Revision of Durbadnus Pasteels, 1954, with Notes on Other Afrotropical Blennocampinae and Allantinae (Hymenoptera: Symphyta: Tenthredinidae)

ABSTRACT Three species are included in this revision of the Afrotropical sawfly genus Durbadnus Pasteels, 1954: D. chubbi (Forsius, 1930), D. obscuripes (Forsius, 1931) comb. n. and D. taegeri sp. n. A new junior synonym of D. chubbi is Blennocampa bensoni Forsius, 1931 (syn. n. ). The genus Durbadnus is restricted to South Africa, with D. taegeri being the first recorded member of the Tenthredinidae in the winter-rainfall zone of western South Africa. A key is provided to the five Afrotropical genera of Blennocampinae (Aethiocampa, Distega, Durbadnus, Tesslinia and Trisodontophyes). Kivua Forsius, 1934 is transferred from the Blennocampinae to the Allantinae. New junior synonyms of Kivua are Bicrista Pasteels, 1949 (syn. n.), Tumura Pasteels, 1949 (syn. n.) and Urocerca Pasteels, 1949 (syn. n.). The following new combinations result: Kivua bicolor comb. n. (Bicrista bicolor Pasteels, 1949), Kivua acuticornis comb. n. (Erythraspides acuticornis Pasteels, 1953), Kivua afra comb. n. (Erythraspides afra Pasteels, 1953), Kivua cribrifrons comb. n. (Urocerca cribrifrons Pasteels, 1949), Kivua pallipes comb. n. (Urocerca pallipes Pasteels, 1949) and Kivua tenuis comb. n. (Urocerca tenuis Pasteels, 1949). The specific epithet luteiventris is preoccupied in Kivua by K. luteiventris Pasteels, 1949, so that new names are proposed for two further species group taxa transferred here to Kivua: Kivua violae nom. n. for Tumura luteiventris Pasteels, 1949 and Kivua cara nom. n. for Urocerca luteiventris Pasteels, 1949. Neacidiophora quadrifoveata Koch, 1998 is a junior synonym of Kivua incrassata Pasteels, 1949 (syn. n.).


INTRODUCTION
The arid to semiarid winter-rainfall zone of Namibia and South Africa stretches West of a line from Swakopmund on the Atlantic coast of Namibia to Mossel Bay on the Western Cape coast of South Africa. The northern part of the zone is characterized by the Succulent Karoo Biome and the southern part by the Fynbos Biome with its richer floral diversity. Our very poor knowledge of the Symphyta of the winter-rainfall zone is explained by the restricted diversity of the sawfly fauna, the low abundance of these few taxa, and by the lack of field work so far specifically targeted on this group. Hitherto, 23 species of Argidae (Tenthredinoidea) have been the only indigenous Symphyta known from this area: nine species of Triarge Forsius, 1931(Koch 2006 and 14 species of Arge Schrank, 1802 (Pasteels 1953(Pasteels , 1955(Pasteels , 1963Koch & Goergen 2010). Additionally, three Tenthredinidae (Caliroa cerasi (L., 1758), Fenusa dohrnii (Tischbein, 1846) and Nematus oligospilus Förster, 1854), as well as the siricid Sirex noctilio Fabricius, 1793 are established, introduced species. During recent studies on the sawfly fauna of this area, one previously undescribed and probably endemic species of the family Tenthredini dae (Tenthredinoidea) was discovered. Initially, the new species described below was thought to belong to Distega Konow, 1904. Further research showed that it should be placed in Durbadnus Pasteels, 1954. A revision of Durbadnus became necessary to resolve the identity of the new species, which in turn revealed problems, here briefly outlined, affecting the characterisation of other Afrotropical genera previously that were placed in the Blennocampinae, as well as the distinction of Allantinae from Blennocampinae in the Afrotropical fauna. Here, a first attempt is made to review and key the Afrotropical genera of Blennocampinae.
The terminology of wing venation follows Goulet (1992). Except for Durbadnus spe cies, original descriptions are not cited in full. Detailed citations are to be found in Taeger et al. (2010).

