Scoterpes Cope

Shear, William A., 2010, 2385, Zootaxa 2385, pp. 1-62 : 38-40

Genus Scoterpes Cope

Scoterpes Cope, 1872:414 ; Loomis, 1943:92; Shear, 1972:279; Hoffman, 2000:235.

Type species: Spirostrephon (Pseudotremia) copei Packard 1871 , by original designation.

Included species: In addition to the genotype, S. austrinus Loomis 1943 , S. nudus Chamberlin 1946 , S. ventus Shear 1972 , S. syntheticus (Shear) 1972 , S. sollmani Lewis 2000 , and the new species alabama , blountensis , hesperus , jackdanieli , musicarustica , stewartpecki , tombarri , tricorner and willreevesi . Scoterpes dendropus Loomis 1939 has been transferred to Causeyella Shear ( Shear 2003) .

Diagnosis: Scoterpes species lack coxal trichomes, which separates them from all other US trichopetalid genera. Blind, unpigmented, and with 30 segments, as in Causeyella and Zygonopus ; in Causeyella the left and right gonopod angiocoxites are closely appressed to each other, while in Scoterpes they are widely separated; Causeyella also displays large, lamellate, fimbriate colpocoxite branches—in most Scoterpes these are thin with filamentous branches; in Zygonopus the fimbriate branch of the colpocoxite is relatively small and arises from its anterior surface, while in Scoterpes it arises apically and is a prominent feature of the gonpod.

Animals ( Fig. 56) range in size from 6 to 12 mm, generally larger than Trichopetalum individuals, but not so large as those of Causeyella species , which are up to 15 mm long. Segmental setae are very long and prominent ( Fig. 55), from 80% of to slightly longer than the body width, longer than in any other genus of chordeumatidan millipeds. The lateralmost segmental setae are borne on prominent, paratnota-like knobs, the paralateral setae similarly but on a shorter knob, and the mesal setae on still lower mounds near the midline. Frequently blobs of secretion can be seen on the segmental setae ( Fig. 57). Pregonopodal legs are only slightly enlarged or not different from postgonopodal legs. Legpairs 1–5 are notably shorter than subsequent pairs and the length diffierence is mostly in the femur, so that legpairs 6 and 7 appear abruptly longer. Legpair 6 sometimes has a slight ventral swelling on the femora ( Fig. 58). Legpair 9 is typical for the subfamily ( Fig. 59), with spindle-shaped telopodites that frequently have a vestigial claw, or rarely a few vestigial distal segments. This suggests that the telopodite article represents a coalesence of all articles distal to the prefemur, and not just the prefemur itself. The gonopods are typical of the family but with more obvious fusion of the coxae with the sternum and with each other. Coxal setae are in two groups, the mesoproximal group with three setae on each coxa, the distolateral group of variable number but always more than three. The angiocoxites are divided into two branches, mesal and lateral; the lateral branches are often complexly folded at their tips ( Figs. 60, 61, 65, 67). The mesal branches may be reduced and sometimes are absent entirely. The colpocoxites are large, ovoid, and poorly sclerotized, with a single prominant fimbriate branch at the apex.

The name Scoterpes is compounded from the Greek words skotos and herpeton, and means “a crawler in darkness.” It is masculine in gender.

Distribution: Always in caves as troglobionts, ranging through southern Indiana, the interior plateaus and karst basins of Kentucky and Tennessee, the southern Blue Ridge in Tennessee and Georgia, and the southern Ridge and Valley Province in Alabama. Outliers are S. sollmani in southern Indiana, the only species occurring north of the Ohio River, but an obvious close relative (if not a synonym) of S. copei across the river in northern Kentucky, and S. blountensis n. sp., isolated far to the east of the main distribution in Blount Co., Tennessee, and of uncertain relationship. In western Kentucky and northcentral Tennessee, the Cumberland River bounds the distribution. South of where the Cumberland turns northwestward, the boundary is the Tennessee River, but these rivers are also nearly at the limit of the westward extension of karst (map in Barr 1961). Southeastward, S. syntheticus occurs farthest south, just north of Birmingham, Alabama, and S. nudus is found in the Etowah River valley, a little distance southeast of Rome, Georgia.

