Clark's Cave her- strietor, a species that lives in the area today. This petofauna is also listed in the Table. change was based on the fact that Fay was unable The Saltville site, Smyth County, Virginia, has to find characters to separate isolated Mastico- yielded vertebrate fossils for more than two cen- phis and Coluber vertebrae, and on modern zoo- turies (McDonald, 1984). The site lies at an el- geographic grounds. Fay also changed the iden- evation of about 1720 f.eet (525 m). A description tifieation of Crotalus ad,amanteus to Crotalus of the site and previous research there is given horrid,us, the latter speeies also being a modern by McDonald (1984). Amphibians and reptiles resident of the area. This change was based on from the Saltville site were collected mainly by the low vertebral neural spine of the fossil, a field parties from Radford University which were character whieh readily separates C. horridus directed by J. N. MeDonald during 1980 to 1984. from C. adamanteus. A report on amphibians and reptiles of the Salt- Only three sites in Virginia thus far have ville site (Holman and MeDonald, 1985) detailed yielded significant published herpetofaunas. two salamanders, two anurans, two turtles, and These are the Natural Chimneys site in Augusta two snakes. The Saltville herpetofauna is also County, Clark's Cave in Bath County, and the listed in the Table. Saltville site in Smyth County, Virginia. The Natural Chimneys site, 1 mile (1.6 km) west DISCUSSION of Mt. Solon, Augusta County, Virginia, was excavated by Carnegie Museum field parties Late Pleistocene amphibians and reptiles have from 1949 through 1961. The site lies at an been studied mueh less thoroughly than mam- elevation of 1329 to 1476 feet (405 to 450 m). The mals. This is probably the result of two faetors. bones are thought to have been derived from the First, fossil amphibians and reptiles often are 38 VIRGINIA DIVISION OF MINERAL RESOURCES

Table. The Late Pleistocene Herpetofauna Of Virginia. *

Scientific Name Vernaeular Name Natural Clark's Saltville Chimmeys Cave C rg ptobr anchus alle g aniensis Hellbender 0 0 + + N otophthalmus airid e s c ens Eastern Newt + + Amby stoma (maculatum group ) Mole Salamander + + 0 Ambystoma opacunx Marbled Salamander 0 0 AmbEstomatigr'inum Tiger Salamandel * 0 Desmognathus sp. Dusky Salamandel * 0 Plethodontin'i sp. Woodland Salamandel * 0 Ga rinophilus p or p hy riticus Spring Salamander 0 0 Pseudotriton sp. Red Salamander 0 0 Seaphiopus holbrooki Eastern Spadefoot + 0 Bufo americanus American Toad + 0 + + B uf o w o o dhous ei f ow leri Fowler's Toad HEla chrysoscelis or H. uersicolor Gray Treefrog + 0 Hgla crucifer Spring Peeper 0 0 Rana catesbeiana Bullfrog + 0 Rana clamitans Green Frog + 0 Rana palustris Pickerel Frog + 0 Rana ef. Rana pipiens Northern Leopard Frog + + Rana sylaatica Wood Frog + 0 Cheladra serpentina Snapping Turtle + + Graptemgs geographica Map Turtle 0 0 Terrapene carolina Eastern Box Turtle + 0 Chrysemys picta Painted Turtle 0 + Sceloporus undulatus Eastern Fenee Lizard + 0 Scincella lateralis Ground Skink + 0 Eumeces ? Eumeces laticeps Broadhead Skink 0 0 Nerodia erythrogaster Plainbelly Water Snake + 0 Nerodia sipedon Northern Water Snake + + ? Regina septemuittata Queen Snake 0 0 Stor eria oc cipitomaculata Redbelly Snake + 0 Storeria dekagi Brown Snake 0 0 Thamnophis sauritus Eastern Ribbon Snake + 0 Thamnophis sirtalis Common Garter Snake + 0 Virginia aaleriae Earth Snake 0 0 Carphophis arnoenus Worm Snake + 0 Diadophis punctatus Ringneck Snake + 0 Heterodon platyrhinos Eastern Hognose Snake 0 0 ? Tantilla eoronata Southeastern Crowned Snake 0 0 Coluber sp. or Masticophis sp. Raeer or Coaehwhip Snake + 0 Opheodrys aestiaus Rough Green Snake 0 0 Opheod,rys oernalis Smooth Green Snake 0 0 Cemophora coccinea Scarlet Snake 0 0 Elaphe guttata Corn Snake + 0 Elaphe obsoleta Rat Snake + + Lampropeltis getulus Common King Snake + 0 L ampr op eltis c allig aster Mole Snake 0 0 L ampr op eltis triangulum Milk Snake + 0 Crotalus homidus Timber Rattlesnake + 0

*Indeterminate genera are not listed in one locality if they are specifically identified in one of the other localities. + = present; 0 = absent. PUBLICATION 75 39

represented by small, fragile bones that may be fauna that the deposit was heterochronic. This rare or overlooked at fossil sites. Second, herpe- reasoning undoubtedly grew out of the ingrained tological fossils are more difficult to identify than belief that solution features in the southeast quite the mammalian jaws and teeth traditionally stud- often contained mixtures of Miocene, Pliocene, ied by vertebrate paleontologists. For the second or Pleistoeene fossils, or mixtures of fossils of reason many workers have hesitated to identify various Pleistocene ages. The faet is that most herpetologieal remains for fear of making mis- solution deposits in the southeast eontain rather takes. Nevertheless, herpetological information restricted Pleistocene faunas, occasionally Plio- is quite important in the study of late Pleistocene cene faunas, and rarely Miocene faunas; very paleoecology. rarely is there a scrambled mixture of these Hibbard (1960) first realized the significanee faunas of different ages. The only mixed fossils of crocodilians and large land tortoises in the that regularly occur are the readily identifiable interpretation of Pliocene and Pleistocene cli- invertebrate fossils redeposited from the lime- mates. He also believed (personal communica- stones that previously eontained these forms. tion) that information gathered from the study Webb (1974) and Holman (1976) have shown that of smaller amphibians and reptiles would yield almost all known southeastern Pleistocene faunas important data about Plioeene and Pleistocene in solution features such as caves and sinkholes climates. He reasoned that these ectotherms are not heteroehronic but represent fairly res- would be much more sensitive to environmental tricted units of time. changes than mammals and thus would be much The fact that solution features in the southeast- more indicative of climatic change. Another point ern United States almost always contain fossil is that most amphibians and reptiles of the late faunas representing restricted units of time Pleistocene represent living species whose eco- merely follows the basie paleontological principle logical toleranees are fairly well known. that in order to have fossilization, some type of Unfortunately, herpetologieal evidence from immediate burial must occur. The seenario that Pleistocene faunas has sometimes been misinter- occurs is as follows. (1) Animal remains are preted. Many Pleistocene faunas contain herpe- trapped in or carried into solution features by tological species that today would not be ecolog- raptors or carnivores, and (2) rapid sedimenta- ically compatible with the mammalian species of tion oecurs to lead to the fossilization of the bones. the fauna (Lundelius and others, 1983). Workers This would explain why some solution features sometimes disregarded these herpetological spe- contain vertebrate fossil remains, while other cies, believing that they were modern intrusions seemingly identical ones are barren of fossils. into the fauna, and based their paleoclimatic Again, there is no evidence that the late Pleis- interpretations on the mammalian component tocene deposits in Virginia that have amphibians only. and reptiles reported from them contain hete- Actually, one usually can reeognize quite easily rochronic faunas, and therefore we must conclude intrusions of herpetological elements into a fauna, that a herpetofauna of such a deposit was con- and I wish to emphasize that none of the late temporaneous with the mammalian fauna. Pleistoeene Virginia amphibians and reptiles Another important factor to consider (Russell listed here are considered to be modern intru- W. Graham, personal communication) with re- sions. Intrusive herpetological forms may be rec- gard to faunas with ecologically incompatible ognized on the basis of three criteria that follow. members is that there are practically no mam- (l) Intrusive forms are almost always burrowers malian faunas from archeological or paleontolog- such as Mole Salamanders of the genus AmbEs- ical sites in the Holocene that have ecologically toma or bumowing anurans such as Seaphiopus. incompatible members, yet these assemblages (2) Bones of intrusive forms are usually of a are presumably subjected to the same taphonomic differentcolor and densitythan those ofthe fossils processes as the late Pleistocene ones. Unfortu- of the deposit. (3) Intrusive forms usually consist nately, there is practically no published informa- of suspiciously complete skeletons, unlike fossils tion on the herpetofaunas of the Holocene. of a deposit which usually consist of isolated The realization by Hibbard (f960) that faunas bones. that he studied from the High Plains, especially Another mistake that was made, especially in those from southwestern Kansas, were unit fau- the interpretation of southeastern Pleistoeene nas, and had ecologically incompatible forms in herpetofaunas, was to reason that when eeolog- them,led him and subsequent authors (Dalquest, ically ineompatible forms were present in the 1965; Dalquest and others, 1969; Hibbard, 1970; 40 VIRGINIA DIVISION OF MINERAL RESOURCES

Hibbard and Taylor, 1960; Martin and Neuner, (personal communication). The other extralim- 1978; Slaughter, 1967, 1975; Slaughter and ital f orm s, Gr aptemg s g eo gr aphic a, S c'inc ella late - Hoover, 1963; and Taylor, 1965) to form a Pleis- ralis, Nerod'ia erythrogaster, Tantilla coronata, tocene "climatic equability hypothesis". The cli- and Cemophora coccinea oec:ur south of the Vir- matic equability hypothesis explained the former ginia Pleistocene localities (Table; Conant, 1975). co-occurrence of extralimital northern and south- Thus, based on the herpetofauna alone, one would ern forms that are considered climatically incom- assume that the climate was nearly the same or patible today, as follows. It was reasoned that perhaps slightly warmer than today in the late mild winters would account for the presence of Pleistocene of Virginia extralimital southern species, and that cooler On the other hand, the mammalian fauna of summers would account for the presence of ex- the Natural Chimneys, Clark's Cave, and Saltville tralimital northern species. This hypothesis was contains so many boreal species that previous based mainly on mammals from the Pleistocene workers have postulated a boreal climate and of the central and southern Great Plains, but boreal vegetation for Virginia in the late Pleis- giant tortoises of the genus Geochelone were con- tocene (Guilday, 1962; Guilday and others, 1977). sidered especially important in indicating very Based on the herpetological evidence from these mild winters when these reptiles were present sites, I cannot emphasize enough that we must in faunas.' take a critical look at the environmental recon- The term "disharmonius communities" re- structions based on the mammals. cently has been used widely to describe commu- First, a boreal-type summer climate seems nities that have species in them that would be impossible. Stuart (1979) pointed out that a mean ecologically incompatible today. Lundelius and July temperature well in excess of 18oC appears others (1983) provide a comprehensive discussion necessary for the European Pond Tortoise, EmEs of disharmonious eommunities and their relation- orbicularis, to breed successfully in Europe. This ship to the Pleistocene climatic equability hypoth- is why no turtle populations are able to maintain esis based on terrestrial vertebrate faunas in the themselves in southern England, which has a cool United States from 25,000 to 10,000 years before equable climate that lacks enough warm weather the present. in the summer for turtle eggs to hatch. Certainly With specific reference to fossil herpetofaunas, the four species of late Pleistocene Virginia tur- Holman (1976) suggested that ecologically incom- lles (Cheludra serpentina, ChrEsemgs picta, Grap- patible amphibians and reptiles in late Pleisto- temys geographica, and especially Terrapene ca' cene deposits in Florida, Georgia, and Mississippi rolina) would not have been able to have could be explained on the basis of the climatic successful hatching of eggs in an arba with a equability hypothesis. A more comprehensive re- boreal summer climate. Moreover, the large num- view of Pleistocene herpetofaunas of eastern and ber of egg-laying snakes found in the three Vir- central North America (Holman, 1980) did not ginia late Pleistocene faunas would have had support the classical concept of alternating similar problems reproducing in a boreal warm-dry, cool-moist glacial and interglacial cli- summer climate. These species include Carpho- mates. On the contrary, the herpetological evi- phis amoenus, Diadophis punctatus, Heterodon dence generally indicated Pleistocene climates platyrhinos, Tantilla coronata, Coluber constric- south of the glacial limits were warmer or more tor, Opheodrgs aest'iuus, Opheodras uernalis, Ce- equable than today until at least very late Pleis- mophora coccinea, Elaphe guttata, Elaphe ob- tocene times. soleta, and Lampropeltis getulus. The egg-laying But what do the amphibians and reptiles thus I izards S celop orus undulatus, Scincella later alis, far identified from the Pleistoeene of Virginia and Eumeces laticeps provide further evidence indicate regarding the climate of the time? With that summers must consistently have had periods only a few exceptions, the late Pleistocene am- of much warnner temperatures than occur in phibians and reptiles occur in the same areas boreal areas today. today and none of them is extinct. Two of the To repeat, these reptiles are not intrusive ele- extralimital forms, AmbEstoma tigrinum and ments in the faunas and were contemporaneous R an a cf . p ip'iens (T able ) are p robab ly ab se nt f rom with the fossil mammal species of Natural Chim- the area today because of local edaphic changes neys, Clark's Cave and Saltville (Fay, 1984; Hol- (Fay, 1984). These forms also may indicate the man and McDonald, 1985). Therefore, it would individualistic response of species to their own seem appropriate to suggest other climatic and tolerances, as emphasized by Russell W. Graham vegetational models to replace the boreal one that PUBLICATION 75 4l

has been based almost entirely on the mammalian localities. Moreover, the important Saltville site fauna. continues to yield herpetological fossils. Holoeene Russel W. Graham (personal communication) herpetofaunas are espeeially in need of study to believes that the mammalian forms did not react determine whether amphibians and reptiles show to environmental change as "eommunity units", modern community structure as do the mammals. but rather that each species responded to envi- In future studies, more taphonomic work needs ronmental modifications according to their own to be done such as establishing whether the tolerance limits. Moreover, he believes that north- accumulation of bone results from raptor or pred- ern mammals forced into southern communities ator activities, hibernating aetivities of the amby the glaciation became intergrated with the phibians and reptiles involved, or from other resident biota, creating the disharmonius com- causes. munity. He believes an equable climate would More fossil material is needed from existing then maintain this assemblage. The herpetofaunal units to confirm some of the previous fauna, on the other hand, at the latitude of Vir- herpetological identifications. Most snakes have ginia, would not appear as mixed as the mam- been identified on the basis of isolated vertebral malian fauna, for it contains no strictly northern remains, and small natricine and colubrine herpetological species. Thus a boreal climate is genera are espeeially difficult to identify on the not indieated for the late Pleistocene of Virginia. basis of such material (Holman, 1981). Dreimanis (1968) suggested that ehanges in Finally, data on minimal critical temperatures vegetational structure and composition rather necessary for various reptile eggs to hatch would than shifting ecological zones may have occurred be most enlightening and helpful in the study of in the late Pleistocene. This composite vegeta- the ecology and distribution of late Pleistocene tional situation may have allowed for the simul- reptile species. It would appear that these data taneous occurrence of northern and temperate would provide an unprecedented chance to pos- species. tulate critical mean summer temperatures in the A seeond part of the herpetological dilemma late Pleistocene. is that although many late Pleistoeene mammalian genera and species became extinct in Vir- ACKNOWLEDGMENTS ginia (and elsewhere in the world), thus far no extinct amphibian or reptile taxon has been re- Leslie P. Fay kindly allowed me to cite the late ported from the late Pleistocene of Virginia (Ta- Pleistocene herpetofaunal lists from the Natural ble). In fact, only a very few extinet herpetologieal Chimneys and Clark's Cave sites from his doctoral species are known from the late Pleistoeene of dissertation at Michigan State University, and the entire United States, and the status of at least he provided thought-provoking conversations one of these is in doubt (Lundelius and others, about paleoherpetological problems. Letters and 1e83). fossil material sent by Neil D. Richmond in the Certainly numbers of some herpetological spe- late 1950's initiated my thinking about late Pleis- cies have dwindled in historic times, and some tocene herpetological problems in the Appala- forms today may be on the verge of extinction chian Region. Ralph E. Eshelman, Fred Grady, because of the destruction of habitats by humans. Russell W. Graham, and Jerry N. McDonald read But it is important to note that amphibians and drafts of this paper and provided valuable com- reptiles appear to have been able to cope admi- ments and materials. I thank all of these gen- rably with the same conditions that drove many tlemen for their help without which this paper mammalian taxa to extinction. Whatever biolog- could not have been written. ieal attributes these cold-blooded species have, they evidently have fitted them well for changes that took place in the late Pleistocene and Hol- ocene. REFERENCES More Late Pleistocene Studi,es are Need,eil.-It is clear that the study of late Pleistocene-early Conant, R., 1975, A field guide to reptiles and Holocene amphibians and reptiles in Virginia is amphibians of eastern and central North yet in its infaney. It is beyond the seope of this America: Boston, Houghton Mifflin, 429 p. paper to list all of the known localities in Virginia Dalquest, W. W., 1965, New Pleistocene forma- containing herpetofaunas, but Ralph E. Eshel- tion and local fauna from Hardeman County, man has provided me with a list of 17 such Texas: J. Paleontol., vol. 39, p. 63-72, 42 VIRGINIA DIVISION OF MINERAL RESOURCES

Dalquest, W. W., Roth, E., and Judd, F., 1969, and Webb, S. D., 1983, Terrestrial vertebrate The mammal fauna of Schulze Cave, Edwards faunas, inPorter, S. C., ed., Late-Quaternary County, Texas: Bull. Florida St. Mus., vol. 13, environments of the United States: The late p.205-276. Pleistocene: Minneapolis, Univ. Minnesota Dreimanis, A., 1968, Extinetion of mastodons in Press, p. 311-353. eastern North America: testing a new cli- Martin, L. D. and Neuner, A. M., 19?8, The end matic environmental hypothesis: Ohio J. Sci., of the Pleistocene in North America: Trans. vol.68, p.251-271. Nebraska Acad. Sci., vol. 6, p.lL7-126. Fay, L. P., 1984, Late Wisconsin Appalachian MeDonald, J. N., 1984, The Saltville, Virginia herpetofaunas: stability in the midst of locality: A summary of research and field trip change: Ph.D. Dissertation, Michigan St. guide: Charlottesville, Virginia, Division of Univ.,68 p. Mineral Resources, 45 p. Guilday, J. E., 1962, The Pleistocene local fauna Peterson, O. A., 1917, A fossil-bearing alluvial of the Natural Chimneys, Augusta County, deposit in Saltville Valley, Virginia: Ann. Virginia: Ann. Carnegie Mus., vol. 36, art. 9, Carnegie Mus., vol. 11, p.469-474. p.87-122. Ray, C. E., Cooper,8.N., and Benninghoff, W.S., Guilday, J. E., Parmalee, P. W., and Hamilton, 1967, Fossil mammals and pollen in a late H. W., 1977, The Clark's Cave bone deposit Pleistocene deposit at Saltville, Virginia: J. and the late Pleistocene paleoecology of the Paleontol., vol. 41, p.608-622. central Appalachian Mountains of Virginia: Richmond, N. D., 1963, Evidence against the Bull. Carnegie Mus. Nat. Hist., no.2,87 p. existence of croeodiles in Virginia and Mary- Hibbard, C. W., 1960, Pliocene and Pleistocene land during the Pleistocene: Proc. Biol. Soe. climates in North Ameriea: 62d Ann. Rept. Washington, vol. 7 6, p. 65-68. Michigan Acad. Sci., Arts, Letts., p. 5-30. Slaughter, B. H., 1967, Animal ranges as a clue 1970, Pleistocene mammalian local faunas to late-Pleistocene extinctions, iro Martin, P. from the Great Plains and Central Lowlands S., and Wright, H. 8., eds., Pleistocene ex- -, provinces of the United States, ,i.n Dort, W., tinctions, the search for a cause: New Haven, Jr., and Jones, J. K., Jr., eds., Pleistocene and Yale Univ. Press, p. 155-167. Recent environments of the eentral Great 1975, Ecologieal interpretation of the Plains: Spec. Publ. Dept. Geol. Univ. Kansas, Brown Sand Wedge local fauna, in Wendorf, No.3, p.395-433. F., and Hester, J. J., eds., Late Pleistocene Hibbard, C. W., and Taylor, D. W., 1960, Two environments of the southern High Plains: late Pleistoeene faunas from southwestern - Publ. Fort Burgwin Res. Cent. Ranchos de Kansas: Contrib. Mus. Paleontol. Univ. Mich- Taos, New Mexico, p. L79-192. igan, vol. 16, no. 1, p. 1-233. Slaughter, B. H., and Hoover, B. R., 1963, The Holman, J. A., 1976, Paleoelimatic implications Sulphur River Formation and the Pleistoeene of "ecologically incompatible" herpetological mammals of the Ben Franklin local fauna: species (late Pleistocene: southeastern United Jour. Grad. Res. Cent., So. Methodist Univ., States): Herpetologica, vol. 32, p. 290-295. vol. 31, p. 132-148. 1980, Paleoclimatic implications of Pleis- Stuart, A. J., 1979, Pleistoeene oceurrences of the tocene herpetofaunas of eastern and eentral European Pond Tortoise (Emus orbicularis -, North Ameriea: Trans. Nebraska Acad. Sci., L.) in Britain: Boreas, vol. 8, p. 359-371. vol. 8, p. 131-140. Taylor, D. W., 1965, The study of nonmarine 1981, A review of North American Pleis- mollusks in North America, inWright, H.8., tocene snakes: Publ. Mus. Michigan St. Univ. and Frey, D. D., eds., The Quaternary of the -, Paleontol. Ser., vol. 1, p. 261-306. United States: Princeton, Princeton Univ. Holman, J. A, and McDonald, J. N., 1985, A late Press., p. 597-611. Quaternary herpetofauna from Saltville, Vir- Webb, S. D., 1974, Chronology of Florida Pleis- ginia: (in press) Brimleyana. tocene mammals, in Webb, S. D., ed., Pleis- Lundelius, E.L., Graham, R. W., Anderson, E., tocene mammals of Florida: Gainesville, Guilday, J., Holman, J.A., Steadman, D. W., Univ. Florida Press, p. 5-31. PUBLICATION 75 43

QUATERNARY VERTEBRATE LOCALITIES OF VIRGINIA AND THEIR AVIAN AND MAMMALIAN FAUNA

Ralph Eshelmanr and Frederick Grady2

"The varying degrees in our understanding of the evolution of the Late Wisconsin fauna is a function of the geographie and chronologic density of the fossil loealities. The known faunas are not evenly distributed spatially or temporally but are clustered in those areas and time periods on which intensive researeh programs have been focused. Thus, the late Pleistocene [and Holocene] faunas of some areas and times are better known than those ofothers" (Lundelius and others, 1983, p. 311).

CONTENTS

Page

Depositional environments and Virginia Quaternary vertebrate loealities .... 49 Distribution of Quaternary verteb

Quaternary environmental reconstruction based on mammalian fossils ...... 64 Disharmonious faunas and their possible causes ...... 65 Age interpretation of Virginia's Quaternary localities ...... 66

ILLUSTRATIONS

Figure

2. Lacation of avian and mammalian sites in Virginia .. 49 TABLES

1. Virginia localities from which Quaternary birds and mammals have been identified ...... 44 2. Quaternary depositional environments where bird and mammal fossils have been or

3. Taxonomic list of Quaternary birds and mammals of Virginia showing distribution of

4. Comparison of the terrestrial mammalian fauna of Virginia in the late Pleistocene and in the 16th Century, with inferred feeding habits ...... 61

rCalvert Marine Museum, Solomons, Maryland, 20688, and Researeh Associate, Department of Paleo- biology, National Museum of Natural History, Washington, D. C.,20560.

2Department of Paleobiologf,, National Museum of Natural History, Washington, D. C., 20560. 44 VIRGINIA DIVISION OF MINERAL RESOURCES

ABSTRACT Virginia that have been studied or are under study, to establish the late Pleistocene and 16th Century There are 77 Quaternary vertebrate localities in faunas compiled from these localities, and to discuss the state of Virginia from which 100 taxa of birds the zoogeography, age, and environmental reeon- and 93 taxa of mammals have been identified. struction based on the evidenee provided by these While many loealities are represented by single faunas. The term "16th Century fauna" as used here specimens, there are several significant local fau- refers to the fauna that existed in Virginia, just nas containing numerous specimens representing prior to the arrival of the first European settlers both marine and terrestrial environments. Karst in the Middle Atlantic region. features, such as caves and fissures, account for The number of Quaternary localities (published 74 percent of the sites. The depositional environ- and unpublished) in Virginia reported in this paper ment, age, and distribution of Virginia's Quater- is 77. These sites are listed in Table 1, which also nary vertebrate localities are discussed, and Vir- provides information about their location, deposi- ginia's late Pleistocene and 16th Century tional environment, agent(s) of accumulation, age mammalian faunas are identified and compared. and method of dating, and sources of data on them. Figures 1 and 2 show the locations of these sites. INTRODUCTION We follow the formatof Lundelius and others (1983) to facilitate comparison with records for the rest Early European settlers in Virginia occasionally of the United States. The Quaternary amphibian found Quaternary vertebrate fossils. In 1686 John and reptilian fossils from Virginia have been stud- Banister exhibited large bones and teeth from the ied by Fay (1984; and this volume) and Holman interior of Virginia to John Clayton (Goode, 1901; (this volume). Simpson, L942,1943; Ewan and Ewan, 1970). Os- The following institutional designations are used born (1939) noted that Samuel Maverick in 1636 throughout: ANSP, Academy of Natural Sciences reported fossil bones from 18 feet below the surface of Philadelphia; CMNH, Carnegie Museum of Nat- of the James River 60 miles above its mouth: ural History; NJSM, New Jersey State Museum; however, these specimens were probably Miocene RU, Radford University; and USNM, United in age. Over one hundred years later, Thomas States National Museum, Department of Paleobi- Jefferson (1787) noted proboscidean remains from ology. the now well-known Saltville locality near the North Fork of the Holston River (Ray and others, DEPOSITIONAL ENVIRONMENTS AND 1967). VIRGINIA QUATERNARY VERTEBRATE During the 19th Century, fossil finds consisted LOCALITIES for the most part of isolated specimens, some of which are preserved in the U. S. National Museum Table 2 is a list of depositional environments from of Natural History (USNM) collections. The first which Quaternary birds and mammals are known record of a marine mammal of Quaternary age or could be expected to be recovered. This table from Virginia was recorded in a scholarly publi- includes all localities in Table 1, listed by their cation by Mitchill and others (1827). Cope (1867) provided the first published record of Quaternary vertebrates from a cave within the present borders of the state of Virginia. However, it was not until the middle of the 20th Century that systematic searches for fossil vertebrates were made in Virginia. Two institutions, the Carnegie Museum of Natural History (CMNH) and USNM, made concerted efforts to recover Quater- nary vertebrates. Even so, most of the fossils that have been collected remain unstudied and the COASTAL PLAIN localities unpublished. Kurt6n and Anderson's (f980) Ple'istocene Mammals of North America noted only three significant localities in Virginia Figure 1. Map of Virginia, showing the three major and made passing reference to four others. physiographic regions of Virginia and the counties The purpose of this paper is to record the avian (shaded) from which Quaternary birds and mam- and mammalian fossil sites of Quaternary age in mals have been recovered. PUBLICATION 75 45

Table 1. Virginia localities from which Quaternary birds and mammals have been identified.

Depositional Agen(s) of Age and Method Notes and Local Fauna lpcation Environment Accumulation of Dating References

1. Abingdon Washington Co. Salt spring? Natural accumu- Late Pleistocene, Prout, 1860; Hay, 1923 lation extinct taxon 2. Arcadia Cave #1 Botetourt Co. Cave Owl Early Holocene, CMNH coll. and #2 based on faunal unpublished data assemblage 3. Baek Creek Cave #1 Bath Co. Rockshelter, Owl Late Pleistocene/early Guilday and others, (aka Cook Cave) owl roost Holocene, based on 1977; CMNH coll. faunal assemblage unpublished data 4. Back Creek Cave #2 Bath Co. Rockshelter, Owl Late Pleistocene/early Guilday and others., (aka Sheets Cave) owl roost Holocene, based on 19ZZ; Jones and, faunal assemblage others.,1984; CMNH coll. , unpublished data 5. Baek Creek Cave #3 Bath Co. Holoeene CMNH coll. unpublished data 6. Bill Neff Cave Rockingham Co. Holocene? USNM coll. (aka Susan's Big unpublished data Mouth Cave) 7. Blacksburg Montgomery Co. Handley and Patton, t947 8. Cedar Hill Cave Rockingham Co. Cave USNM coll. unpublished data 9. Chesapeake City Late Pleistocene, USNM coll. extinct taxon unpublished data 10. City Point Prince George Natural accumu- Late Pleistocene, Hay, 1923 Co. lation extinct taxon . 11. Clark's Cave Bath Co. Cave Owl Late Pleistocene, Guilday and others, based on faunal 1977; Jones and assemblage others,1984 l?.Cliff Cave#2 Giles Co. Cave Owl Early Holocene, CMNH coll. (Station B) based on faunal unpublished data assemblage 13. Comers Cave Page Co. Cave Recent? Guilday, 1982b 14. Continental Shelf Offshore Ocean Bottom Natural accumu- Late Plesitocene, Whitmore and others, lation extinct taxa 1967: USMN coll. unpublished data 15. Covington Alleghany Co. Natural accumu- Late Plesitocene, Hay, 1923 lation extinct taxa 16. Crab Orehard Cave Tazewell Co. Cave Cultural Holocene, based on Clark, 1979 accumulation faunal assemblage unpublished data and human association 17. Darty Cave Scott Co. Cave Owl Level 1= late CMNH coll. Pleistocene/Holocene?; unpublisheddata level 2-4 = early Holocene, based on

18. Deep creek pit cityor Auuvium Natural accumu- *:ir::il::::' stuckenrath and Chesapeake lation based on laC dates Mielke, 1972; of 18,780 t 630 and USMN coll. 16,830 yr B.P. on unpublished data Manmwt Mnsricannttm, tusk fragment. Wood date 0.3 m above tusk at ' 38,000 yr B.P. 46 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 1. (cont.) Depositional Agen(s) of Age and Method Notes and Iocal Fauna location Environment Accumulation ofDating References

19. Denniston Halifax Co. Late Pleistoeene, Hay,1923 extinct taxon

20. Early's Cave Wythe Co. Cave Mid Pleistocene?, based Cope, 1869a, b; (aka Early's Pit, on faunal assemblage Hay, 1923; Ivanhoe?) Guilday, 1962b; Jones and others, 1984

21. Eclipse City of Marine sedi- Natural accumu- Pleistocene? Olsen and Blow, Suffolk ments lation USNM coll. unpublished data

I 22. Edinburg Fissure Shenandoah Fisbure Late Pleistocene USNM coll. Co. (Illinoian?), based on unpublished data faunal assemblage 23. Edom Rockingham Co. Natural accumu- Late Pleistoeene, Hay,1923 lation extinct taxon 24. Egglesion Fissure Giles Co. Fissure Late Pleistocene, CMNH coll. based-on faunal unpublished data assemblage

25. Ely Cave Lee Co. Natural accumu- Holocene?, based on Allen,1885 (aka Gilley?) lation/cultural human association accumulation 26. Fairview Beach King George Co. Plesitocene, based'on NJSM coll. ' (aka Belvedere) extinct taxa unpublished data 2?. Gardner's Cave Wythe Co. Cave Late Pleistocene, Guilday,1962b; based on faunal Eshelman, USNM coll. assemblage unpublished data 28. Gillespie's Cliff Tazewell Co. Cave Owl Mixed - late CMNH coll. Cave Pleistocene and unpublished data Holocene, based on faunal assemblage 29. Glade Cave Augusta Co. Cave Holocene Eshelman, unpublished data

30. Holston Vista Washington Co. Cave Owl? Late Pleistocene/ CMNH coll. Cave Holocene, based on unpublished data faunal assemblage

31. Hot Run Frederick Co. Spring/ Natural aecumu- Pleistocene, based on Ott and Weems, alluvium lation uC of 11,550 + 165 yr this volume B.P. on Mamtm'dmolar 32. Hot Springs Bath Co. Spring Natural accumu- Late Pleistocene, Hay,L923 lation extinct taxon

33. Jasper Saltpeter Lee Co. Cave Mid Pleistocene Guilday, lg?9; Cave (Illinoian?), based on Jones and others, faunal assemblage 1984; CMNH coll. unpublished data

34. Keyser Farm Page Co. Cultural Late llolocene Manson and others, aecumulation (Woodland Period), 1943 based on faunal assemblage and human association 35. Klotz Quany Cave #5 Giles Co. Cave Owl Late Pleistoeene, CMNH coll. based on faunal unpublished data assemblage

36. Kumis Roanoke Co. Handley and Patton, 194?

37. Lane Cave Scott Co. Cave Late Pleistocene, Holsinger,1967 based on extinct taxon PUBLICATION 75 47

Table 1. (cont.) Depositional Agen(s) of Age and Method Notes and Iocal Fauna Location Environment Accumulation of Dating References

38. Limeton Warren Co. Clark, 1938, 1939 39. Loop Creek Russell Co. Cave Owl Late Pleistocene/early CMNH coll. Quarry Cave Holocene, based on unpublished data faunal assemblage 40. Luray Caverns Page Co. Cave Hovey,1882 41. Meadowview Cave Washington Co. Cave Late Pleistocene, Guilday and Parmalee, based on faunal 1972: Jones and assemblage others, 1984; Klippel and Parmalee, 1984; CMNH coll. unpublished data 42. Metomkin and other Aceomack Co. Ocean beach Natural accumu- Late Pleistocene, Mitchill and others, barrier island lation extinct taxon 1827;Hay,1923; USMN coll. unpublished data 43. Natural Chimneys Augusta Co. Cave, owl Mid-Pleistocene?, Guilday and (aka Brown's Cave and based on faunal Bender, 1960; Cave of the Wooden assemblage Wetmore, 1962; Steps Guilday,1962a; Guilday and Parmalee, 1972: Jones and others, 1984; Fay, 1984; Anderson, 1984 44. New Castle Craig Co. Handley and Patton, L947 45. New Quarry Cave Smyth Co. USNM eoll. unpublished data 46. Newport Giles Co. Handley and Patton, 1947 47. Pembroke Railroad Giles Co. Cave Late Pleistocene, CMNH coll. Cave #1 based on faunal unpublished data assemblage 48. Pembroke Railroad Giles Co. Late Pleistoeene. CMNH coll. Cave #2 based on faunal unpublished data assemblage 49. Paul Penleys Cave Bland Co. Cave CMNH coll. unpublished data 50. Rass Hole Highland Co. Cave Lite Pleistocene. CMNH eoll. (aka Corbett's) exiinct taxon unpublished data 51. Ripplemead Quarry Giles Co. Fissure Owl, and Late Pleistocene/ Weems and Higlins, other natural Holocene. based on 19?7: Jones and aceumulation faunal assemblage others, 1984 52. Ruffner's Cave #2 Page Co. Cave Holocene, based on Grady, USNM coll. faunal assemblage unpublished data 53. Saltville Smyth Co. Alluvial, Natural accumu- Late Pleistocene, based Peterson, 1917; lacustrine, lation on nine raC dates Pickle; 1946, Ray paludal ranging from 27,000 to and others, 1967; 10,050 yr B.P. Guilday and Parmalee, 1972; Guilday, 1982a; McDonald and Bartlett 1983; McDonald 1984: CMNH, RU and USMN coll. unpublished data 48 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 1. (cont.) Depositional Agent(s) of Age and Method Notes and Iocal Fauna I,ocation Environment Aecumulation of Dating References

54. Shenandoah Caverns Shenandoah Co. Cave Grady, USNM coll. unpublished data

55. Shires Saltpeter Craig Co. Cave Late Pleistocene/ CMNH coll. Cave Holocene, based on unpublished data faunal assemblage

56. Skyline Caverns Warren Co. Cave Late Pleistocene, USNM coll. based on faunal unpublished data assemblage

57. Spurey Cave Madison Co. Cave Natural accumu- Holocene?, based on Eshelman, USNM coll. lation faunal assemblage unpublished data 58. Staunton Augusta Co. Fissure Natural accumu- Late Pleistocene, Hay,1923 lation extinct taxon 59. Strait Canyon Highland Co. 'I lssure Natural accumu- Late Pleistocene, based Guilday,1982b; lation on raC date of 29,870 Anderson,1984; +1800/-1400 yr B.P. Fay,1984; Jones and others, 1984; Guilday and others, unpublished data CMNH and USNM colls. unpublished data 60. Stratford Hall Westmoreland Fluvial? Natural accumu- Late Pleistocene, USNM coll. Co. Iation extinct taxon unpublished data 61. Thessalia Giles Co. Handley and Patton, 1947

62. Toms Brook Shenandoah Co. Cave Holocene USNM coll. unpublished data

63. Unknown Clarke Co. Cave? Late Pleistocene, based USNM coll. exact on faunal assemblage unpublished data proveience unknown 64. Unknown Augusta Co. Late Pleistocene, Hay,1923; ANSP coll. exact extinct taxon unpublished data provenience unknown 65. Unnamed Cave Smyth Co. Cave Late Pleistocene/ RU coll. Holocene, based on unpublished data faunal assemblage 66. Unnamed fissure Warren Co. Fissure Carnivore den? Holocene, based on Haynes, 1978 (Archeological site faunal assemblag:e 44WR61) 67. Varner's Cave Highland Co. Cave Gr4dy, USNM coll. unpublished data

68. Vickers Cave Washington Co. Cave Pleistocene Guilday,1971a; (Illinoian?), based on CMNH coll. faunal assemblage unpublished data 69. Virginia Beach (near) Exact Late Pleistocene, USNM coll. provenrence eitinct taxon unpublished data

unknown . 70. Virginia Caverns Rockingham or Cave Holoeene, based on McGill,1933 (near New Market) Shenandoah Co. human association 71. Wade Site Halifax Co. Cultural accumu- Late Holocene Stevens, 1979 lation (Woodland Period), based on faunal assemblage and cultural material PUBLICATION 75 49

Table 1. (cont.) Depositional Agen(s) of Age and Method Notes and Local Fauna location Environment Accumulation of Dating References

72. Warrenton Fauouier Co. Late Pleistocene, Hay, 1923 extinct taxon 73. Will Farleys Cave Washington Co. Cave Owl? Late Pleistocene?, CMNH coll. based on faunal unpublished data assemblage 74. Williamsburg (near) York Co. Natural accumu- Late Pleistocene, Mitchell, 1818; lation extinct taxon Hay,1923 75. Williamsburg (near) York Co. Late Pleistocene, based G.H. Johnson, on extinct taxon unpublished data 76. Winding Stairs Cave Scott Co. Cave Late Pleistocene, based CMNH coll. on faunal assemblage unpublished data 77. Womack Pit City of Marine rip- Natural accumu- Late Pleistocene, based Ray and others, 1968 Virginia Beach tide channel lation on faunal assemblage

Note: fourteen specimens from the USNM colleetion listed in Table 2 are not included here because their exact provenience is unknown. There are also another 17 cave faunas in the CMNH collections for which we have no faunal data and these localities are, therefore, not included in this table.

respective depositional environment, as well as 14 raptor pellets, packrat middens, naturally trapped isolated specimens listed by catalog number for animals, and fluviafcolluvial deposits. Regurgi- which no loeality number has been assigned. Eleven tated owl pellets eontaining the remains of small depositional environments are recorded in order vertebrates collected in the immediate area of the of decreasing frequency of representation at fossil- roost are probably the most important sources of bearing localities in Virginia. Though no fossii cave bone. Natural Chimneys, Clark's Cave, and records are known from eolian deposits such as the Back Creek caves are examples of Virginia beach dunes, they are to be expected. caves that contain signifieant bone aecumulations Caves, fissures, and, to a lesser degree, roekshel- deposited by predatory birds. Caves, other than ters are the sites that have been studied most natural traps, also present a taphonomie bias in systematically in Virginia. These same features that small vertebrates are usually well represented probably provide the most promise for new Qua- whereas large vertebrates are frequently repres- ternary vertebrate localities since Virginia has no ented by one or only a few individuals whose less than 2300 known limestone caves, fissures, and remains are widely scattered over the site (Lun- rockshelters (Holsinger, 1975). Caves commonly delius and others, 1983). contain abundant fossil vertebrates because they Woodrats or packrats are well known for their provide good environments of preservation and habit of collecting odd objects, including bones and accumulate bones from such sourees as roosting- teeth, and often including them in their cave nests. Woodrats may be responsible for the few large bones and teeth from Natural Chimneys. Guilday and Hamilton (1978) suggested similar woodrat accumulation for some West Virginia caves. Strait Canyon and Edinburg Fissure, at least in part, consist of natural trap deposits, while Gardner's Cave and Rass Hole contain bones of large mammals that probably washed into the deposit by natural processes. Other eaves, such as Ely, and rnany rockshelters contain what might be, in part, accumulations attributable to human activities.. Springs are known to attract vertebrates (Shiner, 1983). The marshy environment common to many springs may serve as a trap, and at springs, prey Figure 2. Map of Virginia, showing the loeation of s'pecies may be. attacked by predators which have sites listed in Table 1. staked out such strategie locales. Salt licks,such 50 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 2. Quaternary depositional environments where bird and mammal fossils have been or might likely be recovered in Virginia.

Isolated Specimens not Depositional Environmentr Representative Iocalities2 given Locality Numberss A. Terrestrial Caves (42) 2, 4,5,6,9, tl,12,13, 16, 17 USNM 1645 MUIohVtts nasutus 20,25,27,29, 29, 30, 33, 35, USNM 1820t Equussp. 37, 38?, 39, 40, 41, 43, 45,47, 48, 49, 50, 52, 54, 55, 56, 57, 62, 63?, 65, 67, 68, 70, 73, 76 Fissures (6) 22,24,51,58, 59, 66 Rockshelters (2) 3,4 Springs (including hot and 1?, 31, 32, salt springs) (3) Fluvial (streams and river deposits) (4) 18,26?,53,60? Lacustrine or paludal deposits (1) 53 Eolian (sand dune, loess deposits) (0)

Marsh or peat (1) 104 USNM 8808 Mammuf a,mer.conncrn (swamp) Archeological context (4) L6,25,34,71 USNM S2lConissp. B. Marbw Shallow sea bottom (3) 14,21,77 USNM 23787 Synbos sp. USNM 187576 Mamm.ut empricannwn USNM 214824 Manwwt amsrinawrn USNM 2L49 43 M ammut amprinanum USNM 243748 Bison'sp. USNM 2437 53 Mammathus sp. USNM 245926 M omm:u,thrs sp. USNM 243927 Momm,uth,us sp. USNM 2997 ll M ammut amprinanum Beach (barrier island) (2) 42,69?

