number 5 /I j JUL2 0 19ttl FIELD MUSEUM llBRftBY EDITORIAL STAFF John E. Cooper, Editor Alexa C. Williams, Managing Editor John B. Funderburg, Editor-in-Chief Alvin L. Braswell, Curator of Lower Vertebrates, N.C. State Museum John C. Clamp, Associate Curator ( Invertebrates ), JV. C. State M useum Martha R. Cooper, Associate Curator ( Crustaceans ), N.C. State Museum James W. Hardin, Department of Botany, N. C. State University Board David S. Lee, Chief Curator of Birds and Mammals, N.C. State M useum William M. Palmer, Chief Curator of Lower Vertebrates, N.C. State Museum Thomas L. Quay, Department of ^oology, N. C. State University Rowland M. Shelley, Chief Curator of Invertebrates, N.C. State Museum Brimleyana, the Journal of the North Carolina State Museum of Natural His- tory, will appear at irregular intervals in consecutively numbered issues. Con- tents will emphasize zoology of the southeastern United States, especially North Carolina and adjacent areas. Geographic coverage will be limited to Alabama, Delaware, Florida, Georgia, Kentucky, Louisiana, Maryland, Mississippi, North Carolina, South Carolina, Tennessee, Virginia, and West Virginia. Subject matter will focus on taxonomy and systematics, ecology, zoo- geography, evolution, and behavior. Subdiscipline areas will include general in- vertebrate zoology, ichthyology, herpetology, ornithology, mammalogy, and paleontology. Papers will stress the results of original empirical field studies, but synthesizing reviews and papers of significant historical interest to southeastern zoology will be included. Suitability of manuscripts will be determined by the Editorial Board, and ap- propriate specialists will review each paper adjudged suitable. Final ac- ceptability will be decided by the Editor. Address manuscripts and all cor- respondence (except that relating to subscriptions and exchange) to Editor, Brimleyana, N. C. State Museum of Natural History, P. O. Box 27647, Raleigh, NC 27611. In citations please use the full name — Brimleyana. North Carolina State Museum of Natural History North Carolina Department of Agriculture James A. Graham, Commissioner CODN BRIMD 7 ISSN 0193-4406 The Blancan Carnivore Trigonictis (Mammalia: Mustelidae) in the Eastern United States Clayton E. Ray Department of Paleobiology, National Museum of Natural History, Smithsonian Institution, Washington, D. C. 20560 Elaine Anderson 730 Magnolia Street, Denver, Colorado 80220 AND S. David Webb Department of Natural Sciences, Florida State Museum, University of Florida, Gainesville, Florida 3261 1 ABSTRACT. — Fossils of a large galictine mustelid from Florida, North Carolina, and Maryland, represent Galera macrodon Cope 1868, here assigned to Trigonictis Hibbard 1941 and regarded as a senior synonym of T. idahoensis (Gazin) 1934. The smaller T. cookii (Gazin) 1934, is tentatively identified from Florida on the basis of a single specimen. INTRODUCTION In 1868 Cope described a partial mandible from presumed “postplio- cene” deposits in Charles County, Maryland, as the new species, Galera macrodon, assigning it to the genus of the living neotropical tayra. Fir a (= Galera) Barbara, in recognition of its galictine character. Leidy (l869) added a P-, apparently from the same individual and thus part of the holotype, although not indicated by Cope. In 1886 Nehring transferred the species to Galictis, the genus of the living neotropical grison. No additional material has ever been referred to Cope’s species. Meanwhile, fairly abundant remains of galictine mustelids have been reported from widespread occurrences of Blancan land mammal age in the western states. Gazin (1934) described similar species from Idaho, a larger, Lutravus{l) idahoensis, and a smaller, L.(?) cookii, and in 1937 assigned both tentatively to Canimartes. In 1941 Hibbard described the new genus and species, Trigonictis kansasensis. In recent years larger specimens have been assigned to Trigonictis idahoensis, including T. kan- sasensis, and smaller specimens to T. cookii. In the eastern states Trigo- nictis has been listed only in the Blancan assemblage from the Santa Fe River, Florida (Webb 1974, 1976), but the material has not been de- scribed. Interestingly, although the galictine affinities of both eastern and western forms have been recognized consistently from the first, none of Brimleyana No. 5:1-36. July 1981. 1 2 Clayton E. Ray, Elaine Anderson, S. David Webb the authors who studied the western Trigonictis has restudied Galera macrodon, and only Bjork (1970:28) has mentioned it. The specimen could not be found at the time of Schreuder’s (1935) writing, and the western species now assigned to Trigonictis had only just been described (Gazin 1934) and were not considered by her. Reig (1957) regarded Canimartes{l) cookii and Galictis macrodon as very close, probably representing a new subgenus of Galictis, and C.(?) idahoensis as conge- neric with Trigonictis kansasensis, and referred all of them to the Galicti- nae. Apparently working only from the literature, he did not recognize the extremely close relationship between T. idahoensis and T. cookii, which has been substantiated further by consideration of new material (Zakrzewski 1967; Bjork 1970; Galbreath 1972; Gustafson 1978). In 1971 one of us (Anderson) found the holotype of Galera macrodon in the collections of the Academy of Natural Sciences of Philadelphia, recognized its general affinities, and subsequently concluded that it is conspecific with Trigonictis idahoensis. This was expressed tentatively in Kurten and Anderson (1980:155). Recently Peter J. Harmatuk sent another one of us (Ray) a mustelid mandible found by Mr. Donnie Bailey at Goldsboro, North Carolina. Efforts to identify the specimen prompted a review of the literature and critical examination of the excellent sample of Trigonictis from Idaho in the National Museum of Natural History, of the Florida material in the Florida State Museum, and of the holotype of Galera macrodon in the Academy of Natural Sciences of Philadelphia. All of this led to the conclusion that Trigonictis idahoensis is a junior synonym of Trigonictis macrodon, to which all but one of the eastern specimens of Trigonictis, all western specimens previously assigned to T. idahoensis, and at least some previously assigned to T. cookii, should now be referred. Our purposes here are to review and describe the material from the eastern states, to present our conclusions regarding Galera macrodon, to place all of these in the context of the more abundant, geochronologically better known, western Trigonictis, and to comment briefly on its broader relationships with Palearctic paleotaxa and with neotropical neotaxa. Abbreviations ANSP - Academy of Natural Sciences of Philadelphia BMNH - British Museum (Natural History) F:AM - Frick:American Mammals Collection, American Museum of Natural History KUVP - University of Kansas, Vertebrate Paleontology UF/FSM - University of Florida, Florida State Museum USGS - United States Geological Survey Blancan Carnivore Trigonictis 3 USNM - National Museum of Natural History UM - University of Michigan UO - University of Oregon UW - University of Washington C - canine P - premolar M - molar mybp - million years before present OR - observed range with superscript (upper) or subscript (lower) Galictinae Reig 1957 Trigonictis Hibbard 1941 Type-species. — Trigonictis kansasensis Hibbard 1941. Included species. — T. macrodon (Cope) 1868, T. cookii (Gazin) 1934. Diagnosis. — see Hibbard 1941a:344; 1941b:273; Bjork 1970:22; and below, under T. macrodon. Distribution. — United States. Age. — Blancan Land Mammal Age; also Irvingtonian(?) in Florida. Trigonictis macrodon (Cope), 1868 Figs. 1, 2 (part), 3-6 “Carnivore the size of a fox,” Cope, 27 October 1867, letter to his father, in Osborn 1931:148. Galera, a new carnivore of the genus. Cope 1868a: 138. — Bjork 1970:28. Galera macrodon Cope 1 868b: 1 55. — Leidy 1 869:369, 47 1 , pi. 30, figs. 1-3. — Coues 1877:17 (type description reprinted). — Roger 1887:135. — Gillette and Colbert 1976:33. Putorius macrodon, Wortman 1883:1001. — Cope and Wortman, 1884:5. Galictis macrodon, Nehring 1886:151. — Roger 1896:60. — Trouessart 1897:264. — Nehring 1901:216. — Hay 1902:767. — Trouessart 1904:197. — Reig 1957:41, 44 (possibly distinct subgenus). Grison macrodon. Hay 1919:367. — Hay 1923:220, 347. — Hay 1930:531 (part, not Merrill 1907A). — Schreuder 1935:89. LutravusG) idahoensis Gazin 1934:137-143, fig. 1, table 1. Canimartesl idahoensis, Gazin 1937:363-364. — Hibbard 1941a:346- 347. — Hibbard 1941b:273-274. — Reig 1957:33, 45. — Repenning 1967:297. Trigonictis kansasensis Hibbard 1941a (genotype):344-347, fig. 5a, b. — Hibbard 1941 b:273-274. — McGrew 1944:53. — Reig 1957:45. — Zakrzewski 1967:293-294. 4 Clayton E. Ray, Elaine Anderson, S. David Webb Trigonictis kansasensis, “at least generically identical with Pannonictis pliocaenica,'' Repenning 1967:296. — Thenius 1972:206. Trigonictis idahoensis, Reig 1957:33, 45. — Zakrzewski 1967:293-297. — Shotwell 1970:82, fig. 37M,N. — Bjork 1970:22-24, figs. 10 b,c,e, 12. — Hibbard 1972a: 108-109. — Hibbard 1972b: 128. — Gustafson 1978:39- 41. — Kurte^n and Anderson 1980:155, fig. 2.2. Trigonictis cookii, Hibbard 1972a: 109 (part, F:AM 49163). Trigonictis cooki, Gustafson 1978:39, fig. 22. Remarks. — It is seldom possible to determine exact dates of publica- tion for the Proceedings of the Academy of Natural Sciences of Philadel- phia during the period of concern here, which includes the papers that we cite as Cope 1868a, 1868b (containing the type description of Galera macrodon), and 1868c. They were presented orally at meetings of the Academy on 10 December 1867, 31 December 1867, and 28 July 1868, respectively. Nolan (1913:xiii), the standard source for establishing time of Academy publications, is imprecise in these instances, but for the critical paper, 1868b, evidence indicates that it was published between 1 January and 1 1 May 1868 (see Nolan 1913; Osborn 1930). Holotype. — ANSP 11626, partial right mandibular ramus, lacking anterior and posterior extremities; retaining Pj - Mj- well preserved, pos- terior root of Pj, and with alveoli of canine (incomplete), Pj, and M^; fragment of left maxilla with P- well preserved, posterior root of P-, and partial alveoli of M-. The maxillary fragment was not mentioned by Cope, but it bears the same old number 187 and the same catalog number (ANSP 1 1626), as does the mandible, and its P- was included by Leidy ( 1 869:47 1 , pi. 30, figs. 2,3). Also, its preservation, size, and other charac- teristics are compatible with those of the mandible. Referred material. — From the western United States, all specimens, including the holotypes, heretofore assigned to Trigonictis idahoensis or T. kansasensis by Hibbard (1941a:344), Shotwell (1970:82), Bjork (1970:22-23), Hibbard (1972a: 109), and others. Also partial mandibles, previously assigned to T. cookii, F:AM 49163 from the Sand Draw local fauna, Nebraska (Hibbard 1972a: 109), and UW 41527 from the White Bluffs local fauna, Washington (Gustafson 1978:39). From the eastern United States, the following specimens, not pre- viously described: Santa Fe River, locality VIH A, Gilchrist Co., Florida UF/FSM 14256, right femur, and UF/FSM 16762, left femur, both collected June 1968 by William Hunt; both tentatively referred. UF/FSM 18475, left tibia, collected February 1971 by Kent Ainslie and William Hunt; tentatively referred. UF/FSM 18912, complete left mandibular ramus with Pj - Mp Blancan Carnivore Trigonictis 5 and associated incomplete right ramus with Pj and Pj, and incomplete P^ and Mj-; all teeth on both sides heavily worn, collected by Larry Roberts. Smith Mill Run, near Goldsboro, Wayne Co., North Carolina USNM 306507, left mandibular ramus lacking anterior extremity and part of coronoid process, with Pj - Mj well preserved and with part of canine alveolus, collected spring 1980 by Donnie Bailey. Type locality. — “collected by James T. Thomas near his residence in Charles County, Maryland, not far from the Patuxent River” (Cope 1868b: 138). Cope described several taxa of fossil vertebrates, mostly cetaceans, from the same source, but never published precise field data. Efforts made through the years to determine the exact locality and horizon in Charles County, and the identity of James T. Thomas, have been inconclusive thus far (see Dryden and Overbeck 1948:57, 90; Kel- logg 1955: 143, 1965:48, 1968: 105). Cope ( 1868a, b) referred to the deposit as postpliocene in age based on the fauna, a “common peccary” (later identified as Cynorca proterva Cope, a Miocene species; see Woodburne 1969), a “Manatus” (specimen has not been relocated, but sirenian remains are fairly common in the Calvert Formation but have not been substantiated from later beds in the area; see Kellogg 1966), and the Galera. Beds of the marine Miocene Chesapeake Series are overlain in the area by the Brandywine Formation, also termed “upland deposits,” a thin (10-30 feet, 3-9 m) mantle of gravel, sand and loam, mostly fluviatile in origin, deposited by the southwestwardly migrating channel of the Potomac River, and variously regarded as late Miocene, Pliocene and/or early Pleistocene in age, although lacking diagnostic fossils (Clark 1915; Hack 1955; Schlee 1957; Glaser 1968; Owens and Denny 1979). Although Cope’s other “postpliocene” fossils probably came from the Miocene Calvert Formation, his Galera macrodon more likely came from the overlying “upland deposits,” judging from its taxonomic affinities. It should be noted in passing, especially in view of the more recent recovery of referred material in Florida, that Reig (1957:44-45) mistak- enly ascribed the original material of Trigonictis macrodon to Florida rather than Maryland. Distribution. — In addition to the type locality (Fig. 1, loc. 7) in eastern Charles County, near the Patuxent River, Maryland, the material here referred to Trigonictis macrodon is known from the following localities: White Bluffs, east side of Columbia River, Franklin Co., southcen- tral Washington, some 10-30 mi. (16. 1-48.3 km) N of Richland (Fig. 1, loc. 1). UW locality A3027. Ringold Formation; age very early Blan- Clayton E. Ray, Elaine Anderson, S. David Webb Fig. I. Distribution of galictine mustelids of Blancan land mammal age in North America, l.ocalities I - 9, Trii^onictis niacrocion\ localities 2, 3, 5, 10, T. cookii\ localities 3, 4, Sminthosinis ho\vleri\ locality II, Trigonicfis sp.; locality 12, Caniniartes cuniniinsii\ locality 13, mustelid of the subfamily Galictinae. See text, under Distribution, for details of localities. Blancan Carnivore Trigonictis 1 can. (Gustafson 1978:39, 55, not plotted in his fig. 2; specimen identi- fied as T. cooki). Grand View, Jackass Butte locality, SW of Snake River, Owyhee Co., southwestern Idaho, some 17 mi. (27.4 km) NW of Grand View (Fig. 1, loc. 2). UO locality 2404. Upper part of the Glenns Ferry Formation, age regarded as late Blancan, although Mammuthus has recently been found at the locality. A K-Ar date of approximately 1.4 mybp from a lava above the Glenns Ferry Formation places a min- imum limit on the age of the Grand View local fauna. (Shotwell 1970:7, 82, fig. 8; Neville et al. 1979:519). Hagerman, west side of Snake River, Twin Falls Co., south- central Idaho (Fig. 1, loc. 3), the type locality of Trigonictis idahoensis and T. cookii. Various localities of UM, USGS, Dwight W. Taylor, and USNM. (Bjork 1970:4, 22-23, map 1, chart). Trigonictis occurs in the lower part of the Glenns Ferry Formation at elevations of approx- imately 865 to 990 m in beds dated by K-Ar and magnetic stratigraphy from older than 3.75 mybp to older than 3.2 mybp; thus, the Hager- man fauna is early Blancan in age. (Neville et al. 1979:521-522, fig. 10). Broadwater and Lisco, Morrill and Garden cos., western Nebraska (Fig. 1, loc. 4). Exact locality not published. The Broadwater fauna in the Broadwater Formation is late Blancan in age. (Schultz et al. 1951: table 1; Hibbard 1972b:128, 131-134). Sand Draw, Brown Co., central northern Nebraska (Fig. 1, loc. 5). Frick Prospecting Localities 277, some 5.5 mi. (8.8 km) N of Ains- worth, along Booth Draw, and 278, some 6.5 mi. (10.5 km) ENE of Ainsworth, N of Magill Draw. This fauna in the Keim Formation is late Blancan in age. (McGrew 1944:53; Skinner 1972:30-34; Hibbard 1972a: 109, specimen from 278 identified as T. cookii). Deer Park, Meade Co., southwestern Kansas (Fig. 1, loc. 6). Uni- versity of Kansas Meade County locality 1, some 10 mi. (16. 1 km) SW of Meade, Meade County State Park; SE 1/4, sec. 15, T33S, R29W, Lake Larrabee 7.5-min. quad., USGS. The Deer Park fauna, occur- ring in the Ballard Formation which overlies the Rexroad Formation, is late Blancan in age. (Taylor 1966:103; Hibbard 1972b: 128). Rexroad locality 3, Meade Co. (Fig. 1, loc. 6; distinct from preced- ing locality, but not distinguishable at scale of map). University of Kansas Meade County locality 3, some 11 mi. (17.7 km) SW of Meade; SW 1/4, sec. 22, T33S, R29W, Lake Larrabee 7.5-min. quad. This is the type locality of Trigonictis kansasensis. The Rexroad local fauna in the Rexroad Formation is early Blancan in age, perhaps somewhat older than the Hagerman local fauna. (Hibbard 1941a:344; 1941b:273; 1972b:128; Taylor 1966:101). Smith Mill Run, 2.5 mi. (4 km) N of center of Goldsboro, Wayne 8 Clayton E. Ray, Elaine Anderson, S. David Webb Co., east-central North Carolina (Fig. 1, loc. 8); 35®25'15"N, 7/^59'35" W, Goldsboro 15-minute quadrangle. The specimen was recovered by wet screening of sands redeposited in the bed of Smith Mill Run. At the same locality Mr. Bailey also has collected remains of Cretaceous vertebrates, including a partial femur of a hadrosaur, and of Tertiary vertebrates, including sharks, bony fishes, and marine mammals, mostly as float. Probably all of the Tertiary marine vertebrates are derived from the Pliocene Formation, poorly exposed in the creek banks, and probably assignable to the Rushmere Member, which in Virginia has yielded a glauconite date of 4.4 ± 0.2 mybp (Ward and Blackwelder 1980:D31). Although it is conceivable that the jaw of Trigonictis could have come from the Yorktown Formation, the most probable source is the undifferentiated post-Miocene de- posit, consisting of gravels, sands, and clays, of mostly fluviatile origin, that mantles much of the Goldsboro area to a depth of 30 feet (9 m) (Pusey 1960: 16-17). In detailed studies of the surficial sediments of the region, Daniels and Gamble (1974) identified, between the Kenly and Surry Scarps, a mappable, medium-fine sand unit, well sorted and with little clay, which would have been assigned to the “Sunderland” Formation by earlier workers. It is tempting to suppose that these sediments may be similar in genesis and age to the “Brandy- wine” or “upland deposits” of Maryland, and that the similar speci- mens of Trigonictis came from similar deposits in the two areas. Only additional, stratigraphically controlled collections will resolve the problem. Sante Fe River VIII A, Gilchrist Co., northern Florida (Fig. 1, loc. 9), vertebrate fossil locality in the Santa Fe River about 8 mi. (12.9 km) E of its confluence with the Suwanee River, SW 1 /4, SE 1 / 4, sec. 18, T7S, R16E. Fauna closely resembles those from Santa Fe River sites I A, I B, IV A. The fauna is late Blancan in age, and includes abundant material of Nannippus phlegon (MacFadden and Waldrop 1980). The citations of Trigonictis from Santa Fe River I B (Webb 1974: 17) and I A (Webb 1976:226) were in error, based on specimens from Santa Fe River VIII A. Other North American Galictinae. — Trigonictis cookii (Gazin) is known in Idaho from Grand View, Jackass Butte locality (Fig. 1, loc. 2), and Hagerman, the type locality (Fig. 1, loc. 3); Sand Draw, Frick Pros- pecting Locality 277, near Ainsworth, Nebraska (Fig. 1, loc. 5); and Haile XVI A, Alachua Co., Florida (Fig. 1, loc. 10), reported herein (tentative identification). Kurte^n and Anderson (1980:156) noted the presence of T. cookii also in the Pliocene of Texas and California. Trigonictis sp. has been listed from Cita Canyon, some 15 miles (24.1 Blancan Carnivore Trigonictis 9 km) southeast of Canyon, Randall Co., Texas (Fig. 1, loc. 11); Schultz (1977:126, 129). The Cita Canyon local fauna is thought to be slightly older than the Blanco on the basis of magnetic stratigraphy (Lindsay et al. 1975:1 16), but not on the basis of the faunas (Schultz 1977:123). Sminthosinis bowleri Bjork, 1970, was described from Hagerman (Fig. 1, loc. 3), and Kurt^n and Anderson (1980:156) cited it from the Broad- water fauna (Fig. 1, loc. 4). This small mustelid is closely related to Trigonictis and may be only subgenerically distinct from it (Bjork 1970). Canimartes cumminsii Cope, 1892, a possible galictine, known only from a single specimen, was described from the Blanco fauna, northeast- ern Crosby County, some 40 miles (64 km) east northeast of Lubbock, northwestern Texas (Fig. 1, loc. 12); Schultz (1977: 105, 126, fig. 20). The Blanco local fauna is between 1.4 and 2.4 mybp in age on the basis of fission track dating and magnetic stratigraphy (Lindsay et al. 1975:1 14). A galictine mustelid is recorded apparently from both the Arroyo Seco and Vallecito Creek faunas, in the Palm Spring Formation, Anza Bor- rego State Park, eastern San Diego County, southern California (Fig. 1, loc. 13). These faunas are late Blancan in age, and the latter may extend into the Irvingtonian (Opdyke et al. 1977:323, 325). Geologic Age. — Trigonictis is thought to have reached North Amer- ica as an immigrant from Eurasia in the early Blancan (Repenning 1967:296-297), or perhaps somewhat earlier. Tedford and Gustafson (1977) suggested that dispersal from oriental temperate woodlands resulted in the simultaneous appearance of Trigonictis in North Ameri- ca and of closely related forms in Europe. However, Schultz et al. (1972) believed that Trigonictis migrated from North America to Eurasia at this time. The biostratigraphic significance of Trigonictis is based entirely on its western occurrences where it is found only in faunas of Blancan Land Mammal Age. The western early to late Blancan local faunas including Trigonictis may be arranged tentatively in order of decreasing age as follows: White Bluffs, Rexroad, Hagerman, Broadwater-Sand Draw- Deer Park-Cita Canyon-Arroyo Seco, Blanco, Grand View-Vallecito Creek. Many uncertainties exist in such a sequence; for example. Shot- well ( 1 970: 1 6) could see little faunal basis for a distinction in age between the Hagerman and Grand View local faunas, and Schultz et al. (1978:60) proposed a somewhat different sequence of faunas. Radiometric ages and magnetic stratigraphy bracket the age of these faunas from somewhat less than 4 mybp to somewhat more than 1.4 mybp. In the east only the occurrence at Santa Fe VHI A in Florida is in demonstrably Blancan association. The records from North Carolina and Maryland are regarded as Blancan only by extrapolation, but they do serve to suggest the possibility of correlation of nonmarine and nearshore 10 Clayton E. Ray, Elaine Anderson, S. David Webb Fig. 2. Left upper carnassials (P-'^) of some fossil (A, B) and modern (C, D, E) galictine mustelids, in occlusal aspect, whitened for photography. All figures approximately 4X, scale in mm. A, ANSP 1 1626, Charles Co., Maryland, holo- type of Trigonictis macrodon; B, USNM 23664, Hagerman, Idaho, T. macrodon\ C, USNM 155480, Grisonella cuja', D, USNM 180224, Galictis vittata\ E, USNM 281468, Eira barbara. Blancan Carnivore Trigonictis 11 deposits in the east that have thus far proved intractable to dating. Voor- hies (1974) demonstrated the feasibility of this approach by recovery of two teeth of the Pliocene horse, Nannippus minor, from previously barren upland gravels near the Fall Line in Georgia. Revised Diagnosis. — A large galictine mustelid. Crowded upper and lower premolars. P- single rooted; P- (Figs. 2A, B) triangular in occlusal view with low, strong, conical protocone and well developed cingulum that forms a shelf extending from protocone to posterolingual base of metacone; small hypocone on cingulum; talon with shallow basin bor- dered anterolingually by protocone, posterolingually by hypocone and labially by anterior part of trigon; cingulum continuous around base of crown. M- three-rooted with reduced cingular shelf labial to metacone; tooth rectangular in occlusal outline with no distinct waist. Mandible (Figs. 3F, 4A) relatively short, robust, ventral margin straight. Premolars simple, lacking accessory cusps, cingulate, Pj double-rooted and set obliquely. Mj- trigonid longer than talonid; strong metaconid appressed to protoconid; talonid broad, basined; hypoconid moderately developed. Dental formula | y 5 5. Separated from T. cookii by larger size (see Measurements and Tables 1 and 2). Description of eastern specimens. — In ANSP 11626, the type speci- men from Charles County, Maryland, the P- is slightly wider and more robust, with a broader talon and lesser anterior emargination between the parastyle and protocone than is seen in other P-’s of Trigonictis. The P- (KUVP 4604) from Rexroad 3 has an incipient cusp on the cingulum that is absent in the other specimens. Examination of a large number of Recent galictines shows variations of this type to be quite common, and they are not regarded as significant. The inferior border of the mandibu- lar ramus of the holotype is somewhat more curved in labial aspect than is seen in the other specimens, which have a straighter profile (cf. Fig. 4A,B). The alveoli of Pj show that the tooth was double-rooted and set obliquely in the jaw. Pj is rather small, and there is a space between it and P4. The length of My (12.2 mm) falls near the lower end of the observed range (OR 12.0-14.9, N=24) of Trigonictis macrodon. The talonid of My is rather heavily worn; wear is also seen on the crests of the posterior faces of the metaconid and protoconid and obscures the notch between the protoconid and hypoconid. The left ramus from Smith Mill Run, North Carolina, USNM 306507 (Figs. 3D, 5D) has all the postcanine teeth in place, little worn, and mostly well preserved. Pj is small, double-rooted and set obliquely between the canine and P3. P-3 is more typical for Trigonictis than is that of the type of T. macrodon, as it is more robust and slightly overlaps the adjacent teeth. All of the premolars have strong cingula, which com- pletely encircle the crowns of Pj and P^. My is relatively large, elongate. 12 Clayton E. Ray, Elaine Anderson, S. David Webb Fig. 3. Mandibles of Trigonictis macrodon in occlusal aspect, whitened for photography. A, USNM 25026, Hagerman Idaho; B, USNM 12030, Hagerman, Idaho holotype of T. idahoensis; C, USNM 23560, Hagerman, Idaho; D, USNM 306507, Smith Mill Run, North Carolina; E, UF/FSM 18912, Sante Fe River, Florida; F, ANSP 1 1626, Charles Co., Maryland, holotype of T. macrodon. 1 CM Blancan Carnivore Trigonictis 13 Fig. 4. Mandibles of Trigonictis macrodon in labial aspect, whitened for photography. A, ANSP 1 1616, Charles Co., Maryland, holo- type, reversed; B, UF/FSM 18912, Santa Fe River, Florida. 14 Clayton E. Ray, Elaine Anderson, S. David Webb Fig. 5. Mandibles of Trigonictis macrodon in labial aspect, whitened for photography. A, USNM 25026, Hagerman, Idaho; B, USNM 12030, Hagerman, Idaho, holotype of T. idahoensis’, C, USNM 23560, Hagerman, Idaho; D, USNM 306507, Smith Mill Run, North Carolina Blancan Carnivore Trigonictis 15 and narrow with a distinct lingual indentation in its occlusal outline between the paraconid and metaconid; complementary crests on the pos- terior slope of the protoconid and the anterior slope of the hypoconid meet at the notch separating the trigonid from the talonid (contrast the worn condition of these crests in the holotype, Figs. 3D, F, 4A, 5D); the labially placed hypoconid is the highest part of the talonid; the talonid basin is bordered continuously posteriorly and lingually by a regular, crenulated crest; the talonid basin is floored by the smoothly and gently sloping internal wall of the hypoconid. The left half of the mandible from Santa Fe River VIII A, UF/FSM 18912, is essentially complete and one of the best preserved specimens of Trigonictis known (Fig. 4B). It is characterized by short, robust propor- tions, relatively straight ventral profile, low articular process, erect, broad coronoid process, deeply incised masseteric fossa terminating pos- terior to the level of the M2 alveolus, and narrow symphyseal region with crowded incisive alveoli. The strongly molded bony surface with some rugosity and the heavily worn dentition indicate advanced individual age. The left Pj was lost during life and its alveoli nearly obliterated. The teeth in the associated incomplete right ramus show comparable wear. Oth- erwise the characteristics of the postcanine teeth fit well within the varia- tion of the sample of Trigonictis, and similar patterns of extreme wear and dental anomalies were observed among modern galictines in the USNM. The three postcranial elements from Santa Fe VIII A are assigned only tentatively to the genus and species. The left (UF/FSM 16762) and right (UF/FSM 14256) femora are similar enough in all respects to represent a single individual. Thus only the left was described, measured and illus- trated (Fig. 6), along with the left tibia (UF/FSM 18475). The tibia seems compatible with the femora in size, maturity, morphology, and preserva- tion and could also represent the same individual. Unfortunately, in the present state of knowledge, very little can usefully be done with these dissociated elements beyond placing them on record for future reference. There is only a single partial skeleton of Trigonictis known, UM V49819, T. cookii, described and illustrated by Bjork (1970:25-26). He referred several isolated postcranial elements to T. idahoensis, including a femur and tibia, but did not describe (except for comments on the tibia), mea- sure, or illustrate them. The femur of T. cookii illustrated by Bjork (1970:fig. 14e) appears to be very similar to those from Florida, except for its smaller size. Among modern mustelids, the tibia is longer than the femur and is curved in mustelines, is slightly shorter than the femur in galictines, and is curved in Galictis and Grisonella. The tibia is nearly straight in Eira, but both it and the femur are relatively slender. If the femora and tibia from 16 Clayton E. Ray, Elaine Anderson, S. David Webb Fig. 6. Left femur and tibia, whitened for photography, of Trigonicfis macrocion, Santa Fe River, Florida. All figures approximately IX; scale in mm. Femur, UF/ FSM 16762, in cranial (A) and caudal (B) aspects; tibia, UF/ FSM 18475, in cranial (C) and caudal (D) aspects. Blancan Carnivore Trigonictis 17 Florida do in fact represent a single individual, or at least reflect the true relative lengths of the elements, they are subequal in Trigonictis, as indi- cated also by the associated skeleton of T. cookii (cf. Table 3). In relative length, absolute size, proportions, and morphology, the Florida specimens compare most favorably with those of Pannonictis pilgrimi from the Villafranchian of Hungary described and illustrated by Mottl (1941), and further compared and the illustrations reproduced by Ficcarelli and Torre (1967). Measurements. — See Tables 1-3 for measurements and statistics. The distinction in size between Trigonictis macrodon and T. cookii is not clear-cut, yet all authors, including us, agree that the range in size is too great to be accommodated in a single species, or at least in a single penecontemporaneous population. We have followed Bjork (1970) in placement of the boundary, as reflected in his statistics presented in our Table 2, modified slightly by the addition of new specimens. Incidentally, there is an anomalous point for an Mj- assigned to T. cookii in Bjork’s figure 13, plotted at approximately 12.2 mm length and 4.8 mm width; this specimen does not appear in the statistics for either species. Gustafson (1978) presented a case for a chronocline of increasing size upward in the Hagerman beds. Presumably because the length of Mj- (12.0 mm) of the single specimen from the older White Bluffs fauna fell in the gap between the two species, he moved the boundary upward and referred the specimen to T. cookii. However, in the other variates pre- sented, the specimen falls within the observed range of T. macrodon, and we have referred it to the larger species. Gustafson’s measurements for UW 41527 include: length of P^, 6.5 mm; width of P^, 3.7; length of Mj-, 12.0; width of Mj-, 5.3; depth of jaw below Mj-, 12.7 (cf. our Table 2). Bjork (1970) did not give depth of jaw below M^ for T. cookii, but two specimens from Hagerman in USNM measure 10.6 and 10.8 mm. The inferior border of the ramus is missing in the holotype (USNM 12606), but its depth could scarcely have exceeded 1 1 mm. Similarly, on the basis of size (length of Mj-, 12.5 mm), we regard F:AM 49163 from the Sand Draw local fauna as T. macrodon rather than T. cookii, as assigned by Hibbard (1972a: 109). His measurement of 12.6 for “greatest length P^ - Mj-” is in error, and should be 18. 1. Based on our measurements from his figures, Shotwell’s (1970:82, fig. 37J-N) specimens from the Grand View local fauna are correctly assigned, although the larger is in some dimensions slightly larger than other published specimens of Trigonictis, for example in length of Mj-, 15 mm, and depth of jaw below Mj-, 17 mm. However, there is a fragment of a jaw, USNM 25027, from Hagerman, almost certainly representing Trigonictis, that would have been as large or even larger. 18 Clayton E. Ray, Elaine Anderson, S. David Webb However, the separation in size is not absolute. For example, the holo- type of T. macrodon falls within the observed range for T. cookii in width of P3, but widths of anterior premolars have slight diagnostic value. Development of lower premolars is highly variable in modern galictines. We have noted variations in Eira barbara, for example, ranging from several specimens with Pj absent or tiny and spicular, some with spaces between the premolars, to others with extremely large, broad, overlap- ping premolars. Relationships. — The closest relative of Trigonictis macrodon is its only congener, T. cookii, distinguished by slightly smaller size. Having two carnivores so similar in morphology and size apparently coexisting (in the Hagerman, Broadwater, Sand Draw, and Grand View local fau- nas) has vexed everyone who has studied these animals. Gazin (1934) considered and rejected the possibility of a single species with extraordi- narily high sexual and/or individual variation, as did Zakrzewski (1967), and Bjork (1970). We applied Zakrzewski’s (1967:205-206) method of assessing sexual dimorphism to a series of Eira barbara from Panama in the USNM and a sample of Grisonella cuja from Argentina and Brazil in the BMNH, as the tayra and the little grison show dimorphism in size and should offer a more relevant guide to Trigonictis than would the more distantly related mustelines used by Zakrzewski. His “sex size ratio” for a given variate is obtained by dividing the mean value for males by that for females. Our results are as follows: Eira barbara N X OR Toothrow length male 10 31.54 29.6-34.0 female 12 28.77 26.6-30.2 sex size ratio 1.10 Length of Mj male 14 10.33 9.2-11.0 female 12 9.74 9.1-10.3 sex size ratio 1.06 Blancan Carnivore Trigonictis 19 Grisonella cuja N X OR Toothrow length male 14 26.19 22.0-28.4 female 11 23.99 22.0-25.4 sex size ratio 1.09 Length of Mj- male 14 7.75 6.4-8.6 female 11 6.89 6.3-7.8 sex size ratio 1.12 With the exception of the sex size ratio for toothrow length in Mustela vison, our values are less in Eira barbara and Grisonella cuja than for these variates in the four species of mustelines tested by Zakrzewski, suggesting a weaker secondary sexual dimorphism in the tayra and little grison. Galbreath (1972) suggested the possibility of a chronocline of decreas- ing size from T. idahoensis in the lower beds at Hagerman to T. cookii and Sminthosinis bowleri in the higher beds, whereas Gustafson (1978:39-40) suggested a chronocline of increasing size from T. cookii to T. idahoensis based on the relatively small sample of suitable, stratigra- phically placed mandibles. The trend shown by Gustafson for the Hag- erman beds is suggestive, but far from conclusive, and the evidence from other local faunas seems to weaken the case. For example, Gustafson assigned his White Bluffs specimen, thought to be very early Blancan in age, to T. cookii, but it seems to us to fall more comfortably within the variation of the larger T. macrodon, based on his measurements. Also, the specimens from the early Blancan Rexroad local fauna are large, and the specimen from the Irvingtonian Haile XVI A fauna is small. T. cookii and T. macrodon occur together in the late Blancan Broadwater, Sand Draw, and Grand View local faunas. Thus, no clear trend of changing size through time can be demonstrated on the basis of present collections and correlations. In any case, the eastern specimens fall comfortably on one side or the other of the indicated specific boundary and thus present no problems of assignment. Should all Trigonictis ultimately prove to belong to a single heterogeneous species with great individual, sexual, temporal, and geographic variation, then T. macrodon would be the senior name. Beginning with Cope (1868) the fossils now assigned to Trigonictis have always been compared appropriately to the modern neotropical Galictinae. The complex nomenclatural history of this small group has been confusing, and its systematic arrangement remains unresolved. We 20 Clayton E. Ray, Elaine Anderson, S. David Webb Table 1. Measurements (mm) of P- in some specimens of Trigonictis and Sminthosinis. Length Width Trigonictis macrocion ANSP 1 1626, holotype KUVP 4604, from 1 1.1 7.6 Hibbard (1941a:347) UM V49728 (cast, in USNM), 1 1.9 7.8 Hagerman local fauna 11.1 6.2 USNM 23664, Hagerman local fauna 11.4 7.2 USNM 23665, Hagerman local fauna 11.6 — Trigonictis cookii UM V49819, Hagerman local fauna. from Bjork (1970:26) 9.6 5.7 Sminthosinis howler i Hagerman local fauna, from Bjork (1970:28) UM V52868, holotype 8.01 5.35 UM V53344 7.70 5.08 UM V55214 7.93 5.00 UM V55952 7.93 5.15 follow Cabrera (1958:258-265), except that we accord Grisonella full generic status, on the basis of absence of the metaconid in My, in agree- ment with Bjork (1970:28), in recognizing the following taxa: Galictis vittata for the gxhon.Grisonella cuja for the little grison, and Eira bar- bara for the tayra. Available to us for this study were 16 specimens of Galictis, 10 of Grisonella, and 49 of Eira at USNM, and 27 of Grisonella and 62 of Eira at BMNH. Trigonictis certainly requires close comparison to this group, but its precise affinities within it and with fossil galictines of Eurasia, as well as their interrelationships, are less clear. There seems to be merit in the opinion that the affinities of Eira could be musteline rather than galictine (cf. Pocock 1921, and Ficcarelli and Torre 1967), but we adhere to the more traditional arrangement, pending the necessary comprehensive review of fossil and modern material. In 1867 Cope probably had available at ANSP no specimens of Galic- tis vittata, two skulls and jaws of Grisonella cuja, and four of Eira (= Galera) barbara. Cope’s concept of Galictis, and assertion that it lacked the metaconid in Mj- (1868b: 156, footnote), obviously were based on Grisonella, then and now generally regarded as congeneric with Galictis vittata. With the exception of the obscure Putorius ardeus Bravard whose affinities were recognized only recently (see Viret 1954), none of the European fossils was known at that time; hence. Cope’s (1868b: 155-156) Blancan Carnivore Trigonictis 21 assignment of his new species to Galera was remarkably perceptive and his brief discussion of neotropical affinities of the North American Pleis- tocene fauna resoundingly modern. He indicated that his species differed from the living tayra in having an apparently “more slender muzzle,” relatively larger Mj-, and double-rooted Pj, among other characters. Nehring (1886:151-152) pointed out that, unknown to Cope, the grison has a metaconid in Mj-, and indicated that in relative size of this tooth and form of masseteric fossa. Cope’s fossil jaw more closely resembles that of the grison, and therefore transferred the species to Galictis. He later reiterated this conclusion (1901:215-216), stating that the specific charac- ters distinguishing Galictis macrodon from the living grison are the two- rooted Pj, the space between Pj and P^, and the somewhat greater dimen- sions of the teeth and jaw bone. Although Nehring was aware of Leidy’s (1869) illustration of the P- of Galera macrodon, he obvously based his reassignment to Galictis on Cope’s type description, based only on the lower jaw, as reprinted by Coues (1877), and on Leidy’s illustration of the lower jaw. Until now, only Schreuder (1935:89) has considered the P- of Galera macrodon, but, as the specimen was not available at that time, she was limited to Leidy’s illustration. She supported Nehring’s conclusions. The P- of Galictis (and Grisonella), however, differs fundamentally from that of Trigonictis macrodon in having a well developed basin in place of the protocone, bordered posteriorly by a large hypocone, especially in Galictis (Fig. 2C,D), a contrast noted by Zakrzewski (1967:294). Eira barbara is more similar to Trigonictis in development of the protocone of P-(Fig. 2E), but differs markedly in shape of the tooth through having the protocone set off from the trigon by a narrow neck (as typically devel- oped in mustelines and apparently as in Canimartes cumminsii). The Mj- of Grisonella is significantly different from those of Trigonic- tis, Galictis, and Eira in lacking the metaconid. Although Mj- shows considerable variation in the modern series available for Galictis and Eira, in Galictis it is relatively large, as in Trigonictis, but generally has a narrower talonid on which the hypoconid is more prominent and cen- trally placed and the basin absent or incompletely enclosed lingually; in Eira the Mj- is relatively small, but otherwise resembles that of Trigonictis in having a broad talonid with labially placed hypoconid and distinct basin, though generally not as fully enclosed as in Trigonictis (USNM 104546 and many other specimens of Eira have essentially Trigonictis- like talonids). Schreuder (1935:88-89) quantified the relative size of Mj- by expressing its crown length as a multiple of that of P^ (length of My /length of Py)^ We have used her technique to compare relative size of Mj- in the series of modern galictines in USNM to that of Trigonictis, with the following results: 22 Clayton E. Ray, Elaine Anderson, S. David Webb N X OR Eira barbara 43 1.34 1.23-1.53 Grisonella cuja 10 1.73 1.55-1.88 Galictis vittata 14 1.85 1.73-1.98 Trigonictis macrodon ANSP 11626, holotype 1.67 USNM 306507, North Carolina 1.99 KUVP 4604, holotype of T. kansasensis (from Hibbard 1941a:346) 1.85 UW 41527, Washington (from Gustafson 1978, table 14) 1.85 Hagerman sample in USNM 5 1.81 1.73-1.86 T. cookii Hagerman sample (from Bjork 1970, table 8) 3 1.94 1.81-2.03 UF/FSM 27509 1.89 Thus Trigonictis clearly is similar to Grisonella and Galictis in relative length of Mj-, and unlike Eira, which has a reduced Mj-. However, as noted above, in all features other than relative size, M j- of Trigonictis is more similar to that of Eira than to that of Grisonella or Galictis. The overall profile of the mandible of Trigonictis macrodon as seen in labial aspect in the specimen from Florida (Fig. 4B) is much more like that in Eira than in Galictis, in that the ventral border is straight, not convex ventrally, and the coronoid process is broad and blunt, not tapering and pointed. On the other hand the ventral border is curved and the coronoid pointed in the Florida specimen tentatively referred to T. cookii (Fig. 7B). and the shape of the coronoid is reversed in specimens assigned to the two species from Idaho (Shotwell 1970, Fig. 37J,N). Similarly, we have not been able to identify consistent characters in the shape and extent of the masseteric fossa. The “narrower muzzle” of Galera macrodon, as com- pared to the tayra, noted by Cope, does seem to be reflected in Trigonictis in its narrower symphyseal region with more crowded incisive alveoli, as preserved in UF/FSM 18912. In summary, Trigonictis be about as similar to (and as different from) the tayra, as to (and from) the grison. The tendency of American authors to refer to Trigonictis as a grison, an extinct grison, a grison ancestor, or as grison-like is not justifiable, and is in fact misleading, as a case equally strong and probably equally errone- ous can be made for the tayra. We are not able to follow Bjork’s (1970:28) suggestion that T. cookii might be ancestral to Galictis vittata (and Sminthosinis bowleri to Grisonella cuja), especially as there are coeval fossil galictines in Eurasia much more like the modern Galictis. Blancan Carnivore Trigonictis 23 The close relationship between some Neogene palearctic mustelids and modern neotropical galictines, especially Galictis, has long been recog- nized (for example. Major 1902:626). A comprehensive discussion of these Old World forms is beyond the scope of this paper and is in any event not feasible, as it must be based upon restudy of original materials and resolution of a welter of morphological, taxonomic, and nomencla- tural problems (Viret 1954, and Ficcarelli and Torre 1967, provide the best introduction to the subject). However, a few comments are appropriate here insofar as the Eurasian material may be instructive in resolution of North American problems. Synonymy of New" and Old World forms has been suggested, for example as follows, '"Trigonictis kansasensis appears to be at least generi- cally identical with Pannonictis pliocaenica" (Repenning 1967:296; fol- lowed by Thenius 1972:206 and by Kurten and Anderson 1980:156). We feel that this is at least premature if not wrong. Pannonictis pliocaenica has a basined talon in P- similar to that in the grison (Kormos 1931:169; Schreuder 1935:83). Pilgrim (1932:854-855) also called attention to the importance of this feature, as follows, “The protocone of P- in Enhydric- tis is backwardly expanded. . .but the pronounced cusp of Trochictis retained in Tayra has been replaced by a cup-shaped protocone sur- rounded by a rim, which is identical with that of Pannonictis and Mus- telid gen. indet., sp. n., Zdansky and almost so with GrisonT Also according to Pilgrim P^ is single-rooted in Enhydrictis, as in living galictines, but not in Trigonictis. The P- of Trigonictis closely resembles that of the geologically older Trochictis, which also retains a double- rooted Pj. Trochictis is clearly more primitive than Trigonictis, but may well have been ancestral, whereas Trigonictis is in turn less advanced than Enhydrictis, Pannonictis, Galictis, and Grisonella . Kormos (1934:131), concerned about the great variation in size of specimens referred to Pannonictis pliocaenica, measured a series of lower dentitions. He found that the length of Mj- fell into three groups, as follows: 8 specimens with a range of 10.7-12.0 mm and a mean of 1 1.5 mm; 17, range 12.6-14.5, mean 13.5; 15, range 15.0-16.2, mean 15.6. He concluded that the largest size group represented males of P. pliocaenica, the middle group females, and the smallest a new species, P. pilgrimi. However, after completing the manuscript he discovered a skull of the new species which proved to be more like the skull of Enhydrictis than like that of the type species, Pannonictis pliocaenica, in spite of its Pannonictis-\\\it dentition. Transfer of pilgrimi to Enhydrictis, probably as a junior synonym of E. ardea (cf. Viret 1954:83), however, is not to say that Pannonictis is a synonym of Enhydrictis as some American authors apparently have supposed. Discovery of skulls of Trigonictis macrodon, T. cookii, and Sminthosinis howleri may prove to be similarly enlighten- Table 2. Measurements of some jaws and lower dentitions of Trigonictis, compared with statistics for samples from the Hagerman local fauna, from Bjork (1970) except toothrow length, distance from posterior margin of canine alveolus to posterior margin of Mj alveolus, from Zakrzewski ( 1967). Measurements in brackets are minima, based on specimens with missing, broken, or deeply worn teeth. The minimum length of M- for specimens assigned to T. macrodon is for UW 41527, from Gustafn Qi (N D, s cd O >77 OO oo r-' rd od rd c/5 CnI — (N ' — ' * ""iw C3 o c ? lx >40 O Ov oo rd =s Ai m 5 z m — o ■(^^5 ‘^U The-9hf:^lh6l) PJ^^qqiH ujojj tJ- vO J — duopo.uniu \i —) sisu<)snsunyj \i ‘adAtojoq 1 >ri TOqhdAT'M rr^ 4—1 -C r- oo ON ^ , o3 .5? rd 1 rf (T) (N T3 — X X ~ ti, OJ rd 1 1 1 m •nui|0JT?3 qjJOfsj 57 T0590e fMNSn rd r<3 r-' rn i_i ' — ' •puniAjn[/\i (N ‘ UOpO.IJDLH SUJlUOi^Li± SO So' >77 C*~) 1 ^3dAio|oq ‘9391 i dSNV 00 c X X 4— > — 4-* 00 00 c c — X o u. Iri 4-> 00 x: 4— • |4J- CL c (U o 1 o Irs) |r j 1 r V~ Cl Cl CL Blancan Carnivore Trigonictis 25 rn oo rsi m rd id rd sd 1 • 1 6.1 (N lO ON »d rd id 1 i/^ 1 r- rd Tf- rd id ro oo rd rd id — m rn r<^ w T3 00 c "O 00 c _o 1 r^. a. — 1 1 Cl ir^ o CnI O rd •n X ■*— • 00 •o c c _o -i rd rs Ci, 05 -«i** Vj C o dj a -CJ 05 VJ <3 )4. -C) CN S .c-» :!!!!