Orwellium, a new Valdivian genus of Platygastridae (Hymenoptera)

Orwellium enigmaticum is described as a new genus and species. Th is species is endemic to the Valdivian forests of Chile. Th e genus is hypothesized to be the sister-group of the subfamily Platygastrinae sensu lato (i.e., including both Platygastrinae and Sceliotrachelinae).


Introduction
Th e parasitoid wasp superfamily Platygastroidea traditionally has been divided into two families, the Platygastridae and the Scelionidae (e.g., Masner 1993Masner , 1995Austin et al. 2005). Both groups are found worldwide, with the exception of Antarctica. All scelionids, as far as is known, are idiobiont egg parasitoids, attacking hosts in a range of insect orders as well as spiders. A small number of genera of Platygastridae are also egg parasitoids, a few are known to attack nymphal Sternorrhyncha, but the vast majority of host records for species in the family are koinobiont egg-larval or larval parasitoids of gall fl ies (Diptera: Cecidomyiidae) (Austin et al. 2005).
Th e formal classifi cation implied that these families represent two distinct lineages. However, a recent analysis of relationships within the superfamily (Murphy et al. 2007) based on sequence data from three genes -18S rRNA, 28S rRNA, and cytochrome oxidase I -came to the tentative conclusion that the platygastrids were nested as a monophyletic group within Scelionidae, thus rendering the latter paraphyletic. Th e details of the intergeneric relationships varied depending upon the analysis used, and low levels of support were found for many of the deeper nodes. Additionally, the analysis included representatives of only 58 genera, less than one fourth of the number of genera currently recognized to be valid within the superfamily. On the basis of these considerations, no changes were proposed in the classifi cation. Sharkey (2007), however, apparently considered this evidence and analysis to be suffi cient for he formally placed the family Scelionidae Haliday, 1839 as a junior synonym of Platygastridae Haliday, 1833. Sharkey (2007 made no mention of the status of the fi ve subfamilies within the two former families. Th us, in clearing up the classifi cation by eliminating one putatively paraphyletic taxon, the internal classifi cation of Platygastridae sensu lato was rendered even more uncertain. Presumably, the monophyletic Platygastridae sensu stricto should be treated as a single subfamily. In lieu of concrete and well-supported hypotheses of relationship, for the purposes of this paper we will continue to treat the three scelionid subfamilies -Scelioninae, Teleasinae, and Telenominae -as valid taxonomic units. Th us, the family Platygastridae will be considered to contain four subfamilies. Th e Platygastrinae have long been recognized as a discrete unit. Th is is evident not only in sequence data, but by a number of morphological characters. Th e number of antennomeres is reduced to a maximum of 10; the sex-segment in the male antenna is found on A4 or rarely on A3; cerci are absent; the second metasomal segment is signifi cantly larger than all others; the netrion is lacking and the second fl exor of the fore wing arises on the mesopleuron (Mikó et al. 2007); and the fore wing tracheate veins are strongly reduced to, at a maximum, only a single truncated R. Th ese are largely reductional characters that are also observed in genera of other subfamilies. Th e combination of characters, however, and the predominant use of cecidomyiids as hosts, has generally been adequate to defi ne the group for the past 170 years.
In the analysis of Murphy et al. (2007) the sister group of Platygastrinae is not conclusively identifi ed. Th e genera Archaeoteleia Masner and Neuroscelio Dodd are basal to all other platygastroids in both the maximum parsimony and Bayesian analyses. In the latter, the next branching point is an unresolved polytomy comprised of the branch leading to the platygastrines, the genera Sparasion Latreille and Sceliomorpha Ashmead, and the monophyletic group composed of all of the remaining genera. In the maximum parsimony analysis, Sparasion groups together with Archaeoteleia and Neuroscelio. None of these potential sister taxa is intuitively obvious. In general, individuals in these genera are large in body size, coarsely sculptured, and characterized by having higher values in meristic characters (such as the number of sex segments in the male antenna). Platygastrines, in contrast, are uniformly small to minute, fi nely sculptured, and char-acterized by reductions in most meristic characters. Th us, the subfamily appears to stand apart as a morphological and biological outlier.
Th e platygastroid fauna of southern Chile has both a number of unique elements as well as others that are consistent with transantarctic distributional patterns. Th e genera Romilius Walker, Almargella Masner & Huggert, Magellanium Masner & Huggert, and Nanomerus Masner & Huggert are unique to Chile (Masner and Huggert 1989;unpublished data); Proplatygaster Kieff er is found in Chile and Australia (Masner and Huggert 1989); Archaeoteleia is known only from Chile and New Zealand (Early et al. 2007). Whether such patterns represent vicariant events, widespread extinction, long-range dispersal, or simply inadequate sampling is unknown. However, the oldest known fossils are in Lebanese amber dated 112 -122 mya (Johnson et al. 2008c); thus, the age of the superfamily is at least consistent with the possibility of vicariance.
Among recent collections from Valdivian forest in Chile, one of us (L. Masner) discovered numerous specimens of an enigmatic platygastroid, puzzling in that it seemed to fi t well in neither the Scelionidae nor the Platygastridae. Th e purpose of this paper is to describe this species, placing it in a new genus and species, and to discuss its possible position within the superfamily Platygastroidea.

