Baeus Halliday Baeus seminulum Haliday Anabaeus Ogloblin Baeus (Anabaeus) ventricosus Ogloblin Baeus Haliday Angolobaeus Kozlov Parabaeus machadoi Risbec Hyperbaeus Foerster Baeus seminulum Haliday Baeus Haliday Paraneurobaeus Risbec Paraneurobaeus arachnevora Risbec Psilobaeus Kieffer Baeus curvatus Kieffer Systematics, distribution and biology of the Australian ' micro-flea' wasps, Baeus spp. (Hymenoptera: Scelionidae): parasitoids of spider eggs. Stevens, N. B. Austin, A. D. Zootaxa 2007 1499 1 45 62FJR urn:lsid:biosci.ohio-state.edu:osuc_concepts:455 Insecta Scelionidae Baeus CoL Animalia Baeus Haliday Hymenoptera 6 Arthropoda genus   Baeus Halliday, 1833: 270. -Ashmead, 1893: 167; Kieffer, 1926: 146; Galloway & Austin, 1984: 86; Austin, 1988: 88; see Johnson, 2004, for complete bibliography [Type species, by monotypy, Baeus seminulum Haliday, 1833].  Anabaeus Ogloblin, 1957: 440 -[Type species, by designation, Baeus (Anabaeus) ventricosus Ogloblin, 1957]. Proposed as a subgenus of Baeus Haliday.  Angolobaeus Kozlov, 1970: 218. -Masner, 1976: 67; Johnson, 1992: 336 [Type species, by designation, Parabaeus machadoi Risbec, 1957]. Syn nov.  Hyperbaeus Foerster, 1856: 144. -Masner, 1976: 67; Johnson, 1992: 457 [Type species, Baeus seminulum Haliday] Unnecessary replacement name for Baeus Haliday.  Paraneurobaeus Risbec, 1956: 821 - Masner, 1976: 67; Johnson, 1992: 457 [Type species, by designation, Paraneurobaeus arachnevora Risbec, 1956]. Syn nov.  Psilobaeus Kieffer, 1926: 151 - [Type species, by monotypy, Baeus curvatus Kieffer, 1910] Synonymy by Masner (1965).  Diagnosis. Female. Head much wider than mesosoma, closely abutted to pronotum, marginally wider than metasoma (Fig. 1 1D); lateral ocellus much closer to posterior margin of eye than to median ocellus (Fig. 4B); hyperoccipital carina distinct along dorsal-posterior margin of vertex; occiput vertical and concave, not normally visible dorsally unless head flexed forward, always forming acute angle with vertex; posterior margin of vertex concave medially; in dorsal and lateral views, firons rounded; in anterior view, head ovoid in shape; antenna 7-segmented (Fig. 6A), with 4 funicle segments, F1 about twice as long as each of F2-4 which are transverse, clava unsegmented (fine suture lines present on dorsal surface only), densely pilose (Fig. 6A), scape not reaching to median ocellus; frontal carina incomplete, not reaching to median ocellus (Fig. 13D); antennal insertion positioned below ventral margin of eyes, between ventral margin of firons and dorsal margin of clypeus; clypeus with transverse furrow just dorsal to its ventral margin, furrow with 6 bristles along dorsal margin; mandibles tridentate; maxillary and labial palps 1-segmented, labial palps wart-like, bearing one bristle, maxillary palps clavate, having three bristles apically; dorsally firons coriarious; malar region with cristulations seemingly emanating from anterior tentorial pits (Fig. 13D); gena broad and distinct (Fig. 4A and 9E). Mesosoma. Compact, higher and wider than long (Fig. 10A and 11D); both fore- and hindwings reduced to minute sclerotized plates that are often overlooked because they neatly fit into recesses beneath lateral margins of mesoscutum and mesoscutellum respectively, only marginally larger than tegula (Fig. 4A and 7B-F); pronotum visible latero-dorsally, but not medio-dorsally, dorso-posterior corner of pronotum bearing protuberance(Fig. 7E); mesoscutum much wider than long; mesoscutellum transverse, approximately one-third length of mesoscutum (Fig. 4B); metanotum greatly reduced, hidden beneath mesoscutellum so not visible (cf. Figs 4A and B to Figs 5A and B); posterior surface of propodeum vertical, dorsal surface forming thin transverse band, posterior margin extending over anterior margin of T2 (Figs 4A and B); dorsal metapleuron fused with propodeum to varying degrees (Figs 7A-F); coxae large, anterior surface densely covered with stout bristles; legs longer than mesosoma and metasoma combined, hind leg longest; hind leg femoral spine present or absent (sometimes difficult to see even at high magnification) (Figs 6C and D).  FIGURE 4. Female Baeusspp.: A, lateral habitus of B. arthuri; B, dorsal habitus of B. murphyi. Scale lines, A & B = 100 µm. fc = frontal carina; fw = forewing remains; ge = gena; hw = hindwing remains; hyo = hyperoccipital carina; lo = lateral ocelli; lt = laterotergite; mo = medial ocellus; mp 1 & 2 = mesopleuron & metapleuron repectively; ms = mesoscutum; pn = pronotum; pr = propodeum; pr sp = propodeal spiracle; sc = mesoscutellum; tg 1 & 2 = forewing & hindwing tegula, resectively; T2, 3 & 4 = tergite 2, 3 & 4 respectively. Metasoma. Short, broadly abutted against vertical posterior surface of propodeum so body appears fused (Figs 4A and B); T1 and S1 forming vertical anterior surface of metasoma, T1 not visible dorsally (Fig. 3A); T2 largest tergite, occupying>0.6 of dorsal surface of