REVISION OF THE WORLD SPECIES OF ARADOPHAGINI (HYMENOPTERA: SCELIONIDAE)

Abstract Seven species of Aradophagini are classified in three genera. Aradophagus Ashmead includes A. fasciatus (USA, Canada, Europe) A. pulchricornis n. sp. (Mexico, USA), and A. microps n. sp. (South India). Ladora n. gen. includes L. brunnea n. sp. (type-species: Morocco, Mallorca, Gambia), L. maura n. sp. (South Africa), and L. trjapitzini n. sp. (Central Asia). Abuko n. gen. with A. sarotes n. sp. (type-species: Gambia). The tribe Aradophagini is redefined and its taxonomic position in the family Scelionidae is discussed. Keys to genera and species of Aradophagini are given.

The tribe Aradophagini was proposed by Kozlov (1970) to contain the genus Aradophagus Ashmead. Kozlov followed Ashmead (1893), Kieffer (1926), and Muesebeck and Walkley (1951) in classifying Aradophagus in the subfamily Telenominae. However, Masner (1976~) expressed the view that Aradophagus is evidently a highly derivative member of the subfamily Scelioninae, and is only convergent with some members of the Telenominae. The latter view is further substantiated in the present paper by several important discoveries. Extensive material of Aradophagus fasciatus became available in 1976 (Masner 1976b), permitting the much-needed dissection of the metasoma. The discovery of Ladora and Abuko, two new genera related to Aradophagus, has helped to clarify the higher classification of Aradophagini.
The principal morphological difference between the subfamily Telenominae, and the subfamilies Scelioninae and Teleasinae, is in the structure of the metasoma (Masner 1 9 7 6~) .
The sternal plates of the telenomine metasoma are solid, undivided sclerites, not differentiated laterally into laterosternites. Consequently, there is no locking system with the wide laterotergites (Fig. 2). On the contrary, the sternal plates of the Scelioninae and Teleasinae show distinct laterosternites usually locked into corresponding folds of the narrow laterotergites (Fig. 1). While the presence or absence of laterosternites is the essential distinction between these two groups of scelionid wasps, there are various secondary modifications within each of them.
The metasoma in Aradophagus is strongly depressed dorsoventrally, so as to be almost foliaceous in dry specimens, with the sternal plates thin, considerably less sclerotized than in other scelionids. This condition makes it difficult to determine the presence of laterosternites on dry, mounted specimens. The previous rarity of individuals of Aradophagus in collections did not allow for dissections. However, numerous individuals of A . fasciatus collected during 1976 were preserved in alcohol, and had the metasoma conveniently inflated to expose the delicate but distinct laterosternites at the sides of each sternal plate. These laterosternites were more conspicuous, due to the colour, in melanic males than in xanthic females. However, the wide laterotergites do not lock with the narrow laterosternites but rather with the main sternal plates (Fig. 3), i.e., the submarginal groove so typical for the Scelioninae is not developed in Aradophagini. The presence of laterosternites together with the antenna1 formula 12-12 and the position of the ocelli (described below) substantiate in our view the classification of Aradophagini in the Scelioninae rather than in the Telenominae. Kozlov 1970, Aradophagini Kozlov, Rev. Ent. U.R.S.S. 49: 215. 1976, Aradophagini: Masner, Mem. ent Soc. Can. 97: 52. Body moderately to considerably depressed dorsoventrally; foramen magnum of head located quite high, almost adjacent to top of vertex, leaving narrow occiput and unusually long distance to hypostomal arc; ocelli in a curved line or a very low triangle, lateral ones distinctly separated from inner orbits by at least one diameter; FIGS. 1-5. Schematic cross-section of (1) generalized metasoma in subfamilies Scelioninae and Teleasinae, frons without depression, with a short, blunt keel above antennal insertion; mandibles small, bidentate; palpi extremely short, formula 2-1; radicle remarkably elongate, about 113 of scape length; antennae in both sexes elongate, slender, subclavate in females, antennal formula 12-12; prepectus and skaphion absent; notauli abbreviate anteriorly, running parallel, or absent; scutellum narrow, almost strip-like, strongly transverse; metanotum and propodeum unarmed, the latter with more or less well defined median carina and lateral plicae; fore wings with 1 or 2 dark transverse bands; marginal vein at least as long as stigma1 vein, postmarginal vein long; hind wings with complete submarginal vein reaching to hamuli; metasoma strongly flattened dorsoventrally, moderately elongate, T2 or T3 the longest tergites; T7 in females extruded with ovipositor; laterotergites relatively wide, locking loosely on sternal plates, leaving only narrow space between them; submarginal groove not distinct.

