The first fossil leptofoenine wasp (Hymenoptera, Pteromalidae): A new species of Leptofoenus in Miocene amber from the Dominican Republic

Th e fi rst fossil of the pteromalid subfamily Leptofoeninae Handlirsch is documented. Leptofoenus pittfi eldae sp. n. is described and fi gured from a single male preserved in Early Miocene (Burdigalian) amber from the Dominican Republic. Th e fossil is compared with its modern congeners and comments generally provided on the taxonomy of the subfamily, including a key to the two genera.


Introduction
Th e infrequently encountered wasps of the subfamily Leptofoeninae are remarkable giants among the otherwise more diminutive members of the Pteromalidae, a family that is likely paraphyletic, if not polyphyletic (e.g., Desjardins et al., 2007).With species ranging in size from 11-27 mm in females and 5-12 mm in males and their elongate, almost tubular body forms with elongate legs, leptofoenines closely resemble parasi-toids of the Pelecinidae, Gasteruptiidae, and particularly the Stephanidae (Figs. 1-2), similarities refl ected in the epithets for various species.When Leptofoenus peleciniformis Smith was fi rst described, Smith (1862) left open the familial assignment, refl ecting this challenge in the name of the genus and species, the generic name implicating a placement near gasteruptiids and stephanids and the specifi c epithet indicating peleci- nids.Westwood (1868) subsequently described a separate genus and species, Pelecinella phantasma Westwood, for what would later be recognized as a synonym of Smith's species, but he recognized the chalcidoid affi nities of the group.Nearly 30 years later, Ashmead (1895) identifi ed Westwood's genus as belonging to the Cleonymidae (today a subfamily of Pteromalidae) and established the subfamily Pelecinellinae therein for this group.Handlirsch (1924), following Ashmead (1895), recognized Pelecinellinae within the Cleonymidae, but placed Leptofoenus as a distinct family near the Stephanidae.Various authors treated Smith's and Westwood's genera as separate until 1924 when Brues recognized them to be synonyms and brought them together into a single group, employing Handlirsch's family-group name, Leptofoeninae, for the subfamily within Cleonymidae, a usage apparently followed by all subsequent authors (e.g., Bouček 1958Bouček , 1988;;Hedqvist 1961), albeit sometimes at tribal rank.
Despite their conspicuous size nothing is known of leptofoenine biology aside from anecdotal notes that they have been collected from dead or fallen trees.Th eir structure is strikingly similar to wood-inhabiting stephanids, with elongate, tubular body; head with numerous tubercles on frons and vertex; long legs; mesopleural impression for reception of legs; elongate ovipositor; etc. Th e body form and structures are presumably adaptive for living in wood and being parasitoid on such boring insects as beetles and wood wasps, so it seems reasonable to suspect that leptofoenines have a similar biology.Two genera are presently recognized in the subfamily -Leptofoenus Smith, with fi ve modern species in the New World ranging from northern Argentina to the southwestern United States, and Doddifoenus Bouček, with two species in Papua New Guinea and northeastern Australia (Table 1).
While fossils of the Pteromalidae have been recovered from various localities (e.g., Darling 1996;Gibson 2003), none have been hitherto found of the subfamily Leptofoeninae.Herein I described the fi rst fossil leptofoenine from a male recently discovered in Early Miocene (Burdigalian) amber from the Dominican Republic (Fig. 3).Th is is also the fi rst leptofoenine recorded from the West Indies proper.Morphological terminology and format for the description generally follows that of LaSalle and Stage (1985) and Bouček (1988).Photomicrographs were prepared using a Nikon D1x digital camera attached to an Infi nity ® K-2 long-distance microscopic lens.
Diagnosis.Large, elongate species (5-27 mm in length); head subcubic, with parascrobal crests; inner orbits of compound eyes parallel or slightly divergent toward lower tangents; antenna 13-segmented (except see below), without true anellus, toruli situated above lower tangent of compound eyes; malar space short; mandible blunt, with large truncate apex, sometimes with a small ventral truncation or tooth; mesosoma elongate, with pronotum dorsally twice as long as wide; legs slender and elongate; petiole elongate; ovipositor elongate.