_ ’S P 0 >Ci Z !>a ? P 8^ s; c c2 3 v. P Galictis vittata USNM 395079, adult male 69.4 15.8 4.5 12.0 Blancan Carnivore Trigonictis 27 ing. Meanwhile, synonymizing any of the three is not justifiable. The status and relationships of Canimartes cumminsii can be assessed only after addition of new material. For the present it seems best to regard the Galictinae as including at least the genera Trigonictis and Sminthosinis in North America, Enhy- drictis and Pannonictis in Eurasia, and Galictis, Grisonella, Eira, and Lyncodon in South America. Of these, Trigonictis apparently is the most primitive, standing closest to Trochictis, the probable ancestor to the group. Trigonictis cookii (Gazin), 1934 Fig. 7 Lutravus(?) cookii Gazin 1934:142-143, fig. 2, table 1. Canimartes ? cookii, Gazin 1937:363-364. Canimartes(?) coocki, Reig 1957:33. Caminartes(?) cooki, Reig 1957:41. Canimartes(?) cooki, Reig 1957:42,44. Trigonictis cookii, Zakrzewski 1967:293-297. — Hibbard 1972a: 109 (part). — Kurten and Anderson 1980:156. Trigonictis cooki, Shotwell 1970:82, fig. 37J-L. — Bjork 1970:24-26, fig. 14. — Galbreath 1972:786. — Hibbard 1972a: 128, fig. 50. — Gustafson 1978:39-41 (part). Holotype. — USNM 12606, partial right mandibular ramus, lacking anterior and posterior extremities and most of ventral margin; retaining Pj - Mj- well preserved except part of metaconid missing from Mf, retain- ing alveolus and partial root of Mj, partial alveoli of anterior and poste- rior roots of Pj with part of posterior root, and posterior wall of alveolus of canine. Referred material. — From the western United States, the specimens assigned to the species by Zakrzewski (1967), Shotwell (1970:82), and Bjork (1970:24-25), and one specimen, F:AM 49160, assigned to the spe- cies by Hibbard (1972a: 109, fig. 50); possibly as yet undescribed speci- mens from Texas and California (Kurten and Anderson 1980:156); also USNM 25128, incomplete left mandibular ramus with Pj, from Hager- man, Idaho, collected by C. L. Gazin, 1934. From the eastern United States, tentatively referred herein, UF/FSM 27509, well preserved right mandibular ramus lacking anterior extremity, with Pj and M^; collected May 1973 by UF/FSM staff at Haile XVI A, Alachua Co., Florida. From the same locality in Florida, there is a complete right humerus, UF/FSM 27510, of a mustelid, rather similar in its distal half to that of T. cookii reported by Bjork (1970:26, fig. 14a), though slightly smaller and lacking the entepicondylar foramen, characteristic only of the skunks 28 Clayton E. Ray, Elaine Anderson, S. David Webb Fig. 7. Mandibles of Trigonictis cookii in labial (A-C) and occlusal (D-F) aspects, whitened for photography. Labial aspect: A, USNM 25138, Hagerman, Idaho; B, UF/FSM 27509, Haile XVI A, Florida; C, USNM 12606, Hagerman, Idaho, holotype. Occlusal aspect: D, USNM 25138, Hagerman, Idaho; E, UF/FSM 27509, Haile XVI A, Florida; F, USNM 12606, Hagerman, Idaho, holotype. Blancan Carnivore Trigonictis 29 among mustelids. The specimen cannot be assigned to Trigonictis, at least until more is known of its postcranial osteology. Type locality. — Hagerman (one mi. S of Plesippus quarry), west side of Snake River, Twin Falls Co., south-central Idaho (Fig. 1, loc. 3). Distribution. — In addition to the type locality (Fig. 1, loc. 3), in the Hagerman fauna of early Blancan age, T. cookii is known from the Jackass Butte locality, Owyhee County, Idaho (Fig. 1, loc. 2), in the Grandview fauna, and from Frick Prospecting Locality 277, Booth Draw, Brown County, Nebraska (Fig. 1, loc. 5) in the Sand Draw fauna, both of late Blancan age. See Distribution under T. macrodon for addi- tional details. Haile XVI A, Alachua County, northern Florida (Fig. 1, loc. 10), lies near the center of the Haile limestone mining district about 1 km north- east of Haile XV A (Robertson 1976) at an elevation of about 26 m. The fossil vertebrates occur in a large fissure carved from the late Eocene Ocala limestone and filled with massive dark, gray-brown, silty clays. The fissure was at least 8 m wide, and possibly of equal depth; unfortunately the site was destroyed by mining operations before its dimensions could be determined with accuracy. The sediments are high in dark organic content and show no signs of oxidation, unlike most fissure-fill sites in the Haile district. The vertebrate fauna from Haile XVI A is currently under study by Michael Frazier who generously permitted us to mention this local fauna. Its age is Irvingtonian, probably older than the Coleman II A local fauna of late Irvingtonian age (Martin 1974). The aquatic elements of the fauna include Rana, urodeles, abundant kinosternid turtles, Deirochelys, Chrvsemys, Alligator, and a natricine snake. The diverse terrestrial fauna, in which Eremotherium and Crotalus are unusually abundant, includes representatives of both mesic and xeric habitats. Geologic age. — Blancan Land Mammal Age, with the exception of the one record of Irvingtonian Land Mammal Age from Haile XVI A, if the specific identification is correct. See Geologic age under T. macrodon for additional details regarding western occurrences. Diagnosis. — As for T. macrodon, but smaller. Description of eastern specimen. — The right mandibular ramus from Haile XVI A, Florida, UF/FSM 27509, is complete and well preserved except for the anterior extremity, including the symphyseal region (Fig. 7B,E). The coronoid process is triangular in outline, with a rather acute apex, in contrast to the lobate outline with rounded apex in UO 16352 (Shotwell 1970:fig. 37J,K). The incomplete alveoli of Pj reflect crowding between canine and Pj, with the tooth oriented obliquely and the poste- rior root situated medial to the anterior root of Pj. P^ is relatively broad, with a marked posterior heel, completely cingulate, and with a slight 30 Clayton E. Ray, Elaine Anderson, S. David Webb protuberance in the position of a metaconid, and an undulation on the posterior crest of the principal cusp. A small pressure facet on the ante- rior margin of the crown reflects contact with Pj in life, Mj- is robust, with a strong, discrete metaconid and hypoconid; its deep, closed basin is bordered posteriorly and lingually by a crenulated crest of uniform height. A posterior pressure facet evidences contact with Mj, the alveolus of which indicates a large tooth with a root of circular cross section, in contrast to that in other specimens of T. cookii, with root cross section antero-posteriorly enlongate and narrower posteriorly. Measurements. — See Tables 1-3 for measurements and statistics of T. cookii, and see Measurements under T. macrodon for discussion of dis- tinction in size between the two species. Table 2 shows that UF/FSM 27509 falls within the observed range of virtually all available measure- ments for the meager series of T. cookii, and below that for T. macrodon. Nevertheless, the toothrow of the Florida specimen is crowded overall and the teeth broad in comparison to specimens from Idaho (Fig. 7D-F). Relationships. — For a discussion of the relationships of T. cookii see this heading under T. macrodon. The relationships of the specimen from Haile XVI A remain uncertain. Its morphological peculiarities, the possi- ble pertinence of the humerus from the same deposit, and the Irvingto- nian age of the fauna, suggest the possibility of a late, derived species. However, the difficulty experienced by others in interpreting series of fossil galictine mandibles in the absence of skulls recommends caution. Therefore, we prefer to assign this isolated jaw to T. cookii pending recovery of supplementary material. ACKNOWLEDGMENTS. — Knowledge of the later Cenozoic terres- trial mammalian fauna in the eastern United States is extremely limited as compared with that in the western states. Since the opening of the vast fossil fields of the west more than a century ago, professional paleo- mammalogists understandably have devoted much less attention to the relatively unproductive eastern states, where there are no vast exposures of fossiliferous strata and where results of collecting are apt to be meager. However, a modest renaissance is underway in the east, largely as a result of the energetic and effective efforts of an increasingly sophis- ticated and numerous cadre of dedicated amateurs. In North Carolina, these hobbyists recently organized themselves as the North Carolina Fos- sil Club; and in Florida they have formed the Florida Paleontological Society. The present report provides an example of their already signifi- cant accomplishments and an inkling of future prospects. Thus, we wish to thank two members of the North Carolina Club: Donnie Bailey, whose discovery and contribution of the specimen from North Carolina stimulated the investigation leading to preparation of this paper, and Peter J. Harmatuk, who as usual recognized the specimen as unusual and Blancan Carnivore Trigonictis 31 brought it to the attention of one of us. Likewise, we are grateful to the Florida amateurs, especially to William Hunt and Kent Ainslie for their collecting at Santa Fe VIII and to Ben Waller, now President of the Florida Paleontological Society, for pioneering underwater paleontology on the Santa Fe River. Gay Vostreys and Charles Smart made specimens available from the collections of the Academy of Natural Sciences of Philadelphia, as did G. B. Corbet from the British Museum (Natural History), and Michael Fraz- ier from the Haile XVI A fauna. Victor Krantz made the photographs, and Lawrence B. Isham prepared the figures. The manuscript was reviewed critically by Jessica Harrison. Robert W. Purdy, Charlotte Hol- ton, and Carol M. Spawn aided in efforts to learn more about the James T. Thomas collection from Charles County, Maryland. LITERATURE CITED Bjork, Philip R. 1970. The Carnivora of the Hagerman local fauna (late Pliocene) of southwestern Idaho. Trans. Am. Philos. Soc., n.s., 60(7). 54 pp. Cabrera, Angel. 1958. Catalogo de los Mamiferos de America del Sur. I (Metatheria - Unguiciilata - Carnivora). Rev. Mus. Argent. Cienc. Nat. ‘Bernardino Rivadavia’ Inst. Nac. Invest. Cienc. Nat. Zool., 1957, 4(1). iv + 308 pp. Clark, William B. 1915. The Brandywine Formation of the middle Atlantic Coastal Plain. Am. J. Sci., 4th ser., 40(239):499-506. Cope, Edward D. 1868a. Remarks on the contents of caves of South-western Virginia. Proc. Acad. Nat. Sci. Phila. (1867)/9(4):137-138. . 1868b. An addition to the vertebrate fauna of the Miocene period, with a synopsis of the extinct Cetacea of the United States. Proc. Acad. Nat. Sci. Phila. (1867)7P(4):138-156. . 1868c. Second contribution to the history of the Vertebrata of the Miocene period of the United States. Proc. Acad. Nat. Sci. Phila. (1868)20(3):- 184-194. , and J. L. Wortman. 1884. An account of the mammalian fauna of the post-Pliocene deposits in the state of Indiana. Indiana Dep. Geol. Nat. Hist. Annu. Rep. 14(2). 62 pp. Coues, Elliott. 1877. Fur-bearing animals: a monograph of North American Mustelidae. Dep. Interior, U. S. Geol. Surv. Terr. Misc. Publ. 8. xiv + 348 pp. Daniels, Raymond B., and E. E. Gamble. 1974. Surficial deposits of the Neuse-Cape Fear divide above the Surry scarp, North Carolina, pp. 88-101 in Oaks, Robert Q., Jr., and J. R. DuBar (eds.). Post-Miocene stratigraphy central and southern Atlantic Coastal Plain. Utah State Univ. Press, Logan, viii + 275 pp. Dryden, Lincoln, and R. M. Overbeck. 1948. Detailed Geology, pp. 29-127 in The physical features of Charles County. Md. Board Nat. Resour., Dep. Geol. Mines Water Resour., Baltimore, xiv + 267 pp. Ficcarelli, Giovanni, and D. Torre. 1967. II mustelide Enhydrictis galictoides del Pleistocene della Sardegna. Palaeontogr. Ital. (53:139-161. Galbreath, Edwin C. 1972. [Review of] The Carnivora of the Hagerman local fauna (late Pliocene) of southwestern Idaho. J. Paleontol. 4(5(5):785-786. 32 Clayton E. Ray, Elaine Anderson, S. David Webb Gazin, C. Lewis. 1934. Upper Pliocene mustelids from the Snake River basin of Idaho. J. Mammal. 75(2): 137-149. . 1937. Notes on fossil mustelids from the upper Pliocene of Idaho and Texas. J. Mammal. /5(3):363-364. Gillette, David D., and E. H. Colbert. 1976. Catalogue of type specimens of fossil vertebrates. Academy of Natural Sciences, Philadelphia. Part II: terrestrial mammals. Proc. Acad. Nat. Sci. Phila. /25(2):25-38. Glaser, John D. 1968. Coastal Plain geology of southern Maryland. Md. Geol. Surv. Guidebook 1. iv + 56 pp. Gustafson, Eric P. 1978. The vertebrate faunas of the Pliocene Ringold Formation, south-central Washington. Mus. Nat. Hist. Univ. Oreg. Bull. 23. 62 pp. Hack, John T. 1955. Geology of the Brandywine area and origin of the upland of southern Maryland. U. S. Geol. Surv. Prof. Pap. 267-A. 43 pp. Hay, Oliver P. 1902. Bibliography and catalogue of the fossil Vertebrata of North America. U. S. Geol. Surv. Bull. 779:868 pp. . 1919. On the relative ages of certain Pleistocene deposits. Am. J. Sci., 4th ser.,47(281):361-375. . 1923. The Pleistocene of North America and its vertebrated animals from the states east of the Mississippi River and from the Canadian provinces east of longitude 95°. Carnegie Inst. Wash. Publ. 322. viii + 499 pp. . 1930. Second bibliography and catalogue of the fossil Vertebrata of North America. Vol. H. Carnegie Inst. Wash. Publ. 390(2). xiv + 1074 pp. Hibbard, Claude W. 1941a. New mammals from the Rexroad fauna, upper Pliocene of Kansas. Am. Midi. Nat. 2b(2):337-368. . 1941b. Mammals of the Rexroad fauna from the upper Pliocene of southwestern Kansas. Trans. Kans. Acad. Sci. 44:265-313. . 1972a. Class Mammalia, pp. 77-125 in Skinner, Morris F., and C. W. Hibbard. Early Pleistocene pre-glacial rocks and faunas of north-central Nebraska. Bull. Am. Mus. Nat. Hist. 745(1): 1-148. . 1972b. Sand Draw local fauna. Correlation, age, and paleoecology. pp. 1 25-1 34 m Skinner, Morris F., and C. W. Hibbard. Early Pleistocene pre-glacial rocks and faunas of north-central Nebraska. Bull. Am. Mus. Nat. Hist. 745(1):1-148. Kellogg, Remington. 1955. Three Miocene porpoises from the Calvert cliffs, Maryland. Proc. U. S. Natl. Mus. 705(3354): 101-154. . 1965. The Miocene Calvert sperm whale Orycterocetus. U. S. Natl. Mus. Bull. 247(2):47-63. . 1966. New species of extinct Miocene Sirenia. U. S. Natl. Mus. Bull. 247(3):65-98. . 1968. Miocene Calvert mysticetes described by Cope. U. S. Natl. Mus. Bull. 247(5): 103-132. Kormos, Theodor. 1931. Pannonictis pliocaenica n.g., n. sp., a new giant mustelid from the late Pliocene of Hungary. Ann. Inst. Reg. Hungar. Geol. 29(3): 167-177. . 1934. Neue und wenig bekannte Musteliden aus dem ungarischen Oberpliozan. Folia Zool. Hydrobiol., Riga 5(2): 129-158. Kurten, Bjorn, and E. Anderson. 1980. Pleistocene mammals of North America. Columbia Univ. Press, New York, xvii + 442 pp. Leidy, Joseph. 1869. The extinct mammalian fauna of Dakota and Nebraska, including an account of some allied forms from other localities, together with a Blancan Carnivore Trigonictis 33 synopsis of the mammalian remains of North America. J. Acad. Nat. Sci. Phila., 2nd ser., 7:23-472. Lindsay, Everett H., N. M. Johnson and N. D. Opdyke. 1975. Preliminary correlation of North American land mammal ages and geomagnetic chronology. Univ. Mich. Mus. Paleontol. Pap. Paleontol. 72:111-119. MacFadden, Bruce J., and J. S. Waldrop. 1980. Nannippus phlegon (Mammalia, Equidae) from the Pliocene (Blancan) of Florida. Bull. Fla. State Mus. Biol. Sci. 25(l):l-37. Major, C. I. Forsyth. 1902. Exhibition of, and remarks upon, the skull of a new fossil mammal {Enhydrictis galictoides). Proc. Zool. Soc. Lond. 1901. 2(2):625-628. Martin, Robert A. 1974. Fossil Mammals from the Coleman IIA fauna, Sumter County, pp. 35-99 in Webb, S. David (ed.). Pleistocene mammals of Florida. Univ. Presses Fla., Gainesville, x + 270 pp. McGrew, Paul O. 1944. An early Pleistocene (Blancan) fauna from Nebraska. Field Mus. Nat. Hist. Geol. Ser. 9(2):33-66. Mottl, Maria. 1941. Untersuchungen an Pannonictis-Extremitaten. Mitt. Jahrb. Kgl. Ungar. Geol. Anstalt J5(2):42-72. Nehring, Alfred. 1886. Either Lutra brasiliensis, L. paranensis, Galictis crassidens und Galera macrodon. Sitzungsber. Ges. naturf. Freunde Berlin 1 886(10): 144- 152. . 1901. Ueber Galictis canaster Nelson, Gal. crassidens Nhrg. und Gal. Allamandi Bell. Sitzungsber. Ges. naturf. Freunde Berlin 1901(9):209-216. Neville, Colleen, N. D. Opdyke, E. H. Lindsay and N. M. Johnson. 1979. Magnetic stratigraphy of Pliocene deposits of the Glenns Ferry Formation, Idaho, and its implications for North American mammalian biostratigraphy. Am. J. Sci. 279(5):503-526. Nolan, Edward J. 1913. An index to the scientific contents of the Journal and Proceedings of the Academy of Natural Sciences of Philadelphia, 1812-1912. Acad. Nat. Sci. Phila. xiv + 1419 pp. Opdyke, Neil D., E. H. Lindsay, N. M. Johnson and T. Downs. 1977. The paleomagnetism and magnetic polarity stratigraphy of the mammal-bearing section of Anza Borrego State Park, California. Quat. Res. (N.Y.)7(3):3 16-329. Osborn, Henry F. 1930. Biographical memoir of Edward Drinker Cope, 1840-1897. Natl. Acad. Sci. U.S. A. Biogr. Mem. 7i(3): 127-317. . 1931. Cope: master naturalist. Princeton Univ. Press, Princeton, xvi T 740 pp. Owens, James P., and C. S. Denny. 1979. Upper Cenozoic deposits of the central Delmarva Peninsula, Maryland and Delaware. U. S. Geol. Surv. Prof. Pap. 1067-A. iii + 28 pp. Pilgrim, Guy E. 1932. The genera Trochictis, Enhydrictis, and Trocharion, with remarks on the taxonomy of the Mustelidae. Proc. Zool. Soc. Lond. 1932(4):845-867. Pocock, Royal 1. 1921. On the external characters and classification of the Mustelidae. Proc. Zool. Soc. Lond. 1921(4):803-837. Pusey, Richard D. 1960. Geology and ground water in the Goldsboro area. North Carolina. N. C. Dep. Water Resour. Div. Ground Water Bull. 2. ix + 77 pp. Reig, Osvaldo A. 1957. Un mustelido del genero Galictis del Eocuartario de la Provincia de Buenos Aires. Ameghiniana 7(l-2):33-47. Repenning, Charles A. 1967. Palearctic-Nearctic mammalian dispersal in the late Cenozoic. pp. 288-31 1 in Hopkins, David M. (ed.). The Bering land bridge. 34 Clayton E. Ray, Elaine Anderson, S. David Webb Stanford Univ. Press, Stanford, xiii + [iii] + 495 pp. Robertson, Jesse S. 1976. Latest Pliocene mammals from Haile XV A, Alachua County, Florida. Bull. Fla. State Mus. Biol. Sci. 20(3): 1 1 1-186. Roger, Otto. 1887. Verzeichniss der bisher bekannten fossilen Saugethiere. (Zusammengestellt von Dr. Otto Roger.). Ber. Naturwiss. Ver. Schwaben Neuburg (a. V.), Augsburg 29:1-162. . 1896. Verzeichniss der bisher bekannten fossilen Saugethiere. (Neu zusammengestellt von Dr. Otto Roger.). Ber. Naturwiss. Ver. Schwaben Neuburg (a. V.), Augsburg 22:1-272. Schlee, John. 1957. Upland gravels of southern Maryland. Bull. Geol. Soc. Am. b«5(10):1371-1410. Schreuder, Antje. 1935. A note on the Carnivora of the Tegelen clay, with some remarks on the Grisoninae. Arch. Neerl. Zool. 2(l):73-94. Schultz, C. Bertrand, G. C. Lueninghoener and W. D. Frankfurter. 1951. A graphic resume of the Pleistocene of Nebraska (with notes on the fossil mammalian remains). Bull. Univ. Nebr. State Mus. i(6):41 pp. , L. D. Martin, L. G. Tanner and R. G. Corner. 1978. Provincial land mammal ages for the North American Quaternary. Trans. Nebr. Acad. Sci. 5:59-64. , L. G. Tanner and L. D. Martin. 1972. Phyletic trends in certain lineages of Quaternary mammals. Bull. Univ. Nebr. State Mus. 9(6): 183-195. Schultz, Gerald E. 1977. Blancan and post-Blancan faunas in the Texas panhandle, pp. 105-145 in Schultz, Gerald E. (ed.). Guidebook, field conference on late Cenozoic biostratigraphy of the Texas panhandle and adjacent Oklahoma, August 4-6, 1977. West Texas State Univ. Killgore Res. Center Dep. Geol. Anthropol. Spec. Publ. 1. vii + 160 + 3 + [ii] pp. Shotwell, J. Arnold. 1970. Pliocene mammals of southeast Oregon and adjacent Idaho. Mus. Nat. Hist. Univ. Oreg. Bull. 17. 103 pp. Skinner, Morris F. 1972. Geology, pp. 14-35 in Skinner, Morris F., and C. W. Hibbard. Early Pleistocene pre-glacial rocks and faunas of north-central Nebraska. Bull. Am. Mus. Nat. Hist. 745(1): 1-148. Taylor, Dwight W. 1966. Summary of North American Blancan nonmarine mollusks. Malacologia 4(1): 1-172. Tedford, Richard H., and E. P. Gustafson. 1977. First North American record of the extinct panda Parailurus. Nature 2b5(5595):621-623. Thenius, Erich. 1972. Grundzuge der Verbreitungsgeschichte der Saugetiere. Eine historisch^Tiergeographie. Gustav Fischer, Stuttgart, x + 345 pp. Trouessart, Edouard-Louis. 1897. Carnivora, Pinnipedia, Rodentia 1. (Protrogo- morpha et Sciuromorpha.) Fasciculus II. pp. 219-452 in Catalogus Mammalium tarn viventium quam fossilium. Nova editio. 1897-1899. R. Friedlander and son, Berlin, v + 1469 + vi + v pp. . 1904. Primates, Prosimiae, Chiroptera, Insectivora, Carnivora, Pinni- pedia. Fasciculus I. pp i-iv + 1-288 in Catalogus Mammalium tarn viventium quam fossilium. Quinquennale supplementum, anno 1904. 1904-1905. R. Friedlander and son, Berlin, iv + 929 + vii pp. Viret, Jean. 1954. Le Loess a Bancs Durcis de Saint-Vallier (Drome) et sa Faune de Mammiferes Villafranchiens. Nouv. Arch. Mus. Hist. Nat. Lyon 4. 200 pp. Voorhies, Michael R. 1974. The Pliocene horse Nannippus minor in Georgia: geologic implications. Tulane Stud. Geol. Paleontol. 7/(2): 109-1 14. Ward, Lauck W., and B. W. Blackwelder. 1980. Stratigraphic revision of upper Miocene and lower Pliocene beds of the Chesapeake Group, Middle Atlantic Blancan Carnivore Trigonictis 35 Coastal Plain. U. S. Geol. Surv. Bull. 1482-D. iv + 61 pp. Webb, S. David. 1974. Chronology of Florida Pleistocene mammals, pp. 5-31 in Webb, S. David (ed.). Pleistocene mammals of Florida. Univ. Presses Fla., Gainesville, x + 270 pp. . 1976. Mammalian faunal dynamics of the great American interchange. Paleobiology 2(3):220-234. Woodburne, Michael O. 1969. Systematics, biogeography, and evolution of Cynorca and Dyseohyus (Tayassuidae). Bull. Am. Mus. Nat. Hist. 141(2): 271-356. Wortman, Jacob L. 1883. [Remarks on Galera macrodon.] Am. Nat. 77(9): 1001. Zakrzewski, Richard J. 1967. The systematic position of Canimartesl from the upper Pliocene of Idaho. J. Mammal. 45(2):293-297. Accepted 18 March 1981 ADDENDUM After this paper was submitted for publication, Richard H. Tedford and Henry Galiano of the American Museum of Natural History very kindly called to our attention the following unreported specimens of Trigonictis: F:AM 63107, partial left maxilla with P-, P-, and M- lacking labial part of crown. We assign the specimen to T. macrodon on the basis of size; length of P- 11.2, width 7.8. In addition P- shows an interesting combination of features tending to bridge the morphological gap between this tooth in the holotype and in referred specimens from the west. The occlusal outline is broadly triangular with a wide talon, as in the holotype, but the anterior margin is indented more as in the Hagerman specimens. The cuspation of the talon is reminiscent of that in the holotype of T. kansasensis (Hibbard 1941a:fig. 5b) and is more complex than in any other specimen seen, with a strong hypocone, without a domi- nant protocone, having in its place a group of interconnected cus- pules in part expressed as crenulations of the cingulum, of which the anterior-most is strongest. The specimen is from Matthew Wash, west of Safford, Graham County, southeastern Arizona, and, though not strictly referable to any of the published Blancan local faunas in the vicinity of Safford, is clearly Blancan on the basis of associated fauna and magnetostratigraphic studies in progress (Tedford, pers. comm.). Also from the same area (southwest corner of Bear Springs Flat, west of Safford) is a left mandibular ramus lacking teeth, F:AM 63109, almost certainly referable to T. macrodon. F:AM 62740, nearly complete left mandibular ramus with canine, Pj, Pj, and Mj-. Measurements as in our Table 2: toothrow length 32.1; Pj length 4.1, width 2.6; Pj length 5.3, width 3.6; Mj- 36 Clayton E. Ray, Elaine Anderson, S. David Webb length 12.3, width 5.1, length of trigonid 8.4; depth of jaw below Mj- 13.0, width 6.8. In most of these dimensions the specimen lies near or below the lower extreme of the OR in T. macrodon; Pj and Pj in particular are smaller than in any other specimen. As pointed out, however, these teeth are highly variable in size and are of low diagnostic value. Although this specimen slightly extends the OR in several variates, we assign it to T. macrodon. The specimen is from Bevins Pit 2, near Channing, Old- ham County, in the northwestern part of the Panhandle of Texas. This unpublished locality is within the Rita Blanca Creek drainage that includes the Blancan Red Corral local fauna (Schultz 1977:131, fig. 5), and has produced a Blancan fauna (Tedford, pers. comm.). Also in the collections of the American Museum of Natu- ral History are casts of five mandibular rami of Trigonictis from the University of Nebraska State Museum. Four of these are from the Broadwater fauna, Morrill County, Nebraska, and clearly represent T. cookii (length of Mp 10.7 or less), not T. macrodon. If the pub- lished records (Schultz et al. 1951: table 1; Hibbard 1972b:128, 131- 134) for the larger species are correct, they must have been based on other specimens. The fourth specimen, from Hooker County, west- central Nebraska, represents T. macrodon (length of Mj- 13.4). UF/FSM 18085, isolated right P-, length 9.9, width 6.6. Its size does not serve to allocate the tooth unequivocally to either T. macrodon or T. cookii, and we record the specimen only as Trigonictis sp. The specimen is from Inglis I A, Citrus County, Florida, of early Irvingtonian age (Webb 1974:29, table 2.1, fig. 2.1). This late occurrence and the size and robustness of the specimen are compatible with the record for Haile XVI A, and reinforce the suggestion that a derivative Trigonictis extended into the Irvingto- nian of Florida. Pseudanophthalmus from Appalachian Caves (Coleoptera: Carabidae): The Engelhardti Complex Thomas C. B ARR, J R. School of Biological Sciences, University of Kentucky, Lexington, Kentucky 40506 ABSTRACT. — The engelhardti complex includes about 55 species of the cave beetle genus Pseudanophthalmus , here arranged in 7 species groups with the following new species described and illustrated. (1) ENGELHARDTI group: deceptivus, VA; wallacei, TN; nortoni, TN; fastigatus, GA; nickajackensis, TN / AL\ sequoyah, AL‘, steevesi, AL. (2) TEN NESSEENSIS group: pusillus, TN; paynei, TN; unionis, TN. (3) HIRSUTUS group: sericus, VA; paulus, TN; ventus, TN; assimilis, AL. (4) HUBRICHTI group: sanctipauli, VA; paradoxus, TN. (5) JONESl group: scutilus, TN\rogersae, seciusus, \ A\pallidus, TN; longiceps, VA/TN; thomasi, VA; cordicollis, VA. (6) HYPOLITHOS group: hypolithos, KY; scholasticus, KY; frigidus, KY; calcareus, KY; praetermissus, VA. (7) ALABAMAE group: georgiae, GA. Other tax- onomic changes are: P. delicatus Valentine, new status (full species), and P. virginicus, new combination for Aphanotrechus virginicus Barr. The species of the engelhardti complex occupy the Appalachian valley from Giles Co., Virginia southwest through east Tennessee to northwest Georgia and north Alabama, including karst “islands” near the eastern Allegheny front — Pine Mountain (KY/TN), Grassy Cove (TN), Sequatchie Valley (TN), Lookout Valley (TN/GA), and Wills Valley (AL). The hypothesis is offered that a majority of species groups of Pseudanophthalmus are descendants of lineages evolving in the Alle- gheny plateau and surviving in an Allegheny refugium during the ( Pleistocene. The trechine cave carabids of eastern United States include 6 genera (Barr 1972) of which Pseudanophthalmus Jeannel, with an estimated 200 species, is by far the largest and most widely distributed. The other 5 genera — Neaphaenops, Darlingtonea, Ameroduvalius, Nelsonites, and Xenotrechus — encompass a total of only 10 or 11 additional species. This is the first in a series of detailed papers in which I intend to mono- graph Pseudanophthalmus . It deals with the species of the southern part of the Appalachian Valley and Ridge Province (hereafter called simply the “Appalachian Valley”) and of associated karst “islands” within but near the eastern edge of the Allegheny Plateau. About one-fourth of the genus (approximately 55 species) consists of an array of apparently related species which may be called the "''engel- hardti complex.” In this array presumed relationship among its compo- nent species is suggested by two apomorphic characters: (I) the apical recurrent groove of the elytron is elongate and bisinuate, and 2) the Brimleyana No. 5:37-94. July 1981. 37 38 Thomas C. Barr, Jr. transfer apparatus within the internal sac of the aedeagus consists of a canoe-shaped right dorsal sclerite partly enfolding an elongate-triangular, minutely spinulose, left ventral piece. The species of the engelhardti com- plex are widely distributed in the Appalachian valley from Giles County, Virginia, southwestward through eastern Tennessee, southeastern Ken- tucky (Pine Mountain), and northwest Georgia, to DeKalb and Blount counties, Alabama, then following the Tennessee River valley westward across north Alabama, finally northward again to Decatur, Perry, and Wayne counties in west-central Tennessee. (Seven trans-Allegheny spe- cies belonging to the engelhardti group will be treated in a subsequent paper.) No classification of the genus into species groups and no attempt to understand the internal phylogeny of Pseudanophthalmus is possible without a reasonably clear picture of this large block of related species. The classical conflict between the dual goals of taxonomy — organiza- tion and retrieval of information versus expression of phylogenetic rela- tionships — arises in any attempt to regroup the engelhardti complex. In this paper I propose seven species groups, three of them previously sug- gested by Valentine ( 1 932) or Jeannel ( 1 949), and four new groups. There are eleven subsets of related species; four of these are placed in the engelhardti group as defined here (species clustered around P. engel- hardti, P. holsingeri, P.fulleri, and P. loedingi, respectively) and two in the hubrichti group (species clustered around P. hubrichti and P. egberti). However, I regard all seven groups as internally monophyletic. A brief list of the five species groups of the northern Appalachian valley (not in the engelhardti complex) is given at the end of this paper. Four- teen other groups have been proposed for the numerous species west of the Allegheny Plateau (see Jeannel 1949: Valentine 1932; Krekeler 1973). Most of the caves cited in this paper have been described by Douglas (1964) or Holsinger (1975) for Virginia, Barr (1961) or Matthews (1972) for Tennessee, and Varnedoe (1973) for Alabama. This paper is dedicated to the memory of my late friend and colleague. Dr. Walter B. Jones, for many years State Geologist of Alabama and founder and Director of the Alabama Museum of Natural History. In the 1940s and 1950s he explored many caves in Alabama, Kentucky, and Tennessee with Dr. J. Manson Valentine, discovering many new taxa of cave trechines. He would have delighted in the descriptions of the new species in the present paper, especially those, I think, from Alabama. I am particularly pleased that one of the most interesting of the new species groups {jonesi group) bears his name. His zest for life, his generosity, and his assistance to those of us who shared his never-ending curiosity about the subterranean world earned him a major role in the development of North American biospeleology. Appalachian Pseudanophthalmus 39 Pseudanophthalmus Jeannel ENGELHARDTI COMPLEX Integuments generally pubescent, although dorsum of head and prono- tum disc subglabrous in a few species. Elytral microsculpture always more or less transverse, usually a little confused, often obsolescent on central disc, rarely forming distinct meshes; elytra not pruinose. Anterior discal puncture at or near level of 4th umbilicate puncture. Apical recur- rent groove elongate, oblique or subparallel, usually bisinuate (except alabamae group) and usually connected to apex of 3rd stria with or without a crosier. Humeri setulose, rarely serrulate. Mentum tooth short, usually emarginate. Mesosternum simply declivous. Profemur unmodi- fied. Last abdominal sternite of males with apical margin entire. Transfer apparatus consisting of an elongate, canoe-shaped right piece (sometimes with dorsal keel) partly enfolding a smaller, usually shorter, elongate- triangular, spinulose left piece. Distribution: — The species of the engelhardti complex occur primar- ily in the Appalachian Valley and Ridge Province from the New River valley in Giles County, Virginia, southwest through east Tennessee, northwest Georgia, and north Alabama, and also in the Tennessee River valley in west-central Tennessee as far north as Decatur and Perry coun- ties. They also inhabit karst “islands” near the eastern edge of the Alle- gheny Plateau — Grassy Cove, Tennessee; Pine Mountain, Tennessee and Kentucky; Lookout Valley, Tennessee and Georgia; Wills Valley, Alabama; and Sequatchie Valley, Tennessee and Alabama. Southwest of Tazewell County, Virginia, all known species of cave trechines in the Appalachian valley or in karst islands at the edge of the Allegheny Pla- teau belong to the engelhardti complex. Limited overlap (but not sympa- try in the same caves) occurs with species of the petrunkevitchi group (all species with circular, well-formed eye rudiment), which are known from caves in Washington, Smyth, Wythe, Pulaski, Tazewell (Burkes Garden), and Bland counties, Virginia. Ranges of engelhardti complex species {P. quadratus, P. egberti) and pusio and gracilis group species in Giles County are mutually exclusive (see Barr 1965). Pleistocene Refugia Cave species of Pseudanophthalmus are presumably descendants of edaphobitic ancestors which became extinct outside of caves during warmer, drier interglacials (Jeannel 1949; Barr 1967a). Jeannel (1949) proposed a Unaka refugium hypothesis: ancestral beetle stocks spread over the Appalachian valley and Interior Low Plateaus during glacial maxima from an interglacial refugium in the higher mountains along the Tennessee-North Carolina border. A pulse theory of cave colonization 40 Thomas C. Barr, Jr. was thus implied, each “pulse” corresponding to colonizations at the beginning of each interglacial. Polytypic Trechus hydropicus and T. cumberlandus show present distribution patterns suggesting how a Unaka-based stock might have spread west and north (Barr 1969, 1979a). Study of the engelhardti complex, coupled with concurrent examina- tion of cave trechines west of the Allegheny plateau (Barr, in prepara- tion), suggests a different interpretation. 1) With the notable exception of the engelhardti group, species groups of Pseudanophthalmus on opposite sides of the Allegheny Plateau are quite distinct and mutually exclusive. The small genera of eastern cave trechines — Neaphaenops (see Barr 1979b), Darlingtonea, Ameroduvalius, and Nelsonites (see Valentine 1952) — are limited to Kentucky and part of Tennessee west of the Alleghenies; they have no counterpart in the Appalachian valley. 2) In Appalachian valley caves the majority of Pseudanophthalmus species are concentrated near the Allegheny front and become increasingly rare toward the Unaka front. 3) No edaphobitic species of Pseudanophthal- mus have been found in the Unaka region, despite considerable searching during the past 20 years. In fact, the only edaphobitic, noncavernicolous species of Pseudanophthalmus thus far known from eastern United States is P. sylvaticus (Barr 1967b), from Pocahontas County, West Vir- ginia, in the heart of the Allegheny Plateau. 4) The pattern of species distribution in caves is not what one would expect from repeated expan- sion and contraction of an ancestral stock from a Unaka source. Only the engelhardti group shows this sort of pattern — a series of closely similar species strung out as if on a chain of islands in an archipelago, with representatives on both sides of the Alleghenies. All other species groups in the genus differ substantially across the Allegheny Plateau. The caves of two karst islands — Grassy Cove and Pine Mountain — are inhabited by species of the jonesi and hypolithos groups, other species of which are found in Appalachian valley caves to the east. The significance of the Allegheny Plateau as an interglacial refugium for Pseudanophthalmus ancestors has been neglected in favor of Jean- nel’s Unaka refugium hypothesis. As an alternative view I offer the possi- bility that the source of many, perhaps most, of the species groups of Pseudanophthalmus was the upland forest of the Allegheny Plateau. The diversity within the genus and the difficulty of relating species groups on opposite sides of the Allegheny Plateau — only the gracilis group (east) and inexpectatus group (west) seem clearly related (Krekeler 1973) — suggest that considerable precave diversification may have taken place within the plateau, and that only those groups near karst areas survived the warm, dry, climatic regimes of four successive interglacial periods. In fact, the level of intrageneric diversity opens up the possibility that Pseudanophthalmus is much older than the Pleistocene. Jeannel (1949 Appalachian Pseudanophthalmus 41 and elsewhere) demonstrated a close relationship between the Eurasian genus Trechoblemus (eyes and wings usually present) and Pseudanoph- thalmus and Neaphaenops (I would add Nelsonites), postulating a Euro- pean introduction ancestral to these cave beetles. In my judgment Lasio- trechus discus, from eastern Canada and northeastern United States, belongs within the same phyletic series. Trechoblemus was recently reported from Oregon (Barr 1972). The notion of a Unaka refugium should not be totally discarded: the distribution of the engelhardti group (one of seven groups in the engel- hardti complex) is compatible with the hypothesis of successive pulses from a Unaka refugium, and the distribution of the petrunkevitchi group (Virginia; not treated in detail in this paper) suggests an ancestral refu- gium in the Blue Ridge. But new and more complete distributional data for cave trechines in eastern United States, including Pseudanophthal- mus, strongly support the hypothesis that the Allegheny Plateau played a major role as a refugium during Pleistocene time and perhaps even ear- lier. At least two other introductions must be postulated to account for a) Darlingtonea and Ameroduvalius, of unknown affinities within the Tre- chini; and b) Xenotrechus, whose two Missouri species are unquestiona- bly very close to Chaetoduv alius, a genus of the Carpathians and Tran- sylvanian Alps. KEY TO SPECIES GROUPS OF THE ENGELHARDTI COMPLEX 1. Apex of aedeagus without constriction before apex (dorsal view) and without hatchet shape or sharp ventral cusp (lateral view) 2 Apex of aedeagus arrow-shaped in dorsal view because of subapical constriction and/or hatchet-shaped with sharp ventral cusp in lateral view (Figs. 5-11) engelhardti group 2(1). Pronotum disc with 1-4 long setae each side in addition to normal pubescence 3 Pronotum disc pubescent but without longer setae 5 3(2). Small, slender, usually depressed species (3. 1-4.5 mm) 4 Larger (4.4-6. 2 mm), slender, often convex species; aedeagal apex pro- duced and slender or (Rye Cove, VA) button-shaped (Figs. 48-54) jonesi group (new group) 4(3). Aedeagus with apex slender, moderately produced and attenuate, with very finely truncate knob in lateral view (Figs. 12-15) tennesseensis group (new group) Aedeagus with apex feebly produced and bluntly rounded at tip, median lobe weakly arcuate, basal bulb rather small, left copulatory piece very small (Figs. 20-25) hirsutus group 5(2). Aedeagus with apex slender and considerably produced and attenuate . .6 Aedeagus with apex briefly produced, either bluntly rounded or finely truncate, in some species also deflexed (Figs. 32-33) hubrichti group (new group) 42 Thomas C. Barr, Jr. 6(5). Aedeagus with apex strongly deflexed, falciform, slightly knobbed (Fig. 39); apical groove of elytron subparallel to suture; northwest GA, northeast AL alabamae group Aedeagus with apex not knobbed, not or but slightly reflexed at tip (Figs. 34-38); Pine Mountain, KY, and Scott Co. (Hunter Valley), VA hypolithos group (new group) Occasional specimens of P. alabamae and P. georgiae {alabamae group) have 1-3 long setae on the pronotum, so specimens from northwest Georgia or northeast Alabama should be checked against Fig. 39. engelhardti group Aedeagus more or less constricted before apex, which is arrow-shaped in dorsal view. Length 3. 6-5.4 mm, mean lengths about 4.0-4. 