{l W H H ‘m ‘ (21:: 8 ‘31 [Ill ”1} H! H THESE; LIBRARY Michigan State 2. " University This is to certify that the dissertation entitled LATE NISCONSINAN APPALACHIAN HERPETOFAUNAS: STABILITY IN THE MIDST OF CHANGE presented by Leslie Porter Fay has been accepted towards fulfillment of the requirements for Doctor of degree in Geological Philosophy Sciences fl/VZ‘WQP essor ' JT'AE::1Holman Datexo Tug;/7c?§’:/ MSU is an Affirmative Action/Equal Opportunity Institution 0-12771 )VISSI_] RETURNING MATERIALS: Place in book drop to LIBRARJES remove this checkout from n ' your record. FINES will be charged if book is returned after the date stamped below. LATE NISCONSINAN APPALACHIAN HERPETOFAUNAS: STABILITY IN THE MIDST OF CHANGE By Leslie Porter Fay A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Geological Sciences 1984 ABSTRACT LATE NISCONSINAN APPALACHIAN HERPETOFAUNAS: STABILITY IN THE MIDST OF CHANGE By Leslie Porter Fay Most fossil herpetofaunas from eastern North America do not exhibit the dramatic Wisconsinan-Holocene range adjustments characteristic of associated birds and mammals. Mammal and bird lists consist of current residents of the fossil locality plus a sizable component of taxa that range to the north or west. Herptile lists are composed of area residents with only a very few now-extirpated forms. Factors respon- sible for the varying responses to climatic change among vertebrate classes have not yet been adequately identified. Late Nisconsinan Appalachian herpetofaunas neither support nor contradict the climatic equability hypothesis, as almost no extralimital records previously reported are valid. Although the ranges of plants and other animals changed in response to the glacial-interglacial climatic shift, herptiles in eastern North America were relatively unaffected by these events. ACKNOWLEDGEMENTS I thank J. Alan Holman for initiating me into the study of fossil herptiles. The passage has been smoother for his good counsel and good cheer. Terrance J. Martin and Thomas C. LaDuke offered stimulating discussion of natural history in general and bones in particular. The efforts of Robert L. Anstey, Marvin M. Hensley, Grahame J. Larson, and Richard J. Seltin in improving this paper are greatly appreciated. Kingston Saltpeter Cave was excavated by Larry 0. Blair, James H. Henry, Joel and Carol Sneed, who have encouraged me to decipher the Georgia herpetofaunal record. Ronald C. Wilson loaned the Kingston material, a very welcome addition to the study. Gwen Fay has remained steadfast in her support and belief that I would eventually finish my schooling. John E. Guilday graciously permitted me to begin this project. The finished product is not all it could have been, for lack of time to discuss the work with him. ii TABLE OF CONTENTS Page LIST OF TABLES ........................... iv LIST OF FIGURES .......................... v LIST OF ABBREVIATIONS ....................... v11 INTRODUCTION ............................ 1 REGIONAL SETTING .......................... 4 HISTORY AND PURPOSE OF STUDY .................... 6 PROCEDURES ............ ' ................. 8 FAUNAL LISTS AND DESCRIPTIONS ................... 9 New Paris No. 4, Bedford County, Pennsylvania ......... 9 Natural Chimneys, Augusta County, Virginia .......... 14 Clark's Cave, Bath County, Virginia .............. 18 Baker Bluff Cave, Sullivan County, Tennessee ......... 20 Kingston Saltpeter Cave, Bartow County, Georgia ........ 24 SYSTEMATIC DESCRIPTIONS ...................... 25 RESULTS .............................. 48 DISCUSSION ............................. 52 CONCLUSIONS ............................ 62 SUMMARY .............................. 62 REFERENCES CITED .......................... 64 LIST OF TABLES Table ' Page 1. Checklist of herptiles from five Late Wisconsinan Appalachian herpetofaunas. Taxa from all levels are included for New Paris No. 4 and Baker Bluff Cave ................ 10 2. Three ratios used to distinguish Storeria dekayi from Storeria occipitomaculata ................... 41 3. Three ratios used to distinguish Virginia striatula from Virginia valeriae ....................... 43 4. Taxonomic sampling (fossil recovery) rate expressed as percentage of fossil taxa to taxa currently ranging within 50km of fossil locality .................... 49 iv LIST OF FIGURES Figure Page 1. Location of five Late Nisconsinan fossil herpetofaunas in the central and southern Appalachians. Base map shows paleovege- tation of 10,000 years ago (map redrawn after Delcourt and Delcourt, 1981, p. 149) .................... 5 2. Sympatry maps for New Paris No. 4 fossil herpetofaunas. Unit A <11,300 rcybp, Unit B >11,300 rcybp. Squares indicate fossil locality ........................ 13 3. Sympatry maps for (A) Natural Chimneys and (B) Clark's Cave fossil herpetofaunas. Squares indicate localities ....... 17 4. Sympatry maps for Baker Bluff Cave fossil herpetofaunas. (A) upper levels [0 - 5ft <10,560 rcybp], (B) mid-levels [5 - 7ft 10,560 to 11,640 rcybp], (C) lower levels [7 - 10ft 11,640 to 19,100 rcybp]. Circles indicate locality. . 23 5. Sympatry map for Kingston Saltpeter Cave fossil herpetofauna. Square indicates locality ................... 26 LIST OF FIGURES Figure Page 6. Histogram of modern range terminations for fossil mammals and herptiles of four Late Wisconsinan local faunas along Guilday's Appalachian faunal gradient ............. 60 vi BB CC CL ft KSC NAN NC NP4 NSL POPR rcybp ypb ZN LIST OF ABBREVIATIONS Baker Bluff Cave Clark's Cave greatest centrum length feet kilometers Kingston Saltpeter Cave meters millimeters least neural arch width Natural Chimneys New Paris No. 4 neural spine length greatest postzygapophyseal to prezygapophyseal length radiocarbon years before present years before present (determined by non—radiometric method) greatest zygosphene width vii INTRODUCTION The time period from the Late Nisconsinan glacial maximum 20,000 years ago to the present is marked by extensive climatic change. Deglaciation, the Climatic Optimum, and introduction of large-scale agriculture by European immigrants have had dramatic impact on the biota of eastern North America. The effects of these changes have long been a subject of study. Until recently, reconstructions developed from fossil evidence depended on the assumption that current (that is, pre- settlement) patterns of vegetational communities have existed during most of the Quaternary. Nith allowance for areas covered by glacial ice and perhaps a severe "periglacial" fringe, the modern communities were mapped as if they migrated intact south and north in response to glacial—interglacial cycles (Brunnschweiler, 1962; Dillon, 1956; Flint, 1971; Martin, 1958). The boundaries of these dislocated communities were delineated on scanty pollen evidence with reinforcement from scat- tered vertebrate discoveries and often from the author's "intuition". The animals currently associated with each community were assumed to move along with the plants (Martin, 1958). A more detailed picture has emerged in the past two decades with the advent of sophisticated research techniques and equipment. Pollen analysis gained a precise time frame in radiocarbon dating. By tracking rates of tree species dispersal across the increasing number of dated fossil pollen localities, Davis (1976, 1983) has shown that 1 2 the communities we now recognize did not exist as such during the Wisconsinan Stage. Plant species dispersed at different rates dependent on the ability of populations to expand. Multiplied by the dozen or more glacial-interglacial cycles now documented, only portions of the communities we know can be expected to have coexisted continuously through the Quaternary. With the use of screen washing technique for recovery of small macroscopic fossils (Hibbard, 1949; McKenna, 1962) it has become theo- retically possible to recover every specimen of every taxon from a fossiliferous deposit. Small animals provide more information useful in making paleoclimatic reconstructions because, as a group, they are more closely associated with particular microhabitats and less likely to migrate seasonally than large vertebrates. Thorough collecting with the screen washing technique and careful stratigraphic documentation make it possible to distinguish between synchronously deposited and heterochronic fossil accumulations that contain species not now sympatric. Our present understanding of Wisconsinan and Holocene environments is based mostly on the fossil record of plants and mammals. Only in the last fifteen years have other organisms been subjected to detailed study. Did these groups respond to climatic change in similar fashion? Beetles seem to have reacted to Late Nisconsinan glaciation in much the same manner as did mammals. The translocation is even more dramatic than that of mammals, as most Late Nisconsinan beetle faunas near the ice front in midwestern and eastern North America are dominated by boreal parkland (open ground) forms (Morgan, Morgan, Ashworth, and Matthews; 1983), while contemporary mammal faunas include both boreal and temperate animals. Late Nisconsinan terrestrial molluscs are known mostly from loess deposits in the Midwest. These animals apparently did maintain community integrity while being dislocated by the ice margin (Frest and Fay, 1980), although several forms now restricted to the Rocky Mountains ranged east across the Great Plains (Frest and Rhodes, 1981). Study of non-mammalian vertebrates has lagged behind other groups for several reasons. Bird, herptile, and fish bones can be more diffi- cult to identify than mammal teeth. They are also more fragile and less diagnostic after abrasion and breakage. Birds and fish are poorly rep- resented in some accumulations because their habit and habitat require- ments render them less available for fossilization than terrestrial vertebrates. In addition, fewer paleontologists specialize in so-called lower vertebrates than in mammals. Herptiles have been noted in eastern North American local faunas since the earliest studies (Cope, 1899). However, they frequently appear in the literature as faunal lists with little or no interpreta- tion or evaluation in comparison with other vertebrates. This report gathers together previously unstudied Late Wisconsinan Appalachian herpetofaunas, along with several which have received partial examina- tion. The primary question asked is whether or not herptiles responded to the Nisconsinan-Holocene climatic shift in the same manner as mammals. PreVious investigations implied one of two possibilities. One, the known fossil herptiles do not match the zoogeographic and ecologic characteristics of mammals from the same local faunas and are therefore intrusive at a later time. This also leads to the conclusion that 4 either there were no herptiles present when the mammals accumulated, or that herptiles were present but not preserved. Two, the known fossil herptiles do not match the zoogeographic and ecologic characteristics of mammals from the same local faunas because some unknown factors of habitat, physiology, or behavior allowed reptiles and amphibians to remain in an area that was seemingly unsuitable (if presence of boreal mammals be taken literally to infer boreal climate). A second question may then be posed if the latter possibility holds. What factors allowed vertebrates of differing (current) habitat requirements to coexist during the Late Wisconsinan? REGIONAL SETTING The five herpetofaunas of this study were recovered from localities within the Ridge and Valley Province of the central and southern Appala- chian Mountains (Figure 1, see faunal lists for map coordinates). This geomorphic setting has probably remained relatively stable since at least early Quaternary time. No glaciations or sea-level fluctuations have directly impacted the region. Proximity of the glacial limit to the north and northwest may well have been controlled as much by struc- ture as climate. Upon crossing the present location of Lakes'Erie and Ontario, the Nisconsinan ice encountered a regional slope rising to the south and east onto the Appalachian Plateau, which the glacier appar- ently was unable to surmount (Mickelson, Clayton, Fullerton, and Borns; 1983). Direct physical evidence of alpine glaciation on Appalachian peaks is lacking. Supposed glacial grooves in North Carolina (Berkland and Raymond, 1973) are, in fact, cable grooves from logging or quarrying nixeo Hanowooos Figure 1. Location of five Late Nisconsinan fossil herpetofaunas in the central and southern Appalachians. Base map shows paleovegetation of 10,000 years ago (map redrawn after Delcourt and Delcourt, 1981, p. 149). 