of Neurepyris Kieffer , 1905 ( Hymenoptera , Bethylidae ) , a new synonym of Pristocera Klug , 1808

The genus Neurepyris Kieffer, 1905 is revised. The adult male holotypes N. rufiventer Kieffer, 1913 from Eritrea and N. tagala (Ashmead, 1905) from the Philippines are redescribed and illustrated. Both species are transferred from the subfamily Epyrinae to Pristocerinae because they have the metanotum well developed medially. Neurepyris rufiventer is transferred to Pristocera Klug, 1808 because the hypopygium is deeply divided into two apical lobes, the subdiscoidal and cubital veins do not reach the border of the forewing, the pronotal disc has the anterior region slightly elevate medially, and the stigma is elongate. Neurepyris tagala is transferred to Apenesia Westwood, 1874 because the basal tooth of mandible is not curved inward, the median lobe of clypeus is not depressed near the antennal insertions, and the aedeagus consists of one lamina. The genus Neurepyris is considered a junior synonym of Pristocera because its type species is N. rufiventer.


Introduction
created Neurepyris Kieffer, 1905 to accommodate a species of Epyrini, which has a longitudinal carina on the propodeal disc absent, unidentate tarsal claw and lanceolate stigma.Kieffer & Marshall's key (1904-1906) indicated that there was one species from Eritrea known from one female, but they did not described in that time.This species was formally described years later as N. rufi venter Kieffer, 1913by Kieffer (1913).Kieffer (1914a) transferred a doubtful Philippine species Rhabdepyris tagala (Ashmead, 1905) to Neurepyris.Both species are known only from their types.However our initial analyses indicate that both species are males of Pristocerinae because both of them have the metanotal well developed medially, propodeum with convex sides, metasoma depressed and genitalia with complex aedeagus.
In this study, we aimed to reconsider the genus, presenting a new view about both species here redescribed and illustrated.

Material and methods
The material studied was borrowed from the Museo Civico di Storia Naturale 'Giacomo Doria' -MCSN (F.Penati) and the National Museum of Natural History -USNM (D.Furth).
Measurements and indices used in this study are as follows: body length measured from the apex of clypeus to the posterior margin of the last metasomal segment; LFW, length of forewing; LH, length of head, measured in frontal view, from vertex crest to median apical margin of clypeus; WH, width of head, measured in frontal view, its maximum width including eyes; WF, width of frons, measured in frontal view, its minimum width, usually about bottom of eyes; HE, height of eye, measured in lateral view, across its maximum height (length); OOL, ocelli-ocular line, measured in latero-dorsal view, the shortest distance from eye top to posterior ocellus; WOT, width of ocellar triangle, measured in frontal view, maximum width, including ocelli; DAO, diameter of anterior ocellus, measured in frontal view; distance of ocellar triangle to vertex: measured in dorso-posterior view, distance from posterior ocellus to vertex crest; VOL, vertex-ocular line, measured in dorsal view, distance from eye top to vertex crest.
WINGS. (Fig. 5).Forewing with radial vein long and evenly curved; R1 short than half length of stigma; M+Rs (basal vein of Evans) concave; cu-a (transverse vein of Evans) angulate; Cua (discoidal vein of Evans) tubular for short distant, intersecting M+Rs.Hind wing without basal hamuli, with six equidistant median hamuli.
METASOMA.(Fig. 6A-B).Not petiolate; anterior half mostly polished, posterior half mostly very weakly coriaceous; progressively more hairy posterad.Sternite VII rectangular, without median stalk, lateral stalk very short, posterior margin with median narrow semicircular concavity.Hypopygium (Fig. 6A-B) strongly concave, but not divided into two lobes, inner margin badly concave, inner surface setose apically, median stalk about 0.6x as long as plate, lateral stalk very short, dorsal surface short, surrounding inner margin of lobe, progressively higher basad, posterior margin with perpendicular median lamina directed upward.
GENITALIA.(Fig. 6C-D): paramere subvertical, about as long as basiparamere; placed dorsally, outer surface mostly hairy, bilaminar, except basal-inner quadrant; apex rounded; wide, dorsal margin much developed mesad, especially basally, basal half convex, apical half concave, with median angulate callus; ventral margin straight.Basiparamere with dorsal margin inclined and truncate in lateral view.Basivolsella completely outlined and separated from basiparamere, short, basal half subquadrate.Digitus laminar, wide, rounded, dorsal surface convex and denticulate.Cuspis laminar, base completely angled upward, somewhat narrow, apex rounded.Aedeagus bulging, divided into dorsal and ventral parts, apex of dorsal one with three pair lobes, dorsal lobe completely angled ventrad, tubular, apex deeply bifurcated; median lobe stout, apical margin with delicate fi lament apicad, ventral margin denticulate, inner margin membranous and little hairy, ventral lobe with large expansion directed downward; ventral surface laminar expansion, angled laterad with apex rounded and base with ventral loop of irregular margin; ventral part of aedeagus laminar with base wider than apex, inner surface vertical, inner lower part with ridge, apex divergent.Genital ring subrectangular.Basal ring narrow.Apodeme narrow, extending beyond genital ring.