5 mm in a majority of species. In many species the aedeagal apex bears a sharp, ventral cusp when examined in lateral view (Figs. 5, 7-1 1); in others the apex is more or less attenuate and produced, with or without a terminal knob (Fig. 6). Form subdepressed, pronotum more or less cordiform and moderately to strongly transverse, pronotal disc without long setae in addition to normal pubescence (except in west-central Tennessee). Type species: P. engelhardti (Barber). Distribution. — Lee Co., VA; Anderson, Campbell, Claiborne, Decatur, Hamilton, Hancock, Marion, Perry, Rhea, Union cos., TN; Blount, Colbert, Dekalb, Jackson, Lauderdale, Lawrence, Madison, Marshall, Morgan cos., AL; Dade, Walker cos., GA. The group is appar- ently bicentric in distribution with several species along the Virginia- Tennessee border and the remainder in the lower Tennessee River valley, from Chattanooga westward. Pseudanophthalmus engelhardti (Barber) Anophthalmus engelhardti Barber 1928:195. Length 3.6-4. 1, mean 3.8 mm. Head rounded; pronotum subglabrous, 5/6 as long as wide, cordiform, greatest width in apical fifth just behind anterior marginal setae, sides subparallel in basal fifth, hind angles large, about right, sharp; elytra narrow, 1.7 times longer than wide, humeri not sharply angular, stria 1-3 well impressed but intervals essentially flat, apical groove elongate and bisinuate, joining apex of 3rd stria via crosier at level of 7th umbilicate puncture, usually 2 irregular rows of short but moderately dense pubescence each interval; aedeagus rather strongly arcuate, apex rather narrow, sharp ventral cusp slightly reflexed, para- meres slender with 3 apical setae. This species is known only from English Cave, Claiborne County, Tennessee (see Barr 1961 for location, description, and map), where it is Appalachian Pseudanophthalmus 43 relatively abundant on wet, rotting wood of old bridges and walkways. It is sympatric and syntopic with two more widely distributed species, P. rotundatus {engelhardti group) and P.pallidus {jonesi group). At present this is the only known case in the entire Appalachian valley of three- species sympatry in the same cave. Normally (three earlier visits) P. engelhardti is 15 to 20 times more abundant than P. rotundatus, but on 16 September 1979, at the end of an unusually wet summer, I took nine beetles in English Cave, five of which were P. rotundatus and four P. engelhardti. Pseudanophthalmus pallidus is known from small series in the Cedar Fork area a few kilometers south of English Cave, but collec- tions by at least five" different biospeleologists over the past fifty-two years have produced only a single specimen of P. pallidus from English Cave. Although Valentine (1945) suggested some form of sympatric specia- tion to explain coexistence of P. engelhardti and P. rotundatus, the more extensive range of P. rotundatus indicates allopatric speciation with sub- sequent dispersal and range overlap. The process of geographic specia- tion outlined by Barr (1967a) still appears to be the best explanation for species multiplication and divergence in Pseudanophthalmus . Invocation of other modes of speciation (e.g. chromosomal mutations resulting in aneuploid gametes in heterozygotes) is conceptually difficult and seems unnecessary to explain known distributions of related, sympatric species of anophthalmids. Pseudanophthalmus deceptivus, new species Fig. 5 Etymology. — Latin, deceptivus, “deceptive.” Diagnosis. — Closely similar to P. engelhardti in habitus, pronotum cordiform with approximately right hind angles, humeri somewhat rounded, longitudinal striae rather feebly impressed, intervals weakly convex; aedeagus much more arcuate than that of P. engelhardti, apical blade deflexed and quite narrowly truncate in lateral view, much con- stricted and broadly expanded in dorsal view. Description. — Length of holotype and one paratype 4.0-4. 4 mm. Form moderate, about as in P. engelhardti', rufotestaceous, shining, pronotum disc with sparse micropubescence, elytral disc with dense but rather short pubescence, elytral microsculpture finely transverse. Head rounded or a little longer than wide (L/W 1.02-1.14), labrum singly emarginate; last segment of maxillary palp 1.3- 1.4 as long as penultimate segment; antenna 0.6 body length. Pronotum cordiform, 0.85 as long as wide, anterior angles small and subdued, sides quite shallowly sinuate in basal fifth, hind angles large and about right, basal angles small and 44 Thomas C. Barr, Jr. Figs. 1-4: Pseudanophthalmus spp., engelhardti and tennesseensis groups. 1) P. wallacei, n.sp. 2) P. nortoni, n.sp. 3) P. fastigatus, n.sp. 4) P.pusillus, n.sp. All scales in mm. Appalachian Pseudanophthalmus 45 Figs. 5-11: Aedeagi of Pseudanophthalmus spp., engelhardti group left lateral view. 5) P. deceptivus, n.sp. 6) P. wallacei, n.sp., a-apex ventral view. 7) P. nortoni, n.sp. 8) P. fastigatus, n.sp. 9) P. nickajackensis, n.sp. 10) P. sequoyah, n.sp. 1 1) P. steevesi, n.sp. rounded; widths at apex and base subequal, about 0.7 greatest width, which occurs in apical fourth behind anterior marginal setae. Elytra elongate-oval, 1.6 times longer than wide, disc subconvex; prehumeral borders oblique, humeri rather prominent but rounded; inner 4 longitud- inal striae finely but rather shallowly impressed, intervals weakly convex, striae 5-8 progressively obsolescent; apical groove elongate, oblique, bi- sinuate, joining apex of 3rd stria at level of 7th umbilicate puncture. Aedeagus of paratype 0.55 mm long, strongly arcuate, basal bulb bent at nearly right angle to median lobe, apex briefly produced and deflexed, apical blade quite narrow in lateral view, conspicuously constricted and very broad in dorsal view. Type series. — Holotype female (American Museum of Natural His- tory), Fisher Cave, near the top of Newmans Ridge, between Blackwater and Kyles Ford, Fee Co., Virginia (Kyles Ford IV2 Quadrangle, 36°37'22" X 83'^03'40", elevation 480 m), 17 July 1979, T. C. Barr, Jr., T. C. Barr, III, and J. R. Holsinger. One male paratype (teneral), same cave, T. C. Barr, Jr., 30 September 1979. Measurements (mm). — Holotype, total length 4.37, head 0.84 long X 0.82 wide, pronotum 0.82 long X 0.98 wide, elytra 2.41 long X 1.49 wide, antenna 2.69. 46 Thomas C. Barr, Jr. Discussion. — This rare species coexists in Fisher Cave with P. longi- ceps (jonesi group), which is much more abundant (19:2). On 17 July, six beetles were collected, one of which was the female holotype of P. decep- tivus. On 30 September, 14 beetles were collected, one of which was the teneral male paratype, which was found under rocks at the bottom of the steep talus slope just inside the entrance of the cave. Morphologically P . deceptivus is very similar to P. engelhardti, which inhabits English Cave 45 km west of Fisher Cave. At present 10 pairs of Pseudanophthalmus species are known to coexist in the same cave systems in the Appalachian valley (excluding the ecologically and geologically very different Green- brier valley of West Virginia), in addition to the unique case of three species coexisting in English Cave. Pseudanophthalmus wallacei, new species Figs. 1,6 Etymology . — Patronymic honoring Mr. Richard L. Wallace, Knox- ville, Tennessee, for his assistance in collecting this rare species. Diagnosis. — Closely similar to P. engelhardti, differing in the sharp, right (rather than obtuse) hind angles of the pronotum and broader elytra (L/ W 1.6 vs. 1.7 in P. engelhardti)', aedeagal apex narrowly produced and truncate, without ventral cusp. Description. — Length 4. 3-4. 6, mean 4.5 mm (N = 4). With general form, color, microsculpture, and pubescence of P. engelhardti. Head rounded, labrum rather deeply and singly emarginate; antenna about 0.6 body length. Pronotum transverse-cordiform, 0.85 as long as wide, wid- est in apical fifth at level of anterior marginal setae, anterior angles a little more prominent than in P. engelhardti, hind angles large, sharp, about right, base behind angles perpendicular to midline (rather than inclined forward as in P. engelhardti) and distinctly though shallowly emarginate between angles. Elytra 1.6 times longer than wide, broader and slightly more convex than in P. engelhardti, margins a little more rounded; pre- humeral borders oblique, humeri rounded but still rather prominent; apical groove elongate, bisinuate, with crosier running to 3rd stria at or slightly in advance of 7th umbilicate puncture. Aedeagus of paratype 0.59 mm, strongly arcuate, apex as in P. holsingeri. Type series. — Holotype female (American Museum of Natural His- tory) and one paratype female. Weaver Cave, 3.0 km N Clinton, Ander- son Co., Tennessee on the southeast (left) side of Clinch River (Norris 71/2' Quadrangle, 36°08'19" X 84°06'55"), 3 February 1980, R. L. Wallace, One female paratype, same cave, 19 April 1965, J. A. Payne; one female paratype, same cave, 1 October 1980, R. L. Wallace; one male paratype, same cave, 1 March 1981, R. L. Wallace. Measurements (mm). — Holotype, total length 4.32, head 0.84 long X Appalachian Pseudanophthalmus 47 0.76 wide, pronotum 0.76 long X 0.90 wide, elytra 2.30 long X 1.48 wide, antenna 2.70 long. Discussion. — This rare species is probably closest to P. holsingeri, with which it shares the same aedeagal pattern. It is remarkable that another species {P. paynei, tennesseensis group) exists in Moores Bridge Cave only 1.3 km south of Weaver Cave, on the same side of Clinch River, but P. paynei has not yet been encountered in Weaver Cave, nor has P. wallacei been taken in Moores Bridge Cave; the two species are potentially sympatric. From P. wallacei, P. paynei is distinguished by several pronotal characters: the hind angles are obtuse, the sides are convergent and not sinuate, and the disc bears two pairs of long setae. Pseudanophthalmus rotundatus Valentine Valentine 1932:271. For many years a very rare species, P. rotundatus was originally de- scribed from English Cave, also the type locality of P. engelhardti\ it is more robust than P. engelhardti, with more angular humeri, the prehu- meral borders perpendicular to the midline, deeper elytral striae, and a broader apical blade on the aedeagus. Jeannel (1949) reported the species from Parkeys Cave, Hancock County, Tennessee, farther up the Powell River valley. J. R. Holsinger and T. C. Kane recently took P. rotundatus and P. delicatus {hirsutus group) by trapping in Smith Cave, near Rose Hill, Lee County, Virginia. In July 1979, J. R. Holsinger, T. C. Barr, HI, and I collected 33 specimens of P. rotundatus in Subers Cave, 2.7 km east-northeast of Hopewell (Coleman Gap IVi Quadrangle, 36°34'32" X 83‘^25'52"), Hancock County, Tennessee, in a locality intermediate between English Cave and Smith Cave. The beetles occurred beside a small stream, under rocks, sticks, and leaves. The range of this species thus extends along the Powell River valley for approximately 30 km, overlapping the range of P. delicatus at the northeast end and coexisting in English Cave with P. engelhardti and P. pallidus. Pseudanophthalmus sidus Barr Barr 1965:64. This rare species is known only from three specimens taken in Mere- dith Cave, at Shanghai Boat Dock, Campbell County, Tennessee, about 45 km southwest of English Cave in the lower Powell River valley. It is smaller (3.6-4. 1 mm) than either P. engelhardti or P. rotundatus, subpar- allel, depressed, polished (microsculpture obsolescent at center of elytral disc); prehumeral borders perpendicular to midline, humeri angular; pronotum 0.8 as long as wide, hind angles very large and right. The aedeagus is shorter (0.51-0.55 mm) and less arcuate than that of P. rotun- 48 Thomas C. Barr, Jr. datus, with moderately large apical blade, the apex unusually narrow and rather feebly arrow-shaped when seen from above. The holotype male, originally in my collection, has been deposited in the American Museum of Natural History. The species was described on a pair of specimens, and only one additional male has been taken (September 1979, R. Wallace and T. Barr). Pseudanophthalmus holsingeri Barr Barr 1965:63. Length 4.2-5. 4, mean 4.8 ± SD 0.3 mm (N = 38). Head 1.2 times as long as wide, labrum singly emarginate or (often) with low and wide median lobe. Pronotum transverse-cordiform, greatest width in apical fourth behind anterior marginal setae, hind angles large, more or less right (but often slightly obtuse), secondary basal angles small but conspicuous. Elytra 1.6 times longer than wide, prehumeral borders oblique, humeri prominent though a little rounded; disc subconvex, flatter near middle, 3 longitudinal striae present and traces of outer striae, intervals flat to weakly convex; apical groove elongate, bisinuate, running to 3rd stria apex via crosier at level of 7th umbilicate; pubescence dense but rather short. This species is known only from the Young-Fugates Cave system near Gibson Station, western Lee County, Virginia, where it is relatively abundant. From the majority of species of ihQ engelhardti group it differs in having the aedeagal apex deflexed and simply and finely truncate (lateral view), but without a sharp ventral cusp. Pseudanophthalmus nortoni, new species Figs. 2, 7 Etymology. — Patronymic honoring Mr. Russell M. Norton, discov- erer of this and many other species of cave beetles. Diagnosis. — Similar to P.fulleri, differing in the perpendicular pre- humeral borders of the elytra, deeply impressed elytral striae, and strongly arcuate aedeagus. Description. — Length of unique holotype 4.28 mm. Form robust and subconvex. Head rounded, as wide as long; labrum strongly trilobed in anterior margin; genae subglabrous; 3rd segment of maxillary palp 0.75 as long as 4th segment; antenna 0.6 body length. Pronotum transverse- subquadrate, 0.9 as long as wide, anterior angles prominent, apex 0.9 as wide as base, base 0.8 greatest width, which occurs in apical fourth behind anterior marginal setae; sides subparallel in basal fifth, hind angles slightly more than right, sharp and slightly produced, secondary angles of base prominent; disc with one long seta each side, otherwise Appalachian Pseudanophthalmus 49 with very sparse and short pubescence. Elytra 0.67 times longer than wide, elongate-oval, prehumeral borders perpendicular to midline, humeri serrate and setose; all longitudinal striae impressed, intervals subconvex, with 1-4 rows of pubescence, apical recurrent groove bisinuate, conspicu- ously oblique to suture, connected to 3rd stria by crosier; microsculpture obsolete on center of disc. Aedeagus 0.57 mm long (holotype), strongly arcuate, apex with ventral cusp. Type series. — Unique holotype male (American Museum of Natural History), Grassy Creek Cave, 0.7 km S Washington, Rhea Co., Tennes- see, 3 July 1967, R. M. Norton. Measurements (mm)." — Holotype, total length 4.28, head 0.80 long X 0.80 wide, pronotum 0.86 long X 0.95 wide, elytra 2.30 long X 1.38 wide, antenna 2.63 long, aedeagus 0.57. Discussion. — The type locality cave lies between the Allegheny front (Waldens Ridge) and the west (right) bank of Tennessee River; it is developed along a sinking section of Grassy Creek (Evensville IVi Quad- rangle, 35^3 T40" X 84°54"44"). Judging from overall habitus and aedea- gal morphology, P. nortoni is closer to P.fulleri, P. nickajackensis, and other “southern” species of the group than to the more geographically remote species along the Virginia-Tennessee border. Pseudanophthalmus fulleri Valentine Valentine 1932:272. Barr 1965:66. Length 4.0-4. 5 mm; head rounded, labrum singly emarginate; prono- tum transverse, almost subquadrate, with shallow lateral sinuation and large, obtuse hind angles; elytra with angular humeri, widest at middle, subconvex, striae very shallow, intervals flat to weakly convex, apical groove elongate, oblique, bisinuate; apex of aedeagus hatchet-shaped, with sharp ventral cusp, moderately arcuate. In a previous paper (Barr 1965) I reported a 20 km range extension for this species, from Tennessee Caverns, Hamilton County, Tennessee (type locality), to Howards Waterfall Cave, 3.5 km southwest of Trenton, Dade County, Georgia, predicting that future collecting in Dade County caves “will eventually yield more P.fulleri and perhaps even the less well known P. digitus. ...” In 1967 and subsequently, J. R. Holsinger, S. B. Peck, and their associates collected P. fulleri in the following Dade County caves, all of them in Lookout Valley on the west side of Lookout Mountain: Byers Cave, 2.5 km SW Rising Fawn; Johnsons Crook Cave, 7.5 km NE Rising Fawn; Howards Waterfall Cave (additional speci- mens); Morrison Cave, 3.5 km E Trenton; Sittons Cave, 3.5 km SE Trenton; Deerhead Cove Cave, near Trenton (T. lies and A. Dobson, •eg). 50 Thomas C. Barr, Jr. In Tennessee Caverns, Byers Cave, and Johnsons Crook Cave P.fulleri coexists with P. digitus {hirsutus group). Related vicar species occur in a series of en echelon anticlinal valleys nearby: P.fastigatus in the valley of West Chickamauga Creek to the east, P. sequoyah in Wills Valley to the southwest, and P. nickajackensis still farther west in the Tennessee River valley. Pseudanophthalmus fulleri was the first of a series of closely similar, allopatric species to be discovered in southeastern Tennessee, northwest Georgia, north Alabama, and west-central Tennessee. This ^"fulleri series” includes all species of the engelhardti group in this southern area except the rather different P. loedingi. All species of the series possess a hatchet- shaped aedeagal apex with sharp ventral cusp, resembling the northern species P. engelhardti, P. rotundatus, P. sidus, and P. deceptivus in this respect. The elytral humeri are minutely serrulate in iht fulleri series, and occasional specimens have a single long seta on each side of the pronotal disc. The aedeagal apex in P. loedingi, in contrast, is deflexed, produced, attenuate, and finely but distinctly truncate; polytypic P. loedingi occu- pies caves of the Cumberland plateau north of the Tennessee River. South of the Tennessee River near and west of Guntersville, caves are occupied by polytypic P. meridionalis Valentine, a member of the fulleri series. Pseudanophthalmus fastigatus, new species Figs. 3, 8 Etymology. — Latin fastigatus, “sloping down, tapered.” Diagnosis. — Resembles P.fulleri but more polished, humeri much sharper, greatest width of elytra at apical third, elytral apexes elongate- attenuate (not rounded), longitudinal striae deeper, intervals subconvex; aedeagus smaller, apical blade larger, wider, and more deflexed. Description. — Length 4.4-4. 5 mm (N = 3). Form moderately slender and elongate, subconvex; rufotestaceous, very polished, pubescent; ely- tral microsculpture finely and obsoletely transverse. Head 1.2 times longer than wide, labrum singly emarginate; last 2 segments of maxillary palp subequal; antenna two-thirds body length, segments rather thick. Prono- tum transverse, more than 0.8 as long as wide, greatest width in apical third behind level of anterior marginal setae (as in P.fulleri), width at apex only seven-eighths width at base, which is about 0.85 greatest width; anterior angles subdued, sides very shallowly sinuate in basal fifth, hind angles large and slightly obtuse, secondary angles of base broad, low, rounded. Elytra highly diagnostic: 1.6 times longer than wide, prehu- meral borders perpendicular to midline, humeri sharply angular, greatest elytral width in apical third (rather than at middle) with sides behind Appalachian Pseudanophthalmus 51 gradually convergent, apexes attenuate; disc subconvex; short scutellar stria present, longitudinal striae moderately deep, intervals subconvex, inner 3 striae fairly regular, outer striae progressively obsolescent; apical groove elongate, oblique, feebly bisinuate, with anterior crosier at level of 7th umbilicate puncture. Aedeagus of holotype 0.55 mm, similar to that of P. fulleri but a little smaller, apical blade larger, wider, and more deflexed. Type series. — Holotype male (American Museum of Natural His- tory), and one male paratype. Horseshoe Cave, 7 km SW Chickamauga, Walker Co., Georgia, 10 June 1967, J. R. Holsinger, S. B. Peck, A. Fiske, and R. Baroody. One paratype male, same cave, 21 June 1967, S. B. Peck and A. Fiske. Measurements (mm). — Holotype, total length 4.40, head 0.92 long X 0.76 wide, pronotum 0.83 long X 0.98 wide, elytra 2.36 long X 1.48 wide, antenna 2.84, aedeagus 0.55. Discussion. — This small, highly polished species with tapered elytra was discovered in a cave in the valley of West Chickamauga Creek, on the opposite (east) side of Lookout Mountain from caves occupied by P. fulleri and P. digitus. It is probably closest to P. fulleri but is readily distinguished by the characters cited in the diagnosis. Horseshoe Cave is stratigraphically and structurally isolated from the caves of Lookout Valley. Pseudanophthalmus nickajackensis, new species Fig. 9 Etymology. — Geographic place name. Diagnosis. — A large species with cuspate aedeagal apex, moderately transverse pronotum, oblique prehumeral borders, and 3 feebly impressed elytral striae, the apical groove oblique and bisinuate, running to apex of 3rd stria; aedeagus strongly arcuate, apex produced, recurved ventral cusp highly diagnostic. Description. — Length 4. 1-5.0, mean 4.5 ± SD 0.6 mm (N = 40). Head slightly longer than wide, dorsum subglabrous; anterior margin of labrum with low median lobe; 3rd segment of maxillary palp 0.8 as long as 4th segment. Pronotum 0.8 as long as wide, apex and base widths subequal and 0.75 greatest width, which occurs in apical fourth; sides sinuate in basal fifth, hind angles about right, base with prominent secondary angles; disc subglabrous. Elytra two-thirds longer than wide, elongate-oval, depressed, prehumeral borders oblique to midline, humeri nevertheless rather prominent; 3 inner striae shallow, intervals flat, outer striae obsolescent, 3-5 irregular rows of fine pubescence per interval, apical recurrent groove elongate, oblique to suture, bisinuate, connected to 3rd stria via crosier. Aedeagus 0.61-0.65 mm long, strongly arcuate. 52 Thomas C. Barr, Jr. apex produced and cuspate ventrally, cusp sharply recurved. Type series. — Holotype male (American Museum of Natural His- tory), and 87 paratypes, Nickajack Cave, 1.0 km S Shellmound Station near the mouth of Nickajack Cove, Marion Co., Tennessee, 2 July 1967, S. B. Peck and A. Fiske; 10 paratypes, same cave, 29 July 1967, S. B. Peck, A. Fiske, J. E. Cooper; 8 paratypes, same cave, 1 1 November 1967, J. E. Cooper, J. R. Holsinger, M. Richmond. Measurements (mm). — Holotype, total length 4.72, head 0.88 long X 0.84 wide, pronotum 0.88 long X 1.06 wide, elytra 2.65 long X 1.59 wide, antenna 2.08, aedeagus 0.65. Discussion. — Pseudanophthalmus nickajackensis is readily diagnosed by the recurved cusp on the ventral side of the aedeagal apex. The species is known only from the type locality, a large stream cavern which was unfortunately flooded late in 1967 by the waters of Nickajack Reservoir and is no longer accessible. The cave was described by Barr (1961:305), and a map of the cave was published by Matthews (1971: 133). Nickajack Cave is on the south (left) side of the Tennessee River; most of the cave lies in Marion County, Tennessee, but a small part extends under Jack- son County, Alabama. Pseudanophthalmus sequoyah, new species Fig. 10 Etymology. — Place name, noun in apposition. Diagnosis. — Resembling P. fulleri, but differing in slightly longer head, pronotum hind angles right, elytral humeri less angular and inter- vals convex, aedeagus more arcuate with shorter copulatory pieces. Description. — Length 4. 5-5.0, mean 4.7 ± SD 0.1 (N = 10). Moder- ately robust, subconvex, pubescent, rufotestaceous; microsculpture of head rather strongly isodiametric, satiny-smooth isodiametric at center of pronotum apex and base but very finely transverse elsewhere on disc, very fine and transverse without forming meshes on elytral disc. Head one-eighth longer than wide, labrum singly emarginate, maxillary palp with last segment a little longer than penultimate segment; antenna two- thirds body length. Pronotum transverse-subquadrate, a little less than 0.9 as long as wide, width at apex and base subequal and about 0.75 greatest width, which occurs in apical fifth; anterior angles very promi- nent, produced; sides subparallel in basal fourth, hind angles large, about right, somewhat reflexed, and blunt; base with conspicuous secondary angles; disc glabrous. Elytra almost subparallel, 1.6 times longer than wide, prehumeral borders not quite perpendicular to midline, humeri prominent though rounded, disc subconvex; very short scutellar stria present, inner 3 longitudinal striae well impressed, outer striae progres- sively feebler and irregular, intervals convex with 3-4 rows of pubescence; Appalachian Pseudanophthalmus 53 apical groove very long and oblique, feebly bisinuate, running to 3rd stria via crosier well in advance of 7th umbilicate puncture; elytral apexes separately rounded. Aedeagus 0.57 mm long (paratype), closely similar to that of P.fulleri but considerably more arcuate. Type series. — Holotype male (American Museum of Natural History) and 6 paratypes, Ellis (=Sequoyah) Cave (AL 333), 3.3 km SW Sulphur Springs, T4S/R10E/S20, Dekalb Co., Alabama, 25 August 1965, S. B. Peck. Three paratypes, same cave and collector, 9 August 1965. Measurements (mm). — Holotype, total length 4.64, head 1.06 long X 0.94 wide, pronotum 0.94 long X 1. 10 wide, elytra 2.73 long X 1.67 wide, antenna 3.08. Discussion. — This species occupies a cave at the northwest end of Wills Valley, west of the Wills anticline; it is thus geologically isolated from most other Dekalb County caves, which are east of the anticline axis and inhabited by P. alabamae. Pseudanophthalmus sequoyah coex- ists with P. assimilis, a smaller and rarer species belonging to the hirsutus group. Alabama caves are assigned a sequential number as they are added to the official list of caves in that state (Varnedoe 1973). For example, Ellis Cave is AL 333, and nearby Kudzu Cave, discovered and explored much later, is AL 734. Pseudanophthalmus steevesi, new species Fig. 1 1 Etymology . — Patronymic honoring Mr. Harrison R. Steeves, Jr., Birmingham, Alabama. Diagnosis. — A member of the fulleri series, with subquadrate prono- tum and very shallow elytral striae; aedeagus closely resembling that of P. fulleri but more sharply recurved at apex. Description. — Length 4. 2-4.6, mean 4.4 ± SD 0.2 mm (N = 7). Form subparallel, depressed, pubescent, rufotestaceous; elytral microsculpture obsolescent-transverse, shining. Head slightly wider than long but basi- cally rounded; labrum with very low and wide median lobe; last 2 seg- ments of maxillary palp subequal; antenna two-thirds body length. Pronotum subquadrate, 0.8 as long as wide, widths at apex and base subequal and about 0.7 greatest width, which occurs at apical fifth; sides subparallel in basal sixth, anterior angles not prominent, hind angles large and more or less right (in some specimens slightly produced and very sharp), secondary basal angles prominent but rounded. Elytra sub- parallel, rather depressed, prehumeral borders slightly oblique and humeri less prominent than in other species of the fulleri series (especially in specimens from the type locality cave); longitudinal striae very shal- 54 Thomas C. Barr, Jr. low, 3 or 4 inner striae complete and outer striae obsolescent, intervals flat; scutellar stria very short, vestigial; apical recurrent groove fairly long, oblique, feebly bisinuate or not, joining 3rd stria behind level of 7th umbilicate puncture. Aedeagus 0.55-0.59 mm long, about as in P.fulleri but not quite as arcuate, median lobe more slender, apex with ventral cusp more strongly recurved, parameres with 4 quite short apical setae. Type series. — Holotype male (American Museum of Natural History) and one paratype male, Randolph Cave, 1.7 km SW Blount Springs, Blount Co., Alabama (AL 414, T13S/ R3W/S12), 15 June 1967, Harri- son R. Steeves, Jr., and Thomas C. Barr, Jr. Female paratype, same cave, 12 December 1965, S B. Peck. Measurements (mm). — Holotype, total length 4.25, head 0.73 long X 0.76 wide, pronotum 0.78 long X 0.98 wide, elytra 2.39 long X 1.45 wide, antenna 2.84, aedeagus 0.59. Discussion. — This species has also been collected in Rickwood Cav- erns (AL 236; one male and one female, 24 November 1972, W. W. Torode), Bryant Cave (AL 355; female, 19 March 1966, S. B. Peck), and Horse Cave (AL 721; male, 11 April 1970, W. W. Torode). Rickwood Caverns and Bryant Cave are both in T13S/R3W near Randolph Cave, and Horse Cave is in T12S/R1W/S10, all three caves in Blount County. A total of 7 specimens of this species was examined, including the type series. tennesseensis group (new group) Aedeagus with apex not constricted before base, not arrow-shaped in dorsal view; apex more or less finely truncate and slightly knobbed; copulatory pieces as in engelhardti group, left piece large and spinulose. Length 3. 6-4. 9 mm, mean lengths 3.8-4. 1 mm. The group includes four depressed, pubescent species with 2 long setae each side of the pronotum disc. The species occupy caves in Roane, Anderson, Knox, and Union counties west and north of Knoxville, in the Appalachian valley of east Tennessee. Type species: P. tennesseensis Valentine. Distribution. — Roane, Anderson, Knox, Union cos., TN. Discussion. — All known species of the group are closely similar, reflecting relatively recent common ancestry, and they are best differen- tiated by aedeagal structure. Barriers separating the species include both structural/ stratigraphic ones and the Clinch and Powell rivers. From other small species with discal setae on the pronotum the species of the tennesseensis group are separated by the large left copulatory piece (very small in hirsutus group) and produced, apically knobbed aedeagus (bluntly rounded in hirsutus group and in P. pallidus, a small species of the jonesi group). Appalachian Pseudanophthalmus 55 Figs. 12-15: Aedeagi of Pseudanophthalmus spp. tennesseensis group, left lateral view. 12) P. tennesseensis Valentine 13) P.pusillus, n.sp. 14) P.paynei, n.sp. 15) P. unionis, n.sp. Pseudanophthalmus tennesseensis Valentine Fig. 12 Pseudanophthalmus tenesensis Valentine 1937:98 (emendation). Length 3. 7-4. 2, mean 3.9 ± SD 0.2 mm (N = 12). Slender and depressed, somewhat parallel, pubescent. Head rounded. Pronotum wid- est in apical seventh, apex one-eighth wider than base, sides convergent and shallowly sinuate before somewhat obtuse, sharp, and slightly reflexed hind angles; base with prominent secondary angles; disc subgla- brous except 2-3 long setae each side. Elytra with humeri prominent, prehumeral borders perpendicular to midline, striae shallow, intervals nearly flat, apical groove short, joining 3rd stria behind level of 7th umbilicate puncture; pubescence of disc dense and not in rows. Aedeagus 0.61-0.65, mean 0.63 mm, moderately arcuate, apex with small button. Pseudanophthalmus tennesseensis occupies caves in a long strike band of the Knox group which lies south of the Copper Creek fault in Knox and Roane counties, Tennessee. It is known from four caves: Grand Caverns (type locality, also known as Atomic Caverns), George Light Cave (1.7 km S Solway), and Rock Hill Cave (3.3 km S Heiskell), all in Knox County; and Eblen Cave (1.7 km SE Bradbury), Roane County. The known geographic range is approximately 40 km long. In all the caves P. tennesseensis was collected from under rocks on damp silt, not beside streams. Total length measurements given by Barr (1965:66) were 0. 1-0.2 mm too small; the aedeagal lengths I previously reported are incorrect because of a typographical error. 56 Thomas C. Barr, Jr. Pseudanophthalmus pusillus, new species Figs. 4, 13 Etymology . — Latin pusillus, “very little.” Diagnosis. — Closely similar to P. tennesseensis, differing in the much less prominent anterior angles of the pronotum, less subparallel form of the elytra, less dense pubescence, and smaller aedeagus. Description. — Length 3. 7-3.9, mean 3.8 ± SD 0. 1 mm. Closely similar to P. tennesseensis, differing as follows: Head a little longer than wide, labrum margin without median lobe; pronotum with anterior angles subdued, hind angles more sharply reflexed; elytra with less prominent humeri, margins more arcuate (less subparallel), striae feebler, pubes- cence less dense. Aedeagus smaller, 0.56-0.57 mm long, basal bulb much less sharply reflexed, apex a little more attenuate and apical knob slightly larger. Type series. — Holotype male (American Museum of Natural History) and one female paratype, Martin Cave, 7.2 km SW Clinton beside Southern Railroad tracks, Anderson Co., Tennessee, 23 November 1963, R. M. Norton; two male paratypes, same cave, 5 January 1966, R. M. Norton. Measurements (mm). — Holotype, total length 3.67, head 0.74 long X 0.67 wide, pronotum 0.74 long X 0.76 wide, elytra 2.06 long X 1.23 wide, antenna 2.43, aedeagus 0.57. Discussion. — P. pusillus is a vicar species of P. tennesseensis, from which it is geographically separated by the Copper Creek fault. It occu- pies caves in Ordovician limestones on the west (right) side of Clinch River, and is known only from the type locality. This and the next two species share a relatively recent ancestry with P. tennesseensis. Pseudanophthalmus paynei, new species Fig. 14 Etymology. — Patronymic honoring Dr. J. A. Payne, discoverer of this species. Diagnosis. — Closely similar to P. tennesseensis, differing in nonsinu- ate, simply convergent sides of pronotum, less prominent humeri, and more slender and more produced apex of median lobe of aedeagus with less sharply defined terminal knob. Description. — Length 3. 6-4. 9, mean 4. 1 ± SD 0.3 mm (N = 13). Head about 1.15 longer than wide; labrum trilobed; antenna less than 0.6 body length. Pronotum 0.85 as long as wide, greatest width at apical fourth behind anterior marginal setae, sides convergent to obtuse hind angles, secondary basal angles moderate, 2 long setae either side of disc. Elytra 1.6 times longer than wide, sides feebly rounded, humeri rather promi- Appalachian Pseudanophthalmus 57 nent but not sharply angular (as in P. tennesseensis), prehumeral borders slightly oblique, disc a little depressed; 5 or more shallowly impressed longitudinal striae, intervals flat or feebly convex, apical groove rather short, bisinuate, running to 3rd stria at level of 8th umbilicate puncture, without crosier. Aedeagus of paratype 0.63 mm long, arcuate, apex grad- ually attenuate and very finely truncate, median lobe more slender and apical knob less pronounced than in P. tennesseensis. Apex thicker and less abruptly attenuate and apical knob less pronounced than in P.pusil- lus or P. unionis. Type series. — Holotype male (American Museum of Natural His- tory), 1 male, and 7 female paratypes, Moores Bridge Cave, 1.3 km N Clinton on east (left) side of Clinch River, Anderson Co., Tennessee (Norris IV2' Quadrangle, 36°0736" X 84°06'55"), 4 July 1967, R. M. Norton; one male and 2 female paratypes same cave, 22 July 1965, J. A. Payne. Measurements (mm). — Holotype, total length 4.10, head 0.84 long X 0.73 wide, pronotum 0.75 long X 0.88 wide, elytra 2.08 long X 1.33 wide, antenna 2.30 long. Discussion. — The type locality cave lies a short distance upstream from the old Moores Bridge across Clinch River, just north of Clinton on the Norris road (Tennessee Route 61). A steep descent of 45 m opens into a strike gallery 9 to 15 m wide and 180 m long, with high, irregular ceiling and damp floor. Beetles occurred about “200 ft. from entrance under rotten log and stones” (Payne) and near the foot of the entrance drop on “damp . . . floor with rotting wood” (Norton). This species is quite close to P. tennesseensis, differing primarily in the convergent sides of the pronotum, slightly less angular humeri and more rounded sides of the elytra, and the slightly more arcuate and apically less sharply knobbed aedeagus. One specimen of this species was collected by R. M. Norton in Flowstone Cave (Anderson County Park) and one specimen in Norris Quarry Cave No. 2; both caves are near Norris, in Anderson County, 10-12 km northeast of Moores Bridge Cave. Pseudanophthalmus unionis, new species Fig. 15 Etymology . — Geographic place name, from Union County, Tennessee. Diagnosis. — Closely similar to the preceding three species, differing in the sparser pubescence of the elytral disc, oblique prehumeral borders, less prominent anterior angles of pronotum, less subparallel elytra, and shorter and thicker aedeagus. Description. — Length 3. 7-4. 2, mean 4.0 mm (N = 3). Head one-sixth longer than wide, sides feebly rounded; labrum with low median lobe in 58 Thomas C. Barr, Jr. anterior margin. Pronotum a little transverse, 0.8 as long as wide, ante- rior angles and secondary basal angles less prominent than in P. tennes- seensis, margins feebly and briefly sinuate before small, obtuse, but sharp and slightly reflexed hind angles. Elytra 1.6 times longer than wide; in contrast to P. tennesseensis prehumeral borders slightly oblique to mid- line, disc a little more convex, sides slightly rounded, pubescence sparser and confined to rows on intervals; striae shallow, somewhat irregular, only inner 4 complete. Aedeagus short and thick, 0.55 mm long in holo- type, basal bulb bent at right angle to less arcuate median lobe, apex very briefly produced and finely knobbed. Female unknown. Type series. — Holotype male (American Museum of Natural His- tory), Wright Cave, Union Co., Tennessee, 31 December 1963, R. M. Norton; male paratype, same cave and collector, 6 January 1965; male paratype. Wolf Cave, Union Co., Tennessee, 22 August 1972, J. R. Hol- singer and D. C. Culver. Both caves were described and located by Barr (1961). Measurements (mm). — Holotype, total length 4.10, head 0.88 long X 0.71 wide, pronotum 0.71 long X 0.86 wide, elytra 2.31 long X 1.35 wide, antenna 2.77 long, aedeagus 0.55 long. Discussion. — The species is known only from two caves 2.3 km apart, both on the north side of Clinch River in the “Peninsula” (the interfluve between Clinch and Powell rivers). The principal barrier between the ranges of P. unionis and P. paynei is apparently the Clinch River. Although P. pusillus and P. unionis are on the same side of Clinch River, they are separated by Powell River and are in different strike bands of limestone. hirsutus group Aedeagus weakly arcuate, not constricted before apex, which is bluntly rounded or very feebly knobbed; copulatory sclerites very small, especially left piece. Length 3. 1-4.5 mm, mean lengths about 3. 7-4.2 mm. Form slender and depressed, pronotum transverse-cordiform with 2-4 long setae each side of disc, elytra more or less subparallel. Type spe- cies: P. hirsutus Valentine. Distribution. — Lee and Scott cos., VA; Hamilton, Marion, and Mon- roe cos., TN; Dade Co., GA; Dekalb Co., AL. Pseudanophthalmus hirsutus Valentine Fig. 20 Valentine 1931:252. Barr 1965:46 (in part). Length 4.0-4.5, mean 4.2 ± SD 0.2mm (N = 6). Head longer than wide; pronotum sides convergent to obtuse hind angles, anterior angles rather prominent; disc with 3 long setae each side; elytra elongate-subparallel. Appalachian Pseudanophthalmus 59 Figs. 16-19: Pseudanophthalmus spp., hirsutus group. 16) P. delicatus Valen- tine 17) P. sericus, n.sp. 18) P. paulus, n.sp. 19) P. digitus Valentine 60 Thomas C. Barr, Jr. striae very fine and regular, intervals flat, at least anterior to posterior discal puncture (in 2 specimens striae deepen toward apex); apexes atten- uate, not separately rounded. Aedeagus of a male from Saltpeter Cave 0.47 mm long, similar in general form to that of P. delicatus but a little larger. This species is known only from caves in the band of Newman (Missis- sippian) limestone at the base of Cumberland Mountain (the Allegheny front) in western Lee County, Virginia. J. F. Quinlan (pers. comm.) has traced subterranean stream flow from a point near Ewing, Virginia, to Cudjos Cave, the type locality for P. hirsutus, a distance of about 25 km. In July, 1979, J. R. Holsinger and Virginia Tipton collected six specimens of this species in Cumberland Mountain Saltpeter Cave, about 2.5 km east of (and probably connected with) Cudjos Cave. Pseudanophthalmus delicatus Valentine, new status Figs. 16, 21 Pseudanophthalmus hirsutus delicatus Valentine 1932:270. Barr 1965:46 (in part). Compared with P. hirsutus, smaller (P<0.02) 3. 1-4.2, mean 3.7 ±SD 0.3 mm (N = 31); pronotum a little more transverse with less prominent anterior angles, sides slightly but distinctly sinuate before sharp right or slightly acute hind angles; elytral striae deeper. Elongate, subparallel, depressed, pubescent. Head a little longer than wide, labrum with low median lobe, last segment of maxillary palp one-fourth longer than penultimate segment; antenna rather short, 0.6 body length. Pronotum 0.75-0.80 as long as wide, widest about apical sixth behind anterior mar- ginal setae, anterior angles fairly prominent (but less so than in hirsutus), sides convergent but shallowly and barely perceptibly sinuate just before more or less right hind angles (obtuse in hirsutus)', disc with 3 long setae each side. Elytra 1.7 times longer than wide, apexes not separately rounded, intervals weakly subconvex, apical groove elongate, bisinuate, joining 3rd stria via crosier. Aedeagus 0.39-0.45, mean 0.40 ± SD 0.02 mm long (N = 8), slender and weakly arcuate, basal bulb not sharply set off from median lobe, apex very briefly attenuate and bluntly rounded; copulatory pieces about as in P. engelhardti but left piece quite small; parameres with 3 apical setae. In a previous study of Appalachian valley cave beetles (Barr 1965) I misinterpreted a series of P. delicatus from Jones (=Ewing) Saltpeter Cave as P. /2. hirsutus. Actually P. hirsutus is confined to the cave system at the base of Cumberland Mountain, and P. delicatus inhabits a number of caves in Ordovician limestones along the Powell River valley in central Lee County, Virginia. Furthermore, there is a distinct geographic gap Appalachian Pseudanophthalmus 61 between known ranges of the two species in addition to profound struc- tural and stratigraphic barriers. Young-Fugate Cave, near Gibson Sta- tion, lies between Cumberland Mountain Saltpeter Cave and Jones Salt- peter Cave; more than 50 specimens of Pseudanophthalmus have been collected in Young-Fugate Cave, but all of them are P. holsingeri (engel- hardti group). The ubiquity of this small species in the Powell River valley of Lee County is surprising. From a point near Ewing northeast to Dryden (both in Lee County) the range is approximately 45 km long and 8 to 9 km wide, lying north of Wallen Ridge and south of Cumberland and Stone mountains. Caves from which the species is known include those developed in middle Ordovician limestone exposed in broad belts along the Powell Valley as well as caves in the Chepultepec member of the Knox group. At the southwest corner of its range P. delicatus is sympat- ric with P. rotundatus {engelhardti group). It is a species of pool and stream margins or (more often) muddy areas with rotting wood. The distribution as described is based on examination of 58 specimens from 13 Lee County caves: Gilley (type locality, 13 topotypes seen). Bowling, Cattle, Gallohan No. 1, Garrett, Jones Saltpeter, Molly Wagle, Poor Farm, Seal Pit, Smith, Spangler, Sweet Potato, and Unthanks caves. The majority of these specimens were collected between 1962 and 1979 by J. R. Holsinger, D. C. Culver, T. C. Kane, R. M. Norton, and T.C. Barr. Most of the caves were located and described by Holsinger (1975). Figs. 20-25: Aedeagi of Pseudanophthalmus spp., hirsutus group, left lateral view. 20) P. hirsutus Valentine 21) P. delicatus Valentine 22) P. sericus, n.sp. 23) P. digitus Valentine 24) P. ventus, n.sp. 25) P. assimilis, n.sp. 62 Thomas C. Barr, Jr. Pseudanophthalmus sericus, new species Fig. 17, 22 Etymology. — Latin sericus, “silken.” Diagnosis. — Resembles P. delicatus in general form of body and aedeagus and in weakly subconvex elytral intervals; differs in having greatest pronotum width at apical 4th (instead of apical 6th), singly emar- ginate labrum, and separately rounded elytral apexes. Description. — Length 3.7-4. 1, mean 3.9 ± SD 0. 1 mm (N = 10). Form slender, subparallel, depressed, pubescent; rufotestaceous; elytral micro- sculpture obsoletely transverse, shining. Head 0.15 longer than wide; labrum singly emarginate; maxillary palp with last 2 segments subequal in length; antenna rather short, about 0.6 body length. Pronotum cordi- form, 0.8 as long as wide, widest in apical fourth, apex subtruncate and anterior angles subdued, sides barely perceptibly sinuate in basal fifth, hind angles obtuse, secondary basal angles evanescent; disc with 2 long setae each side. Elytra rather subparallel, depressed, prehumeral borders oblique, humeri a little rounded; no scutellar stria; longitudinal striae feebly impressed (about as in P. delicatus, i.e. deeper than P. hirsutus), intervals weakly subconvex, inner 2 striae a little deeper and vaguely punctulate, striae 3 and 4 shallower, outer striae obsolete; apexes sepa- rately rounded; apical groove elongate, bisinuate, joining 3rd stria via crosier slightly in advance of 7th umbilicate puncture; disc moderately pubescent, 2 or 3 rows per interval. Aedeagus 0.42-0.44 mm long, similar to that of P. delicatus, but apex somewhat less produced and conspicu- ously deflexed; parameres with 3 setae. Type series. — Holotype male (American Museum of Natural His- tory), 4 male and 5 female paratypes. Lane Cave, in the valley of Moc- casin Creek, Scott Co., Virginia (Gate City IVi Quadrangle, 36°39'48" x 82°36'46"), 7 October 1967, J. R. Holsinger and George Titcomb. Measurements (mm). — Holotype, total length 4.10, head 0.84 long X 0.73 wide, pronotum 0.69 long X 0.82 wide, elytra 2.20 long X 1.29 wide, antenna 2.53, aedeagus 0.44. Discussion. — Lane Cave is developed in the same strike band of limestone (Maryville limestone, Ordovician) as Blair-Collins and Coley No. 2 caves, which lie a few kilometers northeast in Moccasin Creek valley. However, the latter two caves are occupied by P. thomasi ijonesi group), and only P. sericus has been collected in Lane Cave. Future collecting may demonstrate these two species to be sympatric. There seems little doubt, because of morphological similarity, that P. sericus, P. hirsutus, and P. delicatus are allopatric vicar species descended from a common ancestor. Appalachian Pseudanophthalmus 63 Pseudanophthalmus paulus, new species Fig. 18 Etymology. — paulus, “little, small.” Diagnosis. — Related to P. hirsutus and P. digitus', pronotum only 0.7 as long as wide, hind angles right, greatest width in apical third behind anterior marginal setae; elytral apexes individually rounded, only 2 com- plete longitudinal striae, inner 3 intervals subconvex, medial umbilicate punctures widely spaced: distance between 4th and 5th punctures about 1.2 times distance between 5th and 6th punctures. Description. — Length 3. 