.SOCIETY Journal of Hymeno Research Volume 7, Number 1 April 1998 ISSN #1070-9428 CONTENTS ABD-RABOU, S., and M. M. ABOU-SETTA. Parasitism of Siphoninns pbillyreae (Homoptera: Aleyrodidae) by aphelinid parasitoids at different locations in Egypt 57 ENGEL, M. S. A new species of the Baltic amber bee genus Electrapis (Hymenoptera: Api- dae) 94 ENGEL, M. S., and R. W. BROOKS. The nocturnal bee genus Megaloptidia (Hymenoptera: Halictidae) 1 van der ENT, L.-J., and S. R. SHAW. Species richness of Costa Rican Cenocoeliini (Hy- menoptera: Bracondiae): a latidudinal and altitudinal search for anomalous di- versity 15 JOHNSON, N. F., and L. MUSETTI. Geographic variation of sex ratio in Pelecinus polyturator (Drury) (Hymenoptera: Pelecinidae) 48 KIMSEY, L. S., and M. S. WASBAUER. Revision of the American Hphiid genus Quemaya Pate (Hymenoptera: Tiphiidae: Brachycistidinae) 38 KURCZEWSKI, F. E. Territoriality and mating behavior of Sphex pensylvanicus L. (Hyme- noptera: Sphecidae) 74 NEWMAN, T. M., and D. L. J. QUICKE. Sperm development in the imaginal testes of Aleiodes coxalis (Hymenoptera: Braconidae: Rogadinae) 25 RUST, R. W. The effects of cavity diameter and length on the nesting biology of Osmia lignaria propinqua Cresson (Hymenoptera: Megachilidae) 84 SHAW, S. R., P. M. MARSH, and J. C. FORTIER. Revision of North American Aleiodes Wesmael (Part 2): the apicalis (Brulle) species-group in the New World (Hymenop- tera: Braconidae, Rogadinae) 62 (Continued on back cover) INTERNATIONAL SOCIETY OF HYMENOPTERISTS Organized 1982; Incorporated 1991 OFFICERS FOR 1998 James M. Carpenter, President Andrew D. Austin, President-Elect James B. Woolley, Secretary John T. Huber, Treasurer E. Eric Grissell, Editor Paul M. Marsh, Assistant Editor Subject Editors Symphyta and Parasitica Aculeata Biology: Mark Shaw Biology: Sydney Cameron Systematics: Donald Quicke Systematics: Wojciech Pulawski All correspondence concerning Society business should be mailed to the appropriate officer at the following addresses: President, Department of Entomology, American Museum of Natural His- tory, Central Park West at 79th Street, New York, New York 10024; President-Elect, University of Adelaide, Glen Osmond, Australia; Secretary, Department of Entomology, Texas A&M University, College Station, Texas 77843; Treasurer, Eastern Cereal & Oilseed Research Centre, Agriculture Canada, K. W. Neatby Building, Ottawa, Ontario, Canada KIA 0C6; Editor, Systematic Entomology Laboratory, USDA, % National Museum of Nahiral History, NHB 168, Washington, D.C. 20560. Membership. Members shall be persons who have demonstrated interest in the science of ento- mology. Armual dues for members are US$35.00 per year (US$40.00 if paid after 1 May), payable to The International Society of Hymenopterists. Requests for membership should be sent to the Treasurer (address above). Information on membership and other details of the Society may be found on the World Wide Web at http://www.IRIS.biosci.ohio-state.edu/ish/. Journal. The Journal of Hymenoptera Research is published twice a year by the International Society of Hymenopterists, Vc Department of Entomology NHB 168, Smithsonian Institution, Washington, D.C. 20560, U.S.A. Members in good standing receive the journal. Nonmember subscriptions are $60.00 (U.S. currency) per year. The Society does not exchange its publications for those of other societies. Please see inside back cover of this issue for information regarding preparation of manuscripts. Statement of Ownership Title of Publication: Journal of Hymenoptera Research. Frequency of Issue: Twice a year. Location of Office of Publication, Business Office of Publisher and Owner: International Society of Hymenopterists, Vo Department of Entomology, NHB 168, Smithsonian Institution, Wash- ington, D.C. 20560, U.S.A. Editor: E. Eric Grissell, Systematic Entomology Laboratory, USDA, % National Museum of Nat- ural History, NHB-168, Washington, DC 20560. Managing Editor and Known Bondholders or other Security Holders: none. This issue was mailed 14 May 1998 J. HYM. RES. Vol. 7(1), 1998, pp. 1-14 The Nocturnal Bee Genus Megaloptidia (Hymenoptera: Halictidae) Michael S. Engel and Robert W. Brooks (MSE) Department of Entomology, Comstock Hall, Cornell University, Ithaca, NY 14853 USA; (RWB) Division of Entomology, Natural History Museum, Uruversity of Kansas, Lawrence, KS 66045, USA Abstract. — The nocturnal augochlorine bee genus Megaloptidia is revised and three species rec- ognized; Mt'galopttidia contradicta (Cockerell), M. nocturna (Friese), and M. saulensis Engel and Brooks new species. The genus is newly diagnosed, distributional data given, and the male com- pletely described for the first time. The first flower record for a species of this genus is given as Dichorisandra iilei (Commelinaceae). A lectotype and two paralectotypes are designated for M. contradicta. Megalopta angiisticeps Friese is considered a junior synonym of Megaloptidia nocturna (new synonymy). Megalopta pmllitarsus Friese, previously considered to be a species of Megaloptidia, is in fact a junior synonym of Megommalion insigne (Smith) (new synonymy). The neotropical halictine genus Megal- optidia Cockerell consists of three, rarely collected, nocturnal bees of the tribe Au- gochlorini. Individuals of Megaloptidia can be captured at lights during the night, however, aside from their nocturnal habit, the biology of Megaloptidia species remains unknown. One female of Megaloptidia iwc- tunia (Friese 1926), however, has been col- lected at flowers of the monocotyledon Di- chorisandra iilei (Commelinaceae). Mem- bers of Dichorisandra are nectarless and it is therefore believed that pollen is the only reward for floral visitors. Dichorisandra species possess poricidal anthers, suggest- ing "buzzing" as the means of freeing the pollen (i.e., the bee grasps the anther and buzzes its wings, thus translating the vi- bration to the anther and releasing the pollen through the apical pore). Most au- gochlorine species which have been stud- ied are known to be buzz pollinators (e.g., Michener 1962, Rego and Albuquerque 1989, Renner 1989) and, in fact, the related D. hexajidra is recorded to be buzz polli- nated by the diurnal augochlorines Aitgo- chloropsis (Paraugochloropsis) cupreola (Cockerell 1900) and Pseudaugochlora gra- minea (Fabricius 1804) (Sigrist and Sazima 1991). Within the Commelinaceae, Dichor- isandra species are among the only ones to be open during the night (e.g., D. hexandra opens before dawn in southern Brazil) (Sigrist and Sazima 1991), making them suitable targets for nocturnal or crepus- cular insects. Among related bee genera, floral associations are only known for Me- gommation (s. str.) (Moure 1943) which has been recorded at species of Bauhinia (Fa- baceae) (Friese 1926) and Ipomoeae (Con- volvulaceae) (Jorgensen 1912). There is nothing about the floral morphology of any of these plants which seems to readily explain the peculiarly modified mouth- parts of these genera, which are extremely narrowed with a distinctly pointed galeal apex. Megaloptidia was originally described by Cockerell (1900) as a subgenus of the com- mon nocturnal augochlorine genus Megal- opta Smith (1853) owing to its superficial similarity to this group. Cockerell distin- guished his subgenus on the basis of wing venation. Moure (1958) redescribed the group, according it generic status for the first time, and recognized its affinity with Journal of Hymenoftera Research his genus Megommation. Eickwort (1969), in a revision of the genera and subgenera of Augochlorini, provided a more thor- ough description of the genus; however, he did not have the opportunity to ex- amine the male terminalia thus the distal male sterna and genitalia remained un- described. The genus is closely allied to the genera Ariphaiiarthra (Moure 1951), Megonmiation (s. lato), and Micromniation (Moure 1969). All of these genera share the distinctly narrowed labiomaxillary complex, pointed galeal apex, orthogonal epistomal sulcus, absence of the male la- bral distal process, narrow spiculum on the male eighth sternum, and the presence of a parapenial lobe in the male genitalia. The phylogenetic position of Megaloptidia will be further elaborated on in a forth- coming paper concerning the entire tribe (Engel in prep). Herein we provide new descriptions for the genus and all included species. The male is thoroughly described for the first time and a key is presented for the iden- tification of all three species. MATERIAL AND METHODS The following abbreviations are used for institutions where material used in this study is deposited: American Museum of Natural History, New York, New York, J.G. Rozen, Jr. (AMNH); Carnegie Muse- um of Natural History, Pittsburgh, Penn- sylvania, R.L. Davidson (CMNH); Field Museum of Natural History, Chicago, Il- linois, P.P. Parrillo (FMNH); Cornell Uni- versity Insect Collection, Ithaca, New York, J.K. Liebherr and E.R. Hoebeke (CUIC); Natural History Museum of Los Angeles County, Los Angeles, California, R.R. Snelling (LACM); Instituto de Inves- tigacion de Recursos Biologicos, Alexan- der von Humboldt, Santa Fe de Bogota, Colombia, F. Fernandez (UNCB); Instituto Nacional de Pesquisas da Amazonia, Ma- naus, Amazonas, Brazil (INPA); Museo de Invertebrados 'G.B. Fairchild', Universi- dad de Panama, Estafeta Universitaria, Panama City, Panama', D. Quintero (MIUP); Museu Entomologico de Univer- sidade Federal de Vi^osa, Vigosa, Minas Gerais, Brazil (UFVB); Museum fiir Natur- kunde, Humboldt-Universitat, Berlin, Ger- many, F. Koch, A. Kleine-MoUhoff (ZMHB); United States National Museum, Smithsonian Institution, Department of Entomology, Washington, D.C., R.J. McGinley (USNM); Division of Entomol- ogy, Natural History Museum, University of Kansas, Lawrence, Kansas, R.W. Brooks (SEMC); G.A.R. Melo collection (GARM); Philadelphia Academy of Natural Sci- ences, Philadelphia, Pennsylvania, D. Azuma (ANSP). A total of 67 specimens of Megaloptidia were located and examined in the course of this study. All measurements were made using an ocular micrometer on a WILD-M5a microscope. The abbreviation "o.d." in the descriptions refers to "ocellar diameter" and relates the approximate length of setae to the diameter of the me- dian ocellus. SI, Tl, and Fl are given as abbreviations for first metasomal sternum, first metasomal tergum, and first flagello- mere respectively. Genus Megaloptidia Cockerell Megalopta (Megaloptidia) Cockerell 1900: 373. Type species: Megalopta {Megaloptidia) coii- tradicta Cockerell 1900, by monotypy and original designation. Moure 1958: 180. Eick- wort 1969: 442. Diagnosis. — Individuals of Megaloptidia are robust bees which superficially resem- ble in general habitus species of the larger nocturnal genus Megalopta and the sole species of Megommation (s. str.). Megalop- tidia differs from the former genus in the narrowed labiomaxillary complex, the ser- rate inner hind tibial spur, the irregular spacing of the distal hamuli, the narrow gena, the unmodified labral distal keel, the orthogonal epistomal sulcus, and the ab- sence of lateral notches on the male S4. Megommation differs from Megaloptidia in the acute marginal cell apex, the deeply Volume 7, Number 1, 1998 concave clypeal apex, the extremely short and weakly border basitibial plate, and the medioapical processes on the male S3-4. Description. — The following description is based on all three included species. FE- MALE: Epistomal sulcus orthogonal. Clypeus and supraclypeal area strongly bowed, as in Megommation (Fig. 2). Malar space short. Mid-region of face gently sunken around antermal socket. Preoccip- ital ridge rounded. Inner orbit of com- pound eye strongly emarginate; eyes greatly enlarged, much broader than gena in profile (Figs. 2, 8, 14, 19); eye hairs short. Ocelli greatly enlarged (Figs. 1, 7, 13, 18), without impressed line running between lateral ocelli. Vertex extremely short, barely an ocellar diameter in length, usually less. Labral basal elevation orbic- ular, protuberant in profile, distal process narrowly triangular, labral teeth absent (Fig. 1). Mandible broad, subapical tooth well defined (Fig. 1). Hypostomal ridge carinate, not projecting beyond posterior margin of head, anterior angle rounded. Mouthparts greatly narrowed, approxi- mately 10 times longer than wide (Fig. 4). Galeal base reaching to base of stipes, apex pointed, inner strip with setae; galeal comb absent; maxillary palp not greatly lengthened. V-shaped brace of salivary plate absent; combined length of labial palp segments 2 and 3 shorter than 1. Pronotal lateral angle obtuse, not pro- duced; lateral and dorsal ridges rounded. Mesoscutum broadly rounded anteriorly; mesoscutal lip low and rounded. Tegula rounded. Propodeal dorsal ridge rounded; lateral ridge carinate, ridges slightly di- vergent; propodeum slightly narrowed posteriorly; pit of posterior face narrow. Wing hairs long; apex of marginal cell truncate (Fig. 3). Anterior basitarsal brush present. Inner hind tibial spur serrate, ser- rations sharp (Fig. 5). Basitibial plate nar- rowly rounded, all borders well defined. Division of T5 narrow. MALE: As for the female with the following modifications. Antenna of moderate length, reaching to posterior border of mesoscutum; scape long; F2 shorter than Fl; sensory plate ar- eas present. Labral basal elevation absent; distal process absent. Gradulus of T7 ab- sent. S4 with depression along posterior margin. S8 with posterior margin pro- duced into a short, blunt median projec- tion; spiculum narrow. Gonobasal bridge narrow; dorsal lobes strong. Dorsal gon- ostylus a long thin process. Basal process of gonostylus absent. Parapenial lobe present. Ventral prong of penis valve pres- ent. KEY TO SPECIES OF MEGALOPTIDIA 1. Propodeal lateral surface strongly punctured, punctures almost contiguous; length of pro- podeal triangle approximately equal to that of metanotum; mesepisternum strongly and closely punctured, punctures separated by less than a puncture width, integument between smooth (excluding hypoepimeron); mesoscutum punctured contiguously, integument be- tween imbricate M. contradicta (Cockerel!) - Propodeal lateral surface imbricate with punctures separated by 2-3 puncture widths; length of propodeal triangle little to much greater than that of metanotum; sculpturing of mesepisternum and mesoscutum variable, but mesoscutum never strongly punctured .... 2 2. Labrum and clypeus amber; mesepisternum minutely punctured, punctures separated by 3^ times puncture width, at least on posterior half, integument between smooth or im- bricate (excluding hypoepimeron); mesoscutum minutely punctured, punctures separated by 2-3 times puncture width; lateral ocellus almost touching compound eye, ocellocular distance 0.25 o.d. in female (Fig. 1), 0.33 o.d. in male (Fig. 7); frontal line ver>' weakly produced in profile in female (Fig. 2); median ocellus wider than interantennal distance in female (Fig. 1); scopal hairs of hind femur and tibia amber M. noctuma (Friese) Journal of Hymenoptera Research Figs. 1-8. Me^nlopthiin nocturna (Friese), 1-6 female, 7-8 male. 1, 2, head, front and side views respectively. 3, fore wing. 4, mouthparts excluding labium, ventral view, mp = maxillary palps, c = cardo, m = maxilla, hf = hypostomal fossa. 5, inner hind tibial spur. 6, hind leg and metasoma, side view, circle enlargement shows scopal hairs on S2 and S3. 7, 8, head, front and side views, respectively. Volume 7, Number 1, 1998 5 - Labrum and clypeus dark brown; mesepisternum closely and weakly punctured, punctures separated by less than a puncture width (excluding hypoepimeron); mesoscutum punc- tured, punctures weak and separated by 2-3 puncture widths centrally, laterally becoming separated by less than a puncture width; lateral ocellus removed from compound eye 0.5 o.d. (Fig. 18); frontal line protuberant in profile (Fig. 19); median ocellus narrower than interantennal distance (Fig. 18); scopal hairs of hind tibia black (those of femur amber) . . M. saulensis n. sp. Megaloptidia contradicta (Cockerell) (Figs. 13-17, 20) Megalopta {Megaloptidia) contradicta Cockerell 1900: 373. Schrottky 1902: 407. Friese 1926: 124. Megaloptidia contradicta (Cockerell); Moure 1958: 180. Diagnosis. — Labrum and apical % of clypeus amber. Mesoscutum closely and coarsely punctured, punctures separated by less than a puncture width, except over central disc punctures becoming v^^eaker and separated by 1-3 times a puncture width. Mesepisternum closely punctured, punctures separated by less than a punc- ture width, integument between punc- tures smooth; hypoepimeron smooth with a few widely spaced minute punctures. Scopal hairs of hind leg fuscous. Propo- deal lateral surface strongly punctured, punctures nearly contiguous; length of propodeal triangle equal to that of meta- notum. Only three specimens of this spe- cies were located, all from Cockerell' s type series. Description. — The following description is based on the lectotype. MALE: Total body length 11.3 mm; forewing length 8.6 mm. Head wider than long (Fig. 13). Dis- tal margin of clypeus barely projecting be- low lower tangent of compound eye (Fig. 14); supraclypeal area wider than long, 0.59 times length of clypeus. Frontal line carinate from just below antennal sockets to just above sockets, ending at this point. Measurements of head of lectotype in mm: width 3.0; length (to apex of clypeus) 2.5; clypeal length 0.8; lower interorbital dis- tance 0.9; upper interorbital distance 1.2; interantennal distance 0.4; antennocellar distance (to median ocellus) 0.6; between lateral ocelli 0.4; median ocellus to lateral ocellus 0.1; lateral ocellus to compound eye 0.16; prementum length 1.9, width 0.18. Scape reaching past lateral ocellus, length 1.1; pedicel slightly longer than wide, length 0.2, width 0.18; Fl longer than wide and longer than F2, length 0.26, width 0.2; F2 as long as wide, length- width 0.22; F3-9 each longer than wide, individual lengths 0.24, widths 0.22; FIO longer than wide and longer than imme- diately preceding flagellomeres, length 0.26, width 0.22; Fll longer than wide and longer than all previous flagellomeres, length 0.42, width 0.22. Median line strongly impressed; parapsidal lines weak. Intertegular distance 2.2. Propodeal triangle about as long as metanotum, much shorter than scutellum. Basal vein distad cu-a by 2.5 times vein width; Ir-m basad Im-cu by vein width; 2r-m distad 2m-cu by 4.5 times vein width, 2r-m weakly curved. First submarginal cell lon- ger than second and third combined; sec- ond narrowed anteriorly; anterior border of third along Rs almost 2 times as long as anterior border of second, posterior border of third 2 times longer than ante- rior border; marginal cell length 1.2, width 0.3. Distal hamuli arranged 3-1-1-3. S5 api- cally emarginate and bent ventrally (Fig. 15). S6 apically emarginate and mediolon- gitudinally concave (Fig. 15). Male termin- alia as in figures 16 and 17. Clypeus and supraclypeal area finely imbricate, with a few weak punctures sep- arated by 1-3 puncture widths on lateral borders. Face minutely and closely punc- Journal of Hymenoptera Research Figs. 9-17. Megaloptidia noctiirnn (Friese), 9-12 male; Megaloplidia coiitrndicla (Cockerell), 13-17 male. 9, me- tasoma, apical half, oblique view, numbers refer to sterna. 10, S7 and S8, dorsal view is right half and ventral view is left. 11, genital capsule, dorsal view is right half and ventral view is left, dl = dorsal gonostylar process, pi = parapenial lobe, vp = ventral prong of penis valve. 12, S6, ventral view. 13, 14, head, front and side views, respectively. 16, genital capsule, dorsal view is right half and ventral view is left. 17, S7 and S8, ventral view is right half and dorsal view is left. Volume 7, Number 1, 1998 tured, integument between imbricate. Ver- tex smooth and impunctate. Gena and postgena impunctate and finely imbricate. Pronotum finely imbricate. Mesoscutum closely and coarsely punctured, punctures separated by a puncture width or less, in- tegument between punctures imbricate, medially punctures becoming weaker and more widely spaced, separated by 1-3 punctured widths. Scutellum with minute punctures separated by 1-2 puncture widths, integument between smooth. Metanotum sparsely and weakly nodu- late, integument otherwise smooth. Pre- episternum and mesepisternum closely punctured, punctures separated by a puncture width or less, integument be- tween smooth. Hypoepimeron smooth, with a few minute punctures separated by 4-5 puncture widths. Metepisternum smooth. Propodeal triangle imbricate; lat- eral surface closely punctured, punctures separated by less than a puncture width; posterior surface smooth. Terga and ster- na imbricate. Head dark brown with a few weak blue-green highlights; labrum and apical three-quarters of clypeus amber; mandible dark brown; antennae brown. Mesosoma and metasoma dark brown; tegula amber and translucent along outer margins; legs dark brown, except amber on protarsus, apical four mesotarsomeres, and apical 2 metatarsomeres. Pubescence generally pale or fuscous, fuscous hairs mostly on mesosoma and apical terga of metasoma. Scattered simple hairs (1 o.d.) on face, with suberect branched hairs (0.5 o.d.) along inner mar- gins of compound eyes. Similar suberect and simple hairs on gena. Postgena with a few simple hairs (1.5-2 o.d.) on lateral and posterior borders. Hairs of face, gena, and postgena pale, those of vertex becom- ing fuscous. Pronotum with scattered sim- ple hairs (1 o.d.) and laterally with ex- tremely short, appressed branched hairs not obscuring the surface. Mesoscutum with scattered simple hairs (1 o.d.) and shorter (less than 0.5 o.d.) branched hairs on lateral and posterior borders. Scutel- lum and metanotum with sparse simple hairs (1.5 o.d.). Pleura with scattered sim- ple pale hairs (1-1.5 o.d.) and shorter (0.5 o.d. or less) suberect branched hairs. Hairs of forelegs pale or golden, outer margins with simple hairs (1-1.5 o.d.); mid legs with pale hairs except fuscous on basitar- sus and outer margin of tibia, hairs sim- ple, small tuft of closely packed hairs (1 o.d.) on inner basal margin of femur; hair of hind legs mostly fuscous, inner surface of basitarsus with stiff black hairs (1-1.5 o.d.). Terga with scattered simple hair (1- 1.5 o.d.), becoming longer on lateral mar- gins and more distal terga. Sterna with sparse simple hair (1-2.5 o.d.); hairs of S4 longest and concentrated on borders of apical depression and medially towards basal border; S5 with pair of apical sub- median hair tufts (Fig. 15); hairs of 56 most numerous, more dense laterally (Fig. 15). FEMALE: Unknown. Type material.— BRAZIL: Para: Lecto- type #345, male, Benevides [1°22'S, 48°15'W], July (CMNH). Paralectotype, male, Santarem (CMNH). Paralectotype, male, Santarem, but lacking all metasomal segments except Tl and SI and both an- tennae leaving only the scape, pedicel, Fl and F2 (ANSP). Remarks. — In 1957 Padre Jesus S. Moure examined the material listed above and placed lectotype labels on the specimens, however, he failed to publish lectotype designations for this species. We now des- ignate lectotype and paralectotypes for Megaloptidia contradicta using the same specimens. The lectotype specimen now bears a new label reading "LECTOTYPE: Megalopta (Megaloptidia) contradicta Cock- erel!, 1900; desig. M. Engel & R. Brooks", and the paralectotypes now have similar labels. In Eickwort's (1969) study of the Au- gochlorini he referred to a male of M. con- tradicta (the paralectotype from ANSP des- Journal of Hymenoptera Research ignated above) along with a series of fe- males he considered as possibly belonging to M. contradicta. This series of females from the AMNH are, in actuality, all of M. nocturna. Megaloptidia nocturna (Friese) (Figs. 1-12, 20) Megalopta nocturnn Friese 1926: 127. Megalopta angiisticeps Friese 1926: 127. New synonymy. Megaloptidia iwctuma (Friese); Moure and Hurd 1987: 242. Megaloptidia angiisticep's (Friese); Moure and Hurd 1987: 241. Diagnosis. — Labrum and clypeus amber. Mesoscutum minutely punctured, punc- tures separated by 2-3 times a puncture width. Mesepisternum minutely punc- tured, otherwise integument smooth or imbricate; hypoepimeron as on mesepi- sternum although punctures more widely spaced. Scopal hairs of hind femur and tibia amber. Propodeal lateral surface im- bricate with punctures separated by 2-3 times a puncture width; length of propo- deal triangle little to much greater than that of metanotum. This is the most com- mon of the three species with 62 speci- mens examined during the course of this study. Description. — The following description is based on Friese's holotypes. MALE: To- tal body length 13.3 mm; forewing length 8.3 mm. Head wider than long (Fig. 7). Distal margin of clypeus barely projecting below lower tangent of compound eye (Fig. 8); supraclypeal area wider than long, 0.68 times length of clypeus. Frontal line carinate from just below antennal sockets to just above sockets, ending with acute projection (Fig. 8). Measurements of head of holotype in mm: width 2.8; length (to apex of clypeus) 2.5; clypeal length 0.8; lower interorbital distance 0.8; upper in- terorbital distance 1.1; interantennal dis- tance 0.3; antennocellar distance (to me- dian ocellus) 0.5; between lateral ocelli 0.3; median ocellus to lateral ocellus 0.04; lateral ocellus to compound eye 0.08; pre- mentum length 2.1, width 0.2. Scape reaching past lateral ocellus, length 1.0; pedicel slightly longer than wide, length 0.2, width 0.18; Fl longer than wide and longer than F2, length 0.24, width 0.18 (Fig. 7); F2 as long as wide, length-width 0.2; F3 and F4 each longer than wide, in- dividual lengths 0.22, widths 0.2; F5 lon- ger than wide, length 0.24, width 0.2; F6 longer than wide, length 0.26, width 0.2; F7 and F8 longer than wide, individual lengths 0.28, widths 0.2; F9 and FIO longer than wide, individual lengths 0.3, widths 0.2; Fll longer than wide and longer than preceding flagellomeres, length 0.44, width 0.2. Median line strongly im- pressed; parapsidal lines weak. Intertegu- lar distance 1.8. Propodeal triangle longer than metanotum, slightly shorter than scu- tellum. Basal vein distad cu-a by vein width; Ir-m distad Im-cu by 5 times vein width, intersecting second submarginal cell about half way through cell; 2r-m dis- tad 2m-cu by 2 times vein width, 2r-m weakly curved. First submarginal cell lon- ger than second and third combined; sec- ond narrowed anteriorly; anterior border of third along Rs about as long as anterior border of second, posterior border of third 2 times longer than anterior border; mar- ginal cell length 2.5, width 0.5. Distal ham- uli arranged 3-1-3. Male terminalia as in figures 10 and 11. Clypeus and supraclypeal area finely imbricate, with a few weak punctures sep- arated by 1-2 puncture widths on lateral borders. Face imbricate. Vertex smooth and impunctate. Gena sparsely and weak- ly nodulate, otherwise smooth. Postgena smooth. Pronotum smooth. Mesoscutum minutely punctured, punctures separated by 2-3 times puncture width, integument between punctures smooth and shining. Scutellum as on mesoscutum. Metanotum sparsely and weakly nodulate, integument otherwise smooth. Pre-episternum and mesepisternum minutely punctured, punctures separated by 3-4 times punc- Volume 7, Number 1, 1998 ture width, integument between smooth; hypoepimeron as on mesepisternum ex- cept minute punctures separated by 4 times puncture width. Metepisternum smooth. Propodeal triangle imbricate; lat- eral surface imbricate, with punctures sep- arated by 2-3 puncture widths; posterior surface smooth. Terga and sterna imbri- cate, except Tl which is smooth. Head dark brown with strong metallic green highlights; labrum and clypeus am- ber; mandible amber except red at apex; antennae brown. Mesosoma dark brown with strong metallic green highlights, ex- cept pronotum light brown and without such highlights. Legs pale brown. Meta- soma dark brown. Pubescence golden. Scattered simple hairs (1-1.5 o.d.) on face, with suberect branched hairs (0.5 o.d.) along inner mar- gins of compound eyes. Similar suberect and simple hairs on gena, branched hairs becoming longer towards postgena. Post- gena with scattered branched hairs (2 o.d.). Pronotum with scattered simple hairs (1 o.d.) and laterally with extremely short, appressed hairs. Mesoscutum with scattered simple hairs (1 o.d.) and shorter (less than 0.5 o.d.) branched hairs on lat- eral and posterior borders. Scutellum and metanotum with sparse simple hairs (1.5- 2 o.d.) and more numerous shorter simple hairs (0.5 o.d.). Pleura with scattered sim- ple pale hairs (1-1.5 o.d.) and shorter (0.5 o.d. or less) suberect branched hairs; me- tepisternum without longer simple hairs. Pubescence of legs simple (1-2 o.d.) Terga with scattered simple hair (1 o.d.), becom- ing longer on lateral margins and apical terga. Sterna with sparse simple hair (0.5- 1.5 o.d.); S4 with a few simple hairs (0.5 o.d.) concentrated on borders of apical de- pression (Fig. 9); hairs of S6 most numer- ous, laterally more dense than median area. FEMALE: As for the male with the modifications indicated below. Measure- ments from female type of M. angusticeps. Total body length 12.2 mm; forewing length 8.7 mm. Head as in figures 1 and 2. Supraclypeal area 0.8 times length of clypeus. Head width 3.0; length (to apex of clypeus) 2.7; clypeal length 0.8; lower interorbital distance 0.9; upper interorbital distance 1.1; interantennal distance 0.3; an- tennocellar distance (to median ocellus) 0.6; between lateral ocelli 0.3; median ocel- lus to lateral ocellus 0.04; lateral ocellus to compound eye 0.08; prementum length 2.2, width 0.2. Scape reaching past lateral ocellus, length 1.4; pedicel slightly longer than wide, length 0.2, width 0.18; Fl lon- ger than wide and longer than F2, length 0.24, width 0.22; F2 wider than long, length 0.2, width 0.22; F3 and F4 each as long as wide, individual lengths-widths 0.22; F5 and F6 longer than wide, lengths 0.24, widths 0.22; F7 and F8 longer than wide, lengths 0.26, widths 0.22; F9 longer than wide, length 0.3, width 0.22; FIO lon- ger than wide and longer than previous flagellomeres, length 0.44, width 0.22. In- tertegular distance 2.1. Wing as in figure 3; basal vein distad cu-a by vein width; Ir- m confluent with Im-cu; 2r-m distad 2m- cu by 5 times vein width, 2r-m weakly curved. First submarginal cell longer than second and third combined; second nar- rowed anteriorly; anterior border of third along Rs slightly longer than anterior bor- der of second, posterior border of third 2 times longer than anterior border; margin- al cell length 2.6, width 0.6. Distal hamuli arranged 3-1-3. Mesoscutum minutely punctured, punctures separated by 3^ times punc- ture width, integument between punc- tures weakly imbricate and shining. Terga and sterna weakly imbricate, except Tl which is smooth. Head dark brown with strong metallic green highlights; labrum and clypeus am- ber; supraclypeal area light brown medi- ally; mandible amber except black at apex. Mesoscutum, scutellum, metanotum and propodeum brown with strong metallic green highlights; pronotum and pleura amber with metallic green highlights 10 Journal of Hymenoptera Research fainter than those of mesoscutum. Legs amber. Terga amber, except apical mar- gins brown; sterna amber. Pubescence golden. Scattered simple hairs (1-1.5 o.d.) on face. Gena and post- gena with branched hairs (1.5-2 o.d.). Pu- bescence of fore- and mid-legs as in male; hind legs with scopa formed of plumose hairs (2.5-3.5 o.d.) on trochanter, femur and imier margin of tibia. Terga with scat- tered simple hair (1-1.5 o.d.), becoming longer on lateral margins and apical terga. Sterna with sparse simple hair (1-1.5 o.d.). Type material. — BRAZIL: Amazonas: Holotype, male. Faro, 23 January 1910, Ducke (ZMHB). Additional material. — BRAZIL: Amazo- nas: 1 male, Parintins, 9 October 1919, Par- ish (FMNH). 1 male, Beruri, Rio Purus, 15 October 1991, G.A.R. Melo, sitting under leaf (GARM). 6 females, Hwy. BR 174, 86 km N Manaus, ZF6 km 9, 2°16'S, 59°56'W, 3 July 1986, M.V.B. Garcia, attracted to propane lantern at night (UFVB); 5 fe- males, same except 4 July 1986 (UFVB); 2 females, same except (SEMC). 3 females, Reserva Ducke, 26 km NE Manaus, Itacoa- tiara Hwy., 12-23 May 1972, Munroe (SEMC). 2 males, Autaz-Mirim, Faz. Sao Lucas, 25 August 1994, Joao Vidal, malaise trap (INPA); 2 males, same except (SEMC); 1 male, same except 28 October 1994 (INPA). 1 female. Campus Universi- tario, Manaus, 7-24 September 1988, M. Castilho, J. Elias Bindo, Shannon trap, baited with feces (INPA). 2 females, Mun. Itacoatiara, Faz. Aruana, AMOIO km 215, 19-21 September 1990, C Motta, R. An- dreazze, R. Ferreira (INPA); 2 females, same except 18-19 September 1990, light trap (INPA). 1 female, Reserva Ducke, 9- 22 September 1994, J. Rafael, J. Vidal, sus- pended malaise trap, 20 m (INPA). 1 fe- male, Manaus, Campus Universitario, 21 June 1982, J. Rafael, malaise trap (INPA). 1 female, 26 km NE Manaus, Reserva Ducke, 6 October 1988, J. Raphael, sus- pended malaise trap, 10 m (INPA); 1 fe- male, same except 9 May 1978, light trap. J. Arias, N. Penny (INPA). 1 female, Rio Jau, Meriti, Mun. Novo Airao, 4-10 June 1994, J. Raphael (INPA). Para: 4, females, Val de Cans. Belem., 20-21 November 1963, Oliveira, Wygodzinsky, at light (AMNH). 1 male. Baker (LACM). CO- LOMBIA: Dpto. Guaviare: 1 female, 1 male, Nukak Maku, Cerro Moyano, 200 m, 12 November 1995, malaise trap, F. Fernandez (UNCB). ECUADOR: Sucum- bios: 1 female, 0.5°S, 76.5°W, 290 m, Sacha Lodge, 13-23 June 1994, P. Hibbs, malaise trap (LACM). FRENCH GUIANA: 1 fe- male (holotype of Megaloptidia angusti- cep^s), Nouveau Chantier, Collection le Moult, June (ZMHB). 1 female, Saint-Lau- rent, du Maroni (AMNH). 1 female, Saiil, 3°37'N, 53°12'W, 13-18 April 1983, S. Mori, ex: light trap (AMNH). 2 males, Roura, 18.4 km SSE, 240 m., 4°36'38"N, 52°13'25"W, 25-29 May 1997, J.S. Ashe and R.W. Brooks, FG1AB97-081, ex: flight in- tercept trap (SEMC). 1 male, Roura, 27.4 km SSE, 280 m., 4°44'20"N, 52°13'25"W, 23-24 May 1997, J.S. Ashe and R.W. Brooks, FG1AB97-022, ex: flight intercept trap (SEMC). 1 male, Roura, 27.4 km SSE, 280 m., 4°44'20"N, 52°13'25"W, 10 June 1997, J.S. Ashe and R.W. Brooks, FG1AB97-177, ex: flight intercept trap (SEMC). 1 male, Matoury, 41.5 km SSW on Hwy N2, 50 m., 4°37'22"N, 52°22'35"W, 29 May-9 June 1997, J.S. Ashe and R.W. Brooks, FG1AB97-170, ex: flight intercept trap (SEMC). GUYANA: Mazaruni-Potaro District: 1 female, Kartabo Point, 25-27 December 1983, W. Steiner, J. Byrd, J. Hill, F. Holtzclaw, malaise trap at edge of sec- ondary forest and farmed field (USNM). PERU: Cuzco: 1 male. Quince Mil, 750 m, September 1962, L. Pefia (CUIC). Huanu- co: 1 female, Tingo Maria, 21 February 1972, J. Schuster, on flower of Dichorisnii- dra iilei, 5:53am (CUIC). Madre de Dios: 1 male, Estacion Pakitza, Reserva Manu, 18 June-4 July 1993, R. Cambra, T. Amorilla (MIUP). 1 female, Rio Tambopata Res., 30 air km SW Pto. Maldonaldo, 290 m, 16-20 November 1979, J. Heppner, subtropical Volume 7, Number 1, 1998 11 19 Figs. 18-19. Mcgnloi'tkiia saulensis Engel c& Brooks n. sp., 18, 19, female head, front and side views respec- tively. moist forest (USNM). SURINAME: 3 fe- males, Raleigh Vallen-Voltzberg Research Foengoe 4°43'N, 56°12'W, 26 January-15 February 1982, J. Carpenter, D. Trail (CUIC). VENEZUELA: Amazonas: 2 fe- males, San Carlos de Rio Negro, 1°56'N, 67°3'W, 6-12 December 1984, R. Brown (CUIC). 3 females. Camp IV, 0°58'N, 65°57'W, Cerro de la Neblina, 760 m, IS- IS March 1984, O.S. Flint, Jr., malaise trap over dry stream channel (USNM). Flower records. — One female of M. noc- tunia has been collected on Dichorisandra iilei (Monocotyledonae: Commelinaceae), although there was no pollen in her scopa. This is the first floral association recorded for any species of Megaloptidia. Variation. — As typical of nocturnal bees, Megaloptidia nocturna is lightly melanized throughout most of its Amazonian range. However, at the western edge of its dis- tribution the male specimen from Colom- bia is light yellow brown and noticeably lighter than the brown Brazilian male specimens. This is odd since a female caught in the same trap on the same day is darker than the normal color of the Bra- zilian female specimens. The male speci- men from Peru, Madre de Dios is almost black and much darker than the Brazilian material. Megaloptidia saulensis Engel and Brooks, new species (Figs. 18-20) Diagnosis. — Labrum and clypeus dark brown. Mesoscutum weakly punctured, punctures separated by 2-3 times a punc- ture width over central disk, laterally punctures separated by less than a punc- ture width. Mesepisternum closely and weakly punctured, punctures separated by less than a puncture width; hypoepi- meron with minute punctures separated by 2-3 times a puncture width. Scopal hairs of hind tibia black, those of femur amber. Propodeal lateral surface imbricate with punctures separated by 2-3 times a puncture width; length of propodeal tri- angle little to much greater than that of metanotum. This species is only known on the basis of two females. Description. — The following description is based on the holotype. FEMALE: Total body length 12.4 mm; forewing length 8.8 mm. Head wider than long (Fig. 18). Dis- tal third of clypeus projecting below lower tangent of compound eye (Figs. 18, 19); 12 Journal of Hymenoptera Research supraclypeal area wider than long, 0.58 times length of clypeus. Frontal line cari- nate from below antennal sockets to just above sockets, ending with an acute pro- jection and protuberant in profile (Fig. 19). Measurements of head of holotype in mm: width 3.4; length (to apex of clypeus) 2.8; clypeal length 1.0; lower interorbital dis- tance 1.1; upper interorbital distance 1.4; interantennal distance 0.4; antennocellar distance (to median ocellus) 0.6; between lateral ocelli 0.4; median ocellus to lateral ocellus 0.08; lateral ocellus to compound eye 0.2 mm; prementum length 2.0, width 0.2. Scape reaching past lateral ocellus, length 1.5; pedicel slightly longer than wide, length 0.24, width 0.18; Fl longer than wide and longer than F2, length 0.26, width 0.22 (Fig. 18); F2 as long as wide, length-width 0.2; F3-6 each longer than wide, individual lengths 0.22, widths 0.2; F7 and F8 each longer than wide, individ- ual lengths 0.24, widths 0.22; F9 longer than wide, individual lengths 0.26, widths 0.24; FIO longer than wide and longer than preceding flagellomeres, length 0.4, width 0.24. Median line strongly impressed; par- apsidal lines weak. Intertegular distance 2.5. Propodeal triangle slightly longer than metanotum, about half length of scutel- lum. Basal vein distad of cu-a by vein width; Ir-m basad Im-cu by vein width; 2r-m distad 2m-cu by 3 times vein width, 2r-m straight. First submarginal cell lon- ger than second and third combined; sec- ond narrowed anteriorly; anterior border of third along Rs as long as anterior bor- der of second, posterior border of third 2.5 times longer than anterior border; margin- al cell length 2.4, width 0.6. Distal hamuli arranged 4-1-1-1-4. Clypeus and supraclypeal area finely imbricate, with a few weak punctures sep- arated by 1-2 puncture widths on lateral borders. Face imbricate with a few weak punctures along border with compound eye. Vertex, gena, and postgena smooth and impunctate. Pronotum weakly imbri- cate. Mesoscutum punctured, punctures separated by 2-3 times puncture width, integument between punctures imbricate and shining, except laterally punctures separated by less than a puncture width. Scutellum with sparse, minute punctures, integument smooth. Metanotum sparsely and weakly nodulate, integument other- wise weakly imbricate. Pre-episternum and mesepisternum punctured, punctures separated by less than a puncture width, integument between smooth; hypoepime- ron as on mesepisternum except minute punctures separated by 2-3 times punc- ture width. Metepisternum smooth with a few minute punctures sparsely scattered. Propodeal triangle imbricate; lateral sur- face imbricate with punctures separated by 2-3 puncture widths; posterior surface smooth. Terga and sterna imbricate. Head dark brown with metallic blue- green highlights, except labrum, mandi- ble, clypeus, supraclypeal area without such highlights. Mesosoma dark brown without highlights, except scutellum, metanotum, and propodeal triangle am- ber. Legs brown. Metasoma dark brown. Pubescence pale to golden. Scattered simple hairs (1-1.5 o.d.) on face, with sub- erect branched hairs (0.5 o.d.) along inner margins of compound eyes. Similar sim- ple hairs on gena, hairs becoming longer towards postgena; postgena with scat- tered simple hairs (2 o.d.). Pronotum with scattered simple hairs (1 o.d.) and laterally with extremely short, appressed hairs. Mesoscutum with scattered simple hairs (1 o.d.) and shorter (less than 0.5 o.d.) branched hairs on lateral and posterior borders. Scutellum with sparse simple hairs (1.5-2 o.d.) and more numerous shorter simple hairs (0.5 o.d.). Longer hairs of mesoscutum and scutellum fus- cous. Metanotum with sparse simple hairs (1.5-2 o.d.). Pleura with scattered simple hairs (1-1.5 o.d.) and shorter (0.5 o.d. or less) suberect branched hairs; metepister- num with longer simple hairs more dense on ventral half; pubescence of pre-epister- num fuscous, remainder pale. Pubescence Volume 7, Number 1, 1998 13 Fig. 20. Distribution of Mcgnloptidia species. Circles represent locality records for M. iioctunuj (Friese), squares for M. contradictn (Cockerell), and the single triangle for M. saulensis Engel & Brooks. of legs simple and amber (1-2 o.d.), except tarsomeres 1 and 2 (of all legs), apex of protibia, and entirety of mesotibia and metatibia, including tibial scopa. Terga with scattered simple hairs (1 o.d.), be- coming progressively longer on lateral margins and more distal terga. Sterna with sparse simple hairs (1.5-2.5 o.d.), most with a few branches. MALE: Unknown. Type material.— FRENCH GUIANA: Ho- lotype, female, Saiil, 3°37'N, 53°12'W, 1983, S. Mori, #158 (AMNH). 1 paratype female, Saul, 3°37'N, 53°12'W, 1983, S. Mori, #157 (AMNH). Etymology. — The specific epithet is de- rived from the type locality of the species. SPECIES MISTAKENLY PLACED IN MEGALOPTIDIA Moure and Hurd (1987), in a catalog of the Western Hemisphere halictid bee spe- cies, tentatively placed Megalopta pallitar- siis Friese as a species of Megaloptidia. In actuality, examination of the holotype male for M. pallitarsus (ZMHB) reveals that this species is a junior synonym of Megommatioii insigne (Smith 1853). We therefore propose the following new syn- onymy: Megommation insigne (Smith 1853) Halictus hisigiiif Smith 1853: 65. Halictus cherazon Vachal 1904: 113. Megalopta I'irgili Friese 1911: 454. Megalopta {Megaloptella) iponweae Schrottky 1912: 271. Megalopta pallitarsus Friese 1926: 125. New syn- onymy. Megaloptidia pallitarsus (Friese); Moure and Hurd 1987: 242. ACKNOWLEDGMENTS We gratefully acknowledge the curators listed above for the sundry loans of material which made 14 Journal of Hymenoptera Research this study possible. We are thankful to D.B. Baker, E.E. Grissell, R.J. McGinley, and B.B. Norden for crit- ically reviewing previous versions of this paper. Their valuable criticisms served to improve the pre- sentation of this study. C.R. Hardy graciously shared with us his knowledge of Commelinaceae biology, morphology, and systematics. MSE was supported by a National Science Foundation Predoctoral Fellow- ship during this work. The generic description pre- sented herein is taken from a monograph of the gen- era and subgenera of Augochlorini being completed by MSE as part of a Ph.D. dissertation at Cornell Uni- versity. This paper is contribution Nr. 3186, Division of Entomology, Snow Entomological Collection, Uni- versity of Kansas Natural History Museum. LITERATURE CITED Cockerell, T. D. A. 1900. Descriptions of new bees collected by Mr. H.H. Smith in Brazil. 1. Proceed- ings of the Academy of Natural Sciences of Pliiliulel- phia 52:356-377. Eickwort, G. C. 1969. A comparative morphological study and generic revision of the augochlorine bees (Hymenoptera: Halictidae). University of Kansas Science Bulletin 48:325-524. Fabricius, J. C. 1804. Systenia piezatorum secundum or- dines, genera, species adiectis synonymis, locis, oh- servationihus, descriptionibus. Reichard, Brunsvi- gae, xiv + [15]-439, 1-30 pp. Friese, H. 1911. Neue Bienen aus Siid-Amerika (Hym.). Deutsche Entomologische Zeitschrift 1911: 453-156. Friese, H. 1926. Die Nachtbienen-Gattung Megalopta Sm. Stettiner Entomologische Zeitung 87:111-135. Jorgensen, P. 1912. Beitrag zur Biologie einiger sii- damerikanischer Bienen. Zeitschrift fiir loissen- schaftliche Insektenbiologie 8:268-271. Michener, C. D. 1962. An interesting method of pol- len collecting by bees from flowers with tubular anthers. Revista de Biotogia Tropical 10:167-175. Moure, ]. S. 1943. Notas sobre abelhas da cole^ao de Zikan (Hym. Apoidea). Revista de Entoiiiologia 14: 447^84. ' Moure, J. S. 1951. Ariphanarthra, um novo genero de Halictidae (Hymenopt. — Apoidea). Dusenia 2: 137-140. Moure, J. S. 1958. On the species of Megalopta de- scribed by F. Smith (Hymenoptera, Apoidea). journal of the New York Entomological Society 66: 179-190. Moure, J. S. 1969. Micrommation, novo genero de Hal- ictidae do Parana (Hym. Apoidea). Atas Sociedade de Biologia 12:247-249. Moure, J. S., and P. D. Hurd, Jr. 1987. An annotated catalog of the halictid bees of the Western Hemisphere (Hymenoptera: Halictidae). Smithsonian Institution Press, Washington, vii + 405 pp. Rego, M. M. C, and P. M. C. de Albuquerque. 1989. Comportamento das abelhas visitantes de Muri- el, Byrsonima crassifolia (L.) Kunth, Malpighi- aceae. Boletim do Museu Paraense Emilio Goeldi, Zoologia 5: 179-194. Renner, S. S. 1989. Floral biological observations in Heliamphora tatei (Sarraceniaceae) and other plants from Cerro de la Neblina in Venezuela. Plant Systematics and Evolution 163:21-30. Schrottky, C. 1902. Ensaio sobre as abelhas solitarias do Brasil. Revista do Museu Paulista 5:330-613. Schrottky, C. 1912. Beschreibung von Megalopta (Me- galoptella) ipomoeae n. sp. Zeitschrift fur Wissen- schaftliclien Insektenbiologie 8:271-272. Sigrist, M. R., and M. Sazima. 1991. Biologia floral e poliniza^ao por vibra^ao em duas especies sim- patricas de Dichorisandra (Commelinaceae). XLII Congresso Nacional da Sociedade Botanica do Brasil, Sociedade Botanica do Brasil, Universidade Federal de Golds 1991:484. Smith, F. 1853. Catalogue ofhymenopterous insects in the collection of the British Museum, part 1: Andrenidae and Apidac. British Museum, London, 197 pp., 6 pis. Vachal, J. 1904. Etude sur les Halictiis d'Amerique (Hym.). Miscellanea Entomologica 12:9-24, 113- 128, 137-144. J. HYM. RES. Vol. 7(1), 1998, pp. 15-24 Species Richness of Costa Rican Cenocoeliini (Hymenoptera: Braconidae): a Latitudinal and Altitudinal Search for Anomalous Diversity Leendert-Jan van DF.R Ent and Scott R. Shaw Department of Plant, Soil and Insect Sciences, P.O. Box 3354, University of Wyoming, Laramie, WY 82071, USA Abstract. — Latitudinal patterns of species diversity of New World Braconidae have been scarcely surveyed to date. Such patterns may be of biogeographical and ecological interest because some literature data suggest that some braconid subfamilies do not show an increase in species diversity towards the equator despite an increase of potential host species (i.e., "anomalous diversity"). In the present study, species diversity of a "presumptive" anomalous diverse braconid taxa, Ceno- coeliini, was surveyed in Costa Rica. The results were compared with published distribution data of North American Cenocoeliini. Also, species richness and abundance of Cenocoeliini from sea- level to 3400 m altitude in Costa Rica were analyzed to compare latitudinal and altitudinal gra- dients in species diversity. Costa Rican Cenocoeliini were five times more speciose than those in Canada and USA combined. The increase in estimated species richness per unit area towards the equator of North American Cenocoeliini was similar to that of their most common hosts, Cer- ambycidae and Scolytidae, but exceeded that of the potential hosts in Costa Rica. Diversification in Costa Rican Cenocoeliini was partly influenced by adaptation to different host families and host substrates. Most species and individuals of Cenocoeliini were found at low altitudes (<500 m) in Costa Rica. Cenocoeliini were not encountered above 1600 m in Costa Rica, this being in contrast to their most likely hosts, Cerambycidae and Scolytidae, which also occurred at high altitudes. Larger-sized Costa Rican Cenocoeliini were often brightly colored suggesting an apo- sematic function towards visually-oriented predators. New World Cenocoeliini appeared to be tropical lowland-centered and this is expected to be rather an effect of temperature requirements than an effect of host-limitation. Anomalous diversity is defined as a pat- the tropics may provide biased results for tern in species richness "counter to the large-sized parasitoids (>3 mm) in areas prevalent trend of increasing species num- where small-sized parasitoids may be ber in a taxon with decreasing latitude" abundant (Hespenheide 1979; Morrisson (Rathcke and Price 1976). Owen and et al. 1979). More recently, results of Mal- Owen (1974) were the first investigators to aise trap sampling revealed that some show anomalous diversity for parasitic subfamilies of Ichneumonidae displayed Hymenoptera of the family Ichneumoni- anomalous diversity (Gauld 1986, 1987, dae, despite an increase of numbers of po- 1995b). Several theories to explain latitu- tential host species towards the equator, dinal patterns in species diversity of par- Janzen and Pond (1975) found a similar asitic Hymenoptera have been formulated: pattern in species richness for parasitic resource fragmentation (Janzen and Pond Hymenoptera; they were less or equally 1975; Janzen 1981), predation on hosts speciose in Costa Rica compared to those (Rathcke and Price 1976), predation on occurring in a meadow in England. Sev- parasitoids (Gauld 1987), and the "nasty" eral investigators noted that sweep-sam- host hypothesis (Gauld ct al. 1992; Gauld pie studies of parasitic Hymenoptera in and Gaston 1994). According to these the- 16 Journal of Hymenoptera Research ories, different ecological groups of para- sitic Hymenoptera are expected to show different patterns in latitudinal species di- versity (Hawkins 1994). To date, anomalous diversity has not been fully analyzed for New World Bra- conidae. Quicke and Kruft (1995) found some subfamilies of Braconidae (e.g., Aly- siinae, Aphidiinae, Cheloninae) to be less speciose from northern (zone 3, 4) towards southern (zone 5) latitudes in the USA. This suggests anomalous diversity, how- ever, these analyses did not include neo- tropical regions. Species richness of anoth- er group, Cenocoeliinae (i.e., Cenocoeli- ini), increased from northern to southern temperate regions, but did not increase from middle to lower latitudes in the USA (Quicke and Kruft 1995). Also, species richness of Cenocoeliini in southern tem- perate regions was equal to that of north- ern and central Mexico combined and to that of southern Mexico (data from Saffer 1982). To date, not more than two species of Cenocoeliini have been described from Costa Rica (Saffer 1977, 1982). Previous data, therefore, suggest a pattern of anom- alous diversity in Cenocoeliini. The distri- bution of Central American Cenocoeliini, however, is less well documented than that of North America (Saffer 1982). Cenocoeliini are diurnally active endo- parasitic koinobionts of endophytic beetle larvae (van Achterberg 1994; Saffer 1982). In North America, Cenocoeliini were found to parasitize Cerambycidae (68% of recorded host species), Scolytidae (22%), and Buprestidae (Saffer 1982). Koino- bionts are parasitoids which let their hosts continue to be mobile and defend them- selves for a while after being parasitized. Koinobionts are expected to have more narrow host ranges than idiobionts (Askew and Shaw 1986; Gauld 1987; Haw- kins 1994) and are often referred to as "specialist" parasitoids. Sheehan and Hawkins (1991), however, noted that com- parisons of average host ranges between koinobionts and idiobionts have to be evaluated carefully. Specialist parasitoids (i.e., koinobionts) associated to endophytic hosts may show, as predicted by the the- ories of resource fragmentation and pre- dation on parasitoids, a (very) strong de- crease in species richness from the tem- perate to tropical zones (Table 4.2, Haw- kins 1994). At the other hand, these parasitoids may also show, as predicted by the theory of predation on hosts and that of the "nasty" hosts, a weak decrease to increase in species richness towards the equator. Thus, based on parasitizing be- havior, contrasting predictions could be made for species richness of Costa Rican Cenocoeliini. Species richness along altitudinal gra- dients may demonstrate patterns similar to those of latitudinal gradients (Brown 1988; Stevens 1992). Noyes (1989) noted that in general species diversity of para- sitic Hymenoptera in an Indonesian rain forest was the highest at low altitudes (< 1000 m). Also the highest diversity of their hosts, e.g. Lepidoptera, was found at low altitudes (Holloway 1986). This suggests that, if patterns in latitudinal and altitu- dinal species richness are similar, anoma- lous diversity should not occur among parasitic Hymenoptera. Results of an alti- tudinal transect study in the Venezuelan Andes by Janzen et al. (1976) showed that species richness of parasitic Hymenoptera was as high at 200 m as at 1600 m, but that it was lower at high altitudes (3550 and 3600 m). The decline of species rich- ness of parasitic Hymenoptera between 1600 and 3550 m was smaller than the de- cline of species richness of most other groups of insects such as other Hymenop- tera (ants, bees, aculeate wasps) and other insect orders (e.g., Coleoptera and Lepi- doptera). This suggests that some groups of parasitic Hymenoptera show anoma- lous diversity along altitudinal gradients. At a lower taxonomic level, tropical alti- tudinal species diversity of Ichneumono- idea was found to differ among subfami- lies (Gauld 1985; Gaston and Gauld 1993; Volume 7, Number 1, 1998 17 Gauld and Hanson, in press). Also, Ophioninae were found to have different patterns in altitudinal species diversities between tropical regions (Gauld and Han- son, in press). As noted earlier, anomalous diversity has not been completely analyzed for New World Braconidae. Literature data suggest that the braconid tribe Cenocoeliini may show a decrease, or at least no increase, in species diversity towards the equator. Therefore, the aims of the present study were: 1) to determine whether species richness and a-index of diversity of the Cenocoeliini in temperate regions were higher than those in tropical regions, 2) to determine whether potential hosts of Cen- ocoeliini increased more in species rich- ness towards the equator than their para- sitoids, 3) to relate latitudinal species rich- ness with altitudinal species richness, and 4) to determine whether abundance and species richness of Cenocoeliini were higher at intermediate and high than at low altitudes in Costa Rica. This research is the first in a series of analyses of differ- ent groups of Braconidae to be examined for species diversity in Costa Rica in re- lation to altitudes. MATERIALS AND METHODS The tribe Cenocoeliini is a monophyletic group in the Helconinae (Shaw 1995), al- though several authors placed the Ceno- coeliini as the main tribe in a separate sub- family, Cenocoeliinae (van Achterberg 1984, 'l993; Shaw and Huddleston 1991; Wharton 1993). Before the generic revision by van Achterberg (1994), most Cenocoe- liini species were considered to be within the genus Cenocoelius. Individuals of Cenocoeliini were col- lected with Malaise traps (85%) and the rest by hand netting in Costa Rica, mostly during the last 10 years. Hand collected specimens for this study were from H.A. Hespenheide (University of California, Los Angeles), F. Parker (University of Utah) and J.A. Ugalde (INBio, Instituto Nacional de Biodiversidad, Costa Rica). Malaise traps were located in different habitat types and at different altitudes throughout Costa Rica as described by Gauld (1991, 1995a). Our analyses includ- ed ca 70 Malaise trap-years of samplings at ca 60 different sites. Therefore, sample coverage was expected to be reasonably representative for the Costa Rican fauna. From these samples, Braconidae were sorted and sent to the University of Wy- oming for identification. Cenocoeliini were sorted and individuals were deter- mined to morphospecies using the follow- ing set of characters: hindwing venation (relative length vein IM-l-CU to IM, and IM to Ir-m); color patterns of head, me- sosoma, metasoma, legs, ovipositor sheaths and antennae; body size; oviposi- tor length relative to forewing length; number of flagellomeres; and shape of apical flagellomeres. Additional charac- ters, like sculpture patterns on the pro- episternum, apex of the propodeum, and vertex, were included to distinguish among presumptive sibling species com- plexes. Data by Saffer (1982) were used to compare species richness of Braconidae in North America with that in Costa Rica. Sample coverage and sample intensity for Mexico was relatively low and less rep- resentative for the area than those of Can- ada, USA and Costa Rica. Two formulae were used to estimate ex- pected species richness of the faunae (S') based on the numbers of individuals per species in a sample: 1. S^' = S-/(S - S,) (S, = total number of species with one individual) 2. S"- = S + ((S,)V2S,) (S; = total number of species with two individuals) The first formula of expected species rich- ness (S'') is derived from the formula of sample coverage (1-(N,/I): Fagen and Goldmann 1977). In this formula, I is the total number of behavior types observed and N| is the total behavior types ob- 18 Journal of Hymenoptera Research served only once. I was substituted with S and N, with S,. S""' was calculated as the inverse of the sample coverage multiplied by the number of observed species (S). The second formula of expected species richness (S'"'), is that of Chao 1 as de- scribed in Colwell and Coddington (1994). In addition to species richness, a-index of diversity of the logarithmic series was cal- culated because of its good discriminant ability and its low sensitivity to sample size (Magurran 1988). To estimate species richness per unit area the formula S, = x/a°-^ was used (S, = number of species per unit area, x = number of observed species in country or region, a = area of country or region (10' km-): MacArthur and Wilson 1967, Gaston et al. 1996). We also estimated species richness per unit area for the most important temperate hosts of Cenocoeliini, the Cerambycidae (data from Monne and Giesbert 1994) and the Scolytidae (data from Wood 1982; Wood et al. 1991). To examine the effect of altitude on spe- cies richness of Cenocoeliini, we defined 4 altitude classes: low (0-500 m), low inter- mediate (500-1500 m), high intermediate (1500-2500 m) and high (>2500 m) alti- tudes. These altitude classes reflect the distribution of different habitat types as described by Gauld (1995a). We assumed that there was a linear relationship be- tween sample effort (i.e.. Malaise trap- months) and number of individuals caught in Malaise traps. Because seasonal variation in abundance of neotropical in- sects occurs (Owen and Chanter 1970; Wolda 1988, 1989; Wolda and Wong 1988), only Malaise traps which operated three or more consecutive months were includ- ed in the analysis, and abundances of Cen- ocoeliini were summed for several year- round Malaise trap samples. We estimated expected numbers of Cenocoeliini per al- titude class by multiplying the total num- ber of observed Cenocoeliini with the pro- portion of number of Malaise trap-months of a particular altitude class to the total Table 1. Numbers of individuals (N), species rich- ness (S), two estimates of expected species richness (S"', S'-; see materials and methods), a-index of di- versity, numbers of localities (loc) and averaged local species richness (S,,^) of Cenocoeliini from Canada and USA combined, Mexico and Costa Rica. Canada + USA MfMO. e i.sla Rica N 1108 51 301 S 11 13 57 S" 11 24 96 S"= 11 22 123 a-index 1.665 5.291 20.822 loc 233 13 48 s„„ 1.1 ± 0.3 1.5 ± 1.0 2.8 ± 4.4 (range) 1-3 1-4 1-25 number of Malaise trap-months. We test- ed whether observed and expected num- bers were equal between low (<500 m) and higher altitude classes combined us- ing a chi-square test (Sokal and Rohlf 1981). RESULTS In total, 290 individuals of Cenocoeliini were found from 80,000 to 90,000 Bracon- idae sampled in Costa Rica. The individ- uals of Cenocoeliini were divided into 55 morphospecies. The two described Costa Rican Cenocoeliini species by Saffer (1977, 1982) did not match our individuals and were added to our morphospecies result- ing in a total of 57 species and 301 indi- viduals (Table 1). From this set of 57 spe- cies, 44 species belonged to Capitoiiius, and 7 species to Ceiiocoelius. Of the remaining species, 4 species may belong to Capitonius but are rather distinctive and 2 species be- long to a new genus of Cenocoeliini. The species varied in size from 3 to 12 mm. The ovipositors were relatively long, 0.7 to 1.7 times the forewing length. About half of the individuals and species were rela- tively large, mostly >5 mm, and bright yellow-orange to red-orange and often partly black in color, with partly to com- pletely darkened wings. Most other indi- viduals were relatively small, <7 mm, and blackish, brownish or black and dull red Volume 7, Number 1, 1998 19 Table 2. Estimated species richness per unit area (S,) of parasitoids (Cenocoeliini) and their most common temperate hosts (Cerambycidae and Scolytidae) for different geographical regions of North and Central Amer- ica, using S., = x/a"-' as a standard species-area relationship (see text). Between parentheses are the relative richnesses when the estimated species richness of Canada was set at 1.0. Geographical region Area (10' km-') Cenocoeliini S. Cerambycidae S," ScolvHciae Canada ( + Alaska) 11,49b 0.39 (1.0) 29.4 (1.0) 17.3(1.0) USA (-Alaska) 7,828 1.17 (3.0) 85.8 (2.9) 50.3 (2.9) Mexico 1,969 1.95 (5.0) 201.2 (6.8) 90.8 (5.2) Costa Rica 5\ 21.36 (54.8) 260.1 (8.8) 155.3 (9.0) to reddish brown in color with clear wings, sometimes with brownish spots on the forewings. Highest species richness was found in Costa Rica and lowest species richness in Canada and USA combined (Table 1). Both estimators of expected species rich- ness (S''' and S'-) predicted that in Mexico and Costa Rica about half of the total number of species have been caught to date. More species were found in 2 Costa Rican Malaise trap sites (Heredia Prov- ince, Biological Station of OTS La Selva, and Puntarenas Province, 24 km west of Piedras Blancas) than in all trap sites of Mexico or Canada and USA combined. The highest diversity (a-index) was found for Costa Rican Cenocoeliini, the lowest for those of Canada and USA combined. The increase of estimated species rich- ness per unit area for North American Cenocoeliini was similar to that of their most common hosts (Table 2). On average, 3 times more species per unit area oc- curred in the USA than in Canada and about twice as many species per unit area occurred in Mexico compared to the USA. In contrast, Costa Rican Cenocoeliini were 11 times more speciose per unit area than in Mexico, meanwhile the potential hosts increased only about 1.5 times in species richness per unit area for the same areas. Cenocoeliini were most speciose at low altitudes in Costa Rica (Table 3). Speci- mens were not encountered above 1600 m and Cenocoeliini were, significantly, more abundant at low (<500 m) than at higher (>500 m) altitudes. DISCUSSION The increase of species richness of Cen- ocoeliini from Canada to Mexico was equal to that of their common hosts and exceeded that of their potential hosts in Costa Rica. Hence, anomalous diversity could not be shown for New World Cen- ocoeliini. This is in contrast to the predic- tion by resource fragmentation (Janzen and Pond 1975; Janzen 1981) suggesting that tropical host trees and their associat- ed host beetles are too widely distributed to be exploited by koinobiont parasitoids. To illustrate this fragmentation, one ha of tropical lowland rain forest in Costa Rica contained 3 to 4 times more tree species than one ha of the most diverse North American forests (Hartshorn 1983; Whit- tacker 1965). Also a 50 ha moderate di- verse lowland rain forest in Panama con- tained an equal number of tree species as the whole of western North America, north of Mexico (Condit et al. 1996; Little 1980). Janzen's (1981) caveat to resource fragmentation was that tropical koino- biont parasitoids had developed a broader host range compared to temperate koino- bionts, or that they had become very effi- cient in looking for sparcely distributed hosts. In the present study, this could not be analyzed for neotropical Cenocoeliini, partly due to the lack of host records. North American Cenocoeliini, on average. 20 Journal of Hymenoptera Research Table 3. Total species richness (S), number of Malaise trap-months {# tm), observed (N„i,J and expected (N^.,j,) numbers of individuals of Cenocoeliini per altitude class in Costa Rica. Only Malaise traps which operated more than three consecutive months were included in the analysis. It was tested if the observed and expected numbers of Cenocoeliini were equal at low altitudes (<500 m) and at higher altitude classes com- bined using chi-square test. Altitude cl ass S # tm N N,..„ xMdf = 1) S500 m 48 546 219 155 78.27 500-1500 m 16 71 8 20 (P < 0.001) 1500-2500 m 5 138 7 15 39 79 >2500 m 0 72 0 20 Totals 57* 827 234 234 * Twelve species occurred at the two lowest altitude classes; one species occurred from 0-1600 m. Individuals were not found above 1600 m altitude. parasitize one to 4 different host species (Saffer 1982). This "narrow" host ranges would classify the North American Cen- ocoeliini as specialists, however, we have to be careful to generalize, because these host records were not complete (Shaw 1994). Results of the present study showed that the increase in species richness per unit area of Cenocoeliini from Mexico to Costa Rica was 6 times larger than that of their potential hosts (Table 2). This sug- gests that Costa Rican Cenocoeliini para- sitize a larger proportion of the Ceram- bycidae and Scolytidae, than those in North America. At the other hand, some Costa Rican Cenocoeliini may have adapt- ed to other beetle families of beetles with a different biology. The only known host record from Cenocoeliini in Costa Rica was that of a species attacking seed-boring beetle larvae of the family Curculionidae (Saffer 1977). Long-term research on seed- boring beetles in Costa Rican dry forest (Janzen 1980), however, did not reveal ad- ditional observations of Cenocoeliini on seed-boring beetles (Janzen, pers. comm.). Five percent of our surveyed specimens of Cenocoeliini from Costa Rica were collect- ed from treefalls by hand (Hespenheide, unpublished data) and other micro-habi- tats not were indicated on the collecting labels. We expect, therefore, that wood- and bark-boring insects are the most likely host for Costa Rican Cenocoeliini. In the present study, we found that few- er species of Cenocoeliini occur at inter- mediate than at low altitudes (Table 3). This is comparable to the reduction in spe- cies richness towards the equator (Table 1). But the decline in species richness of Cenocoeliini with altitude in Costa Rica was higher than expected. As shown in the present study, Cenocoeliini were not encountered at 2000 m in Costa Rica. At latitudes in temperate North America with a similar mean yearly temperature, however, 5 to 8 species of Cenocoeliini oc- curred (Quicke and Kruft 1995). In Costa Rica, seasonal changes in temperature are usually smaller than diurnally fluctuation in temperature (Gauld 1995a). In temper- ate regions this is often reversed. These differences in temperature regimes may affect species richness of New World Cen- ocoeliini. Individuals of Cenocoeliini were not en- countered at altitudes higher than 1600 m, even though a total of 150 Malaise trap- months located at 10 different sites were surveyed. This was an unexpected obser- vation as potential hosts of Cenocoeliini have been observed at high altitudes. Cer- ambycidae are most abundant at low al- titudes in Costa Rica but occur also at high altitudes up to timberline (3200-3400 m; Volume 7, Number 1, 1998 21 Lezama, pers. comm.). Scolytidae are known to be regularly encountered at high altitudes in Costa Rica (Wood et al. 1991). It is unlikely, however, that alter- native hosts for Cenocoeliini, like larvae of seed-boring beetles do occur at high alti- tudes. Legume trees, of which the fruits are among the most frequently attacked by seed-boring beetles (Janzen 1980), are scarce at intermediate and absent at high altitudes (Holdridge et al. 1971). Also, Gas- ton and Gauld (1993) noted that Pimplinae (Icneumonidae) were more abundant at high altitudes than at low altitudes. In case species of Cenocoeliini would have been present at high altitudes in Costa Rica, they likely would have been collect- ed in the Malaise traps. This suggests that host presence does not explain absence of Cenocoeliini at high altitudes in Costa Rica. In the present study, it could not be de- termined if Cenocoeliini were scarce in Costa Rica or that it is difficult to sample them by using Malaise traps. On average, one individual of Cenocoeliini was caught per 3 to 4 Malaise trap-months. Trap effi- ciency for Cenocoeliini was twice as high at low altitude rain forests than at low middle altitude rain forest or low altitude dry forest. Also Cenocoeliini were most frequently caught during the dry season (Feb. -May; unpublished data). But even in the optimal habitat type and season, Cen- ocoeliini was never found to be abundant suggesting that they occur in low popu- lation densities. Another remarkable result was that 77% of the Mexican and 60% of the Costa Rican species of Cenocoeliini were represented by one or 2 individuals. Estimated species richness of Mexican Cenocoeliini may have been underestimated as sample cov- erage by the Malaise traps in Mexico was low compared to those in Costa Rica and the USA and Canada combined. Analyses of geographical distribution of Costa Rican Cenocoeliini could not be jus- tified due to low numbers of individuals in Malaise traps. The observed 24 Ceno- coeliini species in the Pacific lowland rain forest around Golfo Dulce in Puntarenas Province shared 9 species with the 25 Cen- ocoeliini species in the Atlantic lowland rain forest of La Selva in Heredia Prov- ince. This species distribution of Cenocoe- liini may suggest that many Costa Rican Cenocoeliini have a restricted geographi- cal distribution according to Rapoport's Rule (Stevens 1989, 1992). It may also be a sample artifact due to the low number of individuals. In the present study it was found that half of the Cenocoeliini and especially those larger than 5 mm were bright or- ange and black colored, often with partly or completely darkened wings. Saffer (1982) described 2 similar bright colored Cenocoeliini from southern tropical Mex- ico, but no such colored species from tem- perate North America. Bright colors are common in tropical parasitoids (Quicke 1986a; Shaw 1995). Bright colors occur in several other neotropical braconid sub- families such as Agathidinae and Bracon- inae and they are characteristic for larger sized (>5 mm) diurnally active Braconi- dae with long ovipositors, which likely parasitize concealed hosts (Shaw 1995). In general, bright colors are characteristic for lowland insects where it occurs in up to 25% of insects and do not occur at high altitudes in Costa Rica (Janzen 1973). Quicke (1986b) noted that in general bright colors have a warning function to- wards visually-oriented predators (i.e., aposematic coloration). Bright colored tropical parasitic wasps may mimic sting- ing aculeates and some larger sized para- sitoids are capable of stinging by them- selves (Quicke 1986b). Other authors hy- pothesized that parasitoids may mimic unpalatable hosts such as Chrysomelidae (Gauld, pers. comm.) or Symphyta (anon, rev., pers. comm.). Gauld and Gaston (1994), suggested that parasitoids with bright colors may be unpalatable for pred- ators after sequestering "nasty" tasting 22 Journal of Hymenoptera Research secondary plant chemicals from their hosts. If the latter is true and the "nasty" host hypothesis has validity for Cenocoe- liini, bright colored Cenocoeliini may at- tack seed-eating or phloephageous beetle larvae, rather than wood-living Scolytidae or Cerambycidae. Quicke (1986a) found homeochromatic assemblages for some large sized Braconidae and their potential hosts (i.e., Cerambycidae). Hespenheide (1996) showed that color patterns of Chrysomelidae may be substrate- related. More research is needed to elucidate the underlying defense mechanisms or other meaning of bright colors in large sized neotropical wood-boring braconid parasit- oids. The smaller sized Costa Rican Cenocoe- liini from our survey were mostly less conspicuous colored than the larger sized ones. Many of these smaller sized Ceno- coeliini may be ant-mimics, which is ex- pected to occur frequently in neotropical Braconidae (Shaw 1995). This color pat- tern of presumably ant-mimics also oc- curred among North American Cenocoe- liini (Saffer 1982). To date, possible behav- ioral and olfactorial cues involved in ant- mimicry of Braconidae have not been doc- umented. In conclusion, the results of the present study showed that the species diversity of the Cenocoeliini increases towards the equator. This is the normal pattern in lat- itudinal species richness (Fisher 1960; Pianka 1966; Stevens 1989; Wilson 1992). Thus, anomalous diversity was not ob- served for Cenocoeliini. The increase in species richness per unit area of Cenocoe- liini from temperate North America to the Neotropics exceeded that of their potential hosts. In Costa Rica, species and individ- uals of Cenocoeliini were not found above 1600 m altitude, this in contrast to their potential hosts which also occur at high altitudes. This suggests that species rich- ness in Cenocoeliini is not host limited. The Cenocoeliini apparently is a tropical lowland-centered group of which a limit- ed number of species have adapted to year-round cool conditions at lower mon- tane rain forests and none to montane for- ests. Some other species have adapted to climatological conditions in temperate regions and evolved overwintering mech- anisms (Saffer 1982). ACKNOWLEDGMENTS We would like to thank Paul Hanson and Ian Gauld for setting up the Malaise trap network in Cos- ta Rica and for having the University of Wyoming involved in identification of Braconidae. We would like to thank Ian Gauld, David Kazmer, Dave Legg, Jeff Lockwood, Mark Shaw and an anonymous re- viewer for their critical comments on earlier versions of the manuscript. LITERATURE CITED Achterberg, C, van. 1984. Essay on the phylogeny of Braconidae (Hymenoptera: Ichneumonoidea)£ii- tomologic TuMrift 105: 41-58. Achterberg, C, van. 1993. Ilustrated key to the sub- families of Braconidae (Hymenoptera: Ichneu- monoidae). Zoologische Verhamkiiiigen 283: 1-189. Achterberg, C, van. 1994. Generic revision of the sub- family Cenocoeliinae Szepligeti (Hymenoptera: Braconidae). Zoologische Verhandelingen 292: 1-52. Askew, R. R. and M. R. Shaw (1986). Parasitoid com- munities: their size, structure and development. Pages 225-264, in J. Waage and D. Greathead (eds) /»str( Parasitoids. 13th Symposium of the Royal Entomological Society of London. Aca- demic Press, London. Brown, J. H. 1988. Species diversity. Pages 57-90, in A. A. Myers and P. S. Gilles (eds) Analytical Ge- ography. Chapmann and Hall. London, New York. " Colwell, R. K. and ]. A. Coddington. 1994. Estimation terrestrial biodiversity through extrapolation. Philosophical Transactions of the Royal Society Lon- don (series B) 345: 101-118. Condit, R., S. P. Hubbell, J. V. Lafrankie, R. Sukumar, N. Manokaran, R. B. Foster and P. S. Ashton. 1996. Species-area and species-individual rela- tionships for tropical trees: a comparison of three 50-ha plots, journal of Ecology 84: 549-562. Fagen, R. M. and R. N. Goldmann. 1977. Behavioural catalogue analysis methods. Animal Behaoioiir 25: 261-271. Fischer, A. G. 1960. Latitudinal variation in organic diversity. Evolution 14: 64-81. Gaston, K. J. and I. D. Gauld. 1993. How many spe- cies of pimplines (Hymenoptera: Ichneumoni- dae) are there in Costa Rica? journal of Tropical Ecology 9: 491^99. Volume 7, Number 1, 1998 23 Gaston, K. J., I. D. Gauld and P. Hanson. 1996. The size and composition of the hymenopteran fauna of Costa Rica. ]ournal cf Bwgeogrnp!iy 23: 105-113. Gauld, I. D. 1985. A preliminary survey of Ophioni- nae (Hymenoptera: Ichneumonidae) of Brunei. Brunei Museum Journal 6: 169-188. Gauld, I. D. 1986. Latitudinal gradients in ichneu- monid species-richness in Australia. Ecologicnl Entomolog}/ 11: 155-161. Gauld, I. D. 1987. Some factors affecting the compo- sition of tropical ichneumonid faunas. Biological loiirnal of the Linnaean Society 30: 299-312. Gauld, I. D. 1991. The Ichneumonidae of Costa Rica, 1. Memoirs of the American Entomological Institute. No. 47. Gauld, I. D. 1995a. Introduction. Pages 1-19, in P. E. Hanson and I. D. Gauld (eds) The Hymenoptera of Costa Rica. The Natural History Museum, Lon- don. Oxford Scientific Publications. Gauld, I. D. 1995b. Ichneumonidae. Pages 390^31, in P. E. Hanson and I. D. Gauld (eds) The Hyme- noptera of Costa Rica. The Natural History Muse- um, London. Oxford Scientific Publications. Gauld, I. D., K. J. Gaston and D. H. Janzen. 1992. Plant allelochemicals, tritrophic interactions and the anomalous diversity of tropical parasitoids: the "nasty" host hypothesis. Oikos 65: 353-357. Gauld, I. D. and K. J. Gaston. 1994. The taste of ene- my-free space: parasitoids and nasty hosts. Pages 279-299, in B. A. Hawkins and W. Sheehan (eds) Parasitoid Community Ecology. Oxford Scientific Publications. Gauld, 1. D. and P. Hanson. In press. Distribution pat- terns of cloud forest Hymenoptera in Costa Rica. In N. Kadkarni and N. Weelright (eds.) Montev- erde: natural history and conservation. Oxford Uni- versity Press. Hartshorn, G. S. 1983. Plants: Introduction. Pages_ 118-183, in D. H. Janzen (ed) Costa Rican Natural History. University of Chicago Press. Hawkins, B. A. 1994. Patterns & Process in Host-Para- sitoul Interactions. Cambridge University Press. Hespenheide, H. A. 1979. Are there fewer parasitoids in the tropics? The American Naturalist 113: 766- 769. Hespenheide, H. A. 1996. The role of plants in struc- turing communities of mimetic insects. Chapter 6. Pages 109-126, in A. C. Gibson (ed) Neotropical biodiversity and conservation. Occasional Publica- tion of the Mildred E. Mathias Botanical Garden 1. University of California, Los Angeles. Holdridge, L. R., W. C. Grenke, W. H. Hatheway, T. Liang and ]. A. Tosi Jr. 1971. Forest environments in tropical life zones: a pilot study. Pergamon Press, Oxford. Holloway, J. D. 1986. Macrolepidopteran diversity in the Indo-Australian tropics.: geographic, biotopic and taxonomic variations. Biological Journal of the Linnean Society 30: 325-341. Janzen, D. H. 1973. Sweep samples of tropical foliage insects: effects of seasons, vegetation types, ele- vation, times of day, and insularity. Ecology 54: 687-708. Janzen, D. H. 1980. Specificity of seed-attacking bee- tles in Costa Rican deciduous forest. Journal of Ecology 68: 929-952. Janzen, D. H. 1981. The peak in North American Ich- neumonid species richness lies between 38° and 42° N. Ecology 62: 532-537. Janzen, D. H., M. Ataroff, M. Farinas, N. Rincon, A. Soler, P. Soriano and M. Vera. 1976. Changes in the arthropod community along an elevational transect in the Venezuelan Andes. Biotropica 8: 193-203. Janzen, D. H. and C. Pond. 1975. A comparison, by sweeping, of the arthropod fauna of secondary vegetation in Michigan, England and Costa Rica. Transactions of the Royal Entomological Society Lon- don {Series B) 127: 33-50. Little, E. L. 1980. National Audubon Society field guide to North American trees- western region. Alfred A.- Knopf, New York. MacArthur, R. H. and E. O. Wilson. 1967. The theory of island biogeography . Princeton University Press, New Jersey. Magurran, A. E. 1988. Ecological diversity and its mea- surement. Princeton University Press, New Jersey. Monne, M. A. and E. F. Giesbert. 1994. Checklist of the Cerambycidae of the western hemisphere. Wolfgar- den Books, Burbank California, USA. Morrison, G., M. Auerbach and E. D. McCoy. 1979. Anomalous diversity of tropical parasitoids: a general phenomenon? The American Naturalist 114: 303-307. Noyes, J. S. 1989. The diversity of Hymenoptera in the tropics with special reference to Parasitica in Sulawesi. Ecological Entomology 14: 197-207. Owen, D. F. and D. O. Chanter. 1970. Species diver- sity and seasonal abundance in tropical Ichneu- monidae. Oikos 21: 142-144. Owen, D. F. and J. Owen. 1974. Species diversity in temperate and tropical Ichneumonidae. Nature 249: 583-584. Pianka, E. R. 1966. Latitudinal gradients in species diversity: a review of concepts. The American Nat- uralist 100: 33-46. Quicke, D. L. J. 1986a. Preliminary notes on homeo- chromatic associations within and between the Afrotropical Braconinae (Hvmenoptera, Bracon- idae) and Lamiinae (Coleoptera, Cerambycidae). Entomologi.it's Monthly Magazine 122: 97-109. QuickC, D. L. J. 1986b. Warning coloration and mim- icry, in Betts (ed.) The Hymenopterist's Handbook, 2th edition The amateur Entomologist 7: 74-80. Quicke, D. L. J. and R. A. Kruft. 1995. Latitudinal 24 Journal of Hymenoptera Research gradients in North American braconid wasp spe- cies richness and biology. Journal of Hi/iiieiioftern Research 4: 194-203. Rathcke, B. ]. and P. W. Price. 1976. Anomalous di- versity of tropical ichneumonid parasitoids: a predation hypothesis. The American Nntiirnlist 110: 889-902. Saffer, B. 1977. A new species of Cenocoelius from Cos- ta Rica. Proceedings of the Entomological Society of Washington 79: 593-596. Saffer, B. 1982. A systematic revision of the genus Cenocoelius (Hymenoptera, Braconidae) in North America including Mexico. Polskie Pismo Ento- mologiczne 52: 73-167. Shaw, M. R. 1994. Parasitoid host ranges. Pages 111- 144, in B. A. Hawkins and W. Sheehan (eds) Par- asitoid Community Ecology. Oxford University Press. Shaw, M. R. and T. Huddleston. 1991. Classification and biology of braconid wasps (Hymenoptera: Braconidae). in W. R. Dolling and R. R. Askew (eds) Handbook for identification of British Insects. Vol. 7, Part iii. Royal Entomological Society of London. Shaw, S. R. 1995. Braconidae. Pages 431-463, in P. E. Hanson and 1. D. Gauld (eds) The Hymenoptera of Costa Rica. The Natural History Museum, Lon- don. Oxford University Press. Sheehan, W. and B. A. Hawkins. 1991. Attack as an indicator of host range in metopiine and pimp- line Ichneumonidae (Hymenoptera). Ecological Entomology 16: 129-131. Sokal, R. R. and F. J. Rohlf (1981). Biometry. 2nd edi- tion. W.H. Freeman and Company. San Francis- co. Stevens, G. C. 1989. The latitudinal gradient in geo- graphical range: how so many species coexist in the tropics. The American Naturalist 133: 240-256. Stevens, G. C. 1992. The elevational gradient in alti- tudinal range: an extension of Rapoport's's lati- tude rule to altitude. The American Naturalist 140: 893-911. Wharton, R. A. 1993. Bionomics of Braconidae. An- nual Reviezv of Entomology 38: 121-143. Whittacker, R. H. 1965. Dominance and diversity in land plant communities. Science 147: 250-260. Wilson, E. O. 1992. The diversity of life. Harvard Uni- versity Press, Cambridge, Massachusetts. Wolda, H. 1988. Insect seasonality: Why? Annual re- view of Ecology and Systematics 19: 1-18. Wolda, H. 1989. Seasonal cues in tropical organisms. Rainfall? Not necessarily? Oecologia 80: 437^142. Wolda, H. and M. Wong. 1988. Tropical insect diver- sity and seasonality. Sweep-samples vs. light- traps. Proclaniaties Koninklijke Nederlandse Akade- mie voor Wetenschappen 91: 203-216. Wood, S. L. 1982. The bark and ambrosia beetles of North and Central America (Coleoptera: Scoly- tidae) a taxonomic monograph. The Great Basin Naturalist Memoirs No.6. Brigham Young Univer- sity, Provo, Utah. Wood, S. L., G. C. Stevens and H. J. Lezama. 1991. Los Scolytidae de Costa Rica: clave de generos y de subfamilia Hylesinae (Coleoptera). Rei'ista Biologica Tropica 39: 125-148. J. HYM. RES. Vol. 7(1), 1998, pp. 25-37 Sperm Development in the Imaginal Testes of Aleiodes coxalis (Hymenoptera: Braconidae: Rogadinae) Terence M. Newman and Donald L. J. Quicke' Department of Biology, Imperial College of Science, Technology and Medicine, Silwood Park, Ascot, Berks SL5 7PY, UK Abstract — Spermatogenesis and spermiogenesis in the imaginal testes of the long-lived braconid wasp, Aleiodes coxalis (Spinola) has been investigated. The cyclostome group of braconid subfam- ilies, to which Aleiodes belongs, are considered to have unspecialised sperm, but in Aleiodes several aspects of spermatogenesis and morphology differ from that reported in other Hymenoptera, suggesting that reinterpretation of previously published studies may be necessary. In particular, the centriolar adjunct is found to lie between the nucleus and one of the pair of mitochondrial derivatives, resulting in the mitochondrial derivatives being offset longitudinally, giving the im- pression that the mitochondrial derivatives are of different lengths. A projection extends from the centriolar adjunct to the one mitochondrial derivative which abuts the nucleus. The shape change undergone by the nucleus during spermatogenesis is associated with an uneven distribution of peri-nuclear microtubules (microtubular manchette). These are not found over two extra-nuclear electron-dense regions, ('lateral plates') which appear to add increased rigidity to the nucleus at these points. Very little is known about the sperm of Hymenoptera despite the fact that this is one of the largest and most diverse of in- sect orders (Jamieson 1987), and descrip- tions of spermatogenesis are limited to a handful of taxa, mostly bees and ants (Hoage & Kessel 1968, Hogge & King 1975, Cruz-Landim & Beig 1980). A recent survey of mature sperm morphology and ultrastructure across the order has re- vealed considerable variation between higher taxa that may be important for un- derstanding phylogenetic relationships within the order (Quicke et al. 1992), some of the most extreme modifications occur- ring in the parasitic wasp family, Bracon- idae. In order to interpret the structures and modifications found in some braconid subfamilies it is first necessary to describe the 'normal' situation in that family. The Braconidae is a large family of par- asitic wasps comprising some 40 subfam- ' To whom correspondence should be addressed. ilies that can be broadly divided into two groups, the cyclostomes and the non-cy- clostomes, which differ in morphology of the mouth and in their biology (Shaw & Huddleston 1991). In an initial survey (Quicke et al. 1992), it was found that in members of the cyclostome group of brac- onid subfamilies (e.g. Braconinae, Doryc- tinae and Rogadinae, the latter including Aleiodes) the sperm have a normal appear- ance, being between 80 and 170 |xm long with the head (nucleus + acrosome) com- prising 10 to 20% of the sperm's total length. However, nearly all non-cyclo- stomes have apomorphic sperm with total length between 10 and 20 |xm, of which the head constitutes nearly 50%. Further, whilst the axonemes of the cyclostome braconid sperm have a normal 9-1-9-1-2 ar- rangement of microtubules, as found in most insects, those of the short apomor- phic non-cyclostome sperm frequently have only one or no central microtubules. These features appear to be highly phy- 26 Journal of Hymenoptera Research logenetically informative and, in order better to understand their ontogenies, we have investigated spermatogenesis in a range of braconids. Here we report on that leading to the apparently unmodified sperm of the rogadine, Aleiodes coxalis (Spinola), which will form the basis for fu- ture comparisons. Despite the normal ap- pearance and ultrastructure of the mature sperm in Aleiodes, spermatogenesis in this taxon exhibited several structures that do not appear to have been reported in other insects. It has long been appreciated that the timing of spermiogenesis and spermato- genesis in insect testes is a reflection of the adult life span which may in turn be as- sociated with the occurrence of sib-mating (Phillips 1970). Thus imaginal testes of species with short lived adults, such as mayflies (Needham et al. 1935) and caddis flies (Ross 1944), or those in which males almost invariably mate with their nearby sisters upon emergence such as many par- asitic wasps, typically contain spermatids and spermatozoa, but no gonial and mei- otic divisions, which would be completed in the pupal or nymphal stages. In con- trast, meiotic and even gonial divisions of- ten continue well into the imaginal stages in testes of long lived species (e.g. many beetles and dragonflies) which usually search for and mate with several unrelat- ed females. Most rogadine braconid wasps, such as the Aleiodes species inves- tigated here, are solitary parasitoids of Lepidoptera larvae and belong to the sec- ond category. Males fed on honey water will often live a month and sometimes considerably longer, and will mate readily throughout much of their adult life (M. R. Shaw pers. comm.). Indeed, some species are reluctant to mate immediately after eclosion and only start to mate once they have fed. There are no papers on the ultrastruc- ture of spermatogenesis in any member of the Ichneumonoidea, and the only para- sitic Hymenopteran studied to date is the pteromalid chalcidoid Nasoiiia vitripennis (Walker) (Hogge & King 1975). MATERIALS AND METHODS Adult Aleiodes coxalis (Spinola) were col- lected at Silwood Park (Imperial College) in 1995. Testes were dissected under 2% glutaraldehyde in phosphate buffered sa- line (pH 7.2), and fixed for two hours. Tis- sue was transferred to 2% osmium tetrox- ide in cacodylate buffer (pH 7.2) for 2 hr. After another buffer wash, tissue pieces were dehydrated to 50% ethanol and then further fixed with saturated uranyl acetate in 50% ethanol prior to complete dehydra- tion, embedding in Epon resin and poly- merisation overnight. Large silver sections were picked-up onto high resolution grids, stained with uranyl acetate and lead citrate and examined using a Phillips EM400 electron microscope. RESULTS Adult testes of Aleiodes coxalis contain follicles (seminal tubules) with multiple cysts present at various levels of maturity. Although neighbouring cysts can exhibit different stages of development, generally the more posterior the position along the testes the greater the proportion of later stages. Although most cysts contain sper- matids, it is also possible to find in the adult indications of earlier stages of de- velopment. Primary spermatocytes still appear to be present some without any sign of a normal spindle and with a con- centric arrangement of the endoplasmic reticulum around the chromosomes (Fig. 1 long arrow). Centriolar pairs are present in these cells {arrowheads), as are large numbers of small mitochondria. This ap- pearance is considered indicative of an abortive first meiotic division; the result of incomplete centriolar polarisation. Centri- oles are then extruded after meiotic mul- tiplication by cytoplasmic blebbing (Fig. 2), where centrioles, attached to microtu- bules (Fig. 2a), are found incorporated into rounded portions of the plasma mem- Volume 7, Number 1, 1998 27 brane. This blebbing is also seen in sec- ondary spermatocytes (which also exhibit loss of plasma membrane in larger whorls), although not all blebs exhibit cen- trioles (Fig. 2b arrow). Large numbers of polyribosomes are also present at this stage. In the secondary spermatocytes it is dif- ficult to find evidence of synaptonemal- like structures. At this stage the cells are considered to form a syncytium because of incomplete cytogenesis during the ear- lier mitotic division. The intercellular bridges (Fig. 3) linking the spermatocytes are difficult to identify as the cells can be very closely apposed. It is only during lat- er spermiogenesis, with cell elongation, that the connections become apparent (Fig. 3a) with evidence of organelle conti- nuity and possibly even intercellular movement of vesicles and granules (Fig. 3b). This has been suggested as being a method of transferring organelles from large to small spermatids where there is unequal division after the second meiotic division. No noticeable size difference was seen in Aleiodes. Nebenkern formation occurs concomi- tantly with the above processes (Fig. 4). Mitochondria begin to accumulate in one area of the cell (Fig. 4a) where they fuse to form the beginnings of the nebenkern (Fig. 4b), a highly convoluted membra- nous organelle that is the precursor of the two mitochondrial derivatives. The devel- opment of the nebenkern into two labyrin- thian networks has been linked with the appearance of microtubules in the cyto- plasm. However, in Aleiodes microtubules appear to be a consistent component of the cytoplasm from earlier stages. Flagellum growth (Fig. 5) is evident pri- or to nuclear condensation, commencing with the appearance of a basal body (Fig. 5a arrowhead) in a small depression of the nucleus (Fig. 5a arrow). The nuclear enve- lope thickens at this area (Fig 5b) and the doublet microtubules of the flagellum grow from the triplet microtubules of the basal body (Fig 5b arroio). The flagella ini- tially do not have accessory tubules (Fig. 5c). These appear to develop from sub-fi- bre b of each doublet as previously re- ported. The flagella /mitochondria axis is developed before this takes place (Fig. 5d arroiv). Although spermiogenesis is ad- vanced at this stage, separation following the last meiotic division in many cases is not complete and the nucleus still appears spherical. Nuclear condensation begins with the appearance of polarity in the nu- clear membrane in the region closely ap- posed to the spermatid cell membrane (Figs 5d, 6 arrowheads) where the sperma- tid is attached to the wall of the cyst (a single epithelial layer with supportive and secretory functions). Microtubules appear around the nucleus but these are not even- ly located; an arrangement which proba- bly assists the complicated shape change that the nucleus undergoes. The acrosome is detectable in early spermatid stages as an accumulation of a large number of small particles termed the proacrosomal granule (Fig. 7). During spermiogenesis the proacrosomal granule becomes located between the plasma membrane and the nucleus at the region of nuclear polarity (Fig. 8a). The acrosome is formed by a flattening of this structure and an indenting on the side apposed to the nucleus (Fig. 8b). A small protrusion is found anteriorly to the acrosome in the early spermatid (Fig. 8b arroiv). This may represent an apical expansion of extra-ac- rosomal material, similar to that found in Locusta migratoria L. (Szollosi 1974). How- ever, the structure in Aleiodes appears to contain electron dense material. Whether this could be a separate structure, or rep- resents a later stage of fusion of Golgi de- rived products, as found in the proacro- somal granule, is not clear. The situation is further complicated by the impression that the is distinct coated membrane which surrounds the acrosome and ex- tends posteriorly over the anterior portion of the nucleus during development, at 28 Journal of Hymenoptera Research 3b =*■•*?;^■ ;4b J, ;■ Figs. 1-4. 1, Primary spermatocyte from the testes of nn adult male Aleiodes, with anomalous metaphase plate formation, characterised by the absence of a normal spindle and a concentric arrangement of the en- doplasmic reticulum around the chromosomes (armii'). Large numbers of small mitochondria and a centriolar pair (anvu'liciidi), are also present. Scale bar = 0.5 |j.m. 2, Membrane blebbing during spermiogenesis: (a) loss of centrioles (attached to microtubules) through membrane blebbing; (b) membrane blebbing (iinoiv) without loss of centrioles. Scale bar a = 0.6 |i,m; b = 0.85 jjim. 3, Syncytium formed by incomplete cvtogenesis during mitotic division: (a) Intercellular bridges (imou'lwntl) link spermatocytes; (b) linkage provides a method for organelle movement between spermatocytes. Scale bar = 0.75 jjim. 4, Nebenkern formation in the spermato- Volume 7, Number 1, 1998 29 least initially does not appear to be contin- uous with this granular tip. As the acro- some elongates an acrosomal rod (or per- foratorium) develops, and extends from a depression in the nucleus and into the su- bacrosomal space formed during folding (Fig. 8c, arrozi'). The acrosomal rod be- comes hollow as it develops (Fig. 8d), but does not occupy all of the large sub-acro- somal space (Fig. Be). At this stage the complete structure is referred to as the ac- rosomal complex. With the formation of the acrosomal complex and cell elongation, a distinct centriolar-adjunct appears in the sper- matocyte (Fig. 9). It develops at the pos- terior nuclear pole adjacent to the forming basal body (Fig 9a). Unlike the situation in some other insect species (Gatenby & Tah- misian 1959, Breland et al. 1966) it does not appear to be forming around the cen- triole that is perpendicular to the axis of the flagellum. Instead, the basal body is located between the centriolar adjunct and the 'perpendicular' centriole; the ultimate fate of this second centriole is not clear. In transverse section, the axoneme is associ- ated with one mitochondrial derivative and the sectioned centriolar adjunct (which in the past has been interpreted as a darkened mitochondrial derivative). In longitudinal section the centriolar adjunct can be seen to lie between the posterior pole of the nucleus and one of the mito- chondrial derivatives (Fig. 9c), and is slightly more than 2|xm long. The mito- chondrial derivatives are thus offset lon- gitudinally, with the other member of the pair abutting the nucleus. This may result in the appearance of the adult mitochon- drial derivatives, where one member of the pair often seems to extend further pos- teriorly, possibly giving the erroneous im- pression that in fact the mitochondrial de- rivatives are of different lengths. This is a different arrangement from that previous- ly reported for other related species. Fur- thermore, anteriorly, it can be seen that for part of the length of the centriolar adjunct, where the centriolar adjunct and one of mitochondrial derivatives lie side by side, there is an extension of the centriolar ad- junct which surrounds the mitochondrial derivative on the side facing the centriolar adjunct (Fig. 9d). Interestingly in a rare ter- atological sperm-tail with two axonemes the centriolar adjunct was found to extend to enclose two mitochondrial derivatives. This contact is most evident in the region of the basal body and it is not clear if it actually extends into the flagellum proper as defined by the possession of both a ring of doublet microtubules and a central pair of microtubules. The change in shape of the nucleus (Fig. 10) that occurs with the appearance of peri-nuclear microtubules is also marked by the formation of two extra-nuclear elec- tron-dense regions (Fig. 10a, see also Fig. 6). Peri-nuclear microtubules are unevenly distributed and are not found over the ex- tra-nuclear electron-dense regions, which we have called lateral plates, nor do they occur in the area of the nucleus that will be convex during the elongation process. The lateral plates appear to be the locus for shape changes that occur during the elongation of the nucleus. Condensation of the chromatin into coiled fibrillar threads then follows (Fig. 10b). In trans- verse section, as the nucleus elongates, the threads are found associated with that por- tion of the nuclear membrane that is ad- jacent to the region lacking microtubules (Fig. 10c). The side of the nuclear mem- brane surrounded by the full complement cyte: (a) Large numbers of small mitochondria accumulate in one area of the spermatocvte; (b) mitochondrial fusion occurs to form the labyrinthine network of the nebenkern, which will itself then give the two mito- chondrial derivatives of the mature sperm. Scale bar a = 1.4 (j.m; b = 1.15 \xm. 30 n Journal of Hymenoptera Research 5b 4^ w ni '"^if'/ '':'t^ f ■■: 5c ...^/ r t Figs. 5-7. 5, Flagellum formation during spermiogenesis: (a) tlu' Iviscil hcidv (iiirowhciht) forms from the lateral centriole by insertion into a depression of the nucleus {nnow); (b) doublet microtubules {iinou') grow from the triplet microtubules of the basal body; (c) accessory tubules arc absent at this stage, although a central pair of microtubules form in the developing axoneme post-basal body; (d) the flagella/ mitochondria axis is de- veloped (arrow) before the change in shape of the nucleus is found, although polarity in the nuclear membrane can be found, n = nucleus. Scale bar a, b = 0.7 jjim; c = 0.8 jj.m; d = 3 (xm. 6, Polarity of the nucleus develops prior to shape change with increased electron density of the nuclear membrane including the appearance of two extra-nuclear densities (lateral plates). Microtubules (nrroivlicads) appear around the nucleus. Scale bar = 1 \xm. 7, Proacrosomal granule formed from small Golgi derived vesicles. Scale bar = 1 jim. Volume 7, Number 1, 1998 8a :##^:; 8 b / 31 _ l.t_t_>.,iiii> U.>.atod be- tween the plasma membrane and the nucleus (urrozc); (b) the acrosome indents to enfold the nucleus, and a small electron density is found anteriorly; (c) an acrosomal rod ('perforatorium' arrow) occupies the subacro- somal space formed during folding; (d) the perforatorium (arrpic) is seen in transverse section to be hollow; (e) the subacrosomal space itself can be seen to be large and is not filled by the perforatorium for its entire length, n, nucleus. Scale bar a, c, d = 0.5 \j.m; b, e = 0.65 (j.m. 32 Journal of Hymenoptera Research 9d -^ "^ *r\ >.'■-■ «v; V ' "yV^ ':M^ i?^wiK*r^ p:^^^^ 9c 1 K* ^^ ^ f ' 1 ^ B^ : ^^^ * „ -tf >*'! 1^ /|, r ^^H_L * ^■ 3 ■ ^ 1 r ■ '*■»■'«« ■^ ^-. Fig. 9. Centriolar adjunct formation during spermiogenesis: (a) arraii^unuiit nl iLiitriolar adjunct {arnnr, 'a'), basal body (anmu, 'b') and perpendicular centriole (iirroic, 'c'); (b) in transverse section the centriolar adjunct (nrrou') and a single mitochondrial derivative {arroiclwad) lie in parallel with the axoneme; (c) in longitudinal section the centriolar adjunct (c) abuts the nucleus anteriorly at the nuclear plate (largo nrrcw) and a mito- chondrial derivative (curved arrozv) posteriorly. Microtubules (arrcwf) are evident lining the nucleus (n) and the developing axoneme; (d) at the level of the basal body (arrows) the centriolar adjunct (iinoic/icHife) extends to contact and enclose the single mitochondrial derivative. Posteriorly to the centriolar adjunct two mitochon- drial derivatives are found {arrou'). Note also the absence of the central pair of microtubules in the basal bodies, n, nucleus. Scale bar a = 0.5 jjim; b = 0.3 ixm; c = 0.6 |j.m; d = 0.9 |j.m. VoLUMK 7, Number 1, 1998 33 10f 10d r--- 10e. *. ^I^B i .. ' ■. ■--• % - ,,: -; .^^^%l ^s^fc*^ •-**' ^ Ktm^J^i''* »i wi^^ :'m>^frm^'- Fig. 10. change in shape of nucleus during spermiogenesis: (a) extra-nuclear electon-dense region (lateral plate', arnnc) are produced upon nuclear shape-change; (b) condensation of the chromatin into coiled fibrillar threads (arwxv); (c) the threads divide and become associated with that portion of the nuclear membrane on either side of the region lacking microtubules (nnmi's); (d) the side of the nucleus with the full complement of microtubules {lar;^c arrows) flattens, and then infolds with the lateral plates (arrowheads) providing a locus for shape change; (e) the chromatin becomes a layer on the outer surface of a horse shoe (arrowheads) that infolds to surround elements of the endoplasmic reticulum; (f ) as the spermatocyte elongates the groove begins to disappear from the anterior and posterior ends; (g) the dense nuclear material then begins to redistribute more evenly around the nuclear membrane as the width decreases. Scale bar a = 40 |xm; b = 1 jxm; c, d, e = 0.6 |j.m; f = 1.2 |j.m; g = 1.7 |j.m. 34 Journal of Hymenoptera Research of microtubules then flattens (Fig. lOd ar- row), with the chromatin becoming a layer on the outer surface of the developing horse-shoe shape; as viewed in section (Fig. lOd arrowhead). The lateral plates be- come the linear portions of the outer sur- face of the arms of the 'horse-shoe' (Fig. lOe arrowheads); at the most extreme, the lateral arms (in transverse section) almost come into contact with one another. As the nucleus infolds, so elements of the endo- plasmic reticulum, which surround the microtubules, become largely enclosed by the nucleus (Fig. lOe). It is not clear how this highly folded nucleus reaches the cy- lindrical shape of the adult sperm. As the sperm elongates it appears that the in- verted side is flattened with the result that the grove begins to disappear from the an- terior and posterior ends (Fig. lOf). The dense nuclear material then begins to re- distribute more evenly around the nuclear membrane as the width decreases (Fig. lOg); however, intermediate stages are dif- ficult to identify. The axoneme itself develops into the 9 + 9 + 2 arrangement (Fig. 11a); 9 outer sin- gle accessory tubules, 9 doublets and 2 central single microtubules. Intratubular material is abundant (with radial spokes — Afzelius rays) and indications of the inner and outer dynein arms. Two deltoid bod- ies, (also referred to as triangular rods; Lensky et al. 1979) develop, but their exact derivation is not clear, although a mem- brane origin is apparent and a close as- sociation with the mitochondrial deriva- tives is probable. This is further supported by the observation that only one deltoid body is found at the level of the centriolar where there is only a single mitochondrial derivative. At the level of the two deltoid bodies is a single central rod, as previous- ly reported in ant spermatozoa (Wheeler et al. 1990). This arrangement is main- tained in mature sperm (Fig. lib). By com- parison with that of the spermatocyte, the nucleus of the mature sperm is homoge- neous (Fig. lie). The acrosome develops a distinctive glycocalyx (Fig. lie arroivheads) which extends over the anterior portion of the nucleus. The two mitochondrial deriv- atives show the offset arrangement that may be the result of the centriolar adjunct overlaying one of them anteriorly (Fig. lie arrows). In addition an end piece with no mitochondrial derivatives is found (Fig. lid arrow). DISCUSSION In the testes of imaginal male Aleiodes we have found structures indicative of early stages of sperm development. In particular, evidence for an abortive first meiotic division of the primary spermato- cyte, a major feature of hymenopterous in- sects in which males are haploid, has been found. This appears to involve anomalous metaphase plate formation, probably caused by the absence of a normal spindle. This has been previously described in the drone of the bee Scaptotrigona postica by Cruz-Landim & Beig (1980), where it is characterised by the presence of a concen- tric arrangement of the endoplasmic retic- ulum around the chromosomes and the absence of a normal spindle. In Aleiodes, as in S. postica, this is connected to an anomalous movement of spermatocyte centrioles, which, instead of assuming po- lar locations, migrate to the cell surface and are lost by cytoplasmic blebbing (Hoage & Kessel 1968). The presence of such early stages of development in the adult supports our supposition that gonial development in this species could to be re- lated to the life style of the adult. The rel- atively long adult life time of the male would seem to obviate the need for syn- chronous maturation of all the individual sperm cysts. The manchette of microtubules that as- sembles around the developing spermatid nucleus appear to be important for nucle- ar compression and elongation (Baccetti 1972). The nuclear shape change found during development in Aleiodes, and the asymmetric distribution of microtubules. Volume 7, Number 1, 1998 35 •11a 5i^>•^^•^ ' '■■liiiS^Sife^ %^^H^' lib/- | ■*f r^ 'h t* Fig. 11. Axoneme structure and the mature sperm:(a) The arrangement of 9 outer, singlet accessory micro- tubules (at), 9 doublet (d) and two single central microtubules (t) is evident in the later stages of spcrmiogen- esis, with intratubular material (i) and prominent radial spikes (s); (b) two deltoid bodies {arroivs) develop in association with the mitochondrial derivative except at the level of the centriolar adjunct where only one is found; (c) the mature sperm illustrates the uneven length of the mitochondrial derivatives (arrow) and the glycocalyx that develops around the mature acrosome {arrowhmds); (d) the tail piece at the posterior of the sperm (arrow) contains only the 9+2 arrangement of doublet and single microtubules. Scale bar a = 100 |a.m; b = 0.25 (xm; c = 0.75 (xm; d = 0.4 jjLm; p = perforatorium. have not previously been described for parasitic wasps. A gutter-shaped devel- opmental stage of the nucleus occurs in some bugs (Lee & Lee 1992) but nuclear elongation is not accompanied by concav- ity in many other insects (e.g. SzoUosi 1975, Friedlander 1993, Wolf & Joshi 1995). In some caddis flies, in which the 36 Journal of Hymenoptera Research spermatid nucleus transiently assumes a sickle shape, the microtubular manchette surrounding the nucleus is interrupted (Wolf & Klein 1995), as it is in the drag- onfly, Aeshna grandis L., in which the elon- gating nucleus becomes locally com- pressed (Kessel 1966). The structures we have termed lateral plates, which occur between groups of microtubules of the mi- crotubular manchette, do not appear to have been described previously in any in- sect. The present study of spermiogenesis in Aleiodes has implications for published studies of the centriolar adjunct. The pres- ence of a centriolar adjunct overlying only one of the mitochondrial derivatives has not to our knowledge previously been re- ported. In particular the connection from the centriolar adjunct to the mitochondrial derivative that abuts the nucleus is novel. This arrangement may be responsible for observations in related species suggesting that anteriorly there may be a particularly electron dense mitochondrial derivative (Quicke et al. 1992) or a region of overlap between the axoneme, nucleus and mito- chondrial derivative (Chauvin et al. 1987) — i.e. some studies may have misin- terpreted the centriolar adjunct as a par- ticularly electron dense mitochondrial de- rivative. Further, the longitudinal offsetting of the mitochondrial derivatives caused by the interposition of the centriolar adjunct between one of them and the nucleus could be responsible for the appearance of some sections through the posterior re- gion of the sperm, just before the tail piece, in which only one mitochondrial derivative is found. The presence of a sin- gle mitochondrial derivative in sections through this region in other species prompted previous authors to conclude that mitochondrial derivatives are of dif- ferent lengths, which has even been con- sidered to be a characteristic, not just of the Hymenoptera, but of all holometabo- lous insect orders (Wheeler et al. 1990). However this result could be explained simply by the arrangement of the centrio- lar adjunct. In the light of this finding in Aleiodes, it would be interesting to re-ex- amine these previously reported groups for the presence of a similarly located cen- triolar adjunct. A large centriolar adjunct has been reported in ant spermatozoa (Wheeler et al. 1990), and was thought to distinguish this group from other Hyme- noptera, such as bees, which appear to lack a homologous structure (Lensky et al. 1979, Cruz-Landim & Beig 1980). The ar- rangement of the centriolar adjunct in the ant spermatozoa, at least in longitudinal section, is similar to that of Aleiodes. How- ever, in these the centriolar adjunct can be seen to overlie both mitochondrial deriv- atives. The connection to the mitochondrial de- rivative is particularly interesting as it oc- curs at the level of the basal body where the two central microtubules are absent. The origin and exact function of the cen- tral pair is not known. A structural role for the central pair could imply a similar role for the centriolar adjunct and the lat- ter could therefore be compensating for the absence of the central pair at the re- gion of the basal body. However, some studies have provided evidence that the centriolar adjunct consists partly of RNA (Baccetti et al. 1969) and how this would be related to a structural role is uncertain. The exact relationship between centri- oles, the basal body, the centriolar adjunct and the developing axoneme in Aleiodes is not clear. In another parasitic wasp, the pteromalid Nasouia vitripeunis (Walker), the distal centriole is reported to lengthen to form the basal body of the flagellum, and then, with the proximal centriole, form the ring centriole (Hogge & King 1975). This then associates with a spherical reticulate dense body, which forms adja- cent to the nuclear membrane opposite a nuclear pore at the secondary spermato- cyte stage. As maturation proceeds the proximal centriole is reported to become surrounded by granular material and Volume 7, Numbhr 1, 1998 37 break down, with loss of material from the spherical, reticulate dense-body. The gran- ular material, with presumably the re- mains of the proximal centriole is then thought to contribute to the centriolar ad- junct. However, in Aleiodes such a straight- forward relationship is not clear; in fact it appears as if the centriolar adjunct may develop prior to loss of the proximal cen- triole. Thus this may not contribute to the formation of the centriolar adjunct at all. In discussions of the centriolar adjunct in ant spermatozoa, Wheeler et al. (1990) suggest caution when applying this term, as the developmental homology to such structures in other species, such as mam- mals, has not been proven. ACKNOWLEDGEMENTS The authors gratefully acknowledge the invaluable assistance of Mr Ian Morris (Electron Microscopy Unit, Biology Department, Imperial College). This work was supported by the NERC Initiative in Tax- onomy. LITERATURE CITED Baccetti, B. 1972. Insect sperm cells. Advances in Insect P/ji/sio/iW/ 9: 315-397. Baccetti, O. P., R. Dallai, and F. Rosati. 1969. The sper- matozoon of Arthropoda III. The lowest holomet- abolic insects. Journal of Microscop}/ 8: 233-248. Breland, O. P., C. D. Eddleman and J. J. Biesele. 1968. Studies of Insect Spermatozoa. Entomological News 79: 197-216. Chauvin, G., M. El Agoze, C. Hamon and J. Huig- nard. 1987. Ultrastructure des spermatozoides di- ploides de Dimlroinus inilchellus Wesmael (Hy- menoptera: Ichneumonidae). International ]onrnal of Insect Morphology and Embryology 17: 358-366. Cruz-Landim, C. and D. Beig. 1980. An electron mi- croscopical study of spermatogenesis in the drone of Scaptotrigona postica (Hymenoptera, Ap- idae). International Journal of Invertebrate Repro- duction 2: 272-283. Friedlander, M. 1993. Phylogenetic position of rhy- acophiloid caddisflies (Insecta, Trichoptera): a spermatological analysis of Rliyacophilidae and Glossosomatidae. Zoologica Scripta 22: 299-304. Gatenby, J. B. and T. N. Tahmisian. 1959. Centriolar adjunct, centrioles, mitochondria and ergosto- plasm in orthopteran spermatogenesis. An elec- tron microscopical study. La Cellule 60: 103-134. Hoage, T. R. and R. G. Kessel. 1968. An electron mi- croscopical study of the process of differentiation during spermatogenesis in the drone honey bee with special reference to replication and elimi- nation. Journal of Ultrastructure Research 24: 6-32. Hogge, M. A. F. and P. E. King. 1975. The ultrastruc- ture of spermatogenesis in Nasonui vitripennis (Walker) (Hymenoptera: Pteromalidae). journal of Suhmicroscopical Ci/tolog}/ 7: 81-996. Jamieson, B. G. M. 1987. The Ultrastructure and Phy- logeny of Insect Spermatozoa. Cambridge Univer- sity Press, Cambridge, 320pp. Kessel, R. G. 1966. The association between microtu- bules and nuclei during spermiogenesis in the dragonflv. Journal of Ultrastructure Research 16: 293-304. Lee, Y. H. And C. E. Lee. 1992. Ultrastructure of sper- matozoa and spermatogenesis in Nepomorpha (Insecta: Heteroptera) with special reference to phylogeny. Zoological Science 9: 971-981. Lensky, Y., E.Ben-David and H. Schindler. 1979. Ul- trastructure of the spermatozoan of the mature drone honeybee. Journal of Apiculture Research 18: 264-271. Needham, J. G., ]. R. Traver, and Y. C. Hosu. 1935. The Biology of Mat/flies. Comstock Publishing As- sociates, Ithaca, New York. Phillips, D. M. 1970. Insect sperm: their structure and morphogenesis. Journal of Cell Biology 44: 243- 277. Quicke, D. L. J., S. N. Ingram, H. S. Baillie and P. V. Gaitens. 1992. Sperm structure and ultrastructure in the Hymenoptera (Insecta). Zoologica Scripta 21: 381-402. Ross, H. H. 1944. The caddis flies or Trichoptera of Illinois. Illinois Natural History Survey Bulletin 23: 1-326. Shaw, M. R. and T. Huddleston. 1991. Classification and biology of braconid wasps (Hymenoptera: Braconidae). Handbooks for the Identificatum of Brit- ish Insects 7(11): 1-126. Szollosi, A. 1974. Ultrastructural study of the sperma- todesm of Locusta migratoria migratoroides (R.F): ac- rosome and cap formation. Acrida 3: 175-192. Szollosi, A. 1975. Electron microscope study of spermi- ogenesis in Locusta migratoria (Insecta Orthoptera). Journal of Ultrastructure Research 50: 322-346. Wheeler, D. E., E. G. Crichton and P. H. Krutzsch. 1990. Comparative ultrastructure of ant sperma- tozoa (Formicidae: Hymenoptera). /oMniii/ of Mor- phology 206: 343-350. Wolf, K. W. and H. C. Joshi. 1995. Microtubule or- ganization and the distribution of 7-tubulin in spermatogenesis of a beetle, Tenehrio nuttitor (Te- nebrionidae, Coleoptera, Insecta). Journal of Celt ScH-nce 108: 3855-3865. Wolf, K. W. and C. Klein. 1995. Development of the sperm head in the caddis fly Potainophylax rotun- dipennis (Insecta, Trichoptera). Zoomorphology 115: 109-115. J. HYM. RES. Vol. 7(1), 1998, pp. 38-47 Revision of the American Tiphiid Genus Quemaya Pate (Hymenoptera: Tiphiidae: Brachycistidinae) Lynn S. Kimsey and Marius S. Wasbauer Bohart Museum of Entomology, Department of Entomology, University of California, Davis, CA 95616, USA Abstract. — The brachycistidine genus Quemaya is reviewed and six new species are described from southern California, Arizona, northern Mexico and Costa Rica, confusa, costaricensis, eu- rekaensis, megalops, mexicana, and sonorensis. The genus Quemaya Pate (1947) is an ob- scure group of nocturnal tiphiids original- ly known from five species described from the southwestern deserts of North Amer- ica, as far south as Sonora, Mexico. In this paper we more than double the known species in this genus, and extend the known distribution hundreds of miles. Additionally, intensive collecting in Costa Rica by the Instituto Nacional de Biodi- versidad and by Frank Parker has turned up a new species from the arid north- western region of Costa Rica. This region appears to be the southernmost extension of this arid nearctic desert habitat. These wasps are easily overlooked be- cause of their small size, dark coloration and nocturnal habits. However, males are strongly attracted to ultraviolet light sources at night. Females are as yet unknown. The new Costa Rican species, costaricensis, has some characteristics unusual for Quemaya, including the tiny ocelli and sculptured pro- podeum, but still has the diagnostic features for the genus discussed below. Quemaya is characterized by a combina- tion of primitive and derived features. The wing venation is reduced (Figs. 10-17). The forewing has only one, or less commonly two, submarginal cells, one discoidal and one subdiscoidal cell, and the marginal cell is separated from the costal wing margin. In the hindwing the cubital vein is nearly straight meeting the transverse cubital vein at an angle much greater than 135°. Que- maya species lack the ventrally "tailed" an- tennal socket carina, forecoxal stridulatory structure, scrobal sulcus, mandibular cari- na, basal "ring" carina of the first gastral tergum, and carinate gastral sterna char- acteristic of other brachycistidine genera. Members of Quemaya all have a distinctly modified epipygium, that has sublateral carinae and an emarginate, truncate or con- vex apex (Figs. 18-21). MATERIALS Specimens in this study were obtained from: DAVIS — Bohart Museum of Ento- mology, University of California, Davis (S. L. Heydon); ITHACA — Department of En- tomology, Cornell University, Ithaca, New York (J. Liebherr); LAWRENCE— Snow Entomological Museum, University of Kansas, Lawrence (R. Brooks); LOGAN — Department of Entomology, Utah State University, Logan (T. Griswold); OTTA- WA— Canadian National Insect Collec- tion, Agriculture Canada, Ottawa (L. Mas- ner); RIVERSIDE— Department of Ento- mology, University of California, River- side (S. Triapitsyn); SAN FRANCISCO— California Academy of Sciences, San Fran- cisco (N. Penny); SANTO DOMINGO— Instituto Nacional de Biodiversidad, Santo Domingo de Heredia, Costa Rica (C. M. Rodriguez); TUCSON — University of Ari- zona, Tucson; WASHINGTON— U.S. Na- Volume 7, Number 1, 1998 39 tional Museum of Natural History, Smith- Two abbreviations are used below for sonian Institution, Washington, D. C. (A. the sake of brevity; MOD = midocellus di- S. Menke, K. V. Krombein). ameter and F = flagellomere. KEY TO SPECIES OF QUEMAYA (MALES) 1 Forewing with two submarginal cells (Figs. 10, 15, 16) 2 - Forewing with one submarginal cell (Figs. 11-14, 17) 5 2 Mandible with two apical teeth; distance between midocellus and closest eye margin less than 1.5 MOD (Fig. 1) arenicola Wasbauer - Mandible with three apical teeth, one may be very small; distance between midocellus and closest eye margin more than 1.5 MOD 3 3 Mesopleuron densely punctate, with punctures evenly dispersed over entire surface and 1-2 puncture diameters apart; propodeum coarsely sculptured with irregular demarcation between dorsal and posterior surfaces indicated by rugae confusa, new species - Mesopleuron sparsely punctate, with punctures generally 4—6 puncture diameters apart, denser dorsally than posteroventrally, surface above midcoxa nearly impunctate; propo- deum smooth, without distinct punctation or rugosities, and no demarcation between dorsal and posterior surfaces 4 4 Midocellus separated from eye margin in frontal view by more than 2 MOD; epipygium apically bidentate perpunctata (Cockerell) - Midocellus separated from eye margin in frontal view by more than 1 MOD but less than 2 MOD (Fig. 6); epipygium apically flat and broadly triangular (Fig. 19) ... megalops, new species 5 Gular carina with basal tooth-like projection near mandible (Fig. 25); clypeus transversely indented, without medial projection, arcuately raised apicomedially (Fig. 5); epipygium apex truncate (as in Fig. 21) inermis (Malloch) - Gular carina simple; clypeus medially bulging or with medial tooth; epipygium various 6 6 F-1 and II subequal in length, about twice as long as broad or longer; epipygium apex strongly convex and lip-like (Fig. 20) ntarcida (Bradley) - F-I shorter than II, and between 1.2 and 1.8x as long as broad; epipygium apex medially emarginate or truncate 7 7 Midocellus larger, separated by 2 MOD or less from nearest eye margin; F-I 1.6X as long as broad or longer 8 - Midocellus small, separated by 3 or more MOD from nearest eye margin; F-I 1.5x as long as broad or shorter 9 8 Midocellus separated from nearest eye margin by 1.5 MOD or less; epipygial apex convex; clypeus with broad, blunt medial projection, bulging and strongly subtriangular in profile; forewing Rl vein strongly angulate near costal margin, marginal cell broadly parallel- sided (as in Fig. 15) mexicana, new species - Midocellus separated from nearest eye margin by 1.8-2.0 MOD; epipygial apex bidentate; clypeus with narrow noselike or almost digitate medial projection; forewing Rl vein nar- rowly separated from stigma, curved or indistinguishable near costal margin paupercula (Bradley) 9 Midocellus five or more MOD from nearest eye margin (Fig. 3); F-I 1.2x as long as broad; F-II 1.4X as long as broad; epipygium apex truncate (as in Fig. 21); propodeum coarsely sculptured, with irregular, partial transverse carina (Fig. 24) ... . costaricensis, new species - Midocellus 3.0-3.5 MOD from nearest eye margin; F-I 1.5x as long as broad; F-II 1.7- 1.9 X as long as broad; epipygium apex truncate or emarginate; propodeum smooth with- out sculpturing or transverse carina 10 40 Journal of Hymenoptera Research 10 Clypeus with small, sharp medial projection, apical truncation 1.5 MOD wide (Fig. 4); epipygial apex medially emarginate, with sublateral carina eurekaensis, new species - Clypeus with broad, transverse medial projection subtended by discrete declivity, apical truncation 2 MOD wide (Fig. 9); epipygial apex truncate sonorensis, new species Quetnaya arenicola Wasbauer (Figs. 1, 10, 22) Quemaya arenicola Wasbauer 1967:169. Holotype male; California: Inyo Co., 6 mi w Glamis (SAN FRANCISCO, type No. 9306). Male. — Body length 4-6 mm; clypeus with narrow apical truncation 0.6 MOD wide (Fig. 1); mandible with two apical teeth; gular carina simple; F-I and II length twice breadth; interantennal distance 0.3 MOD wide; midocellus separated from nearest eye margin by 1.1-1.2 MOD; fore- wing with two submarginal cells, second cell triangular or subquadrate, completely underlying the first (Fig. 10); mesopleural punctures 0.2-0.5 puncture diameter apart; epipygium with sublateral carina 4. eurekensis 5 inermis 7. marcida ^ mexicana Figs. 1-9. Front view of m,ik' fan', k'ft iintcnna removod. 6 megalops 9. sonorensis Volume 7, Number 1, 1998 41 10. arenicola 12. eurekaensis 14. marcida 16. perpunctata 22. arenicola 23. marcida 1 1 . costaricensis 13. inermis 15. megalops 17. sonorensis 24. costaricensis 18. confusa 19. megalops 20. marcida 21. eurekaensis 25. inermis Figs. 10-25. Figs. 10-17. Forewing venation. Figs. 18-21. Apex of epipygium (last gastral tergum). Figs. 22- 23. Lateral view of male genital capsule. Fig. 24. Posterior view of male propodeum. Fig. 25. Oblique posterior view of male head, showing genal tooth. 42 Journal of Hymenoptera Research ending in an apical lobe, apical margin medially emarginate; paramere subapi- cally expanded, apical margin linear ter- minating in acute ventral angle (Fig. 22). Body color reddish brown; antennae pale red; wing veins clear and at most faintly tinted with dark brown stigma. Material examined. — 84 specimens from California: Imperial, San Bernardino, Riv- erside and San Diego Co.; Arizona: Yuma Co.; Mexico: Baja California Norte and So- nora. Discussion.— This is one of the few Que- maya species with two submarginal cells. It can be immediately distinguished from the others with similar venation by having only two mandibular teeth and large ocelli narrowly separated from the eye margin. Quemaya confiisa Kimsey and Wasbauer, new species (Figs. 2, 18) Male. — Body length 4-6 mm long; face (Fig. 2): clypeus medially projecting and nose-like in profile, apical truncation me- dially emarginate or slightly trilobate and 1.3 MOD wide; mandible with three apical teeth, two subsidiary teeth considerably smaller than apical one; gular carina sim- ple; F-I and F-II 1.7x as long as wide; dis- tance between midocellus and closest eye margin 2 MOD; interantennal space 0.5 MOD wide; mesopleural punctures 0.5- 1.0 puncture diameter apart; forewing with one submarginal cell; marginal cell rhomboid, apically broad and Rl angu- late; epipygium apicomedially emargin- ate, with two sublateral carinae forming an ovoid, slightly sunken, medial area and terminating in apical lobes (Fig. 18); par- amere nearly parallel-sided, only slightly widened subapically before acute ventral angle. Body color black to dark brown; leg brown; antenna bicolored, paler ventrally than dorsally; wing veins pale brown-tint- ed, stigma dark brown; wing membrane untinted. Etymology. — This species is named "confusa" because of the strong similarity between it and other species of Quemaya found in the same localities. Typw material. — Holotype male: MEXI- CO: Sonora, 6 km nnw San Carlos, 11-15 July 1983, E. Fisher, malaise trap (DAVIS). Paratypes, 55 males (DAVIS, SAN FRAN- CISCO, RIVERSIDE, LAWRENCE, OT- TAWA, WASHINGTON, TUCSON): two— USA: CALIFORNIA: Riverside Co., Blythe, 21 June 1963, F. D. Parker and L. A. Stange; one — Deep Canyon, 24 June 1964, E. I. Schlinger; one— Millard Cyn., 20 June 1963, E. 1. Schlinger; two — Winches- ter, 10 July and 14 Aug. 1967, W. Icenogle; four — Imperial Co., 3 mi n Glamis, 15—16 Sept. 1977, M. Wasbauer & A. Hardy; one— HoltviUe, 8 July 1965, R. A. Flock; three — Chocolate Mts., Ogilby Rd., 3 mi s Jet. Hwy. 78, 16-21 Oct. 1977, M. Was- bauer; five — San Bernardino Co., 10 mi w 29 Palms, 27 May 1966, M. Wasbauer; two— NEW MEXICO: Dona Ana Co., Las Cruces, 2 June 1965, R. M. Bohart; one — 2 mi. e Mesilla; ten— ARIZONA: Pinal Co., Marana, 6 July 1955, Butler & Werner; one — Maricopa Co., Gila Bend, Bohart & Butler, 12 Aug. 1954; one — Wickenburg, 10 Aug. 1950, H. K Lloyd; one— Pinal Co., 10 mi. s Toltec, 21 June 1953, T. R. Haig; nine — 4 mi. se Casa Grande, 18 June 1964, Smith & Baker; one — Pima Co., Continen- tal, 17 July 1966, H. K. Court; one— TEX- AS: Brewster Co., Chisos Mts. 10-12 June 1908, Mitchell & Cushman; one— MEXI- CO: Sonora, Cocorit, 23 May 1968, Parker & Stange; one — 11 June, F. D. Parker; two— 18 mi. e El Puerto, 7 Aug. 1960, Ar- naud, Ross & Rentz; four — 40 mi. n Her- mosillo, 8 Aug. 1960, Amaud, Ross & Rentz; one — 6 km nnw San Carlos, 11-15 July 1983; one — 4.9 mi n Magdalena, Rt. 15, 25 Aug. 1964, M. E. Irwin. Discussion. — Quemaya confusa is another species with two submarginal cells. It can be distinguished from other species with this kind of wing venation by the densely and evenly punctate mesopleuron, coarse- ly sculptured propodeum, and mandibles with the normal three apical teeth. This Volume 7, Number 1, 1998 43 species is probably most similar to perpiiii- cata and megalops. Quemaya costaricensis Kimsey and Wasbauer, new species (Figs. 3, 11, 24) Male. — Body length 4.0-5.5 mm; face (Fig. 3): clypeus flat in profile, with apical truncation 2.5 MOD wide; mandible with three apical teeth, preapical teeth subequal in length; gular carina simple; F-I length 1.2X breadth; F-II 1.4x as long as broad; interantennal distance 0.9 MOD wide; dis- tance from midocellus to closest eye mar- gin 5.7 MOD; mesopleuron with punc- tures 1-2 puncture diameters apart; fore- wing with one submarginal cell; marginal cell small, more than 3x as long as broad, widely separated from costal margin (Fig. 11); propodeum coarsely rugose dorsally, with broken irregular transverse ridge separating dorsal from posterior surfaces (Fig. 24) epipygium apicomedially trun- cate, with two sublateral carinae above apex forming an ovoid, slightly sunken medial area; paramere gently tapering, apicoventral angle narrowly rounded. Body color black; antenna bicolored, paler ventrally than dorsally; wing veins pale brown, stigma dark brown; wing mem- brane slightly brown tinted. Etymologi/. — This species is named after the country of collection, which is the southernmost record for the tiphiid sub- family Brachycistidinae. Ti/pe material. — Holotype male: Costa Rica: Guanacaste Prov., 14 km s Caiias, F. D Parker, 17 Feb. 1989 (LOGAN). Para- types, 52 males (DAVIS, LOGAN, SANTO DOMINGO): eight— same data as holo- type; one: 4-5 Mar. 1989; two— 24 Feb. 1989; one— 28-29 Jan. 1989; three— 18 Feb. 1989; three— 28 Feb. 1989; one— 9 Mar. 1989; one— 11-13 Jan. 1990; one— 16 Feb. 1989; one— 15-24 Feb. 1990; one— 5 Mar. 1989; two— 1-11 Feb. 1990; one— south of Cafias, 9-14 Feb. 1989, F. D. Parker; one— 25-28 Feb. 1989; one— 25 Feb.-8 Mar. 1989; twenty-one — Santa Rosa Natl. Pk., 21 Feb.-ll-14 Mar. 1981. Discussion. — There are many distinctive features of this species, which will imme- diately separate it from all other Quemaya, including: the tiny ocelli and small eyes, coarsely sculptured propodeum and short, broad flagellomeres. It is closest to sono- reiisis and eurekaensis based on the small ocelli, single submarginal cell and short broad flagellomeres. Quemaya eurekaensis Kimsey and Wasbauer, new species (Figs. 4, 12, 21) Male. — Body length 3.5^.0 mm; face (Fig. 4): clypeus with small, sharp medial projection, apical truncation 1.5 MOD wide; mandible with three apical teeth; gular carina simple, without tooth-like projection; F-1 length 1.5x breadth; F-II length 1.9 X breadth; interantennal dis- tance 0.5 MOD wide; midocellus separat- ed from nearest eye margin by 3.2 MOD; mesopleural punctures 3-6 puncture di- ameters apart; forewing with one submar- ginal cell (Fig. 12); epipygium with sub- lateral carinae each ending in an apical lobe, epipygial apex medially emarginate (Fig. 21); gonostylar shape apically nar- rowed into single acute apical angle. Body dark reddish brown, antennal and leg col- or pale reddish brown; wing veins untint- ed, except stigma dark reddish brown. Etymology. — This species is known pri- marily from Eureka Valley, thus the name. Type material. — Holotype male: Califor- nia: Inyo Co., Eureka Valley, 13 July 1975, F. Andrews & A. Hardy (DAVIS). Para- types, eight males (DAVIS, SAN FRAN- CISCO); six — same data as holotype; one — 19 June 1972 Derham & Guiliani; one — Saline Valley dunes, 30 km e Inde- pendence, 26 May 1993, D. E. Russell, malaise trap. Discussion.— The diagnostic features for eurekaensis are the small ocelli, separated from the eye margin by more than 3 MOD, medially emarginate epipygium 44 Journal of Hymenoptera Research apex, and short F-I and II. This species is closest to sonorensis and less so costaricen- sis, but can be immediately distinguished from sonorensis by the sharp clypeal pro- jection and emarginate epipygium. The larger ocelli and smooth propodeum will separate eurekae)isis from costaricensis. Quemaya inermis (Malloch) (Figs. 5, 13, 25) Brachycistis inermis Malloch 1924:23. Holotype male; Arizona: Higley (WASHINGTON). Male. — Body length 2>-^ mm; face (Fig. 5): clypeus transversely indented, without medial projection, apical truncation 0.9 MOD wide; mandible with three apical teeth, preapical teeth subequal; gular ca- rina with tooth-like projection near base of mandible (Fig. 25); forewing with one sub- marginal cell (Fig. 13); F-I and II length 1.7X breadth; interantennal distance 0.8 MOD wide; midocellus separated from eye margin by 2 MOD; mesopleuron an- terior part with punctures 1-3 puncture diameters apart, posteriorly nearly im- punctate and polished; epipygium with- out sublateral carinae, only slightly in- dented apicomedially; paramere gently ta- pering, apicoventral angle narrowly rounded. Body color reddish brown; an- tenna paler than body; wing veins and stigma nearly colorless. Material examined. — 239 specimens from: California: Riverside, San Bernardino, and Imperial Cos.; Arizona: Yuma, Santa Cruz and Maricopa Cos. Nevada: Lincoln Co.; Mexico: Sonora and Baja California Norte. Discussion. — This is the only Quemaya species with a tooth-like projection on the gular carina; a feature typically found in members of the genus Brachycistis. Que- maya inermis can be immediately recog- nized by the transversely medially indent- ed clypeus, and the apically truncate epi- pygium. Quemaya marcida (Bradley) (Figs. 7, 14, 20, 23) Bracln/cisttfi marcida (Bradley) 1917:283. Holo- type male; USA: California, Imperial Co. (ITHACA). Male. — Body length 3-5 mm; face (Fig. 7): clypeus with sharp medial projection, appearing nasiform in profile, apical trun- cation 1.7 MOD wide; mandible with three apical teeth; gular carina simple; F-I and F-II twice as long as broad; interantennal distance 0.6 MOD wide; midocellus sepa- rated from eye margin by 2.5 MOD; meso- pleuron sparsely punctate, punctures 4-6 puncture diameters apart or more; fore- wing with one submarginal cell; marginal cell nearly parallel-sided, Rl vein strongly angulate and joining stigma at or below costal margin (Fig. 14); epipygium apico- medially rounded and lip-like, with sub- lateral carina beginning above lip (Fig. 20); paramere apically truncate with apical margin concave (Fig. 23). Body reddish brown to brown; antenna and legs paler reddish brown than body color, forewing veins pale brown tinted; stigma brown. Material examined. — 873 specimens from CALIFORNIA: San Bernardino, Riverside and Imperial Cos.; Arizona: Yuma and Co- chise Cos.; Nevada: Nye; Mexico: Sonora. Discussion. — As with the majority of Quemaya species, marcida has one submar- ginal cell in the forewing and a simple gu- lar carina. However, of these species only marcida, arenicola and perpuncata have the first two flagellomeres twice as long as broad. Quemaya marcida is the only one that has the epipygial apex strongly con- vex and lip-like. Additionally, the ocelli are large and narrowly separated from the nearest eye margin. Quemaya megalops Kimsey and Wasbauer, new species (Figs. 6, 15, 19) Male. — Body length 4-5 mm; face (Fig. 6): clypeus with large tuberculate medial projection, strongly nasiform in profile, apical truncation 0.9 MOD wide; mandible with three apical teeth, subsidiary ones considerably shorter than primary tooth; gular carina simple; F-1 and 11 length 1.9- 2.0 X breadth; interantennal distance 0.5 MOD wide; midocellus separated from eye Volume 7, Number 1, 1998 45 margin by 1.6 MOD; mesopleuron nearly impunctate; forewing with two submargin- al cells, the second large, nearly rectangular and Rs aligned with 2rs-m; marginal cell large and parallel-sided with Rl vein strongly angulate before joining stigma at costal margin (Fig. 15); epipygium apically truncate, with distinct sublateral carinae, flat medially (Fig. 19); paramere lanceolate, narrowly tapering apically. Body reddish brown to brown with yellow legs and an- tenna; wing veins and stigma transparent and lightly yellow tinted. Ehpiiologi/. — mega = large, ops = eyes; i. (Gr.). The name refers to the greatly en- larged eyes and ocelli. Ti/pe material. — Holotype male: Califor- nia: Riverside Co., 5 mi nw Desert Center, 22 Oct. 1955, M. Wasbauer (DAVIS). Para- types: 7 males (DAVIS, RIVERSIDE, SAN FRANCISCO): one — same data as holo- type; two: Indio, Aug. 1977, Allen and Duffy (DAVIS); one — Magnesia Cyn. 2 July 1952; one — Inyo Co., 7 mi nne Pana- mint Spr., 15 May 1969, P.Rude; one — Ar- izona: Yuma Co., 4 mi w Salome, 8 June 1958, D & G. MacNeill; one — Mohawk, Apr. 1963, Timberlake. Discussion. — This species most closely resembles perpuncata as both have two submarginal cells, a sparsely punctate mesopleuron and smooth propodeum. Quemaya megalops can be distinguished from perpuncata by the much larger ocelli, separated from the nearest eye margin by less than two midocellus diameters, pale stigma, and the very distinctive epipy- gium, which appears broadly triangular with the apex narrowly truncate. The wing venation is also diagnostic, with the second submarginal cell much larger than in other Quemaya species. Quemaya mexicana Kimsey and Wasbauer, new species (Fig. 8) Male. — Body length 2.5-4.5 mm; face (Fig. 8): clypeus with broad, blunt medial projection, bulging and strongly subtrian- gular in profile, apical truncation 1 MOD wide; mandible with three apical teeth; gular carina simple; F-I-II 2.0-2.4 X as long as broad; interantennal distance 0.4 MOD wide; midocellus separated from eye margin by 1.5-2.0 MOD; mesopleuron sparsely punctate, punctures 4-6 puncture diameters apart or more; forewing with one submarginal cell; marginal cell nearly parallel-sided, Rl vein strongly angulate and joining stigma at or below costal mar- gin; epipygium apicomedially convex, otherwise similar to sonorensis; gonostylar apex truncate, with apical margin linear. Body brown to dark brown; antenna and legs paler reddish brown than body color, forewing veins pale brown tinted; stigma brown. Etymology. — The name refers to the pre- ponderance of specimens having been col- lected in northern Mexico. Material examined. — Holotype male: MEXICO: Baja California Norte, 1 km s El Rosario, 24-25 July 1992, D. E. Russell, MT (DAVIS). Paratypes: 46 males (DAVIS): 37— Diablo Cyn, Dry Lake, 16 July 1979, D. Giuliani; four — same data as holotype; three — eastern base of Sierra de Juarez be- low Rumorosa, 11 Sept. 1961, I. L. Wig- gins; one — USA: California, Andrade, 4 Aug. 1966, M. Wasbauer; one — Calexico, 19 June 1969. Discussion. — This species closely resem- bles marcida as discussed under that spe- cies. However, mexicana can be readily distinguished from marcida by the shorter flagellomeres, broad blunt clypeal projec- tion, and distinctively convex epipygial apex. Quemaya pauperciila (Bradley) Brachycistis pauperciila Bradley 1917:282. Holo- type male; California: Calexico (ITHACA, type No. 129.1) Male. — Body length 3-4.5 mm; clypeus with medial nose-like projection in profile, apical truncation 1.1 MOD wide; mandible with three apical teeth; gular carina sim- 46 Journal of Hymenoptera Research pie; F-I and II length 1.6X breadth; inter- antennal distance 0.3 MOD wide; mido- cellus separated from eye margin by 1.9 MOD; mesopleural punctures almost ab- sent; forewing with one submarginal cell; marginal cell narrow, closed on costal margin; epipygial apex truncate, subapi- cally concave; paramere nearly parallel- sided, apex abruptly truncate, with acute ventral and dorsal angle and apical mar- gin somewhat concave between. Body col- or dark brown; antenna paler than body; wing veins nearly colorless, except stigma dark brown. Material examined. — 109 specimens from California: Riverside, Imperial, Kern, San Bernardino and San Diego Co.; Arizona: Coconino, Yuma Co.; Nevada: Lincoln Co.; Texas: Brewster and and Presidio Co.; New Mexico: Dona Ana Co.; Mexico: Baja California Norte and Sonora. Discussion. — Queniaya paupercula is one of the most commonly collected species of the genus. It is most readily confused with eurekaensis, which has a very restricted distribution. Both species have a medially emarginate epipygial apex, short flagel- lomeres, two submarginal cells and a sim- ple gular carina. However, the larger size of the ocelli will readily separate pauper- cula from eurekaensis. Quemaya perpunctata (Cockerell) (Fig. 16) Brachycistis perpunctata Cockerell 1896:291. Ho- lotype male; New Mexico: Las Cruces (PHIL- ADELPHIA). Male. — Body length 4.0-5.5 mm; clypeus with low medial projection, apical margin 1 MOD wide; mandible with three apical teeth; gular carina simple; F-I length twice breadth; F-II 2.2X as long as broad; inter- antennal distance 0.5 MOD wide; mido- cellus separated from eye margin by 2.8 MOD; forewing with two submarginal cells, second cell triangular or subquad- rate, completely underlying the first (Fig. 16); mesopleural punctures 1-2 puncture diameters apart, nearly impunctate above midcoxa; epipygium with sublateral cari- nae each ending in an apical lobe, epipy- gial apex medially emarginate; paramere expanded subapically, apex truncate, api- cal margin linear, between two acute an- gles. Head and body dark brown to black, except antennae, clypeal apex, palpi and tegula reddish brown, mandibles and tarsi yellow, femora and tibiae brown; stigma dark brown, nearly black, veins brown. Material examined. — 249 specimens from California: San Bernardino, Riverside, Im- perial and San Diego Cos.; Nevada: Lin- coln Co.; Texas: Brewster Co.; Arizona: Yuma Co.; Mexico: Baja California Sur. Discussion. — The most striking feature of this species is the wide range of varia- tion in coloration. This variation appears to be geographically correlated. In Califor- nia specimens some red is always present on at least the head. The prothorax and often the entire thorax may also be red, giving the specimen a strongly bicolored appearance. Otherwise, perpuiwtata can be recognized by having two submarginal cells, three mandibular teeth, propodeum without distinct sculpturing, and small ocelli. Quemaya sonorensis Kimsey and Wasbauer, new species (Figs. 9, 17) Male. — Body length 3.5^.0 mm; face (Fig. 9): clypeus with broad transverse me- dial projection subtended by discrete de- clivity, apical truncation 2 MOD wide; mandible with three apical teeth; gular ca- rina simple; F-I length 1.5x breadth; F-II length 1.7X breadth; interantennal dis- tance 0.7 MOD wide; midocellus separat- ed from eye margin by 3 MOD; mesopleu- ron smooth, punctures 1-4 puncture di- ameters apart becoming sparser above midcoxa; forewing with one submarginal cell, venation as in paupercula (Fig. 17); epipygium apex truncate; paramere as in costaricensis. Body and antennal color pale Volume 7, Number 1, 1998 47 reddish brown; wing veins untinted, ex- cept stigma reddish brown. Etymologi/. — The type series of this spe- cies is from Sonora, Mexico; thus the name. Type material. — Holotype male: MEXI- CO: Sonora, 6 km nnw San Carlos, 11-15 July 1983, E. Fisher, MT (DAVIS). Para- types: 11 males (DAVIS), same data as ho- lotype. Discussion. — The small ocelli, short fla- gellomeres, single submarginal cell, trun- cate epipygial apex and lack of a gular tooth distinguish sonorensis from most oth- er Qiiemaya species, except eurekaensis and costaricensis. Quemaya sonorensis can be distinguished from these two species by the larger ocelli and flagellomeres I and II 1.5 X as long as broad or longer, not short- er as in costaricensis, narrow clypeal trun- cation and broad transverse clypeal pro- jection. ACKNOWLEDGMENTS This study was made possible by the assistance of all of the collections and curators who provided spec- imens. Thanks also to Maria Garcia who did the final preparation of the ink drawings, and to the editors for their suffering. LITERATURE CITED Bradley, J. C. 1917. Contributions toward a mono- graph of the Mutillidae and their allies of Amer- ica north of Mexico. Transactions of the American Entomological Society 43:247-290. Cockerell, T. D. A. 1896. Descriptions of new Hyme- noptera. Transactions of the American Entomologi- cal Society 22:289-297. Malloch, ]. R. 1924. A new species of the genus Bra- chycistis. Bulletin of the Brooklyn Entomological So- ciety 19:23. Pate, V. S. L. 1947. A conspectus of the Tiphiidae, with particular reference to the nearctic forms. journal of the New York Entomological Society 54: 115-145. Wasbauer, M. S. 1967. A new species of Quemaya from the Colorado desert of California. Proceed- ings of the Biological Society of Washington 80:169. J. HYM. RES. Vol. 7(1), 1998, pp. 48-56 Geographic Variation of Sex Ratio in Pelecinus polyturator (Drury) (Hymenoptera: Pelecinidae) Norman F. Johnson and Luciana Musetti Department of Entomology, Museum of Biological Diversity, 1315 Kinnear Road, The Ohio State University, Columbus, OH 43212-1192, USA Abstract. — The relative abundance of males and females of Piieciiius poh/turafor (Drury) (Hy- menoptera: Pelecinidae) was examined on the basis of specimens held in natural history collec- tions. The species may be divided into two groups of populations. Those in the United States and Canada (between 28°N and SFN) are primarily thelytokous: males form only 4% of the total number of individual specimens. Populations from localities 23°N and southward have a sub- stantially higher frequency of males (36%). Within each group of populations, there is no demon- strable change in sex ratio with latitude. Male emergence dates generally precede those of females, but there seems to be no significant difference in the time period in which the two sexes are flying. Within the U.S. and Canada, the uncommon males are not randomly distributed. The parasitoid wasp Pelecinus polytura- tor (Drury) (Hymenoptera: Pelecinidae) is a large and familiar inhabitant of moist deciduous forests in the Nearctic. Its range extends well beyond this, generally from southeastern Canada to central Argentina (Muesebeck 1979, Masner 1993). Surpris- ingly little is known of Pelecinus biology despite its relative abundance. The few published host records indicate that this wasp is an internal larval parasitoid of soil-dwelling Scarabaeidae (Coleoptera) (summarized in Lim et al., 1980). One as- pect of its biology that has received wide comment, however, is that males are ex- tremely uncommon in the northern por- tion of the range. Brues (1928) cited Pelecinus polyturator as an example of the phenomenon of geo- graphic parthenogenesis. Arrhenotokous parthenogenesis is the most common mode of reproduction in the Hymenop- tera, in which males are usually produced from unfertilized eggs. Thelytoky is not at all rare, being known from at least twenty families (Stouthamer et al. 1990). Pelecinus appears to demonstrate both modes: the- lytoky in the north temperate region and arrhenotoky elsewhere. At the time that he wrote his paper Brues admitted that he had seen very few male specimens of P. polyturator from the United States or Canada. His analysis of the sex ratio of the species was anecdotal. His sampling of Pelecinus was never de- scribed and must have been quite limited (see below). Neither did he actually quan- tify observed sex ratios in tropical and temperate regions. Our purpose here is to quantify the issue of sex ratio variation in this wasp throughout its geographical range. MATERIALS AND METHODS Specimen data. — Our information on the distribution and relative abundance of males and females is based upon an ex- tensive survey of the holdings of Pelecinus in natural history collections. Material for this study was borrowed or data acquired from 91 collections around the world (see Appendix). The information associated with specimens varies extensively in ac- curacy and completeness, especially given Volume 7, Number 1, 1998 49 the small size of the typical label attached to a specimen. The data were transcribed and stored in a specimen-level relational database. The table structure is slightly modified from the information model developed by the Association of Systematics Collections (1993) and is implemented in the Oracle?