Discussion
According to the most recent Bethylidae catalog (Gordh & Móczár 1990), Neurepyris type species, Neurepyris tagala, was considered lost but recently Fabio Penati found it in the MCSN collection.Thus, we were able to access it and study both species of the genus.Our results confi rm the initial hypothesis that both species of Neurepyris are Pristocerinae.Our analysis also indicated that the only two species belong to distinct genera within Pristocerinae.A historical review of nomenclatural acts of the genus indicated that the type species designation was wrongly perpetuated by Gordh & Móczár (1990) and its sex was misidentifi ed.Kieffer (1913) misidentifi ed the sex of the holotype of Neurepyris rufi venter as a female.However, the specimen is clearly a male.There are two identifi cation labels attached to the type, one is Epyris rufi venter and the other is Neurepyris rufi venter.However this specimen has never been formally named as Epyris rufi venter.That indicates a probable confusion about the concept of the genus within Epyrini by the author species, although both Neurepyris and Epyris are classifi ed as Epyrini.
Neurepyris rufi venter is easily recognized as belonging to Pristocerinae mostly because it has the metanotum well developed medially.It also has the hypopygium deeply divided into two lobes.Such a condition is found in fi ve genera of this subfamily: Dicrogenium Stadelmann, 1894, Neodicrogenium Benoit, 1957, Diepyris Benoit, 1957, Kathepyris Kieffer, 1907 and Pristocera.Among these genera, the fi rst two have a large spine on the genal area which makes them distinct.Diepyris has the mandible sickle-shaped with only two apical teeth, whereas Kathepyris and Pristocera have mandibles that are wide apically, with four or fi ve apical teeth.However, Kathepyris has the forewings with subdiscoidal and cubital veins reaching the apical margin of the wing whereas Pristocera does not.Neurepyris rufi venter should certainly be assigned to Pristocera because it has the combination of the following characters: the mandibles wide apically with fi ve apical teeth, the genal area without spine, the forewings with subdiscoidal and cubital veins barely visible and not reaching apical margin of the wing, and the hypopygium divided into two lobes.Because of that we propose the transference of Neurepyris rufi venter to Pristocera and the subsequent new combination (P.rufi venter comb.nov.).
Neurepyris tagala was originally described as a species belonging to Epyris Westwood, 1832 (see Ashmead 1905).However Kieffer (1908) transferred it to Rhabdepyris and indicated a possible confusion about the generic identity of this species.Finally it was transferred to Neurepyris by Kieffer (1914a).Nevertheless, both transferences remained within Epyrini genera.However, what fi rst called our attention is the fact N. tagala presents the metanotum well developed medially as it occurs in all Pristocerinae.Actually, this condition is pointed out as the main synapomorphy of this subfamily (Sorg 1988;Terayama 1996).Neurepyris tagala has the basal tooth of the mandible not curved inwardly, the median lobe of clypeus not depressed near the antennal insertions, and the aedeagus consisting of one lamina.This combination of characters addresses this species to Apenesia Westwood.Because of that we propose the transference of Neurepyris tagala and the subsequent new combination (Apenesia tagala comb.nov.).
According to Gordh & Móczár (1990), N. tagala was designated the type species of Neurepyris by 'subsequent monotypy'.However, this assumption does not fi t with the nomenclatural history of the genus.
Neurepyris was established by Kieffer (1905) in a key.He did not provide any description nor include any species on it.Curiously, Kieffer & Marshall (1904-1906) mentioned that Neurepyris was created to accommodate one species from Eritrea known only from one single female (see the footnote in Kieffer & Marshall's key, page 251).However, there was no species formally addressed to Neurepyris in that time.Kieffer (1908) formally described Neurepyris and pointed out that "l´espèce… sur lequel ce genre est etabli provient de l'Erythree et n'a pas encore ete decrit" (= this genus was established based one species from Eritrea not described yet).Kieffer (1913) formally provided the description of the fi rst species of Neurepyris, N. rufi venter, the species from Eritrea.Kieffer (1914a) transferred Rhabdepyris (Trichotepyris) tagala (Ashmead, 1905), previously described in Epyris, to Neurepyris.He also indicated Neurepyris rufi venter as the type species of this genus.
Given this scenario, we offer three conclusions.First it is clear N. rufi venter is the type species by subsequent monotypy in Kieffer (1913), second the indication of type species by Kieffer (1914a) was unnecessary, and third the citation of N. tagala as type species by Gordh & Móczár (1990) is not adequate.So we assume that the correct type-species of the genus is Neurepyris rufi venter.Thus it has to be treated as a new junior synonym of Pristocera.