Distinction between Blennocampinae and Allantinae
The classification of the Blennocampinae and Allantinae has been treated in many different ways, as reviewed by Lacourt (2003). In the European and North American faunas, Allantinae can be clearly identified by possession a fully developed (closed) and barely medially constricted fore wing anal cell, which in Blennocampinae is basally reduced (open, or petiolate), except in Lycaotini (Goulet 1992), where it is complete but strongly constricted. This distinction breaks down in the Afrotropical fauna, where the development of the fore wing anal cell varies considerably within some genera: particularly Kivua (Pasteels 1949b), but very rarely also in Distega, especially D. bevisi Forsius, 1930 (Fig. 2). The phylogeny and classification of the approximately 112 described genera of Allantinae and 106 genera of Blennocampinae in the world that are currently considered to be valid (Taeger et al. 2010), are in great need of intensive study. As already noted by Benson (1938), both subfamilies as presently constituted are probably polyphyletic. For the purposes of the present work, as a purely pragmatic and provisional solution to permit identification, Afrotropical Blennocampinae are distinguished from Allantinae as follows: -Anal cell of fore wing complete, or if veins 2A+3A are reduced, then combined length of flagellomeres 4-7 less than half as long as remainder of flagellum, flagellomere 4 usually much shorter than flagellomere 3 (maximally 0.6 as long) and ventral surface of all or any of flagellomeres 4-7 with a discrete patch (sometimes concave, or paler than rest of flagellomere) bearing setae shorter than those on remaining surface .  The modified underside of the apical flagellomeres and the general pattern of flagellomere length ratios in Afrotropical Allantinae, but not in any Afrotropical Blennocampinae, are characters that are usually considered to characterize the Waldheimini (e.g. Goulet 1992). The Waldheimini as currently defined are generally, however, re garded as belonging to the Blennocampinae. The phylogenetic significance of these cha racters is at present unclear. Abe and Smith (1991) and Taeger et al. (2010) listed nine valid genera of Blennocampinae as occurring in the Afrotropical Region: Aethiocampa Pasteels, 1949 Pasteels, 1949) and Urocerca Pasteels, 1949(type species: Urocerca luteiventris Pasteels, 1949. The type species of all these genera have been exa mined, and it was concluded that Bicrista (syn. n.), Tumura (syn. n.) and Urocerca (syn. n.) are junior synonyms of Kivua. These taxa had been distinguished from Kivua (s. str.) by single character states. Urocerca was characterized by the form of the cerci, as described by Pasteels (1949a), suggesting that the species concerned may differ in ovi position habits from other Kivua species. Bicrista and Tumura were erected mainly on the basis of extremes of development of the anal cell in the fore wing. This character is of little significance in the Afrotropical fauna because of its extreme variability. Kivua is provisionally removed to the Allantinae (new placement), because the fore wing anal veins 2A+3A are more or less distinctly indicated. In most other characters, Kivua is closely similar to Neacidiophora Enslin, 1911 (Allantinae), revised by Koch (1998). Eight Kivua species were recognized as valid by Taeger et al. (2010) and eight are added here through recombination, with two species removed to Durbadnus, so that the genus now contains the following 14 species:
Five Afrotropical genera are retained in the Blennocampinae. Aethiocampa is characterized by absence of the crossvein Rs+M and a very long, petiolate anal cell (1A) of the hind wing (Fig. 4). The basal lobe of the apically bifid tarsal claws is enlarged (Fig. 5). The head of the only known species in this genus, Aethiocampa aethiopica (Enslin, 1913), especially its frons, is conspicuously transversally microridged and dull. The me dial tooth of each mandible is strongly developed (Fig. 25).
Distega is the largest genus of Afrotropical Blennocampinae, with 25 species currently recognized as valid (Taeger et al. 2010). According to Pasteels (1949a), placement of Distega in the Blennocampinae is not necessarily correct, because he sporadically found specimens of different species with variable reduction of the fore wing veins 2A+3A. In fact, apart from a basal stub of 2A+3A (Fig. 1) and in very few cases a very small vestige of these veins at the base of the anal crossvein (Fig. 2;Koch 2000, fig. 34), particularly in D. bevisi, these veins are almost completely obliterated (Fig. 1) in all other species. A distinctive apomorphy of Distega is the continuous suture that divides the upper and lower halves of the mesepisternum (Fig. 24).
Durbadnus, now with three species, is revised below. Tesslinia is currently considered to be monotypic. The genus possesses a distinctive combination of characters: fore wing vein Rs+M is weakly curved near its junction with R (resembling Selandriinae, but none of these has a basally incomplete fore wing anal cell); the anal cell of the hind wing is very short or nearly impetiolate (Fig. 6); the basal lobe of the apically bifid tarsal claws is flattened (Fig. 7); the medial tooth of the rather slender mandible is very small (Fig. 26); and the head is micropunctate and shiny.
Durbadnus species have so far only been found in South Africa, whereas species of the morphologically similar genera Tesslinia Pasteels, 1951 andAethiocampa Pasteels, 1949 have an Equatorial distribution. The larvae and host plants of Durbadnus are unknown.

Redescription:
Female. Head black; labrum light brown, apical half of mandible reddish brown. Thorax black; pronotum, tegula, lateral lobes of mesoscutum except for a large black medial spot, lateral margin of medial lobe of mesoscutum, postspiracular sclerite, dorsal half of mesopleuron, and anepimeron orange. Legs blackish; tibiae and tarsi whitish, only at extreme apices of tibiae and tarsomeres 5 somewhat darker.Wings subhyaline; costa, subcosta and base of venation blackish, stigma and rest of venation light brown. Abdomen black. Head narrowed behind eyes. (Antenna as long as head and thorax together; vide Forsius 1931 for Blennocampa bensoni. Flagella of Monophadnus chubbi holotype missing; vide Forsius 1930.) POL:OOL = 1.0:0.8. Postocellar area: width:length = 1.0:0.7, late ral furrows slightly divergent posteriorly and convex; a longitudinal medial furrow shallowly developed, especially at the anterior margin. Supra-antennal grooves large and deep, no furrow developed towards torulus. Interantennal area with very small me dial groove. Eyes converge downwards, lower interocular distance as long as eye length. Sup raclypeal furrow deep. Head more or less micropunctate, shiny; pubescence on head pale, somewhat shorter than the diameter of lateral ocellus. Thorax sparsely micropunctate, shiny; pubescence pale, shorter than that on head. In fore wing, cell 2Rs as long as 1R1 and 1Rs united; stub of 2A+3A more or less straight at apex (Fig. 10). Tergum 1 with moderately wide but deep medial excision. Terga transversally microridged, dull. Sawsheath in lateral view pointed at apex (Fig. 11), in dorsal view very narrow towards apex. Lancet as in Fig. 12, with scattered, conspicuous longer setae at centre towards base; with approx. 27 serrulae; serrulae at centre very flat, truncate apically, with about 8-11 very small posterior denticles and 3 anterior denticles (Fig. 13). Length 6.7-7.3 mm.