7-3. 8 mm. Head 0.1 longer than wide; ante- rior margin of labrum with low median lobe, 3rd segment of maxillary palp 0.75 as long as 4th segment; antenna 0.6 total length. Pronotum 0.7 as long as wide, more transverse than in other species of hirsutus group, apex about 0.1 wider than base and 0.7 greatest width, which occurs in apical third behind anterior marginal seta, sides subparallel in basal fifth, hind angles right, secondary angles of base obsolescent; disc with usual sparse, short pubescence and 2 or 3 long setae each side, but these less conspicuous than in other species of the group. Elytra 1.8 times longer than wide, depressed, prehumeral borders nearly perpendicular but humeri not unusually prominent, apexes individually rounded, with slight subapical emargination; only inner 2 longitudinal striae complete, but inner 3 intervals subconvex, each with 2-4 rows of fine pubescence; apical recurrent groove elongate, bisinuate, slightly oblique, connected to 3rd stria by crosier. Male unknown. Type series. — Holotype female (American Museum of Natural His- tory) and one female paratype, Nobletts Cave, 4.8 km W Sweetwater on the west side of Watson Ridge, Monroe Co., Tennessee, 4 January 1967, Russell M. Norton. Measurements (mm). — Holotype, total length 3.73, head 0.82 long X 0.73 wide, pronotum 0.67 long X 0.99 wide, elytra 1.99 long X 1.13 wide, antenna 2. 19. Discussion. — The cave in which P. paulus occurs is 0.8 km south of Tennessee Highway 68 and only 0.5 km east of the McMinn County line (Niota IVi Quadrangle, 35°35'33" X 84°31'02"). It is a small, muddy Stream cave about 1 15 km northeast of the range of P. digitus and 215 km southwest of that of P. hirsutus. Pseudanophthalmus digitus Valentine Figs. 19, 23 Valentine 1932:67. Length 3.6-4. 1, mean 3.9 mm (N = 9). Slender, subparallel, depressed. 64 Thomas C. Barr, Jr. Head about 1.15 times longer than wide, last segment of maxillary palp one-third longer than penultimate segment. Pronotum almost as long as wide, greatest width slightly behind anterior marginal setae, apex and base widths subequal, sides shallowly sinuate in basal fourth, hind angles acute, secondary angles of base prominent, disc with 2 long setae each side. Elytra nearly 1.8 times longer than wide, subparallel, depressed, humeri prominent even though prehumeral borders a little oblique, apex rounded; disc densely pubescent, inner 4 striae feebly impressed, traces of 5th and 7th striae discernible, apical groove elongate, bisinuate, subparal- lel to suture. Aedeagus 0.47-0.48 mm long, evenly arcuate, closely similar to that of P. hirsutus, with simple apex. Compared with P. hirsutus, the pronotum of P. digitus is longer with acute (rather than obtuse) hind angles, and the elytral apexes are con- spicuously more rounded. Neither Valentine (1932) nor Jeannel (1949) was able to examine a male of this species, but they correctly surmised that it is closely related to P. hirsutus. Examination of the aedeagus in freshly collected material confirms this supposition. The species coexists in caves of Lookout Valley, Hamilton County, Tennessee, and Dade County, Georgia, with P.fulleri {engelhardti group). The type locality is Tennessee Caverns, Hamilton County. J. R. Holsinger and S. B. Peck collected a total of 9 specimens from Johnsons Crook and Byers caves, near Trenton, Dade County. Pseudanophthalmus ventus, new species Fig. 24 Etymology. — Latin ventus, “wind.” Diagnosis. — Resembles P. digitus in size, habitus, rounded apexes of elytra, and aedeagal form; differs in longer head, shorter and wider pro- notum with convergent (=nonsinuate) sides and right hind angles, and more convex elytra. Description. — Length 3. 8-4. 2, mean 4.0 ± SD 0.1 mm (N = 6). Head 1.3 times longer than wide, labrum with low median lobe, last segment of maxillary palp one-third longer than penultimate segment; antenna 0.75 body length. Pronotum about 0.9 as long as wide, widest in apical sixth, sides convergent, not sinuate or barely sinuate before hind angles, which are right or slightly more than right; apex and base widths subequal and 0.75 maximum width, secondary basal angles subdued; disc with 2-3 long setae each side in addition to sparse, short pubescence. Elytra 1.8 times longer than wide, moderately convex, elongate-oval, prehumeral borders slightly oblique to midline, humeri nevertheless prominent, apex rounded, with slight subapical sinuation; longitudinal striae all discernible (outer striae obsolescent in one specimen only), at least at base, intervals feebly Appalachian Pseudanophthalmus 65 convex, with 2-3 rows of pubescence per interval; apical groove elongate, bisinuate, oblique to suture, connected to 3rd stria by crosier. Aedeagus 0.49-0.53 mm long, evenly arcuate, apex simply rounded as in P. hirsutus and P. digitus, slightly larger than in P. digitus. Type series. — Holotype male (American Museum of Natural History) and 5 paratypes. Blowing Cave, in town of Sequatchie, Marion Co., Tennessee (Sequatchie IVi Quadrangle, 35°07T4" x 85®35"39"), 29 August 1968, S. B. Peck. Measurements (mm). — Holotype, total length 3.96, head 0.92 long X 0.70 wide, pronotum 0.73 long X 0.86 wide, elytra 2.23 long X 1.25 wide, antenna 2.88. Discussion. — This species is known only from the type locality, the outlet of a subterranean stream on the west side of the Sequatchie Valley. It is closely similar to P. digitus, which occurs about 25 km southeast in another, parallel, anticlinal valley. The species is a little more densely pubescent than P. digitus, and there is a slight subapical sinuation in the elytral margin. More conspicuous character differences are cited in the diagnosis. Pseudanophthalmus assimilis, new species Fig. 25 Etymology. — Latin assimilis, “similar.” Diagnosis. — Similar to P. digitus, differing in less deeply sinuate pronotum sides with obtuse hind angles, secondary basal angles less prominent, outer longitudinal striae of elytra less regular, apical groove more deeply impressed, and aedeagus less arcuate, middle part of median lobe straighter. Description. — Length 4. 2-4.4 mm (N = 4). Form slender, depressed, pubescent; rufotestaceous, shining; elytral microsculpture very fine, dense, transverse-obsolescent. Head about 1.15 times longer than wide, labrum margin with prominent, broad, median lobe; maxillary palp with penultimate segment 0.8 as long as last segment; antenna 0.7 body length. Pronotum 0.9 as long as wide, cordiform, sides barely sinuate in basal fifth, hind angles obtuse and sharp, secondary angles of base subdued; greatest width in apical fifth, base width slightly greater than apex width and about 0.7 maximum width; 2-4 long setae on each side of disc in addition to sparse micropubescence. Elytra subparallel, depressed, 1.7 times longer than wide, prehumeral borders slightly oblique to midline, humeri prominent though somewhat rounded; inner 3 longitudinal striae well impressed, intervals subconvex, outer striae very irregular, 4th vestigial; no scutellar stria; apical groove deeply impressed, bisinuate, with crosier to 3rd stria or (1 of 4 specimens) running to 5th stria; discal 66 Thomas C. Barr, Jr. pubescence rather long, 2-3 rows per interval. Aedeagus 0.47-0.49 mm long, similar to that of P. digitus but middle portion of median lobe straighter; apex finely attenuate. Type series. — Holotype male (American Museum of Natural History) and one paratype female, Ellis (=Sequoyah) Cave (AL 333), 3.3 km SW Sulphur Springs in T4S/ RlOE/ S20, Dekalb Co., Alabama, 9 August 1965, S. B. Peck. One paratype male, same cave and collector, 25 August 1965. Measurements (mm). — Holotype, total length 4.24, head 0.82 long X 0. 73 wide, pronotum 0.78 long X 0.84 wide, elytra 2.33 long X 1.35 wide, antenna 2.86, aedeagus 0.49. Discussion. — From P. sequoyah, which also occurs in Ellis Cave, P. assimilis is easily distinguished by smaller size, cordiform pronotum with 2-4 discal setae each side, subdued anterior angles and sharper hind angles, and shorter and more clearly bisinuate apical groove; the aedea- gus is very different. A single female of this species was collected March 1, 1970, in Kudzu Cave (AL 734, T4S/R10E/S4), Dekalb County, by W. W. Torode. The specimen agrees closely with the type series but was not made a paratype. Ellis Cave lies on the west side of the Wills anticline in Wills Valley; caves on the east side of the anticline are inhabited by P. alabamae {alabamae group). Presumably P. assimilis is more closely related to P. digitus than is P. ventus, judging from external characters, but the less arcuate aedeagus of P. assimilis is distinctive. hubrichti group (new group) Aedeagal apex not arrow-shaped in dorsal view, but briefly attenuate and finely truncate {hubrichti) or bluntly rounded (egberti) or bluntly rounded and deflexed (vicarius). Mostly small species, 3. 4-4. 4 mm, but P. paradoxus ranges between 4. 3-4. 9 mm. Pronotum subcordiform, disc without long setae; form depressed, moderately slender. Type species: P. hubrichti Valentine. Distribution. — Giles, Russell, Scott cos., VA; Hawkins Co., TN. Pseudanophthalmus hubrichti Valentine Valentine 1948:13. Barr 1965:47 (in part). Length 3. 8-4.0 mm. Form a little subparallel, depressed; head rounded, labrum singly emarginate; pronotum transverse-cordiform, hind angles obtuse, sides very shallowly sinuate before base, which is oblique behind angles and shallowly emarginate between; elytra 1.6 times longer than wide, prehumeral borders oblique and humeri rounded, striae shallow and punctulate, intervals flat. Apical groove elongate, bisinuate, running via crosier to 3rd stria in advance of 7th umbilicate puncture, 5th and 6th Appalachian Pseudanophthalmus 67 \imbilicates widely separated; aedeagus 0.61-0.63 mm long, gently arcuate, apex briefly produced and bluntly truncate at tip. This species is known only from Dougherty Cave, 3.3 km northwest of Lebanon, Russell County, Virginia, on the left bank of Cedar Creek. The record of P. hubrichti from Banners Corner (=Big Spring) Cave in west- ern Russell County (Barr 1965:48) applies to the next species. Pseudanophthalmus sanctipauli, new species Fig. 26, 32 Pseudanophthalmus hubrichti: Barr 1965:47, NOT Valentine 1948:13. Etymology. — From town of Saint Paul, Virginia, near the type locality. Diagnosis. — Differs from P. hubrichti, to which it is closely similar, in more slender and more subparallel elytra with perpendicular prehumeral borders and deeper longitudinal striae; aedeagus longer, less arcuate, apex more slender but similarly truncate at tip. Description. — Length 3. 8-4.0, mean 3.9 mm (N = 5). Form and microsculpture as in P. hubrichti except for more slender elytra and more angular humeri. Head a little longer than wide, genae weakly convex; labrum with quite low but distinct median lobe; last segment of maxillary palp 1.3 times longer than penultimate segment; antenna two-thirds body length. Pronotum seven-eighths as long as wide, transverse-cordiform, apex and base widths subequal and each 0.8 greatest width, which occurs in apical sixth at level of anterior marginal setae; sides sinuate in basal sixth, hind angles obtuse, sharp, finely and minutely produced; base emarginate in middle, oblique behind angles, secondary angles small, sharp, slightly produced. Elytra narrow, 1.8 times longer than wide (1.6 in P. hubrichti), more nearly subparallel than in P. hubrichti, prehumeral borders perpendicular to midline, humeri angular but a little rounded, disc depressed; inner 4 longitudinal striae moderately impressed but a little irregular, outer striae obsolete, intervals weakly convex, each with about 3 rows of fine pubescence; apical groove elongate, bisinuate, directed toward apex of 3rd stria but without crosier or actual connec- tion; distance between 5th and 6th umbilicate punctures 1.2 times greater than between 4th and 5th umbilicates. Aedeagus 0.59 long in holotype, longer and less arcuate than in P. hubrichti, apex more slender and produced, gradually attentuate, finely truncate at tip; parameres rela- tively short and thick, each with 3 apical setae. Type series. — Holotype male (American Museum of Natural His- tory), 1 male and 1 female paratypes. Banners Corner Cave, near St. Paul, Russell Co., Virginia (Moll Creek IV2 Quadrangle, 36‘^52T7" X 82^ir55"), 12 April 1962, T. C. Barr. 68 Thomas C. Barr, Jr. Figs. 26-31: Pseudanophthalmus spp., hubrichti and hypolithos groups. 26) P. sanctipauli, n.sp. 27) P. paradoxus, n.sp. 28) P. hypolithos, n.sp, 29) P. scholasticus, n.sp., right elytral apex 30) P.frigidus, n.sp., right elytral apex 31)P. calcareus n.sp., right elytral apex. Appalachian Pseudanophthalmus 69 Measurements (mm). — Holotype, total length 3.84, head 0.83 long X 0.70 wide, pronotum 0.67 long X 0.80 wide, elytra 2.77 long X 1.54 wide, antenna 2.56 long, aedeagus 0.59 long. Discussion. — This species was erroneously determined as P. hubrichti in an earlier paper (Barr 1965). The caves inhabited by the two species are in different strike bands of limestone and are separated by major faults. Head and pronotum are closely similar in the two species, but the nar- rower elytra with more angular humeri and deeper and impunctate striae offer useful diagnostic characters. The aedeagi of the two species are rather different but suggest close relationship. J. R. Holsinger collected two males from Greears Sweet Potato Cave, Scott County, Virginia (Fort Blackmore IVi Quadrangle 36^49'03" X 82^3 IT 8"), which I have assigned to this species but have not made paratypes; this small cave is about 13 km southwest of Banners Corner Cave in the same band of Maryville limestone. Pseudanophthalmus egberti Barr Barr 1965:49. Length 3. 4-4.2, mean 4.0 ± SD 0.1 mm (N = 9). Slender, depressed, shining. Head a little longer than wide, labrum singly emarginate. Prono- tum strongly cordiform with deep antebasal sinuation, sides slightly di- vergent at base, hind angles correspondingly large, sharp, and acute; apex truncate, base rectilinear. Elytra slender, depressed, prehumeral borders a little oblique yet humeri fairly prominent, striae deep, intervals convex, only striae 1-3 complete but traces of 5-8 present; apical groove subparal- lel, rather short, running to 3rd stria, clearly bisinuate in 6 of 10 speci- mens examined. Aedeagus 0.61-0.65 mm long (greater lengths in the original description are a typographical error), moderately arcuate with well-formed basal bulb, apex briefly attenuate and slightly deflexed. Pseudanophthalmus egberti is known only from two caves in the New River valley of Giles County, Virginia: Starnes Cave (type locality) and Giant Caverns (=Hopkins Cave). Pseudanophthalmus quadratus Barr Barr 1965:60. Closely similar to P. egberti but a little less slender, hind angles right, elytral striae shallow, intervals less convex, apical groove bisinuate; aedeagal apex produced and bluntly rounded, not deflexed. Length 3.4- 3.7 mm, aedeagus 0.66-0.67 mm. This species was erroneously placed in \hQ gracilis group (Barr 1965) but is much closer to P. egberti and belongs here. It is known from three males taken in Straleys Cave in the valley of 70 Thomas C. Barr, Jr. New River at Eggleston, Giles County, Virginia, and is the northernmost representative of the engelhardti complex. Pseudanophthalmus vicarius Barr Barr 1965:48. Length 3. 5-4.4, mean 3.9 mm. Similar to P. egberti but more robust and more convex. Pronotum with apex truncate, hind angles large and slightly acute, antebasal sinuation deep, basal angles weakly developed, base slightly concave. Elytra broader, subconvex, intervals less convex, 4 complete inner striae, apical groove bisinuate, running to 3rd stria (no crosier) at level of 7th umbilicate. Labrum margin somewhat variable, often with low median lobe. Head more rounded than in P. egberti, tenerals and late tenerals with small, lemon-shaped eye rudiment visible. Aedeagus similar to that of P. egberti but apex more slender, briefly produced, and deflexed. This species occurs in caves of the Maiden Spring area, Tazewell County, Virginia: Hugh Young Cave (type locality), Bowen Cave, Fallen Rock Cave, Gully Cave, and Lost Mill Cave (see Douglas 1964 and Holsinger 1975 for locations and descriptions). Pseudanophthalmus paradoxus, new species Figs. 27, 33 Etymology. — Latin paradoxus, “paradox.” Diagnosis. — A large species with greatly reduced dorsal pubescence, the aedeagal apex deflexed and bluntly rounded at the tip, as in P. vicarius. Description. — Length 4. 3-4.9, mean 4.7 ± SD 0.2 mm (N = 8). Form rather robust and depressed; shining, microsculpture on elytral disc obso- lescent and finely transverse; dorsally subglabrous, pubescence very short on pronotum and elytral disc, denser internal to elytral margins. Head longer than wide, a little depressed, sides moderately rounded, labrum singly emarginate; last segment of maxillary palp one-sixth longer than penultimate segment. Pronotum transverse-cordiform, 0.83-0.85 as long as wide, widest in apical fourth, sides convergent to prominent antebasal sinuation, apex wider than base and 0.75 maximum width; hind angles large, sharp, nearly right, basal impressions short, deep, oblique. Elytra 1.7 times longer than wide, elongate-subparallel, depressed; humeri moderate, finely serrulate, setose; inner 4 striae moderately impressed and vaguely punctulate, intervals weakly convex; apical groove elongate, bisinuate, connected to 3rd stria via crosier at level of 7th umbilicate puncture. Aedeagus 0.95-1.00 mm long, basal bulb large and flexed at Appalachian Pseudanophthalmus 71 less than right angle to median lobe, which is swollen near base of inter- nal sac; apex briefly attenuate, deflexed, slightly knobbed, rounded at tip; parameres rather broad, with 3 or 4 long apical setae. Type series. — Holotype male (American Museum of Natural His- tory), Sensabaugh Saltpeter Cave, about 8 km W Kingsport, Hawkins Co., Tennessee (Church Hill IVi' Quadrangle, 36^33'57" X 82''38'05"), May 1968, Sam Taylor and Dick Powers, leg. One paratype female, same cave, 16 June 1963, L. G. Conrad. Six paratypes, same cave, 17 July 1979, T C. Barr, Jr., T. C. Barr, III, and J. R. Holsinger. Measurements (mm). — Holotype, total length 4.92, head 0.94 long X O. 88 wide, pronotum 0.85 long X 1.03 wide, elytra 2.73 long X 1.58 wide, antenna 3.21, aedeagus 0.95. Discussion. — The form of the aedeagus is strikingly similar to that of P. vicarius. At present P. paradoxus is the only known Tennessee Pseudanophthalmus species in the Holston Valley and the only known species south of Clinch Mountain. The type locality cave opens near the top of a wooded ridge in a subdivision of Kingsport; from the mouth a steep, muddy slope drops about 40 m to a small stream. The specimens collected in July 1979 were taken from under rocks and rotten wood near a ceiling drip, and no specimens were encountered along the stream. The microhabitat is thus similar to that of P. longiceps (jonesi group), another species of rather large size which inhabits caves high on a ridge. Sensa- baugh Saltpeter Cave is also known as Moffitt Saltpeter Cave or simply “Saltpeter Cave;” another cave in the vicinity, described as Click Creek Cave by Barr (1961), is sometimes locally called “Sensabaugh Cave” because it is near the Sensabaugh (railroad)Tunnel. jonesi group (new group) Aedeagus with apex produced, long and slender, not arrow-shaped in dorsal view, in lateral view tapering to a point (P. cordicollis), with a terminal button {P. seclusus), or simply bluntly rounded. Most species large to medium-large (4. 2-6. 2 mm), although one species {P. pallidus) is 3. 9-4. 3 mm. Form slender to very slender, convex, appendages elongate; pronotum elongate-cordiform with sides shallowly sinuate before small hind angles, disc with 1-4 long setae each side. Type species: P. jonesi Valentine. Distribution. — Lee, Scott, and Wise cos., VA; Harlan Co., KY; Campbell, Claiborne, Cumberland, and Hancock cos., TN. 72 Thomas C. Barr, Jr. Figs. 32-39: Aedeagi of Pseudanophthalmus spp., hubrichti, hypolithos, and alabamae groups. 32) P. sanctipauli, n.sp. 33) P. paradoxus, n.sp. 34) P. hypo- lithos, n.sp. 35) P. scholasticus, n.sp. 36) P. frigidus, n.sp. 37) P. calcareus, n.sp. 38) P. praetermissus, n.sp. 39) P. georgiae, n.sp. Pseudanophthalmus jonesi Valentine Fig. 48 Valentine 1945:645. Barr 1965:62. Length 4. 6-4. 8 mm. Form elongate, slender, moderately convex. Head 1.2 times longer than wide; labrum with distinct median lobe. Pronotum 0.9 as long as wide, cordiform, greatest width in apical fourth just behind anterior marginal setae; base width and apex width subequal, about 0.7 greatest width; sides subparallel in basal fifth, hind angles large, approx- imately right, sharp, secondary basal angles prominent; disc with 2 long setae each side. Elytra 1.7 times longer than wide, widest behind middle, rather convex, prehumeral borders oblique to midline but humeri angu- lar and very prominent; microsculpture transverse-obsolescent on disc; short scutellar stria present; inner two longitudinal striae finely impressed, intervals subconvex, outer striae gradually obsolescent; apical groove elongate, not clearly bisinuate, subparallel, running to 3rd stria or not. Aedeagus 0.73-0.76 mm long, basal bulb large and set off from weakly arcuate median lobe, apex gradually attenuate, finely and bluntly rounded at tip; parameres with 4-5 apical setae. Appalachian Pseudanophthdlmus 73 Pseudanophtahlmus jonesi is known from three caves in Grassy Cove, a large polje along the plunging axis of the Sequatchie anticline in Cum- berland County, Tennessee: Grassy Cove (=Brady) Saltpeter Cave, the type locality; nearby Mill Cave; and Blowhole (=“The Gouffre”). Grassy Cove is a karst island in the interior of the Cumberland Plateau; caves in the Appalachian valley to the east are occupied by species of the engel- hardti and hirsutus groups. In common with P. scutilus and P. rogersae, which occur in another karst island. Pine Mountain, P. jonesi has the elytra widest behind the middle and the apical groove is not distinctly bisinuate. This is the only species of the group previously described. Valentine (1945) quite rightly noted its unusually attenuate form and speculated that it was a representative of a more widely distributed stock. Pseudanophthalmus scutilus, new species Figs. 40, 49 Etymology . — Latin scutilus, “very lean.” Diagnosis. — Resembling P. jonesi in the indistinctly bisinuate apical groove and slightly ventricose elytra, but head proportionately larger, pronotum more cordiform, and elytral striae obsolete. Description. — Length 4.3-5. 4, mean 4.8 ± SD 0.3 mm (N = 1 1). Form elongate and slender, convex, with elongate appendages; color pale rufo- testaceous, shining but a little dull; elytral microsculpture finely trans- verse, a little confused. Head rather large, a fourth longer than wide, maxillary palps with penultimate segment about 0.7 as long as fourth segment, labrum with obsolete median lobe. Pronotum slender and cor- diform, only slightly wider than long (L/W 0.94-0.96), greatest width in apical fifth at level of anterior marginal setae, apex width slightly greater than base width and about 0.8 greatest width; disc convex, with 1 or 2 long setae each side in addition to sparse micropubescence; anterior angles subdued, sides shallowly sinuate in basal seventh, hind angles sharp and a little more than right, secondary basal angles only suggested. Elytra elongate-oval, 1.8 times longer than wide, convex, deplanate around scutellum, prehumeral borders a little oblique to midline, humeri somewhat rounded, finely setose but not serrulate; discal pubescence short, fine, rather sparse; longitudinal striae obsolete, inner 3 striae barely visible, intervals flat, outer striae absent, no scutellar stria; apical groove rather short and deep, hardly bisinuate but with slight suggestion of anterior flexure, ending blindly at level of anterior apical puncture. Aedeagus 0.71-0.73 mm long, basal bulb not greatly enlarged, median lobe slender, moderately and evenly arcuate, apex finely attenuate but without knob, truncation, cusp, or other special modification; parameres short, with 4 terminal setae. 74 Thomas C. Barr, Jr Figs. 40-43: Pseudanophthalmus spp., 70^7^5/ group. 40) P. scutilus, n.sp. 41) P. rogersae, n.sp. 42) P. seclusus, n.sp. 43) P. pallidus, n.sp. Appalachian Pseudanophthalmus 75 Type series. — Holotype male (American Museum of Natural History) and one female paratype, New Mammoth Cave, 1.5 km E Elk Valley on the north side of Pine Mountain, Campbell Co., Tennessee (Ivydell IV2 Quadrangle, 36°29'03'' x 84°13'46"), 8 August 1979, T. C. Barr, Jr. Two paratypes, same cave, 16 September 1979, T. C. Barr, Jr. Seven para- types, same cave, 21 November 1979, T. C. Barr, Jr., and J. R. Holsinger. Measurements (mm). — Holotype, total length 4.80, head 0.98 long X 0.78 wide, pronotum 0.84 long X 0.88 wide, elytra 2.67 long X 1.49 wide, antenna 3.31, aedeagus 0.71. Discussion. — This species, isolated within the Allegheny Plateau but near its eastern front like P. jonesi, appears more closely related to P. jonesi than to other species of the group. Grassy Cove and Elk Valley are approximately 90 km apart. New Mammoth Cave has a large, rather dry upper level several hundred meters long; small crevices through break- down lead down to a lower stream passage at various points. The cave is more extensive than indicated by the map reproduced in Barr (1961:92). The beetles are very hygrophilous, all of them taken from very wet, soggy, rotten wood at the edge of the stream. One small log, its end actually immersed in the stream, yielded a beetle on each of three collecting visits to the cave; a large colony of collembolans, Folsomia Candida, was found just inside the outer layers of the log and was a probable food source for the beetles. Pseudanophthalmus rogersae, new species Fig. 41 Etymology . — Patronymic honoring Mrs. Mary Rogers, Pine Moun- tain Settlement School. Diagnosis. — Resembles P. scutilus in the singly emarginate labrum, indistinctly bisinuate apical groove, and slightly ventricose elytra, differ- ing in cordiform pronotum with abrupt lateral sinuation and large hind angles, regular and shallowly impressed elytral striae, and subconvex intervals. Description. — Length of unique holotype 4.65 mm. Form elongate, slender, convex, appendages elongate; rufotestaceous, dull shining, pubescent; elytral microsculpture confused, isodiametric in striae and weakly trans- verse on intervals. Head a fourth longer than wide, labrum singly emar- ginate, maxillary palp with fourth segment about 0.2 longer than penul- timate segment; antenna 0.7 body length. Pronojum cordiform, 0.9 as long as wide, widest in apical third behind anterior marginal setae, widths 76 Thomas C. Barr, Jr. at apex and base subequal and 0.75 maximum width; sides rather abruptly sinuate at basal 0.4, then convergent to large, obtuse hind angles, base oblique behind angles, secondary angles small and inconspicuous; disc rather convex, with 3 long setae each side and scattered pubescence. Elytra 1.75 times longer than wide, widest behind the middle, prehumeral borders oblique to midline, humeri somewhat rounded, antapical sinua- tion shallow; disc convex, deplanate near scutellum, rather densely pubescent; striae regular and shallow but deeper than usual for jonesi group, intervals subconvex, apical groove elongate and subparallel to suture, running into 3rd stria. Male unknown. Type series. — Unique holotype female (American Museum of Natural History), Sawmill Hollow Cave, 2.0 km NNW Nolansburg and 600 m ESE Pine Mountain Settlement School on the northwest slope of Pine Mountain, elevation 700 m (Nolansburg IVi Quadrangle, 36°56'50" x 83^10'31"), 23 August 1979, T. C. Barr, Jr., and T. C. Barr, HI. Measurements (mm). — Holotype, total length 4.65, head 1.01 long X 0.80 wide, pronotum 0.83 long X 0.92 wide, elytra 2.63 long X 1.50 wide, antenna 3.21. Discussion. — Even though no male is available, the species is readily differentiated on external characters alone. It is evidently related to P. jonesi and P. scutilus, differing in the pronotal and elytral strial charac- ters cited in the diagnosis. Sawmill Hollow Cave is a small cave (length about 300 to 400 m) consisting of muddy crawlways on two or three levels, with a small stream and a pool in the lowest level; the cave is in Newman limestone dipping southeast at about 32^. The unique holotype was collected at the lower end of a section of the cave called “The Emper- or’s Palace”, a small stream channel intersected by a 10-m dome. Pseudanophthalmus seclusus, new species Figs. 42, 50 Etymology. — Latin seclusus, “remote, separated.” Diagnosis. — Distinguished from other species of the group by the following combination of characters: labrum with prominent median lobe, elytra subconvex with subparallel sides and broadly rounded apexes, head nearly 1.3 times longer than wide, elytral striae shallowly impressed and inner intervals subconvex, aedeagus with apex briefly reflexed and obliquely truncate. Description. — Length 4.2-5. 0, mean 4.6 ± SD 0.2 mm (N = 40). Pale rufotestaceous, form rather elongate and subconvex, pubescent. Head 1.25-1.28 times longer than wide, labrum with prominent median lobe in anterior margin (doubly emarginate), antenna 0.6 times body length. Pronotum 0.85 times as long as wide, greatest width in apical fourth Appalachian Pseudanophthalmus 11 behind level of anterior marginal setae, widths at apex and base subequal and 0.75 greatest width; sides feebly sinuate in basal fifth, anterior angles subdued, hind angles large, acute, slightly produced and elevated, second- ary angles of base prominent but broadly rounded; disc with 1-3 (usually 2) long setae each side in addition to sparse, rather long pubescence. Elytra 1.7 times longer than wide, subparallel but rather broad, humeri prominent but not sharp, prehumeral borders oblique to midline, apexes broadly rounded; scutellar stria long, attaining suture at level of 4th umbilicate puncture, striae 1-3 shallowly impressed, intervals subconvex with 2-3 rows of pubescence, outer striae irregular and progressively obsolescent, apical recurrent groove elongate and bisinuate, joining 3rd stria via crosier in advance of anterior apical puncture. Aedeagus 0.67- 0.74, mean 0.69 ± 0.03 mm long (N = 8). moderately arcuate, basal bulb large but not sharply set off from median lobe; apex of median lobe slightly swollen, then briefly reflexed and obliquely truncate, buttonlike, at tip; right copulatory piece canoe-shaped, partially enfolding elongate- triangular, minutely spinulose, shorter left piece; internal sac weakly armed with minute spines; parameres rather stout, with only 3 apical setae. Type series. — Holotype male (American Museum of Natural History) and 45 paratypes, Flannery Cave, Scott Co., Virginia (Clinchport IVi Quadrangle, 36°42'28" x 82°43'30"), 1 August 1964, S. B. Peck and J. R. Holsinger. Measurements, (mm) — Holotype, total length 4.64, head 1.02 long X 0.76 wide, pronotum 0.86 long X 0.92 wide, elytra 2.55 long X 1.45 wide, antenna 2.96, aedeagus 0.71. Discussion. — Pseudanophthalmus seclusus is known from a series of caves in the Rye Cove karst near Clinchport, Scott County, Virginia, including Flannery (type locality), McDavid (traversed by the Flannery Cave stream). Alley No. 2, Cox Ram Pump, Pond, Hill, and Kerns No. 1 caves. Most of the caves are developed along the eastern flank of the Rye Cove syncline and were described by Holsinger (1975). A total of 54 specimens was seen. The species is abundant on mud banks along the stream in Flannery Cave. Four specimens were taken in McDavid Cave and only one or two each in the other caves cited above. The form of the aedeagus, with reflexed and obliquely truncate, buttonlike apex, is highly diagnostic. Superficially P. seclusus is closest to P.pallidus, which occurs along the Powell River in Claiborne County, Tennessee, but it is larger, has a longer pronotum, and the elytral apexes are quite broadly rounded rather than attenuate as in P. pallidus. 78 Thomas C. Barr, Jr. Pseudanophthalmus pallidus, new species Figs. 43, 51 Etymology. — Lsitin pallidus , “pale.” Diagnosis. — A moderately depressed and slender species of small size (3. 9-4.3 mm), the elytral striae obsolescent and the aedeagal apex simple. Description. — Length 3. 9-4.3, mean 4.2 ± SD 0.2 mm (N = 8). Rather pale testaceous, moderately slender and subconvex, dull shining, moder- ately pubescent. Head 0.2 longer than wide; labrum with prominent median lobe in anterior margin; antenna two-thirds body length. Prono- tum 0.87 as long as wide, cordiform, apex 0.1 wider than base and 0.75 maximum width; anterior angles prominent, sides evanescently sinuate and almost convergent from apical third to obtuse, sharp, slightly ele- vated hind angles, base oblique behind angles, small secondary angles broadly rounded; disc subconvex, 3 long setae each side. Elytra 0.75 longer than wide, elongate-oval, prehumeral borders oblique to midline, humeri rather prominent; 4 inner striae and trace of 5th vaguely discernible, intervals flat, apical groove elongate and bisihuate, joining 3rd stria via crosier. Aedeagus 0.52-0.53 mm long, smaller than in any other known species of the jonesi group; apex simply and briefly attenuate. Type series. — Holotype male (American Museum of Natural His- tory), 1 male and 3 female paratypes, Chadwell Cave, 6 km NE Tazewell, 425 m S of Cedar Fork Road, and 1000 m N of Henderson Knob, in a wooded sink at elevation 400 m (Tazewell IV2 ' Quadrangle, 36‘^29'46" x 83^3 r23"), 18 July 1979, T. C. Barr, Jr., T. C. Barr, III, and J. R. Holsinger; one paratype male, same cave, 1 March 1964, S. B. Peck. Measurements (mm). — Holotype, total length 4.34, head 0.89 long X 0.73 wide, pronotum 0.77 long X 0.89 wide, elytra 2.36 long X 1.35 wide, antenna 2.88, aedeagus 0.53. Discussion. — Chadwell Cave, developed in the Harrison Bridge and Martin limestones (Ordovician), has approximately 600 m of readily tra- versable passages. Most of the specimens of this species occurred on wet silt under rocks near the stream. In addition to Chadwell Cave, P. palli- dus has been taken in nearby Buis Saltpeter Cave (6.7 km northeast of Tazewell, two females) and English Cave (Barr 1961:115), both in Clai- borne County. The single female from English Cave, taken in July 1965 by J. Holsinger and C. Rippy, establishes the first known case of sympa- try of three species of Pseudanophthalmus in the Appalachian valley. Much more abundant in English Cave are P. engelhardti and P. rotunda- tus, the former known only from English Cave and the latter extending up the Powell River valley to small caves near Rose Hill, Lee County, Virginia. A troglobitic milliped, Pseudotremia nodosa Loomis, and a troglobitic pseudoscorpion, Kleptochthonius affinis Muchmore, are Appalachian Pseudanophthalmus 79 found in English Cave as well as in the caves of the Cedar Fork area. There is no obvious structural geologic barrier separating these caves. Pseudanophthalmus longiceps, new species Figs. 44, 52 Etymology. — Latin longiceps, “long head.” Diagnosis. — Differs from other species of the group in the slender form and unusually elongate head (1.5 times longer than wide), the small, acute, pronotal hind angles, and the shallow elytral striae with subconvex intervals. Description. — Length 4. 6-6. 2, mean 5.3 ± SD 0.3 mm (N = 17). Form elongate, head very narrow, pronotum small and rather convex, elytra convex; rufotestaceous, sparsely pubescent, elytral microsculpture finely transverse, obsolescent. Head nearly 1.5 times longer than wide, sides subparallel, dorsum depressed; anterior margin of labrum with low but sharp median lobe; frontal grooves feeble behind, barely extended onto sides of head, subaphaenopsian. Pronotum cordiform, 0.9 as long as wide, convex, disc with 4 long, somewhat irregular setae each side; widths at apex and base subequal and 0.75 greatest width, which occurs in apical sixth at placement of anterior marginal setae; anterior angles prominent, a little produced, sides feebly and very shallowly sinuate in basal sixth, hind angles small, sharp, a little less than right; basal impressions deep and oblique, secondary basal angles subdued. Elytra broadly convex, slightly ventricose, prehumeral borders oblique to midline, humeri prom- inent but slightly rounded, setose and feebly serrulate; disc deplanate in scutellar region, with longitudinal rows of rather sparse pubescence; inner 4 striae feebly impressed, outer striae obsolescent, intervals subcon- vex; apical recurrent groove elongate and bisinuate, joining 3rd stria at level of 7th umbilicate puncture. Appendages all unusually slender and elongate; antenna 0.8 times body length; last segment of maxillary palp 1.15 times longer than penultimate segment, Aedeagus 0.60-0.63 mm long, gently arcuate in left lateral view, apex produced and very slightly knobbed, transfer apparatus as in P. seclusus. Type series. — Holotype male (American Museum of Natural His- tory), Fisher Cave, near the top of Newmans Ridge, between Blackwater and Kyles Ford, Lee Co., Virginia (Kyles Ford IVi Quadrangle, 36^37'22" X 83^03'40", elev. 480 m), 15 August 1969, J. R. Holsinger; 5 paratypes, same cave, 17 July 1979, T. C. Barr, Jr., T. C. Barr, III, and J. R. Holsinger; 13 paratypes, same cave, 30 September 1979, T. C. Barr, Jr. Measurements (mm). — Holotype, total length 4.78, head 1.13 long X 0.76 wide, pronotum 0.86 long X 0.95 wide, elytra 2.66 long X 1.59 wide, antenna 3.76, aedeagus 0.62. 80 Thomas C. Barr, Jr. Discussion. — This remarkable species is unusually slender and elon- gate, almost aphaenopsian in form. The frontal grooves are nearly obso- lete at the sides of the head, the pronotum is small and convex, and the elytra are convex with marked deplanation around the scutellum. At 6.2 mm, two members of the type series are the largest cave carabids known from the Appalachian valley. The species was found under rocks and rotting wood on crumbly soil, walking across a wet, sloping rock surface, and under large flat stones near the bottom of the long talus slope extend- ing down from the entrance, where it coexists with P. deceptivus (engel- hardti group). In addition to the type locality cave (description in Holsin- ger 1975:133), P. longiceps is also known from Panther Creek Cave, Hancock County, Tennessee (description in Barr 1961:255), where J. R. Holsinger collected six specimens on 20 November 1979. Both caves are in Newman limestone high on Newmans Ridge, and both are damp with no permanent stream. Pseudanophthalmus thomasi, new species Figs. 45, 53 Etymology . — Patronymic honoring Thomas C. Barr, III, who col- lected the only two males known of this rare and unusual species. Diagnosis. — Resembling P. longiceps in trilobed labrum margin, gen- eral form, and aedeagal structure, differing in less slender head, larger hind angles, and obsolescent elytral striae. Description. — Length 4. 7-5. 2 mm (N = 4). Form elongate, slender, subconvex, appendages elongate; color pale rufotestaceous, dull shining, microsculpture of elytral disc very fine, transverse, obsolescent. Head 1.3 times longer than wide and nearly as wide as pronotum; labrum with distinct, broad, median lobe in anterior margin; last segment of maxillary palp 1.2 times longer than penultimate segment. Pronotum cordiform, convex, as long as wide, greatest width in apical third, width at apex a little more than base width and about 0.8 maximum width; sides deeply sinuate in basal sixth, then subparallel or slightly divergent to prominent, sharp, reflexed, slightly acute to right hind angles; base oblique or emar- ginate behind angles, secondary angles of base conspicuous but rather broadly rounded; anterior marginal setae at apical ninth, well forward of maximum width, posterior marginals placed well forward of hind angles; disc glabrous except for 4 long setae each side. Elytra subconvex, elongate-oval, 1.8 times longer than wide; prehumeral borders oblique to midline, humeri somewhat rounded, antapical sinuation very shallow; apical recurrent groove elongate, deep, bisinuate, running into apex of 5th stria well in advance of anterior apical puncture; disc with rather Appalachian Pseudanophthalmus 81 Figs. 44-47: Pseudanophthalmus sp^.Jonesi and alabamae groups. 44) P. longi- ceps, n.sp. 45) P. thomasi, n.sp. 46) P. cordicollis, n.sp. Al)P. georgiae, n.sp. 82 Thomas C. Barr, Jr. short and moderately dense pubescence. Aedeagus of holotype 0.74 mm long, basal bulb not sharply set off from evenly arcuate median lobe, which is gradually attenuate and rounded at apex. Type series. — Holotype male (American Museum of Natural History) and one male paratype, Blair-Collins Cave, Scott Co., Virginia (Gate City 71/2' Quadrangle, 36''40'27" x 82°33"27", elevation 490 m), 17 July 1979, T. C. Barr, HI; one female paratype, same cave, 6 November 1966, J. R. Holsinger; one female paratype from nearby Coley Cave No. 2, Scott County, Virginia, 3 June 1967, J. R. Holsinger. Measurements (mm). — Holotype, total length 4.97, head 1.04 long X 0.80 wide, pronotum 0.86 long X 0.86 wide, elytra 2.75 long X 1.53 wide, antenna 3.34, aedeagus 0.74. Discussion. — This species is known only from the Blair-Collins Cave system and from Coley Cave No. 2, which are developed in the Maryville limestone (Ordovician) along Moccasin Creek near Gate City, Scott County, Virginia (Holsinger 1975:267, 275). All four specimens in the type series were taken on mud banks in stream passages. The two males were collected in the lower stream passage of Blair-Collins Cave by Thomas C. Barr, III, at age 12 already an accomplished and indefatigable cave trechine collector, who reported that one of the beetles was observed seizing a predatory mite just before it was caught. Pseudanophthalmus cordicollis, new species Figs. 46, 54 Etymology . — Latin cor, “heart”, + collum, “neck.” Diagnosis. — A large, elongate species with slender appendages, elon- gate head, small pronotum with convergent sides and obtuse hind angles, obsolescent elytral striae, and aedeagal apex simply attenuate. Description. — Length 4. 7-5.0, mean 4.8 mm (N = 4). Form slender and elongate, subconvex, appendages elongate and slender; pale rufotes- taceous, pubescent; elytral microsculpture finely transverse, shining. Head 1. 1-1.3 times longer than wide, mandibles unusually large, long and slender; labrum with prominent median lobe; maxillary palps with last segment about 0.4 longer than penultimate segment; antenna 0.6 body length. Pronotum cordiform, 0.9 as long as wide, widest in apical fifth; apex subtruncate, anterior angles rounded, sides convergent without sinuation, hind angles obtuse, secondary angles of base very small and rounded; disc with 3 long setae each side, otherwise subglabrous. Elytra subparallel, subconvex, 1.6 times longer than wide, widest at middle; prehumeral borders slightly oblique, humeri rather prominent but not angular, apexes attenuate; longitudinal striae feeble and irregular, inter- vals very flat (feebly subconvex in 1 of 4 specimens), with 2 or 3 rows of long pubescence per interval; apical groove elongate, subparallel, faintly Appalachian Pseudanophthalmus 83 to strongly bisinuate, running directly into 3rd stria, no crosier; all fixed setae hypertrophied, unusually long. Aedeagus of holotype 0.61 mm long, moderately arcuate, apex finely and simply attenuate. Type series. — Holotype male (American Museum of Natural History) and 3 female paratypes. Little Kennedy Cave, Wise Co., Virginia (Appa- lachia IVi' Quadrangle, 36°53'24' x 82''45'01'), 26 November 1970, J. R. Holsinger, Roger Baroody, and R. M. Norton. Measurements (mm). — Holotype, total length 4.88, head 0.88 long X 0.78 wide, pronotum 0.84 long X 0.90 wide, elytra 2.49 long X 1.57 wide, antenna 3.00, aedeagus 0.61. Discussion. — Together with P. longiceps and P. thomasi this species forms a trio of rather large, unusually attenuate species with long append- ages. Little Kennedy Cave, near the edge of the Allegheny front in Newman limestone (Mississippian), was described by Holsinger (1975: 402). The cave is about 40 km north of the type locality of P. thomasi (to which P. cordicollis is most closely similar) and 65 km northeast of the type locality of P. longiceps. hypolithos group (new group) Aedeagus with apex long and slender, not arrow-shaped in dorsal view, bluntly rounded or slightly falciform at tip. Size small (3. 