metasoma (Figs 4A and B), glabrous band present along posterior margin; laterotergites wide, ventral margins free, not incised into a submarginal groove (cf. Fig. 3A to 3B). Male. Body not rounded and fused, division between mesosoma and metasoma obvious (Fig. 5); dorsal surface of mesosoma well above that of metasoma; metasoma petiolate; sculpturing of all dorsal sclerites similar to female, usually more pronounced, except for propodeum and anterior margin of T1 which are visible and generally confused-rugulose and crenulate-costate, respectively; pilosity generally longer than female conspecifics but not as dense; colour similar to female conspecifics. Head. Wider than mesosoma, but not as wide in relation to mesosoma as in female, closely abutted to pronotum; hyperoccipital carina distinct along dorsal-posterior margin of vertex; occiput vertical and concave, often forming acute angle with vertex (as in female); eyes not as large as in female, but more distinctly bulging from face; occelli more prominent; antenna 11 or 12-segmented, F9 and F10 sometimes fused or separated(Fig. 6B). Mesosoma. More quadrate than transverse, length greater than width and height, though only marginally (Fig. 5A and B); mesoscutum wider than long; mesoscutellum often semi-ovoid in dorsal view, approximately half length of scutum, projecting posteriorly above metanotum; metanotum transverse, visible posterio-dorsally; propodeum with exposed oblique sculptured surface in anterior dorsal part, posteriorly surface more vertical and abutted against T1; sculpturing of dorso-lateral regions of meso- and metapleuron confused; wings present and macropterous (Fig. 2); forewing narrow so that anterior and posterior margins almost parallel; submarginal (Sc+R) + marginal (R1) veins only reaching about 0.4 times along wing length, marginal vein short, stigmal vein (r) much longer than marginal vein, basal vein (Rs+M) present as thickened infuscate band; setal fringe present and long except for proximal posterior margin; hindwing extremely narrow, setal fringe on posterior margin much longer than maximum wing width. Metasoma. Petiolate, short, usually shorter than mesosoma; T1 clearly visible dorsally; T2 the largest tergite but not as dominant as in female; ventral surface often collapsed inwards so concave; laterotergites free and wide as in female, but difficult to see when metasoma has collapsed ventrally.  Monophyly and relationships of Baeus The monophyly of Baeusis supported by eight critical characters associated with the female sex: 1) the female metasoma is short, convex, wide anteriorly and broadly abutted against the vertical surface of the mesosoma so the body appears to be rounded and fused (Figs 4A & B), 2) T1 and S1 are wafer thin and form the semi-vertical anterior surface of the metasoma (Fig. 3A); because of this arrangement, T1 is not visible dorsally (Fig. 4B), 3) T2 is the largest tergite, comprising at least 0.6 times the length of the metasoma (Fig. 4A & B), 4) the laterotergites of the metasoma wide and free, not incised into a submarginal groove as in Mirobaeoides(Fig. 3A cf. 3B), 5) the propodeum is vertical and forms the whole of the posterior surface of the mesosoma, 6) the metanotum is reduced and hidden below the mesoscutellum, 7) the maxillary and mandibular palps are reduced to a single segment, and 8) the wings are micropterous, i.e. both fore- and hindwings reduced to minute sclerotized plates, the forewing only slightly larger than the tegula, therefore, often referred to as ‘wingless’ (Fig. 4A, 7B-F). In addition, the male is characterized by having the forewing narrow so that the anterior and posterior margins are almost parallel, the submarginal + marginal veins only about 0.4 times the length of the wing, and a basal vein is usually present (Fig. 2). Of the female characters above, a rounded, apparently fused body (character 1) is found in the genera Mirobaeoides, Neobaeusand two undescribed baeine genera from southern Africa (in CNC); the laterotergites being wide and free (character 4) is found independently in Neobaeus, Tiphodytes Bradley( Thoronini) and all members of the Telenominae(Masner 1976), while the forewings being reduced to a tiny sclerotized flaps (character 8) is found in numerous taxa across the Scelionidaeand Platygastridae, particularly those associated with soil and litter habitats (Austin et al. 2005). The other characters listed above are all apparently unique to Baeus. The phylogeny of the Baeiniwas first examined by Iqbal and Austin (2000a) using a morphological dataset for exemplar species representing virtually all described genera as well as five putatively undescribed generic level taxa. Parsimony analysis resolved Baeusas monophyletic but only with the inclusion of Apobaeus Ogloblin, a potential synonym of Baeusfrom the Neotropics. To date we have been unable to examine the type species of Apobaeus( Tetrabaeus insularis Ogloblin), and