Hymenoptera ) of Madagascar 1 Trigonalidae ( Hymenoptera ) of Madagascar

Seven species of the primarily hyperparasitoid family Trigonalidae are reported from Madagascar: Orthogonalys brevis Smith & Tripotin, sp. n., Orthogonalys gigantea Benoit, 1951; O. hova Bischoff, 1933; O. maculata Bischoff, 1933; Orthogonalys parahova Smith & Tripotin, sp. n., O. seyrigi Bischoff, 1933; and Trigonalys natalensis Kriechbaumer, 1894. Diagnoses and a key to species are given.


Introduction
Other than original descriptions and catalogs, no work has been done on the trigonalid fauna of Madagascar.In the most recent list by Carmean and Kimsey (1998), three species of Orthogonalys Schulz and one species of Trigonalys Westwwod are listed.About 200 specimens were collected during the survey of Madagascar by the California Academy of Sciences, San Francisco.Study of this material, as well as the types and type series of the described species, has resulted in the recognition of six species of Orthogonalys and one species of Trigonalys.Two main problems in Orthogonalys included the validity O. gigantea Benoit, which appeared to be only a large form of O. hova Bischoff (Carmean and Kimsey 1998), and O. seyrigi form maculata Bischoff, which has been considered as a synonym or as a subspecies of O. seyrigi Bischoff (Weinstein and Austin 1991).Here, we recognize both O. gigantea and O. maculata as valid species.In addition, two new species of Orthogonalys are described.Illustrations and a key to species for Madagascar are provided.
Trigonalidae are hyperparasitoids mostly of Ichneumonidae (Hymenoptera) and Tachinidae (Diptera) parasitizing Lepidoptera caterpillars and social wasps (Weinstein and Austin 1991, Carmean 1991, Carmean and Kimsey 1998, Murphy et al. 2009).However, two species of Taeniogonalos in Australia have been reared directly from sawfly hosts (Raff 1934, Carne 1969).The only records of hosts for Madagascar are by Benoit (1951), who mentioned that Orthogonalys hova was obtained from rearing of a limacodid moth, and Schultz (1910), who stated that Trigonalys natalensis was reared from an unidentified lepidopteran pupa.
Trigonalids have been recorded from the provinces of Antsiranana, Fianarantsoa, Majunga, Toamasina, and Toliara and probably occur throughout Madagascar.