7-4. 3 mm, most species means about 4 mm), form moderately slender and depressed; pronotum transverse-cordiform, sides shallowly sinuate before hind angles, disc without long setae. Type species: P. hypolithos, new species. Distribution. — Pike, Harlan, Bell, and Whitley cos., KY (Pine Moun- tain); Scott Co., VA. Discussion. — Four of the small species of this group occupy different caves in Pine Mountain, Kentucky, a large fault block in which the Newman (Mississippian) limestone is exposed on the northwest face. The fifth species {P. praetermissus) occurs in a cave in Hunter Valley, Scott County, Virginia, near the base of the Allegheny front but in the edge of the Appalachian valley. Pseudanophthalmus hypolithos, new species Figs. 28, 34 Etymology. — Greek hypo-, “under,” + lithos, “rock.” Diagnosis. — Distinguished from other members of the group by the rather deep and complete elytral striation, convex intervals, reduced pubescence, and falciform aedeagal apex. Description. — Length 3.9-4. 3, mean 4.0 mm (N = 4). Form elongate, rather slender, depressed; rufotestaceous, shining, head and pronotum subglabrous, elytra with short and sparse pubescence, limited to 1 or 2 84 Thomas C. Barr, Jr. rows per interval; elytral microsculpture transverse-obsolescent, forming meshes. Head rounded, as long as wide; labrum singly emarginate; last segment of maxillary palp about 0.2 longer than penultimate segment; antenna rather short, 0.6 body length. Pronotum transverse-cordiform, 0.8 as long as wide, width at base and apex subequal and 0.75 greatest width, which occurs in apical fourth behind anterior marginal setae; anterior angles prominent, sides shallowly sinuate, subparallel or conver- gent in basal fifth, hind angles large and sharp, right or slightly obtuse; secondary angles of base prominent but rounded. Elytra elongate-oval, 1.8 times longer than wide, subconvex; prehumeral borders a little oblique, humeri prominent but slightly rounded, apexes bluntly rounded; short scutellar stria present, longitudinal striae deep, inner 3-4 complete with clear traces of 5,6,7, and 8, intervals convex, apical groove elongate, bisinuate, running to 3rd stria without crosier at level of 7th umbilicate puncture. Aedeagus of paratype 0.55 mm long, moderately arcuate, apex slender and produced, slightly falciform; parameres with 4-5 apical setae. Type series. — Holotype male (American Museum of Natural His- tory), 1 male and 2 female paratypes. Old Quarry Cave, 1.8 km SSE Ashcamp, Pike Co., Kentucky (Hellier IVi Quadrangle, 37^15'09" X 82°25'29''), 2 August 1979, T. C. Barr, Measurements (mm). — Holotype, total length 3.92, head 0.72 long X 0.70 wide, pronotum 0.69 long X 0.84 wide, elytra 2.23 long X 1.22 wide, antenna 2.42. Discussion. — This species, near the northeast end of Pine Mountain, is immediately differentiated by the unusually deep elytral striation and reduced pubescence. Old Quarry Cave is situated 100 m east of an old, abandoned limestone quarry at an elevation of 620 m. The beetles were taken from under rocks at the back of the entrance room and in the lower of two subparallel crawlways, all within 20-60 m of the entrance. Abun- dant cave rat debris was present. Pseudanophthalmus hypolithos and the next three species are re- stricted to caves in Pine Mountain, a huge fault block 125 km long, extending from Elk Gap, Campbell County, Tennessee, to Breaks Inter- state Park near Elkhorn City, Kentucky. The Newman limestone (Missis- sippian) crops out on the northwest face of Pine Mountain, dipping to the southeast at about 25-35°. Caves of Pine Mountain are of the Appala- chian pattern, with major passages along the strike and passages connect- ing different levels along the dip. Pseudanophthalmus scholasticus, new species Figs. 29, 35 Etymology. — Latin scholasticus, “scholastic.” Appalachian Pseudanophthalmus 85 Diagnosis. — Differs from other species of the group in the attenuate elytral apexes, deflexed aedeagal apex, and somewhat more robust form. Description. — Length 3.9 mm (in 2 specimens of the type series). Form as in P. hypolithos but more robust; pubescence moderate on genae and pronotum disc, dense on elytra, longer than in P. hypolithos, 2-3 rows per interval. Head rounded; labrum singly emarginate; last seg- ment of maxillary palp 0.3 longer than penultimate segment; antenna 0.6 body length. Pronotum 0.8 as long as wide, apex and base widths sub- equal and about 0.7 greatest width, which occurs in apical fourth; sides conspicuously sinuate and subparallel in basal fifth, anterior angles prominent, hind angles large, sharp, slightly obtuse, secondary angles of base low, broad, rounded. Elytra elongate-oval, prehumeral borders a little oblique, humeri prominent though somewhat rounded, apexes attenuate; no separate scutellar stria, longitudinal striae finely impressed, intervals weakly subconvex, inner 4 striae present, outer striae obsolete; apical groove elongate and bisinuate, running to 3rd stria via short cro- sier. Aedeagus of holotype 0.55 mm long, apex produced, slender, and conspicuously deflexed, finely and bluntly rounded at tip. Type series. — Holotype male (American Museum of Natural History) and one female paratype. Sawmill Hollow Cave, 2 km NNW Nolansburg and 600 m ESE Pine Mountain Settlement School on the northwest side of Pine Mountain, Harlan Co., Kentucky (Nolansburg IVi Quadrangle, 36°56'50" X 83°I0'31"), 23 August 1979, T. C. Barr, Jr., and T. C. Barr, III. Measurements (mm). — Holotype, total length 3.95, head 0.69 long X 0.70 wide, pronotum 0.73 long X 0.87 wide, elytra 2.33 long X 1.29 wide, antenna 2.25, aedeagus 0.55. Discussion. — This species coexists with P. rogersae (jonesi group) in Sawmill Hollow Cave, the only known case of sympatry in Pine Moun- tain caves. While P. scholasticus occurs in the upper level near the entrance, P. rogersae was collected in the lower level beside a small stream. A similar microhabitat segregation was observed in the same cave between a troglophilic (upper level) and a troglobitic species of Pseudo- tremia millipeds. The elytral apexes in P. scholasticus are more attenuate than in the other species of the group, and the deflexed aedeagal apex is also diagnostic. Pseudanophthalmus calcar eus, new species Figs. 31, 37 Etymology. — Latin calcareus, “pertaining to limestone.” Diagnosis. — Distinguished from other group species by the elongate head, narrower pronotum, and wider elytra with rounded apexes and 86 Thomas C. Barr, Jr. only 2-3 clearly impressed inner striae; apical recurrent groove ending blindly, not connected to 3rd stria. Description. — Length 4.0-4. 1 mm (N = 3). Form about as in hypolithos but head and pronotum narrower and elytra wider, pubes- cence rather dense. Head 1.15-1.25 longer than wide; labrum with sug- gested very low and wide median lobe; last segment of maxillary palp 0.3 longer than penultimate segment; antenna 0.6 body length. Pronotum 0.85-0.87 as long as wide, anterior angles somewhat rounded, hind angles acute (2 of 3 specimens) or obtuse (1 specimen), large and sharp, basal angles as in P. hypolithos', apex width slightly more than base width and 0.75 greatest width, which occurs in apical fifth at level of anterior margi- nal setae; disc irregularly with 1 long seta each side. Elytra short, 1.6 times longer than wide, elongate-oval, prehumeral borders oblique, humeri slightly rounded, apexes much rounded; inner 2 striae more deeply impressed than 3rd stria, but all very fine, intervals essentially flat, outer striae obsolete; apical groove elongate, bisinuate, ending blindly at level of 7th umbilicate puncture. Aedeagus 0.63 mm long in paratype, median lobe less arcuate, apex produced, slightly sinuate in lateral view, narrowly and bluntly rounded at tip. Type series. — Holotype male (American Museum of Natural History) and two male paratypes. Limestone Cave, on the northwest slope of Pine Mountain, 2.5 km NNW of the common corner of Whitley Co., Ken- tucky, and Campbell and Claiborne cos., Tennessee, in Whitley Co., Kentucky (Jellico East IVi Quadrangle, 36‘^36'38" X 84°01'10", elev. 420 m), 8 August 1979, T. C. Barr, Jr. Measurements (mm). — Holotype, total length 3.95, head 0.81 long X 0.73 wide, pronotum 0.72 long X 0.84 wide, elytra 2.14 long X 1.32 wide, antenna 2.57. Discussion. — Pseudanophthalmus calcareus is readily distinguished by the slender head and pronotum and relatively short, broad elytra which are widest just behind the middle; the widest part of the pronotum is at the level of the anterior marginal setae. The three specimens of the type series were collected in the entrance room of Limestone Cave under rocks on damp silt in areas rich in organic debris (cave rat nest debris, rotting wood, etc.). This seems to be a pattern for species of this group — close to food sources near entrances, at least in summer. J. R. Hol- singer and I carefully searched the entire cave in November 1979, finding no beetles; apparently cold, dry air moving into the cave from the exte- rior rendered accessible microhabitats in the entrance room unsuitable. Pseudanophthalmus frigidus, new species Figs. 30, 36 Etymology . — Latin frigidus, “cold.” Appalachian Pseudanophthalmus 87 Diagnosis. — Resembles P. calcareus in rounded elytral apexes and flat intervals, differing in wider head and pronotum, greatest width of elytra near middle, and aedeagal apex not as slender nor as greatly produced. Description. — Length 3. 7-4.3 mm(N = 2). Head rounded, only slightly longer than wide; labrum singly emarginate; last two segments of maxillary palp subequal; antenna two-thirds body length. Pronotum about as in P. scholasticus, 0.8 as long as wide, widths at apex and base subequal and about 0.75 greatest width, which occurs in apical fourth behind anterior marginal setae; sides subparallel in basal fifth, anterior angles prominent, hind angles sharp and about right, basal angles small and rounded. Elytra 1.65 times longer than wide, apexes rounded, prehumeral borders a little oblique, humeri slightly rounded, greatest width near middle; scutellar stria present; striae 1-3 finely and regularly impressed, intervals flat, outer striae obsolescent, apical groove elongate, bisinuate, oblique, running to 3rd stria via anterior crosier at level of 7th umbilicate puncture. Aedeagus 0.64 in holotype, compared with P. calcareus basal bulb more sharply deflexed and apex less slender and less produced; parameres shorter, with 3 apical setae. Type series. — Holotype male (American Museum of Natural History), Icebox Cave, 25 m above L & N railroad tracks on north side of Cumberland River in the town of Pineville, 1.0 km SE of the courthouse. Bell Co., Kentucky, 26 October 1963, S. B. Peck. One female paratype, same cave, 22 August 1979, T. C. Barr, Jr. and T. C. Barr, HI. Measurements (mm). — Holotype, total length 4.34, head 0.86 long X 0.80 wide, pronotum 0.80 long X 0.98 wide, elytra 2.48 long X 1.50 wide, antenna 2.91, aedeagus 0.64. Discussion. — The type locality is a small but very well known cave; a small entrance leads into an antechamber from which a low, rocky crawlway extends 8 m to a muddy strike gallery 75 m long. The holotype was found at the edge of a temporary pool (S. B. Peck, pers. comm.) and the paratype under a rock among wet stalactities. Pseudanophthalmus praetermissus, new species Fig. 38 Etymology. — Ediim praetermissus, “overlooked.” Diagnosis. — A robust species closest to P. scholasticus, differing in more rounded elytral apexes, presence of a short scutellar stria, no crosier connecting apical groove to 3rd stria; size slightly larger; aedeagus similar with deflexed apex, but a little larger. Description. — Length 4.0-4. 3, mean 4.1 mm (N = 4). Form moder- ately robust and depressed; rufotestaceous, dorsum of head and prono- 88 Thomas C. Barr, Jr. turn subglabrous, elytral disc densely pubescent; elytral microsculpture rather coarse, vaguely transverse, no meshes, dull-shining. Head as long as wide; labrum singly emarginate; last segment of maxillary palp elon- gate, 1.5 times as long as penultimate segment; antenna 0.6 body length. Pronotum transverse-cordiform, 0.8 as long as wide, apex and base sub- equal and 0.7 greatest width, which occurs in apical fifth; anterior angles prominent, sides subparallel to slightly divergent in basal sixth, hind angles slightly produced, acute, secondary basal angles effaced. Elytra elongate-oval, 1.6 times longer than wide, prehumeral borders slightly oblique, humeri somewhat rounded, disc depressed and heavily pubes- cent; short scutellar stria present, striae 1 and 2 very shallow though regular, intervals nearly flat, striae 3-5 shallower; apical groove elongate, a little oblique, bisinuate, connected medially to 3rd stria but without crosier. Aedeagus 0.61-0.63 mm, broadly arcuate, apex slender, pro- duced, deflexed, finely and bluntly rounded at tip. Type series. — Holotype male (American Museum of Natural History) and one paratype male, Kern’s Cave No. 1, Scott Co., Virginia (East Stone Gap IVi Quadrangle, 36°47'20" X 82°39''33"), 26 July 1969, J. R. Holsinger and James Beck. Two male paratypes, same cave, J. R. Hol- singer, 6 October 1979. Measurements (mm). — Holotype, total length 3.98, head 0.92 long X O. 92 wide, pronotum 0.94 long X 1. 14 wide, elytra 2. 17 long X 1.35 wide, antenna 2.54. Discussion. — This is the only species of the hypolithos group known outside Pine Mountain, Kentucky, but it inhabits a cave near the base of the Allegheny front. On each of two visits to Kern’s Cave No. 1, J. R. Holsinger collected three beetles: one P. seclusus (jonesi group) and two P. praetermissus. Externally the two species are closely similar and in the same general size range (4.0-4. 3 mm for P. praetermissus and 4. 2-5.0 mm for P. seclusus). The aedeagi are quite different and highly diagnostic; in addition, the labrum in P. seclusus bears a low but distinct median lobe, there are usually two or three long setae on each side of the pronotal disc, and the elytra are slightly longer and more subparallel. alabamae group Aedeagus with apex slender, produced, distinctly and briefly deflexed at tip, apex not arrow-shaped in dorsal view. Medium-sized species, 4.0- 4.8 mm, mean lengths 4.4 and 4.5 mm, respectively for the two known species. Pronotum a little transverse and strongly cordiform; apical recur- rent groove highly diagnostic: subparallel to suture (not bisinuate), ante- rior end directed toward apexes of 4th and 5th striae, sometimes joining (no crosier), sometimes ending blindly. Type species: P. alabamae Valentine. Appalachian Pseudanophthalmus 89 Distribution. — Dekalb Co., AL; Chattooga and Walker cos., GA. Pseudanophthalmus alabamae Valentine Valentine 1932:273. Barr 1965:67. This distinctive species is known from eight caves in Little Wills Valley, Dekalb County, Alabama, all on the east side of the Wills anticline: Bartlett (AL 251, T9S/ R7E/ S13); Cherokee (AL 806, T6S/R9E/S26); Kelly Girls (AL 252, T9S/R7E/S13); Lykes (AL 239, T9S/R8E/S13); Manitou (type locality, - AL 13), in Fort Payne: Talley (AL 443, T5S/R10E/S32); Stanley-Carden (AL 730, T6S/ R9E/S13); and Section 26 (AL 804, T4S/ RlOE/ S26). I have examined 54 specimens including 16 topotypes; the length of this series is 4.0-4.8, mean 4.4 ± SD 0.4 mm. In P. alabamae the humeri are less sharply angular and the tip of the deflexed aedeagal apex is falciform rather than knobbed as in the only other known species of the group, described below. Aedeagi of four topotypes measured 0.57-0.63, mean 0.60 mm long. Some individuals have 1 or 2 long setae on each side of the pronotum disc. Figs. 48-54: Aedeagi of Pseudanophthalmus spp., jonesi group. 48) P. jonesi Valentine 49) P. scutilus, n.sp. 50) P. seclusus, n.sp. 51) P. pallidus, n.sp. 52) P. longiceps, n.sp. 53) P. thomasi, n.sp. 54) P. cordicollis, n.sp. 90 Thomas C. Barr, Jr. Pseudanophthalmus georgiae^ new species Figs. 39, 47 Etymology. — Geographic name. Diagnosis. — Resembles P. alabamae in the short, subparallel apical groove and slender, attenuate, deflexed aedeagal apex; differs in the more angular humeri and more slender, less arcuate aedeagus with finely knobbed apex. Description. — Length 4. 1-4.8, mean 4.5 ± SD 0.1 mm (N = 13). Form slender and moderately elongate, subconvex, pubescent, rufotestaceous; elytral microsculpture transverse, not forming meshes. Head a little longer than wide; labrum with broad, prominent, median lobe; mandibles unusually long and slender, terebral teeth large and very conspicuous; last segment of maxillary palp one-third longer than penultimate seg- ment; antenna two-thirds body length. Pronotum transverse, strongly cordiform, 0.85-0.89 as long as wide, convex, disc glabrous (very sparse micropubescence only), a few individuals with 1-3 discal setae; width at base slightly greater or equal to width at apex and about 0.8 greatest width, which occurs in apical fourth behind level of anterior marginal setae; anterior angles moderate, sides arcuate, very shallowly (or not) sinuate at basal fifth, hind angles sharp, usually a little less than right, secondary basal angles small but conspicuous. Elytra elongate-oval, 1.6 times longer than wide, prehumeral borders slightly oblique, humeri prominent and angular, apexes attenuate, disc moderately convex with slight deplanation around scutellum; striae fairly deep, intervals convex, inner 3 striae deeper but striae 4-7 still discernible; apical groove subpar- allel to suture, hardly bisinuate at all, joining or directed toward juncture of 4th and 5th striae at level of anterior apical puncture without anterior crosier. Aedeagus 0.60-0.65 mm long, less arcuate than that of P. alaba- mae, angle between basal bulb and median lobe greater, apex more slender, slightly knobbed or finely truncate and deflexed; 5-6 setae at apexes of paramers. Type series. — Holotype male (American Museum of Natural His- tory), 1 male and 8 female paratypes, Blowing Spring Cave, 4 km NE Cloudland and 1.6 km NW Chelsea at the east base of Lookout Moun- tain, Chattooga Co., Georgia (Dougherty Gap IVi Quadrangle), 21 June 1967, S. B. Peek and A. Fiske. Measurements (mm). — Holotype, total length 4.59, head 0.83 long X 0.77 wide, pronotum 0.80 long X 0.92 wide, elytra 2.45 long X 1.53 wide, antenna 3.00, aedeagus 0.63. Discussion. — This species obviously shares a relatively recent common ancestry with P. alabamae, which occurs in several caves on the east side of the Wills anticline in Little Wills Valley, Alabama. The subparallel Appalachian Pseudanophthalmus 91 apical groove, occasional 1-3 long setae on each side of the pronotal disc, and slender, nonconstricted aedeagal apex suggest a more remote rela- tionship to the Grassy Cove and Pine Mountain species of the jonesi group (P. jonesi, P. scutilus, P. rogersae). The known range of P. geor- giae includes three Georgia caves: Blowing Spring Cave, the type local- ity; Pettijohn Cave, 8 km southwest of Lafayette and 2.3 km northwest Bronco and a similar distance south-southeast of Atwood Point on Pigeon Mountain ( an east spur of Lookout Mountain) (Estelle IVi Quadrangle); and Mt. Cove Farm Cave, 2.5 km east of Lookout near the head of McLemore Cove, 0.9 km north-northwest of Dougherty Gap (Dougherty Gap IVi Quadrangle). The latter two caves are in Walker County. Other Appalachian Valley Species Groups Although northern Appalachian Pseudanophthalmus species will be treated in detail in a subsequent paper, availability of much fresh material has substantially altered my views on the relationships between the approximately 35 known species in the region. Five rather well defined species groups exist, and the distributions of four of them are in accord with an Allegheny refugium hypothesis. With respect to my earlier paper (Barr 1965) on Appalachian valley Pseudanophthalmus, the following classificatory changes are now indicated. Subspecies citations have been omitted, but some of the Greenbrier valley species of the grandis group are polytypic. grandis group grandis Valentine 1931:254 (WV) fuscus Valentine 1931:254 (WV) sylvaticus Barr 1967b: 167 (WV) montanus Barr 1965:52 (WV) krekeleri Barr 1965:52 (WV) hypertrichosis Valentine 1932:266 (WV) virginicus (Barr 1960:66 (NEW COMBINATION for Aphanotrechus virginicus Barr) Collection of a male of P. virginicus by J. R. Holsinger and R. M. Norton made possible the determination that virginicus is only an aber- rant species of iht grandis group near P. hypertrichosis. It is known only from Hugh Young Cave, in the Maiden Spring area, Tazewell County, Virginia. hubbardi group Same as hubbardi group of Valentine (1945) and Jeannel (1949); equiv- alent to hubbardi subgroup of Barr (1965). 92 Thomas C. Barr, Jr. pusio group Same as pusio group of Valentine (1932), Jeannel (1949), and Barr (1965), readily identified by the elongate, cylindrical median lobe of the aedeagus. I differ with Jeannel (1949) in including P. higginbothami Valentine (WV) in the pusio group and placing P. fuscus in the grandis group. gracilis group (new group) Small species in which the males have an apical emargination in the last abdominal sternite; aedeagus long and slender, copulatory pieces subequal and rather short. The group is closely related to the inexpecta- tus group of central Kentucky (Krekeler 1973). gracilis Valentine 1931:253 (VA) hadenoecus Barr 1965:53 (WV) petrunkevitchi group (new group) Small species with a distinct, circular eye rudiment (as opposed to irregular, lunate scars in P. vicarius and some species of the grandis group); aedeagal apex slender and produced, transfer apparatus of two elongate pieces fused into a cylinder at the base of the internal sac. petrunkevitchi Valentine 1945:652 (VA) hoffmani Barr 1965:58 (VA) hortulanus Barr 1965:60 (VA) Four of these species groups {grandis, hubbardi, pusio, gracilis) have species in karst islands in the eastern part of the Allegheny Plateau as well as in the Appalachian valley proper. These distributions are consonant with the hypothesis that they are relics of ancient lineages which were edaphobitic in a Pleistocene (or earlier?) refugium within the Allegheny Plateau. In the grandis group, P. sylvaticus remains an edaphobite living in the forest floor of mountains above the Greenbrier valley caves where P. fuscus, its closest known relative, occurs. The petrunkevitchi group includes species at the eastern edge of the Appalachian valley and no species are known from the Allegheny karst islands. Although this group could have dispersed eastward from an Allegheny source, a Blue Ridge refugium is equally likely, or perhaps more likely in view of distribution of its known component species. ACKNOWLEDGMENTS. — Many persons have contributed spec- imens, given advice, or assisted me in the field over the years, and this paper would not have been possible without their help. I single out for Appalachian Pseudanophthalmus 93 special thanks T. C. Barr, III, J. E. Cooper, J. R. Holsinger, the late W. B. Jones, C. H. Krekeler, J. A. Payne, S. B. Peck, R. M. Norton, W. W. Torode, J. M. Valentine, and R. L. Wallace. Some of the beetles de- scribed in this paper were collected with the assistance of grants from the National Science Foundation (G-18765 and GB-5521) and the Kentucky Nature Preserves Commission. LITERATURE CITED Barber, Henry S. 1928. Two new cave beetles related to Anophthalmus pusio Horn. J. Wash. Acad. Sci. 75:195-196. Barr, Thomas C., Jr. 1960. A new genus of cave beetle (CarabidaeiTrechini) from southwestern Virginia, with a key to the genera of Trechini of North America north of Mexico. Coleopt. Bull. 74:65-70. 1961. Caves of Tennessee. Tenn. Div. Geol. Bull. 64, Nashville. 567 pp. 1965. The Pseudanophthalmus of the Appalachian valley (Coleoptera: Carabidae). Am. Midi. Nat. 75:41-72. 1967a. Observations on the ecology of caves. Am. Nat. 707:475-492. 1967b. A new Pseudanophthalmus from an epigean environment in West Virginia (Coleoptera:Carabidae). Psyche 74:166-174. 1969. Evolution of the Carabidae (Coleoptera) in the southern Appala- chians. pp. 67-92 in P. C. Holt (ed.). The distributional history of the biota of the southern Appalachians, Part I: Invertebrates. Res. Div. Monogr. 1, Va. Polytech. Inst., Blacksburg. 295 pp. 1972. Trechoblemus in North America, with a key to North American genera of Trechinae (Coleoptera:Carabidae). Psyche 75:140-149. 1979a. Revision of Appalachian Trechus (Coleoptera:Carabidae). Brim- leyana. 2:29-75. 1979b. The taxonomy, distribution, and affinities of Neaphaenops, with notes on associated species of Pseudanophthalmus (Coleoptera:Carabidae). Am. Mus. Novit. No. 2861:1-20. Douglas, Henry H. 1964. Caves of Virginia. Va. Cave Survey, Falls Church. 761 pp. Holsinger, John R. 1975. Descriptions of Virginia Caves. Va. Div. Mineral Resour. Bull. 85. 450 pp. Horn, George H. 1868. Catalogue of Coleoptera from south-west Virginia. Trans. Am. Entomol. Soc. 2:123-128. Jeannel, Rene. 1949. Les coleopteres cavernicoles de la region des Appalaches. Etude systematique. Notes Biospeol. 4 (Publ. Mus. Nat. Hist. Nat., Paris, no. 12):37-104. Krekeler, Carl H. 1973. Cave beetles of the genus Pseudanophthalmus (Coleop- tera:Carabidae) from the Kentucky Bluegrass and vicinity. Fieldiana (Zool.) (52:35-83. Matthews, Larry E. 1971. Descriptions of Tennessee caves. Tenn. Div. Geol. Bull. 69. Nashville. 150 pp. 94 Thomas C. Barr, Jr. Valentine, J. Manson.1931. New cavernicole Carabidae of the subfamily Trechinae Jeannel. J. Elisha Mitchell Sci. Soc. 4(5:247-258. 1932. A classification of the genus Pseudanophthalmus Jeannel (fam. Carabidae) with descriptions of new species and notes on distribution. J. Elisha Mitchell Sci. Soc. 4<^:261-280. 1937. Anophthalmid beetles (fam. Carabidae) from Tennessee caves. J. Elisha Mitchell Sci. Soc. 53:93-100. 1945. Speciation and raciation in Pseudanophthalmus (cavernicolous Carabidae) Trans. Conn. Acad. Arts Sci. 3(5:631-672 1948. New anophthalmid beetles from the Appalachian region. Geol. Surv. Ala. Mus. Pap. 27:1-20. 1952. New genera of anophthalmid beetles from Cumberland caves (Cara- bidae, Trechinae). Geol. Surv. Ala. Mus. Pap. 34:1-41. Varnedoe, William W., Jr. 1973. Alabama caves and caverns. Alabama Cave Survey, Huntsville. Accepted 12 December 1980 Records of Leatherback Turtles, Dermochelys coriacea (Linnaeus), and Other Marine Turtles in North Carolina Waters Davids. Lee and William M. Palmer North Carolina State Museum of Natural History, P.O. Box 27647, Raleigh, North Carolina 2761 1 ABSTRACT. — New information is presented on the occurrence of five species of marine turtles in North Carolina waters. Dermochelys coria- cea and Caretta carettd, the two most commonly occurring species, are emphasized. Thirty-three unpublished records of Dermochelys for North Carolina, and information from other sources, indicate that in North Carolina at least, this turtle typically occurs throughout the warmer months in relatively shallow shelf waters. It may not be an open-ocean wanderer. Information on seasonal distributions of marine turtles in North Caro- lina’s offshore waters is fragmentary. Since 1975, one of us (DSL) has been regularly surveying seabirds in this area (Lee and Booth 1979) and making incidental observations of marine turtles. These records, com- bined with others in the files of the North Carolina State Museum of Natural History (NCSM), contribute considerably to our knowledge of sea turtle occurrences off the coast of the state. All five species known from the Atlantic have been encountered. Nearly all of the 85 offshore trips departed either from Oregon Inlet or Hatteras Inlet, Dare County, North Carolina. Each daylong outing (ca. 10-1 1 hours) typically followed predesignated transects of 20 to 50 miles (32 to 80 km) from the point of departure and into the Gulf Stream. Data accompanying sightings are, unfortunately, not uniform because of (1) lack of LORAN equipment on some charter boats, (2) abbreviated record keeping necessitated by conditions at sea, and (3) concentration of pri- mary field effort on seabirds, which sometimes made it impossible to record maximum data on turtles. Furthermore, surveys of marine turtles from boats are difficult, since surface conditions and angle of view nor- mally provide a narrow corridor of visibility. Variability of surface condi- tions from one trip to the next makes comparisons of trip-by-trip numbers observed meaningless. The difference in numbers of turtles seen from boats and numbers observed in aerial surveys, to be discussed later, is striking. Information from NCSM files was compiled from numerous inde- pendent records accumulated during the past 20 years. Units of meas- urements used here are, for the most part, expressed as originally reported to us; few have been converted to metric units. Brimleyana No. 5:95-106. July 1981. 95 96 David S. Lee and William M. Palmer Dermochelys coriacea coriacea (Linnaeus). Atlantic Leatherback Ernst and Gilroy (1979) summarized the 25 known occurrences of Dermochelys from the central Atlantic states (VA-4 records, MD-4, DE- 3, NJ-14) and concluded that, although these reptiles were generally believed to make long, open ocean journeys and are rare along the middle Atlantic coast, many remain close to shore during migrations and are seasonally common along the central Atlantic states. Lazell (1980) docu- mented regular occurrences in New England waters. Schwartz (1977) noted that in North Carolina this species is known from only seven adults, and juveniles from one presumed nesting at Cape Lookout, Car- teret County, reported to him in June 1966. There apparently are only three published records for the species in South Carolina waters (De Sola and Abrams 1933; Schwartz 1954; Pritchard 1976). Our 33 additional records (Table 1) seem significant when compared to the total number of leatherbacks reported from New Jersey south to South Carolina, and especially when compared to the modest number of loggerheads observed from boats offshore during the same survey period. Also, many of the 36 previous New Jersey to South Carolina records represent animals found dead on beaches or possibly unhealthy individuals, and may not reflect normal seasonal movements for the western North Atlantic Dermochelys “population.” Therefore, our observations may be of importance in even- tually understanding such movements. Four live individuals observed by DSL were floating just below the surface with only their heads protruding. When breathing, the entire head was exposed for 5 to 10 seconds. The animals were not as easily fright- ened as were most of the loggerheads observed, and we were able to keep the boat within 10 to 15 m of them for several minutes. Even then, turtles could not be seen until we were extremely close to them. Sightings were made only on days when the water surface was calm. Consequently, even on relatively calm days, leatherbacks would be much more difficult to census from a boat than would loggerheads, which float with part of the carapace above the surface. Several Oregon Inlet boat captains and mates insisted that leatherbacks were as common or more common offshore than loggerheads, but they could be seen only on calm days. Captain John Booth saw five in one day during the summer of 1976. At sea, observations and trawler catches of leatherbacks are from the relatively shallow waters over the continental shelf (>100 fathoms). Even though most of Lee’s observation time was spent along the edge of the continental shelf at or near the inner edge of the Gulf Stream, leather- backs were not seen there. Boat captains confirmed that turtles they saw were well inshore (10-30 miles) of the edge of the shelf. Atlantic trawler fishing also takes place in shallow (normally 20 to 30 m) shelf water. The North Carolina Marine Turtles 97 Fig. 1. Locations of Dermochelys sightings off North Carolina’s Outer Banks, 1976-1980. Dots indicate specific sightings. Dashed line encloses approximately 8+ sightings made from charter boats not equipped with LORAN instrumenta- tion. X indicates turtles found dead on beach. Contour lines are in fathoms. South Carolina specimen reported by Schwartz (1954) was trawled from water 25 to 30 feet deep, as were a significant number of the North Carolina turtles. In the 1979 aerial surveys of the North Carolina Wildlife Resources Commission, only two leatherbacks were sighted. During 225 hours of flight time, the spotters seldom ventured more than two to four 98 David S. Lee and William M. Palmer miles from the beach (see Caretta discussion). We conclude that, with few exceptions, leatherbacks confine their activities to the shallow waters over the continental shelf but normally remain well away from the beach, although long range migrants may travel more direct open-water routes. LazelPs (1980) records were all from well inside the 183 m contour, but he considered this to be a result of the distribution of food organisms {Cya- nea). Pritchard (1976) noted, however, that although his recapture localities made this species appear to be a coastal form, individuals cap- tured had ventured atypically close to shore. He also noted that the integument of the leatherback is so delicate that prolonged or preferred residence in shallow water is improbable, since even occasional contact with rocks or coral would likely cause extensive damage to the animals. Our limited data indicate that Dermochelys, at this latitude, tends to remain inshore over the continental shelf and not to venture regularly into water over 500 fathoms deep (see Figure 1 and Table 1). We have, though, only limited experience past the 1000 fathom contour. Since this turtle is assumed to be a surface feeder, water depth would appear to be inconsequential. In further support of our belief that the leatherback normally frequents shallow shelf waters rather than those of the open sea, we provide the following records from Bermuda waters, kindly supplied by David B. Wingate, Conservation Officer for Bermuda: 14 July 1835 (7 ft. long, estimated weight 1200 lb.); 8 August 1967 (81/2 ft. long from head to tip of tail, estimated weight 1 100 lb.; tangled in a fishing net and drifting help- lessly); 9 December 1972 ( 4 ft. 9 in. from tip of head to tip of tail, estimated weight 400 lb.; length of carapace 3 ft. V/z in.; width of shell around curvature at widest point 2 ft. 9 in.; span of forelimbs tip to tip 6 ft.; washed ashore dead on Coopers Point, Bermuda). Wingate noted that other individuals are occasionally sighted by fishermen, but no dates have been recorded. In view of the long involvement the citizens of Bermuda have had with the sea in general and sea turtles in particular, the limited number of records of Dermochelys from Bermuda indicates that it infre- quently occurs in open sea areas there. Additionally, Ralph W. Harvard and Howard E. Inspahr (Sula Pelagic Expeditions), and Charles D. Duncan, University of Alabama, Bir- mingham, reported to us records of 14 live and 3 dead leatherbacks in the Northern Gulf of Mexico (Daulphin Island, Alabama, area). Two of the three dead individuals were believed by them to have drowned in shrimp nets. Period of occurrence ranged from May to September (1975-80). “All sightings were in shallow green water (20-150 ft.) and always near a tide line. The food seems to be large jellyfish.” David Rupke, Louisiana State University, reported three individuals from 500 m off Horn Island, Mississippi, on 23 June 1980. They were in approximately 4 m of water. North Carolina Marine Turtles CT3 C U o z B o a a • •««* o -s: o Q o S 03 H U U Qi. D O on on U qC < UJ qC oC w H < X H Cu u Q B o jn ■i-t w u < [JU cC D on U UJ Ql! < W > H B o U* r- r-- r- r- r- r- 2: o os os N ^ N N N -t-i *-> ■t-j ss Ui k. Ui 03 03 03 U j= -C cn o CJ CJ on cn on cn cn -3 -3 33 U rj O O iz: 'w on cn on C. Z Uh u (D (U 3 3 kH u -*-> >> >> Xi X) ■«-> ■<-> 3; x: 00 00 3 3 3 3 0 0 I I I I I I 1 I os rn os in 00 — (N (N “ co3 o3 03 3 i 1 03 03 0 0 0 0 1 ! 0 0 H X -J h- iT) u2 00 00 00 00 00 00 00 00 M -xi ' N W 0 »n so 0 0 0 so 1—1 — 0 so 00 o3 cn > m m rrs k- _ 3 B OQ cn 1- B m 3 00 0 PQ Om U . os aj 3 03 > 0 kJ D, H < os CU 3 3 '—5 00 SO r- n 3 cc3 so < -> o X CL ■*-» o X Cl -«-> o X 3 (U CU (D CU M (U Ou X w X X ■*-» ."3 ."3 o O o o o o <4-1 o o o o o o X X X S QQ QQ PQ QQ QQ QQ •4-> (L> (U -•-> (U C/5 c/5 C/5 3 3 3 3 3 3 3 3 3 U u U X O X O <—> X o X O X O •—3 X O 3 (U c (U c lU TD - cd o iS (U % cd X) ■<-> JS oo 3 cd o (U cd Jc o iS c o ao _c -*-> cd O C ID ^ CJ •O Cd ^ >Ov cd — H u- 4j D C B o c B 3 T3 u ^ O > C*-l LTt 3 o (U u. O X c/3 3 0 m rsi tN fN r<3 '3- TT 1 1 1 Tt fNl 3 CJ 3 ■0 3 3 3 CJ QQ lU X 0 _CJ 3 X 00 X 0 »o '3- 0 c 0 0 0 X '3‘ r~ 00 r- (U •N r> 00 Ui V 00 3 3 Son X ON ON >3 3 S U-3 Tt On rsi — o 0 1 ON r- ON ON aj 0 X S g X r-- g g X 3 '—3 '—3 ON ON 3 (U 0 w c/5 C/5 c/5 0 (U 00 3 < July- 1977 July- 1977 July- 1977 July-Sept. 1977 1 35°50',74°53^ — ca 24-25 50 Kenneth W. Pekrum 17 June 1978 1 0.8 km offshore, be- ca. 24 5.5 caught by trawler NCSM file tween Emerald Isle Pier and Rogue Inlet — North Carolina Marine Turtles x: OS) C/3 2 TD c/5 ^ u S Z Si c/3 e "w ;z: fS rt (U O ^ x: OS) • c/3 S c/5 U Z c B (U 1-1 o Um c OJ B B a> o .“t: U "rt in -i? (u O u S . (U C/3 E u c/5 CQ Z T3 Uh O o u Xl .2? ' n S C/5 U z l-l cS X o (U cd O (U o X , oJ 0 <4-, c T3 D u ■1^ C,— 1 (U • - E C/5 cf2 <4-1 0 X 0 c/5 l-l 3 T3 U 3 c iH 0 U l-l 3 0 4-> 4-1 cd C.I-- 3 O c/3 c o ^-1 o 00 c iX C X c X (U c/3 1 c/3 TD X D, c/3 Cy c c/3 3 C cd Ui lU > r3 Ov X c/3 # rN CJ CJ c/3 M _4J "c X o c/3 X ON 1- 3 C/3 1^ 3 cd X 0 Tf OQ 0 0 c/3 c ^ 0^ 0 *4 0 C3 w u D m 0 in 00 m 0 in m 1-, 3 CJ — -4 0 0 Cl, 0 0 73 ‘C <0, 2 S - u z 10 20 30 40 BODY LENGTH mm Fig. 1. Length-frequency histograms at seasonal intervals of P. postica larvae from Greenbottom Swamp. M = March, D = December, N = November, and number = sample size. months. Wiggins (1973) reported living, first-instar, phryganeid larvae (of unknown genus), still within a gelatinous matrix, beneath wet leaves in a pool basin in Algonquin Park, Ontario in October 1960. Wiggins (1977) reported final instars of P. ocellifera in a littoral region of a British Columbia lake from November to June. Larval case length also showed seasonal variation (Fig. 4): fall (November) x = 23 mm, R = 18-33 mm; winter (December) x = 39.3 mm, R = 22-55 mm; spring (March) x = 50.0 mm, R = 33-69 mm. The correla- tion between case length and body length was high in fall (Y = 11.65 + 0.86X, r= 0.72) and winter (Y = -10.61 + 2.25X, r= 0.87) (Fig. 5), but low in spring (Y = 27 + 0.74X, r = 0.37). In these equations, Y = case length (mm) and X = total body length (mm). These data indicate that body length and case length are closely related in fall and winter, but in spring case construction continues as larval growth slows down. Once the larva of P. postica has begun to add a new ring at the anterior end of the case, it completes this addition rather rapidly, often with segments of the same leaf. The larva may pause for several days, or even SPRING M 43 Caddisfly Life History 111 2.8- 2.6- 2.4- 2.2- E E 1.8- 1.6- 1.4- 1.2- 1.0- 75% rh IV.-V* 14% FALL WINTER SPRING N D M 20 11 43 Fig. 2. Seasonal variation of head width in P. postica larvae. Horizontal lines = means; vertical lines = ranges; open rectangles = one standard deviation; shaded rectangles = standard error of the mean; letter = month; number = sample size; and % = growth rate. weeks, however, before adding another anterior ring (Fankhauser and Reik 1935). The larvae in our study were largely carnivorous early in their life cycle, shifting to a more omnivorous diet in later stages (Table 1). The diet of larvae collected in the fall consisted primarily of the copepod, Cyclops sp. (x= 4.82 per foregut, 85 percent frequency of occurrence (% FO) ). Ostracods (x = 2.1 per foregut, 50% FO), diatoms (x = 0.42 per foregut, 20% FO), and detritus (x = 11.55 per foregut, 100% FO) were also found. In winter, the larval diet consisted primarily of Cyclops sp. and ostracods (x = 4.33 and 2.33 per foregut, respectively, 55% FO for each). Diatoms (x = 1.19 per foregut, 73% FO) and detritus (x = 18.39 per foregut, 100% FO) were also found in winter. The larval diet shifted in spring and consisted primarily of ostracods and midge larvae (x = 4.5 and 2.0 per foregut, respectively, 20% FO for each). Detritus (x = 27. 12 per foregut, 100% FO) and diatoms (x = 0.90 per foregut, 62% FO) were also found during this season. 112 Mary Beth Roush and Donald C. Tarter Instar Head width ranges mm I II III 0.96-1.07 IV 1.38 - 1.73 V 2.16 -2.66 Fig. 3. Head width of all P. postica larvae versus number of larvae. Wiggins (1977) stated that phryganeid larvae generally are omni- vorous, with a few predaceous for at least part of their life cycle. Merritt and Cummins (1978) classified Ptilostomis larvae into two func- tional feeding groups: (1) shredders and detritivores, and (2) engulfers. Shapas and Hilsenhoff (1976) reported Ptilostomis spp. from the Mecan River in Wisconsin that were exclusively carnivorous. Wiggins (1977) reported that late instars of P. ocellifera from a British Columbia lake were largely predaceous. Pupal stage. — Fifty-two pupae were collected in April and May, the first on 23 April and the last on 16 May. Most were found embedded in logs. April measurements were: x length = 21.84 mm, R = 17-27.5 mm; x head width = 2.77 mm, R= 2.4-3 mm. May measurements were: x length = 22.23 mm, R = 18-23 mm; x head width = 2.62 mm, R = 2. 3-2. 7 mm. Adult stage. — The first adult was collected on 12 May and the last on 27 May; 31 were collected after dusk. In one hour, 1, 12, 14 and 4 adults were collected on May 12, 17, 20 and 27, respectively. Twenty females and 1 1 males were collected. A chi-square test showed no significant deviation from a 1:1 sex ratio at the 0.05 confidence level. No eggs were observed in adult females or in the field. According to Wiggins (1973), O. S. Flint Jr. reported that adult Ptilostomis in Michi- gan deposited eggs in August, 10 to 20 cm above the water surface of streams. Gelatinous masses of eggs were attached to the underside of a log and an overhanging bank. Caddisfly Life History 113 70- 60- 50- E ^ 40- O) 1 30- 0) «/) o ^ 20- 10- 0-1 20 FALL WINTER SPRING N D M Fig. 4. Seasonal variation of case length in P. postica larvae. Horizontal lines = means; vertical lines = ranges; open rectangles = one standard deviation; shaded rectangles = standard error of the mean; letter = month; and number = sample size. ACKNOWLEDGMENTS. — We would like to express our thanks to Miss Vickie Crager for typing the manuscript. Also, we thank the follow- ing gradutate students for assisting in the field work: Kerry Bledsoe, Bill Cremeans, Pam Dolin, Paul Hill, Steve Lawton and Joan Schramm. 114 Mary Beth Roush and Donald C. Tarter Table 1. Seasonal foregut analysis of Ptilostomis postica krvae from Green- bottom Swamp, Cabell County, West Virginia. X = mean, %FO = percent frequency of occurrence. N = November (20 larvae), D = December (1 1 larvae), M = March (43 larvae). Fall Category Taxa X %FO (N) Copepods Cyclops sp. 4.82 85 Ostracods 2.1 50 Cladocerans 1.0 15 Midge larvae Chironomus sp. 1.0 5 Rotifers 1.0 5 Filamentous algae 1.0 20 Unidentified 1.0 30 Grid Detritus 11.55 100 method: Diatoms 0.42 20 Winter (D) Copepods Cyclops sp. 4.33 55 Ostracods 2.33 55 Cladocerans 1.75 36 Midge larvae Chironomus sp. 1.0 9 Rotifers 11.0 9 Dipterans 1.0 9 Filamentous algae Microspora 1.0 9 Unidentified 1.0 18 Grid Detritus 18.39 100 method: Diatoms 1.19 73 Spring (M) Ostracods 4.5 20 Midge larvae Chironomus sp. 2.0 20 Rotifers 9.0 5 Filamentous algae 1.0 5 Unidentified Odonata 1.0 5 Grid Detritus 27.12 100 method: Diatoms 0.90 62 Caddisfly Life History 115 Total length Total length mm mm Fig. 5. Regression analysis of body length and case length for P. postica larvae. Number in parentheses = sample size in fall and winter. LITERATURE CITED Betten, Cornelius. 1934. The caddis flies or Trichoptera of New York state. N. Y. State Mus. Bull. 292. 576 pp. Fankhauser, Gerhard, and L. E. Reik. 1935. Experiments on the case-building of the caddisfly larva, Neuronia postica Walker. Physiol. Zool. (^(3):337-359. Hill, Paul L., D. C. Tarter, B. Cremeans and M. B. Roush. 1978. State records of family Phryganeidae in West Virginia (Insecta; Trichoptera). Proc. W. Va. Acad. Sci. 50(1):24. Merrill, Dorothy. 1969. The distribution of case recognition behavior in ten families of caddis larvae (Trichoptera). Anim. Behav. /7(3):486-493. Merritt, Richard W., and K. W. Cummins. 1978. An introduction to the aquatic insects of North America. Kendall/ Hunt Publ. Co., Dubuque, Iowa. 441 pp. Ross, Herbert H. 1944. The Caddis Flies, or Trichoptera, of Illinois. 111. Nat. Hist. Surv. Bull. 23. 326 pp. Shapas, Theodore J., and W. L. Hilsenhoff. 1976. Feeding habits of Wisconsin’s predominant lotic Plecoptera, Emphemeroptera, and Trichoptera. Great Lakes Entomol. 