the generic status of this taxon remains unclear. In this analysis, virtually all ‘wingless’ taxa formed a monophyletic group, with Baeus+ Apobaeusbeing placed apically and sister to Neobaeusand then Mirobaeoides. However, removal of wing size character states from the dataset resulted in a largely comb-like tree, and pointed to the fact that in the absence this character, which is notoriously homoplastic, there were few informative characters to infer relationships among genera. However, as partly outlined in the Introduction, a recent molecular study (Carey et al. 2006), including23 baeine species representing seven genera and eight non-baeine scelionids, resolved Baeusas monophyletic but placed it in a more basal position as sister to Odontacolus+ Hickmanella+ Idris+ Ceratobaeus, with the latter two genera being polyphyletic. Further, in this analysis the Baeiniwas not monophyletic as Mirobaeiodes+ Neobaeuswere resolved with other scelionid genera. These results have been confirmed within a broader sampling of taxa, and point to the fact that the Baeiniis not monophyletic and that scelionids have probably exploited spider eggs independently at least twice (Murphy et al. in press).  Comments on non-Australian genera Prior to the commencement of the current study, one of us (ADA) was able to examine the types of two monotypic baeine genera, Angolobaeus Kozlovand Paraneurobaeus Risbec. Both of these genera are here proposed as junior synonyms of Baeus. Kozlov (1970) erected Angolobaeusfor Parabaeus machadoi Risbecfrom Angola (Risbec 1957) (holotype in MRAC) on the basis that the type species has the vertex and firons with two “mound-like projects” on each. Parabaeus Kiefferis morphologically highly convergent with Baeus-likescelionids, but the genus belongs to the Platygastridaeas recognized by Masner (1976) (see also Austin 1988; Manser & Huggert 1989). Kozlov (1970) correctly placed P. machadoias a scelionid and a member of the Baeini. However, there is no justification for the generic status of this species; it is congeneric with Baeusin every respect except for the cephalic projections, which we propose are simply autapomorphic and of importance only at the species level. A range of cephalic processes also occur on the head of some Telenomus( Telenominae) (Mineo 1979; Johnson 1980; Huggert 1983) and the vertex and eyes of some Platygastridae(Austin 1984), where they are also only of species level importance. Risbec (1956) described Paraneurobaeusfor a species reared from spider eggs collected at Garoua (Cameroon) that appeared to have a 6-segmented, rather than a 7-segmented, antenna. The type species, Paraneurobaeus arachnevora Risbec, is represented by a syntypic series (MNHN) mounted on two microscope slides. Unfortunately, Risbec’s description is inaccurate and his drawings look nothing like the slide mounted specimens that are in reasonable condition. Examination of the type series shows in fact that the antenna is 7- segmented and that three (not two) tiny funicle segments are present in addition to the larger basal funicle segment, as is the case in Baeusand most other members of the tribe. In every respect, this species is congeneric with Baeus. The specimen on one slide with the head attached we here designate as the lectotype of P. arachnevora, and the remaining specimens that are slightly broken as paralectotypes.  Biology Although there are numerous series of Baeusreared from spider eggs in world collections, in most cases the host spider has not been identified. Austin (1985) summarised all available host information for parasitoids associated with spider eggs and, worldwide, 22 species of Baeushave host data recorded to at least family level. Although the data are limited, it does point to this genus parasitising a narrower spectrum of spider host families compared with the largest genus of baeines, Idris. Where Baeushas been reared from six spider families, more than 70% of which come from just two families ( Araneidae- 9 records and Theridiidae- 7 records), Idrishas been reared from 11 host families (22 records). Interestingly, the spider hosts for Baeushave uniformily more complex structured eggsacs compared with the hosts of Idris, in that they have either dense flocculent silk walls or have multiple layers of dense and flocculent silk (types 2 and 3, viz. the eggsac classification proposed by Austin 1985). Further, it has been postulated that the highly modified body shape of Baeusmay also function as an adaptation for the female wasp to burrow through the silk wall of the eggsac to reach and oviposit into the eggs within (Austin 1988; Austin et al. 2005). However, some caution should be exercised in not over-interpreting the above apparent difference as there is likely to be some bias in the host data for Baeusin that most records come from spider hosts collected from above-ground vegetation (where eggsacs are easy to locate and collect), rather than from soil and leaf-litter where Baeusappear to be more abundant. For Australia, Baeushave been reared from three spider families (Table 2); Araneidae, Lynyphidaeand Theridiidaewith four of the six records being from the araneid genera Araneus, Argiope, Celaeniaand Cyrtophora, all of which produce eggsacs attached to vegetation. Individual host data are given below for two Baeusspecies, however for the other four host records the material was not available to examine and the species had not been identified.  