Redescription of Brochopeltis mjoebergi Verhoeff, 1924 and description of a second Brochopeltis species from Australia (Diplopoda, Polydesmida, Paradoxosomatidae)

Abstract Brochopeltis mjoebergi Verhoeff, 1924 is redescribed from type and new material, a lectotype is designated and Brochopeltis mjoebergi queenslandica Verhoeff, 1924 is synonymised with Brochopeltis mjoebergi. Brochopeltis mediolocus sp. n. is the first native paradoxosomatid described from Australia’s Northern Territory.


Introduction
Paradoxosomatid species in the Australian genera Brochopeltis Verhoeff, 1924, Helicopodosoma Verhoeff, 1924and Tholerosoma Mesibov, 2006 have unbranched gonopod telopodites with the prostatic groove opening at the telopodite tip. The taxonomic placement of these genera is uncertain. Brochopeltis and Helicopodosoma species have a medial process on the male leg 1 femur, a diagnostic feature of the subfamily Australiosomatinae. Both genera were placed in the tribe Antichiropodini within Australiosomatinae by Jeekel (1968). However, Jeekel (1979: 652) later suggested that Brochopeltis, Helicopodosoma, Australodesmus Chamberlin, 1920 andMjoebergodesmus Verhoeff, 1924 were distinctive enough to merit each being placed in its own tribe.
Tholerosoma species lack a process on the male leg 1 femur (Mesibov 2006), and I left the genus without a subfamily assignment in my original description.
The three genera are hard to place because the gonopod is so simple in structure. There are no obvious clues to intergeneric relationships in the non-gonopodal characters, and gonopod simplification could have occurred in more than one ancestral lineage. Verhoeff (1924: 33) clearly recognised this problem when discussing Brochopeltis: "...it is wrong to judge genera solely according to the gonopods, particularly when these, as is here the case, are so simplified through secondary regression of the tibiotarsus, that very similar organs could also arise in this way in related genera independently of one another" (my translation).
Gonopod simplification also makes it difficult to place a recently discovered species from Australia's Northern Territory in a genus. In this paper I tentatively assign this species to Brochopeltis because of similarities to B. mjoebergi Verhoeff, 1924 in both gonopod and paranota structure. I also redescribe B. mjoebergi and designate a lectotype for it.

Materials and methods
"Male" and "female" in the text refer to adult individuals. Specimens are stored in 70-80% ethanol in their respective repositories. Gonopods were cleared in 80% lactic acid and temporarily mounted in a 1:1 glycerol:water mixture for optical microscopy. Body measurements were estimated with a Nikon SMZ800 binocular dissecting microscope using an eyepiece scale. Colour images were manually stacked using a Canon EOS 1000D digital SLR camera mounted on the Nikon SMZ800 fitted with a beam splitter, then processed with Zerene Stacker 1.04. Figs 5B and 5C were captured as screenshots from the output of a 1.3 megapixel digital video eyepiece camera mounted in one ocular tube of a Tasco LMSMB binocular microscope. Preliminary gonopod drawings were traced from prints of screenshots captured in the same way. Images and drawings were prepared for publication using GIMP 2.8.

