A REVISION OF THE NEW WORLD SPECIES OF LEPTOTELEIA KIEFFER (HYMENOPTERA: SCELIONIDAE), EGG PARASITES OF CRICKETS

Abstract The genus Leptoteleia of the New World is revised. Twenty-five species are recognized, of which 23 are new to science: L. alexandrae ♀ (Brazil), L. americana ♀ ♂ (USA), L. andreai ♀ (Cuba), L. annarum ♀ (Brazil), L. ferdinandi ♀ (Dominican Republic), L. jarmilae ♀ (Ecuador), L. josephi ♀ (Dominican Republic), L. kareli ♂ (Jamaica), L. lubomiri ♀ ♂ (Dominican Republic), L. majkae ♀ (Panama), L. marcelae ♀ ♂ (Brazil, Paraguay), L. mariae♀ ♂ (Argentina), L. marketae ♀ (Brazil), L. martae ♀ (USA), L. miladae ♀ ♂ (Brazil), L. monicae ♀ (Dominican Republic), L. normani ♀ ♂ (USA), L. petrum ♀ (Bélize), L. radeki ♀ (Dominican Republic), L. stani ♀ ♂ (Dominican Republic), L. vaclavi ♀ (USA), L. verae ♀ (Brazil), and L. zdenae ♀ ♂ (Peru). The two previously described species, L. oecanthi (Riley) ♀ ♂ and L. arndti Dozier ♀, are redescribed. Generic diagnosis of Leptoteleia and the key to New World species are given. The higher classification, interrelationships, and distribution of Leptoteleia species are discussed.


INTRODUCTION
H~STORIC AL The name Leptoteleia was proposed by Kieffer (1908) as monotypic to contain the Nearctic Baryconus oecanthi Riley. Masner and Muesebeck (1968) selected the lectotype of Riley's species and the present study represents the first redescription. The interpretation of Leptoteleia, however, was not quite clear at all times. Muesebeck and Walkley (1951) synonymized it with Baryconus Foerster because of supposed isogenotypy. Following the clarification of the confused designation of type-species of Baryconus (Muesebeck and Walkley 1956) the name Leptoteleia was resurrected (Muesebeck in Krombein et al. 1958). Szabo (1962) and Masner (1964) upheld the resurrection of Leptoteleia. In 1931 Nixon described Thelepte without comparing it with Leptoteleia. Masner (1976) synonymized Thelepte with Leptoteleiaand redefined the latter genus.
Six species have been placed in Leptoteleia but only four are still in the genus (Masner 1976). Riley (in Ashmead 1893) described the Nearctic L. oecanthi. Dodd (1913Dodd ( , 1914 described two species from Australia, viz. L. australica and L. aurea, but later (Dodd 1916) transferred the former species to Baryconus (auct. nec Foerster). Galloway (1976) studied Dodd's types and transferred his two species to Duta Nixon thus leaving no described species in Australia. Dozier (in Arndt and Dozier 1931) described the first Caribbean species, L. arndti from Haiti. Nixon (1931) described Thelepte serapis from South Africa and finally Risbec (1956) added L. antsingyi from Madagascar. Masner (1976) studied the types of all but Dodd's species and transferred Thelepte serapis to Leptoteleia. Masner (1976) pointed out that Leptoteleia is represented in the Oriental region by several, and in South America, by numerous undescribed species.

BIOLOGY, DISTRIBUTION, and RELATIONSHIPS
The biology is known only in two species, namely L. oecanthi and L. arndti. Both species are solitary endoparasites of cricket eggs, the former of Oecanthus niveus (DeG.) and 0. nigricornis quadrimaculatus Beut., the latter of Chremon repentinus Rehn. Since crickets occasionally are injurious to a variety of crops (e.g. C. repentinus on coffee trees) the egg parasites may play an important role in the biological control. Ayers (1884) published a detailed account on postembryonic development of an egg parasite of 0. niveus; the parasite was referred to by Ayers as Teleas sp, reflecting the specialized derivative, the arndti-group. There are two distinct chromatic subgroups, viz. the melanic oecanthi (s. str.) subgroup with six species (alexandrae, annarum, mariae, martae, oecanthi, and zdenae) and the xanthic marcelae-subgroup with five species (jarmilae, majkae, marcelae, rniladae, and verae).