C. Unl*wum(L3l) 7, 9, 15, lg, 23, 36, 44, 46, USNM 336566 Artiodactyl vertebra 61,64,72,74,76

r Number of included sites shown in parenthesis. 2 Numbers refer to localities listed in Table 1. 3 Depositional environment known but exaet provenience unknown. a Seb Hay(1923, p.352)

as the one at Saltville and thermal springs such Virginia's barrier islands have produced a few as those at Hot Springs are highly attraetive to fossil vertebrates, including walrus (Odoberrus) and, vertebrates, especially mammals. Salt is an attrac- giant beaver (Castoroid,es). The latter (USNM tant in its own right while thermal springs may 215021) is believed to have washed from a drowned provide refuge and water during harsh winters. peat deposit just offshore of Metomkin Island. Classic examples of such environments are Big Clam, scallop, and bottom-fish dredge-boats eom- Bone Lick, Kentucky and the geyser basins of monly pull up fossil mammalian bones from the Yellowstone National Park, Wyoming. Reeves continental shelf (Whitmore and others, f967). To (1932) noted more than 90 thermal springs from date, bison (Bison), musk ox (Sgmbos), mammoth northwestern Virginia and adjacent parts of West (Mamrnuthus), and mastodon (Ma.,mmut) remains Virginia; water temperatures range up to 1050 F have been dredged from the continental shelf. in these springs. Eshelman recovered an artiodactyl vertebra frag- PUBLICATION 75 5t

ment (USNM 336566) in water 142 feet deep just of ten taxa. MacCord (1973a) reports deer and off the Cherrystone Channel in Chesapeake Bay, turkey from the Hirsh Site, and elk, deer, bear, Virginia. Such finds, no doubt, date from full fox, raecoon, and wolf (dog?) from the Hidden glacial periods when sea level was lower and the Valley Rockshelter (MacCord, 1973b). From Bath continental shelf and the bottom of what is now County alone, Clark reported 32 prehistoric Indian Chesapeake Bay were exposed. Borrow pits within sites (Guilday and others, 1977,p.78). The potential the Coastal Plain in places reveal marine and for studies ofvertebrates in such areas is rich. Clark fluvial deposits containing Quaternary vertebrates (1979) reported 26 known Indian burial caves and (Ray and others, 1968). 40 known prehistoric roekshelter sites in Virginia. Finally, Semken (1983) points out the potential of archeological sites for increasing information ZOOGEOGRAPHY about the history of Holoeene vertebrates (Rose, this volume). The growing interest in fossils and con- Our discussion on zoogeography will be limited tinuing construetion activity within Virginia no primarily to the mammalian fauna. Virginia falls doubt will result in the discovery of additional within the Symbos-Ceraalces Pleistocene Faunal Quaternary fossil localities in the future. Province of Martin and Neuner (1978) characterized by the presence of. Sumbos, Cerualces, Casto- DISTRIBUTION OF QUATERNARY roid,es, and Sangamona. Graham (1979) also rec- VERTEBRATE LOCALITIES ognized Pleistocene faunal provinces and included Virginia as a part of the Eastern Mammalian For the present discussion Virginia is divided Region, which is characterized by Blari,na sp., into three physiographic regions: the Coastal Plain, Tuxnias striatus, Synaptomys cooperi, Castoroid,es the Piedmont, and the Appalachian Mountains olui,oensis, Sambos sp., Bootherium sp., and, MgIo- (Figure 1). Quaternary fossils have been recovered Ir,yus sp. from the Coastal Plain and especially from karst It is diffieult to make paleozoogeographical and features in the Appalachian Mountains. Although paleoenvironmental comparisons between late Hay (1923, p. 352) reported that "none are known Pleistocene and Recent mammalian faunas for an anywhere in the Piedmont Plateau," he did note area as large as Virginia. Virginia has 40,815 an unidentified horse tooth from Denniston, Hal- square miles of land ranging in elevation from sea ifax County (Table 1: locality 19), in the Piedmont. level to over 5700 feet. The maximum dimension Eaeh county from which Quaternary birds and of the State is over 500 miles, ranging from open mammals have been recovered is shaded in Figure ocean to mountains. The late Pleistocene fauna 1. Of Virginia's 95 counties and 41 independent spans perhaps 500,000 years, assuming the Illin- cities, 32 eounties and 4 eities are known to have oian fauna is represented in known colleetions, yielded Quaternary fossils. There are 65 localities while the 16th Century fauna spans only 100 years. in the Appalachian Mountain region, 11 in the Yet, by contrasting the late Pleistocene and Eu- Coastal Plain and 1 in the Piedmont region. Nearly ropean contact (16th Century) faunas of Virginia every county within the Appalachian Mountains and by looking at extirpated forms as listed in Table from which caves are known, has yielded fossils. 3, some general trends can be discerned. Late Only a few eounties in the Coastal Plain have so Pleistoeene taxa whose Recent distributions, based far produced Quaternary vertebrate fossils, and on Hall's (1981) work on mammals and Robbins' only one county in the Piedmont has done so. and others (1966) work on birds, do not range into Though fossil records are rare from the Coastal Virginia today and are considered extirpated. Guil- Plain and Piedmont, the potential is great for day (1982a) lists four categories for the late Pleis- adding to the Quaternary vertebrate history of tocene fauna of the mid-Appalachian region: (1) these regions through the discovery of new local- extinct species, (2) present-day boreal forms, (3) ities. Certainly there are more sites like the Wom- present-day midwestern forms, and (4) species still ack borrow pit (Ray and others, 1968) to be found. living in the area. Though it was not the intent of this paper to do Extirpated forms which now live in more north- a systematic search of faunal studies from archeo- erly areas inelude: sharp-tailed grouse (Ped,iocetes logical sites, additional data about the presence and phasianellus), spruce grouse (Dend,rogapus cuna- distribution of prehistoric vertebrates in Virginia d,ensis), gray jay (Perisoreu,s canad,ensis)' magpie can be gleaned from the study of such localities. (Pi,ca pica), rock ptarmigan (La,goptus cf. m,utus)., Keyser Farm (Table 1: locality 34; Manson, Mac- pine grosbeak (P'inicola ef. errueleator), glaueous Cord, and Griffin, 1943)yielded an associated fauna sull (Laru,s ltEperboreus), common murre (Uri,a 52 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 3. Taxonomic list of Quaternary birds and mammals of Virginia showing distribution of species by locality.

Scientific Name Common Name Iocality number (s)

AVES BIRDS

Gaviiformes loons

l. Gaaiasp. Ioon sp. unidentified 2l

Podicipediformes grebes

2. Po d,iLEmUus p o d'i.cep s pied-billed grebe 11, 59

3. Podiceps aurifus horned grebe 44

Pelecaniformes pelicans, etc.

4. Morus bassarrus gannet 2L,77

5. P ha\acr o cor an au,ritu s double-crested cormorant 77

Ardeiformes herons

6. B otaums lenii,gimsus American bittern 1l

Anseriformes waterfowl

7. OIar columbiomts whistling swan 34

8. Brarrtabenvi,cla brant 77

9. Ano,s cf. A. platgrhynclws or mallard or black duck 11 A. ruhripes

10. Anos discors blue-winged teal 43 cf. Anas d:iscors 1l ll. cf. Anq,s crecca green-winged teal 11

12. hrcephala albeola bufflehead 43 cf. Bu,cephnLasp. 59

13. Lophod,Ates at cullalus hooded merganser 11 cf . Lophod,ytes arcullatus 59 t4. Mergus sp. merganser 11

15. O rg ur a j o,mni,cmsis ruddy duek 11 Anatid ducks, unidentified L2,59

Accipitriformes diurnal raptors

16. Cathartes atna turkey vulture 34

17. Accipiter cooperi,i cooper's hawk 3

18. Accipiter striatus sharp-shinned hawk 11, 43, 59

19. Buteo lineahts red-shoulder hawk 43 Buteo cf. B. lineatus 59

20. Buteo jomaicmsis red-tailed hawk 43 cf. Buteo jamuicensis 59

2I. Buteo pLatgptnrus broad-winged hawk 43 Butuo cf. B. platEpterus 11

22. Haliaeetus Leu,cocephalus bald eagle 34 PUBLICATION 75 53

Table 3. (eont.).

Scientific Name Common Name locality number (s)

23. Faleo spamerius American kestrel ll Accipitrid hawks, unidentified 11

Galliformes gamebirds

24. C o lin us uir gbai,a,nu"s bobwhite quail 4, 11, 43, 59

25. #cf. Dend,ra,gayus catnd,msis spruce grouse 4, tt, 43,59

26. Bonosaumbelhrc ruffed grouse 4, 11, 43, 59 cf. Banasaumbellus 3 D. canaderwis or B. urnbellus 4, 11

27. #Logopus cf. mnrtus rock ptarmigan 4, tL #Logoptts sp. ptarmigan 4

28. # P ed;io cete s phasionc ll:ns sharp-tailed grouse 3,4,43 ef . # P edio cete s phasion elhts 4, 11 Tetraonid grouse, unidentified ll

29. Meleo,gris gallopauo wild turkey 11, 30, 34, 43, 71 cf.. M eleagris g allop aa o 59

Ralliformes marsh birds

30. #Cirus americena whooping crane 43

31. Rolhw l;irwicola Virginia rail 59 cf. RaIIus limi,cola 1l

32. Porzana carol:ina sora 11

33. Gall;i,tt uLa chloropus common gallinule 11

Rallid rails, unidentified 11

Charadriiformes shore-birds, gulls and auks

34. Pluuiolis dutwi,wica American golden plover 1l

35. C har ad.riu,s a ociferu s killdeer 43

36. cf. Actitis mnm,larta spotted sandpiper 11, 59

37. Eroliarnimti,lla least sandpiper 43

38. C o,pella gallina,go common snipe 11

39. Philohelam:hwr American woodcock 11, 43, 59

40. cf. Tri,nga solijtartn solitary sandpiper 11

41. Bartramia longiaauda upland plover 43

42. cf. Limnsa,? sp. godwit 11

43. C aptoptropharus sernipalrlatus willet 43 Scolopacid sandpiper/plover unidentified 11

44. #Larus Werbareus glaucous gull 77

45. + Pinwiwus impennis great auk 77

46. #Uria aal,ge common murre 77 Alcid auks, unidentified 2l 54 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 3. (eont.).

Scientific Name Common Name Iocality number (s)

Columbiformes pigeonlike birds

47 . + E ctopistes migratoriu"s passenger prgeon 3, 4, 11, 43,59

Cueuliformes cuckoos

48. Coccgzus sp. cuekoo, unidentified 1l

Strigiformes owls

49. Otus asin screech owl 4, Ll

50. Buho vi,rginianus; great horned owl 11, 59 Stria? or Bubo? owl 3

51. Asio cf.. A. flammeus or A. otus short-eared or long-eared owl; 1l cf. lsio sp. eared owl unidentified 59

52. Aegolius aaad;ims saw-whet owl 4 Strigid owls, unidentified 4

Caprimulgiformes goatsuckers

53. C hord,ei.Ies rwinor common nighthawk 11

54. Clweturapela,gica chimney swift 1l

Coraciiformes rollerlike birds

55. MegonuVle alcyon belted kingfisher 11,43

Piciformes woodpeckers

56. Colaptes auralus flicker 4, 11, 59 cf.. C olaptes aural,tts 4,43

57 . Dtgocopus pi,lnofas pileated woodpecker 11

58. Melonerpes carolinus red-bellied woodpecker 43 cf.. Melanapes earol'i,mn l1

59. Melonerpes erytlvocephalus red-headed woodpecker 4,43

60. Sp h.g r o,pims u arins yellow-bellied sapsucker 11

6L. Dmd,racopos ttillosu,s hairy woodpecker 4,tL,43

62. Dmd,rocopos puheEcens downy woodpeeker 11 Picid woodpecker, unidentif ied 4,Ll

Passeriformes perching birds

63. Swomisphoebe eastern phoebe 43

64. Empiflonaasp. flycatcher 11

65. Conf,optts uirms eastern wood pewee 43

66. E renwphi.Ia alp e stri s horned lark 11

67. Petrochelidpn pWrhmwta cliff swallow 11,43

68. # Perisoreus canodenci^s grayjay 11, 43

69. Cuonaai,tta aristufn blue jay 4,11,43 cf. Cganoc'ifra cristufn 11 PUBLICATION 75 oo

Table 3. (eont.).

Scientific Name Common Name Iocality number (s)

7Q. Conrus Ura,cWhwwhos common crow 11 cf. Conrus sp. crow 3, 59

71. #Pica pi,ca black-billed magpie 43,59

72. Pontssp. chickadee, unidentified 11

73. Parus lvi,colw tufted titmouse 11

74. Sitta canadmsis red-breasted nuthatch 11,43

7 5. C erthi,a famil;iaris brown creeper 11

76. Cistotharu,s cf. C. platznsis short-billed marsh wren l1

77. Torostama rulum brown thrasher 43 cf. Toxastoma rufum 11

78. Cothonts sp. thrush 11

79. cf.. Sialia sialia eastern bluebird 11

80. Twdus rnigraf,eri,us robin 11, 43

81. Hsloeinhlasp. thrush 43

82. ef.. Anth.us Witwl,etta water pipit 11

83. Bom,Aacilla cedrqrum cedar waxwing 11

84. cf.. Dmd,roeia coronata myrtle warbler 11

85. Seiums sp. waterthrush 1l

86. V ernviu or a chry s optnra golden-winged warbler 4 Parulid warblers, unidentified 11

87. Dolinlumgn bobolink 11

88. cf.. Sturttellasp. meadowlark 1l

89. Agelahw phoewineus red-winged blackbird 11,43, 59

90. Molotlurus atnr brown-headed eowbird 43

91. M elosp'iaa mel,od.i.a song sparrow 43

92. Zonotrinhia albicollis whitethroated sparrow 43

93. ef.. Icterus sptri,us orchard oriole t1

94. Pirangasp. tanager 11

95. ef.. Carpod,asus Wrpurqw purple finch 3 96. cf. #Piwicola marleator pine grosbeak 4,lL

97.l'oriasp. crossbill 4, tl

98. Pooecetes gromi,nats vesper sparrow 11

99. Junco lryemahis slate-colored junco 4,tl Jurwo sp. junco 43

100. Passerellailiaca fox sparrow 43 Fringillid sparrows, unidentified 11 Passeriform passerines, unidentified 4 Aves birds, unidentified 3,4,17,20,24,28,35 56 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 3. (cont.).

Scientific Name Common Name Iocality number (s) MAMMALTA MAMMALS

Marsupialia pouch-bearing mammals

l. Did.elphis uirgriniana common opossum 34,36,44,46,66,7r Inseetivora shrews and moles 2. #Saren wct:icus Arctic shrew 4, 11, 30, 43

3. Sorer cineretn masked shrew 3, 4, 11, 30 Swer cf. S. cinpreus 43 Soren c,inereus or S. lonEirostris masked or southeastern shrew 59

4. Sorer d:ispar long-tailed shrew 4, 11, 30, 59

5. Sorerfumeus smoky shrew 3, 4, 11, 30,43, 59 Sorer ef. S, fumeus 33

6. Sorer palwtris water shrew 11, 30,43 Sorec sp. tt,12,24,28,30,35,47

7. Microsorer hayt pygmy shrew 11, 30, 35,43, 47

8. Blarina Ureticaud,a short-tailed shrew 3, 4, 11, 30, 33,43,59 Blarina cf. B. fuwicaud,a 20 Blarina sp. t2, 17,22,24,30,35, 41, 47, 56,73

9. Cryptotis paraa least shrew 43 Soricid shrews, unidentified 8,22,39 t0. Par as c alop s Ar eu eri hairy-tailed mole 3, 4, 11, t2, 17, 30,33, 43, 59, 73 ll. Scalopu,s a4ntizus eastern mole 4, 11, 43,66 Scalnpu^s sp. 30

12. C ond,E lura cristata star-nosed mole 3, 4, tL,30, 33, 39, 43, 59 Chiroptera bats L3. Mgoti^s htcifugus or M. sod,alis little brown bat 4, tL Mgotis cf. M. lucilugs 30,43

14. Myot:is lcemii Keen's bat 11,43 Myotis kemii or M. grrisescerx 4,30

15. MEot:is gtrisescens gray bat 4

16. MEtois leibii small-footed bat 4, Ll Myotis sp. 3, 4,30, 33, 35, 54

17 . Pipistrellus subfl atrus eastern pipistrelle 11, 33, 43, 59, 6? Pipistrellus sp. 4

18. L asion g ctnris nn et'in ag ans silver-haired bat 51

19. Eptesicu,s fuscus big brown bat 3,4, 11,30, 33 Eptes'inus cf. E. fuscus 43 Eptesicus sp. 12, 30, 35, 43

20. Pl,ecotus ef. taumsmdii big-eared bat ll Plecotus sp, 4, 33 21. Losiurus borealis red bat l1 Las'ium,s ef. L. bmealis 43 Chiropteran bats, unidentified 6, t2, 20, 22, 40, 47, 56, 70 Edentata armadillos and sloths 22. + M eg alang u j eff er s anii Jefferson's ground sloth 37 +Megalanar cf. M. jeffersonii 20,53 PUBLICATION 75 D'J

Table 3. (cont.).

Scientific Name Common Name Iocality number (s)

23. #Dasapus cf.*D. bellus beautiful armadillo 4t

Lagomorpha hares and rabbits

24. #Ochotonasp. pika 33

25. cf. Sglui.la.gu.s transitilyn alis New England cottontial 11

26. Sa luilo,gu,s floridantus Eastern cottontail 34,44,46 Syhtilo,gus sp. cottontail rabbit 43,52,54,66 cf. Sgluila,Ns sp. t2

27, Lepsuwricmrus snowshoe hare 4, tt,22,43 Lepus sp. 30 Leporid hares and rabbits, unidentified 3, 4, 8, 16, L7, 27, 28, 30, 95, 4L, 43, 47 , 48, 57 ,62, 63, 73 Rodentia rodents

28. To.mias striatus eastern chipmunk 3, 4, 11, 33,43, 54 Tatnias cf. T. strintu,s 20,30 Tamios sp. chipmunk, unidentified t2

29. #Eutamios mini.mns least chipmunk 11 Eutomias sp. 4

30. Marmatamunan woodchuck 3,4,6, 11,30,33, 34, 43,44,63,66 Mo,rm,ota cf. M. mana.r 20 Mo,rmntasp. 12, L7,24,28,35,41,76

31. # Spermnplyilus trid.ecemlineatus thirteen-lined ground squirrel 3,4, 11, 30, 43 Sperm,ophilus sp. 30 cf.. Spermophihssp. 39

32. Sciurus carol:inensis gray squirrel 33, 34, 43 Sciurus ef. S. carolirwnsi,s I4 Sciunts carolinens'is or S. niger gray or fox squirrel 3,30 Sciurus sp. 17, 30

33. Tamiasc'iunts hudsonicus red squirrel 3,4, Lt, t2,33 T amios ciunt s huds unicus Pleistoeene red squirrel 43 *cf. T. h. tsnd.uid,ens

34. Glaucomys aolans southern flying squirrel 11, 43 Glaucamgs cf. G. aolans 3,4,t7,20,30

35. Glau,comy s saArirlus northern flying squirrel 1t, 30, 43 Glaucom.y s cf . saUrinus 3,4 Glaucomgs sp. flying squirrel, unidentified 3, L2,20,28,33,47 Tamias or GlaucornEs squirrels, unidentified 11

Sciurid squirrels, unidentified 2, t7 , 22,28, 35, 39, 4t, 47 , 48, 63, 73

36. Castor canadm,sis beaver 3, t7, 20, 22, 34, 43, 46, 5t, 62

37 . + C astor oide s o hi.o ens'is giant beaver 42,43 Castorid beaver, unidentified 47

38. Per amE s cus man icul ahts deer mouse 3,4,lt,43 Perom.yscns ef.. P. maniculatus 43,59

39. Perom,E scus leunopn s white-footed mouse 3, 4,7 , ll Perom,gscns cf.. P. leu,copw 30,43

Peromqscu,s sp. 3, 4, 6, 12, t7 , 22,24,35, 43, 59, 62, 63, 67 , 73 58 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 3. (cont.).

Scientific Name Common Name Ioeality number (s)

40. Neotmnaflorid,ana eastern woodrat 9,4,ll, B0, gg, 48, bg, 6? Neotorna ef. N. florid,ana Z0 gb, Neotama sp. woodrat unidentified z, 4, 12, L7 , zz, 24,29, 99, 47 , 49, 62, 79, 76

41. Clethrionom..gs gapperi red-backed vole 9,4, ll,12,17,24,30, BB, 85, 99, 4L,49, 47, 5l-, 59, 73

42. Pheno,cam.gs intermed,ius heather vole B, 4, 11, bg Phma.com.gs cf. P. intermedius 4B Phm,acurnEs sp. L7, g0, gg,4l 43. Minrotus pmnsylaanicus meadow vole g, 4, ll,1g, 24, 90, gB, 4g M'inrotus cf. M. pmnsylaan'inus Z0,4l, Eg 44. Mi,crotus chroturrhintrc rock vole 3,4, Lt,24,30,89,49 M. pmnsElaaniarc or M. chrotanhinus 3, 4,17,33, 3b, 43

45. #Microtus uanthognathw yellow-cheeked vole B, 4, 11, 22,28,90,49,69 46. Minrotu,s pirutarum woodland vole g, 4, LL,67 Mi'crohrc p'inntsrum or M. ochrogsatnr woodland or prairie vole L7,24,30,33, 39, 41, 43, 47,59 M'i*rotus sp. voles, unidentified 6 ,8, 12,22,28, 30, 33, 35, 47, 5G, b7, 62, Gg

47 . #Neofiher *leonard,i extinct round-tailed muskrat 59 Microtid microtine, unindentified z

48. Ond,atra zibethinus muskrat 3, 11, 29, 34, 43, 59,7I Ond,alrasp. l7,z\,5g

49. funaptamEs cooperi southern bog lemming 3, 11, 17, 24,30,35,35, 39, 43, 4?, 59 Synaptornys cf. S. cooperi 41 b0.#Synaptangsberealis northernboglemming 3,4, 11, 12,24,33,43,59 Sytnaptomys ef.. S. borealis 30 Sgnaptomgs sp. bog leming, unidentified 28, Bg, gb

51. Zapus hu.dsanims meadow jumping mouse 11, 43, 59 Zapus sp. jumping mouse, unidentified 80, BB 52, Napa,eozapts in*ignis woodland jumping mouse 11,24,43,59

53. Eretlvizan,dorsaftu,m porcupine 11, 17,30,43,59,66

Cetacea whales and porpoises

5 4. # E s chrinhli.us g laum,s gray whale 42 Cetacean whales and porpoise, unidentified 77

Carnivora flesh-eaters

55. Cmvis cf.*C. d,irus dire wolf 11

56. Can;is lupus gray wolf 43, 51

57. Cafti"s latrans coyote 22

58. Canisfamilitris domestic dog 34 Conis sp. dogs, undentified 3,22,26,40,66 59, Urocyon ei,nereoargenteus gray fox . 34,51,66

60. Vulpes rrulpes American red fox 4 Vulpes sp. 43 Canid dogs and foxes, unidentified 25,35,47,59,66 PUBLICATION 75 59

Table 3. (cont.).

Scientifie Name Common Name Locality number (s)

6L. Ursus a:rnnricqnus black bear 3,6, 11,22,27,34 Ursu,s cf.. U. amcrbanus 43 Ursas sp. 20,26,30,38,4r,44,55 Ursid bears, unidentified 35, 40, 48

62. Procyon lotor raccoon 3, 11, 34, 43, 46, 52 54, 66 Procyancf, P. Iotar t7,20,30,73

63. Martes americann marten tL, L7, 43 Martes cf. M. am.ericana 39

64. Martes pmnart:i fisher 3, L7,51 Martes cf. M. pmnanti 43

65. Mustelarinosa least weasel 3, 4, 11, 43

66. #Mustela emwinea ermine 11

67. #Mustnlafrmata long-tailed weasel 34 Mustela cf. M. frmata 43 Mustela errwinpa or M. frenala 11

68. Mu,stela visun mink 4, tL, L2, t7,43 Mustela sp. weasels, unidentified L2

69. Mephitis mnptuitis striped skunk 8,11,34, 43,44,46,66

7 0. Spilogale put oritts eastern spotted skink 20,5t

71. Irutra canodnnsis river otter 34

72. +Puntlwra? or Fel,is*atror indeterminate large cat 2,38,40

73. Felis cannolor mountain lion 34

74, I'an* rufus bobcat 34 Lww? sp. indeterminate small cat 17, 39,43

7 5. # O dnb m:us r o srnal.'tls Atlantic walrus 42,77

7 6. Hal;iclnerus grvpus gray seal 77 Carnivorans earnivores, unidentified 20,77

Proboscidea mastodonts, gomphotheres and elephants 7 7 . + Mam.mnt o:tn eri.canum american mastodon 1, 10, 14, 15, 18,23,31,32,53, 59,65,74

7 8. + M o,m,mathus pri,mi g mhrc woolly mammoth 14, 53 +Mamrruthussl,. mammoths, unidentified t4,72

Perissodaetyla odd-toed hooved herbivores * 79, E qws carryli.c,atus complex-toothed horse 1, 59 Eqws cf.. * E, cam.plicatus 20

80. Equus cf.*8. fratemus brother horse 31 Equussp. horses, unidentified 9, 18, 19, 20,5t,53,58 60, 65, 73

81. +Tapiras cf.*T. aeroensis vero tapir 59 Tapirus sp. tapirs, unidentified 20,22,38,5r,73,76

Artiodactyla even-toed hooved herbivores 82. +MulobVw M,sutkt"s long-nosed peccary 59 Mglohgus sp. peccaries, unidentified 3, 20, 22, 26, 43, 47, 51, 73

83. + Plotagonus conlwes fl,s flat-headed peecary 20, 31,66 60 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 3. (cont.).

Scientific Name Common Name Locality number (s)

84. +Platagorrus aetus Leidy's peceary 68 PlatEganu,s sp. peccaries, unidentified 27, 45,54,56,65

85. #Cenrus elaphus elk 11, 34, 61 Ceruus cf. C. elaphus 59 cf. Certrus sp. t7

86. O d,o c oileus ui.r gin ianus white-tailed deer 4, t6,20,31,34, 43,53, 59, 66, 71 cf . O do coileus uirgin i an:u,s 11, 40 Odocoileus sp. deer, unidentified 2, t7,24, 35, 47 , 49, 62,73

87. +Songomona? sp. fugitive deer 47,76

88. #cf. Alces alces moose 59

89. +?Cerualces sp. extinct moose 22,53

90. # Ron gifer tarond,us caribou t7,24,35,53

91. #Biso sp. bison t4,20,26,31, 53, 69 Bovid 20

92. +Sambos caoifrons woodland musk ox 53, 59 +Suzabos sp. I4

93. +Booth,erium sp. Harlan's musk ox 53 Ovibovine extinct musk ox, undetermined 14, 53 + extinctgenus * extinct species # extirpated from Virginia aalge), gannet (Moru,s bassarrus), arctic shrew roensis), long-nosed peccary (Mylohyus nasutus), (Soren arcticzs), yellow-cheeked vole (Microtus nan- fl at-headed peeeary (Platggoms conxpresslt s), fugi- thognathus), northern bog lemming (Sgnaptom.Es tive deer (Sangamona fugi,tiua) (Churcher, 1984 borealis), heather vole (Phenacom,As intermed,'ius), questioned the validity of this taxon), extinct moose pika (Ochotana princeps), least chipmunk (Euta- (Ceraalces scotti), woodland muskox (Symbos cotui- miu,s minhtru.s), ermine (Mustela ertni'nea), walrus frons), and Harlan's muskox (Bootheriumsp.). (Odobenus rosnxaras), gray seal (Halichoerus gry- All of the taxa listed above became extinct in pzs), moose (A\ces alces), and caribou (Rangifer the late Pleistocene except the great auk and pas- tarandus). Extirpated forms which now live in the senger pigeon which became extinct during the west include the whooping crane (Grus americona), historic period. Though no fossil record of the l3-lined ground squirrel (Spermophilus tridecem- species exists in Virginia, the Carolina parakeet lineatus), and bison (Bisonbison).In addition, pika, (Corru,ropsi,s carolinensis) is another species that least chipmunk, sharp-tailed grouse, and gray jay, became extinct during the historic period. Not listed as northern forms, also can be listed as surprisingly, the late Pleistocene mammalian and western. avian fauna of Virginia show a marked association Animals now extinct which lived in Virginia in with the Recent northern fauna. Perhaps surpris- late Pleistocene time include: great auk (Pingui,nus ing, however, is the presence of forms with southern 'impennis), passenger pigeon (Ectopistes ruigtrato- affinity - the beautiful armadillo (DasUpus bellus) rfus), Jefferson's ground sloth (Megalonyr jefferso- and the roundtail water rat (NeoIiber leonard'i). nii), beauliful armadillo (Dosypus belhus), roundtail Table 4 compares Virginia's late Pleistocene water rat (Neofiber leonnardi); dire wolf (Cani,s mammals with the State's "potential" 16th Century d'irus), extinct large cat (Panthera or Fel'is atror), mammals. Using this approach one can negate the American mastodon (Mommut anxericunu,nx), effects of European introduetion of exotic taxa such woolly mammoth (M a,mmuthus prim'igenizrs), com- as Rattus and Mus as well as possible extirpation plex-toothed horse (Ery,us com,plicatus), brother of original faunal elements such as Ectop'istes, horce (Eqrus fraternus), Vero tapir (Tapirus ue- Conuropsis, Certlus, and B'ison Guilday (1963, PUBLICATION 75 61

Table 4. Comparison of the terrestrial mammalian fauna of Virginia in the late Pleistocene and in the 16th Century, with inferred feeding habits.

Late Pleistocene Fauna Feeding Habits Potentiall 16th Century Fauna omnivore Di.d,elphi,s uirgi,niana Soren arct'ints insectivore Sorea langrirostris Ssrex ainereus insectivore Swer c'inereus Swer d'ispar insectivore Sorer dispar Swerfumeus insectivore Sorerfumzus Sorer palu,stris insectivore Sorer palustris Microsorer lwyi insectivore M'i,croswer havi insectivore BLarinatalmal,estes Blarina brwi,caud,a insectivore Blarina breuicaud.a Cryptot:is panta insectivore Cryrptot:is po,nta Parascal,ops Ureweri fossorial carnivore Parascalops brewert Scalopus a,qunrinls fossorial carnivore Scalopu,s aqualbus Cond,glu,ra cristata fossorial earnivore Candalura cristata Mgot:is luti,fu,gtrs aerial insectivore MEot:is luti,fugu^s Mgot:is kemii aerial insectivore Maoti.s keewi,i Mgottis grisescms aerial insectivore Mgot:is grrisescms Maotis lei.bii aerial insectivore Maoti,s lei,bii aerial insectivore MEotis sod,alis Pipistrellus wbflatus aerial insectivore Pipi,streLhs subflauus

La,si,ang cteris nn cti,u a,g ans aerial insectivore Lasi,m,g cteris nn ct'iu ag ans Eptes'im,s fuseus aerial insectivore Eptesi,cus fuscus Plecotus tnwnsend:i:i aerial insectivore I'a,si.ums cincreu,s Lasiurus borealis aerial insectivore I'aEi,uru^s borealis

aerial insectivore N g cti.ce'ius hwn er aL is aerial insectivore Plecotus raf'hesqwii MegalanAr jeffersawii large insectivore Dasupus belhrc medium insectivore Ochotanasp. small grazer Suhi,lo,gus palustris

Sahi,Lo,gtrs transi.ti,un alis small grazer SU lttilogus tr on sit/i,onali s SAlui,lwusflsri.darrus small grazer Sului,lo,gus flurid,anus Lepus am,erinanu,s small grazer Inpus o,meri,comts Tomins strintus small granivore Tamios striottts Eutamios miwimas small granivore Mormntam,snar medium semi-fossorial herbivore Martnotamonan Spenrwphi.las tri.d,ecemlinpatus small semi-fossorial herbivore Seixrrw carolinpnsis arboreal granivore Sciunts carol'h,ewis arboreal granivore Sciunnwi,ger T am iasciurus hud,s anims arboreal granivore Taminscinrus h:ud,san;ims GlaurwnEs uolans arboreal g""*uo". GlaunomEs aolans Glau,c.om,ys safutnus arboreal granivore Glau,curnys saArbws Costor canad,ensis semi-aquatic herbivore Castor conadensis C ostqroid,es ohioensis semiaquatic herbivore small granivore Reith,rodmfiom,g s lwnulis

Peramg s cus manim,latus small herbivore Perotryscus m,onbulahrc small herbivore Ochrotom.gs mft.alli 62 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 4. (cont.)

Late Pleistocene Fauna Feeding Habits Potentiall 16th Century Fauna Perom,yscus lewopn small herbivore Peromyscus leunopus

small herbivore Perontg seus gostyp'irw,s small herbivore Si$wdfrnWidis Ph,eno,cam,y s hrqmedhrc small herbivore Orazows palustris Neotqrnafl,ori.ilona small herbivore Neotom,aflorid,ann Clethrianam.Es garyeri small herbivore Clnthrionom,gs gapperi'

Minrohrc pmnsylaanicus small herbivore M'inroh$ p enn sa la oni,cus

M icr ohrc ch,r ot orrhirrus small herbivore Mi*rotus ch,rotorrluitt us Microfus aanlhogmathus small herbivore Mierotqs pirwtorurn small herbivore Mi,crotus p'inetorum Neofiber Leunard;i semi-aquatic hbrbivore 0nd,atra zibeth'i,cus semi-aquatic herbivore Onl,atra zibetluinrc Synaptom.gs cooperi herbivore Sytaptam,ys coaperi, Swaptows borealis herbivore ZaWs hud,soaims herbivore Zap,s hudsouim,s Napaeozapus insignis herbivore N ap apozapus in,sigtvi,s Eretbizon dorsolurn herbivore ?Erethiwn d,orsahm, CMrk dirlts medium-sized earnivore Can:i,s lupu,s medium-sized earnivore Canis Lupus Conis latrans medium-sized earnivore

U ro cg on c'hwr eo ar gerfin'ust medium-sized carnivore Uro ey on eirwr eo arg mt eu,s Vulpes rruLpes medium-sized earnivore ?Vulpes rrulpes Ursus am,eri,canas large omnivore Urws amprinarws Proeyan lotor medium-sized carnivore Procyon l,otar Martes pmnanti small carnivore Martes pennant;i Mustelariaosa small carnivore Mustelarisosa Mu,stela erm'inea small carnivore Mustelafrmata small carnivore Mustnlafrmata Mustzlaui^sqn small carnivore Mustnlavisan small carnivore Itutracanolewb Mepbi,t;is mephif;is medium-sized omnivore Mephitis mppluit:is

Spi,l,o g ale puf,ariil,s medium-sized omnivore Spilagal.e yutari,us Fel:is sp. medium-sized omnivore Feli,s cannol,or Lynn rufus medium-sized omnivore Itwnrufus M o,m.mnt o,mpricanil,nx large herbivore M o,mtnutluus pri.migmius large herbivore Eguus cam,pl'i,c,ahrc large herbivore Equusfratcraus large herbivore Tapiru,s aeroms'is large herbivore Mgl,ohgus nasuhts medium-sized omnivore Platy g on us com,pre sv.Lg medium-sized omnivore Ceruus eLo,phus large herbivore Cenns elo,phtts

O do coileus uirgini,curus large herbivore O d,o coileus ui,r ginim.us Sangam,anasp. large herbivore Al,ces alces large herbivore Centalces sp. large herbivore PUBLICATION 75 63

Table 4. (cont.)

Late Pleistocene Fauna Feeding Habits Potentiall 16th Century Fauna Rangifer taranihs large herbivore Bisansp. large herbivore Bison bison Symbos cavifruns large herbivore Bootlwriumsp. large herbivore ?Hanw sapiens large herbivore Homa so,piens

rBy potential is meant the fauna to be expected today without the effect of colonization by Europeans.

1971b, and 1982a) thought that bison, which are late Pleistocene fauna and only six in the 16th absent from archeological sites in Virginia, were Century fauna. Microtines are characteristic of transient in the Appalachian Mountains. We do not more northern or boreal environments than are dispute the possibility that Indians affected the cricetines. The pattern of microtine occurrence in prehistoric mammal fauna; we are only trying to Virginia is consistent with the inferred climatic negate the effects of European eulture on the fauna differences between the late Pleistocene and the and identify a "potential" natural mammalian Holocene. Eshelman (1975) demonstrated similar fauna. Table 4lists each taxon by inferred feeding changes in cricetine/microtine proportions that habit. The late Pleistocene mammalian fauna con- were correlated with cooler and warmer climates sists of 85 taxa including Paleo-Indians (Homo of Pleistocene glacials and interglacials, respec- sapiens). The 16th Century fauna, including man, tively. consists of 73 taxa (Guilday and others, 1gZ?, list The New Paris No. 4 fauna in Pennsylvania 68 and Guilday, l982a,listed ?b mammalian taxa (Guilday and others, 1964) illustrates this pattern from the mid-Appalachian region). Fifteen taxa of changing cricetine/microtine proportions eorre- from the late Pleistocene fauna are extinct, these lated with climatic changes in the region. Micro- include twelve herbivores, two carnivores and one tines account for 80 percent of the total individuals omnivore. Twenty-six taxa in the late Pleistocene identified in the lower levels, but only 34 percent fauna are not present in the 16th Century fauna; in the younger, upper levels (dated at 11,300 + 100 twenty of these are herbivores. Thirteen taxa listed yr B.P.). The associated faunas suggest that boreal for the 16th Century fauna are not known for the climate existed when the lower levels were dep- late Pleistoeene of Virginia; of these, four are bat osited, and that this changed to a more temperate species and five are cricetid species. Additionally, climate when the upper levels were deposited. four species, otter (Lutra cqnadensis), fox squirrel Graham (1976) shows that the taxonomic diversity, (Sc'iurus ntger), southeastern shrew (Soren lonry- of microtines during the late Pleistocene decreases rostris), and Indiana Myotis bat (Myotis so,d,alis) from the northeast to the southwest across the U. are possibly present in the late Pleistocene fauna S. This trend parallels this same cline seen today, of Virginia, but they have not yet been identified but the group was less diverse in the late Pleistocene positively. than it is today. Graham (1976) notes that the Guilday (1982a) notes that for the mid-Appala- diversity of shrews also increased over this time chian area the late Pleistocene mammalian fauna period. However, for Virginia there are eight shrew was more diverse, especially for the browsing and taxa in the late Pleistocene and nine in the 16th grazing herbivores, than the Recent fauna. Table Century fauna (Table 4). This count on shrew 4 indicates that this statement appears to be true species could reflect bias in eollecting and iden- for Virginia. tification errors. Assuming that the greater number of cricetine Though the opossum (Didelphis ui,rgin'iana) is rodent species in the 16th Century fauna, as com- listed from six localities in Table 3, three oceur- pared to the late Pleistocene fauna (seven versus rences (96, 44, 46) are from sites with uncertain two),js not an artifact of collecting bias or iden- age determinations. Localities 66 and 71 are as- tif_@tion error, a glaring disparity in cricetine sumed to be Holocene sites based on associated reiiresentation during these two periods is evident. fauna. The sixth site, Keyser Farm (Table 1: locality In contrast, there are ten microtine rodents in the 34), is a Woodland Period archeological site which 64 VIRGINIA DIVISION OF MINERAL RESOURCES is dated between 1000 BC and 1600 AD. The parkland Equus fratermts, E. complicafus, Booth- opossum may not have been present in Virginia erium sp., Sambos cauifrons, and' Bison sp. would during the late Pleistocene. In addition to the six have been prevalent. localities listed in Table 3 for the porcupine (,Ere- It was also during this time that sea level was thizan dorsatum), Ray and Lipps (1970) refer to more than 300 feet below its present level, thus three unpublished occurrences of porcupine in Vir- exposing much of the continental shelf from where ginia (Ray, personal communication, indicates that Mam,ruu,t and Mam,ruu'thus teeth have been reco- specimens from these localities are under study at vered by bottom-dredge fishing boats. The area of the University of Tennessee). Guilday (1962a) states the Chesapeake Bay would have contained the large the spotted skunk (Spilogale) and the opossum valley of the proto-Susquehanna River and its (Didelphis) are suspeeted of being post-Pleistocene tributaries. Here Paleo-Indians, no doubt, found additions to Virginia. Table 3 concurs with Guil- shelter and food. Post-glacial warming changed day's observation for the opossum, but the oecur- this setting remarkably. Sea level slowly rose to rence of Spilogale at Clark's Cave indicates that its present stand, and the primarily coniferous the spotted skunk was present in the late Pleis- forest of Virginia described above was succeeded tocene. by a deciduous forest blanket. By early Holocene The reduction in number of mammalian species time the State was covered by a vast, closed-canopy from the late Pleistoeene to the Holocene has been deciduous forest in which the mammalian fauna noted by several researchers (Martin, 1967; Martin consisted dominately of forest product eonsumers, and Webb, 1974; Graham, 1976, 1979). Semken insectivores, and carnivores (Guilday, f984) (Table (1983, p. 192) summarized this phenomenon as 4). follows: "The Pleistocene faunaof the United States The forest-product consumers consisted of was characterized by a combination of (1) extinct browsers such as the large herbivores: Certvus megavertebrates and (2) extant temperate mega- elaplrus, Odocot'leus uirginianns, and Alces alces. vertebrates and microvertebrates in association Mast feeders (nut eaters) included the arboreal with (3) now disjunct large and small northern granivores: Sciurus carol;inensis, Tamiasc'i'urus species. The Holocene fauna of temperate regions hudsonicus, Glaucomys aolans, and Glau,conrys sa- generally is eomposed of the seeond category only." brirrus. Grazing animals (grass eaters) were res- The data provided here indicate that this general tricted to what were mainly temporary grasslands. observation is accurate for Virginia. These were small rodents and rabbits including: SEluitags transitionalis, Sylailagws floridonus, QUATERNARY ENVIRONMENTAL Lepus americonus, SEnaptomys cooperi, Sana'p- RECONSTRUCTION BASED ON tom,ys bwealis, Microtus pennsglntaniats, and Mi- MAMMALIAN FOSSILS crohrc p'i,netoram. B'ison sp. was probably transient or relatively uncommon because of restricted grass- During maximum Wisconsinan glaeiation from lands during the Holocene. approximately 18,000 yr B.P. to just prior to the Insectivores whieh fed upon insects of the decid- Pleistocene/Holocene boundary at 10,000 yr B.P., uous forest consisted of shrews and bats and in- the environment of Virginia was boreal. It was cluded Sorer dispar, M't'crosorer hoyi, Blarinabre- during this period that the large extinct herbivores uicauda, Cruptotis po,r"ua, Myotis keen'ii, Maoti's listed in Table 4 such as Mammnt americanttnn, leibii, Pipistrellus subflatrus, Lasiong ct'eris no;ctiu- Mammuthus pri,migenius, and specialized herbi- o,gutLs, Eptesicus fusctts, Plecottts tunn'smd'ii, La- vores such as Megalongr jeffersoruii, Platygorrus siu,rus borealis, and, Dasgpus bellus. Moles, such cornpresslrs, Mglohyus nasu,hts, and T apirus aeroen- as Sculopus aquaticus and Condglwra cristata, sis existed. Smaller extirpated boreal forms listed which are listed in Table 4 as fossorial carnivores, in Table 4 include Soren arcticus, Ochotunasp., and eat insects. Microtus nanthogrnathus. These forms existed, pos- Fresh-water mammals of late Pleistocene/early sibly sympatrically, within a sprucefiackpine/fir, Holocene time were Castor canadernis, Castoroides birch parkland interspersed with woody shrubs ohioensis, Ondatra zibethicus, and Neofiber leo- and grasses which eovered the entire State (Del- nardi. Finally, carnivores whieh preyed upon many court and Delcourt, f981). This unique, extinct, of the above species were: Cani,s dirus, Conis luptts, coniferous parkland has no modern counterpart. Vulpes rrulpes, Martes pennantti, Mustela uisan, In mountain areas of the State, tundra may have Felis sp., and funn rufus. existed. There, Rongi,Iertdrond'us would have been It should be pointed out, that the late Pleistoeene more plentiful. In the grassy areas of the boreal fauna of Table 4 did not coexist sympatrically at PUBLICATION 75 65

any one time nor did all these forms inhabit every is in need of restudy and thus there must remain part of what today is Virginia. some question as to the specific identifications of certain forms, such as species within Equus and DISHARMONIOUS FAUNAS AND THEIR Mam,mnthus. POSSIBLE CAUSES A second plausible explanation for seemingly disharmonious faunas could be faunal eompression Lundelius and others (1983, p. 311) states: "Var- caused by continental glaciers moving south in the ious taxonomic groups of vertebrates from a given late Pleistocene. "That the fauna of glaciated ter- site frequently suggest different and sometimes ritory was forced southward or eliminated does not contradictory paleoenvironmental reconstructions; imply that the boreal fauna eompletely replaced for example, in the Appalachian area mammalian a more southern faunal type or that the area south faunas and some birds indicate'boreal' conditions, of the ice front was appreciably colder than at but the herpetofaunas and other birds from the present. This view holds that an admixture of same locality indicate a climate similar to that of ecologically segregated boreal and austral elements today or even milder." existed, possibly almost side by side, near the ice Fay (1984) studied the herptiles of Strait Canyon front but within unglaciated territory" (Smith, (Table 1: locality 59) and notes that climatic con- L957, p.207). ditions 30,000 years ago were nearly identical to Guilday (1971a), however, argue that latitudinal those of modern western Virginia. and altitudinal shifting of life zones in response Holman (this volume) states: "Although Virgi- to Quaternary climatic oscillations cannot fully nia's late Pleistocene mammalian fauna contains explain the disharmonious faunas'known from the extinct forms and forms that suggest boreal con- late Pleistocene. For example, Guilday (1971a) ditions, the herpetofauna appears to be almost points out that using the present distribution of identical to what it is today." Why would amphi- the round-tail water rat (Neofiber allewi) as a bians and reptiles (ectotherms) characteristic of southerly climatic indicator could lead to "serious mild environments coexist with birds and mam- error". Its present range may be restricted by mals (homoiotherms) many of which are charac- eompetition with the larger muskrat (Ondatra teristic of boreal environments? Following is a zibethicus), rather than by climate. In the Pleisto- discussion which may help to explain this disparity. cene, however, closely related water rodents were The identification of some fossil specimens might apparently sympatric. During Kansan, Sangamo- be inaecurate because of the fragmentary nature nian, and Illinoian times the extinct muskrat (Oro- of the specimens and the lack of adequate modern d,atra annectens) and a species of water rat, Neofiber eomparative materials. Bird remains that are eom- sp., were evidently sympatric at Trout and Hamil- monly misidentified include hawks, eagles, and ton caves, West Virginia and O. Iuiatid,ens and grouse (Lundelius and others, 1983). Grouse are Neofiber sp. were sympatric at Port Kennedy Cave, commonly cited as being indicative of "boreal" Pennsylvania. conditions. While such errors of identification Ray and others (1968) point out that the gannet might contribute to the problem of apparent ec- (Morus bassamus\ and gray seal (Hal'i,choerus gng- ological incompatibility among species, within a pus) apparently were breeding in southeastern given fauna, the identification of remains of many Virginia, far south of their present range, during boreal mammal species (e.g., caribou, pika, yellow- the late Pleistocene. These records might suggest cheeked vole, and rock vole) is comparatively easy. that coastal Virginia had cooler temperatures dur- Guilday and Parmalee (1972) and Guilday and ing the late Pleistocene, being similar, perhaps, to Hamilton (1978) notes that the heather vole is easily that of Nova Scotia today. But Ray and others (1968) identified even in very young individuals. Though point out that more recent natural and human some misidentifications might have been made, influences on the region (e.g., destruction of favor- most identifications probably are reliable. Classic able coastal breeding habitat by sea level stabi- exceptions to this expected standard are Prout's lization, port aetivity, etc.) also could have been (1860) description of a juvenile mastodon tooth from important eontributing factors to their extirpation King's Salt Works, near Abingdon, Virginia, as from Virginia. that of a titanothere and Peterson's (1917) iden- Today the pika (Ochotona princeps) is confined tification of a juvenile proboseidean tusk from to the northern Roeky Mountains, yet the genus Saltville as the tooth of a crocodillian (Ray and is known from six sites in the Appalachian Moun- others, 196?). Additional confusion can be expeeted tains, including Jasper Saltpeter Cave (Table 1: beeause the taxonomy of some mammalian groups locality 9). Guilday (1979) points out that pikas 66 VIRGINIA DIVISION OF MINERAL RESOURCES

