Gasterogramma Jeekel, 1982

Mesibov, Robert, 2003, The millipede genus Gasterogramma (Diplopoda: Polydesmida: Dalodesmidae) in Tasmania, Australia, with descriptions of seven new species, Memoirs of Museum Victoria 60 (2), pp. 207-219 : 207-211

Gasterogramma Jeekel, 1982

Gasterogramma Jeekel, 1982: 10 .— Shelley et al., 2000: 102.

Type species. Gasterogramma psi Jeekel, 1982 , by original designation.

Included species. G. austrinum sp. nov., G. imber sp. nov., G. extremum sp. nov., G. plomleyi sp. nov., G. psi Jeekel, 1982 , G. rusticum sp. nov., G. tarkinense sp. nov. and G. wynyardense sp. nov.

Remarks. Jeekel (1982: 10) gave the following diagnosis for somatic features of Gasterogramma : “Medium-sized Dalodesmidae with 20 somites and a normal poreformula. Head without particulars; antennae of moderate length, clavate, the 6th antennomere longer than the 5th. Collum without paranota; the sides almost perpendicular. Somites very weakly constricted, almost cylindrical; the waist broad. Metatergites without sculpture. Paranota weakly developed, ridgelike, only dorsally demarcated by a furrow in which the pores are situated. Sternites much longer than wide, without particular modifications. Legs of moderate length, incrassate in the male and with dense setation of short stiff bristles on ventral side of podomeres 2 to 4 and dense granulation on ventral side of podomeres 5 and 6. Anal somite without particulars.” This general description also applies to the seven new species described here but the following generic features should be included: integument well calcified; narrow but clearly distinguishable paranota on second segment, lower on body than the lateral collum margins and the paranotal traces on third segment; gonopod aperture ovoid with long axis transverse, rim not raised in front and variably raised at rear; epiproct blunt, prominent, extending caudad past the anal valves.

Jeekel’s diagnosis of the Gasterogramma gonopod (1982: 10) is here amended: telopodites elongate, loosely joined medially; solenomerite arising at one-half to two-thirds the length of the telopodite, each solenomerite a small, distally tapering rod twisted in the following sense: a right-handed helix (clockwise turning away from the observer) on the right gonopod, a left-handed helix on the left gonopod.

In an attempt to identify (i.e., homologise) the elements of the G. psi gonopod, Jeekel (1982: 11) hypothesised that “the course of the spermal channel [here called “prostatic groove”] is homologous in all polydesmoid millipedes, and that the location of the base of each process in relation to the spermal channel defines its morphological identity.” He named the process arising caudad of the prostatic groove the tibiotarsus, and the structure arising cephalad the femoral process. If the distal portion of the G. psi gonopod is untwisted, in imagination, then the lateral process with a tip shaped like a bird’s head (fig. 6) is identifiable as the femoral process; the anterior, laterally fringed process bearing a rod-like structure is a prolongation of the prefemur; and the rod-like structure is the tibiotarsus. In the notch between the femoral and prefemoral processes arises the solenomerite.

An alternative interpretation is that the distal end of the telopodite is not torted, and that the ‘femoral process’ is a growth from the caudomedial side of the gonopod which has displaced the prostatic groove laterally. It is important to note that the course of the prostatic groove in G. psi is actually two opposed spirals. From the base of the telopodite of the right gonopod, the groove first runs distally in a left-handed spiral to encircle the “femoral process”, then in a right-handed spiral upon entering the solenomerite. The cheirality of the spirals on the left gonopod mirrors that on the right. In the three species here assigned to Gasterogramma which lack a femoral process ( G. austrinum , G. rusticum and G. wynyardense ), the basal spiral is missing but the solenomerite spiral persists.

Jeekel’s hypothesis sought to make each gonopod process in polydesmoids a homologue of either the prefemur, femur or tibia+tarsus of the millipede leg, and uses a “fixed” solenomerite as a morphological landmark. If the “no torsion” interpretation of the Gasterogramma gonopod is correct, then either the simple notion of homologisation is inadequate, or the solenomerite is not “fixed.” Unfortunately, polydesmoid gonopods develop suddenly from unsegmented primordia in the ultimate moult, and there are no visible intermediate stages which would allow us to unambiguously identify individual gonopod processes with prototypical leg segments, in Gasterogramma or any other polydesmoid. A correct interpretation of gonopod structure may be possible in future with advances in the developmental genetics of arthropods. For the present, Jeekel’s labelling of the gonopod processes in Gasterogramma is a useful hypothesis, and is adopted for descriptive purposes in this paper.

Distribution. Western and north-eastern Tasmania (fig. 13); not recorded from King I. or the Furneaux Group in Bass Strait.

Habits. The eight species of Gasterogramma are burrowing millipedes typically found in moist, organic-rich soil, in deep moist leaf litter and in and under wet rotting logs. As noted by Jeekel (1982: 9) for G. psi , Gasterogramma are “sluggish” and at first sight they resemble “a worm or a dipteran larva rather than a millipede.” They are often locally abundant in loose, mixed-age aggregations. All Gasterogramma species release a pungent defensive secretion which presumably includes hydrocyanic acid, although this is not the most prominent odour component. The pinkish-purple Gasterogramma species have been called “stinky pinkies” by local collectors impressed by the pungency of the secretion. G. plomley i has the strongest smell in the genus, and can be detected by treading heavily through its wet forest habitat and sniffing attentively. Unfortunately, no somatic characters have been so far noted which can reliably be used to distinguish the seven western species (i.e., Gasterogramma spp. other than G. plomleyi ), and at many western sites two Gasterogramma species are present. I have not, therefore, assigned non- plomleyi females and juveniles to species; these individuals are recorded in the separate specimen data table as “ Gasterogramma sp. ” Males and females have been found in copula in all months of the year, but mating is most frequent in the austral spring; the only two pairs so far preserved in copula ( G. psi, QVM 23:8086 and QVM 23:40560) were collected in late November. Females are often found curled around an egg mass; it is unclear whether such females are in the process of constructing an egg chamber (as expected for Polydesmida ) or are brooding their eggs.

Relationships. Jeekel (1982: 12) noted that Gasterogramma and the Chilean Semnosoma Silvestri, 1903 were the only dalodesmid genera in which the prostatic groove had been reported to be helical, and “It seems likely, therefore, that the new genus is closely related to Semnosoma . Eventually, the two may form the nucleus of a tribe for which the family-group name Semnosomatidae Brölemann, 1916 is available.” The illustrations of Semnosoma gonopods in Demange and Silva (1976a, b) show the prostatic groove running first caudally, then laterally, then anteriorly as in G. psi . Further, species in the Semnosoma group of Chilean dalodesmids (i.e. Semnosoma , Anaulacodesmus Attems, 1898 , Chilorus Chamberlin, 1957 and Tsagonus Chamberlin, 1957 ) all have, like Gasterogramma , greatly reduced paranota ( Chamberlin, 1957: 19). The argument for a Gondwanan dalodesmid lineage including the Chilean and Tasmanian genera thus appears reasonable.

Relationships within Gasterogramma are unclear. Grouping phenetically on gonopod form, we might have ( G. extremum + G. imber + G. plomleyi + G. psi + G. tarkinense ) with a femoral process and ( G. austrinum + G. rusticum + G. wynyardense ) without such a process, but relationships within these groups are far from obvious. Both groups are widely dispersed across Tasmania, suggesting that the two might have differentiated independently in the landscape. However (see below) the two groups are not independently distributed: the seven non- psi species currently form an allopatric/parapatric mosaic.