The arndti-group (Fig . 7). This highly apomorphous group is characterized as follows: prothorax dorsally strongly developed both in cervical and humeral regions, as long as scutellum; body extremely elongate, more or less flattened dorsally, and in lateral aspect almost in one level; epomia absent; occipital carina fine, complete, or nearly obsolete; head usually subglobose (i.e. not cubical); A2 in females distinctly shorter than A3, A4, or A5, and clava not abrupt, sub-tetramerous, distinctly more slender than in the other two groups. The arndti-group seems to be peculiar to the Neotropical region, with two species in the Caribbean and one in South America. Three species are in this group (arndti, marketae , and monicae).
The americana-group. All members of this group are distinct: occipital carina incomplete; pentamerous club in female antennae abrupt; head cubical; epomia of prothorax usually sharply angulate; A2 in females at least as long as A3, A4, or A5 or longer; prothorax dorsally short, if longer (e.g. in andreai, petrum, and vaclavi) then never as long as scutellum (different from arndti-group). The americana-group is not known in South America, but is dominant in the Caribbean subregion (7 spp.) to lesser extent in Central America and Nearctic region (4 spp.). This group is considered quite apomorphous as compared with oecanthi-group largely because of the obliterated occipital carina and abrupt pentamerous antennal club in females. There are 11 species in this group (americana, andreai, ferdinandi, josephi, kareli, lubomiri, normani, petrum, radeki, stani, and vaclavi).
MORPHOLOGY AND MEASUREMENTS Morphological terms used in this paper are identical with those used by Masner (1976). The only new term used is the parapsidal carinae, referring to a pair of longitudinal keels (carinae) situated outside of the notauli, i.e. where the true parapsidal furrows occur in some other groups of Hymenoptera.
Measurements given in "relative proportions" refer to maximal length divided by maximal width (antennal segments, metasomatic tergites, etc.). This, rather than indexes, gives the reader a chance to compare various parts of the body.  Ashmead, 1893. Very slender, elongate species with cylindrical body, in some species slightly depressed dorsoventrally. Head almost cubical or slightly transverse or sub-ellipsoidal.
Frontal depression more or less developed; rarely closed by keels from all sides, usually only above, in some species upper kee1 only faintly indicated; depression with transverse sculpture (except in L. serapis and L. marketae). Eyes usually hairy, rarely glabrous or with short scattered hairs. Mandibles tridentate. Malar groove developed or absent. Clypeus small, not protruding. Radicle of antenna short, usually shorter than one-sixth of scape. A3-A5 in females usually elongate, clava 5-, rarely 4or 6-segmented, abrupt or semiabrupt. Skaphion absent. Prepectus well developed. Notauli usually absent, rarely fine and then abbreviate. Mesoscutum usually with distinct parapsidal carinae indicated in posterior one-third by shining keels. Metanotum medially expanded into lamina of various shape, usually transverse, sinuate, bilobate, subtriangular or semielliptic, rarely tridentate. Propodeum often partly overlapped medially by metanotal lamina, usually excavate medially in females to house hump of TI, its dorsal carinae not produced into spines. Submarginal vein not "broken" before joining marginalis; basalis and medialis absent; marginalis elonagate, distinctly longer than stigmalis, postmarginalis longer than marginalis. Metasoma spindle-like usually widest at anterior one-third or at the middle. T1 in female usually with a hump of various size fitting more or less into propodeal cavity; hump not developed in one species. T7 in females always external, well sclerotized, articulating with T6 and not with ovipositor, not extruded with ovipositor. Ovipositor sheaths extruded and sclerotized in most species.