(Oehotona) from the Pleistocene of the eastern Graham (1976, 1979) and King and Graham United States may be a separate species with (1981) noted that animal distributions are prima- environmental requirements different from the rily determined by either the distribution of plant western form and/or that modern pikas are con- foods on which they depend or on the distribution fined to their present niche by interspecific com- of prey species. King and Graham (1981,, p. 133) petition, thus its present range is limited by "com- stated "animals show varying tolerances by the petitive faetors rather than physiological or range of habitats in whieh they can exist. . . If anatomical strictures" (Guilday, 1979, p. 442). each species reacted individually in response to Weems and Higgins (1977) attempt to explain environmental change through time as they have the disharmonious Ripplemead Quarry (Table 1: been shown to do through space, then plant com- locality 51) fauna as an accumulation that included munities would be constantly changing and their Pleistocene forms (including tapir, horse, and pec- species eomposition would not necessarily be the cary) that had survived into the Holocene, thus same as modern analogs." Thus, because individual allowing a mixture of boreal and temperate species. species of a community react independently to Semken (1983) and Mead and Meltzer (1984) re- environmental changes, it is not always appropriate viewed purported records of Holoeene faunas con- to apply characteristics of modern communities to taining Pleistocene taxa that have been dated ra- the reconstruction of past environments. This diogenically and argue that reeords of such model was elaborated and termed "coevolutionary postulated late survivals are rare if valid at all. disequilibrium" by Graham and Lundelius (1984). Guilday (197la) and Lundelius and others (1983) Thus Holman's (this volume) dilemma of why the pointed out that several taxa from the Appalachian Virginia herpetofauna does not suggest the boreal region show a positive Bergmann's Response, that environmental interpretation that many of the is, the tendency for members of a speeies to increase birds and mammals do, may be explained by their in size with increasing latitude through its range. individual response to change. Amphibians and A southward shift of cooler elimates is inferred for reptiles (ectotherms) apparently were little af- Pleistocene glaciations. Several species, such as fected by the modern plant eommunity changes Ondatra (Nelson and Semken, 1970) and Blarina from the late Pleistocene to Recent, while birds and (Graham and Semken, 1976; Jones and others, mammals (homoiotherms) were more affected. 1984), were larger during late Pleistocene than they Semken (1983) points out that several species are at the same areas today; the late Pleistocene including Erethizon d,orsatum, Martes pennanti, forms from the U. S. are comparable to those living Sperm,ophilus tridecemlineo,tus, Clethrionom,ys in Canada today. Guilday (1984, p.254 noted "nine gapperi,, and Tam;i'asciu'rus hudsotutms apparently species of mammals have undergone a Holocene had more extensive ranges until a few hundred size reduction, while four have increased in size years ago. The modern faunal distribution cannot within the [Appalachian] area during the last be attributed wholly to climatic change associated 11,000 years, paralleling, for the most part, modern with deglaciation since many small mammals were latitudinal size clines." Thus size analyses may strongly affected by (a) climatie change during the support a claim of overlap of southern and boreal Holocene, (b) pre-Columbian cultural patterns, and types. (c) European settlement(Semken, 1983). Thus, even Many researchers have attempted to explain within relatively short periods of geological time incompatible, "allopatric" or "disharmonious" fau- and in the absence of major climatic changes, range nas (Semken, 1974) by the "climatic equability changes among vertebrate species can occur suf- hypothesis" which maintains that more equable ficiently to create new disharmonious faunas. climates (eooler summers and wat'mer winters) in the Pleistocene allowed seemingly unnatural asso- AGE INTERPRETATION OF VIRGINIA'S ciations of northern, western, and southern forms QUATERNARY LOCALITIES (Hibbard, 1960; Dalquest, 1965; Slaughter, 1967, 1975; Dalquest and others, f969; Hibbard, 1970; Table 3 lists all of the Virginia fossil vertebrate Holman, 1976, 1980; Martin and Neuner, 1978). localities giving their presumed ages as either late Dreimanis (1968) argued that late Pleistocene Pleistocene (Wisconsinan) or Holocene. A few of changes in the structure and composition of plant the human-associated sites, such as Keyser Farm communities rather than shifting ecological zones (Table 1: locality 34), can be placed within the may have allowed the sympatry of northern and Woodland Indian Period based on associated cul- temperate mammals and temperate herpetological tural material. Gardner's Cave, Edinburg Fissure, forms. and possibly part of Jasper Saltpeter Cave and PUBLICATION 75 67

Strait Canyon Fissure suggest early Wisconsinan Steadman, USNM, and especially Jerry McDonald, or even Illinoian ages. Virginia has not yielded a RU, for reviewing and making helpful suggestions rich middle Pleistocene fauna such as Trout Cave, which have greatly improved this paper. West Virginia; Cumberland Cave, Maryland, or Port Kennedy Cave, Pennsylvania. Jones and others REFERENCES (1984) listed Early's Cave (Table 1: locality 20) as mid-Pleistocene and Strait Canyon (Table 1: local- Allen, J. A., 1885, On an extinct type of dog from ity 59) as early Wisconsinan. More detailed study Ely Cave, Lee County, Virginia: Memoirs Mus. and more radiometric age determinations of the Comp. Zool., Harvard Univ., vol. 10, no. 2, 13 known faunas of Virginia (only 4 of the 77 localities p. reported in this paper have been dated radiomet- Anderson, E., 1984, Review of the small carnivores rically) are needed before better chronologic inter- of North America during the last 3.5 million pretations can be made. years, 'inGenoways, H. H., and Dawson, M. R., eds., Contributions in Quaternary vertebrate ACKNOWLEDGMENTS paleontology: A volume in memorial to John E. Guilday: Carnegie Mus. Nat. Hist., Spec. First and foremost, acknowledgment must be Pub. no. 8.,p.257-266. given to the late John Guilday under whose direc- Churcher, C. S., 1984, Sangamana,: Furtive Deer, tion no less than 44 Virginia Quaternary fossil in Genoways, H. H., and Dawson, M' R., eds., localities have been investigated, ten of which had Contributions in Quaternary vertebrate paleon- been published formally or mentioned in publica- tology: A volume in memorial to John E. Guil- tions prior to his untimely death. The Carnegie day. Carnegie Mus. Nat. Hist., Spec. Pub. no. Museum of National History for which John 8, p. 316-331. worked has led in the systematic collecting of Clark, A. H., 1938, Some Pleistoeene mammals vertebrate fossils in Virginia. Carnegie Museum from Warren County, Virginia: Science, vol.88, personnel Harold Hamilton and Allen McCrady, no.2273,p.82. as well as numerous volunteers such as Robert 1939, Some Pleistocene mammals from Whittemore and Lois Miller, helped John and Warren County, Virginia: The Raven, vol. 10, CMNH compile this impressive collection of pa- nos.4 and 5, p. 6-7. leontological data. The authors are most grateful Clark, W. E., 19?9, An assessment of the archeo- to Alice Guilday who provided us with John's - logical resources in the caves of Virginia: Ap- unpublished data, without which this paper would pendix V, Report of the Virginia Commission be painfully incomplete. on the conservation of eaves to the governor The United States National Museum of Natural and the general assembly of Virginia, House History also conducted fossil cave studies in Vir- Document No. 5, p.22-34. ginia, principally through the efforts of Clayton Cope, E. D., 1867, Remarks on the contents of caves Ray and Albert Myrick. Paul Drez, Herman Wil- in southwestern Virginia: Proc. Acad. Nat. Sci. bur, and Ron Clay are among the hundreds of Philadelphia, vol. 19, p. 1,37-138. amateur fossil collectors who have contributed to 1869a, Synopsis of the extinct mamma- the Quaternary history of Virginia. Without their lia of the cave formations in the United States cooperation and diligent effort, Virginia's Quater- with observations on some Myriapoda found in nary vertebrate record would be less known. and near the same, and on some extinct mam- The authors are indebted to the following insti- - mals of the caves of Anguilla, W. I., and of other tutions and respective curators for permission to localities: Proc. Amer. Phil. Soc. vol. 11, p. 171- use their collections and unpublished data: CMNH, t92. Alice Guilday, Harold Hamilton; USNM, Clayton Ray, Warren Blow, Storrs Olson; ANSP, Peter tone caves of Virginia: Proe. Acad. Nat. Sci. Feigley; NJSM, David Parris; and RU, Jerry Philadelphia, vol. 21, p. 3. McDonald. Bob Purdy, USNM, was most helpful Dalquest, W. W., 1965, New Pleistocene formation in solving many of the early mammal locality and local faunafrom Henderson County, Texas: identifications in the USNM collections. Wayne J. Paleont. vol. 39, p.63-72. Clark, Patterson Park and Museum, St. Leonard, Dalquest, W.W., Roth, E. and Judd, F., 1969, The Maryland, provided helpful information on archeo- Mammal fauna of Schultze Cave, Edwards logical cave and rockshelter sites in Virginia. The County, Texas: Bull. Flordia State Mus., vol. authors are most greatful to Clayton Ray and Dave 13,p.205-276. 68 VIRGINIA DIVISION OF MINERAL RESOURCES

Delcourt, P.A., and Delcourt, H. R., 1981, Veget- from Wythe County, Virginia: Annals Carnegie ative maps for eastern North America: 40,000 Mus., vol.36, p. 77-86. yrs. B.P. to the present, irz Romans, R. C. ed., 1963, Evidence for buffalo in prehis- Geobotany II: New York, Plenum Press, p. 123- toric Pennsylvania: Pennsylvania Archaeol., 165. vol.33, no.3, p. 135-139. Dreimanis, A., 1968, Extinction of mastodons in l97la, The Pleistocene history of the eastern North America: testing a new climatic - Appalachian mammal fauna, p. 233-262, in environmental hypothesis: Ohio J. Sci., vol. 68, Holt, P. C., ed., The distributional history of p.25t-27L. the biota of the southern Appalachians, part Eshelman, R. E., 1975, Geology and paleontology - III: vertebrates: Research Division Monog. 4, of the early Pleistocene (late Blancan) White Virginia Polytechnic Institute and State Uni- Rock fauna from north-central Kansas, in Stud- ver. ies on Cenozoic paleontology and stratigraphy, 1971b, Biological and archaeological Claude W. Hibbard memorial vol. 4: Univ. analysis of bones from a 17th Century Indian Michigan, Papers on Paleontology, no. 13, 60 village (46Pu31), Putnam County, West Vir- p. ginia: West Virginia Geol. and Econ. Survey, Ewan, J. and Ewan, N., 1970, John Banister and Rep. Archaeol. Invest., no. 3, 64 p. his natural history of Virginia, 1678-1692: Ur- 1979. Eastern North America Pleisto- bana, Univ. Illinois Press,486 p. cene Ochotona (Lagomorpha: Mammalia): An- Fay, L. P., 1984, Mid-Wisconsinan and mid-Hol- nals Carnegie Mus., vol. 48, art.24, p. 435-44. ocene herpetofaunas of eastern North America: 1982a, Appalachia 11,000-12,000 years A study in minimal contrasts, iz Genoways, H. ago: a biological review: Archaeolory of eastern H., and Dawson, M. R., eds., Contributions in N. Amer., vol. 10, p.22-26. Quaternary vertebrate paleontology: a volume 1982b, Dental variations in Microtu,s in memorial to John E. Guilday: Carnegie Mus. - ranthogmathus, M. chrontwrluirru,s, and M. pen- Nat. Hist., Spec. Pub. no. 8, p. 14-19. rcsyluoni,cus (Rodentia: Mammalia): Annals Goode, G. 8., (1886) 1901, The beginnings of natural Carnegie Mus., vol. 51, art. LL,p.2LI-230. history in America: U. S. Nat. Mus. Rept., 1897, - 1984. Pleistocene extinction and envi- pt.2, pp.357-406. ronmental change, case study of the Appalach- Graham, R. W., 1976, Late Wisconsin mammal ians: in Martin, P. S., and Klein, R. G., eds., faunas and environmental gradientsof the east- Quaternary extinctions, a prehistoric revolu- ern United States: Paleobiology, vol. 2, p. 343- tion: Tucson, Univ. Arizona Press, p.250-258. 50. Guilday, J. E., and Bender, M. S., 1960, Late 1979, Paleoclimates and late Pleisto- Pleistocene records of the yellow-cheeked vole, cene faunal provinces in North America, in M'icrotus nanthognathns (Leach): Annals Car- Humphries, R.L., and Stanford, D., eds., Pre- negie Mus., vol.35, p. 315-330. Llano cultures of the Americas: paradoxes and Guilday, J. E., and Hamilton, H. W., 1978, Eco- - possibilities: Washington, D. C., Wash. An- logical significance of displaced boreal mam- thropol. Soc.,p. 49-69. mals in West Virginia caves: J. Mammal., vol. Graham, R. W., and Lundelius, E. L., Jr., 1984, 59, p. 176-181. Coevolutionary disequilibrium and Pleistocene Guilday, J. 8., and Parmalee, P. W., 1972, Qua- extinctions, 'in Martin, P. S., and Klein, R. G., ternary periglacial records ofvoles ofthe genus eds., Quaternary extinctions, A prehistoric rev- Phenacomys Merriam (Cricetidae: Rodentia): olution: Tucson, Univ. Arizona Press, p. 223- Quat. Res., vol.2, p. 170-175. 249. Guilday, J. 8., Martin, P. S., and McCrady, A. D., Graham, R.W., and Semken, H. A., 1976, Paleo- 1964, New Paris No.4: Pleistoeene cave deposit ecological significance of the short-tailed shrew in Bedford County, Pennsylvania: Bull. Natl. (B\arina), with a systematic discussion of. Blar- Speleol. Soc., vol. 26, p.121-194. ina ozarkensis: J. Mammal., vol. 57, p.433-449. Guilday, J. E., Parmalee, P. W., and Hamilton, H. Guilday, J. E., 1962a, The Pleistocene local fauna W.,1977, The Clark's Cave bone deposits and of the Natural Chimneys, Augusta County, the late Pleistocene paleoecology of the central Virgina: Annals Carnegie Mus., vol. 36, Art. Appalachian Mountains of Virginia: Bull. Car- 9,p.87-122. negie Mus. Nat. Hist., no.2,87 p. Notes on Pleistocene vertebrates Hall, E. R., 1981, The mammals of North America:

-1962b, PUBLICATION 75 69

New York, John Wiley and sons, 2 vols., 1181 tions: Amer. Antiq., vol.46, no. 1, p. 128-L42. p. Klippel, W. E., and Parmalee, P. W., 1984, Arma- Handley, C. O., Jr., and Patton, C. P.,1947, Wild dillos in North American late Pleistocene con- mammals of Virginia: Commission of Game texts,'i,n Genoways, H. H., and Dawson, M. R., and Inland Fisheries, Commonwealth of Vir- eds., Contributions in Quaternary vertebrate ginia, 220 p. paleontology: a volume in memorial tp John E. Hay, O. P.,1923, The Pleistocene of North America Guilday. Carnegie Mus. Nat. Hist. Spec. Pub. and its vertebrated animals from the states east no.8, p. 149-160. of the Mississippi River and from the Canadian Kurten, B., and Anderson, E., 1980, Pleistocene provinces east of longitude 95 degrees: Carne- mammals of North America, New York, Co- gie Inst. Washington, Pub. no.322,499 p. lumbia Univ. Press, 442p. Haynes, G., 1978, Excavation of a small Virginia Lundelius, E. L., Jr., Graham, R. W., Anderson, cave: Quart. Bull., Archeol. Soc. Virginia, vol. E., Guilday, J. E., Holman, J. A., Steadman, 32, no.3, p. 37-41. D. W., and Webb, S. D., 1983, Terrestrial Hibbard, C. W., 1960, Pliocene and Pleistocene vertebrate faunas, 'in Portet, S. C., ed., Late- climates in North America: Annual Rept. Mich- Quaternary environments of the United States: igan Acad., Sci., Arts and Letters, vol. 62, p. vol. 1, the Late Pleistocene: Minneapolis, Univ. 5-30. Minnesota Press. p. 311-353. 1970, Pleistocene mammalian local fau- MacCord, H. A., l973a,The Hirsh site: Quart. Bull., nas from the Great Plains and Central Low- Archaeol. Soc. Virginia, vol. 27, no. 3, p. 151- lands provinces of the United States, 'in Dort,, r57. W., Jr., and Jones, J. K., Jr., eds., Pleistocene MacCord, H. A., 1973b. The Hidden valley rock- and Recent environments of the Great Plains: shelter, Bath County, Virginia: Quart. Bull. Kansas Univ., Dept. Geol., Spec. Pub. no. 3, p. Archaeol. Soc. Virginia, vol. 27, no. 4, p. 198- 395-433. 228. Holman, J. A., 1976, Paleoclimatic implications of Manson, C., MacCord, H., and Griffin, J. B., 1943, "ecologically incompatible" herpetological spe- The culture of the Keyser Farm Site: Papers, cies (late Pleistocene: southeastern United Michigan Acad. Sci., Arts and Letters, vol.29, States): Herpetologica, vol. 32, p.290-295. p.375-418. 1980, Paleoclimatic implications of Martin, L. D., and Neuner, A. M., 1978, The end Pleistocene herpetofaunas of eastern and cen- of the Pleistocene in North America: Trans. tral North America: Trans. Nebraska Acad. Nebraska Acad. Sci., vol. 6, p.ll7-L26. Sci., vol. 8, p. 131-140. Martin, P. S., 1967, Prehistoric overkill, 'inMartin, -Holsinger, J. R., 1967, Some bones and shields from P. S., and Wright, H. 8., Jr., eds., Pleistocene acave in SWVirginia: Natl. Speleol. Soc. News, extinctions: the search for a cause: New Haven, vol.25, p.198-201. Yale Univ. Press, p.75-120. 1975, Descriptions of Virginia caves: Martin, R. A., and Webb, S. C., 1974, Late Pleis- Virginia Division of Mineral Resources, Bull. tocene mammals from the Devil's Den fauna, 85, 450 p. Levy Counly, in Webb, S. C., ed., Pleistocene Hovey, H. C., 1882 Celebrated American caverns: mammals of Florida: Gainesville, Univ. Presses - Cincinnati, R. Clark and Co., 228 p. Florida, p. 114-45. Jefferson, T., 1787, Notes on the state of Virginia: McDonald, J. N., 1984, The Saltville, Virginia London, John Stockdale,382 p. locality: a summary of research and field trip Jones, C. A., Choate, J. R., and Genoways, H. H., guide: Symposium on the Quaternary of Vir- 1984, Phylogeny and paleobiogeography of ginia, September 26-29, 1984, Charlottesville, short-tailed shrews (genus Blari,na),'i,n Geno- Virginia, Division Mineral Resourees,4S p. ways, H. H., and Dawson, M. R., eds, Contri- McDonald, J. N., and Bartlett, C. S., Jr., 1983, An butions in Quaternary vertebrate paleontology: associated musk ox skeleton from Saltville, a volume in memorial to John E. Guilday: Virginia: J. Vert. Paleont., vol.2, p. 453-470, Carnegie Mus. Nat. Hist., Spec. Pub. no.8, p. McGill, W. M., 1933, Caverns of Virginia: Virginia 56-148. Geol. Survey, Bull. 35, 187 p. King, F. 8., and Graham, R. W., 1981, Effects of Mead, J. I., and Meltzer, D. J., 1984, North Amer- ecological and paleoecological patterns on sub- ican late Quaternary extinctions and the ra- sistence and paleoenvironmental reconstruc- diocarbon record, in Martin, P. S., and Klein, 70 VIRGINIA DIVISION OF MINERAL RESOURCES

R. G., eds., Quaternary extinctions, a prehis- ican Quaternary Assoc., 3rd Biennial Meeting, toric revolution: Tucson, Univ. Arizona Press, Univ. Wisconsin, Madison, p.25. p.440-450. 1983, Holocene mammalian biogeo- Mitehill, S. L., 1818, Observations on the geology graphy and climatic change in the eastern and of North America; illustrated by the descrip- central United States, in Wright, H. E., ed., tion of various organic remains found in that Late-Quaternary environments of the United part of the World, in George Cuvier, Essay on States: vol.2, the Holocene: Minneapolis, Univ. the theory of the Earth, with mineralogical Minnesota Press, p. 782-207. notes, and an account of Cuvier's geological Shiner, J. L., 1983, Large springs and early Amer- discoveries, by Professor Jameson: New York, ican Indians: Plains Anthropol., vol. 28, p. 1- Kirk and Mercein, p.319-431. 7. Mitchill, S. L., Smith, J. A., and Cooper,W.,L827, Simpson, G. G., 1942,The beginnings of vertebrate Discovery of a fossil walrus in Virginia: Annals paleontology in North America: Proc. Amer. Lyc. Nat. Hist., New York, p. 107-114. Phil. Soc., vol. 86, p. 130-188. Nelson, R. S., and Semken, H. A., 1970, Paleoeco- 1943, The discovery of fossil vertebrates logical and stratigraphic significance of the in North America: J. Paleont., vol. 17, no. 1, muskrat in Pleistocene deposits: Geol. Soc., p.26-38. America Bull., vol. 81, p. 3733-3?38. Slaughter, B. H., 1967, Animal ranges as a clue Osborn, H. F., 1939, Cope: master naturalist: - to late Pleistocene extinctions,'inMartin, P. S., Princeton, Princeton Univ. Press. and Wright, H. E., eds., Pleistocene extinctions, Prout, H. A., 1860, Remarks of Dr. Prout on a fossil the search for a cause: New Haven, Yale Univ. tooth found at King's salt works, near Abing- Press, p. 155-167. don, Virginia: Acad. Sci. St. Louis Trans., p. Slaughter, B. H., 1975, Ecological interpretation 699. of the brown sand wedge local fauna, inWen- Peterson, O. A., 1917, A fossil-bearing alluvial dorf, F., and Hester, J. J., eds., Late Pleistocene deposit in Saltville Valley, Virginia: Annals environments of the southern High Plains: Fort Carnegie Mus., vol. 11, nos. 3-4, art.19, p. 469- Burgwin Research Center, Pub. no. 9, p. 179- 474. 92. Pickle, R. W., 1946, Discovery of folsom-like arrow- Smith, P. W., 1957, An analysis of post-Wisconsin point and artifacts of mastodon bone in south- biogeography of the prairie peninsula region west Virginia: Tennessee Archaeol., vol. 3, no. based on distributional phenomena among ter- 1, p. 3-6. restrial vertebrate populations: Ecologa, vol. Ray, C. E., Cooper, 8.N., and Benninghoff, W.S., 38, p.205-218. 1967, Fossil mammals and pollen in a late Stevens, T., 1979, The \{ade Site (44H34): a flood Pleistoeene deposit at Saltville, Virginia: J. damaged Woodland site: Quart. Bull. Archeol. Paleont., vol. 41-, p.608-622. Soc. Virginia, vol. 33, no.4, p. 148-59. Ray, C. E., and Lipps, L., 1970, Southerly distri- Stuckenrath, R., and Mielke, J. E., 1972, Smith- bution of porcupine in eastern United States sonian Institution radiocarbon measurements during late Quaternary time: Georgia Acad. VII: Radiocarbon, vol. 14, p.401-472. Sci. Bull., vol.28, no.2. Weems, R. 8., and Higgins, B. B., 1977, Post- Ray, C. E., Wetmore, A., Dunkle, D. H., and Drez, Wisconsinan vertebrate remains from a fissure P., 1968, Fossil vertebrates from the marine deposit near Ripplemead, Virginia: Nat. Spe- Pleistocene of southeastern Virginia: Smithson- leol. Soc. Bull., no. 39, p. 106-108. ian Misc. Coll., vol. 153, no. 3, publ. 4742,25 Wetmore, A., L962, Birds, in J. E. Guilday, ed., p. the Pleistocene local fauna of the Natural Chim- Reeves, F., 1932, Thermal springs of Virginia: neys, Augusta County, Virginia: Annals Car- Virginia Geol. Survey Bull. 36,124p. negie Mus., vol.36, p.87-122. Robbins, C. S., Brun,8., and Zim,H. S.,1966, Birds Whitmore, F. C., Jr., Emery, K. O., Cooke, H. B. of North America: New York Golden Press. 340 S., and Swift, D. J. P., 1967, Elephant teeth p. from the Atlantic Continental Shelf: Seience, Semken, H. A., Jr.,1974, Mieromammal distribu- vol. 156, no. 3781, P.1477-1481. tion and migration duringthe Holocene: Amer- PUBLICATION 75 7l

EARLY MAN IN VIRGINIAI

Ben C. McCary2

CONTENTS

Page

Clovis culture in Virginia 72 Age of Clovis culture tt

ILLUSTRATIONS Figure 1. Map showing number of Clovis points reported from Virginia cities and counties ...... 72 2. Clovis points from three areas of eoncentration 74 3. Clovis tools from the Williamson site (A) 75 4. Clovis tools from the Williamson site (B) 76

ABSTRACT INTRODUCTION

The Archeological Society of Virginia's survey The distinctive fluted point associated with the offluted projectile points, carried out since 1947, Clovis culture has been found in every state in has recorded more than 700 such points from the eontinental United States and in parts of within the State. Over 85 pereent of these records Canada. No prototype of it has been found in are from the Piedmont and Tidewater areas south Europe or Asia. Therefore, it apparently orig- of the James River, as are most of the known inated in the present forty-eight contiguous Clovis sites in Virginia. A significant outlying United States. Clovis points usually have parallel eoncentration of projectile point records focuses or slightly convex sides. The base in most cases upon Smyth County, southwestern Virginia. Ra- is concave. One major characteristic is the pres- diocarbon dated Clovis sites from eastern North enee of flutes on the proximal or basal faces which America fall between 11.120+180 and 10.1901300 extend about one-fourth to one-half of the length yrs B.P. of the points. There are no notehes in the basal area. Based on raC-dated sites, the Clovis culture rln this artiele the terms Early Man and Paleo- seems to have appeared soon after 12,000 yrs B.P. Indian have the same meaning and refer specif- According to Haynes (1970), there is some meager ically to the Clovis complex, the earliest eulture evidence of pre-Clovis migrations to America, but found in Virginia, which dates to between 12,000 the representative tools of these possibly earlier and 10,000 yrs B.P. migrants are too poorly known to make comparisons with Clovis tools. 2Professor Emeritus of Modern Languages, Lec- Interest in Paleo-Indians in Virginia was stim- turer in Anthropology, College of William and ulated to a great extent by various archeological Mary, Williamsburg, Virginia 23185 finds made in the western part of the United 72 VIRGINIA DIVISION OF MINERAL RESOURCES

States ftom 1927 to the 1950's. The Dent site, For example, one collection of approximately found near Dent, Colorado in 1932, is generally 20,000 projectile points had only 6 Clovis points. accepted as the first site which yielded, in 1936 The Archeological Society of Virginia was or- and 1937, Clovis fluted points found in association ganized in 1940. A few years later it decided to with mammoth remains. The Blackwater No. 1 undertake the publication of a Quarterlg Bulletin. locality, New Mexico, in 1936 and 1937 yielded With this established, I announced in the De- the type specimens from which is derived the cember 1946 issue (Volume 1, Number 2) of the name Clovis fluted point (Wormington, 1957). As Quarterla Bulletin that in collaboration with the more sites were found in the western part of the Society, I would begin a continuing survey of United States it became apparent that these early Clovis points found in Virginia. The entire thirty- hunters covered much territory in pursuit of four pages of the September 1947 issue of the large animals, such as the mammoth, mastodon, Quarterla Bulletin was devoted to the first fluted- camel, horse, and other members of the late point survey report, which listed and described Pleistocene fauna. Associated with the skeletal 126 Clovis points found in Virginia. The results remains atthe kill sites were distinctive projectile of 23 surveys now have been published in the points which, as already stated, became known Quarterlg Bulletin, givin g pertinent inform ation as Clovis fluted points. on more than 700 Clovis points found in Virginia (Figure 1). The interest aroused by the surveys in the CLOVIS CULTURE IN VIRGINIA Paleo-Indian and his tool kit has led to the discovery of twelve small Clovis sites in Virginia. My main objective with this paper is to review Six of them have been reeorded, and the remain- much of whatwe have learned since the late 1940's ing six are being investigated by their finders about the wide but, in most cases, thinly spread and will be recorded at later dates in the Quar- Clovis culture in Virginia. In the early 1930's, terly Bulletin. I began examining collections of Indian artifacts The largest of the six small sites already re- throughout Virginia. I was amazed to learn that corded has produced approximately 100 Paleo- the number of Clovis points, even in large col- Indian tools and 12 Clovis points. The smallest lections of artifaets from Virginia, was so small. of the six sites has yielded about 20 Paleo-Indian

a7 ilUilBER OF CLOVIS POINTS REPORTEO m AREA OF GREATEST CONCENTRATION a SMALL CLOVIS SIT€S w WILLIAilSON SITE s SALTVILLE

".---|.-l*..

Figure 1. Map of Virginia counties and larger Tidewater cities showing number of Clovis points reported for each political unit. Notiee the high concentration of points in the southern Piedmont and Tidewater and in and near Smyth County in southwest Virginia. PUBLICATION 75 73 tools and 2 Clovis points. The six recorded sites materials on the Williamson site. When the site are located in Isle of Wight (1), Dinwiddie (1), was diseovered in 1949 it was eovered with thou- Chesterfield (2), and in Sussex (1) eounties and sands of unutilized flakes and with many large the City of Suffolk (1). Adjacent to the Sussex and small cores. The materials provided by the County site, two small outcrops of chert and outcrops consist, in layman terms, mainly of chalcedony were found. The six small Paleo- variegated chalcedony and chert with cream, Indian sites which are being investigated and gray, brown, and blue colors dominating. A have not been reeorded in the Quarterla Bulletin bluish-black chert also occurs there. Small are located in the following counties: one in Ame- poekets of minute crystals are often found in the lia, one in Greensville, three in Dinwiddie, and chalcedony and chert, and they doubtlessly led one in Frederick. The Paleo-Indians wanted to to breakage when flaking was attempted. Flakes be near good outcrops ofcryptocrystalline quarlz, and cores of white quartz, and of gray and brown and they made efforts to find them. More than quartzite, found in small quantities on the site, half of the twelve small sites just mentioned were could have been of local provenanee. Rock crystal, located near an outcrop of the preferred lithic found sporadically as flakes and implements, was materials. probably brought in from sourees higher up in One of the largest and most prolific Paleo- the Piedmont area. Petrified wood, represented Indian sites in the United States, the Williamson by very few flakes, eould have had its origin in site, was discovered in the summer of 1949 in Prince George County which adjoins Dinwiddie Dinwiddie County, Virginia (Figure 1). From County. 1946 to l-949 three survey reports had appeared The proximity of rivers, creeks, springs, and in the QuarterlA Bulletin, and the five persons swamps was desirable to the Paleo-Indian, and primarily responsible for the discovery of the the presenee of large game played an important Williamson site had profited by the helpful in- part in the selection of a site. The Williamson formation provided by the survey reports. I was site probably met most of these requirements. fortunate in having been associated with the Little Cattail Creek was likely in existence and finding and identification of the site whieh has flowed by the site. Stony Creek was slightly under now become recognized as one of the largest two miles and Rowanty Creek slightly over two Paleo-Indian workshop and settlement sites in miles away. The Nottoway River was about ten North Americd. The Williamson site is about 4.5 miles to the south and the Appomattox was ap- miles east of Dinwiddie Courthouse. It lies on an proximately the same distance to the north. The elevated stretch of land between Little Cattail large game, if not the mastodon, was presumably Creek on the north and a very small ereek on the deer, the small black bear, and the elk. It the south known as Health Meadows. The cul- should be remembered that no mastodon or mam- tivated land on which the site is widely scattered moth bones have been found in an eastern North constitutes more that 75 acres. American Paleo-Indian habitation site (Gramly, I prepared a report on the site that was pub- 1e82). lished in American Antiquitg in July, 1951. At Our surveys indicate that there was also a that date, two years after its discovery, the site sizable coneentration of Paleo-Indians in Mec- had yielded 33 Clovis points, 174 end scrapers, klenburg County, Virginia, which is about 55 214 side scrapers, 16 preforms, several gravers, miles southwest, as the crow flies, of the Willi- and many rough cores (McCary, 1951). Today, one amson Paleo-Indian site. Mecklenburg County might estimate conservatively that the site has has yielded about 142 Clovis points found on produced 200 Clovis points. One might also es- farms near Chase City and Clarksville. However, timate that the site has yielded approximately no large eommunity site similarto the Williamson 1200 to 1400 end scrapers, I"000 side scrapers, site has been found in Mecklenburg County. and other tools in smaller numbers including It should be noted that the concentration of knives, perforators, gravers, spokeshaves, Clovis points in Mecklenburg County extends into choppers, wedges, drills, and hammerstones (Fig- Granville County, North Carolina which adjoins ures 2-4). Mecklenburg County. A. D. Capehart of Oxford, As already stated, the Paleo-Indian preferred North Carolina reported to me in September, cryptocrystalline quartz (chalcedonies, flints, 1948, that over a period of years he had collected cherts, and jaspers) to other materials for tool 62 Clovis points from Granville County (McCary, making. Evidence indicates that there were prob- 1948). Since that time the number of points from ably several outcrops of these preferred lithic Granville County has increased to 96. Alamanee 74 VIRGINIA DIVISION OF MINERAL RESOURCES

Figure 2. Clovis points from three areas of concentration in Virginia. (1-3) Mecklenburg County; (4-6) Saltville, Smyth County; (7-20) Williamson site, Dinwiddie County. 74 VIRGINIA DIVISION OF MINERAL RESOURCES

Figure 2. Clovis points from three areas of concentration in Virginia. (1-3) Mecklenburg County; (4-6) Saltville, Smyth County; (7-20) Williamson site, Dinwiddie County. PUBLICATION 75 75

Figure 3. Clovis tools from the Williamson site (A). (1-3) large biface blade knives; (4) side scraper; (5, 6) spokeshaves; (?) distal end of broken Clovis point converted to end scraper; (8) end scraper; (9) twist drill; (10) expanding drill; (11)twist drill; (12-14) end scrapers with gravers; (15) end scraper. PUBLICATION 75 75

Figure 3. Clovis tools from the Williamson site (A). (1-3) large biface blade knives; (4) side scraper; (5, 6) spokeshaves; (7) distal end of broken Clovis point converted to end scraper; (8) end scraper; (9)twist drill; (10)expanding drill; (11) twist drill; (12-14) end scrapers with gravers; (15) end scraper. 76 VIRGINIA DIVISION OF MINERAL RESOURCES

Figure 4. Clovis tools from the Williamson site (B). (1) preform; (2) broken preform; (3) graver; (4) triangular knife; (5, 6) broken preforms; (7) wedge; (8) side scraper; (9) humpback scraper; (10) triangular knife; (11) wedge; (12) two spokeshaves on a flake; (13-15) perforators; (16) scraper with two spokeshaves; (17) humpback scraper. 76 VIRGINIA DIVISION OF MINERAL RESOURCES

