Macrosternodesmidae Brölemann 1916

Shear, William A. & Shelley, Rowland M., 2007, The milliped genus Tidesmus Chamberlin, 1943 (Polydesmida: Macrosternodesmidae), Zootaxa 1656, pp. 51-68 : 52-54

Family Macrosternodesmidae Brölemann 1916

Macrosternodesmini Brölemann, 1916:585.

Macrosternodesmidae : Hoffman, 1980:177; 1982:722; 1999:457. Kevan, 1983:2969. Blower, 1985:209. Shelley, 1988:1650; 2000:190; 2002 a:1870. Simonsen, 1990:64, 81–82. Lewis, 2002:103. Kime, 2004:39.

Diagnosis: (adapted from those of Hoffman (1982) and Simonsen (1990)). Generally colorless to reddishbrown, small to moderate-size Trichopolydesmoidea (6–12 mm long but occasionally ca. 30 mm long) with 20 segments including epiproct; collum narrower than head, not overlapping epicranium; segments 3–4 generally smaller than 2 and 5; metaterga with transverse sulci and three or four rows of variably rounded to subconical pustules giving rise to clavate setae; paranota small but distinct; limbi smooth to irregularly scalloped and ragged, one termination occasionally elongated and spiniform; tarsi ca. twice as long as next longest podomeres, male prefemora swollen and convex dorsally; sphaerotrichomes present on at least ambulatory tibiae and tarsi. Gonopodal aperture large, broadly ovoid, completely filling metazonite, not extending onto prozonite but sometimes spreading caudad between 9th legs. Gonocoxae large, completely filling respective halves of aperture, excavated mediad to accommodate telopodites; prefemora horizontal or angling ventromediad, giving rise to acropodite and additional projection homologous to process B of Nearctodesmidae (terminology of Shelley (1994)); acropodite part distal to origin of solenomere (distal zone) variably configured, sometimes folded, flattened, and not recognizable as such; solenomere long and narrow, arising subterminally, without hairpad and ampulla, prostatic groove opening terminally.

Components. Chaetaspis Bollman, 1887 ; Ophiodesmus Cook, 1896 ; Macrosternodesmus Brölemann, 1908 ; Tidesmus Chamberlin, 1943 . Additional genera, currently assigned to Polydesmidae , Trichopolydesmidae , or other families, may also belong here.

Caucasodesmus inexpectatus, Golovatch 1984 , from North Ossetia, Russia, was originally placed in the Macrosternodesmidae but from published figures obviously does not belong here as the family is now understood. The prefemora are globose, not transverse; process B is missing; and there is no solenomere branch. Indeed, according to Golovatch (1984), there is no prostatic groove and no solenite (= cannula), conditions almost unique in the Polydesmidea and Trichopolydesmidea. On the other hand, the gonopod plan of the longestablished genus Archipolydesmus Attems, 1898 ( Spain and North Africa; see Abrous-Kherbouche and Mauriès 1996), appears close to those of a series of undescribed macrosternodesmid genera from Arizona, USA, though largely by tradition, Archipolydesmus is formally included in the Polydesmidae . Furthermore, Mauriès (1980) re-established the family Mastigonodesmidae Attems, 1914, for Mastigonodesmus Silvestri, 1898 , an action that seems justified because the genus cannot be considered a polydesmid in the classical sense as it lacks the torted prostatic groove (= seminal canal) and characteristic cuticular fimbriae surrounding the distal opening of the latter, which we consider defining characters of the Polydesmidae . Lack of process B seems to eliminate Mastigonodesmus from the Macrosternodesmidae .

Distribution. Holarctic. Indigenous to northern coastal Croatia; western Europe (southern Sweden, Denmark to the Pyrenees of France); eastern US east of the Central Plains (central Oklahoma to central North Carolina, southern Illinois and southcentral Indiana to northern Arkansas & Alabama and northcentral Mississippi); southwestern US (Arizona, California,?New Mexico); northwestern Mexico (Baja California Norté,?Sonora) ( Jeekel 1953; Enghoff 1973, 1974; Shelley 1978, 2000; Hoffman 1980, 1982, 1999; Blower 1985; Simonsen 1990; Geoffroy 1996; Lewis 2002; Kime 2004); introduced to Newfoundland, Canada ( Palmén 1952; Kevan 1983; Blower 1985; Shelley 1988, 2002 a; Hoffman 1999). Simonsen (1990:88, Map 3) depicts the distribution in Europe and part of that in the eastern US.