Comments.Th e family-group name based on Pelecinella predates that based on Leptofoenus and accordingly holds priority for usage.Although Pelecinella is a subjective junior synonym of Leptofoenus (Brues, 1924;LaSalle and Stage, 1985), this does not aff ect the availability of the family-group name based upon it (ICZN, 1999: Art.40.1).However, reverting to Pelecinellinae for the subfamily is not in the general interest of nomenclatural stability given that a family-group name based on Leptofoenus has been in universal usage since at least 1958 (e.g., Bouček 1958).Accordingly, the name Leptofoeninae is retained as the name for this subfamily in accordance with the spirit, if not the precise letter, of Art. 40.2 (ICZN, 1999).LaSalle and Stage (1985) provided a revision of the species of Leptofoenus [including at that time Doddifoenus australiensis (Dodd)], while Bouček (1988)  Diagnosis.Th e new species can be readily recognized from its modern and mainland congeners by the combination of a thinly sclerotized band posterior to M+Cu in the forewing (Fig. 5), the presence of a sclerotized spot in the forewing at the junction of M+Cu and the basal vein (Fig. 5), M+Cu being comprised of a series of 14 setae, the basal cell containing 22 setae, the anterior margin of the costal cell sclerotized with a dense fringe of short fuscous setae, the hind wing with only three hamuli, the apicalmost antennal article apparently composite and resulting in an overall 11-segmented antenna, and the absence of a rasp-like structure on the inner surface of the metatibia.
Description.Male: Total length 8.8 mm; forewing length 4.7 mm; head 1.0 mm in length; mesosoma 3.3 mm in length, petiole 1.2 mm in length; gaster 3.3 mm in length.Integument dark brown throughout (Fig. 3).Head with scrobal cavity carinate laterally; maxillary palpus 4-segmented, apical palpomere elongate, with numerous elongate black setae dorsally; clypeus bare except for two erect setae laterally near oral fossa (visible in left, oblique view); antenna apparently 11-segmented, apical fl agellar article elongate and apparently with lines indicating possible fusion of two very minute apical articles (resulting in an otherwise 13-segmented antenna); parascrobal crests and vertex strongly tranversely striate, forming rasp-like structure dorsally.Mesosoma with pronotum and mesoscutum strongly, transversely strigate (Fig. 4), strigae increasingly bearing spicules posteriorly on mesoscutum; pronotum elongate, anterior margin rounded, posterolateral surface with striolate area (Fig. 4), striolate area ovoid (distinctly longer than wide; more rounded in L. rufus LaSalle and Stage), striations in striolate area fi ner and more evenly spaced anteriorly; mesoscutellum strongly transversely strigate, with numerous spicules; mesopleuron coarsely and irregularly punctate except in impressed region largely glabrous; propodeum laterally densely and contiguously punctate, with sparsely scattered setae.Procoxa with strong carina along dorsolateral margin; metacoxa sculptured as on lateral surface of propodeum; metafemur longitudinally striate ventrally except reticulate basally; metatibia without longitudinal rasp-like structure on inner surface; metabasitarsus twice as long as second tarsomere (tarsi pentamerous).Forewing with anterior margin of costal cell sclerotized (Fig. 5), with dense fringe of short fuscous setae; basal cell with 22 setae; sclerotized spot at junction of M+Cu and basal vein (Fig. 5); M+Cu composed of series of 14 setae and posteriorly by thinly sclerotized band leading to sclerotized spot (Fig. 5); hind wing with three hamuli.Metasoma with petiole approximately 7 times longer than wide (direct dorsal view not possible so this is estimated); petiole weakly transversely strigate dorsally; gaster imbricate.Etymology.Th e specifi c epithet is a matronym honoring Ms. Morgan Pittfi eld, enthusiast of all things adventurous when it comes to amber and niece of Keith Luzzi, who generously donated the holotype and permitted its study.

