NovitatesAMERICAN MUSEUM PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET NEW YORK, N.Y. 10024 U.S.A. NUMBER 2682 AUGUST 29, 1979

THOMAS C. BARR, JR. The , Distribution, and Affinities of , with Notes on Associated Species of (Coleoptera, Carabidae)

AMERICAN MUSEUM Novitates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2682, pp. 1-20, figs. 1-13, table 1 August 29, 1979

The Taxonomy, Distribution, and Affinities of Neaphaenops, with Notes on Associated Species of Pseudanophthalmus (Coleoptera, Carabidae)

THOMAS C. BARR, JR.'

ABSTRACT Neaphaenops Jeannel includes one polytypic spe- close common ancestry with Pseudanophthalmus of cies, N. tellkampfi, from caves of the Pennyroyal the pubescens species group, and P. princeps, new plateau and adjacent upland in west-central Ken- species (Kentucky and Tennessee) may possibly rep- tucky. Four subspecies are described and illustrated: resent an intermediate evolutionary stage. Taxonomic tellkampfi tellkampfi (Erichson), t. viator, new sub- changes are prQposed for Pseudanophthalmus ciliaris species, t. henroti Jeannel, and t. meridionalis Barr. orlindae Barr, new combination, and P. loganensis A key to subspecies and a distribution map are Barr, new status. given. It is suggested that Neaphaenops shares a INTRODUCTION The monobasic cave genus Neaphae- around the sides of the head to delimit the nops was established by Jeannel (1920) for cervicum, a facies which Jeannel (1926-1930) Anophthalmus tellkampfii Erichson (1844), a called "aphaenopsian." The species is also no- large, eyeless trechine carabid first collected by table for its unusually extensive distribution and Dr. Theodor Tellkampf on a visit to Mammoth relative abundance. This study is based on ex- Cave, Kentucky. Erichson compared it with amination of approximately 1000 specimens of Anophthalmus schmidti Sturm, described earlier Neaphaenops from 95 Kentucky caves, the in the same year from caves in Yugoslavia, and great majority of specimens collected by me placed it in the same genus. The Mammoth between 1955 and 1979. Cave species thus became the first blind Jeannel (1926-1930, 1949) placed both Nea- trechine to be discovered in North America and phaenops and Pseudanophthalmus Jeannel the world's second described species of this (1920; approximately 185 cave species and one group of cave . Unlike most other Amer- edaphobitic species in eastern United States) in ican cave trechines, Neaphaenops tellkampfi is the '7rechoblemus series" (see also Barr, unusually elongate and slender (fig. 1); with 1972) but did not comment on the precise rela- elongation of the head, the frontal grooves end tionship between the two genera. Neaphaenops blindly on the vertex instead of continuing tellkampfi has been the subject of recent eco-

'Research Associate, Department of Entomology, American Museum of Natural History; Professor of Biological Sciences, the University of Kentucky, Lexington.

Copyright © American Museum of Natural History 1979 ISSN 0003-0082 / Price $1.55 2 AMERICAN MUSEUM NOVITATES NO. 2682

0

FIG. 1. Neaphaenops t. tellkampfi (Erichson), Diamond Caverns, Barren County, Kentucky. 1979 BARR: NEAPHAENOPS 3 logical, behavioral, and genetic studies, par- plateau where it occurs (Barr, 1968; see also ticularly in Mammoth Cave National Park and Giuseffi, Kane and Duggleby, 1978). its immediate environs (Barr and Kuehne, 1971; In Mammoth Cave National Park N. tell- Kane, Norton and Poulson, 1975; Norton, Kane kampfi is the most abundant and most nearly and Poulson, 1975; Giuseffi, Kane and Dug- ubiquitous of terrestrial troglobites. Protection gleby, 1978). In contrast to most species of of habitat and encouragement of ecological re- Pseudanophthalmus (including all 12 species search by the National Park Service will proba- with which it is sympatric), Neaphaenops feeds bly result in further investigations of the heavily on eggs of the rhaphidophorine "cave- biology of this species. The present paper at- cricket" subterraneus (Scudder), tempts to relate nominate tellkampfi to other which it digs from the moist silt of cave floors geographic races of the species and to formu- and ledges where they are deposited (Barr and late a theory of the evolutionary origin of Nea- Kuehne, 1971). This behavior presumably rep- phaenops based on morphological and resents an adaptive shift away from the usual biogeographic considerations. There is no guar- generalist feeder niches occupied by antee that the ecological and behavioral char- Pseudanophthalmus species, taking advantage acteristics of N. t. tellkampfi in the Mammoth of a new food and shortening the food chain Cave region will be duplicated in all of its from the surface to the food-poor cave environ- geographic races. Comparative studies of the ment. Crickets feed outside the caves at night, other subspecies and of presumably related spe- and their guano and eggs constitute an impor- cies of Pseudanophthalmus may shed light on tant fraction of the limited food available to the the evolution of the shift in feeding strategy cave community (Park and Barr, 1961; Barr and that appears to have proven so successful in Kuehne, 1971; Barr and Stoneburner, MS). Nea- Neaphaenops. phaenops has not become so specialized that it Taxonomic treatment of cave trechines in has relinquished its feeding on other prey, but the interconnected solutional networks of the it has greatly broadened its niche by expanding Mississippian plateaus (see Barr, 1967a) is it to include cricket eggs as an abundant and complicated by cases of recent divergence. additional energy source. A close correspond- Pairs of closely similar taxa which cause inter- ence exists between the ranges of Neaphaenops pretive problems fall into four categories: (1) no tellkampfi and Hadenoecus subterraneus (Barr, overlap of ranges, usually with strong geo- 1968; Hubbell and Norton, 1978). logical evidence for extrinsic isolation; (2) With its exclusive predation strategy Nea- ranges overlapping (a) broadly (several caves) phaenops tellkampfi has far surpassed most spe- or (b) narrowly (one or two caves only), with cies of Pseudanophthalmus in abundance. In the overlap zone occupied by apparent hybrids; Mammoth Cave it is so common on damp silt and (3) ranges contiguous but not overlapping, and sand fills that cave guides and explorers no evidence of extrinsic isolation (see Barr, know it by the common name, "sand beetle." 1962a, for an interpretation of (3) based on From modified Lincoln-Peterson estimates of parapatric exclusion, also Wilson, 1975, for a local population densities in Mammoth Cave, theoretical basis for such exclusion). In Nea- Barr and Kuehne (1971) speculated that the phaenops the relationships between the four total population of the species in the Mammoth taxa include (1); (2a); and (2b), respectively Cave system alone might exceed 750,000 indi- implying (1) no gene flow; (2a) moderate gene viduals. The range of N. tellkampfi (fig. 5), flow, and (2b) very limited gene flow. In the which is more extensive than that of any other absence of genetic data I have taken a conser- North American cave trechine, is attributable to vative approach and treated all four taxa as two factors: (a) the high vagility of this large, subspecies, but it is quite possible that henroti mobile, foraging species, and (b) the openness is an allopatric sibling species and meridionalis of solutional networks in the highly cavernous is a semispecies. The wide distribution and Mississippian limestones of the Pennyroyal abundance of Neaphaenops populations offer an 4 AMERICAN MUSEUM NOVITATES NO. 2682