6 operations (McKeon, 1974; Hack and Newell, 1974). Sorted patterned ground in the central and southern Appalachians may indicate cold climate at some time during the Quaternary (Clark, 1968), with at least discon- tinuous permafrost as far south as northern West Virginia (Péwé, 1983). The northeast-southwest trending ridges and valleys provide not only myriad potential microhabitats at various slope attitudes, aspects, and elevations, but potential corridors for dispersal in response to regional climatic change (Guilday, Hamilton, Anderson, and Parmalee; 1978). Interdigitating temperate valley floor situations and boreal to tundra ridgetop environments would allow a mixed or mosaic biota to develop. As many of the Appalachian fossil deposits are the result of raptorial bird activity, the admixture of organisms now inhabiting spatially discrete areas would be enhanced in the resultant fossil record. HISTORY AND PURPOSE OF STUDY Neil Richmond (1964) published the first herpetofaunal paper con- cerning a Late Nisconsinan-early Holocene local fauna from the Appala- chians. The Frankstown Cave (Blair County, Pennsylvania) collection is poor in amphibians, but the seven snake taxa are found in most subse- quently examined local faunas. All animals reported as fossils from Frankstown live near the locality now. A number of local fauna reports include some interpretation of the herptiles recovered. Richmond contributed herptile identifications for the New Paris No. 4 (Guilday, Martin, and McCrady; 1964) and Natural Chimneys (Guilday, 1962) papers. Helen McGinnis and John Guilday 7 studied the anurans and lizards, respectively, from Clark's Cave (Guilday, Parmalee, and Hamilton, 1977). Herptiles of Baker Bluff Cave were examined by Van Dam (1978), and a long-term study of the slightly older Ladds Quarry local fauna, Bartow County, Georgia, has produced two herpetofaunal reports (Holman, 1967; Wilson, 1975). Many other Quater- nary Appalachian local faunas reposited at the Carnegie Museum of Natural History contain modest numbers of herptile remains which have not yet been studied (J. E. Guilday, written communication, 9 April 1982). I became interested in studying Quaternary Appalachian herpeto- faunas in 1982 and, with John Guilday's cooperation and encouragement, began to examine four herpetofaunas (New Paris No. 4, Natural Chimneys, Clark's Cave, and Baker Bluff Cave) with relatively high diversity and potential for meaningful comparison with the co-occurring mammals and birds. The herptiles from Kingston Saltpeter Cave were subsequently made available to me by Ronald C. Wilson, Museum of History and Science, Louisville, Kentucky. This locality adds information from the southern- most portion of the Appalachians as well as the opportunity to contrast a full-glacial local fauna (Ladds) with a nearby late-glacial local fauna (Kingston). The purpose of this investigation is to evaluate the Late Wiscon- sinan herpetofauna of the central and southern Appalachians. Two main objectives are to (1) document regional relationships among reptiles and amphibians at the end of the most recent glacial episode and (2) compare inferred climatic effects on the herptiles with previously interpreted effects on mammals and birds of the same local faunas. PROCEDURES Most specimens used in this study were gathered from previously collected local faunas in order to make comparisons with conclusions already drawn from the mammals and birds. All material except that from Kingston Saltpeter Cave was collected by field parties under the direc- tion of the Vertebrate Fossils Section, Carnegie Musem of Natural History and is reposited at that institution. The Kingston material was col- lected by the Kingston Saltpeter Cave Study Group, Clayton County Cavers, National Speleological Society, Grotto No. 285 and is currently housed at the Museum of History and Science, Louisville, Kentucky. Modern skeletal material used to identify the fossils is reposited in the Herpetology & Ichthyology Division, The Museum, Michigan State University. The fossil localities are all from upland areas in the Appalachian region and are of Late Wisconsinan to early Holocene age (RancholaBrean Land Mammal "Age"). To streamline discussion of identification techni- ques and zoogeographic analyses, it is assumed that no extinct taxa need be considered, nor modern forms which do not occur east of the Mississippi River. Although eastern species currently restricted to coastal plain (austroriparian) environments were considered in making identifications, none were found and so these forms are also not dis- cussed. In nearly all cases, subspecific assignments are not made for fossil herptiles because sufficient osteological criteria have not been defined. Size clines according to Bergmann's Response, which have proven 8 9 most illuminating for the mammals of these local faunas, are not docu- mented for herptiles. Detailed explanations of identification techniques are provided if the method has not previously appeared in the literature. Many herptile bones have not been identified below generic or even familial levels in prior studies. No major advances in osteological identification have resulted from this work. Both qualitative and quantitative methods are employed, as neither is satisfactory alone. Particular means of identi- fication are described in the Systematic Descriptions section. Taxonomy above the generic level follows Dowling and Duellman (1978). Genera, species, common names, and geographic distributions are after Conant (1975). FAUNAL LISTS AND DESCRIPTIONS In this section each of the five herpetofaunas is treated. The faunal list includes type of fossil accumulation, map coordinates, a list of taxa including person responsible for each identification, and taxa rejected and reassigned. A short description of the herpetofauna follows each list. Table 1 permits faunal comparisons among all five local faunas. New Paris No. 4, Bedford County, Pennsylvania fissure fill altitude - 465m. latitude - 40°05'N. longitude - 78°39'W. 10 Table 1. Checklist of herptiles from five Late Wisconsinan Appalachian herpetofaunas. Taxa from all levels are included for New Paris No. 4 and Baker Bluff Cave. TAxggg, WP4 NC cc BB KSC Crggtobranchus alleganiensis X turus maCu osus X Rate fitfialmus vlrldescens x x l I liEystoma maculatun ‘ X I . IICU atun-group x x Eli; 1L sp. 1 Bison nathus sp. X x Disnognatfilnae 'Plethodontiinl' X X gxzinophilus h iticus seu otr ton ru r '0' 3p. Ska hio us holbrooki 5570 anerlcanus o wooafiousel fowleri R la cruel er H. chr soscelis versicolor ”3‘8”“ I” N” I)!" I “I xxx xxxxx H I)!" . ens-group I I. s Ivatlca l Chel are se ggntina nostern ac Chr sen s gicta ra ten 5 e0 ra hica X t Icrrapene carolina ! Ei§didae (unditermlnod) Anolis carolinensis X Scelo rus undulatus X Eumeceg Tasciatus I E. flatice s ShinEE) a ateralis lTeroJia flthr’gaster a, slpedon . sp. 7h in: se tenvittata .toreria e a! ' occipitomgcu at. .. 2" j?! R 43:; E as xuxxxx x ”I” it KKK I i». ll sp. 11mm JMS sauri tus I ”XXI t—E rtalis Hrg n a striatula ' !3LJEEL!E IL sp. La ho his amoenus DiasopEls B¥0Cta£%§ leterodbn p atzrh nos ITantilFa coronata CBTUBer or Mastico his heodr s aestivus . verna is Eggogfiara cocclnea a e guttata E. 0550 eta EL 59. f In ‘Nssiuflnflat . etu us f-tdapmh; Pitta h 5 me anoleucus 9* strodon contortrlx frotalus horridus I“ I”)! RN 3‘ X” nua- it XXX XX xx XX X as xxx xxxxxxxxxx I I“ xxx N xxxxxx “I use X xxxxx xx ' .. __-'.T L m. ~"f-“T-ww- t ' l‘fi—i ‘[ 11 Unit A (upper levels) Notophthalmus viridescens+ Ambystoma maculatum-group++ (A, jeffersonianum or A, maculatum) Ambystoma sp.+ Desmognathus sp.+ "Plethodontiini"++ Bufo a, americanus+ Bufo woodhousei fowleri++ Hyla crucifer+ Rana cf. 3. pipiens+ Rana sylvatica+ Rana sp.++ Nerodia sipedon+ ?Regina septemvittata++ Storeria dekayi++ Storeria occipitomaculata+ Thamnophis sauritus++ Thamnophis sirtalis+ Virginia valeriae++ Carphophis amoenus+ Diadophis punctatus+ Coluber or Masticophis++ *Opheodrys aestivus++ Elaphe guttata++ ETaphe obsoleta+ Lampropeltis calligaster++ Lampropeltis getulus++ Lampropeltis triangulum+ Pituophis melanoleucus+ Agkistrodon contortrix+ Crotalus horridus+ Unit 8 (lower levels) Notophthalmus viridescens+ Amb stoma maculatum-group++ (g, jeffersonianum or A, maculatum) Amb stoma sp.+ esmo nat us sp.+ "Pletaodontiini"++ Bufo a. americanus+ *Bufo americanus cf. 8, a, copei+ H la crucifer+ Rana sp.++ Storeria dekayi++ Storeria occipitomaculata+ Thamnophis Sirtalis+ Virginia valeriae++ Diadophis punctatus+ Elaphe obsoleta+ Lampropeltis getulus++ ?Pituophis melanoleucus+ Agkistrodon contortrix+ Crotalus horridhs+ 12 * no longer present at New Paris + identified by Neil 0. Richmond ++ identified by Leslie P. Fay Taxa dropped from list by Fay: Eur cea? assigned to "Plethodontiini" Plethodon cf. glutinosus assigned to "Plethodontiini" Coluber cf. constrictor assigned to Coluber or Masticophis Neil 0. Richmond (in Guilday, et al.; 1964) reported twenty fossil herptile taxa from New Paris No. 4. Three species identifed by Richmond were not included in the published report. I have identified ten addi- tional forms and modified three of Richmond's published determinations. New Paris No. 4 is dated at 11,300i1000 rcybp (Y-727) on charcoal from Unit A, five feet above the assigned boundary between the two faunal units. No significant zoogeographic changes have resulted from the addi- tions to the herpetofaunal list. The newly-identified species live in the area today and are ecologically compatible with the rest of the herp- etofauna. With the exception of Bufo americanus copei, the New Paris No. 4 herptiles are now widely distributed throughout eastern North America. All New Paris No. 4 herptiles are small enough to have been prey items of raptorial birds. Many of the species live in or near cave mouths and other smaller fissures. It is difficult to distinguish whether herptiles in a fossil accumulation were contributed through predation, accidental falls, or residence at the accumulation site. Unit A vertebrates comprise a temperate fauna with an area of sympatry through much of the central Appalachians for the mammals (Guilday, et al.; 1964, Map A) and a restricted area within the mammal distribution for herptiles (Figure 2). The boreal aspect of the older utut B Figure 2. Sympatry maps for New Paris No. 4 fossil herpetofaunas. Unit A <11,300 rcybp, Unit B >11,300 rcybp. Squares indicate fossil locality. 