(1) Cervical part of pronotum (from pro-mesonotal suture to foramen occipitalis) remarkably elongate, as long as or longer than scutellum; and female A2 distinctly shorter than A3, A4, or A5, antennal clava sub-tetramerous and slender; extremely elongate forms with body more or less flattened dorsally (  Frons between upper margin of depression and anterior ocellus finely granular-punctate, with no other sculpture; A6 almost twice as long as wide (Jamaica) . . . . . . . L. kareli n. sp.
Frons between upper margin of depression and anterior ocellus with transverse ridges or rugae (rather fine in L. stani); A6 at most 1.5 times as long as wide . . Female (Lectotype, USNM). Length 2.5 mm. Melanic species; colouring slightly lighter than in original description due to age of the specimen; head black, mesosoma dark brown, metasoma chestnut brown, antennae dark brown, legs light brown, with lighter apices; wings clear.
Mesosoma slightly higher than wide; epomia of prothorax not angulate (dorsal aspect); mesoscutum slightly distorted due to presumed teneral condition of the lectotype, finely reticulate, with no apparent punctation; scutellum with same sculpture as mesoscutum; metanotal lamina sinuate, excavate medially; dorsal propodeal carinae running diagonally from posterolateral comers of propodeurn up to underneath metanotal lamina, propodeal cavity hence large to accommodate the sizable hump of T1; sides of prothorax reticulate, mesopleura reticulate in its lower part, smooth in its upper depressed part; metapleura almost entirely smooth; fore wings reaching to about middle of T5; marginal vein almost twice as long as stigmalis; postmarginalis almost twice as long as marginalis; hind basitarsus as long as T2-T5 combined.
DISTRIBUTION. Sporadically from Pennsylvania to Nebraska and Arizona; it is not recorded from Canada but expected in southwestern Ontario.
BIOLOGY. Reared from eggs of tree crickets, Oecanthus niveus (DeG.) and 0 . nigricornis quadripunctatus Beut. (Muesebeck and Walkley 1951). Riley (in Ashmead 1893) and Kieffer (1926) cite Ayers' (1884) paper dealing supposedly with L. oecanthi. This record is quite problematic (see Introduction, p. 353). L. oecanthi may not be the rare species it appears to be from the available material. The adults stay most probably in bushes and trees where they get seldom swept by hymenopterists. It is possible that the wasp shares the entire distribution with its hosts, the Nearctic tree crickets.
VARIABILITY. The metanotal lamina seems to vary in females from only slightly emarginate to deeply sinuate. The hump on T1 in females may vary in size and sculpture; it is quite large in the lectotype and almost smooth, whereas in females from Michigan and Pennsylvania it is shorter and with very delicate reticulation at the top. Similarly, the wings may reach almost to the apex of T5 in some females.
REMARKS. L. oecanthi belongs to the group of dark (mostly black) members of the oecanthi-group of Leptoteleia. Its relatives are L. mariae (Argentina), L. martae
Male. Very similar to that of L. oecanthi but differing from it by shape of metanotal lamina which is remarkably wider, entire, i.e. not excavate medially; similarly as in female the longitudinal rugulosity on T1-T4 is more obscured by reticulation. REMARKS. Whether alexandrae is an independent species or just a variety of annarum can not be determined at this moment. Some characters (notably the shape of propodeal carinae) as well as the huge distance between the two type localities are reasons why I prefer to consider these two as valid species. Dedicated to my charming sister-in-law, Mrs. Alexandra Masner (n6e H&) of Dielsdorf-Zurich.
Mesosoma as wide as high; epomia of prothorax not developed; mesoscutum and scutellum scaly-reticulate, notauli absent but almost inconspicuous depressions still exist in front of scuto-scutellar suture, parapsidal carinae well defined; metanotal lamina large, deeply sinuate to house top of hump; propodeum excavate (like a bowl), the dorsal carinae pushed to extreme sides; sides of prothorax scaly-reticulate; mesopleura partly, metapleura entirely smooth and shining; fore wing not surpassing middle of T5; hind basitarsus slightly longer than tarsomeres 2-5 combined.