unusual opportunity to assess the extent of gene puncture opposite second or third umbilicate; flow between local populations of a terrestrial anterior apical puncture absent; humeri minute- troglobite, employing both conspicuous mor- ly setulose, not serrate; prehumeral borders phological characters and electrophoretic strongly oblique; apical recurrent groove ves- allozyme techniques. Such studies require a tigial. Mandibles long and slender; last segment sound taxonomic framework and a detailed of maxillary palp four-fifths as long as penulti- knowledge of the distribution of putatively in- mate segment; mentum tooth long and bifid; fraspecific taxa. Giuseffi, Kane and Duggleby mentum and submentum fused; prebasilar setae (1978) have made an interesting preliminary 2 + 6. Antenna three-fourths body length. study of six loci in populations of nominate Legs very long and slender; protibia pubescent tellkampfi; their results suggest that all their on anterior face. Aedeagus 1.28-1.45 mm. samples (several caves in the Mammoth Cave long, basal bulb large and sharply bent at right region) are from the same large gene pool. angle or more to straight, thick, apically trun- cate median lobe; basal opening and keel both ACKNOWLEDGMENTS quite small; transfer apparatus anisotopic: left (ventral) piece laterally compressed, with sub- This investigation was supported in part by parallel margins, heavily sclerotized, rounded grants from the National Science Foundation or slightly knobbed at apex; right (dorsal) piece (no. 18765 and no. GB-2011) and the National elongate, tentlike in cross-section, apex hyaline Park Service. I thank Dr. J.M. Valentine, Mr. and slightly twisted at apex, a little longer than Leslie Hubricht, Mr. C.J. Gray, and Mr. R.M. and partially enfolding left piece at base; para- Norton for contribution of specimens; and I am meres elongate and slender, with four or five grateful to many individuals who assisted me in long, apical setae. the field, especially Mr. W.M. Andrews, Mr. Neaphaenops (and its single species, N. tell- J.A. Hinton, Dr. J.R. Holsinger, Dr. R.A. kampfi) is defined by a series of patently ap- Kuehne, Dr. T.G. Marsh, Mr. R.M. Norton, omorphic characters, three of which are and Dr. S.B. Peck. generically diagnostic: (1) last segment of max- illary palp shorter than penultimate segment TAXONOMY (these segments are usually subequal in Pseud- NEAPHAENOPS JEANNEL anophthalmus and other American trechines); (2) anterior pair of supraorbital setae and punc- Neaphaenops Jeannel, 1920, p. 154. Type species, tures absent; and (3) frontal grooves not ex- Anophthalmus Tellkampfii Erichson, by original designation. Jeannel, 1931, p. 469; 1949, p. 88. tended onto sides of head. Other features are paralleled to a greater or lesser degree in vari- Size large (6.3-7.5 mm.); form elongate, ous species of Pseudanophthalmus: (a) prono- slender; subglabrous, rufotestaceous, shining, tum convex and hind angles strongly reflexed; elytral microsculpture coarsely and somewhat (b) overall dorsal pubescence reduced; (c) irregularly transverse. Head one-fourth longer humeral serrations absent, humeral setae sparse than wide, sides scarcely rounded and sparsely and minute; (d) elytra very convex with corre- pubescent to glabrous; eyes absent; frontal sponding deplanate area near scutellum; (e) loss grooves incomplete; one pair (posterior) of su- of the anterior apical puncture; and (f) rudimen- praorbital setae. Pronotum slightly longer than tation of the apical recurrent groove. wide, widest in apical one-third to one-half, Neaphaenops tellkampfi is distributed along base one-sixth wider than apex and three- the Pennyroyal plateau and adjacent upland in fourths maximum width; basolateral impres- west-central Kentucky from Meade and sions moderate, hind angles small, usually Breckinridge counties in the north, near the sharp, more or less right, and strongly reflexed. Ohio River, to Simpson and Allen counties in Elytra two-thirds to three-fourths longer than the south, near the Tennessee border. Four sub- wide, strongly convex, deplanate around species are recognized in the present study. scutellum, apically attenuate; anterior discal Published descriptions of N. t. henroti (Jeannel, 1979 BARR: NEAPHAENOPS 5

1949) and N. t. meridionalis (Barr, 1959) are subconvex; pronotum sides distinctly not diagnostic in the light of the material re- though shallowly sinuate in basal fourth cently examined. These subspecies are re- (figs. 1, 2) ...... 3 described, and a new subspecies is described Elytral striae obsolescent, intervals flat; pro- from the eastern part of the species' range. The notum sides not sinuate (fig. 3); punctures deep, discrete; southeast Meade, eastern four subspecies may be differentiated by the Breckinridge, western Hardin, and north- following key (which does not include tell- west Hart counties ...... kampfi x viator or tellkampfi x meridionalis ...... tellkampfi henroti Jeannel hybrids). 3(2). Punctures of elytral striae relatively fewer, shallower and less discrete, striae usually KEY TO SUBSPECIES OF NEAPHAENOPS impunctate in apical half; easternmost Hart, TELLKAMPFI (ERICHSON) northwest Metcalfe, and western Green counties. tellkampfi viator, new subspecies 1. Elytra with single row of fine, very short Punctures of elytral striae more numerous, pubescence on each interval, intervals flat deeper, and fairly discrete, usually extend- or feebly subconvex, striae punctate; la- ing well into apical half; southwest Hart, brum singly emarginate, vertex without southeast Edmonson, northwest Barren, median pit ...... 2 eastern Warren, and northern Allen count- Elytra virtually glabrous, intervals convex, ies .e ellkampfi tellkampfi (Erichson) striae impunctate; labrum doubly emargi- nate, vertex with median pit between frontal grooves; southern Warren, eastern Neaphaenops tellkampfi tellkampfi (Erichson) Logan, northern Simpson, northwestern Figures 1, 6, 10 Allen counties...... Anophthalmus Tellkampfii Erichson, 1844, p. 384...... tellkampfi meridionalis Barr Type locality, Mammoth Cave, Kentucky; type 2(1). Elytral striae shallow but distinct, intervals not seen. Packard, 1888, p. 73, pl. 18(1).

FIGS. 2-4. Head and pronotum of Neaphaenops tellkampfi subspecies. (2) N. t. viator, new subspecies, Brush Creek Cave, Green County, Kentucky. (3) N. t. henroti Jeannel, Sig Shacklett's Cave, Meade County, Kentucky. (4) N. t. meridionalis Barr, Hoy Cave, Simpson County, Kentucky. 6 AMERICAN MUSEUM NOVITATES NO. 2682