Male.
Colouration similar to that of female. Mesoscutum completely black, only lateral margin of lateral lobe orange. Fore femur dirty white, mid and hind femora dirty white at apices. Head conspicuously narrowed behind eyes. Antenna 1.3 times as long as maximum width of head; flagellomere 1 0.9 times as long as flagellomeres 2/3 combined. POL:OOL = 1.0:0.6-0.8. Other features as for female. Penis valve as in Fig. 14; parapenis and harpe as in Fig. 15.  Distribution: South Africa (Fig. 34). Remarks: The type material of M. chubbi was examined and compared with that of B. bensoni. It was not possible to find any significant differences between these specimens, and therefore they are considered to be conspecific. Variability in D. chubbi occurs in the shape of the medial furrow of the postocellar area. This is distinctly developed in the female holotype of M. chubbi, weaker in the female holotype of B. bensoni, and absent in all males examined.
Durbadnus chubbi differs from other Durbadnus species in the large cell 2Rs of the fore wing, which is as long as cells 1R1 and 1Rs united (Fig. 10). In D. taegeri (Fig. 20) and D. obscuripes, 1Rs and 2Rs are about equal in their lower length; in D. chubbi, 1R1 is conspicuously shorter. Additionally, in the fore wing, the stub of 2A+3A of D. chubbi is more or less straight at the apex, whereas in D. taegeri and D. obscuripes it is apically furcate (Fig. 20).
Remarks. The colour pattern of D. obscuripes is very similar to that of D. taegeri. However, the latter species differs in having yellowish fore tibiae and because of the smaller black medial spot on the lateral lobes of mesoscutum. In D. taegeri, flagellomere 1 is as long as flagellomeres 2/3 combined, whereas in D. obscuripes, flagellomere 1 is conspicuously shorter. In particular, D. obscuripes is distinguished by the shape of serrulae, which are flatter than in D. taegeri. Furthermore, the lancet of D. obscuripes, especially at its base, bears very large sensilla. In the paratype, these sensilla are also present, but somewhat fewer in number. It is possible that they break off during oviposition.

Description:
Female. Head black; apical half of mandible reddish brown. Thorax black; pronotum, tegula, lateral half of lateral lobe of mesoscutum, medial lobe of mesoscutum at extreme angle, and postspiracular sclerite orange-yellow. Legs black; anterior surface of apical half and apex of posterior surface of fore femur yellow, fore tibia and fore basitarsomere at base dirty whitish. Wings uniformly slightly infuscate; stigma, costa, subcosta and rest of venation blackish. Abdomen black. Head very slightly narrowed behind eyes. Antenna 1.4 times as long as maximum head width; flagellomere 1 as long as flagellomeres 2/3 combined. POL:OOL = 1.0:0.7. Postocellar area: width:length = 1.0:0.7, lateral furrows slightly convex. Supra-an ten nal grooves large and deep, with conspicuous furrow at lateral margin of torulus. In terantennal area with rounded groove and small longitudinal furrow medially, ending at front margin of frontal area. Eyes converge downwards, lower interocular distance 1.1 times the length of the eye. Supraclypeal furrow deep. Head sparsely micropunctate, shiny; pubescence on head brown, longer than diameter of lateral ocellus. Thorax more densely punctate, shiny; pubescence yellow, shorter than that on head. In fore wing, cells 1Rs and 2Rs subequal in length; stub of 2A+3A furcate (Fig. 20). Tergum 1 with wide and deep medial excision. Terga 1/2 smooth and shiny, following terga transversally microridged and micropunctate, shiny. Sawsheath in lateral view narrowly rounded at apex (Fig. 21), in dorsal view narrowed toward apex. Lancet as in Fig. 22, having about 23 serrulae; serrulae at centre moderately prominent, with about 4 or 5 irregular posterior denticles and 3 or 4 irregular anterior denticles (Fig. 23 Distribution: South Africa (Fig. 34). Remark: Durbadnus taegeri is the first recorded member of the Tenthredinidae in the winter-rainfall zone of western South Africa.
Key to species of Durbadnus 1 Cell 2Rs in fore wing as long as cells 1R1 and 1Rs united (Fig. 10)