® environment on a Silicon Graphics (UNIX) workstation. This combination of hard- ware and software was chosen to deal with the large numbers of specimens in insect collections and for its ability to in- terface with other software (geographic in- formation systems, mapping software, and World Wide Web servers). The data- base stores all of the information on spec- imen labels (place, time, method of collec- tion, etc.), characteristics of the specimens (e.g., sex, color pattern), source of materi- al, and literature references. The relational structure allows the development of ad hoc queries unconstrained by the format of the original data. As such, the system is not only suited to the questions we ask here, but is also applicable to collection management, diversity assessments, taxo- nomic studies, host-parasitoid biology, etc. The database (Johnson & Musetti 1996), intended to represent the sum of documented geographic and temporal in- formation available, contains data from 7,188 specimens of the genus Pelecinus. Latitude and longitude of collecting lo- calities were added where these could be determined with relative confidence. These are stored in two separate tables, for points and polygons, reflecting the level of accuracy of the cited collecting locality and our ability to locate the sites in atlases and gazeteers. Only those classed as "points" were used in the analyses below. In practice, this means that the collecting records for points consist of localities identified with specific populated places, recreational areas, manmade features (e.g., monuments), or geographic features such as mountains and lakes. Brues (1928) pointed out that Neotrop- Table 1. Numbers of specimens of males and fe- males of Pelecinus polyturator used in analyses of sex ratio. latitude Ul localities Xumber of 6£ [%1 Number o\ S 9 (%] T,it.il mini her 25°N-5rN 38°S-25°N Total 119 (4.2] 616 |38.4] 735 116.5) 2723 [95.8] 990 161.6] 3713 183.5] 2842 1606 4448 ical specimens of Pelecinus exhibit notable variation in color patterns. Many of these were described as distinct species in the early 19th Century, but the present taxo- nomic consensus (dating from Schletterer 1890) is that only a single species, P. po- lyturator, is recognized as valid. To avoid confounding data from possible distinct species, we chose to include in our anal- yses only those specimens conforming to the color pattern of specimens from the U.S. and Canada: the head, mesosoma, and metasoma are uniformly black or very dark brown, and the fore wings are clear or gradually infuscate toward the apex. A summary of the numbers of specimens used is presented in Table 1. Analyses. — Coordinates of latitude and longitude of collecting localities were ex- tracted from the database by sex. Brues (1928) asserted (as does conventional wis- dom) that males and females are not iden- tically distributed. This was tested by comparing the cumulative relative fre- quency distribution of specimens by 1° of latitude using the Kolmogorov-Smirnov test (Sokal & Rohlf 1995). If there is variation in sex ratio among sites, especially over the vast area occu- pied by this species, one reason may be that northern females are substantially longer-lived as adults than males, thus leading to an overabundance of females in collection records. To examine this possi- bility, we sorted the collecting date re- cords for specimens by sex and combined them into groups for each 10 degrees of latitude (from 50°N to 40"S). Collecting dates were expressed in terms of polar co- 50 Journal of Hymenoptera Research Fig. 1. World distribution of Pticciiius polyturnlcr. ordinates by Julian date (1-365) and mean and standard deviation of the dates were calculated for each sex in every 10° band. Sex ratio data are expressed as the pro- portion of males in the total population of specimens. Specimens for every 5° band of latitude were pooled and the relationship between sex ratio and latitude of collec- tion examined using regression. For rea- sons developed below, these data were partitioned into two groups north and south of 25°N. The two partitions then were separately analyzed for relationship between latitude and sex ratio. Finally, on the basis of this data partition, we exam- ined the hypothesis that males in the U.S. and Canada are randomly distributed. Specimens were pooled from blocks of 5° of latitude and longitude and the ob- served number of specimens was com- pared using a x' test with that expected using the observed sex ratio of all speci- mens north of Mexico. RESULTS The documented range of Pelecinii^ poh/- turator extends from a maximum of 51°N in Quebec and Ontario south to 38°S in Argentina (Fig. 1 ). The species does not occur in Chile or the West Indies (includ- Volume 7, Number 1, 1998 51 100 >■ o s K IL III > m 80 60 20 max diff = 0 367 P< 0001 f j^fflO-*'""*'** -40 -20 0 20 40 LATITUDE, DEGREES 60 Fig. 2. Cumulative relative frequency distribution (%) of male and female specimens of Pelkiinis poly- luratcr by latitude. Localities grouped for every 1° of latitude. Maximum difference and probability of identity of distributions using Kolmogorov-Smirnov test. ing Trinidad). Specimens are otherwise found throughout South America, al- though material from the Amazon Basin is very scarce. In the United States, the range of the species extends west to 106° in Colorado and New Mexico. The species is apparently absent from peninsular Flor- ida south of the Gainesville area. The cumulative relative frequency dis- tributions of the two sexes, pooled into 1° increments of latitude, is presented in Fig. 2. The comparison of these two distribu- tions clearly leads to rejection of the null hypothesis that the two are identically dis- tributed throughout America. Specimens of males are clearly more abundant out- side of temperate North America. Two possible sources of sampling error that could lead to the observation of high- ly skewed sex ratios are (1) the samples were taken in particular years in which one sex is either very rare or extraordinar- ily abundant, and (2) if populations are strongly protandrous or adult females sur- vive much longer than males, then the ap- parent rarity of males in some areas may be a collecting artifact. The frequency dis- 300 S 200 O ui a. w III a s 3 100 I ,l.li 11 187S 1900 192S 1950 1975 2000 YEAR Fig. 3. Frequency distribution of collecting years of PelecinKS polyturator. Three specimens collected be- tween 1804 and 1875 are not illustrated. tribution of collecting dates by year is il- lustrated in Fig. 3. The distribution of col- lecting dates for each sex were pooled into groups by 10° of latitude (Fig. 4); statistics for each sex in the latitudinal bands is pre- sented in Table 2. The average collecting dates for males generally precedes those of females, but do not differ significantly. Further, it seems to be possible to find males through most of the flight time of females. Figure 5 demonstrates a significant negative relationship between latitude and sex ratio (with southern latitude ex- pressed as negative numbers). The ob- served sex ratio varies from 0.0-0.60, with an average of 0.19. Even cursory exami- nation of the data reveals that this is not a continuous decrease in the frequency of males, but that the change to spanan- drous populations (with <10% males) oc- curs rather abruptly. We have no knowl- edge of any specimens of Pelecinus col- lected between 23° and 28°N latitude, and the sex ratios on either side of this gap differ strongly. Therefore the data were partitioned into two components at 25°N (Fig. 6). Separate regression analyses re- sult in a change in slope from negative to 52 N 0°-10°183 10°- 20° o or>o 20°- 30 30°- 40' 40°- 50° Fig. 4. Collecting dates for males and females of Piiecinus polyturator. Localities grouped for every 10° of latitude. Julian day 1 = 1 January; 92 = 2 April; 183 = 2 July; 274 = 1 October; N: north latitudes; S: south latitudes. Journal of Hymenoptera Research s cf 9 positive, but neither significantly differs from a slope of 0, i.e., there is no demon- strable relationship between latitude and sex ratio in the two groups. The average sex ratio for the southern populations is 0.36 (0.20-0.60), and that for the northern populations is 0.04 (0.00-0.06). Figure 7 maps the abundance of males and fe- males in 5° blocks of latitude and longi- tude. Pooling specimens by latitude or Volume 7, Number 1, 1998 53 Table 2. Dates of collection of Pelecinin pclyturalor specimens, localities within 10' bands of latitude pooled together, v: mean Julian date of collection (calendar date); sd: standard deviation; N: number of specimens. Degrees latitude \orth South f'' 0°-10° X 144.7 (25 May) 153.4 (2 Jun) 22.7 (23 Jan) 344.8 (11 Dec) sd 7.1 8.5 20.7 26.3 N 137 222 16 27 10°-20° .V 181.0 (30 Jun) 185.7(5 Jul) 344.0 (10 Dec) 24.3 (24 Jan) sd 4.8 5.4 0.5 17.9 N 133 390 2 11 20°-30° X 209.3 (28 Jul) 195.9 (15 Jul) 39.0 (8 Feb) 37.5 (7 Feb) sd 3.4 6.9 12.2 18.5 N 26 30 107 189 30°-t0° .Y 213.2 (1 Aug) 222.4 (10 Aug) 48.3 (17 Feb) 25.7 (26 Jan) sd 2.3 3.4 0.8 22.0 N 18 897 3 8 40°-50° X 228.1 (16 Aug) 230.7 (19 Aug) sd 2.4 2.8 N 88 1277 70^ y = -0 45x + 27 25 r = 0.334 LATITUDE, DEGREES Fig. 5. Sex ratio for all specimens of Pdcciiins ;'o/i/- lurntcr (proportion of males in total) as a function of latitude, with regression and 95% confidence limits. Individual localities pooled for every 5° of latitude; south latitudes expressed as negative numbers. longitude (Table 3) reveals that males do not appear to be randomly distributed through the U.S. and Canada. DISCUSSION Collecting records are not random sam- ples and we recognize a number of pos- sible biases in the data. Female Pelecinus are large and "unusual," easily identifia- ble, relatively slow fliers, and are often found resting on vegetation at heights ac- cessible to collectors. Thus, females are commonly found in the holdings of even small collections and may be overrepre- sented. On the other hand, their numerical abundance and the fact that only a single species is recognized may cause experi- enced collectors to ignore them. Males may be relatively scarce in collections be- cause their abundance in some areas may be truly low, or they may be overrepre- sented precisely because of their rarity, at least in the eyes of North American col- 54 Journal of Hymenoptera Research 80 70 I- y = 0406x-21 17 r = 0.088 -40 -20 0 20 40 LATITUDE, DEGREES Fig. 6. Sex ratio for Peleciuus polyturator (proportion of males in total) as a function of latitude, with re- gression and 95% confidence limits. Data partitioned into two groups: localities north of 25°N, and locali- ties south of 25°N. Individual localities pooled for ev- ery 5^ of latitude; south latitudes expressed as nega- tive numbers. lectors. The magnitude and net effect of these biases are impossible to quantify. However these collections represent the material foundation upon which every- thing we know about this species is based Table 3. Test of null hypothesis that male speci- mens in Fig. 7 are distributed randomly among blocks in the U.S. and Canada. When pooling by lon- gitude, the specimens from the two westernmost col- umns of cells and the three easternmost columns of cells were summed to obtain expected numbers great- er than five. Expected numbers of males based on overall sex ratio in America north of Mexico: 4.0% males. ": probability < 0.01; *: probability < 0.05. specimens in 5" blocks pooled by lati- tude " 38.6" 2 specimens in 5° blocks pooled by longi- tude 10.2* 4 and is the only sample available from which to estimate the sex ratio. We believe that we can fairly judge the hypothesis presented by Brues with cautious use of the specimen data from collections. Our survey of collections produced only 83 male specimens of Pelecinus that we are certain would have been available to Brues (i.e., collected in 1928 or earlier; a further 75 males have no year of collection on the label). Even on the basis of such limited data, it appears that Brues gener- - 2d' 94? 2.1% 8^ 143? 5.3% Od- 26? 0.0% Od- 4? 0.0% 27 d' 89? 23.3% Od' 36? 0.0% Od* 4? 0.0% Oo- 3 ? 0.0% 37.^ 399?*'-^ 8.5% 1 cf 15? 6.3% 1 rf 55? 1.8% 2d' 91 ? 2.2% 17 0- 292 ? 5.7% 90" 80? iai% 15 d" 343? 4.2% lOd' 482? 2.0% 80" 19 ? 29.6% Oo- 5 ? 0.0% 63 o' 1382 $ 4.4% 30* 30? 12? 5^ 70? 6.7% Od- 110? 0.0% 1 d' 203? 0.5% Sd" 344?. 1.4% Oo- 7 ? 0.0% So* 23 ? 25.8% Oo- 8 9 0.0% 14 d' 776? 1.8% 9.1% 0.0% Orf 2? 0.0% Od- 12? 0.0% Oo" 38? 0.0% Od- 110? 0.0% Oo" 3? 0.0% lOd- 525 ? 1.9% Od- 165 ? 0.0% 4^ 45? 8.2% 163? 1.8% 15d" 316 ? 4.5% 170* 466 ? 3.5% lOd* 397 ? 2.5% 47 0" 779? 5.7% 114o' 2722? 4.0% Fig. 7. Numbers of males, females, and sex ratio (proportion of males In total) of thclytoknus populations of Pclcciiius polyturator in blocks of 5° latitude and longitude in the U.S. and Canada. Volume 7, Number 1, 1998 55 ally described the true situation: males are very scarce in temperate North America, and elsewhere they occur in numbers con- sistent with a 0.50 sex ratio. There is an abrupt transition between the two popu- lations that corresponds with a geograph- ical disjunction in southern Texas and northern Mexico. Further focussed collec- tions are needed to determine whether this disjunction is real and, if not, what happens to the males in that area. We cannot yet identify any diagnostic morphological differences between the northern thelytokous populations and the bisexual populations to the south. Speci- mens from southern Mexico (Chiapas) and Central America are often distinguishable, but typical black specimens are found from Mexico to Argentina. There is pre- cious little information on the biology of Pelecinus, but Aguiar (1997) has recently described the copulatory behavior of in- dividuals in Brazil, consistent with the idea that males in the tropics and south temperate regions are functional, i.e., that females do indeed mate. Brues (1928) additionally speculated on the genetics of Pelecinus. He asserted that the largest specimens were found in the north and suggested that these may be tet- raploid. The largest specimens we have seen, however, are from Argentina and a great range of sizes may be found even in single populations in the U.S. The size variation could be better explained by variation in host size than by invoking unexplored genetic mechanisms. The discovery of the nonrandom distri- bution of males in the United States and Canada was particularly surprising. This could be the result of a statistical artifact or something unusual may be occurring in some populations. The most notable of these is a population near Ottawa, Ontar- io: male specimens have been consistently collected in this single site over a period of ten years. Young (1990) described a po- tentially similar situation in southern Wis- consin in which he suggested that a bisex- ual population may have replaced the the- lytokous strain. The Ottawa population could be a promising one upon which to focus in order to better understand the role of males in the northern temperate populations of Pelecinus. ACKNOWLEDGMENTS Thanks to the many curators who made their ma- terial and data available to us; to A. P. Aguiar, A. Sharkov, and J. W. Wenzel for constructive comments on the manuscript. This material is based in part upon work supported bv the National Science Foun- dation under Grant No. DEB-9521648. LITERATURE CITED Aguiar, A. P. 1997. Mating behavior of Peteaiuis po- lyturatcr (Hymenoptera: Pelecinidae). Entomolog- ical News 108: 117-121. Association of Systematics Collections. 1993. An in- formation model for biological collections. Re- port of the Biological Collections Data Standards Workshop, August 18-24, 1992. URL: gopher:// kau\keil.ukans.edu:70/ll/staniiards/asc. Brues, C. T. 1928. A note on the genus Pelecinus. Psy- che 35: 205-209. Johnson, N. F. and L. Musetti. 1996. The Pelecinus pro- ject. URL: http://iris.hiosci.ohio-state.edu/Pelecinus. Lim, K. P., W. N. Yule, and R. K. Stewart. 1980. A note on Pelecinus polyturator (Hymenoptera: Pe- lecinidae), a parasite of Phyllophaga anxin (Cole- optera: Scarabaeidae). The Canadian Entomologist 112: 219-220. Masner, L. 1993. Chapter 13. Superfamily Proctotru- poidea. Pages 537-557, in H. Goulet & ]. Huber, eds., Hymenoptera of the world: an identification guide to families. Research Branch, Agriculture Canada, Publication 1894/ E. 668 pp. Muesebeck, C. F. W. 1979. Superfamily Pelecinoidea. Pages II 19-1 120, m K. V. Krombein, P. D. Hurd, Jr., D. R. Smith, and B. D. Burks, Catalog of Hy- menoptera in America north of Mexico. Smith- sonian Institution Press, Washington, DC. 3 vol., 2735 pp. Schletterer, A. 1890. Die Hymenopteren-Gattungen Stenophasmus Smith, Monomachus Westw., Peleci- nus Latr. und Megalyra Westw. Berliner Entomo- togische Zeitschrift 33: 197-250. Sokal, R. R. and F. J. Rohlf. 1995. Biometry. The prin- ciples and practice of statistics in biological re- search. Third edition. W. H. Freeman and Com- pany, New York. 887 pp. Stouthamer, R., J. D. Pinto, G. R. Platner, and R. F. Luck. 1990. Taxonomic status of thelytokous forms of Trichogramma (Hymenoptera: Tricho- 56 Journal of H'imenoptera Research grammatidae). Annals of the Entomological Societi/ of America 83: 475-481. Young, D. K. 1990. Distribution of Pciecinus polytura- tor in Wisconsin (Hymenoptera: Pelccinidae), with speculations regarding geographical par- thenogenesis. Tlie Great Lakes Entomologist 23: 1-4. APPENDIX Sources of material. American Entomological Insti- tute, Gainesville, FL; American Museum of Natural History, New York, NY; Academy of Natural Sci- ences, Philadelphia, PA; Buffalo Museum of Science, Buffalo, NY; California Academy of Sciences, San Francisco, CA; Albertson College of Idaho, Caldwell, ID; Carnegie Museum of Natural History, Pittsburgh, PA; Canadian National Collection of Insects, Ottawa, ON; Colorado State University, Fort Collins, CO; Cor- nell University, Ithaca, NY; Cambridge University Museum of Zoology, Cambridge, UK; Deutsches En- tomologisches Institut, Eberswalde, Germany; Uni- versity of New Hampshire, Durham, NH; College of Environmental Science & Forestry, Syracuse, NY; Denver Museum of Natural History, Denver, CO; Es- cuela Agricola Panamericana, Zamorano, Honduras; Estacion de Biologia "Chamela", UNAM, San Patri- cio, Mexico; North Carolina Department of Agricul- ture; University of California, Berkeley, CA; Utah State University, Logan, UT; University of Wyoming, Laramie, WY; Funda^ao Instituto Oswaldo Cruz, Rio de Janeiro, RJ, Brazil; Field Museum of Natural His- tory, Chicago, IL; Florida State Collection of Arthro- pods, Gainesville, FL; Instituto Miguel Lillo, San Mi- guel de Tucuman, Tucuman, Argentina; Instituto Na- cional de Biodiversidad, Santo Domingo, Costa Rica; Illinois Natural History Survey, Urbana, IL; Instituto Nacional de Pesquisas da Amazonia, Manaus, AM, Brazil; University of Wisconsin, Madison, WI; Uni- versidad Central de Venezuela, Maracay, Venezuela; Kansas State University, Manhattan, KS; Natural His- tory Museum, Los Angeles, CA; Loyola University, Chicago, IL; M.A. Ivie private collection; Milwaukee Public Museum, Milwaukee, WI; Museum of Com- parative Zoology, Cambridge, MA; Mississippi State University, Mississippi State, MS; Museum d'Histoire Naturelle, Geneva, Switzerland; Universidad Nacio- nal de La Plata, La Plata, Argentina; Museum Na- tional d'Histoire Naturelle de Paris, France; Michigan State University, East Lansing, MI; Montana State University, Bozeman, MT; Museu de Zoologia da Universidade de Sao Paulo, Sao Paulo, SP, Brazil; North Carolina State University, Raleigh, NC; North Dakota State University, Fargo, ND; Naturhistorisch- es Museum, Vienna, Austria; Naturhistoriska Riks- museet, Stockliolm, Sweden; New York State Muse- um, Albany, NY; Oklahoma State University, Nor- man, OK; Ohio State University, Columbus, OH; P.K. Lago private collection; Peabody Museum of Natural History, Yale University, New Haven, CT; Pennsyl- vania State University, State College, PA; Purdue University, West Lafayette, IN; Museu Nacional, Rio de Janeiro, RJ, Brazil; Pontificia Universidad Catolica del Ecuador, Quito, Ecuador; Nationaal Natuurhis- torisch Museum, Leiden, The Netherlands; Royal On- tario Museum, Toronto, Ontario, Canada; R.S. Miller private collection; Rutgers State University, New Brunswick, NJ; R. Willis Flowers collection; South Dakota State University, Brookings, SD; Servicio En- tomologico Autonomo, Nicaragua; University of Kansas, Lawrence, KS; Southern Illinois University, Carbondale, IL; Smithsonian Tropical Research Insti- tute, Panama; Texas A&M University, College Sta- tion, TX; T.K. Philips private collection; University of Arkansas, Fayetteville, AK; University of Arizona, Tucson, AZ; Universidad de Concepcion, Concep- cion, Chile; University of California, Davis, CA; Uni- versity of Colorado, Boulder, CO; University of Con- necticut, Storrs, CT; University of California, River- side, CA; University of Delaware, Newark, DE; Uni- versity of Georgia, Athens, GA; University of Louisville, Louisville, KY; University of Massachu- setts, Amherst, MA; University of Mississippi, Ox- ford, MS; Museum of Zoology, University of Michi- gan, Ann Arbor, MI; University of Missouri, Colum- bia, MO; University of Minnesota, St. Paul, MN; Universidad Nacional Autonoma de Mexico, Mexico City, Mexico; University of Nebraska State Museum, Lincoln, NE; National Museum of Natural History, Washington, DC; University of Vermont, Burlington, VT; Virginia Tech University, Blacksburg, VA; Uni- versity of Idaho, Moscow, ID; James Entomological Collection, Washington State University, Pullman, WA; West Virginia University, Morgantown, WV; Humboldt Universitiit, Berlin, Germany; Zoologische Staatssammlung, Munich, Germany. J. HYM. RES. Vol. 7(1), 1998, pp. 57-61 Parasitism of Siphotiinus phillyreae (Homoptera: Aleyrodidae) by Aphelinid Parasitoids at Different Locations in Egypt S. Abd-Rabou and M. M. Abou-Setta Plant Protection Research Institute, Nadi El-Said Street, Dokki, Giza, 12618, Egypt Abstract. — Seven species of aphelinid parasitoids (Hymenoptera: Aphelinidae) were reared from second and third larval stages and pupae of pomegranate whiteflies, Siphoninus phillyreae (Hali- day) (Homoptera; Aleyrodidae) from three locations in Egypt, during a one year survey (June 94- June 95). Three species, Eretmocerus nniiuius Mercet, Encarsia davuii Viggiani and Mazzone, and £. galilea Rivnay, were reported from Arish (Northeast Egypt). Parasitism was greatest (45.7%) in Arish during September 1994, out of which 38% was caused by E. mundiis. Eiicarsia inaron (Walker) was the dominant parasitoid of S. phillyreae in Giza (Central Egypt) and Assiut (Upper Egypt), with average parasitism rates of 38 and 46.5% over the year, respectively. In Giza, total parasitism reached a maximum of 80% during August 1994, with Encarsia inaron being responsible for 66.1%. In Assiut, parasitism peaked at 93.1% in August 1994 with E. inaron accounting for 78% of the total. Eretmocerus diversicilatus Silvestri and Encarsia lutea (Masi) were reported only from Giza. Eretmocerus corni Haldeman was reported only from Assiut. These species had much less impact than E. inaron in both locations. The presence of different parasitoids at different locations was attributed to geographical factors as well as tolerance of the parasitoids to weather factors and probably availability of alternative hosts. A key for the reported parasitoids is presented. The pomegranate v^'hitefly, Siphoninus piiillyreae (Haliday) (Homoptera: Aleyrod- idae), is the most important pest of pome- granate in Egypt (Priesner and Hosny 1932). This insect's host range is restricted to deciduous fruit crops (Byrne et al. 1990). In Egypt, the host range of S. phil- lyreae includes apple, Pyrus mains L.; pear, Pyrus comnntnis L.; quince, Pyrus cydonia L., and pomegranate, Puuica granatum L. (Abd-Rabou 1990). Pomegranate orchards extend from the north coast to southern Egypt. Pomegranate leaves heavily infest- ed with S. phillyreae have the demand for fluid transport substantially increased be- yond the tree's normal capacity to re- spond. The loss of phloem fluids certainly represents a loss of potential productivity and probably contributes to the reported reduction in fruit size (Costacos 1963). El- wan (1982) studied the biology of S. phil- lyreae and showed that its developmental period varied according to temperature and relative humidity. Various aphelinid parasitoids (Hymenoptera: Aphelinidae) have been reported from S. phillyreae, in- cluding Encarsia inaron (Walker), £. siplion- ini Silvestri, £. galilea Rivnay, £. hispida De Santis, £. pseudopartenopea Viggiani and Mazzone and Eretmocerus corni Haldeman (Priesner and Hosny 1940; Mentzeloz 1967; Viggiani and Mazzone 1980a,b; Viggiani and Battaglia 1983; Rivnay and Gerling 1987; Polaszek et al. 1992). Puuica grana- tum is a deciduous fruit crop which defo- liates in winter. The purpose of this study was to mon- itor the rate of aphelinid parasitism on S. phillyreae infested pomegranate leaves in three distinctive regions in Egypt, to as- sess the impact of biological control on the whitefly, and to determine in what ways it can be improved. MATERIALS AND METHODS Second and third larval stages and pu- pae of Siphoninus phillyreae were sampled 58 Journal of Hymenoptera Research Table 1. Percent parasitism of Siphoiiiinis pyhillyreae by different aphelinid parasitoids in Arish, Egypt. JJASONDJFMAMJ 1994 1995 Month Fig. 1. Monthly mean maximum and minimum temperatures and percent relative humidity at the three locations over the survey period. on pomegranate leaves collected monthly (30 infested leaves per sample) from one site representing each of three distinctive regions in Egypt. The number of trees var- ied by location. Arish is located in North- east Sinai (i.e., coastal area), Giza is locat- ed south of Nile Delta and Assiut in southern Egypt. Environmental data were obtained from local weather stations and mean monthly values were plotted in Fig. 1. Pomegranate trees in the three locations did not receive any chemical treatments except at the Assiut area, which was sprayed for stem borers during defoliation (i.e., January-February 1995). Defoliation time was longer in the Arish area than at Giza or Assiut. No data for the whitefly is presented during the defoliation periods. Pomegranate leaves were transferred to the laboratory in well-ventilated boxes. S. phillyreae eggs and first larval stages were Whitefly - individuals/ leaf Percent parasitism Er. miindiis En. En. June 1994 July August September October November December January 1995 February March April May June 41.6 66.6 110.0 100.0 70.4 35.0 16.6 17.0 23.0 31 38 27 10 5 1.5 4.5 8 6.5 3 0.5 0 0 0.3 0.8 1.2 2.0 0.5 0 32.6 13 1.1 0.8 — Data was not available because of defoliation. eliminated as well as other insects. Total number of S. phillyreae individuals in each stage were recorded per leaf. Each leaf was stored in well-ventilated glass emer- gence tube and monitored daily for para- sitoid emergence. Parasitoid adults were slide mounted in Hoyer's medium and identified to species, and a diagnostic key was constructed for their identification. RESULTS Parasitoid species emerging from sam- ples of S. phillyreae on pomegranate varied according to the area from which they were collected in Egypt. In the Northeast- ern Sinai, where Arish is located, Erctmo- cerus miimius Mercet, Encarsia davidi Vig- giani, and E. galilea parasitized an average of 25% of the total S. phillyreae population. Eretmocerus miindus was responsible for 20.5% of the total parasitism and the other 4.4% by the other two parasitoids (Table 1). In the Arish area, parasitism peaked at 45.7% in September (Table 1). The parasit- oid species reported from Arish area (i.e. £. mundus, E. davidi and E. galilea) were not recovered from samples in Giza or Assiut. In Giza (Central Egypt), S. phillyreae was parasitized by £. iiiaroii, Eretinocents di- versicilatus Silvestri, and £. liitea (Masi) at Volume 7, Number 1, 1998 59 Table 2. Percent parasitism of Siiilwmnnf phiUyrecic by different aphelinid parasitoids in Giza, Egypt. Table 3. Percent parasitism of Siyhouiuui phillyrcae by different aphelinid parasitoids in Assiut, Eg\'pt. Whileflv individu^ils/ leal Percent parasitisir 1 Date Whitefly individuals/ — leaf Percent parasitism En. iiiiiron Er iluvn- £.1 Fn imlrim Fr conn D.iti- June 1994 July 70.1 50 0 9.0 12.1 June 1994 27.4 24.1 1.5 0.5 113.8 73.0 July 33.7 39.0 4.5 2.0 August 189.0 78.0 15.1 August 56.3 66.1 8.1 6.0 September 177.8 66.0 17.1 September 68.9 69.0 6.5 3.5 October 140.4 45.1 19.1 October 66.4 53.1 3.0 1.5 November 100.0 20.0 14.0 November 30.6 9.0 0.5 !.0 December 20.7 18.0 10.0 December 10.6 3.1 0.0 0.0 January 1995 — — — January 1995 — — — — February — — — February — — — — March — — — March — — — — April 10.3 0.0 0.0 April — — — — May 40.3 32.2 4.1 May 17.3 0 0 0 June 21 37.1 5.1 June 22.1 25.0 1.1 1.0 — Data was not available beca use of defoliation. Data was not available because of defoliation. average rates of 38, 4.5, and 2.2%, respec- tively (Table 2). The maximum rate of par- asitism reached 80% during August 1994, of which £. itiaron was responsible for 66.1% of the total. In Assiut, parasitism averaged 46.5% by £. inaroti and 12% by Eretmocerus corui (Ta- ble 3). Parasitism peaked to 93.1% during August 1994, where £. inaron was respon- sible for 78% of the total. All reported parasitoids were primary parasitoids except for Encarsia males that are known to be hyperparsites on females of their own species or other parasitoid species (Viggiani 1981). No hyperparasi- tism was observed. DISCUSSION The three areas of Egypt surveyed were distinctive in their locations as well as their weather (Fig. 1). The Arish area is located in Northeast Sinai and can be characterized by colder and longer win- ters, and higher relative humidity than in Giza or Assiut. Also, Arish is isolated from the other two locations by the vast desert area of the Sinai, which may ac- count for difference in the S. pliillyreae par- asitoid complex. The parasitoid species collected from Arish area were not report- ed in the other two locations. £. inaron was the dominant parasitoid of S. phillyreae in both Giza and Assiut. Both of these areas are in the Nile River Valley, south of Nile Delta with Assiut about 300 km south of Giza. Encarsia lutea and £. diversicilatus were obtained only in samples from Giza, while £. corni was recovered only from Assiut. Higher temperature in Assiut may correlate to both higher whitefly popula- tions and higher rates of parasitism. Encarsia davidi was recorded by Abd-Ra- bou (1994) as a parasitoid of Aleurolobus niloticus Priesner and Hosny on Ziziphus spinachristi in Egypt. Encarsia galilea was recorded by Abd-Rabou (1994) as a para- sitoid of S. phillyreae on P. granatum. Pries- ner and Hosny (1932) recorded Encarsia in- aron as a parasitoid of S. phillyreae on P. granatum and indicated a rate of parasit- ism as high as 80%. In our survey, the peak of parasitism by £. inaron occurred in September (69%) and in August (78%) in Giza and Assiut, respectively. Encarsia lutea (Masi) was recorded by Abdel-Fattah et al. (1984) as a parasitoid of Bettiisia tabaci (Genn.) on tomato plants, whereas Eretmocerus corni was recorded by Priesner and Hosny (1940) as a para- 60 Journal of Hymenoptera Research sitoid of B. tahaci on Lantana camara. Eret- mocerus diversicilatus was recorded for the first time from Egypt associated with B. tahaci by BChalifa and El-Khidir (1965), and Eretmocerus mundus has been recorded from Egypt on B. tahaci by El-Helay et al. (1971). In the present study, £. mundus was the dominant species in the Arish lo- cation, with the peak parasitism occurring in September (38%). Parasitoid tolerance to different ranges of temperature and relative humidity is not clearly known. The colder and longer winters of Arish, as well as lower humid- ity in Assiut, may be limiting factors con- trolling the presence and absence of each parasitoid species of S. phillyreac. In addi- tion, the availability of alternate hosts for the parasitoids in the same area is proba- bly another factor. Terminology used in the key follows that of Polaszek et al. (1992). KEY TO APHELINID PARASITOIDS OF SIPHONINLIS PHILLYREAE FROM EGYPT Antennal flagellum 3-segmented in female, club one elongate segment, tarsi 4- segmented, male flagellum 1-segmented Genus Eretmocerus 2 Antennal flagellum 6-segmented in female, club 2-3 segments, tarsi S-segmented, male flagellum 5 or 6-segmented Genus Encarsia 4 Mesoscutum with 2 pairs of setae, male pedicel dark brown 3 Mesoscutum with 3 pairs of setae, first funicle segment triangular, club 6-7 times as long as wide, male pedicel yellow £. comi Haldeman First funicle segment quadrate, second funicle segment longer than wide, submarginal vein with 3 seta E. mundus Mercet First funicle segment triangular, second funicle segment transverse, submarginal vein with 2 setae E. diversicilatus Silvestri Head and mesosoma dark brown to black with 6-10 pairs of setae, metasoma yellow, antennal club 2-segmented E. inaron (Walker) Entire body yellow or orange, or nearly so, metasoma tergite I often infuscate, antennal club 3 segmented 5 Valvular III dark brown, first funicle segment quadrate or wider than long, male funicle segments F1-F3 expanded 6 Valvular III yellow, first funicle cylindrical, 1.6-2.3 as long as wide, male unknown .... £. galilea Rivnay Valvular III short, as long as width at base and 0.3 times as long as ovipositor, tibia II 1.0-1.1 times as long as ovipositor E. lutea (Masi) Valvular III elongate, 2 times as long as width at base and 0.4 times as long as ovipositor, tibia II 0.8 times as long as ovipositor E. davidi Viggiani and Mazzone LITERATURE CITED Abd-Rabou, S. 1990. Taxonomic studies of whiteflies of Egypt (Homoptera: Aleyrodidae). M.Sc. Tlwsis, Fac. of Sciena; Aiii SImins Univ., 193 pp. Abd-Rabou, S. 1994. Taxonomic and biological stud- ies on the parasites of whiteflies (Hemiptera: AI- eyrodidae) in Egypt. Ph.D. Thesis, Fac. of Science, Cairo Univ., 83 pp. Abdel-Fattah, M. I., A. Hendi, M. O. Koliab and A. El-Said. 1984. Studies on Prosfmllelln lulca Masi (Hymenoptera: Aphelinidae), a primary parasite of the cotton whitetly, Bcniisin tabnci (Genn.) in Egypt (Hemiptera: Aleyrodidae). Bulletin ile la So- ciete Entoinolo\;upie ti'Egypte 65: 119-129. Byrne, D. N., T. S. Bellows and M. P. Parrella. 1990. Whiteflies in Agricultural Systems, 227-261, in: Gerling, D. (ed.) "Whiteflies: their Bionomics, Pest Status and Management." Intercept Ltd., 348 PP- Costacos, T. A. 1963. On a severe attack bv Siphomiiiis phiUyreae (Hal.) (Hemiptera: Aleyrodidae) sub- species inneqiinlis Gautier on fruit trees and its control. Geoponika 105; 3-7. El-Hclay, M. S., A. Y. El-Shazli and F. H. El-Gayar. 1971. Biological studies on Beniisia tnlnici (Genn.) Volume 7, Number 1, 1998 61 (Hemiptera: Aleyrodidae) in Egypt. Zeitschrifl feur Angewnndtc Entomologie 69(1): 48-55. Elwan, E. A. 1982. Biological and ecological studies on the pomegranate whitefly, Siphoiiiniis phiUy- rene (Hal.) (Hemiptera: Aleyrodidae). Mij.';/nologiqiie d'Egi/pte 24: 58-70. Rivnay, T. and D. Gerling. 1987. Aphelinidae para- sitoids (Hymenoptera: Chalcidoidae) of white- flies (Hemiptera: Aleyrodidae) in Israel with de- scription of 3 new species. Entonwp'haga 32: 463- 475. Viggiani, G. 1981. The role of hyperparasitism on bi- ological control: A symposium, pp. 19-26. Uni- versity of California, Division of Agricultural Sci- ences. Viggiani, G. and Battaglia, D. 1983. Le specie italian del genera Eretnwcerus Haldeman (Hymenop- tera: Aphelinidae). Bollettino del Laboratorio di En- tcnwlogia Agraria Fdippo Sdvestri 40: 97-101. Viggiani, G. and P. Mazzone. 1980a. Encarsia ptseudo- parlenopea n. sp., parasita di Siphoninus phillyreae (Haliday) (Homoptera: Aleyrodidae). Bollettnw del Laboratorio di Entomologia Agraria Fdippio Sil- vestri 37: 9-12. Viggiani, G. and P. Mazzone. 1980b. Le specie Pa- leartiche di Encarsia del gruppe Lutea (Masi) (Hy- menoptera: Aphelinidae), con descrizione de due nuove specie. Bollettino del Laboratotio di Entomo- logia Agraria Fdippo Silvestrii 37: 51-57. J. HYM. RES. Vol. 7(1), 1998, pp. 62-73 Revision of North American Aleiodes Wesmael (Part 2): the apicalis (Brulle) Species-group in the New World (Hymenoptera: Braconidae, Rogadinae) Scott R. Shaw, Paul M. Marsh, and Joseph C. Fortier (SRS, JCF) Department of Plant, Soil, and Insect Science, P.O. Box 3354, University of Wyoming, Laramie, Wyoming 82071, USA; (PMM) Cooperating Scientist, USDA Systematic Entomology Laboratory, c/o U.S. National Museum of Natural History, NHB-168, Washington, D.C. 20560, USA (correspondence address; P.O. Box 384, North Newton, Kansas 67117, USA) Abstract. — The Aleiodes apicalis (Brulle) species-group is defined to include the following previ- ously described species: apicalis (Brulle, 1832), grandis Giraud, 1857 { = Rhogas malaisei Shestakov, 1940 NEW SYNONYMY), parasiticus Norton, 1869, atriceps Cresson, 1869, abdorninalis Cresson, 1869, rileyi Cresson, 1869, molestus (Cresson, 1872) NEW COMBINATION, schirjajewi Kokujev, 1898, convexiis van Achterberg, 1991 ( = Chelonorhogas rufithorax Enderlein, 1912), and brethesi Shenefelt, 1975 NEW COMBINATION. One newly described species, flavitarsiis Marsh and Shaw, is also included. The apicalis species-group is regarded as monophyletic based on the pres- ence of dense setal mats on the apical metasomal terga of males. The genus Dimorphomastax Shenefelt, 1967 is synonymized under Aleiodes, and the species Dimorphomastax peculiaris Shenefelt, 1979 is regarded as a junior synonym of Aleiodes atriceps Cresson. Rogas rufocoxalis Gahan, 1917 is newly synonymized as a junior synonym of Aleiodes molestus (Cresson). A key to the New World species of the apicalis species-group is provided, and species treatments are given for Nearctic species, including diagnostic characters, distribution, and biological information. The rogadine braconid genus Aleiodes Wesmael is worldwide in distribution, but is particularly species-rich in the Holarctic region. Aleiodes is diverse in North Amer- ica, with at least 90 species in the United States and Canada (Shaw et al. 1997). This paper is the second contribution in a series of planned papers on Aleiodes species- groups, intended to provide a complete revision of the genus for North America. In this paper we treat a distinctive mono- phyletic assemblage, the species of the Al- eiodes apicalis species-group, as it occurs in the New World region. All members of this group have dense setal mats on male terga 4-7 (Fig. 1). Our definition of the species-group includes all species known to us, worldwide. However, because our main intent is to provide a revision for North American species, species treat- ments are limited to the Nearctic fauna. Since only one other New World species is known to us, this is included in the key for convenience. The European species are currently being revised by Kees van Ach- terberg and Mark Shaw. Aleiodes species are koinobiont endopar- asitoids of lepidopteran larvae, especially macrolepidoptera of the superfamilies Noctuoidea and Geometroidea, and to a lesser extent, Arctioidea, Sphingoidea, and Papilionoidea (Shaw et al. 1997). Members of the apicalis group, as far as known, are mostly parasitoids of Noctuidae. The method of parasitism, unique to the tribe Rogadini, is noteworthy: the Aleiodes larva completes its feeding and pupates within the shrunken and mummified remains of the host caterpillar. The form of the mum- my caused by a particular Aleiodes species is usually characteristic for that host and parasitoid, so the mummified remains are Volume 7, Number 1, 1998 63 of considerable diagnostic value and should be retained with the parasitoid, when reared. For more complete discus- sions of Aleiodes biology, readers may re- fer to Shaw (1983, 1994), Shaw and Hud- dleston (1991), Shaw (1995), and Shaw et al. (1997). METHODS Species covered in this paper can be identified as members of the subfamily Rogadinae using the keys of Shaw and Huddleston (1991), van Achterberg (1993), or Shaw (1995). Our definition of Aleiodes follows that of van Achterberg (1991), Shaw (1993), and Shaw et al. (1997). Spec- imens can be determined as Aleiodes using the keys of Marsh et al. (1987), van Ach- terberg (1991), or Shaw (1997). Specimens keyed through Marsh et al. (1987) will key to couplet 185, at which point they can be separated from Rogas by the presence of a discrete median carina on the propodeum, the lack of a foveate sternaulus on the mesopleuron, and the lack of a blunt basal tooth on the tarsal claw. In practice, more than 99% of U.S. and Canadian specimens encountered will be Aleiodes, as Rogas s.s. is only infrequently encountered north of Mexico (but increases in species richness in the neotropics). The species-groups of North American Aleiodes can be keyed us- ing the key provided in Shaw et al. (1997). The species treated in this paper were for- merly assigned to the diictor Thunberg species-group by Shaw et al. (1997) follow- ing a recent interpretation of that species by Papp (1985). However, Kees van Ach- terberg (pers. comm.) has indicated to us that previous interpretations of ductor are not correct, and that the species treated here are better called the apicalis species- group. Terminology follows that used for Aleio- des by Shaw et al. (1997), Shaw (1993) and Marsh (1989). Microsculpture terminology follows that of Harris (1979). Wing vena- tion terminology (see Fig. 16) follows that of Shaw (1997) and Shaw et al. (1997). Abbreviations for museums are as fol- lows: ANSP, Academy of Natural Sci- ences, Philadelphia; AEI, American Ento- mological Institute, Gainesville; AMNH, American Museum of Natural History, New York; CAS, California Academy of Sciences, San Francisco; CNC, Canadian National Collection, Ottawa; CUI, Cornell University, Ithaca; FSCA, Florida State Collection of Arthropods, Gainesville; HNHM, Hungarian Natural History Mu- seum, Budapest; INHS, Illinois Natural History Survey, Urbana; MISU, Michigan State University, East Lansing; MSSU, Mississippi State University, Mississippi State; OKSU, Oklahoma State University; RMNH, Nationaal Natuurhistorisch Mu- seum, Leiden; TAMU, Texas A. & M. Uni- versity, College Station; UCD, University of California, Davis; UMSP, University of Minnesota, St. Paul; RMSEL, Rocky Mountain Systematic Entomology Labo- ratory, University of Wyoming, Laramie; USNM, U.S. National Museum of Natural History, Washington, D.C. Authorship of species is attributed to the senior authors (PMM and SRS) in the order indicated. ALEIODES APICALIS SPECIES-GROUP Included species: apicalis (BruUe, 1832), grandis Giraud, 1857 ( = »;fl/rt(St'/ Shestakov, 1940 new synonymy), parasiticus Norton, 1869, atriceps Cresson, 1869 revised com- bination { = Diiuorphoi)iastax peculiaris Shenefelt, 1979 new synonymy), abdomi- nalis Cresson, 1869 ( = lectus Cresson, 1869), rilei/i Cresson, 1869 revised combination, niolestus (Cresson, 1872) new combination ( = rufocoxalis (Gahan, 1917) new synony- my), schirjajezvi Kokujev, 1898, coiwexus van Achterberg, 1991 ( = Chelonorliogas ruf- ithorax Enderlein, 1912), hrethesi Shenefelt, 1975 new combination (replacement name for nigriceps Brethes, 1909, preoccupied by iiigriceps Wesmael, 1838), and flavitarsus Marsh and Shaw, new species. Diagnostic characters. — Ocellar diameter small, ratio of ocellar diameter to distance 64 Journal of Hymenoptera Research Figs. 1-6. Fig. 1. Densely setose metasomal tergum 4. A. alniomiiiaUs, male. Figs. 2-3. Mandible and oral space. 2. A. atriceps, male. 3. A. abdominalis, male. Figs. 4-6. Mesonotal sculpture. 4. A. abdominalis. 5. A. parasiticus. 6. A. hrcthesi. between lateral ocellus and compound eye less than 1; occipital carina meeting hy- postomal carina laterally; dense setal mat present on male terga 4-7 and subdivided medially (Fig. 1); medial ridge extending down frons 0.55 or more of distance from line between base of scape to clypeus; mesonotal disc sculpture finely granulate to smooth (Figs. 4-6), sparsely or not se- tose; tarsal claws strongly pectinate (Figs. 12-14). Remarks. — A moderate-sized, circumpo- lar and neotropical monophyletic group, associated mostly with noctuids (see Fig. 15). There are some recorded associations with geometrids, lymantriids, pyralids, and even sawflies, but these need confir- mation and the latter seems unlikely. The dense setal mats on male terga 4-7 are undoubtedly synapomorphic. The function of the dense setal mats on male terga 4—7 is unknown, but perhaps they Volume 7, Number 1, 1998 65 may serve to disperse pheromones during {Aleiodes, Cheloiwrhogas, Diniorphojnastax, courtship and mating. and Ro;^as). Chiionorliogas was synonymized Even fairly recently (Shenefelt, 1975; with Aleiodes by van Achterberg (1991), but Marsh, 1979), species belonging in this retained Cheloiwrhogas as a valid subgenus, group have been classified in several genera to which the apicalis-group is assigned. KEY TO NEW WORLD SPECIES OF THE ALEIODES APICALIS SPECIES-GROUP 1 Mandible of male with a large, curved, tusk-like accessory tooth situated near the pos- terior condyle (Fig. 2), or female with a small accessory tooth situated near the posterior condyle; oral space unusually large; southwestern U.S. and Mexico . , A. atriceps Cresson - Mandible normal and unmodified, lacking a large, tusk-like tooth (males) or small tooth (females) near the posterior condyle (Fig. 3); oral space smaller (Fig. 3) 2 2(1) Mesosoma (excluding legs) entirely black 3 - Mesosoma color varying from entirely orange, to orange with extensive black markings posteriorly and laterally, but pronotum, mesonotum, and scutellar disc always colored bright orange 4 3(1) Hind tarsomeres orange to brown, similar to color of tibia; body length 6.0-8.0 mm; fore wing vein Icu-a beyond vein IM by less than 3 times its length; hindwing vein m- cu present (Fig. 16) A. abdominalis Cresson - Hind tarsomeres 1-4 yellow, much lighter than color of tibia; body length 4.5-5.5 mm; fore wing vein Icu-a beyond vein IM by 3 times its length; hindwing vein m-cu absent (Fig. 17) A. flavitarsiis Marsh & Shaw, new species 4(3) Body entirely orange to orangish brown; mesonotum granulate and dull A. rileyi Cresson - Body only partly orange, head entirely black, legs, mesosoma, and apex of metasoma with variable black markings; mesosoma sculpture variable, but always somewhat shin- ing and polished 5 5(4) Mesosoma entirely orange, legs entirely black; mesonotal disc entirely smooth and high- ly polished, virtually devoid of setae (Fig. 6); South American species A. brethesi (Shenefelt) - Mesosoma usually orange and black (rarely entirely orange), leg color orange and black, coxae at least always orange; mesonotal disc not so completely smooth, with numerous setal pits (Fig. 5); North American species 6 6(5) Hind femur and tibia banded, orange on basal half, black on apical half; northern species associated with boreal forests A. parasiticus Norton - Hind femur and tibia entirely black; southern species frequently associated with agroe- cosytems A. molestus (Cresson) Aleiodes atriceps Cresson, revised palpi yellow, mesosoma varying from en- combination tirely black to entirely orange, tegula yel- (Fig. 2) low, metasoma orange, legs orange, wings hyaline, veins brown except vein C + Sc + R Aleiodes atriceps Cresson, 1869, Amer. Ent. Soc. ^^^ ^^-^^^ yellow; body length, 6.0-7.0 Trans. 2:380. .,.„-,,.„ mm; 46-53 antennomeres; malar space Dn«orptom«stop.c«/mns Shenefelt, 1979, Proc. ^^ ^^^^, ^^^^^^ ^^ mandible; oral Ent. Soc. Wash. 81:133. New synonymy " . .... ^ . i i , , .,. c\. < ^ openme circular, diameter equal to malar based on exammation of holotype. ^ P , , , , ., . space m female and greater than malar Diagnosis. — Body bicolored, head in- space in male; ocelli small, ocellocular dis- cluding antenna black, mandibles and tance about twice diameter of lateral ocel- 66 Journal of Hymenoptera Research lus; mandible in female with short trian- gular tooth at condyle, in male with large sickle-shaped tooth at condyle; occipital carina meeting hypostomal carina; head entirely coriaceous, sometimes smoother in male; pronotum rugose laterally; meso- notum and scutellum coriaceous, notauli weakly scrobiculate; mesopleuron smooth, subalar sulcus rugose, sternaulus absent; propodeum rugose, median carina com- plete; metasomal terga 1 and 2 costate-ru- gose, median carina complete; third meta- somal tergum longitudinally costate over basal 0.25 to 0.5, smooth posteriorly; terga 4—7 in male with dense row of short yel- low setae at base, and dense patches of yellow setae on each side of mid-line; fore wing with vein Icu-a beyond vein IM by distance nearly twice length of Icu-a, hind wing with marginal cell gradually broad- ening to apex, vein RS straight on basal half and slightly bent downward on apical half, vein m-cu present; tarsal claws strongly pectinate with 6-7 stout teeth, apical 3 teeth of pectin about 0.75 as long as apical claw, remaining teeth gradually shorter towards base of claw; apical tibial spurs of male blunt. Type material examined. — Aleiodes atriceps Cresson, holotype female, Mexico (ANSP). Diniorphomastax peculiaris Shenefelt, para- type female. Portal, Arizona (AEI). Distribution. — Mexico, Arizona, and Texas. Biology. — Unknown. Adults have been collected from July through December. Conwients. — Shenefelt (1975) classified atriceps under Rogas, but we are moving it here back to its original combination with Aleiodes. Although the unusual and dis- tinctive shape of the mandible (Fig. 2) prompted Shenefelt (1979) to create a new genus for this species, we consider it to be only a specialized species of Aleiodes with peculiar mandibles. The species is, in most other characters, a typical Aleiodes and we consider the mandible shape, as well as the blunt tibial spurs of the male, to be autapomorphies. Dimorphomastax Shene- felt, 1969 is therefore considered a junior synonym of Aleiodes, new synonymy. The unusual accessory mandibular tooth of this species is strongly sexually dimorphic (small triangular tooth in the female, large sickle-shaped projection in the male) suggesting a possible role in courtship and mating. The sexual dimor- phism raises doubts about whether it may serve any primary cutting function, such as assisting in escape from the host m.um- my. Aleiodes abdominalis Cresson (Figs. 1, 3, 4, 7, 12, 16) Aleiodes abdominalis Cresson, 1869, Amer. Ent. Soc. Trans. 2:379. Aleiodes lectus Cresson, 1869, Amer. Ent. Soc. Trans. 2:379. Possible synonymy with Aleio- des abdominalis Cresson indicated by Cresson, 1869. Synonymized by Muesebeck & Walk- ley, 1951. Diagnosis. — Body bicolored, head and mesosoma black, metasomal terga 1-3 or- ange, remainder of terga orange to black, antenna and legs orange, wings lightly dusky, veins brown, tegula yellow; body length, 6.0-8.0 mm; 54-63 antemiomeres; malar space longer than basal width of mandible; face costate with distinct raised ridge between antennae, frons, vertex and temple coriaceous; oral opening circular, diameter about equal to basal width of mandible; propleuron rugose; mesonotum and scutellum finely coriaceous, notauli weakly scrobiculate and meeting in small rugose are before scutellum; mesopleuron smooth, subalar sulcus rugose, sternaulus absent; propodeum rugose-coriaceous, median carina on basal half only; first and second metasomal terga costate, median carina complete, third tergum costate at base, remainder smooth; apical terga of males, especially terga 4-7 densely cov- ered with silvery setae, except along me- dian line; fore wing with vein Icu-a wing beyond vein IM by distance greater than length of Icu-a; hind wing with marginal cell gradually widening, vein RS slightly Volume 7, Numbkk 1, 1998 67 s ^i y^Bfli^^B^ "^ Figs. 7-12. Figs. 7-8. Propodeal sculpture. 7. A. ahdominalis. 8. A. parasitictis. Figs. 9-10. Metasomal sculpture. 9. A. brethesi, first tergum. 10. A. brethesi, border of second and third metasomal terga. Figs. 11-12. Tarsal claws. \\. A. parasiticus. 12. A. abdominalis. curved downward, vein m-cu present; tar- sal claws strongly pectinate with 6-7 stout teeth, apical 3 teeth of pectin about 0.75 as long as apical claw, remaining teeth grad- ually shorter towards base of claw. Type material examined. — Aleiodes abdom- inalis Cresson, holotype female, Pennsyl- vania (ANSP). Aleiodes lectiis Cresson, ho- lotype male, Illinois (ANSP). Distribution. — Widely distributed in eastern North America from Quebec and Ontario south to North Carolina, west to South Dakota and Arizona; more com- monly encountered in the eastern parts of its range. Biologi/. — Unknown. One specimen from Maryland is associated with an un- determined noctuid. The mummy is dark brown, smooth, and about 1 cm long. Adults appear in early June in the north- 68 Journal of Hymenoptera Research m Figs. 13-15. Figs. 13-14. Compound microscope photographs of slide-mounted tarsal claws. 13. A. hkyi. 14. A. niclcstiis. Fig. 15. Mummified plusiine noctuid host larva parasitized by A. tnolestus. ern parts of its range; the specimens from Arizona were collected in late September. Comments. — This is the most common member of the species-group with a most- ly black body. The only other North American species in the group with an all- black mesosoma is flavitarsus, from which abdominalis can be distinguished by its larger body size (6.0-8.0 mm), fore wing vein Icu-a beyond vein IM by less than 3 times its length, and hindwing vein m-cu present (Fig. 16). The European species, A. apicalis, is also similar in body color and general appearance. However, abdominalis can be distinguished by its finely coria- ceous to granulate mesonotum (Fig. 4), while the disc of the mesonotum is smooth-punctate in apicalis. Superficially, abdominalis is similar in color pattern (black and orange) to the very common species A. terminalis Cresson, but terminalis is a member of a different species-group, and can be easily separated by the species- group key provided in Shaw et al. (1997). Males of terminalis do not have densely se- tose metasomal terga 4-7. Aleiodes flavitarsus Marsh and Shaw, new species (Fig. 17) Female. — Body color: head black, man- dibles and mouthparts yellow, antenna light brown; mesosoma black, propleuron sometimes brown or orange; tegula yel- low; legs yellow except apical tarsal seg- ments, apical 0.25 of hind femur and api- cal 0.5 of hind tibia which are black; meta- soma with segments 1-3 yellow-orange except tergum 1 black medially and ter- gum 3 black apically, segments 4-8 black; wings hyaline, veins light brown, tegula yellow. Body length, 4.5-5.5mm. Head: malar space short, equal to basal width of mandible and about eye height; temple narrow, about 0.5 eye width; occipital ca- rina not quite meeting hypostomal carina; oral space small and oval, width equal to malar space and about 0.5 face height; 46- 49 antennomeres, all flagellomeres slightly longer than wide, first slightly longer than second; ocelli small, ocellocular distance equal to or slightly greater than diameter of lateral ocellus; face rugulose-coriaceous with median carina between antennae; frons coriaceous; vertex and temples cori- aceous; occiput smooth and shining; max- illary palpus not swollen; mandibles small, tips not overlapping when closed. Mesosoma: propleuron rugose, often smooth medially; mesonotum and scutel- lum coriaceous; notauli scrobiculate, meet- ing in small rugose area before scutellum; mesopleuron smooth and shining, rugose Volume 7, Number 1, 1998 69 Figs. 16-17. Wings showing venation terminology. 16. A. abdominaU^. 17. A. flnvitarfus. dorsally and in subalar sulcus; sternaulus absent; propodeum rugose-coriaceous dorsally, coriaceous laterally, median ca- rina obscured apically. Legs: tarsal claws strongly pectinate with 4-5 stout teeth, apical 3 teeth of pectin about 0.75 as long as apical claw, remaining teeth gradually shorter towards base of claw; inner spur of hind tibia slightly less than 0.5 length of hind basitarsus; hind coxa smooth dor- 70 Journal of Hymenoptera Research sally. Wings: hyaline; fore wing with vein r 0.5 length of 3RSa, vein Icu-a beyond IM by nearly 3 times length of 1 cu-a, vein ICUa slightly longer than ICUb; hind wing with vein RS straight, cell IRl grad- ually widening to wing apex, vein r-m slightly shorter than IM, vein M + Cu slightly longer than IM, vein m-cu absent. Metasoma: first tergum rugulostriate, me- dian carina complete, length equal to api- cal width; second tergum rugulostriate, median carina complete; third tergum ru- gulostriate basally, smooth apically, me- dian carina absent; fourth and following terga smooth; ovipositor short, about 0.5 length of hind basitarsus. Male. — Essentially as in female. Holotype.— Female: MICHIGAN: Mar- quette County, August 14, 1959, R. And K. Dreisbach. Deposited in USNM. Parnti/pes.— CANADA, BRITISH CO- LUMBIA: 2 females, Gagnon Rd., 6 mi W Terrace, June 20, 1960, J. C. Chillcott, W. W. Moss; 4 females, Hixon, July 11, 1965- July 7, 1966, E. D. A. Dyer; 1 female. Ter- race, July 9, 1960, W. R. Richards; 1 fe- male, 10 km S Fernie, July 29, 1980, D. Williams. MANITOBA: 2 females. Big Eddy, em March 13, 1953, ex Aiitographn sp. NEW BRUNSWICK: 2 males, Kouchi- bougnac N.P., July 21, 1977, S. J. Miller. NEWFOUNDLAND: 1 male, Agr. Exp. Sta., St. John's, July 16, 1967, J. F. Mc- Alpine. ONTARIO: 1 female. One Sided Lake, July 12, 1960, S. M. Clark; 1 male. North Branch, July 23, 1960, S. M. Clark; 1 male, Cumberland, June 13, 1975, L. Ling. QUEBEC: 1 female, 1 male, Parke Reserve, Kam. Co., July 5-12, 1957, G. E. Shewell. UNITED STATES, MAINE: 1 fe- male, Oxford Co., Bryand Pond, July 23, 1976, leg. Heinrich. MICHIGAN: 1 female. Delta Co., August 6, 1959, R. & K. Dreis- bach; 1 male, Houghton Co., August 20, 1959, R. & K. Dreisbach; 1 male, School- kraft Co., August 5, 1959, R. & K. Dreis- bach. MINNESOTA: 1 female, Itasca St. Park, September 1927, S. Carthside. WASHINGTON: 1 female. Lake Cush- man, July 22, 1917, A.L. Melander. Depos- ited in CNC, HNHM, MCZ, RMNH, RMSEL, USNM. Distribution. — Widely distributed across Canada and the northern United States. Biology. — Two specimens from Manito- ba were reared from an unknown species of the genus Autographa (Noctuidae). Adults are active from late June through September. The specimens from Manitoba were labeled as emerging in March, but were probably collected during the pre- vious summer and emergence may have been under laboratory conditions. Comments. — This species is similar to ah- dominalis, but differs by its smaller body size (4.5-5.5 mm), vein Icu-a of the fore wing being beyond vein IM by three times its length (Fig. 17), by having hind tarsomeres 1^ yellow, and by the absence of vein m-cu in the hind wing (Fig. 17). Etymology. — The specific name is from the Latin flaviis meaning yellow, in refer- ence to the yellow hind tarsus. Aleiodes molestus (Cresson), new combination (Figs. 14, 15) Rogas molestus Cresson, 1872, Trans. Amer. Ent. Soc. 4:188. Rogas rufocoxalis Gahan, 1917, Proc. U.S. Natl. Mus. 53:207. New synonymy based on ex- amination of holotype. Diagnosis. — Body bicolored, mesonotum orange, mesopleuron and propodeum varying from entirely black to entirely or- ange, with various intermediate forms oc- curring, metasomal terga 1-3 always or- ange, rest of terga varying from orange to black, legs beyond coxae dark brown or black, wings slightly dusky, veins brown, tegula orange; body length, 5.0-7.0 mm; 45-47 antennomeres; malar space long, slightly greater than basal width of man- dible; face, frons and vertex rugulose, tem- ple smooth; mesonotum and mesopleuron smooth; propodeum rugose, median cari- na complete; first and second metasomal terga strigate-rugose to costate, median Volume 7, Number 1, 1998 71 carina complete; fore wing with vein Icu-a beyond IM by distance nearly twice length of Icu-a; marginal cell of hind wing narrowest basally, gradually widening to- ward wing apex; tarsal claws strongly pectinate with 6-8 stout teeth, apical 3 teeth of pectin about 0.75 as long as apical claw, remaining teeth gradually shorter towards base of claw. Type material examined. — Rogas molestus Cresson, holotype female, Texas, G.W. Belfrage collection, [USNM]. Rogas rufo- coxalis Gahan, holotype female, Colorado, Rocky Ford [USNM]. Distribution. — South Dakota south to Arkansas, Louisiana, Texas, and Mexico, west to Wyoming, Utah, Arizona, and southern California. Biology. — Although A. molestus is com- monly collected by Malaise trap, sweep net, or at lights, verified rearing records are less common. In the USNM collection there are single specimens labelled as reared from several plusiine and noctuine species including Autoplusia egena (Gn.), the soybean looper, Pseudoplusia includens (Wlkr.), the cabbage looper, Trichoplusia ni (Hbn.), and the variegated cutworm, Per- idroma saucia (Hbn.). The known hosts are all generalist feeders on a variety of low vegetation including numerous crop spe- cies. Adults have been collected from late April through mid-September in the southern parts of its range. Comments. — A. molestus is a fairly com- mon midwestern and southern species fa- voring open fields, low vegetation, and agroecosystems. It belongs to the parasiti- cus assemblage, comprising parasiticus Norton, molestus (Cresson) and brethesi (Shenefelt), all of which have an orange mesonotum that is smooth and shining (as in Figs. 5-6) and well-developed longitu- dinal sculpture on tergum 2 -(-3 (as in Fig. 10). A. molestus differs from parasiticus by having the legs entirely dark brown or black beyond the coxae (the femorae and tibiae are banded in parasiticus). It differs from brethesi by having orange coxae. while the legs are entirely black in that South American species. There is considerable variation in the extent of dark coloration on the mesopleu- ron and propodeum in molestus, conse- quently rufocoxalis, which differs only by having an entirely orange mesosoma, can- not be held as a valid species. In his de- scription of rufocoxalis, Gahan (1917) stat- ed that "it would not be surprising if it {rufocoxalis) would ultimately turn out to be merely a color variety of Cresson's spe- cies (molestus)." Indeed, although some- times the mesopleuron and propodeum are entirely black, intermediates with less dark color and orange patches showing through are fairly common, so it would appear that Gahan's prediction is correct. Aleiodes parasiticus Norton (Figs. 5, 8, 12) Aleiodes parasiticus Norton, 1869, Trans. Amer. Ent. See. 2:327. Diagnosis. — Body bicolored black and orange, head black, antenna orange ba- sally to black apically, mesosoma orange except mesopleuron below subalar sulcus and propodeum black, first and second metasomal terga orange, third tergum or- ange on basal 0.5, black on apical 0.5, re- mainder of terga black, fore and middle legs orange except apical tarsomeres black, hind coxa and trochanters orange, hind femur orange on basal 0.66, black on apical 0.33, hind femur yellow on basal 0.5, black on apical 0.5, hind tarsus or- ange except apical tarsomere black, wings lightly dusky, veins brown, tegula yellow; body length, 5.0-7.0 mm; face costate, frons and vertex rugose, temple punctate, occipital carina scrobiculate; malar space longer than basal width of mandible and about 0.5 eye height; ocel- locular distance longer than diameter of lateral ocellus; pronotum rugose; meso- notum smooth, mesopleuron smooth me- dially, subalar sulcus and sternaulus ru- gose; propodeum rugose dorsally, punc- 72 Journal of Hymenoptera Research tate laterally, median carina complete; first and second metasomal terga costate- rugose, median carina complete, third tergum costate on basal 0.5, smooth on apical 0.5, median carina absent, remain- der of terga smooth, fourth-seventh terga of male with dense patches of setae on apical 0.5 on each side of mid-line; fore wing with vein Icu-a beyond IM by dis- tance twice length of Icu-a, marginal cell of hind wing gradually widening, vein RS straight, vein m-cu absent; tarsal claws strongly pectinate with 5-6 stout teeth, apical 3 teeth of pectin about 0.75 as long as apical claw, remaining teeth gradually shorter towards base of claw. Type material examined. — Aleiodes parasi- ticus Norton, holotype female, Connecticut (MCZ). Distribution. — Eastern Canada and Unit- ed States south to Maryland, west to North Dakota, Wyoming, and Colorado. Biology. — We have examined speci- mens from the CNC and USNM collec- tions labelled as reared from the plusiine noctuids Anagrapha falcifera (Kby.) and Syngrapha epigaea (Grt.). The former is a generalist on low plants including blue- berries and clover, while the later is a generalist feeding on conifers including pines, spruces, and firs. The mummy formed is typically cream-colored and fairly smooth. Norton (1869) and Shene- felt (1975) listed this species as having been reared from the diprionid sawfly Neodiprion abietis (Harris) on Abies, but this seems very unlikely. Adults of A. par- asiticus are active from late May to early September. Comments. — A. parasiticus is similar to molestus, but parasiticus differs in having the posterior femur and tibia banded (or- ange on basal 0.5, black on apical 0.5). A. parasiticus is a distinctly northern species associated with boreal forests, while mo- lestus is a midwestern and southern spe- cies favoring drier and more open habi- tats, including agroecosytems. Aleiodes rileyi Cresson, revised combination (Fig. 13) Aleiodes rileyi Cresson, 1869, Trans. Amer. Ent. Soc. 2:382. Diagnosis. — Body unicolored orange or honey yellow, flagellum black, wings hy- aline, veins brown, stigma yellow to light brown; body length, 5.5-8.0 mm; 53-55 antennomeres; malar space short, about equal to basal width of mandible and 0.25 eye height; ocelli large, ocellocular dis- tance equal to or slightly less (about 0.75) diameter of lateral ocellus; oral operung small and circular, diameter about equal to basal width of mandible; face rugose, frons smooth, vertex and temple coria- ceous; occipital carina not meeting hypos- tomal carina; pronotum procate; mesono- tum and scutellum coriaceous; mesopleu- ron smooth, subalar sulcus rugose, ster- naulus absent; propodeum rugose dorsally, coriaceous laterally, median ca- rina complete; first and second metasomal terga rugose costate, median carinae com- plete; third tergum costate on basal 0.33, smooth or weakly coriaceous on apical 0.66, sometimes entirely smooth or weakly coriaceous, median carina absent; remain- der of terga weakly coriaceous; terga 4—6 in male with lateral patches of dense gold hair; fore wing with vein Icu-a beyond IM by distance slightly greater than length of Icu-a; hind wing with vein RS arched in middle, marginal cell narrowest in mid- dle; tarsal claws strongly pectinate with 7- 8 stout teeth, apical 3-4 teeth of pectin about 0.75 as long as apical claw, remain- ing teeth gradually shorter towards base of claw. Type material examined. — Aleiodes rileyi Cresson, holotype female, Missouri (ANSP). Distribution. — Connecticut south to Flor- ida, west to Michigan, Kansas, and Sas- katchewan. It probably occurs throughout the eastern half of North America. Biology. — We have examined specimens Volume 7, Number 1, 1998 73 from the CNC, INHS, and USNM collec- tions labelled as reared from the noctuids Acrom/cta oblinaia (J.E. Sm.), Melanclirn pic- ta (Harr.), and Nep^heloiies minians Gn., the lymantriid Dasychira vagans (B. & McD.), and possibly the pyralids Ostriiiia obiim- bratalis (Led.), and O. penitalis (Grt.). The known hosts are all generalist feeders on a variety of low vegetation, shrubs, and low trees such as willows. Adults of Alcio- des rileyi are active in Florida as early as January, but in northern parts of its range (Illinois) adult activity is in late summer (August). Comments. — Marsh (1979) classified rile- yi under Rogas, but we are moving it here back to its original combination with Al- eiodes. A. rileyi is quite distinctive by being the only member of the species-group that is entirely orange; all other North Ameri- can species in the apicalis-group have at least some black coloration on the body. ACKNOWLEDGMENTS This research was supported by grant DEB-930- 6314 from the National Science Foundation. Addi- tional support was provided by supplemental Re- search Experience for Undergraduates (REU) grants in 1994, 1995, and 1996. Support was also provided by a Faculty Grant-in-Aid from the University of Wy- oming Research Office, U.W. Experiment Station Pro- ject WYO-256-90, and a CANACOL Foundation grant to the junior author (JCF). Additional thanks are due to Ms. Teresa Williams, of the Western Research In- stitute, for assistance with the Environmental Scan- ning Electron Microscope and photography. We also thank the curators of the museums mentioned in the Methods section for the loan of specimens used in this study. LITERATURE CITED Achterberg, C. van. 1991. Revision of the genera of the Afrotropical and W. Palaearctical Rogadinae Foerster (Hymenoptera: Braconidae). Zoolcgische Verhandelingen 273: 1-102. Achterberg, C. van. 1993. Illustrated key to the sub- families of Braconidae (Hymenoptera: Braconi- dae). ZiX'logisclic Vcrhamlchngcn 283: 1-189. Cresson, E. T. 1869. List of the North American spe- cies of the genus Aleiocia Wesmael. Transactions of the American Entwnological Societi/ 2: 377-382. Gahan, A. B. 1917. Descriptions of some new parasitic Hymenoptera. Proceedings of the United States Na- tional Museum 53: 195-217. Harris, R. A. 1979. A glossary of surface sculpturing. Occasional Papers in Entomology 28: 1-31. Marsh, P. M. 1979. Family Braconidae. Pp. 144-313, In K. V. Krombein, P. D. Hurd Jr., D. R. Smith, and B. D. Burks (eds.). Catalog of Hymenoptera in America North of Mexico, Smithsonian Institution Press, Washington, D.C. Marsh, P. M. 1989. Notes on Braconidae (Hymenop- tera) associated with jojoba (Snnmomisia chmen- sis) and descriptions of new species. Pan-Pacific Entomologist 65: 58-67. Marsh, P. M., S. R. Shaw and R. A. Wharton. 1987. An identification manual for the North American genera of the Family Braconidae (Hymenoptera). Memoirs of the Entomological Society of Washington 13: 1-98. Norton, E. 1869. American Hymenoptera. Catalog of the described Tenthredinidae and Uroceridae of North America. Transactions of the American En- tomological Society 2: 321-368. Papp, J. 1985. Contributions to the braconid fauna of Hungary, VII. Rogadinae (Hymenoptera: Bracon- idae). Folia Entomologica Hungarica 46: 143-164. Shaw, M. R. 1983. On[e] evolution of endoparasitism: the biology of some genera of Rogadinae (Bra- conidae). Contributions of the American Entomolog- ical Institute 20: 307-328. Shaw, M. R. 1994. Parasitoid host ranges. Chapter 7, pp. 112-144, In B.A. Hawkins and W. Sheehan (eds.), Parasitoid Community Ecology, Oxford Uni- versity Press, Oxford. Shaw, M. R. and T. Huddleston. 1991. Classifcation and biology of braconid wasps. Handbooks for the Identification of British Insects 7: 1-126. Shaw, S. R. 1993. Systematic status of Eucystomastax Brues and characterization of the Neotropical species. Journal of Hymenoptera Research 2: 1-11. Shaw, S. R. 1995. Braconidae. Chapter 12.2, pp. 431- 463, In P.E. Hanson and l.D. Cauld [eds.j. The Hymenoptera of Costa Rica, Oxford University Press, Oxford. Shaw, S. R. 1997. Subfamily Rogadinae s.s. Journal of Hymenoptera Research, Special Publication 1997: 4()2-412. Shaw, S. R., P. M. Marsh and J. C. Fortier. 1997. Re- vision of North American Aleiodes Wesmael (Part 1): the pulchripes Wesmael species-group in the New World (Hymenoptera: Braconidae, Roga- dinae). Journal of Hymenoptera Research 6: 10-35. Shenefelt, R. D. 1975. Braconidae 8: Exothecinae, Ro- gadinae. Pp. 1115-1262, In van der Vccht and R.D. Shenefelt, (eds.), Hymenopterorum Catologiis (novo cditio), W. Junk B.V., The Hague. Shenefelt, R. D. 1979. Some unusual Braconidae (Hy- menoptera). Proceedings of the Entomological Soci- ety of Washington 81: 125-134. J. HYM. RES. Vol. 7(1), 1998, pp. 74-83 Territoriality and Mating Behavior of Sphex pensylvanicus L. (Hymenoptera: Sphecidae) Frank E. Kurczewski Environmental and Forest Biology, State University of New York College of Environmental Science and Forestry, Syracuse, New York 13210-2778, USA Abstract. — Daily observations were made on nine individually marked males of Sphex pensyl- vanicus in upstate New York during 25 July-8 August 1982. Males occupied territories on or near a grate atop a storm sewer drain in which 12 females nested. They obtained honeydew at a maple tree and slept and fed on white sweet clover growing on a nearby hillside. The following main- tenance and reproductive activities of males were defined: (1) perching at stations, (2) swivelling, (3) cleaning, (4) wing raising, (5) defecating, (6) spontaneous flights, (7) feeding flights, (8) pounc- ing on conspecific males, (9) pursuit flights, (10) grappling, (11) trailing conspecific females, (12) clasping conspecific females, and (13) copulation. Aggressive interactions between territorial con- specific males occupied more time than all other maintenance and reproductive activities com- bined. Almost nothing was known about the behavior of male solitary wasps until Lin's (1963) study of male territoriality in Splie- cius speciosus (Drury), the cicada killer. Re- cent interest in male behavior surfaced fol- lowing the revival of Darwin's (1859) sex- ual selection theory, especially as cham- pioned by Trivers (1972) and his contemporaries. The activities of male sol- itary wasps are primarily aimed at obtain- ing matings. Males feed on nectar, rest on plants or in burrows, or bask in the sun when not in pursuit of females (Evans and O'Neill 1988). Males of most species of Sphecidae are free from parental duties and their reproductive success is solely determined by the number of eggs they fertilize. In other words, males of most species contribute little more than genes to their offspring. Male solitary wasps tend to emerge be- fore the females, a phenomenon known as protandry (Evans 1966). Circumstantial evidence suggests that females of most digger wasps mate only once during their lifetime (Alcock et al. 1978). Although nesting may extend for several weeks in certain species of Sphecidae (Hager and Kurczewski 1986, Kurczewski 1997), cop- ulations in most species take place only during the first week or two. The majority of contacts between males and females do not end in successful copulation (O'Neill 1979). Once females are actively nesting, they rebuff all attempted matings by males (Evans and O'Neill 1988). The rela- tive low fecundity of the females may mean that they actually gain little from additional matings (O'Neill 1985). Alcock et al. (1978) address the costs and benefits associated with multiple matings in spe- cies of aculeate Hymenoptera. Territoriality, as defined by spacing, maintenance of stations, and aggressive encounters between conspecific males, does not occur in all sphecid wasps (Hag- er and Kurczewski 1985). But, it can be ex- aggerated in some species (Minkiewicz 1934, Lin 1963, Evans and O'Neill 1988). Males may establish stations or territories near prominent landmarks on the ground (Astata, Minkiewicz 1934; Tachysphex, Kur- Volume 7, Number 1, 1998 75 czewski 1966) or emergence holes {Splie- ciiis, Lin 1963), or they may scent mark plants to attract females to a site (Philan- thus, Evans and O'Neill 1988). The estab- lishment and maintenance of stations or territories clearly facilitates mating be- tween resident males and females in the vicinity (Alcock et al. 1978). Studies on the behavior of male solitary wasps lag far behind those on female nest- ing behavior (Evans 1966, Alcock et al. 1978, Evans and O'Neill 1988). Male soli- tary wasps are often smaller, shorter lived, and do not maintain a nest making them more inconspicuous to the average ob- server (Kurczewski 1966, Hager and Kur- czewski 1985). Among the species of Sphe- cinae (R. Bohart and Menke 1963, 1976), many of the studies on male behavior in- volve the genus Ammophila (Turner 1912, Baerends 1941, G. Bohart and Knowlton 1953, Olberg 1959, Powell 1964, Hager and Kurczewski 1985). Little is known about territoriality and male behavior in the ge- nus Sphex including the nearctic S. pensyl- vauicus Linnaeus, the Great Black Wasp of John Bartram (Rau 1944). Rigley and Hays (1977) noted dominance, grappling, and attempted copulation in a few males of S. peiisi/lvanicus. Gillaspy (1962) described mating behavior in the nearctic S. tepane- cus Saussure. Janvier (1928) noted perch- ing and grappling in the neotropical S. la- treillei Lepeletier. Surprisingly, nothing is known of male behavior in a common nearctic species, S. ichneutiioneus (Linnae- us)(Brockmann 1980). My paper presents new and interesting information on the behavior of males of S. pensi/lvanicus and provides a verbal out- line of the mating system employed by this species. The literature on reproductive behavior in sphecids is depauperate and observations such as those reported herein are needed to rectify this situation. Terri- toriality and mating behavior in this spe- cies were observed concurrent with an in- vestigation on the sequential daily activi- ties of the females (Kurczewski 1997). 1 chronologically sequenced and described individual male behaviors, especially those connected with spacing, aggression, and copulation, and attributed possible significance to them, thereby adding to the overall knowledge of male behavior in the family Sphecidae. METHODS Males of S. pensylvanicus were observed on a daily basis, weather permitting, from 25 July to 8 August 1982 from 0730 to 2100 hrs (EDT). The study area was examined before and after these dates and there was no sign of male activity. One must as- sume, therefore, that the males had not emerged before 25 July and either had died or left the area after 8 August. Nine males each were color-coded by placing a drop of Tester's model paint on their me- soscutum with a tiny paint brush from which had been removed most of the hairs. The activities of these males were observed and recorded at or near a storm sewer grate for a total of 62.5 hrs during 10 of the 15 days that they were in evi- dence. The individual life spans of the nine males was only 10-14 days, with three males living an entire two-weeks- long period. I similarly marked all 12 fe- males from this aggregation thus yielding a secondary sex ratio of 3:2 in favor of fe- males. One male and one female each were collected before individually mark- ing them and placed as voucher speci- mens in the insect museum of the State University of New York College of Envi- ronmental Science and Forestry, Syracuse, New York. Territoriality in this species facilitated simultaneous observation of all nine males; however, some time was allocated to observe certain focal males more exten- sively during fixed observation periods. Individual male behaviors were described and chronologically sequenced. Particular emphasis was placed on reproductive be- havior as it occurred near a storm sewer drain in which the females nested. Less 76 Journal of Hymenoptera Research emphasis was placed on male mainte- nance behaviors such as feeding, sleeping, and basking in the sun that took place away from this site. EMERGENCE AND LOCATION OF STUDY Both males and females emerged from soil near a broken tile at the bottom of the sewer drain situated in an asphalt drive- way beside the Marcellus Senior High School, Marcellus, Onondaga County, New York. Three males emerged on 25 July 1982, two days before the first female appeared. The emergences of the other six males shortly thereafter (27-31 July 1982) were interspersed among those of early emerging females. Behavior that repre- sented aspects of territoriality such as spacing, maintaining stations, and aggres- sion between conspecific males was fo- cused on a 70 X 70 cm grate atop the 70 cm-deep drain in which the females nest- ed (Fig. 1, Kurczewski 1997). Males also visited a maple tree 13 m south of the sew- er drain in order to obtain honeydew, and slept and fed at a stand of Melilotus alba (white sweet clover) 55 m away (Fig. 2, Kurczewski 1997). Both males and females fed on the flowers of other plant species nearby but not as frequently as at the flo- rets of white sweet clover, probably be- cause this species was more abundant in the vicinity (Kurczewski 1997). The nine males left their sleeping roosts on white sweet clover daily on warm sun- ny days, occasionally fed on nectar or basked in the sun, and then made low cir- cling or figure-8 flights over the sewer grate as early as 0828-0837 hrs (EDT) at an air temperature as low as 16°C. They stationed themselves on and /or near the grate from 0845 to 1643 hrs, and once as late as 1818 hrs, at air temperatures of 16.5-26°C. Males interrupted this territo- rial activity to feed on the flowers of Mel- ilotus alba from 1125 to 1800 hrs at air tem- peratures of 23-26°C. They circled the drain in flight between 1819 and 2019 hrs. alit briefly on or entered through openings in the sewer grate, but then quickly exited and flew to sleeping roosts. Males did not maintain stations at or near the grate dur- ing this time. They did not make extensive and intricate flights for the purpose of ori- entation to the immediate environs as did the provisioning females (Kurczewski 1997). MALE ACTIVITY The following activities of male S. pen- sylvanicus were delineated: (1) perching at stations, (2) swivelling, (3) cleaning, (4) wing raising, (5) defecating, (6) spontane- ous flights, (7) feeding flights, (8) pounc- ing on conspecific males, (9) pursuit flights, (10) grappling, (11) trailing conspe- cific females, (12) clasping conspecific fe- males, and (13) copulation. Definition and significance of these activities are as fol- lows: (1) Perching at stations. — Males perched on or near the sewer grate with mid- and hindlegs outstretched yet raised and wings folded flat on the dorsum. They pe- riodically moved their head or anterior body from side to side. While perching, males moved their antennae or held them still in an upward and outward attitude. The forelegs were either "fishhooked" me- dially or held backward beneath the body, raised above the substrate, and moved back and forth slowly or in short rapid bursts. Positioning by certain males close to the grate openings gave them a decided advantage over males stationed farther away with respect to access to entering or exiting females. These males subsequently obtained more claspings and copulations than males stationed farther from the sew- er grate (see below). Following emergence and through the first week of nesting, males maintained stations near the female nesting site. Males spent much time at a preferred site or sta- tion; however, most males maintained several stations in close proximity moving from one to another throughout the day. Volume 7, Number 1, 1998 77 On 27 July 1982, five males perched at sta- tions on or near the sewer grate. Two of the males occupied opposite ends of the grate, being only 50-70 cm apart. The oth- er three males were located near the grate at cardinal points of the compass, each less than 1 m from the two males. The males near the grate were not allowed to approach the males on the grate any closer than this without being chased away. On the following day, four additional, newly emerged males took up stations near the grate less than 1 m from the older males. Nine males now maintained one or more stations within a diameter of 2 m from the grate. The two males stationed on the grate and a third male nearest the grate perched at their stations for periods of 20- 150 min (x = 37.9 min, n = 37 observa- tions), except for chasing away conspecific males, females, and other insects, and making periodic flights with no discern- ible function. The six most peripherally situated males, on the other hand, perched for an average of only 3.8 (a- nccus Saussure. Bulletin of the Brooklyn Entomolog- ical Society 57: 15-17. Hager, B. J. and F. E. Kurczewski. 1985. Reproductive behavior of male Ammophila harli (Fer- nald)(Hymenoptera: Sphecidae). Proceedings of the Lntomological Society of Washington 87: 597- 605. Hager, B. J. and F. E. Kurczewsl P > 0.50). The mean number of cells per nest was significantly different between the years with more cells in 1991 than in 1992 (F = 4.51, P = 0.03). Cell distribution paralleled nest distribution with significantly more cells in longer cavities (F = 15.42, P < 0.001) but a similar number of cells in all cavity diameters (F = 0.4, P = 0.66). Nests in 230 mm length cavities contained sig- nificantly more cells than did nests in 140 mm and they contained more than did nests in 80 m cavities (Table 1). The distribution of adults by cavity length was 69% in 230 mm cavities, 28% in 140 mm cavities, and 3% in 80 mm cav- ities and by cavity diameter was 32% in 6 mm cavities, 30% in 7 mm cavities, and 38% in 8 mm cavities. Five hundred and forty-nine adults (317 males and 232 fe- males) were produced (Table 2). In 1991, 312 adults (166 males and 146 females) were produced, and in 1992, 248 adults (151 males and 86 females) were pro- duced. The distribution of males or fe- males by cavity length was similar to the distribution of all individuals. However, the distribution of males or females by cavity diameter showed different patterns with the number of males decreasing with increasing cavity diameter (6 mm — 44%, 7 mm— 30%, and 8 mm— 26%) and the number of females increasing with in- creasing cavity diameter (6 mm — 17%, 7 mm— 30%, and 8 mm— 53%). Both male and female weights showed a significant year by cavity length by cav- ity diameter interaction (males F = 3.81, P = 0.02; and females F = 3.88, P = 0.02) and were thus separated by year to ex- Volume 7, Number 1, 1998 87 Table 2. Male and female production by Osmia //sf- nnrm propmqun in cavities with 6, 7, and 8 mm di- ameters and 80, 140, and 230 mm lengths in 1991 and 1992 at Lake City, California. Males Females Didnieter IMUI 1W2 Ti.t.il I9»l loo: Ti.tal 80 mm length 6.0 0 5 5 0 1 1 7.0 0 1 1 0 0 0 8.0 1 2 3 2 5 7 Total 8 8 140 mm length 6.0 27 26 53 7 3 10 7.0 18 13 31 16 5 21 8.0 14 4 18 13 8 21 Total 102 52 230 mm length 6.0 46 35 81 20 8 28 7.0 34 28 62 24 24 48 8.0 26 37 63 64 32 96 Total 206 Totals 172 166 151 317 146 86 232 amine the effects of cavity length and di- ameter on individual vv^eights (Table 3). In 1991, male weights differed signifi- cantly among cavity lengths (F = 23.5, P = 0.0001), cavity diameters (F = 7.79, P = 0.001), and the cavity length by diameter interaction (F = 130.5, P = 0.01) (Table 3). Males in 80 mm length cavities were heavier than males in 140 mm or 230 mm cavities. Males in 8 mm diameter cavities v^ere heavier than males in 7 mm or 6 mm cavities. However, in 1992 male weights were not significantly different in cavity lengths, cavity diameters, or cavity length by diameter interaction. In 1991, female weights in were signifi- cantly different for cavity lengths (F = 6.75, P = 0.002), cavity diameters (F = 2.72, P = 0.07), and the cavity length by diameter interaction (F = 2.94, P = 0.06) (Table 3). Females in 80 mm cavities were significantly heavier than females in 140 mm or 230 mm. Females in 6 mm cavities were significantly lighter than females in 7 mm or 8 mm cavities. However, in 1992 female weights were not significantly dif- ferent among cavity diameters or in cavity lengths and or their interaction. The sex ratios based on mean weight of all males and females ranged from 1.2:1 to 1.9:1 (males: females) and were slightly different between years (Table 4). Sex ra- tios increased in both years with increas- ing cavity lengths (1.2:1 to 1.8:1) but not diameters (1992—1.6:1 to 1.9:1 and 1991— 1.6:1 to 1.5:1). The amount of mud used for each cell per nest was significantly different be- Table 3. Mean weights and standard deviations (mg) of male and female Ounia lignnrui prop'inqua produced in cavities with 6, 7, and 8 mm diameters and 80, 140, and 230 mm lengths in 1991 and 1992 at Lake City, California. * equals one individual. M. ales Females Parameter HMI IW2 IWl igq: Cavity diameter 6.0 35.8 ± 5.7 34.2 ± 4.8 56.8 ±11.2 54.8 ± 7.6 7.0 36.0 ± 7.3 35.1 ± 5.8 63.3 ± 9.1 59.7 ± 11.6 8.0 44.9 i 9.5 32.6 ± 6.9 Cavity length 68.9 ± 12.7 61.7 ± 12.6 80 70.9' 32.6 ± 5.6 83,7 ± 5.5 53.2 ± 4.1 140 40.6 ± 6.5 35.3 ± 5.1 67.9 -t 10.2 62.2 ± 12.1 230 33.7 ± 5,2 33.4 ± 6.1 Total 61.8 ± 11.7 60.6 ± 12.3 Total 37.7 ± 5.1 33.9 ± 5.8 64.1 ± 10.1 60.3 ±11.4 88 Journal of Hymenoptera Research Table 4. Sex ratio of Osiiiin Ui;iuuiii prcpiiiqun pro- duced in cavities with 6, 7, and 8 mm diameters and 80, 140, and 230 mm lengths in 1991 and 1992 at Lake City, California. Table 5. Percent mortality in Osiiiin Ugiuirin pircpin- qua cells produced in cavities with 6, 7, and 8 mm diameters and 80, 140, and 230 mm lengths in 1991 and 1992 at Lake City, California. P.ir.inicttT mm um: DianifltT |W1 IMu; IoCjI Cavity diameter 80 mm length 6.0 1.6:1 1.6:1 6.0 100.0 14.3 71.4 7.0 1.7:1 1.7:1 7.0 — 83.3 83.3 8.0 1.5:1 1.9:1 8.0 0.0 30.0 23.1 Cavity length 140 mm length 80 1.2:1 1.6:1 6.0 20.9 23.7 22.2 140 1.7:1 1.8:1 7.0 12.8 18.1 14.7 230 1.8:1 1.8:1 8.0 42.5 7.6 35.0 Totals 230 mm length Total 1.7:1 1.8:1 6.0 34.6 21.8 30.1 7.0 17.1 16.1 16.6 8.0 23.1 11.5 18.5 tween years (F = 7.32, P = 0.008) with less mud used in 1991 (mean 92.2 ± 33.2 mg Total for: Total for: TOTAL 80 6 mm 60.0, mm 31.0, 28.1 140 mm 23.7, 7 mm 18.1, 18.5 230 8 mm 21.7 mm 22.4 24.2 Tjor-mic IQQO- mc^r\ 1 C;c; T -1- Ql^ ' I t-v-\rr\ \ A^^A investment per cell by cavity length and diameter was also significantly different (cavity length F = 6.46, P = 0.002, cavity diameter F = 16.4, P = 0.001). There was more mud per cell in the 80 mm cavities (163.3 ± 85.0 mg) when compared to the 140 mm (116.9 ± 77.8 mg) and 230 mm (117.7 ± 73/2 mg). The 6 mm diameter cavity nests contained less mud per cell (88.7 ± 36.9 mg) than either the 7 mm (146.0 ± 68.9 mg) or 8 mm (145.6 ± 102.9 mg). Bees using 230 mm length cavities placed the first cell not at the bottom of the cavity but at an average of 21.4 ± 35.2 mm from the bottom of the cavity. Bees nesting in 80 or 140 mm length cavities place the first cell at the bottom of the cav- ity. The last cell in a nest was significantly closer to the entrance in 80 mm cavities (mean 33.7 ± 15.8 mm) than in either 140 mm (mean 50.7 ± 27.5 mm) or 230 mm (mean 75.1 ± 54.4 mm) cavities (F = 5.17, P = 0.007). For both the placement of the first and last cell in a nest, neither year, cavity diameter, or any of the interactions were significant. The percent mortality averaged 24.2% for all cells produced and was not signif- icantly different between years or among cavity lengths, cavity diameters, or any of the interactions (Table 5). Chalk brood {Ascosphaera torchioi Youssef and Mc- Manus) caused the greatest loss in 1991 (15.6%) whereas egg death or failure to hatch caused the greatest loss in 1992 (7.5%) (Table 6). Chalk brood was the overall highest mortality factor (10.9%). Nesting parameters of ten species of megachilid bees have been reported on where they were provided with a choice of cavity lengths (Table 6). There was a positive relationship between female mean body width and mean preferred cavity diameter (Y (cavity diameter) = 1.10 -H 1.50X (body width); F = 2.62, P = 0.144). The relationship became significant when Osmia marginata was removed from analysis (Y (cavity diameter) = -0.48 + 1.84X (body width); F = 16.6, P = 0.005). Osmia marginata does not place cells in a linear series in large diameter cavities, but fits cells to the cavity dimension that al- lows a maximum use of the space (Tepe- dino and Parker 1983). Female mean body length and mean cavity length also showed a significant positive relationship Volume 7, Number 1, 1998 89 Table 6. Female size and nest parameters of megachilid bees from studies where different length cavities were available for nesting. Female size Nest characteristics Mean wn dth Mean length Mean diameter Mean lenj^th Mean » Spi-vK's (mm) (mm) (mm) (mm) lit cells Ketereiue Osiriin iinaii 3.2 8.7 5.3 9.9 8.9 Maeta 78 tiiunii 4.1 11.0 6.8 14.9 9.0 Maeta 78 cornifrons 3.5 9.7 5.8 14.8 8.0 Maeta 78 pctiicornis 4.3 12.5 6.7 14.6 6.8 Maeta 78 L'xcazmta 3.8 10.7 6.5 15.0 7.2 Maeta 78 corntita 4.0 12.0 8.0 17.1 4.0 Bosch 94 iiuu\;iniit{i 3.5 9.0 9.0 9.0 8.2 Tepedino & Parker 83 lt};iuirm 3.8 11.5 6.9 18.2 Hoptlitis 6.6 present fulgida 3.5 9.7 6.0 8.5 Mff;iuhik 4.9 Tepedino & Parker 84 rotundata 3.0 8.5 4.9 7.8 6.3 Gerber & Klost. 72 (Y (cavity length) = -9.95 + 2.22X (body length); F = 17.18, P = 0.003). There was no relationship between female body width or length and the mean number of cells per nest. DISCUSSION Females of Osmia I. propinqua chose sig- nificantly deeper drilled holes in trap- nests when presented with an equal dis- tribution of cavity lengths for nesting. These deeper holes were filled with more cells and offspring. However, females showed no preference for a particular di- ameter trap-nest from within the range available. This nest selection pattern was observed in both years. In similar studies, Bosch (1994) found a significant prefer- ence in Osmia cornuta for longer cavities for nesting (12, 15, or 21 cm, cavity di- ameter was 8 mm) and produced more cells in them. Osmia marginata Michener females preferred nest-traps in drilled el- derberry {Snmbuciis spp.) stems with the longest (90 mm) and widest (9 mm) cavi- ties (Tepedino and Parker 1983). They pro- duce significantly more cells in them. Te- pedino and Parker (1984) observed the same selection pattern in Hoplitis fulgida (Cresson) nesting in drilled elderberry stems. The opposite cavity usage pattern was observed in the completed nests of Megachile rotundata (F.) which decreased from 100% in 1.25 and 2.5 cm length cav- ities to 16% in 15 cm cavities (Stephen and Osgood 1965). However, nest utilization increased as cavity diameter increased from 4.0 to 6.0 mm. Gerber and Kloster- meyer (1972) also found Megachile rotun- data to use more short (4 cm compared to 8, 12 or 16 cm cavity lengths) trap-nests for nesting in a three years study. How- ever, unlike the Stephen and Osgood (1965) results, Gerber and Klostermeyer (1972) found more cells were produced in 8 to 16 cm length cavities. Maeta (1978) provided Osmia imaii Hir- ashima, O. taurus Smith, O. cornifrons (Ra- doszkowski), O. pedicornis Cockerell, and O. excavata Alfken with a broad selection of reed cavities for nesting (4 to 11.9 mm in diameter and 3 to 33 cm in length). His presentation of nesting materials attempts to represent what is most likely available to the species in nature and consequently the usage preference observed best repre- sents the species natural usage patterns (Table 6). 90 Journal of Hymenoptera Research The general pattern of cavity choice suggests that female size dictates her choice in both the diameter and length of cavity. This assumes that vk'ithin the nat- ural habitat there exists a variety of cavi- ties for nesting and that females visit sev- eral cavities before making a selection. Se- lection of a short cavity will require find- ing a second or even third cavity to continue the nesting process. Selection of a long cavity simply requires the female to initiate cell construction at the "aver- age" depth to maximize the cells or off- spring produced. The bee is leaving un- used cavity behind the first cell. The pres- ence of an empty space behind the last cell was evident in only in the 230 mm long cavities. Maeta (1978) found an increasing percentage of empty spaces with increas- ing cavity length in all five Osmia species studied. Cavity selection for nesting affects the general population structure of the spe- cies. All studies show that selection of long wide cavities allows females to pro- duce more cells with larger offspring, and more females (Stephen and Osgood 1965, Gerber and Klostermeyer 1972, Maeta 1978, Tepedino and Parker, 1983, 1984, Frohlich and Tepedino 1986, Tepedino and Torchio 1989, Sugiura and Maeta 1989, Bosch 1994). In bee species that ex- cavate a tunnel in a twig or stem and then construct and provision linear series of cells there is no relationship between tun- nel diameter or tunnel length and the sex ratio of the offspring (Garofalo et al. 1981 for Lithiirgiis, Johnson 1988 for Ceratina, and Watmough 1983 for Xylocopa). Why should short narrow cavities be se- lected by any individual females? The an- swer is variation both in body size and cavity size. Females are selecting from an large assortment of potential nest cavities the cavity that best matches their body width configuration. Selection of short cavities is difficult to interpret. Why should a female invest more time in for- aging for and constructing cell partitions and nest plugs than in offspring produc- tion? Jayasingh and Taffe (1982) and Rust (1993) have reported on the greater cost to produce offspring in short cavities. Rust (1993) has also shown that nest plugs cost more to produce than cell partitions in O. /. pwpinqiia and Osmia rihifloris Cresson. He suggested that a nest should contain four or more cells to equalize the extra cost of nest plug. Individuals nesting in the 80 mm length cavities were producing on average only 2.8 cells and used signif- icantly more mud per cell than in the oth- er cavity length nests. The selection of a short cavity also im- plies that the female must spend addition- al time searching for a second and perhaps third cavity for nesting. Naturally occur- ring nest sites must be considered as clumped; beetle borings in dead trees and logs, shrubs with hollow stems, etc. This clumped distribution suggests that new site searching may be minimal. Tepedino and Torchio (1989) showed no pattern or preference for a given diameter nest when O. I. pwpinqiia searched for and initiated a second or third nest. Parasite or predator load may be a strong selective factor favoring females that select several different nest sites. Both parasite and predator build-up can be- come a serious problem with high mortal- ity in commercial populations of cavity nesting bees (Torchio 1970, 1972; Stephen and Undurraga 1978; Eves et al. 1980). Fe- males selecting one long cavity will be at a disadvantage in a high density parasite or predator site. The overall immature mortality in O. I. propinqua was low, less than 30%, and is similar to other reports on cavity-nesting, non-social bees and wasps (Krombein 1967, Danks 1971, Raw 1972, Cross et al. 1975, Freeman 1977, Maeta 1978, Taffee 1979, Smith 1979, Jayasingh and Freeman 1980, Tepedino and Frohlich 1982, Tepe- dino and Parker 1983, 1984). The various mortality agents or factors were unrelated to either nest diameter of length in the Volume 7, Number 1, 1998 91 present study. Chalk brood was the only agent to showed a substantial yearly change. Rust and Torchio (1991) also re- ported extreme year to year variations in chalk brood mortality within populations of O. I. propinqua. Tepedino and Parker (1983) reported a significantly greater mortality due to developmental failure in large diameter, long nests of O. marginata. They suggest the reason to be a departure in cell construction from a linear array of cells to an array of cells perpendicular to the long axis of the cavity. There was no difference in parasite or predator attacks in the various nests. In Hoplitis fulgida, Te- pedino and Parker (1984) found signifi- cantly less mortality in the short, least used nests. Since several species of Osmia and Megachile rotuiidata have been developed for commercial pollination (Torchio 1987, 1990), the choice of the appropriate cavity size is paramount to maximize pollinator production in a management strategy. The economics of producing effective commer- cial nest cavities requires the availability of materials and tools to manufacture the "average" cavity for a commercial popu- lation. This cavity may not be the opti- mum for the species. The choice of the standard length (15 to 17 cm) drill bits and the difficulties of obtaining wood with grain pattern suitable for the manufacture of "bee boards" with many straight, close, deep holes (greater than 17 cm) resulted in the production the commercial nest cav- ity for O. /. propinqua of a 7 X 170 mm paper soda straw inserted into a 8 X 170 mm hole in redwood (see Torchio 1982a, 1982b for details). This nest cavity allows for the production of sustainable popula- tions of O. /. propinqua for both apple and almond pollination. My study suggests that holes deeper than 170 mm should be provided for O. I. propinqua for maximize its offspring production even in a com- mercial situation. ACKNOWLEDGMENTS I thank Phil Torchio and Jordi Bosch for discus- sions that lead to this research and the development of the manuscript. Two anonymous reviewer inputs were most helpful and appreciated. The research would not have been possible without the use of the orchards and kind assistance of the Kent and Libby Upchurch and Lynn Nardella and Jason and Sophia Sheppard, all of Lake City, California. LITERATURED CITED Bosch, J. 1994. Osniii! coniutn Latr. (Hym., Mcgachili- dae) as a potential pollinator in almond orchard. Journal of Applied Entomology 117:151-157. Bosch, J., M. Bias and A. Lacasa. 1992. Osmia cornuta (Hymenoptera; Megachilidae), un nuevo polini- zador para los almendros. Fniticiilturn Profession- al 44:65-71. Charnov, E. L. 1982. The theory of sex allocation. Prince- ton University Press, Princeton, New Jersey, USA. Charnov, E. L., R. L. Los-den Hartogh, W. T. Jones and J. van den Assem. 1981. Sex ratio evolution in a variable environment. Nature 289:27-33. Cody, R. P. and J. K. Smith. 1991. Applied Statistics and the SAS programming language. Prentice-Hall, Englewood Cliffs, New Jersey, USA. Cowan, D. P. 1981. Parental investment in two soli- tary wasps Ancistrocerus adiahatus and Euodyne- rus foraminatus (Eumenidae: Hymenoptera). Be- hai'orial Ecotog}/ and Sociobiology 9:95-102. Cross, E. A., M. G. Smith and T. R. Bauman. 1975. Bionomics of the organpipe mud-dauber, Trypox- ylon politum (Hymenoptera: Sphecoidea). Annals of the Entomological Society of America 68:901-916. Danks, H. V. 1970. Biology of some stem-nesting acu- leate Hymenoptera. Transactions of the Ro\/at Eii- tomologi/ Society of London 122:323-399. Danks, H. V. 1971. Populations and nesting-sites of some aculeate Hymenoptera nesting in Rulnis. Journal of Animal Ecology 40:63-77. Evans, H. E. and M. J. West-Eberhard. 1970. The waspjs. University of Michigan Press, Ann Arbor, Michigan, USA. Eves, J. D., D. P. Mayer and C. A. Johansen. 1980. Parasites, predators, and nest destroyers of the alfalfa leafcutting bee, Megachile rotundata. West- ern Regional Extension Publication 32:1-15. Fisher, R. A. 1958. The genetical theory of natural selec- tion. 2nd. Dover, New York, New York, USA. Freeman, B. E. 1977. Aspects of the regulation of size of the Jamaican population of Sceliphron assimile Dahlbom (Hymenoptera: Sphecidae). journal of Animal Ecology 46:231-247. Freeman, B. E. 1980. Parental investment, maternal size and population dynamics of a solitar)' wasp. American Naturalist 117:357-362. 92 Journal of Hymenoptera Research Frohlich, D. R. and V. J. Tepedino. 1986, Sex ratio, parental investment, and interparent variability in nesting success in a solitary bee. Ei'clntioii 40: 142-151. Garofalo, C. A., E. Camillo, M. J. O. Campos, R. Zuc- chi and J. C. Serrano. 1981. Bionomical aspects of Lithurgus conimbae (Hymenoptera, Megachili- dae), including evolutionary considerations on the nesting behavior of the genus. Brazilian Jour- nal of Genetics IV 2:165-182. Gauld, 1. and B. Bolton. 1988. Tlw Hymcuoptcra. Ox- ford University Press, Oxford, Great Britain. Gerber, H. S. and E. C. Klostermeyer. 1972. Factors affecting the sex ratio and nesting behavior of the alfalfa leaf-cutter bee. Washington Agricultural Ex- periment Station Tcchanical Bulletin 73:1-11. Jayasingh, D. B. and B. E. Freeman. 1980. The com- parative population dynamics of eight solitary bees and wasps (Aculeata; Apocrita; Hymenop- tera) trapnested in Jamaica. Biotropica 12:214-219. Jayasingh, D. B. and C. A. Taffe. 1982. The biology of the eumenid mud-wasp Pachodynerus nasidens in trapnests. Ecological Entomology 7:283-289. Johnson, M. D. 1988. The relationship of provision weight to adult weight and sex ratio in the soli- tary bee, Ceratina calcarata. Ecological Entomology 13:165-170. Johnson, M. D. 1990. Female size and fecundity in the small carpenter bee, Ceratina calcarata (Robertson) (Hymenoptera: Anthophoridae). Journal of the Kansas Entomological Society 63:414-419. Krombein, K. V. 1967. Trap-nesting wasps and bees: life histories, nests, and associates. Smithsonian Insti- tution, Washington, D.C., USA. Maeta, Y. 1978. Comparative studies on the biology of the bees of the genus Osmia of Japan, with special reference to thier managements for pol- linations of crops (Hymenoptera: Megachilidae). Tohoku National Agricultural Expvriment Station Bulletin 57:1-221. Malyshev, S. I. 1935. The nesting habits of solitary bees. Eos (Spain) 11:201-309. Michener, C. D. 1947. A revision of the American spe- cies of Hoplitis (Hymenoptera, Megachilidae). Bulletin of the American Museum of Natural History 89:259-317. Mitchell, T. B. 1962. Bees of the eastern United States. Volume II. North Carolina Agricultural Experinwnt Station Techincal Bulletin 152:1-557. Peters, D. S. 1977. Systematik und Zoogeographie der west-palaarktischen Arten von Osmia Panzer 1806 s. str., Monosmia Tkalcu 1974 und Orientos- mia n. subgen. (Insecta: Hymenoptycra: Mega- chilidae). Senckenbaergutna Bwlogia 58:287-346. Phillips, J. K. and E. C. Klostermeyer. 1978. Nesting behavior of Osmia lignaria propinqua Cresson (Hymenoptera: Megachilidae). journal of the Kan- sas Entomological Society 51:91-108. Rau, P. 1928. Field studies in the behavior of the non- social wasps. Transactions of the Academy of Sci- ence of St. Louis 25:325-489. Rau, P. 1937. The life history of Osmia lignaria and Osmia cordata, with notes on Osmia conjuncta. An- nals of the Entomological Society of America 39:324- 342. Raw, A. 1972. The biology of the solitary bee Osmia rufa (L.) (Megachilidae). Transactions of the Royal Eutoniological Society of London 124:213-229. Rust, R. W. 1974. The systematics and biology of the genus Osmia, subgenera Osmia, Chalcosmia, and Cephalosmia (Hymenoptera: Megachilidae). Wass- inann Journal of Biology 32:1-93. Rust, R. W. 1993. Cell and nest construction costs in two cavity-nesting bees {Osmia lignaria propinc]ua and Osmia ribifloris bicdermannii) (Hymenoptera: Megachilidae). Annals of the Entomological Society of America 86:327-332. Rust, R. W. and P. F. Torchio. 1991. Induction of As- cosphaera (Ascomycetes: Ascosphaerales) infec- tions in field populations of Osmia lignaria pro- pinqua Cresson (Hymenoptera: Megachilidae). Pan-Pacific Entomologist 67:251-257. Sandhouse, G. A. 1939. The North American bees of the genus Osmia. Memoirs of the Entomological Society of Washington 1:1-167. SAS. 1990. SAS/stat users guuie. Volumes 1 and 2. SAS Institute, Cary, North Carolina, USA. Smith, A. 1979. Life strategy and mortality factors of Sceliphron laetum (Smith) (Hymenoptera: Spheci- dae) in Australia. Australian Journal of Ecology 4: 181-186. Sokal, R. R. and F. J. Rohlf. 1969. Biometry. Freeman, San Francisco, California, USA. Stephen, W. P., G. E. Bohart and P. F. Torchio. 1969. The biology and external morphology of bees. Oregon Agricultural Experiment Station, Corvallis, Ore- gon, USA. Stephen, W. P. and C. E. Osgood. 1965. Influence of tunnel size and nesting medium on sex ratios in a leaf-cutter bee, Megachile rotundata. Journal of Economic Entomology 58:965-968. Stephen, W. P. and J. M. Undurraga. 1978. Chalk brood disease in the leafcutting bee. Oregon State Universiti/ Agricultural Experinwnt Station Bulletin 630. Sugiura, N. and Y. Maeta. 1989. Parental investment and offspring sex ratio in a solitary mason bee, Osmia corniforns (Radoszkowski) (Hymenoptera: Megachilidae). Japaiwse Journal of Entomology 57: 861-875. Taffe, C. A. 1979. The ecology of two West Indian species of mud-wasps (Euenidae: Hymenoptera). Biological Journal of the Linnean Society 11:1-17. Tepedino, V. J. and D. R. Frohlich. 1982. Mortality factors, pollen utilization, and sex rat'O in Mega- chile pugnala Say (H:ymenoptcra: Megachilidae), Volume 7, Number 1, 1998 93 a candidate for commercial sunflower pollina- tion, journal of the New York Entomological Society 90:269-274. Tepedino, V. J. and F. D. Parker. 1983. Nest size, mor- tality and sex ratio in Osinia marginata Michener. Southwest Entomologist 8:154-167 Tepedino, V. ]. and F. D. Parker. 1984. Nest selection, mortality and sex ratio in Hoplitis flugiila (Cres- son) (Hymenoptera: Megachilidae). Journal of tlie Kansas Entomological Socu-t\/ 57:181-189. Tepedino, V. J. and P. F. Torchio. 1982a. Temporal variability in the sex ratio of a non-social bee, Osmia lignaria propinqua Cresson: Extrinsic deter- mination or the tracking of an optimum? Oikos 38:177-182. Tepedino, V. ]. and P. F. Torchio. 1982b. Phenotypic variability in nesting success among Osmia lig- naria propinqua females in a glasshouse environ- ment (Hymenoptera: Megachilidae). Ecologial En- tomology 7:453^62. Tepedino, V. J. and P. F. Torchio. 1989. The influence of nest-hole selection on sex ratio and progeny size in Osmia lignaria propinqua (Hymenoptera: Megachilidae). Annals of the Entomological Society of America 82:355-360. Torchio, P. F. 1970. The biology of Sapi/ga pumila Cres- son and its importance as a parasite of the alfalfa leafcutter bee, Megachile rotundata (Fabricius). Ar- kansas Agricultural Experiment Station Miscella- neous Publication 127:84-90. Torchio, P. F. 1972. Sapyga pumila Cresson, a parasite of Megachile rotundata (F.) (Hymenoptera: Sapy- gidae; Megachilidae). II Methods for control. Me- landeria 10:23-30. Torchio, P. F. 1976. Use of Osmia lignaria Say (Hy- menoptera: Megachilidae) as a pollinator in an apple and prune orchard. Journal of the Kansas Entomological Society 49:475-t82. Torchio, P. F. 1982a. Field experiments with Osmia lignaria propinqua Cresson as a pollinator in al- mond orchards: II, 1976 studies, journal of the Kansas Entomological Society 54:824-836. Torchio, P. F. 1982b. Field experiments with the pol- linator species, Osmia lignaria propinqua Cresson as a pollinator in almond orchards: III, 1977 (Hy- menoptera: Megachilidae). journal of the Kansas Entomological Society 55:136-144. Torchio, P. F. 1984a. Field experiments with the pol- linator species, Osmia lignaria propinqua Cresson, in apple orchards: III, 1977 studies, journal of the Kansas Entomological Societ}/ 57:517-521. Torchio, P. F. 1984b. Field experiments with the pol- linator species, Osmia lignaria propinqua Cresson, in apple orchards: IV, 1978 studies, journal of the Kansas Entomological Society 57:689-694. Torchio, P. F. 1985. Field experiments with the polli- nator species, Osmia lignaria propinqua Cresson, in apple orchards: V (1979-1980), methods of in- troducing bees, nesting success, seed counts, fruit yields (Hymenoptera: Megachilidae). journal of the Kansas Entomological Society 58:448-464. Torchio, P. F. 1987. Use of non-honey bee species as pollinators of crops. Proceedings of the Entomolog- ical Society of Ontario 18:1-32. Torchio, P. F. 1989. In-nest biologies and develop- ment of immature stages of three Osmia species. Annals of the Entomological Society of America 82: 599-615. Torchio, P. F. 1990. Diversification of pollination strategies for U.S. crops. Environmental Entomol- ogy 19:1649-1656. Torchio, P. F. and V. J. Tepedino. 1980. Sex ratio, body size and seasonality in a solitary bee, Osmia lignaria propinqua Cresson (Hymenoptera: Mega- chilidae). Evolution 34:993-1003. Watmough, R. H. 1983. Mortality, sex ratio and fe- cundity in natural populations of large carpenter bees {Xylocopa spp.). journal of Animal Eai/n^n/ 52: 111-125. Yasumatsu, K. and Y. Hirashima. 1950. Revision of the genus Osmia of Japan and Korea. Mushi 21: 1-21. J. HYM. RES. Vol. 7(1), 1998, pp. 94-101 A New Species of the Baltic Amber Bee Genus Electrapis (Hymenoptera: Apidae) Michael S. Engel Department of Entomology, Comstock Hall, Cornell University, Ithaca, NY 14853, USA Abstract. — Electrapis stilbonota, a new species of fossil bee is described and figured from two female specimens preserved in a single piece of Eocene Baltic amber. The species is assigned to a new subgenus, Melikertes n. subg., characterized by the sparse hairs of the scutellum, outwardly curved scape, few distal hamuli, absence of hind tibial spurs, tear-drop shaped tegula, and absence of setal bands on the apices of the metasomal terga. The specimens of E. stilbonota are morpho- logically workers and were presumably from a highly eusocial colony. The classification of Elec- trapis among apines is briefly discussed, and the subtribe Electrapina proposed to accommodate the genus. The proposal that Electrap>is and its presumed sister. Apis, coexisted in time is briefly examined and found to be unsupported. The Eocene Baltic amber contains a fas- cinating, although uncommon, bee fauna. Those few specimens that are known pres- ent the picture of an assemblage of groups unlike anything seen today. Of the species represented in the Baltic amber only one is currently assigned to a modern genus, this being Andrena zvrisleyi Salt (1931), al- though the generic assignment of this spe- cies is of considerable question and it is possibly a melittid (Michener and Poinar 1996). The remainder, however, are as- signed to extinct genera whose affinities are difficult to ascertain and in some cases cannot confidently be placed to tribe. By comparison, bees of the Dominican amber, which is Oligo-Miocene in age (Grimaldi 1995), are referable to modern day genera or extinct groups closely allied to extant genera (Engel 1995, 1996, 1997, Michener and Poinar 1996, Rozen 1996). In 1909 Prof. Theodore D.A. Cockerell described a number of Baltic amber Hy- menoptera among which was the genus Electrapis (1909a). The genus is a member of the corbiculate apine tribe Apini which contains only one other genus, the familiar honey bees (Apis L.). Electrapis was erected to accommodate the type species Apis me- liponoides Buttel-Reepen (1906) which, as the specific epithet suggests, possessed characters both Buttel-Reepen and Cock- erell took to be intermediate between the Apini and their sister tribe, the Meliponini (the stingless bees). Since the time of its description, Electrapis has acquired a total of nine species segregated into three sub- genera. Table 1 summarizes the current classification of the known species. Herein I describe a tenth species of Elec- trapis and assign it to a new subgenus, Me- likertes. In the descriptions the following abbreviations are used for morphological terms: F, flagellomere; S, sternum; T, ter- gum. All measurements were made using an ocular micrometer on a WILD-M5a mi- croscope and are in millimeters. All mea- sures are approximate since the best po- sition for viewing a specific structure was not always achievable owing to the cur- vature of the amber surface. Measure- ments which were not possible to make for a given specimen are indicated by an asterisk (*). Values given in the specific de- scription are for the holotype with the cor- responding measure of the paratype indi- cated in brackets. Volume 7, Number 1, 1998 95 Table 1. Current classification of £/t'cfra;ns species. ElL'ctnipis fnoarii (Menge) is tentatively included in the subgenus Mclikertes. Subgenus Species KfU'rtTUf Eleclnipis s.str. nielipoiioidcs Buttel-Reepen 1906 Electrapis s.str. tornquisti Cockerell 1909b Electrapis s.str. apciites Manning 1960 Electrapis s.str. ijiiinitn Kelner-Pillault 1970 Electrapis s.str. hoinhusohlcs Kelner-Pillault 1974 Protohvnbiis inciecisiis Cockerell 1909a Protobomhus tristcllus Cockerell 1909a Roiiss}/ana pmlmiiickeuensis Roussy 1937 Melikcrtes stilhonota present study 7 proava Menge 1856 Genus Electrapis Cockerell Melikertes Engel, new subgenus Diagnosis. — Roussyana-like species with- out dense pubescence covering the scutel- lum; hairs generally sparse, never obscur- ing integument. Clypeus flat. Labrum U- shaped, broader than long, with apical fringe of simple hairs, hairs laterally short, becoming longer by middle. Labial palpus four segmented, basal segment longest, al- most as long as following three segments combined. Minute flabellum at apex of glossa. Antennae set well below mid-line of face; scape slightly curved outwards at apex, inner concave surface without pu- bescence; pedicel longer than Fl; Fl as long as F2 and F3 combined. Compound eyes bare. Face relatively flat. Vertex scarcely elevated above ocelli; preoccipital ridge with a weak carina behind vertex, becoming rounded by gena (Fig. 3). Meso- scutal anterior border weakly rounded, nearly straight; median line and notauli not apparent, parapsidal lines faintly evi- dent. Tegula tear-drop shaped, with blunt apex pointing posteriorly. Scutellum not bulging, surface flat, not reaching back to obscure portions of the metanotum; ante- rior border nearly straight medially, pos- terior border broadly rounded. Strigular concavity set on a slightly protuberant shelf; velum and malus simple, velum not divided. Mesocoxae separated medially. Hind tibia without inner apical spurs; cor- bicula comprising apical three-quarters of hind tibia, posterior apical border round- ed; corbicular surface glabrous and not strongly depressed, with a few sparsely scattered long simple hairs, otherwise hairs restricted to tibial lateral borders, such hairs long and branched; inner sur- face of tibia with a dense medial field of extremely short, simple hairs; strong ras- tellum on inner apical surface of hind tib- ia, extending the full width of tibial apex; penicillum absent; hind basitarsus rough- ly quadrangular, about as broad at base as at apex; attachment to tarsomere 2 set forth on short distal process on anterior border; auricle present; inner surface of hind basitarsus with series of stiff setal rows (as in Apis); claw with minute tooth on lower third of irmer margin. Pterostig- ma small; distal wing venation strong; posterior border of second submarginal cell extended posteriorly (Fig. 4), greatly narrowed anteriorly; basal vein basad cu- a, vein relatively straight; cu-a curved out- wards, not orthogonal with Cu or A; less than 10 hamuli on anterior margin of hind wing; jugal lobe present, just over one half length of vannal lobe; jugal and vannal in- cisions shallow. Metasoma without distal rows of hair on terga (Fig. 3). Ti/pe species. — Electrapis (Melikertes) stilhon- ota Engel, new species, present designation. Etymology. — ^The subgeneric name is taken from the mythology of ancient Crete. Meli- kertes (meaning "honey-cutter") was associ- ated with Corinth and Melissa, priestess to the mother-goddess Demeter. Remarks. — Electrapis proava is possibly mis- placed in Roussyana and should be included in Melikertes. Official transfer of this species will have to wait, however, until £. proava can be studied in more detail. Electrapis (Melikertes) stilhonota Engel, new species (Figs. 1-5) Description. — FEMALE (male un- known): Body form Trigona-\\ke. Total 96 Journal of Hymenoftera Research ^0S-- Figs. 1-2. Eleclrapis (McUkcrtes) stilbouota new species. 1, holotype, dorsal view. 2, paratype, lateral view. body length 3.76 [3.72]. Head wider than long (length 1.10 [1.12], width * [1.32]). In- ner margins of compound eyes straight, nearly parallel; upper interorbital distance 0.84 [0.86]; lower interorbital distance * [0.82]; eye length 0.74 [0.76], width 0.32 [0.30]. Gena width 0.28 [0.24]. Mandible with two blunt denticles on upper half of apical margin, length 0.50 [0.44]; malar space length 0.04 [0.06]. Labrum length (median) 0.20 [0.24], width (basal) 0.44 [0.44]; clypeus length * [0.18], width * [0.64]; clypeoantennal distance 0.08 [0.08]. Scape length 0.36 [0.36]; pedicel length 0.10 [0.06]; flagellum length 0.88 [0.86]; Fl length 0.10 [0.10]; F2 length 0.04 [0.04]; F3 length 0.04 [0.04]; FIO length 0.18 [0.18]; F6-10 with dense sensillar plates on inner surfaces; interantennal distance * [0.20]; antennal-ocellar distance 0.58 [0.56]. Me- dian ocellus diameter 0.12 [0.12]; distance between lateral ocelli 0.26 [0.26]; distance between median ocellus and lateral ocel- lus 0.08 [0.08]; distance from lateral ocel- lus to eye 0.28 [0.28]; distance from lateral ocellus to occiput 0.16 [0.16]. Mesosoma length 1.34 [1.32]; intertegular distance 0.82 [0.80]; mesoscutum length 0.64 [0.60]; scutellum length 0.26 [0.28], width 0.48 [0.44]; metanotum length 0.04 [0.04]; pro- podeal triangle over four times longer than metanotum, length 0.18 [0.18]. Me- socoxae separated by more than mesocox- al width; hind tibia length (median) 1.02 [1.02], width (basal) 0.12 [0.12], width (api- cal) 0.28 [0.30]; hind basitarsus length 0.36 [0.32], width 0.28 [0.28]. Wings hyaline, all veins brown and strong; basal vein basad cu-a by 2 times vein width; pterostigma small; Im-cu bisecting second submargin- al cell; 2r-m distad 2m-cu by 1.5 times vein width; marginal cell length 0.84 [0.86], width 0.24 [0.22]; first submarginal cell shorter than second and third combined; length of anterior border of second sub- marginal cell one-tenth that of posterior Volume 7, Number 1, 1998 97 border; length of anterior border of third submarginal cell half of that of posterior border, just over 3 times length of anterior border of second submarginal cell; fore- wing length 3.00 [3.16]; venation of fore- wing depicted in figure 4; six distal ham- uli on outer margin of hind wing; cu-a of hind wing orthogonal to M + Cu; hind wing length 2.08 [2.16]; venation of hind wing depicted in figure 5. Metasoma length 1.32 [1.28]. Integument over entire bee smooth and glabrous, except on metanotum where the integument is apparently rugulose. S3-6 apparently weakly nodulate, nodules scat- tered over the surface, integument be- tween nodules smooth and shining as on previous sterna. Color not well preserved, apparently dark brown to black, metallic and shining, without any maculations. Pubescence generally pale. Hairs of face widely scattered, simple, and short. Such hairs becoming longer by vertex. Gena with simple, short, suberect hairs. Postgena with long, simple hairs sparsely scattered over integument. Pronotal collar without pubes- cence; pronotal border with mesoscutum with short, simple hairs; lateral surface with similar minute hairs, such hairs appressed to surface, not obscuring integument. Me- soscutum with scattered simple hairs, more sparsely scattered and shorter over central disc, those hairs on anterolateral borders with a few short branches. Scutellum like that of mesoscutum except pubescence lon- ger and restricted to posterior border. Meta- notum with dense, minute, simple hairs, not obscuring the surface. Hypoepimeron with- out pubescence; mesepistemum with scat- tered simple hairs, becoming longer ven- trally, central disc, however, without hairs. Propodeal triangle and posterior surface without pubescence; lateral surface with scattered long simple hairs and shorter, ap- pressed hairs, partially obscuring the sur- face. Pubescence of fore- and midlegs gen- erally simple and scattered, except inner surfaces of midtrochanter and femur with- out pubescence, and outer surface of mid- tibia with dense, branched hairs. Inner sur- face of hind femur and trochanter without pubescence, except apical quarter of femur with dense field of minute hairs similar to those on inner surface of hind tibia (see ge- neric diagnosis). Eight comb rows on inner surface of hind basitarsus; outer surface with scattered, long, simple hairs. Tl with- out hairs over central disc, a few simple hairs on lateral borders. T2 as on Tl, except a few simple hairs on posterolateral bor- ders. T3 with simple hairs, longer than those of Tl-2, sparsely scattered over cen- tral disc, more concentrated on lateral mar- gins. T4-6 similar to T3. Sterna with sparse- ly scattered simple hairs. Material examined. — Holotype: female (Figs. 1 & 3), worker, Samland, Eocene Baltic amber, specimen In. 17778, Depart- ment of Palaeontology, the Natural His- tory Museum (British Museum, London). Paratype: female (Figs. 2, 4-5), worker, same piece of amber and same accession information as holotype. Preservation. — The bees presented herein are exceptionally well preserved. The only hindrance to their examination is the un- even surface of the amber and the small block of storage media they are preserved in. A few small fracture planes arising from the wings do not obscure any im- portant structures, although one small fracture near the face of the holotype spec- imen obscures some features of the clyp- eus and lower face. A bit of mold on the legs along with the remains of what might have been collected pollen in the corbicu- la, while slightly demoting them from per- fect specimens, does not detract in any sig- nificant way from examining their mor- phology. The inner teeth of the claws are minute and difficult to see. The best view of these structures can be achieved by back-lighting the specimens and examin- ing the extended hind legs. Etymology. — The specific epithet is de- rived from stilbo (Gr. shine) and noton (Gr. back), and is a reference to the glabrous 98 Journal of Hymenoptera Research Figs. 3-5. Electrapis (Melikerks) stilbonota new species. 3, close-up of holotype, dorsal view, showing integ- ument of metasoma and propodeal triangle as well as vertex and preoccipital ridge. 4, left forewing of para- type. 5, left hind wing of paratype. integument of the dorsum, in particular that of the propodeal triangle. DISCUSSION Bees of the tribes Apini and Meliponini are all advanced eusocial, except for a few parasitic forms among the stingless bees (Michener 1974), and, based on their sis- ter-group relationship (Chavarria and Carpenter 1994), presumably inherited this aspect of their biology from a com- mon ancestor that was similarly eusocial. The phylogenetic position of the genus Electrapis within the tribe Apini suggests that species of Electrapis were also highly eusocial with a well developed caste sys- tem. Also suggestive of eusociality among Electrapis species is the fact that the spec- imens described herein are morphologi- cally workers. As in many eusocial species the loss of ovarian development in the worker caste results in a greatly reduced metasoma, a feature seen in both speci- mens. Similar lines of evidence were used to make the inference that the oldest known fossil bee, Trigoua prisca, was a worker of a similarly advanced eusocial society (Michener and Grimaldi 1988a, b). While Electrapis runs to the tribe Apini in Michener's (1990) key to the corbiculate bee tribes (treated as subfamilies of Api- dae in that work), there are significant enough differences between Electrapis and its sister Apis which would more than jus- tify placing Electrapis in a tribe of its own. Recognition of a separate tribe for these bees, however, would obscure the rela- tionship of Electrapis with members of the Apini as both possess a jugal lobe, bifid claws, complete distal wing venation, and a marginal cell apex gently pulled away from the anterior wing margin. A more practical approach to the problem is the recognition of subtribes, retaining a broadly defined Apini. The separation of VoLUMK 7, Number 1, 1998 99 Table 2. Brief summary of the subtribal classification of Apini Latreille (based on worker caste). An elab- oration of each character is given in the text. Apin.1 Eye hairs: Labral apex: Mandible: Vertex: Scutellum: Propodeum: Mesocoxae: Marginal cell: Marginal cell: Basal vein: present concave without dentition short bulging short, dcclivious nearly meeting reaching wing apex not tapering distad cu-a HIeclrj Lipinj absent convex with or without dentition long weakly convex or flat long, not declivious well separated not reaching wing apex gently tapering basad to just distad cu-a these subgroups is as follows (a summary of the differences is given in Table 2): Electrapina (new subtribe containing only the typical genus Elect rapis): Com- pound eyes without hairs; labral apex con- vex; mandible with or without dentition; malar space extremely short, much less than basal width of mandible; vertex as long as ocellar diameter, or more; scutel- lum not bulging, surface weakly convex to flat; propodeal triangle with defined sur- face, not declivious; mesocoxae separated by at least their width; marginal cell not reaching to wing apex, gently tapering over its length; basal vein basad to just distad cu-a, never strongly distad (7 times vein width or more). Apina: Compound eyes covered with long hairs; labral apex gently concave; mandible lacking dentition; malar space as long as, or longer than basal width of mandible; vertex extremely short, much less than ocellar diameter; scutellum strongly convex and bulging, obscuring metanotum and propodeal triangle; pro- podeal triangle extremely short and decli- vious; mesocoxae nearly meeting medial- ly; marginal cell long, nearly reaching wing apex, not gently tapering over its length; basal vein confluent (in some fossil Apns) to strongly distad cu-a (over 7 times vein width), never basad cu-a. Arillo et al. (1996) have recently sug- gested that Apis and Electrapis overlapped in geologic time. Specimens of Electrapis are only known from the Baltic amber which is Eocene in age (Kosmowska-Cer- anowicz 1987, Kosmowska-Ceranowicz and Miiller 1985) while Apis species are unknown until the middle Oligocene (Cul- liney 1983, Engel in press, Michener 1990, Ruttner 1988, Zeuner and Manning 1976). Thus, the little available evidence in no way suggests that these taxa were coinci- dent in time. Arillo et al. (1996) are correct, however, in their assertion that there is no reason to believe Electrapis is the direct an- cestor of the true honey bees. Apis, as has been done by some earlier authors (e.g., Statz 1931, Zeuner and Manning 1976). Lastly, these authors have peculiarly used invalid family-group names for bees. For example, they refer to Michener (1986) for the recognition of Rophitidae in place of Halictidae. In fact, Michener (1986) clearly advises the use of Halictidae (even in his fairly short abstract), a proposal which was later supported by Michener (1991) and validated by the International Com- mission on Zoological Nomenclature (1993). Therefore, these authors should not be followed in their use of family-group names for bees. ACKNOWLEDGMENTS I am thankful to A. J. Ross for hosting me during my visit to the Department of Palaeontologv, the Nat- ural History Museum (British Museum, London), and for bringing these specimens to my attention. J. K. Liebherr kindly allowed use of his camera-micro- scope mount for the production of the plates. I am 100 Journal of Hymenoptera Research grateful to G. Chavarria, E. E. Grissell, and J. G. Roz- en, Jr., for kindly reading the manuscript prior to publication. Their constructive criticisms helped im- prove the final draft. Support for this work was pro- vided by an Ernst Mayr Award from Harvard Uni- versity's Museum of Comparative Zoology and by a National Science Foundation Predoctoral Fellowship. LITERATURE CITED Arillo, A., A. Nel, and V. M. Ortuiio. 1996. Two fossil bees from the Oligocene of Izarra (Alava, Spain) (Hymenoptera, Apoidea). Bulletin de la Societe en- tomohgique de France 101:59-64. Buttel-Reepen, H. v. 1906. Apistica. Beitrage zur Sys- tematik. Biologic, sowie zur geschichtlichen und geographischen Verbreitung der Honigbiene (Apis mellifica L.), ihrer Varietaten und der iiber- igen Apis-Arten. Mitteilungen aus dem Zoologisch- en Museum in Berlin 3:117-201. Chavarria, G., and J. M. Carpenter. 1994. "Total evi- dence" and the evolution of highly social bees. Oadistks 10:229-258. Cockerell, T. D. A. 1909a. Descriptions of Hymenop- tera from Baltic amber. Sclmften der PJn/sikalisch- okonomischen Ceiellschaft 50:1-20. Cockerell, T. D. A. 1909b. Some additional bees from Prussian amber. Schriften der Physikalisch-dkon- oinisclien Gesellschaft 50:21-25. Culliney, T. W. 1983. Origin and evolutionary history of the honeybees Apis. Bee World 64:29-38. Engel, M. S. 1995. Neocorynura electra, a new fossil bee species from Dominican amber (Hymenoptera: Halictidae). journal of the Neu< York Entonwlogical Society 103:317-323. Engel, M. S. 1996. New augochlorine bees (Hyme- noptera: Halictidae) in Dominican amber, with a brief review of fossil Halictidae. journal of the Kansas Entomological Society, Supplement 69:334- 345. Engel, M. S. 1997. A new fossil bee from the Oligo- Miocene Dominican amber (Hymenoptera: Hal- ictidae). Apuiologie 28:97-102. Engel, M. S. In press. Fossil honey bees and evolution in the genus Apis (Hymenoptera: Apidae). Api- dologie. Grimaldi, D. A. 1995. The age of Dominican amber, pp. 203-217. In Anderson, K.B., and J.C. Crelling, eds.. Amber, Resinite, and Fossil Resins. American Chemical Society Symposium Volume, Washing- ton, D.C. I.C.Z.N. 1993. Opinion 1713: Some bee family-group names (Insecta, Hymenoptera): names based on Colletes Latreille, 1802, on Paracolletes Smith, 1853, on Halictus Latreille, 1804, on Anthuiium Fa- bricius, 1804 and on Anthoptiora Latreille, 1803 given precedence over some senior names. Bul- letin of Zoological Nomenclature 50:85-89. Kelner-Pillault, S. 1970. L'ambre Balte et sa faune en- tomologique avec description de deux apoides nouveaux. Annates de la Societe entomologique de France 6:3-24. Kelner-Pillault, S. 1974. Etat d'evolution des apides de l'ambre balte. Annates de la Societe entomolo- gique de France 10:623-634. Kosmowska-Ceranowicz, B. 1987. Mineralogical-pet- rographic characteristics of the Eocene amber- bearing sediments in the area of Chlapowo, and the Palaeogene sediments of Northern Poland. Biuletyn Instytut Geologii 356:29-50. Kosmowska-Ceranowicz, B., and C. Miiller. 1985. Li- thology and calcareous nannoplankton in amber bearing Tertiary sediments from boreholes, Chla- powo (Northern Poland). Bulletin of the Polish Academy of Sciences 33:119-129. Manning, F. J. 1960. A new fossil bee from Baltic am- ber. Proceedings of the 11"' International Congress of Entomology, Vienna 1:306-308. Menge, A. 1856. Lehenszeichen vortu'eltUcher im Bern- stein eingeschlossener Thiere. Programm Petrischu- ler Danzig. Michener, C. D. 1974. The social behavior of the bees: a comparative study. Harvard University Press, Cambridge. Michener, C. D. 1986. Family-group names among bees. Journal of the Kansas Entomological Society 59: 219-234. Michener, C. D. 1990. Classification of the Apidae (Hymenoptera). University of Kansas Science Bul- letin 54:75-164. Michener, C. D. 1991. Case 2535. Proposed prece- dence of some bee family-group names (Insecta, Hymenoptera): names based on Colletes Latreille, 1802, on Paracolletes Smith, 1853, on Halictus La- treille, 1804, on Anthidium Fabricius, 1804 and on Anthophora Latreille, 1803 to have precedence over some senior names. Bulletin of Zoological No- menclature 48:227-235. Michener, C. D., and D. A. Grimaldi. 1988a. A new Trigona from Late Cretaceous amber of New Jer- sey (Hymenoptera: Apidae: Meliponinae). Amer- ican Museum Novitates 2917:1-10. Michener, C. D., and D. A. Grimaldi. 1988b. The old- est fossil bee: apoid history, evolutionary stasis, and antiquity of social behavior. Proceedings of the National Academy of Sciences. USA 85:6424-6426. Michener, C. D., and G. O. Poinar, Jr. 1996. The known bee fauna of the Dominican amber, four- nal of the Kansas Entomological Society, Supplement 69:353-361. Roussy, L. 1937. Contribution a I'etude de I'abeille tertiaire de ses parasites et de ses ennemis. Ga- zette apicole 388:49-72. Rozen, J. G., Jr. 1996. A new species of the bee Hct- erosariis from Dominican amber (Hymenoptera: Andrenidae; Panurginae). journal of the Kansas Entomological Society, Supplement 69:346-352. Volume 7, Number 1, 1998 101 Ruttner, F. 1988. Biogeography ami taxonomy of honey- bees. Springer- Verlag, Berlin. Salt, G. 1931. Three bees from Baltic amber. Bernstein- Forscluingen 2:136-147. Statz, G. 1931. Eine neue Bienenart aus Rott am Sie- bengebirge: ein Beitrag zur Kermtnis der fossilen Honigbienen. Wissenschaftlichen Milleihingen ties Vereinsfiir Natiir- und Heimatkunde, Koln 1:39-60. Zeuner, F. E., and F. ]. Manning. 1976. A monograph on fossil bees (Hymenoptera: Apoidea). Bulletin of the British Museum of Natural History (Geology) 27:149-268. NOTE ADDED IN PROOF A paper has recently reached me concerning a Middle-Eocene bee from Germany which is attributable to Electrapis [H. Lutz. 1993. Eekfeldapis eleetrnpotdes nov. gen. n. sp., eine "Honigbiene" aus dem Mittle-Eozan des "Eckfelder Maares" bei Manderscheid/Eifel, Deutschland (Hymenop- tera: Apidae, Apinae). Maimer naturwissensclniftliches Arehiv 31:177-199]. This bee is clearly a species of Electrapis s. str. and, based on the wing venation, appears most similar to £. apoides. I, therefore, here synonymize Eekfeldapis (new synonymy) with Electrapis, and place its only included species as a species of the latter: Electrapis (Electrapis) electrapoides (Lutz), new combi- nation. It must also be noted that Lutz's figure 3h, labeled as the wing venation of Apis inellifcra, should be disregarded as it resembles very little the venation of this species (particularly in the position of the basal vein and cu-a). J. HYM. RES. Vol. 7(1), 1998, pp. 102-115 Neotropical Eucoilidae (Cynipoidea) Associated with Fruit-infesting Tephritidae, with New Records from Argentina, Bolivia and Costa Rica R. A. Wharton, S. M. Ovruski, and F. E. Gilstrap (RAW, FEG) Texas A&M University, Department of Entomology, College Station, TX 77843- 2475, USA; (SMO) CIRPON, C. C. C. 90 4000, San Miguel de Tucuman, Argentina Abstract. — Host and distribution records are presented for five species of Neotropical Eucoilidae (Hymenoptera: Cynipoidea) reared in association with fruit-infesting Tephritidae. All previously recorded tephritid host associations for New World species are critically reviewed with the con- clusion that several of these records are doubtful. Members of the genera Aganaspis and Odonto- sema are confirmed as parasitoids of Anastrepha and Ceratitis. Based on isolated puparia, Dicera- taspis and Lopheucoila are recorded as parasitoids of Drosophilidae and Lonchaeidae, respectively. It is suggested that Dicerataspis is unlikely to attack Tephritidae, and records of Lopheucoila from Tephritidae require confirmation. One new species reared from Tephritidae, Aganaspis nordlan- deri Wharton, is described. The Eucoilidae are solitary endoparasi- toids that oviposit in the larval stage of cyclorrhaphous Diptera and emerge as adults from the host puparium. Several eucoilid species have been implicated as important natural enemies of different phytophagous dipteran species (Wishart and Monteith 1954, Harding 1965, Valla- dares et al. 1982, Johnson 1987), and two species have been used for the biological control of fruit fly pests (Clausen 1978). The Asian species Aganaspis daci (Weld) has been introduced to the New World, and the Neotropical endemic Aganaspis pelleranoi (Brethes) was reared and re- leased from 1941 to 1945 in several areas of Tucuman, Argentina (Nasca 1973). At present, A. pelleranoi is being mass-pro- duced in Metapa de Dominguez, Chiapas, Mexico (Ruiz et al. 1996). Members of the genus Aganaspis are the only eucoilids thus far utilized in biologi- cal control efforts against Tephritidae. The two species involved, A. daci and A. pel- leranoi, are also the only eucoilids attack- ing tephritids for which biological infor- mation other than host records has been published. Different aspects of the basic biology of A. daci were studied in the lab- oratory associated with programs directed against Bactrocera dorsalis (Hendel) in Ha- waii (Clausen et al. 1965) and Anastrepha suspensa (Loew) in Florida (Nunez-Bueno 1982). This species has also been intro- duced to Mexico (Jimenez-Jimenez 1956) and Costa Rica (Wharton et al. 1981, Jiron and Mexzon 1989). Establishment in Mex- ico and Costa Rica is doubtful, but in Flor- ida it is established on Anastrepha suspensa (Loew), though in low numbers (Bara- nowski et al. 1993). A detailed biology of A. pelleranoi was given by Ovruski (1994a, 1994b). The aim of this note is to provide pre- liminary information on the diversity of eucoilid species associated with tephritid fruit flies in the Neotropics, and clarify the status of species previously recorded as te- phritid parasitoids. While several species have been associated with tephritids, few of these have been reared from puparia that were sufficiently isolated to enable Volume 7, Number 1, 1998 103 verification of the host. Data presented here are based on surveys of tephritid par- asitoids and on literature records. MATERIALS AND METHODS Fruits damaged by tephritid larvae were collected from 1991 to 1994 in Tu- cuman, Catamarca and La Rioja provinces in northvv'estern Argentina, and from Au- gust, 1979 through November, 1982 in the provinces of Alajuela, Cartago, Guana- caste, Heredia, Limon, Puntarenas, and San Jose in Costa Rica. Additional eucoilid specimens were also received from Boliv- ia, and all reared material housed in the U. S. National Museum of Natural Histo- ry, Washington, D. C. (USNM) was ex- amined. Samples collected in Argentina and Costa Rica consisted of fallen fruit and fruit still on the tree. In Argentina, fruit samples were placed in styrofoam boxes with damp sand in the bottom as a pupation substrate. Fruit fly puparia were recovered weekly and transferred to a closed wooden box for holding until emer- gence of flies or parasitoids. The proce- dure differed slightly for the samples from Costa Rica (Wharton et al. 1981), where 81,279 puparia were isolated (most of them in individual vials) for verification of host records. Tephritids of the genera Anastrepha Schiner and Ceratitis MacLeay (or their parasitoids) accounted for 69,012 of these puparia, with C. capitata repre- senting 64.8% of the total tephritids. Lon- chaeidae (also discussed below) were rep- resented by 4583 puparia. Specimens reported on here are housed at Museo de La Plata, Argentina (MLP), Museo de Ciencias Naturales Bernardino Rivadavia, Buenos Aires, Argentina (MBR), Instituto Fundacion Miguel Lillo, Tucuman, Argentina, Texas A&M Univer- sity, College Station (TAMU), and USNM. Measurements for the description of the new species are as described by Nordlan- der (1978, 1982). RESULTS AND DISCUSSION Several species of Eucoilidae, represent- ing at least five genera, have been reared in association with fruit-infesting tephri- tids. Most of the species are almost cer- tainly attacking other Diptera associated with ripe and decomposing fruit (e.g. Dro- sophilidae, Lonchaeidae, Phoridae, Neri- idae). We have verifiable host records from Tephritidae for Agauaspis and Odon- tosetna. The published records for Dicera- taspis, Lopheucoila, Rhoptoiiicris, and Tri/- bliographa attacking tephritids in the New World need confirmation. Species of the genus Leptopilina, well-known parasitoids of Drosophilidae, may also be reared com- monly from rotting fruit. Though Droso- philidae tend to colonize fruit after te- phritids, fallen, broken fruit may have more rapidly decaying portions inhabited by Drosophilidae at the same time as more sound portions still inhabited by tephritid larvae. Since fruit is often collected and reared in bulk, it is easy to obtain parasit- oids of both Drosophilidae and Tephriti- dae, for example, from the same sample. Unless puparia are isolated individually, correct host associations cannot be made. These seven genera may be separated by the characters in Table 1. Additionally, Lopheucoila is unique within this group of genera in having a small spine on the dor- sal plate of the scutellum and longitudinal ridges on the mesoscutum. See also papers by Weld (1952), the updated classification by Nordlander (1978, 1980, 1981) and the description of Aganaspis by Lin (1987). Agartaspis Lin The genus Aganaspis was relatively re- cently described (Lin 1987) to accomodate four species from southeast Asia (Taiwan and Malaysia). One of these species, A. daci (Weld), was originally described in Tn/hliographa (Weld 1951b) but its generic placement had always been problematic (Kerrich and Quinlan 1960, Nordlander 1981). Aganaspis daci is the only one of the 104 Journal of Hymenoptera Research Table 1. Genera of Eucoilidae reported from fruit-infesting Tephritidae in the Neotropics compared with Lqjtopili}ia, parasitoids of Drosophilidae commonh' reared from the same fruits. Scuteilar disc posteriorly in dorsal view Fore wing Posterior-dorsal margin of pronota! plate 4th antennal segment {2nd flagellomere) of male Anterior and posterior parts ot pronotal plate fused or separate laterall\' Agciimspis Dicerataspis Lopheucoila Odontoscma Rhoptrcineris TrybUographa LL'ptopilinn rounded or truncate or weakly bi- lobed bifurcate, with 2 tooth-like lobes directed posteriorly weakly dentate, with 4 lobes; somewhat truncate distinctly setose distinctly setose bare or nearly so bifurcate, with 2 bare or nearly tooth-like so lobes directed posteriorly rounded or truncate distinctly setose rounded or truncate distinctly setose rounded or truncate distinctly setose protruding above anteri- or margin of mesoscutum; deeply con- cave medially weakly protrud- ing above an- terior margin of mesoscu- tum; undu- lant, with 4 rounded lobes protruding above anteri- or margin of mesoscutum; deeply con- cave medially protruding above anteri- or margin of mesoscutum; deeply con- cave medially not protruding above anteri- or margin of mesoscutum; evenly round- ed or very weakly con- cave medially not protruding above anteri- or margin of mesoscutum; evenly round- ed or very weakly con- cave medially not protruding above anteri- or margin of mesoscutum; evenly round- ed or very weakly con- cave medially 3 not bent 4 < 3 not bent 4 < 3 not bent 4 < 3 not bent 4 > 3 bent out- wardly 4 > 3 usually bent widely to nar- rowly sepa- rated contiguous or narrowly sep- arated fused or contig- uous fused or contig- uous fused 4 £ 3 not bent fused widely separat- ed ' Volume 7, Number 1, 1998 105 four originally included species for which hosts have been recorded. Nordlander (in lift.) suggested that the New World spe- cies pellemiwi should also be placed in Aganaspis. This transfer was made by Ovruski (1994a), bringing the total num- ber of species in Aganaspis to five. All four of the Old World species have distinctly setose eyes (more noticeable in the fe- male), while those from the Neotropics do not. This is one of the easiest ways to rec- ognize A. daci in those areas of the New World where it has been introduced. The placement of pelleranoi and nonilan- deri, n. sp. (described below) in Aganaspis broadens the limits of this genus, and opens up the possibility that several of the Neotropical species formerly placed in ei- ther Ttybliographa or Pseudeiicoila may ac- tually belong here. As indicated by Nor- dlander (1981, Table 2), several generic names are available for these species, and placement of most of the previously de- scribed species will not be possible with- out a revision of the entire group. Collec- tion records (e.g.. Weld 1932) and label data on specimens in the USNM suggest that several of these species have been reared from tephritids, but most records are not sufficiently precise to preclude the possibility that the actual hosts may be other fruit-inhabiting flies. See further dis- cussion below under Tryhliographa. Aganaspis pelleranoi (Brethes) De Santis (1965) placed Ganaspis carvalhoi Dettmer, 1929 as a junior subjective syn- onym of Eucoila pelleranoi Brethes, 1924. The senior author has confirmed this synonymy through comparison of one of Dettmer's syntypes in USNM with one of the syntypes of pelleranoi from the Brethes collection (MBR). Both types match the specimens we reared from Argentina and Costa Rica, con- firming their identity as pelleranoi. Known hosts and distribution records of A. pelleranoi are as follows: Hosts. — A. ludens (Loew), A. obliqua (Macquart) (Aluja et al. 1990), A. serpentina (Wiedemann) (Costa Lima 1940), A. striata Schiner (Clausen 1978), A. distincta Greene (Katiyar et al. 1995), A. fraterculus (Wie- demann) (Brethes 1924), Ceratitis capitata (Wiedemann) (De Santis 1965), and Rhag- oletis turpiniae Hernandez-Ortiz (Hernan- dez-Ortiz 1993). The records from "Lon- chaea sp." and "L. pendula Bezzi" (Borg- meier 1935 and Costa Lima 1948, respec- tively) are suspect, and require verification. As noted by McAlpine and Steyskal (1982), the name pendula has been misapplied on numerous occasions to Neotropical lonchaeids of the genus Neo- silba McAlpine that have been reared from fruit. Our own records suggest that A. pel- leranoi may only rarely attack lonchaeids. Of 295 specimens of A. pelleranoi that we reared from isolated puparia in Costa Rica, 191 (64.7%) were from C. capitata (mostly in coffee), 81 from Anastrepha (nearly all A. striata), and only two were from a lonchaeid (Neosilba hatesi (Curran), new record). Twenty-one other specimens of Aganaspis were also reared from Neosil- ba in Costa Rica, but these are only tenta- tively assigned to pelleranoi because of slight but consistent color differences rel- ative to the other specimens of A. pelleranoi reared during these studies. We reared A. pelleranoi from infested fruits of the following species: Casimiroa ediilis Llave and Lex., Citrus aurantiifolia (Christm.) Swingle, Coffea arabica L., Ficus carica L., Juglans australis Grisebach, Pru- nus domestica L., Prunus persica (L.) Batsch, Psidium guajaim L., P. friedrichsthalianum (O. Berg) Niedenzu, P. littorale Raddi ( = cattleianum), Syzygium jambos (L.) Al- ston, and Terminalia catappa L. These data, together with previously published re- cords, suggest that A. pelleranoi, like sev- eral of the other commonly encountered tephritid parasitoids, has little or no host plant preferences. See additional com- ments below under discussion of Odonto- sema. Distribution. — Argentina: Buenos Aires, Misiones, Salta, Tucuman, Jujuy, Corrien- 106 Journal of Hymenoptera Research tes (De Santis 1967, Diaz 1986), La Rioja and Catamarca (new records); Bolivia: Santa Cruz de La Sierra (new record); Peru (Clausen 1978); Brazil (Dettmer 1929); Venezuela (Katiyar et al. 1995); Co- lombia (Yepes and Velez 1989); Costa Rica (Wharton et al. 1981); El Salvador (Ovru- ski et al. 1996); and Mexico (Aluja et al. 1990). There are also specimens in the USNM from Panama, Guatemala, and Be- lize (all new records). During the survey for tephritid parasit- oids in Costa Rica (Wharton et al. 1981), several species resembling A. pelleraiwi were reared. Only one of these was re- peatedly reared from tephritids, and it is described next. Aganaspis nordlanderi Wharton, new species (Figs. 1, 3, 4, 6-8, 10, 12) Quantitative measurements, based on 5 females and 2 males, are presented either as ranges or means to the nearest 0.05. Female (Fig. 1). — Body length 2.3-3.5 mm; fore wing length 2.3-3.1 mm. Head in dorsal view 1.05-1.15 times wider than mesoscutum; 1.75-1.90 times wider than long, when length measured in profile; temples strongly receding in dor- sal view. Frons, vertex, and occiput bare; eyes without visible setae at 50 x. Face shining, unsculptured; malar sulcus a weak, narrow groove, without additional striae; antennal base elevated, especially laterally, forming shallow depression be- tween antemia and eye. Posterior ocelli widely separated: distance between them 1.2-1.3 times distance between posterior ocellus and eye. Antenna (Fig. 4) short, about 2.5 times height of head; without distinct club, the segments gradually broadening distally, first 4-5 flagellomeres weakly clavate: each slightly broader sub- apically than medially; first flagellomere slightly shorter than second, relative lengths, first 8 flagellomeres: 1.0:1.1:1.1: 1.05:1.0:1.0:0.95:0.95; ratio of length to maximum width, first 8 flagellomeres: 2.4: 2.3:2.2:2.0:1.9:1.8:1.65:1.55; flagellomeres each with 3-5 whorls of setae. Mesosoma 1.25 times longer than high; 1.6 times longer than wide; 1.3 times high- er than wide. Pronotal collar (Figs. 1, 6) distinctly protruding above anterior mar- gin of mesoscutum in lateral view, deeply bilobed; median bridge wider than ante- rior ocellus; lateral arms of anterior and posterior parts narrowly but distinctly separated; posterior part of pronotal plate with a few, completely decumbent setae, otherwise bare and polished, 2.3-2.8 times wider than median bridge. Mesoscutum divided into three parts of approximately equal width by two longitudinal rows of 5-6 decumbent setae per row; margin with scattered setae separated from each other by their own length; short, shallow, crescentic grooves present on posterior half directly anteriad lateral bars of scu- tellum. Scutellar disc (Figs. 7, 8, 10) dis- tinctly reticulate, with dorsoposteriorly- directed setae somewhat longer than in pelleranoi; in dorsal view (Fig. 8) disc slightly excavated medially, and thus weakly bilobed, the lobes protruding slightly beyond cup; posterior margin of disc sinuate in profile; height of posterior margin of scutellum about 1.5 times length (in lateral view) of flat portion of cup; scutellar cup (Fig. 8) large, broadly tear-drop shaped, with posterior margin weakly rounded, nearly truncate, anterior margin short, not extending through scu- tellar fovea, surface nearly flat, not droop- ing posteriorly, shallowly excavated me- dially, with only one or two punctures and 2^ short, erect setae laterally; width of cup 0.5-0.6 times width of disc. Fore wing (Fig. 12) 2.6-2.9 times longer than wide; marginal setae short, longest sub- apical seta 0.08-0.1 times maximum width of wing; radial cell deep, completely open along wing margin, second radial abscissa 2.7-3.1 times longer than first; costal cell densely setose, ventral surface with 3-4 longitudinal rows of setae. Mid and hind coxae (Fig. 1) with extensive patches of Volume 7, Number 1, 1998 107 Figs. 1-5. Agimaspna spp. 1, A. nonilaiideri, habitus. 