Neaphaenops Tellkampfi: Jeannel, 1931, p. 471, figs. 1.7 miles SE Hinesdale; Hidden River 92-103. (=Horse) Cave, in town of Horse Cave; Hogan Neaphaenops Tellkampfi Tellkampfi: Jeannel, 1949, Caves, 0.4 mile NE Ronalds Cave; Logsdon p. 89, figs. 90-98. Barr, 1959, p. 23 (in part); Valley Cave, 4 miles W Munfordville; Mam- 1962b, p. 281. moth Onyx Cave, 2 miles N Horse Cave; DESCRIPTION: Length 6.7-7.3, mean 7.0 Ronalds Cave, 2.5 miles N Cave City on N mm. Head without median pit on vertex; la- side Interstate 65 in Hatcher Valley; Three brum singly emarginate. Pronotum sides shal- Springs Cave, in Three Springs; Webb Cave, lowly sinuate before hind angles. Elytra with 0.2 mile NW Bear Wallow; unnamed cave at fine pubescence, at least at sides; striae shallow Horse Cave interchange on Interstate 65. War- but distinct, intervals feebly subconvex, punc- ren County: Crump (=Lisenby) Cave, 0.6 mile tures usually fairly deep and discrete and ex- NE Smiths Grove; Oakland Pit, 1000 feet S tending well beyond middle of elytra. Oakland School; Pruett Saltpeter Cave, 0.8 Aedeagus 1.29-1.45, mean 1.37 mm. long, left mile SE Anna. copulatory piece knobbed. MATERIAL SEEN: Over 600 specimens from DISTRIBUTION: KENTUCKY: Allen County: 58 caves. The distributional list was compiled Buchanan Cave, 0.8 mile W. Gainesville. Bar- from specimens in my private collection. ren County: Brown Cave, 2.2 miles SE Coral DISCUSSION: Nominate tellkampfi is widely Hill; Brushy Knob Cave, 1.9 miles ENE Park represented in most large museums and many City; Buck Creek Cave, 1.3 miles SE Railton; private collections, and the majority of speci- Burnet Cave, 0.5 mile W Park City; Cave City mens in those collections are from Mammoth (=Railroad) Cave, in Cave City; Crawhorn Cave or caves in the immediate vicinity of Cave, 2.3 miles NW Rocky Hill; Crystal Onyx Mammoth Cave, Cave City, or Park City Cave, 1.5 miles SW Cave City on Prewitts (=Glasgow Junction of older literature). Some Knob; Diamond Caverns, 1.7 miles NNW Park of the locations cited in the literature (for ex- City; Hanson Cave, 2.7 miles SW Hiseville; ample, by Packard, 1888) are difficult or im- Neals Chapel Cave, 1.3 miles SE Lecta; Parker possible to identify by the cave names given, Cave, 1.7 miles SSW Park City; Slick Rock which have changed with time; however, the Cave, 0.3 mile W Slick Rock on S bank Bea- distributional data presented above are far more ver Creek; Twyman Cave, 1.4 miles N Hise- complete and almost certainly include all the ville; Vance Cave, 0.4 mile NE Park City; older locations within the boundaries delimited. Walnut Hill Cave, 1.7 miles S and slightly W In Mammoth Cave the beetles are commonly Park City; Wonderland Cave, 1.5 miles SW observed abroad on silty floors and ledges in Turleys Corners. Edmonson County: Mammoth damp spots throughout the system, absent only Cave (type locality, also extending under parts in the driest and dustiest sections. They are of Barren and Hart counties); other caves in occasionally found eating eggs or first instar Mammoth Cave National Park as follows: nymphs of cave crickets or dead or disabled Blowing Spring, Bluff, Cathedral (=Buzzard), individuals of their own species. The nymphs Dixon, Gothic, Great Onyx, Hickory Flat, Jim, are invariably very pale and have probably Little Beauty, Little White, Long, Martin, been captured as they emerged from the egg. Pagoda, Proctor, Running Branch, Smith Val- As noted by Jeannel (1949), the beetles occur ley, White, Y Camp, also spring cave above in much drier microhabitats than those fre- River Styx outlet, also freshly opened sinkhole quented by sympatric Pseudanophthalmus spe- in Woolsey Valley; Coach (=Hundred Domes) cies; however, "drier" in this context means an Cave, 1.8 miles WNW Park City; James Cave, atmosphere not saturated with water vapor but 1.3 miles WNW Park City; Short Cave, 2.1 with relative humidities still well above 90 per- miles NNW Park City. Hart County: Bald cent. Beetles also occur, usually at much lower Knob Pit, 1.8 miles W and slightly N Hardy- population densities, in the lowest levels of the ville on W side Bald Knob; Barnes Smith system on mud and silt deposits along the Cave, 3.4 miles N Hinesdale; Bert Burd Sink, banks of the cave rivers, where they feed on 0.85 mile E Seymour; Buckner Hollow Cave, small, limicolous oligochaetes and possibly 1979 BARR: NEAPHAENOPS 7 other small invertebrates. High population den- cur from May to October (Barr, unpublished sities were observed in stream passages in data). Details of energy flow between aquatic Brown Cave, Barren County, and Buckner Hol- and terrestrial components of the cave commu- low Cave, Hart County. nity are far from clear, but Neaphaenops (and Larvae and pupae are found, usually under Pseudanophthalmus menetriesi) abound in areas silt-embedded rocks, in the drier microhabitats, of high TOC input. At the very least one can where the frequency of juvenile stages peaks in conclude that the propensity of N. tellkampfi late winter and early spring. Norton et al. for feeding on cave cricket eggs has not en- (1975) have correlated seasonal peaks in abun- tirely replaced its ability to utilize other food dance of juveniles and teneral adults with the sources and thus to carve out subniches that are fall peak in Hadenoecus oviposition; the less optimal than the principal niche occupied autumn abundance of food provides a rich en- by the majority of populations of the species. ergy source for egg production in female Nea- Where suitable microclimatic conditions pre- phaenops, leading to a spring peak in juveniles vail, N. tellkampfi may occur very close to and a summer peak in tenerals. The life span is entrances of caves, even in the twilight zone. probably one to two years. Winter and early In Cave City Cave, Barren County, specimens spring samples contain slightly more females were observed walking about in daylight at the than males; Norton et al. (1975) suggested that bottom of the entrance sink, where a stream this is the result of greater longevity in fe- crosses the floor. Roger Sperka dug open a males. small sinkhole in Woolsey Valley, Mammoth Although I believe that most local N. t. Cave National Park, on a warm day in Novem- tellkampfi populations follow the life history ber 1974, and found both Neaphaenops tell- pattern described by Norton, Kane and Poulson kampfi and Pseudanophthalmus menetriesi (1975) there appear to be other ecologies and (Motschulsky) within 50 cm. of the surface, other patterns, in all probability based on other where they had probably been attracted by prey. Ludwig (1950) predicted on theoretical abundance of food. grounds that selection would favor invasion of Neaphaenops t. tellkampfi is sympatric with less-than-optimal subniches, and the ecological five species of Pseudanophthalmus, all of versatility of N. tellkampfi can be viewed as a which occur in the Mammoth Cave system. possible example of the Ludwig effect. The Three species-P. menetriesi, P. striatus low density populations of base level stream (Motschulsky), and P. pubescens (Horn)-are conduits are likely to be quite different, but relatively abundant throughout most of its annual flooding makes continued observation range. Two smaller species-P. audax (Horn) throughout the year difficult, if not impossible. and P. inexpectatus Barr-are rather rare and Domepit populations are easier to follow. A quite limited in distribution (Barr, 1962b). All series of 45 beetles collected February 10, six trechine species coexist in White Cave (0.5 1979, in Cathedral Domes, Mammoth Cave (a mile southwest of the Historic Entrance to wet microhabitat), included 19 tenerals (42%), Mammoth Cave), although they have never indicating an oviposition peak about June 1978. been taken in the same season. To the east, N. On the same date a visual sample of Neaphae- t. tellkampfi intergrades with N. t. viator, and nops in Sophie's Avenue and the Radio Room to the south it intergrades with N. t. merid- (areas typical of the microhabitat category stud- ionalis. ied by Norton, Kane and Poulson) included only eight tenerals out of a total of 171 speci- Neaphaenops tellkampfi viator, new mens (4.7%). The principal energy sources at subspecies Cathedral Domes are rotting wood of old stairs Figure 2 and moderately high organic content in the stream of water falling into the Dome. From ETYMOL,OGY: Latin viator, "traveler." total organic carbon (TOC) analyses of domepit DESCRIPTION: Length 6.7-7.3, mean 7.0 water elsewhere in Mammoth Cave, I deter- mm. Head without median pit on vertex; la- mined that peak TOC values (7-13 p.p.m.) oc- brum singly emarginate. Pronotum sides shal- 8 AMERICAN MUSEUM NOVITATES NO. 2682 lowly sinuate in basal fourth before hind is no sharp break between the ranges of the two angles, as in nominate tellkampfi. Elytra with subspecies. Some of the specimens from fine pubescence; striae shallow but distinct, in- Crump Cave, Hart County, have a minute de- tervals feebly subconvex, punctures quite shal- pression in the middle of the vertex, but it is low, fewer and less discrete than in nominate much smaller than that which occurs in tell- tellkampfi, punctuation seldom extending kampfi meridionalis. beyond middle of elytra. Aedeagus 1.28-1.43, mean 1.38 mm. long, similar to that of tell- Neaphaenops tellkampfi henroti Jeannel kampfi tellkampfi, left copulatory piece slightly Figure 3 knobbed at apex. TYPE SERIES: Holotype male (American Mu- Neaphaenops Tellkampfi subsp. Henroti Jeannel, seum of Natural History) and 18 paratypes, 1949, p. 90. Type locality, Sig Shacklett's Cave, Brush Creek Cave, 0.8 mile east and slightly Meade County, Kentucky; type in Museum Na- north of Lobb on the west side of Brush Creek, tional d'Histoire Naturelle, Paris (not seen). in western Green County, Kentucky, September Neaphaenops tellkampfi tellkampfi: Barr, 1959, p. 23 28, 1963, T.C. Barr, J.R. Holsinger, and R.M. (in part). Norton. DESCRIPTION: Length 6.6-7.5, mean 7.0 MEASUREMENTS: Holotype, total length 7.2 mm. Head without median pit on vertex; la- mm., head 1.24 mm. long x0.99 mm. wide, brum slightly emarginate. Pronotum sides not pronotum 1.33 mm. long xl.30 mm. wide, sinuate before hind angles (except for minute elytra 4.12 mm. long x2.48 mm. wide, an- emargination at site of posterior marginal seta, tenna 4.99 mm. long. also present in other subspecies). Elytra with DISTRIBUTION: KENTUCKY: Green County: fine pubescence; striae obsolescent, intervals Brush Creek Cave (type locality); Aetna Cave, flat, punctures fairly deep and discrete, usually 1.3 miles E Eve; Camp Branch (=Gentry) extending well beyond middle of elytra. Cave, 1.0 miles NE Grab just SE forks Camp Aedeagus 1.28-1.40, mean 1.36 mm. long, Branch; Milby Cave, 1.8 miles NW Gabe; Salt- about as in nominate tellkampfi, left copulatory peter Cave, 1.2 miles NE Whickerville on W piece slightly knobbed at apex. side Little Barren River; Scott Cave, 1.9 miles DISTRIBUTION: KENTUCKY: Breckinridge ESE Eve on E side Brush Creek. Hart County: County: Glass Cave, 1.2 miles E Custer; Peni- Crump Cave, 1.6 miles W Eve in E wall valley tentiary Cave, 1.4 miles E Clifton Mills; Thorn- of Green River; Turner Cave, 3.7 miles SE hill Cave, 2.1 miles WSW Big Spring. Hardin Magnolia. Metcalfe County: Devils Den, 3 County: Belt Cave, 3.5 miles NW Howe Val- miles S Center in deep sink W North Metcalfe ley; Bland Cave, 1.1 miles NE Spurrier in School. Akers Valley; Patterson Cave, 2.5 miles NW MATERIAL SEEN: Eighty-seven specimens, Stephensburg; Saltpeter Cave, 2.0 miles S and including the type series of 19 beetles, from a slightly W Flaherty; Turkey Hollow Cave, 1.8 total of nine caves. miles S and slightly E Old Stephensburg; Pick- DISCUSSION: This subspecies is close to handle Wilmoth Cave, 2.1 miles SW Franklin nominate tellkampfi but differs consistently in Crossroads; Wonderland Cave, 1.5 miles SE the feebler punctuation of the elytral striae. Old Stephensburg. Hart County: Cooch Webb Specimens from Bald Knob Pit and Bert Burd Cave, 2.3 miles W Priceville on Ky. 728; Sink, adjacent to the range of tellkampfi viator Copelin Cave, 2 miles E Millerstown; Puckett as given above and listed as nominate tell- Cave, 0.65 mile WSW Priceville; Rough Cave, kampfi, are transitional in punctuation, but two 2.6 miles WSW Priceville on Ky. 728; Saltpe- late tenerals from Three Springs Cave, which is ter (=Mushroom) Cave, 1.5 miles NE Lone 3 miles east of Bert Burd Sink, are strongly Star; Riders Mill Cave, 2.5 miles N Priceville punctured. Intergradation between t. tellkampfi on Roundstone Creek. Meade County: Sig and t. viator occurs gradually in eastern Hart Shacklett's Cave (type locality), 3.0 miles N County along a zone about 5 miles wide; there Big Spring in Stith Valley, E corner intersec- 1979 BARR: NEAPHAENOPS 9 tion Ky. 1238 and Ky. 1735; Scott Cave, near labrum doubly emarginate. Pronotum sides not Guston (Jeannel, 1949). sinuate before hind angles. Elytra virtually MATERIAL SEEN: Eighty-nine specimens glabrous, 0.06-0.10 wider than in other sub- from 17 caves, including 13 topotypes from Sig species; intervals convex, impunctate. Shacklett's Cave. Aedeagus 1.28-1.36, mean 1.32 mm. long, left DISCUSSION: Jeannel (1949) described this copulatory piece not knobbed, simply rounded subspecies as having more pronounced humeral at apex. angles, the prehumeral borders supposedly DISTRIBUTION: KENTUCKY: Allen County: being less oblique than in t. tellkampfi. This is Jack Johnson (=Howell) Cave, 0.75 mile SSE not the case. Previously I confused Hart and Claypool in hollow S Bays Fork. Simpson Hardin county series of henroti and t. tell- County: Bennett Cave, 2.0 mile WNW Gold kampfi (Barr, 1959). Although the subspecies is City on W side Lick Creek, 800 feet E 1-65 here diagnosed on a quite different basis from (J.A. Hinton, legit). Hoy Cave (type locality), that originally proposed, there are no constant 1.9 miles N county courthouse in Franklin, un- differences between the Meade-Breckinridge der US 31W; Old Smoky Cave, 1.4 miles ENE and Hart-Hardin series, and the available name Salmons; Slates Cave, at Spout Spring 1.8 henroti applies to all of them. miles E slightly N Gold City. Warren County: As with viator and meridionalis, the series Blue Level Cave, 2 miles SSE Blue Level of henroti are smaller than those available for (C.J. Gray, legit); Bypass Cave, in south nominate tellkampfi. At least part of this is the Bowling Green between Smith Drive and US result of greater homogeneity of habitat in the 31W bypass (cave now closed); Horseshoe smaller caves inhabited by the peripheral sub- (=Vales) Cave, in C.W. Lampkin Park in west species. Most of the northern caves in which Bowling Green; McGinnis Cave, 2.5 miles SW henroti was collected are wet or damp. Dry Bowling Green (J.M. Valentine, legit); Moats microhabitats similar to those in the upper lev- Cave, 0.8 mile W Claypool; Plano Saltpeter els of the Mammoth Cave system were present Cave 0.5 mile SSW Plano (C.J. Gray, legit). only in Belt, Thornhill, and Wonderland caves, Logan County: Wheeler Cave, 2.0 miles NE Hardin County. In Bland Cave, Riders Mill South Union (R.M. Norton, legit). Cave, and Sig Shacklett's Cave the Neaphae- MATERIAL SEEN: One hundred forty-seven nops were collected on silt banks in stream specimens, including the type series of 24 from passages. Hoy Cave, from a total of 11 caves. A prominent sandstone ridge and complex DISCUSSION: This subspecies is readily rec- fault system run transversely across Hart ognized by the vertex pit and the broad, County at the southern limit of the range of glabrous elytra with convex intervals and im- henroti. Close morphological similarity be- punctate striae; the labrum is distinctly trilobed; tween henroti, viator, and nominate tellkampfi mean total length and aedeagal length are argue for treatment of henroti as a subspecies, slightly smaller; pronotum sides are convergent but there are no known cases of intergradation rather than sinuate; and the left copulatory with t. tellkampfi or t. viator, consequently piece is not knobbed at the apex. Traces of a henroti could well be an allopatric sibling spe- vertex pit (more feeble than in meridionalis) cies. were seen in a few specimens from series of nominate tellkampfi in Pruetts Saltpeter Cave Neaphaenops tellkampfi meridionalis Barr and tellkampfi viator in Crump Cave. There are Figures 4, 7, 11 inaccuracies and omissions in the original de- Neaphaenops tellkampfi meridionalis Barr, 1959, p. scription of meridionalis (Barr, 1959): the total 23. Type locality, Hoy Cave, Simpson County, length range given in 1959 does not include Kentucky; type in the American Museum of Nat- outstretched mandibles and is thus too low, the ural History. hind angles are not smaller as stated, and the DESCRIPTION: Length 6.3-7. 1, mean 6.8 striae are not "less profound" but actually mm. Head with distinct median pit on vertex; deeper. 10 AMERICAN MUSEUM NOVITATES NO. 2682