Distribution. The greatest diversity of Australian Baeusspecies exists within the peripheral mesic environments of the continent, a common pattern exhibited by their hosts (Raven 1988). Of the 20 species treated here, 55% are endemic to a single biogeographic subregion, with the highest level of endemism found in the Kosciuskan subregion (25%) (Fig. 1). The eastern seaboard of the continent, from northern Queensland to southern New South Wales, exhibit the greatest species richness (Fig. 1) and share the most number of species (Table 3). The lowest number of species occurs in the South-Western and Eyrean subregions which have no endemic taxa, although these areas are also the poorest surveyed. In addition, we have recognized a further eight likely new species, but have refrained from describing them because of the limited material available and/or their poor condition. Five of these species are from the Kosciuskan subregion, two of which also occur in other subregions, one in the South-West, the other extending into the Torresian subregion. The Tasmanian, Timorian, and Torresian subregions each have a single undescribed species. It is likely that numerous additional species will be discovered in the future as more long-term collecting using a range of techniques is undertaken in specific habitats. Although beyond the scope of the current study, available material of Baeusfrom New Zealand and islands surrounding Australia were also examined. Apparent from this is a close affinity of the neighbouring Pacific fauna with the Australian fauna. For instance, New Zealand has a relatively small fauna comprising five species, two of which also occur in Australia ( B. leaiand B. saliens). Of the remaining three species, two are similar to B. murphyiand B. saliens, and may be indicative of allopatric speciation events. The Baeusspecies found to occur on several neighbouring Pacific islands from which material was available (Norfolk Island, Lord Howe Island, New Caledonia, and Fiji) all belong to species also present in Australia. This close affinity is in stark contrast to the Baeusfauna of Christmas Island in the Indian Ocean, some 2000 km to the north-west of Australia. Of the five species found to occur on Christmas Island, only one, B. tropaeumusbrevisis also present in Australia. Interestingly, B. tropaeumusbrevisin Australia has only been collected from the north-west of Australia, the closest region of the Australian mainland to Christmas Island. This difference between the Christmas Island and Australian fauna’s is hardly surprising given that the islands of the Sunda Arc to the north, particularly Java and Sumatera, are the closest landmasses to Christmas Island, and the direction of the prevailing winds may facilitate colonisation from Indonesia. Study of the Baeusfauna of the neighbouring Indonesian islands may reveal a close affinity with the Christmas Island fauna, as Smithers (1995) has documented for the Psocoptera. The distribution of B. tropaeumusbrevisappears to be an example of colonisation of Australia from the north-west.  Key to females of Australian Baeusspp. Please note, because of the small size of the specimens to be identified, the minimum magnification required to view the characters used in this key is 80x's. In addition, an eyepiece graticule is required because measurements of various characters and their states are relied upon in this key. 1. Hind femoral spine large,> 15 µm in length, clearly visible under stereo-light microscope (Fig. 6C).....2 - Hind femoral spine absent or reduced (<15 µm), not readily visible under stereo-light-microscope (Fig. 6D).................................................................................................................................................3 2(1). Propodeal spiracle opening large, = 20 µm in diameter, forming a volcanic-like cone (Figs 14E & F); frontal carina reaching <0.5 distance to medial ocellus (e.g. Fig. 13B)................. B. vulcanus sp. nov. - Propodeal spiracle not as large as above, <15 µm in diameter, not forming a volcanic-like cone (Figs 11A & B); frontal carina reaching = 0.5 distance to medial ocellus (e.g. Fig. 13D).............................. ................................................................................................................................ B. matthewi sp. nov. 3(1). Posterior margin of mesoscutellum extending to, or beyond anterior margin of T2, so propodeum not visible medio-dorsally (Figs 14A-D); mesoscutellum bearing three complete transverse rows of setae ...............................................................................................................................................................4 - Posterior margin of mesoscutellum not extending over propodeum to anterior margin of T2, so propodeum visible medio-dorsally (e.g. Figs 12A-D); mesoscutellum bearing either one or two complete transverse rows of setae........................................................................................................................ 