Tribe Aradophagini
Three genera are known, viz. Aradophagus (Holarctic, Ethiopian, Oriental), Ladora (Palearctic, Ethiopian), and Abuko (Ethiopian). Representatives of Ladora and Abuko retain some plesiomorphic character states, viz. partly developed notauli, and OOL shorter than LOL. Aradophagus is a more derivative member of the tribe, with no notauli and OOL at least as long as LOL.
The metasoma of Aradophagini resembles superficially the telenomine type in that the laterotergites are relatively wide, not forming the usual submarginal groove on the ventral side. However, similar cases of convergence with Telenominae are known in other tribes of the Scelioninae, e.g. in Baeus Hal. (Baeini) and in Tiphodytes Bradl., Tanaodytes Msn., and Microthoron Msn. (Thoronini).
Host associations are not known with certainty as no representatives of the tribe have been reared. Ashmead's (1893) assumption of parasitism on Aradidae (Heteroptera) was never confirmed, but the present data seem to indicate association of members of Aradophagini with wood inhabiting insects. The flattened habitus of all members of this tribe indicates that the host eggs may be also flattened.

KEY TO GENERA OF ARADOPHAGINI
(1) Notauli absent; OOL slightly to considerably longer than LOL; laterotergites in ventral view wide (as in Fig. 3 Body strongly depressed dorsoventrally, head and mesosoma 2-3 times wider than high, metasoma even more depressed, almost foliaceous; head in frontal view subcircular or elongate; cheeks not striate; eyes glabrous or with minute hairs; OOL slightly to considerably longer than LOL; notauli absent; T2 distinctly longest of all tergites, longer than T3; laterotergites rather wide, not incised into sternites, no submarginal ridge.
8, Abuko sarotes n. sp. 9 , holotype, antenna. 9, A . sarotes n. sp. 9 , holotype, fore wing. MORPHOLOGY. This species was already adequately described by Masner and Kozlov (1965), who figured the female and described the unknown male. Ashmead's (1893) data on the palpal formula is not quite exact. The dissection of mouth parts revealed the labial palpi to be only 1-segmented, almost wart like, not 2-segmented as claimed by Ashmead. An interesting chromatic inversion exists in the colour pattern of the scape, A2 and A3 in females of fasciatus and pulchricornis. In fasciatus the tip of the scape, A2 and extreme base of A3 are light coloured whereas exactly those areas are darker in pulchricornis. The two species also differ in that the sexes are dichromatic in fasciatus (females xanthic, males melanic), but isochromatic in pulchricornis (both sexes xanthic).
DISTRIBUTION. This is a rare but apparently widely spread Holarctic species. It has been recorded from Florida and Massachussetts (Brues 1908), Texas (Muesebeck and Masner 1967), Kansas (Masner 1976b), Switzerland, Hungary and Soviet Union (Masner and Kozlov 1965;Kozlov 1971), and Romania (Fabritius 1975). Masner and Kozlov (1965) considered A . fasciatus as native to the whole Holarctic Region, rejecting the idea of spread by human activities such as commercial shipping, etc. However, new data presented in this paper might indicate at least some interactions in this respect. Over several years in July-August G.W. Byers observed large numbers of individuals of A . fasciatus swarming on windows inside a house in Lawrence, Kansas (Masner 1976b). Three males of A. fasciatus were collected inside a house in Palo Alto, California (see below). D.B. Levin (pers. comm.) found a dead female of A. fasciatus stuck to a window inside a wood-frame house in Waterloo, Ontario (first Canadian record). The closely related A. pulchricornis has been intercepted twice by quarantine procedures (Brownsville, Texas and Baltimore, Maryland).