Despite its very distinct appearance, N. t. habitat was made in Bear Cave (August 1963), meridionalis does apparently hybridize with t. when a Neaphaenops was found eating a small tellkampfi in a narrow zone at the southeast chrysomelid beetle that had washed into the margin of its range, along Trammel Fork and cave through an upper sinkhole entrance. Bays Fork in Warren and Allen counties. Hadenoecus subterraneus coexists with N. t. Elsewhere the approximate geographic bound- meridionalis in all the caves where it has been ary between the two taxa is the Barren River. found. In the laboratory meridionalis specimens A series of 18 specimens from Friendship dug deeply into artificial "egg holes" made by Cave, 1.3 miles southeast of Alvaton, Warren poking an H. subterraneus ovipositor into County, is compared with "typical" t. tell- damp sand, a behavior also observed by Nor- kampfi (Diamond, Pruett) and meridionalis ton, Kane and Poulson (1975) in nominate tell- (Hoy, Plano) in table 1. The hybrid population kampfi. has the greater length of t. tellkampfi and re- sembles meridionalis in the broader elytra and DISTRIBUTION left copulatory piece, but is intermediate be- tween these two taxa in six other characters. A The distribution of Neaphaenops tellkampfi single male from Bear Cave, 3 miles southeast (fig. 5) is essentially coextensive with distribu- of Friendship Cave but in Allen County, is tion of outcrops of the Girkin (or Paoli to the similar (transfer apparatus, fig. 11). Gene flow north), Ste. Genevieve, and St. Louis lime- between the two taxa is probably very limited, stones, all highly cavernous strata. The species and meridionalis may well be a semispecies, range is effectively bounded by the Big Clifty although a conservative treatment is adopted sandstone member of the Golconda formation here. to the west and the Salem and Warsaw (or Microhabitats selected by N. t. meridionalis Harrodsburg in the north) limestones to the do not seem to include the drier areas fre- east. As a troglobite, Neaphaenops is limited to quented by nominate tellkampfi; this may be a subterranean dispersal routes through reason- real behavioral difference, or it may be only ably continuous limestone; however, small the result of greater homogeneity of habitat streams (for example, the Green River) seldom afforded by the shallow, extensive cave sys- constitute barriers to trechine dispersal if there tems which it occupies. Invariably it has been are numerous caves on both sides of the taken in wet areas near water (pools or streams (Barr, 1959; Barr and Peck, 1965). streams); in riparian microhabitats it occurs on Neaphaenops distribution is apparently further walls or ledges or silt banks well above the controlled by the distribution of Hadenoecus stream and not at its edge. The only feeding subterraneus, which in turn appears dependent observation on this subspecies in its natural on moderately high cave density, so crickets