9(4): 175-188. Sibley, Charles K. 1926. Trichoptera. pp. 102-108, 185-221 in A preliminary biological survey of the Lloyd-Cornell Reservation. Bull. 27 Lloyd Libr. Bot. Pharm. Mater. Med. Siltala, Antti J. 1907. Trichopterologische Untersuchungen. Zool. Jahr. Abt. Syst. Oekol. Geogr. Tiere. (Suppl. 1) 9:309-626. 116 Mary Beth Roush and Donald C. Tarter Wesenberg-Lund, Care. 191 1. Uber die biologie der Phryganea grandis und uber die Mechanik ihres Gehausebaues. Int. Rev. Gesamten Hydrobiol. 4.65-90. 1913. Wohnungen und Gehausebau der Susswasserinsekten. Fortschr. Naturw. Forsch. 9:55-132. Wiggins, Glenn B. 1960a. A preliminary systematic study of the North American larvae of the caddisfly family Phryganeidae (Trichoptera). Can. J. Zool. 55:1153-1170. 1960b. The unusual pupal mandibles in the caddisfly family Phryganei- dae (Trichoptera). Can. Entomol. 92(6):449-457. . 1961. The rediscovery of an unusual North American phryganeid, with some additional records of caddisflies from Newfoundland (Trichoptera). Can. Entomol. 95(8):695-702. 1973. A contribution to the biology of caddisflies (Trichoptera) in temporary pools. R. Ont. Mus. Life Sci. Contrib. 88. 28 pp. 1977. Larvae of the North American caddisfly genera (Trichoptera). Univ. Toronto Press, Toronto. 401 pp. Accepted 4 February 1981 Reproduction of the Eastern Cottonmouth Agkistrodon piscivorus piscivorus (Serpentes: Viperidae) at the Northern Edge of its Range Chari.es R. Beem Department of Biology, Virginia Commonwealth University, Academic Divison, Richmond, Virginia 23284 ABSTRACT. — Eastern cottonmouths, Agkistrodon piscivorus pisci- vorus, were studied at the apparent northeastern edge of the species’ range (Hopewell, Virginia). Although biennial reproduction is generally typical of viperid snakes, with approximately 50% of the females pro- ducing young in any given year, 83% of the females collected at the Hopewell site were gravid. Mean litter size (7.7) does not appear to differ from that of populations in other parts of the range. Young cottonmouths at the Hopewell site appear to be smaller than those from other areas and survival is much reduced in years having cold winters. Average snout-vent length of adult females at Hopewell currently is greater and less variable than at locations to the southeast and within the main range of the species. Production of offspring is a function of female size; larger females produce more calories of young and larger offspring. Reproduction in kcal may be accurately predicted from the weight of the female (g) using the equation: kcal = 0.61 weight - 41.20. INTRODUCTION The Eastern Cottonmouth, Agkistrodon piscivorus piscivorus, is one of the more widespread and abundant poisonous snakes in the Southeast. It is generally a lowland species (e.g. Kofron 1978), although in the deep South it may be found in a variety of habitats (Burkett 1966, Mount 1975). In the northern part of its range the cottonmouth becomes more selective in its choice of habitat, and in Virginia it is known only from wetlands in the Coastal Plain. The northeastern extreme of the species’ range apparently is represented by an isolated population at the conflu- ence of the Appomattox and James rivers near Hopewell, Virginia (Fig. 1). At this locality the species occurs in the lower reaches of Swift Creek, Chesterfield County, and along both west and east banks of the Appo- mattox River. No cottonmouths have been collected on the James River in this region. The vegetation of the area, a mosaic of tidal swamps, tidal marshes, and upland forests, will be described elsewhere (Blem and Gutzke, in preparation). This population is at least 60 km from the nearest known locality in the main part of the range, and repeated efforts to find the species in apparently excellent habitat at other sites near Hopewell have been fruitless. Little information is available regarding reproduction of snakes at the northern limits of their distributions and no studies of reproduction of A. p. piscivorus have been published. Studies by Wharton (1966, 1969) in Brimleyana No. 5:1 17-128. July 1981. 117 118 Charles R. Blem North Carolina Fig. I. Localities from which the eastern cottonmouth has been taken in Virginia and northeastern North Carolina. Hollow circle represents the Hopewell site. Florida were of the subspecies A.p. conanti and those by Burkett (1966) and Kofron (1979) involved the western subspecies A. p. leucostoma. Also, there is practically nothing in the literature about the ecology of the cottonmouth in Virginia. In fact, Wood (1954), in a survey of the distri- bution of the snake in Virginia, did not mention the Hopewell site and maintained that cottonmouths did not occur west of the Appomattox River. In this paper I also examine briefly some hypotheses regarding animals at the northern edges of their ranges. For example, environmental condi- tions in such peripheral areas usually are severe relative to the tolerance of a species. Mortality of adults and/or young is expected to be high and occasionally catastrophic; reproduction rates should be adjusted corre- spondingly. A less variable population may be produced through stabiliz- ing or normalizing selection, or, in time, the species may be eliminated entirely. Additionally, the size of young might be adjusted through natural selection in ways that promote survivorship early in life. Eastern Cottonmouth Reproduction 119 METHODS Cottonmouths were observed and collected by hand in more than 500 hrs of field work during April through October 1977-1979. Newly cap- tured snakes were sacrificed, weighed to the nearest 0.1 g, and body measurements (total, snout-vent, and tail lengths; head width and length) were determined to the nearest mm. Each snake was dissected and the ova of the females were counted and their lengths measured to the nearest mm. Fat bodies were carefully removed to exclude obvious connective tissue, weighed to 0. 1 g, and freeze-dried to constant weight. Lipids were extracted from the fat bodies in Soxhlet refluxers in 5:1 petroleum ether:chloroform solution. Loss of weight of the dried sample after extraction represents the amount of extractable lipid. Some obviously gravid female cottonmouths captured in late July and August of each year were kept in captivity until the young were born. Shortly after birth, measurements of young were made as described above. Caloric equiva- lents of lipid, young snakes, and unfertilized eggs were determined by means of a Parr nonadiabatic bomb calorimeter. Scales were counted according to criteria of Burkett (1966), which essentially include Dowling’s (1951) technique for counting ventrals. For comparative purposes ventral melanism was quantified by counting all ventrals from the anal scale anteriad to (but not including) the first three consecutive scales with less than half of their total area covered by black pigment. RESULTS A total of 68 male and 38 female cottonmouths was collected from the study area. Chi-square tests (with Yates continuity correction) indicate that the maleifemale ratio differs significantly from 50:50 (X^=7.93). The difference may be due to intersexual variation in activity, but since most snakes were collected at rest near the entrance of “summer dens” I believe the sexual bias is real. Mainly two types of females were found: those possessing large, yolked follicles (45-50 mm), and those with small, yolk- less ova (8-1 1 mm diameter). This is similar to Rahn’s (1942) findings in a study of Crotalus viridis and those of Tinkle (1962) in an investigation of Crotalus atrox. Apparently, ovarian eggs ripen rapidly and yolk deposi- tion of the set may require only a few weeks at most. Two females captured 13 April 1977 and 13 April 1978 possessed ovarian eggs 16-20 mm long; earliest gravid females were found 31 May 1977. In order to determine the relative occurrence of reproductive activity in mature females within the breeding season, I eliminated from the analysis all those smaller than 700 mm total length (the smallest gravid female was 732 mm), all captured earlier than the first gravid one (3 1 May), and all captured after September 30 (no gravid animals were captured after this 120 Charles R. Blem date; all known birth dates are earlier than this). Only 5 of the remaining 29 were not gravid; the rest contained uterine eggs or produced young in the laboratory. This ratio (24:5=83% gravid) deviates significantly from the 50:50 ratio predicted from published studies (see below, X =11.18). (Note: during research of fat cycles and energetics of cottonmouths in 1980, we collected 5 adult females; all were gravid.) Based on the number of yolked, oviductal eggs or actual number of young born, mean litter size for the 24 gravid females is 7.68 ± 1.77. The mean caloric value of yolked cottonmouth eggs is 6.136 ± 0.067 kcal/g dry weight (N=3); newborn young contain a mean of 5.5 10 ± 0.055 kcal/ g dry weight (N=3). After correcting for water content of young and eggs, I computed the total caloric content of the offspring of 1 1 female cotton- mouths that gave birth in the laboratory. Three of these females pro- duced single infertile eggs along with viable litters. The eggs are counted as part of the litter, and the caloric content computed separately and added to the calories invested in living young. The total represents total caloric output (but not expenditure) in reproduction. All the females in this part of the study were held for relatively short periods (45 days or less) and all births occurred without mortality. The caloric content of the litter is a significant function of the weight of the female shortly after giving birth (Fig. 2): kcal = 0.61 weight (g) - 41.20, r = 0.93 Cottonmouth fat bodies are composed of 26.8% water and the remain- ing dry material is 94.5% lipid. Using the caloric content of lipids extracted from the fat bodies (x=9.227 ± 0.030 kcal/g, N=4), the mean caloric reserves of the fat bodies of female cottonmouths were computed. Assuming that the caloric content of the smallest litter (104.9 kcal) represents a threshold minimum, no cottonmouth possessing fat bodies weighing less than 16.5 g (=105.3 kcal) could be expected to reproduce, especially those collected in early spring or summer. A distinct cycle of lipid deposition is discernible in female cottonmouths at Hopewell, in which reserves are generally low in spring and increase in autumn (Blem, unpublished). In early spring, before the first gravid female was collected, 3 of 6 nongravid females did not possess lipid in excess of this threshold. During the period 31 May - 30 September, 3 of 5 nongravid females had fat bodies below threshold levels, and after 30 September all females (3) had less fat than the amount apparently necessary for reproduction. Seven female cottonmouths collected during mid-September in extreme southeastern Virginia likewise possessed insufficient reserves (fat bodies weighed from 1.5 to 11.0 g) for reproduction. Overall, nonreproductive females possessed slightly more fat than those containing eggs or produc- ing young (“reproductive females”), although the difference is not statis- tically significant (Table 1). Male cottonmouths possessed significantly greater amounts of fat than females in either category, but the difference Eastern Cottonmouth Reproduction 121 Fig. 2. Total energy content (kcal) of offspring as a function of female body weight (g). The equation for the relationship is: kcal offspring = 0.61 weight of female (g) - 41.20; r = 0.93. Eleven different females are included in the analysis. in relative fatness, expressed as a percentage, is not statistically significant. Larger females gave birth to correspondingly larger young (Table 2). Length and weight of young cottonmouths at Hopewell are distinctly smaller than those of newborn cottonmouths at Florida sites (Allen and Swindell 1948, Wharton 1966), but sufficient data are lacking at present for proper analyses of geographic trends. The “heritability” of size of young cottonmouths at the present study location is statistically signifi- cant, but comparative data from other study sites are unavailable. Also, heritability of size is expected to be statistically great when only the female parent is known (Falconer 1960). The physiological/ ecological basis for the relationship between the size of female snakes and their 122 Charles R. Blem young remains to be shown, but it probably is related to aspects of energy mobilization and amount of lipid reserve in larger snakes. Other than the correlations between female size and size of young, the present studies indicate no significant correlation between head length and width, between number of subcaudals and ventrals, or with degrees of melanism. Head lengths and widths and subcaudal counts are sexually dimorphic, and the scatter introduced into a mixed litter of male and female young probably obscures correlations between scale counts of the parent and young. Melanism apparently increases with age, as adult females generally score higher than their young (see Table 2). Lack of a significant correlation between ventral counts of parents and young may be due to low overall variation in ventrals. Table 1. Weight of fat bodies and body weight in adult eastern cottonmouths at Hopewell, Virginia (± 1 standard error). Reproductive females Nonreproductive females Males Mean body weight (g) 430.8 ± 30.4 402.6 ± 28.7 857.2 ± 60.6 Mean fat weight (g) 12.3 ± 2.1 15.4 ± 2.2 33.7 ±4.3 Percentage fat 2.9 3.8 3.9 N 24 14 68 Table 2. Relationship of morphometries of young cottonmouths to female parents. Values are means, r = correlation coefficient, Y = measurement of young, F == measurement of female. Females Young r Equation N 10 80 — — Body weight (g) 363.5 17.3 0.70* Y = 0.018 FT 10.168 Snout-vent length (mm) 747.2 211.3 0.58* Y± 0.061 FT 11.039 Melanism 38.1 17.3 0.40 Ventrals 133.2 132.9 0.40 Subcaudals 41.5 44.6 0.22 Head length (mm) 44.3 20.5 0.34 Head width (mm) 34.9 13.2 0.42 *P<0.05 Eastern Cottoiimouth Reproduction 123 DISCUSSION It is obvious that the limits of a species’ range are reached where mortality regularly exceeds reproductivity. At the Hopewell site we have established that survival of young cottonmouths is quite poor in years in which the winter is unusually cold (Blem and Gutzke, in preparation). The success of females giving birth in the lab indicates that reproduction regularly occurs without undue mortality. In the field, several litters of newborn cottonmouths have also been observed in September. No small cottonmouths (i.e less than 403 mm snout-vent length) have been col- lected at Hopewell in spring or summer, although we regularly collect individuals of 200-500 mm in extreme southeastern Virginia. I believe this is related to the extremely cold winters of 1976-1977 (the coldest January since 1940) and 1977-1978 (the second coldest January of the period 1971-1979). No other environmental variable would seem to account for the disappearance of young overwinter. The more abundant size classes over 600 mm appear to represent greater survivorship over mild winters before 1976. A variety of forces may contribute to death in cottonmouths. Man is probably the major cause of mortality at most sites (Burkett 1966), but starvation during mild winters (Wharton 1969), disease and parasitism (Burkett 1966), and a wide variety of predators (Allen and Swindell 1948, Barbour 1956, Burkett 1966) all take their toll. Mortality from cold seems to be absent in the main range, as tolerance of low temperatures seems well developed in cottonmouths (Wharton 1969). Perhaps only at the northern edge of the range does freezing occur so intensively or for such long periods that survival in the hibernacula becomes difficult. The ability of a species to maintain its distributional limits or expand them depends upon behavioral, physiological, and morphological adap- tations that balance the birth-death ratio. Among the reproductive adjustments that a reptile might demonstrate are: 1) increased litter size, 2) more frequent ovulation, 3) earlier sexual maturation, 4) early ovula- tion or more rapid development of young, 5) production of young that are better adjusted for survival of the environment at the distributional extreme, and 6) niche selection that maximizes the accumulation of energy for reproduction. Few data are available with which to compare litter size at the Hope- well site. Litter sizes in cottonmouths collected in the Everglades (x=6.6 ± 2.5, N=31: Allen and Swindell 1948) do not differ significantly from the data in my study. Mean litter size of females from the Cedar Keys, Florida (5.5 ± 1.5, N=24; Wharton 1966) is significantly smaller (t=4.8, df=46) than that of the Hopewell site. I conclude that there is little evidence at present that indicates any adjustment of the number of young in litters at the northern extreme of the range. 124 Charles R. Blem Reproduction by individual females within many species of snakes apparently is not an annual phenomenon. Evidence indicates that many viperid species produce young at two-year or longer intervals (St. Girons 1957, Tinkle 1962, Fitch 1970, Aldridge 1979). Burkett (1966) and Whar- ton (1966) reported a biennial cycle in A. piscivorus, but an annual cycle has been found in some populations of Louisiana cottonmouths (see Kofron 1979). It is generally believed that snakes at northern latitudes often are not able to accumulate sufficient energy to produce and mature a set of eggs annually (Rahn 1942, Klauber 1972, Tinkle 1957, St. Girons 1957), but my data contradict this generalization. If reproduction is bien- nial or less frequent, 50% or less of the mature females collected should possess ripe eggs in the uterus (as did Florida cottonmouths; see Wharton 1966). However, at Hopewell during 1977-1979, 83% of the females that could have possessed uterine eggs were, in fact, gravid. This ratio is significantly different from 50:50 and indicates that Hopewell cotton- mouths breed more regularly than expected. Frequency of reproduction of viviparous snakes is believed to be a function of fat storage (Tinkle 1962, Wharton 1969, Aldridge 1979). Assuming this is true, the Hopewell population during 1977-1979 seemed to be well prepared for reproduction. Lipid reserves were sufficient for most mature females to produce full litters. Hopewell cottonmouths also exceeded the lipid levels found in all reproductive classes (reproductive and non-reproductive females, males) in Wharton’s (1966) Florida study. A low population density at Hopewell may have reduced competition for food and provided for more productive foraging. At present there are no data with which to assess this hypothesis, but a subjective evaluation indicates that cottonmouths are more abundant in the main part of the range (e.g. extreme southeastern Virginia). The high rate of reproduction found in my study may also be related to winter extremes. Mild winters may cause decreases in fat reserves because of elevated metabolism (Wharton 1966), and conversely, very cold winters may encourage repro- duction through conservation of fat stores. As stated above, the winters during my study were among the coldest of the decade. Few data are available regarding size of female cottonmouths at first reproductive effort. Arny (in Wharton 1966) found that female A. p. leucostoma may mature at a total length of 594 mm. Wharton (1966) noted that 800 mm was the minimum total length of mature females in the Cedar Keys of Florida. At least three Hopewell cottonmouths less than 800 mm total length (732, 770, and 791 mm; =619, 654, and 684 mm snout-vent length) contained oviductal eggs, and two females (707 and 756 mm total length) collected in September had ovarian eggs of 10-11 mm. I conclude that early maturation at the Hopewell site is a viable hypothesis that requires further study. Eastern Cottonmouth Reproduction 125 Although sufficient data are not available for rigorous testing, time of ovulation apparently varies little from one site to the next. Wharton (1966) estimated that Florida cottonmouths ovulated in the period 20 May-8 June, and Burkett’s (1966) data indicate similar timing, as do my data. I have no measurements of growth of young because smaller speci- mens were absent throughout this study. Weight of young at birth could be a fundamental indication of the fitness of offspring. Young cottonmouths at Hopewell are born in Sep- tember (usually around the middle of the month, the mean date = 18 September, range = 5 September-23 September, N=13). They therefore have little opportunity to forage (and accumulate energy stores) before finding a hibernaculum for the winter. In addition, increased body mass may conserve heat and enhance survival at low ambient temperature in the hibernaculum. Cooling rates of large snakes should be slower, allow- ing them to survive short periods of extreme cold or to move to deeper parts of the hibernaculum before cooling to fatal temperatures. Snout- vent lengths and weights of young from other parts of the range largely reflect size of the parental race (see Wright and Wright 1957) and it appears that Hopewell cottonmouths do not produce particularly large young (Table 2). Weights and total lengths of young Florida cotton- mouths (324-343 mm; 28.0-33.7 g) are distinctly greater than those of the Hopewell population (Allen and Swindell 1948, Wharton 1966). Reduced size of young cottonmouths at Hopewell could be interpreted as an evolu- tionary compromise that maintains litter sizes while reducing the expen- diture of energy per litter. This would be advantageous where fitness of small young is high and availability of energy is low. Neither seemed to be the case at Hopewell in 1977-1979, and the significance of small offspring there remains to be shown. Size of mature females at the Hopewell site varies less than in popula- tions in southeastern Virginia or northeastern North Carolina, and mean snout-vent lengths of Hopewell snakes (excluding newborn young) aver- age significantly longer than those in the main part of the range to the southeast (Blem and Gutzke, in preparation). Moreover, within the total range of size, reproductive output is a linear function of weight of the female. Larger females are apparently able to mobilize greater amounts of energy in reproduction, and we have found that, in general, larger individuals have larger fat reserves. The so-called “biennial reproduc- tion”, with 40-60% of females breeding in any given year, may only be an artifact of the collective reproductive rates of a population of females of diverse sizes. Large females reproduce more frequently than small (Bur- kett 1966), and generally have larger lipid reserves; at Hopewell: g lipid = 0.05 snout-vent length (mm) - 19.65, r=0.55. It is therefore likely that larger individuals are capable of collecting and mobilizing larger energy 126 Charles R. Blem reserves and have a greater statistical probability of reproducing in any single year. It appears from this that descriptions of reproductive fre- quency at specific sites are of little value without a description of the sizes of the females involved. The percentage of gravid females in popula- tions composed predominantly of large individuals generally should exceed that of populations containing significant numbers of smaller (but mature) females. The size of young cottonmouths is at least partly determined by the size of the female parent, and as shown, the correlation between female size and that of offspring is fairly large. While selection for large body size in female cottonmouths at Hopewell may be only a phenomenon related to periodic mortality of small snakes, the present situation probably reduces the amplitude of population oscillations at least partly as a result of the relationship between female size and reproductive output. Increase in reproductive output appears to be related to number of mature ova pro- duced and size of offspring (Burkett 1966, and the present study). Elimi- nation of small individuals in extreme years should decrease competition for food among females and might further insure their successful contact with males. Production of young with a relatively high degree of fitness could occur, with most females producing young in breeding seasons following severe winters. Mild winters should enhance subsequent survi- val of young, and decreased lipid reserves and increased competition would result in decreases in the number of females producing offspring. The bias toward large size is even more striking in males (Table 1). No males less than 640 mm total length were captured in 1977-1979 and 18 males were collected that exceeded 1000 mm and 1000 g (the largest was 1345 mm and 2008.5 g). Since Wharton (1966) considered males above 650 mm to be sexually mature, it is likely that the male segments of the 1977-1979 populations were almost entirely mature. It appears that males generally exceed females in size; Wharton (1966) obtained a similar result in his Florida studies. The cottonmouth at Hopewell is much more habitat-specific than in some other parts of its range. For example. Mount (1975) described the cottonmouth in Alabama as “occurring in almost every type of perma- nently aquatic habitat”. At Hopewell it is found very locally and only at sites where tidal marshes and nearby upland provide feeding and hiber- nating sites, respectively. The apparent failure of the species to colonize a number of nearby sites that seem superficially suitable is evidence that relatively inconspicuous factors may be influential in determining habitat suitability for the cottonmouth at the northern edge of its range. Precise habitat selection may be an important factor in enabling female cotton- mouths to obtain sufficient energy for reproduciton. It is obvious that with the present survival rate of young the future of Eastern Cottonmouth Reproduction 127 this population is in jeopardy. Even though small cottonmouths (other than newborn individuals seen in autumn) have begun to appear in 1980, the population structure is skewed heavily toward longer and older indi- viduals. Local people kill many of these adults each year, along with many brown water snakes, Nerodia taxispilota, mistaken for cotton- mouths. I predict the species will become very rare at this site in the future. ACKNOWLEDGMENTS. — I am grateful to Leann Blem, Claire Filemyr, Mike Miller, Cheryl Roeding, and Tom Thorp for assistance in the field. William Gutzke made the project possible through his extensive assistance in the field and laboratory and asked questions that stimulated me to find answers. Val Combs assisted in preparation of the manuscript and John W. Steiner and C. C. Steirly provided important suggestions or observations. Figure 1 is based on specimens loaned to me from the Carnegie Museum of Natural History, National Museum of Natural His- tory, and North Carolina State Museum of Natural History. 1 am indebted to the curators of these collections. Virginia Commonwealth University Computer Center provided computational facilities. Grants from the Virginia Academy of Science and the Virginia Commonwealth University faculty grant-in-aid program provided support for this research. LITERATURE CITED Aldridge, Robert D. 1979. Female reproductive cycles of the snakes Arizona elegans and Crotalus viridis. Herpetologica i5(5):256-261. Allen, E. Ross, and D. Swindell. 1948. Cottonmouth moccasin of Florida. Her- petologica (First Suppl.) 4:1-16. Barbour, Roger W. 1956. A study of the cottonmouth, Ancistrodon piscivorus leucostoma Troost, in Kentucky. Trans. Ky. Acad. Sci. 77(1). 33-41. Burkett, Ray D. 1966. Natural history of cottonmouth moccasin, Agkistrodon piscivorus (Reptilia). Univ. Kans. Publ. Mus. Nat. Hist. 7 7(9). 435-491. Dowling, Herndon G. 1951. A proposed standard system of counting ventrals in snakes. British J. Herpetol. 7(5):97-98. Falconer, D. S. 1960. Introduction to Quantitative Genetics. R. MacLehose and Co., Glasgow. 365 pp. Fitch, Henry S. 1970. Reproductive cycles in lizards and snakes. Misc. Publ. Univ. Kans. Mus. Nat. Hist. 52:1-247. Klauber, Laurence M. 1972. Rattlesnakes: their habits, life histories, and influ- ences on mankind. Second Ed. Univ. California Press, Berkeley. 2 vols. 1523 pp. Kofron, Christopher P. 1978. Foods and habitats of aquatic snakes (Reptilia, Serpentes) in a Louisiana swamp. J. Herpetol. 72(4):543-554. , 1979. Reproduction of aquatic snakes in south-central Louisiana. Her- petologica i5(l):44-50. Mount, Robert H. 1975. The reptiles and amphibians of Alabama. Auburn Univ. Agric. Exp. Stn., Auburn. 347 pp. 128 Charles R. Blem Rahn, H. 1942. The reproductive cycle of the prairie rattlers. Copeia 1942(4); 233-240. St. Girons, H. 1957. Le cycle sexuel chez Vipera aspis (L) dans I’ouest de la France. Bull. Biol. Fr. Belg. 9y(3):284-350. Tinkle, Donald W. 1957. Ecology, maturation and reproduction of Thamnophis sauritus proximus. Ecology 38{\):69-71. . 1962. Reproductive potential and cycles in female Crotalus atrox from northwestern Texas. Copeia 1962(2):306-313. Wharton, Charles H. 1966. Reproduction and growth in the cottonmouths, Agkistrodon piscivorus Lacepede, of Cedar Keys, Florida. Copeia 1966(2): 149-161. . 1969. Cottonmouth moccasin on Sea Horse Key, Florida. Bull. Fla. State Mus. Biol. Sci. 74(3):227-272. Wood, John T. 1954. The distribution of poisonous snakes in Virginia. Va. J. Sci. 5(3): 152-167. Wright, Albert H., and Anna A. Wright. 1957. Handbook of snakes of the United States and Canada. Vol. 2. Comstock Publ. Co., Ithaca. 1 106 p. Accepted 8 October 1980 Habitat Use and Relative Abundance of the Small Mammals of a South Carolina Barrier Island John B. Andre^ Cape Romain National Wildlife Refuge, Route 1, Box 191, Awendaw, South Carolina 29429 ABSTRACT. — Species composition, habitat use, and relative abun- dance of the small mammal fauna of Bulls Island, Charleston County, South Carolina were determined by snap trapping on 17 days between December 1978 and February 1979. Ninety-five mammals of four spe- cies were captured in 2909 trap nights, for an overall trapping success of 3.3%. Oryzomys palustris was found in all habitats and was the only species in the dunes, salt spray forest, and freshwater marsh habitats. Sigmodon hispidus was numerically dominant in the old-field, while Peromyscus gossypinus and Mus musculus were most abundant in the residential area. INTRODUCTION Barrier islands of the southeastern coast of the United States are inter- esting areas for biological studies due to their isolation from, yet close proximity to, the mainland (Gibbons and Coker 1978). Species diversity and composition vary between neighboring islands and between islands and the mainland (Dueser et al. 1979). Patterns of colonization and interisland variability in species characteristics, e.g. habitat preference in association with different species on different islands, may emerge as more data on animals inhabiting different islands are made available for comparisons. Terrestrial vertebrate studies have been conducted on some of the islands of Virginia (Dueser et al. 1979), North Carolina (Engels 1942, 1952), South Carolina (Pelton 1975) and Georgia (Hillestad et al. 1975; Johnson et al. 1974; Teal 1962). However, published information on the vertebrates of many barrier islands is lacking. This paper describes the previously unreported small mammal fauna of a South Carolina barrier island. The objectives of the study were to (1) determine the species composition and relative abundance of the small mammals on Bulls Island and (2) obtain specific information on the presence of a species in specific habitats. THE STUDY AREA AND METHODS The study was conducted on Bulls Island, Charleston County, South Carolina, a 2024 ha barrier island in Cape Romain National Wildlife Refuge. The island is approximately 9.6 km long and 3.2 km wide. Salt Present address; U.S. Fish and Wildlife Service, P.O. Box 87, Kilauea, Kauai, Hawaii 96754 Brimleyana No. 5:129-134. July 1981. 129 130 John B. Andre marsh, tidal creeks, and the intracoastal waterway separate the island from the mainland by a distance of about 3 km. Stalter (1974; manuscript) listed the plant species of Bulls Island and described three major and two minor plant communities. Hosier’s (1975) descriptions of the plant communities of Kiawah Island, Charleston County, although more detailed than Stalter’s Bulls Island studies, showed that the two islands are similar in vegetation. I followed Hosier’s classification to delineate the habitats on Bulls Island, recognizing the following: fore and rear sand dune; salt spray forest; maritime live oak forest; freshwater marsh; salt marsh; and old-field. A residential area consisting of one house and a workshop made up a seventh small mam- mal habitat surveyed. The dominant vegetation of each habitat (excluding the residential area) follows: (1) sand dune. — sea oats, Uniola paniculata; croton, Croton punctatus\ sea elder, Iva imbricata; and panic grass, Panicum amarum. (2) salt spray forest. — live oak, Quercus virginiana\ red bay, Persea borbonia; wax myrtle, Myrica cerifera; and yaupon. Ilex vomi- toria. (3) live oak forest. — live oak; loblolly pine, Pinus taeda; Darling- ton oak, Quercus laurifolia\ and magnolia. Magnolia grandiflora. (4) freshwater marsh. — cattail, Typha sp.\ rush, Scirpus sp.; wax myrtle; cabbage palmetto, Sabal palmetto; and willow, Salix caroliniana. (5) salt marsh. — smooth cordgrass, Spartina alterniflora. (6) old-field — broomsedge, Andropogon sp.; and other grasses and herbaceous plants. Small mammals, i.e. those capturable in mouse and rat snap traps, were collected during 17 days of trapping between 18 December 1978 and 10 February 1979. Trapping was conducted from 18 to 20 December, 9 to 12 and 23 to 26 January, and 7 to 10 February. Trapping effort was not equal in all habitats (see Table 1) due to time and material constraints. Six trapping transects, one per habitat (excluding the residential area), were established. Each had 12 trapping sites, placed at 5 m intervals, and each site included one rat and three mouse snap traps situated 4 m apart along a line perpendicular to the transect. Thus, each transect contained 48 traps and was 12 m wide and 55 m long. Transect width was reduced in the freshwater and salt marsh habitat edges to minimize the chance of trapping animals from adjacent habitats. Twenty-eight mouse and eleven rat traps were selectively placed in and around the residential area to determine small mammal occurrence and the results are included in Table 1. Eight rat traps were placed on trees in the maritime live oak forest to investigate the presence of flying squirrels, Glaucomys volans, but none was captured during the study. All traps were Victor brand, baited with a paste of rolled oats and peanut butter that was replenished as needed. Specimens obtained in this study were donated to the Charleston Museum. Barrier Island Small Mammals 131 Table 1. Summary of small mammal trapping on Bulls Island, December 1978 to February 1979. Habitat Number of days trapped Number of trap nights Number and genera of mam- mals captured Relative abundance value (RAV)"' Fore and rear sand dunes 13 560 9 Oryzomys 1.6 Salt spray forest 13 578 17 Oryzomys 2.9 Maritime live oak forest 13 616 3 Oryzomys 0.5 2 Peromyscus 0.3 Freshwater marsh 4 186 5 Oryzomys 2.7 Salt marsh 10 381 36 Oryzomys 9.4 1 Sigmodon 0.2 Old-field 9 432 1 1 Sigmodon 2.5 3 Oryzomys 0.7 Residential area 4 156 4 Peromyscus 2.6 3 Mus 1.9 1 Oryzomys 0.6 All Habitats 17 2909 74 Oryzomys 12 Sigmodon • • • • 6 Peromyscus 3 Mus ''The percent trapping success for each species in each habitat. 132 John B. Andre Differences in abundance of small mammals among habitats were compared by Chi-square test. Equal probability of capture in each habi- tat was assumed. For each species, expected frequency (E) in each habitat was calculated by dividing the number of trap nights executed in habitat A by the total number of trap nights executed in habitats A and B, then multiplying by the number of individuals captured in habitats A and B. RESULTS AND DISCUSSION Trapping results, habitat use, and relative abundance of the four spe- cies of small mammals found on Bulls Island are shown in Table 1. The percent trapping success for each species in each habitat is the relative abundance value (RAV). Comparisons of the RAVs of different species within a habitat, and of the same species among habitats, reflects the degree of a species’ use of a habitat. The rice rat, Oryzomys palustris, was the only species captured in the sand dune, freshwater marsh, and salt spray forest habitats. It was equally abundant in the freshwater marsh (2.9 RAV) and salt spray forest (2.7 RAV) habitats, and less abundant in the sand dunes (1.6 RAV). The differences were not significant(X^=2.22, 1 df, p>0. 10). The freshwater marsh and salt spray forest habitats are structurally complex, with patches of grass and dense, tangled stands of small trees and shrubs. By contrast, the vegetation of the sand dunes was sparse and structurally simple with only scattered, dense stands of sea oats. Oryzomys prefers habitats with a dense cover of grasses or sedges (Golley 1962). On Bulls Island vegetation of this type was found only in small patches in the dunes, providing little habitat for Oryzomys, The maritime live oak forest produced the fewest mammals of all habitats despite the greater number of trap nights. Oryzomys and the cotton mouse, Peromyscus gossypinus, yielded RAVs of 0.5 and 0.3, respectively. The maritime live oak forest of Kiawah Island was also sparsely inhabited by small mammals, presumably because the forest canopy reduces light penetration to the forest floor, reducing productiv- ity and resulting in little food and cover (Pelton 1975). The three Oryzo- mys I captured in the live oak forest were subadults, as judged by body size. Grant (1971) reported that subadult Microtus pennsylvanicus were more likely to inhabit suboptimal habitat than were adults. A more detailed study is needed to determine whether the same is true of Oryzo- mys on Bulls Island. Trapping in the salt marsh produced 36 O. palustris and 1 hispid cotton rat, Sigmodon hispidus. Oryzomys was significantly more -abun- dant there (9.4 RAV) than in any other habitat during winter (X^= 17.48, 1 Barrier Island Small Mammals 133 df, p<0.005). It feeds on grasses, sedges, insects, crabs, fish, and the eggs and young of birds (Golley 1966; Sharp 1967). Although bird eggs and nestlings are not available in winter, the probable abundance of other food items in the salt marsh could explain why Oryzomys is able to exploit this habitat so successfully. Oryzomys was also common in the salt marshes of Kiawah Island (Pelton 1975) and was the most widely distributed mammal in the Virginia barrier island complex (Dueser et al. 1979). Sigmodon and Oryzomys are sympatric in the salt marsh and old-field habitats, indicating some similarity in habitat requirements, i.e. dense grass and weed cover. Oryzomys prefers stream edges, freshwater marshes, and salt marshes, while Sigmodon prefers drier grass fields and thickets (Golley 1966). Although Oryzomys was found in all habitats on Bulls Island it was most abundant in the salt marsh (9.4 RAV), salt spray forest (2.9 RAV), freshwater marsh (2.7 RAV), and sand dunes (1.6 RAV), while Sigmodon was common only in the old-field habitat (2.5 RAV). Significantly more Oryzomys were captured in the salt marsh than Sigmodon (p<0.005), while more Sigmodon were captured in the old- field habitat than Oryzomys (p<0.05), demonstrating that Sigmodon is found mostly in old-fields. Of the three species captured in the residential area, Peromyscus was most abundant (2.6 RAV), followed by the house mouse, Mus musculus (1.9 RAV), and Oryzomys (0.6 RAV). The difference was not significant (X^=1.8, 1 df, p>0.1) between Peromyscus and Oryzomys. Pelton (1975) found that Peromyscus invaded homes on Kiawah Island, as it has on Bulls Island, and also reported that it was the most abundant small mammal of Kiawah Island. On Bulls Island Peromyscus is less abundant than Oryzomys and Sigmodon. It is interesting to note that Mus (an exotic species) was captured only in the residential area of Bulls Island. Pelton (1975) found no Mus in or around the residential areas of Kiawah Island. He reported, however, capturing Mus in abundance in the sand dune habitat. Dueser et al. (1979) captured Mus on four Virginia barrier islands and found that its distribution was not limited to the vicinity of cabins on three of the islands. Representatives of the small mammal fauna of Bulls Island used all habitats sampled. Mus was found only in the residential area. Peromys- cus inhabited the maritime live oak forest and probably colonized the residential area with migrants from that habitat. Oryzomys captured in the residential area probably also represented migrants from surrounding habitats. In old-fields Sigmodon is numerically dominant to other spe- cies. Oryzomys is a habitat generalist on Bulls Island, but it is most abundant in the salt marsh. 134 John B. Andre ACKNOWLEDGMENTS. — Support for field work was provided by the U.S. Fish and Wildlife Service, Cape Romain National Wildlife Refuge. I thank James A. MacMahon for reviewing an earlier draft of the manuscript, and three anonymous reviewers and Raymond D. Dueser for their helpful comments. LITERATURE CITED Deuser, R.D., W.C. Brown, G.S. Hogue, C. McCaffrey, S.A. McCuskey and G.J. Hennessey. 1979. Mammals on the Virginia barrier islands. J. Mammal. (50:425-429. Engels, W.L. 1942. Vertebrate fauna of North Carolina coastal islands. Am. Midi. Nat. 25:273-304. 1952. Vertebrate fauna of North Carolina coastal islands. II: Shackle- ford Banks. Am. Midi. Nat. ^7:701-742. Gibbons, J.W., and J.W. Coker. 1978. Herpetofaunal colonization patterns of Atlantic coast barrier islands. Am. Midi. Nat. 99:219-233. Golley, Frank B. 1962. Mammals of Georgia: A study of their distribution and functional role in the ecosystem. Univ. Georgia Press, Athens. 218 pp. Grant, P.R. 1971. The habitat preference of Micro tus pennsylvanicus, and its relevance to the distribution of this species on islands. J. Mammal. 52:351-361. Hillestad, H.O., J.R. Bozeman, A.S. Johnson, C.W. Berisford and J.I. Richardson. 1975. The ecology of Cumberland Island National Seashore, Camden County, Georgia. Ga. Mar. Sci. Cen. Univ. Ga. Tech. Rep. Ser. No. 75-5, Skidaway Island. 219 pp. + 4 maps. Hosier, Paul E. 1975. Dunes and marsh vegetation, pp. Dl-D96m Environmental inventory of Kiawah Island. (W.D. Chamberlain, project coordinator.) Environ. Res. Cen., Inc., Columbia. Johnson, A.S., H.O. Hillestad, S.F. Shanholtzer and G.F. Shanholtzer. 1974. An ecological survey of the coastal region of Georgia. Nat. Park Serv. Sci. Monogr. Ser. No. 3:1-233. Pelton, Michael R. 1975. The mammals of Kiawah Island, pp. M1-M45 in Environmental inventory of Kiawah Island. (W.D. Chamberlain, project coordinator.) Environ. Res. Cen., Inc., Columbia. Sharp, Homer F., Jr. 1967. Food ecology of the rice rat, Oryzomys palustris (Harlan), in a Georgia salt marsh. J. Mammal. 45:557-563. Stalter, Richard. 1974. A floristic study of South Carolina barrier islands. ASB Bull. 27:86. Abstract. Manuscript. The flora of Bulls Island, Charleston County, South Carolina. Unpubl. Teal, J.M. 1962. Energy flow in the salt marsh ecosystem of Georgia. Ecology 45:614-624. Accepted 23 February 1981 On the Taxonomic Status, Distribution and Subspecies of the Milliped Pseudotremia fracta (Chamberlin) (Chordeumatida: Cleidogonidae) Richard L. Hoffman Department of Biology, Radford University, Radford, Virginia 24142 A BSTRACT. — Examination of the female holotype of Pseudotremia fracta Chamberlin (1951) shows beyond much doubt that this name is senior to P. cottus (Shear (1972). This widely distributed species is herein considered to contain four subspecies, easily distinguished by form of the median process of the gonopod syntelopodite: P.f. fracta in the Great Smokies; P.f. paynei n. subsp. along the Clinch River, Ten- nessee; P.f. ingens n. subsp. from the Cumberland Mountains in Scott County, Tennessee; and P.f. nantahala n. subsp. in the Nantahala River gorge. North Carolina. The proposal by Shear (1972) to unite fracta, cocytus, and scrutorum in a subgeneric group is reaffirmed. The last- named taxon may in time be shown to be subspecifically related to fracta. Because of the normally restricted ranges of most species of Pseudo- tremia, it is a matter of interest when one is found to occur over a relatively large geographic area that embraces a variety of physiographic provinces and a resultant diversity of biotopes. Such an organism is Pseudotremia fracta (Chamberlin 1951), which appears to be not uncommon in central eastern Tennessee in both epigean and subterra- nean habitats. Since the original description was based upon a female specimen, which was subsequently misplaced and unavailable for study, the status of this nominal species could not be addressed with confidence by the recent monographer of the genus (Shear 1972). With evident perspicuity, however. Dr. Shear suspected (1972:158) that the name fracta was based on an immature specimen of the species that was named Pseudotremia cottus in his revision. During the course of a general renovation of the Chamberlin collection (now on deposit at the Smithsonian Institution), the type oi fracta — among many other misplaced specimens — was recovered but was not closely examined until the recent receipt from Dr. Thomas C. Barr, Jr. of unidentified material showing some degree of affinity with Shear’s P. cottus. With the motivation of having to provide a name for this popula- tion, I essayed an investigation de novo, the results of which are set forth in the following pages. Brimleyana No. 5:135-144. July 1981. 