2, A. pcUcranoi, mid and hind coxae, lateral view, showing small patches of dense setae dorsoposteriorly. 3, A. nordlanderi, male antenna, setae not shown except for 8th flagellomere. 4, A. nordlanderi, female antenna, setae not shown except along margins of 7th and 8th flagel- lomeres. 5, A. pelleranoi, basal 5 antennal segments of female. dense setae, patch on mid coxa extending about half length of coxa from base, patch on hind coxa extending more than half length of coxa. Second metasomal tergum 1.0-1.1 times length of mesosoma. Male. — As in female except as follows: head in dorsal view 1.15-1.2 times wider than mesosoma; antenna (Fig. 3) long, 3.2- 3.3 times height of head; relative lengths, first 5 flagellomeres: 1.0:0.8:0.9:0.95:0.95, ratio of length to maximum width, flagel- lomeres 1-5 and 10: 2.85:2.05:2.25:2.25:2.3: 2.6; mesosoma 1.2 times longer than high; fore wing 3.6-3.75 times longer than wide; second tergum 0.85-0.9 times length of mesosoma. Color. Head and mesosoma black except pedicel and basal flagellom- eres usually dark reddish brown, flagel- lomeres gradually darkening apically, more rarely with antenna entirely dark brown to black; gaster reddish-orange, ex- cept apical terga black posteriorly; legs red-brown (slightly browner than gaster). Hosts. — The specimens forming the type series were reared from individually iso- lated puparia of Ceratitis cnpitntn and Anas- trepha striata, and the puparia from which Figs. 6-11. Aganaspis spp, mesosoma. 6, A. nordlandcn, pronotum in dorsal view. 7, A. nordlanderi, propodeum in posterior view, showing posterior portion of scutellum, setae not shown except portion of basal ring on metasoma. 8, A. nordlanderi, scutellum in dorsal view. 9, A. pelleranoi. scutellum in dorsal view. 10, A. nor- dlanderi, .scutellum in lateral view. 11, A. pelleranoi. scutellum in lateral view. 108 Journal of Hymenoptera Research Volume 7, Number 1, 1998 109 Figs. 12, 13. Agaiuispis, fore wings. 12, A. nordlandcn, showing setal pattern in costal cell, marginal setae, and shape of radial cell, other setae not shown. 13, A. fieUeranoi, showing relatively reduced pattern of setae in costal cell and closed radial cell. they emerged are pinned with the parasit- oids. Holoiype female. — "Costa Rica: Cartago Turrialba, Catie 10.vii.l980 ex: guava R. Chavez." Reared from Anastrepha striata in guava. Deposited in USNM. Paratypes (TAMU), all reared from Ceratitis capitata in Costa Rica: 3 females, 2 males, same lo- cality, 8-vi-1980, 26-vi-1980, 30-ix-1981, and 8-vi-1982, J. Duran and R. Chavez, collectors, from coffee and naranja agria; 1 female, Puntarenas, San Vito, 23-ix-1980, R. Chavez, from guava. Diagnosis. — This species differs from pel- leraiwi in the more densely setose costal cell of the fore wing (with only a single longitudinal row of setae on ventral sur- face in pelleraiwi: compare Figs. 12 and 13), the completely open radial cell, the more extensively furry hind coxa (Fig. 1 vs. Fig. 2), the smaller scutellar cup (width of cup/ width of disc = 0.55 in nordlamieri vs. 0.75 in pelleraiwi; scutellar cup extending nearly to anterior margin of scutellar fo- vea in pelleranoi: compare Figs. 8 and 10 with Figs. 9 and 11), and the relative pro- portions of the flagellomeres (Fig. 4 vs. Fig. 5) and first and second radial abscis- sae. Discussion. — Aganaspis nordlanderi has a distinctly smaller scutellar cup than either pelleranoi or daci, and the surface is not quite as flat as in these two other species. The first flagellomere of the female is also slightly shorter than the second in nordlan- deri but slightly longer than the second in pelleranoi and daci. Otherwise, nordlanderi and pelleranoi share several features which suggest that they are more closely related to each other than either is to daci and the other Old World species of Aganaspis de- scribed by Lin (1987). This relationship be- tween the New World species is based on the absence of distinct setae on the eyes, the shape of the scutellar disc which pro- trudes posteriorly beyond the cup, the lack of a deep, median depression in the posterior margin of the metapleuron, and the more elongate basal flagellomeres and less distinctive club of the female antenna. Until the genera occurring in the Neotrop- ical Region become better known, we pre- fer to retain pelleranoi and nordlanderi in Aganaspis, with the full realization that this placement may need to be reconsid- ered at some later date. Dicerataspis Ashmead There are two described species, and Weld (1952) recorded an additional, ap- parently undescribed species from Flori- da. The genus is known from Mexico and Florida south through the Caribbean to Brazil and Argenhna (Weld 1921, 1952, Diaz 1974). The species have not been re- vised, and most of the reared material in the USNM has been tentatively associated with the name D. grenadensis Ashmead, 1896. At least two species are represented in our material from Costa Rica. The spe- cies differ in the sculpture of the scutel- lum, infumation of the wing, and shape of the radial cell. A revision of the genus is needed before species names can be as- signed with any degree of confidence. Based on records in the USNM, largely from the rearings by Zetek in the Panama Canal Zone, members of this genus have been reared from a variety of fruits (Ficiis, Labatia, Psidiiitn, Carica, and Anacardium) in association with several different spe- 110 Journal of Hymenoptera Research cies of the tephritid genus Anastrepha. Di- cerataspis has also been associated with one species of Rhagoletis from Mexico (Hernandez-Ortiz 1993). However, precise associations have rarely if ever been made. Our own material was reared in associa- tion with Anastrepha fraterculus and Anas- trq}ha sp. from Psidium guajaim (guava) in Tucuman, Argentina, and from Droso- philidae in peaches and guava in Costa Rica. The single specimen from guava in Costa Rica was reared from a carefully isolated drosophilid puparium and repre- sents the first specific, verified host asso- ciation to our knowledge. All of our other material was batch-reared, and could not be directly associated with the individual puparia from which they emerged. The specimens from guava in both Costa Rica and Argentina agree most closely with the description of D. grenadensis. Given the small size of the species of Dicerataspis and their development as sol- itary parasitoids, normal hosts are likely to be drosophilids or other small Diptera in fruit, rather than tephritids. Parasitism of Anastrepha or other tephritids needs verification. Lopheucoila Weld Weld (1951a) described the genus Lo- pheucoila for three New World species, only one of which, L. anastrephae (Roh- wer), has ever been associated with spe- cific hosts. Lopheucoila anastrephae was originally described from specimens sup- posedly reared from Anastrepha sp. in Trinidad (Rohwer 1919). Weld (1951a) subsequently reported this species from A. fraterculus in Panama and Lonchaea sp. in Brazil, as well as from Mexico and Peru (unassociated with hosts). As noted above under the discussion of hosts of A. peller- anoi, the Brazilian record from Lonchaea undoubtedly refers to a species of Neosilba. All of the specimens of Lopheucoila reared from Argentina and Costa Rica during the present study were L. anastrephae, based on comparison with type material of the three described species housed in the USNM. In Argentina, L. anastrephae was ob- tained from guava fruits in association with A. fraterculus and Anastrepha spp. In Costa Rica, it was reared from coffee, gua- va, papaya, sour orange (Citrus aurantium L.), and tangerine (Citrus reticulata Blan- co), and only from Lonchaeidae. All 74 specimens collected in Costa Rica were reared from isolated puparia of Neosilba batesi (Curran), the most abundant of the two species of Lonchaeidae in the Costa Rican samples. Although 69,000 puparia of Anastrepha and Ceratitis were collected from 1979-1982 in Costa Rica, L. anastre- phae was never reared from a tephritid during this period. Odontosema Kieffer Kieffer (1909) based the genus Odonto- sema on a single species from Brazil. Borg- meier (1935) later described a second bra- zilian species, O. anastrephae Borgmeier, collected in association with A. fraterculus from guava. One undescribed species has been recorded from A. striata and A. fra- terculus in guava in Veracruz, Mexico (Hernandez-Ortiz et al. 1994), and an un- identified Odontosema species was reared from A. fraterculus in fruit of four species of Myrtaceae in southern Brazil (Salles 1996). In Costa Rica, O. anastrephae has been reported from C. capitata and Anas- trepha spp. inhabiting several different fruits (coffee, orange and guava) (Wharton et al. 1981). Odontosema is probably widely distributed throughout the Neotropical Region, but at present it has only been re- corded from Brazil (Borgmeier 1935, Costa Lima 1948, Salles 1996), Costa Rica (Whar- ton et al. 1981) and Mexico (Hernandez- Ortiz et al. 1994, Lopez et al. 1996). Odontosema anastrephae was collected only in Costa Rica, and was not found in Argentina during the present study. This species showed distinct host preferences, with 74% of the 193 reared individuals coming from Anastrepha in guavas. An ad- Volume 7, Number 1, 1998 111 ditional 15.5% of the reared individuals were from C. capitata in citrus. Coffee, the most heavily sampled fruit in Costa Rica, yielded only four individuals. Only 2% of the individuals were reared from Lon- chaeidae (all on citrus), with the remain- ing 98% on either Anastrepha or Ceratitis. These results are in direct contrast with those for A. pelleraiioi, which was collected predominantly from C. capitata in coffee in Costa Rica. Slightly less than half as many A. pelleranoi were reared from Anastrqjha in guava as from Ceratitis in coffee. The pattern of host utilization by A. pelleranoi can thus be directly correlated with sam- pling frequency, with the mostly com- monly sampled fruits and tephritids yield- ing the greatest numbers of pelleranoi. Rhoptromeris Forster Only one species of this genus has been associated with tephritid fruit flies. This species, R. hai/zoardi (Blanchard 1947), was originally included in Eucoila and later transferred to Rhoptromeris (De Santis 1967). R. hayzvardi was described from Ar- gentina and Uruguay (Blanchard 1947) and has been reared in association with A. fraterculus and C. capitata (De Santis 1967), as well as Anastrepha spp., from Carica quercifolia Hill, Ficus carica, Phoebe porphyr- ia Gris, Psidium j^uajava, and Primus persica in Tucuman (Turica and Mallo 1961). De- spite intensive surveys of fruit fly parasit- oids that were conducted in several areas of the province of Tucuman for this and related studies (Ovruski 1995) the host fly and host plant associations recorded by Turica and Mallo have not been verified. Nasca et al. (1980) obtained exception- ally large numbers of R. hai/wardi and the diapriid Trichopria anastrephae Costa Lima (nearly 3,500 and 1,000 individuals respec- tively) by using a modified model of Hay- ward's (1940) parasitoid fly-trap. This col- lecting method consisted of a pit in the soil beneath the host plant into which fall- en host-fruit were placed, the pit was then covered with a thin sieve permitting only the capture of parasitoids. It is very likely this method facilitated the production of drosophilids and their parasitoids. For ex- ample T. anastrephae is also known to at- tack drosophilids (Turica and Mallo 1961). This, together with our failure to rear hai/- wardi from tephritid puparia in the prov- ince of Tucuman, lead us to question the recorded host associations of this species. Parasitism of Tephritidae needs to be ver- ified. The generic placement of this species also needs verification following Nordlan- der's (1978) revision of Rhoptromeris. As restricted by Nordlander (1978), Rlioptrom- eris consists of parasitoids of Chloropidae and other small dipterous larvae such as those that commonly breed in the base of grass stems (Poaceae), and it is unlikely that haifwardi belongs to this largely Hol- arctic genus. Trybliographa Forster Several eucoilids associated with fruit- infesting Diptera in the Neotropical Re- gion have been placed in the genus Try- bliographa at one time or another. These in- clude species placed in Pseudeucoila, a ju- nior synonym of Trybliographa (Hellen 1960, Nordlander 1980), as well as various subgenera of either Trybliographa or Pseu- deucoila (Weld 1952). Three of these, in- cluding the species originally described as Trybliographa daci by Weld (1951b), are treated above under the genus Aganaspis. Most of the others, however, cannot be readily assigned to genus at the present time because the Neotropical genera are badly in need of revision. These species appear to belong to what Nordlander (1982) referred to as the Ganaspis group of genera. Nordlander (1981, 1982) discussed some of the problems associated with the Trybliographa and Ganaspis groups of gen- era, presented a list of generic names that he considered valid, and provided useful information on the location and status of the type material. According to Nordlan- der (1981), Trybliographa is predominantly 112 Journal of Hymenoptera Research Holarctic. It is therefore unlikely that any of the species previously associated with fruit-infesting Diptera in the neotropics belong to this genus. We have examined several such specimens in the USNM col- lection of Tri/bliograplia and Pseudeucoila labelled as reared from tephritids or other fruit-infesting flies, and were unable to discover any that belong to Trybliographa s. s. as defined by Nordlander (1981). Aside from the species now transferred to Aganaspis, we are aware of only two other species that have been described, and for which tephritid hosts have been reported (additional records available to us are largely in the form of unpublished label data). These are bmsiliensis von Ihering, 1905 and hookeri Crawford, 1913. Von Ihering's species was first de- scribed as Hexamerocera brasiliensis and lat- er as Eiicoela (Hexamerocera) eobrasiliensis (von Ihering 1914). It was transferred, along with the subgenus Hexamerocera, to Pseudeucoila by Weld (1932). Nordlander (1978), however, treated Hexamerocera as a synonym of Rhoptromeris. Pseudeucoila, as noted above, is now a synonym of Try- bliographa. Von Ihering's brasiliensis does not fit the current definition of either Rhoptromeris or Trybliographa, and thus, like haywardi, remains unplaced in the Eu- coilidae. Von Ihering's brasiliensis is known from Brazil and Panama (Borgmeier 1935) and it was introduced to Puerto Rico during 1935-37 along with other unidentified eu- coilid species to aid in the control of A. obliqua and A. suspensa (Bartlett 1941). The recorded hosts were A. fraterculus, Anas- trepha sp., C. capitafa and Drosophilidae (Borgmeier 1935, Costa Lima 1948). The type material was collected from peaches, where it was thought to be a parasitoid of A. fraterculus (von Ihering 1905). Though von Ihering (1912) provides evidence to support his view of brasiliensis as a para- sitoid of Anastrepha, the actual host was not identified and the possibility that this is a drosophilid parasitoid (because of its small size) cannot be discounted. Hosts for this species thus need verification. Crawford's hookeri was originally de- scribed in Ganaspis, but was placed under Hexamerocera by Weld in his arrangement of the USNM collection (though formal transfer to Pseudeucoila (Hexamerocera) was apparently never published). This is a similarly diminutive species, and there- fore unlikely to be a parasitoid of Anastre- pha, as originally recorded. Both hookeri (from Puerto Rico) and brasiliensis have a complete hairy ring at the base of the sec- ond tergum, unlike similarly-sized species of Leptopilina, in which the second tergum is bare dorso-medially. The scutellar cup is much larger in brasiliensis than it is in hookeri. CONCLUSIONS Information on host specificity in eu- coilids is largely lacking other than for the work by van Alphen, Vet and colleagues on species of Leptopilina attacking Droso- philidae (e.g. van Alphen et al. 1991, Pool- man Simons et al. 1992). There are often many species of Diptera, representing sev- eral families, present in fruit attractive to eucoilids. Thus, in order to assess host specificity accurately, it is essential to ver- ify all records by isolating individual pu- paria or exposing known hosts to ovipos- iting females. For many of the older re- cords (including label data from unpub- lished studies), host associations were based on eucoilids reared from bulk sam- ples of fruits containing pest tephritids. These must be viewed with caution be- cause of the inevitable inclusion of other flies, such as drosophilids and lonchaeids, in these samples. ACKNOWLEDGMENTS We are grateful to the following people for the loan of types and other specimens: Arnold Menke and Da- vid Smith (U.S.D.A., Systematic Entomology Labo- ratory, Washington, D.C.) and A. Roig-Alsina (MBR). We are also indebted to Goran Nordlander for con- siderable assistance in helping us to understand the classification of Eucoilidae. S. Ovruski is deeply Volume 7, Number 1, 1998 113 grateful to N. B. Diaz (Facultad de Ciencias Naturales y Museo de La Plata, Argentina: MLP) for identifying the Dicercitnfpns species from Argentina, C. Pruett (Universidad Autonoma Gabriel Rene Moreno, Santa Cruz, Bolivia) for sending the specimens from Boliv- ia, P. Fidalgo (Instituto Fundacion Miguel Lillo), N.B. Diaz and L. De Santis (MLP) for supplying important bibliographic material. Finally, S. Ovruski wishes to thank Consejo Nacional de Investigaciones Cientifi- cas y Tecnicas de la Republica Argentina for its con- stant support. R. Wharton and F. Gilstrap are most grateful to P. Krauter for organizing the database for the Costa Rican collections and providing printouts for host associations, and to M. Fischel, J. Duran, and R. Chavez for their crucial role in conducting the field work in Costa Rica and isolating individual puparia. R. Wharton is also most appreciative of the assistance provided by C. Yoshimoto in making available notes and other information from L. Weld. Illustrations were provided by B. Flahey. LITERATURE CITED Aluja, M., J. Guillen, P. Liedo, M. Cabrera, E. Rios, G. de la Rosa, H. Celedonio, and D. Mota. 1990. Fruit infesting tephritids (Diptera: Tephritidae) and associated parasitoids in Chiapas, Mexico. E)itoinofJhagn 35: 39-48. Baranowski, R., H. Glenn, and ]. Sivinski. 1993. Bio- logical control of the Caribbean fruit fly (Diptera: Tephritidae). Florida Entomologist 76: 245-251. Bartlett, K. A. 1941. The introduction and coloniza- tion in Puerto Rico of beneficial insects parasitic on west indian fruit flies. The journal of Agricul- ture of the Unit'ersily of Puerto Rico 25: 25-31. Blanchard, E. 1947. Insectos del Uruguay. Comunica- cionei Zoologicas, Museo tie Historia Natural de Montevideo 2(42): 11. Borgmeier, T. 1935. Sobre alguns cynipidos parasiti- cos e cecidogenos do Brasil (Hymenoptera: Cy- nipidae). Archivos Instituto Biologico y Vegetal, Rio de Janeiro 2: 97-124. Brethes, J. 1924. Varios himenopteros de la America del Sur. Nunqunm Otiosus 2: 2-16. Clausen, C. P. 1978. Tephritidae. Pp. 320-325. In: C. P. Clausen (Ed.) Introduced parasites and predators of arthropod pests and weeds: a world review. United States Department of Agriculture, Agriculture Handbook 480. Clausen, C. P., D. W. Clancy and Q. C. Chock. 1965. Biological control of the oriental fruit fly {Dacus dorsalis Hendel). United States Department of Ag- riculture Technical Bulletin 1322: 102 pp. Costa Lima, A. da. 1940. Alguns parasites de moscas de frutas. Analcs da Academia Brasdeira de Scwncias 12; 17. Costa Lima, A. da. 1948. Entomofagos sul amcricanos (parasites e predadores) de insetos nocivos a agricultura. Boletim de Sociedade Brasiliera de Agronomia, Rio de Janeiro 11: 1-82. De Santis, L. 1965. Nota sobre un parasito de la mosca sudamericana de la fruta (Hymenoptera: Cyni- pidae). Rei'ista de la Sociedad Entomolgica Argen- tina 27: 73-74. De Santis, L. 1967. Catdlogo de los himenopteros argen- tinos de la serie Parasitica, incluyendo Bethyloidea. Publicacion de la Comision de Investigacion Cientifica de la Provincia de Buenos Aires 337 PP- Dettmer, H. 1929. Ganaspis carvalhoi n. sp. (Hymenop- tera: Cynipidae), um novo parasita da mosca das frutas {Anastrepha fraterculus Wied.). Boletim Biologica, Sao Paulo 16: 70-74. Diaz, N. B. 1974. Anotaciones sobre cinipoideos ar- gentinos. I. Neotropica 20: 17-20. Diaz, N. B. 1986. Ampliacion de la distribucion geo- grafica de cinipoideos parasitoides en la Repiib- lica Argentina y Brasil. Revista de la Sociedad En- tomoldgica Argentina 44: 32. Harding, J. A. 1965. Parasitism of the leaf miner Lir- iomi/za munda in the winter garden area of Texas. journal of Economic Entomology 58: 442^143. Hayward, K. ]. 1940. Lucha biologica contra las mos- cas de las frutas. Dispositivo que permite la sal- ida de los parasitos beneficos del pozo donde se arroja la fruta atacada. Revista Industrial Agri'cola de Tucumdn 10-12: 230-233. Hellen, W. 1960. Die Eucoilinen Finnlands (Hym. Cyn.). Fauna Fennica 9: 1-30. Hernandez-Ortiz, V. 1993 Description of a new Rhag- oletis species from tropical Mexico (Diptera: Te- phritidae). Proceedings of the Entomological Society of Washington 95: 418-424. Hernandez-Ortiz, V., R. Perez-Alonso, and R. A. Wharton. 1994. Native parasitoids associated with the genus Anastrepiha (Dipt.: Tephritidae) in Los Tuxtlas, Veracruz, Mexico. Entomophaga 39: 171-178. Jimenez-Jimenez, E. 1956. Las moscas de la fruta y sus enemigos naturales. Fitdfito 16: 4-11. Jiron, L. F. and R. G. Mexzon. 1989. Parasitoid hv- menopterans of Costa Rica: geographical distri- bution of the species associated with fruit flies (Diptera: Tephritidae). Entomophaga 34: 53-60. Johnson, M. 1987. Parasitization of Lirumiyza spp. (Diptera: Agromyzidae) infesting commercial watermelon plantings in Hawaii, journal of Eco- nomic Entomology 80: 56-61. Katiyar, K. P., J. Camacho, F. Geraud, and R. Ma- theus. 1995. Parasitoides hvmenopteros de mos- cas de las frutas (Diptera: Tephritidae) en la re- gion occidental de Venezuela. Revista de la Facul- tad de Agronoma (LUZ) 12: 303-312. Kerrich, G. J. and J. Quinlan. 1960. Studies on eucoi- line Cynipoidea (Hym.). Opuscuta Entonwlogica 25:179-196. 114 Journal of Hymenoptera Research Kiefter, ]. J. 1909. Description de nouveaux cynipides zoophages. BuUctiu dc In Societe d'Histoire Natii- rclle, Metz (3)2: 8-96. Lin, K. S. 1987. Aganaspis, a new genus of Eucoilidae (Hymenoptera: Cynipoidea). Taiwan Agricultural Rt'scarf/i Institute, Special Puhlicaticn 22: 67-79. Lopez, M., ]. Sivinski, L. Guillen, C. Ruiz, J. Pinero, and M. Aluja. 1996. Reservorios de parasitoides de moscas de la fruta (Diptera: Tephritidae) en el estado de Veracruz, Mexico. In: 2nd Meeting of the working group on fruit flies of the western hemisphere. Via del Mar, Chile, 1996: 70. McAlpine, J. F. and G. C. Steyskal. 1982. A revision of Neosilba McAlpine with a key to the World genera of Lonchaeidae (Diptera). The Canaiiian Entomologist 114: 105-137. Nasca, A. J. 1973. Parasites de "moscas de los frutos" establecidos en algunas zonas de Tucuman. Re- vista Agn'cola del Noroeste Argentino 10: 31—13. Nasca, A. J., R. V. Fernandez, D. R. Sosa, D. Fernan- dez, and A. De Guerrero. 1980. Eficiencia del pozo trampa en el manejo de enemigos naturales de moscas de los frutos. Adas del 2 ° Congreso Na- clonal de Citricultura, Entre Rios 2: 141-162. Nordlander, G. 1978. Revision of the genus Rhoptrom- eris Forster, 1869 with reference to north-western European species. Studies on Eucoilidae (Hym.: Cynipoidea) II. Entonwiogica Scandinnvica 9: 47- 62. Nordlander, G. 1980. Revision of the genus Lepioptliua Forster, 1869, with notes on the status of some other genera (Hymenoptera, Cynipoidea: Eucoil- idae). Entomologica Scandinavica 11: 428^53. Nordlander, G. 1981. A review of the genus Tryblio- grapha Forster, 1869 (Hymenoptera, Cynipoidea: Eucoilidae). Entomologica Scandinavica 12: 381- 402. Nordlander, G. 1982. Identities and relationships of the previously confused genera Odontcucoila, Co- neucoila and Trichoplasta (Hym., Cynipoidea: Eu- coilidae). Entomologica Scandinavica 13: 269-292. Nunez-Bueno, L. 1982. Trybliographa daci Weld (Hy- menoptera: Cynipidae): biology and aspects of the relationship with its host Anastreplm suspensa (Loew) (Diptera; Tephritidae). Ph. D. Disserta- tion, University of Florida: 152 pp. Ovruski, S. M. 1994a. Comportamiento en la detec- cion del huesped de Agaiuispis pclkranoi (Hyme- noptera: Eucoilidae) parasitoide de larvas de Cer- atitis capitata (Diptera: Tephritidae). Revista de la Sociedad Entomologica Argentina 53: 121-127. Ovruski, S. M. 1994b. immature stages of Aganaspis pclleranoi (Brethes) (Hymenoptera: Cynipoidea: Eucoilidae), a parasitoid of Ceratitis capitata (Wied.) and Anastrepha spp. (Diptera: Tephriti- dae). Journal of H\/menoptera Research 3: 233-239. Ovruski, S. M. 1995. Pupal and larval-pupal parasit- oids (Hymenoptera) obtained from Anastrepha spp. and Ceratitis capitata (Dipt.: Tephritidae) pu- pae collected in four localities of Tucuman prov- ince, Argentina. Entoniopliaga 40: 367-370. Ovruski, S. M., S. Fuentes, F. Nunez, and J. G. Gra- nados Zuniga. 1996. Himenopteros parasitoides de moscas de la fruta (Diptera: Tephritidae) pre- sentes en la Republica de El Salvador. Revista de Ingenieros Agronomos de El Salvador 14: 8-14. Poolman Simons, M. T. T., B. P. Suverkropp, L. E. M. Vet, and G. de Moes. 1992. Comparison of learn- ing in related generalist and specialist eucoilid parasitoids. Entomologia Experimeutalis et Appli- cata 64: 117-124. Rohwer, S. A. 1919. Description of a new cynipoid from Trinidad. Proceedings of the Entomological So- ciety of Washington 21: 156. Ruiz, S. L., J. L. Cancino D., and M. Aluja. 1996. Co- lonizacion de parasitoides nativos para el control biologico de moscas de la fruta. In: 2nd Meeting of the working group on fruit flies of the western hemisphere, Via del Mar, Chile, 1996: 70. Salles, L. A. 1996. Parasitismo de Anastrepha fratercu- lus (Wied.) (Diptera: Tephritidae) por Hymenop- tera na regiao de Pelotas, R.S. Pesquisa Agrope- cuaria Brasileira, Brasilia 31: 769-774. Turica, A. and R. G. Mallo. Observaciones sobre la poblacion de las "Tephritidae" y sus endopar- asitos en algunas regiones citricolas argentinas. IDIA 6: 145-161. Valladares, G., N. Diaz, and L. De Santis. 1982. Tres notas sobre dipteros agronomicidos de la Repub- lica Argentina y sus himenopteros parasitoides. Revista de la Sociedad Entomolgia Argentina 42:319- 330. Van Alphen, J. J. M., G. Nordlander, and I. Eijs. 1991. Host habitat finding and host selection of the Drosophila parasitoid Lcptopilina australis (Hyme- noptera, Eucoilidae), with a comparison of the niches of Eurcipean Leptopilina species. Oecologia 87:324-329. Von Ihering, R. 1905. As moscas das frutas e sua des- trui^ao. 1° edi^ao. Secretaria de Agricultura, Sao Paulo. Pp. 12-13. Von Ihering, R. 1912. As moscas das frutas e sua des- trui^ao. 2° edii^ao. Typographia Brazil, Sao Paulo. 48 pp. Von Ihering, R. 1914. Diagnose de uma Eucoela, par- asita das moscas. Revista Museo Paulista 9: 224- 225. Weld, L. H. 1921. Notes on certain genera of parasitic Cynipidae proposed by Ashmead with descrip- tions of genotypes. Proceedings of the United States National Museum 59: 433-451. Weld, L. H. 1932. Synonymical and descriptive note on Pseudeucoila hrasiUensis (R. v. Ihering, 1905) (Hym., Cynipidae), Revista de Entomologia 2: 24- 27. Weld, L. H. 1951a. New Eucollinae (I hnienoptera, Volume 7, Number 1, 1998 115 Cynipoidea). Pwceeditigs of the Entomological So- ciety of Wmtnngton 53: 223-226. Weld, L. H. 1951b. A new species of Tryhliografilui (Hymenoptera: Cynipidae). Proceedings of the Ha- umiinn Entomological Society 14: 331-332. Weld, L. H. 1952. Cynipoidea (Hym.) 1905-1950. Ann Arbor, Privately Printed. Wharton, R. A., F. E. Gilstrap, R. H. Rhode, M. Fis- chel-M., and W. G. Hart. 1981. Fiymenopterous egg-pupal and larval-pupal parasitoids of Cera- titis capitata and Anastrepha spp. (Diptera: Te- phritidae) in Costa Rica. Entomophaga 26: 285- 290. Wishart, G. and E. Monteith. 1954. Tryhliographa rapae (Westw.) (Hymenoptera: Cynipidae), a parasite of Hylemya spp. (Diptera: Anthomiidae). The Ca- nadian Entomologist 86: 145-154. Yepes, R. and R. Velez. 1989. Contribucion al cono- cimiento de las moscas de las frutas (Tephritidae) y sus parasitoides en el departamento de Antio- quia. Rei'ista de la Facultad Nacional de Agronoini'a de Medellin 42: 73-98. J. HYM. RES. Vol. 7(1), 1998, pp. 116-117 NOTE First Chromosome Record for the Family Dryinidae: The Karyotype of Anteon brevicome Dalman (Hymenoptera: Chrysidoidea) Vladimir E. Gokhman and Kirill A. Kolesnichenko Botanical Garden, Moscow State University, Moscow 119899, Russia Though karyotypes of some members of the superfamily Chrysidoidea have been examined during the last few years (Ho- shiba and Imai 1993, Gokhman and Quicke 1995, Quicke and Gokhman 1996), the family Dryinidae remains totally un- touched by chromosomal investigation. We have studied for the first time chro- mosome number and karyotype of the dryinid, Anteon brevicome Dalman. Chro- mosome preparation was obtained from an adult wasp collected from the wild at the Botanical Garden, Moscow State Uni- versity, Moscow, Russia, in May 1997. Preparation was made according to the previously described protocol (Gokhman and Quicke 1995). Chromosomes were subdivided into four groups — metacen- trics, submetacentrics, subtelocentrics and acrocentrics following Levan et al. (1964) and Imai et al. (1977). The voucher speci- men is deposited in the Zoological Muse- um, Moscow State University, Moscow, Russia. RESULTS Eleven well-spread metaphase plates were obtained from the individual stud- ied, all of them having the same diploid chromosome number, 2n = 10 (Fig. 1). All chromosomes are obviously two-armed and thus arm number (NF) in this species is 20. The karyotype comprises three pairs of submetacentric chromosomes and two pairs of subtelocentric ones. However, all chromosomes differ notably in size, each chromosome pair being at least about 1.5 times longer or shorter than the others. Submetacentrics of the first two pairs are the longest in the set (3^ jxm), those of the third pair are the shortest (0.5 ixm), and subtelocentrics are of intermediate length (1-2 ixm). DISCUSSION The above results, together with accu- mulated data on chromosomes of the oth- er Chrysidoidea, provide qualitatively new karyotypic information for the super- family. First, chromosome number of A. brevicome is the lowest in the Chrysidoi- dea and one of the lowest in all aculeate Hymenoptera. Except for a few ant spe- cies, only four predominantly unrelated members of the Aculeata (although two of them belong to the bee genus, Andrena) were reported to have n values of 5 or fewer (Goodpasture 1974, Hoshiba and Imai 1993). Second, the chromosome set of A. brevicome is highly asymmetric (White 1973), apart from karyotypes of the other Chrysidoidea, where chromosomes show a continuous gradation in length (see for example Hoshiba and Imai 1993, Fig. 2c and 8c, and Quicke and Gokhman 1996, Fig. Ic). Finally, ranges of variation in chromosome number in all studied fami- lies of the Chrysidoidea do not overlap. Specifically, n values of 10-14, 19-21 and 5 were found in the Bethylidae, Chrysidi- dae and Dryinidae respectively. Though it is difficult at present to determine path- ways of karyotype evolution in the Chry- sidoidea, low chromosome number and Volume 7, Number 1, 1998 117 n t I) it Fig. 1. Karyotj'pe of Anteoii brevicorne. Scale bar in- dicates 1 (im. karyotype structure found in the Dryini- dae and perhaps resulted from multiple chromosome fusions suggest that those features may be apomorphic together with the other apomorphies of this specialized group (Brothers and Carpenter 1993). However, further chromosomal investi- gation of the Dryinidae and other Chry- sidoidea will be necessary to confirm this assumption. LITERATURE CITED Brothers, D. J. and J. M. Carpenter. 1993. Phylogeny of Aculeata: Chrysidoidea and Vespoidca (Hy- menoptera). Journal of Hyniciwplern Rcfcanh 2: 227-302. Gokhman, V. E. and D. L. J. Quicke. 1995. The last twenty years of parasitic Hymenoptera karyolo- gy: An update and phylogenetic implications. Journal of Hymenoptera Research 4: 41-63. Goodpasture, C. 1974. Cytological data and classifi- cation of the Hymenoptera. Unpublished Ph.D. thesis. University of California, Davis. 178 pp. Quicke, D. L. J. and V. E. Gokhman. 1996. First chro- mosome records for the superfamily Ceraphron- oidea and new data for some genera and species of Evanioidea and Chrysididae (Hymenoptera: Chrysidoidea). Journal of Hymenoptera Research 5: 203-205. Hoshiba, H. and H. T. Imai. 1993. Chromosome evo- lution of bees and wasps (Hymenoptera, Apo- crita) on the basis of C-banding pattern analyses. Japanese Journal of Entomology 61: 465-492. Imai, H. T., R. H. Crozier and R. W. Taylor. 1977. Karyotype evolution in Australian ants. Chro- mosoma 59: 341-393. Levan, A., K. Fredga and A. A. Sandberg. 1964. No- menclature for centromeric position on chromo- somes. Hereditas 52: 201-220. White, M. J. D. 1973. Animal Cytology and Evolution. Cambridge University Press, Cambridge. 961 pp. ]. HYM. RES. Vol. 7(1), 1998, pp. 118-121 NOTE A Putative Pheromone-gland Associated Modification of the Hind Tibia in Vipio moneilemae (Hymenoptera: Braconidae: Braconinae) Donald L. J. Quicke and Jose Vincent Falco (DLJQ) Department of Biology, Imperial College of Science, Technology and Medicine, Silwood Park, Ascot, Berks SL5 7PY, UK; (JVF) Departamento de Ciencias Ambientales y Recursos Naturales, Universidad de Alicante, Ap. Correos 99, E-03080 Alicante, Spain More is known about the pheromone and other exocrine glands of braconid wasps than any other group of parasitic Hymenoptera, although behavioural evi- dence suggests that they are present in many if not all groups and that their prod- ucts collectively have a variety of roles in- cluding mate location, host marking, de- fence and spacing (Quicke, 1997). The ma- jority of pheromone glands identified to date are located in the metasoma (Wese- loh 1980; Tagawa 1983; Buckingham and Sharkey 1988; Quicke 1991; Field and Kel- ler 1994; Quicke et al. 1996, 1997) although recent studies have indicated that the an- tennae also have a wide range of intrinsic glands that are involved in courtship as well as host assessment (Bin et al. 1986, 1989; Isidoro and Bin 1995; Isidoro et al. 1996). Here we describe and illustrate a modification of the hind tibia of a bracon- ine wasp, Vipio moneilefiiae Gahan, that was noticed during revision of the North American species of Vipio (Inayatullah et al. 1998). The morphology of the structure suggests that it is yet another exocrine gland, and its presence only in males fur- ther suggests that, if the newly described structure is associated with a gland, then this may be a release site for sex phero- mones. It is worth noting that the hind tib- ia of an unrelated braconid, the chelonine Ascogaster reticulatus Watanabe, has been identified as the source of a sex phero- mone, but in this case the glands are pres- ent only in females (Kainoh and Oishi 1993). Males of this species apparently fol- low female pheromone trails as do the males of the aphelinid Aphelinus asychis Walker (Fauvergue et al. 1995). The male hind tibia of V. moneilemae is especially thickened for most of its length (Figs. 1, 3) whereas that of the female (Fig. 2, 4) is essentially unmodified. Along al- most the whole of the dorsal surface of the male hind tibia, there is a deep groove bordered laterally by a distinct ridge on both the inner and outer sides. Within this groove, there is a dense arrangement of setae that point somewhat posteriorly and whose tips converge towards the mid-line of the groove (Figs. 5, 6). In some uncle- aned specimens, the setae are covered by a deposit, perhaps the dried secretion of an associated gland, though this has not been confirmed. We were not able to see any cuticular pores but these may be small and the setae obscured the view of most of the floor of the groove. The arrange- ment and close-spacing of the setae in the tibial groove suggest that they could act as a wick and provide an increased evap- orative area for the release of pheromones; similar evaporative setal structures have been found associated with the metasomal glands of agathidine braconids (Bucking- Volume 7, Number 1, 1998 119 Figs. 1-6. Features of the hind leg of Vipio inonciUinac. 1, 3, 5, 6, male leg showing sivollen tibia with dorsal groove and associated setal arrangement; 2, 4, female showing unmodified tibia. Scale bar: I = 860 ji.m; 2, 3 = 380 ^.m; 4 = 136 (j.m; 5 = 100 |j.m; 6 = 50 M-m. 120 Journal of Hymenoptera Research ham and Sharkey 1988). No similar struc- tures have been found on the legs of any other species of Vipio, nor on those of oth- er braconines. Vipio species are idiobiont larval ecto- parasitoids of concealed hosts living in wood. Unfortunately, very little is knov^n about V. moneilemae apart from the fact that it has been reared from a species of Moneilema (Cerambycidae) in Mexico (Ga- han 1930), and it is known from only a handful of specimens from Mexico and the southern USA (California, Texas, Utah: most of the known specimens are in the United States National Museum of Natu- ral History). The majority of parasitic wasps are protandrous, and the males of many that attack wood-boring hosts con- gregate at sites from which females are about to emerge, and where a variety of competitive tactics may ensue. If the struc- ture seen on the hind tibia of male V. mo- neilemae is indeed associated with a sex pheromone gland, it may suggest a radi- cally different mating strategy. Large putative sex pheromone glands are found in the metasomata of males of many genera of braconines (Quicke 1991) as well as in a range of other braconids such as many opiines, alysiines and agath- idines (Buckingham and Sharkey 1988). Most work, including the chemical analy- sis of glandular products, has been carried out on those of some Opiinae (Williams t'f al 1988) and it has been proposed that these have a role in courtship although a defensive role cannot be excluded. With- out doubt, more work needs to be done on the roles of male exocrine glands in the Braconidae and other parasitic wasps. ACKNOWLEDGEMENTS We would like to thank Rachel Kruft for assistance with electron microscopy and Dave Smith for the 'oan of specimens of V. moneikmnc. This work was supported by the NERC (Natural Environment Re- search Council) Initiative in Taxonomy. LITERATURE CITED Bin, F., S. Colazza, N. Isidoro, M. Solinas and S. B. Vinson. 1989. Antennal chemosensilla and glands, and their possible meaning in the repro- ductive behavior of Trissolcus basalis (Woll.) (Hym.: Scelionidae). Entomcilogicn 24; 33-97. Bin, F., M. R. Strand and S. B. Vinson. 1986. Antenna structures and mating behavior in Trissolciif htis- nlis (Woll.) (Hym.: Scelionidae), egg parasitoid of the green stink bug. In INRA (ed.) Trichogramnia and Other Egg Parasites. 2nd International Sympo- sium, Guangzhou (China), Nov. 10-15 1986. Les Colloques de ITNRA No. 43, Paris, pp. 144-151. Buckingham, G. R. and M. J. Sharkey. 1988. Abdom- inal exocrine glands in Braconidae (Hymenop- tera). pp. 199-242. In V. K. Gupta (ed.) Advances in Parasitic Hymenoptera Research. E. J. Brill Pub- lishing Co., Leiden. 546 pp. Fauvergue, X., K. R. Hopper and M. F. Antolin. 1995. Mate finding via a trail sex-pheromone by a par- asitoid wasp. Proceedings of the National Acadenn/ of Sciences of the U.S.A. 92:' 900-904. Field, S. A. and M. A. Keller. 1994. Localization of the female sex pheromone gland in Cotesia rubecula Marshall (Hymenoptera: Braconidae). journal of Hymenoptera Researcli 3: 151-156. Gahan, A. B. 1930. Synoptical and descriptive notes on parasitic Hymenoptera. Proceedings of the United States National Museum 77: 1-12. Inayatullah, M., S. R. Shaw and D. L. J. Quicke. 1998. The genus Vipio LatreiUe (Hymenoptera: Bracon- idae) of America North of Mexico. Journal of Nat- ural History 32: 117-148. Isidoro, N. and F. Bin. 1995. Male antennal gland of Amitus spinifcrus (Br.thes) (Hymenoptera: Platy- gastridae), likely involved in courtship behav- iour, htternational Journal of Insect Morphology and Embryology 24: 365-373. Isidoro, N., F. Bin, S. Colazza and S. B. Vmson. 1996. Morphology of antennal gustatory sensilla and glands in some parasitoid Hymenoptera with hy- pothesis on their role in sex and host recognition. Journal of Hymenoptera Research 5: 206-239. Kainoh, Y. and Y. Oishi. 1993. Source of sex phero- mone of the egg-larval parasitoid, /l.';cc\\;i!sfi'r re- twulatus Watanabe (Hymenoptera, Braconidae). Journal of Chemical Ecology 19: 963-969. Quicke, D. L. J. 1991. Tergal and inter-tergal glands of male Braconinae. Zoologica Scripta 19: 413—423. Quicke, D. L. J. 1997. Parasitic Wasps. Chapman & Hall, London, 470 pp. Quicke, D. L. J., R. A. Wharton and H. Sittertz-Bhat- kar. 1996. Recto-tergal fusion in the Braconinae (Hymenoptera, Braconidae): Distribution and structure, journal of Hymenoptera Research 5: 73- 79. Quicke, D. L. ]., K.A. Wharton and H. Sittertz-Bhat- kar. 1997. The antero-lateral metasomal scent Volume 7, Number 1, 1998 121 glands of the Braconinae (Hymenoptera, Bracon- idae): Structure & function, journal of Hymcnoy- tera Research 6: 219-230. Tagawa, ]. 1983. Female sex pheromone glands in the parasitic wasps, genus Apanteles. Ap^'Iicd Ento- mologi/ and Zoology 18: 416-427. Weseloh, R. M. 1980. Sex pheromone gland of the gypsy moth parasitoid, Apanteles metanoscelus: Revaluation and ultrastructural survey. Annals of the Entomological Societ}/ of America 73: 576- 580. Williams, H., M. A. Wong, R. A. Wharton and S. B. Vinson. 1988. Hagen's gland morphology and chemical content analysis for three species of parasitic wasps (Hymenoptera: Braconidae). Journal of Chemical Ecology 14; 1727-1736. J. HYM. RES. Vol. 7(1), 1998, pp. 122-123 NOTE A Note on the Mating Behavior of Anoplitis amethystinus Fabricius (Hymenoptera: Pompilidae) Karen R. Sime' and David B. Wahl (KRS) Department of Entomology and Section of Ecology and Systematics, Cornell University, Ithaca, New York 14853, USA; (DBW) American Entomological Institute, 3005 S.W. 56th Avenue, Gainesville, Florida 32608, USA To date, nothing has been reported on the biology of Anopliiis (Notiochares) ame- thystinus Fabricius, a widespread New World pompihd ranging from Argentina to parts of the southern United States (Cal- ifornia, Arizona, New Mexico, Texas, Georgia, Florida) (Wasbauer & Kimsey 1985). A handful of remarks may be found on the biology of related species; to this growing knowledge we add the following observations on A. amethystinus. The encounter occurred in the Riverside Pine Island (longleaf pine (Pinus palustris Miller) woods) of the Ocala National For- est, Florida (29°25'N, 8r47'W). On 11 April 1997, around noon on a lightly over- cast day, we noticed a group of male pom- pilids in the tire rut of a seldom-used, sandy dirt road. Our presence scattered them, but as soon as we stepped back 1- 2 meters, they returned to an oblong area of sand (about 10 by 30 cm) in which it appeared that the surface debris of pine needles, dead leaves, and small twigs had been displaced to the periphery by the ac- tivities of the male wasps. We counted 12- 15 males in this space at any one time, some flying off occasionally to circle the area or to visit nearby shrubs. They walked around in small circles, flicking their wings and drumming their antennae ' Corresponding author on the ground. They did not fight or oth- erwise respond to each others' presence, although in their tight quarters they often touched each other. After watching for a few minutes and capturing some of the males that flew out from the group, we swept aside the loose sand (about 5 mm deep) in the clear area, and then started excavating the compact- ed, cement-like sandy substrate. A few millimeters down we found the opening of a burrow (6 mm diameter) with a pom- pilid female near the top, scraping at the sand in an apparent effort to dig herself out: her wings were open and dry, and she flew directly into the waiting net. When we stepped aside, 4 or 5 males re- turned; several popped headfirst into the burrow, then backed out after a few sec- onds. Further excavation revealed an empty nest similar to those reported for Anoplius (Pompilinus) tenebwsus Cresson, A. (P.) viaticus L. (Aim & Kurczewski 1984), and A. (Arachnophroctotius) apicula- tus pretiosis Banks (Kurczewski & Spofford 1986): exit tunnel straight, ca. 5 cm long, angled 50-60° from surface, with terminus unicellular and widened only very slight- Similar behavior in male Anoplius has been reported in a few other species. Rau (1922) found four males of A. (Notiochares) atramentarius Dahlbom gathered around a Volume 7, Number 1, 1998 123 newly emerged female, and males of A. tenebrosus (Aim & Kurczewski 1984) and A. apiculatus auttimnalis Banks (Evans et al. 1953) have been observed to fly low over sandy areas and occasionally alight. These observations and ours are consistent with the suggestion of Evans et al. (1953) that mating may only be successful with vir- gins in some species (but do not rule out other strategies). Such behaviors also im- ply the use of pheromones in combination with habitat cues in locating female emer- gence areas, and our observations in par- ticular suggest that antennation of the ground may be involved in location of fe- males at short range. Voucher specimens (three males and the female) are deposited at the American En- tomological Institute. We are grateful for the help of Mike McDonald, Rex Rowan, Bill Wcislo, John Wenzel, and several anonymous reviewers, and we thank the staff of the Lake George Ranger District for granting us permission to work in the Ocala National Forest. LITERATURE CITED Aim, S. R., and F. E. Kurczewski. 1984. Ethology of AiiopHiui tenchrosus (Cresson) (Hymenoptera: Pompilidae). Proceedings of the Entomological So- ciet}/ of Washington 86: 110-119. Evans, H. E., C.-S. Lin, and C. M. Yoshimoto. 1953. A biological study of Anopilius apiculatus nutum- nalis (Banks) and its parasite, Evageles mohai'e (Banks) (Hymenoptera, Pompilidae). Journal of the New York Entomological Society 61: 61-77. Kurczewski, F. E., and M. G. Spofford. 1986. Obser- vations on the behaviors of some Scoliidae and Pompilidae (Hymenoptera) in Florida. Florida Entomologist 69: 636-644. Rau, P. 1922. Ecological and behavior notes on Mis- souri insects. Transactions of the Academy of Sci- ence of St. Loins 24(7): 1-71. Wasbauer, M., and L. Kimsey. 1985. California spider wasps of the subfamily Pompilinae. Bulletin of the California insect Sunvy 26: 1-130. Vol. 7(1) J. HYM. RES. 1998, pp. 124-126 BOOK REVIEW Annotated Keys to the Genera of Nearctic Chalcidoidea (Hi/nienoptern). Gary A. P. Gib- son, John T. Huber and James B. Woolley (Editors.). NRC Research Press, Ottawa, 1997. xi + 794 pp. Price: Canada CAN$64.95, other countries the equivalent of US$64.95 (hardcover). ISBN 066016669-0. It has been said that it is dangerous to study the Parasitic Hymenoptera, for those who do are inclined to become al- coholics or end up in mental institutions. Notions of such despair undoubtedly stem from the daunting task that confronts those who venture to unravel the identity of this vast and taxonomically complex as- semblage of enigmatic wasps. The superfamily Chalcidoidea alone comprises close to 19,000 valid species in 2,000 genera, with estimates of the actual number of species ranging from between 60,000 to 100,000 world wide. Compound- ing the difficulties in coming to grips with the identity of this bewildering array of mostly poorly known taxa is the plethora of scattered and often inaccessible descrip- tive literature spanning some 200 years of taxonomic endeavour on the group. The task is made even more arduous by a shortage of essential identification tools such as modern catalogues, revisions, keys and adequately documented refer- ence collections. In Britain, arguably the cradle of Chalcidology, there are for in- stance no modern identification keys for more than half of the 5,000 or so species of Hymenoptera (Weeks et al. 1997, Bull. Ent. Res. 87: 203), many of which belong to chalcidoid groups. What is clearly needed, is to take stock of what is known, collate and synthesize the vast pool of scattered knowledge, much of which is buried in a few inquir- ing minds, and make it readily accessible in such a way that it is both palatable and of practical value to all. Indeed, this is not only imperative in giving credence to the ideals of the Convention on Biological Di- versity, but also in proving the worth of insect systematists in a world of changing attitudes to science. The 17 authors of Annotated Keys to the Genera of Nearctic Chalcidoidea receive full credit for having done exactly that. In harnessing their collective skills to the full they have produced an outstanding syn- opsis of the 19 families and 706 chalcidoid genera known to occur in the region. But the true value of this book lies far beyond that of a manual merely for identifying specimens. It its also an impressive refer- ence work, crammed with invaluable no- menclatural, bibliographical and host in- formation on the Nearctic chalcidoid fau- na. The book comprises 22 chapters. The in- troduction, which contains interesting background information on the history of Chalcidology in North America, numbers of taxa and the rate at which genera have been described from the region through the years, is followed by a chapter on mor- phology. Morphological terms pertaining to the keys and various family diagnoses are conveniently highlighted in bold, clearly explained and depicted in a series of uncluttered line drawings and good quality scamiing electron micrographs. For quick reference, terms are listed al- phabetically in a separate appendix with their abbreviations and reference to fig- ures. A second appendix cross-references the abbreviations with terms. The disad- vantage of having this information sepa- rate, and not in figure legends below the plates, is that some paging is required to determine the meanings of the abbrevia- tions depicted in the illustrations. Chapter three provides, besides a key to Volume 7, Number 1, 1998 125 families, an overview of the Chalcidoidea and covers topics such as recognition of the group, mode of parasitism (feeding types), host ranges and relationships, be- haviour, collecting methods and voucher specimens. Written in clear language, this is essential reading matter which should appeal in particular to the non-specialist seeking general background information on the superfamily at a glance. The family key is primarily a derived combination of those found in A Hand- book of the families of Nearctic Chalci- doidea and Hymenoptera of the World. In recognizing 19 families, the authors have opted for the traditional system of classi- fying chalcidoids at this level. Considering the intended purpose of the book and the instability of the higher classification of these wasps, in which anything from nine to 24 families have been recognized, this is a sensible choice. Preceding the key is a flow chart which gives a pleasing visual impression of how the 19 families are keyed out in the 40 cou- plets. The key itself is visually less pleas- ing. Placing the figures associated with each alternate of a couplet directly above it allows for easy comparison between fig- ures and text, but often results in insuffi- cient space for both alternates on the same page. Consequently, many pages are left with unsightly blank spaces, while the al- ternates of some couplets appear on dif- ferent pages, making comparisons some- what cumbersome. More important though, is the fact that the key works well. Evidently, much careful thought has gone into making it both accurate and workable for the novice and specialist alike. Chosen for diagnostic reliability rather than con- venience, the characters are not necessar- ily "easy" to use, but being appropriately and clearly illustrated there should be no problem in keying specimens out. The bulk of the remaining text is devot- ed to the family treatments. Each of the 19 chapters reviews one family and includes, apart from a generic key, sections on its recognition, systematics, biology, fauna and literature. For larger families a useful index to genera based on couplet numbers is also included. These topics are all well covered, serving as a concise and infor- mative introduction to the 19 families as represented in the Nearctic region. In essence, this book is all about generic keys to Nearctic chalcidoids, and its real value will ultimately be judged by how well these keys perform. So I decided to put some of them to the test, selecting specimens not only from the Nearctic re- gion, but also from further afield. The re- sults were excellent, bearing testimony to the skillful and competent way in which these keys have been constructed. The contents of the couplets are unambiguous and the taxonomic language easy to com- prehend. I was pleased to see that the au- thors have, in all but one of the tested keys, managed to avoid using the handy but dreadful "either/or" method of sepa- rating taxa. All of this adds much to the relative ease with which the user is guided through the keys, even in the case of gen- era which are difficult to define. Although these keys have been designed specifically for the Nearctic region, they will, if used with the necessary insight and care, also serve as a very useful tool in the study of extra-limital forms, as was evident from the large selection of Afrotropical speci- mens which keyed out perfectly well. Greatly enhancing the value of these keys are brief annotations to each genus com- prising references to existing species keys, estimated number of species, known dis- tribution and host range in the region. The use of the keys is facilitated by 1,880 line drawings and scanning electron n\icrographs which, despite the inputs from a variety of illustrators, are of a con- sistently high aesthetic and technical stan- dard. With the intended readership of the book in mind, a few more eye-catching habitus drawings, which are completely lacking for some families, would have added a special touch. 126 Journal of Hymenoptera Research The book concludes with four very use- ful indices. The first one comprises about 130 generic and subgeneric chalcidoid names (with their nomenclatural fate) which have become invalid since publi- cation of the most recent (1979) edition of Catalog of Hymenoptera in America North of Mexico. Following this index are separate lists of plant and animal hosts, and supraspecific chalcidoid names. In summary, I applaud the contents of this book, hence the lack of any serious criticism. Naturally, this is not the last word on the supraspecific identity of Ne- arctic chalcidoids, for our knowledge and understanding of the fauna is far from complete. But the authors have succeeded admirably in synthesizing what is pres- ently known while providing an excellent framework for further revisionary work on the group. 1 recommend this book as an indispensable reference and identifica- tion tool for anyone, specialists and gen- eralists alike, involved in the study of par- asitic wasps. Gerhard Prinsloo, Biosystematics Divi- sion ARC-Plant Protection Research Institute Pretoria, South Africa. ANNOUNCING THE PUBLICATION OF MANUAL OF THE NEW WORLD GENERA OF THE FAMILY BRACONIDAE Edited by Robert A. Wharton, Paul M. Marsh and Michael J. Sharkey Contributing Authors Paul C. Dangerfield, Paul M. Marsh, Donald L. J. Quicke, Michael J. Sharkey, Scott R. Shaw, Cornelis van Achterberg, Robert A. Wharton, and James B. Whitfield Special Publication Number 1 International Society of Hymenopterists This 439 page, multiauthored identification manual presents illustrated keys to the 34 subfamilies and 404 genera of the family Braconidae from the New World. Information about all New World genera described up to 1994 is included. The manual consists of 37 chapters by world au- thorities plus an index to generic names including subgenera and syn- onyms, and an appendix listing all the New World genera alphabetically by subfamily. The first chapter provides an introduction including a re- view of braconid biology, literature, classification, biogeography, collec- tion and curation, and a discussion of the manual format. The second chapter is an extensive illustrated discussion of braconid morphology and terminology used in the keys. The third chapter is a fully illustrated key to the New World subfamilies of the Braconidae. The New World subfamilies are then presented separately in the remaining 34 chapters. For each, a key to the New World genera is included. Each of these keys is annotated to include estimated number of species, distribution, hosts, and critical references for each genus. Each subfamily chapter is exten- sively illustrated and the entire manual contains over 750 line drawings and scanning electron micrographs. A valuable identification tool for biological control workers and museum curators Price: US$20.00 In the United States, prepaid orders are sent postage free. Outside the U.S., postage will be extra based upon destination and method of shipment (i.e., air mail or surface). Send orders (check made out to "Department of Entomology") to: Robert A. Wharton, Department of Entomology, Texas A&M University, College Station, TX 77843 'uSA INSTRUCTIONS FOR AUTHORS General Policy. The loiirnal of HymenopHcra Research invites papers of high scientific quality reporting comprehensive research on all aspects of Hymenoptera, including biology, behavior, ecology, systematics, taxonomy, genetics, and morphology. Taxonomic papers describing single species are unlikely to be accepted unless a strong case is evident, such as importance in economic entomology or with concurrent biology or ecology. Manuscript length generally should not exceed 50 typed pages; however, no upper limit on length has been set for papers of exceptional quality and importance, including taxonomic monographs at generic or higher level. All papers will be reviewed by at least two referees. The referees will be chosen by the appropriate subject editor. However, it would be helpful if authors would submit the names of two persons wno are competent to review the manuscript. The language of publication is English. Summaries in other languages are acceptable. The deadline for receipt of manuscripts is 1 October (for the April issue) and 1 April (for the October issue). Format and Preparation. Three copies of each manuscript, including copies of illustrations, should be submitted on letter size or A4 paper, double spaced, with at least 25 mm margins on all sides. On the upper left of the title page give name, address and telephone and fax numbers of the author to whom all corre- spondence is to be sent. The paper should have a concise and informative title, followed by the names and addresses of all authors. The sequence of material should be: title, author(s), abstract, text, acknowledgments, literature cited, appendix, figure legends, figure copies (each numbered and identified), tables (each numbered and with heading). Each of the following should start a new page: (1) title page, (2) abstract, (3) text, (4) literature cited, (5) figure legends, (6) footnotes. Following acceptance of the manuscript, the author should provide the editor with one copy of the manu- script accompanied by a copy on diskette using DD, double sided computer diskettes — IBM compatible MS DOS 5.25 inch or IBM and Macintosh 3.5 inch diskettes. (Authors who do not have access to a computer should submit three copies of the manuscript.) The paper may be submitted in most PC and Mac word processor prograi sucn as Microsoft Word, FullWrite Professional, WordPerfect, WriteNow, Nisus, Mac- Write, or MacWrite II. It possible, all words that must be italicized should be done so, not underscored. Tables may be formatted in a spread sheet program such as MS Works or MS Excel. Text should be double spaced typing, with 25 mm left and right margins. Tables should be put in a separate file. Diskettes should be accompanied by the name of the software program used (e.g., WordPerfect, Microsoft Word). Authors should keep backup copies of all material sent to the Editor. The Society cannot be responsible for diskettes or text mislaid or destroyed in transit or during editing. Illustrations should be planned for reduction to the dimension of the printed page (14 x 20.5 cm, column width 6.7 cm) and allow room for legends at the top and bottom. Do not make plates larger than 14 x 18 in. (35.5 X 46 cm). Individual figures should be mounted on a suitable drawing board or similar heavy stock. Photographs should be trimmed, grouped together and abutted when mounted. Figure numbers should be on the plate, and it is strongly recommended that names be included after the numbers (e.g.. Fig. 2, ff.vfl>u(s). Include title, author(s) and address(es), and illustration numbers on back of each plate. Original figures need not be sent until requested by the editor, usually after the manuscript has been accepted. Reference to figures/ tables in the text should be in the style "(Fig. 1)" "(Table 1)". Measurements should be in the metric system. All papers must conform to the International Code of Zoological Nomenclature. The first mention of a plant or animal should include the full scientific name including the authority. Genus names should not be abbre- viated at the beginning of a sentence. In taxonomic papers type specimens must be clearly designated, type depositories must be clearly indicated, and new taxa must be clearly differentiated from existing taxa by means of keys or differential diagnoses. Authors are required to deposit all type material in internationally recognized institutions (not private collections). Voucher specimens snould be designated for specimens used in behavioral or autecological studies, and they should be deposited similarly. Acceptance of taxonomic papers will not require use of cladistic methods; however, authors using them will be expected to specify the phylogenetic program used (if any), including discussion of program options used. A data matrix should be provided if the subject is complex. Cladograms must be hung with characters and these should include descriptors (not numbers alone) when feasible. The number of parsimonious clado- grams generated should be stated and reasons given for the one adopted. Lengths and consistency indices should be provided. Adequate discussions should be given for characters, plesiomorphic conditions, and distributions of characters among outgroups when problematical. References in the text should be (Smith 1999), without a comma, or Smith (1999). Two articles by a single author should be (Smith 1999a, 1999b) or Smith (1999a, 1999b). For multiple authors, use the word "and," not the symbol "&" (Smith and Jones 1999). For papers in press, use "in press," not the expected publication date. The Literature Cited section should include all papers referred to in the paper. Journal names should be spelled out completely and in italics. Charges. Publication charges are $10.00 per printed page. At least one author of the paper must be a member of the International Society of Hymenopterists. Reprints are charged to the author and must be ordered when returning the proofs; there are no free reprmts. Author's corrections and changes in proof are also charged to the author. Color plates will be billed at full cost to the author. All manuscripts and correspondence should be sent to: Dr. E. Eric Grissell Systematic Entomology Laboratory, USDA % National Museum of Natural History, NHB-168 Washington, DC 20560 Phone: (202) 382-1781 Fax: (202) 786-9422 E-mail: egrissel@sel.barc.usda.gov CONTENTS (Continued from front cover) WHARTON, R. A., S. M. OVRUSKI, and F. E. GILSTRAP. Neotropical Eucoilidae (Cyni- poidea) associated with fruit-infesting Tephritidae, with new records from Argen- tina, Bolivia and Costa Rica 102 NOTES GOKHMAN, V. E., and K. A. KOLESNICHENKO. First chromosome record for the family Dryinidae: The karyotype of Anteon brevicorne Dalman (Hymenoptera: Chrysidoidea) 116 QUICKE, D. L. J., and J. V. FALCO. A putative pheromone-gland associated modifi- cation of the hind tibia in Vipio moneilemae (Hymenoptera: Braconidae: Bracon- inae) 118 SIME, K. R., and D. B. WAHL. A note on the mating behavior oi Anoplius amethysiinus Fabricius (Hymenoptera: Pompilidae) 122 BOOK REVIEW PRINSLOO, G.— Gary A. P. Gibson, John T. Huber and James B. Woolley (Editors). 1997. Annotated Keys to the Genera of Nearctic Chalcidoidea (Hymenoptera) .... 124 ANNOUNCEMENT International Society of Hymenopterists, Special Publication Number 1: Manual of the New World Genera of the Family Braconidae 127 c,oClETy Journal of Hymenoptera Research ^3 '^M/olume 7, Number 2 October 1998 ISSN #1070-9428 CONTENTS ARCHER, M. E. Worker versus sexual, and sex ratio investments in the social u^asp Vespula vulgaris (L.) (Hymenoptera: Vespinae) in England 257 GESS, F. W. Priscomasaris namihiensis Gess, a new genus and species of Masarinae (Hyme- noptera: Vespidae) from Namibia, southern Africa, with a discussion of its position within the subfamily 296 GIBSON, G. A. P., and V. VIKBERG. The species of Asaphes Walker from America north of Mexico, with remarks on extralimital distributions and taxa (Hymenoptera: Chal- cidoidea, Pteromalidae) 209 LIU, ZHIWEI. A new species of Ihalia from Borneo, with a revised phylogeny and historical biogeography of Ibaliidae (Hymenoptera: Cynipoidea) 149 MACEDO, M. V. DE, M. C. P. PIMENTEL, and R. C. VIEIRA. Response of Pithecellobium tortum Martius (Leguminosae) seeds to the attack of the phytophagous braconid AUorhogas dyspistus Marsh (Hymenoptera: Braconidae) 274 MARTINS, R. P., L. A. SOARES, and D. YANEGA. The nesHng behavior and dynamics of Bicyrtes angiilata (F. Smith) with a comparison to other species in the genus (Hy- menoptera: Sphecidae) 165 NAUMANN, I. D., and H. GROTH. A revision of the Philomastigine sawflies of the world (Hymenoptera: Pergidae) 127 OLIVEIRA, L., R. MELO, and ]. TAVARES. Response of Glypiapanteks militaris (Walsh) (Hymenoptera: Braconidae), a larval parasitoid of the armyworm, Mythimna uni- puncta (Haworth) (Lepidoptera: Noctuidae), to different temperatures 268 (Continued on back cover) INTERNATIONAL SOCIETY OF HYMENOPTERISTS Organized 1982; Incorporated 1991 OFFICERS FOR 1998 James M. Carpenter, President Andrew D. Austin, President-Elect James B. WooUey, Secretary John T. Huber, Treasurer E. Eric Grissell, Editor Subject Editors Symphyta and Parasitica Aculeata Biology: Mark Shaw Biology: Sydney Cameron Systematics: Donald Quicke Systematics: Wojciech Pulawski All correspondence concerning Society business should be mailed to the appropriate officer at the following addresses: President, Department of Entomology, American Museum of Natural His- tory, Central Park West at 79th Street, New York, New York 10024; President-Elect, Uiiiversity of Adelaide, Glen Osmond, Australia; Secretary, Department of Entomology, Texas A&M University, College Station, Texas 77843; Treasurer, Eastern Cereal & Oilseed Research Centre, Agriculture Canada, K. W. Neatby Building, Ottawa, Ontario, Canada KIA 0C6; Editor, Systematic Entomology Laboratory, USDA, % National Museum of Natural History, Washington, D.C. 20560-0168. Membership. Members shall be persons who have demonstrated interest in the science of ento- mology. Annual dues for members are US$35.00 per year (US$40.00 if paid after 1 May), payable to The International Society of Hymenopterists. Requests for membership should be sent to the Treasurer (address above). Information on membership and other details of the Society may be found on the World Wide Web at http://IRIS.biosci.ohio-state.edu/ish. Journal. The Journal of Hymenoptera Research is published twice a year by the International Society of Hymenopterists, % Department of Entomology, Smithsonian Institution, Washington, D.C. 20560-0168, U.S.A. Members in good standing receive the journal. Nonmember subscriptions are $60.00 (U.S. currency) per year. The Society does not exchange its publications for those of other societies. Please see inside back cover of this issue for information regarding preparation of manuscripts. Statement of Ownership Title of Publication: Journal of Hymenoptera Research. Frequency of Issue: Twice a year. Location of Office of Publication, Business Office of Publisher and Owner: International Society of Hymenopterists, % Department of Entomology, NHB 168, Smithsonian Institution, Wash- ington, D.C. 20560, U.S.A. Editor: E. Eric Grissell, Systematic Entomology Laboratory, USDA, % National Museum of Nat- ural History, Washington, DC 20560-0168. Managing Editor and Known Bondholders or other Security Holders: none. This issue was mailed 20 October 1998 J. HYM. RES. Vol. 7(2), 1998, pp. 127-148 A Revision of the Philomastigine Sawflies of the World (Hymenoptera: Pergidae) I. D. Naumann and H. Groth (IDN) CSIRO Division of Entomology, P. O. Box 1700, Canberra, Australian Capital Territory 2601, Australia; (HG) "Timbarrah", M/S 1073, Crows Nest, Queensland 4355, Australia Abstract. — The world genera and species of Philomastiginae (Cerospastus Konow, Ecopatus Smith and Philomastix Froggatt) are revised and an illustrated, identification key is provided. Cewspastus volupis Konow and Ecopatus penai Smith from South America and Philomastix nancarrowi Froggatt, P. macleaii (Westwood) and P. xanthophylax sp. n. from eastern Australia are described or re- described. Larvae of C. volupis feed on foliage of species of Nothofagus (Fagaceae); those of Phi- lonmstix on foliage of Rulms (Rosaceae) oi Alphitonin (Rhamnaceae). All species of Philomastix exhibit maternal guarding of eggs and larvae. The trans-Antarctic distribution of the Philomasti- ginae suggests a Gondwanan origin for the subfamily. Distinctive Gondwanan elements are novi' known within numerous Australian families of Hymenoptera (Naumann 1991). Over a dozen families or subfami- lies and even a few genera are shared by Australia and South America or by Aus- tralia, South America, New Zealand and southern Africa. Examples of such trans- Antarctic distribution patterns can be found among the Symphyta, Megalyro- idea, Evanioidea, Ichneumonoidea, Proc- totrupoidea, Platygasteroidea, Chalcidoi- dea, Vespoidea and Apoidea and include phytophages, parasitoids and predators. Generally these Gondwanan elements are readily recognisable as archaic clades within their respective superfamilies or families. Some, such as the parasitoid ge- nus Monomachus Klug (family Monomach- idae) are morphologically conservative and comprise in total only a few dozen species (Johnson 1992). Some, such as Monomachus, are very rarely collected or are restricted to temperate, forest habitats. In contrast such Gondwanan groups as the thynnine Tiphiidae are spectacularly varied and speciose; the many hundreds of species (Naumann 1991) occur over a wide range of habitats and some of these species are extremely common. Seven subfamilies or tribes of Symphy- ta, the Dereycyrtinae (Xiphydriidae), the Guiglini and Leptorussini (Orussidae), the Scobini (Argidae), and the Pergulinae, Perreyinae and Philomastiginae (Pergi- dae) all exhibit trans-Antarctic distribu- tion patterns. The Dereycyrtinae compris- es the monotypic Austrocyrta Riek in Aus- tralia and three genera totalling sixteen species in South America (Smith 1988; Smith 1995). Larvae of this subfamily are unknown but probably develop within wood. The Guiglini comprise six Austra- lian and two New World species, and the Leptorussini consist just a South American species, five Australian species and a spe- cies in South Africa (Smith 1988). All orus- sid larvae are believed to be ectoparasitic on the larvae of wood-boring beetles and siricid wood wasps. The orussid tribe Ophrynopini, with representatives in south-east Asia and South America might also be considered to be a "trans-Antarc- tic" taxon (N. Springate, pers.comm.). The Scobinini comprise Scobinn Lepeletier and Serville, a common New World genus of 128 Journal of Hymenoptera Research about 50 species, and Antargidium Morice, an Australian genus of six species (Nau- mann 1991). Larvae of two species of Scob- ina feed on Sida Linnaeus (Malvaceae) but nothing is known of the biology of Antar- gidium. The subfamily Pergulinae (Pergi- dae) comprises the genus Haplostegus Ko- now with fifteen species in South America and the monotypic Pergula Morice in south-western, mainland Australia (Smith 1990). The biology of the single, rare spe- cies of Pergula is unknown but larvae of the South American pergulines have been recorded feeding on Myrtaceae, especially Psidium guajava L. (guava). Perreyinae are represented on the western side of the Pa- cific ocean by two species with flightless females (described respectively from Su- lawesi and New Guinea) and east of the Pacific by about 80 species in South Amer- ica (Smith 1990). Larvae of South Ameri- can species feed variously on Asteraceae and Malvaceae. The Philomastiginae con- sists of two monotypic South American genera, Cerospastus Konow and Ecopatus Smith, and three eastern Australian spe- cies of Philomastix Froggatt (Fig. 75). The two previously described species of Phi- lomastix are well known species of tropical and subtropical rainforests. Their larvae feed on the foliage of species of Rubus L. (Rosaceae) (Leask 1944) or Alphitonia Reis- sek ex Endl. (Rhamnaceae) (Jackson 1993) and females are commonly encountered standing guard over egg batches and young larvae (Fig. 72). Members of the subfamily can be recognised using the keys of Smith (1990) and Naumann (1991). The present paper revises the generic diagnoses and key of Smith's (1990) treat- ment of the South American Philomasti- ginae, redescribes all previously named species, describes a new species of Philo- mastix from south-eastern Queensland (Fig. 1), and collates all known biological and distributional data on the subfamily. Morphological and biological studies of the larvae of Philomastix are continuing (Groth and Naumann unpublished). Fig. 1. Philomastix xanthophylax. Scale line = 2mm. TERMINOLOGY The morphological terminology used here follows Goulet and Huber (1993). DEPOSITORIES AMSA Australian Museum, Sydney, Australia ANIC Australian National Insect Col- lection, CSIRO Division of En- tomology, Canberra, Australia BPBM Bernice P. Bishop Museum, Ho- nolulu, Hawaii, U.S.A. CNC Canadian National Collections of Insects, Arachnids and Nem- atodes, Ottawa, Canada BCRI Biological and Chemical Re- search Institute, N. S. W. De- partment of Agriculture, Rydal- mere, Australia BMNH The Natural History Museum, London, U. K. DEIC Institut fur Pflanzenschutzfor- schung der Akademie der Land wirtschaftswissenschaf ten (formerly:Deutsches Entomolo- gisches Institut, Eberswalde, Germany) FCNI State Forests Research Institute, Beecroft, Australia Volume 7, Number 2, 1998 129 HGCN H. Groth Collection, "Timbar- rah/' via. Crows Nest, Australia MVMA Museum of Victoria, Mel- bourne, Australia UMO Hope Department of Entomol- ogy, University Museum, Ox- ford, U. K. QMBA Queensland Museum, Brisbane, Australia QDPl Entomology Collection, Queens- land Department of Primary In- dustries, Brisbane, Australia SAMA South Australian Museum, Adelaide, Australia UQIC Department of Entomology, Uruversity of Queensland, Bris- bane, Australia USNM National Museum of Natural History, Smithsonian Institu- tion, Washington, D. C, U. S. A. PHILOMASTIGINAE Rohwer Pterygophorinae (part): Froggatt 1890b: 696; Ashmead 1898: 231. Ptergophorides (part): Konow 1898: 248; Ko- now 1905b: 37. Philomastiginae Rohwer 1911a: 220; Benson 1935: 224; Benson 1938: 379; Riek 1970a: 891; Riek 1970b: 218; Smith 1978: 159; Naumann 1984: 345, 347; Smith 1990: 13, 21-23; Nau- mann 1991: 934-935; Macdonald and Ohmart 1993:493-^95. Diagnosis. — Antenna 14-21-segmented, filiform in female, serrate in male. Head capsule open (neither postgenae nor hy- postomae forming continuous bridge be- tween occipital and foramen and oral fos- sa). Maxillary palp 4- or 6-segmented; la- bial palp 3- or 4-segmented. Labium api- cally single or tri-lobed. Ventral arms of cervical sclerites pointed, neither meeting nor forming precoxal bridge with proster- num; notauli complete; mesothoracic ster- nopleural suture present; distance be- tween cenchri less than half width of cen- chrus. Mid- and hind tibia each with pre- apical spine; foretibia with 2 apical spines. Costa of forewing narrower than intercos- tal area; forewing without anal cell. Ab- dominal tergum 1 sclerotised. KEY TO GENERA OF PHILOMASTIGINAE 1. Maxillary palp 4-segmented; labial palp 3-segmented, apical segment with large sensory cup occupying more than half length of segment (Fig. 70), maxillary palp with smaller sensory cup; labium single lobed; female without cercus Philomastix Froggatt - Maxillary palp 6-segmented; labial palp 4-segmented, apical segments without sensory cup; labium tri-lobed; female with cercus 2 2. Forewing with 4 cubital cells (IRl, IRs, 2Rs, 3Rs) (Fig. 4); forewing radial cell closed; fourth maxillary palp segment slender (Fig. 9); antenna of female weakly clavate (Fig. 3); meso- scutum conspicuously setose Cerospastiis Konow - Forewing with 2 or 3 cubital cells; fore-wing radial cell open (Fig. 5); fourth maxillary palp segment apically distinctly wider than other segments (Fig. 8); antenna of female not cla- vate (Fig. 2); mesoscutum almost devoid of setae Ecopatiis Smith PHILOMASTIX Froggatt Philomastix Froggatt 1890a: 487^88; Froggatt 1890b:696; Froggatt 1901: 1070; Dalla Torre 1894: 308; Ashmead 1898: 231; Konow 1898; 248, 249; Konow 1905: 36-37; Schulz 1906: 81-82; Rohwer 1911: 87; Morice 1919: 248, 255, 287; Tillyard 1926: 265; Benson 1935: 224; Benson 1938: 379-380; Riek 1970a: 880, 889, 891; Smith 1978: 159; Smith 1990: 21-23; Pag- liano and Scaramozzino 1990: 157; Abe and Smith 1991: 65; Naumann 1991: 926, 933, 935; Macdonald and Ohmart 1993: 493-496; Nau- mann 1994: 414-415. Type species: Perga nan- 130 Journal of Hymenoitera Research 4,5 2,6,7 8,9 Figs. 2-9. Ecopatus penae, Cerospastus volnpis: 2, £. pcnae, female antenna; 3, C.volupis, female antenna; 4, C. vohipis, forewing (part); 5, £. penac. forewing (part); 6, £. pcnac, lateral panel of first abdominal tergum; 7, C. I'olupis, apex of male abdomen, dorsal view; 8, £. penae, maxillary palp palp; 9, C. mhipia, maxillary palp. Scale lines = 1.0 mm. carrowi Froggatt (designated by Rohwer 1911). Perga Leach: Westwood 1880: 372; Kirby 1882: 26. Dalla Torre 1894: 351. (part) Heptncola Konow: Konow 1905a: 167 (part) Female. — Vertex conspicuously setose. Face with some fine microsculpture. Malar space narrower than diameter of anterior ocellus. Antenna 14-19-segmented, weak- ly serrate, not clavate. Right mandible simple (Fig. 69). Maxillary palp 4-seg- mented, filiform (Fig. 70); apical segment with conspicuous sensory cup, this less than half as long as segment. Labial palp 3-segmented, apical segment with con- spicuous sensory cup, this more than half as long as segment. Labium single lobed. Thorax dorsally conspicuously setose. No- tauli, median mesoscutal line deeply im- pressed. Mesoscutellum swollen so that posterior margin concealed from above. Mesepisternum with tubercle. Metascutel- lum in form of transverse band. Forewing with closed radial cell and 4 cubital cells (e.g. Fig. 1). Median, second cubital and third cubital cells each with nygma (small, corneous spot). Abdominal terga dorsally densely setose. Second tergum predomi- nantly smooth. Cercus absent. Ovipositor sheath posteriorly strongly expanded, in posteroventral view with prominent, flat- tened, semicircular surface (e.g. Fig. 21). Head, mesoscutellum, metascutellum orange-yellow. Legs brown to dark brown. Wings with background smokey brown appearance, especially in female. Costal cell of fore wing distinctly more darkened than remainder of wing. Male. — Antenna 15-21-segmented, stron- glyserrate (e.g. Fig. 11). Eighth tergum pos- Volume 7, Number 2, 1998 131 teriorly with moderate or very weak emar- gination (e.g. Fig. 20). Discussion. — Larvae of Philomastix have long been known to feed externally on leaves of various shrubby and scandent species of Riibus (known as blackberries or native raspberries) (Leask 1944). More re- cently (present paper and Jackson 1993) Philomastix larvae have been discovered defoliating trees of the genus Alphitonia (Figs. 72-74). Two species of Alphitonia are attacked: Alphitonia petriei Braid & C. T. White, the pink ash, which is widespread in Queensland and the northern part of the Northern Territory, and Alphitonia ex- celsa (Fenzl) Benth., the red ash, which oc- curs in rainforests and eucalypt wood- lands of Queensland, New South Wales, and the northern parts of the Northern Territory and Western Australia (Francis and Chippendale 1970; Lazarides and Hince 1993). The hosts of Philomastix are thus a shrub and a tree from different plant families. Rubus and Alphitonia do have at least one, ecological characteristic in common — both are pioneer species. Various Rtthiis are well known species of paths, clearings and margins of closed for- ests and A. petriei is the most prominent recolonising tree species in upland rain- forest clearings in north Queensland (Jack- son 1993). A species of "Philomastix" has been re- corded feeding on Eucalyptus sp. at Laun- ceston, Tasmania (Anonymous 1980). This is almost certainly a misidentification: it is the only record of a eucalypt as host and we have seen no authentic specimens of Philomastix from Tasmania in the course of this revision. Philomastix does not occur over the en- tire range of its host plants. The sawfly ge- nus is restricted to the eastern Australian states of Queensland and New South Wales between the latitudes 16°S and 35°S. However Alphitonia occurs beyond this range in the Northern Territory and West- ern Australia and suitable species of Rubus are common in Victoria (Bruzzese 1980). P. nancarwwi and P. macleaii are strictly allopatric (Fig. 76). P. nancarwwi is con- fined to the rainforests and closely adja- cent eucalypt woodlands of the Queens- land Wet Tropics between the northern end of the Atherton Tableland and the Mount Spec-Paluma area. Almost all spec- imens have been collected at altitudes above 500m. P. macleaii is very widely dis- tributed from Eungella in central Queens- land to the lllawarra district in south-east- ern New South Wales. It occurs in tropi- cal, subtropical and temperate rainforest and nearby moist woodlands. In the northern parts of its range (Eungella, Kroombit tops, Bunya Mountains) P. ma- cleaii is not known below 1000m but south of about Brisbane it occurs near sea level. The region of eastern Queensland between Paluma and Eungella which separates P. nancarrowi and P. macleaii is relatively dry; the vegetaHon is dominated by eucalypt woodland and there are no significant patches of rainforest. North from Paluma the so-called "base-of-peninsula" rainfor- est system stretches with some interrup- tions to Cooktown. On the other hand Eungella stands as the northern extremity of a rainforest system that extends, also with interruptions, all the way to southern New South Wales. There are taxonomic discontinuities in several orders of insects (especially mesothermic Odonata, Plecop- tera and Megaloptera) at the gap between Paluma and Eungella (Kikkawa et al. 1981; Watson and Theischinger 1984) and these are thought to reflect Pleistocene fluctuations in climate. During Pleistocene glacial periods, increased aridity resulted in a contraction of the rainforests and a pronounced inhospitable gap between moist refugia north of Paluma and south of Eungella. Presumably vicariance speci- ation occured on either side of this barrier within Philomastix and various odonate, stonefly and alderfly genera. P. xanthophylax has a much more cir- cumscribed distribution pattern (Fig. 76). It has been collected in moist woodlands 132 Journal of Hymenoptera Research rather than rainforest but in the Brisbane nal care. Females stand over or near their area it is sympatric with P. macleaii. There egg masses and young larvae (Fig. 72). is at present no model to explain the ori- When disturbed they shake from side to gin of this species. side and rapidly vibrate their wings to The oviposition behaviour of Philomas- produce a loud buzzing sound. Females tix is unusual for a pergid sawfly. The will guard their offspring, usually from a adult female pierces the leaf of the host position on a petiole or stem, and have plant, pushes each elongate egg through been observed to guide larvae to new the perforation, and attaches one end of leaves (Leask 1944). Females eventually the egg to the underside of the leaf (Mac- die at their post, and sometimes dozens of donald and Ohmart 1993). The other end dried bodies remain hanging on defoliat- of the egg is pushed free of the lower sur- ed host plants. Parental defence of off- face of the leaf. The more typical behav- spring is widespread within the Symphy- iour for pergids is to cut a shallow slit in ta. It has been recorded in three other sub- the host plant and insert the egg into this families of Pergidae (Naumann 1984; Mac- slit. The lancet of most pergids is saw-like donald and Ohmart 1993) and in the in appearance but the lancet of Philomastix Pamphiliidae (Kudo et al. 1992). In those (Figs. 22, 39, 58, 71) is auger-shaped, pre- species which have been studied quanti- sumably to facilitate the "pierce-and- tively it has been shown that female push" oviposition habit. guarding reduces predation on eggs by All species of Philomastix exhibit mater- other arthropods. KEY TO SPECIES OF PHILOMASTIX Female; abdomen with saw-like ovipositor often concealed within sheath (Figs. 1, 21, 46, 55) 2 Male; ovipositor and sheath absent 4 Abdomen orange-yellow (Fig. 1); antennal flagellum black; ovipositor sheath posterome- dially dentate (Figs. 54, 55) xanthophylax sp.n. Abdomen with extensive dark brown, black and cream markings (Figs. 62,64); antennal flagellum orange-yellow; ovipositor sheath not posteromedially dentate (Figs. 21, 46) .... 3 Mesoscutellum usually with posterolateral tubercles (Figs. 30-34); mesepisternal tubercle strong (Fig. 38); abdominal terga 3-5 usually with cream, lateral maculae visible from above (Fig. 62); posterolateral margin of first abdominal tergum weakly curved or straight (Fig. 37), spiracle separated from margin of tergum by a distance 0.9-1.1 times maximum diameter of spiracle macleaii (Westwood) Mesoscutellum without posterolateral tubercles (Fig. 16) although sometimes posteriorly emarginate; mesepisternal tubercle usually weak (Fig. 26); abdominal terga 3-5 without lateral, cream maculae visible from above (Fig. 64); posterolateral margin of first abdominal tergum strongly curved almost angulate (Fig. 19), spiracle separated from margin of ter- gum by a distance 1.6-1.7 times maximum diameter of spiracle nancarroivi Froggatt Mid lobe of mesoscutum with brown to black coloration reaching anterior margin (Fig. 63); diameter of anterior ocellus 0.9-1.0 times minimum distance between antennal sockets; sixth abdominal tergum with lateral, cream or orange-yellow macula visible from above (Fig. 63); mesoscutellum usually with posterolateral tubercles mflc/ffliiXWestwood) Mid lobe of mesoscutum reddish orange anteriorly and brown to black posteriorly or lobe entirely reddish orange; diameter of anterior ocellus approximately 0.7 times minimum distance between antennal sockets; sixth abdominal tergum usually without orange or yellow maculae visible from above; mesoscutellum without tubercles, although posterior margin sometimes distinctly concave 5 Seventh abdominal tergum with widely separated, yellow or orange maculae; posterolat- Volume 7, Number 2, 1998 133 eral margin of first abdominal tergum strongly curved, almost angulate (as in Fig. 19); mesoscutellum posteriorly weakly emarginate (Fig. 17) or straight mmcarrowi Froggatt Seventh abdominal tergum with yellow or orange band across posterior margin, band sometimes narrowly interrupted medially; posterolateral margin of first abdominal tergum weakly curved (as in Fig. 53) or straight; mesoscutellum posteriorly straight or weakly convex xanthophylax sp.n. Philomastix nancarroivi Froggatt (Figs. 10-26, 64-66, 76) Philomastix nancarrowi Froggatt 1890a: 488-489; Froggatt 1890b: 696; Konow 1898: 250; Schulz 1906: 82-83; Rohwer 191 IB: 87; Morice 1919: 287-288, Plate XI, fig. 13, Plate XII, figs 5,6, Plate XIII, fig. 15; Leask 1943: 2; Leask 1944: 1-3; Smith 1978: 159; Bruzzese 1980: 4; Abe and Smith 1991: 65; Macdonald and Ohmart 1993: 493-494; Naumann et al. 1994: 71. Philomastix nancarrowi: Dalla Torre 1894: 308; Konow 1905: 37; Rohwer 1911: 87. (Miss- spelling) nee Philomastix nancarrowi : Forsiusl927: 283. (Misidentification) Female. — Body length 9.5-11.0 mm. Forewing length 10.6-12.3 mm. Distance between antennal sockets 2.8-3.9 times di- ameter of anterior ocellus. Antenna (Fig. 10) 15-18-segmented; apical 2-A segments sometimes fused so that antenna appar- ently 13-16-segmented (traces of interseg- mental sutures sometimes present). First flagellar segment 2.8-3.8 times longer than wide. Second flagellar segment 1.7-1.9 times wider apically than basally. Meso- scutellum without posterolateral tuber- cles, posteriorly rounded or weakly emar- ginate (Fig. 16); anterodorsal surface flat (Fig. 15) to weakly concave. Mesepisternal tubercle weak (Fig. 26). Ovipositor sheath not posteromedially dentate (Fig. 21). Pos- terolateral margin of first abdominal ter- gum strongly curved (Fig. 19), almost an- gulate, spiracle separated from margin by a distance 1.6-1.7 times maximum diam- eter of spiracle. Lancet as in Figs. 22-25, about 8 ctenidia terminating in dorsal tooth, anterior teeth weak Flagellum orange-yellow. Mid lobe of mesoscutum orange-yellow (Fig. 64). Each lateral lobe of mesoscutum with large, dark brown to black macula occupying most of length of lobe. Mesepisternum with broad, brown band adjacent to ster- nopleural suture (Fig. 66). Mesosternum brown with narrow, orange-yellow band along anterior and lateral margins. Meta- scutum orange-yellow to pale brown, usu- ally not much darker than cenchri. Fore- wing with broad, transverse, brown band; cell IM usually completely brown (Fig. 12) or forewing entirely brown (Fig. 13). Ab- domen predominantly dark coloured with some metallic reflections. Abdominal ter- gum 1 orange-yellow to pale brown. Terga 3-5 dark brown, black or metallic blue. Terga 6 and 7 each with cream, lateral macula; maculae narrowly separated mid- dorsally by dark brown or black, or mac- ulae contiguous. Tergum 8 cream. Tergum 9 orange-yellow. Abdominal sterna 2 and 3 at least in part orange-yellow or cream. Sterna 4—7 dark brown to black. Oviposi- tor sheath orange-yellow. Male. — Body length 7.1-9.4 mm. Fore- wing length 7.1-8.1 mm. Distance be- tween antennal sockets 1.7-2.0 times di- ameter of anterior ocellus. Antenna (Fig. 11) 15-17-segmented, apical 2 segments sometimes fused. First flagellar segment 1.6-2.0 times wider than long. Second fla- gellar segment 1.8-2.2 times wider apical- ly than basally. Mesoscutellum sometimes distinctly emarginate posteriorly (Fig. 17); anterodorsal surface usually more con- cave (Fig. 14). Mesepisternal tubercle as in female. Tergum 8 posteriorly with deep emargination (Fig. 20). Genitalia as in Fig. 18, paramere moderately broad, gonolaci- nia not strongly hooked, penis valve trun- cate. 134 Journal of Hymenoptera Research ■12, -10. II 13 1«, 22 23-25 14-17,19,20,21,26 Figs. 10-26. Pliiloinastix luiucanoun: 10, female antenna; 11, male antenna; 12, 13, forewings; 14, male meso- scutellum, surface contour as seen from rear of insect; 15, female mesoscutellum, surface contour as seen from rear of insect, 16, female mesoscutellum, dorsal view; 17, male mesoscutellum, dorsal view; 18, male genitalia, dorsal view; 19, lateral panel of first abdominal tergum; 20, apex of male abdomen, dorsal view; 21, apex of ovipositor sheath, posteroventral view; 22-25, lancet; 26, mesepisternum, profile. Scale lines = 1.0 mm for 10- 17, 19-21, 26; = 0.1 mm for 18, 22-25. VoLUMK 7, Number 2, 1998 135 Clypeus and labrum creamy white. Mid lobe of mesoscutum anteriorly orange-yel- low, posteriorly brown (Fig. 65). Lateral lobe of mesoscutum entirely brown to dark brown. Abdominal terga 5 and 6 without orange-yellow, lateral maculae. Tergum 7 with lemon-yellow, lateral mac- ula. Type. — Holotype female. Cairns (ANIC, on permanent loan from Macleay Muse- um, University of Sydney, examined). Material examined. — Queensland: 1 fe- male, Myola, 1909 (ANIC); 10 females, 22 males, Kuranda, IV. 1902, R. E. Turner (BMNH, USNM, QMBA); 1 female, same locality, 3. V.-20. VI. 1913, R. E. Turner (BMNH);1 male, same locality, 15. IV. 1931, A. N. Burns (MVMA); 1 female, same locality, H. J. Carter (AMSA); 1 fe- male, 2 males, 1.5 km SE Kuranda, 16-17. V. 1980, I. D. Naumann, J. C. Cardale (ANIC); 2 females, Barron Falls, via Ku- randa, 16. VI. 1971, E. F. Riek (ANIC); 1 female. Cairns, 1905, H. Elgner (ANIC); 1 male. Cairns district, F. P. Dodd (SAMA); 1 female, Danbulla (forestry Reserve), 14. V. 1959, G. W. S. (BMNH);1 female, Mt Baldy,approximately 8 km W of Atherton, 13. 1. 1977, M. S. and B. J. Moulds (AMSA); 1 male, 12 miles (19.3 km) from Ivanhoe (?Mine), 5. III. 1961, R. Straatman (ANIC); 1 female, Herberton, 17. 111. 1922 (QMBA); 1 female, same locality, 30. V. 1943, M. F. Leask (QMBA); 4 females, same locality, 4. VI. 1944, M. F. Leask (AMSA); 3 females, Eubenangee, 14. V. 1950, G. Brooks (MVMA); 1 female, Bartle Frere, 5. V. 1928 (QDPI); 1 female, Laceys Creek, Mission Beach, 22. IV. 1970, S. R. Curtis (ANIC); 1 female, Tully, 24. IV. 1931, A. N. Burns (MVMA); 1 female, 4 miles (6.4 km) W Paluma, 13. IV. 1969, I. F. B. Common, M. S. Upton (ANIC); 2 females, Paluma, II.- IV. 1992, R. Jackson (AN1C);18 females. Little Crystal Creek, Mt Spec, 16. V. 1971, E. F. Riek (ANIC). Unlocalised: 2 females, "North Queensland," V. 1944, M. F. Leask (QMBA); 1 female, "Atherton Tableland," 14. III. 1934 (QDPI); 1 female, "Atherton Tableland," 19. 111. 1958, N. H. L. Krauss (USNM); 1 female (QMBA). Distribution. — See Fig. 76. The series split between the BMNH, USNM and QMBA is labelled "Cairns Kur 4. 02." 1 have interpreted this as meaning that the specimens were collected at Kuranda, a well known entomological collecting lo- cality approximately 15 km north west of Cairns. Locality labels for specimens col- lected at Kuranda during the early 1900s commonly bear the additional information "Cairns" or "Cairns district." The collec- tor's name is omitted from the USNM and QMBA specimens. Forsius (1927) recorded P. nancarroivi (and P. macleaii) from speci- mens collected at Dorrigo, in north eastern New South Wales by W. Heron and de- posited in the SAMA. 1 have re-examined these specimens and all are clearly iden- tifiable as macleaii. Discussion. — One male from Mt Lewis, north Queensland listed below under "Other material examined", differs signif- icantly from the description of P. nancar- roivi males given above. In the Mt Lewis male the mesoscutellum is rather truncate posteriorly and thus approaches the tu- berculate condition of P. macleaii; tergum 8 is only weakly emarginate; the dark brown coloration of the vertex does not extend to the posterior surface of the head; and the the clypeus, labrum and meso- scutum are all orange-yellow. The speci- men may represent an extreme of varia- tion in P. nancarroivi or an additional, un- described species. It is perhaps not surprising that an anomalous Pbilomastix occurs on Mt Lewis. The mountain is the most northerly, known locality for the ge- nus. It is part of the Carbine Tableland, a discrete, rainforest-covered upland re- markable for the uniqueness of its fauna. The Carbine Tableland supports the high- est number of endemic vertebrate species in the Queensland Wet Tropics (Nix 1991) and Mt Lewis itself harbours several en- demic species of insects: a stag beetle of the genus Sphaenognathus Buquet which is 136 Journal of Hymenoptera Research known elsewhere only from the Black- down Tableland in central Queensland and Andean South America (Moore 1978; Monteith 1996), a primitive leafhopper representing a tribe otherwise unknown from Australia but recorded from Mada- gascar, New Zealand, Chile and Juan Fer- nandez and a species of flightless dung beetle (Kikkawa et al. 1981). Larval host plants. — Rosaceae: Rubus ros- ifolius Sm. (Leask 1943, 1944); Riilms hillii F. Muell.(Leask 1944; Bruzzese 1980); Ru- biis moluccamis L. (Bruzzese 1980). Rham- naceae: Alphitonia petriei (Jackson 1993). Other material examined. — QUEENS- LAND: 1 male, Mt Lewis, via Julatten, 4. V. 1970, S. R. Curtis (ANIC). Philomastix tnacleaii (Westwood) (Figs. 27^7, 62, 63 76) Perga madeaii Westwood 1880: 372-373, Plate XXXV, fig. 2.; Kirby 1882: 26. Pergn madeayi : Dalla Torre 1894: 351 (unjusti- fied emendation). Heptacola madeayi : Konow 1905a: 167. Philomastix madeaii : Froggatt 1918: 671; Smith 1978: 159; Bruzzese 1980: 4; Smith 1980: 342; Macdonald and Ohmart 1993: 493-494; Nau- mann 1993: 8, 115, 183. Philomastix madeayi : Morice 1919: 248, 265, 287- 288; Tillyard 1926: 265, Plate 21, fig. 2; For- sius 1927:283 (part). Philomastix nancarroivi : Forsius 1927: 283 ( mis- identification). Philomastix glabra Froggatt 1890a: 489-490; Froggatt 1892; 201; Konow 1898: 248; Frog- gatt 1901; 1070; Schulz 1906: 83-84. (Synon- ymised, as glaber, with madeaii by Froggatt 1918: 671.) Philomastix glaber : Froggatt 1890b: 696; Froggatt 1893: 201; Froggatt 1901: 1070, Plate; Froggatt 1907: 73, Plate XI; Froggatt 1918: 671.(Mis- spelling of glabra). Female. — Body length 12.0-14.3 mm. Forewing length 12.2-13.7 mm. Distance between antennal sockets 2.2-2.6 times di- ameter of anterior. Antenna (Fig. 27) 17- 19-segmented; apical 2-3 segments some- times fused so that antenna apparently 15-18-segmented. First flagellar segment 2.5-2.9 times longer than wide. Second fla- gellar segment 1.6-1.9 times wider apical- ly than basally. Mesoscutellum usually with posterolateral tubercles (Figs. 30-34); anterodorsal surface distinctly (Fig. 36) or weakly (Fig. 35) concave. Mesepisternal tubercle strong (Fig. 38). Posterolateral margin of first abdominal tergum weakly curved (Fig. 37) or straight, spiracle sepa- rated from margin by a distance 0.9-1.1 times maximum diameter of spiracle. Ovi- positor sheath not posteromedially den- tate (Fig. 46). Lancet as in Figs. 39^2, about 5 ctenidia terminating in dorsal tooth. Flagellum orange-yellow. Mid lobe of mesoscutum usually with large, brown macula occupying more than half of length of lobe (Fig. 62); mid-lobe some- times uniformly orange-yellow. Lateral lobe of mesoscutum with large, dark brown to black macula occupying most of length of lobe. Mesepisternum usually with broad, brown band adjacent to ster- nopleural suture. Mesosternum usually brown to anterior margin and sternopleur- al sulcus, sometimes entirely orange-yel- low. Metascutum brown to dark brown, usually distinctly darker than cenchri. Forewing (Fig. 47) with narrow, trans- verse, brown band; cell IM rarely com- pletely brown. Abdomen predominantly dark coloured with some metallic reflec- tions. Abdominal tergum 1 dark brown to black. Tergum 2 brown to black or metal- lic blue, on each side with large, cream, macula. Terga 3 and 4 dark brown, black or metallic blue, with smaller, cream lat- eral maculae. Terga 5, 6 and 7 each with large, cream, lateral macula; maculae nar- rowly separated by dark brown or black, or contiguous. Tergum 8 anteriorly dark brown to metallic blue (usually concealed by preceding tergum), posteriorly cream. Tergum 9 dark brown to black. Abdomi- nal sterna 1 and 2 at least in part yellow or cream. Sterna 3-6 dark brown to black. Ovipositor sheath orange-yellow, with or without dark brown markings. Volume 7, Number 2, 1998 137 ,27-29 30-38 Figs. 27-38. P/ii/oHMs/i.v rimcleair. 27, female antenna; 28, male antenna, basal antennomeres, frontal view; 29, male antenna, lateral view; 30-34, mesoscutellum, dorsal view, variation in shape; 35, 36, mesoscutellum, variation in surface contour as seen from rear of insect; 37, lateral panel of first abdominal tergum; 38, mes- episternum, profile. Scale lines =1.0 mm. Male.— Body length 11.0-12.8 mm. Fore- wing length 7.4-11.4 mm. Distance be- tween antennal sockets 1.0-1.3 times di- ameter of anterior ocellus. Antenna (Figs. 28, 29) 18-21-segmented; apical 2 seg- ments sometimes fused (occasionally without trace of sutures) so that antenna apparently 17-21-segmented. First flagel- 138 Journal of Hymenoptera Research Figs. 39^7. Plntiimastix madeaii: 39^2, lancet; 43, 44, apex of male abdomen, dorsal view; 45, male genitalia, dorsal view; 46, apex of ovipositor sheath, posteroventral view; 47, forewing. Scale lines = 0.1 mm for 39-42, 45; = 1.0 mm for 43, 44, 46, 47. lar segment 1.2-1.6 times longer than wide. Second flagellar segment 1.9-2.4 times wider apically than basally. Meso- scutum and mesepisternal tubercle as in female. Tergum 8 posteriorly with shallow emargination (Fig. 43, 44). Genitalia as in Fig. 45, paramere broad, gonolacinia not strongly hooked, penis valve apically rounded. Clypeus and labrum orange-yellow. Mesoscutum entirely brown to dark brown (Fig. 63). Abdominal terga 5 and 6 with orange-yellow, lateral macula. Ter- gum 7 with lemon yellow, lateral macula. Types. — P. madeaii: holotype female, "Australia" (UMO; examined). P. glabra: 4 syntypes, "Australasia," Dunoon, Rich- mond River (ANIC, on permanent loan from Macleay Museum, University of Sydney, examined). Volume 7, Number 2, 1998 139 Other material examined. — Queensland: 1 male, Mt William, Eungella National Park, 1100-1200 m,19. IV. 1979, G. B. Monteith (QMBA); 1 male, Olmara Hills, via Dal- rymple Heights, 1000m, 5. IV. 1975, D. K. McAlpine (AMSA); 1 female, Eungella, 25. IV. 1931, W. A. McDougall (QDPI); 1 fe- male. Three Moon Creek, Kroombit Tops, 3-4 II. 1984, G. B. Monteith, C. Hagen, D. Yeates (QMBA); 1 female, Kroombit Tops, 1000-1100 m, 22-26. II. 1982, G. B. Mon- teith, G. Thompson, D. Yeates (QMBA); 1 female. Forest Station, Bulburin State For- est, 600 m, 12-13. IV. 1974, I. D. Naumann (UQIC); 2 females, Imbil, 13. V. 1937, 30. III. 1938, A. R. Brimblecombe (QDPI); 1 fe- male, Palmwoods, 15. IV. 1911, Miss Ede (QMBA); 1 female, Montville, C. Deane (UQIC); 1 female, 6 males, Mt Kiangarow, Bunya Mountains, 27. 1.1993, K. J. and C. L. Lambkin (QMBA, ANIC); 1 female, Bunya Mountains National Park, 11-13. XII. 1979, M. Schneider (UQIC); 1 female, Mt Glorious, 4. IV. 1959, K. H. L. Key (ANIC); 1 female, same locality, II. 1960, J. Bryan (UQIC); 2 females, same locality, 853 m, 13. III. 1960, R. Straatman (ANIC); 1 female, 1 male, same locality, 5-8. II. 1961, 10. IV. 1962, J. L. and M. Gressitt, Malaise trap (BPBM, USNM); 2 females, same locality, 17. I. 1963, T. Brooks (BPBM, USNM); 1 female, same locality, 1. II. 1968, H. McDougall (UQIC); 1 male, Mt Tenison-Woods, 4. II. 1983, G. Daniels (UQIC); 1 female, Highvale, 12. III. 1960, A. Cameron (UQIC); 1 female, Brisbane, 24. III. 1929, L. F. (MVMA); 1 male, same locality. III. 1953, L. W. Rule (UQIC); 5 fe- males, Brookfield, 30. III. 1990, 2. IV. 1990, 10. IV. 1993, 20. IV. 1994, J. Grigg (UQIC, ANIC); 1 male, Toowoomba, 30. XII. 1917, J. A. Bock (UQIC); 1 female, Mt Tambori- ne, 1893, C. Wild (QMBA); 1 female, same locality, 20. II. 1911, W. W. Froggatt (ANIC); 1 female, same locality, 1923, W. H. Davidson (QMBA); 1 female, same lo- cality, 21. II. 1927, H. Hacker (QMBA); 1 female, same locality, C. Deane (UQIC); 1 female, same locality (QDPI); 1 female. Boonah, 5. FV. 1948 (QDPI); 1 female, Can- ungra, 7. IV. 1928 (QDPI); 1 female, Cun- ningham's Gap, 800m, R. Eastwood (UQIC); 1 female, same locality. III. 1972, R. Baldwin (QMBA); 1 female, 1 male, Beechmont, 1. II. 1972, A. and G. Daniels (AMSA); 1 male, Mt Huntley, 1250 m, 29- 30 I. 1993, G. B. Monteith (QMBA); 1 male. Bald Mountain, 1219 m, 28-31 I. 1972, 1. D. Naumann (UQIC); 3 females. Upper Nerang, III. 1891, H. Tryon (QMBA, QMBA); 1 female. Upper Currumbin, 27. IV. 1932, L. Franzen (MVMA); 3 females, 1 male, McPherson Range, XI. 1928, A. J. Turner (MVMA); 1 male, same locality, H. Tryon (QDPI); 6 females, Lamington Na- tional Park, (some labelled 300 feet = 914 m), 2-3. I. 1921, 26-27. II. 1921, 1-11. III. 1921 (QMBA); 7 females, same locality, XII. 1921, H. Hacker (QMBA, ANIC); 2 males, same locality, 900-lOOOm, 16-18. II. 1964, J. Sedlacek (BPBM); 1 female, same locality, 930 m, 3 II. 1983, W. C. Paine (ANIC); 2 females, same locality, 914 m, 6-7. III. 1980. H. E. Evans, A. Hook (UQIC); 1 female, Binna Burra, I. 1943 (QDPI); 3 females, 2 males, Springbrook, 1930, R. Blackwood (MVMA, ANIC); 1 fe- male, same locality, 2. XI. 1961, 1. C. Cun- ningham (UQlC).New South Wales: 1 fe- male, 2 males, Mt Clunie, 16. XII. 1972, I. D. Naumann (UQIC); 4 females, Tweed River, 1897 and no date (BRI, USNM, BMNH); 1 female, same locality, 1920, W. W. Froggatt (BMNH); 1 female, Wiangaree State Forest, 12. II. 1978, K. Walker (UQIC); 2 males, Richmond River (BRI); 1 male, Huonbrook, 2. III. 1964, D. K. McAlpine (AMSA); 6 females, 1 male. Glen Innes, 17 II. 1974, M. S. Moulds (AMSA, ANIC); 1 female, same locality, 19. I. 1975, R. Gallagher (BRI); 2 females. Platypus Flat camp. Wild Cattle Creek State Forest, 7 IV. 1993, C. Reid (ANIC); 1 female, 1 male, Guyra, II. 1949, A. Dyce (ANIC); 2 females, Ulong, III. 1920, W. Heron (AMSA); 11 females, Brooklana, 1929, W. Heron (BMNH, AMSA, ANIC); 1 female, Coffs Harbour, 11. I. 1950, F. D. 140 Journal of Hymenoptera Research (MVMA); 11 females, Dorrigo, W. Heron (BMNH, MVMA, SAMAA, AMSA); 1 fe- male, same locality, 914 m, G. Heron (USNM); 2 females, same locality (QMBA, BMNH); 1 male, same locality, 914 m, 17. II. 1932, P. J. Darlington (USNM); 1 male, same locality, 12 II. 1968, D. H. Colless (ANIC); 1 female, same locality, 14. II. 1981, D. A. Doolan (AMSA); 1 female, same locality, 13. II. 1984, L. Masner (CNC); 4 females, 1 male. Deer Vale, 12- 13. I. 1931, 13. I. 1933, A. N. Burns (MVMA); 2 females, same locality, 30. I. 1972, G. Daniels (AMSA); 1 female, 4 males, Ebor, I. 1934, F. E. Wilson (MVMA, ANIC); 3 females, same locality, 12. XII. 1962, T. V. Bourke (BRI); 2 females, 3 males, Armidale, 5. II. 1915 (QMBA); 3 fe- males, Bellangry, 2. V. 1894, W. W. Frog- gatt (ANIC, BRI, MVMA, CNC); 4 fe- males. Hanging Rock, 7. I. 1955, K. M. Moore (FCNI); 1 female. Elands, 30. 1. 1928 (BRI); 1 female, Comboyne Scarp, near Upper Lansdowne, 6. IV. 1987, D. K. Mc Alpine, S. Day, R. de Keyzer (AMSA); 1 female. Dingo State Forest, 26-27. II. 1981, G. and T. Williams (AMSA); 1 fe- male. Bay's Hill, Taree, 28. III. 1992, G. Williams (AMSA); 9 females, 2 males, Tuncurry, 21 III. 1931, 15-25. III. 1932, no date, J. Parkes (AMSA, ANIC); 1 female. Upper Allyn River, 6. IV. 1958, R. Mackay (AMSA); 1 female. Upper Allyn, near Ec- cleston, 10 III. 1970, D. K. McAlpine, G. Holloway (AMSA); 1 female, 1 male, same locality, 26. II. 1970, D. K. McAlpine (AMSA); 1 female, same locality, 16. II. 1967, D. K. McAlpine (AMSA); 6 females, Maitland, 1892, W. W. Froggatt (ANIC, MVMA, USNM, BMNH); 1 female, Olney State Forest, 15. III. 1986, J. Grigg (UQIC); 5 females, Ourimbah, IV. 1904, S. W. Jack- son (AMSA); 1 female, Narara, 23 I. 1911 (AMSA); 1 female, same locality, 2. III. 1950, P. C. Hely (BRI); 2 males. Upper Colo, 10. III. 1990, G. R. Brown, M. A. Ter- ras (BRI); 4 females, 8 males, Mt Wilson, 4. I. 1931, A. N. Burns (MVMA); 1 female, same locality, 7 II. 1959, D. K. McAlpine (AMSA); 1 female, same locality, II. 1921 (AMSA); 1 male, same locality, 1067 m, I. 1932, P. J. Darlington (USNM); 1 female. Hartley Vale, 29. III. 1975, G. Daniels (AMSA); 3 females, Mt York, 21. III. 1964, D. K. McAlpine (AMSA, ANIC); 2 males, Katoomba, 26. 1. 1955, K. M. Moore (FCNI); 2 females, 1 male, same locality, 21. II. 1969, G. Hardy (AMSA); 2 females (with eggs), Woodford, 28. II. 1984, M. Hill (BRI); 1 female, 2 males, Beecroft, 18. II. 1968, O. M. Williams (BRI, ANIC); 2 fe- males, same locality, 18. III. 1967, C. E. Chadwick (BRI, ANIC); 5, females. Glen- brook, III. 1994, L. Turton (BRI, ANIC); 1 female, 2 males. Lane Cove, 27. IV. 1946, 30. III. 1947, 3. IV. 1948 (AMSA); 1 male, Sydney, 13. III. 1932, G. A. Waterhouse (MVMA); 1 female, same locality. III. 1949, A. Dyce (ANIC); 1 female, same locality, III. 1977, D. Clyne (ANIC). 1 male. Heath- cote, 20. III. 1952, L. Cascysand (BRI); 1 female, Heathcote National Park, 24. FV. 1994, Cowdrey (AMSA); 2 males, Mt Kei- ra, 23. II. 1983, G. A. Holloway (AMSA); 1 male, Jamberoo Mtn, 16. II. 1963, C. E. Chadwick (BRI); 1 male, Jamberoo, 11. 1. 1950 (AMSA). Unrealised: 1 female, 1 male, W. Heron (BMNH); 2 females, 1 male (QDPI); 1 female (MVMA); 3 fe- males, 1 male (UQIC); 1 female (QMBA); 1 female (ANIC). Distribution. — See Fig. 76. Several fe- males in the ANIC and the BMNH are la- belled "Brooklana Sydney." Since Brook- lana and Sydney are approximately 400 km apart, clearly mislabelling has oc- curred. Reliably labelled specimens indi- cate that P. macleaii occurs at both Brook- lana and Sydney. P. macleaii was not re- corded from Victoria by Bruzzese (1980) in a survey of insects on Rubus. Larval host plants.— Rosaceae: Rubus mol- uccanus L. (Froggatt 1893); Rubus rosifolius Smith. Discussion. — There is some variation in wing venation. In the forewing, Rl mav continue beyond the junction with Rs. In the hindwing cross-vein m-cu may be Volume 7, Number 2, 1998 141 48,49 ■57, 5S 5*)-61 . 50-56 Figs. 48-61. Philomastix xn]itlioplnihi\: 48, female antenna; 49, male antenna; 50, mesoscutellum, dorsal view; 51, mesoscutellum, surface contour as seen from rear of insect; 52, mebepisternum, profile; 5.3, lateral panel of first abdominal tergum; 54, apex of ovipositor sheath, dorsal view; 55, same, posteroventral view; 56, apex of male abdomen, dorsal view; 57, male genitalia, dorsal view; 58-61, lancet. Scale lines = 1.0 mm for 48-56; = 0.1 mm for 57-61. 142 Journal of Hymenoptera Research Figs. 62-66. Philomastix spp.: 62, P. madeaii, female, dorsal view; 63, P. madcaii, male, dorsal view; 64, P. nancarrowi, female, dorsal view; 65, P. nancarroun, male thorax (part), dorsal view; 66, P. nancarrowi, female mesepisternum, lateral view. Scale lines = 1.0 mm. present or absent. The mesoscutellar tu- bercles are usually strong (Figs. 30, 31) but may be weak (Fig. 32), absent on one side (Fig. 33) or absent altogether (Fig. 34). Cell IM is usually at least in part hyaline prox- imally, but in the female from Eungella cell IM is entirely dark and the infuscation extends to cell R. Philomastix xanthophylax Naumann and Groth, sp. n. (Figs. 1, 48-61, 67-74, 76) Female. — Body length 9.0-12.0 mm. Forewing length 9.4—11.5 mm. Distance between antennal sockets 2.2-2.6 times greater than diameter of anterior ocellus. Antenna (Figs. 48, 67) 14-16 segmented; Figs. 67-71. Philoinastix xautlwpln/liix: 67, female, apical antcnnomcres; b8, left mandible; 69, right mandible; 70, labium, maxilla; 71, lancets, ventral view. Scale line = 1.0 mm, for 68-70; see 48, 58 for scale to 67, 71. Volume 7, Number 2, 1998 Figs. 72-74. Pluliiiiiii-.li\ \iintlioplti/liix on Alphitonia excelsa; 72, two adult females, one guaiduii; fgg mass, other guarding batch ot first instar larvae; 73, batch of first instar larvae; 74, third instar larva. Adult sawflies in 72 each approximately 9.5 mm long; larvae in 73, 74 approximately 5 and 17 mm long respectively. apical 2-5 segments sometimes fused so that antenna apparently 12-13-segmented. First flagellar segment 2.9-3.0 times longer than wide. Second flagellar segment 1.8- 1.9 times wider apically than basally. Me- soscutellum without posterolateral tuber- cles (Fig. 50); anterodorsal surface more or less flat (Fig. 51). Mesepisternal tubercle weak (Fig. 52). Posterolateral margin of first abdominal tergum weakly curved (Fig. 53) or straight, spiracle separated from margin by a distance 1.2-1.3 times maximum diameter of spiracle. Ovipositor sheath posteromedially dentate (Figs. 54, 55). Lancet as in Figs. 58-61, 71, about 8 ctenidia terminating in dorsal tooth, antior teeth strong. Flagellum black (Fig. 1). Mid lobe of mesoscutum orange-yellow. Lateral lobe of mesoscutum orange-yellow, sometimes with small, brown macula less than half as long as lobe. Mesepisternum without brown band adjacent to sternopleural su- ture. Mesosternum brown, with orange- yellow band along anterior and lateral margins. Metascutum and cenchri orange- yellow. Forewing with narrow, transverse, brown band; cell IM not completely brown. Abdomen entirely orange-yellow. Male. — Body length 7.3-8.9 mm. Fore- 0 750 150O Kilometres ^^ Fig. 75. Distribution of Philomastiginae: woM distribution of Philomastix, Cerospastus and Ecopatus. 144 Journal of Hymenoptera Research 0 400 800 : Kilometres . 1 1 \ \ i\ Xj '--^_/ 0 Philomastix macleaii .: y^ — h 0 400 800 Kiiometres ■/ \ t 1 A '^^, . *i / "■-■J" '^v^ ^a* ■---. i 9 Philomastix nancarrowl y^ — '■ ^ Philomastix xanthophylax -'-g' Fig. 76. Australian distribution of Philomastix nancarroun, P. macleaii and P. xanthophylax. wing length 6.5-8.8 mm. Distance be- tween antennal sockets 1.3-1.7 times greater than diameter of anterior ocellus. Antenna (Fig. 49) 15-segmented. First fla- gellar segment 1.7-1.8 times longer than wide. Second flagellar segment 2.0-2.1 times wider apically than basally. Mesos- cutellum without posterolateral tubercles; dorsally slightly concave. Mesepisternal tubercle weak. Abdominal tergum 8 pos- teriorly with deep emargination (Fig. 56). Genitalia as in Fig. 57, paramere slender, gonolacinia strongly hooked, penis valve apically rounded. Clypeus and labrum lemon-yellow. Mid-lobe of mesoscutum orange-yellow. Lateral lobe of mesoscutum entirely brown to dark brown. Abdominal terga 5 and 6 without orange-yellow, lateral mac- ulae. Tergum 7 with lemon-yellow, lateral macula. Material examined. — Holotype female. 27.28S 151.56E, 10 km N Toowoomba, Queensland, 12. IV. 1992, H. Groth (ANIC). Paratypes: Queensland: 2 fe- males, same data as holotype (ANIC, HGCN); 8 females, same locality and col- lector as holotype, 3. IV. 1991, 1. III. 1992, 29. III. 1992, 16. III. 1992 (ANIC, BMNH, USNM, HGCN); 4 males, same locality and collector as holotype, but labelled "Highfields," reared from eggs collected V. 1992, emerged as adults 30. III. 1993, 22. IV. 1993, 24. IV. 1993, 26. IV. 1993 (ANIC, HGCN); 3 females, 4 males, Brookfield, 22. III. 1994, 29. III. 1994, 3. IV. 1994, J. Grigg (UQIC, ANIC); 2 females, Bellbird Park, Brisbane, IV. 1994, R. Nattrass (QMBA). Etymology. — The species name is de- rived from the Greek words xanthos, yel- low or golden, and phylax, a guard, with reference to the maternal guarding behav- iour common to all species of Philomastix. Distribution. — See Fig. 76. Volume 7, Number 2, 1998 145 Larval food plants. — Alpliitonia excelsa (Fenzl) Benth. (Rhamnaceae). Discussion. — There is conspicuous vari- ation in wing venation within the type se- ries. In the forewing: (1) Rl sometimes continues as a short spur beyond the junc- tion of Rl and Rs; (2) there may be one, two or no cross-veins between C and R; (3) there may be an incomplete cross-vein distal to 3r-m; and (4) a diagonal vein sometimes defines a small, triangular cell in the anterobasal comer of cell 3M. In the hindwing cross-vein m-cu is rather vari- able: (1) it may be present or absent; (2) it may curve smoothly into CuA, in which case there is no distal abscissa of CuA; (3) it may join M before or after the junction of cross-vein 2r-m and M. CEROSPASTUS Konow Cerospastus Konow 1899: 404--i05; Konow 1905: 36-37; Rohwer 1911: 76. Benson 1935: 224. Benson 1938: 379; Pagliano and Scaramozzi- no 1990: 58; Smith 1990: 21-23; Abe and Smith 1991: 18. Type species: Cerospastus vol- upis Konow (by monotypy). Ceratospastus: Schuiz 1906: 84 (unjustified emen- dation). Female. — Vertex conspicuously setose. Face with some fine microsculpture. Malar space narrower than diameter of anterior ocellus. Antenna (Fig. 3) 14-(-20, Smith 1990) segmented, weakly serrate, weakly clavate. Right mandible simple (Smith 1990). Maxillary palp (Fig. 9) 6-segmented, filiform, without sensory cup. Labial palp 4-segmented, without sensory cup. Labi- um tri-lobed. Thorax dorsally consci- cuously setose. Notauli, median mesoscu- tal line deeply impressed. Mesoscutellum not swollen, posterior margin visible from above. Mesepisternum without tubercle. Metascutellum not band like. Forewing (Fig. 4) with closed radial cell and 4 cu- bital cells; median and second cubital cells each with nygma. Abdominal terga not conspicuously setose. Second tergum pre- dominantly fine transversely striate. Cer- cus present. Ovipositor sheath not strong- ly expanded posteriorly. Male. — Antenna 20-segmented (19-21 according to Smith 1990), strongly serrate. Median cell of forewing without nygma. Eighth tergum posteriorly with very deep, broad emargination (Fig. 7). Cerospastus volupis Konow (Figs. 3, 4, 7, 9, 76) Cerospastus volupis Konow 1899: 404; Konow 1905: 37; Rohwer 1911: 76; Smith 1978: 160; Oehike and Wudowenz 1984: 419; Pagliano and Scaramozzino 1990: 58; Smith 1990: 22- 23; Smith 1993: 11. Female. — Body length 10.0-11.5 mm. Forewing length 11.0-12.0 mm. Distance between antennal sockets 1.2-1.7 times greater than diameter of anterior ocellus. First flagellar segment 2.5 times longer than wide. Second flagellar segment 1.7- 2.0 times wider apically than basally. Me- soscutellum without posterolateral tuber- cles, posterior margin convex; dorsal sur- face weakly convex. Mesepisternum with- out tubercles. Posterolateral margin of first abdominal tergum smoothly curved or slightly angulate, spiracle separated from margin by a distance 0.5-1.3 times maximum diameter of spiracle. Ovipositor sheath not posteromedially dentate. Lan- cet as in Smith (1990: Fig. 34). Flagellum very pale brown, scape and pedicel orange-yellow. Either, head pre- dominantly orange-yellow; upper frons with transverse brown joining upper ex- tremities of compound eyes and encom- passing ocellar triangle; or most of frons, lowermost gena, vertex medially brown. Mandibles brown, remaining mouthparts orange-yellow. Thorax and legs predomi- nantly orange-yellow to cream. Mesono- tum orange-yellow, with brown maculae occupying either most of length of mid and lateral lobes or only anterior half of lateral lobe. Cenchri, metascutum, most of mid and hind tarsi, first abdominal ter- gum and sometimes also ventral surfaces of thorax brown to pale brown. Remain- 146 Journal of Hymenoptera Research der of abdomen orange to yellow. Wings hyaline with faint yellow Hnge. Male. — Body length 7.5 mm. Forewing length 7.3 mm. distance between antennal sockets 0.7-0.8 times greater than diame- ter of anterior ocellus. First flagellar seg- ment 1.3 times longer than wide. Second flagellar segment 2.4 times wider apically than basally. Mesoscutellum, mesepister- num as in female. Genitalia as in Smith (1990: Fig. 31). Head, thorax predominantly black to dark brown. Antenna pale brown. Clype- us orange-brown. Mandibles red-brown. Maxilla, labium orange-yellow. Spiracular lobe of pronotum orange-yellow. Legs or- ange to yellow. Abdomen dorsally dark brown to brown, ventrally orange to yel- low. Wings hyaline with faint brown tinge. Type. — Lectotype female (designated by Smith 1990), Valdivia, Chile, 1897, Loss- berg (DEIC; examined). Other material examined. — 1 female, 1 male, Parque Nac. Conguillio. Province Temuco, Chile, 31. XII. 1976, O. Puentes (USNM). Distribution. — Known only from a few specimens from central western Argentina (Neuquen Province) and central Chile (Cautin, Malleco, Maule and Valdivia). Lariml liost plant. —Nothofagus sp. (Faga- ceae) (Smith 1990). ECOPATUS Smith Ecopatiis Smith 1990: 23-24. Type species: Eco- patus penai Smith (by original designation and monotypy). Female. — Vertex almost bare. Face al- most completely smooth. Malar space very slightly wider than diameter of an- terior ocellus. Antenna (Fig. 2) 18-seg- mented (20-21 according to Smith 1990), filiform, not clavate. Right mandible with 2 inner teeth. Maxillary palp (Fig. 8) 6-seg- mented, 4th segment apically distinctly wider than other segments; apical seg- ment without sensory cup. Labial palp 4- segmented, apical segment without sen- sory cup. Labium tri-lobed. Thorax dor- sally almost bare of setae. Notauli, median mesoscutal line shallow. Mesoscutellum not swollen, posterior margin visible from above. Mesopleuron without tubercle. Me- tascutellum not band like. Forewing (Fig. 5) with open radial cell and 2-3 cubital cells; median and second cubital cells each with nygma. Abdominal terga dorsally not conspicuously setose. Second tergum with faint, reticulate microsculpture. Cer- cus present. Ovipositor sheath not strong- ly expanded posteriorly. Male. — Unknown. Ecopatiis penai Smith (Figs. 2, 5, 6, 8) Ecopatiis penai Smith 1990: 24-25. Female. — Body length 6.0-7.0 mm. fore- wing length 7.7 mm. distance between an- tennal sockets 2.3-2.4 times greater than diameter of anterior ocellus, first flagellar segment 3.6-3.7 times longer than wide. Second flagellar segment 1.1-1.2 times wider apically than basally. Mesoscutel- lum without tubercles; anterodorrsal sur- face convex. Mesepisternum without tu- bercle. Posterolateral margin of first ab- dominal tergum sinuate (Fig. 6), spiracle separated from margin by distance 4.6 times maximum diameter of spiracle. Ovi- positor sheath not medially dentate. Lan- cet as in Smith (1990: Fig. 39). Head, thorax, abdomen predominantly dark brown to black. Following orange, orange-yellow to cream: scape or scape and pedicel, frons just above and below antennal sockets, clypeus, mandible (ex- cluding red-brown teeth), maxilla, labium, posterior margin of pronotum, tegula, ax- illar sclerites, large macula anterodorsal to extremity of mesepisternum, legs (except pale brown extremities of tarsi), abdomen ventrally. Wings uniformly strongly brown tinged. Ti/f^es. — Holotype female, Caramavida, Volume 7, Number 2, 1998 147 Province Arauco, 5-10. II. 1953, L. Pena (USNM; not seen). Paratypes: 2 females, same data as holotype (not seen); 1 female, Curacautin, Rio Blanco, 27-31. I. 1950, L. Pefia (USNM) (examined). Distribution. — Known only from a few specimens from central Chile (Araucan and Curacautin Provinces). Larval host plants. — Unknown. ACKNOWLEDGEMENTS I thank Mr M. Moulds (AMSA), Mr P. Gillespie and Mr J.Macdonald (BCRl), Mr D. G. Notfon (BMNH), Dr H. Goulet (CNC), Drs A. Taeger and S. M. Blank (DEIC), Ms C. Urquhart (FCNI), Dr K. Walker (MVMA), Dr C. O'Toole (UMO), Dr C. Burwell (QMBA), Mr J. Donaldson and Dr M. Elson-Harris (QDPI), Ms J. Forrest (SAMAA), Mr G. Daniels (UQIC) and Dr D. Smith (USNM) for the loan of ma- terial. Mr N. Springate (c/ -Natural History Museum, London, U.K.), Dr Goulet and Mr Macdonald kindly commented on an earlier draft of this paper. Botanical names were checked by Dr L.S. Springate (Royal Bo- tanic Garden, Edinburgh, U.K.). LITERATURE CITED Anonymous. 1980. Insect pest occurrences in Tas- mania, 1978/79. Tasmanian Department of Agri- culture. Insect Pest Suroei/ 12: 1-39. Abe, M. and D. R. Smith. 1991. The genus-group names of Symphyta (Hymenoptera) and their type species. Esakia 31: 1-115. Ashmead. W. H. 1898. Classification of the horntails and sawflies, or the sub-order Phytophaga. Ca- nadian Entonwhgist 30: 225-232. Benson, R. B. 1935. New Australian sawflies. Memoirs of the Queensland Museum 10: 211-229. Benson, R. B. 1938. On the classification of sawflies (Hymenoptera Symphyta). Transactions of the Royal Entomological Society of London 87: 353-384. Bruzzese, E. 1980. The phytophagous insect fauna of Ruhus spp. (Rosaceae) in Victoria, a study on the biological control of blackberry ( Ruhus fructico- sus L. agg.). Journal of the Australian Entoniotogwal Societ]/ 19: 1-6. Dalla Torre, C. G. 1894. Tenthredinidae incl. Urocer- idae (Phyllophaga cSc Xylophaga). Catalogiis Hy- menoftterorum 1:1-459. Forsius, R. 1927. On some sawflies from the Austra- lian region (Hymenoptera Tenthredinidae). Rec- ords of the South Australian Museum 3: 283-308. Francis, W. D. and G. M. Chippendale 1970. Austra- tian Raiu'Forest Trees. Australian Government Publishing Service: Canberra, xvi, 468 pp. Froggatt, W. W. 1890a. Descriptions of a new genus and two new species of Tenthredinidae. Proceed- ings of the Lmnean Society of Neiv South Wales 5: 487^90. Froggatt, W. W. 1890b. Catalogue of the described Hymenoptera of Australia. Proceedijigs of the Lm- nean Society of Nt-ic South Wales 5: 689-762. Froggatt, W. W. 1892. (Notes and exhibits]. Proceed- ings of the Lmnean Society of New South Wales 7: 201. Froggatt, W. W. 1901. The pear and cherry slug (Er- iocampa limacina, Retz), generally known as Selan- dria cerasi, with notes on Australian sawflies. Ag- ricultural Gazette of Neiv South Wales 12: 1063- 1073. Froggatt, W. W. 1907. Australian Insects. Brooks: Syd- ney, i-xiv, 1^149. Froggatt, W. W. 1918. Notes on Australian sawflies (Tenthredinidae). Proceedings of the Linnean Soci- ety of New South Wales 43: 668-672. Goulet, H. and J. T. Huber (1993). Hymenoptera of the World. Agriculture Canada: Ottawa, viii, 668 pp. lackson, R. 1993. Insect herbivory. Program and Ab- stracts Australian Entomological Society Conference and 24th A. G. M.: 46. Johnson, N. F. 1992. Catalogue of world species of Proctotrupoidea, exclusive of Platygastridae (Hy- menoptera). Memoirs of the American Entomologi- cal Institute 51: 1-825. Kikkawa, J., G. B. Monteith and G. Ingram. 1981. Cape York Peninsula: major region of faunal in- terchange, pp. 1695-1742. In A. Keast (ed.) Eco- logical Biogeography of Australia. W. Junk: The Hague. Kirby, W. F. 1882. List of Hymenoptera loith Descriptions and Figures of the typical Specimens in the British Museum. Vol.1. British Museum: London, xxviiii, 450 pp., 16 plates. Konow, F. W. 1898. Uber die Tenthrediniden-tribus Lophyrini. Entomologische Nachrichten. Berlin 30: 248-249. Konow, F. W. 1899. Neue Siidamerikanische Tenthre- dinidae. Anales del Museo Nacional de Buenos Aires 6: 397-417. Konow, F. W. 1905a. Drei neue Syzygoniides aus Australien. Zeilschrift fiir Systematische Hymenop- terologie und Dipterologie 5: 166-169. Konow, F. W. 1905b. Hymenoptera Fam. Tenthredi- nidae. Genera Insectorum 29: 1-176. Kudo, S., K. Maeto and K. Ozaki 1992. Maternal care in the red-headed spruce web-spinning sawtly, Cephalcia isshikii (Hymenoptera: Pamphiliidae). Journal of Insect Behavior 5: 783-795. Lazarides, M. and B. Hince 1993. CSIRO Handbook of Economic Plants of Australia. CSIRO: Melbourne, xiv, 330 pp. Leask, M. F. 1943. Records of two sawflies. The North Queensland Naturalist 11 (69): 2. Leask, M. F. 1944. Habits of sawflies (Philomastix 148 Journal of Hymenoptera Research nancarrowi). Hymenoptera, Tenthredinidae. The North Queensland Nahmilist 11(72): 1-3. Macdonald, J. and C. P. Ohmart. 1993.Life history strategies of Australian pergid sawflies and their interactions with host plants, pp. 485-502. In M. R. Wagner and K. F. Raffa (eds) Snvflij Life His- tory Adaptations to Woody Plants. Academic Press: San Diego, xx, 581 pp. Monteith, G. B. 1996. Recent exciting insect discov- eries. Queensland Museum Association Neiosletter 11(2): 3. Moore, B. P. 1978. A new Australian stag beetle (Co- leoptera: Lucanidae) with Neotropical affinities. Journal of the Australian Entomological Society 17: 99-103. Morice, F. D. 1919. Notes on Australian sawflies, es- pecially the "Authors' Types" and other speci- mens in the British Museum of Natural History and the Hope Collections of the Oxford Univer- sity Museum; with diagnostic synopses of the genera and species, and photographs illustrating their structural characters. Transactions of the En- tomological Society of London 66: 247-333. Naumann, I. D. 1984. An apterous female sawfly (Hy- menoptera: Symphyta) from Papua New Guinea. Systematic Entomology 9: 339-349. Naumann, I. D. 1991. Hymenoptera. pp.916-1000. /;; CSIRO (ed.) The Insects of Australia, 2nd edition. Melbourne University Press: Carlton. Naumann, I. D. 1993. CSIRO Handbook of Australian Insect Names. CSIRO: East Melbourne. 200 pp. Naumann, 1. D. 1994. Systematic and Applied Entomol- ogy. Melbourne University Press: Carlton. 484 PP- Naumann, 1., J. C. Cardale, R. W. Taylor and J. Mac- donald. 1994. Type specimens of Australian Hy- menoptera (Insecta) transferred from the Ma- cleay Museum, University of Sydney, to the Aus- tralian National Insect Collection, Canberra. Pro- ceedings of the Linnean society of New South Wales UA: 69-72. Nix, H. 1991. Biogeography: pattern and process. Ko- warri 1: 11-39. Oehlke, J. and J. Wudowenz. 1984. Katalog der in den Sammlungen der Abteilung Taxonomie der In- sekten des Institutes fur Pflanzenschutzfor- schung, Bereich Eberswalde (ehemals Deutsches Entomologisches Institut), aufbewahrten typen- XXll (Hymenoptera: Symphyta). Beitriige zur En- tomologic 34: 363-420. Pagliano, G. and P. Scaramozzino. 1990. Elenco dei Generi di Hymenoptera del niondo. Memorie del- la Societa Entomologica Italiana 122: 3-210. Riek, E. F. 1970a. Hymenoptera. pp.867-959.//i CSIRO (ed.) The Insects of Australia, 1st edition. Mel- bourne University Press: Carlton. Riek, E. F. 1970b. A redefinition of the subfamily Phy- lacteophaginae with description of a new genus and species (Hymenoptera: Symphyta: Pergi- dae). Journal of the Australian entomological Society 9: 215-218. Rohwer, S. A. 1911a. A classification of the suborder Chalastrogastra of the Hymenoptera. Proceedings of the Entomological Society of Washington 13: 215— 226. Rohwer, S. A. 1911b. II. The genotypes of the sawflies and wood wasps, or the superfamily Tenthredi- noidea. Technical Series. Bureau of Entomology, United States Department of Agruidture 20:69-109. Schuiz, W. A. 1906. Spolia Hymenopterologica. Pader- born. 355pp. Smith, D. R. 1978. Suborder Symphyta (Xyelidae, Par- archexyelidae, Paramphiliidae, Xyelydidae, Kar- ativitidae, Gigasiricidae, Sepulcidae, Pseudosiri- cidae, Anaxyelidae, Siricidae, Xiphydriidae, Pa- roryssidae, Xyelotomidae, Blasticotomidae, Per- gidae). Hymenopterorum Catalogus 14: i-iii, 1-193. Smith, D. R. 1980. Pergidae (Hymenoptera) from New Guinea and Australia in the Bishop Muse- um. Pacific Insects 22: 329-346. Smith, D. R. 1988. A synopsis of the sawflies (Hy- menoptera: Symphyta) of America south of the United States: introduction, Xyelidae, Pamphili- idae, Cimbicidae, Diprionidae, Xiphydriidae, Sir- icidae, Orussidae, Cephidae. Systematic Entomol- ogy 13: 205-261. Smith, D. R. 1990. A synopsis of the sawflies (Hy- menoptera, Symphyta) of America south of the United States: Pergidae. Rei'ista Brasdeira de En- tomologia 34: 7-200. Smith, D. R. 1992. A synopsis of the sawflies (Hy- menoptera: Symphyta) of America south of the United States: Argidae. Memoirs of the American Entomological Society 39: 1-201. Smith, D. R. 1993. Systematics, life history, and dis- tribution of sawflies.pp. 3-32. In M. R. Wagner and K. F. Raffa (eds) Sawfly Life History Adaptions to Woody Plants. Academic Press: San Diego, xx, 581 pp. Smith, D. R. 1995. A new species of Xiphydriidae (Hymenoptera) from Chile. Revista Chilena de En- tomologui 22: 21-24. Tillyard, R. J. 1926. The Insects of Australia and Nezo Zealand. Angus and Robertson: Sydney, xvi, 560 pp. Watson, J. A. L. and G. Theischinger. 1984. Regions of taxonomie disjunction in Australian Odonata and other freshwater insects. Odonatologua 13: 147-157. Westwood, J. O. 1880. A monograph of the sawflies composing the Australian genus Perga of Leach. Proceedings of the Zoological Society of London :359- 379. J. HYM. RES. Vol. 7(2), 1998, pp. 149-156 A New Species of Ibalia from Borneo, with a Revised Phylogeny and Historical Biogeography of Ibaliidae (Hymenoptera: Cynipoidea)^ Zhiwei Liu Department of Entomology, Swedish University of Agricultural Sciences, P.O. Box 7044, S-75007, Uppsala, Sweden Abstract. — Ibalia kalimantanica Liu from eastern Kalimantan, Borneo, is described here as a new species. Reanalysis of a previously published character matrix of the family Ibaliidae with the new species included shows that /. kalimantanica belongs to the subgenus Tremibalia and is the sister species to the clade of ((I.mirabilis, I. japonica), I. hunanica). Biogeographical analysis of the expanded data set strengthened support for an earlier hypothesis concerning the historical bio- geography of the Ibaliidae postulating early radiation of the family in the eastern Palaearctic — Oriental region. The separation of the 7. kalimatanica clade is suggested to have been caused by changes of land area configuration in Southeast Asia as a result of global sea level changes during late Oligocene to early Miocene. The Ibaliidae constitute a small family of parasitic cynipoids comprising nineteen known species, of which all but one are restricted in the Northern Hemisphere. The species belong to the three genera Ei- leenella Fergusson, Heteribalia Sakagami and Ibalia Latreille. They parasitize siricid woodwasps, both in conifers and hard- woods, and some species of Ibalia have been used in the biological control of sir- icid pests in conifer plantations. Ibaliidae is of interest owing to its near-basal phy- logenetic position within the superfamily Cynipoidea (Ronquist 1995). Recently, Liu & Nordlander (1992, 1994) studied the North-American species of the Ibaliidae and presented a review of the world spe- cies of the family, and Nordlander et al. (1996) studied their phylogeny and histor- ical biogeography. In this study, a new species of Ibalia Latreille is described from ' This paper appeared in Acta Uinzrrsitntis Suieciae Sihestria 62(1998) but was not intended for perma- nent scientific record as stated in the Disclaimer pub- lished on page 6 of that publication. This Disclaimer satisfies Chapter III, Article 8, section b, of the ICZN. (Editor) the tropical rain forests of eastern Kali- mantan, Borneo, Indonesia. The character matrix of Nordlander et al. (1996) has been reanalyzed in order to determine the phy- logenetic position of the new species in re- lation to other Ibalia species, and to inves- tigate whether the topology of the phylo- genetic tree of the genus would be thus affected. Species of Ibalia have previously been described only from the Northern Hemi- sphere; the new species represents the first tropical species of the genus. The only oth- er tropical ibaliid species, Eileenella cather- inae, is from New Guinea, and is the sister species to all other ibaliids (Fig. 2). There- fore, the phylogenetic position of the new Ibalia species will provide new evidence for testing the previous biogeographical scenario of Nordlander et al. (1996). MATERIALS AND METHODS Terminology used in this article follows that of Ronquist and Nordlander (1989) and Liu and Nordlander (1994). Only a single female of the new species was available. The character-state coding of the new species was made in compari- 150 Journal of Hymenoptera Research Table 1. Character states tor Ibcjliii knlimatitiDiicn. Characters and character states are coded as in Nordlander et al. (1996), with the following addition of character state: Character 55. Sculpture of mesopleural speculum: (2) distinctly vertically costate with secondary irregular foveolate sculpture. The table is supplementary to the previously published matrix including all the other species. character /. kalimantanka 11??? 11100 10100 01-00 0-01 1 00-10 00120 010-1 son with representatives of the genus Het- erihalia Sakagami and each of the two sub- genera of Ibalia, as well as of the out- groups Liopteridae and Eileenella Fergus- son. The characters and character coding were the same as in the previous cladistic analysis of the Ibaliidae (Nordlander et al. 1996). For characters 31, 34, 35, and 69, a polymorphic condition was coded as a separate, intermediate state and each step was given the weight 0.5, so that a change between non-polymorphic states would count as one step instead of two. Of the multi-state characters, characters 6, 45, 52, 63, 66, 67, and 80 were unordered; the oth- ers were ordered in the sequence 012. The only change in relation to the previous study was that an autapomorphy for /. kal- immitauica required additional state for character 55 (see Table 1). Methods used for phylogenetic and bio- geographical analyses were in general as described in Nordlander et al. (1996), with some minor modifications. As in Nordlan- der et al. (1996), PAUP 3.1 were used for Fig. 1. Ilnilin kalimantaiucii, new species. phylogenetic analysis, but less extensive options were adopted when calculating support indexes for the branches. Options for bootstrapping included heuristic search, random addition sequence, 1000 replications, and for each replication tree search options are simple addition se- quence and tree bisection reconnection (TBR) swapping. The decay index (or Bre- mer support), the number of extra steps needed to break up the group, were ob- tained using branch and bound search and simple addition sequence. For biogeo- graphical analysis, the previously defined distribution area Eastern Palaearctic -I- North-east Oriental was extended to in- clude oceanic Southeast Asia, and defined as Eastern Palaearctic -I- Eastern Oriental. DIVA 1.1 was used in the present study for historical reconstruction (Ronquist 1996, 1997). It is basically the same as the earlier version (DIVA 1.0) used by Nordlander et al. (1996), but with some performance im- provements (Ronquist 1996) that should not affect comparison of the results. Ibalia (Tremibalia) kalimantanica Liu, new species (Fig. 1) Female.— Body length 10.0 mm. Colorn- tion: Head yellow except eyes, upper face, and a narrow longitudinal median strip through lower face, which are black. Fla- gellomeres 4—10 of female antenna white to pale, contrasting with the remaining darker antennal segments. Pronotum pale yellow with anterior plate of pronotum medially, dorsal pronotal area entirely, and lateral pronotal area posteriorly black. Volume 7, Number 2, 1998 Table 1. Extended. 151 41 4h Tl Sh hi 66 71 7h SI 0-012 01111 00102 10100 0-011 11001 ?100? -- Mesothorax mainly black, a small area yellow with dorsal parts of coxa, trochan- posteriorly on mesoscutum, mesopleural ter, and femur yellow to brown. Metaso- triangle entirely, and a broad transverse ma pale yellow with three narrow, trans- band across middle of scutellum yellow, verse to oblique, dark brown strips (Fig 1). Metathorax and propodeum black. Legs Head. — Vertex rather weakly longitudi- X 100/10 95/2 0 2-5 1 X 99/10 10-23 97/5 86/2 1 2-6 73/1 1 1-2 1 2-2.5 5-9.5 64/2 93/4.5 4-4.5 5-6 76/2 3 2-4.5 81/3 1-2 0-/ 78/2 3-5 2-4 10-25 2 7-12 55/2 2-3.5 5-6 52/2 3.5 1 2-13.5 2-7,5 55/1 .5 1-2 1.5-4 48/1 1-2 ^-o 58/1 2 1-3 Liopteridae catherinae Eileenella confluens t nishijimai divergens Heteribalia subtilis aureopilosa , jakowlewi anceps ornata Ibalia mirabilis (Tremibalia) hunanica japonica ^ aprilina rufipes montana kirki Ibalia (ibalia) anzonica ruficollis leucospoides Fig. 2. Shortest tree of interspecific relationships in the Ibaliidae according to the previous analysis by Nord- lander et al. (1996). Figures above and below branches as in Fig. 3. 152 Journal of Hymenoptera Research nally carinate and with rather dense, ad- pressed pubescence. Upper face complete- ly rugose; antennal scrobes indistinctly de- limited by a lateral carina and not distinct- ly depressed. Gena largely glabrate with shallow foveae, postero-ventrally distinct- ly costate. Eye length about 3.1 times length of malar space. Antenna. — Female with 11 flagellomeres; 2nd flagellomere distinctly longer than 1st (F2/F1 = 4/3). Mesoso7na. — Pronotal crest without me- dial incision. Pronotum costate almost en- tirely, covered with dense, adpressed pu- bescence. Propleuron protruding strongly ventrally. Scutellar foveae rugose, separat- ed by median carina. Posterior processes of scutellum raised weakly posteriorly. Distance between outer sides of posterior scutellar processes about 0.7 times maxi- mum width of scutellum. Femoral groove of mesopleuron almost smooth with faint longitudinal carination ventrally; specu- lum vertically striate with secondary ir- regular foveolate sculpture. Metepister- num with vertical costulae. Anterolateral propodeal process distinct; posterior pro- podeal process low; lateral propodeal ca- rina not elevated medially. Wings. — Forewing subhyaline with wide dark strip along outer margin and distinct, narrow, dark band between Rs-I-M and Cul (behind submarginal cell) along outer side of M. Areolet present and very small. Hindwing faintly fuscous along outer margin; with three hamuli. Legs. — Anterior lateral crest of metacoxa rounded and low. Anterior mesotibial spur present. Anterior apical process of 2nd metatarsomere reaching to middle of 4th tarsomere. Metasotna. — Metasoma as long as head and mesosoma combined. Tergum 8 with sparse hairs. Male. — Unknown. B iology. — Unknown . Distribution. — Indonesia: E. Kalimantan, Borneo. Type material. — Holotype, 9, INDONE- SIA: Eastern Kalimantan, Kayan-Menta- rang Natural Reserve, WWF Station, Low- land Dipterocarpus Forest (2°52'N, 115° 49'E), Malaise Trap head, iii.1993 (D.C. Darling and U. Rosichon) (Museum Zool- ogi Bogor, Indonesia). For a comprehensive comparison with other species of the family Ibaliidae, the character matrix of Nordlander et al. (1996) should be consulted. For identifi- cation, /. kali7nnnta)iicn may be keyed out by introducing an additional couplet 3a following the second item of couplet 2 in Liu & Nordlander's (1994) key to the world species of Ibalia: 3a. Female antenna contrastingly colored with flagellomeres 4-10 white and the remaining segments dark. Upper face lacks regular sculpture. Speculum vertically striate. Posterior processes of scutellum posteriorly only weakly raised . . . /. kalimatttanica Liu, new species - Female antenna evenly light yellow to dark brown throughout or becoming darker to- ward apex, but never with contrasting colors. Upper face longitudinally or transversely carinate at least in antennal scrobes. Speculum finely striolate longitudinally. Posterior processes of scutellum raised distinctly posteriorly 3 REVISED PHYLOGENY AND HISTORICAL BIOGEOGRAPHY OF IBALIIDAE The sum of minimum and maximum possible lengths over all characters in the data matrix of Nordlander et al. (1996, Ta- ble 1) with the data of /. kalimantanica add- ed (Table 1) was 99 and 414 respectively. Parsimony analysis using the branch-and- bound algorithm of PAUP resulted in one optimal tree of length 149 (CI - 0.67, RI = 0.84). Compared with Nordlander et Volume 7, Number 2, 1998 153 X 95/2 G 2-5 1 00/1 0 1 X 10-20 86/2 73/1 1 .