TABLE 1 Comparison of Diagnostic Characters in Neaphaenops tellkampfi tellkampfi, Neaphaenops tellkampfi meridionalis and Presumed Hybrids

N. l. tellkampfi tellkampfi x meridionalis N. t. meridionalis (Friendship Cave) Total length 6.7-7.3, mean 7.0 mm 6.9-7.4, mean 7.1 mm 6.3-7.1, mean 6.8 mm Vertex pit Absent Present but feeble Deep Labrum Singly emarginate Feebly trilobed margin Doubly emarginate Pronotum sides Sinuate Weakly sinuate Not sinuate Elytral pubescence Fine but distinct Fine at sides Absent Elytral intervals Nearly flat Subconvex Strongly convex Elytral striae Finely punctured Punctulate Impunctate Elytral LIW index 1.48-1.71, mean 1.63 mm. 1.41-1.69, mean 1.58 mm. 1.54-1.65, mean 1.59 mm. Left copulatory piece Knobbed at apex Apex bluntly rounded Apex bluntly rounded 4.

'I

N D A N A

0 TEL LKAMPF/

a 0 VIA TOR

. U 0 HENROTr

* A MERIDIONAL/S * R E C K R D G E -U0.

T E N N E S S E E FIG. 5. Distribution of Neaphaenops tellkampfi in west-central Kentucky. N. t. tellkampfi x N. t. meridionalis hybrids shown by circled triangles. Small crossed circles represent county seats. Approximate location of St. Louis/Salem and Warsaw contact shown by heavy dashed line; data from United States Geological Survey 71/2' geologic maps. 12 AMERICAN MUSEUM NOVITATES NO. 2682 can disperse from cave to cave. Locally the by way of "microcaverns"-small spaces Salem-Warsaw unit contains a non-cavernous, around tree roots, under rocks, in talus piles, arenaceous to argillaceous siltstone member and so forth-in the soil. The Hart County which may pose a stratigraphic barrier. The ridge also marks the range boundary between few Salem-Warsaw caves occupied by Nea- two species of troglobitic crayfishes, Or- phaenops and Hadenoecus are close to the mar- conectes pellucidus (Tellkampf) and 0. inermis gin of the St. Louis outcrops. Salem-Warsaw Cope (Hobbs and Barr, 1972). A distinct geo- caves are generally smaller and less frequent, graphic race of Pseudanophthalmus menetriesi and the dispersal potential of Hadenoecus and occurs north of the ridge, but two other widely Neaphaenops is correspondingly reduced (Barr, distributed species in the Mammoth Cave re- 1968). Distributional control is probably exerted gion, P. striatus and P. pubescens, have not directly on Hadenoecus subterraneus and indi- penetrated north of the ridge, and P. barberi rectly on Neaphaenops, which is dependent on Jeannel occurs only in caves to the north of it. cricket eggs for a large part of its diet. Several Despite its allopatric distribution, t. henroti is species of Pseudanophthalmus (of the men- morphologically closer to t. tellkampfi and t. etriesi and pubescens groups) range widely viator than is t. meridionalis, yet nominate tell- through many of the Salem-Warsaw caves kampfi and t. meridionalis hybridize in at least southeast of the species boundary of N. tell- one point where their ranges meet. The distri- kampfi, but the caves are devoid of H. subter- butional and geologic evidence indicating that raneus populations. the ridge is an effective barrier to cave beetle Intergradation occurs in geographically inter- dispersal leaves open the distinct possibility mediate populations between (a) t. tellkampfi that both N. t. henroti and the northern race of and t. meridionalis and (b) t. tellkampfi and t. P. menetriesi are allopatric sibling species. viator. Hybridization is certainly secondary in Neaphaenops t. henroti is probably derived the first case and possibly also in the second, from a peripheral population of t. tellkampfi where intergradation is broader and more grad- that worked its way northward through the ual. In contrast, gene flow between t. henroti small, scattered cave systems of the Hart and either of the subspecies to the south is County ridge. The two races may be descended probably severely restricted by a stratigraphic from non-cave populations of ancestral tell- and structural barrier, a heavily faulted ridge kampfi that independently colonized caves north with considerable thicknesses of sandstone and south of the ridge, although close mor- which transversely crosses Hart County (Barr, phological similarity argues against this. Most 1968). Only a few small, isolated caves occur cave trechines are believed to have descended in this area, and if intergrading Neaphaenops from soil- and humus-dwellers that were wide- populations exist, they have not yet been dis- spread near the surface during the cooler, wet- covered. The little caves, some of which are ter climatic regimes of glacial maxima sandstone crevices, are not barriers to (Jeannel, 1949; Barr, 1967b). In any case, dis- Hadenoecus subterraneus, which, unlike Nea- tribution of t. henroti at the northern part of its phaenops, does not require direct connection range, where it seems not to have occupied all between caves, only a reasonably high density the cave areas accessible to it, suggests a of them. Dawson Hollow Cave, 1.5 miles southern origin and gradual subterranean dis- south-southwest of Bonnieville, Hart County, is persal northward. It has not penetrated as far as in the ridge barrier and can serve as an exam- caves along the Ohio River in western Meade ple. It contains a population of thousands of H. County, nor as far west into caves of subterraneus, but no trechines; other terrestrial Breckinridge County as it theoretically could invertebrates include linyphiid spiders and col- have gone. These caves are occupied by lembolans (Sinella, Tomocerus) belonging to Pseudanophthalmus barberi and by Hadenoe- widely distributed, opportunistic species which cus subterraneus, but not by Neaphaenops. can probably disperse through non-cave areas The easternmost subspecies, N. t. viator, 1979 BARR: NEAPHAENOPS 13 has penetrated about as far as it can go without support elevation of meridionalis to semispecies moving into cave systems which are (a) beyond status, but I have retained a conservative treat- the St. Louis limestone outcrops and (b) devoid ment pending availability and analysis of such of Hadenoecus subterraneus. An undescribed data. Pseudanophthalmus species (menetriesi group), whose western species boundary closely coin- cides with that of N. t. viator, is moderately AFFINITIES abundant in all the viator caves and also occurs If one is to look for the affinities and per- farther east in Green County. This suggests that haps the origin of Neaphaenops, it would seem supposed physical barriers to dispersal in this reasonable to begin with Pseudanophthalmus. area may be less important than the absence of Three other genera of troglobitic trechines in- Hadenoecus subterraneus. habit caves of the Cumberland plateau margin Neaphaenops t. meridionalis, the south- in southeastern Kentucky (Valentine, 1952), but ernmost and morphologically the most aberrant Darlingtonea and Ameroduvalius belong to a subspecies, occupies a range about the same different phyletic series (Barr, 1972), and size as that of t. viator. It coexists with two is a fairly clear derivative, non- species of Pseudanophthalmus of the pubescens aphaenopsian, from the robustus-intermedius group and overlaps the range of P. menetriesi section of Pseudanophthalmus. Within the very at Bowling Green. Its eastern boundary is lim- large genus Pseudanophthalmus the most prom- ited by that of nominate tellkampfi, its northern ising line of attack is comparison of male geni- boundary by non-caverniferous sandstones, and talia. Primitively in Pseudanophthalmus there its southeastern boundary approximately by the were two copulatory pieces placed asym- St. Louis-Salem and Warsaw contact. It is not metrically on edge within the internal sac [a clear why this subspecies has not been taken a condition which Jeannel (1926-1930) has called little farther west and south in the Pennyroyal. "anisotopic"]. In most species groups this Throughout its range it is sympatric with P. transfer apparatus is much the same throughout loganensis Barr, which also occurs in great the group, although there are considerable per- abundance throughout the numerous caves of mutations of the basic form in the tenuis and Logan County, Kentucky, and northern Robert- audax groups. The basic form of the aedeagus son and Sumner counties, Tennessee. and its transfer apparatus are normally rather Hadenoecus subterraneus also inhabits many of conservative and of diagnostic utility in distin- these same caves. There are no Neaphaenops- guishing related groups of species. like trechine species there to preempt the niche The aedeagus of Neaphaenops tellkampfi of N. tellkampfi and thus bar it through com- consists of a large basal bulb bearing a small petitive exclusion. The hypothesis that merid- keel and small opening, bent sharply at a right ionalis is of comparatively recent origin and angle or more to the thick, scarcely tapered, has had insufficient time to expand its range and apically truncate median lobe (figs. 6, 7). westward is not wholly satisfactory. The degree The transfer apparatus consists of a heavily of the subspecies' morphological departure sclerotized, flattened, somewhat spatulate left from nominate tellkampfi and its secondary piece, rounded or slightly knobbed at the apex, hybridization with the latter indicate that it is and an elongate, folded right piece which has an older isolate than either henroti or viator. the shape of an inverted V in cross-section and Perhaps the final explanation-or at least a is slightly knobbed and twisted at its thinly more palatable one-could be learned through a sclerotized, hyaline apex. The left piece is comparative study of the ecology and behavior partly enfolded by the right piece near its base of t. tellkampfi and t. meridionalis. The nar- (figs. 10, 11). Among the 20 species groups of rowness of the zone of hybridization implies Pseudanophthalmus the closest parallel in very limited gene flow between these two taxa. aedeagal shape is found in the pubescens I suspect that genetic (allozyme) data would group, especially in P. pubescens itself (fig. 8), 14 AMERICAN MUSEUM NOVITATES NO. 2682

