First host record for the rogadine genus Pholichora van Achterberg (Hymenoptera: Braconidae) with description of a new species and notes on convergent wing venation features

The wasp genus Pholichora is recorded as parasitic on geometrid moth larvae on the basis of a new species, P. subscleroma Quicke and Shaw from South Africa (Natal), reared from Ascotis reciprocaria (Walker) (Lepidoptera: Geometridae, Ennominae). The discovery of this new species necessitates a reassessment of the characters separating Pholichora from Hemigyroneuron. An association between wing venation features (ovoid cells and scleromes) and nocturnality is discussed.


Introduction
The Rogadinae sensu stricto is a moderate-sized subfamily of braconid wasps dominated by the cosmopolitan genus Aleiodes, but it also includes about 35 other genera, some of which have been described only recently (e.g. van Achterberg 1991;Chen and He 1997). However, host records are only available for about half of these, despite the fact that since rogadines mummify the larvae of their lepidopteran hosts they often provide unusually accurate host data (Shaw and Huddleston 1991;Shaw 1994Shaw , 2003. Relationships between the genera of Rogadinae have hardly been discussed in the literature, but recent work on venom apparatus morphology (Zaldivar-Riveron et al. 2004) and current molecular work (M. Mori et al. in preparation) indicate some major groupings, and these appear to accord well with known host associations. Here we provide the first host record for a member of the Afrotropical genus Pholichora van Achterberg, and discuss its implications in the light of what is known about hosts of other rogadine genera and their putative relationships. A new species is described and distinguished from the previously described species of Pholichora, and implications for the diagnosis of the genus, and in particular its separation from Hemigyroneuron Baker, are discussed. Zaldivar-Riveron et al. (2004) provided a set of new characters, based on the detailed anatomy of the venom apparatus, to help resolve genus-level relationships within the Rogadinae. Several putative synapomorphies were found, mostly concerning the structure of the valve connecting the venom gland's secondary duct with the venom reservoir (see also Quicke et al. 1992Quicke et al. , 1997. These data suggest that within the Rogadinae s. s. there is a monophyletic group characterized by a hard, narrow and not-recessed secondary venom duct with a distinct exterior flange, which comprises Pholichora, Bulborogas van Achterberg, Canalirogas van Achterberg and Chen, Colastomion Baker, Conspinaria Schulz, Cystomastacoides van Achterberg, Myocron van Achterberg, Macrostomion Szépligeti, Rectivena van Achterberg, Triraphis Ruthe, Rogas Nees, Spinaria Brullé, Batotheca Enderlein, and Batothecoides Watanabe. This feature is absent in Hemigyroneuron, suggesting that the similarities in wing venation that it shares with Pholichora may be convergent. Hemigyroneuron is a principally Indo-Australian genus with a single known Afrotropical representative, H. certum van Achterberg, from Madagascar. Van Achterberg (1991) indicated that H. apicale Brues, from Natal, South Africa was probably a species of Pholichora, and transferred it to that genus, based on its very thorough original description (Brues 1926), which clearly indicates the presence of two yellowish scleromes in the subbasal cell of the fore wing. He suggested on this basis that it was probably closely related to P. bipanna van Achterberg, though it differs from it in a number of other features. H. apicale differs from the new species described here in having: (1) two scleromes in the subbasal cell, (2) the fore wing vein C+SC+R yellow-brown, (3) the head 36 wider than long, (4) the pterostigma entirely ''bright fulvous'', and (5) the propodeum without a carina or anterolateral angulations. It is also somewhat larger, 10 mm long compared with 7.8 mm in the new species. Hemigyroneuron differs from Pholichora by having: (1) no sclerome or dense patch of setae in the fore wing subbasal cell, (2) vein cu-a of fore wing long, neither curved nor swollen, thus making the first subdiscal cell higher, (3) the occipital carina fully developed and abruptly curved ventrally towards the hypostomal carina. The other characters given in van Achterberg's (1991) generic key (couplet 20) no longer work in the light of the discovery of the new species described below.