Material and methods
Th is work is based upon specimens in the following collections, with abbreviations used in the text: CNCI 1 , Canadian National Collection of Insects, Ottawa, Canada; OSUC 2 , C.A. Triplehorn Insect Collection, Columbus, OH.
In the Material Examined the numbers prefi xed with "OSUC" are unique identifi ers for the individual specimens. Since the label data for all specimens have been georeferenced and recorded in the Hymenoptera On-Line database, details on the data associated with these specimens can be accessed at the following link: http://purl.oclc. org/NET/hymenoptera/hol, by entering the identifi er in the form. Note the space between the acronym and the number. Data for the genus can be accessed at http://purl. oclc.org/NET/hymenoptera/hol?id=238175. Th e description was generated using a database application, vSysLab, designed to facilitate the production of a taxon by character data matrix, and to integrate this with the existing taxonomic and specimen-level database. Th e text output for descriptions is in the format of "Character: Character state(s)." Images and measurements were made using AutoMontage and Cartograph extended-focus software, using JVC KY-F75U digital cameras, Leica Z16 APOA microscopes, and 1X objective lens. Images are archived at Specimage, the image database at the C.A. Triplehorn Insect Collection, and Morphbank. New taxa have been prospectively registered with Zoobank (Polaszek et al. 2005). Biodiversity informatics standards are implemented following Pyle et al. (2008) and Johnson et al. (2008a). External hyperlinks from the text are explicitly cited in the footnotes so that users of the printed version of this paper have access to the same resources. Life sciences identifi ers, LSIDs, may be resolved at the specifi ed URLs or at http://lsid.tdwg.org.   Diagnosis. Th e large second metasomal segment, by far the largest of all segments (Fig. 5), is also found in members of the subfamilies Platygastrinae and Telenominae. Orwellium is distinguished from telenomines by the presence of narrow laterotergites and laterosternites on the metasoma and by the densely setose propodeum and metapleuron. Orwellium is distinguished from platygastrines by the presence of well-devel-oped, tracheate marginal and stigmal veins in the fore wing, the presence of a malar sulcus, and the presence of cerci.

Orwellium enigmaticum
Etymology. Th e name Orwellium honors the author George Orwell, the brilliant and seemingly clairvoyant writer of political science fi ction, and is to be considered to be neuter in gender. Th e specifi c epithet enigmaticum refers to the initial doubts concerning the proper familial placement for this genus.
Discussion. Two outstanding features of Orwellium are the number of antennomeres and the dense pilosity on the metapleuron and propodeum. Although we record here that the female antenna is comprised of ten antennomeres, A3 is extremely long and there is a noticeable constriction in its midlength. Th is seems to indicate the fusion between a former A3 and A4. Th e densely setose metapleuron and propodeum is relatively rarely seen in Platygastroidea, and then principally within the Platygastrinae. In many specimens of Orwellium (such as in Fig. 2), there is an appreciable amount of extraneous material embedded within the setae. Th is may simply be a trap for debris, or it may suggest the presence of an exocrine gland.
Th e 1-2-2 tibial spur formula is uncommonly found in the superfamily outside of the Platygastrinae. It is limited to the genera Archaeoteleia, Neuroscelio, and the tribe Sparasionini (sensu Johnson et al. 2008b). Th ree sparasionine genera were not included in the analysis of Murphy et al. (2007), but the remainder are precisely those taxa that cluster near Platygastrinae at the base of the cladogram of the superfamily. Orwellium has a number of features that are similar to platygastrines. Th e notauli are expanded posteriorly and converge near the transscutal articulation. Orwellium lacks a netrion on the pronotum, the site of origin of the second fl exor of the fore wing (Mikó et al. 2007). Th e mesopleural pit (Fig. 2, mp) is located in a relatively low and posterior position on the mesepisternum, and connected by a groove to the apices of the mesopleural carina and sternaulus anteriorly, and to another pit (unnamed in Mikó et al. 2007) located ventrally near the posterior margin of the mesopleuron. Similar structures are found in a number of platygastrines (see illustrations in Masner and Huggert 1989), but are not yet known in the other subfamilies. Th ese similarities to platygastrines must be reconciled with another set of characters. Th e pronotum lacks the epomial and transverse pronotal carinae. Th e sex segment in the male is on A5, although the fact that the male has one more segment than platygastrines might suggest that the A5 in Orwellium could be homologous with the A4 in platygastrines in which the maximum number of antennomeres is 10. Th e cerci are well developed on the apical metasomatic tergite, but in platygastrines cerci are completely absent. Th e presence of the marginal, stigmal and (short) postmarginal vein (Fig. 6) is completely unknown in any platygastrine. We interpret this evidence to be consistent with the hypothesis that Orwellium is the sister group of the Platygastrinae.
We are currently actively working on the comprehensive morphological and molecular analysis of relationships within the superfamily that is needed to address this question.