1.0 MM

FIGS. 6-9. Aedeagi of Neaphaenops and Pseudanophthalmus species, left lateral view. (6) N. t. tellkampfi (Erichson), Crump Cave, Warren County, Kentucky. (7) N. t. meridionalis Barr, Hoy Cave, Simpson County, Kentucky. (8) P. pubescens (Horn), Crump Cave, Warren County, Kentucky. (9) P. princeps, new species, Plano Saltpeter Cave, Warren County, Kentucky. where there is the same sharply deflexed basal less developed than in other species of the bulb and truncate apex. The transfer apparatus group (compare fig. 10). in all species of the pubescens group consists The male genitalia, then, offer evidence for of a broader, folded right piece, the apex of derivation of Neaphaenops from the pubescens which is often twisted and slightly knobbed group. There are three additional features, no (least so in pubescens), and a heavily spin- one of which alone suffices to establish affinity, ulose, rodlike left piece partly enfolded by the but taken together with the genitalic evidence right piece (fig. 12). The pubescens and Nea- establish a strong case for joint derivation of phaenops left pieces could both be derived the pubescens group and Neaphaenops from a from the simple left piece of the menetriesi common ancestor. (1) The anterior discal seta group, in pubescens becoming spinulose and in on the third elytral interval is normally placed Neaphaenops becoming flatter, blunter, and at or behind the level of the fourth umbilicate more heavily sclerotized. The right piece in (humeral) puncture in Pseudanophthalmus, but Neaphaenops is less folded than in the in Neaphaenops and the pubescens and men- pubescens group, but it does have what may be etriesi groups the anterior discal is placed op- a vestige of an apical knob. In P. pubescens posite the second or third umbilicate. (2) The the right piece has basically the same structure elytral microsculpture in both Neaphaenops and as that of Neaphaenops, and the apical knob is the pubescens group is the same: neither iso- 1979 BARR: NEAPHAENOPS 15

. ,. .

12

FIGS. 10-12. Transfer apparatus of Neaphaenops and Pseudanophthalmus species, left lateral view. (10) N. t. tellkampfi (Erichson), Crump Cave, Warren County, Kentucky. (11) N. t. meridionalis x t. tellkampfi, Bear Cave, Allen County, Kentucky. (12) P. pubescens (Horn), Smith Valley Cave, Edmonson County, Kentucky.

diametric nor finely transverse, but consisting coextensive with part of the highly cavernous of an irregular meshwork of transversely elon- Pennyroyal plateau, the continuity broken only gated polygons (finely transverse in the men- by the Hart County ridge. etriesi group). (3) The geographic distributions Species of five groups of Pseud- of Neaphaenops and the pubescens group over- anophthalmus satisfy the requirement of prox- lap widely in the south Pennyroyal plateau. The imity to the range of Neaphaenops in order to ancestral population from which Neaphaenops qualify as possible joint descendants from a evolved is likely to have existed at one time in common ancestor: the tenuis, audax, inexpec- or quite near some part of the present distribu- tatus, menetriesi, and pubescens groups. There tion of N. tellkampfi. If Neaphaenops evolved is little to suggest that the first three groups from an epigean ancestor, then its present dis- could be closely related to Neaphaenops, but tribution might include several species in differ- the menetriesi group has the anterior discal ent cave areas, because dispersal of the puncture far advanced, opposite the second or ancestor would not have been limited by the third umbilicate puncture, and the apical recur- distribution of continuous cavernous limestone. rent groove is vestigial. Furthermore, P. men- If Neaphaenops evolved from a cave ancestor, etriesi resembles Neaphaenops in the non- which I believe more probable, then its disper- pruinose microsculpture, shallow striae with sal potential would be limited by local stratigra- small, discrete punctures, and (in the Mammoth phy and structure (Barr, 1968). As shown in Cave area only) reduction of dorsal pubescence; figure 5, the distribution of Neaphaenops is all of these are features in which Neaphaenops 16 AMERICAN MUSEUM NOVITATES NO. 2682 differs from members of the pubescens group. trechines that live along the edges of streams These observations indicate that Neaphaenops among leaves and sticks washed underground is probably also related to the menetriesi group, are most likely to have heavy infestations of but because of the microsculpture and the shape Laboulbenia subterranea Thaxter. Selection of of the aedeagus and transfer apparatus, less drier microhabitats by ancestral Neaphaenops closely than it is to the pubescens group. The would make pruinose microsculpture unneces- suggested relationship is origin of the sary and subject to rudimentation if this hy- pubescens and menetriesi groups from a com- pothesis is correct. mon stock, with the Neaphaenops line branch- The theory just presented, that Neaphaenops ing off near the base of the pubescens line. evolved as a cave species and shares a common The pubescens group includes a dozen or ancestry with the pubescens group and a more species (several undescribed) which range slightly more remote relationship to the men- along the Pennyroyal from the Mammoth Cave etriesi group, implies a southern origin in the region south, then westward to Crittenden and Mammoth Cave or Bowling Green area. Jean- Livingston counties, Kentucky. Most of the nel (1949) argued that Neaphaenops is con- species are rather robust, pubescent, and pru- vergent in many characters with Trechopsis inose (when viewed under an oblique light, the lapiei (Peyerimhoff), a European nivicolous elytral disc takes on a frosted appearance as trechine, and that consequently its ancestor light is reflected from many microtrichia pro- must have been a periglacial species that colo- jecting up from the posterior edge of each cu- nized caves following retreat of Illinoisian ice. ticular polygon). The pronotum is typically This view ignores the absence of Neaphaenops transverse-subquadrate, humeri are serrate and from southern Indiana caves and the observa- setose, and the elytral striae are (usually) im- tion that it seems to be in the process of ex- punctate or feebly punctulate. The aedeagus is panding its range northward. Furthermore, the remarkably similar in all species, varying prin- northern origin required by the Jeannel hypoth- cipally in size, the degree of flexure of the esis becomes awkward when it is seen that N. basal bulb, and in minor details of the two t. meridionalis, the southernmost subspecies, is copulatory pieces. The majority of the species morphologically the most aberrant and thus select riparian habitats, appearing in higher lev- probably the oldest isolate. els of caves chiefly during unusually wet weather (McKinney, 1975). Alongside streams, NOTES ON THE PUBESCENS GROUP OF however, they may be exceptionally abundant. PSEUDANOPHTHALMUS Neaphaenops tellkampfi differs from most species of the pubescens group in the larger, Among species of the pubescens group more slender body with elongate appendages; which have been described, there is little to incomplete frontal grooves and loss of anterior suggest any evolutionary tendency in the direc- supraorbitals; smooth humeral margin; reduc- tion of Neaphaenops with the possible excep- tion of elytral pubescence and striae; loss of tion of P. ciliaris orlindae Barr (1959), from pruinose microsculpture; loss of anterior apical caves of the upper Red River valley along the puncture and apical recurrent groove; simpler Kentucky-Tennessee border. This species is in transfer apparatus; and retention of strial punc- the same size range as P. pubescens (5-6 tures. All of these characters are probably ap- mm.) but is a bit more slender and elongate, omorphic except the last two. Pruinose with transverse-cordiform pronotum. However, microsculpture is present in some (undescribed) recent study of trechines collected from caves species of the menetriesi group but absent in near Franklin, Kentucky, has revealed two spe- menetriesi and striatus; its adaptive significance cies of the pubescens group coexisting with is unknown, but may possibly be related to Neaphaenops at the southern limit of its range. resistance to infection by the ectoparasitic as- One species, P. loganensis Barr (1959), is comycetes of the order Laboulbeniales. Cave small (4.3-5.0, mean 4.4 mm.) and depressed. 1979 BARR: NEAPHAENOPS 17