14 Unit B mammal fauna (Guilday, et al.; 1964, Map 8) is in sharp contrast with Unit B herptiles (Figure 2) which show a sympatric distribution nearly identical to the Unit A sample. However, 79% of Unit B herptiles now live in the Quebec-New Hampshire-Vermont area where 83% of Unit B mammals may be found. Natural Chimneys, Augusta County, Virginia raptor roost altitude - 414m. latitude - 38°22'N. longitude - 78°05'W. Notophthalmus viridescens+ Ambystoma maculatum-group++ (A, jeffersonianum or A, maculatum) *Ambystoma tigrinum++ Desmognathus sp.+ 11Plethodontiini"++ ' Scaphiopus holbrooki+ Bufo americanus+ Bufo woodhousei fowleri++ Bufo sp.+ Hyla ?chrysoscelis/versicolor++ Rana catesbeiana+ Rana cTamitans++ Rana ?palustris+ *Rana cf: 3, pipiens++ Rana sylvatica++ Rana sp.+ Chelydra serpentina+ Terrapene cardlina+ Sceloporus undulatus+ *Scincella lateralis++ *Nerodia erythrogaster++ Nerodia sipedon+ Storeria occipitomaculata+ Storeria sp.++ Thamnophis sauritus++ Thamnophis sirtalis+ Virginia sp.+ Carphophis amoenus+ Diadophis punctatus+ Coluber or Masticophis++ O heodr s verna is++ Eiapfie guttata+ Elaphe obsoleta++ Lampropeltis getulus++ 15 Lampropeltis triangulum+ Crotalus horridus+ * no longer present at Natural Chimneys + identified by Neil 0. Richmond ++ identified by Leslie P. Fay Taxa dropped from list by Fay: Masticophis flagellum assigned to Coluber or Masticophis Crotalus cf. adamanteus assigned tE—CFETETUS horridus The Natural Chimneys herpetofauna was first reported by Neil 0. Richmond (in Guilday, 1962) with twenty-seven taxa. I have identified seventeen additional forms and rejected two from the previous list, for a total of thirty-six fossil forms. This expansion was made possible by additional material collected in 1977 as well as re-examination of the original collection. Guilday (1962) claimed that the fossil accumulation was not a tem- poral unit because it may have built up over a great length of time. If this is the case, the amount of time was not so large as to reveal any changes in composition of the local fauna. That is, no disharmonious elements (beyond the boreal-temperate admixture now known to be synchron- ous) are recorded. 0f the four extinct mammals, two are larger subspe- cies of living forms (Tamiasciurus hudsonicus tenuidens - Guilday, et al. 1964; Eptesicus fuscus grandis - Guilday, 1967) and two survived to the Wisconsinan-Holocene boundary (Castoroides ohioensis to 10,2301150, Mylo- hyus nasutus to 9410t155; Meltzer and Mead, 1983). Thus, Guilday's faunal evidence date of earliest Holocene age for Natural Chimneys is supported by radiometric dating of material from other localities. The most important change in the herpetofaunal list is the rejec- tion of Masticophis flagellum and Crotalus adamanteus, which currently range no closer than 300km to the south of the locality. The Natural 16 Chimneys herpetofauna has been thought to be a southern faunal unit, in very strong contrast to the boreal nature of the mammalian component (Guilday, 1962; Lundelius et al., 1983). In the corrected list, all members are found at the latitude of Natural Chimneys, but four species (Ambystoma tigrinum, Rana cf. 3, pipiens, Scincella lateralis, and Nerodia erythrogaster) do not presently inhabit the central Appalachians in Virginia and West Virginia. The Natural Chimneys herpetofauna shows sympatry mapping to be an inexact technique for paleoenvironmental reconstruction. Although all the herptiles may be found as far north as the fossil locality, reliance on areas of sympatry (Figure 3A) yields the same result as the mistaken inclusion of Crotalus adamanteus and Masticophis flagellum--a southern herpetofauna. Pollen evidence indi- cates that the area including the western edge of the Natural Chimneys herpetofaunal sympatry in Mississippi may have been the full-glacial refugium of many Mixed Mesophytic Forest species now distributed through the central Appalachians (Delcourt and Delcourt, 1975). This may prove more useful in reconstructing Wisconsinan environments than supposed southern faunal elements. Most of the fossil herptiles could have been prey items for owls or other raptorial birds. The snapping turtle may have been incorporated by woodrat activity or flooding, two alternatives Guilday (1962) offered for the presence of large mammal remains. The large mammal fossils are fragmentary, as is the turtle. Many of the herptiles are burrowers or occasional inhabitants of caves and could have been living at or near the locality where bones accumulated. The mammals and birds of Natural Chimneys indicate a general cool climate year-round with moister summers (Guilday, 1962). It has been Figure 3. Sympatry maps for (A) Natural Chimneys and (B) Clark's Cave fossil herpetofaunas. Squares indicate localities. 18 stated that the herptiles of this local fauna disagree with the mammals (Lundelius, et al.; 1983). A literal interpretation of increased annual cold for mammals and increased annual heat for herptiles is indeed contradictory, but certainly an oversimplification. All verte- brate classes indicate increased available moisture, which probably acted to increase tolerance limits to temperature extremes for the now-allopatric animals. Clark's Cave, Bath County, Virginia raptor roost altitude - 448m. latitude - 38°05'10"N. longitude - 79°39'25"W. Notophthalmus viridescens# Ambystoma macuTatum-group# (A, jeffersonianum or A, maculatum) Ambystoma opacum# Amb stoma sp.+ Desmognathus sp.# "Plethodontiini"# Gyrinophilus porphyriticus# Pseudotriton sp.#i SEaphiopus holbrooki# Bufo americanus# Bufo woodhousei fowleri# Bufo sp.+ Hyla crucifer+ Rana cateSbeiana+ Rina Clamitans+ Rana pglustris+ *Rina c . R. pipiens+ Rana sylVEtica+ Rana sp.+ *Graptemy§_geographica# Sceloporus undulatus++ Eumeces ?TatiEeps++ Nerodia sipedon# ?Regina septemvittata# Storeria dekayi# Storeria occipitomaculata# Thamnophis sauritus# Thamnophis sirtalis# Virginia valeriaeEi Carph0phis amoenus# 19 Diadophis punctatus# Heterodon platyrhinos# *?Tantilla coronata# Coluber or Masticophis# Opheodrys aestivus# Opheodrys vernalis# *Cemophora coccinea# Elaphe guttata# Elaphe obsoleta# Lampropeltis calligaster# Lampropeltis getulus# Lampropeltis triangulum# Crotalus horridus# no longer present at Clark's Cave identified by Leslie P. Fay identified by Helen McGinnis identified by John E. Guilday I+=u=x~ The Clark's Cave herpetofauna was partially reported in Guilday, et al. (1977). Helen McGinnis identified the anurans plus some of the caudates and John Guilday identified the lizards and sorted the snakes into familial categories. Their list includes fourteen taxa, ranging in rank from families to species. I have made thirty-one identifications (all from families and genera listed by McGinnis and Guilday) for a total of forty-two herptile taxa. The mammal fauna of Clark's Cave is very similar to that of New Paris No. 4, indicating an age assessment for the Clark's Cave fossil accumulation of about 11,000 ybp (Guilday, et al.; 1977). Three reptiles reported here from Clark's Cave (Graptemys geographica, Cemophora coccinea, and Tantilla coronata) plus the Rana pipiens-complex frogs do not presently occur at the locality. The average distance by which they are removed from Clark's Cave is 90km and the farthest distant herptile, the leopard frog, is 125km away. This minor disjunction requires little or no invocation of climatic change. Periodic fluctua- tion of range limits in response to various perturbations or even 20 insufficient collecting, are adequate to explain the extirpations. The Clark's Cave fossil herptiles are presently sympatric in the Great Smoky Mountains (Figure 3B). This distribution likely reflects habitat diver- sity within the area rather than a single set of ecologic conditions suitable for all forms. Fifty-two percent of the Clark's Cave fossil mammals do not presently occur in the central Appalachians, live only at higher elevations in the area, or are of sizes which indicate boreal popula- tions (Guilday, et al.; 1977). Seven percent of the avifauna does not presently spend any part of the year in the central Appalachians. Why can every herptile of the local fauna still be found within 125km after 11,000 years of climatic change that has driven out several birds and many mammals? Baker Bluff Cave, Sullivan County, Tennessee raptor roost altitude - 450m. latitude - 36°27'30"N. longitude - 82°28'00"W. mid-levels Cryptobranchus alleganiensis+ Necturus maculosus+ Ambystoma maculatum+ Ambystoma opacum-group++ Ambystoma sp.+ Desmognathinae++ Bufo‘americanus+ Buio woodhousei fowleri+ Bufo sp.+ Rana sylyatica+ Rana sp.+ Eumeces fasciata+ Nerodia sipedon+ Nerodia sp.+ Thamnophis sirtalis+ Thamnophis sp.+ Diadophis punctatus+ Heterodon platyrhinos+ 21 Coluber or Masticophis+ ?Elaphe sp.+ *Lampropeltis getulus+ Lampropeltis triangulum+ Crotalus horridus+ * no longer present at Baker Bluff + identified by George H. Van Dam ++ identified by Leslie P. Fay Taxa dropped from list by Fay: Ambystoma opacum assigned to Ambystoma opacum-group Desmognathus sp. assigned to Desmognathinae Hyla sp. assigned to Bufo sp. Only minor changes in Van Dam's (1978) herpetofaunal list are made in this study. One anuran (a juvenile Bufo ilium) was misidentified as Hyla sp. Ambystoma opacum is reduced to A, opacum-group because the other member of this form taxon (A, talpoideum) is indistinguishable from A, gpgggm_on vertebral characters used (Van Dam, 1978) and A, talpoideum currently occurs within 100km of the locality. The same conservative reasoning is applied to reduce Desmognathus sp. to Desmognathinae as Leurognathus marmoratus occurs in the vicinity. Baker Bluff Cave is a stratified local fauna ranging in age from 19,1001100 rcybp to 200 ybp (Guilday, et al.; 1978). The "mid-levels" (5 - 7ft) are contemporary with the other local faunas of this study, approximately 10,560 to 11,640 rcybp (Guilday, et al.; 1978). 78% of the Baker Bluff herptile taxa (21 of 27) are present in the mid-levels. Van Dam (1978) asserted that no discernible trends occur in the herpeto- fauna which reflect ecologic or climatic change through the past 20,000 years in northeastern Tennessee. Sympatry maps for the three units (Figure 4) support Van Dam's conclusions. Baker Bluff is distinctive among the five Appalachian local faunas in that only molluscs, anurans, and mammals are relatively high in frequency of occurrence. Bird and 22 .xpwpmuo_ mpmuwucw mmpucwu .Hanzoc ooH.mH ow oeo.HH peoH . NH m_m>mp cmzop Auv .Haaxuc oeo.HH op omm.oH wen . mg mpm>mpuuwa Amy .manxoc ocm.oHv Hem . ca m_m>mF swan: Amu em:_m Loxwm Low mama zepmqsxm .q mesmwd .