-0| 1 1-2 1 94/5 90/4 4 99/10 5-7 4-5 4-6 57/2 2-3 75/4 3-5 53/2 6 2-4 57/2 39/1 2-3 10-22 4-b 1-9 1-3 3-4 8-10 54/2 2-4 5-6 61/3 4 11-13 3-6 57/1,5 1-2 2-4 47/1 55/1 1-2 O Q 2 1-3 Liopteridae catherinae confluens nishijimai divergens subtilis aureopilosa ^ jakowlewi ornata anceps kalimantanica hunanica japonica mirabilis aprilina rufipes montana kirki arizonica ruficollis leucospoides Eileenella Heteribalia Ibalia (Tremibalia) Ibalia (Ibalia) Fig. 3. Shortest tree of interspecific relationships in the Ibaliidae, obtained with the hranch-and-bound al- gorithm of PAUP (length = 149, CI = 0.67, RI = 0.84). Shown above each branch in the tree is the support for the corresponding clade, measured as the percentage with which the clade appeared among the shortest trees in 1000 bootstrap replications of the analysis, followed by the decay index (or Bremer support). Below each branch are the minimum and maximum number of character changes along that branch. Clades marked 'x' were constrained to be monophyletic according to results of Ronquist (1995). al.'s previous optimal tree (length = 141, CI = 0.67, RI = 0.85), the new tree is al- most identical in terms of fitness. Com- pared with the earlier phylogeny of Ibali- idae (Fig. 2) as presented by Nordlander et al. (1996), the topology of the phyloge- netic tree remained unchanged with the addition of /. kalimantanica, except for two local changes within the Tremibalia clade. The first change concerns the relative re- lationship within the clade (/. mirabilis Yasumatsu, /. japonica Matsumura, /. hu- nanica Liu & Nordlander), and the other concerns the relative positions of /. anceps Say and /. ornata Belizin (Figs. 3, 4). The monophyly of Ibalia (Ibalia) became better supported after the inclusion of /. kaliman- tanica in the analysis, the bootstrap value 154 Journal of Hymenoptera Research I I I I I 1 3 8 12 71 Legend: — 0^1 ESS 2^1 ^ l-»0 ^ 0-»2 I I H — jakowlewi 20 35 45 52 1+1-1- 49 64 65 66 I I I I — ornata 57 60 63 73 +++ 25 53 72 35 69 4H- 24 70 anceps 1 [] Q I I g — kalimantanica 2 12 32 40 47 55 ++++ 30 4143 57 -S — hunanica 31 66 20 mirabilis 19 25 72 japonica 27 Fig. 4. The clade of Ibalia (Treniibnlia) of the shortest tree with all character changes that could be mapped unambiguously. increased from 64% to 72% and Bremer support from 2 to 4. As a result of the changes in tree topol- ogy, the previously suggested Eocene-Oli- gocene separation of 7. anceps (ca 33-34 MY A) in the subgenus Tremibalia is now one node further from the base of the tree (Nordlander et al. 1996). The origin of the Ibaliidae, based on node /branch distance calculation, is now estimated to be about 160 MY A, a negligible difference from the previous estimate of 150 MYA with regard to potential calculation errors. The esti- mated time for the origin of the Ibaliidae is still the Late Jurassic. An exact search of the updated distri- bution matrix using DIVA 1.1 resulted in one single reconstruction of the distribu- tion history requiring nine dispersals (Fig. 5). The present reconstruction postulates a center of origin for Heteribalia and Ibalia in the Eastern Palaearctic — North Oriental region, and is in complete accordance with that preferred by Nordlander et al. (1996, Fig. 7). The separation of the /. kaliniaiitaiiica clade probably resulted from a dispersal within the Oriental as early as in the end of Oligocene (29-24 MYA) and its subse- quent isolation from its sister clade. The Volume 7, Number 2, 1998 155 Legend: A = West Nearctic B = East Nearctic C ~ West Polaearctic D — East Polaearctic and East Oriental E = New Guinea E. catherinae H, subtilis H. aureopilosa H, divergens H, nishijimai H. confluens T.) mirabilis T.) japonica T.) hunanica T.) kalimantanica J.) anceps T.) ornata [T.) jakowlewi .) aprilina .) rufipes .) montana .) Wrki .) arizonica .) ruficollis ,) leucospoides Fig. 5. Reconstruction of ancestral distribution of Ibaliidae using DIVA 1.1 resulted in one exact solution that requires nine dispersals. Dispersal events are indicated on the branches, and implied between-area vi- cariance events are indicated by hyphens in the ancestral distributions. land area configuration of the Southeast mained high from Palaeocene through Asia has varied greatly since late Oligo- Oligocene (65-30 MYA). By late Oligocene cene as a result of global sea level changes (29 MYA) there occurred a spectacular fall (Heaney 1991). The global sea levels re- in sea level to about 250 m below the pres- 156 Journal of Hymenoptera Research ent level, and it then recovered to present level by end of Oligocene (24 MYA). From then onwards the sea levels progressively rose, with minor drops, to about 220 m above the present level in the middle Mio- cene (13 MYA). This was followed by sev- eral cycles of fluctuating sea levels (Hutch- ison 1989). During times of low sea levels, Sumatra, Java and Borneo were part of a peninsula projecting south from continen- tal Asia (often referred to as Sundaland) (Morley & Flenley 1987, Heaney 1991), fa- cilitating the dispersal of the stem species of /. kalimatanica and its sister species from the continental Asia to Borneo, and the continuous ancestral distribution was sub- sequently split when sea level rose again. Although, any of these sea level cycles could have been responsible for the spe- ciation of 7. kalimantanica, the many auta- pomorphies of /. kalimantanica and the rather many synapomorphies for its sister group indicates that the event probably occurred rather early. Using the same dat- ing method as in Nordlander et al. (1996), the vicariant event separating I. kaliman- tanica from its sister species was estimated as having occurred at about 21 MYA. This is in general accordance with the sea level recovery since late Oligocene from the late Oligocene dramatic drop, which could have facilitated the dispersal of the ances- tral species. ACKNOWLEDGMENTS I thank Goran Nordlander and Fredrik Ronquist for valuable comments, Chris Darling for presenting me with this pleasant specimen, and Rune Axelsson for photographic assistance. The study was support- ed by a scholarship from the Oscar and Lili Lamm's Memorial Foundation and by a grant from the Swed- ish Natural Science Research Council to F. Ronquist and G. Nordlander. LITERATURE CITED Heaney, L. 1991. A synopsis of climatic and vegeta- tional change in Southeast Asia. In Schneider, S. H (ed.) Cliniiitic dw)i\;c 19(2-2), special i^iuc: Tropj- iail foresti ami climate. Kluwer Academic Publish- ers, Dordrecht, pp 53-61. Hutchison, C. S. 1989. Geological evolution of South- East Asia. Oxford Monographs on Geology and Geo- pliisics, No.l3. Clarendon Press, Oxford. Liu, Z. and G. Nordlander. 1992. Ibaliid parasitoids of siricid woodwasps in North America: two new Ibalia species and a key to species (Hyme- noptera: Cynipoidea). Proceedings of the entomo- logical Societi/ of Washington 94: 500^507. Liu, Z. and G. Nordlander. 1994. Review of the family Ibaliidae (Hymenoptera: Cynipoidea) with keys to genera and species of the World. Entomologica Seandinavica 25: 377-392. Morley, R. J. and J. R. Flenley. 1987. Late Cainozoic vegetafional and environmental changes in the Malay Archipelago. In Whitmore, T. C. (ed.) Bio- geographical evolution of the Malay Archipichigo. Ox- ford Monographs on Geology and Geophisics, No. 4. Clarendon Press, Oxford, pp 50-59. Nordlander, G., Z. Liu, and F. Ronquist. 1996. Phy- logeny and historical biogeography of the cyni- poid wasp family Ibaliidae (Hymenoptera). Sys- tematic Entomology 21: 151-166. Ronquist, F. 1995. Phylogeny and early evolution of the Cynipoidea (Hymenoptera). Systematic Ento- mology 20: 309-335. Ronquist, F. 1996. DIVA version 1.1 computer program and manual available by anonymous FTP from Uppsala University (ftp.uu.se or ftp.systbot.uu.se). Ronquist, F. 1997. Dispersal — vicariance analysis: a new approach to the quantification of the histor- ical biogeography. Systematic Biology 46: 195-203. Ronquist, F. and G. Nordlander. 1989. Skeletal mor- phology of an archaic cvnipoid, Ibalia riifipes (Hy- menoptera: Ibaliidae). Entomologica Scandinaz'ica Supplements 33: 1-60. J. HYM. RES. Vol. 7(2), 1998, pp. 157-164 Taxonomy, Mature Larva, and Observations on the Biology of Gnamptopelta obsidianator (Brulle) (Hymenoptera: Ichneumonidae, Ichneumoninae) Karen R. Sime and David B. Wahl (KRS) Department of Entomology and Section of Ecology and Systematics, Cornell University, Ithaca, New York 14853, USA, (DBW) American Entomological Institute, 3005 S.W. 56th Avenue, Gainesville, Florida 32608, USA Ahstrnct. — The two nominal species in Gnamptopelta, C. obsidiatiator (Brulle) and G. austrina (Cresson), do not merit separate species or subspecies status, and the latter is synonymized under the former; the genus is hence monotypic. Specimens reared from Atnphion floridensis (B.P. Clark) (Lepidoptera: Sphingidae), probably attacked in captivity, represent the first verified host record for Gnamptopelta. A description of the mature larval exuviae is provided. Observations of the wasps in the field and in captivity suggest that G. obsidianator directs its host-searching to grape- vines (Vitis spp.), and that its host range may not include all grapeleaf-feeding sphingids. Gnamptopelta obsidianator (Brulle) is one of the largest ichneumonids in eastern North America and one of the most strik- ing, with its black body, yellow antennae, and, in some southern populations, exten- sive reddish coloring of the head and an- terior mesosoma. While it is commonly collected on the wing and is well repre- sented in collections, no credible hosts have previously been reported. The genus is currently placed in the Trogini, subtribe Callajoppina; nearly all recorded callajop- pine hosts are Sphingidae (Heinrich 1962). Heinrich (1977) noted that G. obsidianator frequents grapevines (Vitaceae: Vitis spp.), which are the food plants for at least 10 sphingids in North America (Forbes 1948; Hodges 1971). Thus, guided by taxonomic inference and a few incidental observa- tions, we set out to identify the hosts of G. obsidianator by rearing Vifis-feeding sphin- gids. We also decided to use this opportunity to clarify the status of species within Gnamptopelta Hopper, in which two spe- cies, obsidianator (Brulle) and austrina (Cresson) were originally included. Sub- sequent authors (Townes and Townes 1951; Heinrich 1962, 1977; Carlson 1979) have considered austrina to be a subspe- cies of obsidianator, although Heinrich (1977) hedged that the subspecies could be "ecologically differentiated species para- sitizing 2 different hosts living on the same plant." The Gnamptopelta specimens examined in this study are in the American Ento- mological Institute (Gainesville, Florida: AEIC) and the Florida State Collection of Arthropods (Gainesville, FL: FSCA). TAXONOMY Heinrich (1962) placed Gnamptopelta in the tribe Trogini, subtribe Callajoppina, which also contains the genera Afrotrogus, Callajoppa, Catadelphus, Conocalama, Dimae- tha, Holojoppa, Pepsijoppa, Stirojoppa, Tme- togaster, Tricyphus, and Yeppoona (Ward and Gauld 1987; Yu and Horstman 1997). All reliable host records indicate that these genera parasitize only Sphingidae. Hopper (1939) erected Gnamptopelta for two North American species, Trogus obsi- dianator Brulle and Trogus austrinus Cres- 158 Journal of Hymenoptera Research son. These were distinguished on the basis of color: G. austrina, found in South Car- ohna, Georgia, and Florida, had the head and anterior mesosoma reddish while G. obsidianator, found in the remaining part of eastern North America west to the Rocky Mountains, had the entire body black ex- cept for the yellowish-white paraocular area. With the exception of Townes (1944), later authors (Townes and Townes 1951; Heinrich 1962, 1977; Carlson 1979) have treated austrina as a subspecies of obsidi- anator. Heinrich (1962) could not decide whether the taxa represented two species or a single species with two color forms. After studying the Florida ichneumonine fauna (Heinrich 1977), he concluded that for Gnamptopelta: 1) "austrina . . . occupies the peninsula of Florida except its most northern part, and that the uniformly black obsidianator occupies the entire east- ern territory of continental North Ameri- ca, including the base of Florida"; 2) "Very sporadically, however, obsidianator also occurs in the territory of austrina, and likewise, sporadically, austrina is found in the southern parts of the territory of obsi- dianator"; 3) ". . .it also must be mentioned that "intergrades" between the 2 forms have never been found" {op. cit., p. 285). Our examination of over 550 Gnampto- pelta specimens reveals that, contrary to Heinrich's assertions, intermediates do oc- cur and all are found in Florida. Most of these individuals are found north of Ocala, in the northern third of the state (above 29°N). Black and intermediate specimens are found throughout the flight period (mid-February through mid-No- vember); there is no sex bias in the color forms. An interesting series of eight males was collected at the American Entomolog- ical Institute in March and April of 1986 [AEIC]. Every variation is present, from uniformly black with no reddish markings to the extreme of an entirely reddish head, reddish anterior mesosoma, and reddish postpetiole. This series alone refutes the notion of two separate entities. We there- fore place austrina as a junior synonym of obsidianator (NEW SYNONYM), noting a particularly apt remark that summarizes the past Gnamptopelta literature: ". . .too much time has been wasted on trying to assemble data, much of it illusory, to maintain already proposed names for sub- species" (Franclemont 1973). All records to date indicate that G. ob- sidianator does not occur west of the front range of the Rocky Mountains. We have, however, seen one specimen from Califor- nia ("Merced Co.; Los Banos National Wildlife Refuge, 4 mi. N. Los Banos; 12- vi-1981; D. Jamieson"; FSCA). It is uni- formly black and identical to specimens from the northeastern part of the range. Queries of the major California collections have not revealed any more specimens from the West Coast. At this time, we con- sider the specimen to be mislabeled or perhaps the result of an import to the area. MATURE LARVA OF GNAMPTOPELTA OBSIDIANATOR Materials and Terminology. — The termi- nology of the cephalic sclerites of the ma- ture larva is that of Gillespie and Finlay- son (1983) and Short (1978), with modifi- cations by Wahl (1990). The epistomal su- ture is the lateral groove between the anterior tentorial pits (Matsuda 1965). It is present as a distinct depression in mature ichneumonid larvae. Although the area is usually unsclerotized, a continuous scler- otized band extends across it in the An- omaloninae, Ichneumoninae, Metopiinae, and Pimplini (and isolated taxa in other subfamilies). Previous ichneumonid work- ers have referred to this area as the "ep- istoma," including in it the portion of the pleurostoma above the superior pleuro- stomal process. This use of "epistoma" should be eschewed. We suggest the fol- lowing descriptive terms: a) epistomal su- ture unsclerotized; b) epistomal suture partially unsclerotized, medially incom- plete (as in Cryptus albitarsis (Cresson); Short 1978: fig. 238); c) epistomal suture Volume 7, Number 2, 1998 159 Fig. 1. Mature larva of Cnainptopclta ohsiduviator: (a) cephalic sclerites, (b) prothoracic spiracle. The scale lines represent 0.1 mm. completely sclerotized forming an episto- mal hand. Methods of larval preparation are those of Wahl (1989). Wahl's notaHon for larval preparations follows the museum acro- nym. It consists of his initials, the day, month, year, and a letter designating the individual preparation. Description. — Epistomal band present, with 14 pores. Labral sclerite, hypostomal spur, stipital sclerite, labial sclerite, and prelabial sclerite, absent. Clypeolabral plates present and triangular in outline, not connected to epistomal band, and not joined or touching medially. Pleurostoma and hypostoma strongly sclerotized. Ac- cessory pleurostomal area moderately sclerotized. Hypostoma straight, posterior end not elongate or upcurved. Accessory reticulate sclerotization prominent. Max- illary palpus with four sensilla and an ad- jacent sclerotized seta; labial palpus with five sensilla and four adjacent setae. Man- dible strongly and uniformly sclerotized; blade about 0.4 X as long as mandible, without denticles. Antenna disc-like, rim moderately sclerotized; central papillus absent. Parietal band present and weakly sclerotized. Prothoracic spiracle as in fig. lb: atrium with width about 0.7x as long as closing apparatus and interior with nu- merous long heavy spines; atrial opening with conspicuous rim; closing apparatus adjacent to atrium. Skin smooth, with widely scattered setae and without spines. Fig. 1 illustrates the cephalic sclerites and prothoracic spiracles of the mature larva; it is a composite of two individuals collected in the Ocala National Forest (USA: Florida; Marion Co.; Lake Delancy campground area, 5.5 miles N-NW Salt Springs, 29° 25'N, 81° 47' W; K.R. Sime & D.B. Wahl; emerged 31 May 1997 [AEIC: DBW l.vi.l997a] and emerged 19 June 160 Journal of Hymenoptera Research 1997 [AEIC: DBW 20.vi.l997a], both from Aviphioii floridensis). The characteristic ab- sence of the labial sclerite in ichneumoni- nes resulted in distortion of the labial re- gion in both preparations. Consequently, placement of the salivary orifice, palpi, and associated setae is only approximate. Discussion. — Short (1978) and Gillespie and Finlayson (1983) provide keys to final- instar larvae of ichneumonine tribes and genera. Townes and Heinrich differed re- garding placement of the Callajoppina. Short follows Townes (Townes et al. 1961) in placing them in the Heresiarchini ("Ich- neumonini" of Short); Finlayson accepts Heirurich's placement of these genera as the subtribe Callajoppina in the Trogini (her "Trogusini"). Both sets of keys will take Gnamptopelta to the correct group of genera. Further use of Short will run Gnamptopelta to Callajoppa; Gillespie and Finlayson will key it to Conocalama. Mod- ification of these keys will not, in our opinion, be useful. Both are inadequate samples of the world ichneumonine fau- na: of approximately 370 genera world- wide (Wahl 1993), Gillespie and Finlayson treat 25 and Short 53. Any attempted cor- rections of the generic keys will risk defeat when the next previously undescribed lar- va is found. BIOLOGY OF GNAMPTOPELTA OBSIDIANATOR Observations. — Field work was conduct- ed in the Ocala National Forest, in north- central Florida (29° 25'N, 81° 47' W), where both grape and G. obsidianator are abundant in the spring (March to May). The study site is part of an extensive stand of longleaf pine (Pinaceae: Pinus paiustris Miller), with an open understory domi- nated by wiregrass (Gramineae: Aristida striata Micheaux) and scattered low shrubs, chiefly scrubby oaks (Fagaceae: Quercus marilandica Muenchhausen and Q. laevis Walter), and pawpaw (Annonaceae: Asimina speciosa Nash and A. obovata Willdenow). Two species of grape are Table 1. Summary of field observations of female obsiiiianator. Identification of plants landed upon with searching times, recorded when possible, in paren- theses. Before 8 April we had not confirmed the iden- tities of Viti:^ species. Wasp No, Searching Sequence 1 (23 March) Vitif (2-3s); V;f/s (2-3s); Vitis (2-3s) 2 (23 March) Quercus (20s); V. aestivalis (10s) 7 (8 April) V. aestivalis (captured after 2-3s) 8 (17 April) V. rotundifolia (>60s) 9 (24 April) V. aestivalis (captured after 2-3s) found here, both extremely common and often covering large expanses: Vitis aesti- valis Micheaux and V. rotundifolia Mi- cheaux. In 1997, most grape plants of both species were fully leafed out by the mid- dle of March, and both continued to put out new leaves throughout the spring. A few G. obsidianator were seen as early as 10 March; the peak of abundance ap- peared to be during the last week of the month, when for a few days 12-15 wasps were seen each day. From then until the middle of May, when observations ceased, we usually saw 1-3 individuals in the course of 4-5 hours spent almost every day in the field. Males fly rapidly, 2-3 meters above the ground, land rarely, and follow wide cir- cuits in patches of forest abounding in grapevines. They were perhaps patrolling areas in which females were likely to eclose or to enter in the course of search- ing for hosts, although we witnessed no encounters between the sexes. The females as well as the males are as- sociated with grape at this field site. The females also fly quite fast, but land often. Table 1 summarizes our observations of searching behavior. Females appear to fly deliberately from one grape patch to the next, less often landing on other common low shrubs, which were usually surround- Volume 7, Number 2, 1998 161 ed by if not overlaid v^ith grapevines. The wasps land on the upper surface of the leaf and tap it several times with the ven- tral surfaces of the antennae; usually, the inspection lasted just a few seconds, and the wasp then proceeded to the next plant, but in some cases the wasps stayed longer and conducted a thorough search of near- by leaves and stems. We suspect that the length of the search might be related to the presence of feeding damage or other trac- es left by host larvae (Vinson 1984). How- ever, we observed no encounters with hosts in the field: the leaves searched, in- cluding those examined for just a few sec- onds, often bore herbivore damage, but we never found larvae in their vicinity. Between 31 March and 2 May 1997 we collected 66 sphingid larvae on Vitis. We reared them in small groups (3-10 larvae) in plastic boxes which we supplied with fresh grape leaves every 1-2 days. Using Forbes's (1948) key, we were able to iden- tify the caterpillars to species when they reached the 4th instar (and confirmed the identifications when adults emerged). All were macroglossine sphingids: Darapsa myron Cramer (33 reared to pupation), Amphiou floridensis B.P. Clark (= nessus Cramer) (5 reared), and Euinorpha achemon Drury (2 reared). Larval mortality was due to braconid parasitism (Aleiodes iex- ensis) killing 4th instars (10 victims, all D. myron) and unknown causes among 2nd- instar larvae (too small to identify). Of the D. myron and A. floridensis found during the first two weeks (about 45 larvae), all but one were small second and third in- stars; later collections included a few 4th and 5th instars. The two £. achemon larvae were found side-by-side on April 26 as very nearly mature 5th instars. All larvae were found on V. aestivalis. Although we often found compelling feeding damage on V. rotundifolia, and this plant was inspected both by us and by Gnamptopelta females, we found no sphin- gid larvae on it: if not coincidence, this could be evidence for an oviposition pref- erence in the adult sphingids, or for higher larval survivorship on V. aestivalis. Fur- thermore, in rearing the larvae, we found that they preferred to feed on V. aestiimlis when both species of grape were in their boxes. This could reflect the habituation of the larvae to the plant on which they ini- tially fed (Jermy 1987), or it could repre- sent an inherently greater palatability of V. aestivalis to the larvae. As most of the larvae were D. myron, these observations may apply only to that species; the num- bers of the other two species collected were too small for useful speculation. In addition, we noted that all larvae fed on full-sized leaves, never on the youngest leaves, which might represent an aversion to the denser pubescence of the youngest leaves (Southwood 1986) or perhaps to higher levels of allelochemicals (Feeny 1976). The larvae of D. myron and A. floridensis have many habits that make them incon- spicuous in the field. Whether feeding or resting, all instars are invariably on the undersides of the leaves, and they usually rest along the leaf veins, which the light- green young instars match particularly well in color and shape. Often the larvae rest on undamaged foliage one, two, or three leaves removed from the feeding site. The younger larvae feed in a distinc- tive fashion, by eating symmetrical holes on either side of the leaf, near the base, a pattern that blends extremely well with the deep rounded lobes and hollows of the leaves of V. aestivalis. Older larvae tend to feed along the edges of the leaf, eating off large pieces. Heinrich (1979) noted similar behaviors in the sphingid Sphecodina ab- botti Swainson feeding on V. vulpina L. (in Minnesota) and suggested that this, along with the habit of staying under the leaves, conceals the caterpillars from predators that hunt by eye, particularly birds. We did not happen to observe any birds in- specting grapevines, but speculate that these behaviors may be somewhat effec- tive against G. obsidianator. As the wasps 162 Journal of Hymenoptera Research land initially on the top surface of the leaf, a caterpillar resting underneath may gain some extra time to fall off without being noticed, and its removal from the feeding site while resting might help frustrate a wasp that was initially attracted (by sight or by odor) to feeding damage (Vet and Dicke 1992). Our observations of G. obsidianator and sphingid larvae in captivity suggest that letting go from the leaves is an effective means by which caterpillars elude wasps. We placed three field-caught females in a small glass aquarium (23-cm cube) with 4 larvae (at a time) feeding on grape sprigs. Some larvae (both D. myron and A. flori- densis) dropped off their leaves as soon as wasps set foot on the leaves; others dropped only after the wasps' antennae or tarsi touched them. If the wasps tried to oviposit, the time it took them to bring their ovipositors into position, after mak- ing antennal contact, was long enough for the larvae to drop down and thus con- found the oviposition attempt. Dropping appeared to be the main defense of small- er larvae; fourth- and fifth-instar larvae, when touched by a wasp, stayed put and swung their heads about with such vio- lence as to parry any attempt at oviposi- tion quite effectively, sometimes hurling the wasp against the side of the cage. The wasps, if not thrown, in many instances had so much difficulty inserting their ovi- positors into the writhing caterpillars that they simply gave up. When exploring the aquarium, the wasps tapped steadily with the extended antennae, lightly touching the apical quar- ter (flattened ventral surface) to the sub- strate. The wasps showed little interest in D. myron larvae, never attempting to sting them even after finding and examining them. In contrast, A. floridensis elicited great excitement: upon encountering a chewed leaf margin or a particle of frass, the wasps slowed their walking pace, vi- brating the antennae much more rapidly and pressing them more firmly against the surface, with more of the antennal area, approximately the apical third, making contact. Upon finding A. floridensis larvae, the wasps tapped them for 1-2 seconds with the antennae and then turned around to sting. We witnessed six apparent ovi- positions, by two females; it appears that the eggs may be inserted anywhere into the host's body apart perhaps from the head capsule. Attempts at oviposition oc- curred with third-, fourth-, and fifth-instar larvae (we did not expose younger lar- vae). The G. obsidianator that we reared came from A. floridensis and resulted from those larvae exposed as 3rd and 5th instars. In total, we had six A. floridensis larvae that grew large enough to be identified: two went to term and emerged as moths, and four others were exposed, of which one died just before and one some time after pupation and two produced the wasps. Six of the D. myron larvae (also 3rd, 4th, and 5th instars) were exposed to the wasps and continuously observed, but no ovipositions were attempted with these larvae and all developed into adult moths. The sphingids, parasitized and not, pu- pated by forming loose cocoons amid leaves and paper towels. The adult wasps emerged by cutting off the anterior 0.1 of the host pupa. Such an emergence hole is apparently characteristic of the Callajop- pina (Mell and Heinrich 1931), and ap- pears to be the primitive condition for the Ichneumoninae (Gillespie and Finlayson 1983). The ichneumonid's rudimentary co- coon, located in the detached anterior sec- tion, consists of a cap of dark brown silk just inside the cut margin. Conclusions. — Although it was the most common grape-feeding sphingid at our field site in spring 1997, our observations suggest that D. myron is not a host of G. obsidianator. Of 33 larvae collected on plants exposed to a considerable popula- tion of G. obsidianator, no wasps emerged; the wasps showed no interest in this spe- cies in captivity (and much interest in A. Volume 7, Number 2, 1998 163 floridensis under the same conditions). It is possible that the larvae we collected were too small for successful oviposition or were not old enough to have spent much time exposed to wasps, thus lowering the probability of parasitism, or that parasit- ism rates are generally very low; however, the combined field and captivity observa- tions argue that D. myrou is at least not a preferred host. That A. floridensis is a natural host is strongly supported by our successful rear- ing in captivity: a koinobiont larva-pupal endoparasitoid, G. obsidianaior is not likely to develop successfully in randomly en- countered moth larvae because it must survive the immunological defenses of the host and also coordinate its development with the onset of pupation in the host (Askew and Shaw 1986; Omata 1984). Am- phion floridensis occurs throughout the geographical range of G. obsidianator and is thus potentially the sole host: its range is described as extending "from Florida, north to Nova Scotia, west to Calgary, Al- berta, and south to Kansas" (Hodges 1971), and our queries of various collec- tions expand this to central Colorado and southern Texas as well. In Florida, at least, the flight period of G. obsidianator coin- cides with the larval stage of A. floridensis (Kimball 1965). However, we can say nothing about £. achemon, because we found only two specimens and did not ex- pose them to wasps; some 10 other sphin- gids feeding on grape in eastern North America (Hodges 1971) also demand in- vestigation. Hopper (1939) lists as a host Papilio po- lyxenes Fabricius (Lepidoptera: Papilioni- dae), referring to a New York list for which voucher specimens are unobtaina- ble. As may be the case for many such iso- lated records (Shaw 1994), this one is al- most certainly specious. This swallowtail is often reared by amateurs and profes- sionals alike, as the caterpillar is quite striking and very often found on garden herbs and common roadside weeds (Um- belliferae). We have seen scores of speci- mens of Trogus pennator Fabricius (Ichneu- moninae) reared from P. polyxenes in col- lections, but no reared G. obsidianator; a life-history study of P. polyxenes in central New York State, where G. obsidianator is common, found that about 10% of 128 pu- pae were parasitized by T. pennator, with a few other parasitoids occurring infre- quently, but no G. obsidianator (Feeny et al. 1985). We conclude that G. obsidianator para- sitizes at least one species of Viffs-feeding sphingid. The degree of specialization within this group is unknown, but our ob- servations of captive wasps suggest that G. obsidianator does not uniformly attack available sphingid species and that it could be restricted, at least locally in Flor- ida, to A. floridensis. Our results support Heinrich's (1962) contention that the gen- era of the Callajoppa group are strictly par- asitoids of Sphingidae. ACKNOWLEDGMENTS We thank Tim Carr, Dan Janzen, Mark Shaw, Bob Wharton, and two anonymous reviewers for advice and assistance. Steve Heydon (U.C. Davis), Wojciech Pulawski (California Academy of Science), Roy Snell- ing (Los Angeles County Museum), Serguei Triapit- syn (U.C. Riverside), and Robert Zuparko (U.C. Berkeley) searched collections for western North American records of Guainptopelln; Robert Brooks (University of Kansas), Richard Brown (Mississippi State University), John Heppner (Florida State Collec- tion of Arthropods), Eric Quinter (American Museum of Natural History), Edward Riley (Texas A&M Uni- versity), and James Whitfield (University of Arkan- sas) checked A. floridensis distribution records. We are indebted to James Wiley for the FSCA Gnamptopelta loan. We thank the staff of the Lake George Ranger District, Florida, for granting permission to work in the Ocala National Forest. This material is based upon work supported under a National Science Foundation Graduate Research Fellowship (KRS) and funded by NSF grant IBN-9600094 to KRS and Dr. P. P. Feeny. LITERATURE CITED Askew, R. R. and Shaw, M. R. 1986. Parasitoid com- munities: their size, structure, and development. In- sect Parasitoids (eds J. Waage & D. J. Greathead), pp. 225-264. Academic Press, London. 164 Journal of Hymenoptera Research Carlson, R. W. 1979. Family Ichneumonidae. Catalog of Hymenoptera in America north of Mexico, vol. 1 (eds K. V. Krombein, P. D. Hurd, D. R. Smith, & B. D. Burks). 1198 pp. Smithsonian Institution Press, Washington, D.C. Feeny, P. P. 1976. Plant apparency and chemical de- fense. Recent Aiizninces in Phytochemistry , Vol. 10 (eds J. W. Wallace & R. L. Mansell), pp. l^W. Plenum Press, New York. Feeny, P. P., W. Blau, and P. Kareiva. 1985. Larval growth and survivorship of the black swallowtail butterfly in central New York. Ecological Monograplis 55: 167-187. Forbes, W. T. M. 1948. Lepiiioptera ofNeiv York and neigh- boring states. Part //. Cornell University Agricultural Experiment Station, Ithaca, New York. Frandemont, J. 1973. Tlie moths of America north of Mex- ico: Fascicle 20.1. Mimallonoidea, Mimallonidae and Bombycoidea, Apatelodidae, Bombycidae, Lasiocantpidae. E.W. Classey Ltd. & R.B.D. Publications, Inc., Lon- don. Gillespie, D. R. and T. Finlayson. 1983. Classification of final-instar larvae of the Ichneumoninae (Hyme- noptera: Ichneumonidae). Memoirs ofttie Entomolog- ical Society of Canada 124: 1-81. Heinrich, B. 1979. Foraging strategies of caterpillars: leaf damage and possible predator avoidance strategies. Oecologia 42: 325-337. Heinrich, G. H. 1962. Synopsis of the Nearctic Ichneu- monidae Stenopneusticae with particular reference to the northeastern region (Hymenoptera): Part VII: Synopsis of the Trogini. Canadian Entomologist sup'- plcment 29: 807-860. Heinrich, G. H. 1977. Ichneumoninae of Florida and neighboring states. Arthropods of Florida and neigh- boring land areas 9: 1-350. Hodges, R. W. 1971. Tlie moths of America north of Mexico, Fascicle 11. Sphhigoidea (Hawkmotlis). E.W. Classey Ltd. & R.B.D. Publications, Inc., London. Hopper, H. P. 1939. A synophcal revision of the tribe Trogini Ashmead of the United States and Canada (Hymenoptera: Ichneumonidae). Transactions of the American Entomological Society LXV: 307-347. Jermy, T. 1987. The role of experience in the host selec- tion of phytophagous insects. Persp'ectiivs in chc- moreception and behavior (eds R. F. Chapman, E. A. Bemays, & J. G. Stoffolano, Jr.), pp. 143-157. Spring- er-Verlag, New York. Kimball, C. P. 1965. The Lepidoptera of Honda: an an- notated checklist. Arthropods of Florida and nciglibor- ing land areas 1:1-363. Matsuda, R. 1965. Morphology and evolution of the in- sect head. Memoirs of tlie American Entomological In- stitute 4:1-334. Mell, R. and G. Heinrich. 1931. Zur Biologie and Sys- tematik der siidchinesischen Ichneumoninae Ashm. (Fam. Ichneumonidae Hym.). Zeitschrift fiir ange- wandte Entomologie 18: 371^03. Omata, K. 1984. Influence of the physiological condition of the host, Papilio xuthus L. (Lepidoptera: Papili- onidae) on the development of the parasite, Trogus mactator Tosquinet (Hymenoptera: Ichneumoni- dae). Applied Entomology and Zoology 19: 430-^35. Shaw, M. R. 1994. Parasitoid host ranges. Parasitoid com- munity ecologi/ (eds B. A. Hawkins and W. Shee- han), pp. 111-144. Oxford University Press, Oxford. Short, J. R. T. 1978. The final larval instars of the Ich- neumonidae. Memoirs of the American Entomological Institute 25: 1-508. Southwood, T. R. E. 1986. Plant surfaces and insects — an overview. Insects and the plant surface (eds B. Ju- niper & T. R. E. Southwood), pp. 1-22. Edward Ar- nold Ltd., London. Townes, H. 1944. A catalogue and reclassification of the Nearctic Ichneumonidae (Hymenoptera). Part I. The subfamilies Ichneumoninae, Tryphoninae, Cryptinae, Paeogeninae and Lissonotinae. Memoirs of the American Entomological Society 11(1): 1^77. Townes, H., and M. Townes. 1951. Family Ichneumon- idae. Hymenoptera of America north of Mexico— syn- optic catalog (eds C. F. W. Muesebeck, K. V. Krom- bein, & H. K. Townes). Agriculture Monograph 2: 1- 1420. Townes, H., M. Townes, and V. K. Gupta. 1961. A Cat- alog and Reclassification of the Indo-Australian Ich- neumonidae: Tribe Trogini. Memoirs of tlie American Entomological Institute 1: 1-522. Vet, L. E. M., and M. Dicke. 1992. Ecology of infodi- emical use by natural enemies in a tritrophic con- text. Annual Rei'ieios of Entomolog}/ 37: 141-172. Vinson, S. B. 1984. How parasitoids locate their hosts: a case of insect espionage. Insect Communication (ed. T. Lewis), pp. 325-348. Royal Entomological Sod- ety, London. Wahl, D. B. 1989. Further notes on preparation of exu- viae of parasitic Hymenoptera. Entomological Neics 100: 181-182. Wahl, D. B. 1990. A review of the mature larvae of Di- plazontinae, with notes on larvae of Acaenitinae and Orthocentrinae and proposal of two new sub- faniilies (Insects: Hymenoptera, Ichneumonidae). Journal of Natural History 24:27-52. Wahl, D. B. 1993. Family Ichneumonidae. Hymenoptera of the World: an identification guide to families (eds H. Goulet & J. T. Huber), pp. 395-442. Canada Com- munications Group, Ottawa. Ward, S. and Gauld, 1. 1987. The callajoppine parasitoids of sphingids in Central America (Hvmenoptera: Ichneumonidae). Systematic Entoniologif 12: 503-508. Yu, D. S. and Horstmann, K. 1997. A catalogue of world Ichneumonidae (Hymenoptera). Memoirs of tin- American Entomological Institute 58 (1-2): 1-1558. J. HYM. RES. Vol. 7(2), 1998, pp. 165-177 The Nesting Behavior and Dynamics of Bicyrtes angulata (F. Smith) with a Comparison to other Species in the Genus (Hymenoptera: Sphecidae) RoGERio Parentoni Martins*, Lourdes Aragao Scares, and Douglas Yanega Laboratorio de Ecologia e Comportamento de Insetos, Departamento de Biologia Geral, ICB-UFMG, Cx. Postal 486, 30.161-970-Belo Horizonte-MG-Brazil; *E-mail: Wasp@dedalus.lcc.ufmg.br Abstract. — Data are presented on the nesting behavior and dynamics of a population of Bicyrtes angulata (F. Smith) found on the campus of the Universidade Federal de Minas Gerais, Belo Horizonte, Minas Gerais, Brazil. Many aspects of the biology of this population of B. angulata are similar to all (or the majority) of the 8 other species in the genus that have been studied in some detail. Other features were atypical in comparison, such as: (1) nesting in sandy soil far from water (2) constructing relatively short burrows (less than 10 cm on average) (3) taking more than two hours to dig a nest. New information is presented on the level of nest failure (roughly a third of all nests initiated are not completed), mortality (roughly 90% of all completed nests fail to produce adult wasps), and variability in egg to adult developmental time (44 to 375 days) due to bimodal prepupal dormancy. Factors affecting mortality are discussed, the most important being termites and ants. The number of nests made per female varied slightly over the year, this ratio being lowest in the dry season (winter and spring), and was not correlated with the total number of nesting females, thus suggesting that there is little or no competition for nesting space in the area. Possible alternative explanations for these patterns are offered, in relation to prey abundance and female mobility and longevity, and we suggest that the availability of suitable patches for nesting may potentially be of importance in the nesting dynamics of this species. Of the 23 described species of Bicyrtes, al species should help address questions 12 are found exclusively in South Ameri- related to patterns of behavioral evolution ca, 8 in North America and 3 occur in both within the genus. continents (see Bohart and Menke 1976 for Bicyrtes angulata (F. Smith) occurs in distributions; Willink, 1947, revised the French Guiana, Paraguay, Argentina, and South American species). Their biology is Brazil (Bohart and Menke 1976). Records largely either unknown or poorly known, for B. angulata in Brazil include some with the exception of the North American states in the North and Northeast (Nasci- B. quadrifasciata (Say) (see Evans 1966 for mento and Overall 1980) and Sao Paulo a review). Some addiHonal information is (Martins 1991), but its biology has been available for B. discisa (Taschenberg), B. unknown until the present, and is com- simillima (Smith) and B. variegata (Olivier) P^''^'^ ^^^^ ^° ^^^^ has been recorded for (Genise 1979, 1982), B. cingulata Burmeis- o*^"" ^P^^^^^ '" 'he genus. This includes ter (Evans and Matthews 1974), and also "°^«^' information on prepupal dormancy, for the Cuban B. spuwsa (Fabricius) (San- ^"'^ ""f ^' ^"'^ ^^"^^^ °^ "^^* ^^''^''^ ^"'^ chez and Genaro 1992). Few of the details "^o^t^lity. of nesting and provisioning behavior have STUDY SITE AND METHODS proven to be uniform across all species We observed an aggregation of B. an- (see Discussion). Information on addition- gulata (147 marked nests in 1993 and 80 in 166 o Neslsin1993 • Nests in 1994 ^■y Bushes Ji* Grasses >!|r WaltherLa Journal of Hymenoptera Research ^»> ^.iff ^ 4fc W *. ^ dk' U, 4t" ^ *#''=«fi»«^^*'*8.„^^i^bo •r-v^f Fig. 1. Diagram of the nesting site of Bicyrtes angulata showing nest distributions in 1993 and 1994. 1994) on the Campus of the Federal Uni- versity of Minas Gerais (UFMG), Belo Ho- rizonte, Minas Gerais, Brazil, from Janu- ary 1993 to June 1994. We spent a total of 556 hours making ad libitum and focal in- dividual observations {sensu Martin and Bateson 1986), mostly between 0900 and 1200 h when the majority of nesting activ- ity occurred, and opportunistically at the beginning and end of some days. The study site, Prefeitura, is a secondary growth, relatively undisturbed area, most- ly covered with grasses, scattered bushes, and trees (including a few exotic and or- chard species; see photo in Gaimari and Martins 1996). Nearby fields of corn, bean and manihot support a substantial popu- lahon of Waltheria americana (Sterculi- aceae), a weed whose flowers are com- monly visited by B. angulata (Macedo and Martins 1998). Most nests of B. angulata were concen- trated in a 35 m long and 0.65 m wide strip of a dirt road (approx. 700 m- total) on a superficially compacted well-drained sandy soil. Nesting soil was completely free of vegetation and exposed directly to the sun (Figs. 1 and 2). We built a tem- ■r' ^ L .y Fig. 2. Nesting site of Bicyrtes nii^uhUn showing the plastic-cup emergence traps. Volume 7, Number 2, 1998 167 porary fence around the nesting area to prevent vehicular and pedestrian traffic. In 1993 the area also contained numerous nests of other sphecid wasps: 193 of Riih- rica uasuta (Christ) (Pimenta and Martins unpublished data), 54 of Ammophila graci- lis Lepeletier (Gaimari and Martins 1996), 4 of Bicyrtes discisa, 4 of Tracln/pus sp., and 1 of Prionyx fervens (Linnaeus). There were also a few bee nests at the same time: 7 of Centris aenea Lepeletier, and 1 of Megnchile iieoxanthoptera Cockerell. We marked and released 180 females (127 in 1993 and 53 in 1994) and 31 males (19 in 1993 and 12 in 1994) that were cap- tured in or near the neshng site while ei- ther searching for a place to begin exca- vation, in the process of excavating a nest, flying in search of prey or in transit to the nest, or visiting the flowers of Waltheria americana (see Macedo and Martins 1998, for armual records of B. angulata visits to this plant). Numerous additional wasps were marked in areas outside the nesting area, but none of these wasps were sub- sequently encountered there. Wasps were restrained and marked individually with fast-drying non-toxic acrylic paint using combinations of 4 colored dots on the cor- ners of the scutum, making it possible to observe individual variation in daily phe- nology, nesting (note that many marked females made no nests in the study site), provisioning, male behavior, longevity, in- teractions with other insects and plants, and movements in the aggregation and its surroundings. Adult longevity was calcu- lated as the length of time between when a wasp was marked and the last time it was seen, and is thus likely to be a very conservative estimate, though many wasps were never seen again after mark- ing (64 females and 21 males), and we did not include these individuals in the aver- age. Differences between male and female longevity were tested using Student's t- test. Nests were marked during interrup- tions in the nest excavation process, using 4 cm-long and 5 mm-wide aluminum ar- rows fixed to the ground with nails. Ar- rows were distinguished by colors corre- sponding to those of the resident female and /or by numbers. After observing a de- finitive nest closure, we nailed an emergence trap (a marked plastic cup) to the ground at the nest entrance (Fig. 2), to record egg-adult development time and the emergence of nest parasites. This also helped us to estimate the number of an- nual generations. Wasps from the first generation that emerged in the field or in the laboratory were marked and released in the study site, so as not to extirpate the population. We excavated all 227 marked nests to determine their architecture and contents, assessing whether they were deserted, parasitized, destroyed by ants or termites, or completed. The status of the nests dur- ing the study was categorized as follows: (1) Deserted nests were those in which we observed a female working but which proved to be empty upon excavation (this class of nests thus includes both abandon- ment and female mortality events) (2) Par- asitized nests were those in which para- sites appeared in the emergence traps (3) Nests destroyed by ants or termites were those in which we found pieces of de- stroyed cocoons and prey remains (4) Completed nests were those that we ob- served being provisioned and in which we found all remains of cocoons and /or prey. Only nests which were not deserted were included in the analysis of nesting phe- nology and counts of nests per females. We counted 307 prey that were con- sumed by larvae in 26 of the nests; it was possible to count the exact number of con- sumed prey because their heads and scuta remained intact in the cell. Eight days af- ter provisioning ceased, seven nests were excavated to measure the time of devel- opment from egg to prepupae. The num- ber of pores of 14 cocoons were counted and the larger and smaller diameter mea- sured for 8 of them. We made casts of 2 168 Journal of Hymenoptera Research Fig. 3. Architecture of the nest of Bicyrtes angulata. completed nests with plaster of Paris to record their shape (Fig. 3), and casts of two incomplete nests. Nesting activity was measured as the number of nests ini- tiated per month. We plotted the number of nesting females per month versus the average number of nests made per female to assess evidence for density-dependent effects. For quantitative measurements we calculated averages with standard devia- tions. Voucher specimens of wasp and prey were deposited at the Laboratorio de Eco- logia e Comportamento de Insetos of the Departamento de Biologia Geral, ICB- UFMG, Belo Horizonte, MG, Brazil. RESULTS Nest excavation. — Over the entire study, a total of 180 females excavated 227 nests, and 80 of these females excavated at least one nest within a week of being marked. In 1993, the final nest density was approx- imately 16.5 nests/m-. We also observed the origin of a new nest site in 1994, when 29 out of the 80 nests of B. angulata found were in an area close to, but separate from, the original site (Fig. 1). Females walk in a slow and meandering path over the nesting site, occasionally an- tennating the ground, until they find a suitable spot (the selection criteria are un- known, but are likely to be tactile) and be- gin to dig. Females dig primarily with their mandibles, using the forelegs to scoop loosened soil backwards, throwing the soil out from under the body. This movement is accompanied by a rapid rocking motion of the body, lifting the ab- domen each time the soil is thrown, and simultaneously tilhng the head down over the soil. The excavated soil accumulates, forming a mound (tumulus) close to the nest entrance. Over the entire period of ex- cavating a single nest, females spend an average of 2.26 ± 0.13 h (n = 4) actually digging the nest, but also 2.39 ± 0.24 h (n = 4) flying to and from the nests, or sim- ply resting in the tunnel. Completed nests (n = 5 with all measurements taken) have a straight tunnel, 9.6 ± 1.1 cm long and 0.8 ± 0.1 cm in diameter, angled from 20° to 35° relative to the soil surface (Fig. 3), and ending in a single ellipsoid cell (only two nests had two cells) measuring 2.6 cm long by 1.2 cm in diameter (all 5 measured cells identical), at 5.6 ± 0.54 cm depth. Fe- males typically complete nest excavation in the afternoon. Afterwards, they close the nest entrance and fly away to feed on nectar, and, from March 16 to April 20, to sleep in mixed male /female aggregations on inflorescences of Panicum maximum (Poaceae), in proximity to adults of Rub- rica nasuta (described below) (Fig. 4). No form of orientation flight was ever ob- served, either after nest construction, af- ternoon closure, or in between provision- ing trips. Nest provisioning. — Out of 227 nests, 147 were succesfully provisioned by 103 dif- ferent females. Provisioning can occasion- ally start the same afternoon that nest ex- cavation is finished (n = 2) but is gener- ally begun the following day. In the latter case, by 0830 to 0900, the females return and open the nest entrances with their forelegs, using motions as described above. Hunting trips occur between 0845 and 1545 h. As they leave the nest for the first and all subsequent hunting trips, they emerge headfirst, and temporarily close the entrance by scooping the tumulus backwards into the entrance. However, sometimes nest entrances stay partially open due to careless and hasty closure. If heavy rains level off the mound, females still have little or no difficulty finding Volume 7, Number 2, 1998 169 Fig. 4. Mixed male-female sleeping site of Bkyrtes angulata. their nest. Sometimes they seem confused initially, but soon find the nest entrance and reopen it. Prey are immatures (sometimes adults) of Megnlotomus sp. and Apidaurus sp. (He- miptera: Coreidae: Alydini [=Alydidae]), accounting for 302 prey items; only one anomalous cell was found, containing 5 pentatomid nymphs. Each prey is stung and partially paralyzed, then the wasp flies back to the nest holding it venter up with her middle legs. She then removes the nest closure with her front legs, keep- ing the prey in the same position, and in this way she enters the nest headfirst. Once, after removing the nest closure, a female dropped the prey at the nest en- trance, entered, came back out headfirst, grabbed the prey in her mandibles, and dragged it down to the cell by moving backwards. Hunting trip lengths are quite variable (45.6 ± 35.9 min., n = 13; range 1-95). After capturing the first prey and placing it in the nest, one egg is attached erect on the metastemum between the middle or hind coxae, exactly as shown by Evans (1966:158). Since the time between the first and second provisioning trips can vary considerably, it appears that the egg may hatch either before or after the second prey is supplied, but in most cases it is probably after two or three prey are al- ready in the nest. Provisioning is progressive and some- what lengthy, lasting approximately 6 consecutive days per nest (6.5 ± 0.53, n = 7), with roughly 12 prey per completely stocked cell (11.88 ± 4.46, n = 26; range 7-24). The final closure of completed nests begins with females scraping the tumulus into the nest entrance, as when making a temporary closure, but is somewhat more thorough, and this is followed by a bout of "hammering," during which a female delivers vigorous blows with the tip of the abdomen until the nest entrance is com- pacted and levelled. The egg-prepupa de- velopmental period is rapid, as we found 170 Journal of Hymenoptera Research in ro E E 3 ■ Number marked D Number nesting F M A M 1994 Fig. 5. Number of marked and nesting females of Bicyrtes angulata in 1993 and 1994. Numbers above the bars indicate the total number of nests constructed in that time interval. prepupae already inside their cocoons (n = 7) when excavated 8 or 9 days after pro- visioning had ceased. Cocoons had 4 to 6 pores (5.5 ± 1.95, n = 14) and averaged 17 ± 0.9 mm long by 5.3 ± 0.06 mm (n = 8) in diameter. The entire cycle for a single nest, from first excavation to final closure, typically takes 8 days (8.5 ± 0.5, n = 6). Seasonal phenology and demographic pat- terns.— Adults are found during the whole year, mainly visiting flowers, but nests are built mainly from February to May (mid- summer and autumn; 85 of 97 provisioned nests in 1993, and all 50 in 1994; Fig. 5). The ratio of nests provisioned per nesting female was generally higher than one ex- cept in the winter and spring (June through November, Fig. 6; in October the ratio was higher because two females pro- visioned five nests in total). Overall, there is no relationship between the number of nesting females per month and the aver- age number of nests per female per month. If we consider the number of nests relative to the total number of marked fe- males, it is obvious that many females were present in the area that never made nests (in 7 of 14 months in which nesting was observed, there were more marked fe- males alive than nests provisioned). De- spite the fact that the range of number of nests per nesting female was one to seven, the average number of nests per marked female only varied from 1.0 to 1.5 in dif- ferent months. Females seem to feed on flowers in ar- eas other than those in proximity to their nests; only three out of 77 females marked Volume 7, Number 2, 1998 4n 171 ■ Nests per marked female D Nests per nesting female FMAMJJASONDJFMAM 1993 1994 Fig. 6. Monthly ratio of number of nests to females of Bicyrtes aji^iilata in 1993 and 1994. on W. americana, close to the study site, were seen nesting in the aggregation. However, 77 out of 103 females that were marked while visibly searching for a place to begin excavation were later seen nest- ing there. Females stayed in the aggrega- tion from 1 to 84 days (24.1 ± 17.5, n = 116). One female was observed nesting until the age of 65 days, with no apparent mandibular or wing wear over this time. The estimate of male longevity (31.4 ± 20.4 days, n = 10; range 3-56) was not sig- nificantly different from that of females (Student's t = -1.25, p > 0.2). There are at least 2 annual generations, though these are staggered and overlap- ping rather than synchronized and dis- crete (see Table 1 ). The range of time from oviposition to adult emergence was 44 to 375 days (200.1 ± 137.1, n = 13; Table 1). The origin of such variability is in the pre- pupal stage, which may remain dormant for widely varying intervals of time. This variability in dormancy can occur not only among but within the broods of individ- ual females; two eggs from one female produced adult males after 55 and 375 days of development. The operational sex ratio at this site was apparently female bi- ased; 15 wasps from the first generation emerged (in the field or from cocoons brought into the laboratory), of which 9 172 Journal of Hymenoptera Research Table 1. Developmental intervals of Bin/rte angulata (arranged by starting date). Nest Nesl starting date Nest closing date Adult emergence date « dav^ egg-adiill Sex 1 3 12 Feb 1993 13 Feb 1993 18 Feb 1993 19 Feb 1993 16 Apr 1993 9 Mar 1994 57 375 F M 2 12 16 Feb 1993 19 Feb 1993 3 Mar 1993 26 Feb 1993 27 Apr 1993 3 Feb 1994' 55 342 M F 4 3 Mar 1993 9 Mar 1993 19 Aug 1993 163 M 10 8 Mar 1993 7 7 Feb 1994 ca. 310 F 13 11 17 Mar 1993 30 Mar 1993 1 Apr 1993 13 Apr 1993 30 Jan 1994 3 Feb 1994* 335 296 F M 15 5 7 6 9 1 Apr 1993 24 Apr 1993 1 May 1993 6 May 1993 13 Oct 1993 13 Apr 1993 6 May 1993 6 May 1993 7 28 Oct 1993 11 Apr 1994* 29 Jan 1994 4 Sep 1993 31 Jan 1994 24 Jan 1994 356 328 88 ca. 310 88 F F M F F 8 28 Oct 1993 5 Nov 1993 24 Jan 1994 80 F 14 11 Feb 1994 22 Feb 1994 8 Apr 1994 44 M * These individuals emerged in thie laboratory. were females (60%). We found 5.7 females per male (85% females) over the period of the study, though males were rarely seen on the aggregation and thus their abun- dance was likely underestimated. Movement pat tents and diel phenology. — Adults of B. angulata spend nights on plants. From March 16 to April 20, marked B. angulata and R. nasuta were seen sleep- ing on inflorescences of Panicum maximum (Fig. 4). The highest number of wasps seen in a single group was 17 B. angulata and 9 R. nasuta (J. F. Macedo, pers. comm.). Fe- males leave the sleeping site between 0800 and 0830 h to take nectar prior to the start of nesting activities (other phenological data above). Usually after 1600 h, or when it is very cloudy, they return to the sleep- ing site. One day they left the sleeping site early, returned around 0900, but then re- sumed nesting activities between 0900 to 1000 h, when the sun began shining again. B. angulata comprised 34% of the individ- uals of the 29 bee and 51 wasp species vis- iting Waltheria americana between 1000 and 1200 h in April 1993 and April 1994 (Ma- cedo 1995). The highest number of B. an- gulata seen on W. americana during one month was 35 females and 7 males in June 1993 at a site some distance from the nest- ing site (despite the availability of flowers, no visits were observed on the same plants in June 1994, reflecting the large variation in the abundance of wasps be- tween years). Only 8 males were captured and marked (of 31 total) while they were pa- trolling the aggregation by flying close to the soil, continuously going back and forth. Occasionally they clashed briefly in mid-air with other males, females, or other insects flying in the patrolling area. The highest levels of patrolling activity were observed in April between 0845 to 1145 h, and 1300 to 1600 h. Despite the patrolling activity, copulation was never seen in the aggregation, nor on flowers of W. ameri- cana or in the sleeping site. Four new adults were observed emerg- ing from their cells between 1045 and 1145 h, and new adults were generally found in the emergence traps in the mornings, suggesting that there is a preferred daily emergence time. Interactions with natural enemies and other insects. — The female actively defends the nest entrance against insects. Once, when an army ant trail crossed the entrance of one nest, the female tried to drive them out by touching them with its abdomen or Volume 7, Number 2, 1998 Table 2. Survivorship and success of Bici/rtcs angulata nests. 173 Nutnber ot nests per year Nest status 1003 IW4 Total Marked 147 80 227 Marked Abandoned 46 30 76 Marked Cast in plaster 4 0 4 Marked Completed 97 50 147 Marked Completed Failed 83 49 132 Marked Completed Survived 14 1 13 hovering above them. However, we never observed females carrying ants into the air and dropping them a few cm away, as Ammophila gracilis did in our study site (Gaimari and Martins 1996). A butterfly that was flying repeatedly over a nest was also driven away by the resident female; in this case the wasp hovered above the butterfly, approaching and touching it several times until it was driven off. Aerial clashes were common between males and females of B. angulata and R. nasiita. Once, a female of B. angulata entered a nest of R. nasuta while the latter was discarding a prey item not consumed by her larva, and when the Rubrica returned, she found the Bicyrtes and drove her away. A few min- utes afterwards, this female started to ex- cavate her own nest close to that of the Rubrica. Table 2 summarizes the fate of the nests iniHated in 1993 and 1994. Out of 227 nests, 147 were completed; 76 were aban- doned and four were cast in plaster; ex- cluding the latter, then, some 34% of nests were abandoned. Apparent causes of nest abandonment were: excavation in hard soil (n = 8); proximity to ant nests (2); heavy rains (2); female disturbed by a fe- male of R. nasuta (1); and the remainder (63) were due to unknown causes. Of 147 provisioned nests, only 15 pro- duced adult wasps; 100 were destroyed by ants (principally Solenopsis sp.) and ter- mites; 16 experienced immature mortality of unknown nature; 7 pupae died due to flooding of the nesting site; 6 nests were parasitized by Ligyra morio (Diptera: Bom- byliidae; 5 adults emerged in 5 traps); one nest was apparently parasitized by Meto- pia n. sp. nr. sinipalpis (Diptera: Sarco- phagidae: Miltogramminae; emerged in one trap), a major parasite of Ammophila gracilis in the study site (Gaimari and Mar- tins 1996); and two nests were excavated prematurely to document egg placement on prey. Excluding the latter two nests, mortality of completed nests was approx- imately 90%. One entire aggregation of 25 marked nests was extirpated in 1991 by ants and termites at the Ecological Station of the UFMG. DISCUSSION AND CONCLUSIONS The data from this study are only ex- ceeded by those available for B. quadrifas- ciata, obtained by several workers in nu- merous localities in the United States (see Evans 1966 for a review). As Evans (1966) has claimed, a detailed comparison of the behavior of the species of Bicyrtes requires that other species like B. angulata are stud- ied in detail. Following Evans' (1966) sum- mary of features of the ethology of the species of Bicyrtes, we have thus con- firmed some general patterns within the few members of the genus that have been studied, as well as added new information (Tables 2, 3 and 4), though, as mentioned, only a few of the details of nesting and provisioning behavior are uniform across all species. Adults of B. angulata are like all the oth- er studied species in the genus in some 174 Journal of Hymenoptera Research Table 3. Comparison of qualitative features of Bicyrtes nesting biology. aiigtiliifa W)St/Sd I'liric^ala ^imillinui qiiaiirifii^cttila Provisioning progressive mass mass mass delaved Prey paralyzation (partial) yes yes yes yes 7 Prey types* N a 4 (+A) N >4 N5 N5 N Prey families** C P P P CPRLSCy Foraging during morning-afternoon yes ? 7 yes yes Mound (tumulus) near nest entrance yes yes no yes no Nests aggregated*** yes no yes no no Accessory burrow no no yes yes 7 Sleep in accessory burrow no no ? yes 7 Sleep in mixed association yes 7 7 no 7 Citation**** present work C1982 G1982 G1982 El 966 * N = nymphs (followed by instar numbers if specified), A = adults, ( + A) indicates adults rarely taken. ** C = Coreidae (s.l.), P = Pentatomidae, R = Reduviidae, L = Lygaeidae, S = Scutelleridae, Cy = Cydnidae, Rh = Rhopalidae, Py = Pyrrhocoridae. *** Statistics not available; category represents subjective evaluation of authors as to dispersion of nests relative to apparent available space. ****G1982 = Genise 1982, E1966 = Evans 1966, E&M1974 = Evans & Matthews 1974, S&G1992 = Sanchez & Genaro 1992. respects: (1) they are frequent visitors at flowers for nectar (they do not feed upon the prey they capture for their larvae) (2) they are solitary ground-nesting predators (3) prey are immature and /or adult het- eropterans (4) nest closure is maintained while outside the nest (5) foraging is pri- marily in the morning and afternoon. Oth- er features which appear similar to the majority of species studied include: (1) nesting in multispecies aggregations (2) sleeping on vegetation (3) partial paralysis of prey (4) nests with one or two cells (5) low nest angles (generally less than 45°). Other features were atypical, such as: (1) nesting in sandy soil far from water (other species except for B. spiiiosa nest along wa- ter courses) (2) relatively short burrows (less than 10 cm on average) (3) more than two hours to dig a nest. The presence of a tumulus can neither be considered excep- tional nor general, and for other features comparative information is lacking. A few of these points merit further dis- cussion. Despite the general tendency of B. angulata, like other species, to intersperse its nests with those of other digger wasp and bees, it tends to aggregate intraspecif- Table 4. Comparison of quantitative features of Bicyrtcs nesting biologv. iitii^iiiiiiii ilisani t'llrif^tllii ^ittnlhttiil i^iiitiinfii^iiitlii Number of cells per burrow 1-2 3 2-5 1 2-3 Number of prey per cell 7-24 16 3-6 11 8-11 Time to dig cell /nest (min) 146 ± 14 30 60 >60 60-120 Angle of entrance burrow 20°-35° 30°-35° 30°-35° 30° 30°-60° Nest depth (cm) 5.6 ± 0.54 5 8-10 20 7 Nest length, burrow plus cell (cm) 9.6 ± 1.1 12 12 24 8-43 Lengt of cell (mm) 25-26 18-20 25-30 17 20-35 Width of cell (mm) 12-13 10-12 12-15 12 5-27.5 Number of pores in cocoon 5.5 ± 1.95 7 5 ? 7 Egg-adult developmental time (days) 44-375 7 7 ? 49 Number of generations per year 2 7 7 7 7 Citations as for Table 3. Volume 7, Number 2, 1998 175 Table 3. Extended. nlr„h~ lo!< 'iVn/.if.; delayed ? progressive progressive ? ? ? no N N + A N + A N( + A) CPRLSCy PS Rh CPLSPy ? ? yes yes no no no yes variable ? ? variable ? ? ? ? ? ? ? 7 7 El 966 E1966 E&M1974 no no no S&G1992 ic nests within these areas. As in its con- geners, B. angulata maintains an outer clo- sure at all times when the female is not in the nest, but no inner closure is ever made. Prey are Coreidae (sensu lato), in- dicating in our case a possibly high level of local specialization, compared for ex- ample with the Cuban generalist B. spinosa that preys upon several species of five het- eropteran families (Sanchez and Genaro 1992). Additionally, the prey is partially paralyzed and carried in flight with the middle legs, clasped tightly beneath the base of the wasp's abdomen, and is not usually deposited on the ground at any time. The egg is laid in a semierect posi- tion on the mid-ventral line of the first prey item. Our observations established that B. angulata is a progressive provision- Table 4. Extended. :y',ti.lh~ h;lu-il< .,.,„;,„„ ,.,„„„, 1-2 3-11 60-120 2-5 10-23 7 1 7 7 1 7 7 45° 4-8 20-30 20-30 45°-60° 8-13.5 12.5-18.5 7 20°-45° 8.5-10.5 12-14 7 45° 4.5-8.5 9.9 ± 2.6 7 8-12 7 10-12 7 ? 7 7 40-42 7 7 7.6 ± 1.54 7 ' -) 7 7 er like B. spinosa and B. cingulata, though mass provisioning and delayed provision- ing are found in other species. Among the novel data for B. angulata is the huge variability in the time of prepu- pal dormancy within the same brood. This variability (sometimes called "parsivoltin- ism"; Torchio and Tepedino 1982) is scarce- ly documented for Nearctic and Neotrop- ical species of solitary wasps and bees (see Evans 1966; Stephen et al. 1969; Evans and West-Eberhard 1970; Torchio and Tepedi- no 1982; Roubik 1989; Wcislo and Cane 1996). A comparable variability was also found for other species of solitary wasps and bees in the same study site or in other sites in the Ecological Station of the UFMG (Martins et al. 1996; Almeida et al. 1997; Pimenta and Martins unpublished data). One of the possible interpretations for such a pattern is given by theories of risk spreading of reproductive effort (or "bet- hedging") in unpredictable environments (Danks 1987; Tauber et al. 1986). The pos- sibility that parsivoltinism may help avoid predation or parasitism in some way seems reasonable, but we cannot confi- dently identify the precise mechanism of selection for such an adaptation in the present case. We do believe, however, that we can exclude variability in the occur- rence of rains (as in deserts where similar insect life cycles occur, e.g. Hanski 1988), because our study site is strikingly season- al and predictable in the occurrence of rains (see Martins and Antonini 1994) and dormancy seems to not correlate with rains (R. P. Martins unpublished data). There is an apparent paradox in the data on number of marked females versus number of nests made (see Fig. 5). Many marked females visiting flowers at the study site did not nest there, so the total number of females exceeded the number of nests in 7 of the 14 months when fe- males were present. Why are there so many active females that are apparently not nesting? This same phenomenon oc- curs in R. nasuta (Pimenta and Martins un- 176 Journal of Hymenoptera Research published data), and for that species, prey are seasonally distributed within the year, and the hypothesis is that females should wait for a burst of prey availability to make nests. It is likely that prey abun- dance for B. angulata also exhibits seasonal variation, because many plant-feeding in- sects in this environment have strongly seasonal life cycles, and a similar expla- nation may apply here. It is alternatively possible, though questionable, that these excess females were nesting elsewhere but all foraging in a limited area together; i.e., the females sampled at the flowers repre- sented the combined population of several nesting areas. However, this was not the only patch of suitable flowers in the area, and there is no obvious reason why the wasps would be concentrated in this par- ticular patch. Because we found no relationship be- tween the number of nesting females and the average number of nests per female, we suggest that crowding has no effect on the number of nests made in the aggre- gation, despite the preference to nest in a somewhat limited area. It is also possible that females made only one or two nests in the aggregation and then dispersed to other areas, given that a female can live for up to 84 days, yet few were in resi- dence this long. That is, rather than inter- preting the small number of nests per fe- male as a result of high mortality, there may be some undetected nesting activity outside of the known aggregation area; further work with marked females might help resolve this. In either case, the avail- ability of adequate patches (whether ag- gregations form in them or not) may be of primary importance in the nesting dynam- ics of this species. In addition to other forces that could potentially influence nesting dynamics is the frequent extirpa- tion of nests or entire aggregations by ter- mites and ants, and the high frequency of nest desertion, due to several causes. Therefore, the patchiness of the environ- ment and the dynamics within the aggre- gations are likely to be crucial in the pop- ulation dynamics of this species. ACKNOWLEDGMENTS Servio Tiilio P. do Amarante identified Bicyrtes an- gulata, B. discisa, B. tricolorata, Rubrica nasuta, and Pri- onyx fervens. Padre J. S. Moure identified Megachile ncoxantJwfJtera and Centrii aenea. Arnold S. Menke identified Amincphila gracilis. ]. A. M. Fernandes iden- tified McgalotcniKS sp. and Apidaunis sp. The late Her- mogenes F. Leitao Filho identified Walthetia amcncana and ]. F. Macedo Panicum maximum. We would also like to thank William T. Wcislo for useful comments and criticisms. The Brazilian Conselho Nacional de Desenvolvimento Cientifico (CNPq) and Fundai^ao de Amparo a Pesquisa de Minas Gerais (FAPEMIG) con- ceded grants, the CNPq also provided research schol- arships to the senior authors (R.P.M. and L.A.S.), a Visifing Researcher Fellowship (301019/96-7 RN) to the junior author (D.Y.), and the U. S. Fish and Wild- life Service provided some logistical support. This study is a contribution of the Program in Ecology, Conservation and Wildlife Management (ECMVS) of the Universidade Federal de Minas Gerais, Belo Ho- rizonte, MG, Brazil. LITERATURE CITED Almeida, D. A. O., R. P. Marfins and M. L. T. Bus- chini. 1997. Behavior and nesting dynamics of the Neotropical cavity-nesting specialist bee MegachUe assumptinnis Schrottky, with compari- sons to the Nearctic Megadule brcvis Say (Hy- menoptera: Megachilidae). ]ouriud of Hymenoptera Research 6: 344-352. Bohart, R. M and A. S. Menke. 1976. Sphecid Wasps of the World, University of California Press, Berke- ley, CA. 695 pp. Danks, H. V. 1987. Insect Dormancy: an ecological perspective. Biological Survey of Canada Mono- graph Series, No. 1. Ottawa, 439 pp. Evans, H. E. 1966. The comparative ethology and evolu- tion of the sand wasps. Harvard University Press, Cambridge, MA, 526 pp. Evans, H. E. and M. J. West-Eberhard. 1970. The u'asps. University of Michigan Press, Ann Arbor, Ml. 265 pp. Evans, H. E. and R. W. Matthews. 1974. Observations on the nest behavior of South American sand wasps (Hymenoptera). Biotropica 6 (2): 130-134. Gaimari, S. D. and R. P. Martins. 1996. Nesting be- havior and nest distributions of Ammophila grac- dis Lepelefier (Hymenoptera; Sphecidae) in Bra- zil. Journal of Hymenoptera Research 5: 240-248. Genise, J. F. 1979. Comportamiento de nidificacion de Bic\/rtes variegata (Ol.) e Bicyrtes discisa (Tasch.) (Hymenoptera: Sphecidae). Rei'ista da Sociedad Entomologica Argentina 38 (1-4): 123-126. Volume 7, Number 2, 1998 177 Genise, J. F. 1982. Estudios sobre el comportamiento de Bembicini neotropicales. I. Bici/rtes simUlima (Smith) y Bkyrtes discisa (Tasch) (Hymenoptera: Sphecidae) con una revision de los tipos de aprovisionamiento em Bembicini. Physis (Buenos Aires) Secc. C, 40 (99): 93-99. Hanski, I. 1988. Four kinds of extra long diapause in insects: a review of theory and observations. An- nales Zootogici Fennici 25: 37-53. Macedo, J. F. 1995. A importancia de Walthcria indica (Sterculiaceae) para abelhas e vespas. M. Sc. Dis- sertation, Universidade Federal de Minas Gerais, Belo Horizonte, M.G., Brazil. 74 pp. Macedo, J. F. and R. P. Martins. 1998. Potencial da erva daninha Walthcria ainericana (Sterculiaceae) no manejo integrado de pragas e polinizadores: visitas de abelhas e vespas. Anais da Soctedade En- tomologica do Brasil 27: 29-40. Martin, P. and Bateson, P. 1986. Measuring Behavior. Cambridge University Press, Cambridge, U.K. 200 pp. Martins, R. P. 1991. Biologia e Comportamento de Comunidades de Vespas Escavadoras (Hyme- noptera: Aculeata). Ph.D. Dissertation, Universi- dade Estadual de Campinas, Campinas, S.P., Bra- zil. 116 pp. Martins, R. P. and D. A. O. Almeida 1994. Is the bee Megaehile assumptionis a cavity-nesting specialist? lournal of Insect Behavior 7 (5): 754-765. Martins, R. P. and Y. Antonini. 1994. The biology of Diadasina distincta Holmberg 1803, (Hymenop- tera: Anthophoridae). Proceedings of the Entomo- logical Society of Washington 96 (3): 553-560. Martins, R. P., F. G. Guimaraes and C. M. Dias. 1996. Nesting biology of Ptilothnx plumata Smith, with a comparison to the other species in the genus (Hymenoptera: Anthophoridae). Journal of the Kansas Entomological Society 69 (1): 9-16. Nascimento, P. T. R. and W. L. Overall. 1980. Catal- ogo da Cole^ao Entomologica do Museu Goeldi. Hymenoptera: Sphecidae. Boletim do Museu Par- tiensc Eniilio Goeldi, Zoologia 99: 1-14. Roubik, D. W. 1989. The Ecology and Natural History of Tropical Bees. Cambridge Tropical Biological Series. Cambridge University Press, Cambridge, U.K. 514 pp. Sanchez, C. S. and J. A. Genaro. 1992. Observaciones sobre la conduta de nidificacion en esfecidos de Cuba (Hymenoptera). Bicyrtes spinosa (Fabr.) Poy- eana 411: 1-8. Stephen, W. P., G. E. Bohart and P. F. Torchio. 1969. The hiology and external morphology of bees. Agri- cultural Experimental Station, Oregon State Uni- versity, Corvallis, OR. 140 pp. Tauber, M. J., C. A. Tauber and S. Masaki. 1986. Sea- sonal Adaptations of Insects. Oxford Uruversity Press, NY. 411 pp. Torchio, P. F., and V. J. Tepedino. 1982. Parsivoltin- ism in three species of Osmia bees. Psyche 89: 221-238. Wcislo, W. T., and J. H. Cane. 1996. Floral resource utilization by solitary bees (Hymenoptera: Apo- idea), and exploitation of their stored food by natural enemies. Annual Review of Entonwlo 0.2). The sexes did not differ in mean length of the flagella (in mm, females: 3.44 ± 0.46; males: 3.28 ± 0.39; Mann-Whitney U = 348, P > 0.8 ) (Figure 1, bottom).' Males had significantly smaller mean scape length than females (in mm, females: 0.55 ± 0.053; males: 0.43 ± 0.04; Mann-Whit- ney U = 461, P = 0.02; Figure 1, top). Within species, non-parasitic females more frequently had a larger body size than males (two-tailed sign test, P < 0.05, T = 22, N = 26), and had longer scapes (two-tailed sign test, P < 0.05, T = 23, N = 23), but females did not have longer fla- gella (two-tailed sign test, 0.1 > P > 0.05, T = 17, N = 25) (sample sizes differ be- cause ties were eliminated; Conover 1971). Regression equations for non-parasitic males versus females were not significant- ly different for scape or flagellum length as a function of body size (P > 0.05, com- paring y-intercepts or slopes). Male and female parasites did not differ in body size, nor in lengths of scapes and flagella. 2,0 z LU LU < 1 2 3 BODY SIZE, mm 12 3 BODY SIZE, mm Fig. 1. Scape (top) and flagellum (bottom) length of spheciforme wasps as a function of body size (inter- tegular distance). Open circle = non-parasitic female; open triangle = non-parasitic male; closed circle = parasitic female; closed triangle = parasitic male. DISCUSSION Among non-parasitic spheciforme wasps females often are larger than males, and have longer scapes. Small samples pre- clude statistical analyses, but parasitic fe- males are not conspicuously larger than conspecific males, and have similarly sized scapes and flagella. In contrast, a study of 114 bee species showed that 1) female non- parasitic bees of a given body size have 180 Journal of Hymenoptera Research longer scapes but shorter flagella than con- specific males, and 2) for parasitic bees, scapes and flagella are siniilar in size, on average, between males and females of a species (Wcislo 1995; also Miiller 1872). Since bees are closely related to, and prob- ably arose from within spheciforme wasps (e.g., Alexander 1992), the sexual differ- ences are probably derived among bees. Relative to non-parasitic females, female parasites may be more "male-like" in their search behavior, but pertinent data are scarce (refs. in Wcislo 1995). If substanti- ated, these behavioral differences can help explain similarities in sensory structures among males of parasitic and non-parasitic species and females of parasitic species, which differ from non-parasitic females. Antennae have olfactory, gustatory, and tactile sensory receptors; they are used like calipers during nest construction; and they can be used to drum, tap, or stroke parts of the female's body (refs. in Wcislo 1995). The scape of some male spheciforme wasps is broadly expanded (e.g., Dinetus), like males of a parasitic bee, Doeringiella (Roig-Alsina 1989); these expansions may house glands for use during courtship and mating, as known for other Hymenoptera (Isidoro et al. 1996). The functional mor- phology of antennae has been studied for relatively few species, and typically only for one or two sensory modalities. Anten- nae have multiple functions, highlighting the need for more detailed studies relating behavior to an animal's sensory world ("utmvelt"), as pointed out long ago by von Uexkiill (1934). ACKNOWLEDGMENTS I am grateful to Richard Hoebeke for unrestricted access to the Cornell University Insect Collection; and to Arnold Menke for arranging the loan of specimens of two important species from the National Museum of Natural History of the Smithsonian Institution. Hermogenes Fernandez helped with some measure- ments. An anonymous reviewer provided very help- ful comments. General research funds from the Smithsonian Tropical Research Institute are gratefully acknowledged. LITERATURE CITED Alcock, J. and J. P. Alcock. 1983. Male behaviour in two bumblebees, Bonihus nevadeniiis aurkimms and B. grisciccllis (Hymenoptera: Apidae). jcurnal cf Zoology. London 200: 561-570. Alexander, B. A. 1992. An exploratory analysis of cla- distic relationships within the superfamily Apo- idea, with special reference to sphecid wasps (Hymenoptera). lournal of Hymenoptera Research 1: 25-61. Bohart, R. M. and A. S. Menke. 1976. Spheaci Wasps of the Worhl, A Generic Revision. Berkely: Univer- sity of California Press, i-ix + 1-695 pp. Brothers, D. J. and ]. M. Carpenter. 1993. Phylogeny of Aculeata: Chrysidoidea and Vespoidea (Hy- menoptera). ]ourunI of Hymenoptera Research 2: 227-304. Cederberg, B., B. G. Svensson, G. Bergstrom, M. Ap- pelgren, and I. Groth. 1984. Male marking pher- omones in north European cuckoo bumble bees, Psitliyrus (Hymenoptera, Apidae). Nova Acta Re- giae Societatis Scientiarum Upsaliensis, Serie V:C 3: 161-166. Cervo, R. and F. R. Dani. 1996. Social parasitism and its evolution in Polistes, pp. 98-1 12. In: Turillazzi, S. and M. ]. West-Eberhard (eds.). Natural History and Evolution of Paper Wasps New York: Oxford University Press, i-xiv + 400 pp. Conover, W. J. 1971. Practical Nonparametric Statistics. New York: John Wiley & Sons, Inc., i-x + 462 pp. Dusenbury, D. B. 1992. Sensory Ecology. New York: W.H. Freeman & Co. i-xx + 558 pp. Harvey, P. H. and M. D. Pagel. 1991. The Comparative Method in Evolutionary Biology. New York: Oxford University Press, i-iv + 239 pp. Holldobler, B. and E. O. Wilson. 1990. The Ants. Cam- bridge: Harvard University Press, pp. i-xii + 732 pp. Isidoro, N. F., F. Bin, S. Colazza, and S. B. Vinson. 1996. Morphology of antennal gustatory sensilla and glands in some parasitoid Hymenoptera with hypothesis on their role in sex and host rec- ognition. Journal of Hymenoptera Research 5: 206- 239. Menke, A. S. 1997. Family-group names in Sphecidae (Hymenoptera: Apoidea). Journal of Hymenoptera Research 6: 243-255. Miiller, H. 1872. Anwendung der Darwinschen Lehre auf Bienen. Verhandlungen des naturhistorischen Vereines der preiissischen Rheinlande uiul Westphal- ens 29: 1-96. Roig-Alsina, A. 1989. A revision of the bee genus Doeringiella (Hymenoptera, Anthophoridae, No- madinae). University of Kansas Science Bulletin 53: 576-621. Uexkiill, von J. 1934. A stroll through the worlds of animals and men, pp. 5-80 [reprinted and trans- Volume 7, Number 2, 1998 181 lated, 1957]. In: Schiller, C. H. (ed.) Instinctive Be- havior: The Development of a Modern Concept. Lon- don: Methuen & Co. Ltd. Wcislo, W. T. 1987. The roles of seasonality, host syn- chrony, and behaviour in the evolutions and dis- tributions of nest parasites in Hymenoptera (In- secta), with special reference to bees (Apoidea). Biological Reviezva of the Cambridge Philosophical So- ciety 62: 415-443. Wcislo, W. T. 1989. Behavioral environments and evolutionary change. Annual of Ecology & System- atics 20: 137-169. Wcislo, W. T. 1995. Sensilla numbers and antennal morphology of parasitic and non-parasitic bees (Hymenoptera: Apoidea). International journal of Insect Morphology & Embryology 24: 63-81. Wcislo, W. T. and J. H. Cane. 1996. Resource utiliza- tion by solitary bees (Hymenoptera: Apoidea), and exploitation by their natural enemies. Annual Revierv of Entomology 41: 257-286. Wilkinson, L. 1988. SYSTAT. Evanston IL: SYSTAT Inc. APPENDIX List of spheciforme wasp species from which an- tennal measurements were taken. * = parasitic taxon; ? = taxon is probably parasitic, but behavioral data are unavailable. Nomenclature follows Bohart and Menke (1976), as modified by Menke (1997). AMPULICINAE AMPULICINI Ampulex compressa (Fabricius) DOLICHURINI Dolichurus cornicuhis (Spinola) ASTATINAE DINETINI Dinetus pictus (Fabricius) ASTATINI Astata mexicana Cresson PHILANTHINAE EREMIASPHECIINI Eremiasphecium schmiedeknechtii Kohl APHILANTHOPSINl Aphilanthops friguius (Smith) PHILANTHINI Philanthus solivagus Say CERCERINI Cerceris frontata Say PEMPHREDONINAE PSENINI Psenulus pallipes (Panzer) PEMPHREDONINI Stigmus americanus Packard SPHECINAE AMMOPHILINI Ammoplula polita Cresson SCELIPHRONINI Podium nifipes Fabricius SPHECINl Sphex dorsalis ( = singularis} Smith BEMBICINAE HELIOCAUSINI Heliocausus larroides (Spinola) MELLININI Mellinus arvensts (Linnaeus) STIZINI Bembicinus U'heeleri Krombein & Willink *Stizoides unicinctus { = renicinctus) (Say) GORYTINI Gorytes simillimus Smith *NYSSONINI 'Nysson (Epinysson) mellipes (Cresson) *7Synnevrus aequalis (Patton) 'Nyssoii simplicicornis Fox BEMBICINI Bembix texana Cresson CRABRONINAE LARRINI Larra bicolor Fabricius MISCOPHINl Solierella plenoculoides (Fox) PALARINI Palarus latifrons Kohl TRYPOXYLONINl Trypoxylon lactitarse Saussure SCAPHEUTINI Scapheutes brasilianus Handlirsch CRABRONINI Crabro cribrellifer (Packard) OXYBELINI Oxybclus emarginatus Say ]. HYM. RES. Vol. 7(2), 1998, pp. 182-208 Systematics of Costa Rican Meteonis (Hymenoptera: Braconidae: Meteorinae) Species Lacking a Dorsope Nina M. Zitani, Scott R. Shaw, and Daniel H. Janzen (NMZ, SRS) Department of Renewable Resources, University of Wyoming, Laramie, Wyoming, 82071-3354, USA, ninaz@uwyo.edu, braconid@uwyo.edu; (DHJ) Department of Biology, University of Pennsylvania, Philadelphia, Pennsylvania 19104, USA, djanzen@sas.upenn.edu Abstract. — This study of the genus Meteonis (Hymenoptera: Braconidae: Meteorinae) treats the 19 known Costa Rican species that lack a dorsope (a pit on the dorsal surface of the first metasomal tergite). Fourteen new species all attributed to Zitani are described and illustrated: M. alejandro- masisi, M. camilocamargoi, M. coffeatus, M. corniculatus, M. desmiae, M. dos, M. mariamartae, M. megalops, M. micrommatus, M. pseudodimidiatus, M. rogerblancoi, M. sterictae, M. uno, and M. yamijuanum. An identification key to 19 species is provided, including the previously described species: M. cougregatus Muesebeck, M. dimidiatus (Cresson), M. lapln/gnme Viereck, M. papiliovorus Zitani, and M. rubens (Nees). Biological information, host associations, and cocoon- forming behavior are included for M. cougregatus, M. rubens and the new species M. alejandromasisi, M. camUocamargoi, M. desmiae and M. sterictae. This work provides the first record of Meteorus attacking Megalopygidae and Hesperiidae including Chiomara asi/chis (StoU), Pyrgus sp., Stnph\/lus azteca (Scudder), and Staphylus sp. Other new host records include the pyralids Desmia tages (Cram.), Omiodes stigmosalis (Warr.), and Stericta albifasciata (Druce), and the sphingids Manducn sexta (L.), and Unzela japix (Cram.). Meteorus cougregatus, M. dimidiatus, and M. rubens are recorded for the first time in Costa Rica. The Meteorinae is a moderately large, cosmopolitan subfamily with at least 183 described species worldwide. The subfam- ily consists of solitary or gregarious koino- biont endoparasitoids of larval Coleoptera and Lepidoptera (Muesebeck 1923; van Achterberg 1979; Huddleston 1980). The genus Meteorus Haliday is the most di- verse and frequently encountered genus of the Meteorinae {seiisii Shaw 1985, 1995; Maeto 1990b; Shaw and Huddleston 1991), and is the sister-group of the subfamily Euphorinae Foerster s.s. (Shaw 1985, 1988). For a detailed review of the biology of the Meteorinae see Shaw (1997), Zitani et al. (1997), Shaw (1995), and Shaw and Huddleston (1991). There has been some revisionary work on the Meteorus species that occur in North America (Muesebeck 1923), Africa (Nixon 1943), the western Palearctic re- gion (Huddleston 1980), and Australia and New Guinea (Huddleston 1983), but none for the Neotropical region. Although six species of Meteorus have been de- scribed from various localities in the neo- tropics (Shenefelt 1969), and the North American species M. laphygmae Viereck has been recorded from Costa Rica (Shene- felt 1969), the first new species from Costa Rica was described only recently (Zitani et al. 1997). In order to keep the present work at a reasonable length only those species lack- ing a dorsope are treated here. A dorsope is a pit lying anterior to the spiracles on the dorsal surface of the first metasomal tergite; these pits occur in a pair (Fig. 1). Species lacking a dorsope (Figs. 3, 5) rep- resent the majority of the material from the examined Costa Rican collections. This character has been used to separate spe- Volume 7, Number 2, 1998 183 cies of Meteorus by several previous au- thors (Muesebeck 1923; Huddleston 1980; Maeto 1989, 1990a). This group of species was also chosen because considerable bi- ological information was available for 6 included species. METHODS Species treated in this work can be iden- tified as members of the subfamily Meteor- inae using the keys of Shaw (1995) or Shaw and Huddleston (1991). Specimens can be determined as Meteorus using the key of Shaw (1997) or Marsh et al. (1987). Approximately 1,200 specimens of Me- teorus were examined. The following in- stitutions provided specimens for this study: ANS: The Academy of Natural Sci- ences, Philadelphia, PA, USA. INBio: Instituto Nacional de Biodiver- sidad, Santo Domingo de Here- dia, Costa Rica. PAS: The Polish Academy of Sciences, Warszawa, Poland. RMSEL: Rocky Mountain Systematic Laboratory, University of Wyo- ming, Laramie, WY, USA. USNM: United States National Museum, Smithsonian Institution, Wash- ington, DC. UCR: Museo de Insectos, Universidad de Costa Rica, San Pedro, San Jose, Costa Rica. Specimens reared by D.H. Janzen and W. Hallwachs at the Area de Conserva- cion Guanacaste (ACG), Guanacaste Prov- ince, Costa Rica, are identified by a vouch- er number with "SRNP" in the middle. Each Janzen and Hallwachs rearing voucher number is for a single host larva. Information about the rearings can also be found at http:/ /janzen.sas.upenn.edu/ in- dex.html. The majority of the specimens were pre- pared and mounted at RMSEL using the technique of Sharkey (1988). Specimen la- bel data were recorded exactly as they ap- pear on the collection label to avoid mis- interpreting data. However, when avail- able, collection site information is added in brackets []. Authorship of all new spe- cies is attributed to the senior author (NMZ). TAXONOMIC CHARACTERS AND TERMINOLOGY The taxonomic characters used here fol- low the work of Huddleston (1980, 1983), Maeto (1989, 1990a), Shaw (1985), and Zi- tani et al. (1997). Microsculpture terminol- ogy follows that of Harris (1979). Morpho- logical terminology follows Wharton et al. (1997). As noted by Huddleston (1980), many characters of the head are important for distinguishing one species from another, especially the following: the size and de- gree of convergence of the eyes, width of the face in relation to the height, and the amount of twist of the mandibles. In this study, eye size was determined by the length of the eye, in anterior view, relative to the length of the head in anterior view. Eyes were designated as 'small' if the head height was equal to or greater than 1.6X the eye height in anterior view, and 'large' if the head height was less than 1.6X the eye height in anterior view. Eye convergence was designated as 'strongly convergent', 'convergent', or 'nearly par- allel'. Eyes were called 'strongly conver- gent' if the maximum /minimum face width ratio was 1.7 or greater, 'conver- gent' if the maximum /minimum face width ratio was 1.4-1.7, and 'nearly par- allel' if the maximum/minimum face width ratio was less than 1.4. The amount of twist of the mandibles can sometimes be difficult to determine. An untwisted mandible, in anterior view, is flat and has two visible teeth. In a strongly twisted mandible it appears as though the ventral tooth has rotated to a position posterior to the dorsal tooth. Thus the mandible, in anterior view, has only one apparent tooth, and the apical 184 Journal of Hymenoptera Research portion appears twisted (Fig. 9). Also, mandibles that are untwisted are usually larger (longer and broader in anterior view) than strongly twisted mandibles. Moderately twisted mandibles, where the ventral tooth has only partially rotated, are the most difficult to determine but are present in only two species, Meteorus cof- featiis and Meteorus micrommatus (Fig. 10). Moderately twisted mandibles are also longer, and broader at the base, compared to strongly twisted mandibles. The number of flagellomeres and size of the ocelli usually vary only slightly within a species, and these characters, therefore, are also useful (Huddleston 1980). In this text, the diameter of a lateral ocellus is ex- pressed relative to the length of the ocell- ocular distance (distance from edge of lat- eral ocellus to border of compound eye). The acronym OCD is used for ocellar di- ameter, and OCOD for ocell-ocular dis- tance. The term 'small ocelli' is used when the OCOD is greater than or equal to 1.5X OCD, and the term 'large ocelli' when the OCOD is less than 1.5X OCD. The width of the apical flagellomere was measured as the width of the base of the flagello- mere. Wing venation characters generally are not stable within species and therefore should not be used as diagnostic charac- ters (Huddleston 1980). For example, the position of the forewing vein m-cu (whether it is antefurcal, interstitial, or postfurcal) often exists in two of these states within a species. One exception may be the shape of the forewing second sub- marginal cell, which appears to be stable in some species (e.g., M. congregatus, M. papiliovorus). Muesebeck (1923) and Huddleston (1980) have pointed out the importance of: 1) the presence or absence of a dorsope and 2) whether or not the ventral borders of the tergite are joined. Huddleston (1980) noted that when a dorsope is pres- ent it is distinct (Fig. 1), even if it is small, and that the ventral borders of the first ter- gite are never joined beneath, and usually widely separated (Fig. 2). In specimens where the dorsope is absent (Figs. 3, 5), the ventral borders are always touching or nearly touching (Figs. 4, 6), partially fused (Fig. 7) or completely fused (Fig. 8). The length of the ovipositor was mea- sured from the base of the hypopygium to the tip of the ovipositor and was ex- pressed as relative to the length of the first tergite. The term 'long ovipositor' was used when the length of the ovipositor was equal to or greater than 2.0 X the length of the first tergite, and the term 'short ovipositor' was used when the length of the ovipositor was less than 2.0 X the length of the first tergite. As noted by Huddleston (1980), color and body size are among the least stable characters and often vary within a species. The color of a specimen can vary depend- ing on how it was preserved, how long it remained in alcohol before it was pre- pared, and exposure to light. Although the colors 'white' and 'yellow' are used in this text, often the specimens described as hav- ing these colors actually have a clear in- tegument, and it is the internal tissues that provide the color. Overall, color is some- what generalized in the species descrip- tions and should not be used without oth- er diagnostic characters. Specimens can have a variety of meta- somal positions, therefore the body length is a combined measurement of the length from the head to the end of propodeum, added to the length from the base of the first tergite to the end of the metasoma (not including the ovipositor). KEY TO COSTA RICAN METEORUS FEMALES WITHOUT A DORSOPE 1. Mandible strongly twisted (Fig. 9); propodeum rugose (Figs. 15-16) or areolate-rugose (Fig. 17), but never carinate, with a distinct median and transverse carina creating very large, defined areolae (Fig. 18) 2 Volume 7, Number 2, 1998 185 Figs. 1-6. 1, 3, 5. First metasonial Ifrgite, dorsal view. 1, Metcorus sp., dorsope present; 3, Mcteorus rogerblancoi, dorsope absent, surface smooth and rounded; 5, Meteorus pseudodimidiatus, dorsope absent, surface not com- pletely smooth and tergite flattened laterally. 2, 4, 6. First metasomal tergite, ventral view. 2, Meteorus sp., (dorsope present) ventral borders widely separated; 4, Meteorus rogerblancoi, (dorsope absent) ventral borders of first tergite joined completely along basal Vi of segment and suture apparent; 6, Meteorus diwidmlus. (dorsope absent) ventral borders touching for only short distance apically. 2(1). Mandible moderately twisted (Fig. 10) or not twisted, broad and flat at base; propo- deum carinate, with a distinct median and transverse carina creating very large, de- fined areolae (Fig. 18), or rugose (Figs. 15-16) 15 Wings very dark, infused with blackish pigment, and second submarginal cell of forewing strongly narrowed anteriorly; forewing vein 3RSa usually much shorter than r, or some- 186 Journal of Hymenoptera Research Figs. 7-12. 7-8. Metcorus con^rc^atiiff, first metasomal tergite, ventral view (dorsope absent). 7, ventral borders partially fused (separated basally, fused apically); 8, ventral borders fused, no indication of suture. 9-10. Anterior view of head. 9, Mctecruf rogcrbhiiicoi, mandibles strongly twisted, with one visible tooth; 10, Mcfctinis coffciitiis, mandibles moderately twisted, with ventral tooth partially visible. 11-12. Mesoscutum, dorsal view, showing notauli. 11, Mctcorus ivgcrblancoi, notauli not distinct, broad, rugose, and converge posteriorly in a rugose area; mesonotal lobes not well-defined; 12, Meteorus megalops, notauli distinct, linear, foveolate ante- riorly, and converge posteriorly in a rugose area; mesonotal lobes well-defined. times nearly equal; eyes small and nearly parallel; notauli distinct and linear, mesonotal lobes well-defined (Fig. 12) 3 Wings not dark, at most with brownish pigment; second submarginal cell of forewing not strongly narrowed anteriorly; forewing vein r usually much shorter than, or some- times equal to, 3RSa; eyes small or large, convergent or nearly parallel; notauli distinct, linear and mesonotal lobes well-defined (Fig. 12), or not (Fig. 11) 4 Volume 7, Number 2, 1998 187 3(2). Head height 2.0x eye height; occipital carina incomplete, widely separated medially (Fig. 23); first tergite with costae parallel, usually yellowish in color and only slightly darker apically, ventral borders fused (Figs. 7-8); highly gregarious (about 25-250 wasps per larva), attacking Sphingidae (Figs. 25-27) M. congregatus Muesebeck Eyes somewhat larger, head height 1.7x eye height; occipital carina not complete but with only a small separation medially (Fig. 21); first tergite with costae convergent posteriorly, usually yellow basally, nearly black apically, ventral borders joined com- pletely but not fused (Fig. 4); metasoma nearly black dorsally; solitary or gregarious, attacking Papilionidae (1-5 wasps per larva) (Fig. 28) M. papiliovonis Zitani 4(2). Ventral borders of first tergite not joined completely along basal Vi of segment (Fig. 6); notauli distinct, linear, and mesonotal lobes well-defined (Fig. 12), or not (Fig. 11) 5 Ventral borders of first tergite joined completely along basal Vi of segment (Fig. 4) and notauli not distinct, broad, mesonotal lobes not well-defined (Fig. 11) 7 5(4). Notauli not distinct, broad, and mesonotal lobes not well-defined (Fig. 11) M. rubens (Nees) Notauli distinct, linear, and mesonotal lobes well-defined (Fig. 12) 6 6(5). Ovipositor long, equal to, or greater than 2.0 X length of first tergite; malar space about equal to basal width of mandible M. dimidiatus (Cresson) Ovipositor short, less than 2.0x length of first tergite; malar space longer than basal width of mandible M. pseudodimidiatus Zitani, new species 7(4). Ocelli small (OCOD equal to, or greater than 1.5X OCD) 8 Ocelli large (OCOD less than 1.5x OCD) 9 8(7). Occipital carina complete (Fig. 19), hind coxa rugulose . . . M. rogerblancoi Zitani, new species Occipital carina not complete, poorly defined medially (Fig. 21), hind coxa polished, punctate M. uno Zitani, new species 9(7). Ovipositor short, less than 2.0 X length of first tergite; tarsal claws with a well-defined lobe (Fig. 22) or without (Figs. 20, 24) 10 Ovipositor long, equal to, or greater than 2.0 x length of first tergite, and tarsal claws with a well-defined lobe (Fig. 22) 12 10(9). Occipital carina not complete, poorly defined medially (Fig. 21); longitudinal costae of first tergite parallel M. alejandromasisi Zitani, new species Occipital carina complete (Fig. 19); longitudinal costae of first tergite convergent pos- teriorly 11 11(10). Malar space length 0.5x length of mandible basally; frons with a median tubercle anterior to median ocellus; sternaulus foveolate, not broad (Fig. 14); hind coxa ru- gulose, with distinct but very fine carinae M. dos Zitani, new species Malar space greater than 0.5x mandible width basally; sternaulus rugose, broad, and long (Fig. 13); hind coxa rugulose-punctate M. laphygmae Viereck 12(9). First tergite laterally flattened (Fig. 5), and dorsal surface rugulose medially (costae not distinctly linear); eyes not so large and protuberant laterally . . . M. sterictae Zitani, new species First tergite not laterally flattened (Fig. 3), with longitudinal costae distinct, parallel, but sometimes convergent posteriorly and, eyes very large, strongly protuberant ... 13 13(12). Eyes convergent; clypeus somewhat flattened; occipital carina complete (Fig. 19); first tergite with costae somewhat convergent posteriorly; vertex not concave behind eyes M. desmiae Zitani, new species Eyes nearly parallel; clypeus not flattened; occipital carina not complete, poorly de- fined medially (Fig. 21); first tergite with costae parallel; vertex strongly concave be- hind eyes M. camilocamargoi Zitani, new species 14(1). Mandibles moderately twisted (Fig. 10) 15 Mandibles not twisted 16 188 Journal of Hymenoptera Research 15(14). Malar space very short, 0.3 x rnandible width basally, eyes nearly touching base of mandible; eyes strongly convergent M. coffeatus Zitani, new species - Malar space very long, 1.1 x mandible width basally; eyes parallel . . . M. micrommattis Zitani, new species 16(14). Propodeum rugose, scutellar disc coarsely rugose and raised into a sharp point; eyes protuberant M. cortticulattts Zitani, new species - Propodeum carinate, with a distinct median and transverse carina creating very large, defined areolae (Fig. 18); scutellar disc rounded and smooth; eyes protuberant or not 17 17(16). Eyes huge, bulging, and strongly convergent; antennae with 25 flagellomeres; large species, body length greater than 5.0 mm M. megalops Zitani, new species - Eyes large but not huge and bulging; 19-20 flagellomeres, very small species, body length less than 4.0 mm 18 18(17). Notauli linear and converging posteriorly in a distinct V-shape; antennae with two white bands on F1-F3, and F9-F16; first tergite with very faint costae or none at all M. mariamartae Zitani, new species Notauli distinctly linear anteriorly but converging posteriorly in a rugose area (Fig. 12); antennae without white bands; first tergite with distinct lateral costae M. yamijuanum Zitani, new species Genus Meteorus Haliday Meteorus Haliday 1835: 24. Diagnosis for species lacking a dorsope: Body color: variations of w^hite, yellow, orange, reddish-brown, brown, dark brown, or nearly black. Body length: 2.5- 6.0mm. Head: maxillary palpi 6-segment- ed; labial palpi 3-segmented; antenna with 19-34 flagellomeres, antennae filiform, apical flagellomere pointed; head 1.1-1.4X wider than high, head height 1.3-2. OX eye height; eyes nearly parallel, convergent, or strongly convergent in anterior view; mandible strongly twisted, moderately twisted, or not twisted; occipital carina complete or incomplete. Mesosoma: no- tauli and sternaulus present; epicnemial carina present. Legs: tarsal claw lobed or simple. Wings: vein m-cu antefurcal, in- terstitial, or postfurcal. Metasoma: ventral borders of first tergite fused completely or partially, joined completely, but not fused, along basal Vi of segment, or, meeting for only a short distance; dorsal surface of first tergite usually with costae posterior- ly; metasoma polished dorsally, smooth and shining; ovipositor 1. 3-3.6 X longer than first tergite, usually straight but sometimes curved. Meteorus alejandromasisi Zitani, new species (Fig. 21) Holotype female. — Body color: Yellow ex- cept antennae light brown; metanotum brown; propodeum brown dorsally and posteriorly; hind leg with coxa and femur brown apically, tibia and tarsus brown; first tergite brown apically; T 2-1-3 brown laterally, T 4 brown. Body length: 3.2 mm. Head: antenna with 28 flagellomeres; fla- gellar length /width ratios as follows: Fl = 2.5, F2 = 2.5, F3 = 2.5, F24 = 3.0, F25 = 3.0, F26 = 2.0, F27 = 2.0, F28 = 3.0; head 1.2X wider than high, head height 1.6X eye height, eyes small but protuber- ant, slightly convergent in anterior view; maximum face width 1.3X minimum face width; minimum face width 1.4x clypeus width; malar space length 1.3X mandible width basally; ocelli large, OCOD 0.9 X OCD; mandible strongly twisted; occipital carina not complete, poorly defined me- dially (Fig. 21). Mesosoma: notauli not distinct, broad, rugulose, and mesonotal lobes not well-defined (as in Fig. 11); scu- Volume 7, Number 2, 1998 189 Figs. 13-18. 13-14. Mesopleuron, showing sternaulus. 13, Meteoruti layih\igmae, stemaulus broad and rugose; 14, Meteorus yninifuanum, sternaulus foveolate but not broad. 15-18. Propodeum, dorso-posterior view. 15, Mftforus psfiidmiiiiudintiis, showing finely rugose sculpturing; 16, Mctconis rogcrblancoi, showing rugose sculp- turing; 17, M('f('iir».< ct^iigrcgiitiis, showing areolate-rugose sculpturing; 18, Mfffonis niegalofs, showing carinate sculpturing. Notice a distinct median and transverse carina creating very large defined areolae. tellar furrow with 3 carinae; sternaulus ru- gose, long but not broad; mesopleuron polished, shining around sternaulus; pro- podeum areolate-rugose, median depres- sion weakly present. Legs: hind coxa ru- gulose; tarsal claw with a small lobe, strongly curved. Wings: forewing length 3.0 mm; vein m-cu interstitial; vein 3RSa 1.8X length of r. Metasoma: (as in Fig. 3); ventral borders of first tergite joined com- pletely along basal Vi of segment (as in Fig. 4); first tergite dorsally longitudinally cos- tate with costae parallel; ovipositor short, 1.8X longer than first tergite. 190 Journal of Hymenoptera Research Figs. 19-24. 19, 21, 23. Posterior view of head showing occipital carina. 19, Mclcoimr )0\;crMtiiici.n, occipital carina complete; 21, MetccrKS alejuniinviuisisi, occipital carina not complete, poorly defined medially; 23, Mc- tcorus congrcgatus, occipital carina widely separated medially. 20, 22, 24. Tarsal claw. 20, Mctcorni. nicgdlcfv, simple tarsal claw; 22, Mclconis cnmilocnmaigci, tarsal claw with well-defined lobe and strongly curved; 24, Mctcorm pscKtlodiniidintm, tarsal claw with small lobe or basal tooth. Variation of paratype females. — Brown ar- eas vary from nearly black to nearly yel- low; 26 flagellomeres; OCOD 1-1. 3x OCD; forewing vein 3RSa 0.8 X length of r, second submarginal cell slighty nar- rowed anteriorly; median depression of propodeum strongly present; dorsal cos- tae of first tergite somewhat convergent posteriorly. Variation of paratype males. — Brown areas vary from nearly black to nearly yellow; head height 1.8X eye height. Cocoon. — Pale brown to brown, some- what wooly, ovoid; length 3.9-^.1 mm. Volume 7, Number 2, 1998 191 1.6-1.8 mm wide medially; usually termi- nating anteriorly with a rounded nipple- like projection; not stuck together but of- ten in a cluster and pointing outward from interior of caterpillar leaf shelter and next to dead or nearly-dead host; usually with anterior end attached to leaf sub- strate by a thread 0.7-2.0 mm in length which originates just posterior to adult emergence hole (note that thread origi- nates from anterior end of cocoon and not the posterior end as is usual for Meteorus). Material examined. — Holotype female: COSTA RICA: Guanacaste Province: ACG, Sector Horizontes, Vado La Esper- anza, reared from last instar larva of Sta- phylus prob. vulgata (Moschler), 1993, Jan- zen & Hallwachs voucher 93-SRNP-5688. Deposited in RMSEL. Paratypes: Guana- caste Province: 8 females, 1 male, same data as holotype; 4 females, ACG, Sector Santa Rosa, Area Administrativa, reared fr. Chiomara asychis (Stoll), larva coll. 27 July 1993, wasps pupated 31 July, wasps eclosed 9 Aug., Janzen & Hallwachs voucher 93-SRNP-4212; 5 females, 1 male, ACG, Sector Santa Rosa, Cerco de Piedra, reared fr. Pyrgus syi., larva coll. 25 Sept. 1993, wasps pupated 28 Sept., wasps eclosed 5 Oct., Janzen & Hallwachs vouch- er 93-SRNP-6387; 11 females, 1 male, ACG, Sector Cacao, Gongora, reared fr. Staphyliis sp., larva coll. 21 Aug. 1994, wasps eclosed 4 Sept., Janzen & Hall- wachs voucher 94-SRNP-6574; 4 females, ACG, Sector Cacao, Gongora, reared fr. Staphyliis sp., larva coll. 21 Aug. 1994, wasps eclosed 9 Sept., Janzen & Hall- wachs voucher 94-SRNP-6577. Deposited in INBio, UCR, RMSEL. Distribution. — Known only from the ho- lotype and paratype localities in the ACG (Sector Santa Rosa, dry forest, and Gon- gora, intergrade between wet and dry for- est), Guanacaste Province. Biology. — The rearing of this gregarious species (4-13 wasps per larva) constitutes the first record of Meteorus attacking Hes- periidae. All of the parasitized larvae were feeding on low plants in isolated dry for- est old field succession. The type speci- mens were all reared from last instar lar- vae of Hesperiidae as follows: Staphylus prob. vulgata and Staphylus sp. feeding on Amaranthaceae, Pyrgus sp. feeding on Sida rhombifolia L. (Malvaceae), and Chiomara asychis, feeding on Baiiisteriopsis muricata (Cav.) Cuatr. (Malpighiaceae). Wasps pu- pated 3-4 days after collection of hosts and eclosed 7-9 days later. Comments. — Meteorus alejandromasisi shares with M. dos and M. laphygmae the following combination of characters: strongly twisted mandibles, completely joined ventral borders of the first tergite, large ocelli, and a short ovipositor. Meteo- rus alejandromasisi can be distinguished from these species by an incomplete occip- ital carina. Intraspecific variation is high in this species. Etymology. — This species is named in honor of Sr. Alejandro Masis in recogni- tion of his outstanding interest in facilitat- ing the parataxonomists in the Research Program of the ACG, and for being a ma- jor supporter of the Research Program ac- tivities and a Hesperiidae fanatic. Meteorus camilocamargoi Zitani, new species (Fig. 22) Holotype female. — Body color: orange. Body length: 6.1 mm. Head: antenna with 35 flagellomeres; flagellar length /width ratios as follows: Fl = 2.6, F2 = 2.6, F3 = 2.0, F31 = 1.5, F32 = 1.5, F33 = 1.5, F34 = 1.5, F35 = 2.0; head 1.3x wider than high, head height 1.5x eye height, eyes large, nearly parallel in anterior view; maximum face width 1.3x minimum face width; minimum face width = clypeus width; malar space length = mandible width ba- sally; ocelli large, OCOD 0.6 X OCD; man- dible strongly twisted (as in Fig. 9); occip- ital carina not complete, poorly defined medially (as in Fig. 21); vertex concave be- hind eyes. Mesosoma: notauli uncertain (see comments section); scutellar furrow 192 Journal of Hymenoptera Research with 3 carinae; sternaulus rugose, not deep or broad but long; propodeum are- olate-rugose (as in Fig. 17), median de- pression present. Legs: hind coxa punc- tate; tarsal claws lobed, strongly curved (Fig. 22). Wings: forewing length 4.6 mm; vein m-cu antefurcal; vein 3RSa 1.3x length of r. Metasoma: (as in Fig. 3); ven- tral borders of first tergite joined com- pletely along basal V2 of segment (as in Fig. 4); first tergite with longitudinal costae parallel; ovipositor long, 3.5 x longer than first tergite. Variation of parati/pe female. — Body length 5.5 mm. Variation of paratype male. — Unknown. Cocoon. — Pale brown, ovoid; length 7.9 mm, 2.6 mm wide medially; does not ter- minate in a nipple-like projection; formed within the host cocoon; attached to the in- side of the host cocoon by a thread, 0.7 mm in length, which originates from the wasp cocoon just posterior to the adult emergence hole (not from the posterior apex of the cocoon as is usual for Meteor- us). Material examined. — Holotype female: COSTA RICA: Guanacaste Province: ACG, Sector Santa Rosa, Casona, reared from Omiodes stigmosalis (Warr.) prepupa, 1978, Janzen & Hallwachs voucher 78- SRNP-143. Deposited in RMSEL. Para- type: Guanacaste Province: 1 female, same data as holotype except Janzen & Hallwachs voucher 78-SRNP-143.1. De- posited in RMSEL. It is not known to which specimen the cocoon belongs, de- posited in RMSEL. Distribution. — Known only from the type-locality, the dry forest of the ACG, Guanacaste Province. Biology. — All type specimens were reared from larvae of the pyralid, Omiodes stigmosalis, feeding on Ficus oimlis Miq. (Moraceae). The host larva webs and mines through nearly ripe to fully ripe figs and turns them brown. The host larvae were collected 25 December 1978 and be- came prepupae 3 days later. The wasp lar- vae emerged from the prepupae and formed cocoons within the host cocoon. The wasps eclosed 18 January 1979. Comments. — Meteorus camilocamargoi, M. desmiae, and M. sterictae can be recognized by the following combination of charac- ters: strongly twisted mandibles, first ter- gite with completely joined ventral bor- ders, large ocelli, long ovipositor, and lobed tarsal claws. Meteorus camilocamargoi can be separated from these species by the following characters: first tergite not lat- erally flattened, parallel eyes, and a con- cave vertex. The type specimens were pinned through the mesonotum, therefore most of the notauli are destroyed. Eti/mology. — This species is named in honor of Sr. Camilo Camargo in recogni- tion of his extreme enthusiasm for facili- tating the full computerization of the Re- search Program and other programs of the ACG, and being a major supporter of the Research Program activities. Meteorus coffeatus Zitani, new species (Fig. 10) Holotype female. — Body color: brown ex- cept face light brown; mesosoma dark brown except fore and mid coxae yellow; legs brown; metasoma brown except first tergite dark brown; basal % of T 2-1-3 yel- low. Body length: 4.0 mm. Head: antenna with 29 flagellomeres; flagellar length/ width ratios as follows: Fl = 3.5, F2 = 3.5, F3 = 3.0, F25 = 1.5, F26 = 1.5, F27 = 1.5, F28 = 1.0, F29 = 2.0; head 1.2X wider than high, head height 1.6X eye height; eye small, strongly convergent in anterior view; maximum face width 1.7x mini- mum face width; minimum face width = clypeus width; malar space length 0.3 x mandible width basally; ocelli small, OCOD 2.5 X OCD; face polished, punctate; clypeus somewhat more convex than usu- al; antennal bases low on face; occipital ca- rina complete; mandibles moderately twisted, broad and flat at base (Fig. 10). Mesosoma: notauli distinct, linear, and mesonotal lobes well-defined (as in Fig. Volume 7, Number 2, 1998 193 12); scutellar furrow with one median ca- rina; sternaulus not broad or long, foveo- late (as in Fig. 14); propodeum rugose, median depression absent. Legs: hind coxa pohshed, punctate; tarsal claws sim- ple, swollen at base. Wings: fore wing length 3.8 mm; vein m-cu interstitial; vein 3RSa 1.7x length of r. Metasoma: ventral borders of first tergite separated basally, joined apically; first tergite dorsally lon- gitudinally costate; ovipositor long, 2.6 X longer than first tergite. Variation of paratype female. — Forewing vein m-cu postfurcal. Variation of parati/pe males. — T 2-1-3 al- most completely yellow; eyes smaller, head height 1.8X eye height. Cocoon. — Unknown. Material examined. — Holotype female: COSTA RICA: San Jose Province: Zurqui de Moravia, 1600 m, vii. 1990, Paul Han- son. Deposited in RMSEL. Paratypes: San Jose Province: 1 female, Zurqui de Mora- via, 1600 m, vii 1992, P. Hanson; 3 males, Zurqui de Moravia, 1600 m, iii 1991, P. Hanson. Deposited in INBio, UCR, RMSEL. Distribution. — Known only from the ho- lotype and paratype localities in San Jose Province. Biology. — Unknown. Comments. — Meteorus coffeatus and M. micrommatus both have moderately twist- ed mandibles but they can easily be sep- arated by the size and degree of conver- gence of the eyes: Meteorus coffeatus has large convergent eyes whereas M. microm- matus has small parallel eyes. Meteorus cof- featus can easily be distinguished from the species with untwisted mandibles by the sculpturing on the propodeum. Etymology. — From Latin meaning "the color of roasted coffee beans," for the brown body color of this species. Meteorus congregatiis Muesebeck (Figs. 7-8, 17, 23, 25-27) Meteorus congregatus Muesebeck 1939: 86. Diagnosis of females. — Body color: yel- lowish-white except antennae black; head orange dorsally; pronotum orange dorsal- ly; mesonotum orange; fore and mid legs darker with tarsi brown apically; hind leg with tibia dark brown apically, tarsus dark brown; first tergite light brown dor- sally; metasoma light brown dorso-medi- ally; wings very dark, infused with black- ish pigment. Body length: 3.8-4.1 mm. Head: antenna with 27 flagellomeres; fla- gellar length /width ratios as follows: Fl = 2.5, F2 = 2.5, F3 = 3.0, F23 = 3.0, F24 = 3.0, F25 = 3.0, F26 = 2.0, F27 3.0x; head 1.1 X wider than high, head height 2.0x eye height; eyes small, nearly parallel in anterior view; maximum face width 1.1 X minimum face width; minimum face width 1.3X clypeus width; malar space length 2.0 X mandible width basally; face, clypeus polished, minutely punctate; ocel- li small, OCOD 1. 8-2.0 X OCD; mandible strongly twisted; occipital carina incom- plete, usually, but not always widely sep- arated medially (Fig. 23). Mesosoma: no- tauli distinct, deep, foveolate, and meso- notal lobes well-defined; mesonotum pol- ished, punctate, with short carina dorso-medially; scutellar furrow with one median carina; mesopleuron polished, punctate; sternaulus rugose; propodeum areolate-rugose, somewhat coarse (Fig. 17); median depression present. Legs: hind coxa smooth; tarsal claws simple, swollen basally. Wings: forewing length 3.3 mm; vein m-cu interstitial, antefurcal; vein 3RSa 0.6 x length of r; second sub- marginal cell strongly narrowed anterior- ly. Metasoma: (as in Fig. 3); ventral bor- ders of first tergite partially fused (Fig. 7) or completely fused (Fig. 8); first tergite dorsally longitudinally costate, costae par- allel; ovipositor short, thick at base, some- times curved, 1.1-1.5X longer than first tergite. Variation of males. — 32 flagellomeres; head height l.Sx eye height. Cocoon.— (Figs. 25, 27) A cluster of 25- 250 pinkish-beige, wooly cocoons, all massed together next to host caterpillar 194 Journal of Hymenoptera Research Figs. 25-28. 25-27. Meteorus coiigregatus and last instar host, Maudtiai Acxtii, at the ACG. Janzon & Hallvvachs voucher 95-SRNP-7538. 25, Large cocoon cluster next to host cadaver; 26, Close-up of host cadaver showing exit holes of M. congrcgatus larvae; 27, Close-up of cocoons. Notice cocoons terminate with a distinct nipple- like projection. 28. Meteorus papiiliovorus cocoons next to host caterpillars, Papilio nnehisiihies. at Santa Rosa dry forest, ACG. Notice cocoons terminate with a distinct nipple-like projection. Janzen & Hallwachs voucher 90- SRNP-2177. and stuck to leaf substrate; each terminat- ing with a distinct nipple-like projection. Material examined.— COSTA RICA: Guanacaste Province: 51 females, 34 males, ACG, Sector Puente, Est. San Cris- tobal, 2 km W. of Dos Rios, reared from Mandiica sexta (L.) larva on Solamim tor- vum Sw., larva collected 4 Aug. 1995, wasps eclosed 11-12 Aug. 1995, Janzen & Hallwachs voucher 95-7538; 92 females, 162 males, ACG, Sector Puente, Est. San Cristobal, 2 km W. of Dos Rios, reared from Manduca sexta larva on Solatium tor- viim, larva collected 4 Aug. 1995, wasps eclosed 11-12 Aug. 1995, Janzen & Hall- wachs voucher 95-SRNP-7539; 19 females, 7 males, ACG, [Sector El Hacha] Los Al- mendros, reared from Uiizela japix (Cram.) larva, Janzen & Hallwachs voucher 96- SRNP-11208, 1996. Deposited in INBio, UCR, RMSEL. Distributicvi. — Recorded only from local- Volume 7, Number 2, 1998 195 ities in wet forest of the eastern end of the ACG, Guanacaste Province, and in Pana- ma (Canal Zone). Biology. — M. congregatus is a gregarious species attacking sphingid larvae. A large number of wasp larvae (about 25-250) emerge from a single host larva. In Costa Rica M. congregatus was reared from Man- diica sexta feeding on Solamtm tonniiu (So- lanaceae), and Unzela japix feeding on Te- tracern vohihilis L. (Dilleniaceae). Wasp lar- vae emerged from the last instar larva of the host (Figs. 25-27). Comments. — Meteorus congregatus was described by Muesebeck (1939) from spec- imens reared from the larva of a sphingid, Eri)ini/is ello (L.), feeding on papaya, from Panama (Canal Zone). The Costa Rican specimens of M. congregatus show some morphological variation not discussed by Muesebeck in the original description: oc- cipital carina with a wide separation me- dially (Fig. 23); ventral borders of first ter- gite partially or completely fused (Figs. 7- 8); and ovipositor sometimes curved. Meteorus congregatus is very similar morphologically to M. papiliovorus, and shares with M. papiliovorus and one other Neotropical species, M. townsendi Muese- beck, a Brazilian species, which has not been found to occur in Costa Rica, a nar- rowed second submarginal cell of the fore- wing. M. congregatus can be separated from these species by its widely separated occipital carina, fused ventral borders, and parallel costae of the first tergite. M. con- gregatus also has a cocoon that is very dif- ferent from that of M. papiliovorus (Figs. 25, 28), and feeds on another family of Lepidoptera. Meteorus comiculatus Zitani, new species Holotype female. — Body color: body red- dish-orange except antenna yellow basal- ly, brown apically; legs yellow except me- sothoracic leg with femur brown apically; metathoracic leg with femur, tibia, and tarsus brown apically; metasoma dark brown except first tergite orange. Body length: 4.5 mm. Head: antenna with 26 flagellomeres; flagellar length /width ra- tios as follows: Fl = 5.0, F2 = 5.0, F3 = 4.0, F22 = 2.3, F23 = 2.3, F24 = 2.0, F25 = 1.5, F26 = 2.7; head 1.3X wider than high, head height 1 .5 X eye height; eye large and protuberant, convergent in anterior view; maximum face width 1.4X minimum face width; minimum face width equal to clyp- eus width; malar space short, malar space length 0.5 X mandible width basally; ocelli small, OCOD 2.3 x OCD; face, clypeus polished; mandible not twisted. Mesoso- ma: notauli distinct, linear, foveolate, and mesonotal lobes well-defined; mesonotum polished except scutellar disc coarsely ru- gose and raised in a sharp point; scutellar furrow with a distinct median carina, two lateral carinae weakly present; mesopleu- ron rugose; stemaulus rugose, long and broad; propodeum coarsely areolate-ru- gose, median depression absent. Legs: hind coxa polished, punctate; tarsal claws simple (as in Fig.20). Wings: forewing length 3.6 mm; vein m-cu antefurcal; r 0.6 X length of 3RSa; second submarginal cell quadrate. Metasoma: (as in Fig. 3); ventral borders of first tergite joined com- pletely along basal Vi of segment; first ter- gite polished with faint costae posteriorly; ovipositor long, 2.2 x longer than first ter- gite. Variation of paratype females. — Body (ex- cept legs) almost entirely dark reddish- brown. Variation of paratype male. — Antennae broken; head 1.2x wider than high; head height 1.8X eye height. Cocoon. — Unknown. Material examined. — Holotype female: COSTA RICA: Puntarenas Province: San Vito, Las Cruces, 1200 msnm, VIII-IX 1988, Coll. P. Hanson. Deposited in RMSEL. Paratypes: Guanacaste Province: 2 females, ACG, [Sector] Santa Rosa, Bosque Humedo-12-C, Malaise, Janzen & Gauld, 3.viii-24.viii.l985; 1 female, ACG, [Sector El Hacha] Los Almendros, E. Lo- 196 Journal of Hymenoptera Research pez, 1 a 22 July 1992, L-N 334800, 369800, INBIO CRIOOO 735444; Puntarenas Prov- ince: 1 male, Rancho Quemado, 200 m. Peninsula de Osa, Oct 1992, M. Segura, L- S 292500, 511000 INBIO CRIOOO 969047; 2 females, San Vito, Estac. Biol. Las Alturas, 1500 m, vi. 1992, traps #1 & #2, Malaise, Paul Hanson; 1 female, R.F. Golfo Dulce, 3 km SW Rincon, 10 m, vi. 1991, Paul Han- son. San Jose Province: Zurqui de Mora- via, 1600 m, viii. 1995, Malaise, Paul Han- son. Deposited in INBio, UCR, RMSEL. Distribution. — Known only from the ho- lotype and paratype localities in Guana- caste, Puntarenas, and San Jose Provinces. B iology . — Unknown . Comments. — This species has an unusual character found in no other Meteorus spe- cies: the scutellar disc is coarsely rugose and raised into a sharp point. In all other species the scutellar disc is rounded and smooth. The combination of the form of the scutellar disc, untwisted mandibles, and a coarsely rugose propodeum makes this species very easy to recognize. Etymology. — Named for the form of the scutellar disc, from Latin meaning "hav- ing a horn-shaped appendage." Meteorus desmiae Zitani, new species Holotype female. — Body color: yellow-or- ange except head with temples and vertex black, antennae brown; pronotum black dorsally; propleura mostly black; meso- notum black; mesopleuron black anteri- orly; propodeum dark brown dorsally; first tergite with lateral brown spots. Body length: 6.0 mm. Head: antenna with 31 flagellomeres; flagellar length /width ra- tios as follows: Fl = 3.5, F2 = 3.5, F3 = 3.0, F27 = 3.0, F28 = 2.0, F29 = 2.0, F30 = 2.0, F31 = 3.0; head 1.2x wider than high, head height 1.4x eye height; eye large and protuberant, ventral margin of eye nearly touching base of mandible; eyes slightly convergent in anterior view; maximum face width 1.6X minimum face width; minimum face width 0.8 X clypeus width; malar space length 0.5 X mandible width basally; ocelli large, OCOD 0.5 X OCD; face, clypeus punctate; clypeus somewhat flattened; mandible strongly twisted (as in Fig. 9); occipital carina complete (as in Fig. 19). Mesosoma: notauli not distinct, ru- gulose, and mesonotal lobes not well-de- fined (as in Fig. 11); mesoscutum polished, punctate, with median carina anteriorly; scutellar furrow with 3 distinct carinae; mesopleuron polished, punctate; sternau- lus rugose, very long and extending dor- sally; propodeum rugose, with two dis- tinct carinae posterio-ventrad, median de- pression weakly present. Legs: hind coxa polished, punctate; tarsal claws lobed, strongly curved (as in Fig. 22). Wings: forewing length 4.6 mm; vein m-cu ante- furcal; vein 3RSa 1.8X length of r. Meta- soma: ventral borders of first tergite joined completely along basal V2 of segment (as in Fig. 4); first tergite dorsally longitudi- nally costate, costae parallel; ovipositor long, 2.5 X longer than first tergite. Variation of females. — Unknown. Variation of paratype male. — Body length 5.0 mm; eyes nearly parallel, maximum face width 1.3x minimum face width; minimum face width equal to clypeus width; fore wing length 3.6 mm; vein 3RSa 1.3x length of r; hind coxa rugulose. Cocoon. — Ovoid, brown; length 5.9 mm, 2.4 mm wide medially; formed within the remains of the larval leaf roll. Material examined. — Holotype female: COSTA RICA: Guanacaste Province: [ACG, Sector Pitilla] Est. Pitilla, 700 m, 9 km S Sta Cecilia, P. Rios & C. Moraga, Oct 1990, L-N-330200, 380200, INBIO CRIOOO 398209. Deposited in RMSEL. Paratype: COSTA RICA: Guanacaste Province: 1 male with cocoon, ACG, [Sector] Santa Rosa, Casetilla Entrada, reared from Dt's- mia tages (Cram.), 1983, Janzen & Hall- wachs voucher 83-SRNP-574A. Deposited in RMSEL. Distribution. — Known only from the ho- lotype and paratype localities in the ACG, Guanacaste Province. Biology. — The male paratype was reared Volume 7, Number 2, 1998 197 from a pyralid, Desmia tages, feeding on Hatnelia patens Jacq. (Rubiaceae). This py- ralid is a leaf roller /webber. The parasit- oid emerged from the last instar larva. Comments. — Meteonis desmiae shares with M. cnmilocamargoi and M. sterictae the following characters: strongly twisted mandibles, completely joined ventral bor- ders of the first tergite, large ocelli, long ovipositor, and lobed tarsal claws. Meteo- nis desmiae can be separated from these species by its first tergite being not later- ally flattened, convergent eyes, and com- plete occipital carina. Etymology. — This species is named for the genus of the pyralid caterpillar it at- tacks, Destnia. Meteonis dimidiatus (Cresson) (Fig. 6) Perilitus dimidiatus Cresson 1872:83. Diagnosis of females. — Body color: dark brown except antennae brown, orange around eyes, face yellow, pronotum yel- low ventrally, legs yellow except hind leg with tibia and tarsus light brown, meta- soma brown. Body length: 3.6 mm. Head: antenna with 27 flagellomeres; flagellar length /width ratios as follows: Fl = 4.0, F2 = 3.5, F3 = 3.0, F23 = 1.5, F24 = 1.5, F25 = 1.5, F26 = 2.0, F27 = 3.0; head 1.1 X wider than high, head height 1.7x eye height, eye small, convergent in anterior view; maximum face width 1.4X mini- mum face width; minimum face width 1.1 X clypeus width; malar space short, malar space length = mandible width ba- sally; ocelli small, OCOD 2.0 X OCD; man- dible strongly twisted (as in Fig. 9); occip- ital carina not complete. Mesosoma: no- tauli distinct, finely areolate, and meso- notal lobes well-defined; scutellar furrow with one distinct median carina, 4 carinae weakly present; mesopleuron polished, punctate; sternaulus not deep or broad but long; propodeum finely areolate-rugose, with a weak carina medially (as in Fig. 15); median depression weakly present. Legs: hind coxa polished, punctate; tarsal claws simple but swollen at base. Wings: fore- wing length 3.3 mm; vein m-cu post fur- cal; vein 3RSa 1.3 X length of r. Metasoma: (as in Fig. 5); ventral borders of first tergite joined for only a short distance apically (Fig. 6); first tergite with lateral longitu- dinal carinae, rugulose medially; oviposi- tor long, 2.6 X longer than first tergite. Variation of males. — Unknown. Cocoon. — Unknown. Material examined. — COSTA RICA: Cuanacaste Province: 1 female, [ACG, Sector Cacao, Estacion Cacao =] Estac. Mengo, S.W. Volcan Cacao, 1100 m IX-X 1989; 1 female, [ACG, Sector Cacao, Esta- cion Cacao =] Estac. Mengo, S.W. Volcan Cacao, 1100 m, 1988-1989; Puntarenas Province: 5 females, San Vito, Estac. Biol. Las Alturas, 1500 m, xii, 1991, Paul Han- son; 1 female, Estac. Biol. Las Alturas, xi 1991; 1 female, Estac. Biol. Las Alturas, i, 1992; 1 female Estac. Biol. Las Alturas, ii, 1992; San Jose Province: 1 male, Zurqui de Moravia, vi. 1990, 1600 m, P. Hanson; 1 female, Zurqui de Moravia, 1600 m, x- xii, 1990; 1 female, Zurqui de Moravia, iii, 1991; 2 females, Zurqui de Moravia, iii, 1992, P. Hanson. Deposited in INBio, UCR, RMSEL. Distribution. — In Costa Rica Known from Guanacaste, Puntarenas, and San Jose Provinces. It is also widely distribut- ed in North America and recorded from Patagonia (Shenefelt 1969). Biology. — Unknown. Comments. — M. dimidiatus and M. pseu- dodimidiatus are very similar morphologi- cally, but can be separated by the length of the ovipositor and size of the eyes. These are the first records of M. dimidiatus in Costa Rica. Meteonis dos Zitani, new species Holotype female. — Body color: Head yel- low, antenna black; mesosoma white ex- cept pronotum dark brown dorsally, mesonotum dark brown anteriorly and laterally, mesopleuron dark brown ven- 198 Journal of Hymenoptera Research trally, propodeum dark brown dorsally and posteriorly, legs brown apically, hind coxa with dark brown patches laterally; metasoma white except first tergite dark brown apical %, T 2+3 dark brown later- ally. Body length: 4.6 mm. Head: antenna with 28 flagellomeres; flagellar length/ width ratios as follows: Fl = 3.5, F2 = 3.0, F3 = 2.5, F24 = 2.5, F25 = 2.0, F26 = 2.0, F27 = 2.0, F28 = 2.5; head 1.2x wider than high, head height 1.3x eye height, eyes large and protuberant, ventral margins nearly touching base of mandible; slightly convergent in anterior view; maximum face width 1.3X minimum face width; minimum face width = clypeus width; malar space length 0.5 X mandible width basally; ocelli large, ocell-ocular distance 0.8 X greatest diameter of a lateral ocellus mandible strongly twisted (as in Fig. 9) occipital carina complete (as in Fig. 19) frons with a median tubercle anterior to median ocellus. Mesosoma: notauli not distinct, broad, rugulose, and mesonotal lobes not well-defined (as in Fig. 11); scu- tellar furrow with 4 carinae; mesopleuron polished, minutely punctate; stemaulus foveolate, not broad or long; propodeum finely areolate-rugose, median depression weakly present. Legs; hind coxa rugulose, with distinct but fine carinae; tarsal claw with a very small lobe or basal tooth (as in Fig. 24). Wings: forewing length 4.4 mm; vein m-cu post furcal; vein 3RSa 1.7x length of r. Metasoma: (as in Fig. 3); ven- tral borders of first tergite joined com- pletely along basal 1/2 of segment (as in Fig. 4); first tergite dorsally longitudinally costate, costae convergent posteriorly; ovi- positor short, 1.5x longer than first tergite. Variation of females. — Unknown. Variation of males. — Unknown. Cocoon. — Unknown. Material examined. — Holotype female: COSTA RICA: San Jose Province: Zurqui de Moravia, 1600 m. I-II 1989, P. Hanson & 1. Gauld. Deposited in RMSEL. Distribution. — Known only from the ho- lotype locality in San Jose Province. Biology. — Unknown. Comments. — This species has an unusual color pattern of dark brown and white patches, similar to the color pattern of M. uno, giving these two species a superficial similarity. Morphologically M. dos resem- bles M. laphygmae, as they share strongly twisted mandibles, completely joined ven- tral borders of the first tergite, large ocelli, short ovipositor, and a complete occipital carina. Meteorus dos can be further sepa- rated from this species by its very large eyes and median tubercle on the frons. Etymology. — The specific epithet is an arbitrary combination of letters. Meteorus laphygmae Viereck (Fig' 13) Meteorus laphygmae Viereck, 1913, Proc. U.S. Nat. Mus. 44560. Diagnosis of females. — Body color: yel- low. Body length: 3.6-5.4 mm. Head: an- tenna with 31-34 flagellomeres; flagellar length /width ratios as follows: Fl = 5.0, F2 = 5.0, F3 = 2.7, F27 = 1.5, F28 = 1.5, F29 = 1.0, F30 = 1.3, F31 = 2.0; head 1.2X wider than high, head height 1.5X eye height, eyes large and protuberant; nearly parallel in anterior view; maximum face width 1.1 X minimum face width; mini- mum face width = clypeus width; malar space length 0.8-1 .2 X mandible width ba- sally; ocelli large, OCOD 1.0-1.4X OCD; mandible strongly twisted (as in Fig. 9); occipital carina complete (as in Fig. 19). Mesosoma: notauli not distinct, broad, ru- gulose, and mesonotal lobes not well-de- fined (as in Fig. 11); scutellar furrow with 3 carinae; mesopleuron polished, minutely punctate; stemaulus rugulose to rugose, broad and long (as in Fig. 13); propodeum rugose (as in Fig. 16) to areolate-rugose, median depression present, sometimes weak. Legs: hind coxa rugulose; tarsal claw with a small lobe, strongly curved. Wings: forewing length 3.2-4.7 mm; vein m-cu antefurcal or interstitial; vein 3RSa 1. 0-0.4 X length of r. Metasoma: (as in Fig. Volume 7, Number 2, 1998 199 3); ventral borders of first tergite joined completely along basal Vz of segment (as in Fig. 4); first tergite dorsally longitudi- nally costate, costae convergent posteri- orly; ovipositor short, sometimes thick at base, 1.6-1.8X longer than first tergite. Variation of males. — 34 flagellomeres; eyes smaller, head height 1.9x eye height; maximum face width = minimum face width; propleura darker, propodeum brown dorsally, hind leg with tibia and tarsus darker, first tergite of metasoma brown apically.. Cocoon. — No reared specimens were ex- amined from Costa Rica. Material examined.— COSTA RICA: Guanacaste Province: 5 females. Hacien- da El Vieja, Filadelfia, sugarcane, 110 m, V 1989, M. Garcia; 3 females, 1 male, [ACG, Sector Cacao] Volcan Cacao, Cerro Pedregal, 1000 m, Il-IV 1989, 1. Gauld & D. Janzen; 4 females, 3 males, Sotobosque, W side Volcan Cacao, 1100 m, II 1989, 1. Gauld. Deposited in INBio, UCR, RMSEL. Limon Province: 1 female, Los Diamantes, Guapiles, 200m, 20/ V/ 1988, col. Paul Hanson. Deposited in INBio, UCR, RMSEL. Distribution. — In Costa Rica known from Guanacaste and Limon Provinces. Previously recorded from Venezuela, Co- lombia, and North America, and intro- duced into Hawai'i (Shenefelt 1969). Biology. — Unknown. Comments. — Meteoriis laphi/gmae shares with M. dos and M. alejandromasisi the fol- lowing combination of characters: strong- ly twisted mandibles, completely joined ventral borders of the first tergite, large ocelli, and a short ovipositor. Meteorus laphygmae can be separated from these species by a complete occipital carina and a longer malar space. Meteonis mariamartae Zitani, new species Holotype female. — Body color: orange- brown except head dark brown, antenna brown except F1-F3 white and with white annulus F9-F16; legs yellow; metasoma slightly paler than mesosoma. Body length: 2.5 mm. Head: antenna with 20 flagellomeres; flagellar length /width ra- tios as follows: Fl = 5.0, F2 = 5.0, F3 = 4.0, F16 = 2.0, F17 = 2.0, F18 = 2.0, F19 = 2.0, F20 = 3.0; antenna slightly tapered ba- sally; head 1.4X wider than high, head height 1.5X eye height; eye large, conver- gent in anterior view; maximum face width 1.6x minimum face width; mini- mum face width 0.7x clypeus width; ma- lar space short, malar space length 0.6 x mandible width basally; ocelli small, OCOD 3.0 X OCD; mandible not twisted; occipital carina complete (as in Fig. 19). Mesosoma: notauli finely areolate, dis- tinctly linear, converging posteriorly in a distinct v-shape, and mesonotal lobes well-defined; scutellar furrow with one median carina; mesopleuron polished, smooth and shining; stemaulus foveolate, not broad (as in Fig. 14); propodeum car- inate, with a distinct median and trans- verse carina creating very large, defined areolae (as in Fig. 18); median depression absent. Legs: hind coxa rugulose; tarsal claws simple (as in Fig. 20). Wings: fore wing length 1.9 mm; vein m-cu antefurcal; vein 3RSa 4.0 X length of r. Metasoma: ventral borders of first tergite joined al- most completely beneath, open basally for a short distance; first tergite dorsal surface rugulose and without any indication of costae; ovipositor long, 3.2 x longer than first tergite. Variation of paratype females. — Body color dark brown; dorsal surface of first tergite sometimes with very faint costae laterally. Variation of males. — Unknown. Cocoon. — Unknown. Material examined. — Holotype female: COSTA RICA: Puntarenas Province: P[arque].N[acional]. Corcovado, Est. Sir- ena, 50 m, x-xii 1990. Deposited in RMSEL. Paratypes: Guanacaste Province: 1 female, [ACG, Sector Cacao, Estacion Cacao = ] Estac. Mengo, SW Volcan Cacao, 1100 m, 1988-1989. Puntarenas Province: 200 Journal of Hymenoptera Research 2 females, same data as holotype; 1 fe- male, Rancho Quemado, 200 m Peninsula de Osa, Die 1992, M. Segura, L-S 292500, 511000. Deposited in INBio, UCR, RMSEL. Distribution. — Known only from the ho- lotype and paratype localities in Guana- caste and Puntarenas Provinces. Biology. — Unknown. Comments. — Meteonts mariamartae close- ly resembles M. i/amijumium, and M. me- galops. These species have untwisted man- dibles and a carinate propodeum (with large, defined areolae). Meteorus mariamar- tae can be distinguished from these other two species by its smaller eyes, unique no- tauli (converging posteriorly in a distinct V-shape), and small body size. Etymology. — This species is named in honor of Srta. Maria Marta Chavarria Diaz in recognition of her unflagging support for the concept of biodiversity conserva- tion throughout the history of the ACG and INBio. Meteorus megalops Zitani, new species (Figs. 12, 18, 20) Holotype female. — Body color: Dark brown except; antenna brown with white annulus F13-F19; mesosoma with fore and mid coxae pale yellow; metasoma with first tergite pale yellow basally and api- cally, tergite 2 + 3 pale yellow for basal %. Body length: 5.4 mm. Head: antenna fili- form with 25 flagellomeres; flagellar length /width ratios as follows: Fl = 5.0, F2 = 4.0, F3 = 3.0, F21 = 1.5, F22 = 1.5, F23 = 1.5, F24 = 1.0, F25 2.0x; head 1.3x wider than high, head height 1.4x eye height; eye large, strongly convergent in anterior view, bulging anteriorly and lat- erally; maximum face width 1.7X mini- mum face width; minimum face width 0.7x clypeus width; malar space short, malar space length 0.5 x mandible width basally; ocelli small, OCOD 2.0 x OCD; mandible not twisted; occipital carina complete (as in Fig. 19). Mesosoma: no- tauli distinct, foveolate, and mesonotal lobes well-defined (Fig. 12); scutellar fur- row with a distinct median carina, 2 lat- eral carinae weakly present; mesopleuron smooth, polished; sternaulus foveolate (as in Fig. 14); propodeum carinate, with a distinct median and transverse carina cre- ating very large, defined areolae (Fig. 18); median depression absent. Legs: hind coxa polished, punctate; tarsal claws sim- ple (Fig. 20). Wings: forewing length 4.8 mm; vein m-cu antefurcal; r 0.6 X length of 3RSa; second submarginal cell slightly narrowed anteriorly. Metasoma: ventral borders of first tergite separated basally, joined apically; first tergite dorsally lon- gitudinally costate; ovipositor long, thick at base, 3.6 X longer than first tergite. Variation of paratype females. — Mesopleu- ron reddish-orange ventrally; 24 flagel- lomeres, annulus on F9-F14; Variation of paratype males. — 29 flagel- lomeres; antennae without annulus, com- pletely brown; eyes not so large and bulging. Cocoon. — Unknown. Material examined. — Holotype female: COSTA RICA: San Jose Province: Zurqui de Moravia, 1600 m, X-Xll, 1989, col. Paul Hanson. Deposited in RMSEL. Paratypes: COSTA RICA: Guanacaste Province: 1 fe- male, [ACG, Sector Pitilla] Est. Pitilla, 700 m, 9 km S. Sta. Cecilia, P.N. Guanacaste, C. Moraga, 31 mar-15 abr 1992, L-N 330200.380200. Puntarenas Province: 3 fe- males, 8 males, San Vito, Estac. Biol. Las Alturas, 1500 m, vi. 1992, traps #1 & #2, Malaise, Paul Hanson; 1 female, San Vito, Estac. Biol. Las Alturas, 1500 m, v. 1992, in the forest. Malaise, Paul Hanson; 1 fe- male, San Vito, Estac. Biol. Las Alturas, 1500 m, V. 1992, forest border. Malaise, Paul Hanson; 1 female, San Vito, Estac. Biol. Las Alturas, 1500 m, vi. 1992, Mal- aise, Paul Hanson; 1 female. Est. Biol. Las Alturas, 1500 m, Coto Brus, M.A. Zum- bado, Nov 1991, L-S-322500,591300; 1 fe- male. Est. Biol. Las Alturas, 1500 m, Coto Brus, F. Arayo, 23 mar a 2 may 1992, L-S- 322500,591300; 2 females. Est. La Casona, 1520 m. Res. Biol. Monteverde, Ago 1992, Volume 7, Number 2, 1998 201 N. Obando L-N 253250,449700; San Jose Province: 1 female, Zurqui de Moravia, 1600 m lV/1989, col. Paul Hanson. Depos- ited in INBio, UCR, RMSEL. Distribution. — Known only from the ho- lotype and paratype localities in Guana- caste, Puntarenas, and San Jose Provinces in Costa Rica. B iologi/. — Unknown. Comments. — Meteorus megalops very closely resembles M. mariamartae and M. yamijiianum. These 3 species share in com- mon untwisted mandibles and a carinate propodeum (with large, defined areolae. Fig. 18). Meteorus megalops is distinctive because of its huge, bulging, convergent eyes, longer antennae, and large body size. Etymologif. — From Greek meaning "large eyes." Meteorus micrommatiis Zitani, new species Holotype female. — Body color: head black, antennae dark brown; mesosoma orange-brown except propleura dark brown; fore leg with coxa yellow, femur, tibia and tarsus brown; middle leg with coxa yellow, femur, tibia, tarsus brown; hind leg brown; wings dark; metasoma dark brown except first tergite orange ba- sally, brown apically; tergite 2 + 3 yellow basally. Body length: 4.6 mm. Head: an- tennae broken, at least 22 flagellomeres; flagellar length /width ratios as follows: Fl = 3.0, F2 = 3.0, F3 = 4.0; head 1.2x wider than high, head height 1.9x eye height, eye small, nearly parallel in ante- rior view; maximum face width 1.1 x min- imum face width; minimum face width 1.2x clypeus width; malar space length long, 1.1 x mandible width basally; ocelli small, OCOD 3.3 x OCD; mandible mod- erately twisted, broad at base (as in Fig. 10); occipital carina complete (as in Fig. 19). Mesosoma: notauli distinct, foveolate, converging posteriorly in a depressed ru- gose area with 4 distinct carinae, and mesonotal lobes well-defined; scutellar furrow with 3 carinae; sternaulus distinct but not broad (as in Fig. 14); mesopleuron polished; propodeum areolate-rugose, me- dian depression absent. Legs: hind coxa polished, punctate; tarsal claws simple. Wings: fore wing length 3.5 mm; vein m- cu antefurcal; vein 3RSa = length of r. Me- tasoma: ventral borders of first tergite sep- arated but nearly touching apically; dorsal surface of first tergite with lateral costae, rugulose medially; ovipositor short, 1.4X longer than first tergite. Variation of paratype females. — Mesosoma brown; body entirely dark brown; 29 fla- gellomeres. Variation of males. — Unknown. Cocoon. — Unknown. Material examined. — Holotype female: COSTA RICA: Puntarenas Province: Fca. Cafrosa, 1300 m. Est. Las Mellizas, P. In- ternac. La Amistad, M. Ramirez, Jul 1991, L-S-316100, 596100, INBIO CRIOOO 624079. Deposited in RMSEL. Paratypes: Guana- caste Province: 1 female, [ACG, Sector Ca- cao] Est. Cacao, 1000-1400 m, Lado SO Vol. Cacao, C. Chaves, Ago 1991, L-N- 32300, 375700, INBIO CRIOOO 571335; Puntarenas Province: 1 female. Est. La Ca- sona, 1520 m. Res. Biol. Monteverde, J. A. Ugalde, Jul 1991, L-N-253250, 449700, IN- BIO CRIOOO 551283; 1 female. Est. Biol. Las Alturas, 1500m, Goto Brus, F. Araya, 23 mar a 2 may 1992, L-S 322500, 591300, INBIO CRIOOO 792767. Deposited in IN- Bio, UCR, RMSEL. Distribution. — Known only from the ho- lotype and paratype localities in Guana- caste and Puntarenas Provinces. Biology. — Unknown. Comments. — Meteorus micrommatus is unusual in that although it lacks dorsopes, the ventral borders of the first tergite are very slightly separated. It shares with M. coffeatus moderately twisted mandibles but can easily be distinguished by its very small (head height 1.9 X eye height) par- allel eyes. Etymology. — From Greek meaning "small eyes." 202 Journal of Hymenoftera Research Meteorus papiliovorus Zitani (Fig. 28) Meteorus papiliovorus Zitani 1997:181. See Zitani et al. (1997) for a species diagnosis. No new material has been examined. Meteorus pseudodimidiatus Zitani, new species (Figs. 5, 15, 24) Holotype female. — Body color: Dark brown except yellow around ventral bor- ders of eyes, face yellow; pronotum with yellow spot ventrally, mesopleuron yel- low on ventral Vz; legs light brown; meta- soma with T 2-1-3 white on basal Vi, brown apically; T 4, 5 brown, white posteriorly and ventrally. Body length: 4.3 mm. Head: antenna with 29 flagellomeres; fla- gellar length /width ratios as follows: Fl = 4.5, F2 = 4.0, F3 = 3.5, F25 = 2.0, F26 = 1.5, F27 = 1.5, F28 = 1.5, F29 = 2.0; head 1.1 X wider than high, head height 1.8X eye height; eye small, convergent in anterior view; maximum face width 1.5x minimum face width; minimum face width = clypeus width; malar space length 1.5x mandible width basally; ocelli small, OCOD 2.0 X OCD; mandible strong- ly twisted. Mesosoma: notauli distinct and mesonotal lobes well-defined (as in Fig. 12); scutellar furrow with a median carina; stemaulus foveolate, not broad or long (as in Fig. 14); propodeum finely ru- gose, with a weak median carina (Fig. 15), median depression absent. Legs: hind coxa rugulose; tarsal claw with a small lobe or basal tooth (Fig. 24). Wings: fore- wing length 4.1 mm; vein m-cu post fur- cal; r 0.4X length of 3RSa. Metasoma: (Fig. 5); ventral borders of first tergite joined for only a short distance (as in Fig. 6); first tergite dorsally longitudinally costate; ovi- positor short, 1.3x longer than first tergite. Variation of paratype females. — 30 flagel- lomeres. Variation of males. — Unknown. Cocoon. — Unknown. Material examined. — Holotype female: COSTA RICA: Heredia Province: Vara Blanca, Finca Georgina, 2100 m, i-ii 1990, P. Hanson. Deposited in RMSEL. Para- types: Heredia Province: 1 female. Vara Blanca, Finca Georgina, 2100 m, iii-iv 1990, P. Hanson; 1 female. Vara Blanca, Finca Georgina, 2100 m, vi-viii 1990, P. Hanson. San Jose Province: 1 female, 16km S. Empalme, 2600 m, III-IV 1989, P. Hanson & I. Gauld; 3 females, Cerro Muerte, 20 km S. Empalme, 2800 m, xi 88- i 1989, P. Hanson. Deposited in INBio, UCR, RMSEL. Distribution. — Known only from the ho- lotype and paratype localities in Heredia and San Jose Provinces. Biologif. — Based on the holotype and paratype localities M. pseudodimidiatus oc- curs at high elevations. The paratype rec- ord on Cerro de la Muerte, at 2800 m, is the highest elevation record for the known Costa Rican Meteorus. Comments. — M. pseudodimidiatus is based on 7 female specimens and there is almost no morphological variation in the type specimens. This species is very simi- lar to M. dimidiatus, and can be separated by the shorter length of the ovipositor and small size of the eyes. Etymologif. — Derived from Greek mean- ing "false dimidiatus" in reference to the many morphological similarities between this species and M. dimidiatus. Meteorus rogerblancoi Zitani, new species (Figs. 3-4, 9, 11, 16, 19) Holotype female. — Body color: Very dark brown except antennae with white annu- lus F23-28; face light brown, orange around eyes; mesonotum yellow medially, scutellar disc yellow; pronotum white ventrally; propleura white; mesopleuron white medially; metapleura white; fore coxa white, leg light brown; mid coxa white, leg light brown; hind coxa white anteriorly, leg brown; first tergite white basally; T 2 + 3 white medially; metasoma white apically and ventrally. Body length: 4.3 mm. Head: antenna with 28 tlagello- Volume 7, Number 2, 1998 203 meres; flagellar length/ width ratios as fol- lows: Fl = 3.5, F2 = 3.5, F3 = 2.5, F24 = 1.0, F25 = 1.0, F26 = 1.0, F27 = 1.0, F28 = 1.5; head 1.2X wider than high, head height 1.6X eye height, eye small, nearly parallel in anterior view; maximum face width 1.2x minimum face width; mini- mum face width = clypeus width; malar space length 1.5x mandible width basally; ocelli small, OCOD 1.50 x OCD; occipital carina complete (fig. 19); mandible strong- ly twisted, and frons with faint horizontal carinae, clypeus rugulose (Fig. 9). Meso- soma: notauli not distinct, broad, rugu- lose, and mesonotal lobes not well-defined (Fig. 11); scutellar furrow with one distinct median carina, 4 lateral carinae weakly present; mesopleuron polished, punctate; sternaulus shallow but not broad or long; propodeum rugose (Fig. 16); median de- pression weakly present. Legs: hind coxa rugulose; tarsal claw with a small lobe, strongly curved. Wings: fore wing length 4.0 mm; vein m-cu interstitial; vein 3RSa 2.0 X length of r. Metasoma: (Fig. 3); ven- tral borders of first tergite joined com- pletely along basal Vi of segment (Fig. 4); first tergite with longitudinal costae con- vergent ventrally; ovipositor short, 1.9x longer than first tergite. Variation of paratype females. — Head completely yellowish; pronotum and mesopleuron completely white; body length 4.1-4.3 mm; 28-32 flagellomeres; maximum face width = minimum face width; fore wing vein m-cu postfurcal; ovipositor 2.0 X length of first tergite. Variation of paratype males. — Overall body color darker; 31 flagellomeres, an- nulus F27-F31, yellowish; head height 1.7X eye height. Cocoon. — Unknown. Material examined. — Holotype female: COSTA RICA: Puntarenas Province: San Vito, Estac. Biol. Las Alturas, 1750 m, IX- XI 1992, P. Hanson. Deposited in RMSEL. Paratypes: Puntarenas Province: San Vito, Estac. Biol. Las Alturas, 1500 m, P. Han- son; 9 females, 30 males, xi. 1991; 5 fe- males, 16 males, xii 1991; 3 females, 20 males, i. 1992; 4 males, ii. 1992; 9 females, 21 males, forest border, v. 1992; 19 males, in the forest v. 1992; 6 females, 35 males, traps #1 & #2, Malaise, vi. 1992; 1 male, vi-vii 1992; 2 female, 8 males, same data as holotype; 5 females, 1 male, 1700 m, ii- iv, 1993. San Jose Province: 1 female, Zur- qui de Moravia, 1600 m, vii. 1992, P. Han- son. Deposited in INBio, UCR, RMSEL. Distribution. — Known only from the ho- lotype and paratype localities at Biological Station Las Alturas, Puntarenas Province, and Zurqui de Moravia, San Jose Prov- ince. Biology. — Unknown. Comments. — Meteorus rogerblancoi shares with M. imo strongly twisted mandibles, completely joined borders of the first ter- gite, and small ocelli but can easily be sep- arated from this species by its complete occipital carina. This is the most numer- ous of the species collected from Malaise traps, with nearly 200 individuals trapped at Biological Station Las Alturas during 1991-1993, and one individual from Zur- qui de Moravia. Etymologi/. — This species is named in honor of Sr. Roger Blanco in recognition of his tenacious and omnipresent guardi- anship of the coordination of the ACG Re- search Program, from its inception in 1990 to the present. Meteorus rubens (Nees von Esenbeck) For most of this century New World specimens of M. rubens were treated under the name Meteorus imlgaris (Cresson) (Muesebeck 1923) and Meteorus leviventris (Wesmael) (Shenefelt 1969; Marsh 1979). The correct taxonomic status of M. rubens was clarified by Huddleston (1980). This species was redescribed by Huddleston (1980) and Maeto (1990a), and the Costa Rican specimens do not differ from these descriptions. Material examined.