County, North Carolina has the second largest Exeavations were made at the Williamson site concentration of Clovis points in the State; this between L972 and 1975 in an effort to obtain consists only of 12 points. Furthermore, a look suitable material with which to absolutely date at the map showing the distribution by counties the Paleo-Indian occupation. By the end of the of Clovis points found in Virginia (Figure 1) will summer of 1975 the excavation consisted of 11 show that the Paleo-Indians in Virginia preferred completed ten-foot squares and two five-by-ten living south of the James River, assuming the cuts. Artifacts were prolific in most of the pattern of reported Clovis points is indicative of squares. Unfortunately, we found neither recog- the relative distribution of Paleo-Indians in the nizable hearths nor datable charcoal samples. We State. Of the 100 counties in Virginia (ineluding also were unsuccessful in our efforts to find pollen five recently reorganized as cities), 26 counties samples. We were disappointed that we did not south of the James River, including Bedford and obtain our major objective which was to date the Pittsylvania at the western boundary, have pro- site. However, further efforts should be made to vided approximately 85 percent of the fluted obtain radiocarbon dates at some of our more points recorded for Virginia. promising small sites in the State. Announcements were made in two early issues of the Quarterlu Bulletin which aroused great AGE OF CLOUS CULTURE hope and expectation of finding an association between Paleo-Indians and extinct megafauna. Before we look at several radiocarbon dates of R. W. Pickle (see McCary, 1947) of Saltville, Clovis sites in the eastern and northeastern parts Virginia, stated that he had found near Saltville of the United States, we should note that it has the broken base of a Clovis point "in a sand and been pointed out that the ten laC-dated Clovis sites gravel stratum bearing numerous mineralized in the western part of the United States are mastodon bones." Pickle (McCary, 1952) reported between 11,000 and 11,500 yrs B.P. (West, 1983). another find of a Clovis point in Saltville at a Pennsylvania, Nova Scotia, and several states depth of 36 to 40 inches "in a sandy like clay soil." in the northeastern part of the United States have A few days later at a still lower depth "a large Clovis sites which have been radiocarbon dated tooth" identified as that of a young mastodon was in recent years. The Debert site in Nova Scotia found. The article stated that no association be- produced a series of dates with an average date tween the Clovis points and the mastodon bones of 10,600 yrs B.P. (MacDonald, 1969). A Paleo- and tooth was elaimed. We are inelined to believe Indian site in western Connecticut, 6LF21, that it was in both of the finds a case of "so close yielded a date of 10,190+300 yrs B.P. (Moeller, but still so far," and that the potentiality to prove 1980). The Vail site, an encampment site in an "assoeiation" at some future date is still in northwestern Maine, produced a date of Saltville. 11,120+180 yrs B.P. (Gramly, 1982). The Shaw- The surveys of Clovis points add proof to the nee-Minisink site, near Stroudsburg, Pennsylva- early opinion that Saltville and the surrounding nia, yielded a series of radiocarbon dates which region attracted the Paleo-Indians. The Quar- averaged about 10,500yrs B.P. (Eisenberg, 1978). terlE Bulletzrc of March 1983, gives the latest count Note that the Debert site and the Shawnee- of the distribution by counties of Clovis points Minisink site are within a few hundred years of found in Virginia (McCary, 1983). It reveals that being as old as the western Clovis sites. More laC 11 Clovis points have been reported from Smyth dates, especially from the east and southeast, County in which Saltville is located. Tazewell might narrow the gap now existing between the County, adjoining Smyth County on the north, has radiocarbon ages of eastern and western Clovis yielded 6 points and Washington County, adjoin- sites. ing Smyth County on the west, has reported 4 points making a total of 21 Clovis points for the CONCLUSIONS three-county area. All the other sixteen counties in the southwestern tip of the State from Craig, Some fifty years of recording Clovis fluted Montgomery, Floyd, and Patrick counties inclu- points discovered in Virginia has shown clearly sive have produced a total ofonly 15 Clovis points. and convineingly that the Paleo-Indian makers Therefore, we eannot diseredit the importance of of these points occurred throughout the State the role that Saltville and its environs might play (Figure L; McCary, 1983). The distribution of one day in furthering our understanding of the points by county and city suggeSts that the south- Paleo-Indian in Virginia. ern Piedmont-southern Tidewater area stands 78 VIRGINIA DIVISION OF MINERAL RESOURCES

apart as a region of relatively great Paleo-Indian environments of the Central Great Plains: activity within the State. The most significant Dept. Geol., Univ. Kansas, Spec. Publ. 3, p. concentration of reported Clovis points from the 77-92. western part of Virginia is from the Smyth- MacDonald, G. F., 1969, Debert: A Paleo-Indian Tazewell-Washington-county area. Site seleetion site in central Nova Scotia: Nat. Mus. Canada, by Paleo-Indians was probably influenced by the Anthropol. Papers, no. 16, 207 p. availability of good water, game, and outcrops McCary, B. C., 1947, A survey and study of of cryptocrystalline quartz. Folsom-like points found in Virginia: Ar- Ongoing and future excavations of known or cheol. Soc. Virginia, Quart. Bull. vol. 2, p.l- suspected Paleo-Indian sites in Virginia should 24. strive to obtain absolute dates for typologically 1948, A report on Folsom-like points found or culturally distinet horizons, and to reconstruct in Granville County, North Carolina: Archeol. the environments utilized by Paleo-Indians as -, Soc. Virginia, Quart. Bull. vol. 3, p. 5-16. much as possible with the use of geomorphic, 1951, A workshop site of Early Man in pollen, and faunal data. Dinwiddie County, Virginia: American An- -, tiq. vol. 17, p.9-17. Survey of Virginia-Folsom points nos. 173-2L9: Archeol. Soc. Virginia, Quart. Bull. -,1952,vol.6, p. 1-7. 1983, The Paleo-Indian in Virginia: Ar- REFERENCES cheol. Soc. Virginia, Quart. Bull., vol. 38, p. -, 43-70. Eisenberg, L., 1978, Paleo-Indian settlement pat' Moeller, R. W., 1980, 6LF21: A Paleo-Indian site tern in the Hudson and Delaware River drain- in western Connectieut: Washington, Conn., ages: Occasional Publ. Northeastern An- Ameriean Indian Archaeol. Inst., 160 p. thropol., no. 4, 159 p. West, F. H., 1983, The antiquity of man in Amer- Gramly, R. M., 1982, The Vail site: A Paleo- ica,inPorter, S. C., ed., The Late Pleistoeene: Indian encampment in Maine: Bull. Buffalo Minneapolis, Univ. Minnesota Press, p. 364- Soc. Nat. Sci., vol. 30, 169 p. 382. Haynes, C. V., 1970, Geochronology of man-mam- Wormington, H. M., 1957, Ancient man in North moth sites and their bearing on the origin of America. Denver Mus. Nat. Hist., Popular the Llano Complex, in Dort, W., Jr., and Series no. 4, Fourth Edition, Revised 1957, Jones, J. K., Jr., eds., Pleistocene and Recent 322p. PUBLICATION 75 79

LATE PREHISTORIC AND PROTOHISTORIC LARGE MAMMAL ZOOGEOGRAPHY

OF VIRGINIA

Robert K. Rosel

CONTENTS

Page

ILLUSTRATION

Figure

TABLES

1. Mammals 2. Mammals from early 3. Distribution of Virginia mammals today ...... 85

ABSTRACT INTRODUCTION

Evidence derived from 18late prehistoric (mid- The purpose of this paper is to evaluate the dle and late Woodland Period) archeologieal sites, distribution of the land mammals of Virginia, from several early historical aceounts, and from especially the large mammals, that were present the current understanding of the distribution of during the late prehistorie and protohistoric pe- Virginia mammals indicates that the large mam- riods and to eompare them with today's Virginia mal fauna of the Commonwealth has not changed mammals. The late prehistorie eomponent of Vir- substantially within the past 4,000 yrs. Some ginia's faunal history, arbitrarily defined as that species (e.g., bison, elk, timber wolf, and moun- representing the 2,000 years before European tain lion) have been extirpated since the settle- settlement, corresponds approximately with the ment of Virginia by Europeans; some previously middle and late Woodland Period. The informa- extirpated species (e.g., porcupine, eoyote, and tion on the mammals of this period eomes from beaver) have been naturally or artificially rein- excavated archeological sites that have been troduced during the historical period, and others dated either by tr6 methods or by distinguishing (e.g., woodchuck and red fox) probably have pottery or other artifacts associated with the sites expanded their distributions as a result of ehang- or strata. The protohistoric (trereafter called ing land-use patterns in the Commonwealth. "early historieal") record is reconstructed from the writings of early Virginia settlers, explorers, or travelers. Early writings date from slightly before A.D. 1600 and extend into the 18th Cen- rDepartment of Biological Sciences, Old Do- tury; the later accounts are from the western minion University, Norfolk, Virginia 23508. sections of the Commonwealth, where European 80 VIRGINIA DIVISION OF MINERAL RESOURCES

settlement occurred much later than it did on the (Didelphis uirginiana), and the coyote (Canis Coastal Plain. The record of modern Virginia latrans), which has reappeared within the past mammals comes from a number of sources: two five years in two western counties (J. Pagels, pers. books on the distribution of Virginia mammals comm.), might be ineluded by some authors. (Bailey, 1946; Handley and Patton, 1947), other However defined, the list of large mammals in compilations (Hamilton and Whitaker, L979; Virginia is short. In this paper, I have operation- Hall, 1981), recent literature of mammalogy, and ally defined "large mammals" as those that were correspondence and conversations with other sufficiently large both to be used as food by the mammalogists within the Commonwealth. Woodland Period Indians and to be observed and recorded in the early historical reeord. With this LARGE MAMMALS method, mammals of squirrel size and larger are included in the study sample. The mammals that were present for the 2,000 years before European settlement were the sur- vivors of the late Pleistocene and early Holocene THE LATE PREHISTORIC RECORD extinctions that occurred in the region. Informa- tion about these extinctions is to be found in other The best direct information for the 2,000 years chapters within this volume, especially the pap- before settlement has been derived from the ers by Eshelman and Grady, and by McDonald. excavations of archeologieal sites of the Woodland Other viewpoints on the late Quaternary dynam- Period. Although there are some difficulties with ics of the mammalian faunas of eastern North the use of these records, this is a fruitful and America are provided by Guilday (1971), Handley expanding area of research. Since 1968, infor- (1971), and by the recent excellent book Pleis- mation from archeological sites has provided a tocene Mammals of North America (Kurten and fair estimate of the mammals that were present Anderson, 1980). The main point from these re- at the time the sites were occupied. However, views is that the large mammal fauna of late several of the papers reviewed here indicate that Quaternary Virginia suffered extensive extinc- only a fraction of the faunal remains recovered tions and, as a result, few species of large mam- from the sites had been evaluated. Thus, the mals survived to the late prehistoric period. information, although accumulating rapidly, is When applied to mammals, any definitions of still fragmentary in terms of its utility in recon- "large" and "small" will have difficulty with structing the distributions of the mammals of the universal acceptance. Kurten and Anderson late prehistoric period. (1980, p. xiv) distinguish between micromam- One of the difficulties with the mammal record mals (orders Inseetivora, Chiroptera, Rodentia, as determined from archeological sites is that and Lagomorpha) and macromammals (orders mammal remains tended to be preserved either Edentata, Carnivora, Artiodactyla, Perissodac- in caves (which are restricted to western Vir- tyla, and Proboscidea). By this system small ginia) or in shell middens in the Piedmont or carnivores such as the weasels, which may weigh Coastal Plain regions. Elsewhere, acid soils gen- as little as 0.1 pounds (50 grams), are macromam- erally have destroyed the evidence that mammals mals. A more common approach is to define an were present. Consequently, there is an uneven arbitrary body weight, for example 11 pounds mammal record within the Commonwealth, with (5 kg) (Snyder, 1978), to differentiate large from several exeellent sites in the Valley and Ridge small mammals. With this method, many of the region, some information from sites in the Coastal small carnivores become small mammals, and Plain, and almost nothing from the Piedmont some rodents (e.g., beaver, Castor canadensis, region (Figure). Although mammals living near and woodchuck, Marmota monan) become large caves or streams may be overly represented, we mammals. Using the body weight criterion, there are more concerned with knowingwhat mammals are only eight species of large mammals in Vir- were present during the 2,000 years before set- ginia today, ineluding woodchuck, beaver, red fox tlement than in knowing where they lived. Fi- (Vulpes uulpes), black bear (Ursus americanus), nally, archeological sites contain samples of mam- raccoon (Proeyon lotor), river otter (Lutra cana- mals biased in favor of large mammals that were densis), bobcat (Felis rufus), and white-tailed taken as food and then preserved in garbage pits, deer (Odocoileusuirginianu.s). The gray fox (Uro= middens, or graves. This bias is not a problem cyon cinereoargenteus), which weighs slightly less here because we are primarily interested in the than 11 pounds (5 kg), the Virginia opossum larger mammals of the Commonwealth. PUBLICATION 75 81

village as was often done with deer (Holland, 1979). Under these eonditions, it is unlikely that large numbers of bones would be returned to the village and to its garbage pits. Elk (Ceraus elephas), which are intermediate in size between white-tailed deer and bison, were present at half of the 18 sites that were evaluated in this paper (Table 1). Beavers and woodchucks were found at most Figure. Location of the 11 montane, B piedmont, of the 18 sites; both are large and easy to hunt and 4 coastal plain late prehistoric archeological because of their fidelity to lodges and burrows sites that have been reviewed in this study. near forest edges, respectively. Foxes, raccoons, and gray squirrels were also common. The flying squirrel (probably southern flying squirrel, Glaucomgs uolans) is a species that might be, but The mammals of the late prehistoric period has not been, found at archeological sites. Despite were similar to those present today (Table 1); each the fact that these squirrels are nocturnal, there can be assigned clearly to an extant species. The are several reports of their presence in the early mammals that were collected at several archeo- historical accounts. logical sites included the Virginia opossum, east- In Table 1, the faunas from archeological sites ern cottontail (Sylui,lagus florid,azzs), seven spe- have been grouped according to geographic lo- cies of rodents (including tree squirrels and cation, starting in the southwest and moving beaver), four canids, one bear, one raccoon, three northward and eastward. The faunas were mustelids, two felids, and two cervids, for a total grouped from west to east in an effort to uncover of 22 species. In addition to the species that were unusual distributional patterns, but none was present at many sites, several other species in- revealed. A few species, such as the woodrat cluding mole, white-footed mouse (Peromgscus (Neotoma floridana), today are restricted to the leucopus), rice rat (Orgzomys palustri,s), meadow western part of the Commonwealth and they v ole (M icr otus p ennsy la anicus), poreup i ne (Ere- might be assumed to have been similarly res- thizon dorsatum), spotted skunk (Spilogale puto- tricted during prehistoric times. However, such rius), and long-tailed weasel (Mustela frenata), a pattern of distribution cannot be detected with were recorded at one or two sites. (The "mole" certainty in this small sample of 18 archeological is probably the eastern mole, Scalopus aquaticus, faunas. the mole with the widest distribution in Virginia In some studies, notably Benthall's (1979) from today.) Dougherty's Cave in Russell County and Barber's The white-tailed-deer is the most common mam- (1981) study of Maycocks Point Shell Midden in mal at these sites. This is not surprising because Prince George County, good informstion is avail- deer probably were abundant throughout Vir- able for long time periods. Dougherty's Cave was ginia, and because of their large size, each suc- a stratified site, with the lower layers dating from cessful hunt provided much food. Conspicuously the middle Archaic Cedar Creek Period (Ben- absent from Table 1 is the American bison, Bison thall, 1979). Barber (1981) commented that nei- bison, although there is one bison tooth known ther the mammals eaten nor their proportions from nearby Currituck County, North Carolina changed in the raC-dated layers at the Maycocks dated about 2,610 yrs B. P. (Painter, 19?8). Ac- Point location; the maximum established age cording to early historical accounts, the bison, difference between the layers was about 500 or American buffalo, was common throughout years. These studies reinforce the observation much of Virginia, except for the Coastal Plain. that no readily apparent changes in the mam- One explanation for the absence of bison from malian fauna within the past 4,000 years have archeological sites is that montane caves and been revealed from the study of archeological Coastal Plain shell middens were far removed sites. from places where bison were common. Another In sum, the archeological record indicates that explanation is that bison would be too large to between 22 and 29 species have been preserved earry back to the Indian villages. Consequently, at one or more sites. White-tailed deer predom- the careass could have been dismembered at the inate almost everywhere, but the Woodland Pe- kill site, and the usable parts earried back to the riod Indians eollected mammals of all sizes and 82 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 1. Species of mamals recorded at late prehistoric archeological sites in Virginia. "X" indicates presence, "O" indicates a conspicuous absence, and "?" refers to a specimen referrable to one taxon but possibly a congener. For scientific names, refer to Table 3.

o o =!!:; ! .: --E';r o tr=?';.:F; d tr ;€E*aA€rEI ;siE.*.cg; 60 Appalachian Highlands E€=E5.}53i€FgEEiErE s a >3.=€ x; E= * k:! E E i sB o <) Q srr rr F frZ 65 a() d m c, 2d 3* dB Benthall 1979 xxxx xxxx xx x x x MacCord 1981 xxx x x x x MacCord & Buchanan 1980 x xxx xxx xxx xx xx x Funk 1976 x xx ?x MacCord 1972 xxx xxxx xx xx xxx x Barber & Baroody 1977 x xx x xx x XX Buchanan 1980 xx x x x xx MacCord 1976 XX X XX X xx Johnson 1979 xxxx xx MacCord 1973a x x xx ?xx xx xxx xxx MacCord 1973b x xx x x x Piedmont Waselkov 1977 xx XX xxx xx xx xxx Stevens 1979 x x xx x Egloff & others 1980 xx xx XXX XX ox x Coastal Plain Barber 1981 x xx xxx xxxxx xxxx Owen 1969 XX X xxx xx x x Barber 1978 xx x xxx X x Geier & Barber 1983 XX xx xx x

of most kinds, even some that might be considered (1) Although some of the early settlers rtrere too small to eat. educated in England and some even knew the English biota fairly well, the 16th- and 17th- THE EARLY HISTORICAL RECORD Century natural philosophers were trained more in Aristotelian principles than in the need for The written accounts of early settlers in the careful observation and description. In faet,"nat- New World would seem to be excellent sources ural history" did not begin in England until the of information about the contemporary fauna publication of White's Natural History of SeI- because they describe species actually seen by the bourne in 1789. William Strachey's 1612 manu- observers. However, there are several shortcom- script, published under the title Historie of Tra- ings with this information. uell in Virgi,nia Britania, is regarded as one of PUBLICATION 75 83

the best historical sources of this period. Quotes metaphors seem to be common features of many from his (and other) writings will indicate the early historical accounts. Comments on the size, qualitative nature of the information in the early ferociousness, and numbers of mammals, espe- historical record: cially large mammals, are often exaggerated. "The Beaver there is as big as an ordinary water "They have diverse beasts fit for provision, the dog but his legs exceeding short, his forefeet like chief are deer, both red and fallow, great store a dog's, his hinder like a swan's, his tail somewhat in the country towards the heads of the rivers, like the form of a raeket, bare without hairs, though not so many amongst the rivers; in our whieh to eat the Savages esteem a great delicate." island about Jamestown are some few, nothing (2) Much of what the early observers described differing from ours in England but that of some was based upon comparison with mammals of of them the antlers of their horns are not so many, their experience in England. Consequently, our our people have seen 200, 100, and 50 in a herd," white-tailed deer is described variously as being (Strachey, L6l2). Clayton's statement that the similar to red, roe, or fallow deer, our rabbits "Rachoone, I take it to be a species of a monkie" to their hares, and so on. Furthermore, some (Berkeley and Berkeley, 1965) hardly instills American mammals, such as opossum and rac- confidence in the modern reader that Clayton was coon, had no counterparts in England or Europe. a trained scientist of his time. After two years These were often called by their Indian names in America, Clayton returned to England and and described as being similar to rats or monkeys, spent the next 40 years of his life as a major and respectively. The descriptions sometimes were influential spokesman in establishing a picture very crude, as seen in another Strachey quote: of the New World in European minds. "There is a beast they eall Aroughcoune, much With these seemingly casual, vague, and flimsy like a badger, tailed like a fox, and of a mingle accounts as the norm, it is remarkable that some black and grayish color, and whieh useth to live wildlife accounts did indicate that eareful obser- on trees as squirrels do, excellent meat, we kill vations had been made. For example, several often of them, the greatest number yet we obtain writers commented on flying squirrels, small by trade." Strachey is describing a raccoon. nocturnal rodents that are difficult to observe. (3) The writer often relied on second-hand Clayton (Berkeley and Berkeley, 1965) reported information or hearsay. This is seen in many seeing two kinds of bats, "one a large sort with accounts, sueh as those of Reverend John Clayton, long ears," and "the other much like the English educated at Oxford University and well known something larger I think, very common." He must for his scientifie investigations before coming to have observed the big-eared bats in the genus America for two years to serve as Rector to James Plecotus. City Parish from 1684-1686. "Elke, I have heard The early historical accounts indieate that the of them beyond the inhabitants . . .", indicating mammals first observed by settlers in the 17th that elk were absent from the Coastal Plain Century, and even a century later, did not differ (Berkeley and Berkeley, 1965). Seeond-hand ac- substantially from the faunal lists based on ev- counts are not surprising from a reetor, who idence from late prehistorie areheologieal sites would have had little reason to travel far and (Table 2). Compared to Table 1, there are some wide, but who would likely be in contact with differences; for example, the woodrat, wood- many people who did travel. We have to rely on chuck, porcupine, and red fox were not reported the good judgment of the rector to sift fact from in historical accounts. Today, the woodrat is fiction, and to reeord the best information pos- restricted to the western part of Virginia and its sible. Accounts are most accurate and most be- absence in early historical aecounts is not unex- lievable when the observations are made by the pected. However, the woodchuck was present at writer. most archeological sites from the western half ( ) The early historical accounts are largely of Virginia, but is absent from the eastern sites from the Coastal Plain, where the earliest set- and from the early historical aceounts. The wood- tlements occurred. Journeys to the south and west chuck would have been easy to observe because brought back information about the wildlife it is large, aetive during the day, prefers forest there, but this information usually became sec- edges and clearings (even near buildings), and ondhand for the writer of the accounts. digs large and noticeable holes in the ground. I (5) None of the early historical accounts was conclude that the woodchuck was not present in written by a woodsman. eastern Virginia during the time of the early (6) For a variety of reasons, hyperbole or faulty settlements. but that it has moved eastward in 84 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 2. Species of mammals recorded in early historical accounts in Virginia. "X" indicates presence and "?" refers to a specimen referrable to one taxon but possibly a congener. For scientifie names, refer to Table 3. 6I ;;.!-Eci- eE i, 'is i€ tta E*a6€aEI Ex€E -. srt 3E sE 5E H 6i 6€r'l E ; fiiEE: 5Ft65 xF;T e ii d=d5 EI:"fiF E ii Clayton 1684 xx x xx xx xxx xx = xx (Berkeley and Berkeley 1965) Hariot 1588 xx xx xxxx Smith 1612 xx xxx XX xxxx x (Arber f967) Beverley 1705 (Dunbar 1964) Argall 1613 (Barbour L972) Byrd 1728 x xx xxx (Boyd 1929) Banister 1678 xx xxx xx x xx (Ewan and Ewan 1970) Strachey 1612 xx xx ? xx xxxxxx, x 17th Century xx x xx xx xx x (Bruce 1927)

association with the clearing of the land for of Virginia, and also at Waselkov's (1977) site in agriculture. In fact, even today populations do Franklin county and Barber's (1981) site in not occur east of Williamsburg. The porcupine Prince George County. The latter site is located probably was common in western Virginia, and in the Coastal Plain, but the elk remains were it is unclear why the presence of this easily seen found in Barber's Zone 4. which was dated A.D. and distinctive mammal was not recorded in an 245t90 yrs. Handley and Patton (L947) state that early account. The red fox, also missing from the last specimen of elk was killed in January these historical aceounts, is another species that 1855 by Colonel Joseph Tuley of Clarke County. requires clearings or openings, and its distribu- Bison were reported in some of the early his- tion too may have expanded rapidly with the torical accounts (Table 2). Handley and Patton spread of agriculture. (1947) contend that "when the settlers first came, The presence of elk was reported only onee in the bison . . . was quite common throughout the the early historical aceounts, and this was the State, at least down to the edge of the Coastal hearsay comment by Clayton. Either elk were not Plain." Dunbar (1964) believed that the 45 place eommon in the Coastal Plain, or else they were names involving "buffalo" are helpful in finding considered to be one of the deer that are similar the former distribution of the bison in Virginia; to "red, roe, or fallow deer." (It is possible that none was located well into the Coastal Plain. elk were mistakenly called red deer, and white- Barbour (1972) reported a June 1613 letter from tailed deer were called fallow deer.) In the ab- Captain Samuel Argall that described seeing "a sence of careful description, we cannot be sure great store of cattle (bison)" grazing along the which explanation is more likely. Elk were pres- banks of the Rappahannoek River "about 65 ent in most archeological sites in the western part leagues" from the Chesapeake Bay. This location PUBLICATION 75 85 was probably close to the Fall Line between the Table 3. Virginia mammals and their likely Piedmont and the Coastal Plain. According to distributions within the Commonwealth, taken Handley and Patton (7947), the last bison was from Hall ( 1981), The Mammals of North America killed in Virginia in the 1790's. and Hamilton and Whitaker (1979) Mammals of Other extirpations recorded by Handley and Eastern North America, using the nomenclature Patton (L947) include the fisher, Martes pennanti ofJones, Carter, and Genoways (1979). (probably in the 1890's), pine marten, Martes americana (probably in the 1830's or 1840's), poreupine (1899), timber wolf, Canis lupus Distribution is statewide for: (19t2), mountain lion, Felis concolor (which were Didelphis u'ir g'in'iana, Virgin ia opossum hunted in the 1880's), beaver (about 1910), and Soren longirostris, southeastern shrew (absent in white-tailed deer in 44 western and central coun- mtns?) ties (about f905). Some of these species, notably Microsoren hoai, pygmy shrew the white-tailed deer and beaver, have been rees- Blarina breuic auda, short-tailed shrew tablished and are doing well today throughout Cryptotis paraa, least shrew the Commonwealth under the protection of reg- Scalopus aquaticus, eastern mole ulated hunting and trapping. C ondy lur a cristata, star-nosed mole Myotis lucifugus,little brown bat MODERN MAMMALS OF VIRGINIA Mgotis keen'ii, Keen's myotis L asiony cteri s no ctiu agozs, s ilve r- hai red b at The current mammalian fauna of Virginia is Pi,pi str ellus subfl aau s, easte rn p i p istrelle similar to that which was present 400 years ago Eptesicus Jusczs, big brown bat when the early European settlements were es- Lasiurus borealis, red bat tablished, minus the extirpations mentioned Las'iurus cinereus, hoary bat above. There have been a few additions, including Nycticeius humeralis, evening bat (east of mtns?) Old World murid rodents (Norway rat, Rattus Ple c otus r afine s quii, Raf i ne sque's b i g-eared b at noruegicus, and house mouse, Mus musculas), the Sa lailagus florid,anus, eastern cottontail hispid cotton rat, Sigmodon hispidzs, which has Tamias striatus,eastern chipmunk (except south- moved northward and eastward during historical ern Va.?) times and was first recorded in Virginia in 1940, Marmota *ono*, woodehuek (except Southeast- and the South American nutria, Myocastor coy- ern Va.) pzs, which has moved into the southeastern Sciurus c ar olinensis, gray squirrel corner of Virginia after escaping from domes- Tamiasciurus hudsonicus, red squirrel (except tication in Louisiana during the 1930's. As dis- Coastal Plain and southern Piedmont) cussed above, it is likely that the woodchuck, and Glaucomys uolans, southern flying squirrel especially the red fox, have expanded their dis- Castor canadensis, beaver (extirpated, reestab- tributions substantially since the onset of Euro- lished) pean settlement. Some other species have been Reithr o dontomy s humulis, easte rn h arvest mouse introduced, such as the Sika deer (Ceruusnippon) Peromyscus leucopus, white-footed mouse and horse (Equus caballus) on Assateague Island Ochrotomys nuttalli, golden mouse (except north- and the black-tailed jackrabbit (Lepus ealiforni- ern Va.) cus) on Cobb Island, but these are considered to Microtus pennsyluanicus, meadow vole be localized exotics and not partofthe local fauna. Microtus pinetorum, woodland vole Table 3 lists the modern mammals of Virginia Ondatr a zib ethicus, mu skrat arranged according to their probable distribu- Rattus noruegicus, Norway rat (introduced from tion within the Commonwealth. Europe) In conclusion, the mammals of late prehistoric, Mus musculus, house mouse (introdueed from early historical, and today's Virginia are similar. Europe) In fact, the mammalian fauna does not seem to Zapus hudsonius, meadow jumping mouse have changed much during the past 4,000 years. Canis lupus, gray wolf (extirpated) This picture of constancy during this period is Vulpes aulpes, red fox (exeept extreme SE Va.?) consistent with the findings of Delcourt and Del- Urocy on cinereoargenteus, gray fox court (this volume) that neither the climate nor Procyon lotor, raecoon the vegetation of Virginia has changed substan- Mustela frenata, long-tailed weasel tially during the last 4,000-5,000 years. Each Mustela aison. mink 86 VIRGINIA DIVISION OF MINERAL RESOURCES

Mephitis mephitis, striped skunk (except extreme Mgocaster coupus, nutria, introdueed from South SE Va.?) America, southern Va. only Lutra canadensis, river otter (now east of mtns.) Fel'is concolor, mountain lion (probably extir- Distribution in Appalachian highlands and south- pated) eastern Virginia only for: Felis rufus, bobcat Sgnaptomys cooperi, southern bog lemming Ceraus elephas, wapiti (or elk) (extirpated) Ursus americanus. black bear O docoileus air ginianu.s, wh ite-tailed deer Bison bison, bison ("buffalo"), except SE Va. (extirpated) species present in the archeological record is readily assignable as a Virginia mammal known Distribution in Appalachian highlands for: from the historical period, although that species Sorer cinereus, masked shrew might now be extinct in Virginia (e.g., the timber Sorer palustris, water shrew wolf). As more and more archeological sites are Soren fumezs, smoky shrew excavated, the former distribution and status of Sorer d'ispar,long-tailed or roek shrew some species is certain to beeome better known. Parascalops breweri, hairy-tailed mole Although some species have been extirpated from Myotis grisescens, gray myotis Virginia sinee European settlement began and Myotis uel'ifer, cave myotis others have expanded their distributions within Myotis sod,alis,Indiana or social myotis Virginia in response to ehanging land-use patt- Pleeotus townsendii. Townsend's big-eared bat erns, the mammalian fauna appears to have re- Syluilagus transitionalis, New England cotton- mained remarkably static during the past four tail millennia. Lepus americanus, snowshoe hare (endangered) Sc'iurus niger, fox squirrel ACKNOWLEDGMENTS Glaucomy s sabrinus,northern flying squirrel (endangered) I thank Randy Turner of the Virginia Researeh Peromyseus maniculafzrs, deer mouse (N. Va.) Center for Archaeology, Mike Barber of the U. Neotomafloridana, eastern woodrat (N. Va.) S. Forest Service. Helen Rountree of Old Domin- Clethrionomy s gapperi,southern red-backed vole ion University, Elizabeth Reitz of the University Microtus chrotomhinas, rock vole (status under- of Georgia, Jim Purdue of the Illinois State Mu- termined) seum, and Charles O. Handley, Jr. of the National Napaeozapus insignis, woodland jumping mouse Museum of Natural History for their help in my Erethizon dorsatum, poreupine (extirpated) searches. I also thank Howard A. MacCord, Sr. Canis latrans, coyote (extirpated, recent immi- for his contributions to Virginia archeology, Jack grant) Cranford, John Pagels, Helen Rountree, Jerry M artes americana. marten (extirpated) McDonald, and Sam Bird for helping to improve Martes pennanti, fisher (extirpated, reeent im- the content and accuraey of this aecount. The migrant) errors that remain, of course, are my own. Debbie Mustelu niu alis. least weasel Miller-Carson of the ODU Center for Instruc- Spilogale putorius, eastern spotted skunk tional Development prepared Figure 1,.

Distribution in southern Piedmont and Coastal Plain for: Blarina c arolinens'is,southern short-tailed shrew REFERENCES Uyzomys palustris, marsh rice rat Sigmodon hispidus, hispid cotton rat Arber, E., 1967, Travels and works of Captain John Smith, President of Virginia, and Ad- Distribution in southeastern Virginia for: miral of New England: New York, Burt Lasiu,rus seminolus, Seminole bat Franklin, Res. Source Works Ser. no. 130, 382 Lasiurus intermediu.s, northern yellow bat p. Sylailagus palustris, march rabbit (Tidewater Bailey, J. W., 1946, The mammals of Virginia: only) Richmond, Williams Printing Co.,416 p. P er omy s cus g o s sE pinus, cotton mouse (T i dewater Barber, M. B., 1978, The vertebrate faunal anal- only) ysis of Jc 27 (James City County, Virginia): PUBLICATION 75 87

an exercise in plow zone archeology: Quart. C., ed., The distributional history of the biota Bull., Archeol. Soc. Virginia, vol. 32, p. 94- of the southern Appalachians, Part III: Ver- 100. tebrates: Blaeksburg, VPI & SU, Res. Div. 1981, The vertebrate faunal utilization pat- Monogr. 4, p.233-262. tern of the Middle Woodland Moekley ceramic Hall, E. R., 1981, The mammals of North Amer- -, users: the Maycock's Point Shell Midden Site, ica: New York, John Wiley & Sons, 1081 p. Prince George County, Virginia. Unpub- Hamilton, W. J., Jr., and Whitaker, J. O., Jr., lished rept., 23 p. 19?9, Mammals of the Eastern United States, Barber, M. B., and Baroody, J.C., 1977, Analysis 2nd ed.: Ithaca, NY, Comstoek Publ. Assoc., of vertebrate faunal remains from the Shan- 346 p. non Site, Montgomery County, Virginia: Handley, C. O., Jr., 1971, Appalachian mammal- Quart. Bull., Archeol. Soc. Virginia, vol. 31, ian geography-Recent Epoch, in Holt, P. C., p. 101-113. ed., The distributional history of the biota of Barbour, P. L., 1972, Further notes on bison in the southern Appalachians, Part III: Verte- early Virginia: Quart. Bull., Areheol. Soc. brates: Blacksburg, VPI & SU, Res. Div. Virginia, vol.27, p. 100. Monogr.4, p.263-303. Benthall, J. L.,1979, Dougherty's Cave, a strat- Handley, C. O., Jr., and Patton, C.P.,1947, Wild ified site in Russell County, Virginia: Unpub- mammals of Virginia: Richmond, Virginia lished rept., Virginia St. Lib., Richmond, Comm. Game and Inland Fisheries, 220 p. Virginia, 130 p. Hariot, T., 1588, A briefe and true report of the Berkeley, E., and Berkeley, D. S., 1965, The new found land of Virginia: London, Bernard Reverend John Clayton: Charlottesville, The Quaritch, 111 p. Univ. Press Virginia, 170 p. Holland, C. G., 1979, The ramifications of the fire Boyd, W. K., 1929, William Byrd's histories of hunt: Quart. Bull., Archeol. Soc. Virginia, vol. the dividing line betwixt Virginia and North 33, p. 134-140. Carolina: Raleigh, The North Carolina Hist. Johnson, L. D., 1979, The Hereules Site, Alle- Comm.,341p. ghany County, Virginia: Quart. Bull., Ar- Bruee, P. A., 1927, Social life of Virginia in the cheol. Soc. Virginia, vol. 33, p. 144-L47. seventeenth century: Lynchburg, J. P. Bell Jones, J. K., Jr., Carter, D. C., and Genoways, Company, Inc.,275 p. H. H., 1979, Common and scientific names for Buchanan, W.T., Jr., 1980, The Hall Site, Mont- the mammals of North America north of gomery County, Virginia: Quart. Bull., Ar- Mexico: Lubbock, The Museum, Texas Teeh cheol. Soc. Virginia, vol. 35, p. 72-100. Univ., Occas. Pap., 17 p. Dunbar, G. S., 1964, Some notes on Bison in early Kurten, B., and Anderson, 8., 1980, Pleistocene Virginia: Quart. Bull., Archeol. Soc. Virginia, mammals of North America: New York, Co- vol. 18, p.75-78. lumbia Univ. Press, 442 p. Egloff, K., Barber, M. B., and Reed, C., 1980, MacCord, H. A., Sr., 1972, Thompson's Shelter, Leggett Site: a Dan River agricultural/rive- Giles County, Virginia: Quart. Bull., Archeol. rine hamlet: Virginia Res. Cent. Archaeol. Soc. Virginia, vol. 27, p.36-57. rept., 45 p. The Hidden Valley Rockshelter, Bath Ewan, J., and Ewan, N., 1970, John Banister and County, Virginia: Quart. Bull., Archeol. Soc. his Natural History of Virginia, 1678-1692: -,1973a,Virginia, vol.27, p. 198-228. Urbana, Univ. Illinois Press,485 p. 1973b, The Quicksburg Site, Shenandoah Funk, T. C., 1976, Excavations at Fort Chiswell, County, Virginia: Quart. Bull., Archeol. Soc. an archeological perspective of Virginia's -, Virginia, vol.27, p. 121-140. western frontier: Univ. Virginia Lab. Ar- 1976, The Thomas Site, Montgomery ehaeol., 116 p. County, Virginia: Quart. Bull., Archeol. Soc. Geier, C. R., and Barber, M., 1983, The Skiffes Virginia, vol. 30, p. 113-135. Creek Site (NN?): a multicomponent Middle 1981, The Sullins Site, Washington County, Woodland Camp in York County, Virginia. Virginia: Quart. Bull., Archeol. Soe. Virginia, Unpubl. rept. to Hampton-Newport News Re- -, vol. 36, p. 99-121. gional Development and Housing Authority, MacCord, H. A., Sr., and Buchanan, W. T., Jr., 266p. 1980, The Crab Orchard Site, Tazewell Guilday, J. 8., 1971, The Pleistoeene history of County, Virginia: Archeol. Soc. Virginia, the Appalachian Mountain fauna, inHolt,P. Spec. Publ. no. 8, 156 p. 88 VIRGINIA DIVISION OF MINERAL RESOURCES

Owen, R. M., Jr., 1969, Martin Farm, New Kent Strachey, W., (MS 7612), The historie of travell County: Quart. Bull., Archeol. Soc. Virginia, into Virginia Britania, London, 1949, edition vol,24, p.81-116. by Louis B. Wright and Virginia Freund from Painter, F., 1978, Bison remains from the Cur- the manuscript at Prineeton Univ.: London, rituck Site: The Chesopiean, vol. 16, nos. 1- Hakluyt Society, 1953. 3, p. 28-31. Waselkov, G. A., 1977, Prehistoric Dan River Snyder, D. P., 1978, Populations of small mam- hunting strategies: M. A. thesis, Dept. An- mals under natural conditions: Univ. Pitts- thropol., Univ. North Carolina at Chapel Hill, burgh, Pymatuning Lab. Ecol., Spec. Pub. no. 146 p. 5,237 p. White, G., 1789, The natural history and antiq- Stevens, T., 1979, The Wade Site (44Ha34): A uities of Selbourne, in the county of Sou- flood-damaged Woodland site: Quart. Bull., thampton: with engravings and an appendix: Archeol. Soc. Virginia, vol. 33, p. 148-159. London, T. Bensley,12 p. PUBLICATION 75 89

ON THE STATUS OF QUATERNARY VERTEBRATE PALEONTOLOGY AND ZOOARCHEOLOGY IN VIRGINIA

Jerry N. McDonaldl

CONTENTS

Page

Quaternary vertebrate paleontology in Virginia 92

ILLUSTRATIONS

Figure 1. Unpublished Quaternary vertebrate faunas in Virginia ...... '. 94 2. Recently published paleontological and zooarcheological localities in Virginia ...... 98

TABLES

Table 1. Unpublishecl Quaternary vertebrate faunas in Virginia ...... 95 2. Vertebrate remains from prehistoric archeological sites reported in the ASY Quarterla