5 4(3). Posterior eye margin contacting hyperoccipital carina (Figs 14A & B); lateral ocelli <0.5 medial ocellus diameter from eye margin; pilosity of mesoscutum and mesoscutellum of medium length (= 10 µm, <30 µm); eye pilosity of medium length (= 5 µm, <10 µm).......... B. tropaeumusbrevis sp. nov. - Posterior eye margin = 0.5 medial ocellus diameter from hyperoccipital carina (Fig. 14C); lateral ocelli = 0.5 medial ocellus diameter from eye margin; pilosity of mesoscutum and mesoscutellum long (= 30 µm); eye pilosity long (= 10 µm)........................................ B. tropaeumusdensus sp. nov. 5(3) Entire dorso-lateral propodeum scrobiculate, except for thin glabrous band along posterior margin (Figs 7E & 13A) ................................................................................................. B. scrobiculus sp. nov. - Dorso-lateral propodeum sculptured otherwise (e.g. Figs 12A-E).......................................................6 6(5). Posterior eye margin = 0.5 medial ocellus diameter from hyperoccipital carina (e.g. Fig. 8 B)..........7 - Posterior eye margin touching, or <0.5 medial ocellus diameter from hyperoccipital carina (e.g. Fig. 8C).........................................................................................................................................................9 7(6). Dorso-lateral propodeum bearing semi-circular ridge that begins near posterior margin of propodeum and extends through ventral region of propodeal spiracle before curving back to end near posterior margin of lateral propodeum (Figs 8A & B, arrowed);..................................................... B. dux Girault - Dorso-lateral propodeum not bearing a semi-circular ridge (e.g. Figs 8C & E);..................................8 8(7). Pilosity of T2 long, = 30 µm; antennal process and ventral frontal carina broadly projecting> 20 µm from ventral surface of firons (Fig. 12D); frontal carina extending <0.3 distance to medial ocellus .... ....................................................................................................................... B. prolatusspissus sp. nov. - Pilosity of T2 shorter, <30 µm in length; antennal process and ventral frontal carina less obvious, projecting<20 µm from ventral surface of frons (Fig. 12C); frontal carina extending> 0.3 distance to medial ocellus.......................................................................................................... B. ocellatus sp. nov. 9(6). Dorso-lateral propodeum bearing distinct crescent-shaped ridge running from ventral margin of propodeal spiracle to posterior margin of lateral propodeum (Figs 7D & 9E, arrowed) ........ B. leai Dodd - Dorso-lateral propodeum not bearing crescent-shaped ridge .............................................................10 10(9). Lateral ocelli, viewed dorso-laterally, touching eye margin or = 0.5 medial ocellus diameter from eye margin..................................................................................................................................................11 - Lateral ocelli, viewed dorso-laterally, much greater than 0.5, often nearly 1.0 x’s medial ocellus diameter from eye margin............................................................................................................................18 11(10). Posterior ocellar line> inter-ocular distance (Fig. 9F)........................................................................12 - Posterior ocellar line = inter-ocular distance ......................................................................................15 12(11). Propodeal spiracle situated close to the edge of the latero-dorsal margin of propodeum (Fig. 13C, arrowed) that is clearly defined by a broad laterally projecting carina that delineates the dorsal propodeum from the lateral propodeum (Figs 12C & D); T2 pilosity mostly long (= 30 µm) but can be of medium length (= 10 µm, <30 µm) in parts; T3 glabrous except a few setae present laterally............. ............................................................................................................................ B. spirolimbus sp. nov. - Propodeal spiracle not situated close to the edge of the latero-dorsal margin of the propodeum as described above either because latero-dorsal