Aradophagus pulchricornis n. sp.
Female. Length 1.3 mm. Xanthic species, generally orange-yellow; tips of mandibles brown; ocelli in dark brown pits; antennae remarkably varicoloured, with apex of scape, A2 and extreme base of A3 infuscate, brownish, rest of A3, A4-A7 and A12 whitish in contrast to dark brown A8-All; fore and hind coxae and trochanters whitish, mid-coxae whitish posteriorly and orange-yellow anteriorly; posterior margins of T2-T5 with darker band, darkest posterolaterally; fore wings predominantly infuscate, with two transverse dark bands separated by whitish band; hind wings-clear; hind tibiae slightly darker than rest of legs.
BIOLOGY. Host unknown. However, the circumstances under which the three specimens were obtained possibly indicate a host associated with human activities.
DISTRIBUTION. Probably Mexico, as one individual was intercepted with medicinal herbs on the Mexican-Texan boundary (Brownsville). The other two specimens were intercepted on a ship in Baltimore harbour. Unfortunately, no more data are available in the latter case.
DISCUSSION. This is the second species of Aradophagus known from the New World. It differs from A. ,fasciatus in several characters mentioned in the key. Furthermore, the sculpture on T2 and following tergites differs in the two species. In pulchricornis the fan of straie on T2 is much shorter, hardly exceeding the basal 113 of the tergite ( 9 8 ) , and T3-T7 are virtually sculptureless. In fasciatus the fan is much longer ( 9 ) extending to distal 314 of the tergite, being coriaceous and not striate distally, and T3-TI ( 9 8 ) show fine but distinct coriaceous sculpture on each of their anterior halves. However, the most striking difference between the two species is chromatic, with the sexes being isochromatic (xanthic) in pulchricornis but dichromatic in fasciatus which has xanthic females and melanic males. Male. Length 1.2 mm. Light brown; head (except for darker vertex), radicle and scape brownish yellow; A2-A12 brown; mesosoma chestnut brown; legs including coxae and prosternum dirty yellow; fore wings with two rather weak darker transverse bands, first below distal end of submarginal vein, second below stigmalis-postmarginalis; metasoma lighter than mesosoma, yellowish brown, T1 lightest.

Figs. 5-7
Female. Body rather flattened dorsoventrally, mesosoma twice as wide as high, metasoma fully 3 times as wide as high; head in frontal view subcircular, in dorsal view strongly transverse; eyes large, with minute scattered hairs; ocelli in curved line, lateral ones distant from inner orbits by more than one diameter; LOL less than twice as long as OOL; subocular suture distinct, carinate, cheeks fan-like striate; antennae 12-segmented, clava indistinct; notauli abbreviate anteriorly, well impressed in posterior half of mesoscutum, running parallel, located in extreme corners of mesoscutum (Fig. 5); scutellum broadly transverse, more than twice as long as metanotum; fore wings with only one dark transverse band in distal half and no tuft of erect bristles below submarginal vein; marginal vein slightly longer than stigma1 vein; postmarginal vein long; metasoma with laterotergites narrower than space between them (ventral view); T3 longest and largest; 7 visible tergites, T7 extruded with ovipositor.

BIOLOGY. Unknown. One species collected on Tamarix-shrubs
ETYMOLOGY. An euphonic word, to be regarded as feminine.
Anterior third of mesoscutum as well as parts lateral to notauli irregularly coriaceous; space between notauli virtually sculptureless, highly shining, with a few scattered setigerous punctures and with very narrow streak of delicate aciculae running along notauli inward; scutellum and metanotum smooth; propodeum mostly rugulose medially, with large smooth field below spiracle; marginalis, stigmalis, and postmarginalis of proportions 9:6:13.
Metasoma as in brunnea except for T3 which is evenly coriaceous, with polygons in median part not distinctly elongate. (USNM), same data a s holotype; 1 9 , paratype (CNC No. 15501), same data a s holotype.