— COST A RICA: Guanacaste Province: 5 females, 3 males, ACG, Sector Orosi, Est. Maritza, reared 204 Journal of Hymenoptera Research from Megalopygidae, larva coll. 10 March, wasps eclosed 23 March 1992, Janzen & Hallwachs 92-SRNP-549. Deposited in IN- Bio, UCR, RMSEL. Cocoon. — Brown, wooly, ovoid; length 3.9^.7 mm, 1.7-1.9 mm wide medially; in mass of 30 (not all adults eclosed) cocoons, most terminating in a nipple-like projec- tion. Distribution. — In Costa Rica known only from the ACG, Guanacaste Province. Biology. — 30 wasp larvae emerged from a last instar megalopygid larva feeding on Ardisia revoluta Kunth (Myrsinaceae). The host larva was collected on 10 March 1992, the wasps pupated on 19 March, and the wasps eclosed 23 March 1992 (only 8 adults eclosed). Comments. — Meteorus rubens is a wide- spread species, Holarctic in distribution, and also recorded from Mexico and Brazil (Shenefelt 1969). Meteorus rubens shares with M. dimidia- tus and M. pseudodimidiatus strongly twist- ed mandibles and ventral borders of first tergite not joined completely. Meteorus rubens can easily be separated from these species by its indistinct and broad notauli, and poorly defined mesonotal lobes. This is the first record of any species of Meteor- us being reared from a megalopygid. Meteorus sterictae Zitani, new species Holotype female. — Body color: yellow ex- cept antennae brown; propodeum brown dorso-medially; hind leg with femur brown apically, tibia and tarsus brown; first tergite dark brown. Body length: 4.6 mm. Head: antenna with 28 flagellomeres; flagellar length /width ratios as follows: Fl = 3.6, F2 = 3.2, F3 = 2.3, F24 = 3.0, F25 = 2.0, F26 = 3.0, F27 = 2.0, F28 3.0; head 1.1 X wider than high, head height 1.5X eye height; eyes large, nearly parallel in anterior view; maximum face width 1.3X minimum face width; minimum face width 1.1 X clypeus width; malar space length 0.7X mandible width basally; ocelli large, OCOD = OCD; face punctate, with faint transverse carinae near base of anten- nae; clypeus rugulose; mandible strongly twisted (as in Fig. 9); occipital carina com- plete (as in Fig. 19). Mesosoma: notauli not distinct, broad, rugulose, and meso- notal lobes not well-defined (as in Fig. 11), mesoscutum polished, punctate laterally scutellar furrow with 3 distinct carinae, mesopleuron polished, punctate; sternau- lus rugose, broad and long (as in Fig. 13); propodeum areolate-rugose, median de- pression present. Legs: hind coxa rugu- lose; tarsal claw lobed, strongly curved (as in Fig. 22). Wings: forewing length 3.4 mm; vein m-cu antefurcal; 3RSa 1.6x length of r; Metasoma: (as in Fig. 3); ven- tral borders joined completely along basal Vi of segment (as in Fig. 4); first tergite with dorsal, lateral, longitudinal costae which converge posteriorly, medially ru- gulose; ovipositor long, 3.0 x longer than first tergite. Variation of paratype female. — yellow ar- eas may be orange; body length 3.9 mm; forewing length 3.3 mm; forewing vein 3RSa 2.0 X length of vein r; ovipositor 3.4 X longer than first tergite Variation of paratype males. — body color as in females and propodeum almost com- pletely brown; 29 flagellomeres; body length 3.8 mm; head height 1.8 X eye height; malar space length 1.5x basal width of mandible. Cocoon. — Brown, ovoid; length 4.6 mm, 1.9 mm wide medially; apparently not at- tached to substrate. Material examined. — Holotype female: COSTA RICA: Guanacaste Province: ACG, [Sector] Santa Rosa, dry forest, reared from Stericta albifasciata (Druce) lar- va, 1996, Janzen & Hallwachs voucher 96- SRNP-9074. Deposited in RMSEL. Para- types: Guanacaste Province: 1 male, same data as holotype; 1 female, same data as holotype except Janzen & Hallwachs voucher 96-SRNP-9103; 1 male with co- coon, same data as holotype except Janzen & Hallwachs voucher 96-SRNP-9102. De- posited in RMSEL. Volume 7, Number 2, 1998 205 Distribution. — Known only from the ho- lotype and paratype localities in the dry forest of the ACG, Guanacaste Province, Costa Rica. Biolog}/. — Meteorus sterictae attacks a py- ralid, Sterictn albifnsciata, which is a gre- garious webber on Ocotea veraguensis (Meisn.) Mez in the Santa Rosa dry forest. Meteorus sterictae is solitary; each wasp emerged from a single last instar larva. Comments. — Meteorus sterictae resembles M. camilocamargoi and M. desmiae, by shar- ing the following characters: strongly twisted mandibles, completely joined ven- tral borders of the first tergite, large ocelli, long ovipositor, and lobed tarsal claws. However, M. sterictae can be distinguished from these two species by its laterally flat- tened first tergite, rugulose sculpturing on the dorsal surface of the first tergite, and smaller eyes. The rugulose sculpturing may give the false appearance of dor- sopes. Etymology. — This species is named after the genus of the pyralid caterpillar it at- tacks, Stericta. Meteorus uno Zitani, new species Holoti/pe female. — Body color: dark brown except head orange, area between ocelli black; pronotum orange dorsally, white ventrally; mesonotum orange; pro- podeum white laterally and posteriorly; fore and mid legs pale yellow; first tergite white basally; metasoma white laterally and ventrally; wings dark, infused with brown pigment. Body length: 4.3 mm. Head: antenna with 28 flagellomeres; fla- gellar length /width ratios as follows: Fl = 2.7, F2 = 2.3, F3 = 2.0, F24 = 1.5, F25 = 1.5, F26 = 3.0, F27 = 2.0, F28 = 3.0; head 1.2x wider than high, head height 1.6X eye height; eye small but protuber- ant, nearly parallel in anterior view; max- imum face width 1.2x minimum face width; minimum face width 1.5X clypeus width; malar space length ].7x mandible width basally; ocelli small, OCOD 1.5X OCD; mandible strongly twisted (as in Fig. 9); occipital carina not complete, poor- ly defined medially (as in Fig. 21); vertex, in dorsal view, descending vertically be- hind lateral ocelli. Mesosoma: notauli ru- gulose, not distinct, and mesonotal lobes not well-defined; scutellar furrow with one median carina; mesopleuron polished, punctate; sternaulus rugulose, broad but not long; propodeum areolate-rugose, me- dian depression absent. Legs: hind coxa polished, punctate; tarsal claw with a small lobe or basal tooth, strongly curved. Wings: forewing length 4.0 mm; vein m- cu post furcal; r 0.6x length of 3RSa. Me- tasoma: (as in Fig. 3); ventral borders of first tergite joined completely along basal Vi of segment (as in Fig. 4); first tergite dorsally longitudinally costate, costae slightly convergent posteriorly; ovipositor short, thick at base, 1.7x longer than first tergite. Variation of females and males. — Un- known. Cocoon. — Unknown. Material examined. — Holotype female: COSTA RICA: Puntarenas Province: San Vito, Las Cruces, 1200msnm, VIII-IX 1988, Coll. P. Hanson. Deposited in RMSEL. Distribution. — Known only from the ho- lotype locality in Puntarenas Province. Biolog}/. — Unknown. Comments. — Meteorus uno shares with M. rogerhlancoi strongly twisted mandi- bles, completely joined borders of the first tergite, and small ocelli, but can easily be separated from this species by its incom- plete occipital carina. Etymology. — The specific epithet is an arbitrary combination of letters. Meteorus yamijuanum Zitani, new species (Fig. 14) Holotype female. — Body color: body dark brown except antenna light brown; pro- notum yellow anteriorly; propleuron yel- low; legs yellow except prothoracic tarsus light brown; mesothoracic tibia and tarsus light brown; metathoracic tibia and tarsus 206 Journal of Hymenoptera Research light brown; metasoma brown except first tergite yellow basally and apically, tergite 2 + 3 yellow basally. Body length: 3.5 mm. Head: antenna with 19 flagellomeres; fla- gellar length /width rahos as follows: Fl = 4.5, F2 = 3.5, F3 = 3.0, F15 = 2.0, F16 = 1.5, F17 = 1.5, F18 = 1.5, F19 = 3.0; head 1.1 X wider than high, head height 1.7x eye height; eye small, not protuber- ant, strongly convergent in anterior view; maximum face width 1.8X minimum face width; minimum face width 0.6 X clypeus width; malar space short, malar space length 0.5 X mandible width basally; ocelli small, OCOD 2.3 x OCD; face, clypeus polished, punctate; mandible not twisted; occipital carina complete (as in Fig. 19). Mesosoma notauli distinct, foveolate, and mesonotal lobes well-defined (as in Fig. 12); mesoscutum polished; scutellar fur- row with a median carina; mesopleuron polished; sternaulus foveolate, not broad (Fig. 14); propodeum carinate, with a dis- tinct median and transverse carina creat- ing very large defined areolae (as in Fig. 18), median depression absent. Legs: hind coxa polished; tarsal claws simple. Wings: forewing length 3.1 mm; vein m-cu ante- furcal; r 0.8 X length of 3RSa; second sub- marginal cell quadrate. Metasoma: ventral borders of first tergite separated basally, joined apically; first tergite polished with costae posteriorly; ovipositor 2.6 X longer than first tergite. Variation of paratype females. — Body length 2.9-3.5 mm; 18 flagellomeres; ovi- positor 2.6-3.0 X length of first tergite. Variation of males. — Unknown. Cocoon. — Unknown. Material examined. — Holotype female: COSTA RICA: San Jose Province: Zurqui de Moravia, 1600 m, vii 1990, Col Paul Hanson. Deposited in RMSEL. Paratypes: Puntarenas Province: 10 females, San Vito, Estac. Biol. Las Alturas, 1500 m, vi 1992, traps #1 & #2, Malaise, P. Hanson; 1 female, San Vito, Estac. Biol. Las Alturas, 1500 m, V 1992, forest border. Malaise, col. P. Hanson; 5 females, San Vito, Estac. Biol. Las Alturas, 1500 m, v 1992, in the forest, Malaise, col. P. Hanson; San Jose Prov- ince: 1 female, Zurqui de Moravia, 1600 m, vii 1990, col. Paul Hanson. Deposited in INBio, UCR, RMSEL. Distribution. — Known only from the ho- lotype and paratype localities in Puntar- enas and San Jose Provinces. Biology. — Unknown. Comments. — Meteorus yamijuanum close- ly resembles M. mariamartae, and M. me- galops. These species have untwisted man- dibles and a carinate propodeum (with large, defined areolae). Meteorus yamijuan- um can be distinguished from these other two species by its smaller eyes and the condition of the notauli. Etymology. — This species is named in honor of Yamilet Acosa and Juan Acosa, in recognition of their hospitality as care- takers of Pitilla Biological Station, ACG, Guanacaste Province in 1995 and 1996. DISCUSSION A close examination of M. congregatus revises our concept of Meteorus. ESEM im- ages show that the ventral borders of the first metasomal tergite are partially fused (Fig. 7) or completely fused (Fig. 8). Pre- viously the ventral borders of the first ter- gite in Meteorus have been known to be completely joined with a suture clearly visible (Muesebeck 1923; Huddleston 1980) (as in Fig. 4). This is the first docu- mentation of the total fusion of the ventral borders of the first metasomal tergite in Meteorus. Also, M. congregatus is unusual in that this character varies within the spe- cies (Figs. 7-8). The biological information for M. con- gregatus, M. alejandromasisi, and M. rubens is also of interest. Most meteorines are sol- itary (Shaw 1995). Meteorus congregatus is unusual in that it is highly gregarious (25- 250 individuals per host larva). Meteorus alejandromasisi is also gregarious (4-13 in- dividuals per host larva). These rearings provide the first records of Meteorus at- tacking the sphingid genera Manduca and Volume 7, Number 2, 1998 207 Unzela, and the family Hesperiidae. Along with M. papiliovorus, M. cotigregatus and M. alejandromasisi suspend their cocoons from very short threads. The new biological information present- ed here, combined with what is known about M. papiliovorus, provides a basis for speculation about a possible phylogeny of Meteorus. Species of Meteonis attacking concealed coleopteran larvae form unsus- pended cocoons in the protected environ- ment of the host, and these species have been hypothesized to be relatively primi- tive (Maeto 1990b). Species that suspend their cocoons also attack exposed hosts, and given that this suspended cocoon makes the parasitoid less accessible to po- tential enemies (Shaw and Huddleston 1991) it can be regarded as an adaptation to attacking exposed hosts. It follows that species that form a suspended cocoon are derived with respect to the more primitive species that form unsuspended cocoons in the host's pupation chamber. Multiple egg-laying gregarious parasit- ism probably evolved from solitary para- sitism in the Hymenoptera (Hanson and Gauld 1995). Gregariousness in Meteorus is presumed to be derived from solitary par- asitism. Gregarious species such as M. papiliovorus, M. congregatus, and M. alejan- dromasisi may have secondarily lost their longer cocoon threads. In addition to the short cocoon threads, these three species have cocoons that terminate with a dis- tinct nipple-like projection. Based on ob- servations of photographs of a highly gre- garious African species that forms a very organized cocoon mass, that species also has cocoons that terminate in a nipple-like projection. These gregarious species with this unique cocoon shape are hypothe- sized to be derived, and probably form a monophyletic group. Of the 4 new species with host associa- tions (M. desmiae, M. camilocamargoi, M. alejandromasisi, and M. sterictae), only one individual (a female of M. desmiae) was found in the examined Costa Rican Mal- aise trap material. This suggests that there may be quite a few more species of Meteo- rus in Costa Rica, and that host rearing, or some other method of trapping, will be necessary to find them. ACKNOWLEDGMENTS We would like to thank the following individuals and institutions: The curators of the institutions who lent specimens; Teresa Williams, Western Research Institute, for assistance with ESEM images; Jeffrey A. Lockwood, Nancy L. Stanton and R. Greg Thorn for helpful comments on the manuscript; The Area de Conservacion Guanacaste for support of much of the rearing and Malaise trapping; INBio and the National System of Conservation Areas (SINAC) for the u.se of their biological resources; the Norwegian Agency for International Development (NORAD) for financing the publication. Special thanks to Paul Hanson for so many years of maintaining Malaise traps and sorting specimens in Costa Rica. The following provided funding for this research: University of Wyoming Office of Research; UW Col- lege of Agriculture and Office of Research; UW De- partment of Plant, Soil, and Insect Sciences; UW In- sect Museum; UW Graduate School; UW McNair Scholars Program; UW Program for the Environment and Nahjral Resources; NSF-BSR 90-24770 and DEB 94-00829 grants to D.H. Janzen. Figure credits: 1-24, N. Zitani, S. Shaw, and T. Wil- liams; 25-28, D.H. Janzen. LITERATURE CITED Achterberg, C. van. 1979. A revision of the subfamily Zelinae auct. (Hymenoptera: Braconidae). Tijdschrift voor Entomologie 122:241—479. Cresson, E. T. 1872. Descriptions of North American Hymenoptera, No. 3. Canadian Entonicilcf;ist. 4:81- 84. Haliday, A. H. 1835. Essay on parasitic Hymenoptera of the Icltncumoncf Adsciti. Entomological Magazine 3:20-45. Hanson, P. E., and Gauld, I. D., eds. 1995. The Hy- inenoptera of Costa Rica. Oxford University Press, New York. 893 pp. Harris, R. A. 1979. A glossary of surface sculpturing. Occasional Papers in Entomolog]/ 28:1-31. Huddleston, T. 1980. A revision of the western Pale- arctic species of the genus Meteonis (Hymenop- tera: Braconidae). Bulletin of the British Museum (Natural History): Entomohgi/ 41:1-58. Huddleston, T. 1983. Meteorus (Hymenoptera: Bra- conidae) of Australia and New Guinea. Si/sfi'm- atic Entomohgi/ 8: 393^20. Maeto, K. 1989. Systematic studies on the tribe Me- teorini (Hymenoptera, Braconidae) from Japan. 208 Journal of Hymenqptera Research V. The pulchricoriiK group of the genus Meteorus (1). Japanese journal of Entcmwiogy 57:581-595. Maeto, K. 1990a. Systematic studies on the tribe Me- teorini (Hymenoptera, Braconidae) from Japan. VII. The groups of Meteorus letericus and M. rub- ens. Japanese Journal of Entomology 58:81-94. Maeto, K. 1990b. Phylogenetic relationships and host associations of the subfamily Meteorinae Cresson (Hymenoptera, Braconidae). Japmnese Journal of Entomology 58:383-396. Marsh, P. M. 1979. Family Braconidae. Pp. 144-313, In: Krombein, K.V., P. D. Hurd Jr., D. R. Smith, and B. D. Burks |eds.]. Catalog of Hymenoptera in America North of Mexico, Smithsonian Institution Press, Washington, D.C. Marsh, P. M., S. R. Shaw, and R. A. Wharton. 1987. An identification manual for the North American genera of the Family Braconidae (Hymenoptera). Memoirs of the Entonu^logical Society of Washington 13:1-98. Muesebeck, C. F. W. 1923. A revision of the North American species of ichneumon-flies belonging to the genus Meteorus Haliday. Proceedings of the United States National Museum 63: 1^44. Muesebeck, C. F. W. 1939. Five new species of Mefco- rus (Hymenoptera: Braconidae). Proceedings of the Entomological Society of Washington 41:83-87. Nixon. G. E. J. 1943. A synopsis of the African species of Meteorus (Hymenoptera: Braconidae). Bulletin of Entomological Rescarcli 34:53-64. Sharkey, M. J. 1988. A taxonomic revision of Alaba- grus (Hymenoptera: Braconidae). Bulletin of the British Museum (Natural History): Entomology 57: 311-437. Shaw, M. R., and Huddleston, T. 1991 Classification and biology of braconid wasps (Hymenoptera: Braconidae). Handbooks for the Identification of Brit- ish Insects 7:1-126. Shaw, S. R. 1985. A phylogenetic studv of the subfam- ilies Meteorinae and Euphorinae (Hymenoptera: Braconidae). Entomography 3:277-370. Shaw, S. R. 1988. Euphorine phylogeny: the evolution of diversity in host-utilization by parasitoid wasps (Hymenoptera: Braconidae). Ecological En- tomology 13:323-335. Shaw, S. R. 1995. The Braconidae, pp. 431-163. In: Hanson, P. and Gauld, I. (eds.). The Hymenoptera of Costa Rica. Oxford University Press, New York. 893 pp. Shaw, S. R. 1997. Subfamily Meteorinae, pp. 326-330. In: Wharton, R. A., P. M. Marsh and M. J. Shar- key, eds. Manual of the New World Genera of the Family Braconidae (Hymenoptera). Special Publi- cation of the International Society of Hymenop- terists. No. 1, 439 pp. Shenefelt R. D. 1969. Hymenopterorum Catalogus, Braconidae 1 Hybrizoninae, Euphorinae, Cos- mophorinae, Neoneurinae, Macrocentrinae, W. Junk, The Hague. 306 pp. Viereck, H. L. 1913. Descriptions of ten new genera and twenty-three new species of ichneumon- tlies. Proc. U.S. Natn. Mus. 44:555-568. Wharton, R. A., P. M. Marsh and M. J. Sharkey, eds. 1997. Manual of the Nezv World Genera of the Family Braconidae (Hymenoptera). Special Publication of the International Society of Hymenopterists, No. 1, 439 pp. Zitani, N. M., S. R. Shaw, and D. H. Janzen. 1997. Description and biology of a new species of Me- teorus Haliday (Hymenoptera: Braconidae, Me- teorinae) from Costa Rica, parasitizing larvae of Papilio and Paruies (Lepidoptera: Papilionidae). Journal of Hynwnoplera Research 6:178-185. J. HYM. RES. Vol. 7(2), 1998, pp. 209-256 The Species of Asaphes Walker from America North of Mexico, with Remarks on Extralimital Distributions and Taxa (Hymenoptera: Chalcidoidea, Pteromalidae) Gary A. P. Gibson and Veli Vikberg (GAPG) Agriculture and Agri-Food Canada, Eastern Cereal and Oilseed Research Centre, Biological Resources Program, K. W. Neatby Building, Ottawa, Ontario, Canada KIA OC6; (W) Liinalammintie 11 as. 6, FIN-14200 Turenki, Finland Abstract. — The species of Asaphes Walker (Pteromalidae: Asaphinae) are reviewed for the world and revised for America north of Mexico. Six species are recognized as valid in the region: Asaphes brevipetiolattis n. sp. (Canada, USA, Finland), A. californkus Girault, A. hirsutus n. sp. (Canada, USA, Mexico, Austria, Czech Republic, Finland, Greenland, Norway, Russia, Sweden), A. petiolatus Zetterstedt, A. suspensus (Nees), and A. indgaris Walker. Asaphes petiolatus (revised status) is removed from synonymy under A. vulgaris. Asaphes indicus (Bhatnagar), described from India, is removed from synonymy under A. vulgaris and along with A. nifipes Brues, A. lucens (Pro- vancher), A. fletcheri (Crawford), and A. americanus Girault is synonymized under A. suspensus (new synonymies). A lectotype is designated for A. americanus. Asaphes hiiebrichi (Brethes) and A. bonariensis (Brethes), described from Argentina and previously synonymized under A. fletcheri and A. lucens, respectively, are treated as nomina inqiiirenda. A key is given to distinguish males and females of the species in the region; diagnostic features are illustrated using scanning electron micrographs. Distribution and host data are summarized for each species in the Nearctic region, including records of A. brevipetiolatus parasitizing Choristoneura fumiferana (Clemens) (spruce bud- worm) and Neodipriou abietis (Harris) (balsam fir sawfly), likely as a hyperparasite. Remarks are also given on extralimital distributions of world species and generic composition of the subfamily, including the existence of an undescribed genus of Asaphinae in the Neotropical region, the likelihood of at least three undescribed species of Asaphes in regions other than the Nearctic, and the likelihood that most or all records of A. suspensus and A. vulgaris from the Neotropical region, and of A. vulgaris from at least the Afrotropical region of Africa, are misidentifications. Species of Asaphes Walker are known might prove to be a synonym of A. s!/fK- matcnnl.-Holoti/pc. female (CNCI, Type No. j^^p^^ National Park, Maligne Road, 7.VI11.50, F.I.S. 22267): CANADA, New Brunswick, F'ton [Frederic- ^o. A615A, ex. Metasyrphus lapponicus (3 5, NFRC). ton), em. June 23, 1966, R.C. Clark, AP.66-10-2, ex. obed, 21.VIII.50, F.I.sl No. A20H9C, ex. Metas\/rp^hus Syrphidae. Allotype, male (CNCI): same data as ho- lapponicus (4 9, NFRC). Rocky Mountain House, 23 lotype. Paratypes: CANADA. Yukon Territory: Ross mi. NW, 17.VIII.53, F.I.S. No. A781D, ex. syrphid. River, 16.IV-31.VIII.84, S.&J. Peck, aspen willow river host prob. aphid (1 9, 1 S, NFRC). Seebe, 13.VI.68, terrace (1 9, 2 S). British Columbia: Manning Pro- ex. Croiiuirtiuin coinivulnu- (iun^us). Pinus contort:! wyr. vincial Park, 2 km N Blackwall Peak, 49°07'N 121' latifolia, 688 1190 03. Manitoba: Picnic Bog, 6.VI.61, Volume 7, Number 2, 1998 217 F.I.S. (1 9, NFRC). Warkworth near Churchill, 29.VI.52, J.G. Chillcott (1 9 ). Quebec: Ct. Jette, RIF '40, 5103B, em. 18.VI1I, ex. Syrphidae sp.? (2 9). Forbes, 26.VI.52, L. Daviault, associated with Choris- toneiira fiimiferana, rearing no. 20 (5 9 ). Laniel, 9.VI1I.40, C.E. Atwood, experiment no. 12131-69, Cn- coecia fumiferana (2 9). Montcalm, em. 19.V1.11 from sp. B [?] worm, parasite of Tortris fumiferana (1 3 ). New Brunswick: same data as holotype (2 9, 3 d; 1 9 and 6 used for SEM). Fredericton, em. 8.VII.47, N.R. Brown, 18839-4d5, ex. Neocnenwiim coxalis (8 9, 1 u ). Newfoundland and Labrador: Gallants, 9.VI1.57, 57-0016(01) BIO (4 9, AFRC), 14.VI1.59, 59- 6009(01) A1-A9 {19,6 6 ), F.I.S., host: Neodiprion abie- tis. Uncertain locality (likely Newfoundland or New Brunswick): 23 [mi. ?) NW Rocky, 17.VIII.53, 53A781D, ex. syrphid (6 9, 8 d). USA. Alaska: Fair- banks, 23.V1.45, 45-19058, par. aphids on wild rasp- berry (1 9, USNM). Colorado: Green Mountain Falls, Canyon, 10.VIII.41, 10,000', H.H. Ross (1 9, INHS). Idaho: Bounds Creek, Fairfield, 5.1X.63, H. Flake & K. Lister, Hopk US no. 50-718a, host: attached pupae (Syrphidae) (8 9, USNM). Maine: Liberty, em. 8.VL48, parasite 48-Cll syrphid fly, beaten ex. Fir (2 9, USNM). Megalloway, 23.VL49, parasite 49-255, Neodiprion abietis, beaten ex. Fir (6 6, USNM). Wash- ington: Yakima Co., Green Lake Road, 31.V1II.94, K.S. Pike, ex. Bhicus sp. or ichneumonid, from Cinara chi- nookiana or unknown on Abies tasiocarpa (subalpine fir) (1 3, WSUC). EXTRALIMITAL— FINLAND (FENNIA, SUOMI). Savonia australis [= Sa, ESJ: Ristiina, 6826:502, 29.V11.83, M. Koponen (1 6, DAZH). Ostrobottnia borealis, N part [= ObN]: Pello, Pentik, 7417:368, 28.VL97, V. Vikberg (1 9, VVPC). Kuusamo [= Ks]: Kuusamo, 7358:596, 27.V1.82, M. Koponen (1 9, DAZH). Etymology. — From the Latin brevis, short, and petiolus, stalk, in reference to the transverse petiole of this species. Female. — Head and mesosoma black with obscure metallic green luster under some angles of light; legs entirely black or, more often, black or dark brown with ex- treme apex of femora, extreme base and apex of tibiae to entire tibiae, and tarsi lighter in color, rufous to yellowish. Head subquadrate in frontal view (Fig. 1), width at most 1.2 times height, and in lateral view lower face abruptly to almost right- angled relative to upper face (Fig. 2); in- terorbital region in dorsal view deeply, broadly concave (Fig. 13); gena as long as or slightly longer than eye width and at least 0.72 eye length; dorsal margin of to- rulus distinctly below lower orbit (Figs. 1, 2). Antenna (Fig. 19) with pedicel length about 3 times greatest width; funicle with fu, ring-like, fu.-fuj subquadrate (Fig. 20), and fus-fu„ increasingly transverse. Me- soscutum (Fig. 43) with mesoscutal lateral lobes broadly bare medially, and with fine engraved net-like sculpture over bare area. Scutellum mostly bare except along extreme anterior and lateral margins (Fig. 43); frenum with distinct, uniform, en- graved net-like sculpture similar to scutel- lum (Fig. 50). Metapleuron with at most 5 setae anteroventrally, these setae mostly in line along base of metapleural flange. Forewing with basal cell evenly setose (Fig. 73); disc with narrow speculum, the speculum closed basally by setae and with one or more lines of setae immediately be- hind submarginal vein (Fig. 73). Petiole in dorsal view distinctly transverse, length 1.4-1.7 times width, reticulate with irreg- ular longitudinal carinae or stronger keels (Fig. 49). Male. — Antenna with scape, pedicel and fu, yellow to yellowish-orange, contrast- ing distinctly in color with rest of dark fla- gellum; color pattern otherwise similar to female except head and mesosoma usually with more distinct metallic green luster; legs usually more extensively light-col- ored, but at least hind leg brownish ba- sally, including trochanter. Scape (Figs. 31-32) with dorsal margin slightly convex, ventral margin sinuate so as to be broad- est subbasally and tapered to apex; inner and outer surfaces uniformly setose and sculptured. Pedicel (Figs. 31-32) length about 3 times width and about 0.40-0.45 scape length. Flagellum length subequal to head width; funicle (Fig. 31) with all segments at least slightly transverse or with one or more of fu,_, quadrate to very slightly longer than wide. Structure oth- erwise similar to female except head in dorsal view with interorbital region even more deeply, broadly concave (Fig. 14); malar space only about 0.77-0.87 eye 218 Journal of Hymenoptera Research width and 0.5-0.6 eye length; petiole usu- ally only slightly transverse (Fig. 55); and sculpture of frenum often distinctly finer than on scutellum (Fig. 55). Setal pattern similar to female except metapleuron with at most 2 short setae. Distribution. — Holarctic; in North Amer- ica transcontinental within the Boreal re- gion and extending south along the Cas- cade and Rocky Mountains into Colorado (Fig. 79). Biology. — Asaphes brei'ipetiolatus is a par- asite of Neocnemodou coxalis (Curran), Meta- syrphus lapponicus (Zetterstedt) and possi- bly other Syrphidae (Diptera) based on as- sociated host pupal remains and label data. The balsam fir sawfly is also indi- cated as a host because two cocoons of Neodiprion abietis (Harris), one with two emergence holes and the other torn in half, are preserved with specimens. Labels from three different rearings in Quebec in- dicate that A. brevipetiolatus is also a par- asite of the spruce budworm, Choristoneu- ra fumiferana (Clemens). Although there are no associated host pupae, one of the two Laniel specimens has an unidentified braconid cocoon preserved with it, indi- cating that A. brevipetiolatus could be a hy- perparasite of the spruce budworm through braconid primary parasites. There is also a single record, based on label data, of the species parasitizing an unidentified aphid on wild raspberry, plus a rearing from either Cinara chinookiana Hottes or an unknown aphid via either Blacus sp. (Bra- conidae) or an unidentified ichneumonid. More than one emergence hole in some syrphid puparia and one balsam fir sawfly cocoon, plus multiple-mounted speci- mens, labelling, or both, for these two hosts and two of the three spruce bud- worm records indicate that A. brevipetiola- tus is often gregarious when parasitizing larger non-aphid hosts. It remains to be demonstrated more conclusively that the balsam fir sawfly and spruce budworm are more than just accidental hosts, and whether A. brevipetiolatus is a primary or hyperparasite of these two species. How- ever, based on current evidence it seems likely that host acceptance for A. brevipet- iolatus includes not only syrphid larvae but other relatively large, oblong, brown- ish pupae or cocoons on coniferous trees. This suggests an evolutionary progression in Asaphes from parasitism of hymenop- terous primary parasites of aphids, to syr- phid-larvae predators of aphids and their parasites, to other hosts that resemble syr- phid larvae. Remarks. — Males are easily distin- guished by structure and color of their scape and pedicel, both are yellowish in distinct contrast to the flagellum beyond fu,. Also, the pedicel is conspicuously long, almost half the length of the scape, and the scape is widened subbasally so that its ventral margin is sinuate (Figs. 31, 32). Females are distinguished by a combi- nation of features, including a sculptured frenum (Fig. 50), dark trochanters and tro- chantelli, closed speculum (Fig. 73), trans- verse petiole (Fig. 49), and unique head structure (Figs. 1, 2). Because the legs are almost uniformly dark basally, females most closely resemble those of A. vulgaris, A. petiolatus and A. hirsutus, but are distin- guished from all three species by de- scribed head structure. Absence of a broad speculum further differentiates females from those of A. vulgaris. Those A. brevi- pietiolatus females with setae on the meta- pleuron are more likely to be mistaken for females of A hirsutus, particularly because some A. hirsutus females have a slightly transverse petiole (Fig. 53) and some have the frenum extensively, though finely sculptured. However, in A. hirsutus the frenum is always more or less broadly smooth immediately behind the frenal sul- cus even though often sculptured poste- riorly (Fig. 54). Only head structure readi- ly differentiates females of A. brevipetiola- tus and A. petiolatus. Volume 7, Number 2, 1998 219 Asaphes califomicus Girault (Figs. 7, 8, 21, 22, 33, 34, 44, 56, 63, 67, 68, 76) Asn/'/ies califomicus Girault 1917[330]: 1. Type data: USA: California [Spreckels], [25] Sep- tember [1916]; reared from [parasitized] Aphis nunicis; [Chittenden No. 1671]; C.F. Stahl [collector]. Sex described: female. Ho- lotype by monotypy; USNM, type no. 21411. Female. — Head and mesosoma dark with varying intensity of olive green me- tallic luster under some angles of light; legs with trochanters and trochantelli al- most always uniformly yellowish to yel- lowish-brown, at least middle and hind femora black except often apically, and tibiae and tarsi usually yellowish to red- dish-brown. Head transverse-triangular in frontal view (Fig. 7), width at least 1.25 times height, and in lateral view lower face evenly curved into upper face (Fig. 8); interorbital region in dorsal view relative- ly shallowly concave; gena length about 0.62-0.80 eye width and 0.50-0.65 eye length; dorsal margin of torulus approxi- mately in line with lower orbit (Figs. 7, 8). Antenna (Figs. 21, 22) with pedicel length at most about 2.5 times width; funicle with fu, ring-like, fu_,-fu4 quadrate to trans- verse, and fu5-fu8 increasingly transverse. Mesoscutum (Fig. 44) with lateral lobes broadly bare medially, and either smooth and shiny or with fine, engraved net-like sculpture over bare area. Scutellum (Fig. 44) mostly bare except along extreme an- terior and lateral margins; frenum smooth and shiny except finely carinate laterally. Metapleuron bare. Forewing with basal cell evenly setose to bare except for single row of setae on dorsal surface (Fig. 68); disc often without distinct speculum, dis- tance between either basal setal line or submarginal vein and first setal line on disc at most about equal to distance be- tween first and third setal lines on disc (Fig. 67) or, if with large and conspicuous speculum, then bare region with at least 2 dorsal setae within bare region, the setae usually separated from basal vein and / or submarginal vein by distance about equal to length of setae or less (Fig. 68). Petiole at least quadrate and usually slightly lon- ger than wide, but less than 1.3 times as long as wide (Fig. 44), reticulate with ir- regular longitudinal carinae (Fig. 44) or stronger keels. Male. — Color pattern similar to female except legs sometimes entirely yellow (more commonly with at least metafemur infuscate to black); antenna usually uni- formly brown to black except sometimes apex of pedicel and fu, lighter in color. Scape (Figs. 33, 34) length about 3.5-5.0 times width, ovoid to spindle-shaped, with dorsal and ventral margins convex to subparallel over most of length; outer sur- face with subbasal, ovoid to elongate-lan- ceolate, microsetose sensory region (Figs. 33, 34a); inner surface with line of ven- trally directed setae, often from along fine ridge, parallelling ventral margin of scape below midline and mediolongitudinal bare, smooth band (Fig. 34b). Combined length of pedicel and flagellum less than 2.5 times scape length (Fig. 33) and at most subequal in length to head width; fu- nicle with all segments at least slightly transverse or with one or more of fu,_s quadrate to very slightly longer than wide. Structure and setal pattern other- wise similar to female except petiole al- ways distinctly (about 1.25-1.80 times) longer tlian wide (Fig. 56) and sometimes entirely reticulate without longitudinal ca- rinae. Distribution. — Restricted to western North America within the Nearctic region, except for one anomalous record from Georgia (see further under 'Remarks') (Fig. 82). CANADA. Yukon Territory, British Columbia, Alberta. USA. Alaska, Arizona, California, Colorado, ? Georgia, Idaho, Kansas, New Mexico, Nevada, Or- egon, Washington, Utah. EXTRALIMI- TAL. MEXICO (BMNH: 9, (5;CNCI: 9,3; EMEC: ?), ARGENTINA (MLPA: 9, S; TAMU: 9, 6); BOLIVIA (USNM: 9), 220 Journal of Hymenoptera Research Table 1. Host information for Asaphcf cnlifoniiciis based on observed specimens; a question mark follows rearings or identifications indicated as questionable on the labels. Unless otherwise footnoted, all 'aphid or other hosts' are Homoptera (Aphidoidea: Aphididae) and all 'associated primary hosts' are Hymenoptera (Braconidae: Aphidiinae). Aphid or other Hosts Associated Primarv Hosts Museum Acronvms Acyrthosiphon lactucae (Pas- serin) A. pisuin (Harris) Amphoroplwrii rubi (Kalten- bach) Aphis sp. A. ccanoDii Clarke A. fabnc Scopoli A. goss\/pii Glover A. hederae pseiniohetierae Theobald A. helianthi Monell A. holociisci Robinson A. rumicis L. A. sambuci L. A. spiraecola Patch A. I'arians Patch Aphthargelia symphoricarpi (Thomas) Brachycauiius caniui (L.) 6. helkhrysi (Kaltenbach) 6. tragopogonis (Kaltenbach) Brachycorynelln aspaiagi (Mordvilko) Braggia sp. B. eriogoni (Cowen) Brevicoryne hrassicae (L.) Capitophorus claeagni (del Guercio) Ciwariella acgopodii (Scopoli) C. paslinacac (L.) Acuiitliociuidus sp. ? Ap^liidius sp. A. alius Muesebeck A. nigripes Ashmead A. smithi Sharma & Subba Rao Diacretus sp. Ephedrus sp. Lysiphlebus {Adialytus) salicaphis (Fitch) L. (Phlebus) testaceipes (Cresson) Pauesia cnlifornicus (Ashmead) Praon sp. T. (Trioxys) ccmplniwtiis Quilis T. (Trioxys) curvicaudtis Mackauer Praon sp. Apliidius sp., A. eroi Haliday, A. smi- tlii Sharma & Subba Rao Praon pcquodorum Viereck Lysiphlebus sp. Aphidius sp.; Lysiphlebus (Phlebus) testaceipes (Cresson) Lysiphlebus (Phlebus) testaceipes (Cresson) Ephedrus californicus Baker; Lysiphle- bus (Phlebus) testaceipes (Cresson); Praon sp. Alhxysta sp. Lysiphlebus (Phlebus) testaceipes (Cresson); Praon sp. Diaeretiella rapae (M'Intosh) Lysiphlebus (Phlebus) testaceipes (Cresson) Diaeretiella rapae (M'Intosh) VVA WSUC CA USNM, UCRC CA UCRC CA UCRC CA USNM, EMEC CA UCRC CA EMEC CA EMEC CA UCRC CA USNM, EMEC CA UCRC CA EMEC CA EMEC WA WSUC CA, OR, WA UCRC, EMEC, USNM, WSUC WA WSUC CA EMEC, UCRC WA WSUC WA WSUC CA, NM EMEC, UCRC, USNM CA UCRC WA WSUC WA WSUC CA USNM WA WSUC WA WSUC WA WSUC WA WSUC WA WSUC CA EMEC WA WSUC WA WSUC Aphidius salicis Haliday Aphidius salicis Haliday WA WA CA, OR, WA WA CA, WA VVA WSUC WSUC EMEC, OSUC, WSUC WSUC CASC, WSUC WSUC Volume 7, Number 2, 1998 Table 1. Continued. 221 Aphid or other Hosts Associated IVimarv Hosts Museum Acronvms Chaeto$ip!wn {Peiilnlriclwpuf) fragarnefolii (Cockerell) Chaitophorus pcpulifotii (Essig) Chaitophorus salkkola Essig Cinara ponderosae (Williams) Diuraphis iioxia (Mordvilko) Dysapliis (Pomapliii) plaii- tagiuca (Passerini)? Elatobium aWietiiium (Walker) Ericaphis gentneri (Mason) Essigella sp. E. californica (Essig) E. pini Wilson Eucalliptcni!. tiliae (L.) Euceraphis punctipcnnis (Zet- terstedt) Flabellomicrosiplnmi knOivl- toni Smith ? Hayhurstia atriplkis (L.) Hyiihpterus pruni (Geoffroy) Hi/peromyzus {Neoiiasonovia) nigrkornis (Knowlton) lllinoia sp. /. nzalene (Mason) /. tirwikndri (Monell) /. niorrisL'iti (Swain) /. sinipsoui (MacGillivrav) /. subviridis (MacDougall) MacrosiphpiikUa ludcivkianac (Oestlund) Macrosip'hum sp. M. dydcsmithi Robinson M. creelii Davis M. euphorhiac (Thomas) M. parvifclii Richards M. rosae (L.) Metopu'lcphiiiKm dirfwdum (Walker) Mkrolophium cariwsuni (Buckton) Myzocallh sp. M. coryli (Goeze) Myzus [Nectarcfiphcn) persi- cae (Sulzer) Nasoucvui nquilegiae (Essig) Aphidius sp. Ephcdnca sp. Pauesia sp. Diaretklla rapae (M'lntosh) Praon uiikuni Smith Aphidius sp. Alloxysta sp.; Prnon uiuciim Smith Diaeretus sp. Aphidius sp.; Diaeretus sp. Priwu sp. Diaretielln rapae (M'lntosh) Praon sp. Praon sp. Alloxysta sp.; Aphidius sp. Praon sp. Aphidius sp. Aphidius sp.; Praon sp. Praon sp. Alloxysta sp.; Aphelinus sp.'; Af/ii- rfius nigripes Ashmead Aphelinus sp.'; Aphidius sp., /4. a/iks Muesebeck; Praon sp. Praon sp. T. (Trioxys) pallidas (Haliday) T. (Trioxys) pallidas (Haliday) Aphidius sp., A. matricarme Haliday; Diaeretus rapae (M'lntosh); Lysiphlc bus (Phlebus) testaceipes (Cresson); Praon sp., P. unicum Smith CA WA WA WA WA WA AB, CA, WA WA OR, UCRC, EMEC WSUC CA EMEC, USNM MT WSUC WA WSUC WA WSUC EC UCRC, NFRC WA WSUC CA EMEC CA UCRC CA UCRC, USNM CA EMEC CA USNM WSUC WA WSUC CA EMEC WA WSUC WA WSUC WA WSUC CA CNCl, EMEC CA EMEC BC EMEC WA WSUC CA UCRC BC, CA, WA CNCI, EMEC, WSUC WA WSUC WA WSUC CA EMEC, UCRC BC CNCI CA, WA CASC, EMEC, UCRC USNM, WSUC WA WSUC WSUC WSUC WSUC CNCI, EMEC, UCRC, USNM, WSUC WSUC 222 Table 1. Continued. Journal of Hymenoptera Research AphKl or i.ther Hostb Associated Priniarv Hosts Museum Acron\'ms Obtuskauiia sp. O. artemisiphita (Knowlton & Allen) ? O. coweni (Hunter) O. filifcliac (Gillette & Palm- er)? Oi'atus cratacgarius (Walk- er)? Phytomyza ilicis Curtis- Phowdon humidi (Schrank) Pseudoepameihaphis tridcnta- tae (Wilson) Rhodobiuni poroiiis (Sander- son) Rhopnloniyzitf (Jiidciikon) /on- iccrac (Siebold) R)ioptiloiiphuiii iiiscrtiini (Walker) R. viaidis (Fitch) R. padi (L.) Schizolachnus piiurndiatnc (Davidson) Sitobium ', Australia (I. Naumann). AFRC Atlantic Forestry Research Centre, Canadi- 240 Journal of Hymenoptera Research ■■i^M ^B||H ^^F ' ' /jM^^^SSf^SS^St^^Ss^^^^^^Ss/^t ^^^ ^S^^^^^^^^^^Kk ^ ^1 H^^9 ^SH^^^Pl^^^H ^^P^^JhR ^^^^^^^^^bN^ ^^ ~ ^^%'I^HhK^ ^^^■^^^H ^^^^^^^^^^^^^^^^^^H^^l I^^^O Figs. 1-6. Head: 1 and 2, /4sflp/ics brevipetiolatus (9); 3 and 4, /I. bivvipctiolntu^ (6); 5 and 6, /^. /!irs»f»s (9). (abbreviation: cly = clypeus; scale bars = microns) Volume 7, Number 1, 1998 241 Figs. 7-12. Head (9): 7 and 8, Aia\>hes catifornicKs; 9 and 10, A. susjvmsks; 1 1 and 12, A. vulgaris, (abbreviation: gc = genal carina; scale bars = microns) 242 Journal of Hymenoptera Research ■ 1 ■ i 1 ■ m ^^^ 1 1 I ■ I t^7J Figs. 13-18. 13-17. Head, dorsal; 13, Asnpluv hrcvifetiotatiif. (9); 14, A. hirvii't'tu^lntiis (d); 15, A. vulgaris (9); 16, A. suspensus (9); 17, A. hirsutus (9). 18. Clypeus and mandibles, A. Iiir^iifii^ (9). (abbreviation: occ = occipital carina; scale bars = microns) Volume 7, Number 2, 1998 243 Figs. 19-24. 19 and 20. Asa^'hef brevipetiolatus (9): 19, antenna; 20, basal 4 flagellar segments. 21 and 22. A. californicus (9): 21, antenna; 22, basal 4 flagellar segments. 23 and 24. A. hir^utuf. (9): 23, antenna; 24, basal 4 flagellar segments, (abbreviation: fu, = first funicular segment; scale bars = microns) 244 Journal of Hymenoptera Research Figs. 25-30. 25 and 26. Asnphcf pftiohitm (V); 25, antenn.i; 26, b.isal 4 fldgoUar segments. 27 and 28. A. suspoi.sKs (9): 27, antenna; 28, basal 4 flagellar .segments. 29 and 30. A. vulgaris {9): 29, antenna; 30, basal 4 flagellar segments, (scale bars = microns) Volume 7, Number 2, 1998 245 Figs. 31-36. 31 and 32. Asaphes hreinpetiolatus (cJ): 31, antenna; 32, scape — a, outer view, b, inner view. 33 and 34. A. californicus (S): 33, antenna; 34, scape — a, outer view, b, inner view. 35 and 36. A. hirsutus (d): 35, antenna; 36, scape — a, outer view (arrow points to flat surface), b, inner view, (scale bars = microns) 246 Journal of Hymenoitera Research Figs. 37-42. 37 and 38. Asaplws petiolatus (cj): 37, antenna; 38, scape — a, outer view, b, inner view. 34 and 40. A. suspensus (6): 39, antenna; 40, scape — a, outer view, b, inner view. 41 and 42. A. vulgaris (cj): 41, antenna; 42, scape — a, outer view, b, inner view, (scale bars = microns) Volume 7, Number 2, 1998 247 Figs. 43-48. Mesosoma, dorsal (9). 43, Asaphes brevipetiolatus; 44, A. califormcuf, 45, A. Imsuhis; 46, A. pct- wlatiis; 47, A. smpciisus; 48, A. vulgaris, (abbreviation: sss = scutoscutellar suture; scale bars = microns) 248 Journal of Hymenoptera Research Figs. 49-54. 49 and 50. Aaiiphes hmnpetiolcitiis (9): 49, frenum to petiole; 50, sculpture of scutellum and frenum. 51 and 52. A. petiolatua (9): 51, frenum to petiole; 52, sculpture of scutellum and frenum. 5.1 and ,54. A. /iirsKfus (9): 53, frenum to petiole; 54, sculpture of scutellum and frenum. (abbreviations: fre = frenum, set = scutellum; scale bars = microns) Volume 7, Number 2, 1998 249 Figs. 55-60. Frenum to petiole (d): 55, A. breinyctiolatus; 56, A. catifonitcus; 57, A. hirsiitus; 58, A. pefiolatus; 59, A. suspensus; 60, A. vulgaris, (scale bars = microns) 250 Journal of Hymenoptera Research Figs. 61-66. 61 and 62. Asaphci liir^iitiis: 61, lateral mcsosoma (9); 62, niL'tapleuron (c!). 63. A. cnUtcniicii-i: lateral mesosoma (d). 64-66. Metathorax to petiole, lateral (9) (arrow points to lateral sulcus ot petiole): 64, A. brevipeliolatu'^; 65, A. petwlnttis; 66, A. s»s;v«s»s. (scale bars = microns) Volume 7, Number 2, 1998 251 Figs. 67-72. Forewing (9): 67 and 68, Aaaphcs cahfonuctis; 69, A. ^uspensus; 70, A. vulgaris; 71, A. htnittus; 72, A. petiolatiis. (abbreviations: be = basal cell, bsl = basal setal line, cc = costal cell, smv = submarginal veini spc = speculum; scale bars = microns) 252 Journal of Hymenoptera Research Figs. 73-78. 73. Forewing: Asaphcs brcvifetiolahis ( 9 ). 74. Gaster: A. himiitii^ ( 9 ). 75. Petiole and base of gaster, ventral view: A. pctiolatus (9). 76. Petiole, ventral view: A. cnlifoniiciis (9). 77 and 78. A. stisff.ifui, base of gaster (9): 77, ventral view; 78, ventrolateral view, (scale bars = microns) Volume 7, Number 2, 1998 253 Figs. 79-80. Nearctic distribution: 79, Asaphes brevi- petiolatus (•) and Asaphes petiolatuf (A). 80. Asaphes hirsutus. Figs. 81-82. Nearctic distribution: 81, Asaphes sus- pensus. 82, Asaphes califcrnicus (•) and A. vulgaris (A). an Forest Service, Insect Collection, St. John's, NF (G. Smith). DCPC BMNH The Natural History Museum, London, En- gland (J. Noyes, Z. Boucek). EDUM CASC California Academy of Sciences, Depart- ment of Entomology, San Francisco, CA (D. EMEC Ubick, W. Pulawski). CISC California Insect Survey, Division of Ento- GNME mology and Parasitology, University of California, Berkeley, CA (R. Zuparko). HAPC CNCI Canadian National Collection of Insects and Arachnids, Ottawa, ON. INHS CUIC Cornell University Insect Collection, De- partment of Entomology, Cornell Univer- MLPA sity, Ithaca, NY (E. R. Hoebeke). DAZH Department of Applied Zoology, Universi- ty of Helsinki, Helsinki, Finland (M. Ko- ponen). Dominique Collet private collection. Ster- ling, AK (D. Collet). Department of Entomology, University of Manitoba, Winnipeg, MB (T. Gallowav). Essig Museum of Entomology, University of California, Berkeley, CA (R. Zuparko). Department of Entomology, Naturhisoriska Museet, Goeborg, Sweden (G. Andersson). Harry Anderson Private Collection, Hun- tington Beach, CA (H. Anderson). Illinois Natural History Survey Insect Col- lection, Champaign, IL (K.R. Zeiders). Museo de la Plata, Universidad Nacional de La Plata, Division Entomologia, La Pla- ta, Argentma (L. De Santis). 254 Journal of Hymenoptera Research MZLU Museum of Zoology, Lund University, Lund, Sweden (R. Danielsson, U. Garden- fors). NFRC Northern Forest Research Centre, Canadian Forest Service, Edmonton, AB (D. Wil- liams). PFRC Pacific Forestry Research Centre, Canadian Forest Service, Victoria, BC (L. Humble). OSUC Collection of Insects and Spiders, Ohio State University, Columbus, OH (A. Shar- kov). OSUO Department of Entomology Collection, Or- egon State University, Corvallis, OR (D.C. Darling). SMDV Spencer Museum, Department of Zoology, University of British Columbia, Vancouver, BC (K. Needham). TAMU Department of Entomology, Texas A&M University, College Station, TX (J. Woolley, E. Riley). UCDC The Bohart Museum, University of Califor- nia, Davis, CA (S. Heydon). UCRC UCR Entomological Training and Research Collection, University of California, River- side, CA (S. Triapitsyn). USNM United States National Entomological Col- lection, U.S. National Museum of Natural History, Washington, DC (E.E. Grissell). UZMH Zoological Museum, University of Helsin- ki, Helsinki, Finland (A. Albrecht, M. Ko- ponen). VVPC Veli Vikberg personal collection, Turenki, Finland (V. Vikberg). WSUC James Entomological Collection, Depart- ment of Entomology Collection, Washing- ton State University, Pullman, WA (K.S. Pike), and Tree Fruit Research Center, Washington State University, Wenatchee, WA (D. Carroll). ZMAS Zoological Museum, Academy of Sciences, Saint Petersburg, Russia (V.A. Trjapitzin). ZMUC Zoological Museum, University of Copen- hagen, Copenhagen, Denmark (B. Petersen). LITERATURE CITED Ashmead, W. H. 1904. Classification of the chalcid flies of the superfamily Chalcidoidea, with de- scriptions of new species in the Carnegie Muse- um, collected in South America by Herbert H. Smith. Memoirs of tlw Carncf^ie Miisi'iiin 1: i-ix + 225-551. Bakkendorf, O. 1955. Notes on Icelandic and Green- landic Chalcidoideous Hymenoptera. Eutomolo- giske Mcddek'lser 27: 135-162. Batulla, B. A. and A. G. Robinson. 1985(1984). Hy- menopterous parasitoids of aphids (Homoptera: Aphididae) in Manitoba. Proccciimgs of the Ento- mological Socicti/ of Manitoba 40: 30-38. Bhatnager, S. P. 1951. Descriptions of new and re- cords of known Chalcidoidea (Parasitic Hyme- noptera) from India. Indian journal of Agricultural Science 21: 155-178. Bocchino, F. J. and D. J. Sullivan, S.J. 1981. Effects of venoms from two aphid hyperparasitoids, Asa- phes tucens and Dendrocerus carpenteri (Hymenop- tera: Pteromalidae and Megaspilidae), on larvae of Aphidius smithi (Hymenoptera: Aphidiidae). Canadian Entomologist 113: 887-889. Bolte, K. B. 1996. Techniques for obtaining scanning electron micrographs of minute arthropods. Pro- ceedings of the Entomological Societi/ of Ontario 127: 67-87. Boucek, Z. 1974. On the Chalcidoidea (Hymenoptera) described by C. Rondani. Redia 55: 241-285. Boucek, Z. 1976. African Pteromalidae (Hymenop- tera): new taxa, synonymies and combinations. journal of the Entomological Societi/ of Southern Af- rica 39: 9-31. Boucek, Z. 1988. Australasian Chalcidoidea (Hymenop- tera). A biosystematic revision of genera of fourteen families, loith a reclassification of species. CAB In- ternational, Wallingford. 832 pp. Boucek, Z. and S. L. Heydon. 1997. Chapter 17. Pter- omalidae. Pages 541-692 in Gibson, G. A. P., J. T. Huber, and J. B. Woolley (eds). Annotated Keys to the Genera of Ncarctic Chalcidoidea (Hymenop- tera). National Research Council Canada, Re- search Press, Ottawa. 794 pp. Boucek, Z. and J.-Y. Rasplus. 1991. Illustrated key to West-Palearctic genera of Pteromalidae (Hymenop- tera: Chalcidoidea). Institut National de la Recher- che Agronomique, Paris. 140 pp. Boucek, Z., B. R. Subba Rao, and S. I. Farooqi. 1978. A preliminary review of Pteromalidae (Hyme- noptera) of India and adjacent countries. Oriental Insects 12: 433-468. Brethes, J. 1913. Himenopteros de la America Merid- ional. Anales del Museo Nacional de Historia Nat- ural de Buenos Aires 24: 35-166. Brethes, J. 1916. Hymenopteres parasites de I'Ame- rique Meridionale. Anales del Museo Nacional de Historia Natural de Buenos Aires 27: 401-130. Brown, N. R. and R. C. Clark. 1960. Studies of pred- ators of the balsam woolly aphid, Adelges picea (Ratz.) (Homoptera: Adelgidae). VIII. Syrphidae (Diptera). Canadian Entomologist 92: 801^11. Brues, C. T. 1908. Notes and descriptions of North American parasitic Hvmenoptera. VII. Bulletin of the Wisconsin Natural History Society 6: 160-163. Burks, B. D. 1958. Superfamily Chalcidoidea. Pages 62-84 (11 Krombein, K. V. (ed.). Hymenoptera of America North of Mexico. Synoptic Catalog. United States Department of Agriculture, Agricultural Monograph No. 2, First Supplement. Washington, D.C, U.S. Government Printing Office. 305 pp. Burks, B. D. 1964 (1963). The Provancher species of Volume 7, Number 2, 1998 255 Chalcidoidea (Hymenoptera). Camuiinn Entomol- ogist 95: 1254-1263. Burks, B. D. 1979. Family Pteromalidae. Pages 768- 835 in Krombein, k" V., P. D. Hurd, Jr., D. R. Smith, and B. D. Burks (eds). Catalog of Hymenop- tera in America North of Mexico. Vol. 1. Symphyta and Apocrita (Parasitica). Smithsonian Institution Press, Washington, D.C. i-xvi, 1-1198. Carew, W. P. and D. J. Sullivan. 1993. Interspecific parasitism between two aphid hyperparasitoids, Dendrocerus carpciiteri (Hymenoptera: Megaspili- dae) and Asaphes luccns (Hymenoptera: Pterom- alidae). Annals of the Entomological Society of America 86: 794-798. Crawford, J. C. 1909. Notes on some Chalcidoidea. Canadian Entomologist 41: 98-99. Curtis, J. 1842. Observations on the natural history and economy of various insects affecting the tur- nip-crops; including the plant-lice, maggots of flies, caterpillers of moths, etc. journal of the Royal Agricultural Society of England 3: 49-78. Dahlbom, A. G. 1857. Svenska Sma-Ichneumonernas familjer och slagten. Kongliga Si'cnska Vetcnskap- sakademiens Handlingnr 14: 289-298. Dahms, E. C. 1978. A checklist of the types of Aus- tralian Hymenoptera described by Alexandre Ar- sene Girault: 1. Introduction, acknowledgments, biography, bibliography and localities. Memoirs of the Queensland Museum 19: 127-190. Delucchi, V. 1955. Notes sur les Pteromalides (Hym., Chalcidoidea). Memoircs dc la Societe royale beige d'Entomologie 27: 171-175. De Santis, L. 1960. Anotaciones sobre Calcidoideos Argentinos. II. (Hymenoptera). Revista de la Fa- cultad dc Agronomia (3a epoca) 36: 109-119. De Santis, L. 1967. Catdlogo dc los Himenoptcros Argen- tinos de la Serie Parasitica, Incluyendo Bethyloidea. Provincia de Buenos Aires Gobernacion, Comi- sion de Investigacion Cientifica, La Plata. 337 pp. De Santis, L. 1979. Catdlogo de los Himenoptcros Chal- cidoideos de America al sur de los Estados Unidos. Comision de Investigaciones Cientificas de la Provincia de Buenos Aires, Publicacion Especial, La Plata. 488 pp. De Santis, L. 1980. Catdlogo de los Himenoptcros Brasi- lei'ws de la Serie Parasitica Incluyendo Bethyloidea. Editora da Universidade Federal do Parana, Cur- itiba. 395 pp. De Santis, L. and P. Fidalgo. 1994. Catalogo de Hi- menopteros Calcidoideos. Serie de la Academin na- cional de Agronomia i/ Vetcrinaria 13: 1-154. Dzhanokmen, K. A. 1987. 5. Familv Pteromalidae (pteromalids). Pages 88-411 in Medvedev, G.S. (editor-in-chief). Keys to the insects of the European part of the USSR. Vol. III. Hymenoptera. Part //. Am- erind Publishing. Pvt. Ltd., New Delhi, i-xii + 1341 pp. lEnglish translation of Medvedev, G. S. (editor-in-chief). 1978. Handbook of the Insects of the European Part of the U.S.S.R. Vol. III. Hi/nwnoptera, Second part. Leningrad, Nauka (in Russian)). Eastop, V. F. and D. Hille Ris Lambers. 1976. Sun'cy of the zoorld's ap'hids. Dr. W. Junk b.v.. The Hague. 573 pp. Farooqi, S. I. and B. R. Subba Rao. 1986. Family Pter- omalidae. Pages 279-306 in Subba Rao, B. R. and M. Hayat (eds). The Chalcidoidea (Insecta: Hy- menoptera) of India and the adjacent countries. Oriental Insects 20: 1^30. Forster, A. 1856. Hymenopterologisclie Studien. II. Heft. Chalcidiae und Proctotrupii. Aachen. 152 pp. Gahan, A. B. and M. M. Fagan. 1923. The type species of the genera of Chalcidoidea or chalcid-flies. Bulletin of the United States National Museum 124: 1-173. Gahan, A. B. and S. A. Rowher. 1917. Lectotypes of the species of Hymenoptera (except Apoidea) de- scribed by Abbe Provancher. Canadian Entomolo- gist 49: 391^00. Gardenfors, U. 1986. Taxonomic and biological revi- sion of Palearctic Ephedrus Haliday (Hymenop- tera: Braconidae, Aphidiinae). Entomologica Scan- dinavica Supplement 27: 1-95. Gibson, G. A. P. 1997. Chapter 2. Morphology and Terminology. Pages 16-44 in Gibson, G. A. P., J. T. Huber, and J. B. Woolley (eds). Annotated Keys to the Genera of Nenrctic Chalcuioidea (Hymenop- tera). National Research Council Canada, Re- search Press, Ottawa. 794 pp. Girault, A. A. 1914(219]. Descriptions of new chalcid- flies. Proceedings of the Entomological Society of Washington 16: 109-119. Girault, A. A. 1917[330]. Descnptumes Hymcnoptero- rum Chalcidoidicarum variorum cum ohservationi- mus V. Private publication, Glendale. 16 pp. Goulet, H. and W. R. M. Mason. 1993. Chapter 4. Use of keys. Pages 60-64 in Goulet, H. and J. T. Hu- ber (eds). Hymenoptera of the world: An identifica- tion guide to famUies. Research Branch, Agricul- ture Canada Publication 1894/ E, Ottawa. 668 pp. Graham, M. W. R. de V. 1969. The Pteromalidae of North-Western Europe (Hymenoptera: Chalci- doidea). Bulletin of the British Museum (Natural History). Entomology 16: 1-908. Graham, M. W. R. de V. 1990. The identity of certain problematic Dahlbom genera of Chalcidoidea (Hym.), some represented by original material in Zoologiska Institutionen, Lund. Entomologist's Monthly Magazine 126(1516-1519): 197-200. Griswold, G. H. 1929. On the bionomics of a primary parasite and of two hyperparasites of the gera- nium aphid. Annals of the Entomological Society of America 22: 438-457. Hagen, K. S. and R. van den Bosch. 1968. Impact of pathogens, parasites, and predators on aphids. Annual Review of Entomology 13: 325-384. Heikinlieimo, O. and M. Raatikainen. 1981. Grid refer- 256 Journal of Hymenoptera Research ences and names of localities in the recording of biological finds in Finland. Notulae Entomohgkae 61: 133-154. [In Finnish]. Kamijo, K. and H. Takada. 1973. Studies on aphid hy- perparasites of Japan, II. Aphid hyperparasites of the Pteromalidae occurring in Japan (Hymenop- tera). Insects Matsurnnrmm, n.s. 2: 39-76. Keller, L. J. and D. J. Sullivan, S. J. 1976. Oviposition behavior and host feeding of Asaphes lucens an aphid hyperparasitoid. Journal of tlw Niic York En- tomological Societi/ 84:206-211. Kurdjumov, N. V. 1913. Notes on Pteromalidae (Hy- menoptera, Chalcidoidea). Reinie Riisse d'Ento- tiwlogic 13: 1-24. Lai, K. B. 1934. Insect parasites of Psyllidae. Parasi- tology 26: 325-334. Le Ralec, A. 1995. Egg contents in relation to host- feeding in some parasitic Hymenoptera. Ento- niofhaga 40: 87-93. Levine, L. and D. J. Sullivan, S.J. 1983. Intraspecific tertiary parasitoidism in Asaphes lucens (Hyme- noptera: Pteromalidae), an aphid hyperparasi- toid. Canadian Entomologist 115; 1653-1658. Lindroth, C. H., H. Andersson, H. Bodvarsson, and S. H. Richter. 1973. Surtsey, Iceland. The devel- opment of a new fauna, 1963-1970. Terrestrial In- vertebrates. Entomologica Scandinavica, Suppl. 5: 1-280. Lundbeck, W. 1897(1896). Hymenoptera Groenlandi- ca. Videnskabelige Meddclelser fra Dansk naturhis- torisk Forening i Kjebenhavn 1896: 220-251. Mackauer, M. 1968. Aphidiidae. Hymenopterorum Catalogus Pars 3: 1-103. Mani, M. S. and G. G. Saraswat. 1974. Part III. Pages 85-107 in Mani, M. S., O. P. Dubey, B. K. Kaul, and G. G. Saraswat. Descriptions of some new and new records of some known Chalcidoidea (Hymenoptera) from India. Memoirs of the School of Entomology, St. John's College, no. 3. 377 pp. McMuUen, R. D. 1966. New records of chalcidoid par- asites and hyperparasites of Psylla p'yricola Fors- ter in British Columbia. Canadian Entomologist 98: 236-239. McMullen, R. D. 1971. Psylla pi/ricola Forster, pear psylla (Hemiptera: Psyllidae). Pages 33-38 in Bi- ological Control Programmes Against Insects and Weeds in Canada 1959-1968. Commonwealth In- stitute of Biological Control Technical Commu- nication 4. 266 pp. Nees ab Esenbeck, C. G. 1834. Hymenopterorum Ich- neumonibus affiniuni monographiae, genera Enro- peaea et species ilhistrantes. Vol. 2. ]. G. Cottae, Stuttgart and Tubingen. 448 pp. Peck, O. 1951. Family Pteromalidae. Pages 534-568 in Muesebeck, C. F. W., K. V. Krombein, and H. K. Townes (eds). Hymenoptera of America North of Mexico — synoptic catalog. United States Depart ment of Agriculture Monograph No. 2. 1420 pp. Peck, O. 1963. A catalogue of the Nearctic Chalcidoi- dea. Canadian Entomologist, Supplement 30: 1- 1092. Peck, O., Z. Boucek, and G. Hoffer. 1964. Keys to the Chalcidoidea of Czechoslovakia (Insecta: Hyme- noptera). Memoirs of the Entomological Society of Canada 34. 120 pp. Philogene, B. J. R. and J. F. Chang. 1978. New records of parasitic chalcidoids of pear psylla (Homop- tera: Psyllidae) in Ontario, with observations on the current world status of its parasitoids and predators. Proceedings of the Entomological Society of Ontario 109: 53-60. Provancher, L. 1887. Additions et corrections a la Faune Hymenopterologicpie de la province de Quebec. C. Darveau, Quebec. 477 pp. Ratzeburg, J. T. C. 1844. Die Ichneumonen der Forstin- secten in forstlicher und entoinologischer Beziehung, vol. 1. Berlin. 224 pp. Ratzeburg, J. T. C. 1848. Die Ichneumonen der Forstin- secten hi forstlicher und entomologischer Beziehung, vol. 2. Berlin, vi 4- 238 pp. Reinhard, H. 1857. Beitrageb zur Geschichte und Syn- onymie der Pteromalinen. Berliner entomologische Zeitschrift 1; 70-80. Risbec, J. 1959. Pteromalidae de Madagascar (Hyme- noptera: Chalcidoidea). Memoires de I'lnstitut Scientifique de Madagascar, serie E II: 129-171. Rondani, C. 1848. Osservazioni sopra parecchie spe- cies di esapodi afidicidi e sui loro nemici. Nuovi Annali delle Scienze Naurali e Rendiconto dei Lai'ori dell Accademia delle Scienze dell'lstituto e delta So- cieta Agraria di Bologna 9: 5-37. Schulz, W. A. 1906. Spolia Hymenopterologica. Pader- bom. 355 pp. Sekhar, P. S. 1958. Studies on Asaphes fletcheri (Craw- ford), a hyperparasite of Aphidius testaceipes (Cresson) and Praon aguti (Smith), primary par- asites of aphids. Annals of the Entomological Soci- ety of America 51: 1-7. Sharma, A. K. and B. R. Subba Rao. 1958. Description of two new parasites of an aphid from North In- dia (Aphidiidae: Ichneumonoidea and Pterom- alidae: Chalcidoidea). Indian Journal of Entomolo- gy 20: 181-188. Sullivan, D. J. 1972. Comparative behavior and com- petition between two aphid hyperparasites: Al- loxysta victrix and Asaphes californicus (Hymenop- tera: Cynipidae; Pteromalidae). Environmental Entomology 1: 234-244. Walker, F. 1834. Monographia Chalciditum. The En- tomological Magazine 2: 148-179. Walker, F. 1846. 1846. List of the specimens of liymenop- terous insects in the collection of the British Museum. Part I. — Chalcidites. E. Newman, London. 100 pp. Waterston, J. 1929. On a chalcidoid bred from a flea larva. Parasitology 21: 103-106. Zetterstedt, J. W. 1838. Sectio secunda. Hymenoptera. Pages [4171-476 in Zetterstedt, J. W.', 1838-1840, Insecta lapponica descripta. Lipsiae. 1139 pp. J. HYM. RES. Vol. 7(2), 1998, pp. 257-267 Worker Versus Sexual, and Sex Ratio Investments in the Social Wasp Vespula vulgaris (L.) (Hymenoptera: Vespinae) in England Michael E. Archer University College of Ripon and York St. John, Lord Mayor's Walk, York Y031 TEX, England, U.K. Abstract. — An attempt has been made to quantify the relative investment in workers versus sexuals, and queens versus males in Vespula luilgaris (L.). In particular I have investigated MacNair's hypothesis that the queen invests equally in queen and male eggs, and that her in- vestment stops after she has laid her eggs. The dry weights and calorific values of the workers, autumn queens, and small-cell, and large-cell, reared males were determined. The lipid content of the autumn queens was 39.9%, and the crop-solid of males was 34.5%, of their dry weight. The seasonal change in worker dry weight was related to the varying work loads of the workers. A simulation model using the compartmental system approach was used to estimate the number of workers, queens and males produced by a successful colony. Investment in workers, either as dry weight or calorific equivalent units, represented nearly 50% of the total colony investment. This is considered a relatively high investment and can be related to the delay of sexual production. Males and queens are produced in about equal numbers but, in terms of dry weight or calorific equivalent units, there is a bias toward queen production. Workers were found to be selectively destroying male brood probably derived from queen eggs. MacNair's hypothesis could not be supported. In England the annual life-history of Vespula vulgaris (Linn.) starts with the emergence of the males and new queens from the mature colonies in the autumn. The sexuals remain in the colonies for a few days before leaving on their mating flights. After a short time the fertilised queens enter over-wintering sites while the males die. When the queens emerge in the spring they each search for a nest site, build a queen-nest and, by early June, rear the first workers. The workers take over the jobs of building the nest and looking after the brood from the queen which con- tinues to lay eggs. At first, the workers build small cells (small-cell colony) in which more workers, and later, males are reared. From the beginning of August, the workers build large cells (large-cell colo- ny) in which the queens and sometimes males are reared. The colony with its queen and workers usually dies by the end of October or early November (Archer 1981a, 1984). With the emergence of the new queens, brood in the small cells is neglected and not well fed (Montagner 1966). The ne- glected larvae and sealed brood are fre- quently pulled out of their cells by the workers and dropped some distance from the nest. This destruction of brood has been frequently recorded (Duncan 1939, Spradbery 1973, Archer 1981a, Greene 1991). The neglected brood also are eaten by dipterous larvae, e.g. Volucella sp. (Archer 1981a). The above life-history in which the re- productive females fly away from their natal colonies and found new colonies in- dependently is called 'Alate Dispersing' (Nonacs 1993). All alate dispersing colo- nies must make two investment decisions. The first is concerned with the proportion of resources devoted either to the produc- 258 Journal of Hymenoptera Research tion of the workers or to the sexuals. The second is concerned with the proportion of resources devoted to either male or queen production (Nonacs 1993). The first decision is a process of max- imising sexual production. Sexual produc- tion could be delayed by the production of more workers for colony maintenance with the trade-off consequence of ensur- ing a greater future sexual production. Brian (1965) produced a general model of colonial growth which showed that queen production increases with increased in- vestment in workers. Pamilo (1991) devel- oped models for perennial colonies, find- ing that greater investment in worker pro- duction was related to a smaller chance of queens founding new colonies and the old colonies surviving. With the annual colo- nies of V. vulgaris, the old colonies do not survive and the queens have a very small chance of founding new colonies (Archer 1984) so worker production relative to sexual production should be higher. The timing of the switch from worker to sexual production varies between the species of vespine social wasps (Archer 1980, 1981a). The second decision concerning propor- tional investments in males and queens fo- cuses on the queen-worker conflict due to asymmetrical genetical relationships (Trivers and Hare 1976, Benford 1978, MacNair 1978, Nonacs 1986, Boomsma 1989, Ratnieks and Reeve 1992). A wide range of variation in sex investment ratios has been found (Crozier and Pamilo 1996), often due to multiple mating by queens and worker reproduction in queenright colonies (Trivers and Hare 1976, Benford 1978). Usually sex ratio investment is mea- sured by determining the dry weight pro- duction of queens and males (Trivers and Hare 1976, Crozier and Pamilo 1996). In a vespine colony with a singly-mated queen and no worker reproduction, it is expected that queens will adjust the sex ratio in- vestment to give equal dry weights of queens and males. Workers, because of asymmetrical genetical relationships, will increase the investments in queens to three times that of males. If the queen has mated with more than one male and /or there is worker reproduction, the sex ratio investment of the queen is unchanged, but the workers will increase their investment in the males, although investment in queens will still be greater. MacNair (1978) argued that queen in- vestment stops after the eggs are laid: fur- ther investment then is carried out by the workers. Thus the queen should distribute her parental investment equally between queen and male eggs to produce a pri- mary sex ratio of 1:1. For workers to skew the investment towards queens, they should try to prevent the queen from lay- ing male eggs or destroy the brood de- rived from the male eggs of the queen. The queens have larger bodies in which to carry sufficient fat bodies to enable them to over-winter. The males do not over-winter but die soon after mating, or attempting to mate, for which purpose a smaller body seems adequate. Thus the departure from equal investment in queen and male eggs by the queen to a relatively greater queen investment does not neces- sarily imply the workers have succeeded in altering the sex ratio investment of the queen. To demonstrate that workers have succeeded in producing a relatively great- er investment in queens, it is necessary to show that male brood derived from queen eggs have been destroyed. Ideally, the determination of worker and sexual production should be carried out di- rectly by visual observations on colonies un- der natural conditions. This is difficult to achieve since colonies of V. vulgaris are sur- rounded by envelopes, are often under- ground, and the workers are aggressive when disturbed during investigations. Vi- sual observations on healthy colonies main- tained in an observation box have so far been restricted to one comb, the lowest comb of the nest (Potter 1964). However, indirect methods can be used Volume 7, Number 2, 1998 259 to estimate the production of workers and sexuals. Colonies can be collected throughout the year and counts made of the brood and adults present. At the end of the last larval stage the gut contents are evacuated to form the black meconium at the bottom of the cell. Counts of these me- conia can be used as an estimate of the number of adults reared. Since a meconi- um is only evidence that a larva has pu- pated, the result could be an over-estimate if incipient adults died during the pupal stage. In addition, the meconial remains do not reveal the sex or caste of the former occupants of the cells. Thus precise counts of worker and sexual production cannot be made from collected colonies, although counts of meconia and queen and male sealed brood can give a first approxima- tion (Archer 1993). To improve the accuracy of estimating worker and sexual production a simula- tion model of a successful colony, i.e. one that rears many queens, has been devel- oped (Archer 1981a, 1981b). The simula- tion model not only used the data from 198 collected colonies of V. vulgaris but also laboratory observations on the length of life of the brood stages and adult work- ers. The model incorporates meconial in- formation for adult production, and rates of cell building, brood stage addition and adult appearance estimated from the col- lected data. Brood neglect and mortality were incorporated into the model to come into effect when adult and brood stages became too numerous in comparison with the data from collected colonies. The mod- el achieves greater realism at the expense of complexity, with the use of 253 param- eters and variables. To handle such a com- plex model the compartmental system ap- proach (Odum 1971) was used. In this paper, I will try to determine the extent of any destruction of queen-derived male brood to test the proposal of Mac- Nair (1978), and derive estimates of the production of workers, males and queens to determine the relative importance of worker production, and the relative in- vestment in males and queens. METHODS Sources of, and treatment of workers. — Workers were bait-trapped during 1970 at Averhams Plantation, between Flaxton and Claxton, about 12 kilometres to the north-east of York, England. Averhams Plantation was an open site with a dense herb layer and recently planted conifers. The trapping station consisted of 16 sub- stations arranged in a square (4 X 4) with two traps at each sub-station. The traps were attached to canes about 70cm above the ground. Each trap consisted of a white polythene container (75mm deep, 75mm diameter) with a 10mm diameter hole in the lid, and contained a jam solution with added yeast. The fermenting jam solution was changed once every two weeks and the catch of workers collected once a week. Trapping was continuous from July until October. The workers were pre- served in 70% ethanol. Samples of 50 workers from each week- ly catch were dried to constant weight in an air oven at 60°C. During July and Oc- tober when the number of workers trapped were smaller all workers were dried. Each weekly sample of dried work- ers was weighed to O.lmg. The weekly dry weight was divided by the number of workers in the sample to give the mean worker dry weight. Workers from seven colonies were col- lected during late June and preserved by deep-freezing. Later the dry weights of these workers were obtained as previous- ly described except that workers were weighed individually so a standard devi- ation could be calculated. Correction for dry weight of workers due to 70% ethanol preservation. — To determine if ethanol-preserved workers lost dry weight, samples of 50 or 55 workers from three colonies were preserved in 70% eth- anol and by deep-freezing. After about six months of such preservation dry weights 260 Journal of Hymenoptera Research were obtained as previously described. Ethanol preserved workers showed a loss of dry weight compared with deep-frozen preserved workers from all three colonies. The percentage dry weight loss was 20.5%. Sources of, and treatment of queens. — Au- tumn queens with fully developed fat bodies were collected from four colonies. From three colonies 172 queens were pre- served by deep-freezing and from one col- ony 47 queens were preserved in 70% eth- anol. Dry weights were obtained as pre- viously described for workers except that queens were weighed individually. Dry weight loss from ethanol preservation was found to be 11.4% based on queens col- lected from one colony when 55 queens were preserved in 70% ethanol and 63 by deep-freezing. Sources of, and treatment of males. — Males with full crops were collected from a col- ony which was rearing males both in the small and large cells. All 102 males were preserved in 70% ethanol. Dry weights were obtained as previously described for workers except that males were weighed individually. The dry weights of the males showed a bimodal distribution indicating a weight difference between males reared in the small and large cells. A method giv- en by Lewis and Taylor (1967) was used to separate the 56 small-cell and 46 large- cell reared males. Males with full crops were collected from three colonies which had reared males only in the small cells. From the three colonies 146 males were preserved by deep-freezing. Dry weights were ob- tained as previously described for workers except that males were weighed individ- ually. Dry weight loss from ethanol preserva- tion was found to be 37.7% based on males collected from three colonies when 201 males were preserved in 70% ethanol and 146 by deep-freezing. Lipid determinations. — Samples of work- ers, autumn queens and males were pre- served by deep freezing and extracted in the Soxhlet apparatus using trichlorethyl- ene as a solvent. Extraction was continued until constant weight was obtained. Nine queens were extracted individually, 27 males in batches of threes, and 74 workers from three colonies in colony batches. Male crop content determinations. — When males left their natal colonies in the au- tumn their crops were found to contain a clear viscous fluid. From one colony 50 males were collected and preserved by deep freezing. Their crop fluid was col- lected by cutting the gaster away from the rest of the body and gently squeezing the gaster so that the crop fluid could be ab- sorbed by a known dry weight of filter pa- per. The filter paper with its absorbed crop fluids were dried in an air oven at 60°C to constant weight. All weightings were made to O.lmg. Some general bio- chemical tests were performed on the crop fluid. Ash content of workers, autumn queens and males. — A sample of 22 autumn queens was collected, preserved by deep-freezing, and dried by freeze-drying. Samples of 74 workers and 27 males were obtained from the lipid-extracted individuals. The sam- ples were heated in a furnace at 500°C for three hours and the residue ash weighed to O.lmg. Calorific determinations. — Samples of workers, autumn queens with fat bodies and males with full crops were preserved by deep freezing, dried by freeze-drying and their calorific values determined with the aid of a Phillipson Oxygen Microbomb Calorimeter (Phillipson 1964). Due to the high lipid content of the au- tumn queens oil was lost when the animal tissue was pelleted in preparation for bomb calorimetry. Attempts to bomb non- pelleted material as suggested by Wood- land et al. (1968) and Howell and Fisher (1977) failed because oil was lost from the sample and found at the bottom of the bomb. The lipids were extracted, as pre- Volume 7, Number 2, 1998 25 124 23 Q ^22 o 5 21 c n) (D 5 20 19 July August ! September Fig. 1. Mean worker dry weight (mg) versus date for Vt'sp'ula viilgnris. viously described, and the calorific deter- minations made on the remaining tissues. Difficulties sometimes were experienced when freeze-drying males in that the vis- cous contents of the crops would not freeze-dry but rather exploded out of the male into the freeze-drying equipment. Calorific determinations were made on crop-extracted males, as previously de- scribed, and on males with crop fluids where freeze-drying was satisfactorily completed. Simulation model of colonial determina- tion.— The development of a simulation model is given in Archer (1981a). The model was developed on the DEC system- 10 computer at the University of York, En- gland. Recently the model has been trans- ferred to a Personal Computer. RESULTS Seasonal variation of worker weights. — The seasonal change in the dry weights of workers from the bait traps is shown in Fig. 1. These dry weights have been cor- rected for dry weight loss during ethanol preservation. From a high mean dry weight during early July there is a de- crease to a low value during August, fol- lowed by an increase during the first half of September, after which there is a slight decrease during late September and Oc- tober. 261 Table 1. The mean dry weights of workers of Ves- puln viil^nrif from late July colonies. Dale No, workers Mean dry weight (m's) standard deviation 17 13 25.1 3.43 20 14 22.0 4.11 29 33 21.2 5.24 30 15 24.6 5.59 30 34 21.9 5.09 30 35 24.4 5.66 The mean dry weights of the workers from the late June colonies are given in Table 1. The mean dry weight from the seven colonies is 23.3mg which is slightly higher than the value for early July (Fig Lipid content of workers, autumn queens and males. — The lipid content of the work- ers is given in Table 2. The higher lipid content corresponds with the higher mean dry body weight found during September. The mean hpid content of a male was 7.4mg (range 4.1-9.1) representing 10.3% of the mean dry body weight. The mean lipid content of a queen was 65.0mg (s.d. 8.2) representing 39.9% of the mean dry body weight. Male crop fluid. — The mean dry weight of the crop fluid was 32.3mg (s.d. 9.98) which represented 60.5% of the wet weight of the crop fluid. The mean dry weight of the crop fluid represented 34.5% of the mean dry body weight. The dry weight of the crop fluid of each male was positively correlated with its dry body weight (n = 50, r = 0.57, p < 0.001). The crop fluid gave a positive result with two general carbohydrate tests: (a) Table 2. The lipid content of adult workers of Ves- puta vulgaris. .\o workers evtracted ".. lipid ot (nig) dry bodv weight 20 Jul. 26 Aug. 23 Sep. 20 25 23.0 16.7 43, S 5.7 3.1 10.1 262 Journal of Hymenoptera Research Table 3. The mean dry weight of autumn queens of Vespula vulgaris. Table 4. The mean dry weight of small-cell reared males of Vespuhi inilgnrif. No, queens Mean dry weight Standard deviation Dale No, males Mean drv weight (mg) Standard de\iation Date 12 Sep. 16 Sep. 23 Sep. 31 Oct. 50 56 50 46 81.1 87.3» 96.3 74.0 12 Sep. 20 Sep. 9 Oct. 31 Oct. 22 47 59 91 163.7 167.1* 167.6 155.3 11.18 11.75 10.43 8.43 14.3 14.7 23.0 13.9 ' Corrected for 70% ethanol preservation. alcoholic thymol and cone, hydrochloric acid and (b) alcoholic alpha-naphol and concentrated sulphuric acid. A positive re- action also was obtained with Benedict's reagent indicating the presence of reduc- ing sugar. Queen dry weight. — The mean dry weights of autumn queens with fat bodies are given in Table 3. The closeness of the deep-freeze and corrected ethanol-pre- served queens gives confidence in the use of the correction factor. The mean of the four samples is 163.4mg. Since these au- tumn queens consisted of 39.9% lipid con- tent the mean dry weight would be made up 65.2mg lipid and 98.2mg non-lipid substances. The dry weight of lipid of each queen was positively correlated with its dry body weight (n = 19, r = 0.79, p < 0.0001). Male dry weight. — The mean dry weights of males reared in the small cells are given in Table 4. The closeness of the deep- freeze and corrected ethanol-preserved males gives confidence in the use of the correction factor. The dry weight of the four samples is 84.7mg. Since the crop sol- id consisted of 34.5% of the body dry weight, the mean dry weight would be made up of 29.2mg crop solid and 55.5mg for the rest of the body. The mean dry weight of males reared in large cells was 130.7mg. This dry weight has been corrected for dry weight loss during ethanol preservation. Assuming 34.5% of this dry weight was crop solid, the mean dry weight would be made up of 45.1mg crop solid and 85.6mg for the rest of the body. Calorific determinations. — The ash con- tent of the adults was so low that it was unlikely to complicate the calorific deter- mination. The ash percentage of mean dry weight for autumn queens was 2.8%, for males 2.9% and for workers 2.5%. The mean calorific values of lipid-ex- tracted queens, males and workers are given in Table 5. Assuming the calorific equivalent of queen lipid to be 9.2 cal/mg (Sawicka-Kapusta 1975) and knowing the lipid percentage of mean dry body weight then the calorific value of the total dry body weight of the queen would be 6.627 cal/mg dry weight. Knowing the mean calorific values of crop-extracted males and males with crops (Table 5), it can be calculated that the crop solid has a calorific value of 4.232 cal/mg dry weight. Such a calorific value indicates that the crop solid was a protein or carbohydrate rather than a lipid sub- stance. Earlier results showed that the Table 5. Calorific values of autumn queens, males and workers of Vespula vulgaris. Sample Cal/mg Standard size dry weight deMJtion Queen (lipid extracted) 10 4.820 0.085 Male (crop extracted) 10 5.201 0.149 Male (with crop) 10 4.872 0.142 Worker— 17Jun. 10 5.092 0.122 Worker — 28 Jul. 10 5.273 0.147 Worker — 24 Sep. 10 5.220 0.109 Worker — combined 30 5.195 0.145 Volume 7, Number 2, 1998 263 Table 6. The number, biomass (dry weight) and calorific equivalent of queens, queen lipid, males, male crop content and workers produced in a simu- lation colony of Vespula vulgaris. Table 7. The dry weights of workers of Vespula vulgaris during the seasonal development of a colony derived from Fig. 1 and Spradbery (1972). CalorilR ^■L]ui\alenl \ umber Biomass (mj;) (cal) Workers 10,248 217,072.3 1,127,690.6 Males Small cells 763 64,626.1 314,858.4 Large cells 261 34,112.7 166,197.1 Total 1,024 98,738.8 481,055.5 Crop solids 34,064.9 144,162.6 Queens 969 158,334.6 1,049,283.4 Lipid 62,985.0 579,462.0 Total 12,241 474,145.7 2,658,029.5 crop solid is a carbohydrate, probably re- ducing sugar. Since the calorific values of workers var- ies little from samples collected during June, July and September a combined val- ue has been calculated (Table 5). The mean calorific values for workers and crop-extracted males are similar, but less than the mean calorific value of au- tumn queens because of the high lipid content of queens. The low mean calorific value of males is due to the high carbo- hydrate content of the crop. The lipid-ex- tracted queens have a lower mean calorific value than the workers and crop-extracted males, as these workers and males have some lipid in their bodies. Production estimates of a successful colo- ny.— The number, biomass and calorific equivalents of the workers, males and queens produced in the simulation model are given in Table 6. The biomass and cal- orific equivalents of the males and queens were calculated by multiplying the num- ber of males and queens produced by the model by the appropriate mean dry weight and calorific values. The crop solid of the males represents 30.0% of their cal- orific equivalent and the lipid of the queens 55.2% of their calorific equivalent. Since the dry weight of workers varies during the development of the colony Dale Dr\' weight (mg) 6 Jun. 11.75 (Spradbery) 16 Jun. 11.75 (Spradbery) 6 Jul. 22.4 11 Aug. 19.5 14 Sep. 24.9 5 Oct. 24.3 (Fig. 1), it is necessary to know the dry weight of the workers on the days that they emerged as adults. These emergence dry weights were assumed to be those of the collected workers, half the length of worker life before the date on which the workers were collected (Archer 1981a). The model calculated the number of adult workers produced each day which was multiplied by the appropriate mean dry weight derived from Fig. 1 and given in Table 7. Increases and decreases in mean dry weight were assumed to be linear. The total worker dry weight produced could be multiplied by worker mean calorific value to give the total calorific equivalent. The relative investment in workers, males and queens of a successful colony is given in Table 8. Just over five workers are needed to rear each sexual, but be- cause sexuals are heavier than workers a greater biomass or calorific equivalent of sexuals is produced than for workers. Workers represent 45.8% by dry weight and 42.4% by calorific equivalent of the to- tal production of the colony. Slightly more males than queens are Table 8. The relative investment in workers, males and queens in a simulation model of a successful col- ony of Vespula vulgaris. Biomass cainnlic Number (dry weight) equivalent Sexuals: Workers Queens: Workers Queens: Males 1:5.14 1:10.58 1:1.06 1:0.84 1:1.37 1:0.62 1:0.74 1:1.07 1:0.46 264 Journal of Hymenoptera Research reared, but, because queens are heavier and of higher calorific value, the sex ratio investment changes markedly in favour of the queens. Because the males have higher metabolic rates than queens, the energetic cost ratio (Boomsma 1989, Bourke and Franks 1995) can be used. The sex ratio investment then becomes one queen to 0.72 male, still indicating an investment in favour of queens. DISCUSSION Worker dry weight. — A similar seasonal change in the dry weight of workers of V. vulgaris from early July until October (Fig. 1) was found by Spradbery (1972) in En- gland and by Malham (1996) in New Zea- land for the equivalent season. Spradbery (1972) also found a very low worker dry weight during June (Table 7): these were queen-reared workers. Brian and Brian (1952) also found that queen-reared work- ers of Dolichovespula sylvestris (Scopoli) had low weights. Malham (1996) found that in areas where insecticide had been used to dras- tically reduce the number of workers, the dry weight of workers during March (equivalent to September in England, Fig. 1) was markedly higher than in untreated areas. This difference was less pronounced earlier in the season and had disappeared by the end of the season. Malham (1996) attributed the difference to food availabil- ity. In treated areas, relatively more food would be available per forager, so larvae would receive more food and produce workers of a heavier weight. The low worker dry weight during Au- gust (equivalent to February in New Zea- land) varied from 12-13mg (Malham 1996) to 17.4mg (Spradbery 1972) and 19.5mg in the present study. The high worker dry weight during September (equivalent to March in New Zealand) varied from 17- 20mg (Malham 1996) to 20.8mg (Sprad- bery 1972) and 24.9mg in the present study. Following Malham (1996) these variabilities in dry weights could be due to shortages in food supply as forager density increases. The variation in worker dry weight also can be related to variation of the work load on workers. The work load will de- pend on the number of larvae to feed (lar- va/worker ratio), the number of cells to build, and the amount of soil to be exca- vated to make the cavity for the nest. Time spent in excavation and building could re- duce the time available to feed the larvae. The low weights of queen-reared work- ers could be a consequence of very high larvae per queen ratio, up to 30 larvae per queen, and high cell building rates, in ex- cess of two cells per day (Archer, unpub- lished). The relatively rapid increase in worker dry weight by late June until early July could be a consequence of workers aiding the queen in brood rearing. At this time the larva /worker ratio rapidly decreases to about 3-4 and the cell building rate to about one cell per worker per day (Archer 1981a). The decline in worker dry weight from early July until August coincides with the development of the small-cell colony dur- ing which a large worker population is reared (Archer 1981a). Workers sampled during August also had the lowest lipid content (Table 2). Despite the exponential growth of the small-cell colony, the work load on workers continues to decrease with larva /worker ratio decreasing to about one, and cell building rate per day per worker approaching zero. However the amount of excavation greatly increases and over 90% of the outgoing workers from a colony may be carrying earth par- ticles (Archer, unpublished). The increase in dry weight from August until September coincides with the devel- opment of the large-cell colony when the future sexuals are reared. Workers sam- pled during September also had the high- est lipid content (Table 2). During this time the worker load remains low. Larva/ worker ratio remains at about one, and Volume 7, Number 2, 1998 265 large cell building rate per worker per day is very low at about 0.04 (Archer 1981a). Soil excavation continues but usually less than 20% of outgoing foragers are carry- ing earth particles (Archer, unpublished). Queen dry weight. — Spradbery (1973) and Harris and Beggs (1995) found that nearly 40% of the dry weight of autumn queens was lipid, which is similar to the value given earlier in this paper. The same authors found that the lipid was used as a food source of which about three-quar- ters was used during the over-wintering period. By dissecting queens of Vespa affin- is (Linn.) from southern Japan, Martin (1993) found that the contents of the fat bodies were used up during the over-win- tering period of four to five months. Harris and Beggs (1995) found the mean dry weight of autumn queens from New Zealand was 121.7mg (range 108.0-154.5), which is about 26% lower than the mean weight reported in the current study. They suggested the low weight of autumn queens was because the fat bodies of the queens had not reached their maximum level of lipid storage. The low weight of queens also could be a consequence of the very high colony densities: up to 33 colo- nies per hectare found in New Zealand, compared with up to about two colonies per hectare in England (Edwards 1980). At lower colony densities relatively more food resources might be available for queen rearing. Male dry weight. — By measuring the wing length of males of Vesya crabro Linn., Potter (1964) also found a bimodal distri- bution of male size. Measurements were carried out on callow adults found in their cells. He also recorded that males of some colonies of V. vulgaris showed a bimodal size distribution but gave no numerical details. Potter's observations support the interpretation of the bimodal size distri- bution of males presented in this paper. The carbohydrate food reserve found in the crops of the males seems to be a new observation. The function of this food re- serve would be to provide a readily avail- able source of energy needed by the males when flying around their mating circuits (Edwards 1980). Worker-sexual ratio investment. — Invest- ment in workers represents nearly 50% of the total biomass investment of a colony (Table 8). By reference to Pamilo (1991, Ta- ble 2) the investment in the workers can be considered to be relatively high. This large investment in workers can be related to the delay of sexual production until September in V. vulgaris, with the conse- quence of a larger output of queens (Brian 1983). Sexual production in Dolichoz>espnila sylvestris (Scopoli) occurs earlier, during July, so this species has a relatively small- er investment in workers, and also a smaller output of queens (Edwards 1980). Sex ratio investment. — The more-or-less equal production of queens and males in V. vulgaris (Table 8) seems rather surpris- ing since at the sealed brood stage, the number of males is usually twice the num- ber of queens (Archer 1981a, Greene 1991). However, Archer (1981b) found that 45% of large-cell male sealed brood was de- stroyed, and observed large-cell male sealed brood and mature larvae were car- ried away from the colonies by the work- ers. If these large-cell sealed brood are in- cluded in the calculation of the primary sex ratio, the ratio becomes one queen to 1.28 males. The simulation model allows for this destruction of large-cell larvae and sealed brood, as well as for the neglect and de- struction of small-cell larvae and sealed brood. The output of the model indicates that 18.4% small-cell larvae and sealed brood will be neglected and destroyed. Since this destruction occurs in the later part of colonial development, most of the destroyed small-cell brood will be males (Archer 1981a). Thus the primary sex ratio would be even more biased towards the males. The interpretation of the above obser- vations and calculations would indicate 266 Journal of Hymenoptera Research that the queen is not laying an equal num- ber of male and queen eggs as MacNair (1978) suggested. Since males are smaller than queens, the queen would seem to be laying relatively more male eggs so that the investment in adult queens and males eventually becomes equal. It is possible that the extra male eggs are derived from the workers and the destruc- tion of the male brood an example of worker policing against male production by other workers (Ratnieks 1988), which is predicted when the queen mates with more than two males. Multiple matings by queens of V. vulgaris are highly likely (Page 1986). However, Ross 1986 and Bourke 1988, failed to find evidence of Vespula workers laying eggs in queenright colonies. The workers would appear to be de- stroying male brood so as to bias invest- ment towards the queens. However the sex-ratio investnient does not reach three queens for every male (Table 8) probably because of multiple mating by the queen (Page 1986). The destruction of males by the workers does not take place until the male brood has reached the mature larval and sealed brood stages. Thus, the queen would seem to be able to disguise the sex of her off- spring during the egg and early larval stages. MacNair (1978) proposed that there would be an evolutionary race be- tween the queens and workers, with genes selected which favour the disguise of the sex of the brood, followed by genes which enable the disguise to be penetrated. The evolutionary race seems to have reached an equilibrium with detection occurring at the late larval and sealed brood stages af- ter the workers have made a considerable investment in rearing males. Brood destruction. — Brood neglect and destruction has been linked to the death or physiological breakdown of the queen, leading to the disintegration of the social life of the colony (Spradbery 1973). How- ever, the queen brood are not neglected but well fed (Montagner 1963), and selec- tive destruction of male brood occurs in the large cells (Archer 1981b). One consequence of the loss of queen influence is the appearance of a domi- nance struggle among the workers (Mon- tagner 1966), whose ovaries start devel- oping (Greene 1991). Workers with devel- oped ovaries lay eggs destined to become males, although due to the lateness of the season it is unlikely these males will be reared, or if reared, will successfully mate (Ross 1985). Probably the capacity of workers to rear males from worker eggs is an adaptive response to the premature death or physiological breakdown of the queen (Ross 1985). About 28% of the col- onies surviving until at least September (Archer, unpublished) are unsuccessful in producing many queens (Archer 1981b), but nevertheless rear males in the small cells. Many of these males could be de- rived from worker eggs. In conclusion the outcomes of the two investment decisions of successful colo- nies of V. vulgaris have been found as fol- lows. The first decision of the trade-off be- tween worker production for colony maintenance and sexual production is to delay sexual production in order to pro- duce more workers and hence to produce relatively more sexuals. In the second de- cision, there is a bias towards queen pro- duction at the expense of male production. This bias depends on worker action in the destruction of male brood probably de- rived from queen eggs. ACKNOWLEDGMENTS Robin Edwards made many helpful comments on this manuscript. LITERATURE CITED Archer, M. E. 1980. Population dynamics: 172-207, in Edwards, R. Social Wasp-,. The Rentokil Library, East Grinstead. Archer, M. E. 1981a, A simulation model tor the colonial development of Pamvcfpuhi vitl^nri^ (Linnaeus) and Dolichovct^puk sy'i'csfn'.s (Scopoli) (Hymenoptera: Vespidae). Mclimdcrm 36: 1-59. Volume 7, Number 2, 1998 267 Archer, M. E. 1981b. Successful and unsuccessful devel- opment of colonies of Vesj.nda vulgaris (Linn.) (Hy- menoptera; Vespidae). Ecohgicnl Eiilciniclogy 6: 1-10. Archer, M. E. 1984. Life and fertility tables for the wasp species Vespnda vulgaris and Dolichoi'espula syhvstris (Hymenoptera: Vespidae) in England. Entomologia Generally 9: 181-188. Archer, M. E. 1993. The life-history and colonial char- acteristics of the Hornet, Vespa crabro L. (Hym., Ves- pinae). Entomologist's monthly Magazine 129: 151- 163. Benford, F. A. 1978. Fisher's theory of the sex ratio ap- plied to the social Hymenoptera. journal of theoreti- cal Biology 72: 701-727. Boomsma, J. J. 1989. Sex-investment ratios in ants: has female bias been systematically overestimated? The American Naturalist 133: 517-532. Bourke, A. F. G. 1988. Worker reproduction in tlie high- er eusocial Hymenoptera. Quarterly Review of Biol- ogy 63: 291-312. Bourke, A. F. G. and Franks, N. R. 1995. Social ei'olution in ants. Princeton University Press, U.S.A. Brian, M. V. 1965. Social Insect Populations. Academic Press, London. Brian, M. V. 1983. Social bisects. Ecology/ and Beltaz'ioural Biology. Chapman & Hall, London. Brian, M. V. and Brian, A. D. 1952. The wasp, Vespula sylvestris Scopoli: feeding, foraging and colony de- velopment. Transactions of the Royal entomological So- cieh/ of London 103: 1-26. Crozier, R. H. and Pamilo, P. 1996. Evolution of social insect colonies. Sex allocation and kin Selection. Oxford University Press, Oxford. Duncan, C. D. 1939. A contribution to the biology of North American Vespine wasps. Stanford Unii'ersity Publications, Biological Science 8: 1-272. Edwards, R. 1980. Social wasps. Their biology and control. The Rentokil Libran', East Grinstead, England. Greene, A. 1991. Dolichovesputa and Vespula: 263-305, in Ross, K. G. and Matthews, R. W. (eds) The Social Biology of Wasps, Cornell Universit)' Press: 263-305. Harris, R. J. and Beggs, J. R. 1995. Variation in the qual- ity of Vespula vulgaris (L.) queens (Hymenoptera: Vespidae) and its significance in wasp population dynamics. Netv Zealand loumal of Zoology 22: 131- 142, Howell, ]. and Fisher, R. C. 1977. Food conversion effi- ciency of a parasitic wasp, Nemeritis connexens. Eco- logical Entomology 2: 143-151. Lewis, T. and Taylor, L. R. 1967. Introduction to experi- mental ecology/. Academic Press, London. MacNair, M. R. 1978. An ESS for the sex ratio in ani- mals, with particular reference to the social Hy- menoptera. Journal of theoretical Biology 70:449-459. Malham, J. 1996. How much does that wasp weigh? Wasp Times 24:2. Martin, S. J. 1993. Weight changes in adult hornets, Ves- pa affinis (Hymenoptera: Vespidae). Insectes sociaux 40: 363-368. Montagner, H. 1963. Relations entre les adultes et le cou- vain chez les guepes sociales du genre Vespa, au moyen d'un radio-isotOf>e. Insectes scoiaux 10: 153- 165. Montagner, H. 1966. Sur le determinisme du couvain abortif dans les nids de guepes du genre Vespa. Compte rendu hebdomadaire des seances de I'Acadeniie des sciences, Paris 263: 826-829. Nonacs, P. 1993. The effects of polygyny and colony life history on optimal sex investment: 110-131, in Kel- ler, L. (ed.) Queen Nundter and Socialihi in Insects. Oxford Scientific Publications, Oxford. Odum, E. P. 1971. Fundamentals of ecology. Saunders, London. Page, R. E. 1986. Sperm utilization in social insects. An- nual Review of Entomology 31: 297-320. Pamilo, P. 1991. Evolution of colony characteristics in social insects. 1. Sex allocation. Tlie American Natu- ralist 137: 83-107. Phillipson, J. 1964. A miniature bomb calorimeter for small biological samples. Oikos 15: 130-139. Potter, N. B. 1964. A study of the biology of the Common Wasp, Vespula vulgaris L., with spvcial reference to the foraging beliaviour. Unpublished Ph.D. thesis. Uni- versity of Bristol, England. Ratnieks, F. L. W. 1988. Reproductive harmony via mu- tual policing by workers in eusocial Hymenoptera. American Naturalist 132: 217-236. Ratueks, F. L. W. and Reeve, H. K. 1992. Conflict in single-queen Hymenoptera societies: the structure of conflict and processes that reduce conflict in ad- vanced eusodal species, journal of theoretical Biology 158:33-65. Ross, K. G. 1985. Aspects of worker reproduction in four social wasp species (Insecta: Hymenoptera: Vespi- dae). Journal of Zoology, London (A) 205: 411424. Ross, K. G. 1986. Kin selection and the problem of sperm utilization in social insects. Nature 323: 798-800. Sawicka-Kapusta, K. 1975. Fat extraction in the Soxhlet apparatiis: 288-292, in Klekowski, G. IBP Handbook No 24, Methods for Ecological Bio-energetics. Black- well, Oxford. Spradbery, J. P. 1972. A biometric study of seasonal vari- ation in worker wasps (Hymenoptera: Vespidae). Journal of Entomology (Series A) 47:61-69. Spradber)', ]. P. 1973. Wasps. An account of the biology and natural history of social and solitary waspis. Sidgwick and Jackson, London. Trivers, R. L. and Hare, H. 1976. Haplodiploidy and the evolution of the social insects. Science 191: 249-263. Woodland, D. J., HaU, B. K. and Calder, J. 1968. Gross bioenergetics of Blaltella germanica. Physiological Zo- ology 41:424-431. J. HYM. RES. Vol. 7(2), 1998, pp. 268-273 Response of Glyptapanteles militaris (Walsh) (Hymenoptera: Braconidae), a Larval Parasitoid of the Armyworm, Mythimna unipuncta (Haworth) (Lepidoptera: Noctuidae), to Different Temperatures L. Oliveira, R. Melo and J. Tavares Universidade dos Azores, Departamento de Biologia, P-9502 Ponta Delgada Codex, Azores, Portugal Abstract. — The effect of four different temperatures (15, 20 25 and 30°C), on biological param- eters of the Azorean population of Gh/ptapantelcs militaris (Walsh) was studied, using Mythimna unipuncta (Haworth) as the host. Thirteen biological parameters of the host-parasitoid interaction were analysed: percentage of hosts that died without producing parasitoids; percentage of larvae parasitized from which parasitoids emerged; percentage of hosts surviving to pupate after the parasitoid's sting; egg-larval development time; pupal period; total developmental time; adult longevity; total number of larvae per host; number of larval parasitoids that fail to emerge from each host; mean number of parasitoids that emerged from host larva but failed to spin a cocoon; mean number of cocoons per host; parasitoid sex-ratio; and finally emergence rate of adult par- asitoid progeny. The percentage of hosts that died without producing parasitoids increased with increasing temperature. Developmental times significantly decreased with increasing temperature. The mean number of cocoons per host; mean number of parasitoids that emerged from each host larva but failed to spin a cocoon; and total number of larvae per host were higher when the temperature was lower. Parasitoid sex ratio and emergence rate of adult progeny were not affected by the temperatures tested. Mythimna unipuncta (Havk^orth) is the most important pest in Azorean pastures. Serious population explosions that require the use of pesticides often occur (Tavares 1992). Glyptapanteles militaris (Walsh) is a larval parasitoid of the armyworm in all islands of the Archipelago (Oliveira 1996). It is desirable to increase the natural pop- ulation of G. militaris by field releases of wasps produced in the laboratory, during the first generation of M. unipuncta. One of the most important abiotic fac- tors that affect insects is temperature. In parasitoids this can influence develop- ment, fecundity, mortality, sex ratio, col- oration and other characteristics in vari- ous species (Kaya and Tanada 1969; Yu and Luck 1988; Klein 1988; Lysyk and Nealis 1988; Spivac et al. 1992). Temperature increases, within a favour- able range, will speed up insect metabo- lism and consequently increase the rate of development. Each species and each stage in the life history may develop at its own rate (Sedlacek et al 1990, Spivac et al. 1992, Gullan and Cranston 1994). A previous study of the effect of two different temperatures on the biological parameters of G. militaris was performed by Oliveira (1991, 1992). In the present study, we analyse the effect of two ex- treme temperatures (15 and 30°C) and two intermediate temperatures (20 and 25°C), on some biological parameters of the Azo- rean population of G. militaris using M. unipuncta as host. MATERIAL AND METHODS Glyptapanteles militaris used in this ex- periment emerged from naturally parasit- Volume 7, Number 2, 1998 269 ized M. unipuncta larvae, collected in pas- tures of Sao Miguel island. Groups of one hundred cocoons were placed in 400 ml glass vials until adult females were re- moved for experiments. Adults were sup- plied with honey solution (10%). We used as hosts M. unipuncta larvae from labora- tory cultures, established from eggs laid by field-collected females. On the third day after adult parasitoid emergence, one isolated female wasp was allowed to parasitize one isolated third in- star larva of M. unipuncta. After the first sting, the host was removed from the par- asitoid and individually kept, until emer- gence of the parasitoids, in a plastic con- tainer (4.5 X 3cm). Each host larva was supplied with a small piece (Icm^) of ar- tificial diet every two days as described by Poitout and Bues (1970) and modified by Oliveira (1991). After parasitization, each group of fifty parasitized host larvae were kept at a different temperature (15±0.5°C, 20±0.5°C, 25±0.5°C and 30±0.5°C), under 75±0.5% R.H. and 16:8 [L:D] photoperiod. After larval parasitoid emergence and construction of the cocoons, each group was maintained in a plastic container (4.5 X 3cm) with a hole covered by nylon tis- sue. Emerged adults were kept in the same conditions and were supplied with honey solution (10%) until their death. Thirteen biological parameters of the host-parasitoid interaction were analysed: percentage of larvae parasitized from which parasitoids emerged; percentage of dead hosts; percentage of hosts surviving to pupate after the parasitoid's sting; egg- larval development time; pupal period; to- tal developmental time; adult longevity; total number of larvae per host; number of larval parasitoids that fail to emerge from each host; mean number of parasit- oids that emerged from host larva but failed to spin a cocoon; mean number of cocoons per host; parasitoid sex-ratio (per- centage of females); and finally emergence rate of adult progeny. The first three parameters were ana- 100 I parasitized hosts B dead hosts []] pupated hosts 2U 2."; Tempei^ture (°C) Fig. 1. Percentages of pupated, dead and parasitized larvae of Mf/thiinna tiinpuncta, at four different tem- peratures (±standard error). A test for multiple com- parison of proportions was used. Each column that is followed by a ciifferent letter is significantly differ- ent (p < 0.05). lysed by a test for multiple comparisons of proportions (Zar 1996), and the others parameters were analysed by non-para- metric "Kruskal-Wallis" and Multiple Comparison tests (p < 0.05) (Scherrer 1984) to compare the results obtained from different temperatures. To compare the effect of temperature on all studied pa- rameters, a discriminant factorial analysis (Thioulouse 1989), was performed. RESULTS AND DISCUSSION Parasitized larvae of M. unipuncta were able to develop normally at the four dif- ferent temperatures, with high percent- ages of hosts that were successfully para- sitized and producing parasite progeny (Fig. 1). However the percentage of para- sitized larvae from which parasitoids emerged differed significantly between 15 and 30°C (t = 1.983, p < 0.05), achieving maximum values at 15°C. The percentage of hosts that died before emergence of the parasitoids increased with temperature. It was least at 15°C, in- termediete and similar at 20 and 25°C, and most at 30°C fFig. 1). A significant differ- ence was observed for the percentage of dead hosts at 15 and 30°C (t = 2.280, p < 0.05). The percentage of surviving hosts after they had been stung by the parasit- 270 Journal of Hymenoptera Research Table 1. Mean {±standard deviation) days of egg-larval development time in Mytlniiiini innpuucla (Dl), pupal period (D2), and the total developmental time (Dl + D2) of Glyp^tapantelcs luilitaris, at four different temperatures. Temperature °C D2 X ± sd 15 40 37.50 ± 4.90a 40 18.25 ± 1.89a 40 55.75 ± 5.87a 20 34 19.00 ± 2.89b 33 07.27 ± 1.15b 33 26.18 ± 3.07b 25 34 13.56 ± 2.14c 31 05.84 ± 0.69c 31 19.36 ± 2.32c 30 31 11.90 ± 2.01c 20 05.45 ± 1.47c 20 17.15 ± 2.01c H value 118.495 96.684 108.677 P value <0.0001 O.OOOl <0.0001 Kruskal-Wallis H and P values and Multiple Comparisons test. Means in each column that are followed by a different letter are significantly different (p < 0.05). oid was very low (2^ %) and therefore they were not statistically analysed. The relation between temperature and rate of development of poikilotherms is an important aspect of ecological studies and basic to the development of pest manage- ment strategies (Lysyk and Nealis 1988). According to several authors, the period between parasitization and parasitoid emergence decreases with increasing tem- perature (Nealis and Fraser 1988; Gould and Elkinton 1990; Allen and Keller 1991; Tillman and Powell 1991). A similar result was obtained in this study with G. militaris and M. unipuncta. We divided the total de- velopment period in two parts: egg-larval development time, and pupal develop- ment (strictly, duration of the cocooned stages). We found similar results for both periods (Table 1), and significant differ- ences were observed between the different temperatures with only one exception (25 and 30°C). Temperature significantly affected the longevity of adult G. militaris obtained in this study except between 20 and 25°C. A similar result was obtained by Allen and Keller (1991) in a study of Cotesia urabne Austin and Allen reared from Uraba lugens Walker. The maximum longevity of G. militaris was obtained at 15°C (8.1 days) and the minimum was at 30°C (2.3 days). At the intermediate two temperatures we obtained 2.9 days. Comparing these val- ues with the 16 days of maximum longev- ity previously observed by Oliveira (1996), we conclude that our results are very low and they may be due to abiotic conditions, such as an insufficient level of ventilation in climatic chambers. The temperature established during the development time of G. militaris can affect the mean number of cocoons found per host, the mean number of parasitoids that emerged from the host larva but failed to spin a cocoon, and the total number of dis- cernable parasitoid larvae. The mean number of cocoons per host decreased with increasing temperature. A significant difference was observed between 30°C and 15 and 20°C (Table 2). Similar results were reported by Oliveira Filho and Foerster (1986) with "Apanteles" muesebecki Blanchard parasitising Pseiidaletia sequax Franclemont. The mean number of para- sitoid larvae that emerged from the host but then failed to spin a cocoon was rela- tively small in each case; the differences found (Table 2), are, though statistically significant, hard to explain in view of the non-linearity of the result and they may not really be informative. The number of parasitoid larvae that failed to leave the host larva was similar at the four temper- atures tested (Table 2). The total number of larvae per host decreased with increas- ing temperature (Table 2). A significant difference was found between the result obtained at 30°C and the other three tem- peratures. Volume 7, Number 2, 1998 271 Table 2. Mean (±standard deviation) number of Gli/ptnfuintcles ruilitiiris cocoons, the number of larvae that failed to spin a cocoon, and that failed to emerge from the host of Mythimna unipuncia, at four different temperatures. lfmptT.itiirf .\ N^ coccons / host Par lar\'ae not pupated X ± sd I'ar, larvae not emerged X ± sd total larvae V ± sd 15 40 55.40 ± 38.24ab 8.50 ± 13.06ab 10.00 ± 16.46a 73.95 ± 31.43a 20 34 50.41 ± 28.71b 2.50 ± 04.86a 8.91 ± 14.83a 61.82 ± 29.25a 25 34 35.12 ± 27.81bc 8.94 ± 13.55ab 12.24 ± 21.84a 56.29 ± 29.21a 30 31 19.42 ± 16.29c 5.55 ± 04.25b 9.19 ± 12.33a 34.16 ± 21.49b H value 27.36 12.84 3.20 30.19 P value <0.0001 0.0050 0.3616 <0.0001 Kruskal-Wallis H and P values and Multiple Comparisons test. Means in each column that are followed by a different letter are significantly different (p < 0.05). The sex ratios (percentage of females) were lower than 19%, in all cases (Table 3). No significant differences as a function of temperature were found. Similar results were obtained by Kolodny-Hirsch (1988) in his study of Cotesia melanoscela (Ratz- burg) and the host Lymantria dispar (L.). In all laboratory cultures of the Azorean G. militaris population we systematically ob- tained low sex ratios (Oliveira 1991, 1992, 1996), when compared with the observed sex ratio of natural populations, usually between 48 and 78% (Oliveira 1991, 1996). This is due to a high percentage of the fe- males that oviposit failing to produce fe- male progeny, indicating a high level of unmated females under laboratory condi- tions. This is an important aspect that will require to be overcome, if we want to do Table 3. Clyptafmntele^ militaris sex-ratio (percent- age of females) and adult emergence rates (±stan- dard deviation), at four different temperatures. Tempera* ture °C Se\ ratio "u Emergence N X ± sd N X ± sd 15 40 0.19 ± 0.29 40 70.90 ± 21.80 20 33 0.16 ± 0.23 34 63.00 ± 25.60 25 31 0.18 ± 0.25 34 70.30 ± 25.80 30 21 0.13 ± 0.22 31 49.80 ± 40.70 H value 0.595 5.012 P value 0.8976 0.1709 Kruskal-Wallis H and P values (p > 0.05). mass releases of G. militaris to control M. unipuncta. The adult emergence rates were not very high, always less than 71%, but no significant differences were observed be- tween temperatures (Table 3). However, an interesting result was observed at 30°C with a high number of adults dying before complete emergence from the cocoons. To compare the effect of the different temperatures on all parameters a discrim- inant factorial analysis was performed. This analysis demonstrated: 1) a super- position of the values obtained at 25 and 30°C; 2) a light separation at 20°C; 3) a complete separation of the results at 15°C (Fig. 2). The parameters that had most in- fluence on this separation were the egg- larval development time, the pupal peri- od, adult longevity, and the number of co- coons per host. Finally, these results confirm that G. mil- itaris is very well adapted to the climatic conditions of the Azores, since the para- sitoid develops better between 15 and 20"C than at 25 and 30°C, and the temper- atures found on the Azores pastures range between 13 and 23°C during the period of highest activity of the parasitoid. ACKNOWLEDGMENTS This research u'as performed on CIRN (Centro para a Investiga^ao dos Recursos Naturais) and fi- nanciallv supported bv the Universidade dos Azores, 272 Journal of Hymenoptera Research ; ^30^ ■■.. \ \25| \ 15° '•, Axis 1 (% of inertia=98.5) Fig 2. Discriminant factorial analysis performed with nine biological parameters of Glyptapanteles mil- itaris: egg-larval development time, pupal period, adult longevity, mean number of cocoons per host, mean number of parasitoids that emerged from host larva but failed to spin a cocoon, mean number of larval parasitoids that failed to emerge from each host, total number of larvae per host, parasitoid sex ratio, and emergence rate of adult progeny. by the Secretaria Regional da Agricultura e Pescas, and by the Program STRIDE (# ST RDB/C/AGR/ 194/92). LITERATURE CITED Allen, G. R. and M. A. Keller 1991. Urabn lumens (Lep- idoptera: Noctuidae) and its parasitoids (Hyme- noptera: Braconidae): temperature, host size, and development. Environmental Entomology, 20 (2), 458-469. Gould, ]. R. and J. S. Elkinton 1990. Temperature-de- pendent growth of Cotesia melanoscela (Hyme- noptera: Braconidae), a parasitoid of the Gypsy moth (Lepidoptera: Lymantriidae). Environmental Entomology. 19 (4), 859-865. Gullan, P. J. and P. S. Cranston 1994. The Insects: An Outline of Entonwlogi/. Chapman & Hall. London. 1-491. Kaya, H. and Y. Tanada 1969. Responses to high tem- perature of the parasite Apmnteles niilitaris and of its host, the armyworm, Pseudaletia unipimcta. Annals of the Entomological Society of America, 62 (6), 1303-1306. Klein, M. 1988. Colour morphs induced under con- trolled environmental conditions in adult Earias insulana (Lepidoptera: Noctuidae). Environmental Entomology , 17: 162-165. Kolodny-Hirsch, D. M. 1988. Influence of some en- vironmental factors on the laboratory production of Cotesia melanoscela (Braconidae: Hymenop- tera): a larval parasitoid of Lymantria dispar. En- vironmental Entomology , 17 (1), 127-131. Lysyk, T. J. and V. G. Nealis 1988. Temperature re- quirements for development of the jack pine budworm (Lepidoptera: Tortricidae) and two of its parasitoids (Hymenoptera). Journal of Econom- ic Entomology, 81 (4): 1045-1051 Nealis, V. G. and S. Fraser 1988. Rate of development, reproduction, and mass-rearing of Apanteles fii- miferanae Vier. (Hymenoptera: Braconidae) under controlled conditions. Cauadiau Entomologist, 120 (3), 197-204. Oliveira, L. 1991. Bioecologia de Apanteles mUitaris (Walsh, 1861) (Hymenoptera, Braconidae). Protias de acesso « categoria de Assistente de bwestiga(;do, Universidade dos Agores, 1-72. Oliveira, L. 1992. Influencia da temperatura no cicio biologico de Apanteles militaris (Walsh) (Hyme- noptera, Braconidae). Actas do V Congresso Ib- erico de Entomologia. Boletim da Sociedade porlii- guesa de Entomologia, 2 (3), 357-362. Oliveira, L. 1996. Apanteles mditaris (Walsh) (Hyme- noptera, Braconidae) parasitoide das larvas de Myihimna unipuncta (Haworth) (Lepidoptera, Noctuidae). Tese de Doutoramento, Universidade dos Azores, 1-196. Oliveira Filho, J. and L. A. Foerster 1986. Ciclo evo- lutivo e preferencia para oviposi^ao de Apanteles muesehecki Blanchard, 1947 (Hymenoptera: Bra- conidae), parasitoide de Pseudaletia sequax Fran- clemont, 1951 (Lepidoptera: Noctuidae). Anais da Sociedade de Entomologia do Brasd, 15 (2), 371-377. Poitout, S. and R. Bues 1970. elevage de plusieurs es- peces de Lepidopteres Noctuidae sur milieu ar- tificial riche et sur milieu artificiel simplifie. Ann Zool Ecol Anim., 2: 79-91. Scherrer, B. 1984 Biostatistique. Gaetan Morin editeur. Quebec, 1-850. Sedlacek, J. D., K. V. Yeargan and P. H. Freytag 1990. Effect of temperature on the development of the blackfaced leafhopper (Homoptera: Cicadelli- dae). Environmental Entomology, 19: 209-214. Spivac, M., A. Zeltzer, G. Degrandi-Hoffman and J. H. Martin 1992. Influence of temperature on the rate of development and colour patterns of queen honey bees (Hymenoptera: Apidae). En- vironmental Entomology, 21 (2): 364-370. Tavares, J. 1992. A importancia econtimica da lagarta das pastagens Mythimna unipuncta (Haworth) (Lep., Noctuidae). Aforeana, 7 (3), 407^14. Tillman, P. G. and J. E. Powell 1991. Developmental time in relation to temperature for Microplitis cro- ceipes, M. demolitor, Cotesia kazak (Hymenoptera: Braconidae) and tlyposoter didymator (Hymenop- tera: Ichneumonidae), endoparasitoids of the to- Volume 7, Number 2, 1998 273 bacco budworm (Lepidoptera: Noctuidae). Etwi- ronmcntal Entoitwlcgy, 20 (1), 61-64. Thioulouse, J. 1989. Statistical analysis and graphical display of multivariate data on the Macintosh. Cominiter Apptlicntions in the Biosciences, 5, 287- 292. Yu, D. S. and R. F. Luck 1988. Temperature-depen- dent size and development of California red scale (Homoptera: Diaspididae) and its effect on host availability for the ectoparasitoid, Aphitis melinus DeBach (Hymenoptera: Aphelinidae). Environmental Entomologii, 17 (2), 154-161. Zar, J. H. 1996. Biostattstical analysis. Prentice-Hall In- ternational Editions, London, 1-662. ]. HYM. RES. Vol. 7(2), 1998, pp. 274-279 Response of Pithecellobium tortum Martius (Leguminosae) Seeds to the Attack of the Phytophagous Braconid AUorhogas dyspistus Marsh (Hymenoptera: Braconidae) Margarete Valverde de Macedo, Marina C. P. Pimentel, and RiCARDO Cardoso Vieira (MVM, MCPP) Laboratorio de Ecologia de Insetos, Departamento de Ecologia, CP 68020, IB, Universidade Federal do Rio de Janeiro, ILha do Fundao, Rio de Janeiro, Brasil, CEP 21941-590; (MVM) Departamento de Zoologia, CP 6109, Universidade Estadual de Campii\as, Campinas, Sao Paulo, Brasil, CEP 13083-970; (RCV) Departamento de Botanica, Universidade Federal do Rio de Janeiro, Ilha do Fundao, Rio de Janeiro, Brasil, CEP 21941-590 Abstract. — This work describes morphological changes on the seeds of Pithecellobium tortum caused by the braconid AUorhogas di/spistus, at Restinga (Costal scrub) of Barra de Marica, Rio de Janeiro State, Brazil. This species was shown to be a gall maker on P. Tortum seeds whose galls result from the proliferation of parenchymatous cells near the tegument but not of the seed coat cells. Its impact on the host plant consists of decreases in plant reproductive potential not only by directly reducing seed viability, but also by contributing to seed mortality via the adult emer- gence hole which allows invasion by pathogenic micro-organisms. The braconid's way of eating the seed, keeping itself in a chamber apart from the seed embryo, which remains alive and there- fore demanding nutrients, accords it the profile of a "manipulative parasite" in the sense of Weis & Abrahamson (1986). Signs of insect herbivory on plants vary greatly. Some are simple feeding marks left on the host plant which normally do not involve any apparent morphological response. Other signs, however, are very complex, resulting from a noticeable mor- phological and /or physiological response of the plant. This response may be defen- sive, pathological or one which benefits the herbivore (Price 1980, Weis & Abra- hamson 1986). Herbivores that are capable of manipulating the response of their host plant for their own benefit have been called "manipulative parasites" (Weis & Abrahamson 1986). Gall makers induce the development of localised growing structures resulting from the abnormal in- crease in number and /or size of plant cells (Darlington 1975). Normally, the galls are induced in undifferentiated tissues, which have their development manipulated (Weis et ah 1988). The gall phenotype is the result of two genotypes: the one of the gall maker, responsible for the stimulus, and the other of the plant, which produces the reaction (Abrahamson & Weis 1987). From an evolutionary perspective, gall morphology is the product of natural se- lection on the insect stimulating the de- velopment of a structure for protection and nutrition and on the plant resisting or trying to avoid the insect stimulus (Weis et al. 1988). Most of the known entomogenous galls are induced by Diptera (especially Ceci- domyiidae), Hymenoptera, Homoptera and Thysanoptera (Meyer 1987; Short- house and Rohfritsch 1992). Within the Hymenoptera the Cynipidae is the most important family, but there are also re- cords for Tenthredinidae, Eurytomidae, Eulophidae, Pteromalidae and Tanaostig- Volume 7, Number 2, 1998 275 matidae (Fernandes 1987). Guimaraes (1957) reared Allorhogas muesebecki Gui- maraes from Aneniopaegnin miraiidutn Alph. DC. (Bignoniaceae) galls. Because this braconid was the only species to emerge the author concluded that this was the gall inducer. Since the first record of a phytophagous braconid (Macedo & Mon- teiro 1989) and its specific description by Marsh (1991) as Allorhogas dyspisttis, Infan- te et nl. (1995) redescribed Monitoriella elon- gata Hedqvist and recorded it as a new case of phytophagy within the Braconidae. This species, like A. dyspistiis, belongs to the Doryctinae, a group which, according to Wharton (1993), should be searched for more cases of phytophagy. Ramirez & Marsh (1996) described two new Psenobol- us species (Braconidae: Doryctinae) which appear to develop as inquilines on plant tissue in fig flowers after their pollination by Agaonidae wasps. More recently, Aus- tin & Dangerfield (1998) recorded the bi- ology of Mesostoa kerri Austin and Whar- ton, a member of the endemic Australian subfamily Mesostoinae, as a new case of galling Braconidae. Pithecellobium tortum Martius (Legumi- nosae) seeds are enclosed in fruits contain- ing about 30 seeds side by side. They are attacked by the braconid Allorhogas dyspis- tiis Marsh, which oviposits directly into immature seeds, when abundant endo- sperm and a small embryo are still pres- ent. After oviposition by the braconid the seed divides internally and externally, re- sulting in an intact region, joined to the funicle, where the seed embryo is usually found (Macedo & Monteiro 1989). In many cases this region of the seed contin- ues growing even after the adult insect has emerged. More than one A. dyspistus can be found in a single seed. In these cases, more than one division occurs and still a single intact region within the seed embryo is found. The main purpose of this study is to describe and discuss the mor- phological seed changes caused by Allor- hogas dyspistus in Pithecellobium tortum and to evaluate this impact of the insect on the host plant. MATERIALS AND METHODS All seeds were collected at Restinga (Coastal scrub) de Barra de Marica (22°57'S and 12° 52'W), Marica county, Rio de Janeiro state, Brazil. To investigate seed tissue morphology, intact and attacked seeds were collected during June 1994, and fixed in 70% alco- hol. Seeds were then laid in 10% ethyle- nediamine for one week, as suggested by Carlquist (1982) to soften them. After washing in distilled water three times for a 2 hours period each time, the seeds were dehydrated in ethanolic series and then embedded in paraffin wax following Jo- hansen (1940). Longitudinal serial 15 |xm sections were obtained with a rotary mi- crotome and stained with basic blue astra- fucsin (Roeser 1962). In 1994, a further sample of 2990 seeds from 150 fruits obtained from five P. tor- tum individuals were collected and dis- sected in order to evaluate the rate of A. dyspistus attack and to check if the at- tacked seeds died or continued develop- ment until complete maturation. In 1995, 327 fruits from 12 plant individuals were also collected and dissected to evaluated A. dyspistus seed attack rate. At the end of the 1995 fruiting season in June, mature fruits of P. tortum were col- lected from their parent plant and from the ground. Attacked and non-attacked seeds were then obtained to perform the tetrazolium viability test (Delouche et al. 1962). This test was performed immediate- ly after the collection of seeds from the plant and from the ground, and also after three and eight months of laboratory stor- age at room temperature for seeds collect- ed from the ground. All tested seeds were cut lengthwise and one half of each seed was completely immersed in 0.5% chlo- ride of 2,3,5-triphenyl tetrazolium solution and the other half boiled before being sub- mitted to this tetrazolium test of viability. 276 Journal of Hymenoptera Research Fig. 1. Longitudinal sections of PitheceUohium tortiini seeds, a) Non-attacked seed with its embryo (em); b) seed attacked by two AUcrlwgaf lii/fpistus individuals, showing two attacked regions (ar); a larva (la) can be observed in one of them. The embryo (em) is present at the center in the preserved region (pr); c) an at- tacked region showing the seed coat palisade tissue (pt) which does not cover all of the region involving the larva. Note the thicker parenchymatous layer (pi). This procedure was necessary in order to eliminate the possibility that the red col- oration of the embryo was due to reduced ions and not to hydrogenases produced during the respiration process of the living embryo. If the dead boiled embryo col- oured, the test would be invalid. All beak- ers were kept in darkness at room tem- perature and the result checked after 12 hours. RESULTS Comparison between non-attacked (Fig. la) and attacked (Fig. lb) seeds shows that the tissue associated with A. dyspistus lar- vae probably results from the proliferation of the fundamental parenchymatous cells near the inside tegument. The gall, the re- gion resulting from tissue proliferation, keeps the insect larva apart from the seed embryo through a clear division of the seed. The seed coat does not cover the en- tire proliferated region where the braconid larva is found (Fig. Ic). This can also be seen by the naked eye because the texture of this attacked region is clearly different from that where the embr^'o is found. Allorbogas di/spistus attacked 55.85% of the dissected seeds. In all, only 6.1%, out of 1670 attacked seeds appeared healthy after insect emergence or death. All other attacked seeds died mainly because of contamination, probably fungus, which probably entered the seed through the A. dyspistus or its parasitoid exit holes. All mature attacked and non-attacked seeds collected from the plant itself were viable according to the tetrazolium test, as well as all mature non-attacked seeds col- lected from the ground (Table 1). A small- er proportion of attacked seeds were via- ble in the three-months stored group and an even smaller proportion of the attacked seeds were viable in the eight-months stored group. It is worth noting that in all groups the embryos from the intact seeds exhibited a clearly darker pink coloration in comparison with the viable embryos from the attacked seeds. No boiled seeds were coloured thus validating the above results. Allorfwgas dyspistus seed attack reached more than 70% in five out of the 12 plants Volume 7, Number 2, 1998 277 Table 1. Percentage of viable seeds according to tetrazolium test of attacked and non-attacked seeds under four situations of collection and storage. The number of tested seeds is in parentheses. Collection / situation of the tested seeds Non-attacked Attacked seeds seeds From plant /immediately after collection 100(50) 100(22) Ground /immediately after col- lection 100(25) 100(18) Ground /after three months of storage 100 (20) 73 (15) Ground /after eight months of storage 100(6) 18(11) Studied and in two of these the rates were very near 100% (Table 2). DISCUSSION Clear cell proliferation, characteristic of the process of gall formation, indicates that A. dysTpistus induces galls in P. tortum immature seeds. This kind of gall is, ac- cording to Gagne (1994), a simple gall, which does not differ from the normal plant tissue except for cell proliferation. This is very different from most of the complex galls {sensu Gagne 1994) on plant vegetative parts {e.g. Redfern & Askew 1992). Monitoriella elongata (Infante et al. 1995) and Mesostoa kerri (Austin and Dan- gerfield 1998), are two Braconidae leaf gallers whose galls are apparently as com- plex as any Cecidomyiidae leaf gall and quite different from the simple galls of A. di/spistiis. As seeds are attacked when im- mature and still in the process of devel- opment, meristematic tissue is available, and this is where galls are primarily pro- duced (Mani 1964, Bronner 1977). Quicke (1997) points out that there is growing ev- idence that at least seed predation actually involves the stimulation of proliferation of some plant cells, that is, a kind of incipient gall formation or simple gall (sensK Gagne 1994). Other insect groups which develop in- side seeds, such as Bruchidae (Coleop- tera), do not promote any similar response Table 2. Numbers of seeds collected (N), attacked seeds and percentage of seed attack by AUorhogai dys- f)is/i(s on each of the 12 plant individuals sampled in 1995. All.ii.ked seeds "n attack .4, i/i/s;'i5M and try it. You may enter the name, city, country or institution of a member and the service will return full contact details in- cluding phone and FAX numbers and a current email address, if we have one. You will find a link to this service at the ISH home page at . Please send any corrections or updated information to the Secretary at the address below. Note also that since the service re- turns details only for those members that are paid up, it functions also as a check on whether or not one has remembered to pay dues for the current year! Jim Wool- ley, Secretary, International Society of Hy- menopterists. Department of Entomology, Texas A&M University, College Station, TX 77843-2475; Phone: (409) 845-9349; FAX: (409) 845-6305; E-mail: JimWoolley® tamu.edu. Call for Bids for the 5th Conference of the International Society of Hymenopterists The conferences of the International So- ciety of Hymenopterists are held approx- imately every four years. The 5th Confer- ence, following that in Canberra next Jan- uary, would therefore occur some time in 2002-2003. The Executive Committee is now calling for bids from potential orga- nizers. You should send your proposal to the Secretary, Dr. Jim Woolley by 1 December 1998 (see address above). One or more bids may be asked to make a short pre- sentation at the Canberra Conference, and a decision will be announced there. Your proposal should include informa- tion relevant to the following: organizing committee, proposed dates, venue, plan for the scientific program, estimated costs of accommodation, possible sponsorship, other costs. We look forward to hearing from you. J. HYM. RES. Vol. 7(2), 1998, p. 306 EDITOR'S NOTE Submission of Manuscripts With this issue, we return to the former procedure of submitting manuscripts di- rectly to the Editor. There has been a slight change in my address, so please use the following: Dr. E. Eric Grissell Systematic Entomology Laboratory, USDA % National Museum of Natural History Washington, DC 20560-0168 It would be helpful to me if authors would follow the "Instructions for Au- thors" on the inside of the back cover. This is especially true when preparing il- lustrations. Individual figures should be mounted on illustration (poster) board or at least a backing that is somewhat rigid. This helps in handling the illustrations and in marking and sending them to the printer. Also, the corresponding author's name and address should appear on the back along with a shortened title and the figure numbers. This helps the press keep materials together. For some submissions from overseas, authors may not want to use heavy board because of postage costs. Please at least consider the use of stiff pa- per (e.g., file folder stock) as backing. If you do not send mounted illustrations, please at least provide me with typed la- bels (gummed or not) to put on the back of plates and I will mount them for you. INSTRUCTIONS FOR AUTHORS General Policy. The loiirnul of Hymencptcrn Rc^earcli invites papers of high scientific quality reporting comprehensive research on all aspects of Hymenoptera, including biology, behavior, ecology, svstomatics, taxonomy, genetics, and morphology. Taxonomic papers describing single species are unlikely to be accepted unless a strong case is evident, such as importance in economic entomology or with concurrent biology or ecology. Manuscript length generally should not exceed 50 typed pages; however, no upper limit on length has been set for papers of exceptional quality and importance, including taxonomic monographs at generic or higher level. All papers will be reviewed by at least two referees. The referees will be chosen by the appropriate subject editor. However, it would be helpful if authors would submit the names of two persons who are competent to review the manuscript. The language of publication is English. Summaries in other languages are acceptable. The deadline for receipt of manuscripts is 1 October (for the April issue) and 1 April (for the October issue). Format and Preparation. Three copies of each manuscript, including copies of illustrations, should be submitted on letter size or A4 paper, double spaced, with at least 25 mm margins on all sides. On the upper left of the title page give name, address and telephone and fax numbers of the author to whom all corre- spondence is to be sent. The paper should have a concise and informative title, followed by the names and addresses of all authors. The sequence of material should be: title, author(s), abstract, text, acknowledgments, literature dted, appendix, figure legends, figure copies (each numbered and identified), tables (each numbered and with heading). Each of the following should start a new page: (1) title page, (2) abstract, (3) text, (4) literature dted, (5) figure legends, (6) footiiotes. Following acceptance of the manuscript, the author should provide the editor witn one copy of the manu- script accompanied by a copy on diskette using DD, double sided computer diskettes — IBM compatible MS DOS 5.25 inch or IBM and Macintosh 3.5 inch diskettes. (Authors who do not have access to a computer should submit three copies of the manuscript.) The paper may be submitted in most PC and Mac word processor programs sucn as Microsoft Word, FullWrite Professional, WordPerfect, WriteNow, Nisus, Mac- Write, or MacWrite 11. If possible, all words that must be italicized should be done so, not underscored. Tables may be formatted in a spread sheet program such as MS Works or MS Excel. Text should be double spaced typing, with 25 mm left and right margins. Tables should be put in a separate file. Diskettes should be accompanied by the name of the software program used (e.g., WordPerfect, Microsoft Word). Authors should keep backup copies of all material sent to the Editor. The Society cannot be responsible for diskettes or text mislaid or destroyed in transit or during editing. Illustrations should be planned for reduction to the dimension of the printed page (14 X 20.5 cm, column width 6.7 cm) and allow room for legends at the top and bottom. Do not make plates larger than 14 X 18 in. (35.5 X 46 cm). Individual figures should be mounted on a suitable drawing board or similar heavy stock. Photographs should be trimmed, grouped together and abutted when mounted. Figure numbers should be on the plate, and it is strongly recommended that names be included after the numbers (e.g.. Fig. 2, tcxniiii^). Include title, author(s) and address(es), and illustration numbers on back of each plate. Original figures need not be sent until requested by the editor, usually after the manuscript has been accepted. Reference to figures/ tables in the text should be in the style "(Fig. 1)" "(Table 1)". Measurements should be in the metric system. All papers must conform to the International Code of Zoological Nomenclature. The first mention of a plant or animal should include the full scientific name including the authority. Genus names should not be abbre- viated at the beginning of a sentence. In taxonomic papers type specimens must be clearly designated, type depositories must be clearly indicated, and new taxa must be clearly differentiated from existing taxa by means of keys or differential diagnoses. Authors are required to deposit all type material in internationally recognized institutions (not private collections). Voucher specimens snould be designated for specimens used in behavioral or autecological studies, and they should be deposited similarly. Acceptance of taxonomic papers will not require use of cladistic methods; however, authors using them will be expected to specify the phylogenetic program used (if any), including discussion of program options used. A data matrix should be provided if the subject is complex. Cladograms must be hung with characters and these should include descriptors (not numbers alone) when feasible. The number of parsimonious clado- grams generated should be stated and reasons given for the one adopted. Lengths and consistency indices should be provided. Adequate discussions should be given for characters, plesiomorphic conditions, and distributions of characters among outgroups when problematical. References in the text should be (Smith 1999), without a comma, or Smith (1999). Two articles by a single author should be (Smith 1999a, 1999b) or Smith (1999a, 1999b). For multiple authors, use the word "and," not the symbol "&" (Smith and Jones 1999). For papers in press, u.se "in press," not the expected publication date. The Literature Cited section should include all papers referred to in the paper. Journal names should be spelled out completely and in italics. Charges. Publication charges are $10.00 per printed page. At least one author of the paper must be a member of the International Society of Hymenopterists. Reprints are charged to the author and must be ordered when returning the proofs; there are no free reprints. Author's corrections and changes in proof are also charged to the author. Color plates will be billed at full cost to the author. All manuscripts and correspondence should be sent to: Dr. E. Eric Grissell Systematic Entomology Laboratory, USDA "■f National Museum of Natural History Washington, DC 20560-0168 Phone: (202) 382-1781 Fax: (202) 786-9422 E-mail: egrissel@sel.barc.usda.goy CONTENTS (Continued from front cover) SHIMA, S. N., F. B. NOLL, R. ZUCCHI, and S. YAMANE. Morphological caste differences in the neotropical swarm-founding Polistine wasps IV: Pseudopolybia vespiceps, with preliminary considerations on the role of intermediate females in the social orga- nization of the Epiponini (Hymenoptera, Vespidae) 280 SIME, K. R., and D. B. WAHL. Taxonomy, mature larva, and observations on the biology of Gnamptopelta obsidianator (Brulle) (Hymenoptera: Ichneumonidae, Ichneumoni- nae) 157 WCISLO, W. T. Sexual dimorphism of wasp antennal structure in relation to parasitic and non-parasitic behavior (Hymenoptera: Sphecidae) 178 ZITANI, N. M., S. R. SHAW, and D. H, JANZEN. Systematics of Costa Rican Meteorus (Hymenoptera: Braconidae: Meteorinae) species lacking a dorsope 182 ANNOUNCEMENTS Membership Search Service Now Available 305 Call for Bids for the 5th Conference of the International Society of Hymenopterists . . 305 EDITOR'S NOTE 306 5OCIETV Journal of Hymenoptera Research JUN 1 0 1999 Volume 8, Number 1 --^.^^BrariES,^ April 1999 ISSN #1070-9428 CONTENTS AZEVEDO, C. O. A key to world species of Scolebythidae (Hymenoptera: Chrysidoidea), with description of a new species of Dominihyihus from Brazil 1 BELOKOBYLSKIJ, S. A. and D. L. ]. QUICKE. A new genus and two new species of bra- chypterous Lysiterminae (Braconidae) 120 COELHO, J. R. and K. WIEDMAN. Functional morphology of the hind tibial spurs of the cicada killer (Sfhecius speciosus Drury) (Hymenoptera; Sphecidae) 6 KIMSEY, L. S. and M. S. WASBAUER. New genera and species of Brachycistidine wasps from Southwestern North America (Hymenoptera: Tiphiidae: Brachycistidinae) . . 65 LaSALLE, J. A new species group and two new species of Euderomphale Girault (Hyme- noptera: Eulophidae) from North America 116 MARSH, P. M. and G. A. R. MELO. Biology and systematics of New World Heterospilus (Hymenoptera: Braconidae) attacking Pemphredoninae (Hymenoptera: Spheci- dae) 13 MARSH, P. M. and S. R. SHAW. Revision of North American Aleiodes Wesmael (Part 5): The melanopterus (Erichson) species-group (Hymenoptera: Braconidae: Ro- gadinae) 98 NEWMAN, T. M. and D. L. J. QUICKE. Ultrastructure of imaginal spermatozoa of sawflies (Hymenoptera: Symphyta) 35 RASNITSYN, A. P., W. J. PULAWSKI, and X. MARTINEZ-DELCLOS. Cretaceous digger wasps of the new genus Bestiola Pulawski and Rasnitsyn (Hymenoptera: Sphecidae: Angarosphecinae) 23 ROMANl, R., N. ISIDORO, and F. BIN. Further evidence of male antennal glands in Aphel- inidae: The case of Aphytis melinus DeBach (Hymenoptera: Aphelinidae) 109 (Continued on back cover) INTERNATIONAL SOCIETY OF HYMENOPTERISTS Organized 1982; Incorporated 1991 OFFICERS FOR 1999 Andrew D. Austin, President John LaSalle, Presideiit-Elect James B. Woolley, Secretary John T. Huber, Treasurer E. Eric Grissell, Editor Subject Editors Symphyta and Parasitica Aculeata Biology: Mark Shaw Biology: Sydney Cameron Systematics: Donald Quicke Systematics: Wojciech Pulawski All correspondence concerning Society business should be mailed to the appropriate officer at the following addresses: President, Department of Crop Protection, University of Adelaide, Glen Os- mond, Australia 5064; Secretary, Department of Entomology, Texas A&M University, College Station, Texas 77843; Treasurer, Eastern Cereal & Oilseed Research Centre, Agriculture Canada, K. W. Neatby Building, Ottawa, Ontario, Canada KIA 0C6; Editor, Systemafic Entomology Labo- ratory, USDA, % National Museum of Nahiral History, Washington, D.C. 20560-0168. Membership. Members shall be persons who have demonstrated interest in the science of ento- mology. Annual dues for members are US$40.00 per year (US$35.00 if paid before 1 February), payable to The International Society of Hymenopterists. Requests for membership should be sent to the Treasurer (address above). Information on membership and other details of the Society may be found on the World Wide Web at http;/ / IRIS.biosci.ohio-state.edu/ish. Journal. The Journal of Hymenoptera Research is published twice a year by the International Society of Hymenopterists, % Department of Entomology, Smithsonian Institution, Washington, D.C. 20560-0168, U.S.A. Members in good standing receive the Journal. Nonmember subscriptions are $60.00 (U.S. currency) per year. The Society does not exchange its publications for those of other societies. Please see inside back cover of this issue for information regarding preparation of manuscripts. Statement of Ownership Title of Publication: Journal of Hymenoptera Research. Frequency of Issue: Twice a year. Location of Office of Publication, Business Office of Publisher and Owner: International Society of Hymenopterists, %, Department of Entomology, NHB 168, Smithsonian Institution, Wash- ington, D.C. 20560, U.S.A. Editor: E. Eric Grissell, Systematic Entomology Laboratory, USDA, % National Museum of Nat- ural History, Washington, DC 20560-0168. Managing Editor and Known Bondholders or other Security Holders: none. This issue was mailed 3 May 1999 J. HYM. RES. Vol. 8(1), 1999, pp. 1-5 A Key to World Species of Scolebythidae (Hymenoptera: Chrysidoidea), with Description of a New Species of Dominihythiis from Brazil Celso Oliveira Azevedo Universidade Federal do Espi'rito Santo, Departamento de Biologia, Av. Marechal Campos 1468, Marui'pe, 29040-090 Vitoria, ES, Brazil Abstract. — Dominihythiis strictus sp. n. from Brazil, the first extant species of the genus, is described and illustrated. Additional specimens of the rarely collected Ch/stopsenelta longivcutris Kieffer 1911 are examined and the species variation analyzed. A world key to the fossil and extant species of Scolebythidae is given. Scolebythidae is a family of Aculeata with five monotypic genera. Evans (1963) proposed this family based on Clystopse- itelln Kieffer 1911, from Brazil, a genus transferred from the Bethylidae, and on Scolehi/thus Evans 1963, from Madagascar. Nagy (1975) added the third genus, Yca- ploca, from South Africa and Australia de- scribing the first male of Scolebythidae. Day (1977) described and illustrated the male genitalia of Clystopseiielln. Evans et al. (1979) reanalyzed Scolebythus with de- scription of the male and sting apparatus. All extant species are exclusive austral, but Prentice et al. (1996) described two fos- sil genera, Libanohythtis and Dominibytlnis, from Lebanese and Dominican amber. Biological data on the species suggest that the species of Scolebythidae are pos- sibly gregarious ectoparasitoids of wood- boring beetle larvae (Evans 1963, Nagy 1975, Day 1977, Evans et al. 1979), but Gauld (1995) pointed out that all published information is questionable. In this paper, Domiiiibi/thus strictus sp. n. from Brazil is described and illustrated. and new taxonomic data are provided for Ch/stopseuella longiventris Kieffer 1911. A world key to the fossil and extant species of this family is given. The examined material used in the pres- ent paper was provided by Canadian Na- tional Collection of Insects, Canada (CNCl, J. T. Huber); Universidade Federal do Pa- rana, Brazil (DZPR, K. Zanol); and Instituto Brasileiro de Geografia e Estatistica, Brasi- lia, Brazil (IBCE, B. Dias). Abbreviations for the main measure- ments used in this study are as follow: LH, length of head; WH, width of head; WF, width of frons; HE, height of eye; OOL, ocello-ocular line; WOT, width of the ocel- lar triangle, including the ocelli; DAO, di- ameter of anterior ocellus; VOL, vertex-oc- ular line; LEW, length of forewing. The nomenclature of the integument follows Eady (1968) for the term coria- ceous and Harris (1979) for the other tex- tures. Terminology generally follows Evans (1963), and the terminology of wing cells and veins follows Gauld and Bolton (1988). KEY TO FOSSIL AND EXTANT SPECIES OF WORLD SCOLEBYTHIDAE 1. Mesoscutum with notaulus absent or at least incomplete; prostemum large, its width at least 2.5 x length of propleuron; forewing with three closed cells, Rs vein shorter than stigma (Domitiibythiis) 2 2 Journal of Hymenoptera Research - Mesoscutum with complete notaulus; prosternum smaller than above, its width at most 2 X length of propleuron; forewing with five or six closed cells, Rs vein much longer than stigma 3 2. Frons with an arched prominence; eye forming the widest part of head; notaulus present anteriorly; occipital carina present dorsally; width of prosternum 2.5 x length of propleu- ron; Cu vein nebulous Dominibythus inopinatus Prentice and Poinar - Frons without prominence; gena forming the widest part of head; notaulus absent; occipital carina absent dorsally; width of prosternum 2.8 X length of propleuron; Cu vein as a short stub Dominibythus strictns Azevedo, new species 3. Pronotal disc enlarged, 1.57 x longer than mesoscutum; parapsidal furrows absent; fore- wing without metacarpus, with five closed cells, marginal cell opened; tibial spur formula 1,1,1 Lybanobythus milkii Prentice & Poinar - Pronotal disc shorter than mesoscutum, about 0.6-0.7 x longer than mesoscutum; parap- sidal furrows present; forewing with metacarpus, with six closed cells, marginal cell closed; tibial spur formula 1,2,2 4 4. Frons with a median prominence between antermal sockets; malar space virtually nonex- istent; forewing with submarginal cell longer than marginal, extending beyond the basal half of marginal cell Ycaploca evansi Nagy - Frons without median prominence between antennal sockets; malar space short; forewing with submarginal cell slightly shorter than marginal, not extending beyond the basal half of marginal cell 5 5. Occipital carina absent; malar space well over half as long as basal width of mandible; apex of marginal cell arched away from anterior margin of forewing; posterior area of metasomal sternite V with two groups of appressed and dense setae Clystopsenella longiventris Kieffer - Occipital carina present; malar space short, less than half as long as basal width of man- dible; apex of marginal cell on anterior margin of forewing; posterior area of metasomal sternite V without special groups of setae Scolebythtts madecassus Evans Dominibythus strictns Azevedo, apical teeth (Fig. 3), and with a shallow new species suture below upper tooth. Clypeus ex- (Figs. 1-7) tremely short, median lobe angulate, with- out median carina. Anterma short, nearly Description of female holotype: length reaching the pronotum; first four antennal of body 3.93 mm; LFW 2.25 mm. segments in a ratio of about 7:4:1:2, fla- Color: body castaneous, head and me- geHomere IX 1.2 X as long as thick, sen- sosoma slightly darker; vertex with two g^ji^g circular. Head slightly narrower be- lighter streaks starting on the crest and ex- j^^ Antennal sockets separated by less tending between the ocellar triangle and ^y^^^ t^gj^ o^vn diameters. Frons weakly eye; malar space, antenna, midtibia and coriaceous, with a very small number of tarsi slightly lighter than head; wings hy- gmall and shallow punctures; frontal ca- aline, veins castaneous. rina Iqw and short. Malar space broad, 1.0 Pubescence: very sparse and long over- x the basal width of mandible. LH 1.27 X all; setae short and dense on antenna, WH; WF 0.5 x WH; WF 0.85 X HE; ocellar some erect; setae noticeably longer on triangle very compact, OOL 1.45 x WOT hindtibia; nearly absent at dorsum of gas- and situated on an imaginary line between tral tergites, and concentrated at posterior eye tops, its frontal angle obtuse; posterior half in gastral sternites. ocelli distant from vertex crest by 2.94 x Head (Figs. 1-2): mandible wholly di- DAO. Eye subtriangular, with rounded rected downward slightly, with two sharp corners; gena forming widest part of head. Volume 8, Number 1, 1999 ^:^ M+Cu Figs. 1-7. Dominiln/thiis sfr;Vf»s. 1. Bodv, lateral. 