135 136 Richard L. Hoffman As might be expected of a milliped that occurs in leaf litter in the spruce-fir forest of the Great Smoky Mountains as well as in limestone caves along the Clinch River 50 miles distant and 4000 feet lower, there is evident geographic variation in structure of the gonopods. This circum- stance was perceived and accounted by Shear (1972:184, Figs. 103, 104), although he did not go so far as to nomenclatorially formalize subspecific differentiation. Since the already manifested differences between the populations of the Tennessee Valley and the Great Smokies have now been further substantiated by the discovery of two new forms of fracta in the Cumber- latid Mountains and in the Nantahala Gorge, I think it desirable to introduce trinomial designations for these four isolates, and to confirm Dr. Shear’s suspicion that cottus might be a junior synonym of fracta. In connection with the citation of material in the following accounts, the standard acronymic abbreviations are used to indicate the location of specimens, thus: MC2 — Museum of Comparative Zoology NCSM — North Carolina State Museum RLH — Personal collection of the author USNM — National Museum of Natural History Pseudotremia Cope 1869:179. Shear 1971:162-193 (monograph). The general structure of the gonopods in Pseudotremia has been adequately described and figured by Shear. I wish here only to express an opinion about one anatomical feature and to suggest a slight refinement in nomenclature. As shown in his Figure 1 (of Pseudotremia hobbsi), the dominant elements of the gonopods (8th pair of legs) are the apparent coxal deriva- tives for which the term colpocoxite (Ribaut 1920) seems applicable. Immediately posterior to the pair of colpocoxites, and attached to them only by connective tissue (no direct musculature has been demonstrated) is a sclerotized, almost medially diastemmate structure (the “bifid lami- nae” of earlier workers) that Shear considers to represent the fused telop- odites of the gonopods. If this interpretation is correct, and I have no better alternative to suggest, perhaps the term “syntelopodite” might be employed in this genus (in the related taxon Cleidogona, the two struc- tures remain discrete or minimally attached at their bases). When the appendages of the genitalic complex are held in a resting position, the syntelopodite is tightly embraced basally by large coxal lobes of the posterior gonopods (9th pair of legs). The anterior (ventral) surface of the syntelopodite is generally modified in Pseudotremia by the development of a central basal structure, referred Milliped Taxonomy and Distribution 137 to by Shear as the “telopodite process” and by him quite properly exploit- ed as a taxonomic character of considerable importance. A wide spec- trum of form is evident, from a low median convexity to enormously elongated and apically bifid blades projecting conspicuously between the colpocoxites. In a small group of apparently related species native to eastern Tennes- see, the syntelopodite process is somewhat enlarged distally, and apically produced into several projections of variable form and length. These species are P. cocytus Shear, P. cottus Shear, P. scrutorum Shear, and P. minos Shear, and the ensemble was referred to by Shear as the Cottus Group. I think he was quite correct in considering it “. . . one of the more coherent species assemblages in the genus Pseudotremia.'' With the pres- ent reinstatement of P.fracta as a senior synonym of cottus, it will be appropriate to likewise alter the group designation. Both geographically and anatomically, P. minos is somewhat disjunct from the others (known from northern Alabama, troglobitic rather than trogophilic, and with the syntelopodite process four- instead of three- pronged) and perhaps warrants assignment to a group of its own. Of the remainder, cocytus occurs in Anderson County, Tennessee, apparently sympatric with fracta but not yet collected at the same locality with it; scrutorum is known only from the type locality in Scott County, Tennes- see; and fracta occurs widely from the crest of the Great Smokies to caves and bluffs along the Clinch River. Pseudotremia fracta Chamberlin The original description of this taxon was based upon a single female from Gatlinburg, Tennessee, and mentioned in four brief sentences color- ation, size, number of ocelli, and form of the metatergal surface sculp- ture. Although the specimen was subsequently misplaced in the collection of the describer, and despite the brevity of the account, it was possible for Dr. Shear to suspect that the name fracta applied to an immature indi- vidual of the species that he (1972:1) named Pseudotremia cottus. Subse- quent to the demise of Dr. Chamberlin in 1968 his myriapod collection was transferred to the National Museum of Natural History and in 1977 the diplopod material was placed in my hands for reorganization. Event- ually the type oi fracta was discovered and Shear’s prophesy fulfilled; it is indeed a female lacking one moult of maturity yet externally identical with specimens identified by Shear as P. cottus (which is, moreover, the only member of the genus so far known from the western slopes of the Great Smokies). The present decision to recognize taxonomically nameworthy popula- 138 Richard L. Hoffman tions within the original concept of cottus engenders the question of their relative status, both to each other and to the closely allied P. scrutorum. Although no decisive answer can be advanced at the present, the point nonetheless merits consideration if for no other reason than the edifica- tion of possible future students of the group. On the basis of differences in dorsal ornamentation (30-40 small, well- defined metatergal tubercles in scrutorum, 4-16 longitudinal rugosities in fracta) and in gonopod structure (median branch of colpocoxite with an acute subapical projection in scrutorum, lacking in fracta), it is probable that these two taxa can be distinguished as two distinct species. Differ- ences in gonopod structure (e.g., lateral colpocoxite branches not caudo- dorsally deflexed in cocytus, prominently decurved in fracta) as well as the apparent sympatry of these taxa, suggest that they, too, differ at the level of species. On the other hand, the general external similarity of the several popu- lations to be recognized within fracta, overall concordance in gonopod structure, and apparent allopatry, suggest that, for the present, subspe- cific status best expresses the degree of relationship. This estimation requires future confirmation by field work in the region south and east of Knoxville to establish whether the present distributional hiatus is real or illusory. If the former be true, a good case might be made for full specific status on the grounds of effectual geographic reproductive isolation of the components. The distributional pattern shown by fracta (i.e., northwest to southeast trend across the Tennessee Valley) is paralleled in a number of other groups. Amongst trechine carabids, which seem to have ecological con- straints like those of pseudotremias, Barr (1969, 1979) demonstrated sim- ilar disjunctions in the case of Trechus tennesseensis Barr, the nominate subspecies of which occurs in a cave in Roane County, Tennessee, and T. t. tauricus in a cave near the Great Smokies in Blount County. In the Trechus schwarzi group, T. schwarzi (several subspecies) is restricted to the southern Blue Ridge, and the closely related vicar, T. cumberlandus, to the Cumberland Plateau in Tennessee and Kentucky. Barr postulated that such patterns developed during the Wisconsin glacial period, when cool or subarctic environments must have been prevalent in the southern Appalachian region generally. Pseudotremia fracta fracta Chamberlin, new status Figs. 1, 5 Pseudotremia fracta Chamberlin 1951:25. Chamerlin & Hoffman 1958:94. Inmature female holotype (USNM) labeled “Gatlinburg Tenn / Cove hardwoods / B/ 6-24-47”; a second label has the notation Milliped Taxonomy and Distribution 139 “SM-145” No collector is given, but the original description states that material was taken by Hugh Hanson, Arizona State College. Pseudotremia cottus Shear (in part) 1972:183, figs. 101-109. Male holo- type (MCZ) from “The Sinks” of Little River, on Tenn. Hy. 73 at the Blount-Sevier County Line, Tennessee; Leslie Hubricht leg. 25 May 1962. New Synonymy! Syntelopodite process of gonopods massive (Fig. 5), lateral apices widely separated, distally curved mesad, extending forward between medial and lateral branches of colpocoxite; median apex relatively long and acute. Lateral branch of coxal region notably broad beyond point of flexure, its outer subterminal divison (x) much larger than inner (y) (Fig. 1). Distal extremities of syntelopodite larger, and curve further anteriad, than in the other three races of fracta. The original description of fracta stated the type locality to be “Gatlin- burg Cove”, obviously a misinterpretation of the collector’s label by Chamberlin, who was probably unaware of the term “cove hardwoods” and so transposed some of the words. The specimen apparently came from the immediate vicinity of Gatlinburg itself. Furthermore, by some inexplicable error during transcribing of original collector’s field data, the name “Cades Cove” was added to the permanent label with the holo- type of P. cottus and was included in citation of the type locality of this species. Actually The Sinks is about 9 miles northeast of Cades Cove. Although P.f. fracta is so far unknown from North Carolina it cer- tainly occurs on the southern slopes of the Great Smokies, since the specimen that I found at Indian Gap was taken less than 100 meters from the state line on the Tennessee side. Aside from the type localities of the two names applied to this race, I have examined the following material (all at present in my collection): TENNESSEE: Blount County. — Big Poplar Trail, Cades Cove, 23 May 1962, L. Hubricht; near Gregory’s Cave, Cades Cove, 25 May 1962, Hubricht. Sevier County. — west side of Indian Gap, 4 August 1958, R.L. Hoffman; between Gatlinburg and Newfound Gap, 25 May 1962, Hubricht. Pseudotremia fracta paynei, new subspecies Figs. 2, 6 Holotype (5, paratype $ and 2 (RLH) from Norris Quarry Cave No. 1, 1 mi. N of Andersonville, Anderson Co., Tennessee; Jerry A. Payne leg. 29 May 1965. 2$ paratypes (NCSM A2599) From Foster’s Cave No. 1, 2 140 Richard L. Hoffman Figs. 1-4. Left colpocoxites, anterior aspect, of four subspecies of Pseudotremia fracta: 1. P.f. fracta Chamberlin, specimen from Cade’s Cove, Tennessee. 2. P. /. paynei, n. subsp., from holotype. 3. P.f. ingens, n. subsp., from holotype. 4. P. f. nantahala, n. subsp., from holotype. All drawings made to same scale. Abbreviations: X, outer division or process of lateral colpocoxite branch; Y, inner division of same. Milliped Taxonomy and Distribution 141 mi. N of Clinton, Anderson Co., Tennessee; J. A. Payne leg. 18 July 1965. Lateral projections of syntelopodite process nearly straight, only a little longer than the median (Fig. 2); all three extend cephalad between median branches of colpocoxites, latter similar to those of fracta but the outer branch distinctly more slender distally and drawn out into a long narrow blade. (Fig. 6). Body size slightly less than in fracta and ingens, length of adult males about 25 mm. Dorsal surface of metaterga nearly smooth, dorsolateral knobs of segments 3-5 not tilted upward. Ocelli of holotype 19 in four rows. Coloration grayish-purple, with lighter areolation as in the other subspecies. As shown by open circles on the map (Fig. 9), this milliped appears to be confined to caves and epigean sites along the Clinch River, west and northwest of Knoxville. Material has been examined from the following localities (all collections from Anderson County by Jerry A. Payne, for whom the subspecies is named; all material at present in my personal collection): TENNESSEE: Anderson County. — Moore’s Bridge Cave, 2 mi. N of Clinton, 22 July 1965; Melton Hill Cave No. 1, 5 mi. S of Oak Ridge, 19 May 1965; Bee Hole Cave, 4 mi. N of Clinton, 30 May 1965; Marie’s Cave, 1 mi. N of Andersonville, 29 May 1965; Wright’s Cave, ca. 2 mi. N of Clinton, 31 May 1965 and 2 August 1965; Wallace’s Cave, ca. 5 mi. N of Clinton, 4 April 1965; Markli’s Cave, 4 mi. S of Clinton, 31 May 1965. Knox County. — Rock Hill Cave, 2 mi. S of Heiskell, 12 June 1965, E. F. Menhinick. Roane County. — Eblen’s Cave near Kingston, 1 August 1955, S. I. Auerbach; Obed River bluff opposite Harriman, 21 May 1961, Leslie Hubricht; wooded hillside 1.3 mi. S of Pine Grove, 30 May 1963, Hubricht. Pseudotremia fracta ingens, new subspecies Figs. 3, 7 Holotype $ and paratype $ (RLH) from New Mammoth Cave, ca. 1 1 mi. NW of LaFolette, Campbell County, Tennessee; T. C. Barr leg. 16 September 1979. 2$ and 1$ paratypes (NCSM A3646) from same local- ity, T. C. Barr and J. R. Holsinger leg. 21 November 1979. Lateral arms of syntelopodite process long and apically divergent, median projection rudimentary (Fig. 7); outer branch (solenomerite) of colpocoxite more slender than in fracta and paynei, its outer subterminal process (x) much smaller than inner (y); undivided basal region of coxa smaller than in the other three races owing to a deeper diastemma 142 Richard L. Hoffman between mesal and ectal branches (Fig. 3). Holotype about 28 mm long (broken); dorsally grayish-purple, meta- terga with large pale oval spot on each side; prozona and legs lighter, nearly white; front of head light brown with four pairs of irregularly- shaped aerolated spots. Ocelli pigmented, 20 in five rows. Metazona of anterior segments with prominent lateral knobs formed by two enlarged ridges, the dorsolateral of which has a seta at front end and is subtended by a larger oblique ridge with seta at posterior end; on segments 3-5 the lateral knobs project upward slightly above level of middorsum, on following segments knobs become progressively smaller and are not present on segments 25-29. Ornamentation of metaterga greatly reduced, middorsal region smooth, a few indistinct elongated tubercles occur near base of dorsolateral ridge. Figs. 5-8. Syntelopodite process of four subspecies of Pseudotremia fracta, ventral aspect, with lateral lobes of syntelopodite also shown: 5. P.f. fracta Chamberlin, specimen from Cade’s Cove, Tennessee. 6. P.f. paynei, n. subsp., from holotype. 7. P.f. ingens, n. subsp., from holotype. 8. P.f. nantahala, n. subsp., from holotype. All drawings made to same scale. Milliped Taxonomy and Distribution 143 This subspecies is known so far only from the type locality, to which the name alludes. Collections from the region between this cave and the type locality of P. scrutorum (northern Scott County, Tennessee) would be of considerable interest in providing information on the taxonomic status of these two taxa vis-a-vis each other. Fig. 9. Eastern Tennessee and extreme western North Carolina, showing known localities for several taxa of Pseudotremia: inverted triangle, P. scrutorum Shear; dot with star, P.f. ingens, n. subsp.; open circles, P.f. paynei, n. subsp.; solid dots, P.f. fracta Chamberlin; circle with dot, P.f. nantahala, n. subsp. Eastern and western boundaries of the Ridge and Valley Physiographic Province are indicated by the two broken lines. Pseudotremia fracta nantahala, new subspecies Figs. 4, 8 Holotype $ (NCSM A2528) from Blowing Springs Cave near Nantahala, Swain Co., North Carolina; S. Platania, P. Hertl, and C. O. Holler leg. 18 March 1979. Mesal and ectal branches of colpocoxites separated by a distinctly U-shaped diastemma, subterminal projection of ectal branch relatively small (Fig. 4); syntelopodite process with stout pedicel, lateral projections 144 Richard L. Hoffman straight, subparallel, extending between branches of colpocoxites, median projection short, broad, apically truncate (Fig. 8). Holotype about 27 mm long (broken), coloration as described for P.f. ingens, ocelli pigmented, 21 in five rows. Dorsolateral knobs of anterior segments less prominent than in the other three races, and not elevated. Metaterga of most segments with two low but distinct convexities between median suture and dorsolaterial ridges, surface of these areas distinctly coriaceous in contrast to the nearly smooth metaterga of other subspecies. Aside from the holotype, I have examined a female (NCSM A2598) from Cliff Ridge near Blowing Springs, 3.3 mi. NE of Nantahala, Swain County, North Carolina; L. Hubricht leg. 16 May 1961. This specimen had been identified by Dr. Shear as P. cottus but was not cited in his monograph (possibly because of the usual uncertainty in the identifica- tion of single females). It is probably from the immediate vicinity of the type locality. ACKNOWLEDGMENTS. — My sincerest thanks are expressed to Dr. Thomas C. Barr, Jr., Dr. Jerry A. Payne, and Mr. Leslie Hubricht for presenting me with most of the material upon which this paper is based. Dr. Rowland M. Shelley kindly loaned the specimens of P. /. nantahala and allowed me to describe it. Dr. Shelley and Dr. J. E. Cooper performed much-appreciated service in removing numerous rough places from the manuscript. LITERATURE CITED Barr, Thomas C., Jr. 1969. Evolution of the Carabidae (Coleoptera) in the southern Appalachians, pp. 67-92 m P. C. Holt (ed.). The distributional history lof the biota of the southern Appalachians, Part 1: Invertebrates. Res. Div. Monogr, 1, Va. Polytech. Inst., Blacksburg. 295 pp. . 1979. Revision of Appalachian Trechus (ColeopteraiCarabidae). Brimleyana 2:29-75. Chamberlin, Ralph V. 1951. On eight new southern millipeds. Great Basin Nat. 77:19-26. ^ and Richard L. Hoffman. 1958. Checklist of the millipeds of North America. Bull. U. S. Natl. Mus 272:1-236. Cope, Edward D. 1869. Synopsis of the extinct Mammalia of the cave formations in the United States, with observations on some Myriapoda found in and near the same, and on some extinct mammals of the caves of Anguilla, W. I., and other localities. Proc. Amer. Philos. Soc. 77:171-192. Ribaut, Henri. 1920. Notes sur les Chordeumoides de France. Bull. Soc. Hist, nat. Toulouse 45:18-34. Shear, William A. 1972. Studies in the milliped order Chordeumida (Diplo- poda): A revision of the family Cleidogonidae and a reclassification of the order Chordeumida in the New World. Bull. Mus. Comp. Zool. 744:151-352. Accepted 16 March 1981 Reptiles and Amphibians of Kiawah and Capers Islands, South Carolina J. Whitfield Gibbons Savannah River Ecology Laboratory, Drawer E, Aiken, South Carolina 29801 AND Julian R. Harrison, III Department of Biology, College of Charleston, Charleston, South Carolina 29401 ABSTRACT. — We obtained information on abundance, species com- position, distribution among habitats, and general ecology of the rep- tiles and amphibians of two South Carolina barrier islands, Kiawah and Capers. The herpetofauna of the islands is greatly reduced compared with that of the adjacent mainland. Kiawah Island has 16% of the mainland’s amphibian species and 43% of the reptile species; Capers Island has only 10% of the amphibian species and 21% of the reptiles. All herpetofaunal species found on Capers occur on Kiawah. The rela- tionship between island size and species numbers on each island is sim- ilar to that reported for other Atlantic Coast barrier islands. Recom- mendations are given for minimizing the environmental impacts on reptiles and amphibians in future recreational development of these or any barrier islands. INTRODUCTION Many coastal islands of the Carolinas and Georgia are being developed for recreational and other purposes. The potential environmental impact of such endeavors is great and has generated concern for how these activities affect certain elements of the fauna and flora. The objective of our studies was to obtain information on abundance, distribution, species composition, and general ecology of the reptiles and amphibians of two barrier islands along the South Carolina coast. Our purpose was to reveal the environmental alterations that must be considered if the impact of development on the herpetofaunal populations is to be minimized. Field studies were conducted on Kiawah and Capers islands during spring and summer of 1978 and 1979. These studies complemented earlier herpetological research on Capers Island during summer and fall of 1975 and on Kiawah Island from spring of 1974 through fall of 1975 (Gibbons and Coker 1978). Barrier Islands Barrier islands occur along 10 to 13 percent of the coastlines of the world (Cromwell 1971). In the United States they margin the entire Atlantic coast from Florida to Canada, and the northern Gulf of Mexico. Brimleyana No. 5:145-162. July 1981. 145 146 J. Whitfield Gibbons and Julian R. Harrison III They are generally of low elevation and parallel the mainland, from which they are separated by salt marsh or, in some instances, by open water. Sand or gravel is the predominant soil texture. Most geologists consider the Atlantic barrier islands of the United States to be Holocene in origin, probably formed within the last 6,000 years. Several theories have been proposed to explain the origin of barrier islands. Among these, the emergence of offshore bars (Leont’yev and Nikiforov 1965; Otvos 1970), the submergence of coastal ridges (Hoyt 1967, 1968), and the isolation of sandspits (Fisher 1968), have been the most prominent. The most appealing theory is that of Schwartz (1971) who compromised by suggesting that a combination and interaction of all three processes was the most suitable overall explanation. However, he indicated that one mode of development or another might be more prevalent in a particular situation or locality. On the southeastern Atlantic coast, those barrier islands with ocean- facing beaches and large enough to support terrestrial vertebrates have certain characteristics in common. Among these are a sand dune complex beginning above the high tide mark. The dune system is frequently mar- gined by a hardwood thicket of shrubs and trees stunted by continual salt spray. Plant communities of the island interiors are generally a combina- tion of hardwoods, palmettos and pines although the composition of a particular forest is the consequence of soil, climate, and recent historical events. One apparent characteristic of the southern barrier islands is that most woody species are evergreen. Fresh water may be abundant on some islands, absent on others. Even the largest freshwater lakes, however, have the potential of completely drying up during long periods of drought. Extensive salt marshes frequently make up the landward borders of the islands, which are then further separated from the main- land by brackish or salt water. This may take the form of tidal creeks or, in some cases, a major sound. A common threat to the ecological integrity of all barrier islands of the United States is the impact of development and other human activities (Dolan et al. 1973). Nonetheless, thorough studies of the flora and fauna of these unique habitats have lagged far behind those of other natural environments. The limited knowledge available about the reptiles and amphibians of the barrier islands of the Gulf and Atlantic coasts was compiled by Gibbons and Coker (1978). Despite the great abundance and diversity of reptiles and amphibians in the Southeast (Conant 1975; Mar- tof et al. 1980), a clear need exists for documentation of the ecology of these animals in the fast-disappearing natural habitats of barrier islands. The Study Areas The islands are located 19 km south (Kiawah) and 22 km north (Cap- ers) of Charleston (80‘^02'W, 32°54'N), Charleston County, South Barrier Island Herpetofauna 147 Carolina. Kiawah Island consists of approximately 3200 ha, half of which « is an extensive salt marsh. Major habitats besides the salt marsh include a 12 km ocean-facing beach and dune system, and maritime thickets and forests that form most of the island’s interior. The maritime thickets are characterized by closely spaced vegetation, predominantly stunted live oaks, Quercus virginiana; yaupon holly. Ilex vomitoria\ and wax myrtle, Myrica cerifera. The interior forests are combinations of pines, Pinus sp.; oaks, Quercus sp.; magnolia. Magnolia grandiflora; and palmettos, Sabal sp.. About a dozen brackish and freshwater lakes that are not under tidal influence occur on Kiawah. They range in salinity from 0 ppt to about 17 ppt depending upon their location on the island and upon recent rainfall. These lakes total approximately 80 ha and in most cases represent impounded salt marshes. In addition, many low-lying areas become flooded during wet parts of the year, creating small, shallow freshwater habitats throughout much of the forested parts of the island. Of major influence on Kiawah Island is the recreational development complex and resort located at the southwest end. This system of cottages, golf courses, and blacktop roads composed more than 10% of the island by 1979. Capers Island consists of about 900 ha, of which approximately 47% is salt marsh. Beach frontage on Capers is 5.3 km, much of which is com- posed of a short, ocean-facing beach with standing dead trees exposed only at low tide. Erosion of the beach and marsh margins has increased in recent years and has altered the size and configuration of impoundment areas. The island vegetation is generally similar to that of Kiawah Island, although the maritime thicket community is greatly reduced in extent. Only one small, freshwater pond, Greene Pond, is believed to have con- tained water continually throughout the study period. This pond, located at the south end of the island, is less than 0. 1 ha in area. An extensive brackish-water impoundment system is also present on the south end of the island. One arm of this impoundment, a large, isolated freshwater pond in earlier years, has been breached by surf erosion. Numerous ditches and low-lying areas throughout the island contain water during early spring and midsummer wet periods. The climate of both Kiawah and Capers Islands is classed as subtropi- cal. Temperatures fall below freezing on fewer than 20 days during the winter (Kjerfve 1974). Mean summer (July) temperatures are 21.5° C min- imum and 31.5° C maximum. Mean winter (January) temperatures are 2.8° C minimum and 15.5° C maximum. Both islands receive approxi- mately 125 cm of rain a year. METHODS Field collecting techniques were diversified, since our objective was to 148 J. Whitfield Gibbons and Julian R. Harrison III document the presence and abundance of as many species of reptiles and amphibians as possible. Baited hoop nets, drift fences with pitfall traps (Gibbons and Bennett 1974), seining, road collecting, muddling in aqua- tic areas, and general collecting were the most extensively used methods. Visual counts were made of alligators. Alligator mississippiensis, on the various lakes during both day and night. Records were kept of road- killed amphibians and reptiles on Kiawah, and of those reported or brought in by residents or workers on either island. Quantitative data were collected in four ways: 1) 50-foot drift fences, with perpendicular fencing at each end, and pitfall traps; 2) visual counts of alligators; 3) equal-effort visual censuses along transects, corresponding to vegeta- tional analysis transects used by Gaddy (in preparation); each transect was walked by a single investigator who noted all reptiles or amphibians seen or heard during a one hour period; and 4) mark-release-recapture sampling of turtles in Greene Pond. Each of these methods and their results will be discussed separately below. RESULTS AND DISCUSSION A total of 31 species of reptiles (24) and amphibians (7) are known to occur on Kiawah and Capers islands (Tables 1 and 2). These represent 20 Table 1. Cumulative summary of reptiles and amphibians from Kiawah and Capers Islands compared to mainland occurrences; excludes Caretta and Malaclemys. Species names are given in Table 2. Mainland Only Kiawah Capers Reptiles Families 1 9 8 Genera 17 20 11 Species 32 24 11 Amphibians Families 5 6 3 Genera 13 6 3 Species 36 7 4 All Herps Families 6 15 11 Genera 30 26 14 Species 68 31 15 Barrier Island Herpetofauna 149 Table 2. Reptiles and amphibians documented from Kiawah and Capers Island, South Carolina. Species names used are those given by Conant (1975). Species Both Kiawah Islands Only Amphibia Anura Bufonidae Bufo terrestris. Southern Toad X Hylidae Hyla cinerea. Green Tree Frog X Hyla squirella. Squirrel Tree Frog X Microhylidae Gastrophryne carolinensis. Narrow-mouthed Toad X Pelobatidae Scaphiopus holbrooki, Spadefoot Toad X Ranidae Rana utricularia (= pipiens or sphenocephala). Leopard Frog X Caudata Plethodontidae Plethodon glutinosus , Slimy Salamander X Reptilia Chelonia Emydidae Malaclemys terrapin. Diamond-back Terrapin Pseudemys scripta. Yellow-bellied Turtle X Terrapene Carolina, Box Turtle X Kinosternidae Kinosternon subrubrum. Mud Turtle X Cheloniidae Caretta caretta. Loggerhead Sea Turtle * Lacertilia Anguidae Ophisaurus ventralis. Common Glass Lizard X Iguanidae Anolis carolinensis, Carolina Anole X Scincidae Scincella laterale. Ground Skink X Eumeces laticeps. Broad-headed Skink ** Eumeces inexpectatus. Southeastern Five-lined Skink X Teiidae Cnemidophorus sexlineatus. Six-lined Racerunner X 150 J. Whitfield Gibbons and Julian R. Harrison III Species Both Kiawah Islands Only Serpentes Colubridae Natrix fasciata. Banded Water Snake X Thamnophis sauritus. Ribbon Snake X Thamnophis sirtalis. Garter Snake X Coluber constrictor. Black Racer X Masticophis flagellum, Coachwhip X Lampropeltis getulus. Chain Kingsnake X Opheodrys aestivus. Rough Green Snake X Elaphe obsoleta. Rat Snake, Chicken Snake X Elaphe guttata. Corn Snake X Tantilla coronata. Crowned Snake X Cemophora coccinea. Scarlet Snake X Viperidae Agkistrodon contortrix. Copperhead X Agkistrodon piscivorus, Cottonmouth X Crotalus horridus, Canebrake Rattlesnake X Crocodilia Crocodilidae Alligator mississippiensis , American Alligator X *Marine or estuarine species known or suspected to nest on both islands. **Eumeces laticeps occurs on Kiawah; some of the skinks observed but not captured on Capers may have been this species. genera and 9 families of reptiles and 6 genera and 6 families of amphibi- ans (Table 2). With the possible exception of Chelydra serpentina, all species found on Capers (11 reptiles, 4 amphibians) have also been reported on Kiawah. (A specimen of C. serpentina was reported on the south end of Capers Island by David Chamberlain and L. L. Gaddy, but has not been seen by the authors.) Barrier islands of the Atlantic (Lewis 1946; Engels 1942, 1952; Quay 1959; Martof 1963; Parnell and Adams 1970; Lee 1972; Johnson et al. 1974; Hillestad et al. 1975) and Gulf (Jackson and Jackson 1970; Blaney 1971) coasts generally have a greatly reduced herpetofauna compared with that of the adjacent mainland (Gibbons and Coker 1978). Kiawah has 16% and Capers 10% of the mainland’s amphibian species. Mainland reptiles are better represented than amphibians on both islands, with Kiawah having 43% and Capers 21%. The abundance of reptiles and amphibians on the islands is in accord with what would be predicted from the amount of woodland area available for habitation by these groups of Barrier Island Herpetofauna 151 animals (Gibbons and Coker 1978). For example, the data points show- ing the relationship between woodland area and number of species on the two islands fall along the same regression lines generated by similar data from eight other islands (Fig. 1). Fig. 1. Relationship between woodland area (based on standard topographic maps) habitable by reptiles and amphibians and actual number of species present. Linear regression equations based on nine Atlantic Coast barrier islands, including Kiawah, are highly significant as reported in Gibbons and Coker (1978). Capers Island data points fall near predicted regression lines. 152 J. Whitfield Gibbons and Julian R. Harrison III Because of its smaller land area, Capers Island would be expected to have fewer species of reptiles and amphibians than Kiawah (MacArthur and Wilson 1967). In addition, Capers has significantly more open water and salt marsh (~2 km; approximately twice the distance for Kiawah) between the island and the mainland. Greater distances from the main- land tend to result in lower numbers of species for oceanic islands (MacArthur and Wilson 1967); however, the factor of distance from the mainland has not yet been evaluated in assessing faunal composition of barrier islands. No endemic species of reptile or amphibian has been reported for any North American barrier island (Gibbons and Coker 1978). The variety of methods used to collect reptiles and amphibians necessi- tates separate consideration of results obtained. The most meaningful approach is to compare results on the two islands by each method. Drift Fences Collections from the four drift fence and pitfall trap areas on the two islands revealed certain characteristics about the herpetofauna, primarily in regard to the occurrence of anurans (Table 3). Speculations about relative abundance of the species should be conservative at this time because of the high variability in trapping success between dates and locations. Gastrophryne carolinensis appears to be a consistent forest floor inhabitant on both islands and is active throughout the summer. Rana utricularia also ventures into the terrestrial environments on occasion, as seen by the sporadic captures, primarily of subadults. Scaphiopus hol- brooki juveniles were the most abundant anurans in the pitfall traps in every month sampled. The absence of S. holbrooki and Bufo terrestris from the Capers Island traps strongly supports our conviction that these species occur only on Kiawah. The absence of Hyla in the traps, although two species occur on both islands, is a consequence of sampling bias; pitfall traps are not effective in capturing tree frogs (Gibbons and Bennett 1974). The Capers Island pitfall traps caught the only Kinosternon sub- rubrum confirmed from this island. Transects Three transects established as part of an interdisciplinary study which included the two islands were used in the herpetofaunal studies. Walking trips were taken along each transect, and all reptiles and amphibians seen or heard were noted (Table 4). No new species were recorded as a result of this method, but a comparison could be made of the animals between islands and among locations on an island. The lowest numbers of reptiles and amphibians on Kiawah were Table 3. Drift fence captures of anurans on Kiawah and Capers Islands during summer 1979 (June 23-September 30). Transect numbers refer to those used by L. L, Gaddy (in prep.) in vegetational analyses on Capers Islands. Y refers to fence at Yellowlegs Marsh on Kiawah Island. Number in parentheses indicates number of trap checks during month. Other numbers indicate individuals of species caught on consecutive trap checks during month indicated. Barrier Island Herpetofauna 153 Dd W PQ U H CU u c/3 H c/3 D O D < m o o o O 00 o 00 - P oo es o ^ -C o 3 cd U D c o W u-i \D fN >0 O o o o VO r- m 4; V) o Z P VO oo IT) H U P c/3 Z < Od 73 73 73 ^ ^ - o ^ ^ - ^ ^ ^ o ^ H H H H H H U-l ccj c/3 l-> -C 03 c/3 Ih J3 cd c/3 Ui < cd (U Cu cd (U O, cd (U O, p cd cd cd c/3 U U U x: cd cd 03 P Oh 03 >3 03 c c o *3 <3 -2 3 3 Qd « o o "o 03 03 V i.. 03 Capers None caught Fence Closed 154 J. Whitfield Gibbons and Julian R. Harrison III Table 4. Numbers of reptiles and amphibians seen or heard during timed trips along established transects (see text). TRANSECT TOTAL TOTAL EXCLUDING TURTLES AND ALLIGATORS Species Individuals Species Individuals No. No./ Hr. No. No./ Hr. KIAWAH 1 (Develop- ed Area) 10 16 2.3 7 8 1.2 2 10 75 11.5 10 75 11.5 3 13 >200 >25 12 >200 >25 CAPERS 1 6 123 22.4 4 19 3.5 2 7 23 6.6 7 23 6.6 3 8 >50 >10 7 >50 >10 observed in the developed section (Table 4). Although the number of species noted was not appreciably different from that along the other transects, the reduction in relative abundance of individuals was signifi- cant. Ten to twenty times as many individuals of terrestrial species (aquat- ic turtles and alligators excluded) were seen in an equivalent collecting time in the forested parts of the island compared with the developed area. The contrast among different areas on Capers Island is less than that on Kiawah. However, fewer species and individuals were present along the two forested transects than on the transect that crossed the major dike complex on the south end of Caper’s Island. The diversity of habitats crossed by this transect presumably accounts for the increase in numbers of species and individuals. The forested areas of Kiawah Island clearly have a higher species diver- sity and relative density of individuals on the basis of these comparisons than do the woodlands of Capers Island. However, even the poorest transect on Capers Island yielded higher numbers of animals observed per hour than did the developed part of Kiawah. Alligator Counts Daytime and nighttime visual counts were made of alligators on both islands (Table 5). Alligators occur in high densities in most freshwater and some brackish habitats on both Kiawah and Capers islands. Except for the presumed absence of large alligators in Sparrow Pond, now part Barrier Island Herpetofauna 155 Table 5. Alligator counts on night observations. Kiawah and Capers Islands; combined day and Aquatic Habitat KIAWAH ISLAND Minimum Number Known Present 1974-75 1979 Sparrow Pond 2 adults 2 juveniles Canals in developed areas no canals present 2 Bufflehead Pond 15 6 Pintail Pond 21 0 Egret Pond 2 - Blue Heron Pond 14 juveniles 1 Canvasback Pond - 15 Miscellaneous, unnamed ~20 ~10 aquatic habitats TOTAL ~64* ~36* Greene Pond CAPERS ISLAND 1 adult. Caper’s Lake 6 juveniles 15 adults South Lagoon 0 Salt Marsh 1 TOTAL (minimal estimate) 23 ♦The reduced number observed in 1979 should not be construed as indicative of overall reduction in number of alligators, since 1974-75 counts were based on maximum observations accumulated over 2 years of the Inn-Hotel complex on Kiawah Island, no difference in abundance was apparent between observations in 1975 and 1979. Bara (1976) reported sighting 12 alligators on Capers Island during a nighttime count in August, 1975. Freshwater Turtle Studies A unique situation permitted a mark-release-recapture study of the Pseudemys scripta population occurring in Greene Pond at the south end of Capers Island. Because the freshwater habitat is restricted, the turtles 156 J. Whitfield Gibbons and Julian R. Harrison III could be collected by hand, marked, and measured on subsequent trips. This permitted an estimate of the population size and was an indication of the demography of P. scripta on Capers Island (Gibbons et al. 1979). Captures of marked P. scripta during six different months from May 1978 to June 1980 resulted in fairly consistent Lincoln index estimates of 54 to 73 (x=60.7, S.E.