Mesosoma only slightly higher than wide; epomia of prothorax not developed; mesoscutum and scutellum reticulate; notauli absent; parapsidal carinae fairly long but not too prominent; metanotal lamina relatively short, entire, not excised or notched posteriorly; dorsal propodeal carinae parallel in their upper part, gap between them very narrow; fore wings reaching to upper margin of T6.
REMARKS. This seems to be the most common and widespread species of Leptoteleia in South America. It is quite distinct chromatically in female sex on account of orange body contrasting to black head. The males are considerably darker with little of the bicolor effect of the females. Named after my beloved wife, Marcela, in recognition of her sincere admiration of my beautiful little wasps.
Mesosoma slightly wider than high, mesonotum rather flattened; epomia absent but frontolateral margins of prothorax sharp; mesoscutum and scutellum reticulate-punctate; notauli wanting; parapsidal carinae prominent; scutellum wide, fully 3 times wider than long; metanotal lamina at about 60" angle, semitransparent, subtridentate, almost touching top of hump on T I ; propodeum deeply excavate to house large hump, dorsal carinae pushed to sides, crowned by semitransparent blades similar to that on metanotum; sides of prothorax reticulate, most of mesopleura and entire metapleura smooth and highly shining; fore wings reaching to middle of T5.
Metasoma almost 5 times as long as wide; tergites in relative proportions 38:32, 37:40, 34:40, 31:37, 25:32, 17:23, 16: 10; top of hump finely net-like reticulate, rest of T I longitudinally costate; T2-T4 longitudinally costate with interspaces mostly smooth and shining; T5 with rather delicate longitudinal elements and some reticulation, T6 and T7 reticulate, T7 fully exposed and not even partly concealed by T6; ovipositor and sheaths protruding. REMARKS. This is a very distinct species of the xanthic subgroup of the oecanthi-group. The unique shape of metanotal lamina together with colour patterns distinguish L. jarmilae at once. It is my great pleasure to name this handsome species after one of the collectors, Dr. Jarmila Kukalova-Peck (Ottawa), the well known expert in insect paleontology.

Leptoteleia miladae n. sp.
Female. Length 3 mm. Xanthic species; orange-yellow with a few darker spots such as tips of mandibles, little spots behind ocelli in inter-ocellar space, A7 light brown, A8-A12 dark brown, minute spot at meson of prothorax near pro-mesothorax suture, anterior margins of T2-T5, posterior margin of T6,entire T7 and ovipositor sheaths dark brown; wings clear.

Leptoteleia majkae n. s p.
Female. Length 3.5 mm. Varicolored species; head black, mesosoma orange-yellow (with darker spot anteromedially on mesoscutum), metasoma dark brown except for lighter TI and the hump which is almost reddish brown; legs including coxae yellow; scape yellow, clava (A8-A 12) black; wings slightly infumate, basalis and medialis indicated by darker lines.
Metasoma distinctly elongate, fully 6 times as long as wide; hump on T1 deeply housed Into propodeum, in lateral aspect top of hump distinctly higher than top of mesoscutum; reaching the level of attachment of middle coxae; top of hump finely reticulate, rest of T1 with longitudinal costae; T2-T4 with strong longitudinal costae, shining, with very little of sculpture between the costae; costae in posterior half of T4 gradually fading away; anterior margin of T5 with slight costae, the rest reticulate; T6 and T7 finely reticulate; T7 entirely external; tergites in relative proportions 47:35, 44:40, 42:40, 40:38, 32:35, 25:25, 20: 12 REMARKS. This is the darkest species of the xanthic subgroup of the oecanthi-group. It differs from other species remarkably by colour pattern. It is closer to L. miladae and L. jarmilae rather than to L. marcelae and L. verae, the main difference being the shape of metanotal lamina and the hump. It is my pleasure in naming this species after my charming cousin, Mrs. Majka Janousek (nCe Pavlik) of Vancouver.