2. Body, dorsal. 3. Mandible, frontal. 4. Mesosoma, ventral. 5. Forewing. 6. Foreleg, lateral. 7. Hindleg, lateral. Scale bars = 0.5 mm. Journal of Hymenoptera Research Vertex rounded, VOL 0.73 X HE; sides of head straight and subparallel and sHghtly convergent below. Occipital carina weak, present only ventrally. Palpal formula 6:4, segment I-III of labial palpi and segments I-IV of maxillary palpi flattened. Mesosoma (Figs. 1-2, 4): thorax weakly coriaceous. Pronotal disc very short, 0.4 X length of mesoscutum; anterior margin of pronotum evenly convex in dorsal view. Mesoscutum without notauli; parapsidal furrows well-impressed, absent on the an- terior half of mesoscutum, paralleled by an outer carina. Scutellum long, only slightly shorter than mesoscutum, prescu- tellar sulcus narrow posterad laterally, crossing entire anterior margin. Metano- tum very narrow medially. Propodeal disc 0.56 X as long as wide; anterior margin of propodeum with a transverse carina, pos- terior half of median propodeal line with a shallow groove; spiracle laterad, decliv- ity without carinae. Propleuron very large, produced strongly forw^ard. Prosternum very long, its width 2.8 X the length of propleuron, and 3 X longer than forefe- mur. Mesopleuron with a scrobal pit above midheight. Mesosterna separated by a wide longitudinal groove. Metaster- num narrow, separating the base of mid- coxa. Legs without spines. Forewing (Fig. 5) with only costal, basal and first discal cells, stigma wide, 0.63 X as long as wide; Rs vein very short, 0.37 X the length of stigma; Icu-a vein nearly vertical; Rs-(-M vein reaching Sc-I-R vein far from stigma; Cu vein as a very short stub. Hindwing with anal lobe distinct. Forefemur 3 X as long as thick (Fig. 6). Hindtibia longer than hindfemur. Hindcoxae closely set. Basitarsus of hindleg very long, longer than half length of hindtibia and longer than remaining tarsomeres together (Fig. 7). Tibial spur formula 1:1:1. Claws simple and arched, dilated basally. Metasoma (Figs. 1-2): coriaceous, wider than both mesosoma and head. Tergite VII with anterior margin deeply concavous. Gonostylus 0.28 x the length of sting. Sting and lancets without teeth; bulbous little dilated; sting 0.63 mm long. Material examined: 1 female holotype BRAZIL, Distrito Federal, Brasilia, Ronca- dor Ecological Station, shrub savanna; 12.ix.l979-21.x.l982; window trap; J. Dal- maceo col. (IBGE). Paratypes, 17 females BRAZIL, 16 females Distrito Federal, same data as holotype except places savanna grassland, shrub savanna, and swamp, and data 12.ix.l979-21.x.l982 (IBGE); 1 fe- male Rio de Janeiro, Rio de Janeiro, Cor- covado; i.l960; Seabra & Alvarenga col. (DZPR). Variation: body wholly dark casta- neous, clypeus distinctly lighter than head; specimen from Rio de Janeiro with clypeus projecting with a distinctly wider median lobe; WH 1.17-1.20 X LH; WF 0.45-0.55 X WH; WF 0.85-0.96 X HE; OOL 1.2-1.44 X WOT; posterior ocelli dis- tant from the vertex crest 2.78-3.52 X DAO; VOL 0.68-0.8 x HE; parapsidal fur- rows not paralleled by outer carina. Remarks: Dominihythus was first de- scribed from a fossil of Late Eocene to Late Oligocene Dominican amber (Prentice et al. 1990). Dominihythus strictiis sp. n. is the first living species of the genus. It may be identified as Domiiiibytlms by the 3 closed cells of the forewing. Dominibytlnis strictus differs from D. inopiuatiis by the absence of a frontal prominence, absence of occip- ital carina dorsally and notauli, and pres- ence of a tubular apical abscissa of the Cu vein, although extremely short, as a stub, while Cu vein in D. inopinatus is entire nebulous. The color pattern of the malar space and the lighter streaks on the vertex of D. strictus resembles that of Clystopse- nella longiventris. Etymology: The name refers to the short Rs vein of the forewing. Clystopsenella longiventris Kieffer This species is first recorded for Bahia and Minas Gerais. The specimens are about 8 mm long and the light color streaks on the vertex are very weak or nearly absent; Volume 8, Number 1, 1999 mandible and malar space distinctly lighter than head, legs lighter than mesosoma; WH 1.06-1.10 X LH; WF 0.61-0.65 x WH; WF 1.10-1.29 X HE; OOL 0.85-0.86 X WOT; posterior ocelli distant from the ver- tex crest 4.0-5.4 x DAO; VOL 0.54-0.61 x HE; pronotal disc about 0.51 X the meso- scutum length; notauli and parapsidal fur- rows complete or nearly so; propodeal disc about 0.5 X as long as wide. New material examined: 8 females BRAZIL, 1 female, Bahia, Encruzilhada; xi.l974; M. Alvarenga col. (CNCI); 4 fe- males Minas Gerais, Pedra Azul; xi.l974; M. Alvarenga col. (CNCI); 2 females Dis- trito Federal, Brasilia, Roncador Ecological Station, shrub savanna; 23.1.1982 and 7.iv.l983; window trap; J. Dalmaceo col. (IBGE); 1 female Sao Paulo, Mogi-gua^u, Campiminas Farm; 3.i.l970; J. M. & G. R. Campbell col. (CNCI). Distribution: Brazil (Bahia, Minas Ger- ais, Distrito Federal, Sao Paulo, Mato Grosso de Sul, Santa Catarina). Gauld (1995) commented that there are a few un- described species of Chjstopsenella in mu- seum collections from Neotropics and Australia, and one or, possibly two species in Costa Rica. ACKNOWLEDGMENTS I wish to thank E. R. Bortolini (UFES) and C. R. F. Brandao (MZSP) for the loan of the camara lucida. and to curators cited in the text for the loan of the material studied here. LITERATURE CITED Day, M. C. 1977. A new genus of Plumariidae from southern Africa, with notes on Scolebythidae (Hymenoptera, Chrysidoidea). Cimbcbasia, Series A, 4:171-177. F.ady, R. D. 1968. Some Illustrations of Microsculp- ture in the Hymenoptera. Proceedings of the Royal Entomological Society of London (A) 43(3-6): 66-72. Evans, H. E. 1963. A new family of wasps. Psyche 70: 7-16. Evans, H. E., C. Kugler, and W. L. Brown Jr. 1979. Rediscovery of Scolebythus madecassus, with a description of the male and of the female sting apparatus (Hymenoptera, Scolebvthidae). Psyche 86:45-51. Gauld, 1. D. 1995. Scolebythidae. p. 468-470. In: Han- son, P. E. and 1. D. Gauld (eds). Hymenoptera of Costa Rica. Oxford, Oxford University Press. 859 PP- Gauld, I. D. and B. Bolton. 1988. The Hymenoptera. Oxford, Oxford University Press. 332 pp. Harris, R. A. 1979. A glossary of surface sculpturing. Occasional Papers m Entomology, Department of Food and Agriculture, California 28:1-31. Nagy, C. G. 1975. A new genus of Scolebythidae (Hy- menoptera) from South Africa and Australia. Journal of t lie Entomological Society of South Africa 38(l):75-78. Prentice, M. A., G. O. Poinar Jr., and R. Milki. 1996. Fossil scolebythids (Hymenoptera, Scolebythi- dae) from Lebanese and Dominican amber. Pro- ceedings of the Entomological Societ\f of Washington 98(4):802-8n. ]. HYM. RES. Vol. 8(1), 1999, pp. 6-12 Functional Morphology of the Hind Tibial Spurs of the Cicada Killer {Sphecins speciosus Drury) (Hymenoptera: Sphecidae) Joseph R. Coelho' and Kimberly Wiedman Department of Biological Sciences, Western Illinois University, 1 University Circle, Macomb, Illinois 61455, USA, (E-mail: JR-Coelho@wiu.edu) Abstrnct. — The functional morphology of the hind tibial spurs in female cicada killers is exam- ined in relation to digging. The spurs are controlled by a passive mechanical system using the corium; when the tarsus is flexed, the spurs are extended. The spurs resist extension, but not flexion. Videography demonstrated that the hind legs are used in burrow excavation to clear soil from the path of the wasp and that the position of the legs makes use of the spurs' resistance to extension. Spur function is not consistent with use during prey carriage, which should cause the spurs to lie flat. Enlarged spurs may have evolved from smaller structures used in grooming, and may reduce the energetic and temporal costs of burrowing. The bodies of winged insects function in both aerial and terrestrial locomotion, which may present conflicting selection pressures on morphology. For example, flight performance may be improved by increasing the ratio of flight muscle mass to body mass (Marden 1987) at the ex- pense of legs or associated muscles, but terrestrial locomotion would be adversely affected. Many insects also dig or burrow, which increases selection pressure for en- hancement of the structures associated with the terrestrial mode. Although fos- sorial adaptation is relatively well under- stood in vertebrates (Hildebrand 1985), lit- tle of the function of specific adaptations of insects to fossorial life has been dem- onstrated. The Hymenoptera are generally excellent flyers, but many also dig bur- rows for nests. Therefore, they can be used to show how animals well adapted to flight handle the requirements of burrow- ing. Various morphological modifications for fossorial life, such as foretarsal rakes. ' To whom all correspondence stiould be ad- dressed. have been described in wasps (Bohart and Menke 1976), but their function has not been ascertained. An exception is provid- ed by Gorb (1996), who examined pretar- sal anatomy and function in a variety of insects. Fossorial species, such as the sphe- cid Benibix rostraia L., have a large ungui- tractor with well-developed microtrichia for the production of strong frictional forc- es with the substrate. Cicada killers {Sphecius speciosus Drury) are the largest (by mass) North American sphecid (Horn 1976), and they are superb flyers. Relative to other Hymenoptera, they have a high ratio of flight muscle to body mass, resulting in a high degree of maneuverability. Females weigh nearly one gram in body mass, which allows them to carry cicadas {Tibiceu spp.) heavi- er than themselves (Coelho 1997). Meta- bolic rate during hovering is high (Joos and Casey 1992), and body temperature during flight is elevated and nearly con- stant (J.R. Coelho, pers. obs.). Female cicada killers also dig extensive burrows. A cicada killer may move up to 1000 times her body mass in dry soil while excavating her burrow (J.R. Coelho and Volume 8, Number 1, 1999 A.J. Ross, pers. obs.). The opening of the burrow is elevated in comparison to the rest of the burrow, and a large tumulus accumulates outside the entrance. The en- trance tends to run at a 35-45° angle downward and then becomes more level (Evans 1966). The burrow branches into an average of 15.8 cells (Dambach and Good 1943), which are used to store paralyzed cicadas. The female lays one egg per cell. On average, cells may vary in diameter from 2 to 3 cm. Burrows vary in length from 30 cm to over 100 cm (Evans 1966). After provisioning the cell, the female ci- cada killer places a wall of dirt across the entrance. A variety of burrow architec- tures, as determined by excavation, are depicted by Riley (1892) and Evans (1966). After the main burrow is dug, the dirt from new cells is apparently used to plug old ones, although soil may be added to the tumulus as new cells are excavated. The digging method uses a variety of body parts and behaviors. The cicada kill- er uses her mouthparts to loosen compact soil to begin excavating. She uses her fore legs to rapidly rake the dirt under and be- hind her body. Finally, she uses her hind legs to push the soil behind her as she backs out of the burrow (Frost 1942; Dam- bach and Good 1943; Evans and West Eberhard 1970). The latter behavior results in the formation of a prominent trough through the tumulus leading away from the burrow entrance (Dambach and Good 1943; Evans 1966). Cicada killer females have a pair of en- larged spurs on the distal end of the hind tibia (see Evans [1966] for line drawings). Males have similar spurs which, by com- parison, are much reduced (Dambach and Good 1943). While the presence of such spurs in bees and wasps has been a useful taxonomic tool (Eickwort and Fischer 1963), their functional significance has rarely been addressed (except by Cane [1979]). Some suggest that the spurs of fe- male S. •^peciosiif are used to support the cicada during prey carriage (Howes 1919; Evans 1962), but our preliminary obser- vations suggest that they are used to move soil during burrow excavation. Hildebrand (1985) lists five require- ments of a fossorial animal to loosen and move resistant material: a digging tool, the capacity to produce and transmit large forces, a transport mechanism for soil, passive resistance to various loads, and the ability to sustain activity. The cicada killer hind leg functions as part of the transport mechanism for soil, and we hy- pothesize that the spurs are morphological adaptations for that mechanism. In this study we investigate the func- tional morphology of the hind tibial spurs in female cicada killers. Their mechanism of action is demonstrated, and their poten- tial function relative to digging and prey carriage is examined. MATERIALS AND METHODS Origin of specimens. — Live cicada killers were obtained from local nesting aggre- gations in McDonough County, Illinois; and the legs of freshly killed individuals were used in all experiments. Unless oth- erwise mentioned, all data are reported as mean ± SEM (N). Dissection. — We first investigated the mechanism for movement of the spurs. Under a dissecting microscope, a section of the exoskeleton was removed from the tibia near the tibiotarsal joint. Muscle tis- sue was examined for any connection to the spurs or to the corium, the unsclero- tized conjunctival membrane which forms the distal end of the tibia and from which the spurs arise. Mechanical linkage. — Five hind legs were obtained. When the tarsus of each was manually flexed, the spurs extended. Sim- ilarly, when the tarsus was extended, the spurs became flexed. The corium was sev- ered by making an incision perpendicular to a line between the spurs and tarsus us- ing a 30-ga syringe needle. The tarsus was flexed and extended again and the results were recorded. Journal of Hymenoptera Research sensor arm /femur/ spurs corium steel bar Fig. 1. Apparatus used to measure the force pro- duced by the spurs. The spurs were extended by low- ering the steel bar, to which the tibia was glued. The stationary force transducer measured the force exert- ed by the spurs during extension. To measure the force exerted during flexion, a similar set-up (not shown) was used, except that the sensor arm was placed above the spurs, and the steel bar was raised. Having found that tarsal flexion causes spur extension, we measured the force generated by the spurs when they were extended in this manner. The tibia of an intact hind leg was glued to a steel bar. The tarsus was manually extended at dif- ferent angles to the tibia, causing the spurs to extend. The spurs were oriented just be- low and perpendicular to the sensor arm of an optical force transducer (Narcosys- tems type A), which was calibrated with weights of known mass and connected to a physiograph (Narcosystems MK-111). The force exerted by the spurs was mea- sured at tibia-to-tarsus angles of 180, 135, and 90° and recorded by the physiograph. Force-displacement curves. — To determine the direction in which the spurs were ca- pable of resisting force, the hind leg of a female wasp was removed and the tibia was glued to a steel bar with the spurs facing upward. To measure the effect of extension, spurs were placed on top of and perpendicular to the sensor arm of a force transducer connected to a physio- graph (Fig. 1). An adjustable ringstand was calibrated so that fractions of a rota- tion of the knurled adjustment ring could be used to move the bar known vertical distances. The angle between the spurs and the tibia was gradually increased by moving the bar downward in 0.07-mm in- crements. Measurements were taken until the ends of the spurs slipped off the force transducer. At this point the force was maximal, but the spurs had not yet bro- ken. In this way, the resistant force exerted by the spurs when they were extended to various degrees was recorded by the phy- siograph. To measure the effects of flexion, the spurs were extended manually, then placed just below and perpendicular to the axis of the force transducer. The bar was moved upward in 0.07-mm incre- ments until the spurs laid flat against the tarsus. Hence, the force exerted by the spurs in resistance to varying degrees of flexion was recorded by the physiograph. Videograpin/. — Cicada killers were re- corded with a Sony VHS-C videotape re- corder while digging. Burrow entrances were plugged with a small amount of soil to induce digging in most cases, but spon- taneous digging was also recorded. The tapes were replayed using slow motion to examine and describe aspects of digging behavior in detail. RESULTS Dissections. — Of the 10 hind legs dis- sected, no muscle or tendon attachment to the spurs or corium was found. The tarsus was traversed by a single tendon arising from the pretarsal muscles in the femur and tibia, an arrangement essentially iden- tical to that described for honey bees (Snodgrass, 1956). Mechanical linkage. — When the tarsus was manipulated and the corium was in- tact, tarsal flexion caused spur extension, and tarsal extension caused spur flexion in every case. The spurs always moved in one plane and resisted lateral movements. After the corium was severed, the spurs Volume 8, Number 1, 1999 1 15 Displacement (mm) Fig. 2. Representative force-displacement curves for extension and flexion of the female cicada killer hind- tibial spurs. no longer moved when the tarsus was ma- nipulated. When the tarsus was held at increasing angles relative to the tibia, the force pro- duced by the extension of the spurs in- creased. When the tarsus was held at 180° to the tibia, no measurable force was de- tected. When the tarsus was held at 135°, an average force of 0.315 ± 0.041(7) mN was detected. Finally, when the angle was decreased to 90°, an average force of 0.615 ± 0.055(7) mN was produced. Force-displacement curves. — Fig. 2 dem- onstrates that as the spurs are displaced, the force resisting extension increases lin- early at first, then levels off near the max- imum, while the force resisting flexion re- mains constant at zero until the spurs lie flat and touch the tarsus. Extension pro- duced a maximum force of 21.68 ± 2.13(8) mN, which is eight times greater than the maximum force resulting from flexion, 2.64 ± 1.04(8) mN. Video records. — Review of videotaped digging behavior in slow motion demon- strated how the female cicada killer uses the hind legs to move soil. A wasp broke off bits of soil from the burrow wall using the mandibles and threw them posteriorly under the body using rapid motions of the fore legs. She then backed out of the bur- row, alternately thrusting her hind legs in arcs that began in a posterior direction and ended in a lateral direction, removing dirt from her path. Both front and rear pairs of legs operated simultaneously, with fore leg raking being repeated during the recovery stroke of each hind leg. The tibiotarsal angle was near 90° at the begin- ning of the thrusting motion when the leg was cocked, then the angle increased as the leg was extended. DISCUSSION In dissections of the cicada killer hind tibia, no muscles or tendons were found leading to the spurs or corium. The spurs were extended with increasing force as the tarsus was flexed at increasing angles, suggesting that the spurs are not under di- rect muscular control, but that their de- gree of extension depends instead upon the angle of the tibiotarsal joint. When the corium was severed, the spurs no longer moved when the tarsus was manipulated. The corium, a flexible, unsclerotized con- junctival membrane, connects the tarsus and spurs. When the tarsus is flexed, it pulls on the lateral margin of the corium, which, in turn, pulls on the medial bases of the spurs, causing them to extend. When the corium is severed, this mechan- ical linkage is broken. Hence, the spurs are controlled by a passive mechanical system using the corium, as suggested by Cane (1979). The spur extension system generates a very small force, less than 1 mN, which is sufficient to extend the spurs, but likely to be of little use otherwise. Hence, resistant forces determine what tasks can be accom- plished by the spurs. The spurs produced a substantial resistant force to being ex- tended outward, but not to being flexed inward (whereupon they simply fold flat against the tarsus). Only behaviors which flex the tibiotarsal joint, using the spurs' resistance to extension, are likely to effec- tively use the action of the spurs. Hymenopteran hind tibial spurs are commonly used in grooming the contra- lateral hind leg (Farish 1972, Cane 1979), and we observed this behavior in captive cicada killers. The tibiotarsal joint is flexed 10 Journal of Hymenoptera Research and brought under the body to the contra- lateral leg, where vertical motions up and down the second leg, held between the two spurs, occur (J.R. Coelho, pers. obs.). Known as L,-L, grooming, this behavior occurs in all 15 superfamilies of bees and wasps tested by Parish (1972). Our data suggest that the downstroke is the most useful in Lj-L, grooming, as it is the power stroke, where the spurs' resistance to ex- tension would be used. The upstroke is simply a return stroke. Although the spurs may be used to groom the ipsilateral wing (Parish 1972), we did not observe this be- havior in cicada killers. The utility of hind tibial spurs for Li-L, grooming suggests that the spurs evolved from smaller setae, but it does not ade- quately explain their exaggerated size in female Sphechis. It seems unlikely that the grooming requirements of the female ci- cada killer would be so much greater than that of a male, or of other Hymenoptera, that such elaborate structures would evolve. Howes (1919) suggests that the spurs of female S. speciosus are used to support the cicada during flight as "she squeezes [the spurs] against the cicada's sides and thus secures her burden during the overland journey to the burrow." When the spurs were removed from one wasp, the next ci- cada was carried in a more vertical posi- tion (Howes 1919). While Howes' (1919) single anecdotal observation weakly sup- ports his suggestion that the spurs have a role in prey carriage, we do not believe the spurs can function in the manner suggest- ed. The hind legs wrap around the cicada at high tibiotarsal angles (J.R. Coelho pers. obs.); therefore, the spurs would not be ex- tended. If squeezed against the sides of the cicada, the spurs will lie flat against the tarsus, as they have almost no capacity to resist flexion. In reviewing videotaped bouts of dig- ging, we noted that although cicada killers are capable of buzzing to produce a pow- erful vibration (Coelho 1998), and many Hymenoptera use vibration to loosen soil (Spangler 1973), cicada killers did not buzz while digging. The compact soil was chewed from the walls of the burrow us- ing the mandibles, and the fore legs were used in rapid motions to rake the loose particles below and behind the wasp. Hence, cicada killers are "rakers" in that they use the fore legs as rakes to move soil (Evans and West Eberhard 1970). Spines on the tarsus of the fore leg of S. speciosus form a pecten (rake), which is believed to augment the efficiency of raking (Evans 1966, Evans and West Eberhard 1970). Al- though raking moves the soil beyond the posterior margin of the animal, it does not necessarily clear it from the burrow or its entrance. As a unique type of "pusher, " the female cicada killer uses the hind legs, as opposed to the abdomen, to move soil out of the burrow and beyond (Evans and West Eberhard 1970). She backs out of the burrow, thrusting each hind leg first pos- teriorly then laterally, removing soil from her path. The tibiotarsal angle appears to be low (near 90°) at the beginning of the motion when the leg is cocked, which ex- tended the spurs. The path of the hind leg during the power stroke caused the spurs to be pushed against the soil in the direc- tion of extension. Hence, the high resistive force of the spurs is used to move additional dirt, thus increasing the efficacy of digging. Since little force is required for flexion, the spurs would not hinder the recovery stroke to complete the motion. This effect is similar to that of the "oars" of aquatic insects as they swim through the water (Gullan and Cranston 1994). When the spurs are fully extended, the effective surface area of the hind leg is substantially increased. Nu- merous setae on the tibia and tarsus prob- ably also contribute to the effective surface area, which presumably allows the female to move more soil per stroke than she could otherwise, increasing the energetic efficiency of digging. The resulting energy savings could be Volume 8, Number 1, 1999 11 used to dig more extensive burrows, to forage for additional cicadas, or for other fitness-enhancing activities. However, per- haps time is more limiting than energy. With an average lifespan of 13 days (Has- tings 1989), they have little time to waste on burrow construction. Observations suggest that they carry out much of their digging at night (Dambach and Good 1943), which would avoid temporal con- flicts with their strictly diurnal foraging bouts. Preliminary data from our labora- tory (J.R. Coelho and A.J. Ross, pers. obs.) suggest that digging rates of cicada killers are consistent with completing a burrow in one night, as suggested by Dambach and Good (1943). Hymenoptera tend to have long, thin, and even delicate legs, while a transport mechanism for soil should be large, broad, firm, and thick (Hildebrand 1985) as it is in mole crickets {Gryllotalpa), cicada nymphs (Magicicnda) and certain scarab beetles (Canthon) (Gullan and Cranston, 1994). Evans (1966) states that the cicada killers' use of hind legs to move soil is "unusual among digger wasps," and Pate (1936) mentions the spurs as taxonomical- ly important structural features of the ge- nus Spheciiis. The large hind tibial spurs of cicada killers appear to have been en- hanced to function in digging while re- taining their original function of groom- ing. Enlarged spurs appear to be an im- perfect solution to the problem of fossorial adaptation in comparison to the highly derived morphology of fully fossorial in- sects. Nonetheless, cicada killers clearly are accomplished burrowers in addition to being excellent flyers. The spurs are light- weight and collapsible, which may make them a suitable compromise between the demands of different locomotory habits. ACKNOWLEDGMENTS This study was supported by a Western Illinois University Research Council Grant to ]RC and by minigrants from the Honors Councils of the Illinois Region and Western Illinois University to KW. We thank Allan J. Ross for field and laboratory assistance, and Jacqueline A. Runestad and two anonymous re- viewers for comments on the manuscript. LITERATURE CITED Bohart, R.M. and A.S. Menke. 1976. Sphcciii waspjs of the world: a generic reinsion. University of Califor- nia Press, Berkeley, 695 pp. Cane, J.H. 1979. The hind tibiotarsal and tibial spur articulations in bees (Hymenoptera: Apoidea). Journal of the Kansas Entomological Society 52:123- 137. Coelho, J.R. 1997. Sexual size dimorphism and flight behavior in cicada killers (Sphecius speciosus). Oi- kos 79:371-375. Coelho, J.R, 1998. An acoustical and physiological analysis of buzzing in cicada killer wasps (Splie- cius speciosus). Journal of Comparative Physiology A 183:745-751. Dambach, C.A. and E. Good. 1943. Life history and habits of the cicada killer in Ohio. The Ohio jour- nal of Science 43:32-41. Eickwort, G.C. and R.L. Fischer. 1963. A morpholog- ical study of the tibial spurs of bees of the sub- family Halictinae (Hymenoptera: Halictidae) of eastern North America. Annals of the Entomologi- cal Society of America 56:348-354. Evans, H.E. 1962. The evolution of prey-carrying mechanisms in wasps. Evolution 16:468-483. Evans, H.E. 1966. The comparative ethology and evolu- tion of the sand wasps. Harvard University Press, Cambridge. 526 pp. Evans, H.E. and M.J. West Eberhard 1970. The wasps. University of Michigan Press, Ann Arbor. 265 PP- Parish, D.J. 1972. The evolutionary implications of qualitative variation in the grooming behaviour of the Hymenoptera (Insecta). Animal Behaviour 20:662-676. Gorb, S.V. 1996. Design of insect unguitractor appa- ratus. Journal of Morphology 230:219-220. Hastings, J. 1989. Protandry in western cicada killer wasps, (Sphecius grandis, Hymenoptera: Spheci- dae): an empirical study of emergence time and mating opportunity-. Behavioral Ecologx/ and Socio- hiology 25:255-260. Hildebrand, M. 1985. Digging of quadrupeds, pp. 89- 109 in Functional vertebrate morphology. Hilde- brand, M., D.M. Bramble, K.F. Liem, and D.B. Wake (eds.). Harvard University Press, Cam- bridge. 430 pp. Horn, D.J. 1976. Biology of insects. W.B. Saunders Company, Philadelphia. 439 pp. Howes, P.G. 1919. Insect behavior. Gorhani Press, Bos- ton. 176 pp. Joos, B. and T.M. Casey 1992. Flight energetics of the cicada killer wasp. American Zoologist 32:54A [ab- stract]. 12 Journal of Hymenoptera Research Neukirch, A. 1982. Dependence of the life span of the cidae: Gorytini). Bulletin of tlic Bwokhin Eiitoiiio- honevbee {Apif iiwUifica) upon flight performance logical Societi/ 31:198-200. and energy consumption. Journal of Comparative Riley, C.V. 1892. The larger digger wasp. Insect life 4: Physiology B 146:35-10. 248-252. Pate, V.S.L. 1936. Studies in the nyssonine wasps. II. Snodgrass, R.E. 1956. Anatomy of the Iwney bee. Com- The subgenera of Spliecius (Hymenoptera: Sphe- stock Publishing Associates, Ithaca. 334 pp. ]. HYM. RES. Vol. 8(1), 1999, pp. 13-22 Biology and Systematics of New World Heterospilus (Hymenoptera: Braconidae) Attacking Pemphredoninae (Hymenoptera: Sphecidae) Paul M. Marsh and Gabriel A. R. Melo (PMM) Cooperating Scientist, USD A Systematic Entomology Laboratory, c/o National Museum of Natural History, Washington, DC 20560 (correspondence address: P. O. Box 384, North Newton, Kansas 67117); (GARM) Snow Entomological Museum, University of Kansas, Lawrence, Kansas 66045 (present address: Departamento de Biologia, FFCLRP-USP, Av. Bandeirantes 3900, 14040-901, Ribeirao Preto, S.P., Brazil) Abstract. — Four new species of the braconid genus Heterospilus are described from Brazil and Costa Rica and a key to the five species known to attack crabronid wasps of the genera Micros- tigmus and Spihvnenn is presented. Biological notes are given for the hosts and the parasitoids and comments on their relationships are also offered. The genus Heterospilus Haliday is one of the largest of the family Braconidae with an estimated 500 or more species in the New World, most of which are unde- scribed. The genus belongs to the subfam- ily Doryctinae characterized by a circular opening present between the clypeus and mandibles (oral cavity) and a row of stout spines along the anterior edge of the fore tibia. It can be distinguished from most other braconid genera by the reduction or absence of fore wing vein 2RS and can be identified by the key to genera in Marsh (1997). We have very little information about the biology of species in the genus Heter- ospilus, which are, as far as known, idio- biont ectoparasitoids (Shaw and Huddles- ton 1991). Most records suggest parasitism of wood boring Coleoptera, especially Scolytidae, but a few species have been reared from stem boring Lepidoptera and one species from stem boring Symphyta. However, in an unusual host association, one species has been described from nests of the crabronid wasp genus Microstigmus Ducke (Richards 1935; Matthews 196'8). Recently, one of us (GARM) reared sev- eral species of Heterospilus from nests of species of Microstigmus and Spilomeua Shuckard that were being studied in Bra- zil and Costa Rica. The purpose of this pa- per is to provide descriptions of four new species of Heterospilus and present a brief discussion of the biologies of the hosts and parasitoids. BIOLOGICAL NOTES The genus Microstigmus constitutes a distinctive group of crabronid wasps, in particular because of the social behavior and the elaborate suspended nests found in several of its species (Matthews 1968; Richards 1972; West-Eberhard 1977; Mat- thews 1991; Melo in press). Microstigmus together with Arpnctophilus Smith, Spilo- meua and Xi/sma Pate form the subtribe Spilomenina of the Pemphredonini (Men- ke 1989). Melo (1994) presented evidence that Microstigmus is closely related to a group of species within Spilomena contain- ing S. nlini Antropov, S. kimseyi Antropov and other related undescribed species. The use also of S. atini as host bv Heterospilus, 14 Journal of Hymenoptera Research reported here for the first time, therefore should not be considered unexpected. Het- erospilus has not been found parasitizing other species of Neotropical Spilomena whose nests have been studied (Melo un- publ.). Very little is known about the biology of Heterospilus attacking Microstignnis wasps. The only available information is provided by Matthews (1968, 1991), Rich- ards (1972) and Melo and Campos (1993). Matthews (1968, 1991) reported on some aspects of the biology of H. microstigmi Richards parasitizing M. comes Krombein, and Melo and Campos (1993) reported M. myersi Turner being parasitized by a then undescribed species of Heterospilus, here named H. mattheivsi, new species. Parasitoid females have never been found inside the host nests and apparently always lay eggs directly from the outside. The positive correlation between the length of the parasitoid ovipositor and the thickness of the host nest wall found among the different species of Heterospilus (see below) provides additional evidence for this behavior. Matthews (1968) ob- served females of H. microstigmi crawling over nests of M. comes and inserting their ovipositors repeatedly into the nest. He was not able to ascertain if these insertions represented only probing or repeated ovi- positions. However, considering that only one egg or small parasitoid larva has been found on each of the attacked host im- matures and that in most Microstigmus nests no more than one immature in the appropriate stage is likely to be found, these insertions may be only for probing or for stinging and paralysis of the host larva. Only host pre-pupae seem to be at- tacked, since eggs and small larvae of Het- erospilus have been found only on imma- tures at this stage. The pre-pupal stage in Microstigmus is relatively short because these wasps are not known to enter dia- pause, although data on the duration of this stage is lacking. The egg is laid di- rectly on the surface of the integument of the host larva. Parasitized larvae seem to have their development interrupted which as probably caused by pre-oviposition stinging. In their Table 1, Melo and Mat- thews (1997) gave a record of a small par- asitic larva, attributed by them to Heter- ospilus, on a male pupa of Microstig77ius fla- vus Melo and Matthews (this record was erroneously printed as associated with nest 308 instead of nest 303). Since no adult Heterospilus has been reared from nests of this species of Microstigmus, it is possible that this larva represents another species of parasitoid (the only record of an unidentified Chalcididae attacking Micros- tigmus was obtained from this species), and not the result of exceptional behavior by a female Heterospilus. The length of the female ovipositor is clearly correlated with the thickness of the host nest wall or the distance of the brood cells from the nest outer surface. The lon- gest ovipositors are found in H. richardsi, new species, a species attacking Spilomena alini and an undescribed species of Mi- crostigmus. In both host species, the brood cells are usually situated deep inside the nests (see account under H. richardsi). The second longest ovipositors are found in fe- males of H. matthewsi, new species, a spe- cies associated with M. myersi. The thick- ness of the nest walls in M. myersi is quite variable (Melo and Campos 1993). This variation seems to be related to the nest age, with young nests having thinner walls and older nests tending to have an extra covering layer of dirt particles. The females of the other Heterospilus species have shorter ovipositors and attack Mi- crostigmus species whose nest walls are relatively thin. Matthews (1991) found a strong corre- lation between the number of brood cells in nests of M. comes and frequency of par- asitism by H. microstigmi; almost all para- sitized nests had eight or more cells. He explained this pattern also in terms of the relation between the length of the parasit- Volume 8, Number 1, 1999 15 oid ovipositor and the thickness of the host nest walls. Cells in nests with few cells tend to be centrally placed and to have thick walls, and therefore are out of reach for the short ovipositor of females of H. microstigmi. New cells are sequentially added toward the periphery of the nest and as their number increases, the outer wall of the nest becomes thinner, making the cells more accessible to the parasitoid's ovipositor. Except for silk secreted by fe- males, addition of new construction ma- terial to the nest as it grows, a behavior observed for example in M. meyersi (Melo and Campos 1993), does not occur in M. comes (Matthews and Starr 1984), the new cells being added by stretching the nest walls and holding the newly created pock- ets (future cells) with silk. One would expect females of Heterospi- lus to have an ovipositor long enough to cope with this kind of variation in wall thickness of the host's nests, since a long ovipositor could reach deep cells as well as those closer to the nest surface. How- ever, considering the situation found in H. niicwstigmi this does not seem to be the case. Apparently these parasitic wasps are under strong selective pressure to have the length of their ovipositor matching very closely the average depth of the ma- jority of their host's cells. As more mate- rial becomes available, it would be inter- esting to investigate the amount of intra- specific variation in ovipositor length. Cocoon spinning by larvae of Heteros- pilus also varies among the species de- scribed here. Matthews (1968) mentioned that H. microstigmi spins an opaque white cocoon, usually near the bottom of the host cell. Melo and Campos (1993) ob- served that larvae of H. mntthezvsi spin very rigid cocoons, which adults of M. myersi are apparently unable to open. Cells of M. myersi successfully parasitized by H. matthezosi become useless after emergence of the adult parasitoid. The lar- vae of H. richardsi spin cocoons somewhat thicker than those of H. microstigmi, but much thinner than in H. matthewsi. On the other hand, mature larvae and pupae of H. brasilophagous, new species, and H. ar- leiophagous, new species, were not encased by any sort of conspicuous cocoons. This apparent absence of cocoon spinning be- havior needs confirmation, however, be- cause cells of Microstigmus are lined with silk and it would have been easy to over- look a very loose cocoon adhered to the cell walls. Newly emerged adults of Het- erospilus leave their host's nests directly to the outside by chewing a hole in the nest wall, without passing through the nest en- trance. In active nests, these emergence holes are sealed later on with silk and par- ticles by the Microstigmus females. Body color also varies in an interesting way among the present species of Heter- ospilus. Species of Heterospihis attacking Microstigmus with a predominately black integument (including S. alini) also have a dark body color, whereas species associ- ated with light colored Microstigmus (pre- dominately or entirely yellow) have a light body color. The significance of this varia- tion in body color is unknown. The species of Heterospilus show a high degree of specificity regarding their hosts. Four of the present species are associated with only one species of Microstigmus (H. matthewsi, H. arleiophagus, H. hrnsilophagus) or at most with a group of closely related species (H. microstigmi). Only H. richardsi is known to attack hosts in different groups, but even in this case the two hosts nest in the same type of habitat. Most of the material used in the present study was collected by the second author while studying the biology of Microstigmus and Spilomena wasps in the region of Vi(josa (Minas Gerais, southeastern Brazil). Ex- cept for H. microstigmi, the remaining four species coexist sympatrically in this local- ity. The hosts of H. microstigmi, Microstig- mus species in the group theridii, are known only from the Amazon basin and Central America. Judging from this degree of specificity, additional collecting will 16 Journal of Hymenoptera Research probably reveal twice as many species of Heterospiliis attacking these crabronid wasps. Additional biological notes are included with the descriptions of the Heterospiliis species presented below. TAXONOMY OF HETEROSPILUS Although the genus Heterospiliis is large and badly in need of study for the entire Western Hemisphere, preliminary study indicates there will be many distinct spe- cies groups. The species in this study fall in a group with the following characters: vertex, mesonotum and mesopleuron co- riaceous (Figs. 9-11); flagellum unicolored without white band or tip; metasoma ter- gum 24-3 with two transverse scrobiculate grooves which enclose a slightly raised median area (Figs. 6-8, see arrow on Fig. 6). The following key is based heavily on characters found in the female although it will work with most males also. The iden- tification of males is best done by associ- ated rearings. KEY TO HETEROSPILUS SPECIES PARASITIC ON MICROSTIGMUS AND SPILOMENA 1. Body of both sexes entirely brown 2 Body of both sexes honey yellow, metanotum and metasomal terga occasionally marked with brown 3 2(1). Ovipositor of female equal to or longer than metasoma; fore wing with vein 3RSa longer than r-m (Fig. 1); hind femur yellow or at most yellowish-brown richardsi Marsh and Melo, new species - Ovipositor Vi to Vi length of metasoma; vein 3RSa equal to r-m (Fig. 2); hind femur brown on apical %, yellow on basal Va matthewsi Marsh and Melo, new species 3(1). Antennal flagellum yellow, first flagellomere longer than second; forewing vein 3RSa longer than r-m (Fig. 5); ovipositor of female usually equal to length of second and third metasomal terga combined microstigmi Richards - Antennal flagellum black, first flagellomere usually equal to second; forewing vein 3RSa equal to or shorter than r-m; ovipositor usually equal to length of first metasomal tergum 4 4(3). Fore wing vein 3RSa equal to vein r-m (Fig. 4); median transverse area of metasomal terga 2 + 3 (between transverse scrobiculate grooves) coriaceous medially, striate laterally brasilophagits Marsh and Melo, new species Fore wing vein 3RSa shorter than vein r-m (Fig. 3); median transverse area of metasomal terga 2+3 entirely coriaceous arleiophagits Marsh and Melo, new species Heterospilus arleiophagus Marsh and Melo, new species (Fig. 3) Female. — Body color: head yellow, palpi light yellow, antennal flagellum black, first flagellomere yellow at base, pedicel black, scape yellow with black longitudi- nal stripe on outer edge; mesosoma yel- low, metanotum, sides of scutellum and propodeum occasionally black; legs yel- low, wings slightly dusky, veins brown; metasoma yellow, first tergum occasion- ally marked with brown laterally, terga 2- 4 marked with brown medially. Body length: 2.5 mm. Head: vertex and frons weakly coriaceous, face and temple smooth; 26-27 antennomeres, first flagel- lomere equal in length to second; malar space about Vi eye height. Mesosoma: pronotum coriaceous dorsally with medi- an scrobiculate groove; mesonotal lobes coriaceous, median lobe with median lon- gitudinal depression, notauli scrobiculate and meeting in a wide rugulose area near- ly as wide as scutellum; scutellum coria- ceous; mesopleuron coriaceous, sternaulus short and weakly scrobiculate; propo- Volume 8, Number 1, 1999 17 Figs. 1-5. Wings of Heterospilus species.: 1, rlchardsi, new species; 2, matthcwsi, new species; 3, arkwplmgus, new species; 4, brasihpljagus, new species; 5, micwsfiginl Richards. deum rugose with small basal lateral co- riaceous spots. Legs: fore tibia with row of 4-6 short spines. Wings: fore wing vein 3RSa shorter than vein r-m (Fig. 3). Me- tasoma: first tergum slightly longer than apical width, carinate rugulose, raised me- dian area not set off by carinae; second tergum carinate, transverse median area between two transverse scrobiculate grooves entirely coriaceous; third tergum coriaceous; remainder of terga weakly co- riaceous; ovipositor as long as first meta- somal tergum. Male. — Agrees with description of fe- male except as follows: scape usually en- tirely yellow; hind wing with oblong stig- ma at base (as in Fig. 1). Holoti/pe. — Female: BRAZIL, Viqosa, MG, January 28, 1990, G. A. R. Melo, col- lected in nest of Microstigmus arlei, nest 330. Deposited in the Museu de Zoologia, Universidade de Sao Paulo, Brazil. Paratypes. — BRAZIL: 2 females, 1 male. same data as holotype with additional dates of August 7, 1990 and February 2, 1992, nests 330, 367, 368, 603. Deposited in the Museu de Zoologia, Universidade de Sao Paulo, Brazil. Biology. — Reared from nests of Micros- tigmus arlei Richards. This Microstigmus species makes suspended petiolated nests. The nests have been found attached only to surfaces other than plant leaves, like in- clined tree trunks or branches, earth bank walls, and structural timber of exposed roofs of huts; the nests reported by Rich- ards (1972) were hanging from the face of a rock. Cocoons have not been observed for this species. Etymology. — The species name is based on the host wasp. Heterospilus brasilophagus Marsh and Melo, new species (Fig. 4) Female. — Body color: head yellow, palpi light yellow, antennal flagellum black. 18 Journal or Hymenoitera Research first flagellomere yellow at base, pedicel black, scape yellow with black longitudi- nal stripe on outer edge; mesosoma yel- low, metanotum, sides of scutellum and scutellar sulcus black; legs yellow; wings slightly dusky, veins brown; metasoma yellow, terga 2-3 marked with brown. Body length: 2.5 mm. Head: vertex and frons weakly coriaceous, face and temple smooth; 26-27 antennomeres, first flagel- lomere equal in length to second; malar space about Vi eye height. Mesosoma: pronotum coriaceous dorsally with medi- an scrobiculate groove; mesonotal lobes coriaceous, median lobe with median lon- gitudinal depression, notauli scrobiculate and meeting in a wide longitudinal cari- nate area nearly as wide as scutellum; scu- tellum coriaceous; mesopleuron coria- ceous, sternaulus short and weakly scro- biculate; propodeum rugose with small basal lateral coriaceous spots. Legs: fore tibia with row of 4-6 short spines. Wings: fore wing vein 3RSa usually as long as r- m (Fig. 4). Metasoma: first tergum slightly longer than apical width, carinate rugu- lose, raised median area not set off by ca- rinae; second tergum carinate, transverse median area between two transverse scro- biculate grooves coriaceous, striate later- ally; third tergum coriaceous; remainder of terga weakly coriaceous; ovipositor as long as first metasomal tergum. Male. — Essentially as in female; hind wing with oblong stigma at base; with 25- 27 antennomeres. Holoti/pe. — Female: BRAZIL, Vi^osa, MG, February 15, 1992, G.A.R. Melo, col- lected in nest of Micwstigmus brasiliettsis, nest 582. Deposited in the Museu de Zoo- logia, Universidade de Sao Paulo, Brazil. Pamti/pes. — BRAZIL: 5 males, same data as holotype with additional date of July 25, 1992, nests 579, 580, 619. Deposited in the Museu de Zoologia, Universidade de Sao Paulo, Brazil. Biology. — Reared from nests of Micw- stigmus brasiliensis Melo. This Microstigmus species builds the type of pendulous nests considered typical for this genus (see Richards 1972, West-Eberhard 1977); its nests were described and illustrated in Melo (1992). Etymology. — The species name is based on the host wasp. Heterospiliis matthezvsi Marsh and Melo, new species (Figs. 2, 6) Female. — Body color: head brown, palpi yellow; scape yellow with brown longi- tudinal strip laterally, pedicel brown, fla- gellum dark brown; mesosoma and me- tasoma dark brown; legs yellow, hind fe- mur brown on apical %; wings hyaline, veins light brown, tegula yellow. Body length: 3 mm. Head: vertex and frons fine- ly coriaceous, temple smooth, face smooth with fine striations laterally; 26-28 anten- nomeres; malar space about V3 eye height; maxillary palpus longer than fore tarsus. Mesosoma: pronotum coriaceous and shining dorsally with median scrobiculate groove; mesonotal lobes finely coriaceous, median lobe with median longitudinal de- pression, scutellum finely coriaceous and shining; mesopleuron coriaceous, sternau- lus short and weakly scrobiculate; propo- deum rugose, median rugae more distinct indicating vague areola, small basal lateral coriaceous spots. Legs: fore tibia with row of 4-5 short spines on anterior edge. Wings: fore wing with vein 3RSa equal in length to or slightly longer than vein r-m (Fig. 2). Metasoma (Fig.6): first tergum slightly longer than apical width, carinate rugulose, median raised area set off by complete distinct longitudinal carinae; second tergum weakly carinate coria- ceous, ending in distinct transverse scro- biculate groove; third tergum with second transverse scrobiculate groove which meets first groove at sides, tergum cari- nate coriaceous before this groove and be- yond to end of tergum; remainder of terga finely coriaceous and shining; ovipositor V3 to V2 length of metasoma. Volume 8, Number 1, 1999 19 Male. — Essentially as in female; hind wing with oblong stigma at base. Holotype. — Female: BRAZIL, Viqosa, MG, February 10, 1992, G. A. R. Melo, col- lected in nest of Microstigmus myersi, nest 586. Deposited in the Museu de Zoologia, Universidade de Sao Paulo, Brazil. Paratypes. — BRAZIL: 3 females, 2 males, same data as holotype with additional dates of March 16, 1992, July 31, 1992, Jan- uary 28, 1990, nests 585, 586, 610, 620, 637; 1 female, Vargem Alta, ES, September 4, 1992, J. N. C. Louzada, collected in nest of MicrosHginus myersi, nest 637; 1 male, Belo Horizonte, MG,' July 18, 1992, J. N. C. Lou- zada, collected in nest of Microstigmus myersi. Deposited in the Museu de Zoolo- gia, Universidade de Sao Paulo, Brazil. Biology. — Reared from the nests of Mi- crostigmus myersi Turner. Some aspects of the biology of H. matthewsi were presented in Melo and Campos (1993); this species was referred to as Heterospilus sp. in their paper. The rigid cocoons spun by its lar- vae set H. matthewsi apart from other Het- erospilus attacking Microstigmus, whose larvae spin only a thin cocoon or no co- coon at all. It would be interesting to in- vestigate the significance of these rigid co- coons. Distribution. — Known only from Brazil. Etymology. — Named for R. W. Matthews who gave the first complete description of the biology of Heterospilus microstigmi (see Matthews 1968). Heterospilus microstigmi Richards (Figs. 5, 7, 9-11) Heterospilus microstignti Richards 1935:131. Ho- lotype female, deposited in The Natural His- tory Museum, London. Female. — Body color: head, mesosoma and metasoma yellow or honey-yellow, mesonotal lobes, metanotum, propodeum dorsally, and metasomal terga 1-4 often marked with brown; antenna varying from entirely brown to scape, pedicel and basal flagellomeres yellow, apical flagel- lomeres brown; legs yellow; wings hyaline or slightly dusky, veins light brown, te- gula yellow. Body length: 2.5-3.0 mm. Head: vertex and frons finely coriaceous (Fig. 11), face and temple smooth; 24-28 antennomeres; malar space about Vs eye height; maxillary palpus longer than fore tarsus. Mesosoma (Figs. 9, 10): pronotum coriaceous and shining dorsally with me- dian scrobiculate groove; mesonotal lobes finely coriaceous, median lobe with me- dian longitudinal depression, scutellum finely coriaceous; mesopleuron coria- ceous, sternaulus short and weakly scro- biculate; propodeum rugose, median ru- gae more distinct indicating vague areola, small basal lateral coriaceous spots. Legs: fore tibia with row of 4-5 short spines on anterior edge. Wings: fore wing with vein 3RSa about equal in length to vein r-m (Fig. 5). Metasoma (Fig. 7): first tergum slightly longer than apical width, carinate rugulose, median raised area set off by short indistinct basal carinae; second ter- gum carinate rugulose, ending in distinct transverse scrobiculate groove; third ter- gum with second transverse scrobiculate groove which meets first groove at sides, tergum carinate rugulose before this groove, strongly coriaceous beyond to end of tergum; remainder of terga coriaceous; ovipositor Vi to % length of metasoma. Male. — Essentially as in female; hind wing with oblong stigma at base. Distribution. — Trinidad, Costa Rica, Bra- zil. For this study we have seen 23 speci- mens from the following localities in Bra- zil: Manaus, AM; Beruri, AM; Vi^osa, MG; Ma to G rosso. Biology. — Previously reared from nests of Microstigmus theridii Ducke and M. comes Krombein (Richards 1935; Matthews 1968, 1991). Specimens from this study have been reared from nests of species of the M. theridii group which includes M. comes (nests 442, 443, 447, 531). This spe- cies group is restricted to northern South America and Central America (Melo un- publ.) 20 Journal of Hymenoptera Research i-iia^i IP'JQ ^^^^^^T''^'?^^* ^ 1B^^H| ■ J^ ■ ■ r-. ^ J H '^ ^^^Bn/^^^^^^^^Ha| HW ''' '^j^M ^^Sl ^^^^^^^^^^H ^^^pHjk ^^3 Figs. 6-11. Body parts of Hcterospilus species: 6, metasoma, iiuittlicwsi, new species (arrow points to raised median area mentioned in the descriptions); 7, metasoma, iiuciosti^iiii Richards; 8, nietasoma, liclmnisi, new species; 9, mesosoma, lateral view, inicroiitignn; 10, mesosoma, dorsal view, inicwstigDii; 11, vertex, iincrostigiin. Heterospilus richardsi Marsh and Melo, new species (Figs. 1, 8) Female. — Body color: head brown, face light brown, palpi yellow; scape yellow with brown longitudinal strip laterally, pedicel brown, flagellum dark brown; me- sosoma and metasoma dark brown; legs yellow; wings hyaline, veins light brown, tegula yellow. Body length: 3 mm. Head: vertex and frons finely coriaceous, temple smooth, face smooth with fine striations laterally; 24-26 antennomeres; malar space about % eye height; maxillary pal- pus longer than fore tarsus. Mesosoma: pronotum coriaceous and shining dorsally with median scrobiculate groove; mesono- tal lobes finely coriaceous, median lobe with median longitudinal depression, scu- tellum smooth and shining; mesopleuron coriaceous, sternaulus short and weakly scrobiculate; propodeum rugose, median rugae more distinct indicating vague are- ola, small basal lateral coriaceous spots. Legs: fore tibia with row of 4-5 short spines on anterior edge. Wings: fore wing with vein 3RSa longer than vein r-m (Fig. 1). Metasoma (Fig. 8): first tergum slightly longer than apical width, carinate rugu- lose, median raised area set off by short indistinct basal carinae; second tergum carinate rugulose, ending in distinct trans- verse scrobiculate groove; third tergum with second transverse scrobiculate groove which meets first groove at sides, tergum carinate rugulose before this groove, smooth beyond to end of tergum; remainder of terga smooth and shining; Volume 8, Number 1, 1999 21 ovipositor as long as or longer than me- tasoma. Male. — Essentially as in female except as follows; scape usually entirely yellow; hind wing with oblong stigma at base. Holotype. — Female: BRAZIL, Araponga, MG, March 22, 1992, G. A. R. Melo, col- lected in Spilomena aliui nest. Deposited in the Museu de Zoologia, Universidade de Sao Paulo, Brazil. Parah/pes. — BRAZIL: 1 female, 1 male, same data as holotype; 1 male, Vi^osa, MG, April 26, 1989, collected in nest of Spilomena alini. COSTA RICA: 3 females, 2 males, Heredia, Estacion Biol. La Selva, 10°25'N, 84°0'W, 80m, June 14, 1996, reared from nests of Micwstigmns sp., nests 677, 680, GAR Melo. Deposited in the Museu de Zoologia, Universidade de Sao Paulo, Brazil and the national Muse- um of Natural History, Washington, DC. Biology. — Reared from the nests of Spi- lomena alini Antropov in Brazil and an un- described species of Microstigmus from Costa Rica closely related to M. xanthos- celes Melo and Matthews. Spilomena alini excavates its nests in earth bank walls and in small soil clumps hanging from rootlets in banks (Carvalho and Zucchi 1989; Melo unpubl.; see Fig. 1 in Melo and Campos (1993) for an illustration of this kind of habitat). Some nests have also been found inside abandoned mud cells of eumenine wasps hanging from roots in banks (in this case, the nest tunnels and cells were dug in the loose detritus filling up the mud cells). Hetewspilus richardsi was reared only from nests of S. alini built in soil clumps. In these nests, some of the host cells are close to the surface, especially in the smaller clumps, and therefore within reach of the parasitoid ovipositor. Nests built within bank walls are apparently protected from parasitism by Heterospilus. The nests of the second host of H. ri- chardsi, Microstigmus sp., are small to me- dium bags (3-12 mm long) built on hang- ing rootlets in earth banks or in tree trunks. The external walls of the nest are made of soil particles (for nests in banks) or particles of dead wood (nests in tree trunks) aggregated with silk from the fe- male's silk glands. The central part of the nest has a sponge-like appearance and is formed by anastomosing pillars and chan- nels; except in the upper part of the nest, this central portion is not in contact with the external walls. The brood cells are lo- cated in the central portion. It seems that only small nests, in which the cells are rel- atively close to the nest surface, are subject to parasitism by Heterospilus. One female and one male of H. richardsi were reared from a nest with six Microstigmus females (nest 677), while two females and one male were reared from a nest with only one Microstigmus female (nest 680). Four additional nests, containing four, five, 11 and 27 adult Microstigmus respectively, produced no Heterospilus. Distribution. — Brazil, Costa Rica. Etymology. — Named for O. W. Richards who described the first species of Heter- ospilus reared from Microstigmus wasps. Heterospilus species We have seen one female (nest 598) reared from the nests of an apparently un- described species of Microstigmus species from the bicolor group in Vi^osa. As in sev- eral other species of the group (West-Eber- hard 1977), this Microstigmus species feeds its larvae progressively with Cicadellidae nymphs. This single female Heterospilus is similar to microstigmi but differs in having the body somewhat more coarsely coria- ceous or punctate and darker wings. More specimens are needed to determine if it is a variation of microstigmi or another spe- cies. Also, we have seen one badly damaged female specimen and one male specimen (nest 403) reared from Microstigmus leuder- ivaldti species group in Manaus, northern Brazil. These specimens are similar to mi- crostigmi but are darker than identified members of that species. Exact placement 22 Journal of Hymenoptera Research of these must wait until more undamaged specimens are collected. ACKNOWLEDGMENTS We thank Tom Huddleston, The Natural History Museum, London, for comparison of specimens of Hetcrospilus micnistigini with the holotype. Additional specimens of H. microstigmi were also provided by Robert Matthews, University of Georgia, Athens. The assistance of Bruce Cutler, EM Laboratory, University of Kansas, Lawrence, is greatly appreciated. Partial funding for the first author's (PMM) studies of the Doryctinae of Costa Rica was provided by National Science Foundation grant DEB-972614L LITERATURE CITED Carvalho, L. M. and R. Zucchi. 1989. Aspectos fen- ologicos de Spilonicnn sp. (Hymenoptera, Sphe- cidae, Peniphredoninae). Revista Brasileira dc Bio- logia 49:799-807. Marsh, P. M. 1997. Subfamily Doryctinae, pp. 206- 233. In: K. A. Wharton, P. M. Marsh and M. J. Sharkey (eds.). Manual of the New World Gen- era of the Family Braconidae (Hymenoptera). Special Publication of the International Societi/ ofHy- menopterists No. 1, 439 pp. Matthews, R. W. 1968. Nesting biology of the social wasp Microstigmuf conies (Hymenoptera: Sphe- cidae, Peniphredoninae). Psyche 75:23-45. Matthews, R. W. 1991. Iivolution of social behavior in sphecid wasps, pp. 570-602. ///: K. G. Ross and R. W. Matthews (eds.). The Social Biology of Wasps. Comstock, Ithaca. Matthews, R. W. and K. C. Starr. 1984. Micwstigmiis comes wasps have a method of nest construction unique among social insects. Biotiopica 16:55-58. Melo, G. A. R. 1994. Origin and diversification of the Mierostignnis wasps (Hymenoptera, Sphecidae), p. 357. In: A. Lenoir, G. Arnold and M. Lepage (eds.), Lcs Insectes Sociaiix (12o Congresso Inter- nacional da lUSSI). Univ. Paris Nord, Paris. Melo, G. A. R. In press. Comportamento social em vespas da familia Sphecidae (Hymenoptera, Apoidea). Oecologica Brasiliensis. Melo, G. A. R. and L. A. O. Campos. 1993. Nesting biology of Microstigmus myersi Turner, a wasp with long-haired larvae (Hymenoptera: Spheci- dae, Pemphredoninae). Journal of Hymenoptera Research 2:183-188. Melo, G. A. R. and R. W. Matthews. 1997. Six new species of Microstigmus wasps (Hymenoptera: Sphecidae), with notes on their biology. Journal of Natural History 31:421-437. Menke, A. S. 1989. Arpactofihilus reassessed, with three bizarre new species from New Guinea (Hy- menoptera: Sphecidae: Pemphredoninae). Inver- tebrate Taxonomy 2:737-747. Richards, O. W. 1935. Two new parasites of aculeate Hymenoptera from Trinidad. Stylops 4(6):131- 133. Richards, O. W. 1972. The species of the South Amer- ican wasps of the genus Microstigmus Ducke (Hy- menoptera: Sphecidae, Pemphredoninae). Trans- actions of the Royal Entomological Society of London 124:123-148. Shaw, M. R. and T. Huddleston. 1991. Classification and Biology of Braconid Wasps (Hymenoptera: Braconidae). Handbooks for the Identification of Brit- ish Insects 7(11):1-126. West-Eberhard, M. J. 1977. Morphology and behavior in the taxonomy of Microstigmus wasps, pp. 123- 125. Proceedings of the 8th International Congress of the lUSSI. Wageningen, Holanda. J. HYM. RES. Vol. 8(1), 1999, pp. 23-34 Cretaceous Digger Wasps of the New Genus Bestiola Pulawski and Rasnitsyn (Hymenoptera: Sphecidae: Angarosphecinae) Alexandr p. Rasnitsyn, Wojciech J. Pulawski, and Xavier Martinez-Delclos (APR) Paleontological Institute, Russian Academy of Sciences, Moscow 117647, Russia; E-mail; rasna@glasnet.ru; (WJP) Department of Entomology, California Academy of Sciences, Golden Gate Park, San Francisco, California 94118, USA; E-mail: wpulawski@calacademy.org; (XMD) Departament de Geologia Dinamica, Geofisica y Paleontologia, Facultad de Geologia, Universidad de Barcelona, 08071 Barcelona, Spain; E-mail: delclos@natura.geo.ub.es Absfrncf. — The new genus Bestiola Pulawski and Rasnitsyn, characterized by a unique wing venation, is described for the following four new species from the Lower Cretaceous: hispanica Martinez Delclos and Rasnitsyn (type species) from northeastern Spain, communis Pulawski and Rasnitsyn from central Mongolia, subpetiolata Pulawski and Rasnitsyn from central Mongolia, and temiipes Pulawski and Rasnitsyn from Eastern Siberia. The genus is assigned to Angaro- sphecinae Rasnitsyn, 1975, new status by Rasnitsyn, which is treated as a paraphylehc Lower Cretaceous subfamily of Sphecidae. The family name Baissodidae Rasnitsyn, 1975, is synonymized with Angarosphecidae Rasnitsyn, 1975, by Rasnitsyn. Sphecid wasps of the archaic subfamily Angarosphecinae (= Baissodinae) are the most abundant taxa among the Early Cre- taceous Hymenoptera, particularly in the middle interval of that epoch, probably af- ter Berriasian and before Albian, 140-113 myr before present (Rasnitsyn et al. 1998). Only a fraction of the material accumulat- ed in the collections has been described (Evans 1969; Rasnitsyn 1975, 1986, 1990; Hong 1984; Zhang 1985, 1992; Darling and Sharkey 1990; Jarzembowski 1991; Ansor- ge 1993; Ren et al. 1995; Rasnitsyn et al. 1998). The Early Cretaceous fossils de- scribed here originate from three distant areas of Eurasia (central Mongolia, East- ern Siberia, and Spain), but have a unique wing venation and are all approximately the same age. Most of the specimens examined were collected in Bon Tsagan, Central Mongo- lia, a rich fossil site 5-8 km north of Bon Tsagan Nuur (= Bon Tsagan Lake), in Bayanhongor Aymag (= Region). Speci- mens were impressed in marl of the Khur- ilt rock unit, Bon Tsagan Series (Sinitza 1993), possibly of Aptian age (Ponomar- enko 1990). This hymenopteran assem- blage is related to the Wealden Super- group of the Southern England (Valangi- nian to Barremian: Rasnitsyn et al. 1998). Siberian material was collected in two localities east of Lake Baikal. One is Bais- sa, a riverside outcrop on the left bank of the Vitim River, 3 km downstream of the former lodge Baissa and 45 air km up- stream of the Romanovka Village, Buryat Republic. The fossils, impressed in marl of the Zaza Formation, are related to the Pur- beck of South England (Berriassian), based on their hymenopteran assemblages (Ras- nitsyn et al. 1998). The other Siberian lo- cality, Semyon, is at Semyon Creek, 3.5 km SW of Elizavetino Village, west of Chi- ta, Chita Oblast'. The age of insectiferous mudstones, disputable within the Early Cretaceous, is correlated either with Bais- sa (Zherikhin 1978) or with Bon-Tsagan (Dmitriev and Zherikhin 1988). The Spanish specimen originates from the Montsec Range, central Lerida Prov- ince, and comes from lithographic lime- 24 Journal of Hymenoptera Research Stones named La Pedrera de Meia, located 5 km W of Santa Maria de Meia in La No- guera comarca, possibly of Berriassian-Va- langinian age (Martinez-Delclos 1995). The following morphological structures, variously termed in the literature, are here defined or redefined as follows for clarity and convenience sake: - mesosoma: thorax and propodeum combined; - metasoma: abdomen excluding the pro- podeum (= gaster of Bohart and Menke 1976); - metapostnotum: propodeal enclosure of Bohart and Menke 1976; - spiracular lobe (as in Rasnitsyn 1988): pronotal lobe of Bohart and Menke 1976; - adlateral lines: parapsidal line of Bohart and Menke (1976); we prefer this term to avoid confusion, as parapsidal lines of other entomologists correspond to notauli of most hymenopterists; - cell l+2r: submarginal cell I of Bohart and Menke (1976); - 3r: marginal cell of Bohart and Menke (1976); - 2rm and 3rm: submarginal cells II and 111 of Bohart and Menke (1976); - Imcu and 2 mcu: discoidal cells 1 and II of Bohart and Menke (1976); - crossveins 2r-m and 3r-m: distal mar- gins of 2rm and 3rm, respectively (as in Richards, 1956, and Gauld and Bolton, 1988), and corresponding to Ir-m and 2r-m of Bohart and Menke, 1976); unlike the latter two authors, we call Ir-m the vein that extends from cell l+2r to cell Imcu (i.e., the vein that separates the basal cell from cell 2rm) in the xyelid genus Pleroiieurn; - vein IRS: a veinlet between basal cell ( = medial cell of Bohart and Menke, 1976) and cell l-(-2r; - vein 2RS: a veinlet between cells l-l-2r and 2rm; - vein 2r-rs: a veinlet between cells l-l-2r and 3r, called 2r by Bohart and Menke, 1976. The abbreviation PIN stands for the Pa- leontological Institute, Russian Academy of Sciences, Moscow, Russia. TAXONOMY Family Sphecidae Latreille Subfamily Angarosphecinae Rasnitsyn, new status Angarosphecidae Rasnitsyn 1975:109. Type ge- nus: Angnrosphcx Rasnitsyn 1975:110. Baissodidae Rasnitsyn:1975:122. Type genus: Bnissodes Rasnitsyn 1975:123. New synonym by Rasnitsyn. Angarosphecinae are archaic Mesozoic wasps that lack the synapomorphies of any extant sphecid subfamily. They may be paraphyletic with respect to other Apo- idea because they are not defined by any synapomorphy. They are treated here as a subfamily of Sphecidae because there is evidence in some specimens of two unique synapomorphies of Apoidea (they also lack any synapomorphy that would ally them with any other Aculeata). In particular, Pompilopterus corpus Rasnitsyn and Jarzembowski has an elongate spirac- ular lobe, pronotum thickened preapical- ly, and an enlarged metapostnotum (Ras- nitsyn, Jarzembowski, Ross 1998, Fig. 36), and Angawsphex myrmicopterus Rasnitsyn has an enlarged metapostnotum (Rasnit- syn 1980, Fig. 172). A large metapostno- tum is also found in Bestiola tenuipes (Fig. 7). None of the Angarosphecinae has plu- mose setae or enlarged hindbasitarsi typ- ical of bees. The form of the adlateral line of Bestiola and other Angarosphecidae dif- fers from that of all extant Apoidea in that it extends to the posterior margin of the mesoscutum, as pointed out to us by M. A. Prentice (oral communication). This in- dicates that Angarosphecinae very proba- bly represent the most basal lineage of known Apoidea. Rasnitsyn (1975) recognized Baissodi- dae based on the presence of a unique me- dian scutal sulcus believed to be lacking in all other non-bethyloid Aculeata (in- Volume 8, Number 1, 1999 25 eluding Angarosphecidae). Subsequently, he (Rasnitsyn 1980) included Angarosphex in the Sphecidae and hypothesized that the Baissodidae were sphecid ancestors. However, lAngarospihex pallidus Rasnitsyn, 1986, from the lowermost Lower Creta- ceous of Mongolia, combines the wing ve- nation of A)igarosphex with the presence of a median scutal sulcus, although the latter is only slightly indicated. Because of this combination, Baissodidae are here synon- ymized with Angarosphecidae. Bestiola Pulawski et Rasnitsyn, new genus Name derivation. — Bestiola, Latin for small beast. Gender feminine. Ti/pe species. — Bestiola hispanica Martinez Delclos et Rasnitsyn, new species. Lower Cretaceous of Spain. Recognition. — Bestiola is easily recognized by its unique wing venation (Figs 1-8). It has three radiomedian (= submarginal) cells, and cell 2rm receives veins Im-cu and 2m-cu (= both recurrent veins). The com- bination of three unusual features differ- entiates it from all other sphecid genera with these characteristics, both extinct and extant: 1. cell 3rm broader on the costal side than on the anal side (as in the North American genus Xenosphex and some Pa- lanis); 2. crossvein 3r-m joining RS near the distal end of the latter (as in the North American genus Eucerceris and some Pala- nis); and 3. crossvein 2r-m equidistant from 2m-cu and 3r-m or closer to the latter (as in the Old World Tacln/sphex brevipennis Mercet, and several other Larrini and some Diploplectron approach this condition). Description. — Size medium to large, length of forewing 5-15 mm. Antenna with no conspicuous modifications, at least basal flagellomeres longer than wide (all flagellomeres in most species). Occip- ital carina almost circular, reaching hypos- tomal carina. Ocelli not modified, distant from eye. Pronotum short, wide, thick- ened preapically, separated from meson- otum by a groove (Fig. 8). Mesoscutum without median scutal sulcus, with long notauli and adlateral lines; mesopleuron with episternal and scrobal sulci, hyper- sternaulus, and possibly omaulus. Meta- notum short, metapostnotum long, trun- cated, with median line. Propodeal spira- cle elongate, slit-like. Forewing: pterostig- ma well defined; basal vein distant from pterostigma, evenly arched; cell 3r acu- minate at wing foremargin; crossvein 2r- rs longer than width of pterostigma; 2r-m sinuate or arching outwardly, closer to 2r- rs than to 3r-m on RS, equidistant from 2m-cu and 3r-m on M or closer to the lat- ter; 3r-m straight or arching outwardly; Im-cu received near base of cell 2rm; M sharply angled at 2m-cu; cu-a interstitial with M or narrowly postfurcal. Hindwing venation complete, cu-a meeting Cu well beyond M-I-Cu fork. Fore and mid femora with well-defined, narrow trochantellus (hindleg condition unknown), but no ob- vious specializations. Metasoma rounded basally (neither petiolate nor peduncu- late). Composition. — Four species from the Lower Cretaceous of Spain, Eastern Sibe- ria, and Mongolia, as described hereafter. Taxonomic position. — Bestiola is a mem- ber of Aculeata s. s. (= Vespoidea -I- Apo- idea) as evidenced by its sexually dimor- phic antennal flagellum of 10 (female) and 11 articles (male). Unlike all Chrysidoidea, it possesses a complete set of forewing and hindwing cells. It belongs to Apoidea because it shares two unique synapomor- phies of the superfamily: an elongate me- tapostnotum and a pronotum thickened preapically. The genus belongs to Spheci- dae because it lacks the enlarged hindbas- itarsus and plumose body setae of Apidae s. 1. The genus shares with most other An- garosphecinae the position of Im-cu which inserts near the RS+M fork. This feature occurs elsewhere only in some oth- er Apoidea. The genus also lacks any syn- apomorphy that would place it in any ex- tant sphecid subfamily and so we attribute it to Angarosphecinae. 26 Journal of Hymenoptera Research KEY TO SPECIES OF BESTIOLA 1. Forewing cell 2rm conspicuously narrowing anterad, its costal margin markedly shorter than crossvein 2r-rs; the latter emerging shortly after pterostigma's midlength (Fig. 8) Bestiola subpetiolata Pulawski and Rasnitsyn, new species - Forewing cell 2rm moderately narrowing anterad, its costal margin about as long as cross- vein 2r-rs; the latter emerging markedly beyond pterostigma's midlength 2 2. Forewing length 5 mm; crossvein 2r-rs markedly shorter than 2RS (Fig. 1); hindfemur and gastral terga with pale spots (Fig. 1); metasoma conspicuously setose (Fig. 1) Bestiola hispanica Martinez Delclos and Rasnitsyn, new species - Forewing length 11.0-14.5 mm; crossvein 2r-rs longer to minimally shorter than 2RS (Figs. 2-7); hindfemur and gastral terga without pale spots (Figs. 2-7); metasoma not setose or less conspicuously setose (Fig. 5) 3 3. Head, wing veins, and legs except coxae light; forefemur slightly more elongate (Fig. 7) Bestiola tentiipes Pulawski and Rasnitsyn, new species - Head, wing veins, and legs dark; forefemur slightly stouter (Figs. 2, 3) Bestiola communis Pulawski and Rasnitsyn, new species Bestiola hispanica Martinez Delclos and Rasnitsyn, sp. n. (Fig. 1) Name derivation. — Hispanica, Latin for Spanish. Recognition. — Bestiola hispanica differs from its congeners by its small size (fore- wing length 5 mm rather than 11.0-14.5 mn\), crossvein 2r-rs markedly shorter than 2RS (longer to minimally shorter in the other species), the presence of pale spots on the hindfemur and gastral terga, and a conspicuously setose metasoma. Description. — Female unknown, male: Fig. 1. Body dark (including wing veins), but hindfemur with pale spot posteroapi- cally and terga 1-lV each with a pair of preapical spots. Head and metasoma con- spicuously setose. Flagellomeres nearly equal in width, longer than wide, becom- ing shorter toward apex. Head relatively small, narrowing toward mouthparts; gena somewhat inflated; malar space half length of eye; clypeal free margin project- ing mesally; hindocellus separated from eye by about its own diameter. Forewing: posterior pterostigmal margin straight; 2r- rs joining pterostigma near the latter's apex, about as long as costal margin of cell 2rm and half as long as 2RS; 3r-m straight; costal margin of cell 2rm about twice as long as anal margin; cu-a almost intersti- tial with M+Cu fork. Hindwing with long row of hamuli (10 preserved) and cu-a an- gling at Cu. Legs not elongate; hindfemur as long as head width, moderately thick subbasally, not attenuated apically, its dorsal margin convex except subapically, ventral margin almost straight (ventral margin slightly convex on right femur, probably due to fossil compression). Hind tibia 1.25 times as long as femur, with no spines but with one spur. Hind- tarsus slightly longer than tibia, basitarsus longer than tarsomeres II-IV combined. Apical gastral segments and genitalia not preserved. Body length ca 9 mm, distance from forewing base to apex of cell 3rm 5.1 mm. Material examined. — Holotype (LP92/ SC/3662): male, Spain: Lerida Province: La Pedrera de Meia 5 km W Santa Maria de Meia (Institut d'Estudis Ilerdencs, Ler- ida, Spain). Bestiola commtmis Pulawski et Rasnitsyn, new species (Figs. 2-6) Name derivation. — Communis, Latin for common. Volume 8, Number 1, 1999 27 Fig. 1. Bestwla lufpauicn Martinez-Delclos and Rasnitsyn, new species, holotype: cly — clypeus; cocc — occipital carina; f — femur; to — occipital foramen; md — mandible; ppl — propleuron; shy — hypostomal suture; sipg — interpostgenal suture. Recognition. — The following combina- tion of characters is unique to Bestiola cotn- miinis: forewing length 11.0-14.5 mm (5 mm in hispanica); costal margin of cell 2rm about as long as crossvein 2r-rs (markedly shorter in suhpetiolata); 2r-rs longer to min- imally shorter than 2RS (markedly shorter in hispanica), emerging markedly beyond pterostigma's midlength (near pterostig- ma's midlength in suhpetiolata); hindfemur and gastral terga without pale spots (with pale spots in hispanica); and forefemur not elongate (slightly elongate in teniiipcs, compare Figs. 2 and 3 and 7). Unlike ten- iiipcs, the body of communis is all dark, in- cluding the wing veins. Description. — Female (Fig. 2), sex un- known in remaining specimens (Figs. 3- 6). Body and appendages uniformly dark, metasoma inconspicuously setose (Fig. 5). Scape about as long as midflagellar arti- cles; pedicel transverse; flagellomeres more than twice as long as wide, flagel- lomere 1 almost as long as II and 111 com- bined, following ones gradually shorter and thinner toward antennal apex, apical 28 Journal of Hymenoptera Research Fig. 2. Bcstiola coniiiniiuf Pulawski and Rasnitsyn, new species, hoiotype: d — discrimen (interpleural suture); liy — hypostoma; hys — hypersternaulus or signum?; Nl — pronotum; oa — omaulus?; ss — scrobal suture; tl — trochantellus; tr — trochanter; other abbreviations as in Fig. 1. flagellomere almost 3 times as long as wide. Head nearly circular in front view; eyes elongate, widest below midheight, with inner margin concave; malar space present; anterior clypeal margin protrud- ing into wide medial lobe that is shallowly emarginate apically. Notauli and adlateral lines complete or nearly so. Mesopleuron with long, almost straight scrobal sulcus and hypersternaulus, possibly also with omaulus. Wing venation as in hispaiiica ex- cept 2r-rs subequal in length to 2RS and 3r-m arched. Fore- and midfemora thick- est subbasally, narrow apically, with dor- sal margin straight and ventral convex; forefemur as long as head width, midfe- mur slightly longer; fore- and possibly midtibiae shorter than respective femora. Metasoma missing in type series but pres- ent in specimen PIN 3559/4526 (that is ex- cluded from type series). Forewing length 13.0-14.5 mm (11.0 mm in the specimen PIN, 3559/4526 not included into the type series. Fig. 5). Material examined. — Hoiotype: Central Mongolia, Bon-Tsagan, bed 87/8 (PIN. 3559/4525, incomplete female specimen with propodeum, hindlegs and metasoma missing). Paratypes: same locality and bed (PIN, 3559/4528, incomplete specimen with most of antennae and legs, part of thorax and all metasoma missing; PIN, 3559/ 4530, two damaged wings mixed with other insect remains in vertebrate drop- ping). Excluded from type series but possibly conspecific: same locality and bed (PIN, 3559/4526, a somewhat damaged speci- men: Fig. 5); Eastern Siberia, Semyon (PIN, 2385/2392, an isolated wing with venation nearly identical to those of coiii- niiinis and tenuipe^, and attributed to the Volume 8, Number 1, 1999 29 Fig. 3. Bestwhi ccmmiiiuf Pulawski and Rasnitsyn, new species, paratype PIN, 3559/4528: a — antennal fora- men; al — adlateral line; h — hypostome; n — notaulus; otherwise as in Figs. 1, 2. former because of its dark wing veins; Fig. 6). Bestiola tenttipes Pulawski et Rasnitsyn, new species (Fig. 7) Name derivation. — Tenuipes, from the Latin words tenuis, thin, and pes, leg; with reference to the elongate forefemur. Recognition. — The wing venation of B. tenuipes is as in communis, but the head, wing veins, and legs are light rather than Fig. 4. Bestiola communis Pulawski and Rasnitsyn, new species, paratype, PIN, 3559/4530. dark. Also, the forefemur is slightly longer (compare Figs. 7 with 2 and 3), although this difference is difficult to quantify. See Recognition of cotnmunis for differences with hispanica and subpetiolata. Description. — Male (Fig. 7). Female un- known. Antenna and mesosoma (possibly in part) dark, metasoma infuscated toward apex, otherwise coloration light (including wing venation). Thorax with well-defined although shallow punctures that are about 1 diameter apart, and with moderately short and moderately dense setae. Basal flagellomeres unrecognizable, remaining flagellomeres 2.0-2.5 times as long as wide, gradually becoming shorter and thinner to- ward apex. Adlateral line of mesoscutum complete; scutellum wide, elongate; metas- cutellum contrastingly short; metapostno- tum trapezoid, with basal impression, me- dian longitudinal line, and rounded pos- terior angles. Propodeal spiracle elongate, narrow, slightly bent S-like. Wing venation 30 Journal of Hymenoptera Research 5mm Fig. 5. Bcsticla ? communis Pulawski and Rasnitsyn, new species, PIN, 3559/4526; N3 — metanotum; pp- propodeum; scl — scutelluni; other abbreviations as in Figs. 1-3. as in communis. Legs relatively long, fore- femur slightly longer than head width, with dorsal margin straight and ventral margin convex; hindfemur elongate, al- most symmetrical, attenuated apically, both dorsal and ventral margins straight (except basally); midtarsus markedly lon- ger than midfemur; midbasitarsus shorter Fig. 6. Bcstiola ? connminis Pulawski and Rasnitsyn, new species, PIN, 2385/2392. than three following tarsomeres combined. Genitalia elongate, ovoid, with smooth contour, with gonostyle apex narrow rounded. Body length 21 mm as preserved, forewing length 11.5 mm. Material examined. — Holotype: Russia: Eastern Siberia: Baissa, bed 31 (PIN, 3064/ 2055, incompletely preserved male). Bestiola subpetiolata Pulawski et Rasnitsyn, new species (Fig. 8) Name derivation. — From the Latin petiol- us, little foot, stalk, stem; and the prefix sub-, Latin for under, somewhat, less than; with reference to the shape of forewing cell 2rm. Recognition. — The following details of Volume 8, Number 1, 1999 31 5 mm Fig. 7. Bestioln teiiuifvs Pulawski and Rasnitsyn, new species, holotype: pN3— metapostnoUim; sp— propodeal spiracle; other synibols as in Figs. 1-5. 32 Journal of Hymenoitera Research 5 mm Fig. 8. B('sf/o/(7 fiihpctiolttta Pulawski and Rasnitsyn, new species, luilotype: es — episternal suture; nipl — me- tapleura; other abbreviations as in Figs. 1-7. the forewing venation distinguish suhpe- tiolata from all of its congeners: vein 2r-rs meeting pterostigma near the latter's mid- length, costal margin of cell 2rm markedly shorter than 2r-rs, and anterior end of vein cu-a closer to wing base than M + Cu fork. In addition, subapical flagellomeres ap- pear to be only slightly longer than wide, thus markedly shorter than in other Bes- tiola. Description. — Sex unknown (Fig. 8). Body moderately dark (including wing veins) but antennal apex, tarsi (except hindbasitarsus basally), and metasoma Volume 8, Number 1, 1999 33 light. Metasoma at least partly setose (se- tae preserved only along hind margin of last preserved tergum). Length of basal flagellomeres about 3 times width, sub- apical ones probably subquadrate. Eye large, elongate, almost symmetrical. Malar space probably short. Pronotum short, spi- racular lobe not elongate. Mesoscutum: notauli and adlateral lines complete or nearly so. Mesopleuron with complete, crenulate episternal sulcus; and with an- teriorly crenulate hypersternaulus. Meta- pleuron wide, crossed by subhorizontal sulcus. Other thoracic structures unrecog- nizable due to deformation. Forewing vein 2r-rs meeting pterostigma near the latter's midlength, meeting RS near 2rm (longer than costal margin of cell 2rm), 2rm weakly arching, 3r-m straight, ante- rior end of cu-a slightly closer to wing base than M-l-Cu fork. Hindwing vein cu- a meeting M relatively close to M-I-Cu fork. Midfemur: dorsal margin straight, ventral margin convex. Hindfemur about as long as head height, widest subbasally, attenuated apically, with dorsal margin concave in apical half and ventral margin straight except basally. Midtibia slightly shorter, hindtibia slightly longer, than re- spective femur. Mid- and hindtarsi longer than respective tibiae, respective basitarsi slightly shorter than following 3 tarso- meres combined. Metasoma somewhat at- tenuated basally, with apex missing, but probably shorter than head and thorax combined. Body length, as preserved, 10 mm, forewing length 7.0 mm Material examined. — Holotype: Central Mongolia, Bon-Tsagan, bed 87/8 (PIN, 3559/4529, incomplete, rather poorly pre- served specimen with somewhat crum- pled thorax). ACKNOWLEDGMENTS We sincerely thank Michael A. Prentice for his con- structive criticism of the manuscript, as well as Eric Crissell and Arnold S. Menke. We are grateful to Robert L. Zuparko for reviewing an earlier draft of the manuscript. The project was partly supported by a grant to A. P. Rasrutsyn from the Royal Society Joint Project with the Former Soviet Union. LITERATURE CITED Ansorge, J. 1993. Bemerkenswerte Lebenspuren und 'fCrftosphcx catnliiniciis n. sp. (Insecta; Hymenop- tera) aus den unterkretazischen Plattenkalken del Sierra del Montsec (Provinz Lerida, NE-Spanien). Neues ]ahrhiicli fiir Ceologie und Paldontologie Moii- (itshefte 190: 19-35. Bohart, R. M., and A. S. Menke 1976. Sphecid wasps of the world. A generic revision. University of Cali- fornia Press, Berkeley, Los Angeles, London, 1 color pi., ix + 695 pp. Darling, D. Ch., and M. J. Sharkey. 1990. Order Hy- menoptera in Grimaldi D. A. (editor). Insects from the Santana Formation, Lower Cretaceous, of Brasil. Bulletin of the Ameriain Museum of Nnt- ural Histori/ 95: 124-129. Dmitriev, V. Yu, and V. V. Zherikhin. 1988. Changes in diversity of insect families as revealed by the method of accumulated appearances, p. 208-215 in A. G. Ponomarenko (editor). Cretaceous hiocen- otic crisis and evolutum of the insects. Nauka Press, Moscow. 230 pp. (in Russian). Evans, H. E. 1969. Three new Cretaceous wasps (Hy- menoptera). Psi/che 7b: 251-261. Gauld, I., and B. Bolton. 1988. The Hymenoptera. Brit- ish Museum (Natural History) and Oxford Uni- versity Press, Oxford, 332 pp. Hong, Y.-c. 1984 New fossil insects of Liayang Group from Laiyang Basin, Shandong Province. Profes- sional Papers of Stratigraplni and Palaeontology No. 11 1984: 31-41. Jarzembowski, E. A. 1991. New insects from the Weald Clay of the Weald. Proceedings of the Ge- ologists' Association 102: 93-108. Martinez-Delclos, X. (editor). 1995. Montsec and Mont- Rat-Alcover, two Konsen'at-Lagerstdtten. Institut d'Estudis Illerdencs, Catalonia, Spain. 97 pp. Ponomarenko, A. G. 1990. Insects and the Lower Cre- taceous stratigraphy of Mongolia, p. 103-108 in Krassilov V. A. (editor). Non-marine Cretaceous of the USSR. The Submission from the Conference of the Soviet Working Group of the IGCP Project 245, Vladivostok, 19SS. Far Eastern Branch of the USSR Academy of Sciences, Vladivostok. 226 pp. [In Russian]. Rasnitsyn, A. P. 1975. Hymenoptera Apocrita of Me- sozoic. Academy of Sciences of the USSR. Transac- tions of the Paleontological Institute. 147: 134 pp. [In Russian]. Rasnitsyn, A. P. 1980. Origin and ei'otution ofhymenop- terous insect.:. Nauka, Moskva, 189 + two un- numbered pp. (in Russian[ Rasnitsyn, A. P. 1986. Vespida (= Hymenoptera) in Rasnitsyn A. P. (editor). In.sects in the Early Cre- taceous Ecosystems of the West Mongolia. Trans- 34 Journal of Hymenoptera Research actions of tlw loint Soviet-Mongolian Paleontolcgical Expedition. No. 28: 154-164 (in Russian). Rasnitsyn, A. P. 1988. An outline of evolution of the hymenopterous insects (order Vespida). Oriental Insects 22: 115-145. Rasnitsyn, A. P. 1990. Hymenoptera in Ponomarenko A. G. (editor). Late Mesozoic insects of Eastern Transbaikalian. Academy of Sciences of the USSR. Transactions of the Paleontolcgical Institute. 239. Nauka Press, Moscow: 177-205. [In Russian]. Rasnitsyn, A. P., E. A. Jarzembowski, and A. J. Ross. 1998. Wasps (Insecta: Vespida = Hymenoptera) from the Purbeck and Wealden Supergroups (Lower Cretaceous) of Southern England and their hiostratigraphical and paleoenvironmental significance. Cretaceous Research 19: 329-391. Ren, D., L. Lu, Z. Guo, and S. Ji. 1995. Faunae and stratigraphy of Jurassic-Cretaceous in Beijing and the adjacent areas. Seismic Publishing House, Beijing. 222 pp. Richards, O. W. 1956. Handbooks to the identificatwn of British Insects. H[inienoptera. Introduction and keys to fanuties. Royal Entomological Society of Lon- don, London. 94 pp. Sinitza, S. M. 1993. Jurassic and Lower Cretaceous of the Central Mongolia. Transactions of the Joint Sox'iet- Mongoltau Palcontological E.\pcdition. No. 42: 1-239 pp (in Russian). Zhang, Jun-feng 1985. New data on the Mesozoic fos- sil insects from Laiyang in Shandong. Geology of Shandong 1: 23-39. Zhang, Jun-feng 1992. Descriptions of two new gen- era and two new species of Baissodidae from China (Sphecidae, Hymenoptera). Acta Entonio- logica Sinica 35: 483-489. Zherikhin, V. V. 1978. Development and changes of the Cretaceous and Cenozoic faunistic assem- blages (Tracheata and Chelicerata). Academy of Sciences of the USSR. Transactions of the Palconto- logical Institute 165: 1-200. J. HYM. RES. Vol. 8(1), 1999, pp. 35-47 Ultrastructure of Imaginal Spermatozoa of Sawflies (Hymenoptera: Symphyta) Terence M. Newman and Donald L. J. Quicke Department of Biology, Imperial College of Science, Technology and Medicine, Silwood Park, Ascot, Berkshire SL5 7PY U.K. Abstract. — We present the first ultrastructural study of sperm from representatives of three superfamilies of sawflies (Hymenoptera: Symphyta): Xyela julii (Xyeloidea), Cephalcia arvensis (Pamphiloidea) and Tremex sp. (Siricoidea), with particular attention being paid to characters that may be phylogenetically informative. Differences in the location of the centriolar adjunct, partic- ularly in relation to the mitochondrial derivatives, would suggest Cephakia has a better claim than Xyela as having sperm that may be representative of a common ancestral form. The centriolar adjunct of Cephalcia overlies both mitochondrial derivatives symmetrically, as found in ants and bees, whereas in Tremex, which its sperm otherwise closely resemble, the centriolar adjunct is located asymmetrically, abutting a single mitochondrial derivative and thus offsetting the pair of mitochondrial derivatives longitudinally. Xyela has radically different sperm in terms of size and both the arrangement and appearance of the organelles, especially acrosomal substructure and lack of an acrosomal rod. Very little is known about sperm ultra- structure among the Hymenoptera com- pared with most other insect orders (Phil- lips 1970; Jamieson 1987; Quicke 1997), and most of the studies that do exist deal largely with common aculeates such as bees and ants (e.g. Dallai and Afzelius. 1990; Wheeler and Krutzsch. 1992). How- ever, an initial study of the spermatozoa of some species (Quicke et al. 1992) re- vealed a considerable number of ultra- structural features that differ between taxa, raising the possibility that such var- iation might provide new phylogenetic in- dicators, as has been possible in many oth- er groups of insects (Jamieson 1987). The phytophagous sawflies (Symphyta) consti- tute a relatively underived basal grade within the order of Hymenoptera. As such they are important for our understanding of the relationships and development of both the social species of the Aculeata (ants, wasps and bees) and members of the paraphyletic group of the ten or eleven currently recognised, extant superfamilies generally referred to as the 'Parasitica'. This is especially so, since the sister group for the Hymenoptera is not at all certain at present (Whiting ei al. 1997), and so it is not possible to make use of outgroup comparison to determine the ancestral sperm morphology of the order (Watrous and Wheeler 1981). Groundplan sperm ul- trastructure may therefore be determined best by considering the sperm of those ex- tant taxa (i.e. the sawflies) which represent the most basal hymenopteran lineages (Gibson 1993; Yeates 1995). The only pre- vious work on sawfly sperm ultrastruc- ture (Quicke et al. 1992) presented data for only two of the six symphytan superfam- ilies, the Tenthredinoidea and the Cephoi- dea. We have therefore examined sperm ultrastructure, and in particular that of cell organelles, in detail in representatives of three further superfamilies, the Xyeloidea, Pamphiloidea and Siricoidea, leaving only the rare, through interesting, Orussoidea unstudied. Two of the superfamilies ex- amined here, the Xyeloidea represented 36 Journal of Hymenoptera Research by Xyela julii (Brebisson) and the Pamphi- loidea, represented by the pamphiliid, Ce- phalcia arvensis Panzer have usually been considered to be among the most primi- tive of sawflies. In contrast, the Siricoidea represented by Tremex sp., are close to the origin of the Apocrita (Rasnitsyn 1980, 1988; Heraty et al. 1994; Vilhelmsen 1997). The results are discussed in terms of the likely plesiomorphic states for various subcellular features in the Hymenoptera. MATERIALS AND METHODS Testes were obtained from adult males of Xyela julii, Ceyhalcia arvensis and Tremex sp., which had been maintained on dilute honey solution for a maximum of 3 days. Xi/ela were collected as adults in Silwood Park, Berkshire, U.K., the Cephalcia were reared from larvae collected in Italy and the Tremex were collected as adults in Cal- ifornia and couriered to the U.K. for prep- aration. Light microscopy. — Vas deferenda and testes were dissected from living sawflies in insect saline and teased apart on a clean microscope slide. After a few minutes to allow the sperm /spermatodesmata to swim free of the disrupted tissue, the slides were dried on an hot plate at c. 80°C. The smear was then flooded with double-filtered, 0.1% w/w toluidine blue in 1% w/w aqueous sodium borate and stained at 80°C until crystallisation of the stain had started. Following washing in distilled water they were permanently stored dry. Transmission electron microscopy. — Geni- talia were dissected out under 2% glutar- aldehyde in phosphate buffered saline (pH 7.2), and fixed for two hours. Tissue was transferred to 2% osmium tetroxide in cacodylate buffer (pH 7.2) for 2 hr. After another buffer wash, tissue pieces were dehydrated to 50% ethanol and then fur- ther fixed with saturated uranyl acetate in 50% ethanol prior to complete dehydra- tion, embedding in Epon resin and poly- merisation overnight. Silver sections were picked-up on to high resolution grids, stained with uranyl acetate and lead. RESULTS Woodwasps of the superfamily Siricoi- dea are considered to be amongst the most advanced of the sawflies, sharing a num- ber of derived morphological features with the Apocrita (Vilhelmsen 1997). As has been reported previously for other sawflies (Quicke et al. 1992), the mature sperm of siricid, Tremex, stored within the vas deferens and seminal vesicles are present in spermatodesmata bundles (Fig. 1 ), though by the time they reach the sper- matheca of the females they have broken up completely and only isolated sperm are present (Naito, personal communication). In our preparation of Tremex from male seminal vesicles, a small proportion of iso- lated sperm were also present but it is not clear whether they were the result of sper- matodesmata fragmentation upon fixation or whether they indicate a normal pre- transfer phenomenon. The sperm heads are inserted through- out the fairly electron-dense and elongate cap of the spermatodesmata, with those sperm located more centrally being insert- ed more anteriorly (Figs 1, 2). As a result, many different levels of sperm are evident in a single transverse section of each sper- matodesmata (Fig. 2). It is therefore pos- sible, in the same transverse section, to lo- cate adjacent sperm sectioned through ac- rosome and acrosomal rod (perforatori- um), through the nucleus, the basal body with centriolar adjunct, and through the axoneme with mitochondrial derivatives. Also, in transverse section, the acrosome is clearly seen to have a membrane around both the outside and around the invagi- nated portion of the structure (Fig. 4c). Be- tween the acrosomal membrane and the plasma menibrane is an electron dense re- gion extending from the acrosomal mem- brane (Fig. 4c, arrou'lieads). This may be comparable to the material reported to surround the acrosome in other sperma- Volume 8, Number 1, 1999 37 Fig. 1. Longitudinal section of a spermatodesmatum in the imaginal testes of the sawfly, Tremex (Siricoidea). A, acrosome; C, centriolar adjunct; M, mitochondrial derivative; N, nucleus; R, acrosomal rod; X, axoneme. Scale bar = 1.0 nm. 38 Journal of Hymenoptera Research Figs. 2-3. Features of spermatodesmata and sperm in the imaginal testes of the sawtlv, Tivincx (Siricoidea). 2, transverse sections through several spermatodesmata at different levels showing that the more centrallv located spermatozoa have their heads inserted more anteriorly; 3, nuclear-associated organelles showing in 3a, the insertion of the acrosomal rod (R) into the anterior of the nucleus (note also the small anterior sac at the head of the acrcisome), and 3b, the position of the centriolar adjunct (arrowed) in relation to the nucleus and striated mitochondrial tierivative. Scale bars: 2 = 2.0 p.m; 3 = 0.5 [im. Volume 8, Number 1, 1999 39 Fig. 4. Organelles of Tremex sperm (Siricoidea) seen in tranverse and oblique section: 4a, showing the cen- triolar adjuncts (small arrows) of a number of spermatozoa, and in the section indicated by the large arrow the centriolar adjunct can be seen to make contact with both mitochondrial derivatives; 4b, section at the level of the centriolar adjunct (arrowed) where it occupies all the extra-axonemal area; 4c, sections through anterior of nucleus and acrosome showing the acrosomal rod (r) fitting tightly into the nucleus but loosely in the sub- acrosomal space (note: the clear membranes surrounding acrosome and nucleus but not the rod; granular material between acrosomal and plasma membranes (arrowheads); putative nuclear 'pore' (large arrow)); 4d, axoneme with small deltoid bodies (small arrows) and central rod (large arrows) (note also mitochondrial derivatives with distinct membrane and internal structure). Scale bar: 4a = 0.66 \Lm; 4b,c = 0.25 ^m; 4d = 0.2 (Jim. 40 Journal of Hymenoitera Research Volume 8, Number 1, 1999 41 tozoa (Quicke et al. 1992), although the lat- ter structures are larger and have a sub- layered appearance in at least some taxa. In longitudinal section (Fig. 3), the ac- rosome of Tremex can be seen to have a large sub-acrosomal space (Fig. 3a) which is partly occupied by the acrosomal rod. This rod extends into the nucleus for al- most the same length again as it does into the acrosome. Unlike the nucleus and ac- rosome, the rod is not membrane bound, but where the rod is inserted into the nu- cleus there is no surrounding space, giv- ing the impression that the rod is being held by the nucleus. The plasma mem- brane surrounding the acrosome extends slightly anteriorly to produce a small ex- tra-acrosomal space. A membrane bound area found within the nucleus (Fig. 4c, ar- row) may represent a longitudinally run- ning pore. The nucleus (Fig. 3b) is abutted posteriorly by the axoneme at the level of the latter's basal body, where the axoneme lacks the central pair of microtubules. A large centriolar adjunct is present, and this in turn contacts the mitochondrial deriv- atives which have very clearly defined membrane bound cristae (Fig. 3b, arrow). There is at least one membrane separating the centriolar adjunct from the nuclear membrane. The exact arrangement of the centriolar adjunct and nucleus, and in par- ticular, how the centriolar adjunct contacts the nucleus, is not always obvious. In many insect spermatozoa this has given rise to confused interpretations of the structure, even to suggestions that the cen- triolar adjunct is absent. From the present study, the relationship becomes clear at higher magnification where several sper- matozoa lie in close proximity (Fig. 4). The centriolar adjunct (Fig. 4a, arrows) contacts two equally sized mitochondrial deriva- tives, and can be seen to extend some way in between the two, forming what appear in transverse sections, tail-like structures (Fig. 4a, large arrow). The mitochondrial derivatives are thus intimately connected to the centriolar adjunct. In transverse sec- tion, certainly at the level of the basal body, this produces an arrangement where the centriolar adjunct and basal body occupy most of the area of section; an appearance which could be mistakenly interpreted as there being an extension of the nucleus overlying the axoneme (Fig. 4b). Although the acrosome is smaller in di- ameter than the nucleus, there is not a great discrepancy. The axoneme itself has the 9-1-9+2 arrangement (Fig. 4d) common to Hymenoptera; 9 outer single accessory tubules, 9 doublets and 2 central single microtubules. Intertubular material is abundant with radial spokes (Afzelius rays) and indications of the inner and out- er dynein arms. Two deltoid bodies, (also referred to as triangular rods, (Lensky et al. 1979) are present, but they are not large. Between the two bodies, and the two mitochondrial derivatives (Fig. 4d) is a single central rod, as previously report- Fig. 5. Cell ultrastructure of Cqilmlcia sperm (Pamphiloidea): 5a, sperniatadesmata with surrounding cap material (arrowed); 5b, centriolar adjunct (small arrow) abuts nucleus at membranous complex, extending beyond the level of the basal body (arrowhead) to abut a mitochondrial derivative (large arrow); 5c, showing that the centriolar adjunct does not overly both mitochondrial derivatives with one of the two derivatives (arrowed) abutting the nucleus at the region of the membranous complex; 5d, transverse section at level of basal body showing that the centriolar adjunct extends to partially enclose the parallel mitochondrial deriv- ative (arrowed); 5e, transverse section through axoneme showing only a single mitochondrial derivative pos- teriorly near the tail piece; 5f, transverse section through midregion of axoneme with two mitochondrial derivatives (note distinct sub-structure at periphery of each mitochondrial derivative); 5g, transverse sections showing acrosomal rod (r) insertions into the nucleus; 5h, transverse sections showing insertion of acrosomal rod (r) into the acrosome. Scale bar: a = 1.3 (xm; b = 0.5 |xm; c, d = 0.42 (xm; e = 0.3 (xm; f = 0.27 (xm; g = 0.7 |xm; h = 0.57 (xm. 42 Journal of Hymenoptera Research Fig. 6. bpLTiiijliitodcsmaUi ol Xi/cUi (Xyeloidea): 6a, showing the extended length o( Uic spcinialodL'.siiuU.il cap; 6b, showing the distinctive arrangement of the acrosome at the region of insertion into the cap with particulate material (small arrows), distinct periodicity in the core material (arrowheads; appears like longi- tudinal striations), and multilayered membrane coat to the acrosome (large arrows). Scale bar: a = 2.5 jjim; b = 0.66 |xm. Volume 8, Number 1, 1999 43 ed in ant spermatozoa (Wheeler et al. 1990). A tail piece is found where the ax- oneme has no associated mitochondrial derivatives. Of the more basally derived sawfly su- perfamilies investigated, the Pamphilo- idea (Cephalcia) represents a slightly more advanced evolutionary lineage than the Xyeloidea although it was once included in the same family. The spermatozoa of Cephalcia are arranged in spermatodes- mata (Fig. 5a) and have heads (nucleus plus acrosome) approximately 28fJLm long, and tail, 75|xm long. Ultrastructurally, they are very similar to those of Tremex, especially in terms of the size of the ac- rosomal rod, its position within the sub- acrosomal space (Fig. 5h), and its insertion into the nucleus (Fig. 5g). The most no- ticeable difference between the two is in the position of the centriolar adjunct. In Cephalcia the centriolar adjianct can be seen to run parallel to one to the pair of mito- chondrial derivatives (Fig. 5b), rather than overlying both as in Tremex, as is evident in the region where one of the mitochon- drial derivatives is found to abut the nu- cleus (Fig. 5c, arroic). For part of its length (at the level of the basal body) the centrio- lar adjunct contacts and even partially en- closes the single mitochondrial derivative that lies parallel to it (Fig. 5d, large arrow). Possibly as a result of this arrangement, a region occurs at the posterior part of the sperm, where there is only a single mito- chondrial derivative lying next to the ax- oneme (Fig 5e); here there is also only a single deltoid body, as opposed to the two found in normal section (Fig. 5f). Similarly at the level of the centriolar adjunct the single mitochondrial derivative has only a single deltoid body. The axoneme is again similar to that of Tremex in the arrange- ment of elements. The ray material is par- ticularly evident, with Afzelius rays hav- ing distinct spoke heads, and with distinct electron-opaque granules between the pe- ripheral singlets (Bairati and Baccetti 1965). Xyela spermatozoa, although present in spermatodesmata (Fig. 6), differ in a num- ber of ways from those of both other saw- flies described here as well as from the tenthredinoids described by Quicke et al. (1992). The spermatozoa are extremely long with the head ( = nucleus plus acro- some) being approximately 60^,m long, and the tail 150 [xm long (Fig. 6a). This elongation compared with sperm of other sawflies, at all levels, viz. the acrosome, nucleus and tail, is also apparent in lon- gitudinal section. At the anterior end the acrosomes can be seen to be asymmetrical and pointed, containing two types of ma- terial: an irregularly granular material and a core material that has an almost crystal- line periodicity, aligned parallel to the long axis of the spermatozoa. In trans- verse section (Fig. 7) the most prominent of the features is an enlargement of one of the mitochondrial derivatives to a diame- ter greater than that of the axoneme (Fig. 7a), with a concomitant enlargement of that mitochondrial derivative's deltoid body (Fig. 7a, right arrowhead). This dis- places the other mitochondrial derivative which, together with its deltoid body, now occupies an area approximately equiva- lent to the other, larger, deltoid body alone. In Xyela, it is not immediately apparent if there is a centriolar adjunct. In some transverse sections, at the position of the smaller mitochondrial derivative /deltoid body a darker structure is present (Fig. 7b, arrows). This does not seem to be simply a denser mitochondrial derivative because it lacks a deltoid body and generally the two occur together (in shape it is actually closer to a deltoid body). In longitudinal section (Fig. 7c) a structure abutting on to the smaller mitochondrial derivative can be found. This closely resembles the situ- ation in Cephalcia. The structure does not however extend to the nucleus like the other centriolar adjuncts found. Instead, at the region of the basal body, identifiable by the absence of the central pair of mi- 44 Journal of Hymenoptera Research Fig. 7. Organelles of Xycla spermatozoa (Xyeloidea): 7a, asymmetric mitochondrial derivatives (arrows) and well-devloped deltoid bodies (arrovi'heads); 7b, in one region the smaller mitochondrial derivative is replaced by a darker structure similar to a centriolar adjunct (arrows) (note also the numerous, smaller nuclear-like cross sections, smaller than other readily identified nuclei); 7c, longitudinal section showing a centriolar ad- junct-like body (arrow) that abuts the smaller mitochondrial derivative; 7d, transverse section at the level of the basal body (arrows) showing that the centriolar adjunct-like organelle and the smaller of the two mito- chondrial derivatives are both absent, and that the larger mitochondrial derivative (m) partiallv co\ers the basal body; 7e, showing asymmetric insertion of a cone of acrosomal (a) material into ihe nucleus isolating a Volume 8, Number 1, 1999 45 crotubules, the larger mitochondrial deriv- ative wraps round to partially enclose the basal body and the smaller mitochondrial derivative /deltoid body is absent (Fig. 7d). If there is a centriolar adjunct, then how it terminates anteriorly and its rela- tionship with the nucleus remains unclear. The nucleus itself appears similar to those of the other sawflies in density and mem- brane organisation at the level of the sper- matodesmata. However, perhaps in ac- cord with its greater length, there appears to be an area, posterior to its insertion into the cap of the spermatodesmata, that has a relatively smaller diameter and where it is significantly smaller than the tail region with its enlarged mitochondrial derivative (Fig. 7b, d). In Xyela the interface of the acrosome with the nucleus also appears different. There is no discernible rod. Instead acro- somes, which have a distinct, paracrysta- line substructure, contact the nucleus and may even be partially enclosed by it (Fig. 7e). This insertion is displaced to one side, and this asymmetry is also present in the acrosome itself (Fig. 7f). A ridge runs down one side of the acrosome (Fig. 7f, iirwzv). Spermatozoa are orientated within the spermatodesmata so that the ridges all point in the same direction. Interestingly, this is also the same side of the sperma- tozoon that the acrosome inserts into the anterior of the nucleus, although at this point the acrosomal material appears to have lost the ridge, and the acrosome at this level only shows the core of 'periodic material'. The ridge itself contains the par- ticulate matter. In some areas the granules surround membranes resembling the mul- tilayered coated complex that surrounds the acrosome itself. There is also an exten- sion of the outer layers of this coat to form a small ridge to one side of the acrosome. This position of this smaller ridge is again consistent amongst the spermatozoa. DISCUSSION At least with regard to the ultrastruc- ture of the spermatozoa, Cephalcia appears to have a better claim than Xyela as having sperm that may be representative of a common ancestral form. Cephalcia sperm are very similar to those of Tremex. It is mainly in the positioning of the centriolar adjunct that Cephalcia varies from Tremex, having an asymmetric location overlying only one mitochondrial derivative. Tremex by comparison has a centriolar adjunct overlying both mitochondrial derivatives. The arrangement of the centriolar adjunct might seem to have possible usage as a phylogenetic indicator. Unfortunately the arrangement of this organelle has often been poorly understood (e.g. Wilkes and Lee 1965), and so it is difficult to draw any conclusion from all previously reported works. From studies of bee sperm Jamie- son (1987) concluded that in bees the cen- triolar adjunct also lies between the nucle- us and one of the mitochondrial deriva- tives. Recently we have described the ul- trastructure of the parasitic braconid wasp, Aleiodes, which appears to have rel- atively underived sperm (Quicke et al. 1992; Newman and Quicke 1998). The ul- trastructure of individual Aleiodes sper- matozoa closely resembles both Tremex and Cephalcia. The mitochondrial deriva- tives are similarly sized and the acrosomal rod is similarly positioned. The centriolar adjunct is, however, asymmetric and hence similar to Cephalcia. Given the prim- itive status proposed for Cephalcia this might be considered to be the archetype arrangement retained through evolution. small amount of membrane bound nuclear material (arrowed); 7(, showing the periodic appearance of the acrosome within the multilayered coat and membrane material (arrowed). Scale bar: a = 0.27 jjim; b, c = 0.5 Jim; d = 0.6 jjim; e = 0.5 |j.m; f = 0.25 p.m. 46 Journal of Hymenoptera Research The case of Tremex, with symmetric cen- triolar adjunct overlying both mitochon- drial derivatives, would have to be con- sidered as an apomorphic development. However, it is not clear at present which of the arrangements of the centriolar ad- junct represents the groundplan for either the Hymenoptera as a whole or for any of the major lineages within it, and more careful study is necessary. As was made clear in the results, the po- sitioning of the centriolar adjunct is not clear in Xyela. It appears to be asymmetric, but unlike the arrangement in other saw- flies it does not appear to abut the nucle- us. Xyela has sperm with a structure that is extremely divergent in a number of oth- er ways; the shape and arrangement of the acrosome, the apparent absence of the ac- rosomal rod and the size difference in the two mitochondrial derivatives. It seems likely that these must represent a response to selective pressures subsequent to the di- vergence of both the other sawflies and the main body of the order of Hymenop- tera. The presence of an asymmetric centrio- lar adjunct in Cqjhalcia appears to cause the mitochondrial derivatives to be offset longitudinally, and this may explain why some sections through the posterior part of the spermatozoa have only a single mi- tochondrial derivative (e.g. Fig. 5e). Where there is only one mitochondrial derivative the deltoid body is also absent suggesting they may be a good marker for mitochon- drial derivative identification. From the present observations, it seems probable that the identification of the coat material surrounding the acrosome in many taxa may be incorrect. In the Hy- menoptera, this material has been referred to as extracellular matrix (Quicke et al. 1992) and in some insect orders (e.g. Or- thoptera) it has been reported that extra- cellular matrix granules accumulate around the plasma membrane of the ac- rosome to form an extracellular cap (Szol- losi 1974). However, these structures are often highly complex, with layered or re- peated substructure (see for example. Fig. 6b in Quicke et al. 1992), and it is not im- mediately clear how such a structure could be secreted if extracellular; the pos- sibility that they are produced by the ep- ithelia of a deferent duct cannot be ex- cluded. Many plasma membranes possess a glycocalyx which comprises the carbo- hydrate portion of integral membrane gly- coproteins and glycolipids together with associated glycosamminoglycans and pro- teoglycans, and these carbohydrates ex- tend from the plasma membrane into the extracellular space. Where organelle mem- branes become glycosilated, as in the case of secretory granules that will eventually fuse with a plasma membrane, the coated face of the membrane that opposes the in- terior of the organelle is the one that will, upon fusion, face the extracellular space. In the sawflies investigated here, it is clear that the coat lies between two membranes, and is not extracellular as previously re- ported. An intracellular origin for this structure would at least allow a more con- ventional, although as yet, completely un- recognised, mechanism for its production. ACKNOWLEDGMENTS The authors gratefully acknowledge the cheerful and patient assistance of Ian Morris (EM Unit, De- partment of Biology). A. Battisti kindly sent us the live Ctyhalcin, and Paul Johnson, the Tremex. This re- search was supported by the NERC (Natural Envi- ronment Research Council) Initiative in Taxonomy. LITERATURE CITED Bairati, A. Jr. and B. Baccetti. 1963. Indagini compar- ative sull'ultrastruttura delle cellule germinali maschili in Dacuf oleae Gmel. ed in Dm^cpliihi me- laiwgn^ter Meig. (Ins. Diptera). 11. Nuovi reperti ultrastrutturali sul filamento assile degli sper- matozoi. Redia 49: 81-85. Dallai, R. and B. A. Afzelius. 1990. Microtubule di- versity in insect spermatozoa: results obtained with a new fixative. Icurmil of StructKni! Biolcgti 103: 164-179. Gibson, G. A. P. 1993. Groundplan structure and ho- mology of the pleuron in H\nienoptera based on a comparison of the skeletonuisculaturo of \ve- lidae (Hymenoptera) and raphiliidae (Neurop- Volume 8, Number 1, 1999 47 tera). Memoirs of the Entomological Society of Can- ada 165: 165-187. Heraty, J. M., J. B. Woolley and D. C. Darling. 1994. Phylogenetic implications of the mesofurca and mesopostnotum in Hymenoptera. Journal of Hy- menoptcra Research 3: 241-277. Jamieson, B. G. M. 1987. The Ultrastructure ami Pliy- logeny of Insect Spermatozoa. Cambridge Univer- sity Press, Cambridge, 320pp. Lensky, Y., E. Ben-David and H. Schindler. 1979. Ul- trastructure of the spermatozoan of the mature drone honeybee, journal of Apiculture Research 18: 264-271, Newman, T. N. and D. L. J. Quicke. 1998. Sperm de- velopment in the imaginal testes of Aleiodes cox- alis (Hymenoptera: Braconidae: Rogadinae). Jour- nal of Hymenoptera Research Phillips, D. M. 1970. hisect sperm: their structure and morphogenesis. Journal of Cell Biology 44: 243- 277. Quicke, D. L. J. 1997. Parasitic Wasps. Chapman & Hall, London, 470pp. Quicke, D. L. J., S. N. Ingram, H. S. Baillie and P. V. Gaitens. 1992. Sperm structure and ultrastructure in the Hymenoptera (Insecta). Zoologica Scripta 21: 381-102. Rasnitsyn, A. P. 1980. Origin and evolution of Hy- menoptera. Trudy Paleontolagicheskogo Instituta 174: 1-190. Rasnitsyn, A. P. 1988. An outline of evolution of the hymenopterous insects (Order Vespida). Oriental Insects 22: 115-145. Szollosi, A. 1974. Ultrastructural study of the sper- matodesm of Locusta migratoria migratorouies (R.F): acrosome and cap formation. Acrida 3: 175- 192. Vilhelmsen, L. 1997. The phytogeny of lower Hyme- noptera (Insecta), with a summary of the early evolutionary history of the order, lournat of Zoo- logical Systematics and Evolutionary Research 35: 49-70. Watrous L. E. and Q. D. Wheeler. 1981. The outgroup comparison method of character analysis. Sys- tematic Zoology 30: 1-11. Wheeler, D. E., E. G. Crichton and P. H. Krutzsch. 1990. Comparative ultrastructure of ant sperma- tozoa (Formicidae: Hymenoptera). Journal of Mor- pholog}/ 206: 343-350. Wheeler, D. E. and P. H. Krutzsch. 1992. Internal Re- productive system in adult males of the genus CampHvu^tus (hymenoptera: Formicidae: Formici- nae). Journal of Morphology 211: 307-317. Whiting, M. F., J. M., Carpenter, Q. D. Wheeler and W. C. Wheeler. 1997. The Strepsiptera problem: Phylogeny of the holometabolous insect orders inferred from 185 and 285 ribosomal DNA se- quences and morpholog\'. Systenuitic Biologi/ 46: 1-68. Wilkes, A. and P. E. Lee. 1965. The ultrastructure of dimorphic spermatozoa in the hymenopteron DahWomnius fuscipennts (Zett.) (Eulophidae). Gi- nadian Journal of Genetics and Cytology 7: 609-619. Yeates, D. K. 1995. Groundplans and exemplars: Paths to the tree of life. Cladistics 11: 343-357. J. HYM. RES. Vol. 8(1), 1999, pp. 48-64 A Review of the Old World Genus Fopius Wharton (Hymenoptera: Braconidae: Opiinae), with Description of Two New Species Reared from Fruit-infesting Tephritidae (Diptera) R. A. Wharton Department of Entomology, Texas A&M University, College Station, Texas 77843 Abstract. — Two new species of the Old World genus Fopius are described: ceratitivonis from Kenya and schlingeri from Queensland, Australia. Both species were reared from fruit-infesting Tephritidae; ceratitivorus from Ceratitis and schlingeri from Bactrocera. Details are provided on dif- ferentiation of the known species of Fopius, with discussion of their hosts, host specificity, and distribution. The parasitoids of fruit-infesting tephritids from Kenya are closely related to those from Madagascar. The most recent comprehensive classi- fication of the Opiinae is the three-volume monograph published by Fischer (1972, 1977, 1987). This collective work estab- lished a basis for more intense scrutiny of the Opiinae, resulting in several subse- quent modifications and additions to the classification, including the description of the Old World genus Fopius (Wharton 1987, van Achterberg and Maeto 1990). The numerous name changes affecting opiine parasitoids of fruit-infesting Te- phritidae were recently reviewed by Wharton (1997b), who also provided sug- gestions for delineation of species groups within the genus Fopius. Additional changes in nomenclature, some of these affecting opiine parasitoids of tephritids, were published by van Achterberg and Salvo (1997) and Quicke et al. (1997). Keys to most of the species of Fopius can be found in Wharton and Gilstrap (1983) and Fischer (1987), with both works treating the species under the generic name Bios- teres. Palacio et al. (1992) provide addi- tional information on separation of males and immatures of two sympatric species. All opiine braconids reared to date are koinobiont endoparasitoids of cyclorrha- phous Diptera, and all emerge from the puparium of their hosts. Hosts are known for one-third of the approximately 1500 described species, with most of these rec- ords pertaining to Agromyzidae and Te- phritidae. All reared species of Fopius are parasitoids of Tephritidae. Summaries of the literature on hosts and biology of the opiine parasitoids of Tephritidae can be found in Fischer (1972, 1977, 1987), Clau- sen (1978), Wharton and Marsh (1978), Wharton and Gilstrap (1983), Gilstrap and Hart (1987), Wharton (1989, 1997a, b). Messing (1996), Sivinski (1996), and Siv- inski et al. (1997). The primary purpose of the work pre- sented here is to facilitate on-going studies in biological control by providing names for two recently discovered, undescribed species. Both species are of interest with respect to tephritid biological control be- cause of their potential for attacking eggs or early instars, and one of these is a na- tive parasitoid of the Mediterranean fruit fly, Ceratitis capitata (Wiedemann). The use of parasitic Hymenoptera for the biologi- cal control of tephritid pests has received considerable attention in recent years (Knipling 1992, Waterhouse 1993, Head- Volume 8, Number 1, 1999 49 rick and Goeden 1996, Purcell, 1998), and there are active programs currently un- derway in several countries. MATERIALS AND METHODS With the exception of C. Granger's type material from the Paris Museum and a single, swept specimen of Fopiiis schliiigeri, n. sp., all material used in the descriptions of the new species was reared from fruit in association with various fruit-infesting Tephritidae. In some cases, parasitoids were reared from bulk fruit samples, with unconfirmed host associations. Most of the material, however, was reared from isolated puparia. In several of the rearings for Fopius cerntitivorus, n. sp., puparia were individually isolated prior to emer- gence. Though this procedure decreases the percent emergence (due primarily to desiccation and /or physical damage), it enables correct association of the wasp with the host from which it was reared. Specimens of the newly described spe- cies have been deposited in the following institutions: University of Queensland, Brisbane (UQBA), Australian National In- sect Collection, Canberra (ANIC), Texas A&M University, College Station (TAMU), Bernice P. Bishop Museum, Honolulu (BPBM), Hawaii Department of Agriculture, Honolulu (HDA), Queens- land Department of Primary Industries, Indooroopilly (QDPI), Nationaal Natu- urhistorisch Museum, Leiden (RMNH), The Natural History Museum, London (BMNH), National Museum of Kenya, Nairobi, International Centre of Insect Physiology and Ecology, Nairobi (ICIPE), and U.S. National Museum of Natural History, Washington, D. C. (USNM). Descriptive terminology follows Whar- ton (1987, 1988, 1997b) and Sharkey and Wharton (1997), and is based largely on the works of Fischer (1972). A tabular summary is presented rather than a di- chotomous key to facilitate assessment of relationships and point out gaps in our knowledge. IDENTIFICATION, RELATIONSHIPS, HOSTS, AND DISTRIBUTION PATTERNS The tephritid parasitoids in the genus Fopius are readily distinguished from oth- er opiines by the possession of crenulate notauli extending posteriorly to the me- sonotal midpit, an oblique ridge ventral- laterally on the propleuron, a short second submarginal cell (3RSa < 2RS), and a long ovipositor (1.3-4.5 X longer than meso- soma). Character states useful for identi- fying species with known tephritid host records are provided in Table 1, and the characters themselves are discussed be- low. Eight closely related species for which host records are lacking have also been included in the table. The table is deemed more informative than a dichot- omous key because it provides prelimi- nary data for phylogenetic analysis, which is beyond the scope and purpose of the present work, as well as providing sup- plemental characters for assistance in identification. More than half of the spe- cies in Table 1 (including several of the tephritid parasitoids) are known only from the type or the type plus a few other specimens. Additional collecting is essen- tial before progress can be made in our understanding of these species. Individual species and species groups are treated fur- ther following discussion of the charac- ters. Cliaracter 1. — Striate sculpture on the sec- ond and third metasomal terga. 0 = stri- ae absent; 1 = striae present on tergum 2 only; 2 = striae present on tergum 2 and at least base of tergum 3. For most species, assignment of either character state 0 or character state 1 is unambig- uous. Fopius deeralensis (Fullaway), how- ever, has weak striae on tergum 2, and the sculpture is not always readily vis- ible (see diagnosis following description of F. schliugeri, n. sp.). Fopius skiniwri (Fullaway), from the Philippines, is the only species in which striae are usually 50 Journal of Hymenoptera Research Table 1. Matrix of coded character states for species in the genus Fopntis (see text for character definition). 1 : T 4 - h - s - 111 11 i: p. 14 dcnticidifcr (van Achterberg & Maeto) 0 0 0 2 5 1 0 0 3 1 0 4 4.1-4.6 0 nwraiigensis (Fischer) 0 0 0 2 5 1 0 0 2b 1 7 4 1.7 7 taiwanicus (Fischer) 0 0 0 2 5 1 0 0 2b 1 7 4 -3.0 0 ruficornis (Granger) 0 1 0 1,2 3 1 0 0 1 2 0 2 1.5 0 ruhrithorax (Granger) 0 1 0 1 2 0 1 1 0 1 0 1.3 1 bei'isi (Brues) 0 0 1 1 0 1 0 0 1 0 2.2-2.3 1 dcsideratui (Bridwell) 0 0 1 1 0 1 0 0,1 1 0 3.0-3.2 1 uigcr (Szepligeti) 0 0 1 1 0 1 0 1 0,1 0 2.6-2.7 1 otfohvnoaiius (Fullaway) 0 0 1 I 0 1 0 1 1 0 2.8 1 rufotcftnccuf (Granger) 0 0 1 1 0 i 1 0 0,1 0 2.0-2.3 1 altcrnatae (Tobias) 0 0 0 0 0 7 0 3 1 7 2 -1.25 7 arisiinus (Sonan) 0 0 0 0 0 0 0 2 0,1 0 3 2.5-2.8 0 caiyomyiae (Silvestri) 0 0 0 0 0 0 0 2a 0 1 2 1.5-1.7 0 myolejae (Tobias) 0 0 0 0 0 7 0 3 1 7 2 -2.0 7 pcrsulcatus (Silvestri) 0 0 0 0 7 0 0 2a 0 7 3 7 7 skinneri (Fullaway) 0 0 0 0 0 0 0 2b 1 0 3? -2.65 0 vniuienhoschi (Fullaway) 0 0 0 0 0 0 0 2b 0,1 0 2 2.5-2.7 0 ceraiitivorus n. sp. 0 0 0 0 5 0 0 0 0 0 0,1 2 1.7-1.9 0 longicnuda (Granger) 0 0 0 0 0 0 0 0 1 0 0 2 2.6 1 pykucthcrax (Fischer) 0 0 0? 7 7 ■? 7 0,1 0 1 7 7 7 7 silivftrii (Wharton) 0 0 0 0 0 0 0 0 0 1 0 2 2.5-2.6 0 decralensis (Fullaway) 1 0 0 0,2 5 0 0 0 4 0 0 1 3 1 schlingeri n. sp. 0 0 0 0 5 0 0 0 4 0 0 3 2.3-2.5 1 caudatus (Szepligeti) 0 0 0 1 4 0 0 0,1 0 1 1 3 7 1 present on tergum 3, though the sculp- turing on tergum 3 is usually not exten- sive. Character 2. — Occipital carina. 0 = present laterally; 1 = completely absent. The oc- cipital carina is absent mid-dorsally in Fopius, but present laterally in nearly all species. There is also some variation among species in the height of the oc- cipital carina, as exemplified by the two species described below, and this vari- ation may eventually prove useful in demonstrating character state transfor- mations leading to complete loss of the occipital carina. Character 3. — Setal pattern on the oviposi- tor sheath. 0 = two or more rows of densely spaced setae; 1 = setae sparse, at most with a row of long, moderately sparse setae basally and short, widely spaced setae apically. This coding is useful for segregating groups of species, but oversimplifies the complexity of the character states that may eventually be useful for delineation of additional spe- cies. Setal rows are difficult to count, however, and many of the specimens examined were in such poor condition that it could not be determined if setae were sparsely arranged or merely bro- ken off. Of the species coded 0 in Table 1, setal density was greatest in F. detiti- culifer (van Achterberg and Maeto) and least in F. schlingeri, n. sp. and F. ruhri- thorax (Granger). Character 4. — Ventral margin of clypeus. 0 = thin, sharp, and evenly convex, with- out median projection; 1 = somewhat thickened medially, and slightly pro- truding, with labrum sometimes par- tially exposed; 2 = with median, ven- trally-directed, tooth-like (i. e. pointed) projection (clypeus completely occludes labrum). Differences between states 0 and 1 may not be apparent without dis- section to reveal the thickened margin. The tooth-like projection is very small in deeralensis, and the margin thinner than Volume 8, Number 1, 1999 51 in members of the F. marangensis (Fi- scher) species group, as reflected by its coding in Table 1. Character 5. — Pattern of sculpture and se- tae on frons. 0 = densely setose and punctate, the punctures tending to coa- lesce to some degree, giving the appear- ance of transversely rugulose lines (Fig. 4), midline longitudinally rugose; 1 = transversely striate and impunctate over middle half of frons, with deep, widely spaced punctures laterally; 2 = laterally as in state 1, but largely unsculptured medially (at most with a few, irregular, very weak wrinkles), midline with sharp carina basally; 3 = smooth, im- punctate, depressed along midline; 4 = broad, transverse band of deep punc- tures extending from ocelli to eye, oth- erwise smooth; 5 = densely setose and punctate, the punctures discrete, with no indication of rugosities as in state 0; for state 5, the punctures are very densely spaced in marangensis, F. taiivan- icus (Fischer) and denticulifer, less so in deeralensis and schlingeri, and least in ceratitivorus (where they are virtually absent basal-laterally). Character 6. — Postpectal carina. 0 = well developed; 1 = weak to absent. Character 7. — Relative length of first two flagellomeres. 0 = first flagellomere about same length as second (ratio vary- ing from 0.9-1.1); 1 = first flagellomere distinctly shorter than second (0.8 X length or less). Character 8. — Shape of petiole. 0 = petiole length equal to or shorter than apical width, strongly widening apically; 1 = petiole appearing more parallel-sided, with length distinctly greater (at least 1.3 times) than apical width. The petiole of f. caiidatus (Szepligeti) is somewhat intermediate, as reflected by its coding in Table 1. The petiole is not necessarily more parallel-sided in state 1 than in state 0 (width at apex may be twice width at base in both), but appears to be so because the petiole is longer in state 1. Character 9. — Geographic distribution. 0 = continental Africa; 1 = Madagascar; 2 = southern Asia (2a = India; 2b = south- east Asia, including Indonesia, Philip- pines, and Taiwan); 3 = Japan, eastern Russia; 4 = northeastern Australia. Dis- tribution patterns given here do not re- flect the successful introductions of F. arisaniis (Sonan) and F. vaiuienboschi (Fullaway) to Hawaii and arisanus to Central America. Character 10. — Color of mesosoma. 0 = largely pale (red, orange, yellow, or brownish-white); 1 = dark black to brown; 2 = pale with large black spots on mesoscutum and mesopleuron. As- sessment of coloration is somewhat problematic due to postmortem chang- es, especially in shades of red, yellow, and orange. Also, there is almost a com- plete continuum in shades of red from pale through nearly black (skinneri is dark reddish-brown). Two of the spe- cies for which there is abundant mate- rial (e. g. arisanus and vandenboschi) are color-variable. Character 11. — Dorsal carinae of petiole. 0 = dorsal carinae extending posteriorly beyond spiracle for at least a short dis- tance as a distinctly elevated ridge; 1 = dorsal carinae not extending past spi- racle as a distinctly elevated ridge. At least three species are variable in this feature, as reflected by the coding in Ta- ble 1. Character 12. — Configuration of ovipositor tip. 0 = distinct double node dorsally; 1 = weak node or swelling dorsally; 2 = parallel-sided at apex, with little or no node; 3 = strongly tapered apically to a fine, smooth point, narrowest subapi- cally; 4 = strongly tapered apically as in state 3, but with tip flattened dorsal- ventrally. States 1 and 2 merely repre- sent different degrees of development of a transverse ridge near the tip of the ovipositor; and states 3 and 4 represent 52 Journal of Hymenoptera Research conditions for which it is hypothesized here that nodes and /or transverse ridg- es have been lost. Detailed SEM work is still needed to elucidate these character states for many of the species. Character 13. — Approximate ovipositor length. Values given are total ovipositor length divided by length of mesosoma. Accurate measurement of ovipositor length often requires dissection, which was not possible for some of the species. Character 14. — Mesopleural setae. 0 = at least some setae present on mesopleu- ron dorsal to the speculum (the dorsal- posterior section of the mesopleuron); 1 = setae completely absent above spec- ulum. Table 1 has been arranged to facilitate identification of both species and species groups. Several of the species groups are quite distinctive and thus readily recog- nizable (Wharton 1997b), and these will be treated first in the following discussion. Since the focus of this paper is on parasit- oids of fruit-infesting Tephritidae, F. riifi- cornis (Granger) and the marangensis spe- cies group are not further discussed be- cause there are no host records and the species are readily identified using Ta- ble 1. The desideratus species group of Fopius consists of bevisi (Brues), desideratus (Brid- well), niger (Szepligeti), ottotomoamis (Ful- laway), and riifotestaceiis (Granger). As noted by Wharton and Gilstrap (1983), the species of the desideratus group are very similar to one another. For example, riifo- testaceus is virtually identical to bevisi, but has the mesosoma red rather than yellow or yellow-orange. Unfortunately, few specimens have been available for study of intraspecific variation in the color pat- terns currently used to differentiate the species of this group. There are published host records (summarized by Wharton and Gilstrap 1983) for all but riifotestaceiis. Most of the specimens examined were reared from Dacus infesting Cucurbita- ceae. Both desideratus and ottototiioauus have been recorded from undetermined species of Dacus in cucurbits (Bridwell 1919, Fullaway 1957), and uiger was reared from D. humeralis Bezzi (Wharton and Gilstrap 1983). The few remaining published records (Bridwell 1919, Clausen et al. 1965) are from Ceratitis aiiouae Gra- ham on Myriauthus arboreus (specimens of desideratus) and Trirhithrum queritum Mun- ro on Stri/chiios usanibarensis (a specimen tentatively identified as bevisi). All mem- bers of this species group have large, sub- apical nodes on the ovipositor. Based on comparisons of ovipositor morphology with species of known biology in the re- lated genus Diachasmimorpha Ashmead, it is suggested here that members of the de- sideratus group attack late instar larvae of their hosts. Members of this group are known from Cameroon, Kenya, Nigeria, Tanzania, South Africa, and Madagascar, and undoubtedly occur throughout sub- saharan Africa. Fopius rubrithorax (Grang- er) is very similar to the other species mentioned here, despite the reduced sculpture on the frons and a slightly more setose ovipositor sheath; and I therefore place it as a basal member of this group. This placement assumes that both the re- duced setal pattern on the ovipositor sheath and the pattern of sculpture on the frons of the five other species of the desi- deratus group are derived relative to the conditions in rubrithorax; this remains to be tested in a more rigorous fashion. The persulcatus species group of Fopius, characterized largely by striate sculpture on the second metasomal tergum, consists of altcniatae (Tobias), arisanus (Sonan), car- ponn/iae (Silvestri), myolejae (Tobias), per- sulcatus (Silvestri), skinneri (Fullaway), and vaiulenboschi (Fullaway). The species are very similar to one another, but differ pri- marily in coloration, length of ovipositor, and configuration of the ovipositor tip. Following their successful introduction to Hawaii during the biological control pro- gram against oriental fruit fly (Clausen et Volume 8, Number 1, 1999 53 al. 1965), arisaniis and vandenboschi were intensively studied, and much is now known about their biology (with most of the early literature on arisaiius published under the name Opiiis oophihis Fullaway). In their native range, centered around Ma- laysia and Indonesia, both are parasitoids of tephritids in the dacine genus Bactwcera Macquart. Unlike skinneri, neither is at- tracted to cucurbit-infesting flies. Other than the original host records little is known about the other species in this spe- cies group, including skinneri. Data on per- sukaius (type material reared from B. car- i/eae Kapoor) are particularly problematic because of widespread confusion regard- ing the identity of this species during the Hawaiian oriental fruit fly program, and the subsequent description of several sub- species (Fischer 1965). The other species (viz. carpomi/iae, tm/olejae, and alternatae) have been reared, respectively, from try- petine tephritids in the genera Carpomya Costa, Myoleja Rondani, and Rhagoletis Loew. Based on the similarities in the shape of the ovipositor tip, all species in this species group preferentially attack ei- ther the egg or early instar larva of their host (though all eventually emerge from the puparium). This biology, however, has only been confirmed for arisanus (attack- ing eggs) and vandenboschi (attacking pri- marily first instars). The species of this group are known from Pakistan and India east through Indonesia and north through Taiwan, Japan, and far eastern Russia. Wharton (1997b) delimited a silvestrii species group containing longicauda (Granger), pyknothorax (Fischer), and sil- vestrii (Wharton). One of the species de- scribed below, ceratitivorus, also belongs here. This group is currently defined largely by the absence of features that de- fine the three species groups already men- tioned: the clypeus lacks a median tooth on the ventral margin, the setae on the ovipositor sheath are not reduced, and the second metasomal tergum is unsculp- tured. Reduction of features on the dorsal valve of the ovipositor suggests either a sister group relationship to the persulcatus species group or a parallel loss relative to the desideratus species group, but this hy- pothesis needs to be tested more rigorous- ly. Both sili>estrii and ceratitivorus have been reared from ceratitine tephritids in- festing coffee, and silvestrii has also been reared from Dacus bivittatus (Bigot) infest- ing squash (Steck et al. 1986, Wharton 1987). Other members of this group have not been reared. Members of the silvestrii species group have much the same distri- bution pattern as those of the desideratus species group, and are differentiated from one another largely by color (silvestrii and pyknothrorax are dark, longicauda and cer- atitivorus are pale) and ovipositor length. The remaining three species, caudatus, deeralensis, and schlingeri, do not readily cluster into distinctive species groups. Identification of deeralensis and schlingeri is discussed below under the diagnosis fol- lowing the description of schlingeri; cau- datus is readily separated from all other species of Fopius by the distinctive band of setae and punctures on the frons. Both deeralensis and schlingeri are from Queens- land, where (as noted below under the de- scription of schlingeri and in Clausen et al. 1965) they have been reared from various species of Bactrocera in a variety of host fruits. Fopius caudatus has thus far been reared exclusively from ceratitines (Steck et al. 1986). It is known from tropical re- gions of both eastern and western Africa, where it has been reared from coffee ber- ries containing the ceratitines Trirhithruin coffeae Bezzi, C. anonae and C. capitata as well as from other fruits containing anon- ae. Specific host records for caudatus need confirmation, in part because of earlier confusion regarding its identity (Wharton 1987). This species resembles members of the desideratus species group in the mor- phology of the clypeus and petiole, but has a distinctly different ovipositor (strongly narrowed, suggesting oviposi- tion in the host egg) as well as several fea- 54 Journal of Hymenoptera Research tures unusual for members of the genus Fopms (Wharton 1997b). A few generalizations can be made about hosts and distribution patterns, even though our current knowledge is somewhat limited. Rearing records from within their native ranges (Clausen et al. 1965, Steck et al. 1986) suggest that the 16 species for which we have host records are restricted to fruit-infesting tephritids, but that there are different levels of specificity. Some are currently known only from a single host, others (e. g. caudatus on cera- titines) have only been reared from a nar- row group of hosts, and several have been reared from hosts in two or three different tribes. Most of the known hosts belong to the tribes Ceratitini and Dacini, both in the tephritid subfamily Dacinae (White and Elson-Harris 1992). Except where they have been introduced for biological con- trol, members of the persukatus group oc- cur outside the range of fruit-infesting Ceratitini, and several of them have been reared from Trypetini in the tephritid sub- family Trypetinae. Where introduced out- side their native range for biological con- trol, arisanus and vandenboschi have been able to attack other fruit-infesting tephri- tids (Clausen et al. 1965, Wharton et al. 1981). Yet, while some of the species of Diachasmimorpha Viereck introduced to Hawaii to control fruit-infesting Tephriti- dae occasionally attack gall-making (but not flower-infesting) Tephritidae, arisanus and vandenboschi do not (Duan et al. 1996). The genus Fopius provides evidence for a close relationship between the fauna of Madagascar and that of adjacent regions of continental Africa (as do its host te- phritids). Although a few of the Madagas- car elements (notably rubrithorax and es- pecially ruficornis) are unique in several respects, both rufotestaceus and longicauda have their closest known relatives (bevisi and ceratitivorus respectively) on the adja- cent mainland. DESCRIPTIONS Fopius ceratitivorus Wharton, new species (Figs. 1, 2, 7, 8, 10, 11, 13-15, 21) Female.— Head: 1.55-1.75 (m = 1.65 ±.07) times broader than long; 1.25-1.35 Hmes broader than mesoscutum; face distinctly punctate throughout, pattern variable but spacing between most punctures about equal to diameter of punctures; setae short, somewhat decumbent; midridge low, pol- ished, more prominent dorsally, extending between antennal bases (toruli) as a low, flat ridge; distance between toruli greater than distance from torulus to eye; frons longitudinally rugulose along midline, highly polished and weakly depressed ba- sally on either side of rugulose midline, deeply punctate elsewhere, the patch of punctures on each side anteriorad ocelli usually more densely spaced, occasionally with punctures coalescent, ocellar triangle almost completely margined by a crenulate sulcus. Occipital carina in lateral view ex- tending dorsally from base of mandible to a point just below top of eye. Clypeus in profile slightly bulging dorsomedially; ventral margin of clypeus thin and evenly convex, not thickened medially; setae on clypeus very sparse, at least twice length of those in middle of face, weakly directed ventrally; clypeus completely concealing labrum when mandibles closed. Eye (at 50x) apparently bare, large, 2.85-3.8 (m=3.2±0.3) times longer temple; temples very weakly receding in dorsal view; width of head at temples slightly less than width at eyes. Antenna 31-37 segmented; roughly 3.0-3.1 times longer than mesosoma; first flagellomere 0.9-0.95 times length of sec- ond. Maxillary palps longer than height of head. Mesosoma: 1.2.5-1.35 (m = 1.3±0.05) times longer than high, 1.55-1.7 (m = 1.65±0.05) times longer than broad. Median lobe of mesoscutum with 2 paral- lel, rugosopunctate, longitudinal grooves extending more than half length of median lobe, median lobe otherwise setose, with Volume 8, Number 1, 1999 55 Figs. 1-6. Heads of Fopniii spp.: 1 and 2, ccratitivonis frons and face; 3, sc/i/diycn face and cljpeus; 4, aiiiHiiHi frons; 5, i^chlingeri face and clypeus; 6, liccrulciifis face and clypeus (most setae broken), arrow = median projection on ventral margin of clypeus. scattered, deep punctures; lateral lobes bare and impunctate medially, with nu- merous, relatively long, inwardly directed setae around margins; notauli broadening posteriorly, distinctly crenulate through- out, with pits usually becoming elongate posteromedially where the ridges between the pits form a small strigose area; space between strigose area and scutellar sulcus with scattered, deep punctures, postero- median area either broadly and very shal- lowly depressed or with a shallow, more discrete midpit. Scutellar sulcus broader medially than laterally, the posterior mar- gin with a distinct median excavation; number of longitudinal ridges in sulcus 56 Journal of Hymenoptera Research variable. Metanotum with relatively low median ridge. Propodeum finely, densely rugose, the sculpture without obvious pat- tern; elevated median longitudinal carina usually distinct only on anterior 0.25-0.35; propodeum laterally not separated from metapleuron by a well-defined pleural ca- rina, the demarcation represented only by the transition to the weakly sculptured dorsal portion of the metapleuron. Stemau- lus broad, deep, crenulate throughout, ex- tending posteriorly roughly 0.7 times dis- tance from anterior margin of mesopleuron to mid coxa; crenulate sculpture extending dorsally along anterior margin of meso- pleuron through subalar depression; pos- terior margin crenulate ventrad speculum, but with unsculptured sulcus dorsally; me- sopleural disc setose throughout; postpec- tal carina present, but variously developed. Wing: SHgma 2.7-2.9 (m=2.85±0.1) times longer than wide, with r arising slightly distad its midpoint; 2RS weakly sinuate, 1.2-1.45 (m=1.3±0.05) times longer than 3RSa; 3RSa 1.55-2.5 (m=2.1±0.3) times lon- ger than r; 3RSb ending slightly but dis- tinctly anteriorad wing tip; (RS-l-M)a sinu- ate; (RS-l-]VI)b present, m-cu nearly always arising distinctly basad 2RS; Icu-a incli- vous, usually postfurcal relative to IM but varying from interstitial to postfurcal by 0.4 times its length. Hind wing m-cu recli- vous, straight or very weakly recurved near wing margin, extending to wing mar- gin or nearly so as well-developed, deeply impressed crease, usually weakly pigment- ed anteriorly. Metasoma: Petiole 0.95-1.05 (m = 1.0 ±0.05) times longer than apical width, apex 1.80-2.15 (m=1.95±0.1) Hmes wider than base; densely and finely striate; dorsal carinae well-developed over basal two-thirds, weaker posteriorly but distinct to posterior margin, carinae very weakly converging, with distance between carinae at posterior margin roughly equal to dis- tance to lateral margin; dorsope present but not extending basally as a deep pit. Metasoma unsculptured beyond petiole. Hypopygium strongly narrowed and pointed posteriorly but short, not greatly attenuate. Ovipositor tip weakly narrowed apically, without distinct dorsal node or ca- rina but with weak ventral serrations; 1.65- 1.95 (m = 1.8 ±0.1) times longer than meso- soma; ovipositor sheath densely setose with multiple rows of at least 30 setae each, the number of rows difficult to distinguish because of density of setae; sheath 1.35- 1.55 (m = 1.45 ±0.05) times longer than me- sosoma. Color: Pale, yellow to orange, the exact hue dependent largely on manner of preservation; ovipositor sheath, veins, and stigma brown; antenna brown, with scape, pedicel, and basal flagellomeres usually yellow to orange medially. Wings hyaline. Male. — As in female except eye 2.9±0.25 times longer than temple; antenna 30-35 segmented; petiole narrowed at apex, 1.1- 1.25 (m = 1.15±0.05) times longer than api- cal width, apex 1.6-1.95 (m=1.75±0.1) times wider than base; dorsope less dis- tinct. Length: 2.0-3.4 (9) and 1.85-2.9 (6) mm. Hosts. — This species has been reared from isolated puparia of fruit-infesting Te- phritidae attacking coffee in central Ken- ya. It has also been reared from bulk sam- ples of coffee. The tephritids from these samples, in order of abundance, were Cer- atitis capitata, C. rosa Karsch, and Trirliith- ruin coffeae. All are members of the tribe Ceratitini, subtribe Ceratitina. Material examined. — Holotype female, "Kenya: Ruiru C. R. F. 17.IX.1996 ex: te- phritid on coffee berries Ref. No. CB03" Deposited in Kenya National Museum, Nairobi. Paratypes (BMNH, BPBM, RMNH, HDA, ICIPE, TAMU, USNM): 5 9, 76, "Kenya: Nairobi 20.V.1997 ex: Ceratitis capitata, coffee M. Ramadan & R. Mess- ing"; 339, 25 9, "Kenya E. Province, Mbeere Distr Mbeti south Rurima 30.iv.97 ICIPE Fruitfly Project ex Fruittly on Coffee berries"; 19, 16, "Kenya: Ruiru 15mi NNE Nairobi 10.iv.l995 ex: coffee No. CB05 ICIPE Collections"; 19, "Kenya: Western Prov. Koru iv.l995 ex. Coffea ca- neophora CAB Collections" and, 1 6 Volume 8, Number 1, 1999 57 Figs. 7-10. fi);iii(s spp.: 7, ccratitiz\^n(^ head, arrows = top of occipital carina ancJ mid-dorsal elevation of clypeus; 8 and 10, cernlitivonis dorsal view of mesosoma; 9, schliiigeri dorsal view of mesosoma. 