Trichopetalum Harger 1872

Shear, William A., 2010, 2385, Zootaxa 2385, pp. 1-62 : 10-11

Genus Trichopetalum Harger 1872

Trichopetalum Harger, 1872:117 . Cook & Collins, 1895:62. Hoffman, 2000:235 (list). Shear, 1972:277, 2003:4 et seq. Tynopus Chamberlin, 1940:57 .

Flagellopetalum Causey, 1951:120 .

Type species: Trichopetalum lunatum Harger 1872 , by subsequent designation of Cook & Collins (1895); for Tynopus , T. dux Chamberlin 1940 ; for Flagellopetalum , F. stannardi Causey 1951 .

Included species: In addition to the type, T. jerryblatti n. sp., T. dickbrucei n. sp., T. dux (Chamberlin) 1940 , T. montis Chamberlin 1951 , T. stannardi (Causey) 1951 , T. uncum Cook & Collins 1895 .

Diagnosis: Differing from members of Zygonopus and Scoterpes in having 28 trunk segments rather than 30, and in having ocelli. The coxal trichomes, or setulae, are abundant, prominent and often elongated and hairlike ( Fig. 5). The single angiocoxite may be entire, divided at the tip, or deeply divided into two branches. There may also be a lateral process. The fimbriate branch arises from the posterior surface of the angiocoxite, or may be a separate articulated element. The colopoxite is well-sclerotized and diagnostic in form for species. Male legpair 2 with a distinct coxal hook ( Fig. 27), legpairs 3–7 variably enlarged, either 4 and 5, or 6 the largest, femur 6 sometimes modified.

Distribution: Eastern North America, from Newfoundland, Maine and New Hampshire west to Michigan, Wisconsin and Minnesota, and thence south to Mississippi. From New York the range extends down the Appalachian Mountains and adjoining Piedmont Region to northern Georgia. Shelley (1988) reports female Trichopetalum from eastern Canada as T. lunatum ; this is probably correct but is not based on males. Trichopetalum lunatum has evidently re-invaded glaciated territory in the last 12,000 years, and lunatum now occurs farther north (in southern Newfoundland) than any other native eastern North American milliped species except Underwoodia iuloides , which gets even farther north in Newfoundland ( Palmèn 1952). How these small, delicate, slow-moving milliped species with their evident requirements for high humidity accomplished this feat of distribution is a matter for speculation. Species of Trichopetalum have unusually large ranges for small millipeds, and records tend to be scattered. The latter effect may be due to the hygrophilous and psychrophilus nature of the species, which makes them hard to find during the usual collecting season, and to their small size, which makes them easy to overlook.

In 1972, I reported Trichopetalum lunatum from British Columbia, but this record has never been confirmed, and the examination of many recent samples of hundreds of chordeumatidan millipeds from the Pacific Northwest, including British Columbia, has revealed no trichopetalids. The original specimens ( MCZ) must have been mislabelled. Causey (1963) implied occurrence in Louisiana and Colorado; I have seen no specimens from Louisiana, but there is a remote possibility of Trigenotyla being found there. The Colorado specimens (examined, FSCA) are juveniles of the conotylid Austrotyla coloradensis (Chamberlin) . The American Museum collection ( AMNH) contained a number of samples from the Dakotas labelled “ Trichopetalum sp. ” These are all female and juvenile specimens of Underwoodia iuloides .

Notes: Shelley (1993) has assigned Harger’s (1872) Trichopetalum glomeratum , from Oregon, to the conotylid genus Taiyutyla , a reasonable conclusion drawn from study of the female holotype in the Peabody Museum of Natural History at Yale University. Trichopetalum iuloides Harger, 1872 is Underwoodia iuloides ( Caseyidae ; Cook & Collins 1895). Polydesmus ocellatus Packard 1883 was tenatively placed in Trichopetalum by Cook & Collins (1895) but as it was described from Oregon, where Trichopetalum does not occur, that is unlikely. Hoffman (2000) speculated that it might be a rhiscosomidid, but the illustrations of the specimen by Cook & Collins show the collum narrower than the head, while in rhiscosomidids the collum is much wider than the head, and other illustrated details do not fit. The illustrations and description are more like a species of Tingupa , but in the absence of the only specimen (which in any case was probably immature) we will never know the identity of this name. Crasepedosoma flavidum Bollman 1888 was also placed by Cook & Collins (1895) in Trichopetalum , but they could only find “...a small yellow female specimen with 26 segments purporting to be the type of this species....” in the USNM. That specimen appears no longer to be in the collection, and the descriptions of both Bollman (1888) and Cook & Collins (1895) are not sufficient to place the species, so this name, too, will be forever in doubt and have to be arbitrarily assigned. If the 26-segmented specimen was indeed a mature female, then it is possibly Branneria carinata (Bollman), 1893 or Buotus carolinus Chamberlin, 1940 ( Shear 2009) — unfortunately both younger and much more-used names.

The species originally described in Zygonopus Ryder, 1881 , were transferred to Trichopetalum by me in 1972, an action I now think was in error, and Zygonopus is reinstated below as a valid genus (in 2003, in the first paper in this series, I suggested subgeneric status under Trichopetalum ). There can be no doubt, however, that the two genera are closely related. The following new synonymies and combinations are proposed: Trichopetalum appropinquo ( Causey, 1969) and T. quadratum ( Loomis, 1966) are both synonyms of T. stannardi ( Causey, 1951) , T. cornutum Cook and Collins, 1895 , is a synonym of T. lunatum Cook and Collins, 1895 , T. subterraneum Causey, 1967 is a synonym of T. dux ( Chamberlin, 1940) , and T. syntheticum Shear. 1972 is a species of Scoterpes .