Mesosoma distinctly wider than high (37:22), generally flattened and perfectly levelled dorsally; prothorax unusually long, its cervical part longer than scutellum (139). reticutatepunctate; humeral part of prothorax also strongly developed; mesoscutum long, more than 4 times as long as the extremely short scutellum, with remarkably conspicuous scaly reticulation; notauli incomplete, indicated as rudiments just below mid-mesoscutum, as long as scutellum: parapsidal carinae present but fine; scutellum short, perfectly flattened, with same sculpture as mesoscutum but somewhat finer; metanotal lamina flattened, subrectangular; propodeurn deeply excavate medially to contain hump, dorsal carinae running diagonally; fore wings not surpassing middle of T4; hind basitarsus longer than following four tarsomeres together.
DISTRIBUTION. Haiti. BIOLOGY. Arndt and Dozier (1931) reared this species from eggs of the coffee cricket, Chremon repentinus Rehn, in Haiti. The cricket is reported injurious to young coffee bushes by ovipositing in twigs, causing defoliation and eventually loss of the plants. However, the cricket is not specific to coffee trees only, ovipositing in about a dozen of other trees and bushes in Haiti. Since the host eggs are deeply tunnelled in the wood, the way the wasp has to reach them is likely very elaborate. The sclerotized tips of ovipositor sheaths present in all Leptoteleia species may perhaps play some role in facilitating this task. Similarly, the pointed T7, not extruded with ovipositor, may very well serve the same purpose.
VARIABILITY. The original description gives much darker colouring than encountered in the type series (slides and dry mounted specimens). This might be explained by the age of the type material.
REMARKS. This is the first detailed redescription of this species. The original diagnosis was quite inadequate, laconic, and, in some respects, also misleading. On the other hand, the author did correctly classify this species in Leptoteleia, a genus not well recognized in the thirties. L. arndti is closely related to the Brazilian L. marketae from which it differs markedly by clear wings, hairy eyes, and complete occipital carina. In fact, the two species are much closer to each other than to L. monicae, which is, to some extent, transitional to oecanthi-group.
Head in dorsal view sub-circular transverse (30:40), in lateral view almost opisthognathous and distinctly flattened, much wider than high (40:25); frontal depression narrow, remarkably deep, closed, sharply margined from all sides, bisected by strong keel running from antenna1 insertion up to upper arc of depression; surface of depression almost smooth and shining; rest of frons, between depression and inner orbits finely reticulate; deep, subcircular, smooth, and highly shining fossa in between upper arc of frontal depression and anterior ocellus; anterior ocellus sloping into the above fossa; ocellar area distinctly reticulate; vertex and temples reticulate-punctate, punctures rather deep; occipital carina not clearly developed although faintly indicated; malar groove absent; eyes glabrous, in dorsal aspect eyes as long as temples behind them, in lateral aspect shorter than temples; temples behind eyes rather bulging, head wider here than across eyes; ocelli in a low triangle, POL distinctly longer than LOL; A9-A11 twice as long as wide.
Mesosoma strongly flattened dorsoventrally, wider than high (40:30); prothorax strongly developed in both humeral and cervical parts; the latter longer than scutellum, slightly-swollen and higher than adjacent part of mesoscutum; sculpture of prothorax rougher than that of vertex, reticulate-punctate; mesoscutum net-like reticulate-punctate, punctation more dense in anterior one-third; notauli absent; parapsidal carinae present; scutellum reticulate, with some fine punctation, one-fourth as long as mesoscutum, longer than metanotal lamina (10:6); metanotal lamina perfectly flattened, slightly convex posteriorly, finely reticulate; propodeum with strong longitudinal costae in its median two-thirds, dorsal carinae running very close to each other, perfectly parallel; mesopleura sculptured, reticulate, metapleura smooth but with some longitudinal striae; fore wings not surpassing apex of T4; legs extremely long and slender; tarsi and hind tibiae with numerous tiny spine-like setae, reminiscent of pompilid wasps, hind basitarsus as long as hind tibia.