58 Journal of Hymenoptera Research "Kenya: E. Province Mbeere District Rur- ima Farm 0°38'29"S, 37°29'49"E 3.X.1997 ex tephritid in coffee Wharton, Kimani, Ov- erholt." This species is known only from central Kenya. Diagnosis. — This species closely resem- bles longicauda, known only from Mada- gascar. The two species are similarly col- ored, have a densely punctate frons, densely setose ovipositor sheath, and identical configuration of the clypeus. Fop- itts ceratitivorus differs from longicauda pri- marily in the possession of a shorter ovi- positor (sheaths at least 2 times longer than mesosoma in longicauda). The median lobe of the mesoscutum and the junction of the notauli (Fig. 10) are also more heavily sculptured in ceratitivorus than in longicauda, and hind wing m-cu is straight- er. Both ceratitivorus and longicauda differ from other Old World species of Fopius ei- ther in coloration, sculpture of the frons, length of ovipositor, and /or shape of the clypeus and its relative degree of conceal- ment of the labrum. From other orange opiines reared from tephritids in coffee in Kenya, ceratitivorus may be readily distin- guished by the short second submarginal cell with fore wing m-cu distinctly sepa- rated from 2RS (Fig. 21) and by the com- pletely sculptured notauli. Discussion. — I place ceratitivorus in the silvestrii species group of Fopius (Wharton 1997b). The ovipositor tip of ceratitivorus, though narrowed, does not have exactly the same morphology as found in arisanus and schlingeri n. sp. Thus, females proba- bly do not oviposit in host eggs, but based on the shape of the ovipositor tip, they may attack early instars. The same may be true of longicauda, the holotype of which appears to have a similar ovipositor. A weak negative correlation was ob- served between body length and relative length of the ovipositor, but the sample size (N = 10) was too small to confirm this apparent trend. Fopius schlingeri Wharton, new species (Figs. 3, 5, 9, 12, 16, 18-20) Female.— Head: 1.55-1.65 (m = 1.6±.05) times broader than long; 1.3-1.4 (m = 1.35 ±.05) times broader than meso- scutum; face distinctly punctate through- out, pattern variable but spacing between most punctures distinctly greater than di- ameter of punctures; setae short, some- what decumbent; midridge low, polished, narrower dorsally, extending between to- ruli; distance between toruli varying from slightly to distinctly greater than distance from torulus to eye; frons with polished, weakly elevated, crenulately margined, triangular projection extending from me- dian ocellus at least half distance to toru- lus; frons otherwise punctate, the punc- tures anteriorad ocellar field dense, with spacing between punctures about equal to diameter of punctures; ocellar triangle margined at least in part by an impressed line. Occipital carina in lateral view ex- tending dorsally from base of mandible to about middle of eye. Clypeus in profile weakly to distinctly bulging dorsomedi- ally; ventral margin of clypeus thin and weakly but evenly convex, not thickened medially nor with median projection; se- tae on clypeus sparse, about twice length of those on face, weakly directed ventral- ly; ventral margin of clypeus not suffi- ciently convex to completely conceal la- brum when mandibles closed. Eye usually with 1-4 minute setae visible in dorsal view, very large, 5.3-7.5 (m = 6.45 ±0.75) times longer than temple; temples weakly receding in dorsal view; width of head at temples about 0.9 times width at eyes. An- tenna 41-47 segmented; roughly 3.5 times longer than mesosoma; first flagellomere equal in length to second. Length of max- illary palps equal to height of head. Me- sosoma: 1.25-1.35 (m = 1.3) times longer than high, 1.75-1.80 times longer than broad. Median lobe of mesoscutum with 2 parallel, unsculptured, longitudinal grooves extending more than half length Volume 8, Number 1, 1999 59 Figs. 11-16. /-iij'/ii-. '-pp.: 11, ccratitworus propodeum; 12, schliiigeri propodeum; 13 and 14, ccrnlili,\'iii> pro- pleuron, armw = oblique carina; 15, ccriititivoriis petiole; 16, .•ic/i//);,sji'r( petiole. 60 Journal of Hymenoitera Research Figs. 17 cind 18. Ovipositors of Fopius spp; 17, deernlensis; 18, sir, rlilingcri. of median lobe, median lobe otherwise se- tose with numerous, very fine, widely spaced punctures; lateral lobes with nu- merous, relatively long setae around mar- gins and more sparsely scattered setae me- dially; notauli distinctly crenulate though- out, meeting posteriorly in a clearly de- fined midpit that often extends narrowly to posterior margin. Scutellar sulcus par- allel sided or nearly so, usually with 3 well-developed longitudinal carinae plus several additional weaker ones. Metano- tum with distinctly elevated median flange posteriorly. Propodeum densely ru- gose, the sculpture largely without obvi- ous pattern though elevated median lon- gitudinal carina distinct on anterior 0.25, and posterior 0.25 often with remnants of the parallel ridges from a median areola; propodeum laterally not separated from metapleuron by a well-defined pleural ca- rina, the demarcation represented only by the transition to the weakly sculptured dorsal portion of the metapleuron. Ster- naulus broad, deep, crenulate throughout, extending posteriorly roughly 0.7 times the distance from anterior margin of me- sopleuron to mid coxa; crenulate sculpture extending dorsally along anterior margin of mesopleuron throughout subalar de- pression; posterior margin crenulate ven- trad speculum, with unsculptured sulcus dorsally; mesopleural disc setose; postpec- tal carina well developed medially. Wing: Stigma 2.6-3.0 (m = 2.8±0.15) times longer than wide, with r arising slightly distad its midpoint; 2RS nearly as sinuate as (RS-HM)a, 1.15-1.3 (m = 1.25±.05) times longer than 3RSa; 3RSa 1.7-2.25 (m = 1.9±0.2) times longer than r; 3RSb ending nearly at wing tip; (RS-l-M)a sin- uate; (RS-(-M)b present and fairly long, Volume 8, Number 1, 1999 61 roughly 0.25 times length of m-cu; Icu-a inclivous, usually slightly postfurcal rela- tive to IM but varying from nearly inter- stitial to postfurcal by 0.5 times its length. Hind wing m-cu strongly reclivous, dis- tinctly recurved near wing margin, ex- tending to wing margin or nearly so as a deeply impressed, completely pigmented crease. Metasoma: Length of petiole 0.85- 0.95 (m=0.9±0.05) times width at apex; apex 2.3-2.5 (m=2.4±0.1) times wider than base; moderately and somewhat ir- regularly striate posteriorly; dorsal carinae well-developed over basal two-thirds, weaker posteriorly, often indistinct at pos- terior margin, parallel to very weakly con- verging posteriorly; dorsope weakly de- veloped. Metasoma unsculptured beyond petiole. Hypopygium strongly narrowed and distinctly pointed at extreme posterior end, but short, not greatly attenuate, length along midline about 0.55 times width at base. Ovipositor tip strongly nar- rowed subapically, without dorsal node or carina, ventral serrations indistinct to ab- sent; 2.3-2.55 (m = 2.4±0.1) times longer than mesosoma; ovipositor sheath mod- erately setose with 3 rows of setae, two of which have 30-35 setae per row with the third row more sparsely setose, distinct tuft of longer setae at apex, sheath 2.0-2.25 (m = 2.09±0.1) times longer than mesoso- ma. Color: Orange; propleuron and pro- podeum often paler, at least in part, some- times nearly white; ovipositor sheath, hind tarsi, flagellum, and sometimes ped- icel dorsally brown to light brown; base of arolium dark brown. Wings weakly to dis- tinctly infumate: more noticeably infu- mate in larger specimens. Male. — As in female except eye distinct- ly smaller, 4.4-5.85 (m = 5.15±0.5) times longer than temple; petiole narrower at apex with length equal to apical width and apex 1.95-2.25 (m=2.15) times wider than at base. Length: 3.05-4.9 mm. Hosts. — This species has been reared from guava (Myrtaceae) infested with Bnc- twccra tri/oni (Froggatt), Rauwenhoffia lei- chardtii (Armonaceae) infested with B. hal- fordiae (Tryon) and B. neohumeralis (Har- dy), Sijzygium hamagense (Myrtaceae) in- fested with B. rufofusciila (Drew and Hancock), and Fagraea cnmbngei (Logani- aceae) infested with B. peninsularis (Drew and Hardy). It thus appears to be able to attack several species of Bactrocera (Tribe Dacini) developing in the fruit of at least three plant families. Material examined. — Holotype female, "Australia: QLD Mt. Glorious 26.1.1994 G. Quimio MG9413 ex: Rauwenhoffia lei- chardtii fruit containing Bactrocera halfor- diae and B. neohumeralis" Deposited in ANIC. Paratypes (ANIC, BMNH, QDPI, TAMU, UQBA, USNM): 179, \76, same data as holotype; 29, 26, "Nambour Qld 24.iv.95 Guava G. Quimio"; 2 9, 13, same data except 29.V.95; 19, "Australia: QLD Wongabel 6 km S Atherton 1-28-1990 R. Wharton; 19, "Malanda NQ 22.xi.1987 M. Elson-Harris Ex Dacus rufofusculus 249"; 2 9, "Sydney, N.S.W., 28-6.1954 G. J. Snowball" one of these with an additional label "37/54" and the other "51/54". Ad- ditional material (not paratypes): 11 9, 3(5, Australia, North Queensland, Balinda, 10. vi. 1993 from fruits of Fagraea canibagei infested with Bactrocera peninsularis. This species is known only from the eastern coast of Australia. Diagnosis. — As with ceratitivorus, schlin- geri also closely resembles the Madagascar species longicauda. The latter has a smaller eye (slightly less than 4 times longer than temple), somewhat more densely setose ovipositor sheath, a smaller gap between clypeus and mandibles when mandibles closed, and the distal portion of the ovi- positor is parallel-sided rather than suba- pically narrowed relative to schlingeri. Of the species known from Queensland, schlingeri most closely resembles deeralen- sis. The latter has the ventral margin of the clypeus distinctly pointed midventrally (Fig. 6, with mid-ventral projection more distinct when head rotated forward), a distinct subapical ridge on the ovipositor 62 Journal of Hymenoptera Research Figs. 19-21. Wings of Fopnus spp.: 19 and 20, f.cJiliiigcri tore and hind wing; 21, cfintitivcnif fore wing. (Fig. 17), and the second metasomal ter- gum is usually weakly striate, at least ba- sally. Discussion. — Wharton (1997b) observed that the hypopygium is strongly attenuate in many of the parasitoids of fruit-infest- ing Tephritidae. He also noted in his re- description of Fopius that the hypopygium varies from weakly to strongly produced posteromedially. Though distinctly nar- rowed and projecting posteriorly in both ceratitivorus and schlingeri, the hypopy- gium is much less strongly produced than it is in members of the Fopius nmraiigensis species group or in other fruit-infesting te- phritid parasitoids such as the members of the Diachasmimorpha longicaudata (Ash- mead) species group or the species of Psi/t- talia. Fischer (1987) includes species with a partially visible labrum in Diachasma Foerster. Wharton (1997b) briefly dis- cussed variation in this character in Fopius and other genera, noting that some of the "variation" can be attributed to angle of view or the degree to which the mandibles are closed on any given specimen (com- pare Figs. 3 and 5). Nevertheless, there are slight differences among species of Fopius in the exposure of the labrum, and schlin- geri provides a good example of a species with a partially exposed labrum (in con- trast to the completely concealed labrum of ceratitivorus). The ovipositor in scliliugeri (Fig. 18) is virtually identical in form to that of arisanus, strongly suggesting a bi- ology similar to the latter in which the Volume 8, Number 1, 1999 63 parasitoid oviposits into the egg of its host. Field observations kindly supplied by Greg Quimio of the University of Queensland support this. ACKNOWLEDGMENTS I am especially grateful to the following collabo- rators for providing reared material for this study: M. Elson-Harris, F. Gilstrap, S. Kimani, C. Lopez-Vaa- monde, S. Lux, R. Messing, W. Overholt, G. Quimio, M. Ramadan, G. Steck, and T. Wong. 1 also thank M. Trostle and S. Kimani for assistance in isolating pu- paria from which FopiKs ccrntitivorus were reared. Ad- ditionally, the following people kindly made material available for examination, without which the com- parative aspects of this work would have been im- possible: Greg Daniels and Gimme Walter (UQBA, Brisbane), Ian Naumann (ANIC, Canberra), David Wahl (American Entomological Institute, Gaines- ville), Valter Raineri (Museo Civico di Storia Naturale "Giacomo Doria," Genoa), Gordon Nishida, Keith Arakawa, and David Preston (BPBM, Honolulu), Dick Drew and all the members of his research group (formerly at QDPl, Indooroopilly), Kaoru Maeto (Shi- koku Research Centre, Kochi), Tom Huddleston (for- merly at BMNH, London), Ermenegildo Tremblay (Universita de Napoli, Naples), Claire Villemant (Mu- seum National d'Histoire Naturelle, Paris), Sergey Belokobylskij (Zoological Institute, St. Petersburg), Zafar Qureshi, Atomic Energy Research Centre, Tan- do Jam, and Paul Marsh and David Smith (Systematic Entomology Laboratory, USD A, Washington, D. C). This work was funded in part by USDA-CSREES Spe- cial Grant No. 96-34135, Tropical and Subtropical Ag- riculture Research (to R. Messing), USDA/NRl, and the National Science Foundation (DEB9712543), with additional travel support in Australia provided bv R. A. 1. Drew and A. Austin. Support of the Texas Ag- riculture Experiment Station, the International Centre of Insect Physiology and Ecology, and The Electron Microscopy Center of Texas A&M University is also gratefully acknowledged. Finally, 1 am indebted to Nando Bin for assistance in obtaining specimens, and Imelda Mercado, Ken Wilks, and Randy Scott for il- lustrations. LITERATURE CITED Bridwell, J. C. 1919. Descriptions of new species of hymenopterous parasites of muscoid Diptera with notes on their habits. Pwceeiiiii^f of the Ha- umiiim EntonwIo^iciU Sccict\/ 4: 166-179. Clausen, C. P. 1978. Tephritidae (Tr\'petidae, Trupa- neidae), pp. 320-335. In: C. P. Clausen (ed.). In- troduced parasites and predators of arthropod pests and weeds: a world review. Uiiitcil Stak? Department of Agriculture Agricultural Handbook No. 480. Clausen, C. P., D. W. Clancy and Q. C. Chock. 1965. Biological control of the Oriental fruit fly {Drtc»s ilor>aUf Hendel) and other fruit flies in Hawaii. Llnited Sffl/cs Department of Agriculture Technical Bulletin 1322: 1-103. Duan, J. J., M. F. Purcell and R. H. Messing. 1996. Parasitoids of non-target tephritid flies in Ha- waii; implications for biological control of fruit fly pests. Eiitonwphaga 41: 245-256. Fischer, M. (1963) 1965. Die orientalischen und aus- tralischen Arten der Gattung Opius Wesm. Acta Entoniologica Mu^ei Nationalif Pragae 35: 197-242. Fischer, M. 1972. Hymenoptera: Braconidae (Opiinae I). Das Tierreich 91; 1-620. Fischer, M. 1977. Hymenoptera: Braconidae (Opiinae Il-Amerika). Das Tierreich 97: 1-1001. Fischer, M. 1987. Hymenoptera: Opiinae III — athiop- ische, orientalische, australische und ozeanische Region. Das Tierreich 105; 1-734. Fullaway, D. T. 1957. A new reared Opius from Africa (Hymenoptera, Braconidae). Proceedings of the En- tomological Society of Washington 59: 98-99. Gilstrap, F. E. and W. G. Hart. 1987. Biological control of the Mediterranean fruit fly in the United States and Central America. United States Department of Agriculture ARS-56: 1-64. Headrick, D. H. and R. D. Goeden. 1996. Issues con- cerning the eradication or establishment and bi- ological control of the Mediterranean fruit tly, Ceratitis capitata (Wiedemann) (Diptera, Tephri- tidae), in California. Biological Control 6: 412—121. Knipling, E. F. 1992. Principles of insect parasitism analyzed from new perspectives; practical impli- cations for regulating insect populations by bio- logical means. United States Department of Agri- culture Handbook 693: 1-337. Messing, R. M. 1996. Status and needs of biological control research for tephritid flies, pp. 365-367. In; McPheron, B. A. and G. J. Steck (eds.). Fruit Fly Pests — A world assessment of their biology and management. St. Lucie Press, Delray Beach, Flor- ida. Palacio, I. P., R. Ibrahim and A. G. Ibrahim. 1992. Identification of immatures and male adults of the opiine parasitoids (Biostercs spp.) of the Ori- ental fruit fly, Bactrocera dorsalis (Hendel). The Philippines Entomologist 8; 1124-1146. Purcell, M. 1998. Contributions of biological control to integrated pest management of tephritid fruit flies in the tropics and subtropics. Integrated Pest Management Rei'iews 3: 1-21. Quicke, D. L. J., K. van Achterberg and H. C. J. God- fray. 1997. Comparative morphology of the ven- om gland and reservoir in opiine and alysiine braconid wasps (Insecta, Hymenoptera, Bracon- idae). Zoologica Scripta 26: 23-50. Sharkey, M. J. and R. A. Wharton. 1997. Morphology and Terminology, pp. 19-37. In: Wharton, R. A., 64 Journal of Hymenoptera Research P. M. Marsh, and M. J. Sharkey (eds.). Manual of the New World Genern of the Ftvnihj Braconidae (Hy- menoptera). Special Publication of the Internation- al Society of Hymenopterists 1. Sivinski, J. 1996. The past and present of biological control of fruit flies, pp. 369-375. In: McPheron, B. A. and G. J. Steck (eds.) Fruit Fly Pests-A worhi assessment of tlieir bioh^i/ and management. St. Lucie Press, Delray Beach, Florida. Sivinski, J., M. Aluja and M. Lopez. 1997. Spatial and temporal distributions of parasitoids of Mexican Annstrepha species (Diptera: Tephritidae) within the canopies of fruit trees. Annals of tlie Entomo- logical Society of America 90: 604-618. Steck, G. J., F. E. Gilstrap, R. A. Wharton and W. G. Hart. 1986. Braconid parasitoids of Tephritidae (Diptera) infesting coffee and other fruits in west-central Africa. Entomopliaga 31: 59-67. Tobias, V. 1. (1977) 1978. The genus Opius Wesm. (Hy- menoptera, Braconidae) as parasites of fruit flies (Diptera, Tephritidae). Entomological Review 56: 132-139. Van Achterberg, C. and K. Maeto. 1990. Two new aberrant species of Braconidae (Hymenoptera) from Japan. Zoologische Mededelingen Eeiden 64: 59-70. Van Achterberg, C. and A. Salvo. 1997. Reared Opi- inae (Hymenoptera: Braconidae) from Argentina. Zoologische Mededelingen Leiden 71: 189-214. Waterhouse, D. F. 1993. Bwlogical Control: Pacific Pros- pects—Supplement 2. Australian Centre for Inter- national Agricultural Research, Canberra. 138 pp. Wharton, R. A. 1987. Changes in nomenclature and classification of some opiine Braconidae (Hyme- noptera). Proceedings oftlie Entomological Society of Washington 89: 61-73. Wharton, R. A. 1988. Classification of the braconid subfamily Opiinae (Hymenoptera). The Canadian Entomologist 121: 333-360. Wharton, R. A. 1989. Classical Biological Control of Fruit-infesting Tephritidae, pp. 303-313. In: Rob- inson, A. S. and G. Hooper (eds.). Fruit Flies, Their Biology, Natural Enemies and Control. Elsev- ier Science Publishers B.V., Amsterdam. Wharton, R. A. 1997a. Subfamily Opiinae, pp. 378- 395. In: Wharton, R. A., P. M. Marsh, and M. J. Sharkey (eds.). Manual of the New World Genera of the Family Braconidae (Hymenoptera). Special Pub- lication of the International Society of Hymen- opterists 1. Wharton, R. A. 1997b. Generic relationships of opiine Braconidae (Hymenoptera) parasitic on fruit-in- festing Tephritidae (Diptera). Contributions of the American Entomological Institute 30: 1-53. Wharton, R. A. and F. E. Gilstrap. 1983. Key to and status of opiine braconid (Hymenoptera) parasit- oids used in biological control of Ceratitis and Dacus s. 1. (Diptera: Tephritidae). Annals of the Entomological Society of America 76: 721-742. Wharton, R. A., F. E. Gilstrap, R. H. Rhode, M. Fis- chel-M. and W. G. Hart. 1981. Hymenopterous egg-pupal and larval-pupal parasitoids of Cera- titis capitala and Anastrepha spp. in Costa Rica. Entomopliaga 26: 285-290. Wharton, R. A. and P. M. Marsh. 1978. New World Opiinae (Hymenoptera: Braconidae) parasitic on Tephritidae (Diptera). Journal of the Washington Academy of Sciences 68: 147-167. White, 1. M. and M. M. Elson-Harris. 1992. Fruit Flies of Economic Significance: Their Identification and Bi- onomics. CAB International, London and ACIAR, Canberra. 601 pp. J. HYM. RES. Vol. 8(1), 1999, pp. 65-73 New Genera and Species of Brachycistidine Wasps from Southwestern North America (Hymenoptera: Tiphiidae: Brachycistidinae) Lynn S. Kimsey and Marius S. Wasbauer Bohart Museum of Entomology, Department of Entomology, University of California, Davis 95616 Abstract. — The new genera Brachymaya, Paraquemaya and Sedomaya are described from southwestern Arizona, Baja California and Sonora Mexico, and Imperial Co., California. Brachy- maya is based on the new species mexicana. Paraquemaya is based on the new species pallida, and Sedomaya is based on the new species glamisensis. Three other new species of Paraquemaya are also described: bitincta, maxima, and bajaensis. Until recently (Kimsey and Wasbauer 1998) the subfamily Brachycistidinae had not received taxonomic attention in nearly three decades. The last major revisions of any kind of members of this group were published by Mickel and Krombein (1942) and later by Wasbauer (1958, 1966 and 1968). Studies of unidentified male bra- chycistidines, which have accumulated in collections in the past 30 years, have re- vealed a variety of new taxa. Six of these undescribed species had a combination of features, which did not fit any of the current generic groupings. Pre- liminary cladistic analyses of the Brachy- cistidinae indicated the need to describe three new genera for these species. Each of these genera is characterized by a unique combination of apomorphic fea- tures, which occur in other brachycistidine genera in various combinations. They in- clude a tailed antennal socket, external mandibular carina, stridulatory structure on the forecoxa, an elongate digitus, changes in the position of the carina en- closing the oral fossa, loss of the hindcoxal carina, loss of the basolateral carinae on the first gastral tergum, and highly re- duced wing venation. The first of these genera, Paraquemaya, has a large number of primitive features, although it is still more highly derived than Quemaya. Para- quemaya has the following apomorphic characteristics: a tailed antennal socket, presence of a mandibular carina, an elon- gate digitus and hindwing venation simi- lar to that seen in Brachycistellus and Had- rocistis. The second new genus, Brachy- maya, appears to be most closely related to Bracliycistina and Hadrocistis, based on the slender, ecarinate mandible and lack of the dorsal carina on the hindcoxa. Finally, the third genus, Sedomaya, belongs to a group of genera with the stridulatory structure on the forecoxa and the first gas- tral sternum with a short medial carina. This group also includes Brachycistis, Bra- chycistellus and Colocistis. As with many of the brachycistidine genera the undoubtedly wingless, and probably nocturnal, females are unknown. Although a number of genera have been described based only on females (Mickel and Krombein 1942) the taxonomy of the Brachycistidinae is essentially based on the males. None of the female-based gen- era described by Mickel and Krombein (1942) are known to occur in the same re- gion as the taxa described below (Fig. 1), although this is no guarantee that there is 66 Journal of Hymenoptera Research Fig. 1. Distribution map showing collection locali- ties of the brachycistidine genera Brachymaya, Para- queniayn and Scdomnya. no geographic overlap. In addition, the genera described by Mickel and Krombein are relatively large, ranging in length from 5-12 mm. The females of the new genera described below should be small, about 2- 3 mm in length, roughly the size of female Quemaya, which are also as yet unde- scribed. Female brachycistidines described thus far are considerably larger than the females of Erachymaya, Paraqiiemaya and Sedomaya should be based on the male to female body size ratio seen in species where both sexes have been described. Additionally, no hosts are known for any of the new species described below. De- scription of these new genera is essential to enable us to finalize phylogenetic ana- lyses of the subfamily Brachycistidinae. MATERIALS Specimens used in this study came from the Bohart Museum of Entomology, Uni- versity of California, Davis, S. L. Heydon (DAVIS); California Academy of Sciences, San Francisco, N. Penny (SAN FRANCIS- CO); Los Angeles County Museum, Cali- fornia, R. R. Snelling (LOS ANGELES); University of California, Riverside, S. Frommer, S. Triapitsyn (RIVERSIDE), and the personal collection of M. S. Wasbauer. Type repositories are indicated by the cap- italized name given in parentheses. Brachymaya Kimsey and Wasbauer, new genus (Figs. 2, 6, 18, 19) Description of male. — Face (Fig. 6): Man- dible with two apical teeth and without longitudinal carina on external surface; palpi long, extending well outside of oral fossa, maxillary palpus 6-segmented; la- bial palpus 4-segmented; inner eye margin converging medially; flagellomeres some- what arcuate, particularly apical articles; antennal carina greatly thickened into ventral subtriangular process, with tail- like carinule; gular carina with large sub- truncate swelling near mandible; clypeus medially evenly convex, apical margin without ventral bevel; forecoxa without stridulatory patch; scrobal pit small and circular; propodeum without longitudinal groove on dorsal surface and without transverse carina; metasternum without medial ridge terminating in two lobes near hindcoxal base; hindcoxa without ventral or dorsal longitudinal carinae; wings (Fig. 2): forewing with one large rhomboid submarginal cell, and smaller second submarginal cell, marginal cell nearly parallel-sided, Rl barely visible and bending away from the costal margin toward the stigma, one discoidal cell, one subdiscoidal cell; hindwing M vein di- verging from Cu + M after cross vein cu- a; gastral segment I, tergum with well-de- veloped lateral carina at base, sternum without longitudinal carina extending from base; epipygium delimited by small sublateral welt, apex truncate or some- what short and apically rounded; genital capsule (Figs. 18, 19): paramere slender VoLUMi; 8, NUMBEK 1, 1999 67 and tapering dorsally; volsella with small submedial lobe externally, inner surface with several rows of denticles; digitus elongate terminating in acutely pointed lobe. Distribution. — This genus is known only from the vicinity of San Augustine, BCN, Mexico. Eti/iuolo^i/. — The name is intended as a nonsense combination of letters, taken from Brachycistis and Quemaya, and is as- sumed to be feminine. Type species. — Bracbymaya mexicana n. sp. Discussion. — Bracbymaya is somewhat similar to Bracbycistiim and Hadrocistis. All three genera are characterized by having a slender mandible, with subsidiary den- tition reduced to one or two very small teeth, and the hindcoxa lacking a dorso- basal carina. As with most brachycistidine genera Brachymaya has the digitus elon- gate and pointed. However, Brachymaya has the antennal socket subtended by a large triangular enlargement of the carina and the gular carina is dilated, forming a truncate projection near the mandibular base. In addition, the wing venation is re- duced, with one discoidal and one sub- marginal cell. Brachymaifa mexicana Kimsey and Wasbauer, new species (Figs. 2, 6, 8, 19) Description of male. — Body length 5-6 mm; face (Fig. 6); clypeus evenly convex medially, apical truncation 0.9-1.0 mid- ocellus diameters wide; interantennal dis- tance 0.7-0.8 midocellus diameters wide; distance between midocellus and nearest eye margin 1.4-1.5 midocellus diameters; flagellomeres I-II 2.2-2.4 times as long as broad; facial and thoracic punctures small, shallow and widely spaced, 4—6 puncture diameters apart; forewing with two sub- marginal cells (Fig. 2); abdominal seg- ments appearing impunctate, integument finely shagreened; genital capsule (Figs. 18, 19). Head dark brown; thorax and legs pale yellowish brown; abdomen darker brown; forewing venation yellow, except stigma brown; hindwing venation yellow; wing membrane faintly yellow tinted. Type material. — Holotype male: Mexico: BCN, San Augustine, 20 Oct. 1956, R. Mat- toni (LOS ANGELES). Paratypes: three- same data as holotype (LOS ANGELES, DAVIS); one — 3 mi. s. San Augustine, 14 June 1973, J. Doyen (DAVIS). Etymology. — This species is named after the country of collection. Discussion. — Although there are cur- rently no other species placed in this ge- nus, species distinctions probably include proportions of the flagellomeres, ocello- cular distances, punctation and coloration of the wing veins. Paraqiiemaya Kimsey and Wasbauer, new genus (Figs. 3, 4, 7-9, 11, 12, 14-17) Description of male. — Face (Figs. 7-9): Mandible with three apical teeth and lon- gitudinal carina on external surface; palpi long, extending well outside of oral fossa, maxillary palpus 6-segmented; labial pal- pus 4-segmented; flagellomeres somewhat arcuate; antennal carina slightly thickened ventrally, but without tail-like carinule; gular carina with tooth-like projection near mandible (Fig. 12); forecoxa without stridulatory structure; scrobal pit small and circular; propodeum with longitudi- nal groove on dorsal surface and no trans- verse carina; metasternum with medial ridge terminating in two small lobes near hindcoxal base; hindcoxa without ventral or dorsal longitudinal carinae; wings (Figs. 3, 4): forewing with one large rhom- boid submarginal cell, and smaller second submarginal cell in several species, mar- ginal cell nearly parallel-sided, Rl barely visible and bending away from the costal margin toward the stigma, one discoidal cell, one subdiscoidal cell; hindwing M vein diverging from Cu -I- M after cross vein cu-a; gastral segment I, tergum with well-developed lateral carina at base, vis- 68 Journal of Hymenoptera Research 5 Sednmaya glamisensis 13 Scdomaya glamisensis Figs. 2-13. Figs. 2-5. Fore and hindwing of males. Figs. 6-11. Front view of male face, with right antenna removed. Figs. 12, 13. Side view of male head. ible in most specimens, sternum with short longitudinal carina extending from base; epipygium delimited by small sub- lateral welt, apex truncate or somewhat rounded; genital capsule (Figs. 14-17): parameres slender and tapering apically; volsella with small rounded submedial lobe, inner margin with small denticles; digitus elongate and awl-shaped. Distribution. — This genus occurs in southern Arizona and Baja California, Mexico. Etytnolo^y. — The generic name refers to the superficial similarity, because of their Volume 8, Number 1, 1999 69 small size, to species of Qufiiun/a; femi nine. Discussion. — Although Paraqucniai/a re- sembles Quemaya in terms of its reduced wing venation and small size, it shares the majority of characteristics with more de- rived brachycistidine genera including Brachycistis and Brncliycistelhis. Features shared with these genera include the elon- gate digitus, carinate mandible, unmodi- fied palpi, basal gastral sternum with a medial carina, and basal tergum with a well-developed lateral carina adjacent to the petiolar insertion. The presence of a gular tooth is a characteristic shared with some species of Brachycistis and one Que- maya species. Additional diagnostic fea- tures include the lack of hindcoxal carinae, no forecoxal stridulatory structure and the configuration of the hindwing venation, which is very similar to that of Hadrocistis and Brachycistellus. KEY TO THE SPECIES OF PARAQUEMAYA (MALES) 1. Forewing with two submarginal cells (Fig. 3); interantennal distance 0.5 midocellus di- ameter wide or less; body usually appearing bicolored with thorax paler than head and abdomen 2 - Forewing with one submarginal cell (Fig. 4); interantennal distance more than 0.5 mid- ocellus diameter; body concolorous 3 2. Shortest distance between midocellus and eye margin 1-1.2 midocellus diameters wide (Fig. 11); flagellomeres I and II subequal, 2.3-2.6 x as long as broad . . maxima new species - Shortest distance between midocellus and eye margin 1.4-1.6 midocellus diameters wide (Fig. 8); flagellomere I shorter than II, 2.2-2.3 times as long as broad and II 2.3-2.5 x as long as broad bitincta new species 3. Flagellomere I 2.0-2.2 times as long as broad (Fig. 9); flagellomere II 2.2-2.4 x as long as broad; interantennal distance 0.6-0.8 midocellus diameters wide; shortest distance between midocellus and eye margin 1.8-2.0 midocellus diameters wide pallida new species - Flagellomere I 1.7-1.9X as long as broad (Fig. 7), flagellomere II length 2.0-2.2X breadth; interantennal distance 0.8-0.9 midocellus diameters wide; shortest distance between mid- ocellus and eye margin 2.1-2.3 midocellus diameters wide bajaensis new species Paraquemaya bajaensis Kimsey and Wasbauer, new species (Fig. 7) Description of male. — Body length 4 mm; face (Fig. 7); clypeus with narrowly acute medial projection, apical truncation 0.9 midocellus diameter wide; interantennal distance 0.8 midocellus diameter; distance between midocellus and nearest eye mar- gin 2.3 midocellus diameters; flagellomere I twice as long as broad; flagellomere II length 2.4 X breadth; facial and thoracic punctures small and widely spaced, 4-6 puncture diameters apart; forewing with one submarginal cell; abdominal segments appearing impunctate, integument finely shagreened. Body color dark brown; fore- wing venation light brown-tinted, except stigma darker brown; hindwing venation faintly brown-tinted; wing membrane un- tinted. Type material. — Holotype male; Mexico, BCS, 14 mi s El Arco, 2 Nov. 1965, W. Ewart & R. Dickson (RIVERSIDE). Para- types: 1 male same data as holotype, ex- cept 5 Nov. 1965; one male: BCN, El Con- suelo, 0-50 ft. elev., 11 Sept. 1983, R. R. Snelling (DAVIS). Etymologi/. — The name is derived from the region of collection, Baja California, Mexico. Discussion. — This species most closely resembles pallida, as both have a single submarginal cell and the body color is 70 Journal of Hymenoptera Research 16. ---^ 17. Paraquemaya bitincta Paraquemaya maxima Brachymaya mexicana 21. Sedomaya glamisensis Figs. 14-21. Male genital capsule. Figs. 14, 16, 18, 20. Ventral view. Figs. 15, 19, 21. Dorsal view. Fig. 17. Lateral view. concolorous. These two also have the void- est ocellocular distance and widest inter- antennal distance. P. bajaensis has the smallest ocelli in the genus, with the mid- ocellus separated from the eye margin by 2.3 midocellar diameters. This species also has the longest basal flagellomeres of any species of Paraqueniai/a. Paraquemaya bitincta Kimsey and Wasbauer, new species (Figs. 3, 8, 16, 17) Description of male. — Body length 5 mm; face (Fig. 8); clypeus with narrowly acute medial projection, apical truncation 1 mid- ocellus diameter wide; interantennal dis- tance 0.3 midocellus diameter; distance Volume 8, Number 1, 1999 71 between midocellus and nearest eye mar- gin 1.6 midocellus diameters; flagello- meres I and II subequal, length 2.3-2.4 X breadth; facial and thoracic punctures small and widely spaced, 4-8 puncture di- ameters apart; forewing with two submar- ginal cells (Fig. 3); abdominal segments es- sentially impunctate, integument finely shagreened; genital capsule (Figs. 16, 17). Body color reddish brown, head and ab- domen darker brown than thorax; fore- wing venation untinted, except stigma brown; hindwing venation untinted; wing membrane untinted. Ti/pe mnterinl. — Holotype male: Arizona, Pima Co., Organ Pipe, 12 April 1947, A. L. Melander (RIVERSIDE). Three paratypes, same data as holotype (RIVERSIDE, DA- VIS). Eti/mologi/. — The name, bitincta, is de- rived from the two-tone brown color of the body. Discussion. — This species can be distin- guished from pallida and bajaejisis by the presence of two submarginal cells. It can be separated from maxima by the smaller body size, shorter flagellomeres shorter, and greater distance between the eye and the midocellus. Paraqiiemaya maxima Kimsey and Wasbauer, new species (Figs. 11, 14, 15) Description of male. — Body length 7 mm; face (Fig. 11); clypeus with acute, narrow- ly hooked medial projection, apical trun- cation 1 midocellus diameter wide; inter- antennal distance 0.5-0.6 midocellus di- ameters; distance between midocellus and nearest eye margin 0.1-1.1 midocellus di- ameters; flagellomeres I and II length 2.6 x breadth; facial and thoracic punctures small and widely spaced, 4-6 puncture di- ameters apart; forewing with two submar- ginal cells; abdominal segment I with broad, shallow irregular punctures and shagreening; segments II-V with sparse shallow punctures, 4-6 puncture diame- ters apart; genital capsule (Figs. 14, 15). Body color: head and abdomen brown, thorax lighter reddish brown; forewing venation pale brown-tinted, except stigma reddish brown; hindwing venation nearly colorless; wing membrane untinted. Ti/pe material. — Holotype male: Mexico, Sonora, 22 km se Quitovac, 14 Nov. 1965, W. Edward & R. Dickson (RIVERSIDE). Two paratypes, same data as holotype (DAVIS, RIVERSIDE). Eti/mologi/. — The name is based on the unusually large body size for this genus. Discussion. — Other than the large body size of this species, the two submarginal cells and long flagellomeres will distin- guish maxima from other species placed in Paraquemaya. Paraqiiemaya pallida Kimsey and Wasbauer, new species (Figs. 4, 9) Description of male. — Body length 3.5-6.0 mm; face (Fig. 9); clypeus with small acute medial projection, apical truncation 1.0- 1.3 midocellus diameters wide; interanten- nal distance 0.6-0.8 midocellus diameters; distance between midocellus and nearest eye margin 2.0-2.2 midocellus diameters; flagellomere I 2.0-2.2 times as long as broad; flagellomere II length 2.2-2.4 X breadth; facial and thoracic punctures small and widely spaced, 3-5 puncture di- ameters apart; forewing with one submar- ginal cell (Fig. 4); abdominal segments with sparse shallow punctures, 2-4 punc- ture diameters apart. Body color pale red- dish brown, rarely darker brown; fore- wing venation slightly brown-tinted, ex- cept stigma darker; hindwing venation untinted. Ti/pe material. — Holotype male: MEXI- CO: Baja California Sur, 4 mi. wsw Mira- flores, 23-24 April 1979, M. Wasbauer (DAVIS). Paratypes, 68 males (DAVIS, SAN FRANCISCO): 25— same data as ho- lotype; four — El Salto, 8 mi ne Todos San- tos, 9 Oct. 1983, D. Faulkner & F. An- drews; twenty-six — Los Barriles, 24 Mar.- 5 April 1984, J. H. Lynch; two — Puerto Es- 72 Journal of Hymenoptera Research condito, 10-14 July 1989, R. Shaver; one— BCN, Sierra Calamajue, 11 km e Chapala, 29°31'N 115°42'E, 23 Aug. 1994, S. Hey- don; one — BCS: El Pescadero, Playa Los Cerritos, 16-17 April 1979, M. Wasbauer; one— BCS, 13 mi nw La Paz, 10 Nov. 1965, W. Ewart and R. Dicksen; one — Sonora, 14 mi w Guaymas, 13 Nov. 1965, Ewart and R. Dicksen. Etymology. — The name, pallida, refers to the pale coloration of the male. Discussion. — Paraquemaya pallida is char- acterized by having one submarginal cell, flagellomeres I and II subequal in length and about 2.2 X as long as broad, and the midocellus separated from the eye by 2 midocellus diameters or slightly less. Sedornaya Kimsey and Wasbauer, new genus (Figs. 5, 10, 13, 20, 21) Description of male. — Face (Fig. 10): man- dible with three apical teeth and longitu- dinal carina on external surface; palpi long, extending well outside of oral fossa, maxillary palpus 6-segmented; labial pal- pus 4-segmented; eye strongly converging medially, inner eye margin slightly in- dented; flagellomeres somewhat arcuate, particularly apical articles; antennal carina slightly thickened ventrally, but without tail-like carinule; gular carina with large subtruncate swelling near mandible (Fig. 13); clypeus medially concave and slightly trilobate apically, apical margin with ven- tral bevel; forecoxa with small stridulatory patch; scrobal pit small and circular; pro- podeum with obsolescent longitudinal groove on dorsal surface and no trans- verse carina; metasternum with medial ridge terminating in two small lobes near hindcoxal base; hindcoxa with dorsal lon- gitudinal carinae, without ventral one; wings (Fig. 5): forewing with one large rhomboid submarginal cell, and smaller second submarginal cell, marginal cell nearly parallel-sided, Rl barely visible and bending away from the costal margin toward the stigma, one discoidal cell, one subdiscoidal cell; hindwing M vein di- verging from M + Cu after cross vein cu- a; gastral segment I, tergum with well-de- veloped lateral carina at base, visible in most specimens, sternum with short lon- gitudinal carina extending from base; api- cal tergum evenly convex, apex truncate or somewhat rounded; short and apically rounded; genital capsule (Figs. 20, 21): paramere slender and tapering apically; volsella with rounded medial lobe, inner margin with numerous small denticles; digitus elongate and apically acute. Distribution. — This genus is known only from the vicinity of Glamis, in the Algo- dones Dunes, Imperial Co., California. Etymology. — The name is a nonsense combination of letters, and is assumed to be feminine. Type species. — Sedornaya glamisensis n. sp. Discussion. — No one genus appears to be the sister group of Sedomaya. It belongs to the group of genera characterized by the presence of a stridulatory structure on the forecoxae, elongate digitus and the first gastral sternum with a short longitu- dinal carina extending posteromedially from the base. Sedomaya, Dolichetropis, Acanthetropis and Colocistis all have a ven- tral clypeal bevel below the apex. How- ever, unlike the latter three genera Sedo- maya has highly reduced wing venation, having only two small forewing submar- ginal cells and one discoidal cell. Sedornaya glamisensis Kimsey and Wasbauer, new species (Figs. 5, 10, 13, 20, 21) Description of male. — Body length 5-7 mm; face (Figs. 10, 13); clypeus medially convex, apex slightly trilobate with medial lobe projecting anteriorly in profile; apical truncation 0.9-1.0 midocellus diameters wide; interantennal distance 0.4-0.5 mido- cellus diameters; flagellomere I-II 2.9-3.1 times as long as broad; facial and thoracic punctures small, shallow and widely spaced, 4-6 puncture diameters apart; forewing with one submarginal cell (Fig. Volume 8, Number 1, 1999 73 5); abdominal segments appearing im- punctate, integument finely shagreened; genital capsule (Figs. 20, 21). Body color pale reddish brown; fore and hindwing venation yellow; wing membrane slightly yellow tinted. Type material. — Holotype male: Califor- nia, Imperial Co., Glamis, 23 April 1972, M. Wasbauer, blacklight (DAVIS). Para- types: twenty-five males (DAVIS); two — same data as holotype; twenty-two — 3 mi n Glamis, 15-16 Sept. 1972, M. Wasbauer and A. Hardy; four— 10 Sept 1974, M. Wasbauer and R. McMaster. Eti/mologi/. — This species is named after the vicinity of the collection sites in and around Glamis, on the edge of the Algo- dones Dunes. Discussion. — As with Brachymaya mexi- cana, species distinctions within this genus will probably be based on flagellar dimen- sions, the size of the midocellus and its distance to the ocular margin, other facial dimensions and perhaps the configuration of the clypeus. ACKNOWLEDGMENTS Our thanks to Karl Krombein for his help with the manuscript, and to the fine collections support and helpful collection managers who made this study possible. LITERATURE CITED Kimsey, L. S. & M. S. Wasbauer. 1998. Revision of the American tiphiid genus Qiicniaya Pate, journal of Hymeiioptcra Rc^cnrcli 71:38-47. Mickel, C. E. and K. V. Krombein. 1942. Glyptometopa Ashmead and related genera in the Brachycisti- dinae with descriptions of new genera and spe- cies. Aincriam Midland Naturalist 28:648-679. Wasbauer, M. S. 1958. A new genus of brachycisti- dine wasps. Pan-Pacific Entoniologtst 34:139-142. Wasbauer, M. S. 1966. Revision of the male wasps of the genus Bracliyctstis in America north of Mex- ico. University of California Publicatioiif in Ento- mology 43:1-96. Wasbauer, M. S. 1968. New genera of male Brachy- cistidinae with a redescription of Brachycistelhis Baker and a key to North American genera. Pan- Pacific Entomologist 44:184-197. J. HYM. RES. Vol. 8(1), 1999, pp. 74-97 Deceptive Similarity in Army Ants of the Genus Neivamyrmex (Hymenoptera: Formicidae): Taxonomy, Distribution and Biology of N. califomicus (Mayr) and N. nigrescens (Cresson) Philip S. Ward Department of Entomology, University of California, Davis, California 95616, USA Abstract. — The army ant Neivamyrmex califomicus (Mayr) is demonstrated to be a distinct species, endemic to California and adjacent Baja California, whose range overlaps that of N. nigrescens (Cresson), with which it has been previously confused. N. nigrescens is widespread throughout the southern Nearctic region, and shows extensive morphological variation in shape, size and sculpture. In the northwestern extremities of its range (i.e., north-central California, Nevada, Utah, and southwestern Colorado) N. nigrescens is convergently similar to N. califomicus in certain as- pects of worker morphology: workers lack the densely punctulate head sculpture typical of this species, and in their more shiny appearance they are superficially similar to workers of N. cali- fomicus. Many records of "N. califomicus" (and the description of its supposed queen) actually refer to this shiny morph of N. nigrescens. That the shiny form is conspecific with, although partially differentiated from, other populations of N. nigrescens is indicated by the occurrence of intermediate populations in a relatively narrow transition zone in the San Gabriel Mountains of southern California. No intermediates have been observed between N. nigrescens and N. califor- nicus. The latter species appears to be more closely related to N. opacitborax Emery than to N. nigrescens. In California N. nigrescens and N. califomicus raid the nests of other ants, including Messor andrei, Pheidole califomica, P. hyatti, Solenopsis molesta, and Formica moki. Field observations, combined with indirect evidence from the contents of ant nest middens, suggest considerable overlap in prey choice and habitat use. Both N. nigrescens and N. califomicus are commonly sym- patric in southern California (where N. nigrescens retains its distinctive granular-punctulate body sculpture), but less frequently so in northern California where they are more similar in appearance, and where N. califomicus tends to be confined to more mesic, coastal areas than N. nigrescens. Army ants in the genus Neivamyrmex given that the reproductive females are a frequent component of ant commu- (queens) of these and other army ants are nities in tropical and warm temperate re- entirely wingless and have quite limited gions of the New World. The group has powers of dispersal (Gotwald 1995), a fac- received considerable attention from tax- tor that favors population differentiation, onomists (e.g., Smith 1942; Borgmeier The present contribution is concerned 1953, 1955, 1958; Watkins 1976, 1982, with clarifying the taxonomy of Neivmnyr- 1985), with the resulting recognition of mex califomicus (Mayr) and several closely about 120 species (Bolton 1995). Taxonom- related species. Earlier treatments of these ic problems persist, however, partly be- species are shown to be misleading. Sim- cause many names are based on either pie morphometric analyses help to resolve workers or males only — and the names for and diagnose N. califomicus and N. iii^res- the two castes are not yet cross-referenced cens (Cresson), two species whose taxo- and synonymized — but also because there nomic distinctness was previously called is substantial and often confusing intra- into question (Watkins 1985). Both these specific variation in morphology (Smith species — but especially N. nigrescens — 1942). Such variation is not unexpected show considerable variation in integu- Volume 8, Number 1, 1999 75 ment sculpture, and this phenomenon misled earlier investigators about species limits. The range of Neivamyrmex californicus is more restricted than previously thought, the species being confined to the Califor- nian floristic province (California and ad- jacent northern Baja California), where it is broadly sympatric with N. uigrescens. The latter is shown to be a highly poly- typic species with an extensive transcon- tinental distribution. MATERIALS AND METHODS Specimens were examined in the fol- lowing collections: CASC California Academy of Sciences, San Francisco, California, U. S. A. CDAE California Department of Food & Agriculture, Sacramento, Califor- nia, U. S. A. JTLC John T. Longino Collection, Olympia, Washington, U. S. A. LACM Natural History Museum of Los Angeles County, Los Angeles, California, U. S. A. MCZC Museum of Comparative Zoolo- gy, Harvard University, Cam- bridge, Massachusetts, U. S. A. MHNG Museum d'Histoire Naturelle, Geneva, Switzerland NHMV Naturhistorisches Museum, Vien- na, Austria RAJC Robert A. Johnson Collection, Tempe, Arizona, U. S. A. UCDC Bohart Museum of Entomology, University of California, Davis, California, U. S. A. USNM National Museum of Natural His- tory, Smithsonian Institution, Washington, D. C, U. S. A. Other collections cited are: ANSP Academy of Natural Sciences, Phil- adelphia, U. S. A. MCSN Museo Civico di Storia Naturale, Genoa, Italy The following measurements and indi- ces were used for workers and queens. All measurements were taken at 50 X magni- fication with a Wild MSA microscope and a pair of Nikon stage micrometers wired to a digital readout. Measurements were recorded to the nearest thousandth of a millimeter. HW Maximum width of head, mea- sured in full-face (frontal) view HL Maximum length of head in fron- tal view, from the midpoint of a line drawn across the posterior margin to the anteriormost point on the clypeal margin. This ex- cludes the thin, lamelliform clypeal apron, which projects forward from the clypeal margin and may be partly hidden by the mandibles. MFC Minimum distance between the frontal carinae SL Length of the scape, excluding the basal neck WL Weber's length: length of the me- sosoma, as seen in lateral view, from the anterior pronotal margin (excluding the collar) to the pos- terior extremity of the metapleu- ron PL Length of the petiole, measured in lateral view along the long axis of the petiole, from the anterior mar- gin (excluding the short peduncle) to the posterior extremity PH Petiole height, measured at right angles to PL, from the summit of the petiole to the petiolar venter, excluding the anteroventral tooth (if present) DPW Maximum dorsal width of the pet- iole, measured in dorsal view PPW Maximum dorsal width of the postpetiole, measured in dorsal view MTL Length of the metatibia, excluding the basal condyle (Fig. 3) QGL Length of the gaster (queen caste only), measured in lateral view from the posterior end of the gas- 76 Journal of Hymenoptera Research ter to the anterior extremity of ab- dominal segment 3, excluding the helcium. This measurement was taken only on non-physogastric queens. CI Cephalic index: HW/HL SI Scape index: SL/HW FCI Frontal carinal index: MFC/ HW PLI Petiole length index: PH/ PL PWI Petiole width index: DPW/ PL MTI Meta tibial index: MTL/HW Neivamyrmex califomictis and allied species. Among the North American army ants of the genus Neivamyrmex three species — N. californicus, N. nigrescens and N. texan- us — can be diagnosed in the worker caste by having a mandible whose basal margin rounds gradually into masticatory margin, without an angular junction (Figs. \, 4, 7). The species have many other traits in com- mon, including: moderately large size (HW 0.60-1.42); well developed clypeal apron; relatively prominent ocellus-like eye with convex surface; weakly devel- oped occipital lobes; lamellate and ven- trally directed lower pronotal flange; pet- iole longer than high (PLI 0.68-0.96) and markedly longer than wide (PWI 0.52- 0.74); and dense punctulate sculpture cov- ering part of the body but without con- spicuous overlying coarse impressions or rugae. These taxa are part of a larger group of species, including N. chamelensis Watkins, N. cornutus Watkins, N. manni (Wheeler), N. opacithorax (Emery) and N. stimidirnsti (Norton), that can be placed together, on the basis of similarities in worker mor- phology and male genitalia, in an assem- blage of Neivamyrmex species termed "Gruppe VI" by Borgmeier (1955: 490). Although the three species treated in this paper are evidently closely related, their exact phylogenetic relationship to one an- other and to these other taxa remains to be resolved. List of species considered here, with known castes (w = worker, m = male, q = queen) and distribution: californicus (Mayr 1870) (w) U. S. A.: Cahfornia Mexico: Baja California nigrescens (Cresson 1872) (w, q, m) U. S. A.: California, Nevada, Utah, Ari- zona, New Mexico, Colorado, and east-central United States Mexico: Baja California, Baja California Sur, Jalisco, Nayarit, Oaxaca, San Luis Potosi, Sonora, Tamaulipas texanus Watkins 1972 (w, q, m) U. S. A.: Arizona, New Mexico, Colo- rado, and east-central United States Mexico: Chihuahua, Durango, Hidalgo, Jalisco, Nuevo Leon, San Luis Po- tosi, Sonora SPECIES ACCOUNTS Neivamyrmex californicus (Mayr 1870) (Figs. 1-3, 10) Eciton cnlifornicum Mayr 1870: 969. Nineteen syntype workers, San Francisco (Schaufuss) (NHMV) [examined]. One syntype worker here designated lectotype. Ecitoii {Acainatus) cnlifornicum Mayr; Emery 1894; 182. Eciton {Acainntiis) cnlifornicum var. obscurn Forel 1914: 265. Two syntype workers, Vista, Cal- ifornia (E. Hindle) (MHNG) [examined]. One syntype worker here designated lectotype. Synonymy by Borgmeier 1955: 517; here con- firmed. Eciton {Ncivnnn/rmcx) cnlifornicum Mayr; Smith Figs. 1-12. Neivamyrmex workers, full-face view of head (1, 4, 7), lateral view of mesosoma, petiole and postpetiole (2, 5, 8), lateral view of metatibia (3, 6, 9), and dorsal view of propodeum, petiole and postpetiole (10-12). 1-3, 10: N. californicus, San Francisco, California, lectotype worker; 4-6, 11: N. nigrescens. shiny form, 2km SB Mt. Vaca, California; 7-9, 12: N. nigrescens, typical form with opaque head, Sevilleta NWR, New Mexico. Fig. 3 indicates measurement of MTL. Volume 8, Number 1, 1999 r? 3(fl)T V MIL y w 78 Journal of Hymenoptera Research 15 1.0 mm Figs. 13-16. Neivcinn/riucx males, dorsal view of head (13, 14) and lateral view of left paramere (15, 16). 13, 15: N. nigrescens, Ash Mountain Kaweah Power Stn.#3, Tulare Co., California; 14, 16: Neivamyrmex species, probably opncithornx, same locality. 1942: 560. First combination in subgenus Nei- vamyrmex, but the material examined and de- scribed by M. R. Smith was N. opncithornx not N. californicus. Neivami/rmex californicus v. obscurus (Forel); Borgmeier 1953: 8. Neivamyrmex californicus (Mayr); Borgmeier 1953:' 11. Neivamyrmex californicus (Mayr); Borgmeier 1955: 517. Neivamyrmex californicus (Mayr); Watkins 1972: 363 (part). Description of queen (p. 364) is that of N. nigrescens, not N. californicus. Neivamyrmex californicus (Mayr); Watkins 1985: 482 (part). Key (p. 482) and distribution map (figure 4, p. 500) refer partly to N. californicus and partly to N. nigrescens. Worker measurements. — (n = 28). HW 0.63-1.10, HL 0.69-1.11, WL 1.03-1.67, MTL 0.68-1.18, CI 0.86-1.00, FCI 0.033- 0.061, SI 0.68-0.80, MTI 1.05-1.14, PLI 0.78-0.88, PWI 0.62-0.73. Worker description. — Body of moderate size (see HW, HL and MTL measure- ments) and somewhat compact (see plot of WL on HW; Figs. 19, 23); head broad, CI approaching 1.00 in largest workers, i.e., those in which HW and MTL > 1.00 mm; mandible with basal margin rounding gradually into masticatory margin (Fig. 1); masticatory margin with a small tooth at the terminus of this rounding, followed by 1^ denticles (tending to increase in size), then a more prominent tooth midway along the margin; distal portion of masti- catory margin generally edentate (a small denticle or two may follow the mid-point tooth) except for the acute apical tooth; an- terior margin of torulus separated from anterior margin of clypeus (ignoring the thin diaphanous clypeal apron) by about 0.2 X the diameter of the torulus; frontal carinae moderately well separated (MFC 0.022-0.061), diverging anteriorly, and protruding very slightly (largest workers) or not at all (most workers) beyond the anterior clypeal margin, when the head is seen in frontal view; anterior clypeal mar- gin more or less straight (weakly convex Volume 8, Number 1, 1999 79 in smallest workers, and slightly concave in largest workers); clypeal apron relative- ly well developed, its anterior margin gen- erally slightly convex or subangulate; clypeal apron extending anteromedially beyond the clypeal margin proper by a distance equal to 0.4-O.5X the torulus di- ameter; each compound eye consisting of a single convex ommatidium, breaking the surface of the head, its maximum diame- ter approximately 0.06-0.08 mm (6-10% of head width); scapes of moderate length, exceeding the eye when held back against the head (SI 0.68-0.80; SI2 0.67-0.75) (see also plot of SL on HW and SI on MTL; Figs. 18, 22, 26); posterior margin of head usually concave, in frontal view; occipital lobes weakly developed, not protruding conspicuously when the head is seen in frontal view; anterior pronotum descend- ing gradually towards the collar, trans- verse ridge weakly developed; lower pronotal flange thin, lamellate, directed more or less ventrally; dorsal profile of mesosoma rather flat, and dorsal face of propodeum only slightly depressed below the level of the mesonotum (Fig. 2); dorsal face of propodeum rounding into, and subequal in length to, the declivitous face; latter flat to weakly concave, in profile; legs relatively short, MTL/HW ( = MTI) < 1.15; petiole short, high, and with short vertical anterior face, followed by a more or less evenly convex dorsal surface (as seen in profile; Fig. 2), or with a slightly steeper posterodorsal than anterodorsal slope; in dorsal view, petiole subrectan- gular, but with somewhat convex sides, about 1.5 times longer than wide; antero- ventral process of petiole a thickened transverse shelf, in lateral view appearing as a relatively small, blunt tooth directed anteroventrally; a similar, less protrusive structure at the anteroventral extremity of the postpetiole; in dorsal view postpetiole subtrapezoidal (Fig. 10), with straight, di- verging sides, broadest posteriorly, and slightly broader than long. Upper surface of mandible finely and densely striate with scattered punctures, subopaque, lat- eral surface of mandible and area imme- diately preceding the masticatory margin smooth and shiny with scattered punc- tures. Head largely smooth and shining, with numerous piligerous punctures sep- arated by several to many diameters; in larger workers (and in more southern populations) the punctures may be coarser and denser, and parts of the intervening shiny integument dulled very slightly by weak reticulation. Mesosoma densely punctulate, opaque, but in smaller work- ers the sculpture weakened laterally on the pronotum (which becomes finely retic- ulate and sublucid) and replaced by smooth shiny areas on the dorsum of the promesonotum. In workers of all sizes, in- cluding heavily sculptured large individ- uals, the center of the mesonotum is near- ly always smooth and shiny, with a few larger piligerous punctures, and contrasts with the predominantly punctulate and opaque dorsal face of the propodeum (ex- ceptions include some heavily sculptured workers from southern California in which the mesonotum center is weakly re- ticulate and sublucid, but still contrasts with the opaque dorsal surface of the pro- podeum; and small, shiny workers from northern California in which even the pro- podeum dorsum loses its punctulate sculpture centrally, so that contrast be- tween the mesonotum and the propo- deum is lessened). Petiole laterally reticu- lo-punctate and subopaque to sublucid, petiolar dorsum mostly shining, but with variable traces of punctulate reticulation; postpetiole and gaster smooth and shiny, with scattered piligerous punctures. Long, fine, golden pilosity conspicuous on body and appendages (scape, funiculus, legs), mostly suberect to subdecumbent. Body light castaneous brown to deep reddish- brown, tending towards a lighter yellow- brown on the postpetiole and gaster. Man- dible medium to dark brown, usually con- trasting somewhat with the lighter head color. 80 Journal of Hymenoptera Research Queen. — Unknown. The "N. califoniicus" queen from Davis, California, described by Watkins (1972), is that of the shiny form of N. nigrescens. A single queen, in rather poor condition, from Monterey, California (31. v. 1963; leg. Roy Johnson; USNM), could be either N. opacithorax or N. californiciis. It is relatively small in size (HW 1.65, WL 2.68, MTL 1.33), with short scapes (SI 0.46), rounded occipital lobes, short legs (MTI 0.81), and a long gaster (QGL/WL = 2.34). The short appendages and elongate gaster are features that dis- tinguish N. opacithorax queens from those of N. nigrescens, but they might also be characteristic of N, californiciis. There are apparently no worker specimens associ- ated with the Monterey queen, leaving its specific identity uncertain. Male. — Unknown. In LACM and CASC there are a series of Neivamyrmex males, collected in California but unassociated with workers, that appear to be N. opaci- thorax on the basis of head shape and (es- pecially) male genitalia (Figs. 14, 16). The proximal flange on the ventral margin of the paramere would seem to be particu- larly characteristic of that species. Never- theless, until worker-associated males of N. californiciis are discovered, one cannot exclude the possibility that some of these males (from Contra Costa and Tulare Counties) represent N. californiciis. Comments. — Workers of N. californiciis can be distinguished from those of N. ni- grescens by their more compact body, shorter legs and scapes, and broader head and petiole (Figs. 1-3). The simplest quan- titative diagnostic is the metatibial index (MTI = MTL/HW) which ranges from 1.05 to 1.14 in N. californiciis (n = 28) com- pared to 1.16 to 1.52 in N. nigrescens (n = 89). A bivariate plot of the relevant mea- surements (MTL and HW) demonstrates the distinction (Fig. 17). Other differences between N. californiciis and N. nigrescens are captured by plots of scape length (SL) on head width (HW) and Weber's length (WL) on head width (Figs. 18, 19, 21, 22). When samples from all populations of N. nigrescens are considered (i.e., including workers from the Southwest and from eastern United States) there is a slight overlap in the distribution of points (Figs. 18, 19); but when confining the compari- sons to populations of N. nigrescens that are sympatric with N. californiciis (i.e., populations from California) these and other bivariate plots produce non-overlap- ping clouds of points (see Figs. 21-28). Be- cause of allometry, neither the cephalic in- dex (CI = HW/HL) nor the scape index (SI = SL/HW) are diagnostic in them- selves, but when plotted against the me- tatibia length (MTL) a sharp distinction is seen between the two species (Figs. 25- 26). Thus, within a given size class (as measured by MTL) there is no overlap in CI or SI; but small workers of N. califor- niciis have scape and cephalic indices that overlap with large workers of N. nigres- cens. Workers of N. californicus also have the dorsal face of the propodeum less de- pressed below the level of the mesonotum than in N. nigrescens (compare Figs. 2 and 5), and this difference is diagnostic. Less tangible are differences in body sculpture: the dense punctulate sculpture that is so prominent in most populations of N. ni- grescens and that imparts a granular ap- pearance to the head and mesosoma is much less developed in N. californicus, such that the head, mesonotum, and post- petiole are largely smooth and shiny (cov- ered only with scattered piligerous punc- tures). But a reliance on sculptural differ- ences led to the past confusion between these two species: populations of N. ni- grescens that are broadly sympatric with N. californicus in central and northern Cal- ifornia also have weakened body sculp- ture and are superficially similar to those of N. californicus. On average, N. californi- cus is still the shinier of the two (see dis- tinctions in key couplet 9A below) but the differences are subtle. The contrasts in body shape and leg length documented Volume 8, Number 1, 1999 81 17 ■e t.2 E, 18 ■ califomlcus A opacrthorax O nigrescens o texanus «gp 05 06 07 08 09 1 11 12 13 14 15 HW(mm) 04 05 06 07 08 09 1 11 12 HW(mm) 13 14 15 19 IS t 16 E, _J 5 14 12 o ■ califomicus A opacrthorax □ 0 ; D nigrescens 0 texanus o o D : ^^■'^ : # ■ .t A 20" ■ ■ califomicus ^ opacrthorax ■ ■■ '•■"■. A 0:^5 <" • A ^^ a'^ ■ ■ 02 ' zA I A ^A ■ 0 15 A 04 05 06 07 08 09 1 HW(mm) II 12 13 14 15 04 05 06 07 0 8 0 9 1 11 12 HW(mni) 21' 1 L> ■ califomicus 1 b - o nigrescens (CA) D - a - a P a a a D ° a ° f ■ ■ 09 08 07 ■ 22' 0 5 0 6 0 7 0 8 0 9 HW(mm) ■ calrfomicus n nigrescens (CA) 0 ; OdoP D * ■ D ■ 1 1 12 05 06 07 08 09 1 11 1.2 HW(lTim) Figs. 17-22. Bivariate plots of various metric measurements in workers ot tour species of Neivamyniifx. 82 Journal of Hymenoptera Research l 3„ ■ califomicus D nigrescens (CA) P P q, a T] a a 5 14 : o ; B A ' ; B. 24' ■ calffomicus D nigrescens (CA) Q " G° ° G ■ G G : °G s G P G a G CP^^GJ"- a ■ B ° ■ 06 07 08 09 1 11 12 HW(mm) 05 06 07 08 09 1 11 1,2 HW(mm) 25 5 09 ■ ■ califoniicus D nigrescens (CA) ; ■ ■ ■ € a ; S ^ G G G D ■;■ ' G a G ^ p° D G P P n ° P ft" ^C?° GP P P 26 G ^ G 0 P ° D p p ° D nigrescens (CA) 095 0 °D n G 09 4d D f^"*; ° P r, ° 085 °G 0° G ° 08 t a ° P "... ° P 075 ■\ 1 ■ cb a ■ ^ Of s a a 0,7 0,8 0.9 1 11 12 1.3 1.4 1.5 16 MTL (mm) 0,6 07 08 09 1 11 12 13 14 1.5 1.6 MTL (mm) 27' ■ califomicus D nigrescens (CA) a - ■ ■ ■ I ■•''1 CD ■ .' ° □ □ a a □ f * D D 28 (^ - : p Q - p ^,« ■ calrfomicus : G : G G D nigrescens (CA) 14 1 35 : pP „a ° i GO ""f^ r on G dWaG, P 125 S 1 2 : 0 1 15 1 1 105 ■ ■«V;.- 06 07 08 09 1 11 12 13 14 15 16 MTL (mm) 0 16 Olfl 0 2 0 22 0 24 0 26 0 28 0 3 0 32 OPW/MTL Figs. 23-28. Bivariate plots of metric measurements and indices in workers of Nctzuiw\fnm'X culifonucui^ and N. nigrescens (California populations only). Volume 8, Number 1, 1999 29 83 ; O nigr»scens (CA, punctate form) ■ T nigrescens (CA, shiny form) 0 D D ▼ 30 0 5 0 6 0 7 0 8 0 9 W 07 . : D nigrescens (CA. punctate form) ■ T nigrescens (CA. shiny form) a C D T : o D »» \ T a o T 0.5 06 07 Oe 09 HW{mm) 31 u ^ A texanus D nigrescens {SW. East) A 015 A A A 01 A ^ A 1 0 D 0.05 AD Do D i^ □ D 0 D D 32' A texanus . D nigrescens (SW, East) A AA * A A A \ A O A A* tU A ^O °'° 0.4 0.5 0 6 0 7 0 8 0 9 1 11 12 13 1.4 1.5 HW(mm) 0.6 0-8 12 1,4 1,6 1.8 o J> 1 1 A texanus D nigrescens (SW. East) A 0 15 lo, 005 A A A A ^ A ^ O^ i a rp ° ° DO 3 34 UD A lexanus 0 55 D nigrescens (SW, East) A A '. OS : A A *° 0 45 " 04 A to ■S" ACA° y n* /. 3 025 □ 05 06 07 08 09 1 11 12 13 14 15 HL (mm) 05 06 07 08 09 1 11 12 13 14 15 HL (mm) Figs. 29-34. Bivariate plots of metric measurements and indices in workers of Ncivntmfrmex tcxauus and various populations oi N. fii<^resce}is. 84 Journal of Hymenoptera Research above are niore reliable for distinguishing these two species. N. californkus differs more obviously from N. texanus. Like N. nigrescens, work- ers of N. texanus have longer legs (MTI 1.18-1.48), longer scapes (SI2 0.75-0.89), and a more gracile body than those of N. californiciis. They also average larger in size, have more widely separated frontal carinae (FCI 0.066-0.125 compared to 0.033-0.061 in N. californiciis), and have a consistently opaque, densely punctulate head and mesosoma such that they are unlikely to be confused with N. californicus workers. As far as known, the ranges of the two species do not overlap (Fig. 36). The bivariate plots of measurements (Figs. 17-19) show that in many respects N. californicus is more similar to N. opaci- thorax than to N. nigrescens or N. texanus. N. opacithorax can be distinguished from N. californicus by the angular basal margin of the mandible; shorter clypeal apron; narrower petiole (Fig. 20); and different pattern of body sculpture (side of prono- tum usually smooth and shiny, and con- trasting with the rugulose-punctulate me- sosoma dorsum). Material examined. — (CASC, CDAE, JTLQ LACM, MCZC, MHNG, NHMV, RAJC, UCDC, USNM) MEXICO Baja California: 28km E Ensenada, 750m (P. S. Ward). UNITED STATES California Cpntrn Costa Co.: 9km ENE Danville, 490m (P. S. Ward); El Dorado Co.: 14km NW Shingle Springs, 340m (P. S. Ward; G. C. Snell- ing); Los Angeles Co.: Arcadia (c.u.); La Verne (A. C. Oberle); Los Angeles (A. Mallis & ]. Schwartz); Mon- terey Co.: Salinas (B. Oliver c& J. Bunch); Orange Co.: Irvine Park (A. Mintzer); Laguna Hills (R. J. Hamton); Limestone Canyon, El Toro Rd., I.8mi E Cooks Cor- ner (A. Suarez); Tonner Canyon (W. P. Mackay); Riv- erside Co.: Riverside (K. Cooper; K. W. Cooper; E. I. Schlinger); Temecula (A. Suarez); San Bernardino Co.: Chino Hills State Park (G. C. Snelling et al.); Chino Hills State Park, 700 ft. (R. A. Johnson); San Diego Co.: 5mi. NE Poway, 600m (M. S. Trepanier); Camp Pen- dleton (J. H. Hunt); Chula Vista, E end, 160m (P. S. Ward); Chula Vista, 70m (A. Suarez); Elliott Reserve, 150m (A. Suarez); La Jolla (c.u.); Mt. Laguna, MSP site, 6050 ft. (J. H. Hunt); National City (K. Ross); San Diego (c.u.); Vista (E. Hindle); San Francisco Co.: San Francisco (Schaufuss); San Mateo Co.: Jasper Ridge, 150m (K. G. Human; N. J. Sanders); Santa Barbara Co.: 9km N Goleta, 490m (P. S. Ward); Cachuma Saddle, 1100m (J. Longino); Santa Clara Co.: 11km S Palo Alto, 490m (P. S. Ward); Santa Cruz Co.: Santa Cruz (K. Brown). Neivamyrtnex nigrescens (Cresson 1972) (Figs. 4-9, 11-12) Lahidiis nigrescens Cresson 1872: 194. Holotype male, Bosque Co., Texas (Belfrage) (ANSP) [not examined]. Eciton nigrescens (Cresson); Dalla Torre 1893: 5. Eciton (Acamatus) schmitti Emery 1894: 183. Syn- type workers, Doniphan, Missouri (Pergan- de) (MSCN) [not examined], MCZC, USNM [examined]. Synonymy by M. R. Smith 1938: 160. Eciton (Labidus) nigrescens (Cresson); Emery 1895: 258. Eciton (Acnrnntus) nigrescens (Cresson); Emery 1900: 187. Eciton sitmichrnsti; Wheeler (nee Norton) 1900: 564. Description of queen (as "£. sumichras- ti"). Eciton (Neivnmi/rmex) nigrescens (Cresson); Smith 1942: 550 (part). Description of worker (part), queen and male. Eciton (Neivamyrmex) californicum; Creighton (nee Mayr) 1950: 70 (part). Neivamyrmex nigrescens (Cresson); Borgmeier 1953: 6. Neivamyrmex nigrescens (Cresson); Borgmeier 1955: 494 (part). Description of worker (part) and queen. Male (p. 496) is that of N. texanus (Watkins 1972). Neivamyrmex sp. e; Borgmeier 1955: 531. De- scription of male. Neivamyrmex nigrescens (Cresson); Watkins 1972:' 358. Neivamyrmex californicus; Watkins (nee Mayr) 1972: 363 (part). Description of queen (as "N. californicus"). Neivamyrmex nigrescens (Cresson); Watkins 1985:482. Neivamyrmex californicus; Watkins (nee Mayr) 1985: 482 (part). Worker tneasiirenients. — (n = 89). HW 0.60-1.31, HL 0.71-1.31, WL 1.10-2.12, MTL 0.76-1.76, CI 0.80-1.00, FCI 0.030- 0.106, SI 0.75-1.01, MTI 1.16-1.52, PLI 0.68-O.91, PWI 0.52-0.71. Volume 8, Number 1, 1999 85 Worker diagnosis. — Moderately large body size (see HW, HL and MTL mea- surements); mandible with basal margin rounding gradually into masticatory mar- gin (Figs. 4, 7); frontal carinae moderately well separated (MFC 0.021-0.138); clypeal apron well developed and, in all but the largest workers, produced anteromedially by an amount subequal to, or greater than, the minimum distance between the frontal carinae (MFC); scapes relatively long, SI > 0.69 (see also plot of SL on HW and SI on MTL; Figs. 18, 22, 26); occipital lobes weakly to moderately developed; anterior margin of pronotum with transverse ridge generally well developed; dorsal (= basal) face of propodeum conspicuously de- pressed below the level of the promesono- tum (Figs. 5, 8), and rounding into the de- clivitous face, the latter more or less flat (or weakly concave) in profile; legs rela- tively long, MTL/HW ( = MTI) > 1.15; pet- iole subrectangular, somewhat variable in shape (see PLI and PWI values), but al- ways longer than high or wide (Figs. 5, 8, 11, 12). Head and mesosoma typically densely punctulate, and having an opaque, granular appearance; populations from northwestern portions of the species' range, however, have the sculpture much weakened such that the head is partly smooth and shining, with scattered pili- gerous punctures and variable amounts of finer reticulate sculpture that partly dulls the sheen; and in this "shiny morph" the mesosoma is partly sublucid, although with at least weak reticulate-punctulate sculpture on most surfaces. In all popula- tions sculpture weakened on the surface of metapleural gland bulla, such that the lower half or more is smooth (or weakly reticulate) and conspicuously shiny. Peti- ole densely punctulate, subopaque, post- petiole tending to be more lightly sculp- tured. Body varying from light castaneous brown to dark reddish-brown, the post- petiole and gaster usually lighter than the rest of body. Queen diagnosis. — Eye distinct, consist- ing of single convex ommatidium. Head as broad as, or slightly broader than, long (CI 0.96-1.02, n = 6). Occipital corners generally angular and projecting, but may be weakly angulate or rounded. Pronotum without posterior dorsolateral projection. Propodeum (and sometimes also meso- notum) with a median longitudinal im- pression. Metatibial index (MTI) 0.89-1.07 (n = 6). Petiole subquadrate, slightly broader than long (PWI approximately 1.1-1.2), with a vertical anterior face that rounds into a flat dorsal face, as seen in lateral view. Petiole width much less than the length of the metatibia, DPW/MTL 0.58-0.72 (n = 6). Length of gaster (non- physogastric) less than twice WL (QGL/ WL = 1.70-1.90, n = 5). Male diagnosis. — Mandibles broad, spat- ulate (not sickle-shaped). Ocelli moderate in size and separated from the upper mar- gin of the compound eye by a distance greater than twice the diameter of the me- dian ocellus. Prominent transverse swell- ing above antennal fossa, discerned most clearly in dorsal view (Fig. 13). Setae on venter of petiole typically short, whitish, and slanted posteroventrally, but in some western populations the setae are longer, golden, and suberect. Gaster typically black or blackish-brown, often reddish- brown in western populations. In profile, paramere (= stipes) linear subrectangular, with an oblique (anterodorsal to postero- ventral) posterior margin, and a straight ventral margin, not produced anteroven- trally (Fig. 15); posterodorsal extremity with a low, triangular projection (Watkins, 1985, Plate 9, fig. 4), tending to become ob- solete in western populations (Fig. 15); volsella unforked, with long, slender up- turned apex; aedeagus ( = sagitta) with an apically upturned posteroventral process, just exceeding the posterodorsal process (in posterior reach). Comments. — The characteristics that dis- tinguish workers of N. nigrescens from those of N. californicus and N. texanus are discussed under those respective species. 86 Journal of Hymenoptera Research N. nigrescens differs from N. opacithorax by the shape of the worker mandibles, by the heavier body sculpture, especially on the side of the pronotum, by the longer scapes, mesosoma and legs (Figs. 17-19), and by the more strongly convex profile of the promesonotum. A remaining question concerns the sta- tus of the nigrescens-Mke populations with shiny worker heads, that are superficially similar to those of N. californicus. This "shiny morph" of N. nigrescens is found in north-central California, parts of the Great Basin, and in upper sections of the Colo- rado River drainage, while the more "typ- ical" form, with densely punctulate and opaque worker head, is distributed widely from southern California, Arizona and ad- jacent regions of Mexico to southeastern United States (Fig. 37). What happens in zones of contact between the two forms? Records are too sparse to answer this question for the Colorado River basin, but collections from southern California reveal a zone of intergradation centered on the north side of the San Gabriel Mountains. In fact patterns of sculpturation are inter- mediate in samples taken from this region, so that the distinction between the two "forms" becomes quite arbitrary. For this reason it seems clear that they must be treated as conspecific, although the pat- terns of distribution are suggestive of a previous period of isolation followed by secondary contact and introgression. Material examined. — (1) Typical form with opaque, densely punctulate head (CASC, CDAE, LACM, MCZC, RAJC, UCDC, USNM) MEXICO Baja California: 28km E Ensenada, 750m (P. S, Ward); 31.7mi WNW Bahia de los Angeles (Hardy, Andrews & Giuliani); 6mi SE Laguna Cha- pala (A. E. Lewis); Baja California Sur: 12mi S Santa Rosalia (Michelbacher & Ross); 15mi S San Domingo (Ross & Bohart); 20mi W La Paz (E. L. Sleeper); 26km NW Santa Rosalia (R. A. Johnson); Coyote Cove, Con- ception Bay (Michelbacher & Ross); Isla San Jose, Imi S Punta Colorada [as "Punta Colorado"] (J. T. Doy- en); San Hilario (E. L. Sleeper); San Ignacio, 140m (M. Bennett); Jalisco: 3mi SE Plan de Barrancas (F. D. Parker &. L. A. Stange); Nayarit: hlas Tres Marias: Isla Cleofas (R. R. Snelling); Isla Magdalena (R. R. Snell- ing); San Luis Potosi: Cd. Valles, El Bariito (J. F. Wat- kins); Sonora: 2km N Bahia de la Cruz, Isla Tiburon, 10m (P. S. Ward); 2km SW Punta Narragansett, Isla Tiburon, 5m (P. S. Ward); 37mi N Hermosillo, 1700 ft. (R. R. Snelling); 5mi S Cananea (V. D. Roth); 8km N Desemboque de los Seris (R. A. Johnson); Los Hor- cones, Rte. 16, 4km E La Colorada (B. Bestelmayer); Tamaulipas: Cd. Victoria (J. F. Watkins). UNITED STATES Alabama: Dallas Co.: Selma (W. H. Patton); Jefferson Co.: Birmingham (R. D. Jordan); Lauttcrdale Co.: Florence (W. Cloyd; F. Moore); Latvrence Co.: King Cove, Bankhead Natl. Forest (E. O. Wilson); Mobile Co.: Kushla (A. C. Sturtevant); Spring Hill [as "Springhill"] (W. S. Creighton); Mor- gan Co.: Decatur (Murphree); Arizona: Cochise Co.: 2mi NE Portal (G. D. Alpert); Chiricahua Mts. (W. & E. MacKay; J. F. Watkins); Chiricahua Mts., 5000 ft. (J. F. Watkins); Chiricahua Mtns., 13mi NW jet. Rte. 80 on FSR 74, 5850 ft. (S. P. Cover); Chiricahua Mtns., Cave Creek Canyon, SW Res. Stn., 5400 ft. (S. P. Cov- er); Copper Canyon, 8. Imi SE Sunnyside, 5900-6000 ft. (R. R. Snelling); Copper Canyon, Huachuca Mtns., 22km SSW Sierra Vista, 1770m (S. G. Brady); Douglas (W. W. Jones); Huachuca Mts., Miller Canyon (A. E. Lewis); Miller Canyon, Huachuca Mtns., 6000 ft. (W. M. Wheeler); Palmerlee, Huachuca Mtns., 5300 ft. (W. M. Wheeler); Paradise Rd., 1.3mi W Portal Rd. Chir- icahua Mts. (G. C. Snelling); Portal (Gotwald); Ram- sey Canyon, Huachuca Mts. (W. S. Creighton); Ram- sey Canyon, Huachuca Mts., 5800 ft. (W. M. Mann); SWRS, Portal, 5600 ft. (W. S. Creighton); Texas Pass, Dragoon (as "Dragon"] Mtns. (W. M. Wheeler); Gila Co.: Pinal Mts., 8000 ft. (R. A. Flock); Sierra Anchas, Hwy. 288 at Exp. Res. Stn., 4800 ft. (R. A. Johnson); Maricopa Co.: Four Peaks Wilderness, nr. Pidgeon Springs, 5600 ft. (R. A. Johnson); Mazatzal Mtns., on Four Peaks Rd., 10.3mi E Hwy. 87, 4000 ft. (R. A. Johnson); South Phoenix Park (P. S. Ward); Mojave Co.: Hualapai Mts., s. of Kingman, 1450m (E. 1. Schlin- ger); Pima Co.: Baboquivari Mtns, Forestry Cabin, 3500 ft. (W. S. Creighton); Buehman Canyon, Santa Catalina Mts., 2900-3000 ft. (R. R. Snelling & G. C. Snelling); Santa Catalina Mts. (M. Chrisman); Tucson (R. H. Crandall); Piiml Co.: Oracle, 4500 ft. (W. M. Wheeler); Sniita Cruz Co.: 6.7mi W 1-19 on Ruby Rd. (G. C. Snelling); Bog Springs Cpgrd., Madera Canyon, Santa Rita Mts. (G. C. Snelling); Madera Canvon (R. H. Crandall); Madera Canyon, Santa Rita Mts. (R. H. Crandall); Nogales (Ehringer); Pajarito Mtns., Ruby Rd., 6.7mi W I-IO, 4000 ft. (R. A. Johnson); Pena Blan- ca Lake (B. V. Brown & D. H. Feener); Ruby Rd., 7mi W Pefia Blanca (G. C. Snelling); Yavapai Co.: 7.2mi E Chino Valley, 4600 ft. (R. A. Johnson); Yuma Co.: Bur- ro Cyn., 2mi SE jet. 24, Kofa Game Refuge (P. Melhop & R. R. Snelling); California: li'.-^ .Angeles Co.: Clare- mont (B. Crow); E fork, San Gabriel R., Angeles N. F. Volume 8, Number 1, 1999 87 (C. Ishida); Eaton Canyon (Sutton?); Eaton Canyon I'k. (R. H. Crandall); Jet. Angeles Crest & Angeles Forest Hwys. (G. C. Snelling); Millard Canyon, San Gabriel Mts. (R. H. Crandall); Placerita Canyon Park (F. T. Hovore); Orange Co.: Tonner Canyon (W. P. MacKay); Rivcr^ulc Co.: Camino Resales {A. Suarez); Margarita Summit (A. Suarez); near Perris (Mallis, Zschokke & Schwartz); Pinyon Flat, 1220m (P. S. Ward); Riverside (K. W. Cooper; M. E. Irwin); San Timoteo Cyn. (M. Wasbauer & A. Hardy); Temecula Cyn., Sta. Margarita R. (E. I. Schlinger); San Bernar- dino Co.: 2mi E Mentone (W. S. Creighton); Aliso Cr., Chino Hills State Park (M. Bennett et al); Chino Hills State Park, 700 ft. (R. A. Johnson); San Diego Co.: 18km E Mt. Laguna, 300m (P. S. Ward); 5mi N Descanso, MSP primary site, 3000 ft. (J. H. Hunt); 5mi. E La Jolla (M. S. Trepanier); Chula Vista (E end), 160m (P. S. Ward); La Jolla (M. S. Trepanier); La Mesa (F. X. Wil- liams); Nate Harrison Rd. nr. Mt. Palomar, 4800 ft. (E. I. Schlinger); nf. La Mesa (F. X. Williams); Point Loma (P. Leonard; A. Suarez); Ramona, 450m (M. S. Tre- panier); San Diego (c.u.); Colorado: Chaffee Co.: Salida (C.U.); Salida, 7050 ft. (W. M. Wheeler); Georgia: Jnek- son Co.: Commerce (Vanderford); Illinois: Adama Co.: Quincy (T. E. Musselman); lohnson Co.: Feme Clyffe St. Pk. (W. S. Creighton); Iowa: Harrison Co.: Little Sioux (W. F. Buren); Woodbiny Co.: Sioux City (C. N. Ainslie; W. F. Buren); Kansas: Douglas Co.: Lawrence (F. X. Williams); Nat. Hist. Reser., Lawrence (C. W. Rettenmeyer); Harvey Co.: Sedgwick [as "Sedwick") (A. J. McCurray); Jefferson Co.: Valley Falls (A. Mattis); McPherson Co.: McPherson (W. Knaus); Pottawatomie Co.: Onaga (F. F. Crevecoeur); Rdei/ Co.: Jardine Terr. 0- F. Watkms); Manhattan (A. J. Mattis; R. C. Smith); no specific locality (F. Marlatt; J. B. Norton; J. F. Wat- kins); Sedgwick Co.: Wichita (J. R. Horton); Kentucky: Marshall Co.: Kentucky Dam (W. L. Brown); Louisi- ana: Acadia Co.: Crowley (C. E. Hood); Beauregard Co.: DeRidder (W. F. Buren); £i!s( Baton Rouge Co.: Baton Rouge (T. H. Jones); Lafai/ette Co.: Lafayette (E. S. Tucker); Madison Co.: Tallulah (McGehee); Plaquemi- nes Co.: Buras y. R. Horton); Naomi [as "Naomie") (c.u.); Rapides Co.: Alexandria (E. S. Tucker); Missis- sippi: Adams Co.: Natchez (G. W. Alexander); Clarke Co.: Quitman (Murphree); Clay Co.: Cedar Bluff (c.u.); Cedar Bluff, Trimcane (G. W. Haug); Harrison Co.: Landon (Murphree); Humphreys Co.: Belzoni (Mur- phree); lones Co.: Laurel (M. R. Smith); Lozvndes Co.: Columbus (Murphree); Monroe Co.: Aberdeen (Mur- phree); Oktibbeha Co.: Maben (L. C. Murphree); Stark- ville (W. W. Love; M. R. Smith); State College [as "Agr. Col. Miss."] (M. R. Smith); Stone Co.: Bond (Murphree); Wiggins (Murphree); Washington Co.: Greenville (G. L. Snodgrass); Missouri: Boone Co.: Co- lumbia (A. C. Cole; L. Haseman; M. Talbot); Butler Co.: Poplar Bluff (D. E. Read); Cape Girardeau Co.: Cape Girardeau (D. E. Read); Cole Co.: Jefferson City (A. C. Burrill); lohnson Co.: Knob Noster State Park (M. B. & J. R. DuBois); Ripley Co.: Doniphan (Pergan- de); Sf. Charles Co.: St. Charles (M. Talbot); St. Louis Co.: Webster Groves [as "Webster Grove") (G. Loef- fel); New Mexico: Colfax Co.: Cimarron Canyon, Ci- marron (A. C. Cole); Dona Ann Co.: 45km NE Las Cru- ces (W. MacKay); University Ranch (C. A. Kay); Grant Co.: 5km NW Silver City, 1900m (P. S. Ward); 1-10, 3mi E Separ (R. A. Johnson); Hidalgo Co.: 4km N Ro- deo, 1250m (P. S. Ward); San Simon Valley, 0.25mi W jet. Rte. 80 on Portal Rd. (NM533), 4250 ft. (K. Helms); San Simon Valley, 0.3mi NE jet. State Line Rd. & Rte. 533, 4250 ft. (S. P. Cover); San Simon Valley, 0.5mi W jet. Rte. 80 on Portal Rd. (NM533), 4250 ft. (D. Gor- don); San Simon Valley, jet. State Line Rd. & Portal Rd. (NM533), 4250 ft. (S. P. Cover); Santa Fe Co.: Santa Fe (W. M. Mann); Santa Fe, 0.5mi N 1-25 on Cerillos Rd., 6200 ft. (S. P. Cover); Sierra Co.: Hillsboro, 1600m (P. S. Ward); Socorro Co.: Sevilleta NWR (M. Kaspari); Torrance Co.: lOmi S Mountainair, 6650 ft. (A. C. Cole); Union Co.: Clayton (W. M. Wheeler); North Carolina: Swain Co.: Great Smoky Mts. Natl. Park, 5500 ft. (E. S. Ross); Oklahoma: Cimarron Co.: Kenton (T. H. Hubbell); Kay Co.: Ponea City (A. C. Burrill); Tennessee: Chester Co.: Henderson (Murphree); Da- vidson Co.: Nashville (W. S. Creighton; A. R. Laskey); near Nashville (L. Wesson); Hazokms Co.: Rogersville (W. S. Creighton); Knox Co.: Knoxville (A. C. Cole); McMinn Co.: Athens (Murphree); Monroe Co.: no spe- cific locality (Jones); Shelby Co.: Memphis (Murphree); Wayne Co.: Clifton (Murphree); Texas: Bell Co.: Bow- mer Ranch (J. F. Watkins); Be.xar Co.: 20mi S San An- tonio (E. S. Ross); San Antonio (E. S. Ross); Calhoun Co.: Port Lavaca (McGehee); Dallas Co.: Dallas (F. C. Bishop; E. W. Laake; W. D. Pierce; Vanderford); Grimes Co.: Shiro (W. Buren); Hall Co.: 6mi SE Turkey (C. W. O'Brien); Harris Co.: Houston (H. C. Milleri- der); Jeff Davis Co.: Fort Davis State Park (J. F. Wat- kins); McLennan Co.: Waco (R. S. Baldridge; J. F. Wat- kins); Montgomery Co.: Willis (J. C. Bridwell); Travis Co.: Austin (W. M. Wheeler); Val Verde Co.: Del Rio (c.u.).; West Virginia: Mason Co.: West Columbia (Murphree). (2) Form with shiny head (CASC, CDAE, JTLC, LACM, MCZC, UCDC, USNM) UNITED STATES California: Amador Co.: 9km WNW Plymouth, 200m (P. S. Ward); Colusa Co.: 1km W Fouts Springs, 600m (P. S. Ward); CoHfra Costa Co.: Danville [as "Dannville"] (F. X. Williams); El Dorado Co.: 14km NW Shingle Springs, 340m (P. S. Ward); 9km SW Pilot Hill, 340m (P. S. Ward); Kern Co.: Lone- tree Cyn., 6.8mi S jet. Randsburg [as "Randsbury") Rd. & Hwy 14 (F Andrews & M. Wasbauer); Lake Co.: 14km ENE Lower Lake, 290ni (P. S. Ward); 19km ESE Lower Lake, 700m (P. S. Ward); 20km ESE Lower Lake (B. L. Fisher); 6km NW Middletown, 490m (P. S. Ward); Nice-Bartlett Sprgs (A. Andrasfalvy); Los 88 Journal of Hymenoptera Research Angtiesi Co.: Los Angeles (A. Mallis); Pearblossom Hwy & Barrel Spr. Rd. (G. C. Snelling); Mcndocnw Co.: Hopland Field Stn., 240m (P. S. Ward); Monterey Co.: 10km SSW Jolon, Fort Hunter Liggett MR, 340m (P. S. Ward); 14km SW Jolon, Fort Hunter Liggett MR, 640m (P. S. Ward); 15km SW Jolon, Fort Hunter Lig- gett MR, 490m (P. S. Ward); Paraiso Springs (c.u.); Napa Co.: 5km ENE Rutherford, 120m (P. S. Ward); 5km W Oakville, 560m (P. S. Ward); N. side Howell Mtn., 3km NNE Angwin, 3%m (H. B. Leech); Placer Co.: 2km E Colfax, 490m (P. S. Ward); RiversiLte Co.: Red Cloud Mine, Chuckwalla Mts., 2700 ft. (G. C. Snelling); San Benito Co.: 16.8mi N New Idria (A. J. Gilbert & N. Smith); San Bernardino Co.: Cima (c.u.); San Luis Obispo Co.: 19km SSE California Valley, Car- rizo Plain Natural Area, 800m (P. S. Ward); 2.5mi. S Arroyo Grande (G. I. Stage); 20km ESE California Valley, Carrizo Plain Natural Area, 800m {P. S. Ward); Santa Barbara Co.: Arroyo Burro, 800m (J. Lon- gino); Cachuma Saddle, 1100m (J. Longino); Cachu- ma Saddle, Los Padres N. F., 930m (P. S. Ward); N end Sedgwick Ranch, 610m (P. S. Ward); N end Sedg- wick Ranch, 730m (P. S. Ward); near top of Las Cru- ces Mts (Heath); Solano Co.: 2km SE Mt. Vaca, 680m (P. S. Ward); Cold Canyon, 19km NNW Vacaville, 120m (D. M. Olson; P. S. Ward); Cold Canyon, 19km NNW Vacaville, 300m (P. S. Ward); Cold Canyon, 19km NNW Vacaville, 360m (P. S. Ward); Cold Can- yon, 19km NNW Vacaville, 420m (P. S. Ward); Cold Canyon, 19kni NNW Vacaville, 600m (P. S. Ward); Pleasants Ridge, 530m (P. S. Ward); Sonoma Co.: 1km NNE Sonoma, 170m (P. S. Ward); 3km N Sonoma, 290m (P. S. Ward); 8km NNW Cazadero, 300m (P. S. Ward); Pepperwood Ranch, 15km N Santa Rosa, 360m (P. S. Ward); Stanislaus Co.: Del Puerto Canyon, 18km WSW Patterson, 300m (P. S. Ward); Tulare Co.: Ash Mtn. Kwh Pwr Stn. #3 (J. A. Halstead); Tuolumne Co.: 2mi SE jet. Hwy. 49 & 120, 1840 ft. (G. C. Snell- ing); Yolo Co.: 6km SW Winters, 45m (D. A. Holway); 8km SE Davis, lOm (P. S. Ward); Davis (J. J. DuBois; A. Mallis); Woodland (E. I. Schlinger); Colorado: Montezuma Co.: Mesa Verde N. P., 6300 ft. (E. V. Gregg); Nevada: U/on Co.: East Walker River, 26km SSE Yerington, 1460m (P. S. Ward); Weeks, 1280m (P. S. Ward); Utah: Millard Co.: White Sage Valley [as "White Valley"! (R- W. Fautin); Washington Co.: Springdale, 0.25mi S Zion Natl. Park (G. C. Snelling). Note. Workers from sites in Los Angeles County, California are intermediate in sculpture between the two forms. Neivamyrmex texanus Watkins 1972 Ncivainynncx texanus Watkins 1972: 353. Holo- type male, Austin, Texas (W. M. Wheeler) (MCZC) [examined]. Neivamyrmex itigrcscens; Watkins 1972: 358 (part). Worker tneasiiretnents. — (n = 22). HW 0.71-1.42, HL 0.81-1.38, WL 1.27-2.26, MTL 0.87-1.68, CI 0.83-1.06, FCI 0.066- 0.125, SI 0.70-0.93, MTI 1.18-1.48, PLI 0.78-0.96, PWI 0.58-0.74. Worker diagnosis. — Rather large body size (see HW, HL and MTL measure- ments); mandible with blunt basal tooth in largest workers, such a tooth becoming in- distinct to absent in smallest workers; bas- al margin of mandible rounding into mas- ticatory margin; frontal carinae well sep- arated (MFC 0.047-0.168); clypeal apron conspicuous but less well developed than in nigrescetis; scapes relatively long, SI > 0.69 (see also plot of SL on HW; Fig. 18); occipital lobes generally well developed; anterior margin of pronotum with trans- verse ridge well developed; dorsal face of propodeum conspicuously depressed be- low the level of the promesonotum, and forming a subangulate juncture with the declivitous face, the latter more or less concave in profile; legs relatively long, MTL/HW ( = MTI) > 1.15; petiole subrect- angular, variable in shape (see PLI and PWI values), always longer than high or wide, although generally shorter and higher than in iiigrescens. Head, mesoso- ma, petiole and postpetiole densely punc- tulate, and having an opaque, granular appearance. Most of the surface of meta- pleural gland bulla densely punctulate and opaque, although sometimes with a very small, thin shiny strip immediately above the flange of the metapleural gland orifice. Body varying from dark reddish- brown to blackish-brown, the gaster and legs lighter. Queen diagnosis. — Eye distinct, consist- ing of single convex ommatidium. Head slightly broader than long (CI 1.05-1.12, n = 5). Occipital corners rounded. Prono- tum with a distinct posterior dorsolateral projection (Watkins, 1972, fig. 11). Meso- notum and propodeum without a median longitudinal impression, but propodeum with a shallow longitudinal concavity. Metatibial index (MTI) 0.81-0.90 (n = 5). Volume 8, Number 1, 1999 Petiole transverse, markedly broader than long (PWI approximately 1.2-1.7) and, in lateral view, with a single convex antero- dorsal face. Petiole width only slightly less than the length of the metatibia, DPW/ MIL = 0.75-0.91 (n = 4). Length of gaster (non-physogastric) approximately twice WL (QGL/WL = 1.88-2.25, n = 3). Male diagnosis. — Mandibles broad, spat- ulate (not sickle-shaped). Ocelli moderate in size and separated from the upper mar- gin of the compound eye by a distance greater than twice the diameter of the me- dian oecllus. No prominent transverse swelling above antennal fossa. Setae on venter of petiole long, golden, erect or suberect. Gaster reddish-brown. In profile, paramere (= stipes) slender, with a trun- cate posterior margin, and a tall, angular posterodorsal projection (Watkins 1985, Plate 9, fig. 5); volsella unforked, with long, slender upturned apex; aedeagus ( = sagitta) with straight posteroventral process, just exceeding the posterodorsal process (in posterior reach). Comments. — Despite the distinctive males and queens of this species, workers of N. texaniis are difficult to distinguish from those of N. nigrescens. As Watkins (1972, 1985) noted, workers of N. texanus have the declivitous face of the propo- deum slightly concave in profile (more or less straight in N. nigrescens), and forming a more distinctive angle with the dorsal face of the propodeum, but the difference is a subtle one. Workers of N. texajius also have the frontal carinae more widely sep- arated than those of N. nigrescens, and plots of MFC (the minimum distance be- tween the frontal carinae) against various measures of body size (e.g., HW, HL, MTL) reveal strong and almost diagnostic differences (Figs. 31-33), when consider- ing those populations of N. nigrescens in the Southwest and eastern United States that are sympatric with N. texanus. Work- ers of N. texanus also tend to have a short- er, higher petiole than those of N. nigres- cens (Fig. 34). Finally, the swelling of the metapleural gland (bulla) is densely punc- tulate and more or less opaque over near- ly all its surface in workers of N. texanus, while in N. nigrescens it presents a more shiny appearance. Differences between N. texanus and N. californicus have been con- sidered under the latter species. Material examined.— {CASC, LACM, MCZC, RAJC, UCDC, USNM) MEXICO Chihuahua: Mpio. Riva Palacio, Bella- vista (W. & E. MacKay); Mpio Chihuahua, 45km S Sueco (Carnada): Durango: 37mi W Durango (W. S. Creighton); Hidalgo: 5mi S Tizacuya (W. S. Ross); Guerrero Mill (W. M. Mann); Pachuca (W. M. Mann); San Miguel (W. M. Mann); Jalisco: Guadalajara (McClendon; W. M. Mann; Cadwallader); Japopan (Cadwallader); Nuevo Leon: Vallecillo (W. F. Buren); Sonora: lOmi S Agua Prieta (V. Roth); 26mi NW Ba- hia Kino (E. M. Fisher). UNITED STATES Arizona: Cochise Cc: 5.8mi SE Sunnyside, 5700 ft. (R. R. Snelling); S.lmi SE Sunny- side, 5950 ft. (R. R. Snelling); Chiricahua Mts., Rucker Camp, T19S R29E Sect.22 (W. MacKay); Douglas (W. W. Jones); Huachuca Mtns., 3mi SW Rte. 92 on Coro- nado Natl. Mon. Rd., 5100 ft. (S. P. Cover); Huachuca Mtns., 3nii S jet. Rt. 92 on rd. to Coronado Natl. Mon., 5100 ft. (S. P. Cover); Huachuca Mtns., Bear Creek, 19km SW Sierra Vista, 1640m (S. G. Brady); Palmerlee, Huachuca Mtns., 5300 ft. (W. M. Wheeler); San Ber- nardino Valley, Imi NW jet. Rt. 80 on FSR 74 (Rucker Canyon Road), 4700 ft. (S. P. Cover); Piiim Co.: 16mi W Tucson (S. Prchal; G. C. Snelling); Baboquivari Mtns, Brown Canyon (A. C. Cole); Baboquivari Mtns, For- estry Cabin, 3500 ft. (W. S. Creighton); Brown Canyon, Baboquivari Mtns. (Menke & Stange); Sabino Canyon (R, H. Crandall); Tucson Mtns. (F. R. Gehlbach); Santa Cruz Co.: Nogales (Buren); Colorado: EI Paso Co.: Col- orado Springs (J. G. Jack); Colorado Springs and vicin- ity (W. M. Wheeler); ]efferson Co.: Prospect Park (c.u.); Florida: Alachua Co.: 29°34.5'N, 82°29'W (R. W. Lund- gren); Austin Carey Forest, Gainesville (G. B. Fair- child); Gainesville (T. H. Hubbell; A. Van Pelt); Pierce Homestead, Gainesville (W. R. Pierce); Escaiiil'ia Co.: Pensacola (R. M. Lhamon); hiilum River Co.: Sebastian (Nelson); Leon Co.: Woodville (D. E. Read); Volusia Co.: Daytona [Beach] (W. F. Buren); New Mexico: Dona Ana Co.: 45km NE Las Cruces (W. Mackay); EMi/ Co.: Los Medanos, T22S R31E, Sect.l5 (c.u.); San Miguel Co.: Las Vegas (W. M. Wheeler); North Carolina: Nczo Han- over Co.: Wilmington (Vanderford); Yancey Co.{7): Black Mts. (c.u.); South Carolina: Oconee Co.: Clemson Col- lege (J. Berly & M. Smith); Texas: Bexar Co.: San An- tonio (E. S. Ross); Croeketl Co.: Ozona (A. C. Cole); Jeff Davis Co.: Fort Davis (T. W. Taylor); Kimhie Co.: Junc- tion 0- F- Watkins); Llano Co.: Llano (A. W. Morrill); McLennan Co.: Baylor Camp Q. F. Watkins); Waco (J. 90 Journal of Hymenoptera Research Briga; S. Davis; S. Eldridge; F. R. Gehlbach; O. L. Nich- Chesterfield Co.: Warwick (Bond); Essex Co.: Imi SE olson; R. W. Plsek; C. A. Rhines; J. F. Watkins); Taylor Durmsville (D. R. Smith); Henry Co.: Chatmoss Plant, Co.: Abilene rest stop (W. S. Creighton); Travis Co.: Martinsville (S. Schaeffer); Loin'sn Co.: 4mi S Cuckoo (J. Austin (W. M. Wheeler); Austin, Brackenridge Field Kloke & D. R. Smith); Norfolk city: Norfolk (Vander- Lab (S. D. Porter); Victoria Co.: Victoria (c.u.).; Virginia: ford). MODIFICATION OF EXISTING IDENTIFICATION KEYS Couplets 5 and 9 of Watkins' (1985) worker key to the United States species of hleivamyrmex need modification to take into account the much greater intraspecific variability in body sculpture in N. nigresceiis. In the updated portions of the key (below) I have also documented addihonal features of N. tcxanus and N. opncithornx which will facilitate their identification. 5 In lateral view declivitous face of propodeum slightly concave and forming a somewhat angular corner with the dorsal (basal) surface; frontal carinae well separated (MFC 0.05- 0.17), as revealed in bivariate plots of MFC on various measures of body size (Figs. 31- 33); petiole relatively short and high (Fig. 34); surface of most of the metapleural gland bulla densely punctulate and opaque, although sometimes with a very small, thin shiny strip immediately above the flange of the metapleural gland orifice ... N. texanus Watkins - Declivitous face of propodeum more or less straight in lateral view and forming a round- ed angle with the dorsal (basal) surface; frontal carinae usually more closely contiguous (Figs. 31-33) and petiole height tending to be lower (Fig. 34); sculpture on surface of metapleural gland bulla becoming obsolete, such that the lower half or more is smooth (or weakly reticulate) and conspicuously shiny N. nigrescens (Cresson) (part) 9 Inner basal margin of mandible with a straight edge which forms an angular corner or small tooth at its juncture with the masticatory margin; side of pronotum usually smooth and shiny, and contrasting with the rugulose-punctulate sculpture that covers at least part of the mesosoma dorsum N. opacithorax (Emery) Inner basal margin of mandible with a convex edge which curves into the masticatory margin without forming an angular corner (Fig. 1); pronotal sculpture variable, but side of pronotum usually at least weakly reticulate-punctulate rather than exhibiting a smooth, shiny surface that contrasts with the sculptured mesosoma dorsum 9A 9A Body, scapes and legs shorter, head broader; MTI 1.05-1.14; see also plots of MTL on HW, SL on HW, WL on HW, CI on MTL and SI on MTL (Figs. 17-19, 21-23, 25, 26); dorsal surface of propodeum only slightly depressed below the level of the mesonotum (Fig. 2); dorsum of postpehole and (typically) center of mesonotum smooth and shining, with scattered piligerous punctures but little or no trace of reticulation or punctulae . . . N. califomiais (Mayr) Body, scapes and legs longer; MTI 1.16-1.52; see also Figs. 17-19, 21-23, 25, 26; dorsal surface of propodeum conspicuously depressed below the level of the mesonotum (Fig. 5); postpetiole and mesonotum varying from opaque to sublucid, with traces of reticu- lation or punctulae usually evident N. nigrescens (Cresson) (part) In the key to United States Neivamyrmcx based on queens (Watkins 1972: 350-351) "cnliforniciis" in couplet 7 should be replaced with "nigrescens (part)". The queen of N. californiciis is not yet definitively known (see discussion above, under that species). Couplet 8 of Watkins' (1982) key to Mexican species of Ncivnmyrmux can be modified as follows to incorporate N. califoniicus and the more recently described N. cluimetcnsis Watkins (1986). The treatment of N. nigrescens requires no change since the shiny form of N. nigrescens is not known to occur in Mexico. 8 Head and gaster of smaller workers blackish brown or reddish brown with a black over- cast, mesosoma reddish brown without a blackish overcast; dorsum of promesonotum Volume 8, Number 1, 1999 91 slightly convex, and dorsum of propodeum as short or shorter than the node of petiole in lateral view N. manni (Wheeler) Head and mesosoma same color (reddish brown), gaster slightly lighter; posterior one- half of promesonotum flattened; dorsum of propodeum variable in length 8A 8A Dorsal face of propodeum shorter than the declivitous face, and conspicuously depressed below the level of the promesonotum, at least in larger workers (Watkins 1986, fig. 2); postpetiole as long as petiole N. chamelensis Watkins Dorsal face of propodeum as long as, or longer than, the declivitous face, and not con- spicuously depressed below the level of the promesonotum (Fig. 2); postpetiole shorter than petiole 8B 8B Inner basal margin of mandible with a straight edge which forms an angular comer or small tooth at its juncture with the masticatory margin; clypeal apron short, extending anteriorly by a distance equal to about 0.2-0.4 times the diameter of the torulus; petiole relatively narrow (Fig. 20); side of pronotum usually smooth and shiny, and contrasting with the rugulose-punchilate sculpture that covers at least part of the mesosoma dorsum N. opacithorax (Emery) Inner basal margin of mandible with a convex edge which curves into the masticatory margin without forming an angular corner (Fig. 1); clypeal apron longer, extending an- teriorly by a distance equal to about 0.4-0.5 times the diameter of the torulus; petiole broader (Fig. 20); pronotal sculpture variable, but side of pronotum usually at least weak- ly reticulate-punctulate rather than exhibiting a smooth, shiny surface that contrasts with the sculptured mesosoma dorsum N. califomicus (Mayr) BIOLOGICAL OBSERVATIONS Habitat Preferences Although N. nigrescens and N. califomi- cus are probably not sister taxa (see below) they appear to have similar ecological preferences in California. In southern Cal- ifornia and northern Baja California, where morphological differences between the two species are greatest, their geo- graphical distributions overlap broadly (Figs. 35, 36) and they occur together in such habitats as chaparral, coastal sage scrub and oak woodland. In northern Cal- ifornia workers of N. nigrescens are more difficult to distinguish from those of N. califomicus, due to convergence in body sculpture. In this region the two species are less frequently sympatric. N. califomi- cus tends to be confined to more mesic, coastal locations than N. nigrescens, al- though the habitats occupied are similar: oak woodland, riparian woodland, chap- arral and grassland. An exception to this is the occurrence of N. califomicus at a xe- ric inland site in the Sierra Nevada foot- hills (14km NW Shingle Springs, El Do- rado County) where it co-occurs with N. nigrescens and N. opacithorax in a distinc- tive chaparral vegetation on gabbro soil. It was the sympatric association at this lo- cality that alerted me to the distinction be- tween true N. califomicus and the "shiny morph" of N. nigrescens. The gabbro site has a number of rare plant species (Hunter and Horenstein 1992) and N. califomicus likewise gives the impression of being an isolated, relic population here (the north- ern-most point in Fig. 36). Throughout California both N. califor- nicus and N. nigrescens are limited to low elevations, essentially below the conifer- ous forest zones. Most recorded popula- tions come from locations below 1500m (N. califomicus: sea level to 1840m; N. ni- grescens: sea level to 1460m). Outside the range of N. califomicus, pop- ulations of N. nigrescens show a wide lat- itude of habitat choice, being found in scrubland, grassland and canyons of the 92 Journal of Hymenoptera Research 35 36 ■ nigrescens (s.s.) ^nigrescens (shiny form) Figs. 35-38. Known distributions in southwestern United States and Mexico of Neivamyrmex nigrescens (25), N. californicus and N. tcxanus (26), shiny and punctate forms of N. nigrescens (27), and N. opntcitliorax (28). N. nigrescens, N. opncitlwrnx and N. tcxanus also occur in eastern United States (see Watkins 1985: 499-500). Sonoran and Chihuahuan deserts, pine- oak-juniper woodland, prairie grassland, and eastern deciduous forest (Gregg 1963; Mirenda et al. 1980; Schneirla 1958; Smith 1942). It is perhaps unsurprising that over this broad range of ecological conditions the species displays considerable intraspe- cific variation in worker morphology. Associations with Messor atidrei In California both Neivamyrmex nigres- cens and N. californicus are often associated with nests of the common seed-harvesting ant, Messor andrei (Mayr). Indeed, one of the most efficient ways to determine the presence of these species at a locality is to examine the chaff piles of Messcr tvuirci Volume 8, Number 1, 1999 93 nests. Such middens frequently contain the corpses of Neivamyrmex workers. There are several possible explanations for this. First, Messor amirei workers appear to be efficient scavengers that collect dead and dying ants (of all kinds) and later discard the dried corpses in their middens. This is indicated by the fact that the remains of other ants, including species of Campono- tus, Creinntogaster, Don/mynnex, Forelius, Formica, Leptothorax, Monomorium, Pheidole, Prenolepis, Solenopsis and Stenamma, are of- ten encountered in M. audrei middens (Ward, pers. obs.). Second, both N. californicus and N. ni- grescens have been observed attacking nests of Messor mnirei. In Amador County, California I observed nocturnal foraging columns of N. nigrescens workers entering two adjacent M. ivuirei nests — one contain- ing a large Messor colony, the other a small incipient colony. The raid on the small nest was partially successfully, with N. nigrescens workers carrying off para- lyzed Messor worker minims, although some Messor workers (and a dealate queen) escaped capture by climbing short grass stalks. No prey were observed to be taken from the larger nest, which con- tained normal-sized Messor workers. Mark Brown (1999) recorded N. californicus workers attacking Messor andrei nests at Jasper Ridge Preserve, near Stanford Uni- versity. These raids on Messor andrei nests by N. californicus and N. nigrescens do not appear to be especially effective, yet in the observed cases the army ants persisted in their assaults. Third, the high frequency of Neivamyr- mex worker corpses in Messor middens may reflect a tendency of Neivamyrmex colonies to temporarily occupy part of the underground chambers of the Messor nests, leading to an increased frequency of interactions between the two species. Some evidence for this comes from an ob- servation that I made in early April 1984 near Rutherford, Napa County, northern California: a colony of N. nigrescens ("shiny form"), containing larvae, was lo- cated in wet but well-aerated soil imme- diately adjacent to an active Messor andrei nest. The N. nigrescens workers were emerging from the soil and slowly milling about on the ground surface during day- light hours, an unusual behavior but one which occurs in spring (March-May) in northern California before N. nigrescens begins its period of summer-active — and predominantly nocturnal — surface raid- ing. Thus, it appeared that this Rutherford colony of N. nigrescens had over-wintered in the soil in abandoned sections of the Messor andrei nest. Schneirla (1963) report- ed the use of pre-empted ant nests (species not specified) by over-wintering colonies of N. nigrescens in Arizona. Interactions with Other California Ant Species Other ants besides Messor andrei that are subject to raids by Neivamyrmex nigrescens in California include Pheidole californica Mayr, P. hyatti, Soloiopsis molesta (Say) and Formica moki Wheeler (Ward, pers. obs.). Mallis (1938, 1941) reported N. nigrescens (misidentified as N. californicus) foraging nocturnally for insects attracted to street lamps on the Davis campus of the Uni- versity of California, and attacking nests of the introduced ant, Tetramorium caespi- tum (Linnaeus). At the same location Mal- lis (1938) also recorded an altercation be- tween Neivamyrmex and Argentine ants, Linepithema humile (Mayr), in which Line- pithema emerged the victor. L. humile is now very abundant on the UC Davis cam- pus and, during 17 years of observation here, 1 have seen no evidence that popu- lations of Neivamyrmex nigrescens survive on campus, although the species occurs 8km southeast of Davis at a site not yet overrun by Linepithema humile. Work by Suarez et al. (1998) in southern California shows that Neivamyrmex species, includ- ing N. nigrescens, are among the first ant species to disappear from patches of rem- nant coastal sage scrub when Linepithema humile invades from adjacent urban habitat. 94 Journal of Hymenoptera Research Table 1. Tests of nest evacuation response in potential prey of Neivnmyrmex. Each test in\'olved placement of one to several live Neivamyrmex workers at the nest entrance of an active test ant colony. All locations are in California. Ncn'ajJii/rTiu'x species Mass evacuation Test species Location (and source population) response? Pheidole desertonim Pinyon Flats nigrescens (Pinyon Flats) Yes Pheidok hyatti Carrizo Plain nigrescens (Carrizo Plain) Yes Pheidole hyatti Del Puerto Canyon nigrescens (Cold Canyon) Yes Pheidole californ ica Cold Canyon cnlifornicus (El Dorado Co.) Yes Pheidole enlifoni icn Cold Canyon nigrescens (Cold Canyon) Yes Pheidole califoniicn Davis nigrescens (Cold Canyon) Variable- Pheidole enlifornicn Davis nigrescens (Carrizo Plain) No Pheidole enlifornicn Del Puerto Canyon nigrescens (Cold Canyon) No Pheidole califoniicn Ventura Co. nigrescens (Carrizo Plain) No Pheidole califoniicn El Dorado Co. opacithorax (El Dorado Co.) No Messor andrei Cold Canyon nigrescens (Cold Canyon) No Doryniynnex bicolor Pinyon Flats nigrescens (Pinyon Flats) No Dorymyrniex insnnus Davis nigrescens (Cold Canyon) No ■' Evacuation response seen in one of two trials. Elsewhere, across its broad transconti- nental distribution, N. nigrescens is report- ed to preferentially raid ant nests, espe- cially those of Pheidole species (Mirenda et al. 1980), but also including colonies of Aphaenogaster, Camponotus, Dorymyrniex, Formica, Leptothorax and Solenopsis (LaMon and Topoff 1981; Mirenda et al. 1980; Schneirla 1958, 1963; Smith, 1927). Both ant brood, adult workers, and (when available) sexual alates are taken as prey, as are occasionally termites and non-social insects. Less is known about the prey preferenc- es of Neivamyrmex cnlifornicus but they ap- pear to be similar to those of N. nigrescens. Mark Brown (1999) observed N. califortii- cus attacking both Messor andrei and Phei- dole californica colonies at Jasper Ridge. There is indirect evidence that Solenopsis molesta and Pheidole hyatti are also preyed upon. In a collection of dead N. californicus workers from a Messor andrei chaff pile at Jasper Ridge (collected by Nate Sanders) several individuals had dead workers of Solenopsis molesta attached (with closed mandibles) to legs and /or antennae. Dead workers of N. californicus (as well as those of N. nigrescens) have been found in the nest middens of Pheidole hyatti (Ward, pers. obs.). This and certain other Pheidole species in California show an enemy-spe- cific nest evacuation response to the pres- ence of N. californicus and N. nigrescens workers. Observations on this nest absconding be- havior in California Pheidole are summa- rized in Table 1. A single Neivatnyrmex worker, held with a pair of forceps at the nest entrance, can be sufficient to cause mass evacuation of workers and brood in Pheidole californica, P. desertorum, and P. hyatti. In Pheidole californica the response is not invariably observed, however (Table 1), and seems to vary with location, ambient conditions, and possibly as a function of previous experience (as documented for another prey species, Aphaenogaster cocker- elli Andre (McDonald and Topoff 1986)). Similar nest evacuation behavior has been studied in Arizona populations of Pheidole desertorum and P. hyatti attacked by N. ni- grescens (Droual 1983, 1984). POPULATION DIFFERENTIATION AND BIOGEOGRAPHY The taxonomic confusion surrounding N. californicus and N. nigrescens can be traced to undue reliance on superficial fea- tures of head sculpture, lack of attention Volume 8, Number 1, 1999 95 to other structural differences, and insuf- ficient appreciation of the degree of vari- ability in N. nigrescens. In Neivamyrmex and in all other genera of New World army ants (which together comprise the subfamily Ecitoninae) the queens are en- tirely wingless and have limited capacities for dispersal (Gotwald 1995). Because of the population viscosity associated with limited female movement — only partly mitigated by gene flow via dispersing, winged males — and perhaps also because of the lower effective population sizes achieved by these highly predacious (and hence higher trophic level) organisms, one expects conspecific allopatric populations of army ants to develop substantial differ- ences. The prediction is borne out, at least among the more widespread species of ecitonine army ants, in which there exists a large amount of geographical variation (Borgmeier 1955, 1956). It seems desirable for taxonomists to be cautious in establish- ing new species of army ants especially when dealing with closely related allopat- ric populations. In the present paper I have refrained from giving a new name to the shiny form of N. nigrescens, because it is essentially allopatric to the more typical granulate-punctate N. nigrescens. The two are not known to both occur sympatricly and retain their distinctness; rather, in geo- graphically intermediate localities (i.e., in the San Gabriel Mountains of southern California) we find morphologically tran- sitional populations. A genetic analysis of the transition zone would be interesting, especially in view of its relative narrow- ness. Thus, the distribution of the two forms of N. nigrescens (Figure 37) suggests a pre- vious fragmentation of the range of N. ni- grescens, and the consequent differentia- tion of populations but not to a degree sufficient to cause reproductive isolation. The Transverse Ranges of California and the upper Colorado River basin emerge as the probable sites of geographic barriers. The divergence between the common ancestor of N. nigrescens (sensu lato) and N. cnlifornicus must have occurred consid- erably earlier. On the basis of structural similarities (Figs. 17-19) N. californicus ac- tually appears to be more closely related to N. opacithorax than to N. nigrescens, with the latter species being more closely relat- ed to N. texanus. This hypothesis could be tested with the study of additional char- acters and (crucially) the inclusion of ad- ditional taxa from Mexico, especially N. chamelensis, N. cornutus, N. manni, and N. sumichrasti. The possibility that the wide- spread N. nigrescens is paraphyletic should also be considered. A final point of biogeographic interest concerns the distribution limits of Neiva- myrmex californicus (Fig. 36) and N. opaci- thorax (Fig. 38) in northern California. Both species appear to be restricted to ar- eas east and south of the Sacramento River and the San Francisco /San Pablo Bays. This drainage system can be expected to be a significant barrier to dispersal in hy- menopterans such as army ants whose re- productive females are wingless. ACKNOWLEDGMENTS I am grateful to the following persons for access to material in collections: Wojciech Pulawski (CASC), John Sorensen (CDAE), Jack Longino OTLC), Roy Snelling (LACM), Stefan Cover (MCZC), Ivan Lobl (MHNG), Stefan Schodl (NHMV), and Ted Schultz (USNM). Additional useful material was received from Melissa Bennett, Sean Brady, Don Feener, Rob- ert Johnson, Mike Kaspari, Nate Sanders, Gordon Snelling, Andy Suarez and Mark Trepanier. I thank Sean Brady, Brian Fisher, Jack Longino and two anonymous reviewers for comments on the manu- script. Research support was received from the Uni- versity of California and the National Science Foun- dation. LITERATURE CITED Bolton, B. 1995. A nciv general catalogue of the ants of Ike zocrhi. Cambridge, Massachusetts: Harvard University Press, 504 pp. Borgmeier, T. 1953. Vorarbeiten zu einer Revision der neotropischen Wanderameisen. Slmlia Entomolo- gica 2: 1-51. Borgmeier, T. 1955. Die Wanderameisen der neotro- pischen Region. Studia Eiitoinologica 3: 1-720. 96 Journal of Hymenoitera Research Borgmeier, T. 1956. Ueher Rassen bei Eciton (Hym. Formicidae). Revista Brasileira tie Entomologia 4; 209-212. Borgmeier, T. 1958. Nachtraege zu meiner Monogra- phie der neotropischen Wanderameisen (Hym. Formicidae). Stiidia Entoinolcgica (n.s.)l: 197-208. Brown, M. J. F. 1999. Nest relocation and encounters between colonies of the seed-harvesting ant Mt'S- sor andrei. Insectes Sociaux, in press. Creighton, W. S. 1950. The ants of North America. Bulletin of the Miiicum of Comparative Zoology 104: 1-585. Cresson, E. T. 1872. Hymenoptera Texana. Transae- tions of the Ainerieau Eiitontological Soeiety 4: 153- 292. Dalla Torre, K. W. v. 1893. Catalogus Hymcnopterorum hucusque descriptorum systemaiicus et synonymicus. Vol. 7. Formicidae Heterogyna. Leipzig: W. En- gelmann, 289 pp. Droual, R. 1983. The organization of nest evacuation in PheidoU- de^ertorKin Wheeler and P. hyatti Em- ery (Hymenoptera: Formicidae). Beluwioral Ecol- ogy and Sociobiology 12: 203-208. Droual, R. 1984. Anti-predator behaviour in the ant Pheidole desertonim: the importance of multiple nests. Animal Behaviour 32: 1054-1058. Emery, C. 1894. Studi sulle formiche delta fauna neo- tropica. Vl-XVI. Bullettino della Societii Entomolo- gica Italiana 26: 137-241. Emery, C. 1895. Beitrage zur Kenntniss der nord- amerikanischen Ameisenfauna. Schluss. Zoolo- gisches Jahrhiicher. Abteilung fiir Systematik, Geo- grapbie und Bioh^gie der Ticre 8: 257-360. Emery, C. 1900. Nuovi studi sul genere Eciton. ^4e- morie della Reah' Accadcn na delle Scicnzc dell'Istituto di Bologna 58: 173-188 [pagination of separate: 511-526]. Forel, A. 1914. Formicides d'Afrique et d'Amerique nouveaux ou peu connus. Bulletin de la Societe Vnudoife des Sciences Naturelles 50: 211-288. Gotwald, W. H., Jr. 1995. Army ants: the biology of so- cial predation, Ithaca, New York: Cornell Univer- sity Press, xviii + 302 pp. Gregg, R. E. 1963. The ants of Colorado, with reference to their ecology, taxonomy, and geographic distribu- tion. Boulder: University of Colorado Press, xvi + 792 pp. Hunter, J. C, Horenstein, J. E. 1992. The vegetation of the Pine Hill area (California) and its relation to substratum. Pp. 197-206 in: Baker, A. J. M., Proctor, J., Reeves, R. D. (eds.) The vegetation of ultramafic (serpentine) soils. Andover, U. K.: In- tercept Ltd., XX + 509 pp. LaMon, B., Topoff, H. 1981. Avoiding predation by army ants: defensive behaviours of three ant spe- cies of the genus Campninotus. Aniniai Behaviour 29: 1070-1081. Mallis, A. 1938. Army ants in California. Scientific Monthly 47: 220-226. Mallis, A. 1941. A list of the ants of California with notes on their habits and distribution. Bulletin of the Southern California Academy of Sciences 40: 61- 100. Mayr, G. 1870. Formicidae novogranadenses. Sit- zungsberichte der Kaiserlichen Akademie der Wissen- schaften in Wien. Mathematisch-Naturwissenschaft- liche Classe. Abteilung I 61: 370-^17. Mcdonald, P., Topoff, H. 1986. The development of defensive behavior against predation by army ants. Developmental Psychobiology 19: 351-368. Mirenda, J. T., Eakins, D. G., Gravelle, K., Topoff, H. 1980. Predatory behavior and prey selection by army ants in a desert-grassland habitat. Behavior- al Ecology and Sociobiology 7: 119-127. Sclineirla, T. C. 1958. The behavior and biology of certain Nearctic army ants. Last part of the func- tional season, southeastern Arizona. Insectes So- ciaux 5: 215-255. Schneirla, T. C. 1963. The behaviour and biology of certain Nearctic army ants; springtime resur- gence of cyclic function — southeastern Arizona. Animal Behavwur 11: 583-595. Smith, M. R. 1927. A contribution to the biology and distribution of one of the legionarv ants, Eciton schmitti Emery. Annals of the Entomological Society of America 20: 401-404. Smith, M. R. 1938. Notes on the legionary ants Eciton, subgenus Acamatus with a record of new specific synonymy. Proceedings of the Entomological Society of Washington 40: 157-160. Smith, M. R. 1942. The legionary ants of the United States belonging to Eciton subgenus Neivamyrmex Borgmeier. American Midland Naturalist 27: 537- 590. Snelling, R. R., George, C. D. 1979. The taxonomy, dis- tribution and ecology of California desert ants. Re- port to California Desert Plan Program, Bureau of Land Management, U.S. Dept. Interior. 335 + 89 pp. Suarez, A. V., Bolger, D. T., Case, T. J. 1998. Effects of fragmentation and invasion on native ant communities in coastal southern California. Ecol- ogy 79: 2041-2056. Watkins, J. F., II. 1972. The taxonomv of Neiiwnyrmex texanus, n. sp., N. nigrescens and N. californicus (Formicidae: Dorylinae), with distribution map and keys to the species of Neivamyrmex of the United States. Journal of the Kansas Entomological Society 45: 347-372. Watkins, j. F., II. 1976. The identification and distribu- tion of Nejv World army ants (Dorylinae: Tormici- dae). Waco, Texas: Bavlor Universitv Press, 102 PP- Watkins, J. F., II. 1982. The armv anls of Mexico (Hv- Volume 8, Number 1, 1999 97 menoptera: Formicidae: Ecitoninae). loiirnnl oftlic Kansas Eiitoiiwlcgical Society 55: 197-247. Watkins, J. F., II. 1985. The identification and distribu- tion of the army ants of the United States of Amer- ica (Hymenoptera, Formicidae, Ecitoninae). journal of the Kansas Entonwlo\;ical Soach/ 58: 479-502. Watkins, J. F., II. 1986. Nnvann/rmcx chamelcnsis. n. sp. (Hymenoptera: Formicidae: Ecitoninae) from Jal- isco, Mexico. Journal of the Kamas Entomological Society 59: 361-366. Wheeler, W. M. 1900. The female of Eciton sumichrasti Norton, with some notes on the habits of Texan Ecitons. American Naturalist 34: 563-574. Wheeler, W. M. 1908. The ants of Texas, New Mexico and Arizona. Part I. Bulletin of the American Mu- seum of Natural Histon/ 24: 399^85. J. HYM. RES. Vol. 8(1), 1999, pp. 98-108 Revision of North American Aleiodes Wesmael (Part 5): The melanopterus (Erichson) Species-Group (Hymenoptera: Braconidae, Rogadinae) Paul M. Marsh and Scott R. Shaw (PMM) Cooperating Scientist, USDA Systematic Entomology Laboratory, c/o National Museum of Natural History, Washington, DC, USA (correspondence address: P. O. Box 384, North Newton, Kansas 67117 USA); (SRS) Entomology Section, Department of Renewable Resources, P. O. Box 3354, University of Wyoming, Laramie, Wyoming 82071 USA Abstract. — The Aleiodes melanopterus (Erichson) species-group is regarded as monophyletic based on the large oral opening and narrow clypeus. It is defined to include 13 Palaearctic and Neo- tropical species plus the following North American species: inandibtilaris (Cresson) new com- bination, megastomus new species, melanopodus new species, mexkanus Cresson, miani new species, and politiceps (Gahan) new combination. A key to the North American species of the melanopterus species-group is presented. The rogadine braconid genus Aleiodes Wesmael is worldwide in distribution, but is particularly species-rich in the Holarctic Region. Aleiodes is well diversified in North America, with at least 90 species in the United States and Canada (S. Shaw et al., 1997). This study is the fifth in a series of planned papers on Aleiodes species- groups, intended to provide a complete revision of the genus for North America (see S. Shaw et al, 1997, 1998a, 1998b; Marsh and Shaw, 1998). The melanopterus (Erichson) group is moderate sized with species occurring in the Palaearctic, Ne- arctic and Neotropical Regions. This is a distinctive monophyletic group with all species having a large oval oral opening and narrow clypeus. Our definition of this species-group includes all species known to us worldwide. However, because our main intent is to provide a revision of the North American species, species treat- ments are limited to the Nearctic fauna. Aleiodes species are koinobiont endopar- asitoids of lepidopterous larvae, especially macrolepidoptera of the superfamilies Noctuoidea and Geometroidea, and to a lesser extent, Sphingoidea and Papiliono- idea (S. Shaw et al, 1997). Very little is known about the biology of the melanop- terus species-group but the few records in- dicate parasitism of noctuid larvae. The method of parasitism, unique to the tribe Rogadini, is noteworthy: the Aleiodes larva completes its feeding and pupates within the shrunken and mummified remains of the host caterpillar. In all known cases, the form of the mummy caused by a particu- lar Aleiodes species is characteristic for that host and parasitoid, so mummified re- mains are of considerable diagnostic value and should be retained with the parasitoid when reared. For a more complete discus- sion of Aleiodes biology, readers may refer to M. Shaw (1983, 1994), M. Shaw and Huddleston (1991), S. Shaw (1995) and S. Shaw et al. (1997). We have host informa- tion for only one North American species, politiceps (Gahan), which has been reared from at least two species of noctuid larvae. METHODS Species covered in this paper can be identified as members of the subfamily Volume 8, Number 1, 1999 99 Rogadinae using the keys of S. Shaw (1995), M. Shaw and Huddleston (1991) or Wharton et al. (1997). Our definition of Al- eiodes follows that of S. Shaw (1993), S. Shaw et al. (1997) and van Achterberg (1991). Specimens can be determined as Aleiodes using the keys of Wharton et al. (1997). The species-groups of North Amer- ican Aleiodes can be identified using the key provided in S. Shaw et al. (1997). Terminology follows that used for Aleio- des by S. Shaw et al. (1997), S. Shaw (1993) and Marsh (1989). Microsculpture termi- nology follows that of Harris (1979). Wing vein terminology agrees with the system adopted by Wharton et al. (1997) and agrees closely with that of Goulet and Huber (1993). A labeled diagram of wing veins was provided by S. Shaw et al. (1997). Acronyms for collections where type material is deposited are as follows: AEI (American Entomological Institute, Gainesville, FL), AMNH (American Mu- seum of Natural History, New York, NY), ANSP (Academy of Natural Sciences, Philadelphia, PA), CNC (Canadian Na- tional Collection, Ottawa, Canada), CAS (California Academy of Sciences, San Francisco, CA), FSCA (Florida State Col- lection of Arthropods, Gainesville, FL), MCZ (Museum of Comparative Zoology, Harvard University, Cambridge, MA), MSU (Montana State University, Boze- man, MT), NCSU (North Carolina State University, Raleigh, NC), NNML (Natio- naal Natuurhistorisch Museum, Leiden, The Netherlands), RMSEL (Rocky Moun- tain Systematic Entomology Laboratory, University of Wyoming, Laramie, WY), TAMU (Texas A&M University, College Station, TX), UCD (University of Califor- nia, Davis, CA), USNM (National Muse- um of Natural History, Smithsonian Insti- tution, Washington, DC). ALEIODES MELANOPTERUS SPECIES-GROUP Included species. — miiuatiis (Herrich- Schaeffer) 1838, new combination, Europe, North Africa; melanopterus (Erichson) 1848, new combination, South America; aestiio- siis (Reinhard) 1863, new combination, eastern Europe, Middle East; mexicanus Cresson 1869, Mexico, southern United States; maiidibularis (Cresson) 1872, new combination, central United States; kruli- koivskii (Kokoujev) 1898, new combination, eastern Europe, Mongolia; veimstulus (Ko- koujev) 1905, new combination, eastern Europe; lucidus (Szepligeti) 1906, new combination, Bolivia; politiceps (Gahan) 1917, new combination, eastern United States, Central America; wadai (Watanabe) 1937, new combination, Japan; agilis (Te- lenga) 1941, new combination, eastern Eu- rope; desertus (Telenga) 1941, new combi- nation, eastern Europe; /(7/jri»^^fn (Telenga) 1941, new combination, Mongolia; glaber (Telenga) 1941, new combination, eastern Europe, Mongolia; riificeps (Telenga) 1941, new combination, eastern Europe; flavis- tignia Shaw 1993, Brazil; megastomus, new species; melaiwpodus, new species; miani, new species. Diagnostic characters. — Oral opening (Figs. 5-9) large and oval, width equal to or greater than height of face, clypeus very narrow; eyes and ocelli large, the ocellar diameter equal to or slighter greater than diameter of lateral ocellus; mesonotum and mesopleuron usually smooth and pol- ished; hind wing vein RS straight, margin- al cell gradually widening to wing apex (Figs. 1^). A discussion of the phyloge- netic relations of the species-groups of Al- eiodes can be found in Fortier (1997). Comments. — This is a moderately sized species-group associated with noctuids and distinguished by the large oval oral opening. The Neotropical species have been reviewed by Shaw (1993) where he placed them in the subgenus Eucystomas- tax. Although the name melanopterus used for this species-group is not the oldest name, it has been used previously by Shaw (1993) in his study of the Neotropi- cal species and Fortier (1997) in his study of Aleiodes phylogeny. For this reason and 100 Journal of Hymenoptera Research because the ICZN does not provide for the naming of species groups, we have decid- ed to continue the usage of melanopterus in this paper. KEY TO THE NORTH AMERICAN SPECIES OF THE MELANOPTERUS SPECIES-GROUP 1. Width of oral opening about equal to height of face (from clypeus to antennal sockets); malar space about equal to basal width of mandible (Figs. 8, 9) 2 Width of oral opening greater than height of face; malar space much less than basal width of mandible (Figs. 5-7) 3 2(1). First and second metasomal terga strongly porcate (Fig. 11); fore wing vein ICUa longer than Icu-a, wings strongly infumated (Fig. 3) politiceps (Gahan) First and second metasomal terga weakly costate rugulose (Fig. 10); fore wing vein ICUa equal in length to Icu-a, wings hyaline or weakly infumated miani Marsh and Shaw, new species 3(1). Body bicolored, at least legs black, differently colored than body 4 - Body unicolored, legs concolorus with rest of body 5 4(3). Head and legs entirely black mexicanus Cresson Head orange, femora, tibiae and tarsi only black melanopodus Marsh and Shaw, new species 5(3). Fore wing vein ICUa as long as or shorter than vein Icu-a (Fig. 4); clypeus flat, narrow, without distinct apical border megastomus Marsh and Shaw, new species Fore wing vein ICUa longer than Icu-a (Fig. 1); clypeus protruding, with distinct apical carinate border mandibttlaris (Cresson) Aleiodes mandibularis (Cresson), new combination (Figs. 1, 6) Rogas mandibularis Cresson 1872:188. Diagnosis. — Body unicolored orange, an- terma brown, wings hyaline or slightly dusky, veins brown; body length, 8.5-10.0 mm; 55-65 antennomeres; oral opening width greater than height of face (Fig. 6); malar space short, equal to or less than basal width of mandible; face costate, frons, vertex and temple smooth; ocelli small, diameter of lateral ocellus less than ocell-ocular distance; pronotum rugose medially; mesonotum, scutellum and me- sopleuron punctate and shining, subalar sulcus and sternaulus weakly rugose; pro- podeum rugose, median carina complete; first, second and basal Vi of third metaso- mal terga costate-rugose, median carina complete to middle of third terga; fore wing vein Icu-a beyond IM by distance greater than length of Icu-a, vein ICUa longer than Icu-a (Fig. 1); hind wing mar- ginal cell narrowest basally, gradually widening apically, vein RS straight, vein m-cu present (Fig. 1); tarsal claws strongly pectinate. Type material examined. — Holotype male of Rogas mandibularis Cresson is not in the ANSP and apparently lost. A neotype has been selected as follows: male, TEXAS, Victoria, September 18, 1904. Deposited in USNM. Distribution. — Scattered collections throughout central United States from Ohio south to Georgia, west to Nebraska and Texas. Biology. — Host unknown. Comments. — Although mandibularis is a large and distinctive species, it is not com- monly collected. Superficially it resembles politiceps because of its large orange body, but the more greatly enlarged oral open- ing, robust mandibles, and hyaline wings will easily distinguish mandibularis from that species. Volume 8, Number 1, 1999 101 Figs. 1^. Wings of Aleiodes species: 1, iiiaiidibulans (Cresson); 2, nichuwpwdiif n. sp.; 3, politicei's (Gahan); 4, iiifgiKtonutf n. sp. Aleiodes megastomus Marsh and Shaw, new species (Figs. 4, 5) Female. — Body color: varying from dark honey yellow to dark brown or black, legs and mandibles always brown; wings hy- aline, veins including stigma dark brown, stigma occasionally lighter brown or yel- low. Body length: 7-9 mm; fore wing length, 7-9 mm. Head (Fig. 5): 57-61 an- tennomeres, flagellomeres beyond first slightly longer than wide, first flagello- mere nearly twice as long as second; malar space short, !/, eye height and % basal width of mandible; temple % eye width; occipital carina not meeting hypostomal carina; oral opening very wide and oval, width three times malar space and slightly greater than height of face; clypeus flat. without distinct apical carinate border; ocelli large, greatest diameter of lateral ocellus slightly greater ('4) than ocell-ocu- lar distance; face rugulose, frequently with median longitudinal ridge below anten- nae; frons smooth, occasionally rugulose behind antennae; vertex punctate; temples punctate, rugulose near base of mandi- bles; maxillary palpus not swollen; man- dibles large, when closed Hps going well beyond middle of oral opening. Mesoso- ma: pronotum rugulose; mesonotum and scutellum smooth but often with conspic- uous hair pits making it appear punctate; notauli weakly scrobiculate, meeting pos- teriorly in triangular rugose area; scutellar furrow wide, scrobiculate posteriorly, separated from mesonotum by transcutal ridge; mesopleuron smooth, subalar groove 102 Journal of Hymenoptera Research scrobiculate; stemaulus absent; propodeum weakly rugulose dorsally, smooth laterally, propodeal carina present but often weak. Legs: tarsal daw weakly pectinate, with 3- 4 short teeth near base; inner spur of hind tibia slightly less than half length of hind basitarsus; hind coxa rugulose dorsally. Wings (Fig. 4): hyaline; fore wing vein r Vi length of 3RSa and of m-cu, vein Icu-a beyond IM by distance less than length of Icu-a, vein ICUa about '4 length of ICUb, vein ICUa about equal to Icu-a; hind wing marginal cell gradually widening from about middle to wing margin, vein r-m equal to or slightly shorter than IM, vein M+CU only slightly longer than IM, vein m-cu present. Metasoma: first ter- gum weakly strigate with raised smooth triangular area at base, as long as apical width, basal width Vi apical width, medi- an carina usually present, often weak and occasionally absent; second tergum weak- ly strigate, often smooth at apex, median carina weak and often absent; third and following terga smooth, third occasionally punctate at base, median carina absent on third tergum; ovipositor sheaths % length of hind basitarsus. Male. — Essentially as in female. Ho/ofi/;'f.— Female: CALIFORNIA, Cor- coran, Kings County, March 10, 1965, F. H. Surber, light trap. Deposited in USNM. Paratypes.— ARIZONA: 2 females. No- gales, January 30, 1954, March 9, 1961, A. C. Valcarce; 1 female, Prescott, April 22, 1936, Cby. and Bish.; 3 females, Tempe, February 5, 1923, February 8, 1926, E. V. Walter; 10 females, 5, males, Tuscon, Jan- uary 27, 1935, February 4, 1935, December 10-20, 1939, January 8, 1940, January 26, 1953, R. H. Crandall, Hubbard, Samual Green, E. C. Narschall, Donald Foote, G. C. Butler, Bryant collectors, some at light; 1 female, Cameron, Coconimo Co., Feb- ruary 28, 1978, R. C. Miller; 1 female, Bea- ver Dam, Mojave Co., April 2, 1969, P. M. Marsh; 1 fenale, 1 male, Baboquivari Mtns., August 1, 1924, O. C. Poling. CAL- IFORNIA: 1 female, Manzanita Lake, Las- sen Nat. Park, May 23, 1941, P. D. Hurd; 1 female, 1 male. May Lake, Yosemite Pk. July 26, 1948, H. K. Townes; 6 females, 5 males, 3 mi. SW Somerset, El Dorado Co., dates ranging from April 20, 1974 to May 20, 1978, R. Wharton; 2 females, 2 males. Thousand Palms, February 14-22, 1955, W.R.M. Mason, J.E.H. Martin; 1 female. Desert Hot Springs, February 14, 1965, J.E.H. Martin; 1 female, Wrightwood, San Bernardino Co., June 16, 1964, J. S. Buck- ett; 2 females, Quincy, Plumas Co., June 5, 1963, G. Jeskey; 1 female, McClure Vly., Kings Co., March 6, 1953, J. C. Hall; 1 fe- male, Huntington Lake, June 26, 1961, A. T. McClay; 5 females, 1 male, Calaveras Co., 4.8 km S West Point, May 1-9, 1981, Stanley C. Williams; 1 female, Napa Co., N. side Howell Mtn., 2 mi NNE Angwin, 1300 ft., April 11, 1978, H. B. Leech; 1 fe- male, Napa Co., 1 mi N. Angwin, May 9, 1964, H. B. Leech; 1 female, Nevada City, May 16, 1930, E. P. Van Duzee; 1 female, Fresno Co., Coalinga, Los Gatos Cn., March 17, 1931, E. P. Van Duzee; 1 female, Mt.. Diablo, April 23, 1939. IDAHO: 1 male, Murtaugh, May 27, 1931, D. E. Fox; 2 females, Lewiston, April 27, 1936, alt. 550 ft., R. E. Miller. KANSAS: 1, female, Manhattan, April 24, 1926, R. T. Cotton, at light. MONTANA: 1 female, Gallatin Co., May 8, 1932. NEVADA: 1 female, Reno, May 14, 1915; 1 female, Kyle Cyn., Chlstn, Mt., May 25, 1940, G. E. Bohart. UTAH: 1 female. Blue Spruce Camp, 18 mi. N Es- calante, Garfield Co., 8000 ft., July 30, 1965, F., P. & M. Rindge; 6 females, Rich- field, May 22, 1929, May 28, 1930, light trap; 2 females, 1 male, Wellsville, May 17, 1961, G. E. Bohart; 1 female. Baker, S14,T13N,R69E, May 4, 1939, T. O. Thatch- er. WASHINGTON: 3 females, 1 male, Yakima, June-September, March 30, 1932, at light, A. R. Rolfs; 1 male, Orville WY- OMING: 2 females, Stratton Expt. Water- shed, nr. Saratoga, May 16-17, 1972, June 4-8, 1972, J. M. Schmid; 26 females, 2 males, Albany Co., T15N R73W, 7500', May 15, 1978, May 20, 1980, C. D. Ferris; Volume 8, Number 1, 1999 103 1 female, Carbon Co., 3 mi. N of WY 130W, between mi. 54 and 55, June 25- July 4, 1991, Malaise trap, Mian; 1 female, Fremont Co., Sinks Cyn., 12.1 mi. SW Lander, June 23, 1980, LT, Mike Pogue, Robert Lavigne; 2 females, 1 male, Albany Co., 1 mi. E Laramie, May 10-18, 1994, June 8, 1996, BLT 2217, Skyview Lane, 7465 ft., J. S. Nordin. SASKATCHEWAN: 1 female, Saskatoon, May 12, 1933, A. R. Brooks. Deposited in USNM, RMSEL, AMNH, TAMU, MCZ, MSU, CNC, CAS, UCD, NNML. Distribution. — This species is common in western North America from Saskatche- wan to New Mexico and west. The single specimen from Kansas is this species and with more collecting it probably will occur throughout the western half of North America. Biology. — No host records are available. Many of the type series were collected at light or in light traps. The general habitus and color of this species is similar to spe- cies of Homolobus which frequent light traps. Comments. — This species is similar to mandibularis because of the wide oral space. It differs in several characters: the ocellus of megastomus is larger than in mandibularis, the clypeus of megastomus is flat rather than protruding as in mandibu- laris, and vein Icu-a in the fore wing of megastomus is closer to IM than in mandi- bularis. Eti/mologi/. — The specific name is from the Greek megas meaning large or great and stoma meaning mouth referring to the large wide oral space. Aleiodes tnelanopodus Marsh and Shaw, new species (Figs. 2, 7) Female. — Body color: head, mesosoma, metasoma, coxae and trochanters orange; antenna, palpi, ovipositor sheaths, and legs beyond trochanters black, trochanter 2 sometimes dark; wings dusky. Body length: 8-9 mm, fore wing length, 7-8 mm. Head (Fig. 7): 53-58 antermomeres, first flagellomere twice as long as wide, flagellomeres 10-30 as wide as long; malar space short, Va eye height and % basal width of mandible; temple bulging, about as wide as eye; occipital carina nearly meeting hypostomal carina; oral opening wide and oval, width 3.5 times malar space and about twice height of face; clyp- eus protruding; ocelli small, greatest di- ameter of lateral ocellus about % ocello- cular distance; face costate, with median longitudinal ridge below antennae; frons smooth, occasionally weakly rugulose be- hind antennae; vertex and temple smooth; maxillary palpus not swollen; mandibles large, when closed tips going beyond mid- dle of oral space. Mesosoma: pronotum rugulose laterally; mesonotum and scutel- lum smooth and shining; notauli weakly scrobiculate, meeting posteriorly in trian- gular rugulose area with central longitu- dinal carina; mesopleuron smooth and shining, subalar sulcus rugulose, sternau- lus short and wide, weakly scrobiculate; propodeum strongly rugose dorsally, smooth laterally with rugosities near hind coxa, median carina complete. Legs: tarsal claws strongly pectinate, with 6-8 teeth on entire inner surface of claw; hind coxa smooth dorsally. Wings (Fig. 2): dusky; fore wing vein r K length of 3RSa and % length of m-cu, vein Icu-a beyond IM by distance about twice length of Icu-a, vein ICUa about X length of ICUb, vein ICUa longer than Icu-a; hind wing marginal cell gradually widening from about middle to wing margin, vein r-m about % length of IM, vein M + CU about 1.4 times longer than IM, vein m-cu present. Metasoma: first tergum costate-rugose, length slightly less than apical width, median carina com- plete; second tergum costate-rugose, me- dian carina strong on basal half; third ter- gum costate on basal %, median carina ab- sent; fourth and following terga smooth, fourth weakly coriaceous on basal half; ovipositor sheaths about 1.5 times longer than hind basitarsus. 104 Journal of Hymenoitera Research Figs. 5-8. Faces of Aleiodes species: 5, mcgnstomiis n. sp.; 6, numdibulnnf (Cresson); 7, nninuopiMiuf n. sp.; 8, iniaiii n. sp. Male. — Essentially as in female. Holoh/pe.— Female: MARYLAND, Cabin John, (label actually states Washington, DC which is in error), September 24, 1922, H. S. Barber collector. Deposited in USNM. Parntypes.— KENTUCKY: 1 female, 1 male. Golden Pond, October 1-8, 1964, September-October, 1965, S. G. Breeland. MARYLAND: 1 male. Cabin John, Octo- ber 1, 1916, R. M. Fouts collector; 1 male. Bethesda, September 23, 1914, R. C. Shan- non collector; 1 female. Prince George's Co., Beltsville, September 19, 1964, Paul H. Arnaud, Jr.; 1 female, Riverdale, Prince George's Co., September 26, 1979, T. Wa- ters; 1 female, Takoma Park, September 29, 1945, H. and M. Townes. LOUISIANA: 1 male. Sunshine, October 27, 1972, V. A. Brou. NORTH CAROLINA: 1 male. Wake Co., October 16, 1959; 1 female. Wake Co., October 3, 1959, O. F. Vargas; 1 male. Volume 8, Number 1, 1999 105 Clayton, Cent. Crops. Res. Sta., 1959, A. Saldarriaga; 1 male, Raleigh, October 3, 1959, L. Self; 1 female, Pender Co., October 9, 1954, D. A. Becker; 1 male. Wake Co., Raleigh, Centienial Campus, October 13, 1992, E. D. Karoly; 1 male. Wake Co., Ra- leigh, November 5, 1988, D. L. Moncol. VIRGINIA: 1 female, Richmond, collection Ashmead; 1 female. Falls Church, Septem- ber 26 (no year). Banks collector. Depos- ited in USNM, RMSEL, MCZ, CAS, NCSU, FSCA, AEI. Distribution. — Known from Maryland, Virginia, North Carolina, Kentucky and Louisiana. Biology. — Unknown. Comments. — This species is very similar to mandibiilaris by the shape of the mouth and the body color. It differs in having black legs, smaller ocelli, and longer ovi- positor. Etymology. — The specific name is from the Greek melanos meaning dark or black and the Greek podos meaning foot refer- ring to the distinctive black legs. Aleiodes mexicanus Cresson Aleiodes mexicamis Cresson 1869:378. Diagnosis. — Body bicolored, head, an- tenna, pronotum, propleuron and legs black, mesosoma and metasoma orange, wings dusky, veins brown, tegula orange; body length, 9 mm; 64 antennomeres; ma- lar space very short, about V2 basal width of mandible; oral opening large, diameter 5 times malar space length and about equal to face height; ocelli large, ocello- cular distance less than diameter of lateral ocellus; frons, vertex and temple smooth, face rugulose; mesonotum and mesopleu- ron smooth; propodeum with dorsal api- cal corners sharp, median carina complete; first, second and basal % of third metaso- mal terga strigate, median carina complete of first and second terga which also have lateral carinate edges; fore wing vein Icu- a beyond vein IM by distance slightly greater than length of Icu-a, vein ICUa slightly longer than Icu-a; hind wing mar- ginal cell narrowest at base, gradually widening to apex, vein RS straight; tarsal claws pectinate, with 4-6 slender spines. Type matericd examined. — Aleiodes mexi- canus Cresson, holotype female (not male as in original description), Mexico [ANSP]. Distribution. — In addition to Mexico, we have seen one specimen from Mississippi. Biology. — Unknown. Comments. — This species is distinctive by its color, short malar space and wide oral opening. The one specimen from Mis- sissippi may be an accidental introduction and the establishment of this species in the U.S. needs to be confirmed by further col- lecting in the southern United States. Shaw (1993) provides a description and key to this species and other related Neo- tropical species. Aleiodes miani Marsh and Shaw, new species (Figs. 8, 10) Female. — Body color: honey yellow ex- cept antenna, ocellar triangle, palpi, pro- pleuron, mesosternum, metanotum, pro- podeum, basal half of first metasomal ter- gum, apex of all femora, fore and middle tibiae, apical half of hind tibia and all tarsi black, wings hyaline, veins dark brown, tegula yellow. Body length: 7.0-9.0 mm; fore wing length, 5.0-7.0 mm. Head (Fig. 8): eyes and ocelli not unusually large; 59- 60 antennomeres, all flagellomeres beyond the first about as wide as long; malar space shorter than basal width of mandi- ble and about Va eye height; temple broad, slightly less than eye width; occipital ca- rina complete on vertex, not reaching hy- postomal carina; oral opening large, cir- cular, width greater than basal width of mandible and face height; clypeus narrow, not swollen; ocelli small, diameter of lat- eral ocellus slightly less than ocellocular distance; face, frons and malar space ru- gose, temple and vertex smooth except for hair pits; maxillary palpus not swollen; 106 Journal of Hymenoptera Research Figs. 9-11. Morphological features of Alciccics species: 4, face ol poltluc^'^ (Cjahaii); 10, metasomal terga of miani n. sp.; 11, metasomal terga oi poUticqps (Gahan). Volume 8, Number 1, 1999 107 mandibles large, tips crossing when closed. Mesosoma: pronotum costate ru- gose laterally, costate ventrally; mesono- tum and scutellum smooth except for hair pits, notauli weakly scrobiculate, meeting in shallow rugose area before scutellum; mesopleuron smooth, subalar sulcus ru- gose, sternaulus carinate; propodeum ru- gose dorsally, smooth laterally, median carina weak but complete. Legs: tarsal claws weakly pectinate; hind coxa punc- tate dorsally. Wings: fore wing vein r about Vs length of 3RSa and m-cu, vein Icu-a beyond IM by distance equal to length of Icu-a, vein ICUa about Va length of ICUb; hind wing vein RS straight, mar- ginal cell gradually broadening to apex, vein r-m shorter than IM, vein M + CU slightly longer than IM, vein m-cu weak and leaving IM before junction of r-m. Metasoma (Fig. 10): first tergum costate rugose, apical width about equal to length, median carina complete; second tergum costate rugose, median carina not quite complete; third tergum costate ru- gose on basal half, coriaceous on apical half, median carina absent; remainder of terga coriaceous; ovipositor less than Vi length of hind basitarsus. Male. — Essentially as in female except mesopleuron and pronotum nearly and first metasomal tergum entirely black. Ho/ofi/p.