region of propodeum is rounded or, if dorsal propodeum is delineated from the lateral propodeum by a laterally projecting carina (e.g. Fig. 10C), then the propodeal spiracle is distant from the latero-dorsal margin of the propodeum (Fig. 10D); T2 pilosity never long, either medium or short (<30 µm); T3 with transverse coriarious band medially, bearing one row of setae..........................................................................................................................................13 13(12). Anterior and posterior genal margins convergent medially when viewed postero-laterally .................. ................................................................................................................................... B. maryae sp. nov. - Anterior and posterior genal margins mostly parallel or divergent medially when viewed postero-laterally...................................................................................................................................................14 14(13). Mesoscutum bearing 5 rows of setae in medial region; body generally not appearing pilose; density of pilosity of T2 being mostly sparse (= 20 µm between setae of same row), some areas of moderate density(<20 &> 10 µm) may be present; dorsal region of the propodeum delineated from the lateral propodeum by a longitudinal carinae........................................................................... B. jenningsi sp. nov. - Mesoscutem bearing 6 or more rows of setae in medial region; body generally quite pilose in appearance; pilosity of T2 dense (=10 µm between setae of same row) about the anterior region, but generally rest of T2 of moderate density (<20 &> 10 µm); latero-dorsal region of propodeum rounded, not delineated from the lateral propodeum by a longitudinal carinae (e.g. Fig. 13B).... B. murphyi sp. nov. 15(11). Eye height <0.5 head height (Fig. 12B)................................................................. B. mymyae sp. nov. - Eye height = 0.5 head height (Fig. 10A).............................................................................................16 16(15). Body not pilose in appearance; pilosity of vertex, mesoscutum and scutellum mostly sparse (= 20 µm between setae of same row), small areas of moderate density (<20 &> 10 µm) can be present; length of setae on vertex, mesoscutum and scutellum a mixture of short (<10) and medium length, (medium length pilosity within 10-15 µm range, not exceeding 20 µm); T2 pilosity of similar density but length mostly short, occasionally setae of medium length do occur, randomly dispersed amongst short setae .................................................................................................................................... B. moorei sp. nov. - Body pilose in appearance; pilosity of vertex, mescutum and scutellum mostly of moderate density (<20 &> 10 µm), with regions of dense pilosity (=10 µm) occurring; setae of medium length overall, including T2, commonly exceeding 20 µm in length, (range being 15-25 µm) ................................17 17(16). Lateral margins of mesosoma, when viewed dorsally, diverging posteriorly from one another (Fig. 9B); dorsal region of propodeum delineated from the lateral propodeum by a large longitudinal carina (Figs 7C, 9A & B)........................................................................................................ B. iqbali sp. nov. - Lateral margins of mesosoma, when viewed dorsally, largely parallel, or converging marginally posteriorly(Fig. 8F); latero-dorsal region of propodeum rounded, not delineated from lateral propodeum by a large longitudinal carina (Figs 7B, 8E & F)...................................................... B. hallarakeri sp. nov. 18(10). Anterior and posterior genal margins strongly convergent medially when viewed postero-laterally (Fig. 4A); anterior genal margin in contact with = 0.3 of ventral eye margin length (Fig. 4A).............. .................................................................................................................................... B. arthuri sp. nov. - Anterior and posterior genal margins parallel or divergent when viewed postero-laterally, (e.g. 14A); anterior genal margin in contact with> 0.4 of ventral eye margin length (e.g. 12B).........................19 19(18). Pilosity of vertex, mesoscutum and scutellum mostly dense (= 10 µm), small patches of moderate density (<20 &> 10 µm) can occur; pilosity of T2 dense in anterior region, gradating to moderate density posteriorly; T3 bearing two complete transverse rows of setae................ B. saliens (Hickman) - Pilosity of vertex, mesoscutum and scutellum mostly of moderate density (<20 &> 10 µm), small dense (=10 µm) patches can occur; pilosity of T2 mostly sparse (> 20 µm) apart from moderate density in anterior region; T3 glabrous except for several setae laterally.................... B. glenysae sp. nov.