Phylogeny and Classification of the Orussidae
Total Page:16
File Type:pdf, Size:1020Kb
Blackwell Science, LtdOxford, UKZOJZoological Journal of the Linnean Society0024-4082The Lin- nean Society of London, 2003? 2003 139? 337418 Original Article L. VILHELMSENPHYLOGENY AND CLASSIFICATION OF ORUSSIDAE Zoological Journal of the Linnean Society, 2003, 139, 337–418. With 119 figures Phylogeny and classification of the Orussidae (Insecta: Hymenoptera), a basal parasitic wasp taxon LARS VILHELMSEN* Zoological Museum, University of Copenhagen, Universitetsparken 15, DK-2100 Copenhagen Ø, Denmark Received November 2002; accepted for publication March 2003 The Orussidae is a small family of parasitic wasps, comprising 75 species worldwide. It occupies a key position within the Hymenoptera, being the sistergroup of the Apocrita, a taxon containing all other parasitic wasps. In total, 163 morphological characters were scored for 74 species of Orussidae and five outgroup taxa. The dataset was analysed under different weighting schemes. The results do not support a single phylogenetic hypothesis, but most relation- ships were retrieved in the majority of the cladistic analyses. Earlier attempts at tribal and subfamily classifications of the Orussidae are not corroborated. Enforcing a strictly cladistic classification at these levels would require the recognition of many redundant taxa without enhancing the information content. It is proposed that formal recog- nition of tribes and subfamilies within the Orussidae be abandoned. The generic concepts of the family are revised. Sixteen genera are recognized; detailed descriptions and illustrations of each are provided, as is a key to the genera. The monophyly of most genera as defined here is well supported, with the exception of Guiglia Benson, 1938. Guiglia is retained because alternatives to monophyly of this genus are not well supported either. Heliorussus Benson, 1955, syn. nov., is incorporated in Orussus Latreille, 1796; the species formerly included in Heliorussus, Orussus schout- edeni Guiglia, 1937, combination reestablished, O. scutator (Benson, 1955) comb. nov. and O. spinifer (Benson, 1955), comb. nov. are transferred to Orussus. Leptorussus kwazuluensis sp. nov. is described; Pseudoryssus emanuelis Guiglia, 1956 syn. nov. is considered a junior synonym of Pseudoryssus henschii (Mocsáry, 1910). An annotated key to the world species of Orussus is included as an appendix. © 2003 The Linnean Society of London, Zoological Journal of the Linnean Society, 2003, 139, 337-418. ADDITIONAL KEYWORDS: cladistic analysis – generic revision – keys – Orussus. INTRODUCTION order ‘Symphyta’), the Orussidae comes out strongly supported as the sistergroup of the Apocrita (Ronquist The Orussidae is a small family of parasitic wasps, et al., 1999; Vilhelmsen, 2001a). This implies that par- unique among the Hymenoptera in being the only non- asitism arose in the common ancestor of Orussidae apocritan parasitoid taxon. They possess a combina- and Apocrita. Therefore, understanding the lifestyle of tion of advanced and plesiomorphic traits (Vilhelmsen the Orussidae is crucial in elucidating the evolution of et al., 2001), the absence of the ‘wasp-waist’ of the Apo- parasitism within Hymenoptera. crita being the most striking example of a retained Biological information for any species of Orussidae plesiomorphy. The Apocrita with more than 100 000 is scarce, most having been obtained from a few spe- described species, the majority of which are parasi- cies of Orussus. Latreille (1811) provided the first toids, comprise the bulk of the Hymenoptera. Indeed, detailed account of the morphology and lifestyle of the main reason for the present day diversity of the any orussid (Orussus abietinus); he noted the close Apocrita is probably their successful exploitation of association with wood (p. 559: ‘…leurs larves vivent the parasitoid lifestyle. In the most recent cladistic certainement dans l’intérieur du bois.’). The interpre- treatments of the phylogeny of the basal tation of the evidence accumulated since then has hymenopteran lineages (previously known as the sub- been contentious. Everyone agrees that orussid larvae are associated with galleries in dead wood made by other woodliving insect larvae. Most workers favour *E-mail: [email protected] an ectoparasitic lifestyle for Orussidae, usually with © 2003 The Linnean Society of London, Zoological Journal of the Linnean Society, 2003, 139, 337–418 337 338 L. VILHELMSEN Buprestidae (Coleoptera) as the (presumed) host (e.g. apomorphic or plesiomorphic features were empha- Burke, 1917; Rawlings, 1957; Powell & Turner, 1975; sized in the classification. Only with the application of Nutall, 1980); in contrast, Cooper (1953) suggested quantitative cladistics did the systematic position of that