Discussion
Monophyly of Leptofoenus relative to Doddifoenus is supported by the presence of a striolate area on the posterior lateral surface of the pronotum (Figs.1-2), a structure of uncertain function but resembling a stridulatory apparatus.Doddifoenus is supported by the apomorphic presence of a mobile, elongate, spike-like ninth metasomal tergum (= eighth gastral tergum).Placement of the fossil species within Leptofoenus is readily supported by the presence of this striolate region (Fig. 4), along with a long basitarsus relative to the second tarsomere.Relationships within Leptofoenus were tabulated by LaSalle and Stage (1985), who considered L. peleciniformis as basal owing to the absence of a sclerotized spot at the M+Cu-basal vein junction in the forewing (within a more restricted Leptofoenus with L. australiensis removed, sensu Bouček, 1988) (Fig. 6).Th e species L. howardi (Ashmead), L. westwoodi (Ashmead), and L. stephanoides (Roman) were considered a clade owing to the presence of a rasp-like structure on the inner margin of the metatibia.Leptofoenus rufus was placed in an intermediate position (Fig. 6), plesiomorphically lacking the rasp-like structure, but sharing with the aforementioned clade the sclerotized spot in the forewing, as well as a medioapical incision on the sixth metasomal tergum (= seventh abdominal tergum, fi fth gastral tergum).In regard to these characters, L. pittfi eldae would be placed in a position similar to L. rufus owing to the presence of a sclerotized spot in the forewing, but lacking the rasp-like structure on the inner surface of the metatibia (Fig. 6).Th e sclerotized anterior margin to the costal cell would appear to be another feature of phylogenetic value in the genus.Leptofoenus peleciniformis lacks a sclerotized anterior margin, as in Doddifoenus, while the anterior margin is sclerotized and bears a fringe of setae in L. rufus, L. pittfi eldae, and the clade of L. howardi, L. stephanoides, and L. westwoodi.
Th e new species will run to couplet 4 in the key provided by LaSalle and Stage (1985) where it then intermingles the characters.Th e fossil species basically agrees with the second half of couplet 4 in all attributes except that, like L. howardi, the vein beneath the basal cell (M+Cu) is represented by a series of setae (8-18 in L. howardi, 14 in L. pittfi eldae).Uniquely for the genus, in addition to numerous setae demarcating M+Cu, this vein is very thinly sclerotized posteriorly along its length leading up to the sclerotized spot at the junction of the basal vein and M+Cu (Fig. 5).Unlike the couplet for L. howardi, L. pittfi eldae has three hamuli on the hind wing and the basal cell has less than 25 setae, albeit not as few as in L. westwoodi and L. stephanoides which have less than 20 each (L.pittfi eldae has 22).Th e thinly sclerotized band along the posterior of M+Cu and the structure of the apicalmost antennal article are unique and also can serve to immediately distinguish L. pittfi eldae from its congeners.
Given that this is the only such specimen known despite the volumes of Dominican amber inclusions that have been scrutinized in the last few decades, it would appear that Leptofoenus occurred in as low abundance in the Miocene as they do today.Leptofoenus pittfi eldae is the only leptofoenine documented from the West Indies.While such a geographic extinction would not be surprising given the number of localized extinctions documented from the Dominican amber fauna, it may not be a safe assumption that Leptofoenus is truly absent from the region.Given the rarity with which leptofoenines are represented in collections and the obviously large gaps in the distributions of the modern species (e.g., L. stephanoides ranging from northern Argentina to southern Mexico, yet not yet documented from several intervening countries such as Guatemala, Nicaragua, Honduras, or El Salvador), the possibility that the subfamily is today found somewhere across the various Caribbean islands cannot be immediately dismissed.Nonetheless, for the moment L. pittfi eldae would appear to represent yet another group of insects which were established within the Miocene fauna of Hispaniola that subsequently became extinct.

Figures 1- 2 .
Figures 1-2.Photomicrographs of representative modern Leptofoenus species and lateral aspects of their pronota.1 Leptofoenus rufus LaSalle and Stage, female. 2 Leptofoenus stephanoides (Roman), male.Specimens from the collection of the Division of Entomology, University of Kansas Natural History Museum.
to David Grimaldi and an anonymous reviewer for insightful commentary on the manuscript.Partial support for this research was provided by U.S. National Science Foundation grants EF-0341724 and DEB-0542909.Th is is a contribution of the Division of Entomology, University of Kansas Natural History Museum.

Table 1 .
Diversity and distribution of leptofoenine wasp species.

Doddifoenus Bouček Genus Leptofoenus Smith Leptofoenus pittfi eldae Engel, sp. n.
Holotype.Male, KU-NHM-Ent DR-019; Early Miocene amber, Dominican Republic, La Toca group of mines northeast of Santiago, excavated in late Autumn 2008, Keith Luzzi coll.; deposited in the Insect Fossil Collection, Division of Entomology, University of Kansas Natural History Museum.