Remarks. Clearly much remains to be learned about relationships within the superfamilies Polydesmoidea and Trichopolydesmoidea, especially with continued discoveries of new taxa in western North America and elsewhere. Major familial realignments seem likely in the near future.

The Macrosternodesmidae has received little attention since it was established by Brölemann (1916) as a tribe in the subfamily Devilleinae, family Leptodesmidae; 64 years later, Hoffman (1980) elevated it to family status. Subsequent anatomical characterizations ( Hoffman 1982, Simonsen 1990) must be modified to accommodate Tidesmus , in which the gonopodal prefemora are not entirely transverse as they are in Chaetaspis and representatives of the Nearctodesmidae in northwestern North America. Macrosternodesmidae may be distinguished from Nearctodesmidae in that it possesses telopodal process B but lacks branch A (terminology of Shelley (1994)) and from the Polydesmidae by the presence of process B and by the essentially direct course of the prostatic groove, which lacks loops and opens apically on a solenomere instead of in a pore surrounded by cuticular fimbriae on the telopodal stem. Our illustrations show that process B arises from the body of the gonopod at the anteromesial base of the prefemur. Characters are lacking to distinguish Macrosternodesmidae from the other New World trichopolydesmoidean family, Fuhrmannodesmidae , whose limits are unknown. It has become a wastebasket taxon for small tropical trichopolydesmoideans, but no US genera are presently assigned to this family.

Golovatch (1991) attempted to resolve polydesmoidean families based on positional homologies of gonopodal branches. However, his scheme is unworkable for the North American fauna because his gonopod terminology differs from those in other studies; unfortunately, old terminology, with established (although possibly incorrect) meanings, was employed. Djursvoll et al. (2000) subsequently changed the meaning of several terms whereas others seem to have been abandoned.

A stable terminology for polydesmidean gonopods is desirable. The quest began in the 19th century with the contribution of Attems (1894), continued in the 20th century (e.g. Hoffman 1974, Golovatch 1991), and still continues in the 21st century (Djursvoll et al. 2000). None of the systems has gained much acceptance, and the most recent one (Djursvoll et al. 2000) was almost immediately abandoned by one of its proposers ( Golovatch 2006). By studying gonopod development, a few authors attempted to homologize parts of the gonopod with ambulatory podomeres; the most recent ( Petit 1976) concluded that no articles distal to the prefemur were represented. Such studies are made difficult by the "metamorphic" transformation of the lumplike primordia to gonopods in the final molt. At present, however, establishing homologies with podomeres, while desirable, is of secondary importance. What is needed is consensus on a set of names, or even alphabetical symbols, for structures in comparable positions on the gonopods of related species (this tactic has worked for similarly complex structures like the palpi of male spiders). Most terminologies mix names (i.e. tibiotarsus), which suggest homologies, and letters. An acceptable terminology must be based on careful studies of as wide a spectrum of genera as possible (Djursvoll et al. 2000 arbitrarily excluded the North American genera Pseudopolydesmus Attems, 1898 , and Scytonotus C.L. Koch, 1847 ) using both compound light microscopy and SEM. It will be an herculean task.

Our gonopod terminology follows Shelley (1994). All structures distal to the coxa are referred to as the telopodite, which is based on the clearly observable homology of this part of the gonopod with the postcoxal (telopodal) elements of walking legs. Petit (1976) showed that the setose basal part of the gonopod telopodite derives from the ambulatory prefemur, and we so designate it here. The prostatic groove originates in a fossa on the medial side of the prefemur into which an articulated and movable coxal projection, the solenite (= cannula), inserts. Process B arises from the anterodistal margin of the prefemur and is U-shaped or strongly curved in Tidesmus , with the outer branch of the “U” being shorter. The telopodite distal to the indistinct boundary of the prefemur is the acropodite, and in light of Petit’s work, we believe the terms “femorite” and “tibiotarsus” should be abandoned. Macrosternodesmid acropodites bear a number of processes, the most easily recognizable being the solenomere, a tubular projection with the prostatic groove opening apically. The base of the solenomere may be thickened and bear one or more subsidiary processes in Tidesmus . The distal zone, which has been labeled “tibiotarsus,” “endomerite,” and “exomerite,” is the part of the acropodite distal to the solenomere; in Tidesmus , it is a blocky structure that often is folded.