¢ mesmwu 23 24 fish species are represented by only one or a few specimens per level. If any pattern of herpetofaunal change progressed through time, it is inadequately recorded. All but one of the herptiles still lives in the immediate area. Lampropeltis getulus has its nearest range limit 64km east of the locality. All the taxa are widely sympatric through the Appalachian Plateau west of Baker Bluff (Figure 4). Only a slight increase in effective precipitation compared to the present condition is indicated by this zoogeographic shift. Kingston Saltpeter Cave, Bartow County, Georgia cave fill altitude - 277m. latitude'— 34'N. longitude - 84°W. Notophthalmus viridescens Ambystoma maculatum Ambystoma opacum-group Ambystoma tigrinum Desmognathinae "Plethodontiini" Pseudotriton ruber Scaphiopus holbrOBki Bufo americanus Bufo woodhousei fowleri Bufo sp. Gastrophryne carolinensis Rana catesbeiana Rana CTamitans Rana utricularia Rana sp. Kinosternidae Chrysemys picta Terrapene cardlina Emydidae, undetermined (?Clemmys insculpta or Pseudemys concinna) (?Pseudemys concinna or PseUdemys scripta) Anolis carolinensis Sceloporus undulatus Nerodia sipedon Storeria dekayi Thamnophis sauritus 25 Thamnophis sirtalis Virginia striatula Diadophis punctatus Coluber or Masticophis Opheodrys aestivus Elaphe guttata ETaphe obsoleta Lampropeltis calligaster Lampropeltis getulus Lampropeltis triangulum Pituophis melanoleucus Agkistrodon contortrix identifications by Leslie P. Fay A preliminary report on Kingston Saltpeter Cave (Blair, Henry, and Sneed; 1981) includes a partial list of vertebrates recovered. The thirty-seven herptile taxa reported here are the first comprehensive listing from the local fauna. A complete study of the paleontology and history of the cave (Blair, Henry, Fay, Martin, Sneed, Sneed, Steadman, and Wilson; in prep.) is nearing completion. The age of the deposit is latest Wisconsinan based on terminal dates for extinct mammals (Meltzer and Mead, 1983). The herpetofauna is sympatric and currently occupies the northern half of Alabama and the vicinity of the locality in northern Georgia (Figure 5). Climatic requirements for the herptiles in the early post-glacial episode of the southern Appalachians were nearly equivalent to modern conditions. SYSTEMATIC DESCRIPTIONS In this section identification methods for each taxon are docu- mented. Listings include name, local fauna occurrence, and method of identification. — 00‘ u .c‘ u r..." 00’ a'?’;’ ° . ’1'. . ' :‘:-\.. . ’0’ ' ‘v \ I ..- . , t s ,. ,' ~ ’ 0.. JO ’0" o' \ f “a ‘ J "' '.\.a“‘-. o \. \ fl ' o \O \ r s \_. \ .‘0. g i z‘ .- ’ " .-.. ..".‘—"”$: -' “’ . I s Figure 5. Sympatry map for Kingston Saltpeter Cave fossil herpetofauna. Square indicates locality. 27 Class Amphibia Order Caudata Salamanders are identified by vertebral morphology. Generic or specific identity of limb and skull elements can often be determined also. Family Cryptobranchidae Cryptobranchus alleganiensis (Daudin) Hellbender Occurrence: BB Hellbender vertebrae are quite distinctive, the only similar North American form is Necturus. Large size serves to distinguish these two from other caudates. Van Dam (1978) noted that fragmentary Cryptobran- gflg§_vertebrae may be separated from Necturus by morphological details of the transverse processes and by greater sculpturing on the former. Family Proteidae Necturus maculosus (Rafinesque) Mudpuppy Occurrence: BB Necturus identification is discussed with Cryptobranchus. These two obligate aquatic salamanders were only recovered from Baker Bluff Cave, indicating a stronger stream faunal component than in the other four local faunas. Family Salamandridae Notophthalmus viridescens (Rafinesque) Eastern Newts Occurrence: NP4, NC, CC, KSC Newts are the only eastern North American caudates with opisthocoelous vertebrae and thick neural spines. Osteological criteria have not been devised to distinguish the terrestrial eft stage from the aquatic adult. 28 Family Ambystomatidae Members of this family may be grouped according to length-width ratios of trunk vertebrae (Tihen, 1958). Vertebral ratios allow sub- division of the genus Ambystoma into species groups, three of which occur in the central and southern Appalachian region. Unequivocal separation of species within each group by vertebral ratios has not been demonstrated. Therefore, for each locality, specimens of a species group are assigned specific names if only one member of the group currently lives in the area. Ambystoma maculatum (Shaw) Spotted Salamander Occurrence: BB, KSC A, maculatum is the only member of its species group living in the southern Appalachians and the fossil specimens are assigned on this basis. A, jeffersonianum has its southern limit 150km north of Baker Bluff. Ambystoma maculatum-group (A, jeffersonianum or A, maculatum) Occurrence: NP4, NC, CC Both species are currently present at these localities. On the Gaspé Peninsula, Quebec, where a different pair of A, maculatum-group species live, I have found two distinct size classes of fossil vertebrae and assigned specific epithets accordingly (Fay, 1984a). For the local faunas of this study, a continuum of sizes was found, so specific assignment was not attempted. Ambystoma opacum (Gravenhorst) Marbled Salamander Occurrence: CC Specific determination is by geographic range. A, talpoideum, the other member of the A, opacum group, does not range within 400km of 29 Clark's Cave today. Ambystoma opacum-group Occurrence: BB, KSC A, gpgggm_is now present at both localities. The continuous range of A, talpoideum is in the Gulf Coast and Mississippi Embayment, but it has several disjunct areas in the southern Appalachians, including a large one between Kingston and Baker Bluff, reaching to about 100km from each fossil locality. Ambystoma tigrinum (Green) Tiger Salamander Occurrence: NC, KSC A, tigrinum is the only member of the A, tigrinum-group that ranges into eastern North America. A, tigrinum is often distinguishable from other Ambystoma by the pronounced upsweep of the posterior portion of the neural arch (Holman, 1969). Particularly robust A, maculatum exhibit this feature as well, especially on anterior thoracic vertebrae (Van Dam, 1978), thus rendering the character diagnostic on quite large A, tigrinum only. Ambystoma sp. Mole Salamanders Occurrence: NP4, CC, BB Tihen's (1958) vertebral ratios do not provide mutually exclusive groupings. Vertebrae which fall within overlaps between species groups and damaged specimens that cannot be measured are placed here. Family Plethodontidae This family probably originated in the Appalachians and maintains its greatest species diversity in the region. Many species are difficult to distinguish when alive and may well prove impossible to separate as disarticulated fossils. As yet, no comprehensive study of plethodontid 3O osteology focusing on aspects necessary for fossil identifications has been published. The two subfamilies may be distinguished with relative ease using trunk vertebrae. Desmognathinae Occurrence: BB, KSC The dusky salamanders have occasionally been elevated to family status based largely on the characteristic opisthocoelous vertebrae (Soler, 1950). There are twelve taxa distributed east of the Mississippi River, nearly all of which overlap in both size and geographic range. The sub- familial term is used for southern Appalachian local faunas where the two genera, Desmognathus and Leurognathus, presently occur. Desmognathus sp. Dusky Salamanders Occurrence: NP4, NC, CC I Members of this genus are the only desmognathines currently distributed in the central Appalachians. Plethodontinae This subfamily includes all non-desmognathine plethodontids. Two genera, Gyrinophilus and Pseudotriton, have distinctive "false opistho- coelous" vertebrae (Soler, 1950). All remaining eastern North American forms are here placed under "Plethodontiini" for convenience. This term is used for morphological separation of vertebrae only and does not imply evolutionary distance among plethodontines. The twenty-eight subspecies in six genera have not been thoroughly examined for distinc- tive osteological morphology. Size and geographic range have been used for identifications (e.g. Holman, 1967; Lundelius, 1967; Richmond, 1964; Richmond in Guilday, 1964) but, as with desmognathines, a great deal of overlap exists. I have chosen not to pursue identification within this 31 artificial grouping at this time. "Plethodontiini" (plethodontines except Gyrinophilus and Pseudotriton) Occurrence: NP4, NC, CC, KSC See plethodontine discussion. Gyringphilus porphyriticus (Green) Spring Salamander Occurrence: CC 9; porphyriticus vertebrae have poorly developed to obsolete neural spines, whereas Pseudotriton have relatively prominent spines (Holman, 1967). A cave-dwelling congener, A, palleucus, is removed from consid- eration because it occurs in scattered localities in the southern Appalachians, 500km from Clark's Cave. Pseudotriton ruber (Sonnini) Red Salamander Occurrence: KSC Distinguishing characters of Pseudotriton and Gyrinophilus are discussed with the latter. Holman (1967) has distinguished between the two species of Pseudotriton by higher neural spines in E, ruber than in E, montanus. The Kingston fossil vertebrae also agree with E, ruber in this character. Pseudotriton sp. Mud and Red Salamanders Occurrence: CC ’ The Clark's Cave Pseudotriton material did not match either neural spine condition described by Holman (1967) sufficiently to warrant specific assignment. Order Anura Toads and frogs are most readily identified by morphology of the distal portion of the ilium. Sacral vertebrae and many cranial elements are also diagnostic to the familial or generic level. 32 Family Pelobatidae Scaphiopus holbrooki (Harlan) Eastern Spadefoot Toad Occurrence: NC, CC, KSC The only member of the family known to occur east of the Mississippi River, the eastern spadefoot is recognized by a fused urostyle-sacral vertebra. It is also distinguished from bufonids by greater ornamenta- tion on skull elements and the presence of maxillary teeth. Family Bufonidae Bufonid ilia have a variously developed ilial prominence with a dorsal protuberance which may or may not exceed the prominence in height (Tihen; 1962a, 1962b). Bufo americanus Holbrook American Toad Occurrence: NP4, NC, CC, 88, KSC B, americanus may be distinguished from Bufo woodhousei fowleri, the only other upland toad east of the Mississippi, by the broader base of the ilial prominence (viewed laterally) of B, americanus (Holman, 1967; Wilson, 1975). The dorsal protuberance is oriented antero-distally in large, mature B, americanus, but perpendicular to the longitudinal axis of the ilial shaft in B, E: fowleri. A, americanus ilia which exceed the modern size range for the species have been reported from Ladds Quarry (Holman, 1967; Wilson, 1975). Kingston Saltpeter Cave, located about 13km from Ladds, yielded the largest ilia encountered in this study. The largest Kingston ilium is just under one-half the size of the largest Ladds ilium. The larger phenon apparently disappeared from northwestern Georgia after the Wisconsinan full-glacial episode. Bgfg_ americanus copei was identified in the lower levels of the New Paris No. 4 section by Richmond (in Guilday, et al.; 1964). The subspecific 33 determination was based on small size and association with boreal mamals. II repeat this listing without further study at this time, although the identification is doubtful. [_3, americanus ilia are extremely variable iti the morphological details used for identification and a comprehensive examination of both modern and fossil specimens throughout the distri- tnttion is needed. 