Materials and methods
Images were obtained using an EntoVision Imaging Suite that included a firewire JVC KY-75 3CCD digital camera mounted to a Leica M16 zoom lens via a Leica z-step microscope stand.Multiple focal planes were merged using Cartograph 5.6.0 (Microvision Instruments, France) software.
Locality data are recorded essentially as they appear on the specimen labels.Characterization of genera and validity of generic names was treated by Carmean and Kimsey (1998).Terminology is based on Huber and Sharkey (1993).
Madagascar previously was divided into six provinces and 22 regions.The 22 regions became the highest subdivisions in 2007 (http://en.wikipedia.org/wiki/Madagascar).All label data, as well as distributions reported in the literature, are according to the original six provinces, which are used in this paper.Frons completely white, not interrupted in center by broad, black band (Fig. 4) or interrupted by a very narrow black line; gena black (Figs 1, 2); paramere of male genitalia short, round, slightly broader than long, brown (Fig. 6), less than half length of gonostipes (Fig. 7); about 5- Description.Female (Fig. 1).Length.7.0 mm.Similar to male except for penultimate metasomal segment and sheath black, and genitalic characters.
Distribution.Madagascar: Antsiranana, Toamasina.Etymology.The specific name is from the Latin brevis, referring to the short parameres of the male genitalia.
Remarks.The short parameres of the male genitalia place this species close to O. seyrigi; however, the parameres of O. brevis are shorter and broader than long whereas those of O. seyrigi are about as broad as long (Figs 6,39) The color of O. brevis is entirely different than in O. seyrigi and more similar to that of O. hova with the almost entirely orange mesosoma and metasoma.The color of the head is consistently different than in other Orthogonalys species with more white across the frons, which is not interrupted with black, except sometime for a very narrow medial line, and the entirely black gena (Figs 4,5).The metasoma of O. brevis is almost entirely orange with only apical segment white and the penultimate segment mostly black.Description.Female (Fig. 8).Length, 14 mm.Head black; inner orbits broadly from top of eye through malar area and gena, clypeus except for median V-shaped stripe, and mandible except apex white (Figs 11,12).Mesosoma orange with hind corners of pronotum, lower anterior corner of pronotum, and metascutellum yellow.Legs dark orange; hind tibia and tarsus darker, nearly black.Metasoma dark orange with black spot at center of 1 st tergite, small faint black spot at center of 2 nd tergite, lateral and posterior bands on 3 rd tergite, and most of tergite 4 except for white posterior margin; tergite 5 white; laterally and ventrally mostly yellow with black spot on 6 th sternite and sheath black.
Remarks.The status of this species has been questionable.Carmean and Kimsey (1998) thought that it might be only a large form of O. hova.Benoit (1951) described the species from a single female.He separated O. gigantea from O. hova by its large size (14 mm) and by the following: teeth of mandibles broad but not very long, the inferior tooth directed downwards; clypeus very thickened; antennal sockets closer together; punctation of scutellum sparser but stronger; propodeum broader than long, with transverse carinae complete and regular (not tangled and less strong); color identical except the white "atténué" on the head and sides of abdomen.We have not seen additional females but have examined one male that compares in color and structure with the holotype female and was collected not far from the type locality, and we consider them to be the same species.
After study of both sexes, we note a number of characters that justify treating O. gigantea as a valid species.Size cannot be regarded as a valid character in trigonalids.However, in comparison to O. hova, the specimens examined are consistently larger.The female holotype of O. gigantea is 14 mm and the male is 15 mm, whereas the largest O. hova female is 9 mm and males 5.0-9.5 mm, with most in the 7 mm range.We are not sure about the clypeus, mandible, punctation of the scutellum, and closeness of the antennal sockets as Benoit described; these do not seem significant.However, we emphasize the following characters: the sculpture of the propodeum of O. gigantea is more transverse (Fig. 10) than the more reticulate sculpture of O. hova (Fig. 16) and other Orthogonalys species; the color of both are similar, except that O. gigantea has the metascutellum yellow and the male has the metasomal tergites mostly black whereas the metascutellum of O. hova is orange and the metasomal tergites of the male are considerably orange; and the parameres of O. gigantea are dark orange and rounded at the apex (Fig. 13), whereas those of O. hova are white and more acute at the apex (Fig. 20).
Host.Benoit (1951) mentioned that a specimen was obtained by Seyrig from rearing of a limacodid moth.
Remarks.See discussion under O. gigantea for characters separating the two species.This is the most commonly collected species in Madagascar and occurs from the lowlands to high altitude rainforests.We have examined about 130 specimens.Description.Female (Fig. 21).Length, 14 mm.Similar to male except metasoma white laterally and ventrally and tergites 1-4 mostly black dorsally, 5 and 6 black anteriorly.