Male.
BIOLOGY. The type series was allegedly reared from an armoured scale; however, this needs to be confirmed considering the known host associations of other Scelioninae.
DISTRIBUTION. South Africa.
DISCUSSION. L. maura differs markedly from the other two species of Ladora by its dark colour. There are subtle differences between maura on the one hand, and brunnea and trjapitzini on the other, in O O L and the sculpture of the propodeum below the spiracles.

Ladora trjapitzini n. sp.
Female. Length 1.2 mm. Chestnut brown; head and mesosoma darker than metasoma, T1 not lighter than rest of metasoma; A1-A5 dirty yellow, A6-A12 as well as legs including coxae light brownish yellow; fore wing with one dark transverse band under stigmalispostmarginalis.
Coriaceous sculpture of mesoscutum better developed than in maura but considerably finer than in brunnea; space between notauli appears almost smooth; however, under higher magnification a delicate coriaceous sculpture is seen; scutellum, metanotum, propodeum, and ratio of veins in fore wing as in brunnea.
Metasoma as in brunnea except that polygons in middle part of T3 are not elongate. Male. Unknown.
DISCUSSION. This species is a n intermediate between L. brunnea and L. maura in sculpturing of the frons and mesoscutum. However, the ratio of A 2 and A 3 will distinguish L. trjapitzini at once from the other two species. The species is named after its collector, the well-known Soviet chalcidologist, V.A. Trjapitzin (ZIN, Leningrad).
Figs. 8, 9 Female. Body considerably flattened dorsoventrally, almost twice as wide as high, metasoma fully three times as wide as high; head in frontal view subcircular, in dorsal view strongly transverse; eyes large, with minute scattered hairs; ocelli in curved line, lateral ones distant from inner orbits by -more than one diameter; LOL more than twice as long as OOL; subocular suture fine, ecarinate; cheeks with only very short, inconspicuous striae at base of mandibles; antennae 12-segmented, clava indistinct; notauli as in Ladora, abbreviate anteriorly, well impressed in posterior half of mesoscutum, running parallel, located in extreme comers of mesoscutum; scutellum broadly transverse, about twice as long as metanotum; fore wings with two dark transverse bands; at about distal 113 of and just below submarginal vein with Volume 11 1 THE CANADIAN ENTOMOL. OGIST 1099 a tuft of strong erect black bristles forming a slanting cone (Fig. 9); marginal vein slightly longer than stigma1 vein; postmarginal vein long; metasoma with laterotergites in ventral view subequal to space between them; T2 slightly longer than T3; 7 visible tergites, T7 extruded with ovipositor.
DISTRIBUTION. The Gambia (West Africa).
ETYMOLOGY. Named after the Abuko Nature Reserve in The Gambia (West Africa); to be regarded as masculine.
DISCUSSION. Abuko in some respects presents a mosaic of characters found in Aradophagus and Ladora. With Aradophagus it shares the sculpture of the cheeks, the 2-banded fore wings, and the general shape of the metasoma. With Ladora it shares the position of the ocelli and the shape of the notauli. However, the tuft of long black erect bristles on the fore wing in Abuko is virtually unique among all genera of Scelionidae, and is reminiscent of a similar structure found in some Chalcidoidea (Pteromalidae, Eulophidae). Similarly, the overall smoothness, body colour, and iridescent eyes in Abuko are also reminiscent of some members of Chalcidoidea (e.g. Cerocephalinae of the Pteromalidae), which are parasites of Scolytidae and Anobiidae. Abuko sarotes n. sp.
Male. Unknown. ETYMOLOGY. Sarotes in Latin means "sweeper" or "broom", indicating the presence of the unique brush on the fore wing.