8, g, charlesmithi is also a small subspecies which does not live near New Paris today, although it is in the southwestern portion of the range, rather than the northern portion, where B. a, _cgpgl occurs. Size and co-occurrence with extra-limital mamals are insufficient criteria fer subspecific identification of the New Paris No. 4»§, americanus. Perhaps the ilia represent small, but mature B, Q, americanus, the zoogeographic implications of which would match the rest of the herpetofauna. Bufo woodhousei fowleri Girard Fowler's Toad Occurrence: NP4, NC, CC, BB, KSC See A, americanus remarks for specific determination. This is one of a very few eastern North American taxa that can be confidently identified to subspecies based on osteologic characters. The height of the ilial prominence ranges from 30% to 45% of the width of the prominence at its base, while in the western subspecies, B, E: woodhousei, this ratio is 45% or greater (Tihen, 1962b). Family Hylidae Hyla crucifer (Wied) Spring Peeper Occurrence: NP4, CC The New Paris No. 4 hylids were listed as Aylg_cf. crucifer by Richmond (in Guilday, et al.; 1964) and securely placed in A, crucifer by Lynch . (1966). Lynch (1966) reported that the New Paris material represents 34 individuals larger than known modern specimens in snout-vent length by 5mm, while fossils from Cumberland Cave, Maryland (Irvingtonian) were within the size range of living peepers. The Clark's Cave individuals are also within the modern size range. A, crucifer is the only Hylg_ that normally has the dorsal protuberance anterior to the anterior margin of the acetabular fossa (Lynch, 1966). Hyla ?chrysoscelis/versicolor Grey Treefrogs Occurrence: NC Grey treefrogs are distinguished by call rate, but no osteological characters have been defined that separate the two. A single ilium, listed as Ayla sp. by Richmond (in Guilday, 1962), resembles the grey treefrogs. The dorsal protuberance is more symmetrically rounded in the fossil than in comparative material at hand, but the ilium does not resemble that of any other North American hylid more closely. Family Microhylidae Gastrophryne carolinensis (Holbrook) Eastern Narrow-mouthed Toad Occurrence: KSC Gastrophryne has a distinctively rectangular acetabular expansion (Wilson, 1975) and is one of the smallest North American anurans. Family Ranidae Ranid ilia have a pronounced dorsal blade running the length of the ilium. With the exception of coastal plain species, seven Aggg_ currently inhabit eastern North America and all but A, septentrionalis have fossil records. McGinnis (in Guilday, et al.; 1977) lists charac- ters that distinguish A, catesbeiana, A, clamitans, A, pipiens, and A, palustris based on ilial morphology. 35 Rana catesbeiana Shaw Bullfrog Occurrence: NC, CC, 88, KSC In addition to the characters cited by McGinnis (in Guilday, et al.; 1977), A, catesbeiana can be separated from A, clamitans by a more rugged (Holman, 1967) and often slightly concave (Fay, 1984a) vastus prominence. Rana clamitans Latreille Bronze and Green Frogs Occurrence: NC, CC, KSC Distinguishing characters of A, clamitans are discussed with A, catesbeiana. Rana palustris Le Conte Pickerel Frog Occurrence: NC, CC A, palustris ilia are similar to thOse of A, pipiens and A, sylvatica. A, palustris and A, pipiens differ from A, sylvatica ilia by a larger, less produced, and less roughened prominence for the origin of the vastus externus head of the triceps femoralis muscle (Holman, 1965) in the former (Holman, 1967). A, palustris has a less steep slope of the posterodorsal border of the ilial shaft that does A, pipiens. I retain Richmond's (in Guilday, 1962) tentative identification for the Natural Chimneys specimens because this character is somewhat variable between the two species. Rana pipiens-group Leopard Frogs No method has been devised that adequately separates ilia of the leopard frog complex (Holman, 1977a). This group consists of four para- patric species (A, berlandieri, A, blairi, A, pipiens, and A, utricu- laria), which are distinguished by soft anatomy and mating calls (Conant, 1975). I make specific assignments within the leopard frog complex when 36 «only one A, pipiens-group species currently occurs at the fossil locality. If a species boundary is near the locality, or if no member of the complex now occupies the area, I refer ilia to Agflg_cf. A, pipiens (Fay, 1984a). Aggg_cf. A, pipiens Occurrence: NP4, NC, CC See Rana pipiens-group remarks. Rana utricularia Garman Southern Leopard Frog Occurrence: KSC See Rana pipiens-group remarks. Rana sylvatica Le Conte Wood Frog Occurrence: NP4, NC, CC, 88 Identification of A. sylvatica is discussed with A, palustris. Class Reptilia Order Testudines Comprehensive guides or keys for the identification of turtle bones do not exist. A complete set of modern comparative specimens from the region under investigation is necessary for accurate determination of fossil specimens. Family Chelydridae Chelydra serpentina (Linné) Common Snapping Turtle Occurrence: NC Holman (1966) describes unique characters of A, serpentina procoracoid- scapulae with which the Natural Chimneys fossil agress. Family Kinosternidae Occurrence: KSC Three members of this family presently occur near Kingston. The fossil 37 elements (humerus and procoracoid-scapula) are not sufficiently diag- nostic for generic identification. Family Emydidae Chrysemys picta (Schneider) Painted Turtle Occurrence: KSC One left peripheral is assigned to A, pigtg_based on pattern of scute impressions, shape, and size. Graptemys geographica (Le Sueur) Map Turtle Occurrence: CC, 88 One carapace fragment from each locality is identified as A, geographica. Van Dam's (1978) criteria were also used for the Clark's Cave fossil. Terrapene carolina (Linné) Box Turtle Occurrence: NC, KSC The fossil material is indistinguishable from I, carolina, but not suf- ficiently distinct for subspecific assignment. Emydidae, undetermined ?Clemmys insculpta (Le Conte) or Pseudemys concinna (Le Conte) Occurrence: KSC One large, eroded proximal left humerus is similar to both A, insculpta and A, concinna, but not distinguishable. Both turtles are present in the nearby Ladds Quarry local fauna (Holman, 1967). ?Pseudemys concinna (Le Conte) or Pseudemys scripta (Schoepff) Occurrence: KSC One eroded right pubis resembles these two Pseudemys, but is too frag- mentary for specific determination. 38 Order Squamata Suborder Lacertilia Lizards, like turtles, have not been subjected to osteological study for the purpose of fossil identification technique. A comparative collection is necessary. There are relatively few lizards in the Appala- chian region, but many are closely related and difficult to separate on osteological characters. Family Iguanidae Anolis carolinensis Voigt Green Anole Occurrence: KSC Holman (1967) lists characters which identify A, carolinensis dentaries. Sceloporus undulatus (Latreille) Fence Lizard Occurrence: NC, CC, KSC . The Natural Chimneys (Richmond, in Guilday, 1962) and Clark's Cave (Guilday, et al.; 1977) specimens were originally listed as Sceloporus cf. undulatus. Along with the Kingston material, I find all the den- taries indistinguishable from §, 2: hyacinthinus, as did Holman (1967) for specimens from Ladds Quarry local fauna. Family Scincidae Eumeces fasciatus (Linné) Five-lined Skink Occurrence: BB Van Dam (1978) identified skink vertebrae as A, fasciatus on the basis of more upswept neural spines than in A, laticeps. Eumeces ?laticeps (Schneider) Broad-headed Skink Occurrence: CC Guilday (Guilday, et al.; 1977) based this identification on direct comparison with modern specimens. 39 Scincella lateralis (Say) Ground Skink Occurrence: NC A sacrum, distinguishable from other skinks by fused condition and minute size, is the first fossil record of this species. Suborder Serpentes Snake vertebrae present an intricate exercise in identification. A large comparative collection is helpful, for the vertebrae are not only morphologically diverse, but variable as well, both within a taxon and within an individual. Adequate criteria are unknown for identifying anterior trunk and caudal vertebrae. The former have hypopophyses, large neural canals, and are generally short and high in overall shape. The latter have paired ventrolateral processes. Only mid-trunk vertebrae are used for identification in this study. Simple morphological criteria often yield artificial groups. For example, water snakes (Colubridae) and rattlesnakes (Viperidae) both have hypopophyses on the trunk vertebrae. Within a subfamily, some species of one genus may seem more closely allied to another genus (Elaphe and Lampropeltis; Auffenberg, 1963). The guidelines offered by Holman (1981) are used in this study to separate subfamilial groups. Vertebral ratios pioneered by Auffenberg (1963) and expanded by Meylan (1982) are utilized along with qualitative characters from the litera- ture and my own experience. Family Colubridae Colubrids comprise three morphological groups: (1) those with laterally compressed hypopophyses less than one-third the width of the condyle (Holman, 1981), the Natricinae; (2) those with essentially no hypopophyses or haemal keels, the Xenodontinae; (3) all other North 40 American forms (subfamilies Colubrinae and Lampropeltinae in this report), which lack hypopophyses but have variously developed haemal keels. Subfamily Natricinae This subfamily may be further divided on the basis of vertebral size. Seminatrix, Storeria, Tropidoclonion, and Virginia vertebrae are less than 3mm CL. Two species of Storeria and two of Virginia inhabit the Appalachian region. Larger vertebrae may be separated according to length-width ratios. Thamnophis vertebrae have a CL/NAW ratio greater than 1.75. For Clonophis, Nerodia, and Aggjgg, this ratio is usually less than or equal to 1.5. Nerodia erythrogaster (Forster) Plain-bellied Water Snake Occurrence: NC The neural spine of this species is as long as high (Holman, 1962; Meylan, 1982). In other living members of the genus, the spine is higher than long (A, cyclopion, A, rhombifera, A, taxispilota) or shorter than long (A, fasciata, N. sipedon) (Holman, 1962; Meylan, 1982). Although now widely distributed in southeastern North America, this is the first report of fossil A, erythrogaster east of the Mississippi River and north of Florida. Nerodia sipedon (Linné) Common Water Snake Occurrence: NP4, NC, CC, 88, KSC A, sipedon vertebrae have lower neural spines than other members of the genus (Holman; 1962, 1967) ?Regina septemvittata (Say) Queen Snake Occurrence: NP4, CC Most members of the genus Regina have vertebral characters intermediate 41 between my "small" and "large" forms. Auffenberg (1963) developed similar subgroups of the Natricinae. According to my criteria of verte- bral ratios and morphological characters, A, septemvittata falls very near the genus Thamnophis. The former has abbreviated neural spines (Auffenberg, 1963) and blunt accessory processes, but is otherwise difficult to separate from the latter based on vertebral ratios of the comparative material available to me. Thus the identifications are provisional at this time. If valid, these are the first fossil records of A, septemvittata, although Richmond (unpublished) tentatively identi- fied one of the New Paris No. 4 specimens. Storeria dekayi (Holbrook) Brown Snake Occurrence: NP4, CC, KSC Storeria may be confused only with Virginia or juvenile specimens of larger natricines. Immature vertebrae have large neural canals, large condyles, and a generally delicate appearance. Storeria is distinguished from Virginia by the ratio of NSL/ZW (Auffenberg, 1963). Storeria has average values greater than 1.10, Virginia less than 1.05. §, ggAgxi can be very difficult to distinguish from §, occipitomaculata. Several ratios in combination have proven useful, but by no means infallible (Table 2). Table 2. Three ratios used to distinguish Storeria dekayi from Storeria occipitomaculata. POPR/CL NSL/ZW CL/NAW Storeria dekayi, >1.08 <1.17 <1.91 Storeria occipitomaculata <1.07 >1.17 >1.94 42 Auffenberg (1963) found that §, ggAgyj_has a "ground off" or flattened dorsal surface on the neural spine. This character was present on less than half the specimens assigned to §, ggAgyj_by vertebral ratios. Storeria occipitomaculata (Storer) Red-bellied Snake Occurrence: NP4, NC, CC Distinguishing features are discussed with §, 925311, The Natural Chimneys fossils fall between the values of the two Storeria, but nearer to §, occipitomaculata. Storeria sp. Brown Snakes Occurrence: NP4, NC Vertebral measurement ratios of these specimens do not allow placement in either species. Thamnophis sauritus (Linné) Eastern Ribbon Snake Occurrence: NP4, NC, CC, 88, KSC Ribbon snakes (I, proximus and I, sauritus) have accessory processes directed at an oblique angle to longitudinal axis of the vertebra (Holman, 1962). A method to separate the two ribbon snakes by osteo- logical criteria has yet to be found, so assignment is customarily based on geographic range (Holman, 1981). Thamnophis sirtalis (Linné) Common Garter Snake Occurrence: NP4, NC, CC, 88, KSC Accessory processes on I, sirtalis vertebrae are at right angles to the longitudinal axis of the centrum (Holman, 1962). Virginia striatula (Linné) Rough Earth Snake Occurrence: KSC Vertebral ratios serve to distinguish A, striatula from A, valeriae (Table 3). 