The other species is large, elongate, and recurrent groove subparallel, joining or directed slender, with markedly convex elytra and cordi- toward apex of fifth stria slightly in advance of form pronotum. I was initially unable to distin- anterior apical puncture. Antenna rather long, guish it from Neaphaenops while collecting in seven-tenths body length. Aedeagus 1.06-1.23, the caves, so close is the superficial re- mean 1.14 mm. long, thick and arcuate semblance. This species, described below, is throughout, basal bulb not strongly deflexed as morphologically intermediate between the in pubescens; left copulatory piece apically at- robust, often depressed species of the tenuate in lateral view and densely spinulose, pubescens group and the large, convex, slender right piece with broad, apically truncate knob; habitus of Neaphaenops. It may represent a parameres with four long apical setae, some- stage through which ancestral Neaphaenops times shorter fifth seta. passed in its evolution from a pubescens-like TYPE SERIES: Holotype male (American Mu- stock. seum of Natural History) and six paratypes, Hoy Cave, 1.9 miles north of county court- house in Franklin under US 31W, Simpson Pseudanophthalmus princeps, new species Figures 9, 13 County, Kentucky, July 21, 1963, T.C. Barr. MEASUREMENTS: Holotype, total length 6.2 ETYMOLOGY: Latin princeps, "first, chief, mm., head 1.02 mm. long xO.87 mm. wide, ruler. " pronotum 0.96 mm. long x1.05 mm. wide, DIAGNOSIS: Distinguished from all other spe- elytra 3.19 mm. long xl.92 mm. wide, an- cies of the pubescens group by large size, nar- tenna 4.25 mm. long, aedeagus 1.08 mm. long. row head and pronotum, and strongly convex DISTRIBUTION: KENTUCKY: Simpson elytra. County: Hoy Cave (type locality); Old Smoky DESCRIPTION: Length 5.7-6.8, mean 6.2 ± Cave, 1.4 miles ENE Salmons; Slates Cave, at 0.3 mm. Form rather slender and elongate, Spout Spring 1.8 miles E and slightly N Gold pubescent, shining rufous to castaneous-testa- City; Steele Caves, 3 miles SE Franklin on E ceous. Head elongate, one-fifth longer than bank West Fork of Drakes Creek. Warren wide; mandibles large, long, and slender; la- County: Plano Saltpeter Cave, 0.5 mile SSW brum distinctly and doubly emarginate; frontal Plano (C.J. Gray, legit). TENNESSEE: Sumner grooves complete; dorsum of head sub- County: Whiteoak Cave, 2.1 miles ENE glabrous. Pronotum transverse-cordiform, nine- Mitchellville near mouth of Grace Creek. tenths as long as wide, width at apex and base MATERIAL SEEN: Thirty-eight specimens, in- subequal and about seven-tenths greatest width, cluding the type series, from a total of six which occurs in apical fourth; sides arcuate caves. apical half then convergent, shallowly but dis- DISCUSSION: Pseudanophthalmus princeps is tinctly sinuate in basal eighth, hind angles clearly a member of the genus Pseud- small and sharp; basolateral impressions moder- anophthalmus and of the pubescens group, but ate, minute secondary angles present on base; its large size, elongate form, and convex elytra disc pubescence sparse but rather long. Elytra demonstrate that at least one population of elongate-oval, two-thirds longer than wide, pubescens group stock was capable of evolving strongly convex on disc and deplanate near toward a Neaphaenops-like habitus. Living scutellum; prehumeral borders clearly oblique beetles in the caves or pinned specimens in a to midline, humeri finely serrate and setose; tray, seen with the unaided eye, resemble small striae finely impressed, outer striae obsolescent, Neaphaenops. With experience it becomes pos- intervals subconvex, each with two or three sible to distinguish the two species in the field. rows of pubescence; microsculpture weakly Neaphaenops t. meridionalis is redder, shinier, transverse and heavily pruinose; elytral and a bit larger, occurring in somewhat drier chaetotaxy normal, except anterior apical punc- and higher microhabitats than P. princeps. On ture placed opposite second umbilicate; apical the other hand, princeps is more yellowish 18 AMERICAN MUSEUM NOVITATES NO. 2682

FIG. 13. Pseudanophthalmus princeps, new species, Hoy Cave, Simpson County, Kentucky. brown, more opaque, slightly smaller, and oc- the laboratory show little digging activity, wan- curs on wet fill near the edge of water; depres- dering about the surface of the substrate in their sions in silt and mud of princeps microhabitats container. They are aggressive feeders, quickly become filled with water seeping in from the locating and eating bits of raw beef or dead sides. Hadenoecus subterraneus much sooner than P. Captive specimens of princeps maintained in loganensis or N. t. meridionalis. 1979 BARR: NEAPHAENOPS 19