ABSTRACT collecting and study of prehistoric Quaternary vertebrates, though, has been increasing since the The first vertebrate fossils found in North middle of the 20th Century and is currently being Ameriea by Europeans were recovered in Vir- conducted on an unprecedented level. Extensive ginia; seientific archeology in North Amerieawas collecting has been carried out in the western part initiated in Virginia, and Thomas Jefferson- of Virginia. Deseriptions of some of the collections possibly the most influential patron of vertebrate have been published, along with a few more paleontology during its formative years-was a extensive analyses of seleeted local faunas. The native son. However, a strong, enduring tradition Carnegie Museum of Natural History and the in the study of Quaternary vertebrates did not Smithsoniam Institution have been most respon- develop in Virginia. Amateur and professional sible for this increased activity. Some institutions in Virginia are becoming increasingly active, but broad-based institutional support within the rP.O. Box 4098, Newark, Ohio 43055 State for the developing professional and lay 90 VIRGINIA DIVISION OF MINERAL RESOURCES interest in Quaternary vertebrates is needed to the study of Quaternary vertebrates is provided. heighten awareness and appreciation of these Several recommendations are made that might resources. improve the conditions for the appreciation and study of Quaternary vertebrates in Virginia. INTRODUCTION HISTORICAL CONTEXT A growth of interest in the study of prehistoric Quaternary vertebrates of Virginia has oceurred Several important discoveries of vertebrate fos- since the middle of the 20th Century, and has sils were made in Virginia during the colonial ineluded significant work in both vertebrate pa- period. The earliest fossil vertebrate remains re- leontology and zooarcheology. The eollecting of covered by Europeans from Anglo-America were faunal materials from various depositional eon- reported in 1636 by Samuel Maverick as having texts has dominated this work within the State been found about 60 miles upstream of the mouth during the last three decades. These collections of the James River, in Virginia. These were re- have indicated clearly and convincingly that sub- portedly whale bones, probably of Miocene age stantial quantities of prehistoric Quaternary ver- (Simpson, L942, 1943; Eshelman and Grady, this tebrate remains are to be found within Virginia. volume). During his fourteen years in Virginia Comparatively few of the faunas that have been (1678-1692), John Banister eollected several ver- eollected sinee 1950 have been deseribed in the tebrate fossils from marine deposits in the Tide- literature, and very few have been the subject water region of the colony, and he obtained "some" of detailed paleoecological, zoogeographieal, or teeth from the interior of Virginia "found in the evolutionary analyses (Eshelman and Grady, this Hills beyond the Falls of the James River at least volume). a Hundred and fifty Miles up into the Country" This paper assesses the status of the study of (Ewan and Ewan, 1970). The latter were probably prehistoric Quaternary vertebrates of Virginia. delivered to Banister at or near "The Falls" (mod- Representatiae aelivity in both vertebrate paleon- ern Richmond) by Indians from the western Pied- tology and zooarcheology, as determined from a montor Appalachian Mountains, although the spec- search of the literature of the last eight years, imens could have been acquired by the Wood or examinations of collections, and personal inter- Batts expeditions to the Valley and Ridge in 1654 views, is the foundation for this assessment. The and 1671, respectively. There is no good reason to assessment is restrieted to work on vertebrate believe, however, that these teeth were of the faunas of the prehistoric period. It is not intended ground sloth Megalanar jeffersoruii as was sug- to be an exhaustive treatment of the history of gested by Goode (1901), since the remains of nu- vertebrate paleontology or zooarcheology within merous other large vertebrates also have been Virginia (see Eshelman and Grady, this volume, found in western Virginia (Ray and others, 1967; and Rose, this volume, for more detailed histor- Eshelman and Grady, this volume). By the 1740's, ical surveys). Rather, this represents an effort the occurrence of vertebrate fossils in Virginia was to deseribe current patterns and conditions of sufficiently well-known to allow Catesby to write work on the Quaternary vertebrates of Virginia. that in "all parts of Virginia, at the Distance of The body of this paper is divided into four Sixty Miles, or more (from the sea). . . are frequently sections. The eharacteristics of the present inter- found the VerfuAras, and other Bones of Sea Anim- est in Quaternary vertebrates are better under- als." (Catesby, t743, vol. 2, Appendix p. vii, quoted stood ifseen in historical context, so a briefreview in Simpson, L942, p. 134). At about this same time, of the place of studies of prehistoric Quaternary in 1739, Charles Le Moyne, the second Baron de vertebrates in Virginia is presented. Current Longueuil, colleeted some proboscidean remains activity patterns in vertebrate paleontology and near the south side of the Ohio River above modern zooarcheology are then treated separately be- Iouisville, Kentueky. Whether Longueuil or sub- cause these fields have had different histories, sequent French and British eollectors of fossils in even though work within each may converge to this area reeovered the first fossils from Big Bone provide complementary insights into faunal patt- Lick is equivocal, but within three deeades of erns and dynamics. Modern activity is influenced Iongueuil's visit, Big Bone Lick-then a part of by the amount and type of institutional support Virginia-had emerged as the most important and provided paleontologists and zooarcheologists, so well-known source of Quaternary vertebrate re- a review of the major facets of institutional sup- mains in eastern North America (Simpson, 1942, port and regulation within the State that affects 1943). Arthur Campbell sent Thomas Jefferson a PUBLICATION 75 91

mastodon tooth from the Saltville Valley in L782, ization of Virginia was not conducive to the de- thereby identifying that loeality, located in what velopment of a social environment in which the is today southwest Virginia, as a source of Qua- study of natural science could beeome well estab- ternary vertebrate fossils (Boyd, 1952; Jefferson, lished. The population of Virginia was dispersed. r787). Towns were few and usually provided only essential If the lengthy record of vertebrate fossil diseov- administrative and eeonomic services. In the ab- eries in Virginia was inadequate to establish a sence of populous, economically complex urban tradition of curiosity about and study of these centers with critical masses of interested persons, objects, the presence, genius, and energy of native broad-based interest in the natural sciences did not son Thomas Jefferson might have effected a focused develop (Ewan, 1976). Virginians with an interest interest in natural history in Virginia such as was in vertebrate paleontology either had to culture this occurring in some of the more northerly British interest on their own or join learned societies that colonies late in the 18th and early in the 19th had become established elsewhere, such as at Cha- centuries. Lacking the educational and fiscal re- rleston, South Carolina, or the principal port cities sources typieal of some European countries, the from Philadelphia to Boston (Ewan, 1976). Since Ameriean colonies/states late in the 18th Century these soeieties were organized to function by the could not compete effectivley with the Europeans direct participation of members rather than by in many academie fields. In natural history, how- publication of transmitted information, relatively ever, American seientists could make original con- few Virginians rffere able to be functional members. tributions and gain international respect in doing Jefferson, perhaps the most visible excaption, was so, and at the same time dispel notions-sueh as able to participate more fully in learned societies, some of those of Buffon, which particularly irri- especially those of Philadelphia, since his duties tated Jefferson-that American biota was inferior as Secretary of State, Vice President, and President to or otherwise less spectacular than thatof Europe of the United States (1?89-1809), and president of (Jefferson, 1787; Oleson, 1976). Jefferson made the American Philosophical Society (1797-1814), many contributions to the development of verte- kept him in or near that city much of the period brate paleontologa in Ameriea: he published the 1?89-1814 (Kiger, 1963; Greene, 1976; Bedini, first scientific paper describing (inaccurately) a 1985). fossil vertebrate from and in Ameriea (Jefferson, After Jefferson's return to Virginia in 1809, he 1?99), he proffered eritical response to Buffon's devoted much time and energa to the establishment notions about American vertebrates (including hu- of a comprehensive public education system in mans) (Jefferson, 1787), and he used the influence Virginia (Patton, 1906; Bedini, 1985). As a result of his high offices to support the collection, dis- of his (and others') efforts, the University of Vir- semination, and study of vertebrate fossils from ginia was established by the Virginia Legislature America-including the eollecting of some of the in 1819 and began instruetion in 1825. The general first fossils by Amerieans from what is now the public education system, however, was not estab- western United States (Osborn, 1929; Simpson, lished until 18?0. In the meantime, various other 1942; Bedini, 1985). Jefferson is also considered the private and institutional programs developed to person first responsible for conducting a system- provide elementary and more advanced education atic, problem-oriented archeological excavation to the general populace of Virginia. For the most and the first to publish results of such an excavation part, however, these were apprenticeships, military (Jefferson, 1787; Brose, 1973; Stoltman, 1973). Jef- schools, or church-sponsored academies or colleges ferson supported the develpment of learned socie- that were concerned either with providing prac- ties coneerned with advaneing natural philosphy, tical vocational training or academic training in exploration in search of new forms of natural subjects other than the natural sciences (Heatwole, objects and areas, and progressive state-supported 1916; Bell, 1930; Buek, L952). Natural philosophv educational systems and eurriculum development and related subjeets were represented for extended wherein, ultimately, most basic training in the periods only at the University of Virginia and the natural sciences in the United States came to be College of William and Mary. The creation of the offered (Simpson, 1942; Bedini, 1985). Virginia Geological Survey in 1835, with William Despite the role of Virginia in the early devel- Barton Rogers as its first director and its location opment of natural scienees in colonial and early at Charlottesville, waa a progressive act that as- national America, however, strong widespread in- sured that geological investigation would be car- terest in this field failed to materialize within the ried out (Roberts, L942). The Survey was concerned Colony and State. The social and economic organ- primarily with deseribing and mapping the geol- 92 VIRGINIA DIVISION OF MINERAL RESOURCES ogy of the state and in determining its economic funded programs but on a smaller scale than in resources, and vertebrate paleontolog:y did not be- some other eastern states and with emphasis on come an important area of investigation. The es- the historical archeology of Jamestown. One out- tablishment of the Virginia Agricultural and Me- growth of the increased visibility of professionally chanical School (modern VPI&SU)-the largest conducted archeology in Virginia during the De- educational institution to be developed during the pression was the creation of the Archeological second half of the 19th Century in Virginia-was Society of Virginia in 1940. Two years later, in 1942, made possible by the federally funded land-grant this Society commenced publication of the Quar- college program and was intended primarily for terly Bullet'in,which, after a tenuous start, has been practical training in agricultural and mechanical the principal journal dealing with archeology in skills (Buck, 1952; Dabney, 1971). Instruction in Virginia. TheQuarterly Bulletinwas joined in 1963 vertebrate paleontology and archeology, therefore, by The Chesopiean publication of the Chesopiean traditionally has not been a part of general edu- Archaeological Association. cation curricula in Virginia. Since the 1960's, archeology in Virginia has The practice of vertebrate paleontology in North become increasingly profession alized. Several eol- America shifted near mid-l9th Century from leges and universities now offer undergraduate learned societies to the newly created Smithsonian training in archeology, although opportunities for Institution (1346) in Washington, D. C. and several graduate training in prehistoric archeology remain universities and other institutions in the northeast- limited. The Council of Virginia Archaeologists ern states (Greene, 1976; Voss, 1976). At this time, was organized in 1975 to further the aims of the exploration of western territories was focusing professional archeology in the State. Another im- the attention of vertebrate paleontologists on the portant development was the creation of the Vir- west (Lanham, 1973). Virginia-along with most ginia Historic Landmarks Commission (VHLC) in of the eastern United States-reeeived relatively 1966. This Commission was responsible for survey- little attention from vertebrate paleontologists during, registering, and encouraging the proteetion of ing the late 19th Century (Eshelman and Grady, Virginia's archeological sites and, through the Vir- this volume). ginia Commissioner of Archaeologly, was charged More attention has been paid to Virginia's fossil with enforcing the provisions of the subsequently localities during the 20th Century, espeeially by passed Virginia Antiquities Act (1977), the prin- the Carnegie Museum of Natural History (Pitts- cipal body of legislation dealing with the prehis- burgh, Pa.), as part of its systematic investigation toric human record in Virginia. The VHLC became of the Quaternary vertebrate paleontology of the the Virginia Division of Historic Landmarks middle and southern Appalachians (e.g., Peterson, (VDHL) in 1985; the Director of the Division is 1917; Guilday,1962; Guilday and others, 19?7). The now charged with enforeing the provisions of the Smithsonian Institution has also shown heightened VirginiaAntiquities Act and the more recent(1979) interest in Virginia's Quaternary vertebrates dur- Cave Protection Act. ing the 20th Century (Ray and others, 1967; Ray and others, 1968). The activity of both institutions QUATERNARY VERTEBRATE increased during the 1960's and in doing so called PALEONTOLOGY IN VIRGINIA attention to the fact that there are many productive paleontologieal sites of Quaternary age in Virginia Recent work on Quaternary vertebrates in Vir- worthy of professional scientific attention. As a ginia has been characterized by the recovery of result, scientists in Virginia have beeome increas- specimens on an unprecedented scale. The Carnegie ingly active in recent years in collecting and des- Museum of Natural History is clearly the leader cribing Quaternary vertebrate remains. in this activity, having collected at numerous lo- Archeology was developing as a distinct disci- cations throughout western Virginia. Several thou- pline during the 19th Century also, but in this field, sand specimens from in and near the Saltville too, Virginia was the scene of much less activity Valley of southwest Virginia have been collected than some nearby states. Virginia sites received by and for the Smithsonian Institution, while other only peripheral attention during the era of Mound smaller collections have been made by workers Builder investigations and, later, during the south- from other institutions. Several local faunas have eastern river basin surveys (Carr, 1877; Fowke, been described. Excavations are underway at at 1894; Brose, 1973; Stoltman, 1973). Professionally least three sites within the State. Overall, profes- directed archeological investigations took place in sional work with Quaternary vertebrates is increas- Virginia during the Depression under federally ing, a eommunication and field support network PUBLICATION 75 93

is developing, and public awareness and interest ferences about choroclinal variation in character is growing. Note is made, however, that many expression. Temporal changes in the composition workers are located at out-of-state institutions, that or intraspecific morphology of stratified faunas there is no functional repository within the State permit inferences about faunal turnover and chro- where specimens are being curated and studied, noclinal variation. The correlation of patterns of and that there is no worker employed by a Virginia composition and morphology among several strat- institution whose professional responsibilities lie ified sites could permit unusual opportunities to primarily with the collection, curation, or study of document and interpret faunal ehanges on a re- Quaternary vertebrates. Also, there is remarkably gional scale during the environmentally dynamic little information available to the public to inform late Quaternary. Major problems facing the anal- them about the animal life of the Quaternary or ysis of these loeal faunas are the lack of manpower to satisfy or stimulate their curiosity about this to study them in a timely fashion, and the lack of subject area. John Guilday, the central figure in precise temporal and compositional control so typ- the Quaternary vertebrate paleontology of the Ap- ical of faunas from karst features which limits the palaehians, died in November 1982, and with him extent to which relatively precise time/space eco- passed much of the momentum, experience, and logic and zoogeographic patterns can be docu- knowledge for continued work in this region. mented. The Carnegie Museum of Natural History has The Saltville Valley,located in Smyth and Wash- a lengthy tradition of interest in the study of ington counties, is a large subaerial site containing Quaternary vertebrates of the middle and southern a number of different fossil-bearing depositional Appalachian Mountains. One of the most ambitious units. This loeality has been known as a source of and far-sighted programs for the study of Appa- remainsof late Quaternarymammals for more than lachian Quaternary vertebrates is the systematic 200 years (McDonald, 1984a). The first record of sampling of local faunas throughout the region by a fossil from this area is contained in a letter from employees and friends of the Carnegie Museum. Arthur Campbell to Thomas Jefferson, dated No- This program was initiated by John Guilday in the vember, 1782 (Boyd,1952), probably a response to 1950's. Field surveying and collecting was carried Jefferson's appeal for information aboutthe natural out by Carnegie staff members-most notably Ha- curiosities of Virginia. Olaf Peterson of the Car- rold Hamilton with the assistanee of his assoeiates, negie Museum obtained a small collection of fossils especially Robert E. Whittemore (Johnson City, at Saltville in 1917. The first scientifically moti- Tn.) and Lois Miller (Pittsburgh, Pa.). This pro- vated exeavation took place in 1966-67 as a joint gram has resulted in the sampling of hundreds of venture between Virginia Polytechnic Institute and local faunas, almost exclusively from karst and the Smithsonian Institution with the paleontolog- karst-like features, from Georgia and Alabama to ical aspect of this work being directed by Clayton Pennsylvania and Ohio. Forty-eight local faunas E. Ray (Smithsonian Institution). Salvage excava- from Virginia have been collected by the Carnegie tions in 19?8 and 1980 by Charles Bartlett, Jr., led Museum, mostly under this program of systematic to the involvement of this author in excavations sampling, and therefore mostly from the western at Saltville in the fall of 1980. In recognition of counties (Table 1, Figure 1). This program is the Saltville locality's potential to yield abundant continuing, and additional sites in Virginia have information on diverse aspects of the late Quater- been identified for sampling (H. Hamilton, per- nary history of the middle Appalachian region, the sonal communication). Saltville Project was initiated in 1982. This project The Carnegie sampling program has produced is a multidisciplinary investigation of the Quater- what is unquestionably the most important collec- nary history of the locality that currently involves tion of late Quaternary faunas from eastern North sixteen specialists from eight institutions in three America with which can be addressed important countries. systematic, ecological and zoogeographical ques- The Saltville locality, defined as the Saltville tions (see, e.g., Guilday, 1962; Guilday and others, Valley and its environs within a radius of about 1977;Fay,1984a, 1984b, this volume; Holman, this 6 miles, contains a large variety of depositional volume). Spatial variations in faunal composition environments, including those of caves, fissures, for any given period will allow inferences about sinkholes, springp (one with an associated tufa latitudinal gradients in species distribution and deposit), streams, a lake-marsh-bog system, and community composition and structure throughout colluvium. Several different features, representing the middle and southern Appalachians. Similar the karst, tufa, colluvium, lotic, and lentic deposits, variations in intraspecific morphology allows in- are known to be fossiliferous. The most productive 94 VIRGINIA DIVISION OF MINERAL RESOURCES

Figure 1,. Unpublished Quaternary vertebrate faunas in Virginia. The eounty locations of faunas listed in Table 1 are identified here by number. deposits discovered to date are the fluvial sediments is significant, therefore, as a proven source of associated with the extinct Saltville River (ca. vertebrate fossils and contemporaneous biophysical 27,000-13;500 yrs B.P.) and the lacustrine deposits information spanning much of the environmentally associated with extinct Lake Totten (ca. 13,500-140 dynamic late Wiseonsin-Holocene transition (see, yrs B. P.). These well-stratified deposits occur over e.g., McDonald and Bartlett, 1983; McDonald, mueh of the bottom of Saltville Valley. The ground- 1984a, 1984c, 1985, in press; Barfield, this volume; water that occurs in the valley bottom is saline, Delcourt and Delcourt, this volume; White, this and that condition, along with the fact that the Lake volume; Holman and McDonald, in press). Totten deposits are massive clays characterized by Other sites in Virginia producing Quaternary low permeability, has resulted in an extremely good vertebrate fossils and currently being studied in- environment of preservation for the entombed fos- elude the Hot Run site in Frederick County and sils. The variety of depositional environments at an unnamed site in Warren County. The Hot Run this locality provides opportunity to recover infor- Site, being worked by James Ott (University of mation collected by various depositional processes, Maryland), appears to be a shallow subaerial de- each with a different sampling bias. This variety posit that includes remains of several extinct taxa of sampling, in turn, should permit a more complete of large mammals, along with those of domesticated and balanced assessment of the late Quaternary forms (Ott and Weems, this volume). Samuel Bird history of the region than would be possible with (Virginia Division of Energy) and Walter Wheeler data from only a single site or type of depositional (University of North Carolina) have recently eol- environment. The lentic and lotic sediments found lected and are now studying remains of small in the valley bottom appear to represent relatively mammals, birds, and fishes from a cave deposit continuous deposition for approximately the last in Warren County (S. O. Bird, pers. comm.). 15,000 years, and more intermittent deposition over Information has been published on faunas from the preceding 12,000 years. These sediments, there- at least six Virginia sites during the last eight fore, potentially offer a continuous biotic record for years. Weems and Higgins (1977) described a fauna the last 15,000 years and a periodic one for the containing at least 33 vertebrate species found in last 27,000 years. Subaerial stratified sites contain- a fissure near Ripplemead in Giles County. Gary ing proven fossil resources are rare in the Appal- Haynes (1978) reported thirteen genera/species, achians, as are sites that span as much time as mostly mammalian, recovered from a small fissure do the Saltville Valley deposits. The Saltville Valley in Warren County. Most of the remains came from PUBLICATION 75 95

Table 1. Unpublished Quaternary vertebrate faunas from Virginiat

Locality CountalCi.tu Est'i,mated Age Repositoryz

1. Metompkin Island Accomack Co. Pleistocene USNM

2. Glade Cave Augusta Co. ?Holocene USNM

3. Natural Chimneys Augusta Co. Late Pleistocene UT

4. Unnamed Locality Augusta Co. (not stated) ANSP

5. Back Creek Cave #1 Bath Co. (not stated) CMNH

6. Back Creek Cave #2 Bath Co. (not stated) CMNH

7. Back Creek Cave #3 Bath Co. Holocene CMNH

8. Clark's Cave Bath Co. Late Pleistocene UT

9. Rope Burn Cave Bath Co. (not stated) CMNH

10. Bane's Cave Bland Co. Late Holocene CMNH

11. Paul Penley's Cave Bland Co. (not stated) CMNH

12. Arcadia #1 Botetourt Co. Pleistocene/Holocene CMNH

13. Arcadia #2 Botetourt Co. Pleistocene/Holocene CMNH

14. Arcadia Little Cave Botetourt Co. ?Holocene CMNH

15. Springwood Bluff Botetourt Co. (not stated) CMNH

16. Deep Creek Chesapeake (City) Pleistocene USNM

17. Unnamed Locality Chesapeake (City) Late Pleistocene USNM

18. Calmes Neck Cave Clarke Co. Holocene CMNH

19. Unnamed Cave Clarke Co. (not stated) USNM

20. Hot Run Irederick Co. Quaternary UM

21. Shires Saltpeter Craig Co. Late Pleistocene/ Cave Holoeene CMNH

22.Cliff Cave#2 Giles Co. Late Pleistocene/ CMNH Holocene

23. Eggleston Cliff Cave Giles Co. Hqlocene CMNH

24. Eggleston Fissure Giles Co. Late Pleistocene CMNH

Zi.Klotz Quarry Cave Giles Co. Late Pleistocene/ Holocene CMNH

26. Pembroke Railroad Giles Co. Pleistocene CMNH Cave #1

27. Pembroke Railroad Giles Co. ?Pleistocene/Holocene CMNH Cave#2

28. Ripplemead Quarry Giles Co. Late Pleistocene CMNH

29. Rass Hole Highland Co. Late Pleistocene/ CMNH (= Corbertt's Cave) Holocene

30. Saltpeter Highland Co. Holocene CMNH (Arbegast) #5 96 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 1. (cont.)

Locality CountalCitA Esti,mated Age RepositorEz

31. Strait Canyon Highland Co. Pleistocene CMNH/ USNM

32. Varner's Cave Highland Co. Holocene CMNH/ USNM

33. Belvedere Beach King George Co. Late Pleistocene/ NJSM Holocene

34. Carter Cave Lee Co. (not stated) CMNH

35. Jasper Saltneter Cave Lee Co. Late Pleistocene/ Holocene CMNH

36. Skylight Cave Lee Co. Holocene CMNH

37. Wallen Creek Cave Lee Co. Holocene CMNH

38. Spurey Madison Co. Modern USNM

39. Lanexa New Kent Co. Pleistocene CMNH

40. Comers Cave Page Co. ?Modern CMNH

41. Ruffner's #2 Page Co. Holocene USNM

42. Collierstown Fissure Rockbridge Co. Holocene CMNH

43. Fishbone Cave West Rockbridge Co. (not stated) CMNH

44. Bill Neff Cave Rockingham Co. Quaternary USNM

45. Cedar Hill Cave Rockingham Co. Quaternary USNM

46. Monger Cave Rockingham Co. Holocene CMNH

47. Round Hill Cave Rockingham Co. Pleistocene CMNH

48. Wheelbargers Cave Rockingham Co. Holocene CMNH 49. Loop Creek Quarry Cave Russell Co. Late Pleistocene/ Early Holocene CMNH

50. River Bend Cave Russell Co. ?Holocene CMNH

51. Darty Cave Scott Co. Early Holocene CMNH

52. Roadcut Cave Scott Co. Pleistocene/ Holocene CMNH

53. Taylor Cave Scott Co. (not stated) CMNH

54. Winding Stairs Cave Scott Co. ?Late Pleistocene CMNH

55. Edinburg Fissure Shenandoah Co. Quaternary USNM

56. Shenandoah Cavern Shenandoah Co. Holocene USNM

57. Toms Brook Shenandoah Co. Holocene USNM 58. New Quarry Cave Smyth Co. Quaternary USNM 59. Saltville Smyth Co. Quaternary CMNH/ USNM

60. Unnamed Cave Smyth Co. Quaternary USNM

61. Eclipse Beach Suffolk (City) ?Pleistocene USNM

62. Crab Orchard Cave Tazewell Co. Modern VRCA

63. Gillespie's Cliff Cave Tazewell Co. (not stated) CMNH PUBLICATION 75 97

Table 1. (cont.)

Localitg Cou,ntalCitu Estimated Age Repository2

64. Virginia Beach Virginia Beach (City) Pleistocene USNM

65. Skyline Caverns Warren Co. Late Pleistocene USNM

66. Unnamed Cave Warren Co. Quaternary VDMR

67. Holston Vista Cave Washington Co. Late Pleistocene/ Holoeene CMNH

68. Meadowview Cave Washington Co. Pleistocene/Holocene CMNH

69. Vicker's Cave Washington Co. (not stated) CMNH

70. Will Fraley's Cave Washington Co. Late Pleistocene CMNH

71. Stratford Hall Westmoreland Co. ?Pleistocene USNM

72. Parson's Cave Wise Co. Holocene CMNH

73. Gardner's Cave Wythe Co. (not stated) CMNH 74. Unnamed Subaerial Site Yorktown (City)/York Co. Late Pleistocene CWM I Most of these faunas are not described in the literature. Some, such as Clark's Cave (Guilday and others, 1977;Fay,1984a), Natural Chimneys (Guilday, 1962;Fay,1984a), Ripplemead Quarry (Weems and Higgins, 1977), Saltville (Peterson, 1917; Ray and others, 1967; McDonald and Bartlett, 1983; Holman and McDonald, in press) and Strait Canyon (Fay, 1984b) have been published in part, but substantial quantities of undescribed specimens are known for all but Clark's Cave and Natural Chimneys. 2 The repositories of unpublished collections are listed in this column, as follows: ANSP = Academy of Natural Sciences of Philadelphia; CMNH = Carnegie Museum of Natural History; CWM = College of William and Mary;NJSM= NewJersey State Museum; UM = Universityof Maryland (James Ottcollection); UT = MeClung Muesum, University of Tennessee; USNM = United States National Museum of Natural History; VDMR = Virginia Division of Mineral Resources; VRCA = Virginia Research Center for Archaeology. the upper 2-3 inches of the excavated sediments. neys, Clark's Cave, and Saltville in Virginia. Their All identified genera and species are extant. The findings are that herptiles, unlike plants and higher presenee of porcupine (Eretluizoz sp.) and red fox vertebrates, appear to have been relatively unaf- (Vulpes sp.) within the fauna is noteworthy for fected by the environmental dynamics that char- zoogeographic reasons, because these taxa might acterized the region during the late Wiseonsin- not have been present in Virginia at the time of Holocene. Although unpublished, much work on the European contact (Pagels, 1979; Rose, this volume). Strait Canyon fauna had been completed by John McDonald and Bartlett (1983) described and Guilday at the time of his death. Hopefully, someone figured the partial associated skeleton of an extinct will soon finish and publish his work on this im- muskox (cf . Sambos sp.) from Saltville, and Holman portant fauna. and MeDonald (in press) have described a herpetofauna eontaining ten taxa from the same locality. The latter paper, which described samples eol- ZOOARCHEOLOGY IN VIRGINIA lected from radiometrically dated stratigraphic units, concluded that the modern herpetofauna of Zooarcheology is a recently-developed subdisci- the Saltville area has probably been in place for pline of archeology which is still undergoing rapid at least 13,500-15,000 years. Fay (1984a, b; this transformation and increased specialization, par- volume) and Holman (this volume) have analyzed ticularly as it relates to regional and cultural-level several late Wiseonsin-Holocene herpetofaunas research design, data analysis and theory building from the middle and southern Appalachians, in- in archeology (Hesse and Wapnish, 1985). Robison cluding those from Strait Canyon, Natural Chim- (1978) reviewed the development of zooarcheology 98 VIRGINIA DIVISION OF MINERAL RESOURCES

in North America, tracing the development of the archeological sites in Virginia are perceived and subdiscipline from an initial eoncern with merely handled. Professional archeologists and many ama- identifying some or all species (especially mammals teur archeologists routinely exeavate faunal re- and easily recognized birds and reptiles) recovered mains as parts of the complete site record, whereas from a site, through increased quantification ofthe at the other extreme some excavators save few if identified remains and application of these data any faunal remains. to specific ecological questions (such as butchering Most published reports on prehistorie archeology techniques and seasonal hunting patterns), to a done in Virginia by professional or amateur ar- modern orientation that integrates zooarcheolog- cheologists appear in the Quarterla Bulletin of the ieal data into the holistic analysis of local and Archeological Society of Virginia. Most of the site regional cultures and their dynamics (sueh as an- reports published in this journal during the last alyses of intersite functions, or class/ethnic com- seven years have included mention of faunal re- position of the population). For all its potential and mains. Many sites, especially open air sites older dynamics, however, zooarcheologa is based upon than middle Woodland age (1,500 yr B.P.), rarely the reeovery and study of faunal remains from yield faunal remains; most such materials, if ever areheological eontexts. Faunal remains must be they were present at a site, have deeomposed in regarded belore eucaaat'i,on as integral parts of the Virginia's humid mesothermic environment. For archeological record, recognized as artifacts dur- cave, rockshelter, and middle or late Woodland ing survey and excavation activities, excavated sites, however, faunal remains often are not col- with the same full provenance data as other ar- lected by excavators or they are not described in tifacts, properly stored, and made aecessible to site reports, possibly because they are not consid- researchers, if there is to be zooarcheological an- ered important, they have not been identified, or alyses or insights at any level of complexity or they are assigned lower priority than other arti- sophistication. There presently exists substantial facts. Site reports that include mention of faunal variation in the way that faunal remains from remains (Figure 2) typically include a short list

Figure 2. Recently published paleontological and zooarcheological localities in Virginia. The localities are:1-Flanary(44Sc13);2-White( Wg37);3-Arrington(44Wg27);4-Sullins(44Wg12);5-Saltville; 6-CrabOrchard(44Tzl);7-RipplemeadFissure;8-Hall(a M$3);9-Hercules(44Ay40);10-Gathright Dam area; 11 - Hidden Valley area; L2 - Clark's Cave; L3 - Strait Canyon; 14 - Natural Chimneys; 15 - Unnamed cave (44Wr61); 16 - Hot Run; 17 - Wade (44Ha34); 18 - Reedy Creek (44Ha22);19 - Hertzler (44PoB); 20 - College Creek.(44Jc27). PUBLICATION 75 99

of species-usually under "subsistence"-and iden- that ony zooarcheological data, if accurately re- tification of tools or other artifacts made from corded and reported, might be useful (Barber, faunal elements. Most faunal lists are biased to- 1981). ward mammals (Table 2) and toward speeies with A small number of analytical and theoretical easily identifiable characters at that. Among recent studies dealing with the aboriginal usq of verte- publications, only in Barber's analysis of the Col- brates in Virginia have been published reeently. lege Creek Site (44Je27), whieh was intended as E. Randolph Turner (Virginia Research Center for a methodological paper, and in Coleman's descrip- Archaeology) described the ecological function of tion of the Reedy Creek Site (44Ha22), do mammals an intertribal buffer zone situated astride the Fall constitute less than one half the listed taxa (Barber, Line during the early part of the 17th Century. l978a,Coleman, 1982). Faunal remains were quan- He presented evidence suggesting that this con- tified in about one third of the surveyed reports tested zone probably served as arn important source containing faunal data (Table 2). Only one surveyed of game (espeeially deer), and further theorized report (Egloff and Reed, 1980) dealt with the thai overexploitation of the deer population in the functional analysis of faunal remains. Stratigra- Tidewater region had possibly oecurred as early phic context was not provided for most reported as the middle of the late Archaic Period (Turner, faunal remains, due in great part-but not exclu- 1978). Holland (University of Virginia) drew upon sively-to the fact that specimens from most sites ethnohistoric sourees to document and analyzethe were collected on or near the surfaee, usually from protohistoric fire hunt, from which analysis-using the plow zone. Buchanan's faunal data for the Hall distribution patterns of pottery from archeological Site (aaMy33) contains the most detailed stratigra- sites-he postulated the existence of other prehis- phic data among the sites listed in Table 2 (Buch- torie buffer zones in Virginia (Holland, 1979). anan, 1980). The major reason that more zooarcheo- McDonald (1984b) studied 97 partial mandibles of logical work is not reported upon probably is whitetail deer (Odocoileus ui'rgini'anus) from the because of the fact that few amateur or professional l?th Century (ca. A.D. 1600-1635) Trigg Site areheologists working on Virginia's prehistory are (a Mv3) (Buchanan, 1984), from whieh were ob- sufficiently well trained in zooarcheology to iden- tained data on the sex and age of the harvested tify and analyze faunal remains. population, the presence of dental abnormalities, Michael Barber (Thomas Jefferson National For- and patterns of ieasonal hunting. The Trigg Site est and Radford University), the only zooarcheol- deer population was relatively old, females out- ogist in Virginia that works regularly with pre- numbered males, and the deer were hunted historic faunal collections, has addressed several throughout the year. The implications of this pat- methodological problems before the Virginia ar- tern are that the deer population was relatively cheological community. In a 1978 paper he pres- large, and hunters could be relatively selective with ented rationale and procedures for developing a a high probability of hunting success at any season. eomparative colleetion, pointing out that the first Henry Miller (St. Marys City, Md.) studied faunal step in faunal analysis was the accurate identifi- assemblages from fifteen early historic sites along cation of the specimens recovered, and that a type the lower James and Potomac rivers in Virginia collection of "core species" was a necessity for and Maryland in order to test hypotheses predicting identifying animal remains (Barber, 1978b). In how human subsistence patterns change in frontier another 1978 paper, Barber attempted tc demon- settings (Miller, f984). He found that diet patterns strate temporal differences in resource exploitation changed sigpificantly during the 17th Century by comparing the faunal assemblages from three from a seasonally changing, generalist reliance different excavation units of middle to late Wood- upon diverse wild and domestic animals to a more land, late Woodland, and historic ages (Barber, specialized, simplified pattern. Although Miller's 1978a). Later, working with three historic site study does not deal strictly with prehistoric zooa? assemblages, Barber challenged the notion that cheology, it documents the presence of certain small sample size could not provide meaningful species in Virginia at or soon after European insight into the ecological dynamics of sites. His contact; it develops a model useful to the study of conclusion that small faunal assemblages could subsistence change among aboriginal populations provide useful information about cultural lifeways, entering a frontier or a new "exploitation zone"; such as yielding a list of the minimum number of and it represents a modern, sophisticated approaeh species used by the site occupants and revealing to the study of faunal remains from archeological subtle details of their culture, was also a declaration sites wherein model building, theory testing, and 100 VIRGINIA DIVISION OF MINERAL RESOURCES

Table 2. Vertebrate remains from prehistoric archeological sites reported in the ASY Quarterly Bulletin, 1978-1984

Site Nomc onil Number Source Subsbtennet Toolse Othnr| Aget M BR/AF 1. Hercules 44Ay40 Johnson, 1979 61r10 0 x number of bones lw 2. Hidden Valley Geier, 1981 3 00/0 0 X lw Rockshelter 44Ba31

3. Gathright Dam-Lake Geier, 1983 7 rL/0 0 lw Moomaw-Hidden Valley area, Bath Co. 4. Reedy Creek44Ha22 Coleman, 1982 8 24124 number of bones by lw level and taxon 5. Wade Site 44Ha34 Stevens, 1979 51r/02 (Red Fox) Indet. (mAlW) 6. College Creek44Jc27 Barber, 1978a 8 15/04 Species abundance, m\{-H MNI;element, side age and sex of deer remains

7. Hall Site 44My33 Buchanan,1980 7 | 3/0 1 relatively good tw stratigraphic record; wet screened 8. Hertzler Site 44PaB MacCord and 5 ou0 0 lw Livesay,1982 9. Flanary Site 44Sc13 MacCord, 1979 2 tLlO 0 x lw 10. Crab Orchard Site Egloff and 51L/0 0 X limited analysis lw 44Tzl Reed, 1980 of butchering 11. Sullins Site 441{g12 MacCord, 1981 7 rr/0 0 number of elements lw and percent of total; number of specimens by species and feature 12. Arrington Site 44Wg27 Bartlett, 1980 31210 0 x e-mW 13. White Site 44Wg37 Bartlett, 1979 100/0 0 x mA-l\{ tM=Mammals;B=Birds; R/A=Reptiles/Amphibians;F=Fish.Thenumberineachcolumnindicates the number of taxa reported for each class/group considered by the authors of the report to have been used as a food resouree by the aboriginal population.

2 An X in this eolumn indicates that tools (or other artifacts) made from faunal elements were identified as sueh in the report.

3 Noteworthy aspects of the collecting, analysis, or reporting procedure, or the content of the fauna. aAgeofthesite:A=Archaici W=Woodland;H=Historic;e=Early;m=Middle;l=Late. PUBLICATION 75 101 the characteristics and dynamics of regional cul- zooarcheology for students in Virginia institutions tures are addressed. is limited, and takes the form of seminars, field courses, special topics courses, and independent INSTITUTIONAL SUPPORT OF STUDIES study. A similar deficiency in institutional interest in Ownership of prehistoric faunal remains in Vir- prehistoric Quaternary vertebrates exibts among ginia lies with the landowner-private citizens or museums in Virginia. Virginia has many museums corporations, or loeal, state, or federal govern- (Virginia Association of Museums, 1984) but, until ments. The intent of both the state and federal 1985, the State was without a museum of natural governments is to protect their faunal resources history. The three largest public museums of sci- in order to assure maximum public benefit from ence in the State are young institutions: The Pen- the removal (if necessary) and study of these mate- insular Science and Nature Center (1966) in Hamp- rials. In order to eollect or excavate specimens from ton, the Science Museum of Virginia (1969) in federal land, permits must be obtained from the Richmond, and the Roanoke Valley Science Mu- federal agency with local jurisdiction. Archeolog- seum (1970) in Roanoke. None of these science ical collecting and excavating on state lands, and museums offers topical exhibits in Quaternary any collecting or excavating in caves (except by vertebrate paleontology or zooarcheology, nor does the owner), must be permitted by the Virginia any collect, store, or support researeh in these areas. Division of Historic Landmarks. In addition, the In fact, these museums normally do not accept state agency with local jurisdiction may require donations of fossils or zooarcheological specimens; a separate permit, and permission of the landowner typically such offerings will be referred to or passed is also required for removieg materials from caves on to another museum-such as the Smithsonian owned by local governments or private citizens. The Institution-or, if accepted, are placed "in the federal and state governments retain ownership of attic." There is no good prospect for this condition materials excavated under federal or state permit. to change in the near future, given the fiscal Ownership of materials originating on private, constraints and administrative orientation of each county, or municipal land lies with the landowner museum. Yet, it must be realized that the regional but may be transferred at the diseretion of the and national museums, such as the Carnegie and landowner. Sinee most land in Virginia is owned National museums of natural history, cannot con- by local governments or privately, there is little tinue to accept specimens and collections that are legal control by the state or federal governments better suited to statelevel museums (see, e.g., over the exploitation or destruction of subaerial Adams, 1985). areheological or paleontological sites. As a result, The most promise for institutionally supported much information of scientific value is lost regu- progress in the study of prehistoric faunal remains larly, especially to construction activities, unscien- in Virginia lies with the new Virginia Museum of tific eollecting procedures, and commereial traf- Natural History. This venture, directed by Noel ficking in specimens, because of inadequate Boaz and located in the City of Martirisville, opened concern for or awareness ofthe significance ofthese to the public on June 2, L985. The conceptual resources on the part of the landowners or collec- strengths of this institution lie in its being directed tors. by an active research scientist and its orientation Responsibility for the professional level collect- toward scientific collecting and research from its ing and study of faunal remains in Anglo-America inception. This museum offers Virginia its first has traditionally rested with seholars affiliated museum of natural history and a complete museum with colleges and universities or museums of nat- in that it will promote collecting, research and ural history. Unfortunately, there is no vertebrate public education in the natural sciences. paleontologist or zooarcheologist position on the The Virginia Research Center for Archaeology faculty of any Virginia institution of higher edu- (VRCA) was ereated as a unit of the Virginia cation, nor is any regularly listed course offered Historic Landmarks Commission in 1966, but did in either of these areas at the college or university not become functional until 1976 (Outlaw, 1982). level. There are a few faculty members that do The VHLC, in conjunction with VRCA, adminis- researeh in the vertebrate paleontology of Virginia, tered the provisions of the Virginia Antiquities Act but this is not their principal academic area of and, later, the Virginia Cave Protection Act (Code responsibility. No full-time faculty member is of Virginia, Title 10, chapters 12.1 and 12.2, re- known to be conducting research on zooarcheology. spectively), the principal bodies of state legislation Academic training in vertebrate paleontology and dealing with the archeological and paleobiological 102 VIRGINIA DIVISION OF MINERAL RESOURCES resources of Virginia. The VDHL, through VRCA, formation available for more detailed or general- now enforces the provisions of these acts. The ized studies. Proportionately few analytieal, me- VRCA holds what is probably the largest publiely- thodological, or theoretical papers, based upon owned comparative collection of faunal specimens studies of prehistoric local faunas of Virginia, have for use by zooarcheologists in the State, but even been published-but these types of studies are so it is of limited value in many respects, being increasing and are to be encouraged. The colleetion strongest on small mammals but weak on other of additional materials is continuing, and-increas- groups and its collections are generally inadequate ingly-information on the associated paleoenviron- for documentation of individual variation, age dif- ments is also being collected. ferences, and sexual dimorphism. VRCA has been Expanded institutional support appears to be faced with substantial fiscal reductions since 1981 necessary if substantial progress is to be realized and, among other adjustments, recently has had in the study, appreciation, and protection of Vir- to limit the development and use of this collection. ginia's prehistorie Quaternary vertebrate resour- The Department of Anthropology, College of Wil- ees. liam and Mary, also has an inactive comparative collection of faunal specimens. The study of zooar- ACKNOWLEDGMENTS cheological material in Virginia, therefore, requires the worker to have a personal collection, to I thank Dennis R. Baker, Michael Barber, Joanne rely upon collections of institutions outside Vir- Bowen, Ralph Eshelman, Alice Guilday, Harold ginia, or to rely upon published identification Hamilton, Mary Ellen Hodges, Thomas Krakauer, guides or personal knowledge of osteological or Howard A. MacCord, Sr., Henry Miller, David dental charaeters. Parris, Paul Parmalee, Margaret T. Peters, Robert Sullivan, William Walton, and Irvine Wilson for SUMMARY providing information that contributed to the final version of this paper. Michael Barber, Samuel O. Many significant discoveries of prehistoric Qua- Bird, Eugene Barfield, Harold Hamilton, James ternary vertebrate remains were made in Virginia Johnston, E. Randolph Turner, G. Richard Whit- during the colonial and early national periods. A teear, and Susan Woodward critically reviewed an strong interest in the natural sciences did not earlier draft of this paper and made many helpful develop in Virginia, however, and as a result,little comments. The opinions and emphasesthat remain, attention was paid to the prehistoric vertebrate however, are my responsibility. record of the state during the 19th and first half of the 20th centuries. The study of prehistoric REFERENCES Quaternary vertebrates from Virginia by paleobi- ologists and zooarcheologists began to intensify Adams, R., 1985, Smithsonian horizons: Smithson- around the middle of the 20th Century. Paleobi- ian, vol. 16, no. 1, p. 14. ologists at the Carnegie Museum of Natural History Barber, M. 8., L978a, The vertebrate faunal anal- and the National Museum of Natural History were ysis of Jc27 (James City County, Virginia): an most responsible for this increased activity, al- exercise in plow zone archeology: Archeol. Soc. though archeologists with the College of William Virginia, Quart. Bull., vol. 32, p. 94-100. and Mary, the University of Virginia, and the 1978b, On the preparation of a faunal Archeological Society of Virginia also made con- type collection: Archeol. Soc. Virginia, Quart. tributions to the development of foundations for Bull., vol. 32, p. 49-55. subsequent zooarcheological studies within the 1981, Insights into site dynamics as State. More recently, support for the study of - reflected in small samples of animal bones: the prehistorie Quaternary vertebrates has come from faunal analysis of three Virginia historic as- additional edueational institutions and government semblages: Archeol. Soc. Virginia, Quart. Bull., agencies loeated within and near Virginia. Most - vol.35, p. 159-165. work since 1950 has consisted of collecting spee- Bartlett, C. S., Jr., 1979, The White Site, Wash- imens from a large number of localities, partic- ington County, Virginia: Archeol. Soc. Vir- ularly in western Virginia. Some of these collec- ginia, Quart. Bull., vol.33, p.87-97. tions have been described, with considerable 1980, The Arrington Site (44Wg27): an variation in rigor, but most are unstudied. Many early to middle Woodland campsite at Abing- more of these collections need to be described in don, Virginia: Archeol. Soc. Virginia, Quart. order to make, in the least, basie descriptive in- Bull., vol. 35, p.35-48. - PUBLICATION 75 103

Bedini, S. A., 1985, Thomas Jefferson and Amer- Carnegie Mus. Nat. Hist., no.8, p. 14-19. ican vertebrate paleontology: Virginia Div. Fowke, G., 1894, Archaeologic investigations in Min. Res. Pub.61,26 p. James and Potomac valleys: U. S. Bur. Amer. Bell, S., 1930, The Church, the State, and education Ethnol., Bull.23, 80 p. in Virginia. Philadelphia, The Scienee Press, Geier, C. R., 1981, A late Woodland burial from 796p. the Hidden Valley Rockshelter, Bath County: Boyd, J. P., ed., 1952, The papers of Thomas Archeol. Soe. Virginia, Quart. Bull., vol. 35, p. Jefferson: Prineeton, Princeton Univ. Press, 113-138. vol. 6. 1983, The continuum of late Woodland/ Brose, D. S., 1973, The northeastern United States, Protohistoric settlement along a section of the irz Fitting, J. E., ed., The development of North Jackson River in the Ridge and Valley Province American archaeology: essays in the history of of Virginia: an initial formulation: Archeol. regional traditions: Garden City, Anchor Press, - Soc., Virginia, Quart. Bull., vol. 38,p.234-276. p.84-115. Goode, G. 8., 1901, The beginnings of natural Buchanan, W. T., Jr., 1980, The Hall Site, Mont- history in America: Annual Rept. Smithsonian gomery County, Virginia: Archeol. Soc. Vir- Inst. 1897, pt. 2, p. 357 -406. ginia, Quart. Bull., vol. 35, p. 72-100. Greene, J. C., L976, Science, learning, and utility: 1984, The Trigg Site, City of Radford, patterns of organization in the early Ameriean Virginia: Richmond, Archeol. Soc. Virginia, Republic, iz Oleson, A., and Brown, S. C., eds., 477 p. The pursuitof knowledge in theearly American Buck, J. L. B., 1952, The development of publie republic: Ameriean scientific and learned so- - schools in Virginia, 1607 -L9l?:Richmond, Com- cieties from colonial times to the Civil War: monwealth of Virginia, State Board of Edu- Baltimore, Johns Hopkins Press, p. 1-20. cation, vol. 35, no. 1. Guilday, J. E., 1962, The Pleistocene local fauna Carr, L., 1877, Report on the exploration of a mound of the Natural Chimneys, Augusta County, in Lee County, Virginia: Repts. Peabody Mus. Virginia: Carnegie Mus., Annals, vol. 36, p. 87- Amer. Archaeol. Ethnol., vol.2, p.75-94. t22. Coleman, G. N., 1982, The Reedy Creek Site, Parmalee, P. W., and Hamilton, H. H., 44Ha22, South Boston, Virginia: Archeol. Soc. 1977, The Clark's Cave bone deposit and the Virginia, Quart. Bull., vol. 37, p. 150-209. late Pleistocene paleoecology of the central Dabney, V., 1971, Virginia: the new Dominion: Appalachian Mountains of Virginia: Carnegie Garden City, Doubleday and Co. Inc., 629 p. Mus. Nat. Hist., Bull., no.2,87 p. Egloff, K., and Reed, C., 1980, Crab Orchard Site: Haynes, G., 1978, Excavations of a small Virginia A late Woodland palisaded village: Archeol. cave: Areheol. Soc. Virginia, Quart. Bull, vol. Soc. Virginia, Quart. Bull., vol. 34, p. 130-148. 32, p.37-41. Ewan, J., 1976, The growth of learned and scientific Heatwole, C. J., 1916, A history of education in societies in the southeastern United States to Virginia: New York, Macmillan,382 p. 1860,'in Oleson, A., and Brown, S. C., eds., The Hesse, 8., and Wapnish, P., 1985, Animal bone pursuit of knowledge in the early American archeology: Washington, D. C., Taraxacum, 132 republic: American scientific and learned so- p. cieties from colonial times to the Civil War: Holland, C. G., 1979, The ramifications of the fire Baltimore, Johns Hopkins Press, p.208-218. hunt: Archeol. Soc. Virginia, Quart. Bull., vol. Ewan, J., and Ewan, N., 1970, John Banister and 33, p. 134-140. his natural history of Virginia, 1678-1692: Ur- Holman, J. A., and McDonald, J. N.,.In press, A bana, Univ. Illinois Press,485 p. late Quaternary herpetofauna from Saltville, Fay, L. P., 1984a, Late Wisconsin Appalachian Virginia: Brimleyana, vol. 12. herpetofaunas: stability in the midst of ehange: Jefferson, T., 1787, Notes on the State of Virginia: Ph.D. Dissertation, Michigan State Univ., 69 NewYork, W. W. Norton and Co.,3L5 p. (Peden, p. W., ed., L972). 1984b, Mid-Wisconsinan and mid-Hol- + 1799, A memoir on the discovery of ocene herpetofaunas of eastern North America: certain bones of a quadruped of the elawed kind a study in minimal contrast, in Genoways, H. in the western parts of Virginia: Amer. Philos. H., and Dawson, M. R., eds., Contributions in Soe., Trans., vol.4, p.246-260. - Quaternary vertebrate paleontology: a volume Johnson, L. D., 1979, The Hercules Site, Alleghany in memorial to John E. Guilday: Spec. Publ. County, Virginia: Archeol. Soc. Virginia, 104 VIRGINIA DIVISION OF MINERAL RESOURCES