43 Table 3. Three ratios used to distinguish Virginia striatula from Virginia valeriae. POPR/NAW CL/NAW NSL/ZW Virginia striatula >2.00 >1.77 >O.97 Virginia valeriae <1.75 <1.55 mo Nfim Nam Aom Ace Rem Amm tapaap_mm cepmmcwx m m>w I 5.: Nam Awm NHH xo Amm Aom a>mupcczflmupaxmm Ame Nmm smm Am Ase Rm“ a>~u m.xcapu Nmm Am“ Amm ARM ewe Ame msacewgu Fataumz P: Amm Amm Ac Ac Ace Aom a .0“ MWLwa 3mz Peach mmucwmemm mwpwpemumg mmcwvzflmmp mc=:¢. mumvzmu .xpmpmuop Pwmmom we Exom smear: mcwmcmc apucmggzu mxmu op axe» Pwmmoe eo mmmpcmogma mm ummmmeaxm mum; Aagm>oumc _wmmomv mcwpasmm uPEocoxMA .e mpnmp 50 largest quantity of fossil herptile material and Baker Bluff Cave the smallest. Turtles have the poorest representation among herptiles. None were recovered from New Paris No. 4 or Baker Bluff Cave mid-levels. Richmond (in Guilday, et al.; 1964) stated that the boreal climate of New Paris No. 4 time precluded the survival of these reptiles at the locality, citing Terrapene carolina from the Holocene New Paris No. 2 & No. 3 to indicate that turtles can be preserved in similar situations. An alternate explanation is that turtles were present at the time of New Paris No. 4, but were not as likely to become fossilized as other vertebrates. Natural Chimneys fossil turtles represent two species, Clark's Cave has one. All three records are based on one carapace fragment each. The Kingston turtle material, composed of shell, bone, limb, vertebral, and skull elements, includes at least four species. Although more fossil species are represented at Kingston than the other localities, northern Georgia today has fewer turtle species (ten) than the Baker Bluff area (twelve) or the Clark's Cave-Natural Chimneys area (thirteen). New Paris is within the range of ten turtle species at present. Lizards are also low in number and diversity, with no more than 33% of the present fauna preserved at any locality. However, lizard species are not numerous in the modern fauna, with only four species in Pennsylvania near New Paris, increasing to ten near Kingston in Georgia. Low diversity and relatively fragile skeletal elements may explain the scarcity of fossil lizards. Snakes are the best represented reptiles and, at all localities except Baker Bluff and Kingston, the best represented herptiles. Each 51 taxon was identified from vertebrae, but at several localities, espe- cially New Paris No. 4, a large number of cranial elements were recovered as well. Fewer anurans than caudates were found at all localities except Natural Chimneys. This is partially due to a dramatically low number of fossil hylid frogs compared to all other amphibians. No hylids are recorded from Kingston or Baker Bluff, only Hyla crucifer at Clark's Cave and New Paris No. 4, and Hyla chrysoscelis/versicolor at Natural Chimneys. Hylids are the most secretive anurans in the region and are usually conspicuous only during the breeding season. Caudates, like snakes, are usually identified by vertebrae. These elements are the most compact and therefore probably most durable bones. Salamander representation in the local faunas is underemphasized because of the lumping of four genera into one form taxon. New Paris No. 4 and Baker Bluff are fissure fills which may have served as animal traps (Guilday, et al., 1964; Guilday, et al., 1978), while the Natural Chimneys and Clark's Cave (Guilday, 1962; Guilday, et al.,; 1977) and perhaps Baker Bluff Cave (Guilday, et al., 1978) and Kingston materials were accumulated by the activities of raptorial birds. Active predators can be expected to sample a larger portion of the fauna than a physical trap. It is likely that any cavity open to the surface may become the repository for both types of samples and the resident subterranean fauna as well. Pockets of snake vertebrae are often encountered in cave and fissure accumulations (Guilday, et al., 1964; Foley, 1982; Holman, personal communication, April 1984), resulting from den mortality during hibernation. The predominance of crotalid and natricine vertebrae over other herptile bones at New Paris No. 4 52 (Guilday, et al.; 1964) indicates that these snakes denned in the fis- sure. A hibernaculum has been proposed for a Holocene fossil fauna dominated by Bufo americanus remains (Fay, 1984a). DISCUSSION Only one of the sixty-two herptiles represented as fossils does not live in the Appalachians at present. The Hudson Bay Toad (AAAQ_ americanus copei) was identified from the lower levels of New Paris No. 4 by Richmond (in Guilday, et al.; 1964). The identification is very tenuous (see Systematic Descriptions), leaving the zoogeographic impli- cations in doubt, also. If this identification is valid, it represents the most striking range shift of any eastern North American herptile and the only notable one for an amphibian. Why is only one (tenuously identified) strictly boreal herptile found in association with a boreal mammalian local fauna? Modern ranges may provide the answer, especially if interglacial distributions are assumed similar so that the Sangamonian pattern may be considered analogous to the Holocene (as yet, there are an insufficient number of pre-Late Wisconsinan herpetofaunas in eastern North America to support or reject the assumption). In the present interglacial, there are seven herptile forms restricted to the area bounded by Hudson Bay and the Great Lakes on the west and on the south by the latitude of central Pennsyl- vania (the northern limit of the central Appalachians as defined here). If these animals were in the same region during the Sangamonian Stage, the subsequent development of Wisconsinan ice forced them southward, presumably into the central Appalachians or the Appalachian Plateau. 53 A fourteen percent recovery rate (one of seven apparently displaced species) is low, tnnznot unusually so. Of the seven northerly distri- buted forms, three have never been reported as fossils (Thamnophis brachystoma, Ambystoma tremblayi, Rana septentrionalis) and four (Tham- nophis brachystoma, Ambystoma tremblayi, A, laterale, Bufo americanus 59231) are difficult or impossible to distinguish from more widely distributed congeners on osteological grounds. Fourteen mammals from the same region, all with fossil records, do not range further south in the Appalachians than central Pennsylvania (Fay, 1984b). Eastern North American boreal mammals are more likely to be recovered than herptiles from the same region. Seven species present in the Natural Chimneys and Clark's Cave local faunas (see Faunal Lists) no longer occur in the central Appala- chians of West Virginia and Virginia, although they range farther north than the latitude of the two localities to the east or west. The envir- onment of the central Appalachians has seemingly become unsuitable for these animals since the end of the Wisconsinan. Older herpetofaunas of the region also contain extirpated forms. Three Irvingtonian local faunas (Cumberland Cave, Allegany County, Maryland [Holman, 1977a]; Trout Cave-Lower Level, Pendleton County, West Virginia [Holman, 1982]; and Hanover Quarry Fissure, Adams County, Pennsylvania [Fay, unpub- lished]) include herptiles that have retracted their ranges to the north, west, or east outside the central Appalachians. This may suggest a gradual shrinking of species distribution throughout the Quaternary, as one of the species (Elaphe vulpina) is also known from the full-glacial Wisconsinan of Georgia (Holman; personal communication, October, 1983) and the Pliocene of Idaho (Holman, 1968) as well as several other now 54 extra-limital localities of later Cenozoic age (see Holman; 1979, 1981 for localities). Some eastern North American endemic herptiles, such as Clemmxs muhlenbergi, Eumeces 2: anthracinus, Pseudotriton montanus, and Scaphiopus holbrooki, have disjunct Appalachian distributions. Two possibilities would explain both these discontinuous ranges and the extirpation of taxa mentioned above. Physical and biological factors of central and southern Appalachian habitats may have been altered beyond the tolerances of these animals, or current distributions may simply be inadequately known and the animals actually occupy the entire region in suitable situations. The latter is quite likely for Clemmys muhlenbergi (Bury, 1979) and perhaps Rana pipiens (McCoy, 1982). Late Wisconsinan herpetofaunas of this study do not show the dramatic Wisconsinan-Holocene range adjustments of birds and mammals from the same local faunas. Only one toad (tenuous identification) may have left the region and no herptiles from these localities are now extinct. Twenty-eight birds and mammals from the combined faunal lists are now extirpated from the the fossil localities and seven species of mammals and one bird are extinct. In 1965, it was possible to state "The fluctuating ecological conditions during the Pleistocene must have modified the distribution of reptiles more than that of mammals or birds living in the same geographic area." (Auffenberg and Milstead, 1965:557). With the increasing number of herpetofaunas now available, this state- ment is no longer tenable, at least for eastern North America. Even with the evidence at hand, herptiles that do not match the bird and mammal reconstructions are dismissed in the current literature; "...the (Clark's Cave) herpetofauna is similar to that of the area today, with 55 several species...that would have been AggAlg_Ag_survive the "boreal" conditions postulated for the area on the basis of the mammalian fossil fauna." (Lundelius, et al.; l983:3l6 & 324; emphasis added). There is little evidence for the implication that herptiles are heterochronic intrusions into the fossil accumulations or that temperate herptiles are ecologically incompatible with "boreal" birds and mammals. In fact, the "boreal" nature of these mammalian local faunas may have been over-emphasized. Although many species represented by fossils now range farther north than the localities, no more than