Description of new species
Pholichora subscleroma Quicke and Shaw, sp. nov.
Metasoma (Figures 7, 8). Densely clothed with short setae. First metasomal tergite 1.16 wider posteriorly than medially long; with longitudinal striation and complete strong midlongitudinal carina. Second metasomal tergite 1.366 wider posteriorly than medially long; largely finely longitudinally striate, with a distinct but fine mid-longitudinal carina but without a mid-basal triangular area. Third tergite largely finely striate becoming densely punctate posteriorly. Ovipositor sheaths distinctly up-curved and expanded distally.
Coloration. Largely yellowish. Antennae black except medial part of scapus which is dark brown. Stemmaticum and ovipositor sheath black. Telotarsi darkened apico-dorsally. Wings hyaline with apical 0.4 distinctly grey; venation yellow-brown except for fore wing vein C+SC+R, anterior half of pterostigma and parastigma which are nearly black.
Notes. The new species does not run easily in the key to Pholichora species by van Achterberg (1991). It can be distinguished from all other species in that key by the longitudinal hind wing vein 2-SC+R (transverse in inopina van Achterberg, bipanna van Achterberg and madagascarensis (Granger)).
Unfortunately, no host remains are associated with the three available specimens, so it is possible that the associated data are incorrect. However, the putative host, the giant looper, Ascotis reciprocaria (Walker) (Lepidoptera: Geometridae, Ennominae), is an important pest of a variety of commercial crops including citrus fruits, and therefore we feel that making this identification known may have some value for biocontrol. A. selenaria Denis and Schiffermü ller (type locality Austria) was treated as a senior synonym of A. reciprocaria (type locality Congo/South Africa) in Janse (1932), but the African taxon is now treated as a separate species, under the name A. reciprocaria, as in the recent checklist of Geometridae (Parsons et al. 1999), though there are apparently still some taxonomic issues to resolve (L. Pitkin, personal communication). When feeding on citrus crops, the host larva eats the outer layer of the fruit and causes a great deal of both cosmetic and actual damage.

Discussion
Pholichora appears to belong to the Rogas group of genera (namely Rogas, Rectivena, Colastomium, Macrostomium, Canalirogas, Myocron, Megarhogas, Cystomasticoides, Conspinaria, Trirhaphis, and Bulborogas) which, together with the putatively related Spinariina, can be recognized by their hardened insertion of the secondary venom duct on to the venom reservoir (Zaldivar-Riveron et al. 2004). Within this group, a subset of genera appears to be specific parasitoids of Zygaenoidea (Limacodidae and Zygaenidae), namely Rogas, Rectivena, Trirhaphis, Conspinaria, Spinaria, and Batotheca . The apparent use of Geometridae as hosts by Pholichora suggests that it may be more closely related to other members of the Rogas group which share the modified secondary venom duct with a flange, but attack non-zygaenoid hosts.
Hemigyroneuron lacks the modified secondary venom duct (Zaldivar-Riveron et al. 2004) and the possibility exists therefore that the similarity in wing venation (a distally expanded and at least partly glabrous fore wing subbasal cell) to that of Pholichora is homoplastic. Further, Hemigyroneuron is reported in the literature as parasitizing Lasiocampidae (Chen and He 1997), and so host associations do not provide any support for a relationship. The existence of another rogadine genus with very similar modified venation, Gyroneuron Kokujev, may also be an indication that the expansion of the subbasal cell is homoplastic within the subfamily. Gyroneuron differs from both Hemigyroneuron and Pholichora in many external features ( van Achterberg 1991). Unfortunately the venom apparatus of Gyroneuron has not been investigated and its host relationships are unknown. If indeed these represent three independent origins of very similar modifications of wing venation, this raises the question as to what purpose they serve and why they occur so frequently among the Rogadinae. Interestingly, similar modifications, involving either apical expansion of the subbasal cell bordered by curved and rather thickened veins, or an expanded, ovoid second subdiscal cell, both of which often also contain scleromes, are found in a number of other groups of Braconidae, including the Aphrastobracon Ashmead group of Braconinae (Quicke 1987;Quicke et al. 2000) and several species of Macrocentrinae ( van Achterberg 1993b).
In the nocturnal Xiphozelinae, there is a sclerome in the fore wing subbasal cell (van Achterberg 1979b). In the nocturnal Amicrocentrinae the fore wing is normal but the hind wing basal cell is distally strongly ovoid ( van Achterberg 1979a). Scleromes in other parts of the wing membrane are also predominantly found in nocturnal and crepuscular taxa, perhaps the best known being in the ophionine ichneumonid genera Enicospilus Stephens and Dicamptus Szépligeti (though they are known in at least a few species in many other ophionine genera-I. D. Gauld, personal communication). As indicated above, all of these are predominantly nocturnal or crepuscular taxa (Quicke 1992), and the question can thus be posed ''what is it about being nocturnal or crepuscular that also leads to the tendency to have ovoid wing cells formed of thickened veins and also, perhaps, scleromes?' ' Gauld and Huddleston (1976) emphasized the existence of an ''ophionoid facies'', a suite of characters that are typically found in nocturnal parasitic wasps (see also Quicke 1997), but did not mention wing features. One of the more obvious aspects of the ophionoid facies is the largely yellow-brown colour, and such pale cuticle is typically softer than dark, more heavily sclerotized cuticle. Thus, one very tentative explanation of the occurrence of oval cells, usually with markedly thickened veins, and scleromes, predominantly in nocturnal taxa, is that they may compensate for otherwise less rigid venation.