Quart. Bull., vol. 33, p. 141-14?. of Virginia: Blacksburg, VPI and SU, p. 603- Kiger, J. C., 1963, American learned societies: 609. Washington, D. C., Publie Affairs Press, 291 Patton, J. S., 1906, Jefferson, Cabell and the Uni- p. versity of Virginia: New York and Washington, Lanham, U., 1973, The bone hunters: New York, D. C., The Neal Publ. Co., 380 p. Columbia Univ. Press, 285 p. Peterson, O. A., 1917, A fossil-bearing alluvial MaeCord, H. A., Sr., 1979, The Flanary Site, Scott deposit in Saltville Valley, Virginia: Carnegie County, Virginia: Archeol. Soc. Virginia, Mus., Annals, vol. 71,p.469-474. Quart. Bull., vol. 34, p.l-32. Ray, C. E., Cooper, B. N., and Benninghoff, W. S., 1981, The Sullins Site, Washington 1967, Fossil mammals and pollen in a late County, Virginia: Archeol. Soc. Virginia, Pleistocene deposit at Saltville, Virginia: J. Quart. Bull., vol. 36,p.94-I2L. Paleontol., vol. 41, p.608-622. and Livesay, J. A., Sr., 1982, The Wetmore, A., Dunkle, D.H., and Drez, - Hertzler Site, Powhatan County, Virginia: Ar- P., 1968, Fossil vertebrates from the marine cheol. Soc. Virginia, Quart. Bull., vol. 3?, p. Pleistocene of southeastern Virginia: Smithson- 86-104. ian Misc. Coll., vol. 153, no. 3,25 p. -McDonald, J. N., 1984a, The Saltville, Virginia, Roberts, J. K., 1942, Annqtated geological bibli locality: a summary of research and field trip ography of Virginia: Richmond, Dietz Press, guide: Charlottesville, Virginia Div. Min. Res., 726 p. 45p. Robison, N. D., L978, Zooarchaeology: its history ..--.....-..-.--.---.----- 1984b, Population analysis of the Trigg and development, 'inBogan,A. E., and Robison, Site whitetail deer, in Buchanan, W. T., Jr., N. D., eds., A history and selected bibliography The Trigg Site, City of Radford, Virginia. of zooarchaeology in eastern North America: Richmond, Archeol. Soc. Virginia, p. 459-476. Tennessee Anthropol. Assoc., Misc. Paper no. 1984c, Paleoecological investigations at 2, p.l-22. Saltville, Virginia: Current Res., vol. l, p.77- Simpson, G. G., L942, The beginnings of vertebrate 78. paleontology in North America: Amer. Philos. 1985, Late Quaternary deposits and Soc., Proc., vol. 86, p. 130-188. - paleohydrology of the Saltville Valley, south- 1943, The discovery of fossil vertebrates west Virginia: Current Res., vol. 2, p.123-124. in North America: J. Paleontol., vol. 17, p. 26- In press, Valley bottom stratigraphy 38. of the Saltville Valley, Virginia, and its paleo- Stevens, T., 1979, The Wade Site (44Ha34): a flood- ecological implications: Nat. Geog. Soc. Res. - damaged Woodland site: Archeol. Soc. Vir- Repts. ginia, Quart. Bull., vol. 33, p. 148-159. and Bartlett, C. S., Jr., 1983, An as- Stoltman, J. 8., 1973, The southeastern United sociated musk ox skeleton from Saltville, Vir- States, inFitting, J. E., ed., The development ginia: J. Vert. Paleontol., vol.2, p.453-470. of North American archaeology: essays in the Miller, H. M., 1984, Colonization and subsistence history of regional traditions. Garden City, - ehange on the 17th Century Chesapeake fron- Anchor Press, p. 116-150. tier. Ph.D. Dissertation, Michigan State Uni- Turner, E. R., 1978, An intertribal deer exploita- versity, 517 p. tion buffer zone for the Virginia Coastal Plain- Oleson, A., 1976, Introduction: to build a new Piedmont regions: Archeol. Soc. Virginia, intellectual order, iz Oleson, A., and Brown, Quart. Bull., vol. 32,p.42-48. S. C., eds., The pursuit of knowledge in the early Virginia Association of Museums, 1984, Virginia American republie: American scientific and museums and historic sites guidebook: Rich- learned societies from colonial times to the Civil mond, Virginia Assoc. Mus.,40 p. War: Baltimore, Johns Hopkins Press, p. fra- Voss, J., 1976, Foreword: the learned society in ma. American intellectual life,'in Oleson, A., and Osborn, H. F., 1929, Thomas Jefferson, the pioneer Brown, S. C., eds., The pursuit of knowledge of American paleontology: Science, vol. 69, p. in the early American republic: Ameriean seien- 410-413. tific and learned societies from eolonial times Outlaw, A. C., 1982, Partnership in the statewide to the Civil War: Baltimore, Johns Hopkins survey: a call for continued assistance: Archeol. Press, p. uii-r. Soc. Virginia, Quart. Bull., vol.37, p. 148-149. Weems, R. E., and Higgins, B. B., 1977, Post- Pagels, J. F., 1979, The changing scene, inLinzey, Wisconsinan vertebrate remains from a fissure D. W., ed., Proceedings of the symposium on deposit near Ripplemead, Virginia: NSS Bull., endangered and threatened plants and animals vol.39, p. 106-108. PUBLICATION ?5 105

EMBRYOGENESIS DYSFUNCTION IN THE PLEISTOCENE/HOLOCENE TRANSITION MAMMALIAN EXTINCTIONS, DWARFING, AND SKELETAL ABNORMALITY

Dewey M. Mcleanr

CONTENTS

Page

Pleistoeene/Holocene transition climatic warming ...... 106 Pleistocene/Holocene transition "greenhouse" conditions ...... fOg Pleistocene/Holocene transition warming-extinction eonnection ... r0g

The hyperthermia-reduced uterine blood flow connection ...... lll The high ambient air temperature-embryogenesis dysfunction connection .....llz Pleistoeene/Holocene dwarfing and skeletal abnormality: a meehanism . . 118 grasslands Expanding and extinctions: a connection ...... 114 Geographical bias in the Pleistocene/Holocene extinetions: an explanation ...... llb 4eat-induced embryogenesis dysfunction: evolutionary implications ...... l16

ILLUSTRATIONS Figure 1. Weichselian climate changes in northwestern Europe ...... l0? 2. Fluctuations in average July temperature in lowland Britain indicated by beetle fauna ...... l0? 3. Variations of atmospheric CO, in air bubbles in the Dome C iee core (Antarctica) plotted against oxygen-I8 isotopic values . 10g 4. Analysis of percent of CO, in air bubbles in the Dome C ice core (Antarctica) plotted

5. Heat production and core temperature of homeotherms relative to environmental

6. Uterine blood flow responses (mVmin) of cows maintained in shade relative to cows denied shade after injection of estradiol ...... ll2 7. Schematic summarizationof a series of sequential physiological events beginning with mating, proceeding and to birth of offspring tL"* ...... ll2 8. Influence of rectal temperature at the time of matitrg"t on embryo survival in sheep

9. Spatial pattern of mean annual temperature differences (northern hemisphere) between

TABLE

Calibration of Pleistocene/Holocene transition climatic warming on different continents relative to times of extinctions and dwarfing ... .. l0g

lDepartment of Geological Scienees, Virginia Polytechnic Institute and State University, Blacksburg, vA 24061. 106 VIRGINIA DIVISION OF MINERAL RESOURCES

ABSTRACT Most large herbivores and their carnivore predators survived those extreme conditions of cold and Then, at the The Pleistocene/Holocene transition, a time of restricted living space and food, etc. rapid cold, glacial conditions followed by rapid global end of the ice age during the period of most and warming at the end of the last ice age, was a time warming, retreat northward of the ice sheets of fascinating paradoxes. During maximum ad- boreal forests, and increasing food source in the grasslands, occurred one vance of the Weichselian (European) and Wiscon- form rapidly expanding earth sinan (North America) ice sheets 20,000-18,000 yr of the most profound extinction events in mammalian ex- B. P., mammalian faunas facing the extreme cold history, the Pleistocene/Holocene semitropical and restricted living space and food reserves seem- tinctions. At a time when ranges of ingly adapted and survived. Then, during the late species should logically have been expanding, (Martin, 1984)' Wisconsinan period of climatie warming and in- ,"ng.t, in fact, were contracting creasing food supply in the form of expanding At a time when environmentally rigorous eondi- grasslands, mammalian faunas experienced one of tions should logically have been relaxing for high mammals ex- the most profound extinction events in earth his- latitude mammalian faunas, those tory, the Pleistocene/Holocene transition extinc- perienced the most severe extinctions. Concurrent and tions, in which many taxa suddenly became extinct with the extinctions were widescale dwarfing and/or dwarfed. Integration of environmental skeletal abnormalities. In this paper, I will address temperatures with modern reproductive physiol- these paradoxes via integrating the Pleistocene/ with mam- ogy identifies a linkage between high ambient air Holocene transition elimatic warming temperatures and reproductive dysfunction and malian reproductive physiology, expanding upon predicts extinctions and dwarfing during periods the linkage between elevated ambient air temper- I proposed of rapid warming, including that at the end of the ature and reproductive dysfunction that for the paradox- last ice age. earlier (Mciean, 1981). I account and skele- Modern mammals experience reproductive dys- ical nature of the extinctions, dwarfing, reduction of blood function in high ambienttemperatures, in environ- tal deformation via heat-induced mental temperatures that they supposedly have flow to the female uterus and consequent damage stages of pregpancy' been adapted to for the past several thousand years. to the embryo during various work is that High environmental temperatures cause hyper- A conclusion that will emerge in this thermally- thermia (elevation of core temperature) in female because of the sensitivity of embryos to mammals can- mammals. Hyperthermia, in turn, reduees the flow induced reduced uterine blood flow, Times of blood to the female uterus. Uterine blood flow not readily adapt to rapid climatic warming' are is a developing embryo's source of water, oxygen' of rapid rise in ambient air temperatures dys- and nutrients and also serves to dissipate heat from almost inescapably times of embryogenesis the uterus. Reduced uterine blood flow is associated function and, if severe enough, extinction. with increased embryo death, dwarfing, and skeletal abnormality. Pleistocene/Holocene transition mammals facing rapidly shifting climatic ex- PLEISTOCENE/HOLOCENE TRANSITION tremes from maximum cooling during the last ice CLIMATIC WARMING age to rapid warming could only have experienced reproduetive dysfunction in the form of reduced About 20,000-18,000 yr B. P., earth was locked uterine blood flow and consequential damage to in eold, ice age conditions (Figure 1). In Europe, maximum developing embryos. Embryogenesis dysfunction the Weiehselian ice sheet reached its America, was almost certainly commonplace during climatic extent about 20,000 yr B. P. In North reached warming in the Pleistocene/Holocene transition. the equivalent Wisconsinan ice sheet also its maximum extent about 20,000 yr B. P' Extreme cold conditions lasted from about 20,000-18,000 yr INTRODUCTION B. P. Following the most intense glacial conditions, earth began to warm up gradually until about began' The Pleistocene/Holocene transition was a time 14,000 yr B. P. when a more rapid warming of fascinating paradoxes. About 20,000-18,000 yr Based on studies of Coleoptera in northwestern warming B. P. terrestrial mammalian faunas experienced Europe, Coope (19?5) suggests large-scale yr B. P.; warming during some of the coldest conditions of the last ice age between 13,000 to 11,000 rate of loC per during maximum advances of the European Weich- early stages may have been at the in Europe selian and the American Wisconsinan ice sheets. decade (Figure 2). Deglacial warming PUBLICATION 75 L07

1983) indicate the same climate change chronology for the equatorial Atlantic, for the North Atlantic and the European continent. Duplessy and others (1981) indicate 6oC warming after 13,000 yr B. P., with a return to cold conditions 11,000-10,000 yr. B. P., and then warming up for good; study at the Fallortd. . Lofe weichselion Bay of Biscay south of the English Channel indi- I n|BrS?oo ro I cates that one-third of the iee melted between 16,000 and 13,000 yr B. P. For Europe, Wright (1977) indicated the first signs of warming about r) 13,000 yr B. P. I the somewhat x For North America, warmingwas at different; the two stages of warming seen in the o Middle Atlantic and European reeords are not evident in llreichselion eastern North America (Tom Webb, personal communication) nor in the continental interior. How- ever, the rate of warming was also rapid. Wright (1977) notes that spruce forests that had dominated the Middle West disappeared during an interval from about 12,000 yr B. P. in the modern prairie region to about 10,000 yr B. P. in the region of modern northern coniferous forests. The transition at any locality was abrupt, oecurring over just a few hundred years. Webb and Bryson (1972) note Figure l. Weiehselian climate changes in north- that the most profound climatic change in the western Europe (adapted from Coope and Sands, Wisconsin-Minnesota region was during the Pleis- 1966). tocene/Holocene within a few centuries of 11,300 yr B. P.; at that time the temperature rose about 3.3oC, and the length of the summer season increased by one month. In the southern United States, the cold, dry YEARS AVERAGE JULY TEMPERATURE B.P. s I . r0 r? rl r0 18 20(.ct Wiseonsinan elimate ended about 14,600 yr B. P. (Watts and Stuiver, 1980). In South Carolina, the / | PosT GLACTAL boreal coniferous forest was replaced by mesic at t0,000 J about 12,800 yr B. P. Georgia was the southern -....---- 1 WINDERMERE _- - TNTERSTADIAL limit for the Late Wisconsinan spruce forest. In 20,000 MAXIMUM EXPANSION OF ICE SHEETS the Texas region, Vaughn Bryant (personal communication) suggests a change from boreal to less boreal forests about 15,000-10,000 yr B. P., with 30,000 UPTON WARREN INTERSTADIAL no dramatic change at any one locality. In Missouri COMPLEX and Tennessee, pollen of deciduous trees began to appear in spruce-dominated areas about 16,000 yr B. P.; the spruce forests continued their retreat Figure 2. Fluetuations in the average July tem- northward, and between 13,500-12,000 yr B. P., the perature in lowland Britain as indicated by beetle spruce disappeared from Missouri, central Illinois, fauna (after Coope, 1977 a, b). and Tennessee (King and Saunders, 1984). King and Saunders (1984) note that between 12,000 and 10,000 years ago, the ice retreated north of the Great Lakes. Following the retreating ice, the early and in the Atlantic Ocean seems to have occurred pioneering plant eommunities of spruce woodland in two stages. For the high-latitude North Atlantic, and tundra were replaced by a brief episode of pine Ruddiman and Mclntyre (1981) report warming and then deciduous species. The shift in spruce to at 13,000-11,000 yr B. P., cooling from 11,000- pine dominance forms a prominent time-transgres- 10,000 yr B. P., and then warming again from sive marker horizon throughout eastern North 10,000-8,000 yr B. P. Mix and Ruddiman (Kerr, Ameriea (Ogden, 1967). 108 VIRGINIA DIVISION OF MINERAL RESOURCES

PLEISTOCE NE/HOLOCENE TRANSITION .GREENHOUSE" CONDITIONS

Cause of the global warming at the end of the 8l8O 7"o -56 last ice age has long been the subject of intense o -52 -48 interest. It is now known that the warming at the end of the last ice age was concurrent with a rapid o buildup of CO, in the atmosphere that did not stop aa until it reached pre-industrial levels about 10,000 yr B.P. (Figure 3). During the last glaciation, atmospheric CO, varied between two levels: (1) 180- a 200 parts per million (ppm) in very cold conditions o and (2) 250-270 ppm in relatively warm periods.

Pre-industrial levels were around 260-280 ppm (the J() pre-1850 level was about 266 ppm; the modern I ro atmosphere contains about 340 ppm). Study of the x Dome Ice Core from Antarctica (Figure 4, after Delmas and others, 1980) records low CO, values (L between 22,000-L5,000 yr B. P., during the coldest cl part of the last ice age; the CO, values of 0.016 o t5 percent were about one-half that of the modern o! value of 0.033 percent. Neftel and others (1982) o indicate that Cor-indueed greenhouse warming 20 might be involved in control of the ice ages, noting that the inereased atmospherie CO, contributes significantly to the global warming at the transition 25 from the last glaciation to the Holocene. The mechanism by which CO, is released into the atmosphere is not known; however, an interesting sequence of events seems involved. The sug- 30 gestion by Imbrie and Imbrie (1980) that variations in the earth's orbit are the fundamental cause of o.ol o.o2 0.o3 o.o4 o/o of COz ( v/v)

Figure 4. Analysis of percent of CO, in air bubbles in the Dome C ice eore (Antarctica), plotted against E o, oxygen-l8 isotopie values. The dashed line repres- CL ents modern atmospheric CO, content (0.033 per- N cent) (after Delmas and others, 1980). ()o the succession of Pleistocene ice ages now seems confirmed. Atmospheric CO, seems to boost effects of orbital variations. Shackleton and Pisias (1985) indicate that atmospheric CO, concentration leads I 23 ice volume over the orbital bandwidth, and is forced ( by orbital changes by a mechanism presently not Age x lOayr BP ) ,understood, with a short response time. They further note thatvariations in atmospheric CO, should Figure 3. Variations of atmospheric CO, over the be seen as part of the forcing of ice volume changes. past 40,000 years (after Neftel and others, 1982). The order of events suggests that orbital variations PUBLICATION 75 109

are the primary cause of climate change and that In fact, nearly two-thirds of North America's CO, exerts intermediate influence. genera of large mammals, along with larger-bodied avian and reptilian genera, became extinct during PLEISTOCE NE/HOLOCENE TRANSITION the late Wisconsin and early Holocene. The late WARMING-EXTINCTION CONNECTION Wisconsinan megafauna includes about 49 genera, of which 33 genera (67 percent) became extinct During maximum advance of the Weichselian during the Pleistocene/Holocene transition (McDo- and Wisconsinan ice sheets, mammals would have nald, 1984). had to adapt to extreme conditions of cold, res- That the steepest gradient of Pleistocene/Holo- tricted living space, and diminished food reserves, cene transition global warming was coupled tem- etc. That most large mammals survived attests to porally to global mammalian extinctions, dwarf- their ability to adapt to cold conditions. Most large- ing, and skeletal abnormality is well documented bodied herbivores in North America about 2b.000- (Table). Graham and Lundelius (1984) indicate that 20,000 yr B. P. and their predators survived the the critieal period of environmental change for late Wisconsin maximum (McDonald, 1984). most continents seems to have been 10,000 to 12,000 Starting about 18,000 yr B. P. came a relaxation yr B. P. For northeastern North America, Marshall of severe ice age conditions. In North America the (1984) notes that between 12,000 to 10,000 yr B. ice sheets that had advanced from Canada into the P., the ice sheet retreated north of the Great Lakes, northern United States began retreating north- pine and deciduous species replaced the spruce ward. Boreal forests that, during coldest conditions, woodland, and that Mawnrut americanum became had migrated as far southward as Georgia in the extinct. Between 11,500 to 10,500 yr B. P., the east to Texas in the west began their rapid retreat American southwest experienced megafaunal ex- northward to their modern ranges. Concurrenfly, tinctions, and the greatest vegetational-climatic grasslands developed in the North American con- change since the Sangamon interglacial. For the tinental interior east of the Rocky Mountains and Appalachian region, the terrestrial mid-Appala- expanded their area of coverage eastward and chian mammalian fauna "crashed" about 10.000 to northward. Yet, during that time of ameliorating 12,000 yr B. P. (Guilday, 1984). Central and South climatic conditions and expanding living space and America and Africa also experienced large mam- food reserves, in what would seem to be an emerg- mal extinctions at about the same time as in the ing "golden age" for mammals, many taxa became Holarctic about 11,000 yr B. P. (Guthrie, 1984). In extinct, and dwarfing and skeletal abnormalities terms of geographical bias, the extinctions were were widespread. McDonald (1984) notes that the more severe in North America than in Europe, limiting intensity of most selecting forces was Africa, and Asia (Martin, 1984). I will address this relaxed after 18,000 yr. B. P., and that the asso- phenomenon in the section titled "Geographical ciated environmental changes should not have pro- Bias In The Pleistoeene/Holocene Extinctions: An duced the cluster of extinctions whieh occurred. Explanation."

Table. Calibration of Pleistocene/Holocene transition climate warming on different continents relative to times of extinctions and dwarfing (modified from Marshall, 19g4).

LAND AREA CLIMATE CHANGE EXTINCTION DWARFING (yrs B. P.) (yrs B. P.) (yrs B. P.)

Africa 12,000-9,000 12,000-9,000 l_2,000 Australia 26,000-15,000 26,000-20,000 26,000-15,000 16,000-12,000 (some 18,000-15,000) Europe 14,000-10,000 12,000-10,000 15,000-12,000 North America 12,000-10,000 12,000-10,000 12,000-10,000 12,000-11,000 11,500-10,500 12,000-11,000 11,500-10,500 11,000-10;000 South America 10,000 12,000-9,000 (known but not well doeumented) 110 VIRGINIA DIVISION OF MINERAL RESOURCES

To summarize to this point, linkage between cause large mammals have relatively small sur- rapid Pleistocene/Holocene transition climatic facelvolume ratios they would have retained warr1ing and extinctions, dwarfing, and skeletal relatively greater body heat than would small abnormalities is firmly established. mammals during the rapid climatic warming. Small body size, or the ability to reduce body size NEW LIGHT ON OLD PARADOXES during the interval of rapid warming by allowing the more ready dissipation of body heat, would have Modern ecological and evolutionary theory pre- been most advantageous. That the warrning was dict that climatic warming following the maximum rapid is indicated by Kins and Saunders (1984) glacial advance should have resulted in an increase who state that the warming terminating the Pleis- in megafaunal biomass and diversity; if competitive tocene was faster than any other temperature exclusion at any trophic level was going to produce change since the beginning at the last interglacial the extinction of any taxon, it should have done stage. Haynes (1984) notes that one of the most so during the maximum glacial advance 20,000 to abrupt climatic changes came at the end of the 18,000 yr B. P. period when selection was most Pleistocene about 11,000 yr B. P. Knowledge of intense and limiting (McDonald, 1984). The fact modern heat-induced reproductive dysfunetion that most mammals survived the cold and suffered predicts that extinctions, dwarfing, and skeletal extinctions during the warming at the end of the abnormalities would have been inescapable during last ice age indicates that we need new ways of rapid global warming and expansion of grasslands examining the Pleistocene/Holocene transition ex- at the end of the last ice age, and especially among tinctions. I attempt to do so by integrating changes the larger mammalian species. in ambient air temperature with reproductive dysfunction via heat-induced damage to developing ELEMENTS OF THERMOREGULATION embryos. The Pleistocene/Holocene transition extinctions In ambient temperatures where nearly eonstant affected mammals large and small, birds, and core temperatures can be maintained, mammals reptiles. However, several researchers indicate that are homeothermic; horvever, the core temperature large mammals were affected most severely of all. can vary with changes in ambient air temperature Merilees (1984) notes that some process bore more (Figure 5). Within the"Zoneof Homeothermy," and heavily on larger than on smaller mammals. Martin at an optimum environmental temperature that is (1984) indicates that large terrestrial mammals variable depending upon the species and its accli- were most affected with the extinet genera equal- matization, body heat production is at a minimum ling or exceeding 44 kg (100 lbs), and that the in the "Zone of Thermal Neutrality." In the "Zone sudden loss of large mammals must mean that large of Thermal Neutrality" a nearly constant core size was a handicap. McDonald (1984) notes that temperature is maintained with minimal effort whatever the cause of the extinctions in North America, the phenomenon was continent wide, was selective principally against large mammals, and likely resulted from some ultimate, ubiquitous HYPO- IYPER- ZONE OF HOMEOTHERMY causal factor that quickly rendered much of the NE.F(M I nEnMI

continent's megafauna "unfit." CORE TEIIPERATURE Two elements are common to the extinctions: global warming and the extreme sensitivity of developing embryos to above-normal heat. High ambient temperature can, by elevating the uterine areas of female mammals, damage developing embryos. Reproductive dysfunction among modern mammals during hot periods, to which they are supposedly adapted, is commonplace (discussion is -owER LmlT IPPER LITIT OF CONTROI. OF COI{TROL included in following sections). The rapid global DEATH BY DEATH BY warming at the end of the last ice age would have COLD HEAT rendered many taxa "unfit." Rapid rise in ambient air temperature during Figure 5. Heat production and core temperature the Pleistocene/Holocene transition would have im- of homeotherms relative to environmental temper- posed especial hardship on large mammals. Be- atures (modified from Bianca, 1968). PUBLICATION 75 111 from an animal's thermoregulatory mechanisms. the core temperature rises (the animal beeomes MeDowell (1968) notes that the desirable thermal hyperthermic). Bianea (1968) indicates the rise in environment for most livestock is between 13 and body temperature in turn brings about an increase 18'C. in the metabolic rate because of the van't Hoff effect Mammals produce body heat. In addition, they which can cause death by heat. The van't Hoff absorb heat from the sun. Bianca (1968) indieates effect, with respeet to mammalian heat production, that cattle may absorb three times as much heat represents the two-to three-fold increase in heat from the sun as they produce by their metabolism. production brought about by a 10"C increase in In ruminants, the microorganisms contained in the tissue temperature; i.e., the temperature coeffi- rumen constitute an extra source of heat. contri- cient, Q10, of mammalian tissues is of the order buting up to 10 pereent of an animal's basal heat of.2 to 3. The upper lethal temperature is only a produetion. few degrees above normal body temperature; in Studies of domestic mammals show that they most species it is 42 to 45oC or only about 3 to 6oC undergo diurnal rhythms of body temperature above normal body temperature. However, as will reaching minimum and maximum values in early be discussed in the following sections, environmen- morning and late afternoon, respectively (Bianca, tal temperatures far below those required to kill 1968). Under natural conditions, heat stress oecurs outright via thermoregulatory failure can have in a rhythmic cyele with maxima lasting for a few great impaet upon mammalian bioevolution. hours each day. The larger the mammal, the greater the amount of heat stored. Heat stress THE HYPERTHERMIA-REDUCED UTERINE under natural conditions occurs in a rhythmic 24- BLOOD FLOW CONNECTION hour pattern with maxima lasting only a few hours around mid-day. To avoid thermoregulatory prob- That high environmental temperatures can cause lems generated by retention of excessive body heat, the core temperatures of mammals to increase is mammals must lose body heat by radiation and well documented (discussed below). Elevated core convection during the relatively cool night hours. temperatures in female mammals can, in turn, Mammals lose heat from the body to the environ- trigger reproductive dysfunction. Such dysfunction ment through sensible heat loss (radiation, convec- is common every summer in temperatures to which tion, and conduction), and by behavorial tactics modern mammals are supposedly adapted. such as seeking appropriate environments, wallow- The mechanism by which high ambient temper- ing, sweating, and panting, etc. atures induce reproductive dysfunction is becom- In ambient air temperatures rising above those ing clear. High ambient air temperatures, by in- in which an animal's heat produetion stays within ducing hyperthermia in female mammals, reduces Ihe "Zone of Therm al Neutral ity, " a thermal steady the flow of blood to the uterus. Uterine blood flow state cannot be maintained and body temperature (UBF) is a developing embryo's source of water, will rise eontinuously, indicating true hyperther- oxygen, nutrients, and hormones (Senger and oth- mia. An animal will then activate defense reactions ers, 1967; Bazer and others, 1969; Barron, 1970). against overheating (vasodilatation, sweating, or In addition, UBF helps to dissipate uterine met- panting). Ambient air temperatures which cause abolic heat (Abrams and others, l97l; Gwazdaus- eore temperatures to rise in most domestic mam- kas and others, 1974). Brown and Harrison (1981) mals are about 28 to 33"C but may be as low as note that, unlike many body tissues that have 2l to 24"C in some breeds of lactating eattle (Bi- autoregulation, the uterus of the pregnant ewe at anca, 1968). McDowell (1968) indicates that thermoneutrality is unable to maintain constant temperatures between 21 and 28oC cause discom- blood flow as perfusion pressure decreases (Greiss, fort for an animal, triggering changes in behavior, 1,966; Ladner and others, 1970). decreased food intake, and less efficient perform- That UBF is affected by environmental heat has ance. been demonstrated in cows fitted with blood flow Hyperthermia occurs at relatively low environ- transducers around a mid-uterine artery, given mental temperatures when air humidity and solar injections of estradiol, and then plaeed in shade/ radiation are high. With additional heat load, no shade eonditions. Cows placed in the shade sweating or panting increase; however, eooling showed a greater resultant rate of increase in UBF derived via these means cannot maintain homeo- than did cows that were denied shade, demonstrat- thermy and body temperature begins to rise. At ing that UBF is affected, by thermal stress (Roman- the "Upper Limit of Control" heat defense meeh- Ponce and others, 1978) (Figrrre 6), In the subtrop- anisms can no longer maintain homeothermy, and ical climate of Gainesville, Florida, shade provided rtz VIRGINIA DIVISION OF MINERAL RESOURCES

THE HIGH AMBIENT AIR TEMPE RATURE.EMBRYOGENE SIS DYSFUNCTION CONNECTION

Haf.ez (1968) indicates that the effects of tem- .s perature on pregnaney and embryonic survival E vary with the species, temperature, period of ex- E posure, length of gestation, and stage of pregnancy; extremely high temperatures lower fertility, de- o= J crease litter size. increase abortion. and increase ll. fetal resorption. Heat can affect male mammals o via damage to spermatozoa, resulting in elevated o roo embryo death (Howarth and others, 1965; Ulberg oJ @ and Burfening, 1967; Burfening and Ulberg, 1968); however, as Bishop (1964) has indicated, most UJ z investigators consider the female responsible for G lrl embryonic death. F Ulberg and Sheean (1973) indicate that infertility f in females subjected to high environmental temperatures is the result of damage eaused by high body temperature acting directly upon the embryo im- 246 mediately after fertilization (Figure 7). Pregnancy HOURS is most sensitive to environmental stress during the period between ovulation and implantation. The Figure 6. Uterine blood flow responses (mfmin) sheep zygote is most sensitive to the harmful effect of cows maintained in shade relative to cows denied of high ambient temperature during the initial shade after injection of estradiol. Cows maintained stage of cleavage while in the oviduct (Dutt, 1963). in the shade had a relatively greater increase in Rectal temperatures taken in sheep and cattle uterine blood flow than did cows denied shade (after about the time the egg was undergoing the first Roman-Ponee and others, 19?8). cell division showed that in both the pregnancy rate decreased with each degree of increase of rectal temperature. during the summer lowered respiration rates and rectal temperatures of laetating dairy cattle; shade improved both milk yield and reproductive per- formance over control cattle denied shade (Roman- TERM Ponce and others, 1977); Roman-Ponce and others (1978) state that reduction of UBF in response to thermal stress may elevate temperature of the reprodirctive tract and prejudice the developing embryo. In the lactating bovine, environmental temperatures above zl.t"C cause deep body temperature to rise (Thatcher, 1974); temperature increase from 21.1 to 35'C caused the conception rate to decline from 40 to 3l percent. High temperatui"es also affect reproduction among smaller mammals by reducing the rate of uterine blood flow. Oakes and others (1976) noted Figure 7. Schematic summarization of a series of a 25-48 percent decrease in UBF of.ewes exposed sequential physiological events beginning with to induced hyperthermia. In warm summer (27"C) mating, and proceeding to birth of offspring at in England, Leduc (t972) noted a 46 percent re- term: it is e'stimated that one-third or more of duction in placental blood flow in rabbits which potential young cease to develop after the embryo was associated with an increase in number of runts enters the uterus (after Ulberg, L. C., and Bur- and dead fetuses. fening, P. J., 1967). PUBLICATION 75 113

Rabbit embryos are also more suseeptible to heat damage during first cell division than during the . a- - a cotlle (35-42 dovs) second. Ulberg and Sheean (1973) suggest that heat o-o sheep (term) load on a female while her fertilized eggs are J undergoing first cleavage may alter the rates of .t.]. metabolism within the embryos, perhaps interfer- ; - \.184) oQ: E (t6)O/< -t-r-^t+r) ing with RNA synthesis. Four-day-old rabbit em- 150 bryos morphologically abnormal because of heat (t, ...{ao) "\ stress may be capable of initiating an implantation zF \o(37) which is prominent gestation. IJ site at twelve days of Q Slight increase in temperature acting directly on E. UJ the embryo disorients the mechanism controlling (L development within the blastomere; thus, whereas 25 cells may continue to develop for a time, perhaps even forming a morphologically normal embryo, they subsequently die. Ulberg and Burfening 37.5 38.O 58.5 39.O 39.5 40.O (1967) note that temperature stress during early RECTAL TEMPERATURE ("C !O.25") developmental stages results in later death of the embryo. Figure 8. Influenee of rectal temperature at the Ulberg and Burfening (1967) indicate that if a time of mating on embryo survival in sheep and slight increase in temperature surrounding the cattle. The value in parentheses represents the embryo at the time of the first few cell divisions number of observations for each temperature eontributes to delayed death, the body temperature group. Note that increase in body temperature of the female near the time of mating should be reduces survival rate (after Ulberg, L. C., and associated with the chances of success of that Burfening, P. J., 1967). mating. Results of their experiments on effects of heat load on reproductive capability of sheep and cows are shown in Figure 8. The authors note that among mammals on a nearly global scale (Table); an association between body temperature at the skeletal abnormalities are also known during the time of mating and embryo survival seems indi- same time period. Marshall (1984) notes that the cated. Alliston and others (1965) note that variation knowledge that extinction and dwarfing are con- of atmospheric temperature by as little as 1 to 2oC current processes may provide insight in extinction may alter the point in the reproductive process that phenomena, and that if we ean identify mechanisms is affected. According to Woody and Ulberg (1964) causing a deerease in body size, we can better eontrol ewes maintained in constant 70'F had understand cause of the extinctions, assuming that significantly higher level of fertility than ewes both extinetion and dwarfing are indeed linked to maintained in a room with a constant.90oF. a common causal factor(s). In fact, Pleistocene/ Frank Gwazdauskas (Dairy Science, VPI&SU, Holocene transition extinctions, dwarfing, and personal communication) indicates that probably skeletal abnormalities seem to have common cause 50 percent of cattle embryos die under natural in the rapid increase in ambient air temperatures conditions, with 10 percent loss in the first 4 days at the end of the last ice age. and 40 percent loss between 6 to 15 days. Hafez (1968) notes that environmental temper- To summafize to this point, if modern hot ature is perhaps the most important climatic com- summer temperatures-to which modern mam- ponent affecting prenatal growth. Miniature calves mals are supposedly adapted-have such profound, are often born to unadapted European breeds of and widespread, impact upon mammalian repro- livestock following a summer of pregnaney in the ductive capability, the cold-adapted mammals of tropics. Pregnant ewes exposed,to high experimen- the Pleistocene/Holocene transition facing rapid, tal temperatures bore miniature lambs, the reduc- and extreme, climatic warming could only have tion in weight being proportional to the Jength of suffered reproductive dysfunction. exposure. Yeates (1958) demonstrated that dwarfing is an effect of temperaturq and not to reduced PLEISTOCENE/HOLOCENE feed intake. Heat-induced dwarfs are lvell-propor- DWARFING AND tioned miniatures, in contrast to the long-legged SKELETAL ABNORMALITY: A MECHANISM thin lambs from underfed ewes. Concurrent with the mammalian extinctions in Brown and Harrison (1981) note that intrauterirae the Pleistocene/ Holocene transition was dwarfing growth retardation (IUGR) resul'ts from exposure LL4 VIRGINIA DIVISION OF MINERAL RESOURCES of ewes to high environmental temperature during reduction. Many surviving European mammals late gestation, and that since uterine blood flow have become smaller during postglaeial times is a major source of nutrients for the developing (Kurten, 1972). Examples include the gigantie fetus, altered placental function because of heat Pleistocene brown bear of Europe (that in post- stress may be a contributing factor in ovine IUGR. glacial times has been reduced in size to the smal- Their experiments rule out maternal nutrition as lest living race of the species Ursus arctos),lion, a primary cause of fetal dwarfing. Roman-Ponce auroch, bison, elk, and red deer, etc. Kurten (1972) and others (1976) have reported redueed mid-uter- reports that dwarfing as a prelude to extinction ine artery blood flow in response to heat stress in has been observed in many cases. nonpregnant ovariectomized ewes. Brown and oth- For the phenomenon of skeletal abnormalities, ers (1976) also reported decreased uterine blood Hafez (1968) notes that climatic stress or hyper- flow during heat stress in pregnant rabbits. thermia during pregnancy may cause fetal mal- Small lambs resulting from maternal heat stress formations in some species. The effects of temper- of 25 or 53 days duration are proportional dwarfs ature depend upon the stage of pregnancy at the and occur independently of level of nutrition time of exposure, since most malformations occur (Brown and others, 1977); dwarfing was not alle- during organogenesis. The severity of the defects viated by increasing the dietary protein of heat is usually less when heat exposure is delayed until stressed pregnant ewes. Dwarfing occurs with ma- later stages of embryonic development. In hot ternal heat stress during only the last month of climates the later maturing parts of the animal pregnancy. Intrauterine growth retardation, and body such as the loin and rump are stunted. At not disproportional fetal dwarfing, results from sexual maturity, such animals will have smaller late gestation heat stress. pelvic cavities than normal and may produee still- Growth of animals after weaning can also be birth calves because the size of the pelvis is too stunted by high environmental temperatures small for a normal delivery. (Hafez,1968), the degree of stunting varying with Abnormalities were common in Pleistocene/Hol- the breed, age, fatness, plane of nutrition, and ocene transition mammalian faunas. Guthrie (1984) relative humidity. Growth of European breeds of suggests that more spectacular than the general cattle is depressed at temperatures constantly postglacial decline in body size was a reduction in maintained above 24"C. High temperatures may size of some body parts. Guthrie (1984) also reports be lethal to newborn mammals whieh are not well thatvirtually all mammoths associated with Clovis adapted to withstand high temperatures early in points in the New World are diminutive and have life; lambs and calves are especially susceptible. redueed tusks. McDonald (1984) reports that an Lambs between two and seven days of age eannot unusually high frequency of skeletal abnormalities survive longer than about two hours at 38oC, or has been found in Bison from 11,000 to 9,000 yr more than three hours of solar radiation exposure. B. P. For the phenomenon of dwarfing during the To summarize, high environmental tempera- Pleistocene/Holocene transition, all modern large tures affect modern mammals of all sizes by re- species in North America seem to have experieneed ducing the flow of blood (UBF) to the female uterus. size deerease during the Holocene (Gilbert and Reduction of UBF, in turn, results in embryogene- Martin, 1984). Dwarfing occurred in species that sis dysfunction that can be expressed in embryo survived and also in those that became extinct, and death, dwarfing, and skeletafmorphological ab- was concurrent with the extinctions (Guthrie, normalities. In light of this knowledge, it would 1984); some species experienced dwarfing just prior seem that dwarfing and skeletal abnormality trig- to extinction. Tchernov (1984) reported that most gered by warming in the Pleistocene/Holocene mammals in the Middle East responded directly transition would have been inescapable. I suggest and quickly to Pleistocene/Holocene transition cli that the Pleistocene/Holocene transition extine- matic warming by drastic diminution of body sizes. tions, dwarfing, and skeletal abnormalities all have For the Israel area, Davis (198f) reported that the common cause in the climatie warming at the end fox, wild boar, gazelle, goat, and probably aurochs of the last iee age. underwent a size reduction at the end of the Pleis- tocene 10,000-12,000 yr ago; most earnivores in EXPANDING GRASSLANDS AND Israel and Lebanon show size reduction in the EXTINCTIONS: A CONNECTION Pleistocene/Holocene transition (Kurten, 1965). For the Natural Trap Cave area of Wyoming, During the Pleistocene/Holocene transition cli- Gilbert and Martin (1984) indicate that climate matie warming and retreat of ice sheets and boreal change was reflected not in extinction but in size forest northward, grasslands began to appear and PUBLICATION ?5 115

most dramatic extinctions' in the North Ameriean interior. According yr B. P., the time of the expand-Guthrie should have been (1984), North American grasslands be- itr. tung. of semitropieal species to eontraeting' About half the gan to expand across the mideontinent in a roughly ..p""aitg and not that became extinct triangular area with the base backed up against families o1 la"g" vertebrates Ameriea during the late Pleis- the Rocky Mountains and the apex heading toward in temperate North low latitudes, malply in the Amer- Ohio andPennsylvania. At the time the grasslands io."tt" to*ive at their greatest rate from 12,000- ican tropics (Webb, 1984). were expanding at that the yr B. P., mammoths, horses, and camels, in figit oi modern climate dynamics, 10,000 mammalian extinctions etc., beeame extinct. Thus oecurred one of the most Pleistolene/Holocene in the higher latitudes faseinating paradoxes of the Pleistocene/Holoeene would have been more severe periods of warming, the During a time of warming, and expand- is predictable. During transition. than do ing grasslands that would have provided an in- ftigi." latitudes heat up relatively.more (Fig:ure 9)' Thus' during food source for many herbivores, extinc- ih6lo*"t latitudinal areas tions"r""iing oceurred. McDonald (1984) touehes upon the paradox by noting that modern ecological and evolutionary theory predicts that most environmental changes during the time of extinctions should have produced an increase in megafaunal biomass and diversity, and not extinctions. However, as noted in the seetion titled "The Hyperthermia-Reduced Uterine Blood Flow Con- nection." mammals denied shade suffer reduction in UBF. Concurrent with the expansion of grasslands, woodland areas, that would have provided shade during a time of general climatic warming, were being reduced in size. The reduction of shade would have triggered hyperthermia and UBF reduction among mammals long adapted to cold and shade. Large mammals, because of their relatively small surface/volume ratios, would have been especially affected. Embrlrcgenesis dysfunction was virtually inescapable during the Pleistocene/Hol- ocene climatic warming and expansion of grasslands.

