Scelomyza Séguy, 1938

ROHÁýEK, JindĜich, 2014, The identity and family affiliation of Scelomyza hirticornis (Diptera: Opomyzidae, Anthomyzidae), with a new checklist of Afrotropical Anthomyzidae, Acta Entomologica Musei Nationalis Pragae 54 (1), pp. 383-398 : 385-388

Genus Scelomyza Séguy, 1938

Scelomyza Séguy, 1938: 338 [feminine]; STUCKENBERG (1980): 635 (catalogue).

Type species. Scelomyza hirticornis Séguy, 1938: 339 (original designation).

Diagnosis. (1) Head distinctly higher than long, rounded in pro¿le. (2) Eye large, irregularly subovoid, with longest diameter subvertical. (3) Frons very short, not projecting in front of eye; (4) frontal triangle long, reaching to anterior margin of frons, shining. (5) Frontal lunule small but distinct. (6) Occiput distinctly concave and large. (7) Ocellar triangle and ocelli very small, shifted posteriorly, behind top of vertex. (8) Antenna strongly geniculate, pedicel capiform, 1st flagellomere anteriorly emarginated and with extremely long pale setae on apex. (9) Arista basally densely haired, distally long-ciliate (subpectinate). (10) Palpus orange, with 1 longer subapical seta and a few setulae on apex. Cephalic chaetotaxy: (11) pvt inclinate but very small and widely spaced; (12) vti, vte and oc probably relatively short and weak; (13) only 1 strong and long ors, situated near vertex, and 2 microsetulae in anterior part of orbit; (14) postocular setulae short, in single row; (15) 1 distinct, strongly curved (inclinate) vi, subvibrissa small, hair-like, not distinguished from peristomals; (16) peristomal setulae small, relatively numerous. (17) Posterior corner of head (postgena) enlarged but rounded. (18) Sexual dichroism of antenna and face cannot be assessed because male is unknown.

(19) Thorax distinctly narrower than head, shining, with very sparse microtomentum. Thoracic chaetotaxy: (20) 1 reduced hu, 2 npl (anterior longer); (21) no prs; (22) 1 sa (probably small), 1 relatively shorter pa; (23) 2 postsutural dc, both long and strong; (24) ac microsetae entirely absent; (25) 2 sc (apical moderate, laterobasal reduced, hair-like); (26) ppl microseta not seen but probably present; (27) 2 distinct stpl (posterior longer). (28) Mid and hind legs unicolourous yellow, fore leg yellow but with tibia and basitarsus blackish brown. (29) f

1

apparently with ctenidial spine; (30) fore basitarsus with very dense ventral setulae and those proximal prolonged; (31) t 2 with short ventroapical seta; (32) armature of male f 3 not described because male sex unknown. (33) Wing relatively short and narrow; (34) wing membrane unicolourous. (35) C with distinct spinulae among ¿ne hairs on Cs 2; (36) R 2+3 long, bent parallel to C; (37) R 4+5 very slightly bent (recurved) and subparallel to M apically, ending in C about equidistant from apices of R 2+3 and M; (38) dm cell long and narrow; cross-vein r-m situated in front of middle of dm cell; (39) distal part of CuA 1 slightly longer than dm-cu, reaching wing margin; (40) A 1 very short, ending far from margin. (41) Anal lobe reduced and alula small, very narrow.

Male abdomen and male genitalia unknown.

(42) Female abdomen shining blackish brown to black, with large, broad terga ( T2 – T6 ) and narrower sterna (S2–S5). (43) Postabdomen relatively short, basally (6th segment) broad, caudally strongly tapered, with well-developed sclerites. (44) T6 very broad, also S6 relatively large and broad. (45) T7 and S7 fused, forming ring-shaped tergosternum T7 +S7; (46) S8 medially longitudinally divided, laterally lobately extended anteriorly. (47) Internal structures of female genital chamber (uterus) weakly sclerotized, formed by 1 pair of partly membranous and unpigmented posterior plates and (48) 1 anteroventral, very thin, compressed and looped annular sclerite. (49) Anterior (distal) part of uterus, including ventral receptacle and accessory glands undescribed (because lost in holotype). (50) Spermathecae (probably 1+1) subspherical, with simple surface and very long sclerotized collar; (51) spermathecal duct apparently long. (52) T10 small, short, transversely bent saddle-like, bare except for 1 pair of dorsal setae; (53) S10 markedly larger than T10 , unusually long, somewhat trough-shaped, with setae and micropubescence both reduced. (54) Cercus relatively short and robust, with a number of longer ¿ne setae .