Orussus is feeding mainly on the frass in the gal- the Orussidae become firmly established. Currently, leries of wood-boring insects. However, the available they are placed in a separate superfamily (Orus- evidence considered as a whole supports the parasi- soidea), the most inclusive formal category below toid hypothesis (see Vilhelmsen et al., 2001 for further order level currently employed by hymenopterists discussion). Remarkable adaptations for ovipositing in (Vilhelmsen, 2001a). wood includes an echolocation system involving the In contrast to the present well corroborated phylo- antennae and the forelegs in the female and a very genetic position of the Orussidae within the long (at least twice the body length) concealed ovipos- Hymenoptera, the internal phylogeny of the family is itor (Vilhelmsen et al., 2001). entirely uncharted. The most recent comprehensive Larval adaptations for the parasitoid lifestyle in a effort in orussid systematics on a global level is confined habitat include the reduction of the sensory Guiglia’s (1965) catalogue of the world species. It is a apparatus (eyes and antenna), mouthparts (maxillae useful embarkation point for further exploration, but and labium), and locomotory apparatus (thoracic legs it was of course not done within a cladistic context, and suranal process (Vilhelmsen, 2003; see Rohwer & and no attempt was made to accumulate material Cushman, 1917 and Parker, 1935 for further informa- for an exhaustive study. Regional surveys of most tion about anatomy of the immature stages of the local faunas exist: Africa (Benson, 1935a; Guiglia, Orussidae). These features are all shared with apocri- 1937a: Chalinus; Guiglia, 1937b: Orussus), Australia tan larvae. An apparently unique feature is the pres- (Benson, 1938b; Riek, 1955), Europe (Guiglia, 1954a; ence of dorsal transverse rows of small cuticular Kraus, 1998), Far East/South-eastern Asia backward curving spines on all thoracic and abdomi- (Yasumatsu, 1954), North America (Ross, 1937; Mid- nal segments throughout larval development. These dlekauff, 1983), Neotropical (Smith, 1988); these are spines have also been reported in late instar larvae of all limited geographically. Schlettererius cinctipes (Cresson) (Stephanidae, see The main obstacle to further progress in orussid sys- Taylor, 1967). In Orussidae, they presumably facilitate tematics is the scarcity of material. Orussidae are moving in galleries inside the wood in order to seek uncommon and rarely collected. The most fortuitous out hosts that the ovipositing female was unable to record is that of the holotype of Orussella dentifrons reach (Vilhelmsen, 2003). Another feature not (Philippi, 1873), which was plucked from a cobweb. The reported for any other Hymenoptera is the presence of somewhat incomplete condition of this specimen led cuticular folds in the hindgut which perhaps allow the Konow (1897a) to conclude that it ‘entschieden besser orussid larva to clamp it shut during development; in den Papierkorb gewandert wäre’ [‘It would have been anatomically, the mid- and hindgut are continuous better to have thrown it in the paper bin.’]. Abuse not- throughout most of larval development, contrary to withstanding, the species is currently regarded as valid the condition in virtually all Apocrita examined (Vil- and placed in its own genus, attesting to the distinc- helmsen, 2003). Shutting the hindgut and the post- tiveness of the taxon. Only a few major collections hold ponement of defecation until just prior to pupation material of more than half a dozen species of Orussidae. prevent fouling and subsequent putrefaction of the Furthermore, many species are represented by only a host; this is thought to be an important adaptation to handful of specimens, spread in collections all over the the parasitoid lifestyle in Hymenoptera, especially world. Significantly, the only previous revisions of the endoparasitism (Gauld & Hanson, 1995). higher-level classification were undertaken by R. B. The unique status of the Orussidae was recognized Benson (1935a, 1955a) who worked in the NHML, at an early stage in higher-level hymenopteran sys- which holds the largest and most diverse collection of tematics. In precladistic days, this was translated by Orussidae in the world by far. some authors into assigning them an ‘intermediate’ The present revision is the most comprehensive position between ‘Symphyta’ and Apocrita. Rohwer & effort in orussid systematics to date: 75 species in 16 Cushman (1917) erected the suborder Idiogastra genera are recognized. Material from about 50 collec- solely to accommodate the Orussidae, and Bischoff tions from all over the world has been examined, (1926) argued that they should be placed in a separate including 57 primary types. This represents all but superfamily among the ‘Symphyta’.