l8 percent of the mammal species (at New Paris No. 4) are restricted to latitudes north of the fossil localities. A number of alternative explanations, several already alluded to in this study, may be advanced based on evolutionary, zoogeographic, and ecologic stability among Quaternary Appalachian herptiles. (I) Herp- tiles did not respond as strongly to glacial-interglacial climatic changes as did other vertebrates. Herptiles have a marked capacity for acclimitization to seasonal climatic change, which may also provide the ability to adapt to progressive, long term change by alterations in physiology or behavior rather than distribution. (2) There are fewer eastern North American boreal herptiles than mammals to be displaced by glaciation. (3) It is apparent from the discussion of plethodontid salamanders that several herptile genera and many species may go unre- cognized in the fossil record due to the similarity of their bones. (4) Much of the herpetofauna of the Appalachians is endemic, reflecting both probable center of origin and restricted microhabitat requirements for many of the species. In other words, the group has endured in suitable local environments since first evolving, in spite of changes 56 in climate and biota. The Appalachian region, once thought to be the refugium for the entire temperate biota (Braun, 1950), held individual species or enclaves of plants (Davis, 1983). Perhaps the region also shielded the herpetofauna in whole or in part from the rigors of glacial climate. The most popular hypothesis to explain the admixture in Quaternary fossil accumulations of faunal elements that are not currently sympatric is that of increased climatic equability compared to the strongly continental climate prevailing over much of North America now. Milder winters are implied, allowing southerly distributed species to expand northward, as well as cooler summers that permitted northerly distri- buted species to disperse southward (Holman, 1976 and references therein). Although giant tortoises were a stimulus to the development of the hypothesis (Hibbard, 1960), only a few attempts have been made to evaluate climatic equability in regard to the Quaternary herpeto- faunal record (Holman; 1976, 1977b, 1980). Rhodes (1982) asserts that it is unnecessary to invoke climatic equability for local faunas containing more boreal, fewer temperate- mesic, and no more-southern forms than occupy the locality today. Increased microclimatic contrast between different slope exposures may have been sufficient to allow climatic amelioration. No "southern" mammalian element, which would indicate milder winters, occurs in the Wisconsinan Appalachian local faunas, although areas of herpetofaunal sympatries are south of the central Appalachian localities. The extinct mammals Dasypus bellus, Aylohyus nasutus, Platygonus compressus, and Tapirus veroensis were descended from tropical or warm-temperate 57 lines, but are of dubious value to indicate equability (Rhodes, 1982). The herpetofaunas also show no true "southern" (i.e. austroriparian) and perhaps one (tenuously identified) boreal form. The only published record of such an association for the Appalachian region is from Ladds Quarry, where Geochelone cf. A, crassus and Bufo terrestris occur with Clemmys insculpta (Holman, 1967, 1980; Wilson, 1975) and Elaphe vulpina (Holman; personal communication, October 1983). Ladds Quarry local fauna dates from the full glacial (17,520i630; Stuckenrath and Miekle, 1980) about six thousand years before the late-glacial local faunas of this study. If climatic equability was common in eastern North America, it had decayed into the present continental pattern before the end of the Wisconsinan Stage. The faunal gradient concept (Guilday, et al.; 1978) expresses the magnitude of change in composition of mammalian faunas along a north- east to southwest transect in the Appalachians. In other words, the number of animals with range limits along the transect may be shown as a percentage of all animals occurring along the transect. Percentages can be compared between modern and fossil transects to illustrate relative "steepness" of the faunal gradient through time. Ten small rodent species occur today in southern Pennsylvania at the north end of a transect, twelve at the south end in northwestern Tennessee (Hall, 1981). Two species present in Pennsylvania (Clethrio- nomys gapperi and Napeozapus insignis) do not range south as far as northeastern Tennessee at the elevation of Baker Bluff Cave, while four of the Tennessee rodents (Ochrotomys nuttalli, Oryzomyg palustris, Reithrodontomys humilis, and Sigmodon hispidus) do not reach the north end of the transect in Pennsylvania. This means six of the fourteen 58 small rodents (43%) have range limits along the transect (Guilday, et al. 1978). For Late Wisconsinan local faunas, one of the fifteen rodents at New Paris No. 4 (Dicrostonyx hudsonius) did not reach Baker Bluff, while all the Baker Bluff rodents also occurred at New Paris. This indicates a steeper faunal gradient at present than during the Late Wisconsinan (43% versus 7%) for small rodents (Guilday, et al.; 1978). For comparison with Guilday's rodent gradient, I chose snakes because they have the most diverse fossil record of herptile groups in this study and many snakes and rodents have a close predator-prey relationship. The snake gradient is extended to northern Georgia (Kingston). Twenty species currently occur within 50km of New Paris, twenty-eight near Kingston. Two snakes at New Paris (Clonophis kirt- landi and Opheodrys vernalis) do not range as far south as Kingston, while ten species from northern Georgia (Agkistrodon contortrix, Cemo- phora coccinea, Lampropeltis calligaster, Masticophis flagellum, Nerodia erythrogaster, A, taxispilota, Pituophis melanoleucus, Sistrurus miliaris, Tantilla coronata, and Virginia striatula), do not reach southern Penn- sylvania today. Twelve of the thirty snake species that occur along the gradient have range limits between the endpoints (40%). Five of the nineteen fossil species at New Paris No. 4 are not found in the Kingston list, and one of the fifteen fossil snakes from Kingston is not recorded from New Paris. Six of twenty snake species along the gradient had apparent Late Wisconsinan range terminations between New Paris and Kingston (30%). Compared to the small rodent faunal gradient, the tread through time for snakes from Late Wisconsinan to present is the same but less steep. That is, more small rodents and snakes came to have range terminations 59 in the central and southern Appalachians since the Late Wisconsinan, but the results are far more dramatic for rodents. This presentation is obviously too simplistic. Either real differences exist in the manner in which rodent and snake distributions expand and contract in response to climatic change, or unidentified zoogeographic characteristics of modern vertebrates and/or taphonomic disparities among groups of fossil vertebrates render comparisons based on faunal gradients invalid. Range terminations between northern and southern fossil localities may be evaluated in a more rudimentary fashion by illustrating the number of fossil taxa with modern distribution limits to the north or south of the transect endpoints (Figure 6). A much greater number of mammals than herptiles do not reach the southernmost locality (in this example, New Paris is the northern point, Baker Bluff the southern). Rana sylvatica is represented for each local fauna, although Baker Bluff is the only locality beyond its range. Bufo americanus copei is included for New Paris No. 4 despite its dubious presence. Opheodrys vernalis, recorded as a fossil at Clark's Cave, reaches the southern Appalachians as several disjunct outliers. More herptiles than mammals fail to reach the north end of the transect. One bat, Myotis grisescens, is the only mammal of "southern" affinities repre- sented. Also included in Figure 6 are mammals now occurring west of the transect (SEermophilus tridecemlineatus at all localities, plus Taxidea taxus at Baker Bluff) and extinct mammals. No herptiles have withdrawn to the west since the Late Wisconsinan and, as mentioned before, no Late Wisconsinan herptile extinctions are known from the Appalachians. 6O .pcmwnmem chame cmpgumpoaa< m.zoupwaw acopm macaw» quop cmcwmzoumwz mum; Lao» mo mmpwuacm; can mmeEme memow com meowumcwsgop omen; sconce we Ememoumm: .o mesa?“ ta; 5:: :3 misc: ”322:; :53: ed: 2:: >3: .223... . 35:0... 303.! . 5:0..m a..oe.z no. one: _o no.- So! 61 The few herptiles not ubiquitous along the fossil transect indicate incomplete sampling (either by paleontologists or by the agent of accumu- lation) or very minor episodes of local extirpation. The boreal mammal component is indicative of dislocation of that biota due to the exist- ence of glacial ice to the north. Speculations for the lack of "boreal" herptiles have been presented earlier in this report. Watts (1973, 1979) and Delcourt and Delcourt (1981, 1983) have reconstructed late Quaternary vegetational dynamics of eastern North America in terms of periods of community stability and instability. This approach emphasizes the intervals between major events (i.e. glacial maximum, Climatic Optimum, etc.). A transect at 85°W. longitude across eastern North America from the Gulf Coast to the Great Lakes tracks the shifting boundaries of vegetational communities (Delcourt and Delcourt, 1983). From 34°N. to 37°N. latitude a stable boreal forest existed from about 19,000 to 16,500 ybp during the full-glacial episode. As the Wisconsinan ice front retreated, the vegetation changed to mixed conif- erous-deciduous forest (16,500 to 12,500 ybp). Species composition remained in flux after boreal elements were eliminated (12,500 to 9000 ybp). In the latest Wisconsinan-earliest Holocene interval, the decid- uous forest in the region containing the five herpetofaunas reached its highest arboreal species richness during relatively cool, equable condi- tions (Delcourt, 1979). While mammalian faunal composition altered to adjust from a full- glacial boreal environment with coniferous forest, through over 7000 years of gradual but persistent vegetational and climatic change, to a post-glacial temperate environment with deciduous forest, the herpeto- fauna seems to have remained relatively undisturbed. The vegetational 62 instability and climatic change were apparently within the tolerance limits of most reptiles and amphibians. CONCLUSIONS A period of climatic equability with disharmonious herpetofaunas seemingly ended with the Wisconsinan fall-glacial episode before 11,000 ybp in the Appalachians. By the Late Wisconsinan-early Holocene inter- val, herpetofaunas were already stabilized in the structure maintained to the present. A few species have been extirpated from areas of the central Appalachians, probably by loss of suitable habitat through edaphic rather than climatic change. These Late Wisconsinan herpeto- faunas neither support nor contradict the climatic equability hypothe- sis, as almost no extralimital records previously reported are valid. Herptile faunal gradients are similar to, but much less dramatic than, those of mammals. This indicates that the concept may not be equally applicable to the various groups of terrestrial vertebrates. Reptiles and amphibians of the southern and central Appalachians seem to have endured the climatic and biotic alterations brought on by the ending of the most recent glaciation with only minor range adjustments. SUMMARY Most fossil herpetofaunas from eastern North America do not exhibit the dramatic Wisconsinan-Holocene range adjustments characteristic of associated birds and mammals. Mammal and bird lists consist of current residents of the fossil localities plus a sizable component of taxa that 63 range to the north or west. Herptile lists are composed of area resi- dents with only a very few now-extirpated forms. Factors responsible for the varying responses to climatic change among vertebrate classes have not yet been adequately identified. Although the ranges of other animals and plants were altered, thus changing the position and compo- sition of communities, herptiles in eastern North America were relatively unaffected by physical and biotic changes of the last 11,000 years. REFERENCES CITED REFERENCES CITED Auffenberg, Walter, 1963. The fossil snakes of Florida; Tulane Studies 2001., 10(3):131-216. and William W. Milstead, 1965. Reptiles in the Quaternary of North America; in The Quaternary of the United States, H. E. Wright and D. G. Frey (eds.)