Order
Male with vertex and frons sparsely setose, clypeus moderately setose; vertigial sulcus distinct, ending at dorsal level of antennal sockets; post-antennal groove shallow; antennal sockets separated by ca 1.3× socket diameter. Antenna filiform, reaching dorsally to rear of ring 3; antennomeres with relative lengths (2=3)>(4=5=6); 6 apically widest. In dorsal view, head narrower than collum paranota; relative ring widths collum < (2=3=4) < (5 to17). Collum with lateral margin strongly produced as paranotum, anterior and lateral margins smoothly convex, posterior margin more or less straight. Paranota on haplo-and diplosegments with margins thickened dorsally, so dorsal paranotal surface appears slightly depressed. Ring 2 paranotum with lateral margin slightly lower than lateral margins of collum and ring 3 paranota (Fig. 1B); posterior corner produced posteriorly as broad triangle. Paranota of rings 3 and 4 curving posteriorly, almost sickle-shaped, posterior corners bluntly rounded. Paranota on diplosegments 5-17 set at ca 3/4 ring height, directed slightly dorsally with posterior corner highest; anterior margin curving smoothly into nearly longitudinal lateral margin, posterior corner strongly produced as bluntly pointed triangle almost reaching waist of next ring; lateral margin thicker on pore-bearing rings. Paranota reduced but still prominent on rings 18 and 19, with strongly produced posterior corners. Pleural keels on rings 2-4 only, on rings 3, 4 reduced to small, rounded, posterolaterally directed processes. Prozonites and metazonites bare and finely textured, giving dull appearance; transverse furrow at ca 1/2 metazonite length, distinct, ca 2/3 ring width, not reaching paranotal base; waist short, shallow, faintly sculptured with low longitudinal ridges; limbus a narrow, thin, continuous sheet. Pore formula normal; ozopore small, round, opening laterally on thickened paranotal margin almost at level of posterior metazonite margin. Spiracles on diplosegments above and just anterior to leg bases; spiracular filters forming rounded fold in inverted, tight U-shape in spiracular opening; anterior spiracle with anterodorsal portion of rim produced to partly cover strongly emergent, anterodorsal portion of filter. Midbody sternites very sparsely setose, as wide as long, transverse impression more distinct than longitudinal impression; no cones or projections on any sternites. Midbody legs slender, relative podomere lengths femur>tarsus >prefemur>(postfemur= tibia); femur ca 1.3× as long as tarsus; anterior leg prefemora not swollen dorsally. Pre-anal ring sparsely setose; epiproct extending past anal valves, in dorsal view tapering and truncate, tip ca 1/4 width of pre-anal ring; hypoproct paraboloid; spinnerets in rectangular array, much wider than long.
Gonocoxa short, the anterodistal surface with low, rounded protuberance bearing sparse, long setae on distal side. Prefemur large, C-shaped, the distal end projecting posterolaterally as rounded extension reaching ca 1/5 telopodite height; long setae on posterior and posteromedial surfaces of prefemur. Cannula small, arising from gonocoxa apex. Telopodite beyond prefemur without branches, the basal portion straight and slightly expanded distally; at ca 2/3 telopodite height, telopodite constricted, flattening and curving anterolaterally in wide spiral to level of starting point of curve and anterior to it, then curving anterolaterally, the apex slightly expanded with distal margin rounded. Prostatic groove (Figs 3A, 3B) running straight on medial surface of basal portion of telopodite beyond prefemur, then following curve of telopodite to open at apex. Female without leg modifications; epigynum 1/4-1/3 ring width, slightly raised in small rounded triangle medially; cyphopods not examined.
Distribution. In forest litter within a range envelope of ca 1500 km 2 on and near the Atherton Tableland, in the Wet Tropics of far north Queensland (Fig. 4).
Remarks. Types. Verhoeff (1924: 35) reported that apart from a few pairs of B. mjoebergi ("[a]usser einigen Pärchen") from Atherton collected in January, he examined three males and four females from Herberton and one male and a juvenile from "Cedar Creek" (Ravenshoe). B. mjoebergi queenslandica was based on two males from Bellenden Ker. All collections were by Erik Mjöberg in early 1913 (Ferrier 2006).
Most of the syntypes are accounted for, with a few discrepancies. The NHRS material comprises parts of two males and eight females from Atherton; parts of four (not three) males, four females and a juvenile from Herberton; and parts of two males from Bellenden Ker. ZSM has parts of at least two males from Atherton (J. Spelda, in litt.), and ZMB has one male labeled "Queensland" collected by Mjöberg in 1913 (J. Dunlop, in litt.). The latter may be the Ravenshoe male but if so the Ravenshoe juvenile appears to be missing.
The lectotypifications by P.M. Johns (see label information in types section, above) were never published. I have designated an entire male from Herberton in the type series as the B. mjoebergi lectotype because the slide-mounted gonopods of Atherton males are distorted (see also below, on subspecies queenslandica).
Species epithet. The spelling mjöbergii on p. 33 in Verhoeff (1924) is apparently a typesetting error. The name is spelled mjöbergi on pp. 35, 133 and 138 with Brochopeltis, and an additional 26 times in Verhoeff (1924) with the new taxa Cyliosoma queenslandicum mjöbergi, Monographis mjöbergi, Poratobolus mjöbergi, Rhinotus mjöbergi and Siphonophora mjöbergi. The spelling mjöbergi is also used on all but one of the handwritten Verhoeff labels I have seen.