GEOGRAPHICAL BIAS IN THE PLEISTOCENE/HOLOCENE EXTINCTIONS: AN EXPLANATION

Another fascinating paradox of the Pleistocene/ Holocene mammalian extinctions is that during relaxation of extremely cold glacial conditions, the higher northerly latitudes should have become relatively congenial to life. Instead, Guthrie (1984) notes that although the late Pleistocene extinctions pattern mean annual temper- exhibit a global pattern, they were mainly Holarc- Figure 9. Spatial of tic and more dramatie in the northern portions ature diffeiences (northern hemisphere) between (after and (Alaska and Siberia), whereas the rest of the world cold- and warm-year groups Wigley oceurs in high was affected more lightly. Webb (1984) indicates others, 1930). Maximum warming than that it has long been evident that glacial cycles latitudes and continental interiors' up to more increase in a and their climatic effects were harsher at higher five times the hemispheric mean Siberia to latitudes. That high-latitude mammalian faunas region from Finland across Russia and America has were more affected than low-latitude faunas during about 90o8. A large part of North or more times the extinctions is expressed by Martin (1984) who positive temperature changes of two notes that during warming from 10,000 to 12'000 the hemispheric mean. 116 VIRGINIA DIVISION OF MINERAL RESOURCES

warming intervals, high latitude mammals have ice age, some other major Cenozoic mammalian a relatively greater heat load to have to adapt to. extinction events seem also to have oceurred at the The cause of the relatively greater degree of end of ice ages. Webb (1984) correlates the late warming in high latitudes is suggested in the work Hemphillian, late Blancan, and late Irvingtonian of Wigley and others (1980). Comparison of com- extinetion events with glacial terminations. Thus, posites of the five warmest years with the five as indieated by the record, climatic warming seems coldest for the period 1g2b to 1924 indicates that to be more lethal to terrestrial mammalian faunas during warm periods maximum warming oecurs than does climatic cooling. in high latitudes and in continental inteiiors. In The sensitivity of mammals to climatic warming a region from Finland to across the northernmost has implications for the future. Today, our burning parts of Russia and Siberia to about 90oE, the of the fossil fuels coal, oil, and gas is eausing the temperature increase is up to more than five times waste product CO, to accumulate in the atmos- above the hemispheric mean temperature increase. phere. Atmospherie CO, is predicted to double over A large area of northern North America also shows pre-industrial levels during the next century, caus- temperature increases of two or more times the ing global warming ("greenhouse" conditions) of hemispherie mean. Apparently, warm periods in perhaps 3oC in the northern hemisphere. Effects the northern hemisphere have intensified high- of future "greenhouse" conditions upon our civili- latitude (50-70'N) westerlies, greater cyclonic ac- zation have been the subject of lively debate. Topics tivity in the arctic and subarctic regions of the most discussed have focussed upon rise in sea level eastern hemisphere, and a westward displacement associated with melting of polar ice, and the po- of _the Siberian High in the winter half of the year. tential shifting of rainfall patterns away from In application to the Pleistocene/Holocene extine- agricultural regions, etc. We even hear that the tions, the regions showing greatest warming dur- positive effects of global warming might be bene- ing modern warming intervals in the ,ro"1hern ficial, and even outweigh negative ones. In the hemisphere eoincide with the regions that expe- sweeping speculations upon our future, the effects rienced the most intense patterns of pleistocene/ of warming operating directly upon mammalian Holocene extinctions in the northern hemisphere. reproductive physiology have been ignored. In light Conversely, some regions of the northern hemi- of the sensitivity of modern mammals to high sphere apparently cool down duringgeneral north- arnbient temperatures this would seem to be a ern hemispheric warming. These areas, that in- serious oversight. clude Japan, much of India, an area about Turkey, I suggest that the Pleistocene/Holocene mammal- the Iberian peninsula, North Africa, and a region ian extinctions provide an example for evaluating in central Asia, etc., seem to have been more lightly the impact of a future climatic warming upon affected by extinctions. modern mammals. Modern mammals are the sur- vivors of the Pleistocene/Holocene extinctions: I HEAT-INDUCE D EMBRYOGENE SIS suggest that many survived primarily by being able DYSFUNCTION: EVOLUTIONARY to reduce surface/volume ratios (i.e., by becoming IMPLICATIONS dwarfed), and thus maintain core temperatures conducive to reproductive success at a time when Grayson (1934) notes that attempts to explain the large size was disadvantageous. Pleistocene extinctions are made more difficult Duringthis modern interglacial we are, in effect, because we have no theory that relates organisms living on a "thermal plateau." We have no evidenees to their environment sufficiently to explain any set that modern mammalian faunas can withstand a of vertebrate extinctions. I suggest that high am- rise in ambient air temperature by several degrees bient temperatures acting directly upon mammal- Celsius over the short span of a few decades. In ian embryos explains the eause of the extinctions, fact, evidences abound suggesting that we are not dwarfing, and skeletal abnormalities during cli- as well adapted to this thermal plateau as we might matic warming at the end of the last ice age. suspeet. Every summer, all about us, are clear signs This natural linkage between climatic warming of reproductive dysfunction during hot summer and vertebrate reproductive dysfunction shows temperatures that we are supposedly adapted to. that mammals are far more sensitive to climate A rapid elevation of temperatures can only further change than has been suspected and may help to aggravate reproductive dysfunction, causing the explain other mammalian extinetion episodes dur- phenomenon to expand in scope and magnitude. ing the Cenozoic. Similar to the most reeent (Ran- If, as Wigley and others (1980) suggest, certain cholabrean) extinction episode at the end ofthe last regions of the northern hemisphere actually heat PUBLICATION 75 Lt7

up several times above the mean hemispheric tem- heat. In light of sensitivity of modern mammals perature increase, the North American land mass to even normally hot summer temperatures, it could be in for an interesting test of theory. History seems that heat-induced embryogenesis dysfunc- shows us that during the last time of rapid warming tion would have been inescapable during the Pleis- (at the end of the last ice age) that extinctions were tocene/Holocene transition. That the ambient air the most severe of all in North America. If the temperature-embryogenesis dysfunction linkage Pleistocene/Holocene transition extinetions have a may have predictive power is hinted at in the lesson to offer, it is that a future rapid elevation several Cenozoic mammalian extinction episodes of temperatures to above modern levels might be that occurred during warming at the end of ice potentially dangerous, and especially so for mam- ages (Webb, 1984). It is interesting that, similar malian faunas of the North American land mass. to the K-T transition, the mammalian extinctions at the end of the last ice age also oecurred during CONCLUSIONS global warming associated with CO, buildup in the atmosphere. In the middle 1970's, I became fascinated with On a final point, I have avoided critical discussion involvement of the carbon cyele in evolution of of other extinction models in this paper. My limited earth's climate and life and began a series of knowledge of the specific Pleistocene/Holocene tax- multidisciplinary studies that I hoped might illu- onomic base preeludes me being a critic. Instead, minate principles operating in that vast, and es- I have approaehed the Pleistocene/Holocene sce- sentially unknown, complex. In my early work, I nario as part of my general studies searehing for foeussed upon the Tertiary/Cretaceous (K-T) tran- plinciples relating to involvement of the carbon sition of about 65 Myr ago because it was rieh in eycle in the evolution of life on our planet. As such, information. I soon eame to believe that the K-T all those researchers who have collected, and in- transition was a time of perturbation of the carbon terpreted, Pleistocene/Holocene data have been my cycle, and associated COr-induced global warming teachers, and have my appreeiation for providing via the so-called "greenhouse" conditions (Mclean, me data, and viewpoints, to integrate in my seareh 1978). By now, I have linked the K-T transition, for "cause" of the most fascinating event I know the CO, buildup, and the global extinctions, to the of in earth history: the Pleistocene/Holocene mam- Decean Traps flood basalt volcanism, seemingly the malian extinctions. largest volcanic event of the Phanerozoic Eon (Mclean, 1985a, 1985b, f985e). ACKNOWLEDGMENTS One area of greatest fascination to me was to attempt to discover potential linkage between I thank Jerry N. McDonald (Department of Ge- changes in ambient air temperature and direct ography, Radford University) and Samuel O. Bird effects upon terrestrial vertebrates that might in- (Virginia Division of Mineral Resources) for invit- fluence bioevolution. Here, I focussed upon the ing me to participate in the 1984 Charlottesville, Pleistocene/Holocene transition as the time interval Virginia, Symposium on the Quaternary of Vir- riehest in information. Several years ago, I pub- ginia, one of the most exciting, and profitable, lished on such linkage (Mclean, 1981);this cunent meetings I have been privileged to attend, and for paper expands upon my original thesis that hot their kindness in allowing me to include this paper environmental conditions directly affect reproduc- in the symposium publication. I also thank Frank tive capacity. Gwazdauskas (Dairy Science, VPI&SU) whose In application of modern principles to the Pleis- counsel and help with the literature on elements tocene/Holocene mammalian extinetions, dwarf- of dairy seience reproductive physiology have been ing, and skeletal abnormalities, I recognize that invaluable to me. the taxa affected reflect vast physiological and morphological diversity, and were spread over vast geographical areas, and variable climatic scenar- ios. Certainly, overall, many factors (including man) were likely involved in the Pleistocene/Hol- REFERENCES ocene scenario. However, two factors come together in a potent mix, the knowledge of whieh may have Abrams, R. M., Caton, D., and Barron, D. H., 1971, value in paleontology. These are the known rapid Temperature differences in reproductive traet Pleistocene/Holocene global warming and the doc- of nonpreg:nant ewe: American Jour. Obstet. umented sensitivity of mammalian reproduction to Gynecol., vol. 110, p.370-375. 118 VIRGINIA DIVISION OF MINERAL RESOURCES

Alliston, C. W., Howarth, 8., Jr., and Ulberg, L. Paleobiology, vol. 7, p. 101-114. C., 1965, Embryonic mortality following cul- Delmas, R. J., Asceneio, J.-M., and Legrand, M., ture in aitro of. one- and two-cell rabbit eggs 1980, Polar ice evidence that atmospheric CO, at elevated temperatures: Jour. Reprod. Fert., 20,000 yr B. P. was 50% of present: Nature, vol. 9, p. 337-341. vol.284, p.155-157. Barron, D. H., 1970, The environment in whieh the Duplessy, J. C., Delibrias, G., Turon, J. L., Pujol, fetus lives: lessons learned since Barcroft, in C., and Duprat, J., 1981, Deglacial warming Prenatal life: Detroit, Wayne State University of the northeastern Atlantic Ocean: correlation Press, p. 190. with the paleoclimatic evolution of the Euro- Bazer, F. W., Ulberg, L. C., and LeMunyan, C. D., pean continent: Palaeog:eogr., Palaeoclimatol., 1969, Altered uterine environment on embryo Palaeoecol., vol. 35, p. L27-144. survival (abs.): Jour. Animal Sci., vol. 28, p. Dutt, R. H., 1963, Critical period for early embryo t44. mortality in ewes exposed to high ambient Bianca, W., 1968, Thermoregulation,'in Hafez, E. temperature: Jour. Animal Sei., vol. 22,p.7L3- S. E., ed., Adaptation of domestic animals: 7L9. Philadelphia, Lea and Febiger, p. 97-118. Gilbert, B. M., and Martin, L. D., 1984, Late Bishop, M. W. H., 1964, Review. A paternal con- Pleistocene fossils of Natural Trap Cave, Wy- tribution to embryonic death: Jour. Reprod. oming, and the climatic model of extinction, Fert., vol. 7, p. 383. in Mafiin, P. S., and Klein, R. G., eds., Qua- Brown, D. 8., Harrison, P. C., Hinds, F. C., Lewis, ternary extinctions: Tucson, Univ. Arizona Jour. M., Wallace, M.H., and Cogburn, L. A., Press, p. 138-147. 1976, Thermal effects on fetal growth and blood Graham, R. W., and Lundelius, E. L., Jr., 1984, flow (abs.): Jour. Animal Sci., vol.42, p. 1340. Coevolutionary disequilibrium and Pleistocene Brown, D. E., Harrison, P. C., and Hinds, F. C., extinetions, inMartin, P. S., and Klein, R. G., 1977, Heat stress on fetal development during eds., Quaternary extinctions: Tucson, Univ. late gestation in the ewe: Jour. Animal Sci., Arizona Press, p. 223-249. vol. 44, p. 442-446. Grayson, D. K., 1984, Explaining Pleistocene ex- Brown, D. 8., and Harrison, P. C., 1981, Central tinctions: Thoughts on the structure of the sympathetic control of uterine blood flow dur- debate, inMarlin, P. S., and Klein, R. G., eds., ing acute heat stress: Jour. Animal Sci., vol. Quaternary Extinetions: Tucson, Univ. Ari- 52, p. LLl4-1121. zona Press, p. 807-823. Burfening, P. J., and Ulberg, L. C., 1968, Embry- Greiss, F. C., 1966, The uterine vaseular bed: onic survival subsequent to culture of rabbit Effects of adrenergic stimulation: Obstet. Gy- spermatozoa at 38 and 40oC: Jour. Reprod. necol., vol.21, p.295. Fert., vol. 15, p.87-92. Guilday, J. 8., 1984, Pleistoeene extinction and Coope, G. R., 1975, Climatic fluctuations in north- environmental change: ease study of the Ap- west Europe since the Last Interglacial, indi- palaehians, 'i,nMartin, P. S., and Klein, R. G., cated by fossil assemblages of Coleoptera,'i,n eds., Quaternary extinctions: Tueson, Univ. Ar- Wright, A. 8., and Moseley, F., eds., Ice ages: izona Press, p. 250-258. Ancient and Modern: Liverpool, Seel House Guthrie, R. D., 1984, Mosaics, allelochemies, and Press, p. 153.168. nutrients: an ecological theory of late Pleisto- L977 a, Fossil coleopteran assemblages cene megafaunal extinctions, in Martin, P. S., as sensitive indicators of climatic changes dur- and Klein, R. G., eds., Quaternary extinctions: ingthe Devensian (Last) cold stage: Phil. Trans. Tueson, Univ. Arizona Press, p. 259-298. Roy. Soc. London (B), vo1.280, p.313-340. Gwazdauskas, F. C., Abrams, R.M., Thateher, W. l9?7b, Quaternary Coleoptera as aids W., Bazer, F., and Caton, D., 1974, Thermal in the interpretation of environmental history, ehanges of the bovine uterus following admin- in Shotton, F. W., ed., British Quaternary istration of estradiol-l?: Jour. Animal Sci., vol. studies: Oxford, Clarendon Press, p. 55-68. 39, p.87-92. -Coope, G. R., and Sands, C. H. S., 1966, Insect Hafez, E. S. 8., 1968, Morphological and anatom- faunas of the last glaciation from the Tame ical adaptations, 'in Hafez, E. S. 8., ed., Adap- Valley: Royal Soc. Proc. B, vol. 165, p.389-412. tation of domestic animals: Philadelphia, Lea Davis, S. J. M., 1981, The effects of temperature and Febiger, p.61-73. change and domestication on the body size of Haynes, C. V., 1984, Stratigraphy and late Pleis- late Pleistocene to Holocene mammals of Israel: toeene extinction in the United States. 'inMar- PUBLICATION 75 119

tin, P. S., and Klein, R. G., eds., Quaternary variations Archean to Present: American Geo- extinctions: Tucson, Univ. Arizona Press, p. phys. Union, Geophys. Mongr. Ser.32, p.493- 345-353. 503. Howarth, 8., Jr., Alliston, C. W., and Ulberg, L. (1985b), Deccan Traps mantle degass- C., 1965, Importanee of uterine environment ing in the terminal Cretaeeous marine extinc- on rabbit sperm prior to fertilization: Jour. tions: Cretaceous Res., vol. 6, p.235-259. Animal Sci., vol. 24, p. L027-1032. (1985c), mantle degassing unification Imbrie, J., and Imbrie, J.2., L980, Modeling the of the trans-Cretaceous-Tertiary geobiological elimatic response to orbital variations: Science, reeord: Evol. Bio., vol. 19, p. 287-313. vol.207, p.943-953. Merilees, D., 1984, Comings and goings of late Kerr, R. A., 1983, An early glacial two-step?: - Quaternary mammals near Perth in extreme Science, vol.22L, p. 143-144. southwestern Australia, in Martin, P. S., and King, J. E., and Saunders, J. J., 1984, Environ- Klein, R. G., eds., Quaternary extinetions: Tuc- mental insularity and the extinction of the son, Univ. Arizona Press, p.629-638. American mastodont, in Maftin, P. S., and Neftel, A., Oeschger, H., Schwander, J., Stauffer, Klein, R. G., eds., Quaternary extinctions: Tuc- B., and Zumbrunn, R., 1982, Ice eore sample son, Univ. Arizona Press, p.315-339. measurements give atmospherie CO, eontent Kurten, 8., 1965, Carnivora of the Palestine caves: during the past 40,000 yr: Nature, vol. 295, p. AclaZool. Fennica, vol. 107, p. L-74. 220-223. L972, The Iee Age: New York, G. P. Oakes, G. K., Walker, A. M., Ehrendranz, R. A., Putnam's Sons. Cefalo, R. C., and Chez, R.A., 1976, Uteropla- Ladner, C., Brinkman, C. R., III, Weston, P., and cental blood flow during hyperthermia with Assali, N. S., 1970, Dynamics of uterine cir- and without respiratory alkalosis: Jour. Appl. - culation in pregnant and nonpregnant sheep: Physiol., vol.41, p. 197-201. American Jour. Physiol., vol. 218,p.257. Ogden, J. G., III, 1967, Radioearbon and pollen Ledue,8.,1972, The effects of hyperventilation on evidence for a sudden change in climate in the maternal placental blood flow in pregnant rab- Great Lakes region approximately 10,000 years bits: Jour. Physiol., vol.225, p. 339-348. ago, in Cushing, E. J., and Wright, H. E., Jr., Marshall, L. G., 1984, Who killed Cock Robin? An eds., Quaternary paleoeeology: New Haven, investigation of the extinction controversy, irc Yale Univ. Press, p. LL7-L27. Martin, P. S., and Klein, R. G., eds., Quaternary Roman-Ponce, H., Caton, D., Barron, D. H., extinctions: Tucson, Univ. Arizona Press, p. Thatcher, W. W., and Wilcox, C. J., 1976, 785-806. Thermal stress effects on uterine blood flow Martin, P. S., 1984, Prehistoric overkill: the global (abs.): Jour. Animal Sci., vol.43, p. 301. model,'i,nMartin, P. S., and Klein, R. G., eds., Roman-Ponce, H., Thatcher, W. W., Buffington, D. Quaternary extinctions: Tucson, Univ. Arizona E., Wilcox, C. J., and Van Horn, H. H., 1977, Press, p.354-403. Physiological and production responses of dairy McDonald, J. N., 1984, The reordered North Amer- cattle to a shade structure in a subtropical iean selection regime and late Quaternary meg- environment: Jour. Dairy Sci., vol. 60, p. 424. afaunal extinctions,inMartin, P. S., and Klein, Roman-Ponce, H., Thatcher, W. W., Caton, D., and R. G., eds., Quaternary extinctions: Tucson, Barron, D. H., 1978, Thermal stress effeets on Univ. Arizona Press, p.404-439. uterine blood flow in dairy eows: Jour. Animal McDowell, R. 8., 1968, Climate versus man and Sci., vol. 46, p. 175-180. his animals: Nature, vol.218, p. 641-645. Ruddiman, W. F., and Melntyre, A., 1981, The Mclean, D. M., 1978, A terminal Mesozoic "green- North Atlantic Ocean during the last degla- house": lessons from the past: Science, vol.201, ciation: Palaeogeogr., Palaeoelimatol., Palaeo- p.40r-406. ecol., vol. 35, p. L45-214. -- 1981, Size factor in the late Pleistoeene Senger, P. L., Lose,8.D., and Ulberg,L.C.,1967, mammalian extinctions: American Jour. Sci., Reduced blood supply to the uterus as a cause vol.281, p.1144-1L52. for early embryonic death in the mouse: Jour. ..-.--...'..-.--._ 1985a, Mantle degassing induced dead Exp. Zool., vol. 165, p.337-344. oeean in the Cretaeeous-Tertiary transition, iz Shaekleton, N.J., and Pisias, N.G., 1985, Atmos- Sundquist, E.T., and Broeeker, W., eds., The pheric carbon dioxide, orbital forcing, and cli- carbon cycle and atmospheric CO, : natural mate, in Sundquist, E. T.,'and Broecker, W., r20 VIRGINIA DIVISION OF MINERAL RESOURCES

eds., The carbon cycle and atmospheric CO, : evolution in North America, in Martin, P. S., natural variations Archean to Present: Amer- and Klein, R. G., eds., Quaternary extinctions: ican Geophys. Union, Geophys. Mongr. Ser.32, Tucson, Univ. Arizona Press, p. 189-210. p.303-317. Webb, T., III, and Bryson, R. A., 1972, Lat,e- and Tchernov, E., 1984, Faunal turnover and extinction postglacial climatic change in the northern rate in the Levant, inMartin, P. S., and Klein, midwest, USA: quantitative estimates derived R. G., eds., Quaternary extinctions: Tucson, from fossil pollen spectra by multivariate sta- Univ. Arizona Press, p.528-552. tistical analysis: Quat. Res. vol.2, p. 70-115. Thatcher, W. W., 1974, Effects of season, climate, Wigley, T. M., Jones, P. D., and Kelly, P. M., 1980, and temperature on reproduction and lactation: Scenario for a warm, high-Co, world: Nature, Jour. Dairy Sci., vol. 57, p. 360-368. vol. 283, p.17-21. Ulberg, L. C., and Burfening, P. J., 196?, Embryo Woody, C. O., and Ulberg, L. C., 1964, Viability death resulting from adverse environment on of one-cell sheep ova as affected by high en- spermatozoa or ova: Jour. Animal Sci., vol. 26, vironmental temperature: Jour. Reprod. Fert., p.571-577. vol.7, p.275-280. Ulberg, L, C., and Sheean, L. A., 1973, Early Wright, H. E., 1977, Quaternary vegetation his- development of mammalian embryos in ele- tory-some comparisons between Europe and vated ambient temperatures: Jour. Reprod. America; Ann. Rev. Earth Planetary Sci., vol. Fert., Suppl. vol. 19, p. 155-161. 5, p. 123-158. Watts, W. A., and Stuiver, M., 1980, Late Wiseonsin Yeates, N. T. M., 1958, Foetal dwarfism in sheep- climate of northern Florida and the origin of an effect of high atmospheric temperature dur- species-rich deciduous forest: Science, vol. 210, ing gestation: Jour. Agric. Sci. (Cambridge), p.325-327. vol.51, p.85. Webb, S. D., 1984, Ten million years of mammal PUBLICATION 75 tzl

PRELIMINARY OBSERVATIONS ON FLOOD- and Potomac rivers as on their tributaries, sug- PLAIN, TERRACE, AND FAN DEPOSITS IN gesting that most sediment eroded from hillslopes THE UPPER POTOMAC RIVER BASIN OF either is stored on the floodplains of the tribu- VIRGINIA AND WEST VIRGINIA. J. Steven taries, or is transported by the larger streams KITE, Alison BELL, and Mark ALLAMONG, directly to the lower Potomac. Department of Geology and Geography, West Vir- A scarcity of cut-bank exposures along the ginia University, Morgantown, WV 26506 Shenandoah and Potomac rivers hampers study of Holocene alluvial history and associated floodplain stratigraphy along these major streams; Three independent research projects concern- proper study of the floodplain will require ar- ing the late Cenozoic stratigraphy and history of tificially created exposures such as backhoe the upper Potomac River basin were begun dur- trenches or archeological excavations, such as ing 1984. Laboratory analyses and compilation those carried out at the Flint Run (Thunderbird) of field data have barely begun, so it is not yet site south of our study area (Segovia and Franco, possible to present detailed discussions of each 1977; Foss, L977). No distinct Holocene deposi- project. However, our work to date warrants a tional units have been delineated in the present preliminary discussion of the alluvial and collu- study. In general, modern Potomac River channel vial deposits in this part of the central Appala- gravel is composed of sandstone and chert, with chian Mountains. minor carbonate and siltstone clasts, whereas One project was a reconnaissance of all types Shenandoah River gravel is composed of quart- of alluvial deposits in the lower Shenandoah zite and sandstone, with minor conglomerate, Valley, including deposits of the Potomac River, greenstone, and carbonate clasts. In the vicinity in Berkeley and Jefferson counties, West Virginia of alluvial terraces, up to one-third of modern and Clarke County, Virginia. In this study area channel clasts show extensive weathering rinds, (Figure), widespread post-settlement alluvium indicating reworking of older alluvium. The over- with little soil profile development commonly bank facies on modern floodplains is predomi- overlies well-developed, buried, pre-settlement nantly sandy loam or silty loam, locally broken soil profiles. The post-settlement alluvium typ- by gravelly splay deposits. Various soil series ically is 1.6 to 6.6 feet (0.5 to 2 m) thick, and occur on the modern floodplains. contains historical artifacts. At one locality the At least two pedologically distinct terrace units buried pre-settlement soil contains late Woodland occur in the lower Shenandoah Valley. Terraces Period pottery. In the study area, post-settlement 16 to 65 feet (5 to 20 m) above modern floodplains alluvium is not as thick along the Shenandoah are characterized by Captina and Ashton soil series. Alluvial gravel in these terraces generally have fewer carbonate and greenstone clasts, but more quartzite and sandstone clasts, than moderri floodplain gravel. High terraces (more than 66 feet,20 m, above modern floodplains) are characterized by Monon- gahela, Braddock, Waynesboro, Thurmont, and Tygart soils; however, at least two areas of Brad- dock soil near Harpers Ferry are formed on local w/ti: colluvium. A few isolated terrace surfaces occur more than 213 feet (65 m) above modern floodplains. These very high terraces are not as well developed in the lower Shenandoah Valley as along the South Fork of the Shenandoah, in Warren County, Virginia (Rader and others, W 1975). A few exposures of high terrace deposits \,wm i\ show a typical sand-over-gravel floodplain se- rrlu{sourn IonK!/ quence, thoroughly modified by pedogenic clay. Most high terrace deposits lack upper sand layers; it is likely the sand was removed by rill or sheet erosion. Other exposures suggest colluvial Figure. Study areas. mixing of sand, gravel, and pedogenic clay. High 122 VIRGINIA DIVISION OF MINERAL RESOURCES terrace gravel is mostly quartzite, sandstone, and creating conditions suitable for long-term pres- conglomerate. ervation of alluvial deposits. Where earbonate The style of terrace destruction is strongly bedrock oecurs, it weathers beneath the alluvium influenced by the underlying bedrock lithology. undergoing pedogenesis, producing a bisequel Terraces formed on shale tend to persist until soil profile. The profile is essentially a Frederick destroyed by lateral migration of the major or Lodi clay residual sub-soil capped by a Mo- stream channel. The alluvial veneers may be nongahela soil developed in alluvium. Other soil almost entirely eroded away, but the underlying series formed from older alluvial parent material shale straths appear resistant to slope processes include the Unison and Shenval pebbly and eob- and first-order stream erosion. High terraces bly loams. formed on carbonate bedrock generally show The third study contributing to this paper is sinkholes and colluvial reworking of terrace sed- a detailed examination of the geomorphology of iments suggesting that old alluvium commonly New Creek Valley in Mineral and Grant eounties, is let down by karst solution. Although terraces West Virginia, a valley containing landforms and formed on shale survive as distinct landforms deposits typical of smaller valleys of the upper longer than their counterparts on carbonate bed- Potomac basin. For most of its length, New Creek rock, it is likely that alluvium reworked by karst flows over shale, which crops out between ridges solution may be considerably older. underlain by resistant sandstone and orthoquart- In the second project, a detailed study along zite. Bedload in New Creek consists of siltstone, the South Fork of the Shenandoah River in Rock- sandstone, and orthoqtrartzite; although shale ingham County, Virginia, field studies show two bedrock underlies the stream bed, no shale clasts types of alluvium: one composed predominantly were found in the channel. Soils found on the of channel gravel and sand, the other composed modern floodplain include the Pope, Chagrin, mostly of overbank sand and silt. Most alluvium Huntington, and Philo series. Drainage anoma- between the Blue Ridge and the modern river lies and wet, clayey depressions in wide areas of channel appears to be old bedload deposits largely the floodplain indicate old meander cut-offs. derived from mountain tributaries. The gravel Several fans occur between the steep slopes of is well exposed and exhibits varied intensities of New Creek Mountain and the stream. These fans pedogenesis, especially clay aecumulation in the are truncated by the stream, and have moderately B horizon. One exposure revealed a deeply wea- developed soils. Fan deposits adjacent to moder- thered gravel with many thick weathering rinds ate slopes on the opposite side of the stream have on sandstone cobbles, overlying more than 3 feet better developed soil profiles. These fans with (1 m) of fine sand and silt, apparently overbank well-developed profiles occur in areas shown as deposits. The soil series associated with these residual soils on soil survey maps. coarser terrace sediments is the Cotacto variant The New Creek Valley contains two individual cobbly loam. terraces 23 to 33 f.eet (7 to 10 m) above modern Higher benches exist on foot slopes of the Blue stream level. These terraces contain alluvium Ridge and yield sandy soils which may be rem- similar to modern channel deposits but altered nant terrace material that has undergone exten- by the pedogenic processes that have formed siye erosion. Alternatively, they may be of allu- Monongahela, Tygart, and Purdy soils. vial-fan origin. Colluvial landforms are found Higher terraces in New Creek Valley are eom- adjacent to these benches and exhibit typieal posed primarily of large rounded sandstone colluvium soil profiles, such as the Laidig soil boulders. One location approximately 131 feet (40 series which displays a fragipan. m) above modern stream level contains hematite- In contrast to the coarse deposits derived from cemented eonglomerate and breccia composed of the Blue Ridge, the valley bottom northwest of rounded sandstone cobbles and pebbles, and an- the river has mostly sand and silt alluvium. Older gular shale and siltstone clasts. Adjacent to the terraces are well preserved when adjacent to Potomac River, at its confluence with New Creek, tributaries which supply resistant sandstone and there are shale straths mantled by a thin veneer quarlzite to the valley floor, produeing a coarser of gravelly alluvium at Lg? to 213 feet (60 to 65 alluvial deposit. Fragipan-bearing Monongahela m) above modern stream level (Dennison, 1963). cobbly and sandy loams are typieal terrace soils Our three independent projects allow a few in this area. generalizations about terraces in the upper Po- Most of the study area is underlain by saprolite tomac River basin. Terrace deposits are best developed from Paleozoic earbonate and shale, developed wherever tributaries that carried an PUBLICATION 75 r23

abundant supply of resistant sandstone or quart- Two distinet types of fans have developed dur- zite clasts emptied into the major river system. ing the Quaternary Period on opposite sides of Although terraee deposits differ from modern the Blue Ridge Mountains. Active fans east of floodplain deposits in several aspects, these dif- the Blue Ridge are construeted by infrequent ferences can be attributed to post-depositional debris avalanehing eaused by intense rainfall. pedogenesis and erosion. Soil survey maps are Inactive fans along the western slopes of the Blue useful in identifying the extent of probable ter- Ridge appear to have been constructed by braided race deposits, but these soils maps must be field stream processes. Rainfall in excess of 23.6 inches checked before they are of use in the compilation (60 cm) in 8 hours during Hurricane Camille in of aceurate surficial geology maps. 1969 triggered hundreds of debris avalanches east of the Blue Ridge in Nelson County (Williams REFERENCES and Guy, 1973). Sediments have accumulated on small debris fans by repeated debris avalanching Dennison, J. M., 1963, Geology of the Keyser and debris flows duiing the Holocene. The debris quadrangle, West Virginia Geol. Survey Geol. fans are elongate beeause of their confinement Map 1. by steep valley walls in Lovingston granite-gneiss Foss, J. 8., L977, The pedological reeord at sev- bedrock. The loeus of deposition on these fans eral Paleoindian sites in the Northeast, in appears to be controlled by the distribution and Newman, W. S., and Salwin, B., eds., Ame- magnitude of upslope failures. Aerial photo- rinds and their paleoenvironments in north- graphs taken after the Camille storm show that eastern North America: Annals New York an average of 25 percent of fan surfaces were Acad. Sci., vol.288, p. 234-244. activated by debris flows. Nelson County debris Rader, E. K., Biggs, T. H., and Nunn, W. E., 1975, fan sediments are very poorly sorted cobbles and Geologic map of the east portion of the Front boulders in a mud matrix and are generally less Royal quadrangle, Virginia: Virginia Divi- than 66 feet (20 m) thick. These mud-matrix sion Mineral Resources Rept. of Inv. 40, plate gravels are thick-bedded and rarely exhibit in- 1. ternal stratification, except for occasional crude Segovia, A. V., and Franco, L., L977, Arehaeo- bedding and basal imbrication. Inverse grading logical implications of the geomorphic evo- is common because of high dispersive stresses lution of the lower Shenandoah Valley, Vir- during emplacement of the sediments by debris ginia, im Newman, W. S., and Salwin, B., eds., flow meehanisms. No significant down-fan facies Amerinds and their paleoenvironments in variations are visible in the debris fan sediments. northeastern North America: Annals New Lateral variations occur because of the lobate York Acad. Sci., vol. 288, p. 189-193. nature of debris deposition on irregular fan surfaces. The 1969 deposits contain the coarsest clasts in all of the stratigraphic sections observed. Pre- QUATERNARY ALLUVIAL FANS IN CEN- 1969 deposits have higher clay content, greater TRAL VIRGINIA. R. Craig KOCHEL and David oxidation, and higher quartz/feldspar ratios in W. SIMMONS, Departmentof Geology, Southern the matrix. Depositional units are separated by Illinois University, Carbondale, IL 62901. erosional contacts. Many of the bed surfaces contain paleosols formed on fan paleosurfaces during Most sedimentological models of modern and the time between successive debris flows. These ancient alluvial fans have focused on fans formed buried soils can be radiocarbon dated. Dating of in arid, humid-glacial, humid-tropical, or humid- buried soils in the Davis Creek debris fans in- periglacial environments. Little attention has dicates that depositional episodes are separated been given to alluvial fans developed in humid- by periods of inactivity ranging from 3,000 to temperate areas sueh as Virginia (Kochel and 4,000 years. Correlation of fan stratigraphies in Johnson, f984). Signifieant differenees exist in Davis Creek show that deposition has occurred morphology, depositional processes, and facies at least three times during the Holocene, most between the Virginia fans and fans described in reeently in 1969 (Kochel and Johnson, 1984). other climates. Coarse gravel fans are the dom- Depositional units can be reeognized and corre- inant depositional landform along piedmont lated based on the identification of buried soil slopes of the Blue Ridge Mountains in west- horizons, bedding character, clast weathering, central Virginia. and changes in matrix mineralogy with depth. t24 VIRGINIA DIVISION OF MINERAL RESOURCES

Table. Comparison of Blue Ridge fans and debris fans. MORPHOLOGY BLUE RIDGE FANS DEBRIS FANS SIZE 10's km2 l-2km2 SHAPE fan, bajada elongate, isolated GRADIENT moderte, concave less than 5o steep, segmented SOILS thick surface soil, very old paleosol more than 5o no surface soils paleosols SEDIMENTOLOGY , many TEXTURE high Gravel+Sand/Mud low Gravel+Sand/Mud MATRIX sand mud FABRIC imbrieated none SORTING poor (locally good) very poor ROUNDNESS subrounded angular BEDDING alt. sand and gravel crudely developed SEDIMENTARY STRUCTURES gravel forsets, channel fills, bars, occasional inverse grading RETURN INTERVAL more continuous discrete (3,000 yrs+) DOWN_FAN FACIES fining down-fan no down-fan trends facies more continuous laterally (sheets) distinct units lobate, discontinuous in plan DOMINANT PROCESSES braided stream debris avalanche flood surge debris flow sheet flood AGE Pleistocene Holocene Discrete sedimentation units within the fans are imentary structures (Figure). Proximal areas of eharacterized by abrupt changes in the sediment individual fans and/or bajadas contain coarse, texture and ratios of quartzlf.eldspar in the ma- poorly sorted, imbricated boulders deposited by trix. Decomposition of granitic clasts with time powerful, competent surges. The mid-fan facies yields inereasing quaftzlf.eldspar ratios as the is usually the thickest part of the fan and is feldspars alter to clay minerals. Bulk texture characterized by imbricated, eross-bedded tends to decrease with age as clasts disintegrate. gravel and sand. In many areas, the gravel ap- Typically, these deeompositional trends are grad- pears to be organized into distinct longitudinal ual with depth in sediments. However, if weath- bars while the sand beds appear to have been ering is interrupted by catastrophic deposition deposited by laterally extensive sheet floods. The (i.e. by debris flows) these trends will display Blue Ridge fans grade distally into terrace grav- abrupt changes with depth. These qualitative els of the South River and Shenandoah River. indicators of age and depositional aetivity rein- Distal fan sediments are dominated by cross- force the correlations based on radiocarbon dat- bedded sand and cross-bedded pebble-to cobble- ing. sized gravel. Scattered deposits of mud occur in Alluvial fans form an extensive bajada along the distal facies and in inter-fan regions where the western flank of the Blue Ridge that has ponding was likely to have occurred. prograded westward over Cambrian carbonates Manganese ore has been commercially mined in the Shenandoah Valley. These Blue Ridge fans from several of the Blue Ridge fans during the are best developed between Grottoes and Vesu- past century (Hack, 1965). The manganese is vius along the southeastern margin of the She- thought to have developed in the residual soils nandoah Valley. No historical depositional activ- formed on the underlying carbonate rocks whieh ity has been observed on these fans. Fan were buried by prograding fan gravel. Most of sediments are sand matrix, clast supported grav- the deposits are located in the proximal fan facies, els composed primarily of Antietam quartzite. in areas where maximal reworking and placer The Blue Ridge fan gravels are well-imbricated, development would have been expeeted. There- contain abundant channel fill sequences, and fore, some of the manganese may have resulted exhibit large-scale planar cross-stratification in from stratiform placer deposition. Gravel, sand, many areas. Locally, gravels are well sorted, but and clay are being mined from midfan, distal, overall sorting is poor. These gravels appear to and inter-fan facies, respectively. have been deposited by powerful braided streams Age relationships for the Blue Ridge fans are with insignificant debris flow activity (Table). somewhat uneertain. No radiocarbon dating has Blue Ridge fans are thick relative to the debris been done on these sediments. Relative ages ean fans to the east and typically range in thickness be determined based on soil geomorphology and between 65 to 492 feet (20 to 150 m). Distinct weathering charaeteristics of the fan sediments. down-fan variations occur in grain-size and sed- Surface soils are well developed under a fairly {th PUBLICATION 75 r25

BLUE RIDGE FANS.BAJADA paucity of paleosols indicates that significant periods of inaetivity, on the order of hundreds to thousands of years, did not occur. A similar interpretation of continuity of depositional activity can be made for the lower gravel below the paleosol. Therefore, we suggest that there have been at least two episodes of fan construction during the Quaternary, separated by a long period of inactivity whieh is marked by the prominent paleosol. In current investigations we are continuing to z refine the facies model for humid-temperate al- ff luvial fans and attempting to determine the cause I c) for marked variation of alluvial fans on opposite sides of the Blue Ridge. {e hope to determine = more precisely the ages of the Blue Ridge fan sediments and to determine how Quaternary climatic changes may have influenced depositional processes on the fans. We speeulate that Pleis- tocene climates may have been more favorable Figure.,Schematic facies model for alluvial fans than Holocene ones for braided stream processes. along the western flank of the Blue Ridge. Distal The retreat of the polar front and renewed in- fining is consistent with the interpretation of trusion of tropical air masses appears to be coin- tractive processes (i.e. braided stream) dominat- cident with, and may be recorded by, the initi- ing over suspended processes (i.e. debris flow). ation of Holocene debris flow activity on the fans east of the Blue Ridge about 11,000 years ago. At several sites east of the Blue Ridge, Holocene dense forest eover. Development of the thick debris fans are entrenched into Pleistocene fan surfaee spodosol probably required thousands of sediments. These Pleistocene sediments are sand years under the present humid-temperate cli- matrix gravel that appear similar to the Blue mate. Its presence indicates that there has not Ridge fan gravel and may record a similar shift been extensive deposition on these fans through- in dominant depositional processes from braided out the late Holocene and perhaps the entire stream to infrequent debris flow episodes. If the Holocene. Several meters below the surface of the sedimentological changes through time on these Blue Ridge fans there is a well developed paleosol. fans are inexplicable by climatic variations, then This paleosol ean be observed in all the sites other causative processes such as neotectonism investigated and is a useful time-stratigraphic will have to be explored. marker. The paleosol separates upper gravels having a tan, sandy matrix and unweathered quartzite clasts from lower gravels with a red REFERENCES matrix and rotted quartzite clasts. To form such an extensive paleosol and to permit such decom- Hack, J. T., 1965, Geomorphology of the Shenan- position of resistant quartzite clasts require ei- doah Valley, Virginia and West Virginia, and ther a very long time (on the order of tens of origin of the residual ore deposits: U. S. Geol. thousands to millions of years) under present Survey Prof. Paper 484, 83 p. climatic eonditions and/or a warmer, more humid Kochel, R. C., and Johnson, R. A., 1984, Geomor- paleoclimate under which conditions the soil phology and sedimentology of humid-temper- would have developed more quickly. The lack of ate alluvial fans, central Virginia, im Koster, paleosols and significant vertical differences in E. H., and Steel, R.J., eds., Sedimentology sediment induration and clast weathering of gravels and eonglomerates: Canadian Soc. throughout the upper gravel indicate that they Petrol. Geol. Mem. 10, p 109-122. were deposited rapidly by high frequency flows. Williams, G. P., and Guy, H. P., 1973, Erosional The frequency of deposition on these fans may and depositional aspects of Hurricane Ca- have been annual, i.e., a seasonal climate, or as mille in Virginia, 1969: U. S. Geol. Survey infrequently as every few tens of years. The Prof. Paper 804, 80 p. L26 VIRGINIA DIVISION OF MINERAL RESOURCES