All the caves from which P. princeps has bridges South Fork of Red River. Its closest been collected are in the drainage basins of approach to the species boundary of P. West Fork and Middle Fork of Drakes Creek, a princeps is in West Cave, 1.0 mile NNE Provi- south tributary of Barren River. Pseud- dence and about 7 miles SW Hoy Cave, in anophthalmus loganensis occurs in all tie caves Simpson County, Kentucky. Overlap with the listed above. Neaphaenops t. meridionalis was ranges of N. t. meridionalis or P. princeps has collected in Hoy, Old Smoky, Slates, and not been demonstrated; both orlindae and Plano Saltpeter caves, but was not taken in princeps are large riparian species with a cruis- Whiteoak Cave or the Steele caves. It would be ing predation strategy, and mutual exclusion interesting to compare ecology and behavior of may determine their species boundaries with the three species of trechines coexisting in respect to each other. these caves. If P. princeps and/or N. t. merid- ionalis are ecologically and behaviorally inter- Pseudanophthalmus loganensis Barr, new mediate between P. loganensis and N. - t. status tellkampfi, such an investigation might reveal Pseudanophthalmus ciliaris loganensis Barr, 1959, some of the selection pressures mediating the p. 7, fig. 2(1). Type locality, Cook (=Savage) adaptive zone shift which led to the evolution Cave, Logan Co., Kentucky; type deposited in of Neaphaenops. the American Museum of Natural History. Two Pseudanophthalmus taxa to which ref- erence is made in the present paper exist in Pseudanophthalmus loganensis is a medium- caves near the southern end of the range of sized (4.6-5.2 mm.) species widely distributed Neaphaenops tellkampfi. A change in tax- in Warren (south of Barren River), Simpson, onomic status is required for each of these and Logan counties, Kentucky, also northeast taxa, which belong to the pubescens group. Robertson and northwest Sumner counties, Tennessee. It is sympatric with four larger Pseudanophthalmus ciliaris orlindae Barr, trechine species: Neaphaenops t. meridionalis, new combination P. menetriesi, P. ciliaris orlindae, and P. princeps. Recognition of the correct status of Pseudanophthalmus orlindae Barr, 1959, p. 7, figs. orlindae as a subspecies of ciliaris demon- 1, 2(3). Type locality, Jesse James Cave, Robert- strates that loganensis is a full species. Princi- son Co., Tennessee; type deposited in the Ameri- pal diagnostic characters for loganensis are its can Museum of Natural History. total length range, moderately robust and feebly This large (5.4-5.8 mm.) form coexists in convex habitus, and aedeagal form: median many caves of Robertson County, Tennessee, lobe 0.83-0.92 mm. long and evenly arcuate, and southern Logan County, Kentucky, with right copulatory piece slender with terminal the smaller P. loganensis. Restudy of older button, left piece slender and pointed at apex, collections and fresh material demonstrates a parameres with lateral shelves peculiarly ex- clear morphological and geographic intergrada- panded into hyaline, curved flanges. Unlike tion between nominate ciliaris and orlindae in pubescens, princeps, or ciliaris orlindae, which Bell Witch Cave, Robertson County, Ten- are cruising predators, loganensis is an intersti- nessee. In both subspecies of ciliaris the tial feeder, almost always encountered at the aedeagus is elongate and rather straight, with edge of cave streams in mud cracks, on rotting slightly produced and deflexed apex (thus dif- wood fragments, or in the interstices of cherty fering from all other species of the group); cobble- and gravel-bars. orlindae differs from nominate ciliaris in larger size and larger aedeagus (1.3-1.4 mm. long vs. LITERATURE CITED 0.9-1.0 mm. long). The range of polytypic Barr, Thomas C., Jr. ciliaris includes caves along Red River from 1959. New cave beetles (Carabidae, ) the vicinity of Clarksville, Tennessee, eastward from Tennessee and Kentucky. Jour. Ten- approximately to the point where Interstate 65 nessee Acad. Sci., vol. 34, pp. 5-10. 20 AMERICAN MUSEUM NOVITATES NO. 2682

1962a. The robustus group of the genus Jeannel, Rend Pseudanophthalmus (Coleoptera: Car- 1920. Notes sur les Trechini. Bull. Soc. Ento- abidae). Coleopt. Bull., vol. 16, pp. mol de France, pp. 150-155. 109-118. 1926-1930. Monographie des . Mor- 1962b. The blind beetles of Mammoth Cave, phologie comparde et distribution d'un Kentucky. Amer. Midland Nat., vol. 68, groupe de Coldopteres. L'Abeille, vol. 32, pp. 278-284. pp. 221-550; vol. 33, pp. 1-592; vol. 34, 1967a. Observations on the ecology of caves. pp. 59-122; vol. 35, pp. 1-808. Ibid., vol. 101, pp. 475-492. 1931. Revision des Trechinae de l'Amdrique du 1967b. A new Pseudanophthalmus from an epi- Nord. Arch. zool. exp. et gdn., vol. 71, gean environment in West Virginia (Cole- pp. 403-499. optera: Carabidae). Psyche, vol. 74, pp. 1949. Les coldopteres cavernicoles de la rdgion 166-174. des Appalaches. Etudes systdmatique. 1968. Ecological studies in the Mammoth Cave Notes Biospdol., fasc. 4, Publ. Mus. Nat. system of Kentucky. I. The Biota. Int. Hist. Nat., Paris, no. 12, pp. 37-104. Jour. Speleol., vol. 3, pp. 147-203, pls. Kane, Thomas C., Russell M. Norton, and Thomas 37-64. L. Poulson 1972. in North America, with a 1975. The ecology of a predaceous troglobitic key to North American genera of beetle, Neaphaenops tellkampfii (Coleop- Trechinae (Coleoptera: Carabidae). tera: Carabidae, Trechinae). Seasonality of Psyche, vol. 78, pp. 140-149. food input and early life history stages. Barr, Thomas C., and Robert A. Kuehne Int. Jour. Speleol., vol. 7, pp. 45-54. 1971. Ecological studies in the Mammoth Cave Ludwig, W. system of Kentucky. II. The Ecosystem. 1950. Zur Theorie der Konkurrenz. Die Annida- Ann. Speleol., vol. 26, pp. 47-96. tion (Einnischung) als funfter Evolu- Barr, Thomas C., and Stewart B. Peck tionsfaktor. Neue Ergeb. Probleme Zool., 1965. Occurrence of a troglobitic Pseud- Klatt-Festschrift, pp. 516-537. anophthalmus outside a cave (Coleoptera: McKinney, Tom Carabidae). Amer. Midland Nat., vol. 73, 1975. Studies on the niche separation in two pp. 73-74. carabid cave beetles. Int. Jour. Speleol., Barr, Thomas C., and Daniel Stoneburner vol. 7, pp. 65-78. MS. The energy budget of a cricket cave. Norton, Russell M., Thomas C. Kane, and Thomas Erichson, Wilhelm F. L. Poulson 1844. (Footnote, p. 384) In Tellkampf, Theodor, 1975. The ecology of a predaceous troglobitic 1844. Ueber den blinden Fisch der Mam- beetle, Neaphaenops tellkampfii (Coleop- muth-Hohle in Kentucky, mit Bemerkung- tera: Carabidae, Trechinae). II. Adult sea- en uber einige andere in dieser Hohle sonality, feeding, and recruitment. Int. lebenden Thiere. Mullers Arch. Anat. Jour. Speleol., vol. 7, pp. 55-64. Physiol., vol. 4, pp. 384-394, pl. 9. Packard, Alpheus S., Jr. Giuseffi, Steven, Thomas C. Kane, and William F. 1888. The cave fauna of North America, with Duggleby remarks on the anatomy of the brain and 1978. Genetic variability in the Kentucky cave origin of the blind species. Mem. Nat. beetle Neaphaenops tellkampfii (Coleop- Acad. Sci. (U.S.A.), vol. 4, 156 pp., 27 tera: Carabidae). Evolution, vol. 32, pp. pls. 679-68 1. Park, Orlando, and Thomas C. Barr, Jr. Hobbs, Horton H., Jr., and Thomas C. Barr, Jr. 1961. Some observations on a cave cricket (Ab- 1972. Origin and affinities of the troglobitic stract). Entomol. Soc. Amer. Bull., vol. crayfishes of North America. II. Genus 7, p. 144. Orconectes. Smithsonian Contrib. Zool., Valentine, J. Manson no. 105, iv + 84 pp. 1952. New genera of anophthalmid beetles from Hubbell, Theodore H., and Russell M. Norton Cumberland caves (Carabidae, Trechini). 1978. The systematics and biology of the cave- Geol. Surv. Alabama, Mus. Pap. 34, 41 crickets of the North American Tribe PP. Hadenoecini ( Saltatoria: En- Wilson, D.S. sifera: : Dolichopodi- 1975. The adequacy of body size as a niche nae). Misc. Publ. Mus. Zool., Univ. difference. Amer. Nat., vol. 109, pp. Michigan, no. 156, pp. 1-124, 5 pls. 769-784.