The central distribution or core area of the genus is the Cumberland Plateau in Tennessee, the Pennyroyal Plateau in Kentucky, and the interior karst basins, such as the Nashville Basin. Eastern Kentucky, arbitrarily defined as that part of the state east of the Kentucky and Rockcastle Rivers (equivalent to the Bluegrass and Eastern Coalfield physiographic regions), has karst but no Scoterpes ; similarly, only the very isolated S. blountensis is to be found in Tennessee east of the Sequatchie River, despite the presence of abundant karst in the Ridge and Valley portion of that state (there is a Meigs Co., Tennessee, record of S. copei , but this is highly questionable).

In contrast, Scoterpes has found its way into the folded Appalachians in Georgia and Alabama, and there two of the most distinctive species are to be found, though I believe they are most closely related to the species of the Cumberland Plateau. The Cumberland Plateau of northern Alabama was recently shown to have the highest diversity of terrestrial troglobionts in North America (Culver et al. . 2006), and at least six species of Scoterpes occur there.

Notes: Nell Causey worked extensively for many years on the large collections of Scoterpes made available to her by Barr, Peck, Holsinger, Steeves and others, but never published anything on them. She noted (in litt. 1972) that the situation was “hopelessly confusing.” In 1960, she estimated as many as 27 taxa in the genus (species and subspecies). Indeed although I have pushed ahead, and here describe nine new species in the genus, many difficulties remain. While species from the southern end of the generic range seem discrete, relatively homogeneous and limited geographically, at least two “species,” S. copei (Packard) and S. ventus Shear , have large ranges that transgress barriers known to separate species, or even genera, in other troglobiont taxa. Significant variation in gonopod anatomy can be detected within these species, but the variants seem to this eye to merge imperceptibly into one another. Based on her labellng, Causey discerned six subspecies of copei (though some labelling suggests she may have at times thought of these as species) and three subspecies of ventus . These hypothetical subspecies had some geographic coherence, but I cannot reliably separate them. Further, Lewis (2000) described S. sollmani from caves in Indiana, north of the Ohio River, but I do not see any significant differences between this species and copei south of the river and well south and east into Kentucky. I have no doubt that both copei and ventus are “superspecies,” groups of closely related populations that have reached the level of reproductive isolation, but a much more detailed analysis than I am willing to embark on would be required to test that hypothesis. With the increasing ease of the use of molecular methods to demonstrate reproductive isolation, the genus Scoterpes , and in particular the “species” copei and ventus , would present an ideal challenge to a student of molecular systematics. It may well be that Causey’s (1960) estimate of 27 species-level taxa in Scoterpes will turn out not to be far from the mark.

Therefore the account of Scoterpes species that follows is not really a definitive revision but only a “first pass” through the genus, in which I provide names for some obviously distinct populations and plant some signposts to mark out problems for future research.

Scoterpes differs from the other trichopetaline genera in the complete lack of the characteristic trichomes on the gonopod coxae and relative lack of modifications of the pregonopodal legs. It is difficult to say if these and other characters are apomorphic and due to a long history of troglobiosis, or represent truly plesiomorphic characters. If the latter, the genus is basal to the subfamily.