QUATERNARY TRAVERTINE-MABL DE- WISCONSINAN HERPETOFAUNAS OF THE POSITS OF VIRGINIA. David A. HUBBARD, Jr., CENTRAL APPALACHIANS. Leslie P. FAY; and William F. GIANNINI, Virginia Division of The Museum, Michigan State University, East Mineral Resourees, P. O. Box 3667, Charlottesville, Lansing, MI 48824-10451 VA 22903; and Michelle M. LORAH; Department of Environmental Science, University of Virginia, Fossil herptiles of the central Appalachian Charlottesville, VA 22903. Mountains do not exhibit the striking Wiscon- sinan-Holocene range adjustments characteristic of some contemporaneous species of birds and Valley and Ridge travertine-marl deposits have mammals. Composite ranges (based on present been observed in ten Virginia counties (Figure). distributions) of the fossil herptiles are harmon- Occurrences are typically located downstream ious (all speeies sympatric) and oecur in the from faulted and fractured carbonate rocks. De- central or southern Appalachian region. Most posits are the result of inorganic and organic fossil mammal composite ranges are disharmon- precipitation of low magnesium calcite from ious'(not all species sympatric at present) with mixed surface and subsurface waters. CaCOr is 83 percent to 92 percent of the species co-occur- actively deposited seasonally, but stream erosion ring in the central or northern Appalachians. The is currently outpacing deposition over the yearly remaining 8 percent to 17 percent (extirpated cycle at each site. Algal cysts eollected near the from the region some time in the Holocene) co- base of one deposit may indicate Pleistocene dep- occur 93 miles (150 km) north of Edmonton, osition. Angiosperm leaf imprints at numerous Alberta, over 1860 miles (3000 km) from the sites are all of modern tree species. A date be- Appalachian fossil localities in an area defined tween 750 and 1000 A.D. was suggested for by the southern limit of Microtus nanthognathus projectile point and pottery fragments found in and the northwestern limit of Spermophilustri'de- the top of a Frederick County deposit. Deposits cemlineatus. This small area of boreal habitat in have been worked for agrieultural lime in four western Canada is also occupied by many wide- counties with operations continuing at one loea- ranging speeies still present in the Appalachians tion. Soil types and topographic signatures show (56 percent to 84 pereent of the fossil faunas, promise for rapid reeognition of these features. boreal extirpates ineluded; Guilday and others, Investigations are continuing into the location 1978, and referenees therein). Thus, the majority of additional travertine-marl deposits. Detailed of mammals and all herptiles present in the studies are being earried out at two sites by the eentral Appalachians have been in residence authors and others. since at least the Mid-Wiseonsinan (Fay, 1984a), but a few mammals have undergone dramatic range shifts since the Late Wisconsinan. The fossil record for birds is less eompletely examined but includes current residents plus a minority of species now removed to the boreal forest or arctic tundra. Why are 14 bird and mammal . TMVERTINE-I{ARL DEPOSIT species of boreal to tundra habitat preference X COUNTY I.IITH ABANDONED }IORKINGS found as fossils in central Appalachian accu- X COUNTY l.lITH ACTIVE }IORKINGS mulations along with representatives of the current temperate fauna? Why were herptiles apparently not affected by the forees that caused other tetrapods to shift their ranges to the north and west from 186 miles (300 km) to 1984-miles (3200 km) at the close of the latest glaciation? Causes responsible for these varying responses

Figure. Distribution of travertine-marl deposits in Virginia. Counties containing sites worked for I Present address: P.O. Box 141, Winnebago, IL agricultural lime are indicated. 61088 PUBLICATION 75 L27

among and between vertebrate classes have not equable" climatic regimes more swiftly than yet been adequately identified. mammals, if indeed herptile ranges beyond ice Several common miseoneeptions about herptile limits were much affected by glacial-interglacial ecology block further understanding of this zoo- eycles. geographic disparity. Disharmonious mammal- Only one Late Wisconsinan Appalachian eq- ian fossils are usually accepted as the result of uable herpetofauna-Ladds, Georgia-has been synchronous deposition, but herptiles from the described. This locality supports the equability same localities that do not match the ecologic and hypothesis with two "northern" and two "south- distributional characteristics of the mammals ern" disharmonious species in an otherwise sym- have often been suspected of being intrusive patrie herpetofauna characteristic of northern (Lundelius and others, 1983). Important Wiscon- Georgia today (Holman, L976, 1980, 1985a, sinan Appalaehian localities [e.g. New Paris No. 1e85b). 4, Pennsylvania (Guilday and others, 1964) and Several possible explanations for the lack of Ladds Quarry, Georgia (Holman, 1967, 1985)l dramatic range changes in Wisconsinan herpe- were judged to be heterochronic assemblages and tofaunas may be advaneed based on evolutionary the significance of the herptile components left stability (few late Quaternary extinctions or origin question. The distributional incongruencies inations) as well as apparent zoogeographic and more likely result from real ecologic or climatic eeologic stability (Fay, 1984b). (1) Herptiles can factors, as no evidence for heterochronic depo- adapt more rapidly to climatic change than ean sition has been demonstrated in most localities. mammals. Reptiles and amphibians have a The often repeated assertion that herptiles marked eapacity for acclimatization to seasonal were more severely affected by glacial-intergla- climatic ehange, which may allow for physiologic cial climatic shifts than were other animals (Auf- rather than range adjustments over glacial-in- fenberg and Milstead, 1965) is no longer sup- terglacial intervals. (2) There are fewer eastern ported. There is no documentation of disloeation North American "boreal" herptiles than mam- of northern or southern forms into the central mals available to be displaced southward into the Appalachians, a maximum 110 mile (f77 km) central Appalachians by glaciation. More than northward migration of "southern" forms, and twice as many mammal (14) as herptile (4) species evidently no resident species were forced to make are restrieted to the area north of the latitude signifieant range adjustments (Fay, 1984b). of the New York-Pennsylvania border and east The assumption that poleward distribution lim- of Lake Winnipeg. (3) As herptiles can be more its of herptiles are governed by cold temperatures difficult to identify to species-level than mam- is an over simplification that may obseure a great mals, climatieally induced range changes may go deal of information useful in paleoenvironmental unreeognized. Of the taxa just mentioned, all 14 reconstruction. Severity of the cold season is less species of mammals are known as fossils, but only critical (assuming suitable hibernacula) than du- one of the four herptiles has been identified as ration of the warm season. All boreal North fossils. (4) Clinal morphological changes in space American amphibians and some reptiles are ovi- and time are difficult to quantify for herptiles parous, limited by warm seasons of sufficient because of the types of bones used in their iden- duration and intensity for growth and develop- tification and because herptiles are charaeterized ment of eggs. Adults can avoid unfavorable con- by indeterminate growth. The record of Berg- ditions during the remainder of the year by mann's Response and its climatic importanee for hibernation. Appalachian mammals (Guilday and others, If the climatic equability hypothesis is aceurate 1978) is unknown and perhaps irrelevant for in predicting reduced seasonal extremes during herptiles. (5) Much of the herpetofauna of the glacial stages with regard to interglacials, the Appalachian region is endemie or has a wide- cooler summers may not allow sufficient time for spread distribution centered on this region, re- incubation of eggs at higher latitudes (as is the fleeting both probable eenter of origin and res- case for turtles in Great Britain now and during tricted microhabitat requirements for many of previous glacial intervals: Stuart, 1979). Yet an the species. The region may have served as a essentially modern herpetofauna existed at least glacial refugium for herptiles as it was postulated as far north as New Paris in southern Pennsyl- to have been for the eastern deciduous forest vania at the end of the latest glaciation (Guilday (Braun,1950). and others, 1964). Apparently, herptiles are able Although the ranges of other animals and to respond to the shift from "equable" to "non- plants changed in response to the glaeial-inter- t28 VIRGINIA DIVISION OF MINERAL RESOURCES glacial climatic shifts, herptiles of the central 1985b, New evidence on the status of Ladds Appalachian Mountains were apparently unaf- Quarry: Natl. Geog. Res., vol. 1, no.4, p. 569- feeted by physical and biotic alterations over the -, 570. past 30,000 years. Lundelius, E. L. and others, 1983, Terrestrial vertebrate faunas, inPorter, S. C., ed., Late Quaternary environments of the United REFERENCES States, Vol. 1, the late Pleistocene: Minnea- polis, Univ. Minnesota Press, p. 311-353. Auffenberg, W., and Milstead, W. W., 1965, Rep- Stuart, A. J.,1979, Pleistocene occurrences of the tiles in the Quaternary of North Ameriea,,in European pond tortoise (Emys orbicularisL.) Wright, H. E., and Frey, D. G., eds., The in Britain: Boreas, vol.8, no.2,p.359-371. Quaternary of the United States: Princeton, Princeton University Press, p. 557-568. Braun, E. L., 1950, Deciduous forests of eastern North America: New York, Hafner Press, LATE PLEISTOCENE FAUNA OF THE HOL. (L974 reprint), 596 p. LIDAYSBURG FISSURE. Shirley S. FONDA Fay, L. P., 1984a, Mid-Wisconsinan and mid- and Andrew P. CZEBIENIAK, Department of Holocene herpetofaunas of eastern North Geosciences, Pennsylvania State University, Uni- America: a study in minimal contrast, ira versity Park, PA 16803 Genoways, H. H., and Dawson, M. R., eds., Contributions in Quaternary vertebrate pa- Excavation of a large vertical fissure 1.2 miles leontology: A volume in memorial to John E. (1.9 km) northeast of Hollidaysburg, Blair Guilday: Carnegie Mus. Nat. Hist. Spec. Publ. County, Pennsylvania, has produced numerous no. 8., p. 14-19. vertebrate and invertebrate fossils of late Pleis- 1984b, Wisconsinan Appalachian herpeto- tocene age. The fissure is exposed in the south- faunas: what about distribution changes, eastern wall of an abandoned quarry within the -, heterochroneity, and seasonal climatic re- Upper Silurian Keyser Limestone, and is lined sponse? (abs.): American Quat. Assoc., 8th with thick drapes of travertine. Descending at Bien. Mtg., Program and Abstraets, p. 42. least 40 feet (13 m), the fissure is 6 feet (2 m) Guilday, J. E., Martin, P. S., and McCrady, A. wide and 18 feet (6.m) long; it was completely D., 1964, New Paris No.4: A Pleistocene cave filled with rubble from cave breakdown and the deposit in Bedford County, Pennsylvania: weathering of surface material. The upper third Natl. Speleol. Soc. Bull., vol.26, no.4, p. 121- of the fissure was filled with limestone blocks. t94. The middle section, containing many blocks em- Guilday, J. 8., Hamilton, H. W., Anderson, E., bedded in red-brown clay, yielded fragments of and Parmalee, P. W., 1978, The Baker Bluff a young mastodon (dated by amino-acid racemi- Cave deposit, Tennessee, and the late Pleis- zation at about 12,000 yr B.P.) and numerous tocene faunal gradient: Carnegie Mus. Nat. smaller animals. The bottom portion was filled Hist. Bull. 11, 67 p. with horizontally-bedded, varicolored elays, cave Holman, .I. A., 1967, A Pleistocene herpetofauna rubble, and well-weathered limestone blocks; sev- from Ladds, Georgia: Georgia Acad. Sci. Bull. eral large pockets of fossils and many isolated 25,no.3, p.154-166. remains were scattered throughout. An adjacent 197 6, Paleoclimatie implieations of "ecolog- small cave eontinuous with the large fissure con- ically incompatible" herpetological species tained a wide variety of well-preserved bones, -, (late Pleistocene: southeastern United including numerous skulls and mandibles, some States): Herpetologica, vol. 32, no. 3, p. 290- of which were still articulated. Also, a number 295. of pockets within the travertine drapes contained 1980, Paleoclimatic implications of Pleis- abundant small vertebrates. toeene herpetofaunas of eastern and central Forty-eight species of mammals were reeov- -, North America: Trans. Nebraska Acad. Sci.. ered, along with unidentified fish, frog/toad, sal- vol.8; p. 131-140. amander, snake, turtle, and bird bones. Inver- , 1985a, Herpetofauna of the Ladds Quarry tebrates include at least three species of land indicates climatic equability in the Late Pleis- snails and some insect exoskeletons. Sediment tocene of Georgia: Natl. Geog. Res., vol. 1, no. was collected at 2-foot intervals for pollen anal- 3, p.423-436. ysis in the future. PUBLICATION 75 L29

The smaller mammals are represented by most Woodland and Arehaic periods. Of the eight of their skeletal parts, whereas the large mam- species of vertebrates present at the site, three mals, mostly juveniles, are represented in some (Mammut americarru,m, Platagonus conxpressu,s, cases only by a few teeth. Juveniles might have Equus cf.. E. fraternus, and possibly Odocoileus been prone to fall into a fissure through either uirg'inianus) are components of the late Pleisto- carelessness or their size, whereas the few teeth cene fauna, while one speeies (Bison bison) is might have been carried in by small animals such known from the late Holoeene. Three species of as pack rats. Other animals probably inhabited domestic mammals (Bos taurus, Felis d,omesticus, the fissure as is suggested by the oceurrence of and domestic? sheep/goat) are also represented several 0.3-0.5 foot (0.1-0.2 m) layers of snake at the site. The investigation of Hot Run was remains and a few pockets of frog bones. These initiated in L977 following the diseovery of the snake bones were commonly articulated, and the site by Mr. Jonas Chamberlain and James Ott. sediment here contained up to 75 percent skeletal The overall research objective is a paleoecological remains. The occurrence of many gnawed bones reeonstruction of the Hot Run Site during the late suggests that small rodents also inhabited the Pleistocene and early Holocene. fissure. The Hot Run Site is located on a 13 acre tract Of the 48 species of mammals, six are now of land owned by the V. V. Cutshaw family, 1000 extinct (Ceraalces sp., Mylohyus nasutus, Tapirus feet (308 m) north of Stephenson, Virginia aeroensis, Mammut americanum, Castoroid,es (39"L4'22" North; 78o6'20" West); it is in the ohioensis, and Canis dirus). Although some (13 Stephenson, Va. Quadrangle, USGS 7.5' series. percent) of the species suggest a boreal habitat, The site is at an elevation of 590 feet (182 m), the majority indicate a temperate woodland- and lies within a shallow depression at the con- meadow environment, somewhat cooler than the fluence of four streams. The site lies east of U. present. The more northerly speeies eould rep- S. Highway 11, south of State Highway 749, and resent relict populations. west of the Baltimore and Ohio Railroad tracks. The mammalian fauna (48 species) is repres- Closure of the 590 foot (182 m) contour line at ented by at least 7866 individuals, distributed the railroad bridge delimits the fossil-bearing (number of species, percent of total individuals) area. Thus defined, the site extends 590 feet (182 as follows: insectivores (6 spp., 23%), bats (2 spp., m) north-south and 900 feet (277 m\ east-west. 0.6%), carnivores (11spp., 2.2%), rodents (20 spp., The Hot Run Site lies upon three Ordovician 68%), lagomorphs (2 spp., 6%), perissodactyls (2 limestone formations: the Lincolnshire, the New spp., 0.03%), artiodactyls (4 spp., 0.17%), and Market, and the Oranda/Edinburg. proboscideans (1 sp., 0.01%). The greatest number The presence of fossil material at Hot Run was of species and individuals were found in the back brought to light through the construction of a portion of the fissure between the 21- and 25-foot pond which now forms the central feature of the (6.5-7.7 m) levels, in the front portion between site. The pond and its immediate environment the 31-and 33-foot (9.5-10.2 m) levels, and in the have a complex recent history involving four adjacent small cave. major modifications between 1963 and 1984. Each modifieation has resulted in considerable change to the physiognomy of the site and each has simultaneously unearthed and buried large THE HOT RUN SITE: A NEW PLEISTOCENE quantities of fossil-bearing sediments. VERTEBRATE LOCALITY IN NORTHERN The original excavation of the pond took place VIRGINIA.James R. OTT and RobertE. WEEMS, circa 1963 and produced a eollecting basin of Department of Zoology, University of Maryland, three acres. The pond contained three islands. In College Park, MD 20742 and U. S. Geological August of 1977 the pond was partially drained Survey, Reston, V A 22092 and dredged and in November of 1980 a large number of the spoil piles created from the 1977 The Hot Run Site is located in Frederick excavation were leveled. The dam was refur- County, northwest Virginia, on the floor of the bished and the pond filled in the spring of.1984. Shenandoah Valley. The site contains a diffuse, This final modification flooded the major fossil- unstratified deposit composed primarily of sand bearing areas of the site. and clay containing a mixture of late Pleistocene Between August 7 and August 24, L977 abun- and Holocene vertebrate material. Uncommon dant and highly visible fossil material was col- elements include lithic artifacts assigned to the lected from the surface of the pond's margins. 130 VIRGINIA DIVISION OF MINERAL RESOURCES

Following the dredging of the pond on August imal metatarsal fragment, and one left lower 24,1977, material was collected both by searching canine tooth, and Equus ef.. E. fraternus, known the surface of the spoil piles and by raking lightly from one complete lower third premolar. Od,o- the surface of the spoil piles and searching the coileusuirginianus is known from three proximal disturbed areas. Surface collecting on the largest antler fragments, two of which possess the basal of the three islands was initiated on August 7, surface which unites with the pedicel. The geo- 1977 and continued until May 10, 1980 when a logical ages of these fragments are presently grid system was established on the island and unknown. a controlled subsurface excavation was initiated. Late Holocene species represented at the site Twenty quadrates were demarcated and nine include Bison bison, known from one left astra- were excavated to a depth of 14 inches (35 cm) galus and one phalanx, and three domestic spe- in the following manner. Quadrates were cleared cies: Felis domesticus, known from one radius; to a depth of 4 inches (10 cm). This material was Bos taurus, known from one metatarsal; and then removed to the periphery of the island and domestic ?goat/sheep, known from one tibia. washed. The underlying area thus exposed was Lithic materials recovered include the distal turned over to a depth of 8 inches (20 cm), raked portion of an early Archaic knife 2.6 inches (6.5 lightly and washed. This step was repeated with cm) in length and 1.6 inehes (4.0 cm) wide, made the plowing depth increased to 14 inches (35 cm). of rhyolite. This tool was found during subsurface Additional fossil material was recovered by walk- collecting on the island in juxtaposition with ing along the streams assoeiated with the site, fragments of Mammut americanumtusk and bone raking the stream channels, and examining the at the 4 to 8 inch (10-20 cm) depth. A number stream's substrate. of worked chert flakes found primarily in black All vertebrate material recovered was washed carbonaceous earth at the pond's southeast corner in water and permanently labeled. With the ex- have been assigned to the Woodland Period. ception of materials destined for radiocarbon or The Hot Run Site appears to be a spring deposit amino acid dating, all items were coated with and contains a mixed assemblage of vertebrate either Duco Cement and Acetone, Elmers glue and li,thic materials spanning at least 11,550 and water, or Carbowax, depending on their years. Further interpretation of the site must condition. The locations of many, but not all, fossil await completion of other studies now underway, finds were recorded. For specimens collected on ineluding amino acid dating of elements repres- the island the depths at which they were found enting the five indigenous speeies found at the were also recorded. Sediment was removed from site and analysis of pollen collected from within the pulp cavities of Mammutamericanummolars the M ammut americanum molats. and from sections of long bones for pollen analysis The authors express their appreciation to the in the future. V. V. Cutshaw family for extending permission From 1977 through April 1984, over 400 teeth, to work on their property for the last 7 years. bones, and skeletal fragments were recovered and The Archaic knife was identified by Kurt Carr, identified when possible. In addition, a large Catholie University. Woodland artifacts and ver- number of fragments too small to label and a tebrate materials were identified by the respec- small amount of lithic material have been re- tive staffs of the Archeology and Vertebrate covered. Of the three species assigned to the late Paleontology divisions of the Smithsonian Insti- Pleistocene, Mammut americanum is the most tution, Washington, D. C. Funds for radiocarbon commonly represented and is known from over dating were graeiously provided by the Sigma 80 complete or partial milk teeth, premolars, and Xi Researeh Society and are gratefully acknowl- molars. In addition, a large number of tooth and edged. tusk fragments, the latter not exceeding 12 inehes (30 cm), and a small number of bone fragments, the largest weighing about 2 pounds, have been recovered. A Mammnt americanaaa molar sub- mitted to the radiocarbon laboratory of Richard Pardi, Queens College, N. Y., yielded a eollagen date of 11,550 +165 14 C yr B. P. Other identified species from the late Pleistocene are Platygonus cot&pressuq represented by one left caleaneus, one right proximal scapular fragment, one prox- PUBLICATION 75 131

A PLEISTOCENE LEGACY: ARCTIC AND Spruce (Picea rubens) and fir (Abies balsamea BOREAL ELEMENTS IN THE CONTEMPO. and A. fraseri), although much reduced in areal RARY BIOTA OF VIRGINIA. Susan L. WOOD- extent by human aetivity, still occur on a few of WARD and William F. RUSKA, Jr., Department the highest peaks and find their best expression of Geography, Radford University, Radford, VA on Whitetop Mountain (5520 feet, 1682 m) and 24142 and Montgomery County Planning Depart- Mount Rogers (5729 feet, 1746'm), respectively. ment, Christiansburg, VA 24073. Associated with these conifers are such boreal forbs as Clintonia borealis, Maianthemunx cana- d,ense, and. O*alis montana. Whitetop is also ap- Arctic and boreal elements comprise a signif- parently the last refugium in the State for the icant proportion of Virginia's contemporary bi- northern flyin g squ i rrel (Glaucomg s s abrinus). ota, with the component species often occurring Open summits provide habitat for several arc- as isolated outliers of more northerly distribution tic-alpine plants. Arenari,a groenlandic a, Poten- areas. Until rather recently, northern affinities tilla tridentata, and, Trientalis borealis thrive on were explained by the migration of arctic and the Blue Ridge monadnock, Buffalo Mountain boreal forms southwards along the Appalachian (3971 feet, 1210 m), in Floyd County. Large Mountain system into a region where broadleaf populations of. Potentilla trid'entafo also occur on deeiduous forests had been stable since the Mes- Whitetop and on Hawksbill Mountain (4049 feet, ozoic. We know now from pollen studies that 1234 m) in Shenandoah National Park. Virginia's vegetation was greatly affected by High elevation sphagnum "bogs" oecur in the Pleistocene climatic changes and that during the Blue Ridge and the Valley and Ridge physiogra- last glaeial maximum the State was covered by phic provinces. In very restricted habitats may boreal vegetation, with tundra probably occur- be found such boreal species as Vacci'niurn rna' ring on the higher Appalachian summits (Del- crocarpon and. Sanguisorba canadensis. Some court and Delcourt, 1981). The plant and animal bogs in the southern Blue Ridge support relict species of these northern communities have, for populations of the bog turtle (Clemmas mahlen- the most part, retreated upslope and northward bersi). during the Holocene. Refugia for aretic and bo- Northern plants are also found outside the more real species persist where fortuitous combina- typical "boreal" habitats (Harvill and others, tions of elevation, geology, and topography har- 19??, 1981). Most of these are quite rare; aretic- bor cold, wet habitats: mountain summits above alpine bearberry (Arctostaphylos uua-ursi), for 4800 feet (1463 m); treeless summits and north- example, has only a single Virginia station. Other facing talus slopes; and high elevation areas with rare northern species inelude Dalibard,a repens, impeded drainage (Linzey, 1979; Ogle, t982, Cornus canad,ensis, and. Streptopus roseus. 1984) (Figure). None of the curuent communities When trying to interpret disiunct distributions in Virginia truly replicate Pleistocene or Recent there is always a challenge in distinguishing arctic and boreal ecosystems; all have been af- between relicts and recent long-distance immi- fected by local extinction, speciation and/or the grants. In studies of mountain top islands in the admixture of austral forms. Great Basin, Brown (f982) concluded that boreal mammals are relicts of the late Pleistocene, whereas boreal bird populations are maintained by presently high immigration rates. A similar pattern is evident in Virginia; northern mammals have suffered a steady attrition throughout the Holocene and only one species (Tamiasciurus hudsonicus) remains today which is not considered rare, threatened, or endangered state-wide (Linzey, 1979). No non-migratory boreal birds oceur in Virginia at present, although some are known from the fossil record (Guilday and others, 1977). Populations of migratory boreal bird species oceur at high elevations, presumably continually replenished by immigration. Migrant birds Figure. Refugia for arctic and boreal species in may also have played a role in dispersing arctic- Virginia. alpine forbs to some of the open summits in the L32 VIRGINIA DIVISION OF MINERAL RESOURCES

State. The age and origin of these treeless areas PROJECTILE POINTS FROM A MULTICOM- are subject to debate and some at least are known PONENT SITE (4aSm51) IN THE SALTVILLE to be post-Pleistocene. VALLEY, VIRGINIA. Ellen H. WHITE, De- partment of Anthropology, University of Vir- ginia, Charlottesville, VA 22903 (Present address:6349 Mershon St., Philadelphia, PA 19149 ACKNOWLEDGMENTS

The authors wish to thank H. S. Adams of Projectile points from archeological site Dabney S. Lancaster Community College (Co- 44Sm51 convey a variety of information concern- vington) and S. Croy and D. Allard of The Nature ing the presence of prehistoric man in southwest- Conservancy (Blacksburg) for providing us un- ern Virginia's Saltville Valley. This preliminary published data on the distribution of spruce for- report-based primarily upon site tests conduc- ests and "bogs" in Virginia. ted in 1984-illustrates the lengthy record of human presence available at the site, as well as this site's ability to significantly enhance the largely unexplored archeological data base of the valley. REFERENCES The Saltville Valley lies in the Valley and Ridge Province of the Appalachian Mountains. Until Brown, J. H., 1982, The theory of insular biogeo- about 135 years ago, part of the valley supported graphy and the distribution of boreal birds a lake. The lake had been drained by 1850, after and mammals: Great Basin Nat. Mem., no. which much of the valley floor was used as pasture 2, p.209-227. land. Salt extraction, the valley's principal com- Delcourt, P.A., and Delcourt, H. R., 1981, Vege- mercial activity since late in the 18th Century, tation maps for eastern North America: and its allied activities-including the construc- 40,000 B.P. to the present, in Romans, R. C., tion of pipelines, furnaces, roads, railroads, faced., Geobotany II: New York, Plenum, p. L23- tories, etc.-have altered the surface sediments 165. over large parts of the valley. Beginning in the Guilday, J. E., Parmalee, P. W., and Hamilton, 1930's, high pressure injection wells were em- H. W., 1977, The Clark's Cave. bone deposit ployed to extract subsurface salt deposits. Salt, and the late Pleistocene paleoecology of the extraction has seriously undermined the bedrock central Appalachian Mountains of Virginia: of the region, causing some areas to subside. Bull. Carnegie Mus., vol.2,87 p. Railroad construction, removal of sediment for Harvill, A. M., Jr., Stevens, C. E., and Ware, D. fill material, and use of the site surface as a M.E., 1977, Atlas of the Virginia flora, Part parking lot for industrial employees, have further I (Pteridophytes-Monocotyledons): Farm- disturbed parts of site 44Sm51. ville, Va., Virginia Bot. Assoe., 59 p. Site 44Sm51 is situated on a terrace-like ero- Harvill, A. M., Jr., Bradley, T. R., and Stevens, sional remnant that is now one of the highest C. E., 1981, Atlas of the Virginia flora, Part surfaces on the valley floor. This feature appears II (Dicotyledons): Farmville, Va., Virginia to be composed of colluvial slope wash from the Bot. Assoc., 148 p. adjacent Little Valley Limestone hills and was Linzey, D. W., ed., 1979, Proceedings of the probably formed during the (?)Sangamonian In- symposium on endangered and threatened terglacial. The site has been probed sporadically plants and animals in Virginia: Blacksburg, by collectors for several decades. Only recently Virginia Polytechnic Institute and State Uni- have professional excavations been undertaken. versity, 665 p. In 1980, a brief test of the site by archaeologists Ogle, D. W., 1982, Glades of the Blue Ridge in from the Virginia Research Center for Archeol- southwestern Virginia, in MeDonald, B. R., ogy produced evidence suggesting that it was a ed., Symposium on wetlands of the ungla- small late Woodland Period hamlet (Egloff, ciated Appalachian region: Morgantown, f981). A group from Radford University inves- West Virginia Univ., p. L43-147. tigated the stratigraphy of the terrace-like de- 1984, Phytogeography of Vaccinium posit in L982, at which time a small (?)storage nxacrocarpozr Aiton in the southern United pit and several lithic and ceramic artifacts were States: Virginia Jour. Sci., vol. 35, p. 31-47. recovered. In 1984. workers from Radford Uni- - PUBLICATION 75 133

versity conducted a test excavation at the site to REFERENCES determine the time span represented by archeological evidence, and to assess the value of the Egloff, K. 1981, Archaeological test excavation site for further exploration. at 44Sm51 in the Saltville Wellfield Vallev: The 1984 excavation consisted of sampling Unpublished site report on file at Virginia twelve 1-m x 1-m quadrants located along three Research Center for Archeology, Yorktown, transects crossing the site. In excess of 5,000 Va. 13 p. artifacts were recovered, including lithic tools, Holland, C. G., 1970, An archeological survey of ceramics, and several specimens of non-human southwest Virginia: Smithsonian Contribu- bone. Included among these artifacts was a wide tions to Anthropology, no. 12,194 P. range of projectile point types. In all, eighty-three MacCord, H. A., Sr., and Hranicky, W. J., 1979, projectile points, most made of chert, were found. A basic guide to Virginia prehistoric projec- Fifty-nine of these points were identified typo- tile points: Archeol. Soc. Virginia Spec. Publ. logically using the standard guides available no. 6, 48 p. (Holland, 1970; MacCord and Hranicky, 1979; Swope, R., Jr. 1982, Indian artifaets of the east Swope, 1982); the remaining twenty-four points and south-an identification guide: York, Pa., were incomplete, unfinished, or otherwise uni- Robert Swope, Jr., 152 p. dentifiable. The identified point types included Hardaway (2), Palmer (2), LeCroy-St. Albans (2), Big Sandy (2), Guilford (1), Stanly (1), Morrow Mountain (2), Lamoka (3), Halifax (2), Brewerton TOOLS AND DEBITAGE: CULTURAL DEBRIS (5), Merom (3), Orient Fishtail (2), Fox Creek (1), FROM A MULTICOMPONENT SITE (aaSm5l) Bare Island (3), Badin (l), Yadkin (9), Madison IN THE SALTVILLE VALLEY, VIRGINIA. EU- (9), Peedee/Pentagonal (2), and Clarksville (7). It gene BARFIELD, Department of Anthropology, is interesting to note that Bare Island, Fox Creek, Indiana University, Bloomington, IN 47401 Orient Fishtail, and Hardaway points are uncommon in the Tennessee River drainage basin of For almost two centuries the Saltville Valley southwest Virginia. Thirty-four of the projectile of Virginia has been known for its contributions points identified represent the Woodland Period, to paleontology. Evidence documenting multi- while twenty-five points represent the Archaic component human occupation of the Saltville Period, ineluding early Archaic types transi- Valley has accumulated for thousands of years. tional from the Paleo-Indian Period. The points The archaeology of this area, however, has been suggest, therefore, that this site was used by largely ignored by professional archaeologists Indians of various traditions over some 10,000 until recently. During the summer of 1984, a years. preliminary investigation of site 44Sm51 was Clearly, 44Sm51 is a multicomponent site. Ad- conducted in order to determine the complexity ditionally, the recent excavations do not imply and age of the site and to obtain greater insight sedentism, nor that use of the site was restricted into the prehistoric lifestyle of the aboriginal to the Woodland Period. On the contrary, the inhabitants of this valley. The following are par- available data suggest that 44Sm51 was a site tial results of the investigations carried out in of transient occupation over an extended period 1984. of time, probably as a hunting and fishing station Site 44Sm51 is located on the west side of from which the valley bottom resources could be Saltville Valley on the top and sides of an ero- exploited easily. Since 44Sm51 has been exca- sional remnant lying some 23 feet (7 m) above vated only on a trial basis, and since field work the adjacent valley bottom which, until the mid- was concluded only a short time before this report dle of the 19th Century, contained a lake and was prepared, most conclusions are preliminary. marsh. Preliminary investigation of the sediment Further excavations are needed in order to better forming this terrace suggests that it accumulated understand the time frame, areal extent, and uses during the ?Sangamonian Interglacial (McDo- of this site. nald, 1984). About 79 inches (200 cm) of This research was conducted by Radford Uni- ?Sangamon-age deposits lie unconformably over versity with support from the Town of Saltville the Maccrady Shale (Mississippian) bedrock. Sev- and the National Geographic Society (grants25L2 eral centimeters of industrial debris-slag from and 2880 to J. N. McDonald). furnaces, cinders from locomotives, gravel from 134 VIRGINIA DIVISION OF MINERAL RESOURCES

parking lots for heavy machinery, as well as Table 2. Lithic materials recovered by quadrant. discards left by collectors-now cover the surfaee. This site presents at least 32,292 ft.z (3000 mz) z of surface suitable for human occupation. The site F Ftx trlp.N a! location provided an excellent view of the water frEr=E{ N3 and wildlife below. paa -- c'E p F Loeal amateur collectors have known of this site r,o'ict F vrt for many years, and valuable information has 62E-86N 305 13 3 -6 been lost through their activities. Areas for in- 62E-97N 1439 47 4 2 44 vestigation in 1984 were selected with this in 62E-99N tL22 58 5 30 10 39 mind. A meter-wide strip was surveyed aeross 62f'- the top and side of the hill from south to north 101N t23 6 2 2- for excavation; another meter-wide strip was 628- surveyed east to west. Excavated squares were 102N 432 1 1 826 1 located where the least disturbance was noted. 98E-61N 1001 24 1 8312 Twelvel-m by l-m squares were opened, but some Totals 4422 186 13 53 85 58 were found to be so badly disturbed by previous collecting or industrial activity that they were abandoned. Ten students from several universities as well recovered in 1984. In addition, 134 potsherds were as volunteers participated in the 1984 Radford found; most were identified as the limestone tem- University summer field-school. More than one pered late Woodland Period Radford Series. thousand person-hours of labor were devoted to Projectile points from 44Sm51 suggests that the the archaeologieal field excavations. About 258 site was occupied intermittently by several dif- 11.2 (7.32 ms) of earth were carefully excavated ferent aboriginal groups over the last 10,000 by trowels, with stratigraphic control, and sifted years (White, this volume). Other tools and lithic through one-quarter-inch-mesh sereens. Numer- residue associated with these oceupations suggest ous piles of spoil left by eollectors also were sifted. that the site was used primarily as a transient Although the artifacts recovered from these spoil hunting station. This eonclusion is supported piles were out of eontext, they provided informa- espeeially by the lithic tools reeovered (Table 1) tion useful for the preliminary evaluation of this and the large number of small lithic flakes (Table multicomponent site. 1), indicating that some tools had been finished Of the twelve one-meter squares tested, six or retouched at the site. The lithic material was were especially productive. These squares primarily medium-gray, dark-gray, and black yielded several thousand stone and ceramic ar- chert-probably obtained from quarries located tifacts. Tables I and 2 describe the categories, nearby in the Holston River watershed. The lithic composition, and horizontal distribution of souree of jasper, the second-most common ma- cultural debris, exelusive of projectile points, terial, is not known, although local occurrences have been reported (Cooper, 1964). Additional analyses will be undertaken to determine if dis- cernible patterns of size and distribution of the Table 1. Categories of lithic materials. debitage can be recognized. This researeh was conducted by Radford Uni- Flakes 4655 versity and was supported by the Town of Salt- Shatter 70 ville and the National Geographic Society (grant Chunks 46 2880 to J. N. McDonald). Nodule fragments 5 Unidentified tools 7 Serapers 10 Drills 5 REFERENCES Hammerstones 4 Utilized flakes l4 Biface I Cooper, B. N., 1964, New fossil finds at Saltville, Virginia: Mineral Industries Jour., vol. L0, p. Total 4817 1-3. PUBLICATION 75 135

McDonald, J. N., 1984, The Saltville, Virginia, expanded and published in New York as a volume locality: A summary of research and field trip of over 300 pages as The Testimony of Modern guide: Charlottesville, Va., Virginia Division Sc'iencetothe Unitg of Manldnd'. This volume also of Mineral Resources, 45 p. reprints an 185? review by Cabell of Ind,igenous Races oI the Earth, by seven authors ineluding Nott and Gliddon, and such notables as Joseph Leidy, Louis Agassiz, and J. Aitken Meigs. VIRGINIA'S FIRST EVOLUTIONIST: JAMES Cabell's motivation in writing these works was LAWRENCE CABELL. N. T. BOAZ, Virginia clearly to counter the polygenic argument of Museum of Natural History, P. O. Box 3991, human origin advanced by Nott and Glidden, Martinsville, VA 24115. which he considered contrary to both natural science and the biblical Mosaie account. But, as The mid-nineteenth century was a confused with Darwin two years later, his arguments were intellectual battleground of ideas, where the tra- based exclusively on scientific data. As he stated ditional doctrines of the church came into confliet (1858, p. 11), "any attempt to fetter the scientific (not for the first or the last times) with scientific inquirer by the supposed meaning of inspiration, notions of man's place in nature. To further is certain to damage the latter in the estimation charge the environment of debate religious and of a numerous class of intelligent and learned scientific arguments were adduced on both sides men." Virginius Dabney (1981, p. 16) eredits of the political questions of slavery and the con- Cabell with being "ahead of Darwin in publishing trasting economic systems of South and North. a work that recognized the theory of evolution" It was undoubtedly difficult for contemporary and by anticipating "by many years Hugo de natural scientists to avoid the implications of Vries's theory of mutations." On the other hand, their work for the politieal questions of the day, a review of the literature reveals that apparently but it has been virtually impossible for modern no modern historian of science has even cited historians of science to analyze the scientific Cabell or his evolutionary work. The present questions without recourse to inferred political analysis indicates that Cabell was indeed an motivations involved, particularly of those indi- important figure in the development of evolution- viduals in the South (e.g. Gould, 1981; Brace, ary thought in America, although it is also clear 1982). As interesting as these questions may be, that, unlike Darwin's and Wallaee's theory, his this presentation deals exclusively with the scien- formulation lacked the key element of natural tific problems. selection as the mechanism of evolutionary James Lawrence Cabell, M. D. was Professor change. of Comparative Anatomy and Philosophy at the Cabell (1858, p. 17-19) recognized a species as University of Virginia for a period of 52 years, a group of beings with "constant defining char- from 1837 until his death in 1889. Like his well- aeters" and which had "sprung from a common known German eontemporary, Rudolf Virchow, stoek." He argued against the concepts of the he combined a strong interest in natural history influential Samuel George Morton and Louis Ag- with important work in public health. Cabell assiz, which held species to be unehangeable, organized the American National Board of "primordial organie forms." He pointed out that Health and served as the president of the Amer- the varieties that were known to exist in species iean Publie Health Association. He was the were clear indications of adaptations to different nephew of Joseph C. Cabell, Jefferson's spokes- environmental conditions. Cabell therefore un- man in the Virginia General Assembly during derstood one of the basic tenets of the Darwinian the legislation enabling the founding of the Uni- paradigm to follow, i.e., that variation exists versity of Virginia (Dabney, 1981, p. 16). For within a species which can act to adapt the species many years he lived in Pavilion II, and is remem- to a particular environment. bered in one of the Chapel windows at the Uni- Cabell's commitment to evolution-that is, that versity of Virginia (O'Neal, 1976 p. 63,72). species change through time-is clear. He states Cabell wrote a lengthy review of the influential (1858, p. 20) that, ". . .certain acquired peculi- book published in 1854 by Josiah C. Nott and arities are often reproduced with perfect regu- George Gliddon, of Mobile, Alabama, entitled larity so as to give rise, within the limits of a Tgpes of Mankind. The review appeared in the single original species, to'uarieties' marked by Protestant Episcopal Reaiew and Church Register characters as'permanent' as those which distin- of January, 1857. This review was subsequently guish the species itself. . ." Yet, he did not continue 136 VIRGINIA DIVISION OF MINERAL RESOURCES

to the Darwinian conclusion that this evolution ciple of particulate inheritance and thus was would lead to the production of new species. In faced with the problem presented by the theory a footnote, Cabell (1858, p. 2l) endorses Dana's of blending inheritance. Cabell noted that in cases definition of a species as "based on a specific of "accidental variations," both sexes must pos- amount or condition of concentrated force in the sess the trait if it were not to be lost through act or law of creation." We may thus conclude admixture in the general population. One must that Cabell accepted evolution, but only within conclude that he considered the inheritance of species, which were themselves created. acquired characters the more important of the Today we would charaeterize Cabell's evolu- forces of evolution. tionary mechanism, the inheritance of acquired In lengthy passages on physiology, history, lin- characters, as "Lamarkian." However, Cabell ap- guistics, and geology that sound remarkably parently derived most of his belief in this thesis "modern," Cabell argued for the basic unity of from Johannes Muller of Berlin, a disciple of the human species. He (1858, p. 317) disputed Goethe and the mentor of Virchow and the an- claims by Sir Richard Owen and Jeffrys Wyman thropologist Hermann Klaatsch. Appendix A in that the human races differ as much as the "three Cabell's book is dedicated to a discussion of species" of orangutan. As is now well known the Muller's ideas. As Cabell (1858, p. 22-28) dis- orangutan consists of a single speeies, Pongo eusses it: pagrnaeus. Cabell's (1858, p. 309) position on the unity of the human species is made clear in a The law of the permanency of primordial passage comparing Sir Isaac Newton with the forms is admitted on all hands to be subject rest of humanity: to qualifications. Within certain limits [the speciesl, it adapts itself to various changes Can such a "celestial genius" one may rea- in the influences under which the race may sonably ask, be of the same original parent- subsist. Thus, by carefully changing the food age with the Bushman, who lives in holes and other agents of vital stimulation, we may and caves, and devours ants'eggs, locusts modify, to an extent sometimes quite con- and snakes? . . . It should not surprise us siderable, the outward structural character that when we contemplate exelusively the of many plants and low anirnal organisms; patent diversities of raees and overlook the and these newly acquired characters may less obvious but more significant evidences then be perpetuated by hereditary transmis- of a common nature, we should shrink from sion, under the influence of the law of as- the conclusion to which a deeper insight into similation between parent and offspring, the facts must yet inevitably conduct us. even though the eauses which originally determined the variation from the primitive Cabell's position, of course, has been amply sus- type have ceased to operate. tained in this regard over the past 126 years. In conclusion, James Lawrence Cabell clearly From this passage it is clear that Cabell in no was ahead of most of his American contempo- way implied differential survival of offspring, raries, including Nott, Gliddon, Agassiz, and although he recognized, as Darwin did, that var- Dana, in correctly reeognizing many of the cen- iability could be inherited. tral points of evolutionary theory as they were Cabell postulated a second manner of variation, eventually explicated by Darwin and \{allaee. As what we term "mutations" and what Darwin Darwin, he was forced to accept the doctrine of termed "sports." He stated, "a similar effect [to inheritance of acquired characters (Darwin's inheritance of aequired charactersl is produced "pangenes") as an answer to loss of inherited in those cases in which a given variation appears change through "blending" in populations, since accidentally in a single individual and is then neither knew of Mendel's work. Yet, Cabell transmitted to his offspring" (Cabell, 18b8, p. 2B). laeked Darwin's insight that races could be in- This idea is the basis for Dabney's (1981) refer- cipient species, and that differential reproduction ence to de Vries's theory of mutation, and it is lay at the base of evolutionary change through an early, though incorrectly synthesized, reali- time. zation that mutation is a major eontributor to The assistance of F. Spencer and C. R. Kaut species' variability. in the preparation of this study is gratefully As Darwin, Cabell lacked the Mendelian prin- acknowledged. PUBLICATION 75 137

REFERENCES ments of physiology, zoology, and comparative philology in favor of the specifie unity and common origin of all the varieties of man: Brace, C. L., 1982, The roots of the race coneept New York, Edward O. Jenkins,344 P. in American physieal anthropology, in Dabney, V., 1981, Mr. Jefferson's university, A Spencer, F., ed., A history of American phys- history: Charlottesville, Univ. Press of Vir- ical anthropology: New York, Academic ginia, 643 p. Press, p. 11-29. Gould, S. J., 1981, The mismeasure of man: New Cabell, J. L., 1858, The testimony of modern York, Norton, 352 p. science to the unity of mankind. Being a O'Neal, W. B., 1976, Pictorial history of the summary of the conclusions announeed by the University of Virginia: Charlottesville, Univ. highest authorities in the several depart- Press of Virginia, 208 p.