=3.05) adult animals. No juveniles were captured. Mark-rlease-recapture studies were also conducted on Kiawah Island in 1974-75 (Gibbons and Coker 1978) and in 1978-79. However, few recaptures were made due to the difficulty of collecting large samples. This was partly because the many freshwater habitats gave greater oppor- tunity for dispersal on Kiawah. Pseudemys scripta from the two islands are characteristically large, show rapid juvenile growth rates, and have populations primarily made up of adults (Gibbons and Coker 1978; Gibbons et al. 1979). Turtle populations on the barrier islands warrant further investigation in an attempt to understand the ecological and evolutionary significance of their uniqueness. Important Species Although several species of reptiles and amphibians encountered on Kiawah or Capers Islands were common, certain ones emerged as the most abundant and apparent forms during the periods of study. Alligator. — The American Alligator is unquestionably the dominant reptile on both Kiawah and Capers Islands. Both adult and young indi- viduals were observed at one time or another in almost every freshwater habitat on both islands. Nests or recent hatchlings also were observed, indicating that populations on the islands are reproductively active and self-perpetuating. Turtles. — Only one species of freshwater turtle, P. scripta, was abundant on either island. This species was restricted to aquatic habitats and was not observed in waters of high salinity. Kinosternon subrubrum may be present on both islands in greater numbers than our observations suggest. Some of the few specimens were associated with brackish marshes, a habitat not intensively sampled during the study period since on the mainland it is not normally frequented by Kinosternon in this region of the country. Lizards. — Of the several species of lizards found on the two islands, those in the family Scincidae are invariably the most abundant and apparent in areas having forest litter of any sort. Scincella laterale is dominant on both islands, especially in forested areas where pine needles are an important component of the forest floor. However, in ubiquity and relative abundance, Anolis carolinensis is the dominant arboreal species. Despite the apparent availability of suitable habitats throughout Barrier Island Herpetofauna 157 Kiawah Island^ Cnemidophorus sexlineatus was surprisingly scarce. This lizard is abundant on some barrier islands along the southeastern Atlantic Coast. Snakes. — The most apparent snake on Kiawah Island was Thamno- phis sauritus. On Capers Island, Agkistrodon piscivorus and Coluber constrictor were the most frequently encountered species. The difficulty of assessing abundance and population levels of snakes is well known to herpetologists. Hence other, more secretive, species may be more com- mon than is apparent. Salamanders. — Only Plethodon glutinosus was found on Kiawah Island. It occupies mesic habitats and is locally abundant in certain for- ested areas. No salamanders were collected on Capers Island although suitable habitat exists there for P. glutinosus. Salamanders are absent on many barrier islands, or are represented by only a few species. On the mainland, salamander activity is greatest during late fall, winter, and early spring; sampling efforts, which were not made on Capers during these seasons, might confirm the presence of P. glutinosus . Frogs and toads. — Almost all of the anuran species on the islands were very abundant, particularly during periods of heavy rainfall. An assessment of relative population levels of the different anuran species is difficult to make due to the large numbers observed of almost every species at one time or another. Hyla squirella occupies diverse habitats and is very abundant, as attested by the very large breeding choruses that develop after heavy spring or summer rains. It is clearly the most widespread terrestrial spe- cies on Capers and Kiawah islands. Two species of toads appear to differ sharply in abundance on Kiawah Island. Scaphiopus holbrooki was common to abundant, especially in mesic, mixed forests as indicated, for example, by the drift fence data from transect 3 (Table 4). In contrast, Bufo terrestris, a common toad on the mainland, occupies similar habi- tats but was less abundant than S. holbrooki. Since most of the S. hol- brooki captured were juveniles, a demographically realistic comparison of abundance is not possible. A third terrestrial anuran, Gastrophryne carolinensis, is common to abundant in diverse habitats on both islands. Rana utricularia is the most apparent anuran species in the more stable aquatic sites; adults are more or less restricted to freshwater marshes and ponds. Very large breeding choruses of this species develop after heavy winter and spring rains. Although individuals sometimes occur in smaller, less stable aquatic sites (ditches, rain-filled depressions, etc.) throughout both islands, these are often postmetamorphic juveniles or subadults that may not survive. Rana utricularia shares the larger fresh- water marshes and ponds with Hyla cinerea, especially those ponds con- taining cattails or other emergent vegetation. Hyla cinerea is somewhat 158 J. Whitfield Gibbons and Julian R. Harrison III less terrestrial than H. squirella. Although relatively large breeding cho- ruses develop on both islands after heavy spring or summer rains, H. cinerea appears to be much less abundant than H. squirella, especially on Capers Island and in the developed parts of Kiawah. Kiawah Island — Before And After Development of Kiawah Island as a resort community has led to cer- tain changes in the herpetofauna of the modified areas. Many species persist in these areas, but their population densities appear to be consid- erably reduced. Sparrow Pond and associated habitats in the vicinity of the present hotel complex previously harbored a rich herpetofauna represented by virtually all of the species known to occur on the island. With development, the pond was enlarged and deepened and most of the emergent vegetation removed, either deliberately or as a result of envir- onmental manipulations. In addition. Sparrow Pond was stocked with largemouth bass, Micropterus salmoides, and sunfish, Lepomis sp., which do not occur naturally on the island. Surrounding habitats were modified chiefly by clearing understory vegetation, removing surface lit- ter and some trees, and constructing buildings and roads. One apparent consequence of these activities is loss of breeding popu- lations of frogs and toads. While a few species still frequent this area (Table 4), population densities are extremely low and may represent recruitment by dispersal from other parts of the island. Often only two or three individuals are heard calling during or after heavy rains (such as Hurricane David in September 1979) whereas prior to development such meteorological conditions triggered choruses involving hundreds of frogs. Large choruses are still heard in some undeveloped parts of the island. We found no evidence (e.g. eggs and/or larvae) that any anuran is now reproducing in the Sparrow Pond area. Failure of reproduction is due at least in part to the presence of introduced predatory fish. Absence of eggs may also be attributed to use of insecticides in the general area; the entire island is treated weekly during the summer months. Insecticides may affect eggs, larvae, or adults directly, or they may limit population sizes by reducing available food sources (Hall 1980). The palmetto forest that previously bordered the northern perimeter of Sparrow Pond had a large population of P. glutinosus. It is doubtful that a population persists in the area today, as no individuals were observed during the recent study period. Removal of organic litter, warmer ground temperatures resulting from increased insolation, and perhaps use of insecticides, may explain this species’ absence. Similar considerations may account for the absence of scincid lizards; none was observed during the recent study period although three species were abundant earlier. Thamnophis sauritus and Agkistrodon piscivorus were previously Barrier Island Herpetofauna 159 common in the grassy areas bordering Sparrow Pond. We saw no T. sauritus and only one A. piscivorus around Sparrow Pond in 1979. Absence of T. sauritus may reflect in part the dearth of suitable habitat and prey, chiefly frogs. Scarcity of A. piscivorus is due in part to selective removal of this venomous species from the area. Several road-killed snakes were encountered during the recent study period; prior to the existence of paved roads and heavier traffic loads such instances were rare. As development proceeds, the incidence of road-kills will almost certainly increase, at least until population sizes are appreciably reduced. Residual turtle and alligator populations remain in Sparrow Pond, but population densities appear to be low. Drainage operations in the pond prior to its modification revealed the presence of 50 P. scripta (Tony Niemeyer, pers. comm.). Small alligators are still present, but larger ones are removed and translocated to ponds in undeveloped areas. At least two other ponds on Kiawah Island have been enlarged, deep- ened, cleared of emergent vegetation, and stocked with freshwater fish. Both ponds are on the otherwise undeveloped part of the island. Removal of cattails and other emergent plants from these ponds probably accounts in part for the apparent reduction in population size of H. cinerea. This species is often abundant in cattail marshes; its apparent scarcity on Capers Island may reflect the relative dearth of such habitat on that island. Considerations for Recreational Development of Barrier Islands The greatest potential human disruptions to the herpetofauna of Cap- ers Island and the undeveloped parts of Kiawah would occur as a result of major alteration of forests and freshwater ponds. Freshwater habitats support dense populations of many species. Changes in water level and salinity, and removal of vegetation and litter, will change the herpeto- faunal character of the areas. Drainage of shallow aquatic sites for mosquito control or other pur- poses would have a severe impact on reproductive efforts of frogs and toads. Most species of frogs reported from the islands have relatively short larval periods (90 days or less) and deposit their eggs in small, rain-filled depressions or in larger, often temporary ponds that lack pred- atory fishes. Only the mosquitofish, Gambusia affinis, has been reported to occur naturally in strictly freshwater habitats on Kiawah Island, and even this species is absent from Capers. Creating permanent ponds may provide opportunities for establishment of frog species with longer larval periods and different breeding strategies; however, intro- duction of freshwater game fishes would create a serious predator impact on all anurans. Recreational development of ponds inhabited by P. scripta would probably not directly alter their survival, but population features such as rapid growth rates and large individuals might be 160 J. Whitfield Gibbons and Julian R. Harrison III affected. The most severe impact to terrestrial reptiles and amphibians could come from removal of essential organic litter and ground cover in the forest systems. Land clearing or large scale removal of ground cover could eliminate some species and would completely alter the natural systems as they now exist. Removal of trees and other forest vegetation would also make the habitats less desirable for many species. However, some type of recreational development such as conservative trail con- struction or small picnic areas and campgrounds, should have minimal impact on these species. Plethodon glutinosus is a possible exception, since it is most common in the more elevated mesic forests, areas that are also the most suitable for picnic areas, campgrounds, and roads. Con- struction of paved roads will lead to an increase in the incidence of road-killed vertebrates. Because of the minimal presence of most species in the open beach, salt marsh, and dune habitats, changes in these areas will not have a major effect on the herpetofauna as a whole. However, particular species, such as Caretta caretta on the beach and Malaclemys terrapin in the marsh, could suffer unless ecological awareness prevails among those fostering development of these islands. ACKNOWLEDGMENTS. — We are indebted to several individuals who assisted in various phases of field work, including installation of drift fences. In particular we thank Joe Schubauer, Judy Greene, Garfield Keaton, Dr. Jan Caldwell, Gary Moran, Brad Gammon, Dr. Bob Ald- ridge, Steve Bennett, Kent Brown, Lou Ann Brown, Dr. Rebecca Sharitz, Dave Bennett, Joe McAuliffe, Bob Parmenter, Ray Semlitsch, Vince DeMarco, Mike McMillan, Linda Whittlesy, Charles Haddock, and Albert Sanders for their contributions. We thank Robert Jeter and the Youth Conservation Corps group for providing transportation to Capers Island and other assistance; and Michael McKenzie and Robert Dunlap of the S. C. Wildlife and Marine Resources Department for maps and other information concerning Capers Island, and for use of facilities on that island. Tony Niemeyer permitted access and provided information about Kiawah Island. Ray Semlitsch, Rebecca Sharitz and Stephen H. Bennett critically read the original manuscript, which was also greatly improved by the comments of two anonymous reviewers. Support for parts of the research was provided by Contract DE-AC09- 76SR00819 between the U. S. Department of Energy and the University of Georgia, by the South Carolina Wildlife and Marine Resources Department, and by a National Science Foundation Grant (DEB- 7904758) to J. W. Gibbons. Barrier Island Herpetofauna 161 LITERATURE CITED Bara, Mark O. 1976. American alligator investigations. Final Study Report. August 1970-December 1975. S. C. Wildl. Mar. Resour. Dep. Blaney, Richard M. 1971. An annotated check list and biogeographic analysis of the insular herpetofauna of the Apalachicola region, Florida. Herpetologica, 27:406-430. Conant, Roger. 1975. A field guide to reptiles and amphibians of eastern and central North America. 2nd ed. Houghton Mifflin Co., Boston. 429 pp. Cromwell, John E. 1971. Barrier coast distribution. A world-wide survey. Abst. Vol., Second Natl. Coastal Shallow Water Res. Conf. 50 pp. Dolan, Robert, P. J. Godfrey and W. E. Odum. 1973. Man’s impact on the barrier islands of North Carolina. Am. Sci. (57:152-162. Engels, William L. 1942. Vertebrate fauna of North Carolina coastal islands. Am. Midi. Nat. 25:273-304. 1952. Vertebrate fauna of North Carolina coastal islands. II. Shackleford Banks. Am. Midi. Nat. 47:702-742. Fisher, John J. 1968. Barrier island formation: Discussion. Geol. Soc. Am. Bull. 791421-1432. Gaddy, L. L. 1979. In preparation. Vegetation studies of Kiawah and Capers Islands. In Carrying capacity of barrier islands. Report for U. S. Heritage, Conservation, and Recreation Service prepared by S. C. Wildl. Mar. Resour. Dep., Columbia. Gibbons, J. Whitfield, and D. H. Bennett. 1974. Determination of anuran terrestrial activity patterns by a drift fence method. Copeia 1974(l):236-243. , and J. W. Coker. 1978. Herpetofaunal colonization patterns on Atlantic Coast barrier islands. Am. Midi. Nat. 99:213-233. , G. H. Keaton, J. P. Schubauer, J. L. Greene, D. H. Bennett, J. R. McAuliffe and R. R. Sharitz. . 1979. Unusual population size structure in freshwater turtles on barrier islands. Ga. J. Sci. 57:155-159. Hall, Russell J., and D. Swineford. 1980. Toxic effects of endrin and toxophene on the southern leopard frog, Rana sphenocephala. Env. Pollut. Ser. A. 23:53-65. Hillestad, Hilburn O., J. R. Bozeman, A. S. Johnson, C. W. Berisford and J. I. Richardson. 1975. The ecology of the Cumberland Island National Seashore, Camden County, Georgia. Tech. Rep. Ser. No. 75-5 Ga. Mar. Sci. Center. 299 pp. Hoyt,, James H. 1967. Barrier island formation. Geol. Soc. Am. Bull. 75:1125- 1136. Hoyt, James H. 1968. Barrier island formation: Reply. Geol. Soc. Am. Bull. 79:947. Jackson, Crawford G., Jr., and M. M. Jackson. 1970. Herpetofauna of Dauphin. Island, Alabama. Q. J. Fla. Acad. Sci. 55:281-287. Johnson, A. Sydney, H. O. Hillestad, S. F. Shanholtzer and G. F. Shanholtzer. 1974. An ecological survey of the coastal region of Georgia. Natl. Park Ser. Sci. Monogr. Ser. 51-233. Kjerfve, B. 1974. Climatology. Section C. pp. 1-34 m A preliminary report on the environmental inventory of Kiawah Island, S. C. Environmental Research Center, Inc., Columbia. Lee, David S. 1972. List of the amphibians and reptiles of Assateague Island. Bull. Md. Herpetol. Soc. 5:90-95 162 J. Whitfield Gibbons and Julian R. Harrison III Leont’yev, Oleg K., and L. G. Nikiforov. 1965. Reasons for the world-wide occurrence of barrier beaches. Oceanology 5:61-67. Lewis, Thomas H. 1946. Reptiles and amphibians of Smith Island, North Carolina. Am. Midi. Nat. 56:682-684. MacArthur, Robert H., and E. O. Wilson. 1967. The theory of island biogeography. Princeton Univ. Press, Princeton. 203 pp. Martof, Bernard S. 1963. Some observations on the herpetofauna of Sapelo Island, Georgia. Herpetologica 79:70-72. , W. M. Palmer, J. R. Bailey and J. R. Harrison, III. 1980. Amphibians and reptiles of the Carolinas and Virginia. Univ. North Carolina Press, Chapel Hill. 264 pp. McKenzie, Michael. 1975. A conceptual “Use Plan” for Capers Island, Charleston County, South Carolina. Spec. Rep. No. 1. S. C. Wildl. Mar. Resour. Dep. Otvos, Ervin G., Jr. 1970. Development and migration of barrier islands. Northern Gulf of Mexico. Geol. Soc. Am. Bull. 57:241-246. Parnell, James F., and D. A. Adams. 1970. Smith Island — a resource capability study — interim report. Wilmington, N.C. 83 pp. Quay, Thomas L. 1959. The birds, mammals, reptiles and amphibians of Cape Hatteras National Seashore Recreational Area. U.S. Dep. Inter. Natl. Park Serv. 88 pp. Schwartz, Maurice L. 1971. The multiple causality of barrier islands. J. Geol. 79:91-94. Accepted 16 January 1981 163 SUBSCRIPTIONS AND EXCHANGES The editors anticipate two issues of approximately 150 pages each annually. Rates for subscriptions for all issues appearing within the calendar year: Individual — United States $ 7.50 Individual — Foreign $10.00 Institution $12.50 Single issue purchase $ 4.50 All subscriptions must be paid in advance. Issues will be available on an exchange basis to organizations and institu- tions publishing general natural history and ecology journals or papers on a fairly regular schedule. Publications received on exchange will be placed in the State Museum’s H. H. Brimley Memorial Library. Address all subscriptions and requests for information on purchase and exchange to Managing Editor, Brimleyana, N. C. State Museum of Natural History, P.O. Box 27647, Raleigh, NC 27611. Back issues are available for $4.50 each. DATE OF MAILING Brimleyana No. 4 was mailed on 13 January 1981. ERRATUM The following error appeared in Brimleyana No. 4: Page 162, running head should read: George A. Feldhamer and Joseph A. Chapman, not Richard F. Collins. 164 TABLE OF CONTENTS 1980 Number 3 Barr, Thomas C., Jr. New species groups of Pseudanophthalmus from the Central Basin of Tennessee (Coleoptera;Carabidae:Trechinae) 85 Biernbaum, Charles K. Occurrence of the “tramp” terrestrial amphipods Talitroides alluadi (Chevreux) and T. topitotum (Burt) (Amphipoda: • Talitridae) in South Carolina 107 Braswell, Alvin L. (see Palmer, William M.) 49 Brown, E. E. Some historical data bearing on the Pine Barrens Treefrog, Hyla andersoni, in South Carolina 113 Burr, Brooks M. A distributional checklist of the fishes of Kentucky 53 Hedges, S. Blair (see Mitchell, Joseph C.) 119 Laerm, Joshua (see Roth, Janet A.) 1 Lee, David S. (see Robinson, Sarah S.) 43 McMillan, Michael A. (see Semlitsch, Raymond D.) 97 Mitchell, Joseph C. and S. Blair Hedges, Ambystoma mabeei Bishop (Caudata: Ambystomatidae); an addition to the salamander fauna of Virginia 119 Maurakis, Eugene G. (see Burkhead, Noel M.) 75 Meyers, Joseph M. (see Laerm, Joshua) 47 Nagel, Jerry W. Life History of the Mottled sculpin, Cottus bairdi, in northeastern Tennessee (Osteichthyes:Cottidae) 115 Redmond, William H. Notes on the distribution and ecology of the Black Mountain Dusky Salamander Desmognathus welteri Barbour (Amphibia: Plethodontidae) in Tennessee 123 Robison, Henry W. (see Williams, James D.) 149 Rossman, Douglas A. and Robert L. Erwin. Geographic variation in the snake Storeria occipitomaculata (Storer) (Serpentes:Colubridae) in southeastern United States 95 Shelley, Rowland M. (see Filka, Marianne E.) 1 Vitt, Laurie J. (see Laerm, Joshua) 47 Williams, James D. and Henry W. Robison, Ozarka, a new subgenus of Etheostoma (Pisces: Percidae) 149 Wiseman, Jeffrey S. (see Gooch, James L.) 133 Palmer, William M. and Alvin L. Braswell. Additional records of albinistic amphibians and reptiles from North Carolina 49 Robinson, Sarah S. and David S. Lee. Recent range expansion of the Groundhog, Marmota monax, in the Southeast (Mammalia: Rodentia) 43 Roth, Janet A. and Joshua Laerm. A late Pleistocene vertebrate assemblage 165 from Edisto Island, South Carolina 1 Semlitsch, Raymond D. and Michael A. McMillan. Breeding migrations, population size structure, and reproduction of the Dwarf Salamander, Eurycea quadridigitata, in South Carolina 97 Shelley, Rowland M. The status of Cleoptoria shelfordi Loomis, with the proposal of a new genus in the milliped family Xystodesmidae (Polydesmida) 31 Number 4 Ashton, Ray E., Jr., Alvin L. Braswell and Sheldon I. Guttman. Electro- phoretic analysis of three species of Necturus (Amphibia: Proteidae), and the taxonomic status of Necturus lewisi (Brimley) 43 Braswell, Alvin L. (see Ashton, Ray E., Jr.) 43 Burkhead, Noel M., Robert E. Jenkins and Eugene G. Maurakis. New records, distribution and diagnostic characters of Virginia ictalurid catfishes with an adnexed adipose fin 75 Chapman, Joseph A. (see Feldhamer, George A.) 161 Collins, Richard F. Stomach contents of some snakes from eastern and central North Carolina 157 Culver, David C. and Timothy J. Ehlinger. Effects of microhabitat size and competitor size on two cave isopods 103 Ehlinger, Timothy J. (see Culver, David C.) 103 Erwin, Robert L. (see Rossman, Douglas A.) 95 Feldhamer, George A. and Joseph A. Chapman. Mandibular dental anomaly in Whitetailed deer 161 Filka, Marianne E. and Rowland M. Shelley. The milliped fauna of the Kings Mountain region of North Carolina (ArthropodaiDiplopoda) 1 Freeman, B. J. (see Laerm, Joshua) 47 Gooch, James L. and Jeffrey S. Wiseman. Morphological and habitat variability in Gammarus minus Say (AmphipodaiGammaridae) 133 Guttman, Sheldon I. (see Ashton, Ray E., Jr.) 43 Jenkins, Robert E. (see Burkhead, Noel M.) 75 Laerm, Joshua, B. J. Freeman, Laurie J. Vitt, Joseph M. Meyers and Lloyd Logan. Vertebrates of the Okefenokee Swamp 47 Logan, Lloyd (see Laerm, Joshua) 47 166 INDEX TO SCIENTIFIC NAMES (New names in italics) Numbers 3: and 4: (1980) New Names Brevigonus shelfordi 3:31-42 Pseudanophthalmus inquisitor 3:94-95 (Ozarka) [Etheostoma] 4:149-156 Pseudanopthalmus productus 3:91-92 Pseudanopthalmus fowlerae 3:88-89 Pseudanopthalmus simplex 3:86-88 Abacion magnum 4:6,9,20,21,32,33, 34,35,36,37 Acantharcus pomotis 4:52 Accipiter cooperii 4:61 striatus 4:61 Acer rubrum 3:121;4:48 Acipenser fulvescens 3:57 Acris gryllus dorsalis 4:54 Actitis macularia 4:61 Agelaius phoeniceus 4:65 Agkistrodon contortrix 4:157,158 contortrix 4: 158 piscivorus 4: 159 conanti 4:57 leucostoma 4:157 piscivorus 4:157,158,159 Aimophila aestivalis 4:65 Aix sponsa 4:60 Ajaia ajaja 4:60 Alligator 3:1 1,22,23 mississippiensis 3:5,8, 10;4:56,71 Alosa alabamae 3:55,61 chrysochloris 3:61 Ambloplites rupestris 3:73 Amblyopsis spelaea 3:71 Ambystoma cingulatum 4:55 mabeei 3:1 19-121 maculatum 3:44 opacum 4:55 talpoideum 3:50,101;4:55 tigrinum 3:1 13;4:55 Amia calva 3:58;4:51 Ammocrypta 4:149 asprella 3:55,74 Clara 3:55,75 pellucida 3:75 vivax 3:55,75 Ammodramus henslowii 4:65 savannarum 4:65 Ammospiza leconteii 4:65 Amphiuma means 3:49;4:55 Anas acuta 4:60 americana 4:60 clypeata 4:60 crecca 4:60 discors 4:60 penelope 4:60 platyrhynchos 4:60 rubripes 4:60 strepera 4:60 Andrognathus corticarius 4:6,8,10,11, 33,34,35,36,37 Andropogon virginicus 4:48 Anguilla rostrata 3:61;4:51 Aniulus orientalis 4:13 Anhinga anhinga 4:59 Anolis carolinensis 4:56 Anthus spinoletta 4:64 167 Apheloria tigana 4:37 Aphredoderus sayanus 3:71;4:52 Aplodinotus grunniens 3:80 Aquila chrysaetos 4:61 Aramus guarauna 4:61 Archilochus colubris 4:62 Arcitalitrus sylvaticus 3:109 Ardea herodias 4:60 Armadillium sp. 3:108 Arundinaria spp. 3:121 Auturus erythropygos 4:8,9,25-26,27,32,33, 34.35.36.37 georgianus 4:26 Aytha affinis 4:60 americana 4:60 collaris 4:60 marila 4:60 valisineria 4:60 Bipalium sp. 3:108 Bison 3:23 antiquus 3:6,8,20,21,24 bison 3:21 Blarina carolinensis 4:68 Bombycilla cedrorum 4:64 Boraria stricta 4:8,9,26,27-28,32,33, 34,35,36,37,39,40 Boreostracon floridanus 3:12 Brachyiulus lusitanus 4:7,8,12,32,33, 35.36.37 pusillus 4: 12 Branneria carinata 4:38 Branta canadensis 4:60 Brevigonus 3:31-42 shelfordi 3:32-42 Botaurus lentiginosus 4:60 Bubo virginianus 4:62 Bubulcus ibis 4:60 Bucephala albeola 4:60 clangula 4:60 Bufo quericus 4:54 terrestris 4:54 woodhousei fowleri 3;50;4:158 Buteo jamaicensis 4:61 lagopus 4:61 lineatus 4:61 platypterus 4:61 Butorides striatus 4:60 Caecidotea cannulus 4:103-1 13 holsingeri 4:103-1 13 Caiman sclerops 4:56 Calidris alba 4:62 alpina 4:62 mauri 4:62 pusillus 4:62 Callitriche 4:153 Cambala annulata 4:6,9,21,22,32,33, 34,35,36,37 Camellia japonica 3:108 Campephilus principalis 4:63,71 Campostoma anomalum 3:62 oligolepis 3:80 Canis 3:23 dims 3:6,8,14,15,24 lupus 3: 15 rufus niger 4:67,70 Capella gallinago 4:61 Capreolus capreolus 4:162 Caprimulgus carolinensis 4:62 vociferus 3:44;4:62 Carassius auratus 3:62 Cardinalis cardinalis 4:65 Carduelis pinus 4:65 tristis 4:65 Carex 4:48 168 Carphophis amoenus amoenus 3:51 Carpiodes carpio 3:68 cyprinus 3:68 velifer 3:68 Carpodacus purpureus 4:65 Carya illinoensis 3:108 sp. 3:108 Casmerodius albus 4:60 Castor 3:22 canadensis 3:5,8, 14;4:67 carolinensis 4:69 Castoroides 3:22,23 ohioensis 3:5,8,13,14 Cathartes aura 4:60 Catharus fuscescens 4:64 guttata 4:64 minima 4:64 ustulata 4:64 Catoptrophorus semipalmatus 4:61 Catostomus commersoni 3:68 Cemophora coccinea 3:49,1 13 copei 4:56 Centrarchus macropterus 3:73;4:52 Certhia familiaris 4:63 Chaetura pelagica 4:62 Charadrius semipalmatus 4:61 vociferus 4:61 Chelydra 3:22 serpentina 3:4,7,8;4:158 serpentina 3:50;4:57 Chen caerulescens 4:60 Chlamytherium 3:23 Chlidonias niger 4:62 Chologaster agassizi 3:71 Chordeiles minor 4:62 Chondestes grammacus 4:65 Chrysemys 3:22 concinna 3:7,8 floridana 3:5,7,8 nelsoni 4:57 scripta petrolei 3:5,7,8,24 Circus cyaneus 4:61 Cistothorus palustris 4:63 platensis 4:63 Cleidogona caesioannulata 4:38 medialis 4:6,8,14-15,33,34,35,36, 37,39 Cleptoria abbotti 3:39 shelfordi 3:31-42 Clethra alnifolia 4:48 Clinostomus elongatus 3:55,62 funduloides 3:63 Cnemidophorus sexlineatus sexlin eatus 4:56 Coccyzus americanus 4:62 erythropthalmus 4:62 Colaptes auratus 4:62 Colinus virginianus 4:61 Coluber constrictor priapus 4:56 sp. 3:5,8,10 Columbina passerina 4:62 Condylura cristata 4:67 cristata 4:68 Contopus virens 4:63 Coragyps atratus 4:60 Corvus brachyrhynchos 4:63 ossifragus 4:63 Cottus bairdi 3:72;4:1 15-121 bairdi 4:120 kumleini 4:120 carolinae 3:72 Coturnicops noveboracensis 4:61 169 Crangonyx antennatus 4: 1 1 1 , 1 33, 1 45 Croatania catawba 4:7,9,26,28,29,30,33, 34.35.36.38.39 Crocodylus acutus 3:11 Crotalus adamanteus 4:57 horridus 3:49 atricaudatus 4:57 Cryptotis parva 4:67 parva 4:68 Ctenopharyngodon idella 3:63 Cyanocitta cristata 4:63 Cycleptus elongatus 3:68 Cylindroiulus truncorum 4:37 Cyprinus carpio 3:63 Cyrilla racemiflora 4:48 Dasypus 3:23 bellus 3:5,8, 1 1,12,24 novemcinctus 3:12;4:68 mexicanus 4:69 Deirochelys reticularia reticularia 4:57 Delophon carolinum 4:20 georgianum 4:6,9,20-21,32,33,35,36, 37.38.39 Deltotaria lea 4:7,9,29,30,32,33,34, 35,36,38,39,40 Dendroica caerulescens 4:64 cerulea 4:64 coronata 4:64 discolor 4:64 dominica 4:64 fusca 4:64 magnolia 4:64 palmarum 4:64 pensylvanica 4:64 petechia 4:64 pinus 4:64 striata 4:64 tigrina 4:64 virens 4:64 Desmognathus fuscus 4:123,1 25, 1 26, 1 27, 1 28, 1 29, 1 30 auriculatus 4:55 monticola 4: 1 23, 1 25, 1 26, 1 27, 129,130 ochrophaeus 4:125,129 welteri 4:123-131 Diadophis punctatus punctatus 4:56 ssp. 3:51 Diathera 4:153 Didelphis virginiana pigra 4:68 Dolichonyx oryzivorits 4:65 Dorosoma cepedianum 3:61 petenense 3:61 Drymarchon corais couperi 4:56,71 Dryocopus pileatus 4:62 Dumetella carolinensis 4:63 Egretta thula 4:60 Elanoides forficatus 4:61 Elaphe guttata 3:49 guttata 4:56 obsoleta 3:49;4:157 obsoleta 3:52;4:157,158 quadrivittata 4:56,158 spiloides 4:56 sp. 3:5,8,10 Elassoma evergladei 4:51,52 okefenokee 4:51,52 zonatum 3:73 Empidonax virescens 4:63 Enneacanthus chaetodon 4:52 gloriosus 4:52 obesus 4:52 Eptesicus fuscus fuscus 4:67,69 Equus 3:23 caballus 3:17 170 sp. 3:6,8,17 Eremotherium 3:23 mirabile 3:5,8,13 Ericymba buccata 3:63 Erimyzon oblongus 3:68 sucetta 3:68;4:51 Eriobotrya japonica 3:108 Esox americanus 3:62;4:51 lucius 3:62 masquinongy 3:62 niger 3:62;4:51 Etheostoma 4:149-156 (Ozarka) 4:149-156 asprigene 3:75 atripinne 3:75 barbouri 3:75 boschungi 4:149-156 bellum 3:75 blennioides 3:75 caeruleum 3:75 camurum 3:75 chlorosomum 3:75 cinereum 3:75 cragini 4:149-156 flabellare 3:76 fusiforme 3:76;4:52 gracile 3:76 histrio 3:76 kennicotti 3:76 maculatum 3:76 microlepidum 3:76 microperca 3:55,76 neopterum 3:76 nigrum 3:76 obeyense 3:77 pallididorsum 4:149-156 parvipinne 3:77 proeliare 3:77 punctulatum 4:149-156 rufilineatum 3:77 sagitta 3:77 smithi 3:77 sp. 3:78 spectabile 3:77 squamiceps 3:77 stigmaeum 3:77 swaini 3:77 tippecanoe 3:77 trisella 4:149-156 variatum 3:78 virgatum 3:78 zonale 3:78 Eudocimus albus 4:60 Eumeces egregius similis 4:56 fasciatus 4:56 inexpectatus 4:56 laticeps 4:56 Eupatorium sp. 3:98 Euphagus carolinus 4:65 cyanocephalus 4:65 Eurycea bislineata bislineata 3:44 cirrigera 3:50;4:55 quadridigitata 3:97-105,1 13;4:55 Falco columbarius 4:61 peregrinus 4:61,71 sparverius 4:61 Farancia 3:1 15 abacura 3:49,1 13 abacura 3:51;4:56 erytrogramma 4:56 Felis 3:23 concolor coryi 4:67,70,71 onca augusta 3:6,8,16 Florida caerulea 4:60 Fulica americana 4:61 Fundulus catenatus 3:71 chrysotus 3:71;4:51,52 cingulatus 4:51,52 171 dispar 3:55 lineolatus 4:51,52 notatus 3:71,72 notti 3:55,72 olivaceus 3:72 Gallinula chloropus 4:61 Gambusia affinis 3:72;4:52 Gammarus bousfieldi 4:133,145 minus 4:107,109,112,133-147 oceanicus 4:133 pulex 4:133 Gastrophryne carolinensis 4:55 Gavia immer 4:59 Gavialosuchus 3:22,24 sp. 3:5,8,10,1 1 Geochelone 3:23,25 sp. 3:5,8,10 Geomys pinetis 4:68 floridanus 4:67 pinetis 4:69 Geothlypis trichas 4:65 Glaucomys volans querceti 4:69 Glossotherium harlani 3:5,8,13 Glyptotherium 3:22,23,25 floridanum 3:5,8,12,24 Gopherus 3:23 polyphemus 4:57 sp. 3:5,8,10 Gordonia lasianthus 4:48 Grus canadensis 4:61 pratensis 4:58,71 Guiraca caerulea 4:65 Hadropterus 4:149 Haliaeetus leucocephalus 4:61,71 Halichoerus 3:22,23 grypus 3:6,8,16 Helmintheros vermivorus 4:64 Hemitremia flammea 3:55,63 Heterandria formosa 4:51,52 Heterodon platyrhinos 4:56 simus 4:56 Hiodon alosoides 3:61 tergisus 3:61 Hirundo rustica 4:63 Holmesina septentrionalis 3:5,8,12 Hybognathus hayi 3:63 nuchalis 3:63 placitus 3:63 Hybopsis aestivalis 3:63 amblops 3:63 dissimilis 3:64 gelida 3:64 gracilis 3:64 insignis 3:64 meeki 3:64 storeriana 3:64 x-punctata 3:55,64 Hydranassa tricolor 4:60 Hydrodchoerus pinckneyi 3:14 Hyla andersoni 3:1 13-1 17 chrysoscelis 4:54 cinerea cinerea 4:55 crucifer bartramiana 4:55 femoralis 4:55 gratiosa 4:55 squirella 4:55 Hylocichla mustelina 4:63 Hypentelium nigricans 3:68 Ichthyomyzon bdellium 3:57 castaneus 3:57 fossor 3:57 gagei 3:57 greeleyi 3:57 unicuspis 3:57 172 Ictalurus 4:75-93 brunneus 4:75,76,77,79-81,82,83,84, 85,86,87,90 catus 3:69;4:75,76,79,81,82,86,87,89 furcatus 3:69;4;75, 76,79,8 1-82, 84.86.89 melas 3:69;4:75,76,77,78,80,81,82-83 85.87.90 natalis 3:69;4:52,76,78,83,85,87,90 nebulosus 3:69;4:52,75, 76,78,82,83, 85.87.90 platycephalus 4:76,77,79,80,83-84,85, 86.87.90 punctatus 3:70;4:52,76,79,8 1,82,84, 86.89.90 serracanthus 4:76,85 sp. 4:158 Icteria virens 4:65 Icterus galbula 4:65 spurius 4:65 Ictiobus bubalus 3:68 cyprinellus 3:68 niger 3:68 Ilex cassine 4:48 glabra 4:49 Iridoprocne bicolor 4:63 Itea virginica 4:48 Ixobrychus exilis 4:60 Junco hyemalis 4:65 Kinosternon 3:22 bauri palmarum 4:57 sp. 3:5, 7,8 subrubrum subrubrum 4:57 Labidesthes sicculus 3:73;4:52 Lachnanthes caroliniana 4:48 Lagochila lacera 3:55,69 Lampetra aepyptera 3:57 lamottei 3:57 Lampropeltis calligaster rhombomaculata 4:56 getulus getulus 4:56 getulus X floridana 4:56 triangulum elapsoides 4:56 Lanius ludovicianus 4:64 Larus argentatus 4:62 atricilla 4:62 Lasiurus borealis borealis 4:69 cinereus cinereus 4:69 intermedius floridanus 4:69 seminolus 4:68,69 Lepisosteus oculatus 3:58 osseus 3:58 platostomus 3:58 platyrhincus 4:51 spatula 3:58 Lepomis auritus 3:73 cyanellus 3:73 gibbosus 3:73 gulosus 3:73;4:52 humilis 3:73 macrochirus 3:74;4:52 marginatus 3:74;4:52 megalotis 3:74 microlophus 3:74 punctatus 3:74;4:52 symmetricus 3:74 Leptolucania ommata 4:51,52 Lespedeza sp. 3:98 Leucothoe racemosa 4:48 Ligustrum sp. 3:108 Limnodromus sp. 4:62 Limnothlypis swainsonii 4:64 Liquidambar styraciflua 3:108,121 Lontra canadensis vaga 4:70 Lophodytes cucullatus 4:60 173 Lota lota 3:55,71 Lynx rufus 4:68 floridanus 4:70 Lyonia lucida 4:48 Macroclemys temmincki 4:57 Magnolia grandiflora 3:108 virginiana 4:48 Mammut 3:23 americanum 3:6,8,17 Mammuthus 3:23 columbi 3:6,8,16 imperator 3:16 Marmota monax 3:43-48 Martes pennanti 3:25 Masticophis flagellum flagellum 4:56 sp. 3:5,8,10 Megaceryle alcyon 4:62 Megalonyx 3:23 jeffersonii 3:5,8,12 Melanerpes carolinus 4:63 erythrocephalus 4:63 Meleagris gallopavo 4:61 Melospiza georgiana 4:65 melodia 4:66 Menidia audens 3:55,72 beryllina 3:55 Mephitis mephitis elongata 4:70 Mergus merganser 4:60 serrator 4:60 Microtus pinetorum parvulus 4:70 Micropterus coosae 3:74 dolomieui 3:74 punctulatus 3:74 salmoides 3:74;4:52 sp. 4:158 Micrurus fulvius fulvius 4:57 Mimus polyglottos 4:63 Minytrema melanops 3:69;4:51 Mniotilta varia 4:64 Molothrus ater 4:65 Morone chrysops 3:72 mississippiensis 3:73 saxatilis 3:73 Moxostoma anisurum 3:69 atripinne 3:69 carinatum 3:69 erythrurum 3:69 macrolepidotum 3:69 Mus musculus 4:67 musculus 4:70 Mustela frenata olivacea 4:70 vison mink 4:70 Mycteria americana 4:60,71 Myiarchus crinitus 4:63 Mylohyus 3:23,25 fossilis 3:6,8,18,19 nasutus 3:19 Myotis austroriparius 4:67 austroriparius 4:69 Myrica cerifera 3:98 Myrophyllum 4:153 Narceus 4:37 americanus 4:6,9,15,16-19,32,33, 34,35,36,37 annularis 4:17-19,37 Nasturtium officinale 4:153 Necturus 4:43-46 lewisi 4:43-46 maculosus 4:43-46 punctatus 4:43-46 Neochoerus 3:22,23,25 pinckneyi 3:5,8,14,24 174 Neofiber alleni 4:67,68 exoristus 4;70 Neotoma floridana floridana 4:70 Nerodia 4:157,159 cyclopion floridana 4:56 erythrogaster erythrogaster 4:56,158 transversa 4:159 fasciata fasciata 3:51;4:56,158,159 pictiventris 4:56 sipedon sipedon 4:158 spp. 4:157 taxispilota 3:5I;4:56,158 Nocomis biguttatus 3:64 effusus 3:64 micropogon 3:64 Nopoiulus minutus 4:7,8,11,33,34,35, 36,37 Notemigonus crysoleucas 3:64 Notophthalmus perstriatus 4:55 viridescens 4:158 louisianensis 4:55 viridescens 3:44 Notropis amnis 3:55,65 ardens 3:65 ariommus 3:65 atherinoides 3:65 blennius 3:65 boops 3:65,81 buchanani 3:65 camurus 3:65 chalybaeus 4:50 chrysocephalus 3:55,65 chrysoleucas 4:50 coccogenis 3:80 cornutus 3:55 emiliae 3:65 fumeus 3:65 galacturus 3:65 heterolepis 3:81 hubbsi 3:81 hudsonius 3:66 leuciodus 3:66 lutrensis 3:66;4:83 maculatus 3:66 petersoni 4:50 photogenis 3:66 procne 3:67 rubellus 3:66 shumardi 3:66 sp. 3:67 spectrunculus 3:67 spilopterus 3:66 stramineus 3:66 telescopus 3:66 umbratilis 3:66 venustus 3:66 volucellus 3:66 whipplei 3:67 Noturus elegans 3:70 eleutherus 3:70 exilis 3:70 flavus 3:70 gyrinus 3:70;4:52 hildebrandi 3:70 leptacanthus 4:51,52 miurus 3:70 nocturnus 3:70 phaeus 3:70 stigmosus 3:70 Numenius phaeopus 4:61 Nuphar luteum 4:48 Nyctanassa violacea 4:60 Nyticeius humeralis 4:67 humeralis 4:69 Nycticorax nycticorax 4:60 Nymphaea odorata 4:48 Nymphoides aquaticum 4:48 Nyssa sylvatica var. biflora 4:48 Obolaria virginica 3:39 175 Ochrotomys nuttalli aureolus 4:70 Odobenus 3:22,23 rosmarus 3:6,8,16 Odocoileus 3:23 hemionus 4:161 virginianus 3:6,8, 20;4:67,68, 161-163 virginianus 4:70 Olor columbianus 4:60 Oniscus sp. 3:108 Onychiurus 3:93 Opheodrys aestivus 4:56 Ophisaurus attenuatus longicaudus 4:56 compressus 4:56 ventralis 4:56 Ophyiulus pilosus 4:37 Oporornis agilis 4:65 formosus 4:65 Orconectes incomptus 3:86 Orontium aquaticum 4:48 Oryzomys palustris palustris 4:69 Osmerus mordax 3:62 Otus asio 4:62 Oxidus gracilis 4:7,9,22,32,33,34,35, 36,37 (02flr/:a) [Etheostoma] 4:149-156 Oxyura jamaicensis 4:60 Pachydesmus 4:38 crassicutis incursus 4:7,9,29-31,32, 33,34,35,36,38,39 Palaeolama 3:23,25 mirifica 3:6,8,19 Pandion haliaetus 4:61 Panicum 4:48 Paramylodon 3:23 Parula americana 4:64 Parus bicolor 4:63 carolinensis 4:63 Passer domesticus 4:65 Passerculus sandwichensis 4:65 Passerella iliaca 4:65 Passerina ciris 4:65 cyanea 4:65 Pelecanus erythrorhynchos 4:59 Perea flavescens 3:78 Percina 4:149 burtoni 3:55,78 caprodes 3:78,79;4:82 copelandi 3:79 cymatotaenia 3:80 evides 3:79 macrocephala 3:79 maculata 3:79 nigrofasciata 4:51,52 ouachitae 3:55,79 oxyrhyncha 3:79 phoxocephala 3:79 sciera 3:79,80 shumardi 3:79 squamata 3:80 sp. 3:80 uranidea 3:55,79 Percopsis omiscomaycus 3:71;4:82 Peromyscus gossypinus 4:68 gossypinus 4:69 polionotus polionotus 4:69 Persea borbonia 4:48 Phalacrocorax auritus 4:59 Phenacobius mirabilis 3:67 uranops 3:67 Phengodes sp. 4:157 Pheucticus ludovicianus 4:65 Philohela minor 4:61 Phoxinus cumberlandensis 3:67 erythrogaster 3:67 Physeter 3:22 sp. 3:6,8,14 Picoides borealis 4:63,71 176 pubescens 4:63 villosus 4:63 Pimephales notatus 3:67 promelas 3:67;4:83 vigilax 3:67 Pinus echinata 4:5 elliotii 3:98;4:48 palustris 4:48 strobus 4:20 taeda 3:44,98,109, 121;4:48 virginiana 4:5 Pipilo erythrophthalmus 4:65 Pipistrellus subflavus 4:68 subflavus 4:69 Piranga olivacea 3:44;4:65 rubra 4:65 Pituophis melanoleucus mugitus 4:56 Plecotus rafinesquii 4:67,69 Plegadis falcinellus 4:60 Plethodon glutinosus glutinosus 4:55 Pleuroloma flavipes 4:38 sp. 4:38 Podiceps auritus 4:59 Podilymbus podiceps 4:59 Poecilichthys punctulatus 4:150 Polioptila caerulea 4:64 Polyodon spathula 3:58 Polyxenus fasciculatus 4:6,8,9,32,33,34, 35,36,37 Polyzonium rosalbum 4:38 strictum 4:6,8,10,1 1,33,34,35,36,37 Pomoxis annularis 3:74 nigromaculatus 3:74;4:52 Pontederia cordata 4:48 Pooecetes gramineus 4:65 Porphyrula martinica 4:61 Porzana Carolina 4:61 Potamogeton 4:153 Procyon 3:23 lotor 3:6,8,15 elucus 4:67,70 Progne subis 4:63 Protonotaria citrea 4:64 Prunus sp. 3:108 Pseudacris nigrita nigrita 4:55 ornata 4:55 Pseudanophthalmus 3:85-96 acherontis 3:93 bendermani 3:94 catherinae 3:92 cumberlandus 3:85,86,90,91,92 farrelli 3:92 fowlerae 3:87,88-89,90,95 Hesperus 3:90,93 inquisitor 3:89,90,94-95 loedingi humeralis 3:90 occidentalis 3:93 productus 3:88,90,91-92,94 robustus 3:92 simplex 3:86-88,89,90,92 tiresias 3:90,92 tullahoma 3:93 valentine! 3:92 Pseudemys floridana floridana 4:57 scripta scripta 4:57 elegans 4:57 Pseudobranchus striatus spp. 4:55 Pseudopolydesmus branneri 4:8,9,23-25,32,33,34,35, 36,37 collinus 4:22,25 serratus 4:22,38 Pseudotriton montanus floridanus 4:55 ruber ruber 3:44 Ptyoiulus 4:20,34,35,37 ectenes 4:7,8, 10, 12-13,32,33,34,35, 177 36.37.39 impressus 4:7,8, 10, 1 3,33,34,35,36, 37.39 sp. 4:32,33 Polydictis olivaris 3:70;4:76,81,84, 86,89 Pyrocephalus rubinus 4:63 Quercus falcata 3:109 laurifolia 3:108,109 niger 4:49 spp. 3:121 virginiana 4:49 Quiscalus major 4:65 quiscula 4:65 Radicula 4:153 Rallus elegans 4:61 limicola 4:61 longirostris 4:61 Rana areolata aesopus 4:55 catesbeiana 4:55,157,158,159 clamitans clamitans 4:55 grylio 3:1 16;4:55 heckscheri 4:55 palustris 4:158 utricularia 4:55 virgatipes 4:55 Ranunculus 4: 153 Rattus rattus 4:67 alexandrinus 4:70 rattus 4:70 Regina alleni 4:56 rigida rigida 4:56 Regulus calendula 4:64 satrapa 4:64 Reithrodontomys humilus humilus 4:69 Rhadinaea flavilata 4:57 Rhinichthys atratulus 3:67 cataractae 3:81 Rhynchospora 4:48 Rubus sp. 3:98 Salmo fontinalis 3:61 gairdneri 3:61 trutta 3:61 Sayornis phoebe 4:63 Scalopus aquaticus australis 4:68 Scaphiopus holbrooki 3:1 13 holbrooki 4:55 Scaphirhynchus albus 3:58 platorynchus 3:58 Sceloporus undulatus 3:49 undulatus 4:56 Scincella laterale 4:56 Sciurus carolinensis 4:68 carolinensis 4:69 niger niger 4:67,69 Scytonotus granulatus 4:8,9,24,25,33, 34,35,36,37 Seiurus aurocapillus 4:64 motacilla 4:64 noveboracensis 4:64 Seminatrix pygaea pygaea 4:57 carolinensis 4:63 pusilla 4:63 Sphagnum 4:48 Sphyrapicus varius 4:63 Spizella arborea 4:65 passerina 4:65 pusilla 4:65 178 Stereochilus marginatus 4:55 Sterna forsteri 4:62 paradisaea 4:62 Sternotherus minor minor 4:57 odoratus 4:57 Stizostedion canadense 3:80 vitreum 3:80 Storeria dekayi victa 4:57 occipitomaculata 4:95-102 hidalgoensis 4:98 obscura 4:57,95,97,102 occipitomaculata 4:97,98,102 Striaria 4:37 causeyi 4:16 sp. 4:6,8,15,16,32,33,35,36 zygoleuca 4:16 Strix varia 4:62 Sturnella magna 4:65 Semotilus atromaculatus 3:68 Serenoa repens 4:49 Setophaga ruticilla 4:65 Sialia sialis 4:64 Sigmodon hispidus hispidus 4:70 Sigmoria latior 4:7,9,30,31-32,33,34,35,36,37,39 latior X hoffmani 4:31 Siren intermedia intermedia 3:50;4:55 lacertina 4:55 Sistrurus miliarius barbouri 4:57 Sitta canadensis 4:63 pacificus 3:1 10 topetotum 3:107-1 1 1 Talitrus sylvaticus 3:110 Tapirus 3:23,25 sp. 3:6,8,18 Taxodium ascendens 4:48 Teniulus 4:39 parvior 4:13 setosior 4:13 sp. 4:7,8,10,13-14,33,34,35,36, 37,39 Terrapene 3:23 Carolina bauri 4:57 Carolina 4:57 putnami 3:5,7,8 Thamnophis sauritus sackeni 4:57 sirtalis sirtalis 4:57 Thryomanes bewickii 4:63 Thryothorus lucovicianus 4:63 Tilia caroliniana 3:108 Toxostoma rufus 4:63 Trechus aduncus coweensis 3:123 howellae 3:123 toxawayi 3:123 Tremarctos floridanus 3:6,8,15 Trichechus 3:22 sp. 3:6,8,17 Sturnus vulgaris 4:64 Sus scrofa 4:70 Sylvilagus 3:23 floridanus mallurus 4:69 palustris 4:68 palustris 4:69 sp. 3:5,8,13 transitionalis 3:25 Tadarida brasiliensis 4:67 cynocephala 4:69 Talitroides alluaudi 3: 107-11 1 decoratus 3:110 Typhlichthys subterraneus 3:71 Tyrannus dominicensis 4:63 tyrannus 4:63 verticalis 4:63 Tyto alba 4:62 179 Umbra limi 3:62 pygmaea 4:51 Urocyon 3:23 cinereoargenteus 3:6,8,15 floridanus 4:70 Ursus americanus 4:68 floridanus 4:67,70,71 Utricularia spp. 4:48 Vaccinium stamineum 3:39 Viburnum prunifolium 3:39 Vireo flavifrons 4:64 griseus 4:64 olivaceous 4:64 solitarius 4:64 Virginia striatula 3:51;4:57 valeriae valeriae 4:57 Trichopetalum dux 4:6,8,16,32,33,35, 36,37 Tringa flavipes 4:62 melanoleucus 4:62 solitaria 4:61 Trionyx ferox 3:57 sp. 3:5 Troglodytes aedon 4:63 troglodytes 4:63 Tsuga canadensis 4:20 Turdus migratorius 4:63 Tursiops 3:22 truncatus 3:6,8,14 Vermivora bachmanii 4:64,71 celata 4:64 chrysoptera 4:64 pinus 4:64 Wilsonia canadensis 4:65 citrina 4:65 Woodwardia virginica 4:48 Xyris smalliana 4:48 Zenaida macroura 4:62 Zinaria brunnea 4:38 Zonotrichia albicollis 4:65 180 NEW BOOKS FROM NCSM ATLAS OF NORTH AMERICAN FRESHWATER FISHES by D. S. Lee, C. R. Gilbert, C. H. Hocutt, R. E. Jenkins, D. E. McAllister, J. R. Stauffer, Jr., and many collaborators This timely book provides accounts for all 777 species of fish known to occur in fresh waters in the United States and Canada. Each account gives a distribution map and illustration of the species, along with information on systematics, distribution, habitat, abundance, size, and general biology. “. . . represents the most important contribution to freshwater fishes of this continent since Jordan and Evermann’s 'Fishes of North and Middle America' over 80 years ago.” — Southeastern Fishes Council Pro- ceedings. 1980 825 pages ISBN 0-917134-03-6 Price: $25. North Carolina residents add 4% sales tax. Please make checks payable in U. S. currency to NCDA Museum Extension Fund. Send to FISH ATLAS, N. C. State Museum of Natural History, P. O. Box 27647, Raleigh, NC 27611. SEACOAST LIFE by Judith M. Spitsbergen Ecology of natural seashore communities of the Middle Atlantic States, with emphasis on North Carolina, is the subject of this field study guide. It also can be used as a classroom text and overall resource book. A general discussion of coastal habitats and their occupants is followed by sections on communities of the Ocean Beach, Sand Dune, Salt Marsh, Tidal Flat, and Rock Jetty and Piling. Each section includes a description of the habitat and its special features, plant and animal adaptations, the community, and typical organisms. The sections are profusely illustrated. Includes glossary and indexes to scientific and common names. 1980 114 pages Price $5.95, plus $1.00 postage and handling ($6.95). North Carolina residents add 4% sales tax. Make checks payable to NCDA Museum Extension Fund. Send to SEACOAST LIFE, N. C. State Museum of Natural History, P. O. Box 27647, Raleigh, NC 27611. INFORMATION FOR CONTRIBUTORS Submit original and two copies of manuscripts to Editor, Brimleyana, North Carolina State Museum of Natural History, P. O. Box 27647, Raleigh, NC 27611. In the case of multiple authorship, indicate correspondent. Manuscripts submitted for publication in this journal should not also be submitted elsewhere. Preparation of manuscript. Adhere generally to the Council of Biology Editors Style Manual, Fourth Edition. Use medium-weight bond paper, 8 Vi X 11 ”, and leave at least an inch margin on all sides. Double space all typewritten material. The first page will be separate and contain the title and the author’s name and address. Where appropriate, the title will indicate at least two higher categories to which taxa belong. Example: Studies of the genus Hobbseus Fitzpatrick and Payne (Decapoda: Cam- baridae). A brief informative abstract on a separate sheet follows the title page, preceding the text. In- dicative abstracts are not^ acceptable. Footnotes will be used only where absolutely necessary, numbered consecutively throughout the paper. Individuality of writing style and text organization are encouraged, but for longer papers the INTRODUCTION, MATERIALS AND METHODS, RESULTS, DISCUSSION and LITERATURE CITED format is preferable, with those headings centered and capitalized. Headings plus sub-headings must be kept to a total of three levels. Scientific names in taxonomic papers will include the author in first usage. Descriptions of new taxa must be in accordance with the requirements of established international codes. Etymology is desirable. Last item in the text will be acknowledgments, with the body of the section preceded thusly: ACKNOWLEDGMENTS. — Authors should verify that persons mentioned in acknowledgments acquiesce in the wording. Appendixes: place after acknowledgments and before literature cited. Form for literature cited: Author’s last name, first name, middle initial. Year. Title. Jour- nal (see BIOSIS list of Serials with Title Abbreviations) volume (number) :pages. Provide total number of pages for books, dissertations, and theses. For second authors use initials followed by last name. Examples: Woodall, W. Robert, Jr., and J. B. Wallace. 1972. The benthic fauna in four small southern Appalachian streams. Am. Midi. Nat. ^5(2):393-407. Crocker, Denton W. and D. W. Barr. 1968. Handbook of the Crayfishes of Ontario. Univ. Ontario Press, Toronto. 158 pp. Authors, not the editor, are responsible for verifying references. Form for citing references in text: parenthetical (Woodall and Wallace 1972:401), page numbers optional, following a colon; for more than two authors use et al. (not italicized). All tables go on separate sheets at the end of the manuscript. Do not use vertical lines in tables. Indicate lightly in pencil in the margin of the original manuscript where tables and il- lustrations would best fit. Preparation of illustrations. Illustrations, including maps, graphs, charts, drawings, and photographs, should be numbered consecutively as figures. They should not be larger than 21.5 X 28 cm (81/2 X 11”). Plates must be prepared and presented as they are to ap- pear, not as groups or large sheets of items for arrangement by the editors. Do not mount in- dividual photographs. The author’s name, title of the manuscript, figure number, and the notation “Top,” should be penciled lightly on the back of every illustration. Lettering on original drawings and maps should be of adequate size to permit proper reduction where needed. Do not type on illustrations. Legends should be typed, double-spaced, on separate sheets. Avoid indicating scale as “X life size.” Consult CBE Style Manual, pp. 39-45, for more complete guidelines. Page charges, reprints and proofs. A per page charge of $20 is expected from authors who have funds available from institutions, grants, or other sources. Those without such funds should so indicate in their correspondence with the Editor. This will not affect accept- ance for normal publication. Contributors who pay full page costs will be furnished 100 free reprints. Reprint order forms will be sent with galley proofs and are to be returned to the Managing Editor. On papers with more than one author, it will be the responsibility of the correspondent to assure that other authors have an opportunity to obtain reprints. Proofs are to be corrected, signed and returned to the Managing Editor within 48 hours. Changes in proofs other than type corrections will be charged to the author. CONTENTS The Blancan Carnivore Trigonictis (Mammalia: Mustelidae) in the Eastern United States. Clayton E. Ray, Elaine Anderson and S. David Webb 1 Pseudanophthalmus from Appalachian Caves (Coleoptera: Carabidae): The Engelhardti Complex. Thomas C. Barr, Jr 37 Records of Leatherback Turtles, Dermochelys coriacea (Linnaeus), and Other Marine Turtles in North Carolina Waters. David S. Lee and William M. Palmer 95 Ecological Life History of Ptilostomis postica (Walker) (Trichoptera: Phryganeidae) in Greenbottom Swamp, Cabell County, West Virginia. Mary Beth Roush and Donald C. Tarter 107 Reproduction of the Eastern Cottonmouth Agkistrodon piscivorus piscivorus (Serpentes: Viperidae) at the Northern Edge of its Range. Charles R. Blem 117 Habitat Use and Relative Abundance of the Small Mammals of a South Carolina Barrier Island. John B. Andre 129 On the Taxonomic Status, Distribution and Subspecies of the Milliped Pseudotremia fracta (Chamberlin) (Chordeumatida: Cleidogonidae). Richard L. Hoffman 135 Reptiles and Amphibians of Kiawah and Capers Islands, South Carolina. J. Whitfield Gibbons and Julian R. Harrison, HI 145 Errata and Miscellany 163 Table of Contents, Nos. 3 and 4 (1980) 164 Index of Scientific Names, Nos. 3 and 4 (1980) 166 JUL 131981