Discussion. At ¿rst glance, Scelomyza hirticornis really does resemble a very small species of Geomyza Fallén, 1810 . No wonder that SÉGUY (1938) placed it in the Opomyzidae , particularly when he observed only 1 ors, the vibrissa absent (both these characters being apomorphies of the family) and the strongly reduced anal lobe of the wing. However, the vibrissa is, in fact, present (although strongly inclinate so that it was overlooked by SÉGUY 1938, cf. Fig. 8 ) in the holotype and the anterior ors are also present though reduced to 2 microsetulae (as often occurs in Anthomyzidae ). Two further apomorphies of Opomyzidae , viz. the setulose prosternum and the female 7th abdominal spiracle absent, also do not occur in Scelomyza (see above). On the contrary, practically all apomorphies of the Anthomyzidae and also of the recent subfamily Anthomyzinae (except those of the male which remain unknown in Scelomyza ) are present in Scelomyza , viz. the convergent pvt (although very reduced and also overlooked by Séguy), no pteropleural setulae, the f 1 with a ctenidial spine (broken off in the holotype, thus not seen by Séguy), the postocular setulae in single row, the mesopleural setae absent, 2 long stpl and 1 pa seta. Consequently, the af¿liation of Scelomyza with Anthomyzidae is beyond doubt; moreover, this is also con¿rmed by the female postabdominal structures of S. hirticornis (see below).

The genus Scelomyza can be best diagnosed by several striking and obviously apomorphic features which appear to be unique in the family Anthomyzidae : the unusual formation of the head with a shortened frons (character 3 in the above diagnosis), the ocellar triangle reduced and shifted posteriorly (7), the 1st antennal flagellomere with extremely long apical setae (8), the single and unusually strong ors (13), the very enlarged postgena (17), the entirely absent ac microsetae (24) but also the lobately enlarged lateral parts of S8 (46), the very long collar of the spermatheca (50) and the unusually long S10. When these autapomorphies are combined with other characters rarely occurring in Afrotropical Anthomyzidae (e.g. strongly reduced subvibrissa and hu, absent prs, ctenidial spine present), Scelomyza seems to be a distinct and well-characterized genus.

However, searching for its relationships is more dif¿cult, particularly when structures of the male genitalia remain unknown. The formation and structures of the female postabdomen of Scelomyza indicate a relationship with the apterous, ant-mimicking Apterosepsis Richards, 1962 (redescribed and placed in Anthomyzidae by ROHÁýEK 1998). This genus, although with yet more highly modi¿ed morphology and chaetotaxy than in Scelomyza , shares with Scelomyza a similarly formed tergosternum T7+S7 and S8 (although the latter is not laterally extended), a short bent T10, a relatively long S10, a simply ovoid spermatheca with a long collar (see ROHÁýEK 1998: Figs 27–32) but also a ctenidial spine, a reduced pvt, a single ors and even the dark brown tibia and basitarsus contrasting with the remaining yellow parts of the fore leg ( ROHÁýEK 1998: Figs 23, 25). Unfortunately, the male is unknown in Apterosepsis as well, but this genus differs markedly from Scelomyza in other head and thoracic chaetotaxies, in the formation of the abdominal sclerites including the circular annular sclerite in the female genital chamber (cf. ROHÁýEK 1998). It should be noted that the complete ring-shaped female tergosternum T7+S7 is also known in another Afrotropical genus, Barbarista Roháþek, 1993 , but this group of species differs markedly from Scelomyza in head (including antennae) and thorax (including fore leg and wing) morphology and chaetotaxy not to mention additional structural differences in the female postabdomen (see ROHÁýEK 1993).

In conclusion, Scelomyza appears to share many more characters with Apterosepsis than with other named Afrotropical genera of Anthomyzidae (for their descriptions see ROHÁýEK 1993, 2004; ROHÁýEK & BARRACLOUGH 2003) and could be the closest relative of the latter. Considering the characters of the female terminalia, particularly as regards the structures of the genital and postgenital segments (including S8, internal sclerotization of female genital chamber, T10, S10), tendencies toward reduction of the ors, bicoloured fore legs and retention of the ctenidial spine, it seems probable that both these genera are related to the Chamaebosca clade (Neotropical-Holarctic) as de¿ned by ROHÁýEK & BARBER (2009), even resembling in habitus the species of the Mumetopia nigrimana group. Note: the latter group is, in fact, most closely allied to the genus Stiphrosoma Czerny, 1928 as rede¿ned by ROHÁýEK & BARBER (2005), and not to Mumetopia Melander, 1913 to which it has hitherto been af¿liated (for more detail see ROHÁýEK & BARBER 2009). To test this proposed close relationship of Scelomyza and Apterosepsis to the Chamaebosca clade more precisely, it is necessary to discover males of S. hirticornis and A. basilewskyi Richards, 1962 , the only known species of these two Afrotropical genera and to study their postabdominal and genital structures.

Species included. Scelomyza hirticornis Séguy, 1938 (redescribed below).

Figures

Figs 5–9. Scelomyza hirticornis Séguy, 1938, female holotype. 5 – head sublaterally, reconstruction; 6 – left mid femur and tibia, posteriorly; 7 – left fore trochanter, femur, tibia and basitarsus, posteriorly, chaetotaxy of femur reconstructed; 8 – original illustration of head by SÉGUY (1938), laterally; 9 – ditto, frontally.All scales = 0.25 mm. For abbreviations see text (p. 385).