The presence of Scoterpes only in caves throughout its wide range, from Indiana south to Alabama and Georgia, and the high level to which they have developed troglomorphy, suggests to me that the most recent events of the Pleistocene, the Wisconsinian glacial advance and retreat, may not be responsible for the isolation in caves of this genus. Some of the species may be very ancient troglobionts. It is also unclear if underground dispersal or repeated isolation of surface populations dominates the history of the genus (see Culver and Pipan 2009, pp. 131–154, for an excellent discussion of these alternatives). In other words, were there numerous species of surface-dwelling proto- Scoterpes that independently became isolated in caves, or only a few species that have, since their isolation, speciated (and perhaps dispersed) underground? Likely either explanation can be invoked in different parts of the generic range—in the interior basins of Kentucky and Tennessee, where limestone strata are flat-lying and extensive and cave systems enormous and interconnected, far-reaching underground dispersal over millenia is at least possible. But in the southern end of the folded Appalachians in Georgia and Alabama, formidable barriers to underground dispersal exist. The dispersal and vicariance models of troglobiont evolution and distribution are not mutually exclusive. Vicariance (extinction of surface populations) may account in most cases for the original isolation of cave populations, and later, dispersal through contininous karst may explain the existing range of the species ( Holsinger 2005; Culver and Pipan 2009).

Alternatively, the broad distribution of the possible “superspecies” S. copei and S. ventus could argue for their having reached this distribution as epigean/troglophilic species whose eipgean populations have become extinct recently enough not to allow for conspicuous morphological (gonopod) change in the isolated cave populations. As an example from another taxon, the cleidogonid milliped species Pseudotremia hobbsi Hoffman, 1950 , is found in the several counties along the Virginia / West Virginia border, both as epigean populations and as troglophiles in caves ( Shear 1972, 2008). Were the epigean populations to be extirpated by climate change, the surviving cave populations would appear to be, after some evolutionary development in the direction of cave-related adaptations, a widespread trogobiotic species whose distribution could not be explained by underground dispersal. Furthermore it could be argued that when populations become isolated in caves, potential selective forces constraining gonopod anatomy are reduced, since such features need no longer isolate populations. In Pseudotremia several of the most troglomorphic species, unrelated to one another, have clearly converged on a highly simplified gonopod structure ( Shear 1972). These are questions that cannot even be approached without more data on genetics, specifically gathered for the purpose. The work of Hedin (1997) on cave spiders of the genus Nesticus and that of Moulds et al (2007) on pseudoscorpions are pioneering preliminaries to more detailed analyses of the genetics of speciation and population structure in cave animals. While continuing to recognize morphological species, Hedin demonstrated that populations within these species were genetically coherent and distinct, and perhaps at some later time deserving of species status themselves. Such species, however, may only be identifiable by recourse to DNA sequences (see Bond and Sierwald 2002, 2003). As sequencing becomes more and more convenient with each passing year, it is not unreasonable to suppose that it might even be possible to do in the field in the relatively near future.

So despite my own dissatisfaction with what follows, I hope that at least some of the populations I hypothesize to be species will actually turn out to be so. But it cannot be denied that there are probably more species of Scoterpes than are named here, most of them concealed within my concepts of copei , ventus , and perhaps one or two others.

Three fairly distinct species groups can be recognized in Scoterpes , and two species have gonopods divergent enough that they cannot be comfortably placed in any of the three groups. The groups are formed largely by my impression of overall gonopod similarity, but a few characterizing remarks can be made. The largest group by far, occupying most of the generic range in Kentucky and Tennessee, is the Copei Group. The broadly similar gonopods of this group usually feature branched and/or folded lateral angiocoxites, and there seems to be a tendency for the median angiocoxites to be reduced. The Alabama Group appears to be a derivative of the Copei Group, in which both sets of angiocoxites have grown quite long and have become simplified; one of the two species in Alabama group has evidently lost the median angiocoxites. One species occurs in southcentral Tennessee, the other in northern Alabama. The Austrinus Group is composed of two species, very closely related to each other, in northern Georgia. Scoterpes willreevesi , n. sp., from northern Georgia and adjacent Alabama, has highly complicated gonopods which, while they are obviously on the Scoterpes plan, are unlike any others. Finally, S. syntheticus Shear is a small species with correspondingly small gonopods that have become extremely simplified; it occurs in northcentral Alabama and is the most southerly distributed of the species.