Metasoma extremely long and slender, almost 9 times as long as wide, all tergites but T7 distinctly elongate; tergites in relative proportions 35:25, 38:27, 38:26, 36:23, 34:20, 27:18, 10:13, 13:lO; T1-T4 with heavy longitudinal costae and fine aciculation in between; costae becoming irregular on T5 and T6, T7 longitudinally aciculate, T8 reticulate, almost triangular, tapered into a spine which is truncate at very apex. REMARKS. L. marketae is closely related to Caribbean L. arndti from which it differs by having wings heavily infuscate in basal half, eyes glabrous, occipital carina incomplete, etc. Moreover, L. marketae is quite unique for its cephalic characters, e.g. the form of frontal depression, the preocellar fossa, as well as for the shape of T8. The unknown female is expected to be extraordinarily long. Named in honour of Mrs. MarKeta Hak of Teplice (Czechoslovakia), mother of my sister-in-law, Mrs. Alexandra Masner.

Leptoteleia monicae n. sp.
Female. Length 2.6 mm. Xanthic species; orange-yellow; head black, apex of metasoma including T7, T6, and T5 dark brown to black, posterior one-third of T4 brownish, little black speck situated medially at posterior margin of prothorax; antennae dark brown except for orange scape; legs orange-yellow, mid and hind tibiae in distal halves and hind femora dark brown; wings clear.
DISTRIBUTION. Dominican Republic. BIOLOGY. Unknown. VARIABILITY. One female caught at type locality on July 12 is not included in type series. It differs from the other three females in sculpture of TI-T3, however, some malformations all over the body indicate that the specimen is probably monstrous. No chromatic or morphological differences were encountered among the three females from the typical series.
REMARKS. L. monicae belongs to the arndti-group, although it is not so flattened as the other two species, arndti and marketae; however, the length of prothorax and the antennal characters will bring it in this group. It gives me great joy to dedicate this gorgeous species to my sweet daughter, Monica Masner (lo), the little admirer of my tiny wasps.
The AMERICANA-GROUP Leptoteleia americana n. sp.
DISTRIBUTION. Eastern USA from Washington, D.C., to Florida. BIOLOGY. Host unknown but most likely a tree cricket (Oecanthus). The paratypes from Virginia are labelled "with Ips grandicollis on Pinus taeda" and the paratype from Washington is said to be reared from Gossyperia spuria (Modeer) (Hom., Eriococcidae). However, neither of the two insects mentioned could be the potential hosts.
VARIABILITY. Little variability was observed in my material. Punctation of head is denser in some paratypes and epomia less angulate than in the holotype.
REMARKS. Sexual dimorphism in sculpture of head is quite remarkable and suggests that the sexes may belong to different species. The closest relatives of this species are the Caribbean L. ferdinandi and L. lubomiri, both from the Dominican Republic.

Leptoteleia ferdinandi n. sp.
Female. Length 2.6 mm. Differs from L. americana only in the following few characters. Legs including coxae bright golden-yellow. Metasoma in posterior half light brown. Ocellar triangle higher, i.e. POL = LOL. Eyes only sparsely hairy. Occiput less excavate so that the temples behind eyes less bulging. Tergites 4-6 gradually longer, T5 almost as long as wide, T6 distinctly longer than wide (
Mesosoma as wide as high; epomia of prothorax rather angulate (dorsal aspect); mesoscutum finely reticulate-punctate, its mid-section in front of scutellum almost smooth and punctate; parapsidal carinae almost obsolete; scutellum finely reticulate-punctate; sides of prothorax and most of mesopleura reticulate; mesostemum smooth; metanotal lamina deeply excavate medially to appear almost bidentate; fore wings reaching up about middle of T4.
REMARKS. I have some doubts about the association of sexes in this species. The male (allotype) cephalic characters and sculpture of TI-T5 are rather different from female; however, other chcracters plus the locality and timing are exactly as in the holotype. More material, preferably reared, will be needed to solve this question. This species is dedicated to m y late uncle, Mr. Stan[islaus] Suk (formerly of Prague), my beloved companion and instructor in the art of fishing.