Subspecies queenslandica. Verhoeff distinguished this subspecies mainly on minor variation in colour pattern, writing "Structure otherwise as in the preceding form, the gonopods also agreeing with those of the other, but the solenomere bends not sharply bent, but totally rounded, thus even more strongly spirally curved" (Verhoeff 1924: 35, my translation). The supposed gonopod difference is an artefact produced during   Verhoeff, 1924 (circle). Inset. Localities for B. mjoebergi Verhoeff, 1924 (triangles) in far north Queensland; type localities are named and buffered with 5 km-radius circles. Mercator projections. slide preparation (Fig. 5). Verhoeff mounted the spiral acropodites of the gonopods of the mjoebergi queenslandica lectotype more or less in the plane of the spiral, while those of the mjoebergi mjoebergi paralectotypes are mounted with the spirals compressed and bent. The intact gonopods of the mjoebergi queenslandica paralectotype spiral in just the same way as gonopods of mjoebergi mjoebergi from Herberton. The subspecies queenslandica is here made a synonym of the nominate subspecies.
Other notes. B. mjoebergi appears to be abundant in rainforest and open forest on the western side of the Atherton Tableland (Fig. 4). I have not seen this large and easily recognised millipede in collections from elsewhere in the Queensland Wet Tropics. The region has been entomologically well sampled since Mjöberg's expedition and is home to many other described and undescribed paradoxosomatids.
Diagnosis. Differs from B. mjoebergi in distal portion of gonopod telopodite bent but directed distally, not curving in wide spiral; with dorsum uniform in colour, not with pale, paramedian longitudinal bands and pale paranota on a darker background; and with tarsal brushes on all male legs except last two pairs, rather than on legpairs 1-7 only.
Description. Male/female approximate measurements: length 29/30 mm, midbody paranota width 3.6/3.8 mm, prozonite width 2.7/3.0 mm, maximum vertical diameter 2.6/3.0 mm. Live, well-coloured animals more or less uniformly dark brown in body colour, shiny (Fig. 6A). In alcohol, body colour brown (Fig. 6B, C), darker on paranota and posterior metazonite margin, lighter ventrally and in pleural keel area; head and antennae dark brown with lighter spot just above antennal socket; legs darker than body.
Leg 1 (Fig. 7C) with small, short, rounded process at ca 1/2 length of medial femur surface, directed mediodistally and slightly anteriorly. Gonopore small, round, opening on short distomedial bulge of leg 2 coxa. Sternal lamella (Fig. 7D) ca 90% of width between leg 4 bases, leaning slightly anteriorly; lateral margins straight, vertical; corners rounded; ventral margin medially incised. Dense brush setae on tarsi of all but last 2 legpairs, on some anterior legs also at distal end of tibia.
Gonocoxa short, the anterodistal surface with low, ridge-like protuberance bearing sparse, long setae on distal side. Prefemur large, C-shaped, the distal end projecting posterolaterally as rounded extension reaching ca 1/4 telopodite height; numerous long setae on posterior and posteromedial surfaces of prefemur. Cannula small, arising from gonocoxa apex. Telopodite beyond prefemur without branches, the basal half straight and slightly expanded distally; at ca 2/3 telopodite height, telopodite flattening slightly and curving anterolaterally, then constricting and bending sharply anteriorly, curving mediodistally and flattening further, the apical margin rounded distally with lateral margin produced as small triangle. Prostatic groove (Fig. 3C, D) running straight on medial surface of basal half of telopodite beyond prefemur, then following bends and curves of telopodite to open at tip of apical triangular projection.
Female without leg modifications; epigynum ca 1/4 ring 2 width, very slightly raised medially in small rounded triangle; cyphopods not examined.
Distribution. So far known only from the type locality, a farm ca 50 km southeast of Darwin in the monsoon tropics of Australia (Fig. 4).
Remarks. I am tentatively assigning this species to Brochopeltis not only because the gonopods are similar, but because B. mediolocus sp. n. and B. mjoebergi share two features which I have not yet noted in other Australian Antichiropodini. One is the lifting and extension of the lateral collum margins as paranota. The second possible synapomorphy is the pronounced dorsal thickening of paranotal margins.
The types were collected on a farm and it is possible that B. mediolocus sp. n. is not locally native, but has been introduced to Middle Point from elsewhere in tropical Australia. The only previous record of Paradoxosomatidae from the northern portion of the Northern Territory (Australia's "Top End") is of the introduced Asian species Orthomorpha coarctata (De Saussure, 1860) in urban Darwin (Jeekel 1982).