Orthogonalys maculata
Male. (Fig. 22).Length, 8.5-12.0mm.Antenna black with antennomere 10 white at apex, 11-12 completely white, and 13-15 partly brown on lower surface and white on upper surface (Fig. 26) or 11-14 completely white.Head white with vertex black, black extending to dorsal margin of eyes, down posterior margin of gena, and anteriorly onto frons to antennal sockets (Figs 24,25).Mesosoma anterior to transscutal articulation red, posterior to transscutal articulation black with mesoscutellum and metascutellum white; pronotum with anterolateral margins white and laterally white with broad black intermediate stripe; mesopleuron and mesosternum mostly white with black on dorsal margin and margins of mesepimeron black; metapleuron mostly white.Legs orange with hind tibia and basal half of basitarsomere darker, hearly black.Wings hyaline; veins and stigma black.Metasoma mostly orange; tergite 1 mostly black, tergites 2-6 orange with black spots at apices of tergites 5 and 6 and sometimes laterally on tergites 2 and 4, and apical tergite white.Male genitalia with gonostipes brown and paramere white.
Remarks.The color of O. maculata is almost identical to O. seyrigi, which probably led to its description only as a form of that species by Bischoff (1933).Color differences in O. maculata include only 2-4 central antennomeres completely white, darker to black hind tibia and tarsus, and mostly black first metasomal tergite.Orthogonalys seyrigi has more antennomeres completely white, usually 6, orange hind tibiae and tarsi concolorus with the rest of the legs, and the first metasomal tergite orange.Other differences in O. maculata are the narrower mesoscutellum, which is slightly as long as broad (Fig. 23), and the paramere of the male genitalia which is long and white (Fig. 27), much more like the parameres of O. gigantea and O. hova, rather than the short, brown paramere of O. seyrigi (Fig. 39).Because of the similar male genitalia, we regard this species more closely related to O. gigantea and O. hova rather than to O. seyrigi.Description.Female.Unknown.Male (Fig. 29).Length, 10.0 -11.0 mm.Antenna black; antennomeres 11 at apex, 12-15, and base of 16 white.Head white with vertex and upper frons black, black extending to dorsal margins of eyes and medially to near antennal sockets (Figs 31,32); clypeus white.Mesosoma orange with posterior lateral angles of pronotum white and metascutellum yellow; pleura and sterna lighter orange than dorsum.Legs orange with hind tibia and tarsus black.Wings hyaline with veins and stigma black.Metasoma with tergite 1 mostly black, tergites 2-4 orange, tergite 5 mostly black, and tergites 6 and 7 white, narrow lateral edges of tergites white and sternites mostly white; genitalia with gonostipes brown and paramere white.
Etymology.From the species name "hova" with the Greek prefix "para" meaning resembling O. hova.
Description.Female.Known by female allotype as designated by Bischoff (1933) in MNHN.The head is missing but otherwise similar to male.
Male. (Fig. 35).Length, 8.0 -9.0 mm.Antenna black with apex of 9 th antennomere and antennomeres 10-15 completely white.Head white with vertex black, black narrowly extending to dorsal margin of eyes and as broad stripe anteriorly to antennal sockets; small medial black spot on dorsal margin of clypeus (Figs 37,38).Mesosoma white laterally and ventrally with central dark stripe on pronotum; area anterior to transscutal articulation red, posterior to transscutal articulation black with mesoscutellum and metascutellum white.Legs orange with hind tibia and tarsus sometimes slightly darker orange than rest of legs.Wings hyaline; veins and stigma black.Metasoma mostly orange, base of first tergite black; lateral margins of tergites and posterior margins of sternites whiter.Male genitalia brown.
Remarks.Bischoff (1933) described the species from both sexes.All types at MNHU are males, and all except one of those at MNHN are apparently males.The one female at MNHN is probably the allotype, but it has the head missing and therefore we cannot characterize the female.We have not seen specimens other than the type series.
The short parameres of the male genitalia (Fig. 39) relate this species to O. brevis (Fig. 6) but in O. seyrigi the parameres are about as long as broad, the genae are white, the white on the frons is interrupted by a broad median black stripe, and the mesosoma is red anterior to the transscutal articulation, and black posterior to it with the mesoscutellum and metascutellum white.
Host.Schultz (1910) stated that this was reared from an unidentified lepidopteran pupa.

Remarks.
We have seen only one specimen and do not characterize this species further.It is readily separated from the Orthogonalys species by its black antennae, mostly black body, and infuscated wings.Two subspecies may occur in Madagascar, the typical subspecies from Madagascar, without locality, and T. natalensis madegassa from only the two localities given by Bischoff (1933).We see no reason to distinguish subspecies at present.