§ Princeton Univ. Press, PP. 557L568. Berkland, James O. and Loren A. Raymond, 1973. Pleistocene glaciation in the Blue Ridge Province, southern Appalachian Mountains, North Carolina; Science, 181:651-653. Blair, Larry D, James H. Henry, and Joel M. Sneed, 1981. A preliminary report on the study of Kingston Saltpeter Cave, Georgia; (abs.), Georgia Jour. Sci., 39(2):66. Braun, E. Lucy, 1950. Deciduous Forests of Eastern North America; Hafner Press, New York (1974 reprint), 596p. Brunnschweiler, Dieter H., 1962. The periglacial realm in North America during the Wisconsin glaciation; Biuletyn Periglacjalny, no. 11: 15-27. Bury, R. Bruce, 1979. Review of the ecology and conservation of the Bog Turtle, Clemmys muhlenbergii; U.S. Dept. Interior, Fish & Wildlife Service, Spec. Sci. Rpt.--Wildlife No. 219, 9p. Clark, G. Michael, 1968. Sorted patterned ground: new Appalachian localities south of the glacial border; Science, 161:355-356. Conant, Roger, 1975. A field guide to reptiles and amphibians of eastern and centrET’North America; Houghton Mifflin Co.,TBoston, 429p. Cope, Edward D., 1899. Vertebrate remains from the Port Kennedy bone deposit; Jour. Acad. Nat. Sci. Phila., 11:193-267. Davis, Margaret 8., 1976. Pleistocene biogeography of temperate deciduous forests; Geosci. and Man, 13:13-26. , 1983. Holocene vegetational history of the eastern United States; in Late Quaternary Environments of the United States, volume 2, The Holocene, H. E. Wright (ed.), Univ. Minnesota Press, pp. 166-181. 64 65 Delcourt, Hazel R., 1979. Late Quaternary vegetation history of the eastern Highland Rim and adjacent Cumberland Plateau in Tennessee; Ecol. Monog., 49(3):255-280. and Paul A. Delcourt, 1975. The Blufflands: Pleistocene pathway into the Tunica Hills; Amer. Midl. Nat., 94(2):385-400. Delcourt, Paul A. and Hazel R. Delcourt, 1981. Vegetation maps for eastern North America: 40,000 yr B.P. to the present; Geobotany II. R. C. Romans (ed.), Plenum Publ. Corp., pp. 123-165. , 1983. Late Quaternary vegetational dynamics and community stability reconsidered; Quat. Res., 19(2):265-271. Dillon, Lawrence S., 1956. Wisconsin climate and life zones in North America; Science, 123:167-176. Dowling, Herndon 6., 1975. A provisional classification of snakes; Yearbook Herpetol., 1:167-170. and William E. Duellman, 1978. Systematic Herpetology: A synopsis of families and higher categories; HISS Publ., Publ. Herpetol. No. 7. Fay, Leslie P., 1984a. Mid-Wisconsinan and mid-Holocene herpetofaunas of eastern North America: A study in minimal contrast; Carnegie Mus. Nat. Hist., Spec. Publ. 8. , 1984b. Wisconsinan Appalachian herpetofaunas: What about distri- bution changes, heterochroneity, and seasonal climatic response?; (abs.), Amer. Quat. Assoc. 8th Bien. Mtg., Abstracts with Program. Flint, R. F., 1971. Glacial and Quaternarineology; John Wiley & Sons, New York, 892p. Foley, Robert L., 1982. The Moscow Fissure, a Late Pleistocene (Wood- fordian Substage) vertebrate assemblage from the Driftless Area of southwest Wisconsin; Unpub. Ph.D. diss., Dept. Geol., Univ. Iowa. Frest, Terrence J. and Leslie P. Fay, 1980. Peoria Loess mollusc faunas and Woodfordian biomes of the Upper Midwest; (abs.), Amer. Quat. Assoc. 6th Bien. Mtg., Abstracts with Program, p. 82. and Richard S. Rhodes 11, 1981. Oreohelix stri osa cooperi (Binney) in the Midwest Pleistocene; The Nautilus, 95(2):?7-55. Guilday, John E., 1962. The Pleistocene local fauna of the Natural Chimneys, Augusta County, Virginia; Carnegie Mus. Nat. Hist., Ann. 36(9):87-122. , 1967. Notes on the Pleistocene Big Brown Bat [Eptesicus grandis (Brown)]; Carnegie Mus. Nat. Hist., Ann. 39(7):105L114. 66 Guilday, John E., Paul S. Martin, and Allen D. McCrady, 1964. New Paris No. 4: A Pleistocene cave deposit in Bedford County, Pennsylvania; Nat. Speleol. Soc. Bull., 26(4):121-194. , Paul W. Parmalee, and Harold W. Hamilton, 1977. The Clark's Cave bone deposit and the Late Pleistocene paleoecology of the central Appalachian Mountains of Virginia; Carnegie Mus. Nat. Hist. Bull. No. 2, 86p. , Harold W. Hamilton, Elaine Anderson, and Paul W. Parmalee, 1978. The Baker Bluff Cave deposit, Tennessee, and the Late Pleistocene faunal gradient; Carnegie Mus. Nat. Hist. Bull., No. 11, 67p. Hack, John T. and Wayne L. Newell, 1974. North Carolina glacier: Evi- dence disputed; Science, 184:89. Hall, E. Raymond, 1981. The Mammals of North America; John Wiley & Sons, New York, v, I & II, 1271p. Hibbard, Claude W., 1949. Techniques for collecting microvertebrate fossils; Univ. Michigan Mus. Paleont. Contr., 8(2):7-19. , 1960. An interpretation of Pliocene and Pleistocene climates in North America; Michigan Acad. Sci., Arts, Ltrs., 64:5-30. Holman, J. Alan, 1962. A Texas Pleistocene herpetofauna; Copeia, 1962(2):255-261. , 1963. Late Pleistocene amphibians and reptiles of the Clear Creek and Ben Franklin local faunas of Texas; Jour. Grad. Res. Ctr., Southern Methodist Univ., 31(3):152-167. , 1965. A Late Pleistocene herpetofauna from Missouri; Trans. Illinois Acad. Sci., 58(3):190-194. , 1966. Some Pleistocene turtles from Illinois; Trans. Illinois Acad. Sci., 59(3):214-216. , 1967. A Pleistocene herpetofauna from Ladds, Georgia; Georgia Acad. Sci. Bull., 25(3):154-166. , 1968. Upper Pliocene snakes from Idaho; Copeia, 1968(1):152-158. , 1969. Herpetofauna of the Pleistocene Slaton local fauna of Texas; Southwest. Nat., 14(2):203-212. , 1976. Paleoclimatic implications of "ecologically incompatable" herpetological species (Late Pleistocene: Southeastern United States); Herpetologica, 32:290-295. , 1977a. The Pleistocene (Kansan) herpetofauna of Cumberland Cave, Maryland; Carnegie Mus. Nat. Hist. Ann., 46(11):157-172. 67 Holman, J. Alan, 1977b. Herpetofaunal evidence for the Pleistocene climatic equability hypothesis in North America; (abs.) X Intl. Quat. Cong. (INQUA), Birmingham, England. , 1979. A review of North American Tertiary snakes; Michigan State Univ. Publ. Mus., Paleont. Ser., 1(6):200-260. , 1980. Paleoclimatic implications of Pleistocene herpetofaunas of eastern and central North America; Trans. Nebraska Acad. Sci., 8:131-140. , 1981. A review of North American Pleistocene snakes; Michigan State Univ. Publ. Mus., Paleont. Ser., 1(7):263-306. , 1982. The Pleistocene (Kansan) herpetofauna of Trout Cave, West Virginia; Carnegie Mus. Nat. Hist. Ann., 51(20):391-404. and Ronald L. Richards, 1981. Late Pleistocene occurrence in southern Indiana of the Smooth Green Snake, Opheodrys vernalis; Jour. Herpetol., 15(1):123-125. Lundelius, Ernest L., 1967. Late Pleistocene and Holocene faunal history of central Texas; in Pleistocene Extinctions: The Search For A Cause, P.S. Martin and H. E. Wright (edsJTZ Yale Univ. Press, pp. 287-319. “ , Russell W. Graham, Elaine Anderson, John Guilday, J. Alan Holman, David W. Steadman, and S. David Webb, 1983. Terrestrial vertebrate faunas; in Late Quaternary Environments of the United States, volume 1, The Late Pleistocene, S. C. Porter (ed.), Univ. Minnesota Press, pp. 311-353. Lynch, John D., 1966. Additional treefrogs (Hylidae) from the North American Pleistocene; Carnegie Mus. Nat. Hist. Ann., 38(11):265-271. Martin, Paul S., 1958. Taiga-tundra and the full-glacial period in Chester County, Pennsylvania; Amer. Jour. Sci., 256(7):470-502. Meltzer, David J. and Jim I. Mead, 1983. The timing of Late Pleistocene mammalian extinctions in North America; Quat. Res., 19(1):130-135. Meylan, Peter A., 1982. The squamate reptiles of the Inglis IA fauna (Irvingtonian: Citrus County, Florida); Florida State Mus. Bull., Biol. Sci., 27(3):111-196. Mickelson, D. M., Lee Clayton, D. S. Fullerton, and H. W. Borns, Jr., 1983. The Late Wisconsin glacial record of the Laurentide Ice Sheet in the United States; in Late Quaternary Environments of the United States, volume 1, The Late Pleistocene, S. C. Porter (ed.), Univ. Minnesota Press, pp. 3-37. 68 Morgan, Alan V., Anne Morgan, Allan C. Ashworth, and John V. Matthews, Jr., 1983. Late Wisconsin fossil beetles in North America; in Late Quaternary Environments of the United States, volume 1, IDS. Late Pleistocene, S. C. Porter (ed.), Univ. Minnesota Press, pp. 354-363. McCoy, C. J., 1982. Amphibians and reptiles in Pennsylvania; Carnegie Mus. Nat. Hist. Spec. Publ. No. 6, 91p. McKenna, Malcolm C., 1962. Collecting small fossils by washing and screening; Curator, 5(3):221-235. McKeon, John B., 1974. North Carolina glacier: Evidence disputed; Science; 184:88-89. Péwé, Troy, 1983. The periglacial environment in North America during Wisconsin time; in Late Quaternary Environments of the United States, volume I, The Late Pleistocene, S. C. Porter (ed.), Univ. Minnesota Press, pp. 157-189. Rhodes, Richard S., 11, 1982. Mammalian paleoecology of the Farmdalian Craigmile and the Woodfordian Waubonsie local faunas, southwestern Iowa; Unpub. Ph.D. diss., Dept. Geol., Univ. Iowa., 132p. Richmond, Neil D., 1964. Fossil amphibians and reptiles of Frankstown Cave, Pennsylvania; Carnegie MUS. Nat. Hist. Ann., 16(20):225-228. Soler, E. Ivan, 1950. On the status of the family Desmognathidae (Amphibia, Caudata); Univ. Kansas Sci. Bull., 33(2):459-477. Stuckenrath, Robert, Jr. and J. E. Miekle, 1980. Smithsonian Institu- tion Radiocarbon Dates VI; Radiocarbon, 12(1):193-204. Tihen, Joseph A., 1958. Comments on the osteology and phylogeny of ambystomatid salamanders; Florida State Mus. Bull., Biol. Sci., 3 1 :1-50. , 1962a. Osteological observations on New World Bufo; Amer. Midl. Nat., 67(1):157-183. , 1962b. A review of New World fossil bufonids; Amer. Midl. Nat., 68(1):1-50. Van Dam, George H., 1978. Amphibians and Reptiles; in The Baker Bluff Cave deposit, Tennessee, and the Late Pleistocene faunal gradient; J. E. Guilday, H. W. Hamilton, E. Anderson, and P. W. Parmalee, Carnegie Mus. Nat. Hist. Bull., No. 11, pp. 19-25. Watts, William A., 1973. Rates of change and stability in vegetation in the perspective of long periods of time; in Quaternary Plant Ecologx, H. J. B. Birks and R. G. West (eds.), Blackwells, Oxford, pp. 195-206. 69 Watts, William A., 1979. Late Quaternary vegetation of central Appalachia and the New Jersey coastal plain; Ecol. Monog., 49(4):427-469. Wilson, Vincent V., 1975. The systematics and paleoecology of two Late Pleistocene herpetofaunas from the southeastern United States; Unpub. Ph.D. diss., Dept. 2001., Michigan State Univ., 67p.