— Female: WYOMING, Carbon Co., 3/4 mi. N of WY 130 W between mi. 56 & 55, mixed forest near water. Malaise, Mian, July 14-27, 1991. Deposited in RMSEL. ParaU/pes.— WYOMING, 1 female, 4 males, same data as holotype with dates of June 25 through August 6, 1991. De- posited in USNM, RMSEL. Distribution. — Known only from Wyo- ming. Biologi/. — Unknown. Comments. — This species is similar to mandihularis but can be distinguished by its different coloration with the black pro- podeum and first tergum and its smaller oral opening. Etymology. — Named for Mian Inayatul- lah who collected the type series. Aleiodes politiceps (Gahan), new combination (Figs. 3, 9, 11) Rogns politiceps Gahan 1917:206. Diagnosis. — Body unicolored orange, an- tenna black, legs orange except apex of hind tibia and all tarsi black, wings black; body length, 7.0-9.0 mm; 60-65 anten- nomeres; malar space shorter than basal width of mandible; oral opening large (Fig. 9), diameter greater than malar space; ocell-ocular distance greater than diameter of lateral ocellus; head smooth and shining, face weakly costate; mesono- tum, scutellum and mesopleuron smooth and shining, propodeum areolate-rugose, median carina complete; first, second and basal half of third metasomal terga strong- ly porcate (Fig. 11), median carina com- plete on first and second terga; fore wing vein Icu-a beyond IM by distance greater than length of Icu-a (Fig. 3); hind wing marginal cell gradually widening, vein RS straight (Fig. 3); tarsal claws pectinate. Type material examined. — Rogas politiceps Gahan, holotype female, Nashville, Ten- nessee (USNM). Distribution. — Virginia south to Florida, west to Arkansas and Texas. Also record- ed from San Jose, Costa Rica (new record), at an elevation of 1300m. Not yet recorded from other parts of Central America. Biology. — The usual hosts appear to be moderately large noctuid larvae, includ- ing Anicla infecta (Ochs.) and Pseudaletia unipimcta (Haw.). Comments. — This is one of the common- est large Aleiodes species occurring in the southern states. It is frequently taken in Malaise traps, and in large numbers at black lights. The entirely dark wings, bright orange body, and strongly porcate metasomal terga make this species quite distinctive. It is one of the few Aleiodes species that can be reliably identified even 108 Journal of Hymenoptera Research without a microscope. The very coarse sculpture of metasomal terga is unique among the species covered in this study. ACKNOWLEDGMENTS We thank the curators of all the collections listed in Methods for the loan of specimens for this study. Kent Hampton, Kansas State University, prepared the scanning electron micrographs and Linda Lawrence, USDA Systematic Entomology Laboratory, prepared the wing drawings. This research was supported, in part, by grant DEB-930-6314 from the National Sci- ence Foundation to S. R. Shaw. Additional support was provided by supplemental REU grants in 1994, 1995 and 1996 (Research Experience for Undergrad- uates). LITERATURE CITED Cresson, E. T. 1869. List of the North American spe- cies of the genus Aleiodes Wesmael. Tnuisactioiis of the Ainericnii Eittcmologicnl Sccieti/ 2:377-382. Cresson, E. T. 1872. Hymenoptera Texana. Transac- tions of the American Entomological Society 4:153- 292. Fortier, J. C. 1997. Cladistics of the Aleioiies lineage of the subfamily Rogadinae (Hymenoptera: Bracon- idae). Unpublished Ph.D. thesis submitted to the University of Wyoming. Gahan, A. B. 1917. Description of some new parasitic Hymenoptera. Proceedings of the United States Na- tional Museum 53:195-217. Goulet, H. and J. T. Huber. 1993. Hytneiioptera of ttie World: an identification guide to families. Agricul- ture Canada Publication 1894 /E, Ottawa. Harris, R. A. 1979. A glossary of surface sculpturing. Occasional Papers in Entomology of the California Department of Food and Agriculture no. 28, pp. 1- 31. Marsh, P. M. 1989. Notes on Braconidae (Hymenop- tera) associated with jojoba {Simmondsia cliinen- sis) and descriptions of new species. Pan-Pacific Entomologist 65:58-67. Marsh, P. M. and S. R. Shaw. 1998. Revision of North American Aleiodes Wesmael (Part 3): the seriatus (Herrich-Schaeffer) species-group (Hymenop- tera: Braconidae, Rogadinae). Proceedings of the Entomological Society of Washington 100(3):395- 408. Shaw, M. R. 1983. On[e] evolution of endoparasitism: the biology of some genera of Rogadinae (Bra- conidae). Contributions of the American Entomolog- ical Institute 20:307-328. Shaw, M. R. 1994. Chapter 7, Parasitoid host ranges. Pp. 112-144, In: Hawkins, B. A. and W. Sheehan [eds.], Parasitoid Community Ecology. Oxford Uni- versity Press, Oxford. Shaw, M. R. and T. Huddleston. 1991. Classification and biology of braconid wasps. Handbooks for the Identification of British Insects 7:1-126. Shaw, S. R. 1993. Systematic status of Eucystomastax Brues and Characterization of the Neotropical species (Hymenoptera: Braconidae: Rogadinae). Journal of Hymenoptera Research 2:1-11. Shaw, S. R. 1995. Chapter 12.2, Braconidae. Pp. 431- 463, In: Hanson, P. E. and I. D. Gauld |eds.|. The Hymenoptera of Costa Rica. Oxford University Press, Oxford. Shaw, S. R., P. M. Marsh and J. C. Fortier. 1997. Re- vision of North American Aleiodes Wesmael (Part 1): the pulchripes Wesmael species-group in the New World (Hymenoptera: Braconidae, Roga- dinae). Journal of Hymenoptera Research 6(l):10-35. Shaw, S. R., P. M. Marsh and J. C. Fortier. 1998a. Re- vision of North American Aleiodes Wesmael (Part 2): the ductcr Thunberg species-group in the New World (Hymenoptera: Braconidae, Rogadinae). Journal of Hymenoptera Research 7(l):62-73. Shaw, S. R., P. M. Marsh and J. C. Fortier. 1998b. Re- vision of North American Aleiodes Wesmael (Part 4): the albitibia (Herrich-Schaeffer) and praetor (Reinhard) species-groups (Hvmenoptera: Bra- conidae, Rogadinae). Proceedings of the Entomolog- ical Society of Washington 100(3):553-565. Wharton, R. A., P. M. Marsh and M. j. Sharkey, eds. 1997. Manual of New World genera of the family Braconidae. Special Publication of the International Society of Hynienopterists. No.l, 438 pp. J. HYM. RES. Vol. 8(1), 1999, pp. 109-115 Further Evidence of Male Antennal Glands in Aphelinidae: The Case of Aphytis melimis DeBach (Hymenoptera: Aphelinidae) R. ROMANI, N. ISIDORO, AND F. BiN Agricultural Entomology Institute, University of Perugia, 06121 Perugia, Italy, E-mail: Fbin@unipg.it Abstract. — Ultrastructural investigations of the elongated male club, antennomere 6, in Aphytis melimis DeBach have shown that a small oval ventral area, bearing numerous minute setae, is not a sensory complex but rather the release site of a glandular complex with unicellular unit type 1. This finding, combined with behavioral observations reported in the literature, strongly indicates that the secretion induces sex recognition during pre-coital phase. This morpho-functional inter- pretation is discussed in other aphelinids exhibiting similar structures on different antennomeres. In the last few years, some peculiar an- tennal structures of parasitoid hymenop- terans have received increasing interest, and have revealed new aspects for func- tional morphology, biosystematics, and phylogeny (Isidoro et al. 1996). In fact, the male antennae, which were previously be- lieved to be only sensory appendages, have instead been demonstrated as also having a secretory function through epi- dermal glands involved in courtship be- havior (Bin et al. 1997). The secretory func- tion has been reported in families Eulo- phidae (Dahms 1984), Scelionidae (Bin and Vinson 1986), Platygastridae (Isidoro and Bin 1995), Ichneumonidae (Isidoro et al. 1997), Diapriidae (Romani et al. 1997; Sacchetti et al. 1997) and Eucoilidae (Isi- doro et al. submitted). Male antennal glands were reported for the first time in Encarsia asterobemisiae Vig- giani et Mazzone (Pedata et al. 1995), a species supplied with two peculiar ventral features, respectively on antennomeres four and five, which likely appear to be used during pre- and post-coital phases (Viggiani and Laudonia 1989). In another aphelinid, Aphytis iiielinus DeBach, an im- portant worldwide biocontrol agent of cit- rus scales, the courtship behavior was also described stressing the importance of an- tennation during the precoital phase (Gordh and DeBach 1978). The occurrence of this behavior, along with the presence of a "specialized sensory area" on the ven- tral side of the male elongated club, has prompted an ultrastructural study of this "small oval area or plate bearing numer- ous minute setae" (Rosen and DeBach 1979). This study proves that in A. melimis such an area, present on A6 in the form of an elongated club, is the release site of ad- jacent integumentary glands. In addition, for the first time in hymenopterans, glands are reported on the apical antennomere. MATERIALS AND METHODS A colony of A. meliuus, laboratory reared on Aonidiella aiirantii (Maskell), was provided by the Entomology Department of the University of California (Riverside). For scanning electron microscopy (SEM) observations, 10 males, newly emerged and anaesthetized in CO,, were beheaded and immediately immersed in 50% etha- nol water solution and kept overnight at 4°C. After dehydration in a graded ethanol series, the heads with antennae were crit- ical point dried in a Balzers Union CPD no Journal of Hymenoptera Research 020 unit, gold coated in a Balzers Union SCD 040 unit, and finally examined with a Philips XL 30. For transmission electron microscopy (TEM) observations, 10 males were anaes- thetized in C02 and immediately im- mersed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer + 5% sucrose, pH 7.2- 7.3. The apical antennomeres were de- tached to aid fixative penetration, and left at 4°C for 2h. After rinsing overnight in a cacodylate buffer, the specimens were postfixed in 1% osmium tetroxide at 4°C for Ih and rinsed in the same buffer. De- hydration in a graded ethanol series was followed by embedding in Epon-Araldite with propylene oxide as a bridging sol- vent. Thin sections were taken with a di- amond knife on a L.K.B. "Nova" ultra- microtome, and mounted on collodium- coated 50 mesh grids. Finally, the sections were investigated with a Philips EM 400T, after staining with uranyl acetate (20 min, room temperature) and lead citrate (5 min, room temperature). RESULTS The geniculate antennae of male A. mel- inus consist of six antennomeres. The api- cal antennomere, A6 or club, is elongated and bears an oval area on the proximal ventral side which is the release site struc- ture (RSS) of the integumentary glands (Fig la). This area is covered by minute, non-socketed microtrichia and bordered by one row of socketed trichoid sensilla (Fig lb); while the former are not inner- vated the latter are provided of one mech- anosensory neuron. SEM observations of the ventral side of A6 show the oval area slightly depressed and partially cut off from the surrounding club portion by in- distinct grooves. Serial longitudinal and cross sections of the apical antennomere reveal a well de- veloped glandular epithelium adhering to the internal wall of the oval area (Fig 2a). This glandular complex consists of nu- merous, unicellular secretory units vary- ing in size and shape. Each glandular cell has a large, round and regularly shaped nucleus which is often located in the basal part of the cell (Fig 2b). Chromatin is not abundant and most of it is apposed to the nuclear membrane. The perinuclear region of the cytoplasm contains mitochondria with conventional cristae and abundant free ribosomes while few signs of granular or smooth endoplasmic reticulum were observed. The basal plasma membrane of the cell has deep, irregular invaginations forming a lacunar system. The apical cell membrane is surrounded by densely packed microvilli, delimiting a narrow ex- tracellular space. The cuticle associated with the glandular epithelium is pierced by numerous tiny pores randomly distrib- uted. The secretory apparatus underneath each pore is formed by a spherical cham- ber, from which numerous cuticular fila- ments radiate (Fig 2c). These filaments, apparently a specialization of the external epicuticle, have a tubular structure and extend deeply in the extracellular space between the microvilli of the apical cell membrane. DISCUSSION The "small oval area or plate bearing numerous minute setae" on the ventral side of male club in A. melinus (Rosen and DeBach 1979) is not a "specialized sensory area" but the release site of epidermal glands. These glands are unicellular secre- tory units belonging to the type 1 gland cell (Noirot and Quennedey 1974, 1991; Quennedey 1998). The cytological features of the secretory cells do not allow us to attempt an interpretation on the nature of the secretion which in other hymenopter- ans acts on contact (Isidoro et al. 1996) or is volatile (Felicioli et al. 1998). The pecu- liar releasing apparatus consists of numer- ous pores so tiny that neither the external openings nor the material secreted can be seen with SEM, contrary to what has been reported for other parasitoids (Bin and Vinson 1986; Isidoro and Bin 1995). The Volume 8, Number 1, 1999 111 Fig. 1. Aplii/tis "iWdiks male: a) \'cntr(.)-latLTal \ievv of antcnnomcre fa (Afa) showing the relcabo bile structure (RSS); b) detail of the RSS covered by numerous microtrichia (MI) and bordered by a single row of tactile setae (TS). conceivable function of no-socketed mi- Encarsia asterobemisiae has two glandular crotrichia is that of increasing the release complexes, respectively on A3 and A4, be- surface while that of socketed trichoid longing to the same type 1 but different in sensilla is that of perceiving tactile stimuli, cytological characteristics and release site 112 Journal of Hymenoptera Research 5 tjm ilmf-^-ar' Fig. 2. Apln/ti-i niiiniu^ male: j) cross section ot antennomere 6 at about half level showing the extension of the glandular area (GA) on the ventral side; b) perinuclear detail of a secretorv cell; c) apical detail of a secretory cell showing the tiny cuticular pores (P). N, nucleus; EF, epicuticular filaments; H, haemocele; M, mitochondrion; MV, microvilli; R, ribosomes. Volume 8, Number 1, 1999 113 features. This could indicate a difference in composition and role of the relative se- cretions (Pedata et al. 1995). From these two ascertained cases it appears that the so called sensorial complexes described in other aphelinids need to be investigated to define their real nature. Morphological and behavioral observations strongly sug- gest in fact that glands are present on dif- ferent antennomeres involving the scape (Al), from one to 3 intermediate anten- nomeres (A3 to A5) or the apical anten- nomere (A6). Some examples are as fol- lows: On Al of Physcus testaceus Masi there are special structures which could be re- lated to a gland (Viggiani et al. 1986) and something similar occurs in Pteroptryx chi- nensis (Howard) (Viggiaru and Ren 1993). Numerous species of Encarsia Foerster have one up to three antennomeres, A3- A5, which may or may not be noticeably enlarged, and ventrally provided of one specialized structure or two different ones. When two or more antennomeres are glanded the release sites are of two types; furthermore, one enlarged antennomere does not necessarily bear a specialized structure, as A3 in the case of £. asterobem- isiae (Pedata et al. 1995). Therefore, some of the following examples having two or three enlarged antennomeres may need to be confirmed: A3 in E)icar^ia albiscutellum (Girault) (Viggiani 1985), A4 in Encarsia aleurotubae Viggiani, £. hemdoui (Girault) (Viggiani 1987), A3-A4 in Encarsia olivina (Masi) (Viggiani and Mazzone 1982), £. g(- gas (Tchumakova), £. optileiita Silvestri and £. perniciosi (Tower) (Viggiani and Laudonia 1989), A3-A5 in Encarsia antiopa (Girault) (Viggiani 1985). As regards the apical antennomere the A6 of several species of Aphytis Howard, as illustrated by Rosen and DeBach (1979), have external features similar to those de- scribed for melinus, and therefore could have glands: A. cochereaui DeBach and Ro- sen, A. fabresi DeBach and Rosen, A. chi- lensis Howard, A. coluinbi (Girault), A. par- amaculicornis DeBach and Rosen, A. dias- pidis (Howard), A. lingnanensis Compere, A. roseni DeBach and Gordh, A. pilosus DeBach and Rosen. Behavioral observations indicate that the antennation mechanism during mat- ing is different in species having one or two glanded antennomeres. While the sin- gle release site seems to be logically relat- ed to the pre-coital phase, e.g. triggering sex-recognition, female stimulation or fe- male sedation (Gordh and DeBach 1978; Viggiani t'f al. 1986), the presence of a sec- ond and different release site has been ob- served in species performing also a post- coital phase, e.g. E. asterobemisiae (Pedata et al. 1995). These chemo-physical mechanisms can only partially explain the sexual isolation since there also are other chemicals in- volved in mating behavior. In some spe- cies of Aphytis a female sex pheromone (Rao and DeBach 1969) and a male aph- rodisiac pheromone (Khasimuddin and DeBach 1975), whose sources are as yet unknown (Gordh and DeBach 1978), have in fact been reported. In another aphelin- id, Aphelinus asychis Walker, the mate finding is mediated by a female trail sex pheromone deposited in a yet unknown way on the substrate while walking (Fau- vergue et al. 1995). As it is being shown in other parasitoid groups (Bin et al. 1997; Isidoro et al. sub- mitted) the glandular nature of the anten- nal structures provides a new perspective to define the sex selection strategies of aphelinids as well as additional characters for taxonomy and phylogeny. ACKNOWLEDGMENTS We are verv grateful to I'rof. R. F. Luck, who kind- ly supplied insects. We also thank C. Dentini for tech- nical assistance with fixation and embedding of the specimens and A. Mommi for film processing and photographic printmg. SEM and TEM pictures were made using the Electron Microscopy Center of Pe- rugia University (CUME). This research was finan- cially supported by the Ministry for University and Scientific Technological Research (M.U.R.S.T. 40%). 114 Journal of Hymenoptera Research The authors have contributed equally to different as- pects of this paper. LITERATURE CITED Bin, F. and S. B. Vinson. 1986. Morphology of the anteruial sex-gland in male Trissolcus basalis (Woll.) (Hymenoptera: Scelionidae), an egg par- asitoid of the green stink bug, Nczara viriiiula (Hemiptera: Pentatomidae). bilcrnationnl journal of Insect Morphologfi and Enilm/ology 15: 129-138. Bin, F., N. Isidoro, R. Romani, and S. B. Vinson. 1997. Antennal functional areas for sex recognition in some parasitoid hymenopterans. Boletin de la Asociacion cspaiiola de Entomologia, Suplemento al Volumen n°2V. 68-69. Dahms, E. C. 1984. An interpretation of the structure and function of the antennal sense organs of Mel- tittohia australica (Hymenoptera: Eulophidae) with the discovery of a large dermal gland in the male scape. Memoirs of the Queensland Museum 21(2): 361-377. Fauvergue, X., K. R. Hopper, and M. F. Antolin. 1995. Mate finding via a trail sex pheromone by a par- asitoid wasp. Proceedings of the National Academy of Siences of the U. S. A. 92: 900-904. Felicioli, A., N. Isidoro, R. Romani, F. Bin, and M. Pinzauti. 1998. Ethological and morphological analysis of mating behaviour in Osniia coriiuta Latr. (Hymenoptera, Megachilidae). Insect Social Life 2: 137-144. Gordh, G. and P. DeBach. 1978. Courtship behaviour in the Aphytis lingnanensis group, its potential usefulness in taxonomy, and a review of sexual behavior in the parasitic Hymenoptera (Chalci- doidea: Aphelinidae). Hilgardia 46: 37-75. Isidoro, N. and F. Bin. 1995. Male antennal gland of Atnitus spiniferus (Brethes) (Hymenoptera: Platy- gastridae), likely involved in courtship behavior. International journal of Insect Morpliolog\/ and Em- bryology 24: 365-373. Isidoro, N., F. Bin, S. Colazza, and S. B. Vinson. 1996. Morphology of antennal gustatory sensilla and glands in some parasitoid Hymenoptera with hy- pothesis on their role in sex and host recognition. journal of Hymenoptera Research 5: 206-239. Isidoro, N., F. L. Wackers, R. Romani, and F. Bin. 1997. Antennal tyloids are release structures of sex recognition pheromone in Pimpla turionellae (L.) (Hymenoptera, Ichneumonidae). Boletin de la Asociacion espanola de Entonwlogia. Suplemento al Volumen «°21: 70-71. Khasimuddin, S. and P. DeBach. 1975. Mating behav- iour and evidence of a male sex pheromone in species of the genus Aphytis. Annals of the Ento- mological Society of America 68(5): 893-896. Noirot, C. and A. Quennedey. 1974. Fine structure of insect epidermal glands. Annual Review of Ento- mology 19: 61-80. Noirot, C. and A. Quennedey. 1991. Glands, gland cells, glandular units: some comments on ternvi- nology and classification. Annates de la Societe en- tomologique de France 27: 123-128. Pedata, P. A., N. Isidoro, and G. Viggiani. 1995. Evi- dence of male sex glands of the antennae of En- carsia asterohemisiae Viggiani et Mazzone (Hy- menoptera; Aphelinidae). Bollettino del Laboratorio di Entonuilogia Agraria "Fdippo Silvestri" 50 (1993): 271-280. Quennedey, A. 1998. Insect epidermal gland cells: ul- trastructure and morphogenesis. Microscopic Anatomy of bwertebrates 11 A: 177-207. Rao, S. V. and P. DeBach. 1969. Experimental studies on hybridization and sexual isolation between some Apthytis species (Hymenoptera: Aphelini- dae). I. Experimental hybridization and an inter- pretation of evolutionary relationships among the species. HUgardia 39: 515-553. Romani, R., N. Isidoro, and F. Bin. 1997. Antennal structures and sex recognition in Trichopria dro- sophilae (Hymenoptera, Diapriidae). Boletin de la Asociacion espafwla de Entomologia, Suplemento al Volumen n'21: 142. Rosen, D. and P. DeBach. 1979. Species of Aphytis of the world (Hymenoptera: Aphelinidae). W. Junk, The Hague, pp. 1-801. Sacchetti, P., A. Belcari, F. Fagnani, N. Isidoro, and F. Bin. 1997. Antennal structures of CopHera occiden- talis mues. (Hymenoptera, Diapriidae) involved in mating behaviour and host recognition. Boletin de la Asociacion espafwla de Entonwlogia. Supleme}i- to al Volumen n° 21: 72. Viggiani, G. 1985. Additional notes and illustrations on some species of aphelinids described by A. A. Girault and A. P. Dodd in the genera Coccophagus Westw., Encarsia Foerst. and Prospaltella Ashm. (Hym.: Chalcidoidea). Bollettino del Laboratorio di Entomologia Agraria "Filippo Silvestri" 42: 233- 255. Viggiani, G. 1987. Le specie italiane del genere En- carsia Foerst. (Hymenoptera: Aphelinidae). Bol- lettino del Laboratorio di Entomologia Agraria "Filip- po Silvestri" 44: 121-179. Viggiani, G. and P. Mazzone. 1982. Antennal sensilla of some Encarsia Foerster (Hymenoptera: Aphel- inidae), with particular reference to sensorial complexes of the male. Bollettino del Laboratorio di Entomologia Agrarui "Filippo Silvestri" 39: 19-26. Viggiani, G., D. Battaglia, and R. Jesu. 1986. L'accoppiamento di Physcus testaceus Masi (Hym. Aphelinidae), con notizie preliminari sulla strut- tura dello scapo antennale maschile. Bollettino del Laboratorio di Entonwlogia Agraria "Filippo Silves- tri" 43: 3-6. Viggiani, G. and S. Laudonia. 1989. Su alcuni com- plessi sensoriali delle antenne maschili di tre spe- Volume 8, Number 1, 1999 115 cie del genere Eiicarsia Foerster (Hymenoptera: Viggiani, G. and H. Ren. 1993. New species and rec- Aphelinidae) e il loro rapporto con le fasi ords of Aphelinidae (Hymenoptera: Chalcidoi- dell'accoppiamento. Bcllettino iU-I Laboratorio di dea) from China. Bollettiiw del Lahoratorio di En- Entcmologm Agiaim "Filippo Silvestri" 45 (1988): tomologia Agraria "FUippo Silvestri" 48 (1991): 67-75. 219-239. J. HYM. RES. Vol. 8(1), 1999, pp. 116-119 A New Species Group and Two New Species of Enderomphale Girault (Hymenoptera: Eulophidae) from North America John LaSalle Unit for Parasitoid Systematics, CABI Bioscience UK Centre (Ascot), Department of Biology, Imperial College at Silwood Park, Ascot, Berks., SL5 7PY, UK Abstract. — Two new species of Euderomphale, E. sinuata and E. suzannae, are described from North America. These represent a distinct species group within the genus Euderomphale, which is here referred to as the sinuata group. £. sinuata is brachypterous, and this represents the first record of brachyptery in the Euderomphalini. Euderomphale is a cosmopolitan genus containing parasitoids of whiteflies. Thir- teen described species in this genus were listed by LaSalle and Schauff (1994). No comprehensive keys to species have been published, but there are keys to limited numbers of regional species (Erdos 1966, for Hungary; Hulden 1986, for Finland). Euderomphale belongs in the tribe Euder- omphalini. This tribe, containing parasit- oids of whiteflies, was treated at the world level by LaSalle and Schauff (1994), who presented keys to, diagnoses for, and sup- port for the monophyly of seven included genera in two genus groups. The present paper describes two new species which represent a new species group within the genus Euderomphale, the sinuata group. Eu- deromphale is thus separated into two spe- cies groups: the flavimedia group (based on the valid name for the type species) con- tains 13 species (LaSalle and Schauff 1994), and the sinuata group contains the two species described in this paper. Biology for the two new species in the sinuata group is unknown, but it is presumed they are parasitoids of whiteflies like all other members of the Euderomphalini. £. sinuata is brachypterous, and this repre- sents the first record of brachyptery in the Euderomphalini. It is also the first record of brachyptery in females of any Entedon- inae, although brachyptery is known in males of Microdonophagus ivoodlei/i Schauff (Schauff 1986). Kerrich (1973) reported the males of Pediobius arnconae Kerrich to be micropterous, although in this species the wing is only slightly reduced and does not have a highly modified shape. KEY TO SPECIES GROUPS WITHIN EUDEROMPHALE AND SPECIES OF THE SINUATA GROUP (FEMALES) 1. Axilla smoothly rounded anteriorly (Fig. 4). Vertex smoothly rounded posteriorly (Fig. 4), with the lateral ocellus placed on the top of the head. Prepecti.is partially fused to meso- pleuron anteriorly flavimedia group (discussed by LaSalle and Schauff 1994) - Axilla sinuate and concave anteriorly (Figs. 1, 3). Vertex with a distinct transverse carina; lateral ocellus behind this carina (Figs. 1-3). Prepectus free, not fused to mesopleuron sinuata group 2 2. Brachypterous; wings short, not extending past first gastral tergite (Fig. 1), infumated. Axilla with anterior margin deepl concave, anteromedial lobe narrower and more distinct (cf. Figs. 1 and 3). Gaster uniformly dark brown £. simiata LaSalle Volume 8, Number 1, 1999 117 Figs. 1-4. Eudcromphale species, head and mesosoma. 1-2. Eudewmphale sinuata. 3. Eiideromphale suzannae. 4. Eudcroniphiilc sp. (flaviiiicdin group). Not brachypterous; wings normal, extending to or slightly past apex of abdomen, hyaline or very lightly infumate. Axilla with anterior margin shallowly concave, anteromedial lobe wider and less distinct (cf. Figs. 1 and 3). Gaster yellow or pale brown, with several dark transverse stripes E. suzamiae LaSalle Characters for the Eiidero»ip}ialc genus group (within the Euderomphalini): Head usually smooth to lightly sculptured. Frontal suture, when present, placed just ventral to median ocellus. Vertexal suture present, either placed between the ocelli. or behind the ocelli. Malar sulcus usually present, although sometimes incomplete. Funicle with 1 or 2 segments. Dorsum of mesosoma more or less flattened and usu- ally smooth to lightly sculptured, gener- ally when distinct sculpture is present it is 118 Journal of Hymenoptera Research incised. Midlobe of mesoscutum usually with only 1 or 2 pairs of setae. Scutellum distinctly wider than long, with 1 or 2 pairs of setae. Axilla either completely separated from mesoscutum by a com- plete suture, or partially fused to meso- scutum. Submarginal vein with one or two setae on dorsal surface. See LaSalle and Schauff (1994) for further discussion and illustration of these characters, and characters to define the Euderomphalini. Characters for Euderomphale: Scrobal su- ture and frontal suture absent. Vertexal suture extending from the eye margin be- tween the median and lateral ocelli. Malar sulcus present, although incomplete and extending away from the mouth margin (postero-laterally from the eye). Antenna with two funicular segments, although the first is reduced to not much larger than the size of an anellus (the small first fu- nicular segment can generally be distin- guished from an anellus by the presence of setae). Mesosoma always black, non- metallic, more or less flattened in lateral view, sculpture on dorsal surface usually either finely incised or smooth. Midlobe of mesoscutum generally with 2 pairs of se- tae near the anterior margin. Scutellum distinctly wider than long, and with two pairs of setae: a larger one near the middle of the scutellum and a smaller one at the posterior margin. Axilla large (as wide as long), and completely separated from me- soscutum by a sulcus. Dorsal surface of submarginal vein with 2 setae (a single seta in the brachypterous sinuata). See LaSalle and Schauff (1994) for further dis- cussion and illustration of these charac- ters. Characters for the flainmedia species group in Eiidewniphale: Vertex smoothly rounded posteriorly. Axilla smoothly rounded anteriorly. Prepectus reduced and partially fused to the mesopleuron. Characters for the sinuata species group in Euderomphale: Vertex with a distinct transverse carina; lateral ocelli placed pos- terior to this carina. Axilla sinuate and concave anteriorly. Prepectus reduced (strongly so in sinuain) but separate, not fused to mesopleuron. Euderomphale sinuata LaSalle, sp.n. (Figs. 1-2) Female. — Length 0.75 mm. Head and mesosoma black, gaster dark brown. Scape yellow, pedicel and flagellum brown. All coxae black; fore and middle femora and tibiae dark brown to black; hind femur brown to light brown, hind tibia brown basally, light brown apically; tarsi brown, with apical segment dark brown. Wings infumated. Head with del- icate, small, regular incised sculpture. Ver- texal suture present near the eye margin; joining or running just anterior to vertexal carina. Mesosoma with delicate, small, regular incised sculpture. Axilla with an- terior margin distinctly concave, antero- medial lobe narrower and more distinct than in suzannae. Prepectus quite reduced, separate from the mesopleuron. Wings greatly reduced, not reaching apex of first gastral tergite. Forewing with a single strong seta on submarginal vein and an- other on marginal vein. Gaster with tergite 5 extending over the telescoped tergites 6 and 7 and short ovipositor. Holotype female, USA, Florida, Levy Co., 5 km SW Archer, 2-15.ix.l987, FIT, turkey oak shrubs, BRC HYM. TEAM (Ca- nadian National Collection, Ottawa). Euderomphale suzannae LaSalle, sp.n. (Fig. 3) Fe?n(7/e.— Length 0.90-0.95 mm. Head and mesosoma black, gaster yellow with transverse brown stripe on tergites 2-5, and a median longitudinal brown stripe on tergite 1. Scape yellow, pedicel and fla- gellum brown. Fore and middle coxae dark brown to black, hind coxa brown to light brown; fore femur dark brown; mid- dle femur light brown, dark brown dor- soapically, hind femur yellow; fore tibia yellow to light brown, dark brown dor- sally, middle tibia dark brown basally. Volume 8, Number 1, 1999 119 yellow to light brown apically, hind femur yellow to light brown; all tarsi yellow to light brown. Wings hyaline or very lightly infumated. Head with small, regular slightly raised sculpture (slightly stronger than in sinuata). Vertexal suture present near the eye margin; joining or running just anterior to vertexal carina. Mesosoma with small, regular incised sculpture (slightly stronger than in sinuata). Axilla with anterior margin moderately concave, anteromedial lobe broader and not as dis- tinct as in sinuata. Prepectus moderately reduced, separate from the mesopleuron. Wings normal. Forewing with two strong setae on the submarginal vein. Gaster with last tergite not extending over the ovipos- itor. Holotype female, USA, Missouri, Wayne Co., Williamsville, 10-26.vi.l987, MT, J. Becker (Canadian National Collec- tion, Ottawa). Paratype female, USA, Mis- souri, Wayne Co., Williamsville, vii.1987, MT, J. Becker (Canadian National Collec- tion, Ottawa). Et^mologi/. — Named for Suzanne Lewis. ACKNOWLEDGMENTS Space and facilities during this study were kindly provided by the Department of Entomology, The Natural History Museum, London; technical assis- tance from the SEM and photography units of the BMNH is also gratefully acknowledged. Special thanks to Suzanne Lewis (BMNH) and Georgina Godwin (CABI) for assistance with the taking of the photomicrographs, and Nick Hayes (BMNH) for the printing of the photomicrographs. LITERATURE CITED Erdos, J. 1966. Nonnulae Eulophidae novae Hungar- icae (Hymenoptera: Chalcidoidea). Annales His- torico-Ntituralci Musci Natiouiilii^ Hungarici, pars Zoologica 58: 395-420. Hulden, L. 1986. The whiteflies (Homoptera: Aley- rododea) and their parasites in Finland. Nctiilae Entoniologtccic 66: 1—10. Kerrich, G.J. 1973. A revision of the tropical and sub- tropical species of the eulophid genus Pediobius Walker (Hymenoptera: Chalcidoidea). Bulletin of the British Museum (Natural Histon/I. Entomology 29: 115-200. LaSalle, J. and Schauff, M.E. 1994. Systematics of the tribe Euderomphalini (Hymenoptera: Eulophi- dae): parasitoids of whiteflies (Homoptera: Al- eyrodidae). Systematie Entomology 19: 235-258. Schauff, M.E. 1986. Microdonophagus, a new entedon- ine genus (Hymenoptera: Eulophidae) from Pan- ama. Proceedings of the Entomological Society of Washington 88: 167-173. J. HYM. RES. Vol. 8(1), 1999, pp. 120-125 A New Genus and Two New Species of Brachypterous Lysiterminae (Braconidae) Sergey A. Belokobylskij and Donald L. J. Quicke (SAB) Zoological Institute, Russian Academy of Sciences, St. Petersburg 199034, Russia; (DLJQ) Unit of Parasitoid Systematics, CABI Bioscience UK Centre (Ascot), Department of Biology, Imperial College at Silwood Park, Ascot, Berks SL5 7PY, UK and Department of Entomology, The Natural History Museum, London SW7 5BD, UK Abstract. — A new genus Neolysitemtus gen. n. and two new species N. tumeri sp. n. and N. spinator sp. n. (tribe Lysitermini), both from South Africa, are described and illustrated. These are the first brachypterous species belonging to the subfamily Lysiterminae. Aptery and brachyptery are known in only a few subfamilies of Braconidae, in- cluding the Alysiinae, Aphidiinae, Blaci- nae, Doryctinae, Hormiinae, Masoninae, Pambolinae and Orgilinae. In the first of these, aptery is probably a result of their hosts dwelling within subterranean or filthy habitats and the associated difficulty in negotiating soil particles or costs of wing fouling, however, in the others, no obvious host habitat association is appar- ent, though many other brachypterous parasitoids are associated with stored products or with tree trunks /tall shrubs. The brachypterous and apterous braco- nids are more or less equally split between ecto- and endoparasitoids and between idiobiont and koinobiont taxa, suggesting that at least in this family when consid- ered at subfamily level, brachyptery is not dependent upon major life history fea- tures. However, formal comparative anal- ysis at species level will be required before firm conclusions can be reached. The Lysiterminae have previously often been regarded as a tribe within either the Rogadinae, Exothecinae or the Hormiinae (van Achterberg 1976, 1982; Quicke & van Achterberg 1990; Wharton 1993; Beloko- bylskij 1993) but was afforded subfamily status by van Achterberg (1993, 1995) and van Achterberg & Steiner (1996) because it shares no obvious synapomorphies with either Rogadinae s.s. or Hormiinae. This arrangement seems best at present since there is considerable doubt about the monophyly of the Hormiinae as treated (conservatively) by some workers (see for example, Whitfield & Wharton 1997). New molecular data, as well as the investiga- tion of more character systems, will be re- quired to resolve this (see Quicke et al. 1992; Belshaw et al. 1998). No apterous or brachypterous species have previously been described in the Lysiterminae. Although little is known bi- ologically about the Lysiterminae, they appear to have diverse host associations including being ectoparasitoids, or pre- sumed ectoparasitoids, of bagworms (Psy- chidae) and web-feeders (Xyloryctidae = Stenomidae), and also apparently endo- parasitoids of Orthoptera (Hedqvist 1963; Wharton 1993; van Achterberg & Steiner 1996). Interestingly, all of these hosts live in retreats involving silk. The Lysiterminae Tobias, 1968 includes only seven genera in the Old World (van Volume 8, Number 1, 1999 121 Achterberg 1995; Belokobylskij 1995; van Achterberg & Steiner 1996). Unfortunate- ly, differences between lysitermine genera mostly concern fore wing venation, though members of the subtribe Triter- mina, with 2 genera {Triiermus van Ach- terberg and Afroiritermus Belokobylskij) differ from the other genera of Lysitermi- nae in having fused, and therefore im- mobile, 1st and 2nd metasomal tergites (Belokobylskij 1993). Because of the major reliance of lysitermine systematics on wing venation, interpreting the relation- ships of Neolysitermus gen. n., with its re- duced wings, is not so straightforward. Apart from the reduced wings, Neolysiter- mus gen. n. has a distinct median emar- gination of the posterior margin of 3rd metasomal tergite — a character not known in any other species of Lysiterminae. TERMINOLOGY AND COLLECTIONS The wing venation terms used largely follow Tobias (1986). The following abbre- viation are used: POL — postocellar line; OOL — ocular-ocellar line; Od — maximum diameter of lateral ocellus. Collections are abbreviated as follows: The Natural His- tory Museum, London (BMNH); Zoologi- cal Institute, Russian Academy of Scienc- es, St. Petersburg, Russia (ZIP). Neolysitenmts Belokobylskij and Quicke, new genus Type species: Neolysitermus turucri new species Etymology. — From "neo" (new) and " Lysitermus" the genus of the subfamily Lysiterminae. Diagnosis. — The position of this new ge- nus is not clear. Differences between gen- era of the tribe Lysitermini are connected mostly with the venation of the fore wing, but species of Neolysitermus gen. n. have very short wings with reduced venation. This new genus differs from other genera of Lysitermini (in addition to the very short wings) by the presence of distinct median emargination on the posterior margin of 3rd tergite, a character that is absent in all described species of this tribe. Description. — Head weakly transverse (Figs 2, 11). Scapus (Fig. 4, 13) thick, with distinct round cut in outer apical side. Fla- gellum filiform. Apical flagellomere with- out apical spine. Maxillary palpi 6-seg- mented, labial palpi 4-segmented; 3rd la- bial segment distinctly shortened. Malar suture absent. Clypeus strongly convex (Figs 3, 12). Eyes glabrous. Occipital carina fused with hypostomal one higher man- dibles. Hypostomal keel distinct. Ocelli small, forming an almost equilateral tri- angle. Mesosoma reduced, but with all su- tures and depressions (Figs 7, 8, 16, 17). Propleura with longitudinal median cari- na on basal half. Pronotum with distinct lateral median corners (see Figs 8, 17). No- tauli deep and wide. Lateral lobes of me- soscutum with oval depressions postero- laterally. Scutellum with high lateral cari- nae. Sternauli rather deep, wide, oblique, coarsely crenulate. Prepectal carina very strong. Postpectal carina absent. Meta- pleural flange long and narrow. Propodeal areola incomplete or absent. Fore wing very short and narrow, stylet-shaped or oval. Hind tibia and tarsus slender. Hind tibial spurs very short. Hind basitarsus 0.9-0.95 X as long as 2nd-5th segments combined. First and second metasomal tergites not fused, mobile (Figs 9, 10, 18, 19). Dorsope of first tergite small. Second suture deep. Dorsal carina of first tergite semicircularly united basally. Third tergite without spines and carina posteriorly, with deep and rather narrow median emargination, with small single tooth ven- tro-laterally (Figs 9, 18). Distribution. — Afrotropical Region (South Africa). 122 Journal of Hymenoptera Research KEY TO SPECIES OF NEOIYSITERMUS Antenna 17-segmented; apical segment dark. Length of mesosoma 1.7-1.8 x height. Scu- tellum flat. Propodeum without lateral spines. Metasoma narrow, its length 2.3-2.4 x max- imum width. First metasomal tergite longer, apically 1.2-1.3 x wider than long N. tumeri sp. n. Antenna 14-segmented; apical segment whitish. Length of mesosoma 1.3-1.4 X height. Scutellum strongly pointedly convex. Propodeum with distinct pointed lateral spines. Me- tasoma wide, its length twice maximum width. First metasomal tergite shorter, apically 1.7-1.8 X wider than long N. spinator sp. n. Neolysitermus tumeri Belokobylskij and Quicke, new species (Figs 1-10) Holotype female. — "S. Africa. R.E. Turner. Brit. Mus. 1924-136", "Port St. John, Pon- doland. 6-25. Feb. 1924" (BMNH). Paratypes. — 1 female, same data as ho- lotype (BMNH); 1 female, "S. Africa. R.E. Turner. Brit. Mus. 1924-97", "Port St. John, Pondoland. Jan. 1924" (ZIP); 1 fe- male, "S. Africa. R.E. Turner. Brit. Mus. 1924-109", "Port St. John, Pondoland. 29.1 - 5.II.1924" (BMNH). Description. — Female. Body length 1.8- 2.3 mm. Head: 1.5-1.6 X wider than me- dially long; 1.7-1.8 X width of mesoscu- tum. Antennae 17-segmented. Scapus 1.5- 1.6 X as long as wide. First flagellar seg- ment 3.7-4.2 X longer than apically wide, slightly longer than 2nd segment. Penul- timate segment 3 X as long as wide, 0.7 X as long as 1st flagellar segment, 0.9-1 X as long as apical segment. Temple roundly narrowed behind eyes. Transverse diam- eter of eye 1.4-1.6 X length of temple (dor- sal view). POL 1.2-1.5 X Od, 0.3-0.4 X OOL. Antennal socket diameter 0.8-1 X distance between sockets, almost twice distance between socket and eye. Eye 1.2- 1.3 X taller than broad. Cheek height 0.7- 0.8 X height of eye, 1.6-1.7 X basal width of mandible. Face 1.2-1.4 X wider than eye height and 1.2 (wider than height of face and clypeus combined. Clypeus with distinct narrow flange along lower mar- gin. Hypoclypeal depression oval, 0.7-0.8 X wider than distance from depression to eye. Head distinctly and roundly nar- rowed below eyes. Mesosoma: 1.7-1.8 X longer than high, almost twice longer than wide. Median lobe of mesoscutum with antero-lateral teeth. Prescutellar depres- sion long, with median carina, granulose- crenulate, 0.5-0.6 X as long as scutellum. Scutellum rather flat. Subalar depression deep, narrow, crenulate. Mesopleura with- out median furrow. Propodeum without lateral spines. Fore wing stylet-shaped, 3.2^.5 X longer than wide, 0.35-0.4 X length of mesosoma. Hind femur 4.5-5 X longer than wide. Hind tarsus 0.8-0.85 X hind tibia. Second tarsal segment 0.3-0.35 X length of 1st segment, almost as long as 5th segment (excluding pretarsus). Meta- soma: Convex and narrow, its length 2.3- 2.4 X maximum width (on the level of middle of 2nd tergite), 1.4-1.5 x as long as mesosoma. First tergite distinctly and roundly narrowed towards base, rather long, its apical width 1.2-1.3 X length, 2.0-2.2 X its minimum width. Second ter- gite almost as long as basally wide, 0.8- 0.9 X maximum width, 1.3-1.4 X length of 1st tergite, 0.8-0.9 X length of 3rd ter- gite. Third tergite regularly and almost linearly narrowed toward apex. Oviposi- tor sheath 0.5-0.6 X as long as metasoma, 0.8-0.9 X as long as mesosoma. Sculpture and pubescence: Head densely granulate, vertex strongly granulate and usually with fine rugae; face finely granulate. Me- sothorax densely and entirely granulate. Propodeum with median carina, which is Volume 8, Number 1, 1999 123 Figs. 1-10. Neoli/sitenuKs tunwri gen. et sp. n.: 1 — head, frontal view; 2 — head, dorsal view; 3 — head, lateral view; 4 — basal and apical segments of antenna; 5 — hind femur; 6 — hind tibia; 7 — thorax, lateral view; 8 — thorax, dorsal view; 9 — metasoma, dorsal view; 10 — metasoma, lateral view. 1-1.4 X as long as furca; basolateral areas densely granulate; posterior half of pro- podeum transversely striate with dense granulation and 2 short submedian lon- gitudinal carinae. Hind legs finely granu- late. First to 3rd metasomal tergites strongly striate, with fine and dense trans- verse rugae between striae; striae at 2nd and 3rd tergites weakly convexly curved. Setae on dorsal side of hind tibia sparse, short and not erect. Colour: Body light reddish brown, sometimes dorsally dark- er. Head yellow. Antenna yellowish brown, submedially slightly darkened, 5- 6 subapical segments whitish, apical seg- ment dark. Palps pale yellow. Legs yel- low. Male. — Ur^known. .-• Neolysitermus spinator Belokobylskij and Quicke, new species (Figs 11-19) Holotype female. — "S. Africa. R.E. Turner. Brit. Mus. 1924-136", "Port St. John, Pon- doland. 6-25. Feb. 1924" (BMNH). Paratypes. — 1 female, same data as ho- lotype (BMNH). Description. — Female. Body length 1.8- 2.1 mm. Head: 1.4-1.5 X wider than me- dially long; 1.6-1.9 X width of mesoscu- tum. Temple roundly narrowed behind eyes. Transverse diameter of eye 1.7-2 X length of temple (dorsal view). POL 1-1.3 X Od, 0.3-0.4 X OOL. Antennal socket di- ameter 1.5-2 X distance between sockets, 124 Journal of Hymenoptera Research <23 ' ' Figs. 11-19. Neolysitermus spinator gen. et sp. n.: 11 — head, dorsal view; 12 — head, lateral view; 13 — basal and apical segments of antenna; 14 — hind tibia; 15 — hind femur; 16 — thorax, lateral view; 17 — thorax, dorsal view; 18 — metasoma, dorsal view; 19 — metasoma, lateral view. almost 1.5 X distance between socket and eye. Eye 1.3-1.4 x as taller than broad. Cheek height 0.8 X height of eye, nearly twice basal width of mandible. Face 1.3 x wider than eye height and 1.25 x wider than height of face and clypeus combined. Clypeus with distinct narrow flange along lower margin. Head distinctly and round- ly narrowed below eyes. Antennae 14-seg- mented. Scapus 1.5-1.6 x as long as wide. First flagellar segment 3.7-4.0 X longer than apically wide, 1-1.2 X as long as 2nd segment. Penultimate segment 2.5-2.7 X as long as wide, 0.8 X as long as 1st fla- gellar segment, as long as apical segment. Mesosoma: 1.3-1.4 X longer than high, 1.5-1.8 X longer than wide. Median lobe of mesoscutum with antero-lateral teeth. Prescutellar depression long, with median carina, granulose-crenulate, 0.6-0.7 X as long as scutellum. Scutellum strongly pointedly convex. Subalar depression deep, narrow, granulose-crenulate. Meso- pleura with fine oblique median furrow. Propodeum with distinct pointed lateral spines and mediobasal flat small lobe. Fore wing oval and short. Hind femur 4.5- 5 X longer than wide. Hind tarsus 0.8 x length of hind tibia. Second tarsal segment 0.3 X as long as 1st segment, nearly as long as 5th segment (excluding pretarsus). Metasoma: Convex and wide, nearly twice longer than maximally wide (at lev- el of middle of 2nd tergite), 1.7-1.9 X longer than mesosoma. First tergite dis- tinctly and roundly narrowed towards base, short, its apical width 1.7-1.8 x length, 2.4 X its minimum width. Second tergite 0.9-1 x longer than basally wide, 0.8 X its maximum width, 1.7 x length of 1st tergite, as long as 3rd tergite. Third ter- gite regularly and roundly narrowed to- Volume 8, Number 1, 1999 125 ward apex. Ovipositor sheath 0.3-0.4 X as long as metasoma, 0.6-0.8 x as long as mesosoma. Sculpture and pubescence: Head largely densely granulate, vertex sometimes (paratype) with rugae, frons striate with granulation, face finely gran- ulate. Mesothorax densely and almost en- tirely granulate. Propodeum with trans- verse median carina between spines; ba- solateral areas densely granulate; posteri- or half rugulose-striate, with fine granulation. Hind legs finely granulate. First to 3rd metasomal tergites strongly longitudinally striate, with fine and dense transverse rugae between striae; striae of 3rd tergite distinctly converging posteri- orly. Setae on dorsal side of hind tibia sparse, short and not erect. Colour: Body light reddish brown, metasoma partly darker. Head yellow. Antenna yellowish brown, submedially slightly darkened, 5 distal segments whitish, apical segment dark. Palps pale yellow. Legs yellow. Male. — Unknown. ACKNOWLEDGEMENTS This work was supported by the Natural Environ- ment Research Council, Initiative in Taxonomy. LITERATURE CITED Achterberg, C. van. 1976. A preliminary key to the subfamilies of the Braconidae (Hymenoptera). TijiUchrift voor Eittcimologic, LchIcii 119: 33-78. Achterberg, C. van. 1982. A new genus of the Roga- dinae-Lysitermini from Kazakhstan (Hymenop- tera, Braconidae). Entonuilogische Beriditcn, Am- sterdam 42: 125-128. Achterberg, C. van. 1993. Illustrated key to the sub- families of the Braconidae (Hymenoptera: Ich- neumonoidea). Zcv/o^'/st'/it' Vcrlininli'liiigcn. Lciiicn 283: 1-189. Achterberg, C. van. 1995. Generic revision of the sub- familv Betvlobraconinae (Hymenoptera: Bracon- idae) and other groups with modified fore tarsus. Zoologische Verhandtiingen, Leiden 298: 1-242. Achterberg, C. van and H. Steiner. 1996. A new genus of Tetratermini (Hymenoptera: Braconidae: Lys- iterminae) parasitic on grasshoppers (Gryllacri- didae). Zoologische Mededelingen, Leiden 70: 249- 259, Belokobylskij S. A. 1992. On the classification and phylogeny of the braconid wasp subfamilies Do- ryctinae and Exothecinae (Hymenoptera, Bracon- idae). Part 1. On the classification, 1. Entonwlogi- cheskoye Obozreniye 71: 900-928. (In Russian.) Belokobylskij S. A. 1993. On the classification and phylogeny of the braconid wasp subfamilies Do- ryctinae and Exothecinae (Hymenoptera, Bracon- idae). Part 1. On the classification, 2. Entomotogi- chcskoye Obozreniye 72: 143-164. (In Russian.) Belokobylskij S. A. 1995. Two new genera and two new subgenera of the subfamilies Exothecinae and doryctinae from the Old world (Hymenop- tera: Braconidae). Zoologische Mededelingen, Leiden 69: 37-52. Belshaw, R., E. Herniou, C. Gimeno, M. G. Fitton and D. L. ]. Quicke. 1998. Molecular phylogeny of the Ichneumonoidea (Hymenoptera) based on D2 ex- pansion region of 28S rDNA. Systematic Entomol- ogy 23: 109-123. Hedqvist, K.-J. 1963. Notes on Hormiinae with de- scription of new genera and species (Hym., Ich- neumonoidea, Braconidae). Entomokigisk Tidskrift 84: 30-61. Quicke, D. L. ]. and C. van Achterberg. 1990. The type specimens of Enderlein's Braconinae (Hymenop- tera: Braconidae) housed in Warsaw. Tijdschrift voor Entomologie 133: 251-264. Quicke, D. L. J., j. Tunstead, J. V. Falco, and P. M. Marsh. 1992. Venom gland apparatus in cyclo- stome braconid wasps with special reference to the subfamily Doryctinae (Insecta, Hymenoptera, Braconidae). Zoologica Scrip'ta 21: 403^16. Tobias, V. 1. 1968. Voprosv klassifikatsii 1 filogenii sem. Braconidae (Hymenoptera). Cliteniya Pamy- ati N. A. Kholodkovskoga (Moscow-Leningrad) [1967]: 3^3. (In Russian.) Tobias V.I. 1986. Order Hymenoptera. Family Bracon- idae. Introduction. In: Medvedev, G.S. (ed.). Opjredelitel' nasekoniykh evropeyskoy chasti SSSR. Pereponchatokn/lye [Keys to the insects of the Eu- ropean part of the USSR. Hymenoptera). Lenin- grad: Nauka. Vol 3, pt. 4. P. 7-15. (In Russian.) Wharton, R. A. 1993. Review of the Hormiini (Hy- menoptera: Braconidae) with a description of new taxa. Journal of Natural History 27: 107-171. Whitfield, J. B. and R. A. Wharton. 1997. Subfamily Hormiinae. In R. A. Wharton, P. M. Marsh & M. ]. Sharkey (eds) Identification manual to the Neiv World genera of Braconidae. Special Publication of the International Society of Hymenopterists 1: 285-301. J. HYM. RES. Vol. 8(1), 1999, pp. 126 NOTE New Host Record of a Ceraphronid (Hymenoptera) in Trichoptera Pupae J. C. LUHMAN, R. W. HOLZENTHAL, AND J. K. KjAERANDSEN (JCL) Minnesota Department of Agriculture, Biological Control Lab, 90 W. Plato Blvd., St. Paul, MN 55107-2094; (RWH) Department of Entomology, Uiiiversity of Minnesota, St. Paul, MN 55108; (JKK) Museum of Zoology, University of Bergen, Museplass, 3 Bergen, N-5007, Norway Pupae of the microcaddisfly Ochrotrichia moselyi Flint (Trichoptera: Hydroptilidae) were collected in Costa Rica that con- tained pupae of Aphanogmiis sp. (Hyme- noptera: Ceraphronidae). The caddisflies were collected in Puntarenas Province, Bellavista River, ca. 1.5 km NW of Las Al- turas, at 1400 m elevation (8.95rN, 82.846°W). Collections were made June 15-17, 1986. All material is housed in the University of Minnesota Insect Collection, St. Paul, Minnesota. Three collections in alcohol yielded one Apbnnogmus pupa each. There were a total of 12 Ochrotrichia moselyi cocoons of which 3 contained Aphanogmus pupae, 6 contained eaten cad- disfly pupae, and the remainder, devel- oped caddisflies. The cocoons with the parasitoid pupae contained only the wings and cast larval skin of the caddisfly. There was one Aphanogmus per parasit- ized cocoon. Dr Paul Dessart, a ceraphron- id specialist in Belgium, confirmed Luh- man's identifications of the Aphanogmus in the cocoons and stated this to be a new ordinal and family host record for Cera- phronidae (pers. comm., 1996). Heretofore recorded hosts of Ceraphronidae included Diptera, Homoptera, Hymenoptera, Neu- roptera, and Thysanoptera (Muesebeck 1979; Hanson and Cauld 1995). The Aphanogmus were discovered inside the cocoons of Ochrotichia moselyi, but out- side of the caddisfly pupa. Only the de- veloped wings remained uneaten. The Ochrotichia cocoons may have been para- sitized when exposed to the surface near the water substrate interface. Ochrotichia pupate within their larval cases in small clusters of individuals on the sides of rocks and boulders. During the dry sea- son, pupae are often exposed or closer to the surface. ACKNOWLEDGMENTS Trichoptera specimens were collected under NSF grants BSR 8512368 and DEB-9400632 to R. W. Hol- zenthal. LITERATURE CITED Hanson, V. E. and I. D. Gauld (eds.). 1995. The Hy- menoptera of Costa Rica. Oxford University Press, NY. Muesebeck, C. F. W. 1979. Superfamily Ceraphrono- idea, pp.1187-1195. In K. V. Krombein, P. D. Hurd, Jr., D. R. Smith, and B. D. Burks eds., Cat- alog of ll\iiucnof'tcra of America Nortli of Mexico, \'ol. 2. Si/mpln/tn and Apocnta, Washington, D.C.: Smithsonian Institution Press JOURNAL OF HYMENOPTERA RESEARCH (ISSN #1070-9428) An Entomological Journal Dedicated to the Study of Bees, Wasps, and Ants Special Offer to New Members and Opportunity to Purchase Back Issues of JHR for Current Members Obtain a Complete Set (Vol. 1-7) for US$110 when you become a New Member As a special incentive to recruit new members, the International Society of Hymenopterists is offering a complete set of back issues (1992-1998) for the sum of US$1 10.00 if purchased at the same time as a new membership ($35; $40 after 1 March 1999). This is less than one-half the current member's price for a complete set. Additionally, for those members who joined the Society after its inception and want to complete their sets, we offer a reduced, graduated charge for back issues, which are priced by the volumes needed in any combination (e.g., any four volumes sell for US$80, see prices below). Be certain to tell us which volumes you are ordering. Prices include surface shipping anywhere in the world. The Journal of Hymenoptera Research is the official publication of the International Society of Hymenopterists, which was organized in 1982 and incorporated in 1991. The journal began publishing in 1992 and issued one number per year until 1996 when it switched to two numbers (average 3 10 pages per year). This is a peer-reviewed journal dedicated to publishing scholarly research of the highest quality. With an editor and four subject editors currently located in the United States, England, and Scotland, the journal promotes the world-wide communication of scientific data in all branches of hymenopteran research, including biology, behavior, ecology, systematics, ta.xonomy, genetics, and morphology. The journal is currently produced by Allen Press and is printed on paper that meets the requirements of ANSI/NISOZ39.48-1992 (Permanence of Paper). The Society also has produced an illustrated A'/(//;j(a/ o/ the New World Genera of the Family Braconidae (439 pages, spiral bound). ORDER INFORMATION Membership 1999 (after 1 March 1999 US$40.00) US$35.00 Volumes 1-7(1992-1998) US$110.00 Back issues for any volume (fill in the volume number(s) needed): 6 vols. US$ 1 00 5 vols. US$90 4 vols. US$80 3 vols. US$60 2 vols. US$40 1 vol. US$20 Special Publication Number 1 Manual of the New World Genera of the Family Braconidae US$20.00 TOTAL You may photocopy this order blank, or submit the required information, and give us your shipping address with checks or purchase orders made payable to "The International Society of Hymenopterists" c/o Dr. John Huber, Eastern Cereal & Oilseed Research Centre, Agriculture Canada, K. W. Neatby Building, Ottawa, Ontario, Canada KIA 0C6 (Email: huberjh@em.agr.ca). For credit card payments please contact Dr. Jim Woolley, Department of Entomology, Texas A & M University, College Station, TX 77843 (Email: jimwoolley@tamu.edu). INSTRUCTIONS FOR AUTHORS General Policy. The ]iiurnal of Hytncnoptera Research invites papers of high scientific quality reporting comprehensive research on all aspects of Hymenoptera, including biology, behavior, ecology, systematics, taxonomy, genetics, and morphology. Taxonomic papers describing single species are unlikely to be accepted unless a strong case is evident, such as importance in economic entomology or with concurrent biology or ecology. Manuscript length generally should not exceed 50 typed pages; however, no upper limit on length has been set for paf)ers of exceptional quality and importance, including taxonomic monographs at generic or higher level. All papers will be reviewed by at least two referees. The referees will be chosen by the appropriate subject editor. However, it would be helpful if authors would submit the names of two persons who are competent to review the manuscript. The language of publication is English. Summaries in other languages are acceptable. The deadline for receipt of manuscripts is 1 October (for the April issue) and 1 April (for the October issue). Format and Preparation. Three copies of each manuscript, including copies of illustrations, should be submitted on letter size or A4 paper, double spaced, with at least 25 mm margins on all sides. On the upper left of the title page give name, address and telephone and fax numbers of the author to whom all corre- spondence is to be sent. The paper should have a concise and informative title, followed by the names and addresses of all authors. The sequence of material should be: title, authorfs), abstract, text, acknowledgments, literature cited, appendix, figure legends, figure copies (each numbered and identified), tables (each numbered and with heading). Each of the following should start a new page: (1) title page, (2) abstract, (3) text, (4) literature dted, (5) figure legends, (6) footnotes. Following acceptance of the manuscript, the author should provide the editor witn one copy of the manu- script accompanied by a copy on diskette using DD, double sided computer diskettes — IBM compatible MS DOS 5.25 inch or IBM and Macintosh 3.5 inch diskettes. (Authors who do not have access to a computer should submit three copies of the manuscript.) The paper may be submitted in most PC and Mac word processor programs such as Microsoft Word, FullWrite Professional, WordPerfect, WriteNow, Nisus, Mac- Write, or MacWrite II. If possible, all words that must be italicized should be done so, not underscored. Tables may be formatted in a spread sheet program such as MS Works or MS Excel. Text should be double spaced typing, with 25 mm left and right margins. Tables should be put in a separate file. Diskettes should be accompanied by the name of the software program used (e.g., WordPerfect, Microsoft Word). Authors should keep backup copies of all material sent to the Editor. The Society cannot be responsible for diskettes or text mislaid or destroyed in transit or during editing. Illustrations should be planned for reduction to the dimension of the printed page (14 x 20.5 cm, column width 6.7 cm) and allow room for legends at the top and bottom. Do not make plates larger than 14 x 18 in. (35.5 X 46 cm). Individual figures should be mounted on a suitable drawing board or similar heavy stock. Photographs should be trimmed, grouped together and abutted when mounted. Figure numbers should be on the plate, and it is strongly recommended that names be included after the numbers (e.g.. Fig. 2, texanus). Include title, author(s) and address(es), and illustration numbers on back of each plate. Original figures need not be sent until requested by the editor, usually after the manuscript has been accepted. Reference to figures/ tables in the text should be in the style "(Fig. 1)" "(Table 1)". Measurements should be in the metric system. All papers must conform to the International Code of Zoological Nomenclature. The first mention of a plant or animal should include the full scientific name including the authority. Genus names should not be abbre- viated at the beginning of a sentence. In taxonomic papers type specimens must be clearly designated, type depositories must be clearly indicated, and new taxa must'be clearly differentiated from existing taxa by means of keys or differential diagnoses. Authors are required to deposit all type material in internationally recognized institutions (not private collections). Voucher specimens should be designated for specimens used in behavioral or autecological studies, and they should be deposited similarly. Acceptance of taxonomic papers will not require use of cladistic methods; however, authors using them will be expected to specify the phylogenetic program used (if any), including discussion of program options used. A data matrix should be provided if the subject is complex. Cladograms must be hung with characters and these should include descriptors (not numbers alone) when feasible. The number of parsimonious clado- grams generated should be stated and reasons given for the one adopted. Lengths and consistency indices should be provided. Adequate discussions should be given for characters, plesiomorphic conditions, and distributions of characters among outgroups when problematical. References in the text should be (Smith 1999), without a comma, or Smith (1999). Two articles by a single author should be (Smith 1999a, 1999b) or Smith (1999a, 1999b). For multiple authors, use the word "and," not the symbol "&" (Smith and Jones 1999). For papers in press, use "in press," not the expected publication date. The Literature Cited section should include all papers referred to in the paper. Journal names should be spelled out completely and in italics. Charges. Publication charges are $10.00 per printed page. At least one author of the paper must be a member of the International Society of Hymenopterists. Reprints are charged to the author and must be ordered when returning the proofs; there are no free reprints. Author's corrections and changes in proof are also charged to the author. Color plates will be billed at full cost to the author. All manuscripts and correspondence should be sent to: Dr. E. Eric Grissell Systematic Entomology Laboratory, USDA ''f National Museum of Natural History Washington, DC 20560-0168 Phone: (202) 382-1781 Fax: (202) 786-9422 E-mail: egrissel@sel.barc.usda.gov CONTENTS (Continued from front cover) t: WARD, P. S. Deceptive similarity in army ants of the genus Neivamyrmex (Hymenoptera: Formicidae): taxonomy, distribution and biology of N. californicus (Mayr) and N. nigrescens (Cresson) 74 WHARTON, R. A. A review of the Old World genus Fopius Wharton (Hymenoptera: Bra- conidae: Opiinae), with description of two new species reared from fruit-infesting Tephritidae (Diptera) 48 NOTE LUHMAN, J. C, R. W. HOLZENTHAL, and J, K. KJAERANDSEN. New host record of a ceraphronid (Hymenoptera) in Trichoptera pupae 126 ANNOUNCEMENT Special opportuniHes for new and current members to purchase back issues of JHR . . 127 ^^OOET,^^ Journal of Hymenoptera Research Volume 8, Number 2 October 1999 ISSN #1070-9428 CONTENTS BARBALHO, S. M., A. M. PENTEADO-DIAS, and P. M. MARSH. Descriptions of new genera from Brazil in the tribes HeterospilLni and Spathiini with similar wing venation (Hymenoptera: Braconidae, Doryctinae) 139 BROOKS, R. W. and C. D. MICHENER. The Chilicola megalostigma species group and notes on two lost types of Chilicola (Hymenoptera: CoUetidae, Xeromelissinae) 132 COOPERBAND, M. R, R. A. WHARTON, G. W. FRANKIE, and S. B. VINSON. New host and distribution records for Leucospis (Hymenoptera: Leucospidae) associated primarily with nests of Centris (Hymenoptera: Anthophoridae) in the dry forests of Costa Rica 154 ENGEL, M. S. The taxonomy of recent and fossil honey bees (Hymenoptera: Apidae; Apis) 165 FORTIER, J. C. and S. R. SHAW. Cladistics of the Aleiodes lineage of the subfamily Rogadinae (Hymenoptera: Braconidae) 204 LEBLANC, L. The NearcHc species of Protarchus Foerster (Hymenoptera: Ichneumonidae: Ctenopelmatinae) 251 LOCH, A. D. and G. H. WALTER. Does the mating system of Trissolcus basalts (Wollaston) (Hymenoptera: Scelionidae) allow outbreeding? 238 NEWMAN, T M. and D. L. J. QUICKE. Ultrashxicture of spermatozoa in Leptopilina (Hymenoptera: Cynipoidea: Eucoilidae) 197 SHAW, M. R. and R. R. ASKEW. Reproductive variability in Pediobius alcaeus (Walker) (Hymenoptera: Eulophidae), a parasitoid of Phyllonorycter (Lepidoptera: Gracillariidae) 127 BOOK REVIEW MUELLER, U. B. — Detlef Mader. Geologische und biologische Entomookologie der rezenten Seidenbiene CoUetes. Volume 1 268 INTERNATIONAL SOCIETY OF HYMENOPTERISTS Organized 1982; Incorporated 1991 OFFICERS FOR 1999 Andrew D. Austin, President John LaSalle, President-Elecf James B. Woolley, Secretary John T. Huber, Treasurer E. Eric Grissell, Editor Subject Editors Symphyta and Parasitica Aculeata Biologi/: Mark Shaw Biologx/: Sydney Cameron Systematics: Donald Quicke Systematics: Wojciech Pulawski All correspondence concerning Society business should be mailed to the appropriate officer at the following addresses: President, Department of Crop Protection, University of Adelaide, Glen Os- mond, Australia 5064; Secretary, Department of Entomology, Texas A&M University, College Station, Texas 77843; Treasurer, Eastern Cereal & Oilseed Research Centre, Agriculture Canada, K. W. Neatby Building, Ottawa, Ontario, Canada Kl A 0C6; Editor, Systematic Entomology Labo- ratory, USDA, % National Museum of Natural History, Washington, D.C. 20560-0168. Membership. Members shall be persons who have demonstrated interest in the science of ento- mology. Annual dues for members are US$40.00 per year (US$35.00 if paid before 1 February), payable to The International Society of Hymenopterists. Requests for membership should be sent to the Treasurer (address above). Information on membership and other details of the Society may be found on the World Wide Web at http://IRIS.biosci.ohio-state.edu/ish. Journal. The journal of Hymenoptera Research is published twice a year by the International Society of Hymenopterists, % Department of Entomology, Smithsonian Institution, Washington, D.C. 20560-0168, U.S.A. Members in good standing receive the journal. Nonmember subscriptions are $60.00 (U.S. currency) per year. The Society does not exchange its publications for those of other societies. Please see inside back cover of this issue for information regarding preparation of manuscripts. Statement of Ownership Title of Publication: Journal of Hymenoptera Research. Frequency of Issue: Twice a year. Location of Office of Publication, Business Office of Publisher and Owner: International Society of Hymenopterists, % Department of Entomology, NHB 168, Smithsonian Institution, Wash- ington, D.C. 20560, U.S.A. Editor: E. Eric Grissell, Systematic Entomology Laboratory, USDA, % National Museum of Nat- ural History, Washington, DC 20560-0168. Managing Editor and Known Bondholders or other Security Holders: none. This issue was mailed 27 October 1999 J. HYM. RES. Vol. 8(2), 1999, pp. 127-131 Reproductive Variability in Pediobius alcaeiis (Walker) (Hymenoptera: Eulophidae), a Parasitoid of Phyllonorycter (Lepidoptera: Gracillariidae) M. R. Shaw and R. R. Askew (MRS) National Museums of Scotland, Chambers Street, Edinburgh EHl IJF, UK; (RRA) 5 Beeston Hall Mews, Beeston, Tarporley, Cheshire CW6 9TZ, UK Abstract. — Separate, but sympatric, sexual and asexual populations of Pediobius akaeus (Walker) (Hymenoptera: Eulophidae), a parasitoid of leaf-mining Phyllonorycter species (Lepidoptera: Gra- cillariidae), are shown to exist and to be associated with hosts on different tree species in Britain. Pediobius akaeus (Walker) is a solitary koinobiont endoparasitoid of many spe- cies of Phi/lloncrycter mining the leaves of deciduous trees and shrubs (Askew & Shaw 1974, 1979). Eggs are laid inside third instar or older host larvae which continue to develop as far as the pupal stage before being killed, and the adult P. alcaeus emerge from the host pupae (Askew & Shaw 1979). In Britain, Phyllon- orycter feeding on Almis gliitinosn, Corylus avellnna and Quercus (both petraea and rob- ur) seem to be particularly liable to attack from P. alcaeus, but the chalcidoid is also frequently associated with PIn/llonon/cter hosts on Fagus sylvatica, Malus spp. and Carpinus betulus (Askew 1994). Our long-standing interest in chalcidoid parasitoids of leaf mining and other in- sects is, in part, manifested by a records database maintained by one of us (RRA), into which are entered all rearing records from our own field-work, from host col- lections made by others who submit par- asitoids to us for identification, and from reared specimens we have otherwise seen (i.e. all records entering the database are of parasitoids identified by us). As they currently stand, our total records continue to indicate that Pediobius alcaeus is exclu- sively associated with Phyllonorycter spe- cies mining the leaves of deciduous trees and shrubs. Our accumulated rearing re- cords of P. alcaeus for which sex was scored are given in Table 1: an earlier, con- densed compilation giving fewer data was published by Askew (1994), but the rear- ing records of P. alcaeus previously ana- lysed by Askew & Shaw (1974) are not all included in Table 1 as not all of those sam- ples were sexed. For most host species the rearing records of P. alcaeus presented in Table 1 come from numerous collections made over a wide geographical spread of localities in Britain and sometimes also continental Europe. As was first noted by Askew (1975), samples of P. alcaeus originating from hosts on different tree genera have one of two distinct categories of sex ratio: either males are well-represented (35-48 per cent), or the sample is virtually entirely unisexual with males comprising at most only about three per cent (Table 1, see also Askew 1994). Samples from Aluus, Carpi- nus and Corylus are sufficiently large to permit their categorization as host tree genera which support unisexual (female) populations of P. alcaeus, whilst bisexual populations are associated with Betula, Fa- gus, Quercus, Malus, Sorbus and probably Acer. Different Phyllonorycter species feed- ing on the same genus of host plants are attacked by P. alcaeus having the same 128 Journal of Hymenoitera Research Table 1. Accumulated total rearing records of Pediobitis akaeus (Walker) of known sex (specimens all seen by the authors). PhyllomrycU-r Country /' alcaeii^ Host plant .• V i-i Acer campestre sylvelia (Haworth) Britain 1 0 A. platanoides plataiwidella (Joannis) Britain 0 1 A. pfeiidof'latnnus genkuklla (Ragonot) Britain 1 0 Alnus glutinosa froelichklla {ZeWer) / kkemanella (Fabricius) Britain 1 62 Belgium 0 1 rajella (Linnaeus) Britain 0 21 Netherlands 0 1 Germany 1 780 stettincnsis (Nicelli) Britain 0 4 Germany 0 43 A. inama strigulatellii (Zeller) Britain 0 20 Germany 3 135 A. cordata rajella Britain 0 4 A. I'iridis alpina (Frey) Switzerland 0 1 Betula spp. ulmifotklla (Hiibner) Britain 13 26 Netherlands 3 4 Carpinus hetulus qidnnata (Geoffroy) Britain 6 219 Netherlands 0 1 tenerella (Joannis) Britain 1 4 Belgium 0 1 Corylus avellana coryli (Nicelli) Britain 0 58 Belgium 0 9 Netherlands 0 2 nketlii (Stainton) Britain 0 47 Crataegus monogyna oxyacanthae (Frey) Britain 0 1 Fagiis si/lvatka maestingella (Miiller) Britain 83 107 Netherlands 2 0 Mains domestka corylifoliella (Hiibner) Britain 0 1 M. sylvestris hlancardelln (Fabricius) Britain 0 3 cydoinella (Denis & Schiffermiiller) Britain 7 5 Quercus robur / petraea quercifolklla (ZeWer) / harrisella (Linnaeus) Britain 110 155 Netherlands 0 1 lautella (Zeller) Britain 10 21 France 2 2 Netherlands 2 0 saportella (Duponchel) Britain 0 1 Salix caprea salkkolella (Sircom ) Britain 0 1 species indet. Britain 0 2 S. aurita viminklla (Sircom) Britain 0 1 Sorhus aucKparia sorhi (Frey) Britain 3 2 Netherlands 1 0 type of sex ratio; further, our limited data indicate that the sex ratio type /host plant relationship probably remains constant across Europe. Careful comparison failed to reveal any consistent morphological differences between P. alcaeus from the unisexual and bisexual populations. In this paper we report the results of sleeving experiments in which virgin fe- male P. alcaeus from either the unisexual or bisexual populations were presented with developing leaf-mines on the para- sitoids' own or on an alien host plant. The aims of this investigation were first, to es- tablish the type of parthenogenetic repro- duction employed by each of the two pop- Volume 8, Number 2, 1999 129 Table 2. Sleeves on Quercus robur and Cori/hi$ avellana into which Pln/llonon/cter species, and later virgin Pediobiu^ alcaeus females from varying sources, were introduced. The number of moths and the number and sex of P. alcaeuf reared in each sleeve are recorded. Sleeve No. Host tree rili/llLilwry^lfr species I'clmbJUi \ no, & source \o ot molhs reared reared 1 2 3 Q. robur Q. robur Q. robur quercifoliella quercifoliella quercifoliella 1, ex P. quercifoliella 1, ex P. coryli 1, ex Phyllonorycter sp. underside 36 40 1 46i 34 9 9 6 99 4 5 6 C. avellana C. avellana C. avellana coryli coryli nicellii miner on Abuts glutinosa 2, ex P. coryli 2, ex P. coryli 2, ex Phyllonorycter sp. underside miner on Alnus glulinosa 4 14 14 29 9 9 40 9 9 23 9 9 7 C. avellana nicellii 2, ex P. con//; 36 14 9 9 ulations, and second, to determine wheth- er the host plant range of the unisexual population could be expanded under ar- tificial rearing conditions. METHODS Seven muslin sleeves were fastened to branches of oak (Quercus robur) (sleeves 1- 3) and hazel {Corylus avellana) (sleeves 4— 7) in the period 2i-25.v.l979. At the same time, or shortly afterwards, reared adult Phi/llonorycter were introduced. The three Quercus sleeves each received two male and two female P. quercifoliella (an under- surface miner), whilst six male and three female P. coryli (an uppersurface miner) were put into sleeves 4 and 5, and four male and two female P. nicellii (an under- surface miner) were introduced to sleeves 6 and 7. Variation in the numbers of moths introduced was partly due to short- age of livestock, but was also experimen- tal against the risk of overstocking sleeves (which might cause leaf abscission as the mines developed). Virgin female Pediobius alcaeus were put in the sleeves during the period 22- 25. vi. 1979, at which time developing third and fourth instar mines could be seen in most sleeves. The Pediobius had been in- dividually reared in tubes and came from three sources: P. quercifoliella mines on Quercus, P. coryli mines on Corylus and from undersurface leaf-mines, probably of P. froelichiella, on Alnus glutinosa. One or two virgin Pediobius were introduced into each sleeve; their host origins are given in Table 2. Variation in stocking levels was entirely due to shortage of livestock. All host and parasitoid material came from sites in the vicinity of Reading, Berkshire where the experiments were performed. The sleeves were brought indoors on 10.vii.l979 (sleeves 1-3), 13.vii.l979 (sleeves 4,5) or l.viii.l979 (sleeves 6,7). All mature and most immature mines were removed and stored separately in plastic boxes for rearing adult insects. When emergence had finished, the numbers and sex of Pediobius and numbers of moths produced in each sleeve were recorded. RESULTS Despite sometimes heavy leaf abscission due to population explosions of enclosed aphids, mines matured in all sleeves, mostly in good numbers but rather sparse- ly in sleeve 3. The first P. quercifoliella, a leaf undersurface miner, and P. coryli, an uppersurface miner, emerged 12- 14.vii.l979, whilst P. nicellii, another un- dersurface miner, emerged somewhat later about 2.viii.l979. Pediobius also com- menced emergence about this latter date. Total emergences of Pediobius and moths are shown in Table 2. 130 Journal of Hymenoptera Research DISCUSSION Sleeve 1, the only sleeve holding P. al- caeus from a bisexual population, yielded only male Pediobius. This unequivocal re- sult demonstrates that the bisexual popu- lation of P. nicaetis on Qiiercus produces males parthenogenetically by arrhenoto- ky, which is usual in haplodiploid Hy- menoptera. In contrast the other sleeves, in which Pediobius from unisexual popu- lations were introduced to their own or alien hosts, all produced only female P. al- caeus. This confirms that the populations which appear to be unisexual from rearing records (Table 1) are indeed thelytokous, and not the result of inseminated females exercising control over fertilisation of their eggs. The successful parasitization of P. quer- cifoliella on Quercus, naturally a host of bi- sexual populations of the parasitoid, by Pediobius originating from unisexual pop- ulations on Con/lus and Ahius (sleeves 2 and 3), shows that hosts on the parasit- oid's 'wrong' host food-plant can support it under artificial rearing conditions. Also, less surprisingly, Pediobius from Alnus un- dersurface mines was reared through un- dersurface mines on Con/lus (sleeve 6), and Pediobius from uppersurface leaf- mines on Corylus was reared through un- dersurface mines on the same tree (sleeve 7). Populations of P. alcaeus therefore differ in their reproductive biology. The bisexual populations reproduce by haplodiploidy and only males result from unfertilised eggs (arrhenotokous parthenogenesis). Re- production in the unisexual populations is thelytokous with females developing from unfertilised eggs. It is not clear whether the occasional males appearing in rearings of normally unisexual populations, as on Alnus and Con/lus (Table 1), are the prog- eny of 'stray' bisexual females, in which case they would presumably be potential- ly reproductively functional in relation to the bisexual race, or whether they are progeny of normally thelytokous females, in which case they may be entirely repro- ductively non-functional. The situation in Pediobius alcaeus is sim- ilar in some respects to that pertaining in Diplolepis spinosissimae (Giraud) (Hym., Cynipidae) (Plantard et al. 1998). This rose gallwasp is thelytokous (up to 4 per cent males) in populations on the Atlantic coast of France, but in two inland populations males comprise 21 and 29 per cent of the populations. The thelytokous populations were found to be infected by the endo- symbiotic bacterium Wolhachia which pre- vents the formation of males. Bisexual populations of D. spiiwsissitnae were free of Wolbacbia. Whether or not a micro-or- ganism is implicated in the thelytoky of Pediobius alcaeus remains to be determined, but even if this were to be the case, the apparent segregation of sexual and thely- tokous populations onto different tree genera, which in Pediobius alcaeus occurs sympatrically and (certainly in Britain) consistently, would still require explana- tion. ACKNOWLEDGMENTS The experimental part of this work was carried out while MRS was supported by a Reading University Research Fellowship. Eric Gissell made helpful com- ment on an earlier version of the manuscript. LITERATURE CITED Askew, R. R. 1975. The organisation of chalcid-dom- inated parasitoid communities centred upon en- dophytic hosts, pp. 130-153. In: Price, P.W. (ed.), Evolnlwmm/ strntcgief^ ofpnrnsitic iiifcctf ami iiiiks. Plenum Press. Askew, R. R. 1994. Parasitoids of leaf-mining Lepi- doptera: what determines their host ranges? pp. 177-202. In: Hawkins, B.A. and W. Sheehan (eds), Pimmtoid cotnmunily ecology. Oxford Uni- versity Press. Askew, R. R. and M. R. Shaw. 1974. An account of the Chalcidoidea (Hymenoptera) parasitising leaf-mining insects of deciduous trees in Britain. Biologicnl journal of the Linncan Socich/ 6: 289-335. Askew, R. R. and M. R. Shaw. 1979. Mortality factors affecting the leaf-mining stages of Plii/llonon/ctfr (Lepidoptera: Gracillariidae) on oak and hirch. 2. Volume 8, Number 2, 1999 131 Biology of the parasite species. Zoological loiininl oky in the rose gallwasp Diphtepis s;ii»i)siss/»ii7i' ()/ the Unnean Society 67: 51-64. (Giraud) (Hymenoptera: Cynipidae), and its con- Plantard, O., J.-Y. Rasplus, G. Mondor, I. Le Clainche, sequences on the genetic structure of its host. and M. Solignac. 1998. Wi);;irtc/;((7-induced thelyt- Proceedings of the Royal Society B 265: 1075-1080. J. HYM. RES. Vol. 8(2), 1999, pp. 132-138 The Chilicola megalostigma Species Group and Notes on Two Lost Types of Chilicola (Hymenoptera: Colletidae, Xeromelissinae) Robert W. Brooks and Charles D. Michener Entomology Division, University of Kansas Natural History Museum, Lawrence, Kansas 66045, U.S.A. Abstract, — The three species of the megalostigma group of Chilicola (Hylaeosoma) are reviewed and a key for their separation is provided. Chilicola (H.) stenocephala Brooks and Michener, new species, is described from Amazonian Colombia. Chilicola (H.) polita Michener is recorded for the first time from Costa Rica. Lectotypes are designated for two Brazilian Chilicola species orig- inally described as Oediscelis huberi Ducke and O. minima Ducke. The xeromelissine genus Chilicola Spi- nola has been characterized by Toro and Moldenke (1979) and Michener (1994, 1995). It consists of small, slender, usually black bees superficially similar to species of Hylaeiis in the subfamily Hylaeinae, but differing notably in the presence of a weak scopa on the posterior femora and tibiae, and on the first three metasomal sterna, the hairs on the second sternum being the best developed part of the scopa. Chilicola is particularly abundant and diverse in temperate South America (Chile, Argenti- na) but ranges north to central Mexico and to St. Vincent in the Lesser Antilles. The subgenus Hylaeosoma, characterized by Michener (1994, 1995), occurs from Bra- zil to central Mexico. It consists of unusu- ally slender species with a depression for the reception of the antennal scape above each antennal alveolus. THE CHILICOLA MEGALOSTIGMA SPECIES GROUP Within the subgenus Hylaeosoma there is a species-group consisting of very smooth, shiny species with a strong, flaring pre- occipital carina. This group, here known as the megalostigma group, which was also characterized as Group B by Michener (1994:83), consists of C. megalostigma Ducke from northeastern Brazil, C. polita Michener from Mexico to Panama, and the new species described below from Colom- bia. A hitherto unrecognized characteristic of the megalostigma group is the modified front tarsus of females, perhaps serving to pull pollen from minute, deep flowers or tubular anthers. The front tarsus of an or- dinary species of Hylaeosoma, Chilicola mexicana Toro and Michener, is illustrated in Figure 1. Figures 2 to 4 show the front tarsi of the three species of the megalostig- ma group. All (including C. mexicana) are bristly, with strong curved bristles as well as straight ones, as shown in Figure 4. This vestiture is omitted in Figures 1 to 3. In females of the megalostigma group, the front basitarsus is shortened if one ignores the apical process, whereas the second segment is relatively large, compared to that of other species groups. More details are indicated in the figures and in the key to species, below. Abbreviations used are the following: S, sternum; T, tergum. KSEM, Entomology Division, Snow Collections, University of Kansas Natural History Museum, Lawrence, Kansas, U.S.A. MPEG, Museu Paraense Emflio Goeldi, Belem, Para, Bra- zil. Volume 8, Number 2, 1999 133 KEY TO THE SPECIES OF THE MECALOSTICMA GROUP OF CHILICOLA Minimum distance between eyes about three-fourths of width of eye on same horizontal line seen from front (Fig. 8); malar area distinct, over half as long as wide (Fig. 9). First and second front tarsal segments of female each ending in long, down-curved, horn-like process with blunt apex (Fig. 4) stenocephala Brooks & Michener Minimum distance between eyes about equal to width of eye on same horizontal line seen from front; malar area short, less than half as long as wide. First and second front tarsal segments of female each ending in long process terminated by curved bristle 2 Frons with pit on frontal line above level of antennae, forming equilateral triangle with the two antennal sockets; first and second front tarsal segments of female each ending in long process terminated by curved bristle coarser than other tarsal bristles (Fig. 3) megalostigma (Ducke) Frons without pit on frontal line; first and second tarsal segments of female each with apical process ending with curved bristle similar to bristles elsewhere on tarsus polita Michener ^==^:=^^^^^S^ Figs. 1^. Outer views of anterior tarsi of females of Chilicola (Hi/taeoscma). All are bristly with both curved and relatively straight bristles and hairs, as shown in Figure 4. 1, C. mexicaiui Tore and Michener; 2, C. politn Michener; 3, C. megahsligmcj (Ducke); 4, C. itenocephii- la n. sp. In Figure 3 the bases of only two large setae are indicated arising from the apices of the processes of the first two tarsal segments of C. megalostigma. Comparable but seemingly more slender setae exist in C. polita (Fig. 2). Chilicola (Hylaeosoma) megalostigma (Ducke) Oeldiscelis] megalostigma Ducke 1908:62; Ducke 1912:83. Oediscelis megalostigma, Nascimento 1979:7. Clhilicola] (Hlylaeosoma]) megalostigma, Michener 1994:87. This species is known from the Serra de Baturite, 4°15'S, 39°05'W, altitude 700 m, in the state of Ceara, Brazil. Although on a mountain, this location in the xeric northeast of Brazil is a very different hab- itat from the moist forest areas where oth- er species of this group are found. The specimens were taken by Ducke on the flowers of Borreria verticillata Mey. (Rubi- aceae). The lectotype, selected but not pub- lished by J. S. Moure and C. D. Michener, was designated (i.e., published) by Nas- cimento (1979); it is in MPEG. Lectopara- types are in MPEG, KSEM, and presum- ably in the collection of Padre J. S. Moure, Universidade Federal do Parana, Curitiba. Specimens are also in the Departamento de Zoologia, Universidade de Sao Paulo, Sao Paulo, Brazil. A front tarsus from two different fe- males was removed and examined in glyc- erin (and preserved in microvials on the pins). The apical processes of the first two 134 Journal of Hymenoptera Research tarsal segments each ends in a curved bristle coarser than the other tarsal bris- tles. Bases of these bristles are shown in Figure 3. A bristle that is nearly as coarse, however, arises from the lower distal end of the third tarsal segment. These bristles are brittle, easily broken off so that they look like pegs even at a magnification of SOX. Chilicola (Hylaeosoma) polita Michener Chilicoln (Hi/laeosomn) polita Michener 1994:87. Except for the characters indicated in the key to species, C. polita appears not to differ from C. megalostigma. When the front tarsus of the female is in its usual slightly down-curved position, the process of the basitarsus is appressed against the under side of the second segment and is unrecognizable, its hairs seemingly arising from the second segment. We recognized the process only when the tarsus was re- laxed and bent under a dissecting micro- scope (Fig. 2). We have not been able to examine the front tarsi of females of C. pol- ita at high magnification in a liquid such as glycerin; there may be no consistent dif- ference between C. polita and C. megalostig- ma in this structure (see key to species). Chilicola polita is known from Veracruz and Oaxaca, Mexico, and Chiriqui Prov- ince, Panama. A new record is as follows: Costa Rica: San Jose Province, San Anto- nio de Escazu (Eberhardt's house) [9°55'N, 84°08'W], May 27, 1996, on Ciiphea (G. Melo, collector). [See Addendum.] The specimen from Costa Rica as well as the holotype from Veracruz are in KSEM. Chilicola {Hylaeosoma) stenocephala Brooks and Michener, new species In its elongate body form (Figs. 5, 6) this species resembles other species of Hylaeo- soma, although the length is exaggerated by the long head which in all specimens available is directed forward in a prog- nathous position (Fig. 7), rather than being hypognathous as in most bees. Other spe- cies of the subgenus have a tendency to the same condition, especially in species with long heads. This is likely to be relat- ed to pushing down into small deep flow- ers or flower clusters. The key to species indicates some of the principal differences between C. stenocephala and its closest rel- atives, C. megalocephala and polita. Male: Body length 6 mm. Coloration: Black, the following parts testaceous: lon- gitudinal band on clypeus (yellowish), la- brum, mandible, malar area (dark), ante- rior part of hypostomal area, antennal scape and pedicel, under side of flagellar segments 2 to 4 (sometimes dark brown), legs (femur, tibia and tarsus of posterior leg infuscated). The following transparent amber: posterior pronotal lobe, tegula, ax- illary sclerites of wings, apical margins of metasomal terga. Bases of T2 to T4 and metasomal sterna except SI brown. Wings transparent with veins and stigma black. Body surface: Highly polished and shin- ing, with widely separated, not sharply defined punctures, except propodeum; dorsal surface of propodeum with fine longitudinal, radiating striae; sides of pro- podeum lineolate with small punctures, rather dull; metasomal sterna lineolate. Hairs: Sparse, simple, erect, dark testa- ceous to dull whitish, notably long on me- tasoma, especially posteriorly; frons with V-shaped pattern of hairs, lower end be- tween antennae, arms extending to upper part of frons (Fig. 8); pronotal lobe mar- gined with short, appressed white hairs; propodeum with short, appressed white hairs that do not hide surface, except for hairless striate dorsal surface. Structure: Head elongate (Figs. 7, 8), proportions as shown in illustrations, paraocular area, with well developed depression for anten- nal scape extending upward toward ocel- locular area from antennal base (as in oth- er Hylaeosoma); medial to depression a lon- gitudinal, impunctate welt (white on left of Fig. 8) extending from antennal base to upper ocular tangent; first flagellar seg- Volume 8, Numbek 2, 1999 135 Figs, 5-7. Clnluohi {Hiiliicosoiiin) stciioccpluihi n. sp. 5, male holotype; 6, temale paratype; 7, dorsal view to show thorax, male. ment shorter than pedicel, about 1.5 times as long as broad, segments 2 and 3 broad- er than long, remaining segments longer than broad, 11 over three times as long as broad; malar area about as long as wide (Fig. 9); ocelloccipital distance about two ocellar diameters measured to apex of high preoccipital carina; genal area above about as wide as eye seen from side, be- low tapering to vanishing point. Maxillary palpus about as long as prementum, near- ly as long as head, first two segments short, remaining four segments long, straight, progressively slightly shortened toward apex, so that segment 3 is longest (also broadest). Pronotum with large dor- sal surface (Figs. 5, 7) about on level with scutum which is over three times as long as mid-dorsal part of pronotum. Legs slender (Fig. 5). Wings as shown in Figure 5; apex of marginal cell minutely truncate. Tl longer than broad, in profile a slight constriction between Tl and T2 (Fig. 5); T7 rather narrowly rounded at apex; poste- rior half of S6 with strongly elevated lon- gitudinal ridge tapering posteriorly, end- ing in narrow, sharp, bristled apex (Figs. 10, 11). T7, T8 and genitalia as in Figures 12 to 14; genitalia with apices of gonofor- ceps much elongated and curved mesad, thus differing from those of the other spe- cies of this group (see Michener 1994, Fig. 10) as well as from all other Chilicola (see Toro and Moldenke 1979). Female: As described for male except as follows: Clypeus black, without testaceous band; flagellum blackish; pronotal lobe dark brown; legs infuscated dark brown except trochanters testaceous; metasoma black, dark brown ventrally. Scopa of nearly erect whitish plumose hairs on SI to S3, longest on S2. Scopal hairs of hind femur plumose, shorter than femoral di- ameter; hairs of hind tibia long and sim- 136 Journal of Hymenoptera Research Figs. 8-14. Male of Chilicola {Hi/laeoacmii) steiwccphain n. sp. 8, face, sculpturing on left, hairs on right; 9, side view of head; 10, S5 and S6; 11, side view of S6, ventral side to the left; 12, genitalia, dorsal on left; 13, S7, right hand half, ventral view; 14, S8, dotted line across extreme base represents actual condition of spec- imen; but the base has been drawn in what is presumably its undamaged shape. pie. Flagellum short, segments broader than long except 8 and 9 which are slight- ly longer than broad and 10 which is about twice as long as broad; last two seg- ments of maxillary palpus missing, prob- ably broken off in the single specimen; an- terior tarsus bristly, segments 1 and 2 with large, apical, down-curved processes lack- ing apical hairs or pegs (Fig. 4); Tl about as long as broad; apex of metasoma un- modified. Holotype male and one male and one female paratype: Colombia: Amazonas: La Chorrera [0°44'S, 73°0rW], 24-31 August, 1976 (M Cooper). One additional male paratype, same data but 3 September, VoLUMb 8, Number 2, 1999 137 1976. All bear the Natural History Muse- um (London) number BM 1976-727, and the specimens are in that Museum except one male paratype in KSEM. NOTES ON TYPES While on the subject of tropical species of Chilicola, it is appropriate to publish in- formation on the types of two species that had been lost. In 1955 Padre J. S. Moure and C.D.M. remounted many specimens in the Ducke collection in the Museu Par- aense Emilio Goeldi in Belem, Para, Brazil (MPEG). Remounting was necessary be- cause of corroded pins that broke at the levels of labels and inside of specimens, sometimes bursting them. Many species were represented by series of syntypes. We selected and labeled lectotypes and lectoparatypes. These were published, and thus formally designated, by Nascimento (1979) in his catalog of hymenopteran types in MPEG. That catalog makes no mention of Oed- iscelis Iniheri Ducke and O. minima Ducke, although types of O. megalostigma Ducke and other species described by Ducke in the same paper were designated. The rea- son has only now become evident, since the lectotypes of O. hiiberi and minima were found by G. Melo in KSEM, where they must have been left by C.D.M. or by Padre Moure. They will be returned to MPEG. The following notations and lec- totype designations are in the style of Nas- cimento's (1979) catalog of types. Oediscelis huberi Ducke 1908:63. Lectotipo. Brasil: Ceara, Serra de Batur- ite [4°15'S, 39°05'W], 600 m, 3- VIII- 08 (Ducke). (Des. Moure e Miche- ner, 1955). The headless female is badly mounted on a paper point. To judge by the wing venation this is a member of the subgenus Hi/laeofoma, where it was placed by Mich- ener (1995). The abundant punctures and unmodified front tarsus exclude it from the megalofitigma group. Oediscelis minima Ducke 1908:63. Lectotipo: Brasil: Ceara, Quixada' [04°58'S, 39°01'W], 4-VII-08 (Ducke). (Des. Moure e Michener 1955). The specimen consists of a female head and anterior half of thorax, with one badly damaged forewing, but no other wings, badly mounted on a paper point. This spe- cies is a member of the subgenus Proso- poides, where it was placed by Michener (1995). ACKNOWLEDGMENTS For the loan of the specimens of Chilicola stenoce- phntn we thank Suzanne Lewis and Christine Taylor of the Natural History Museum [London). This is contribution number 3231 from the Division of En- tomology, University of Kansas Natural History Mu- seum, Lawrence, Kansas 66045, U.S.A. LITERATURE CITED Ducke, A. 1908. Contribution a la connaissance de la faune hymenolopterologic]ue du nord-est du Bre- sil, II Hymenopteres recoltes dans I'etat de Ceara en 1908. Revue d'Entomologie 27: 57-87. Ducke, A. 1912. Die natiirlichen Bienengenera Siida- merikas. Zoologische lahrbucher, Abteilung fiir Sys- tcinatik, Ceograpliie unci Biolngie lier Ticre 34: 51- 116. Michener, C. D. 1994. Mexican and Central American species of Chilicola. Folia Entoiuologica Mexwaiui no. 85 (1992): 77-93. Michener, C. D. 1995. A classification of the bees of the subfamily Xeromelissinae. Journal of the Kan- sas Entomological Society 68: 332-345. Nascimento, P. T. R. 1979. Catalogo de tipos ento- mologicos do Museu Goeldi, Hymenoptera. Bol- etim do Museu Paraense Emilio Goeldi, Zoologia no. 96: 1-18. Toro, H. and A. Moldenke. 1979, Revision de los Xe- romelissinae Chilenos. Anales del Museo de His- toria Natural Valparaiso 12: 95-182. ADDENDUM We have just received four specimens of Chilicola (Hy- laeosoma) polila Midiener that add substantially to its dis- tribution as indicated in the bod\' of this pap>er. Tlie first two, listed below, are from the Centre for Land and Biological Resources, Ottawa, Ontario, Canada. The oth- er two were received through Laurence Packer, York 138 Journal of Hymenoptera Research University, North York, Ontario, Canada. The data are as follows (one specimen from each locality): Guatemala: Zacapa, 3.5 km southeast of La Union, 1500 m altitude, 20-27 June 1993, in flight intercept trap (J. Ashe, R. Brooks). Mexico: Chiapas: Pk. Lago Belgica, 14 June 1989, in flight intercept trap (H. Howden); Puebla: 22 km north of Xicotepec de Juarez, 1070 m altitude, 17 June 1983 (M. Kaulbars); Tamaulipas: Rancho del Cielo near Gomez Farias, 1000 m altitude, 7 Aug. 1983 (M. Kaulbars). The last locality is the northernmost record for the genus Cliilicoln. J. HYM. RES. Vol. 8(2), 1999, pp. 139-153 Descriptions of New Genera from Brazil in the Tribes Heterospilini and Spathiini With Similar Wing Venation (Hymenoptera: Braconidae, Doryctinae) Sandra M. Barbalho, Angelica M. Penteado-Dias and Paul M. Marsh (SMB, AMP-D) Departamento de Ecologia e Biologia Evolutiva, Universidade Federal de Sao Carlos, Via Washington Luiz km 235, Caixa Postal 676, CEP 13-565-905, Sao Carlos, SP, Brazil; (PMM) P. O. Box 384, North Newton, KS 67117 USA (Cooperating Scientist, Systematic Entomology Laboratory, USD A, Washington, DC) Abstract. — Six new genera and 11 new species are described from Brazil. These new taxa all have wing venation similar to that found in the genus Heterospilus, with fore wing vein 2RS absent or weakly indicated. The new genera are placed in the tribes Heterospilini and Spathiini and a key to the New World genera of these tribes is given. The new taxa described are as follows: Amazondonictes Barbalho and Penteado-Dias, n.gen., A. bicolor Barbalho and Penteado-Dias, n. sp., A. ater Barbalho and Penteado-Dias, n. sp.; Canchim Barbalho and Penteado-Dias, n.gen., C. carinatus Barbalho and Penteado-Dias, n. sp., C. enigosus Barbalho and Penteado-Dias, n. sp.; Heterospathius Barbalho and Penteado-Dias, n.gen., H. belokohylskiji Barbalho and Pen- teado-Dias, n. sp., H. petiolatus Barbalho and Penteado-Dias, n. sp., H. silvaticiis Barbalho and Penteado-Dias, n. sp.; Jataiella pilosa Barbalho and Penteado-Dias, n.gen., n.sp.; Leptodonjctes luizi Barbalho and Penteado-Dias, n.gen., n.sp.; Pioscelus austrimis Marsh, n. sp.; Spathiospilus brasiliensis Marsh, n. gen., n. sp. The subfamily Doryctinae is one of the most diverse groups in the Braconidae, es- pecially in the Old and Ne^N World trop- ics. Recent studies by Marsh for the Neo- tropics (see for instance Marsh 1993) and Belokobyl'skij for the Old World tropics (see for instance Belokobyl'skij 1994a, b, 1995) have shown the incredible diversity at the generic level and have lead to re- vised concepts of previously described genera. The genus Heterospilus was described by Haliday in 1836 and characterized by the absence of fore wing vein 2RS, thus the first and second submarginal cells are more or less confluent. In some cases, vein 2RS may be sclerotized but only appears as an infuscate line with no distinct vein edges such as a true tubular vein (see Ma- son 1986 for definition of wing vein types in braconids). This character was unique for the Bracorudae and for many years Heterospilus was the only genus known with such wing venation. Several more genera that were subsequently described with this type of wing venation have been synonymized with Heterospilus (Shenefelt and Marsh 1976; Belokobyl'skij 1992). Muesebeck and Walkley (1951) proposed the genus Pioscelus for two species which differed from the typical Heterospilus in other morphological characters. Hedqvist (1963) described the genus Labania which had this same wing venation; he placed it in the subfamily Hormiinae but stated that it had more affinities to the Doryctinae. The genus Heterospilus is a very large ge- nus with about 500 species in the New World, most of which are undescribed. Thus, most doryctines with the fore wing vein 2RS absent have been placed in this genus. 140 Journal of Hymenoptera Research Recent studies of the Neotropical fauna of the Doryctinae have revealed several forms with this wing venation but which differ in many distinct characters from typical Hetewspiliis. Until recently, these would have all been placed near Hetewspi- liis in the tribe Heterospilini. However, some of these are more closely related to the genus Spathius Nees with its petiolate metasoma. These genera are properly placed in the tribe Spathiini as defined by Belokobyl'skij (1992). The genus Labania is placed in its own tribe, Labaniini, and is more closely related to the tribe Ecphylini. The purpose of this paper is to provide descriptions of several new genera from Brazil in the tribes Heterospilini and Spa- thiini which have this characteristic wing venation. Keys to the New World genera are also provided. This is part of a long term program by the first author (SMB) to study the Doryctinae of Brazil. MATERIALS AND METHODS The subfamily Doryctinae can be iden- tified using the keys provided by Wharton et al. 1997. The New World genera can be identified using the key provided by Marsh (1997) and a key to the tribes of Doryctinae is provided by Belokobyl'skij (1992). Morphological and wing venation characters are based on Wharton et al. (1997). Taxonomic authority for the new genera and species described here is as in- dicated for each taxon. Abbreviations for institutions that provided specimens for this study and where type specimens are deposited are: DCBU, Departamento de Ecologia e Biologia Evolutiva da Univer- sidade Federal de Sao Carlos, Sao Carlos, SP, Brazil; INPA, Instituto Nacional de Pesquisas da Amazonia, Manaus, Brazil; UFPR, Universidade Federal do Parana, Brazil; USNM, National Museum of Nat- ural History, Smithsonian Institution, Washington, DC. TRIBE HETEROSPILINI FISCHER This tribe is distinguished from the Spa- thiini by the following characters: meta- soma not petiolate, the first metasomal ter- gum being usually broad and short with length not much longer than apical width; the acrosternite of the first metasomal seg- ment short, less than K length of the ter- gum and not fusing with the lateral mar- gins (Fig. 13); dorsope on first metasomal tergum usually distinct and deep (see Be- lokobyl'skij 1992). KEY TO THE NEW WORLD GENERA OF THE TRIBE HETEROSPILINI L Fore wing vein r-m absent (Figs. 4, 5) 2 Fore wing vein r-m present (Figs. 1-3) 3 2(1). Hind wing veins cu-a and m-cu present (Fig. 5) Canchim Barbalho and Penleado-Dias, new genus - Hind wing veins cu-a and m-cu absent (Fig. 4) Leptodoryctes Barbalho and Penteado-Dias, new genus 3(1). Hind coxa with a more or less distinct antero-ventral basal tubercle or tooth 4 - Hind coxa round at base, without antero-ventral basal tubercle or tooth 5 4(3). Body densely covered with long white hair (Figs. 7-9) Jataiella Barbalho and Penteado-Dias, new genus - Body usually sparsely covered with short hair Heterospilus Haliday 5(3). Second metasomal tergum with two posteriorly converging grooves (Fig. 15) Pioscelus Muesebeck and Walkley - First and second metasomal terga with two median parallel carinae (Fig. 29) Amazondoryctes Barbalho and Penteado-Dias, new genus Volume 8, Number 2, 1999 141 Figs. 1-9. Figs. 1-6, wings: 1, Jataiella pihsa n. sp.; 2, Amazondon/ctes ater n. sp.; 3, Heterospathiiis pelichitus n. sp.; 4, Leptodoryctes liiizi n. sp.; 5, Canchim cariimtus n. sp.; 6, Spathiospihis brasiliensis n. sp. Figs. 7-9, /. pdofa: 7, mesonotum; 8, propodeum; 9, metasoma. 142 Journal of Hymenoptera Research Amazondoryctes Barbalho and Penteado-Dias, new genus Type species. — Amazondoryctes bicolor Barbalho and Penteado-Dias, new species. Diagnosis. — Face striate; vertex, malar space and temples smooth and shining, frons smooth or slightly striate; propleu- ron with transverse carinae; pronotum weakly granulate with longitudinal scro- biculate groove; mesopleuron smooth; no- tauli distinctly scrobiculate; sternaulus complete; propodeum strongly areolate- rugose; hind coxa round at base, without basal tubercle; fore wing vein 2RS absent, first and second submarginal cells conflu- ent, first subdiscal cell open at apex (Fig. 2); hind wing vein M-I-CU shorter than IM; first, second and base of third meta- somal terga striate, remaining terga slight- ly granulate, first and second terga with strong complete parallel carinae (Fig. 29). Distribution. — Amazonas State of Brazil. One of us (PMM) has also seen an appar- ently undescribed species from Costa Rica. Comments. — This genus is similar to Pioscelus and will run to that genus in the key provided by Marsh (1997) but is dis- tinguished by the parallel, rather than di- verging, carinae on the second metasomal carina. Etymology. — The generic name is in ref- erence to the locality of the two known species from Amazonia. KEY TO SPECIES OF THE GENUS AMAZONDORYCTES 1. First and secorid metasomal terga striate, remaining terga smooth and shining; frons slight- ly excavated A. bicolor Barbalho and Pentado-Dias, new species - First, second and basal half of third metasomal terga striate; frons not excavated A. ater Barbalho and Penteado-Dias, new species Amazondoryctes bicolor Barbalho and Penteado-Dias, new species (Fig. 30) Female holotype. — Head: circular; 28 an- tennomeres, first flagellomere shorter than scape plus pedicel; occipital carina meet- ing hypostomal carina; oral cavity small, diameter equal to malar space and Vs eye height; face striate, 1.2 times longer than eye height; vertex, frons, malar space and temples smooth and shining; frons slightly excavated. Mesosoma (Fig. 30): propleu- ron with transverse carinae; pronotum weakly granulate, with longitudinal scro- biculate groove; mesopleuron smooth; mesonotum angled and declivous anteri- orly; notauli scrobiculate and meeting be- fore scutellum in costate-rugose area; mid- dle mesonotal lobe granulate basally, al- veolate apically, lateral lobes granulate; sternaulus complete, weakly scrobiculate; metapleuron areolate; propodeum areo- late with two longitudinal carinae lateral- ly. Legs: fore tibia with single row of 8 spines on anterior edge; hind coxa without basal tubercle. Wings (as in Fig. 2): fore wing vein 2RS absent, first subdiscal cell open at apex, vein 2cu-a absent or repre- sented by a fuscous spot; hind wing vein M-hCU shorter than IM, Ir-m more than half length of IM, m-cu weak or absent. Metasoma: first and second terga striate, remaining terga weakly granulate; parallel median carinae on first and second terga complete and strong; ovipositor shorter than metasoma. Color: body and antenna dark brown, face yellowish, vertex dark brown, first metasomal tergum black, re- maining terga brown, fore and middle legs yellow, hind coxa and femur brown, tibia yellow basally, brown apically, ovi- positor sheaths yellow with black at apex, Volume 8, Number 2, 1999 143 wings slightly dusky. Body length: 3.1 mm. Male. — Unknown. Holotype female. — BRAZIL: Amazonas, Manaus, ZF3, Km 23, Fazenda Esteio Res. 1112, November 26, 1986, B. Klein col. De- posited in INPA. Eti/mologi/. — The specific name is in ref- erence to the bicolored body with the face yellow and mesosoma and metasoma dark brown or black. Amazondoryctes ater Barbalho and Penteado-Dias, new species (Figs. 2, 29) Female holotype. — Agrees with the de- scription of A. bicolor except as follows: 38 antennomeres; face with converging sculpture; frons not excavated, slightly striate and with rugose sculpture between toruli and eyes; eye height 2.4 times lon- ger than diameter of oral cavity; first, sec- ond and basal half of third metasomal ter- ga striate (Fig. 29); fore tibia with 3 rows of 22 strong spines; head and mesosoma entirely black, first and second metasomal terga black, remaining terga black on bas- al half, light brown on apical half; body length 4.6 mm. Male. — Unknown. Holotype female. — BRAZIL: Amazonas, Manaus, ZF3, Km23, Fazenda Esteio Res. 1208, February 9, 1985, B. Klein col. De- posited in INPA. Paratypes. — BRAZIL: 1 female, Amazon- as, Manaus, ZF3, Fazenda Esteio Res. 1112, February 9, 1985. Deposited in DCBU. Etymology. — The specific name is in ref- erence to the black color. Canchim Barbalho and Penteado-Dias, new genus Type species. — Canchim cnrinatus Barbal- ho and Penteado-Dias, new species. Diagnosis. — Vertex and frons striate or smooth; face only slightly striate; temple and malar space smooth; mesonotum de- clivous anteriorly; notauli and sternaulus complete and scrobiculate; propodeum areolate-rugose; hind coxa rugose and with basal tubercle; fore tibia with row of 8 spines on anterior edge; fore wing veins r-m and 2RS both absent; hind wing vein M-f-CU slightly longer than IM; first and at least basal half of second metasomal terga striate, terga beyond third smooth and shining. Distribution. — Sao Paulo State of Brazil. Comments. — This genus will run to Het- erospilus in the key provided by Marsh (1997) but is distinguished by the absence of fore wing vein r-m, thus all submargin- al cells are confluent. Etymology. — The generic name is in ref- erence to the localities for the known spe- cies, Fazenda Canchim, Sao Carlos, SP, Brazil. KEY TO SPECIES OF THE GENUS CANCHIM 1. First, second and basal 'A of third metasomal terga striate; vertex and frons striate C. carinatus Barbalho and Penteado-Dias, new species - First and basal Vi of second metasomal terga striate; vertex and frons often smooth C. enigostis Barbalho and Penteado-Dias, new species Canchim carinatus Barbalho and Penteado-Dias, new species (Figs. 5, 26-28) Female. — Head (Fig. 28): occipital pre- sent and meeting hypostomal carina; ver- tex and frons striate; face slightly striate; temple and malar space smooth; face height 1.4 times longer than eye height; eye width 2.2 times temple width; oral cavity about equal to malar space; 17 an- 144 Journal of Hymenoptera Research tennomeres; first flagellomere equal to length of scape and pedicel. Mesosoma (Fig. 26): pronotum, mesopleuron and me- sonotum granulate; mesonotum declivous anteriorly; notauli complete and scrobic- ulate; sternaulus scrobiculate, not com- plete; propodeum areolate-rugoae. Legs: hind coxa rugose, with basal tubercle; fore tibia with row of 8 short spines on anterior edge. Wings (Fig. 5): fore wing veins r-m and 2RS both absent, first subdiscal cell open at apex, vein 2cu-a absent; hind wing vein M + CU slightly shorter than IM. Me- tasoma (Fig. 27): length of first metasomal tergum equal to apical width; first, second and basal third of third terga striate, re- maining terga smooth and shining; ovi- positor V4 length of metasoma. Color: head light brown; mesosoma and metasoma dark brown; second tergum sometimes with triangular yellow spot at base; legs yellow; wings hyaline, veins brown. Body length 2 mm. Male. — Unknown. Holotype female. — BRAZIL: Fazenda Canchim, Sao Carlos, Sao Paulo State, Oc- tober 9, 1996. Deposited in DCBU. Parah/pes. — BRAZIL: 2 females, same data as holotype except dates of June 27, 1985 and July 26, 1989. Deposited in DCBU. Etymology. — The specific name is in ref- erence to the sculpture of the head. Canchim erugosus Barbalho and Penteado-Dias, new species (Fig. 25) Female. — Agrees with the description of C. carinatiis except as follows: head smooth and shining, vertex slightly striate (Fig. 25); mesopleuron slightly granulate; first and basal % of second metasomal ter- ga striate; ovipositor about Vi length of metasoma; body length 1.8 mm. Male. — Unknown. Holotype female. — BRAZIL: Fazenda Canchim, Sao Carlos, Sao Paulo State, April 25, 1985, A. S. Soares col. Deposited in DCBU. Paratypes. — BRAZIL: 1 female, Telemaco Borba, PR, September 1, 1986, PROFAU- PAR (Levantamento da Fauna Entomolo- gica do Estado do Parana). Deposited in UFPR. Etymology. — The specific name is in ref- erence to the smooth head. Jataiella Barbalho and Penteado-Dias, new genus Type s/'fcr'fs. — Jataiella pilosa Barbalho and Penteado-Dias, new species. Diagnosis. — Body densely covered with long white hair; face, temples and vertex smooth and shining; frons striate; prono- tum striate laterally; propleuron and me- sopleuron smooth; notauli deep, smooth and meeting in a triangular area sculp- tured (Fig. 7); sternaulus absent; hind coxa with a small rounded tubercle at base; fore wing vein 2RS absent except for short stub, first subdiscal cell open (Fig. 1); hind wing vein M + CU shorter than IM, male with stigma in hind wing; first metasomal tergum (Fig. 9) slightly longer than apical width; metasomal terga 1-3 and base of 4 striate, remaining terga striate at base, granular at apex. Distribution. — Sao Paulo State of Brazil. Comments. — This genus is distinguished from other genera of Heterospilini by the very densely hairy body, which will sep- arate it from Heterospilus in the key pro- vided by Marsh (1997). Etymology. — The generic name is in ref- erence to the locality of the type species. Jataiella pilosa Barbalho and Penteado-Dias, new species (Figs. 1, 7-9) Female holotype. — Head: densely hairy; occipital carina present and meeting hy- postomal carina; face and vertex smooth and shining, frons striate; face height 1.7 times eye height; eye width 1.8 times tem- ple width; malar space Vj oral cavity; first flagellomere equal to length of scape and pedicel combined. Mesosoma: densely hairy; pronotum striate laterally; propleu- Volume 8, Number 2, 1999 145 ron and mesopleuron smooth and shining; middle mesonotal lobe more elevated than lateral lobes, middle lobe with complete median longitudinal groove (Fig. 7); no- tauli deep, smooth, meeting before scutel- lum in area with longitudinal carinae and few cross carinae (Fig. 7); sternaulus ab- sent; propodeum rugose-areolate with two basal smooth areas (Fig. 8). Legs: hind coxa with small tubercle at base; fore tibia with row of 9 spines on anterior edge. Wings (Fig. 1): fore wing vein 2RS absent except for short stub, vein r-m present, first subdiscal cell open at apex; hind wing vein M+CU shorter than IM, r-m less than half length of IM. Metasoma (Fig. 9): first metasonal tergum 1.2 times longer than apical width; terga 1-3 and base of 4 striate, remaining terga striate at base, granular at apex; ovipositor about % length of metasoma. Color: body black, mesopleuron dark brown; wings infuscat- ed. Body length: 5.6 mm. Male. — Essentially as in female except as follows: propodeum smooth; all metaso- mal terga except apical one striate; stigma in hind wing (Fig. 1); only lower part of mesopleuron dark brown. Holotype female. — BRAZIL: Rio Mogi- Guagii, Luiz Antonio, Sao Paulo, Novem- ber 26, 1993, L. A. Joaquim col. Deposited in DCBU. Paratypes. — BRAZIL: 1 male, same data as holotype. Deposited in DCBU. Leptodoryctes Barbalho and Penteado-Dias, new genus Type species. — Leptodoryctes luizi Barbal- ho and Penteado-Dias, new species. Diagnosis. — Head, pronotum, propleu- ron, mesopleuron and mesonotum smooth and shining; notauli meeting well before scutellum, sometimes incomplete; sternau- lus complete and scrobiculate; propodeum rugulose; hind coxa round at base; fore wing veins 2RS, r-m, 2M and 3M absent, vein 2-lA also absent and, thus, first sub- discal cell absent; hind wing veins cu-a and m-cu absent; first metasomal tergum weakly striate, remaining terga smooth and shining. Distribution. — Sao Paulo, Amazonas and Rio de Janeiro States of Brazil. Comments. — This genus is distinguished from most other braconids by the wing ve- nation with the absence of several veins in both fore and hind wings. It can be defi- nitely placed in the Doryctinae by the presence of a row of stout spines along the anterior edge of the fore tibia, presence of a flange at the apico-lateral corner of the propleuron, circular opening between clypeus and mandibles and presence of the occipital carina. The genus will not fit well in the key to genera provided by Marsh (1997) but would possible run to Heterospilus because of the absence of fore wing vein 2RS. Etymology. — The generic name is from the Greek leptos, meaning slender, weak, in reference to the fragile appearance of this genus. Leptodoryctes luizi Barbalho and Penteado-Dias, new species (Figs. 4, 10-12, 31) Female holotype. — Head (Fig. 12): com- pletely smooth and shining; occipital ca- rina present and meeting hypostomal ca- rina; face height 2.3 times eye height; eye width equal to temple width; oral cavity diameter 1.2 times malar space. Mesosoma (Fig. 11): pronotum, propleuron, meso- pleuron and mesonotum smooth and shining; notauli not complete, meeting well before scutellum; sternaulus com- plete and scrobiculate; propodeum ru- gose. Legs: hind coxa round at base; fore tibia with row of 8 spines on anterior edge. Wings (Fig. 4): fore wing veins 2RS, r-m, 2M and 3M absent, first subdiscal cell absent, vein 2-1 A absent; hind wing veins cu-a and m-cu absent. Metasoma (Fig. 10): First metasomal tergum 3.2 times longer than apical width, weakly striate; remain- ing terga smooth and shining; ovipositor as long as entire body. Color: head, me- sosoma, legs and first metasomal terga 146 Journal of Hymenoptera Research Figs. 10-19. Figs. 10-12, Leptodoryctes hiizi: 10, meldhoma; 11, mesosoma; 12, head. Fig. 13, Hctcnvpilui sp., ventral view, first metasomal segment. Fig. 14, Notiospathiiis sp., ventral view, first metasomal segment. Fig. 15, Pioscclus aiistrinus n. sp., metasoma, dorsal view. Figs. 16-19, SpMithiospihis hnisiUcnsis: 16, habitus; 17, face; 18, first metasomal segment, ventral view; 19, mesosoma and metasoma, dorsal view. yellow, remaining terga and hind femur light brown; wings lightly infuscated. Body length: 1.6 mm. Male. — Essentially as in female except as follows: head and mesosoma dark brown, metasoma light brown, legs yellow; fore wing vein lA weak. Holotyye female. — BRAZIL: Amazonas, Manaus, Reserva Ducke, September 6, 1993, M. T. Tavares, col. Deposited in DCBU. Parati/pes. — BRAZIL: 2 males, Ilha Grande, Rio de Janeiro, RJ, May 6, 1997, L. A. Joaquim, col.; 1 male, Esta^ao Experi- mental de Ubatuba, SP, November 15, 1990, L. A. Joaquim, col. Deposited in DCBU. Volume 8, Number 1, 1999 147 Etymology. — This species is named af- ter the collector and our friend Luiz Joa- quim. Pioscelus Muesebeck and Walkley Piosccliis Muesebeck and Walkley, 1951: 180. Pioscelus austrimis Marsh, new species (Fig. 15) Female. — Body color: head honey yel- low; scape, pedicel and basal 3-6 flagel- lomeres honey yellow, remainder of fla- gellomeres brown; mesonotum brown, mesonotum and mesopleuron often light brown; fore and middle legs yellow, hind coxa except at apex and hind femur brown, apex of hind coxa, trochanters, tib- ia and tarsus yellow; metasoma brown, grooves on second metasomal tergum sometimes lighter; wings hyaline, veins including stigma light brown. Body size: 4.0 mm. Head: antenna with at least 19 an- tennomeres (broken in all females of type series), all flagellomeres at least 5 times longer than wide, first flagellomere very slightly shorter than second; vertex and temple smooth and shining; frons exca- vated, mostly smooth and shining but with a few striations medially; face rugu- lose and dull; eyes large, malar space about Vi eye height; ocelli small, ocell-oc- ular distance about twice diameter of lat- eral ocellus; occipital carina complete, reaching hypostomal carina. Mesosoma: pronotum coriaceous with median scro- biculate groove; mesonotum coriaceous, median lobe sharply declivous anteriorly with lateral corners broadly produced, no- tauli scrobiculate, meeting before scutel- lum in narrow longitudinally rugose-car- inate area; mesopleuron smooth medially, subalar area broadly scrobiculate, sternau- lus smooth, as long as mesopleuron; pro- podeum slightly longer than first metaso- mal tergum, not declivous apically, ru- gose-areolate laterally, rugose dorsally with basal lateral areas indistinct, coria- ceous. Wings: fore wing vein 2RS weak or absent, at most indicated by weak infus- cated line apically, vein 3RSa twice as long a vein r, first subdiscal cell open at apex, vein 2cu-a absent; hind wing vein cu-a present, vein M + CU about Vi length of IM. Legs: hind coxa angled at base but without a distinct antero-ventral basal tu- bercle; hind femur short and swollen, about three times longer than width. Me- tasoma (Fig. 15): first tergum with length about twice as long as apical width, strongly longitudinally carinate; second tergum with two converging grooves en- closing a basal semicircular carinate area, tergum carinate laterally; third tergum separated from second by transverse ar- cuate groove, carinate on basal %, coria- ceous apically; remainder of terga coria- ceous basally, smooth apically; ovipositor about Vi length of metasoma. Male. — Essentially as in female; hind fe- mur more greatly swollen, about twice as long as wide; hind wing without stigma. Holotype female. — BRAZIL: Manaus, ZF3, Km23, Faz. Esteio, Res., 1208, B. Klein col., November 5, 1985. Deposited in INPA. Paratypes. — BRAZIL: 2 females, same data as holotype, February 27, 1985, No- vember 18, 1987; 1 female, Cerrado, Can- chim, Sao Carlos, S.P., December 4, 1989, L. A. Joaquim, col.; 1 male, Mata Canchim, Sao Carlos, S.P., April 26, 1996, L. A. Joa- quim, col. Desposited in DCBU. Etymology. — The specific name is from the Latin anstriiuis meaning southern in reference to this species being the first one recorded from South America. TRIBE SPATHIINI FOERSTER This tribe is distinguished from the Heterospilini by the following charac- ters: metasoma petiolate, first tergum usually long and narrow, acrosternite very long, at least Vi length but usually nearly as long as the tergum, fused with lateral margins (Figs. 14, 18); dorsope on first tergum weak or absent (see Belo- kobyl'skij 1992). 148 Journal of Hymenoptera Research KEY TO THE NEW WORLD GENERA OF THE TRIBE SPATHIINI 1. Fore wing vein 2RS absent or not sclerotized (Figs. 3, 6) 2 Fore wing vein 2RS present and sclerotized 3 2(1). Eyes large, malar space very short or absent (Fig. 17); fore wing vein RS+M arched (Fig. 6); hind coxa with small but distinct antero-ventral basal tubercle Spathiospilus Marsh, new genus - Eyes smaller, malar space at least % eye height; fore wing vein RS + M not arched (Fig. 3); hind coxa without basal tubercle (Fig. 21) Heterospathius Barbalho and Penteado-Dias, new genus 3(1). Hind coxa round at base, without basal tubercle 4 - Hind coxa with distinct basal tubercle or tooth at base 5 4(3). Hind wing vein m-cu curved toward wing apex, hind wing vein M+CU % length of vein IM Psetwbolus Reinhard - Hind wing vein m-cu curved toward wing base, hind wing vein M + CU Va length of vein IM Notiospathins Matthews and Marsh 5(3). First metasomal tergum without triangle area at base Spathiiis Nees - First metasomal tergum. with distinct triang.ular area .at. base 6 6(5). Fore wing vein m-cu asising distad of 2RS Ptesimogaster Marsh - Fore wing vein m-cu arising basad or directly in line with 2RS Trigonophasmus Enderlein Heterospathius Barbalho and long and slender, length at least 4 time Penteado-Dias, new genus apical width. rr ,, , ... .■ , . Distribution. — Amazonas, Para, Ron- Tuve spectes. — Heterosvatnius vetwlatus ,,. ,^ „,^ ^t, ■, .< D u iu J r> i. J A- ■ donia and Sao Paulo States of Brazil. Also, Barbalho and Penteado-Uias, new species. ^ , , , D.agnosis.-Diameter of oral cavity °"^ °^ "^ (^^M) ^^^ ^'^^^ ^^^^""^l ""'^^^- about equal to malar space; occipital ca- ^"^^^ ^P^"^^ ^'■°'" *=°^t^ ^'^^' ^° *^ '^•^- rina meeting hypostomal carina; 25-35 an- tnbuHon of this genus is probably over tennomeres; mesonotum declivous anteri- Central and South America, orly; notauli complete, scrobiculate; ster- Comments.— This genus is similar to No- naulus complete, scrobiculate; propodeum tiospathius but is disHnguished by the ab- horizontal for basal Va, usually with two sence of vein 2RS in the fore wing; it will lateral longitudinal carinae; hind coxa ru" ^o Heterospihis in the key provided by without basal hjbercle; fore wing vein 2RS Marsh (1997) but is separated by the shape absent, first subdiscal cell open at apex, of the petiolate metasomal segment, vein 2cu-a absent, hind wing vein M + CU Eti/mology.—The generic name is in ref- shorter than IM, m-cu absent, r-m less erence to the fore wing similarity to Het- than V2 length of IM, no stigma in hind erospilus and the first metasomal segment wing of male; first metasomal segment similarity to Notiospnthiiis. KEY TO SPECIES OF THE GENUS HETEROSPATHIUS 1. Vertex and face strongly striate-rugose; ovipositor about 1.5 times longer than body (Fig. 24) H. belokobylskiji Barbalho and Penteado-Dias, new species Vertex (Fig. 22) and face finely to weakly striate; ovipositor equal to or shorter than body 2 2(1). Ovipositor about V2 length of metasoma H. petiolatus Barbalho and Penteado-Dias, new species Volume 8, Number 2, 1999 149 Ovipositor about equal to body length H. silvaticiis Barbalho and Penteado-Dias, new species Heterospathiiis petiolatus Barbalho and Penteado-Dias, new species (Figs. 3, 20-23) Female holotype. — Head (Fig. 22): occipi- tal carina meeting hypostomal carina; face and vertex striate; frons smooth; temple smooth and shining; face height 2.1 times eye height; face width 1.6-2.1 times eye width; diameter oral cavity about Vj eye height; eye width 1.4-1.7 times temple width; malar space about equal to diam- eter of oral cavity; 25 antennomeres, first flagellomere longer than scape and pedi- cel combined. Mesosoma (Fig. 20): prono- tum rugulose-granulate; mesonotum de- clivous anteriorly; mesonotum rugulose- granulate; notauli scrobiculate, meeting at scutellum in triangular rugose area; me- sopleuron granulate; sternaulus complete, scrobiculate; propodeum areolate-rugose, with distinct longitudinal carinae. Legs: fore tibia with single row of 7 spines on anterior edge; hind coxa (Fig. 21) rugose, without basal tubercle; first tarsomere of hind tarsus 2.3 times longer than second, second equal to length of third and fourth combined, fifth twice as long as fourth. Wings (Fig. 3): fore wing vein 2RS absent or represented by short stub, first subdis- cal cell open at apex, vein 2cu-a absent; hind wing vein M-l-CU much shorter than IM, r-m less than half length of IM. Me- tasoma (Fig. 23): first tergum rugose, slen- der, length 5.7 times apical width, apical width equal to basal width; remaining ter- ga smooth and shining; ovipositor about Vz length of metasoma. Color: head, me- sosoma and first metasomal tergum dark yellow, remaining terga dark to light yel- low; ovipositor light yellow with apex black; apical 6-12 flagellomeres white, re- mainder brown; wings slightly dusly, veins brown, stigma brown with white at basal third. Body length: 3.3 mm. Variation in female. — Face occasionally weakly striate or granular, width 1.6-2.1 times eye width; eye width 1.4-1.7 times temple width; 20-32 antennomeres; fore tibia with row of 7-13 spines; metasoma with apical terga occasionally black; body length 1.5-3.7 mm. Male. — Essentially as in female except as follows: 20-25 antennomeres, apical 2-6 flagellomeres white; diameter of oral cav- ity slightly greater than malar space; face width 1.5-1.9 times eye width; frons smooth or granulate; fore tibia with row of 7-13 spines; metasoma completely light yellow or with apical Va dark brown; no stigma in hind wing. Holotype female. — BRAZIL: Amazonas, Manaus, ZF3, Fazenda Esteio, November, 1984. Deposited in INPA. Parati/pes. — BRAZIL: 2 females, 1 male, Rondonia, Ariquemes, Rio Ji, October 28, 1986, R. A. Rafael, col; 1 female, Fazenda Canchim, Sao Carlos, Sao Paulo state, April 30, 1987, L. A. Joaquim, col; 1 fe- male, Rio Tocantins, Tucurui, Para state, November, 22, 1989, N. Degullier, col.; 1 female, Santarem-Cucuruna, Para state, February, 1996, A. R. Lisboa col.; 2 fe- males, Amazonas, Manaus, ZF3, Faz. Es- teio, Res. 1301, January 22, 1986; January 29, 1986, B. Klein col.; 1 female, 1 male Amazonas, Manaus, ZF3, Faz. Esteio, Res. 1208, October 22, 1986, B. Klein col.; 1 fe- male, Amazonas, Manaus, km 60, PDBFF/ WWF, Res. 1210, November 8, 1984, B. Klein col; 1 female, Rio Tocantins, Tucu- rui, Para state; 2 females, Rio Branco, Acre state, October 25-November 8, 1991, F. Ramos, A. Henriques, I. Gorayeb & N. Bit- tencourt cols.; 1 male, Amazonas, Manaus, ZF3, Faz. Esteio, Res. 1113, January 30, 1986, B. Klein col.; 1 male, Manaus, ZF3, Faz. Esteio, Res. 1113, January 23, 1986, B. Klein col.; 2 males, Manaus, ZF3, Faz. Es- 150 Journal of Hymenoptera Research Figs. 20-31. Fig. 20-23, Hctcrospntliius pctiolntus n. sp.: 20, propodeum; 21, hind coxae; 22, vertex of head; 23, metasoma. Fig. 24, Hcterospatlui{~^ Ivtckoln/hklji n. sp., metasoma. Fig. 25, Cancltiiii criigcaui^ n. sp., vertex of head. Figs. 26-28, C. cnriiititus n. sp.: 26, mesopleuron; 27, metasoma; 28, vertex of head. Fig. 29, Ainazomti'ri/ctcf ntcr n. sp., metasoma. Fig. 30, A. bicvhr n. sp., mesopleuron. Fig. 31, Lcptodorydcti liiizt. habitus. Volume 8, Number 2, 1999 151 teio, Res. 1208, November 20, 1984; Octo- ber 17, 1984, B. Klein col.; Concei^ao do Araguaia, Para state, January 19-31, 1983, R. Nonato col. Deposited in DCBU, INPA, USNM. Etymology. — The specific name is in ref- erence to the petiolate metasoma. Heterospathitis belokobylskiji Barbalho and Penteado-Dias, new species (Fig. 24) Female. — Agrees with the description of petiolatiis except as follows: 35-38 anten- nomeres, apical 12-13 flagellomeres white; face and vertex strongly striate-rugose, vertex occasionally strongly striate; frons striate; temples smooth; face height 1.6 times eye height; fore and mid granulate; propodeum without strong longitudinal carina; ovipositor 1.5 times longer than body (Fig. 24); body entirely brown; body length 5.5 mm. Male. — Unknown. Holoti/pe female. — BRAZIL: Amazonas, Manaus, ZF3, Km32, Fazenda Esteio, Jan- uary, 1986, B. Klein col. Deposited in INPA. Paratypes. — BRAZIL: 1 female, same data as holotype; 1 female. Para, Santa- rem-Cucuruna, February, 1996, A. Pentea- do-Dias, col. Deposited in INPA, DCBU. Etymology. — This species is named in honor of our colleague and fellow re- searcher on the Doryctinae, Sergey Belo- kobyl'skij. Zoological Institute, Russian Academy of Sciences, St. Petersburg, Rus- sia. Heterospathius silvaticus Barbalho and Penteado-Dias, new species Female holotype. — Agrees with descrip- tion of petiolatiis except as follows: 27 an- tennomeres, apical 8 flagellomeres white; face striate; vertex only slightly striate; frons, temple and malar space smooth and shining; face width 2.2 times eye width; face height 1.7 times eye height; diameter of oral cavity slightly greater than malar space; no longitudinal carinae on propo- deum; fore tibia with row of 7 spines on anterior adge; fore and middle coxa weak- ly striate; ovipositor about equal to body length; body length 2.8 mm. Male. — Agrees with female except as follows: entire body light brown, propo- deum dark brown. Holotype female. — BRAZIL: Amazonas, Manaus, ZF3, Km23, Fazenda Esteio, No- vember 8, 1984, B. Klein col. Deposited in INPA. Paratypes. — BRAZIL: 1 male, Amazonas, Sao Gabriel da Cachoeira, April 29, 1982, J. A. Arias, col. Deposited in DCBU. Etymology. — The specific name is from the Latin silvaticus meaning belonging to woods or trees in reference to the collec- tion of this species in the jungle. Spathiospilus Marsh, new genus Type-species. — Spathiospilus brasilieiisis Marsh, new species. Diagnosis. — Cyclostome braconid, oral cavity circular, labrum concave; eyes large, malar space very small or absent; flagellomeres with double row of placodes separated by ridge around middle of fla- gellomere; mesonotum strongly declivous anteriorly; fore wing vein 2RS absent or weakly present apically near vein r, vein RS + M strongly arched, hind wing of male with stigma; for tibia with row of short stout spines along outer edge, hind coxa with small but distinct antero-ventral bas- al tubercle; metasoma petiolate, first ter- gum slender, parallel sided, rest of meta- soma suddenly widened, acrosternum nearly as long as tergum and fused with tergum. Distribution. — Sao Paulo State of Brazil. We have also seen one undescribed spe- cies from Panama. Comments. — This genus is similar to Het- erospathius in the Spathini by the absence of fore wing vein 2RS but distinguished by the strongly arched fore wing vein RS + M and the large eyes. In the key pro- vided by Marsh (1997) it will run to Het- erospnlus but is distinguished by the long 152 Journal of Hymenoptera Research and fused acrostemum of the first meta- somal segment. Etymolog}/. — The generic name refers to the similarity to Hetewspihts by the wing venation and to Spathius by the petiolate metasoma. Spathiospilus brasiliensis Marsh, new species (Figs. 6, 16-19) Female. — Body color: head, mesosoma and metasoma reddish-brown, metasomal terga 2-5 sometimes infused with black; scape and pedicle yellow, flagellum yel- low basally turning brown to apex; legs yellow; wings hyaline, veins yellow at base and apex, stigma and veins across middle of wing brown; ovipositor sheaths yellow, black at tip. Body size: 3-4 mm. Head (Fig. 17): 29-31 antennomeres, fla- gellomeres with two rows of placodes sep- arated by ridge around middle of each fla- gellomere; eyes large, covering most of head; malar space extremely short or ab- sent, lower margin of eye nearly touching base of mandible; face narrow, width less than length from clypeus to antennal sock- ets; oral cavity circular, diameter slightly greater than basal width of mandible; tem- ple very narrow, about K, eye width; ocell- ocular distance shorter than diameter of lateral ocellus; face, frons and vertex ru- gulose-coriaceous, temple coriaceous; oc- cipital carina scrobiculate along vertex and temple side. Mesosoma (Fig. 19): pronotum with scrobiculate grove across dorsal surface extending laterally on each side, bordered laterally by strong carina, dorsally coriaceous; mesonotum strongly declivous anteriorly, mesonotal lobes co- riaceous, notauli scrobiculate, median lobe with short and wide carinate-rugulose area before scutellum and with median raised line extending to pronotum; scutel- lum coriaceous, bordered laterally by ca- rina, scutellar sulcus deep with 5 cross ca- rinae; mesopleuron coriaceous, subalar area rugose, sternaulus scrobiculate; pro- podeum strongly areolate-rugose apically and laterally, with semicular coriaceous areas basolaterally bordered by distinct carinae. Wings (Fig. 6): fore wing vein r V2 length of 3RSa, vein RS-(-M strongly arched; hind wing vein RS absent, vein m- cu curved toward wing apex. Metasoma (Fig. 19): first tergum petiolate, parallel sided with apical and basal widths equal, longitudinally costate, weakly rugulose between costae; second tergum longitudi- nally costate, weakly rugulose between costae, triangular shaped with basal width abiut V2 apical width, weak line between second and third tergum; third tergum longitudinally costate on basal %, coira- ceous on apical Vs; remainder of terga co- riaceous; ovipositor about X; length of me- tasoma. Male. — Essentially as in female; 26 an- tennomeres; stigma present in hind wing. Holotype Female. — BRAZIL: Rio Mogi Gua^u, Luis Antonio, S.P., February 18, 1988, L. A. Joaquim collector. Deposited in DCBU. Paratypes. — BRAZIL: 1 female, 3 males, same data as holotype with additional dates of March 20-27, 1987; 1 female, Luis Antonio, S.P., Reserva Ecol. do Jatai, Feb- ruary 8, 1994, A. S. Scares collector; 1 fe- male, Faz. Canchim, Sao Carlos, S.P., June 20, 1985, A. S. Soares collector. Deposited in USNM, DCBU. Etymology. — The specific names is in ref- erence to the localities of the types series from Brazil. ACKNOWLEDGMENTS We thank the following for the loan of specimens for this study: Dra. Keti M. Rocha Zanol and Dr. Ren- ato Contin Marinoni of the Universidade Federal do Parana (UFPR), the Institute Nacional de Pesquisas da Amazonia (INPA), and the National Museum of Natural History, Washington, DC (USNM). Financial support was provided by the Conselho Nacional de Desenvolvimento Cientifico e Tecnologico (CNPq) and the Fundaijao de Amparo a Pesquisa do Estado de Sao Paulo (FAPESP). Permits to collect specimens in the area of the Canchim Farm were provided by Centro de Pesquisa de Pecuaria do Sudeste (EMBRA- PA, CPPSE, Sao Carlos, SP). Volume 8, Number 2, 1999 153 LITERATURE CITED Belokobyl'skij, S. A. 1992. On the classification and phylogeny of the braconid wasp subfamilies Do- ryctinae and Exothecinae (Hymenoptera, Bracon- idae). Part 1. On the classification, 1. Eiitoinolo- gkiwskof Obi-izrenie 71:900-928. In Russian. (En- glish version published in 1993, Entomological Rc- vieu' 77:109-137). Belokobyl'skij, S. A. 1994a. A review of parasitic wasps of the subfamilies Doryctinae and Exothe- cinae (Hymenoptera, Braconidae) of the Far East. Hymenoptera Insects of Siheria and Far East. No. 3, pp. 5-77. Belokobyl'skij, S. A. 1994b. A new tribe of the sub- family Doryctinae from Papua New Guinea (Hy- menoptera: Braconidae). Zoosi/stematica Rossica 3(1):141-145. Belokobyl'skij, S. A. 1995. Two new genera and two new subgenera of the subfamilies Exothecinae and Doryctinae from the Old World (Hvmenop- tera: Braconidae). Zoologische MededcUngen 69(3): 37-52. Haliday, A. H. 1836. Essav on parasitic Hymenoptera. Entomologists Magazine 4:38-59. Hedqvist, K.-J. 1963. Notes on Hormiinae with de- scriptions of new genera and species (Hym., Ich- neumonoidea, Braconidae). Entomologisk Tidskrift 84:30-61. Marsh, P. M. 1993. Descriptions of new Western Hemisphere genera of the subfamily Doryctinae (Hymenoptera: Braconidae). Contributions of the American Entomological Institute 28(l):l-58. Marsh, P. M. 1997. Subfamily Doryctinae, pp. 206- 233. In: R. A. Wharton, P. M. Marsh and M. J. Sharkey (eds.), Manual of the New World Gen- era of the Family Braconidae (Hymenoptera). Special Publication of the International Society ofHy- menopterists No. 1, 439 pp. Mason, W. R. M. 1986. Standard drawing conventions and definitions for venational and other features of wings of Hymenoptera. Proceedings of the En- tomological Society of Washington 88:1-7. Muesebeck, C. F. W. and L. M. Walkley. 1951. Family Braconidae (pp. 90-184). In: C. F. W. Muesebeck et al., Hymenoptera of America North of Mexico. Synoptic Catalog. U. S. Department of Agriculture, Agricultural monograph 2, 1420 pp. Shenefelt, R. D. and P. M. Marsh. 1976. Braconidae 9, Doryctinae. In: J. van der Vecht and R. D. She- nefelt, eds., Hymenopteroruni Catalogus (new edi- tion), part 13, pp. 1263-1424. Dr. W. Junk, The Hague. Wharton, R. A., P. M. Marsh and M. J. Sharkey. 1997. Manual of the New World genera of the family Braconidae (Hymenoptera). Special Publication of tlie International Society of Hymenopterists Number 1, 439 pp. J. HYM. RES. Vol. 8(2), 1999, pp. 154-164 New Host and Distribution Records for Leucospis (Hymenoptera: Leucospidae) Associated Primarily with Nests of Centris (Hymenoptera: Anthophoridae) in the Dry Forests of Costa Rica Miriam F. Cooperband, Robert A. Wharton, Gordon W. Frankie, and S. Bradleigh Vinson (MFC, RAW, SBV) Department of Entomology, Texas A&M University, College Station, TX 77843-2475; (GWF) Department of Environmental Science, Policy, and Management, University of California, Berkeley, CA 94720 Abstract. — A study conducted in the dry forests of Lomas Barbudal Biological Reserve, Guan- acaste Province, Costa Rica, revealed new host records and elevation data for five species of Leucospis: nztecn, hulbiventris, cnyennensis, egaia, and latifrons. Four species of trap nesting Centris bees (Hymenoptera: Anthophoridae) were attacked, as well as at least 4 species of trap nesting bees in the family Megachilidae. Of 295 Leucospis wasps reared from these trap nesting bees, it was possible to associate 236 with hosts, 189 of which were from nests of Centris bicornuta. Ele- vational data and host species data are summarized for each Leucospis species reared, as well as sex ratios of reared material. Additional notes on biology and behavior are provided, along with a key to the five species reared from trap-nesting Centris bees in Costa Rica. The taxonomic status of L. hulbiventris, a sexually dimorphic species, is discussed relative to L. manaica. Members of the family Leucospidae are among the largest species of Chalcidoidea, and all are parasitoids of solitary and sub- social aculeate bees and wasps. They drill through the hardened cell walls of their hosts' nests with their unusual oviposi- tors, and develop as ectoparasitoids. Ex- cellent summaries of leucospid biology are provided by Clausen (1940), Habu (1962), and Boucek (1974). The family was revised on a world basis by Boucek (1974), who also summarized the known host data. Hosts were recorded for 32 of the 130 species of Leucospidae recognized by Boucek (1974). Prior to Boucek's (1974) re- vision, no species of leucospids had been recorded as parasitoids of Centris bees, though other anthophorids (notably spe- cies of Xylocopa) were known as hosts of at least three species of Leucospis. Subse- 'MFC has relocated to the Department of Entomol- ogy, University of California, Riverside, CA 92521. quently. Chandler et al. (1985) reared two individuals of L. cayennensis Westwood from Centris in Minas Gerais, Brazil. Leu- cospids frequently parasitize megachilid bees in California (GWF, personal obser- vations), and there are several published records of species of Leucospis attacking various megachilids (Boucek 1974, Burgis 1995). Centris is a large genus of bees in the hymenopteran family Anthophoridae, with at least 32 species occurring in Costa Rica (Snelling, 1984). Their nesting habits are diverse. Some species make their nests in the ground while others utilize holes left in wood by other insects; some pro- vision multiple cells per nest, others pro- vision only one cell per nest. Aside from collecting pollen, all of them collect oil from plants that have oil producing flow- ers, and this oil is used in their nesting biology (Vinson et al. 1996). Centris bees, because of their large size. Volume 8, Number 2, 1999 155 are a readily observed component of the Mesoamerican dry forest habitat during the dry season. Trap-nests are very effec- tive means of studying the activity of wood-cavity nesting species (Frankie et nl. 1988, 1993, Vinson et al. 1996). In the area of this study, six species of Ceniris have been found to nest in tree holes, however only four of these are commonly found in the trap nests (Frankie et al. 1988). Centris biconiuta Mocsary is one of the most abun- dant of the trap nesting species in this area (Frankie et al. 1988, 1993, 1997). A number of inquilines and parasitoids can be reared from trap-nests, and in the Guanacaste Province of Costa Rica, leucospids were one of the dominant parasitoids of some of the species of Centris being studied there. Below we record the first specific host associations for Leucospidae on Cen- tris in Costa Rica. MATERIALS AND METHODS Stud}/ sites. — The study was conducted at the Lomas Barbudal Biological Reserve in the dry forest in Guanacaste Province, Costa Rica. Two bundles, each containing six block monitoring units (BMUs) (Fran- kie et al. 1993), were placed in dry forest sites at 100 m, 300 m, 600 m, and 800 m elevations in a transect extending from Hacienda Monteverde (at 100 m, 8 km NW Bagaces) northwards towards Volcan Rincon de la Vieja. All BMUs were hung at eye level on tree trunks in shaded lo- cations, and were monitored at 14 day in- tervals throughout the dry season (late December to May). The 100 m site had a mixture of oak forest and riparian ever- green forest (Frankie et al. 1988), and had been disturbed by agricultural develop- ment. The 300-800 m sites were largely in- tact oak forest with several other scattered tree species. The oak species, Quercus oleo- ides Schlecht. & Cham., was the same at all four sites. Trap nests. — The nesting activities of several species of Centris were monitored using BMUs. These BMUs consisted of 12 small wooden sticks (11.5 X 2.3 X 2 cm) with holes drilled in one end, bundled to- gether to form a block. The wooden sticks (consisting of pine or two local hard wood species) were drilled lengthwise to a depth of between 7.0 to 11.0 cm depend- ing on the hole sizes (diameters of 4.5, 6.5, 8, 9.5 and 11 mm were used). Two sticks representing each of the hole diameters, along with two additional sticks with a hole diameter of 8 mm, were bundled to- gether using fine wire or twine. The 8 mm hole size was doubled as it is the hole size most commonly used by Centris bees (Frankie et al. 1988, 1993). SHcks were lay- ered within the block so that a drilled end of a particular hole size always alternated with a non-drilled end, and small to large hole sizes descended from the top to bot- tom of a block. Emergence. — Each BMU was numbered and identified as to location, altitude, and time and date of placement. Every 14 days the sticks with completed bee nests were replaced with a new stick of the same hole diameter and additional information was recorded on the removed stick, including date removed. Sticks with completed nests were placed in large wire baskets (—20 x 20 X 40 cm) made of 2.5 cm open mesh hardware cloth that prevented access by mammals, but not parasitoids. These bas- kets were hung from nails on nearby trees. During the wet season, glass scintillation vials were taped to the nest entrances to intercept any emergences. Emergences were monitored on a two week basis and parasitoid and host associations recorded. Data presented here were collected during four consecutive years (December 1993 to December 1997). Behavior and development. — All observa- tions on female wasp behavior were made at 100 m sites where 2-4 BMUs were con- tinuously being monitored. Completed nests, 1 to 14 days old, were removed from BMUs, placed in baskets at these sites, and observed daily during daylight hours for 3 weeks. Each basket contained 156 Journal of Hymenoptera Research 12 0) ro o ro O X 5 Chalicodoma Unknown spp ^^g D L- azteca B L- bulbivenths ?ZL. cayennensis S/.. ega/a ■ L. latifrons sv\\v.\\\\\<.v?^?a ^^3 '//////A ^ 0% 25% 50% 75% 100% Percent of Leucospis species Fig. 1. Percent of each Leucospis species found on different hosts. Total for each Leucospis species across all hosts adds to 100%. approximately 100 sticks arranged in three rows of two sticks deep so that each stick had at least one surface exposed. The number of leucospids searching or at- tempting to oviposit on nests was record- ed until the experiment was terminated. One stick from each basket was removed daily during week 3 of the experiment, and opened to monitor bee development, and that of any present parasites, parasit- oids, and diseases. This experiment was replicated 4 times. Additional observations were made on 16 ovipositing female leucospids where, following completion of oviposition and departure of the leucospid, the exact dril- ling location was marked. The marked stick was then removed and dissected to determine the stage of the host attacked and the placement of the leucospid egg. Observations on oviposition behavior were also made on nests removed from BMUs and placed in baskets at the 100 m site. Specimen repositories. — Voucher speci- mens for the Ceritris and megachilid bees are at University of California, Berkeley, and those for the Leucospis species are at UC Berkeley and Texas A&M University (TAMU). Material for comparison, includ- ing primary types, was borrowed from the Philadelphia Academy of Natural Sciences (ANSP), the Natural History Museum, London (BMNH), and TAMU. RESULTS AND DISCUSSION Emergence data. — We reared 295 leucos- pids, representing five species, from at least seven species of wood-hole nesting bees in Guanacaste Province. Of 236 host bees, 189 were Centris bicornuta which was attacked by all five species of Leucospis: L. azteca Cresson, L. bulbiventris Cresson, L. cayennensis Westwood, L. egaia Walker, and L. latifrons Schletterer (Fig. 1). Centris nitida F. Smith and C. vittata Lepeletier each hosted four different species of Leu- cospis (Fig. 1). The most abundant leucos- pid in our samples was L. bulbiventris (56% of reared individuals) (Fig. 2). Discounting L. egaia, represented only by three reared individuals in our samples, all leucospids were more abundant at lower elevations (Fig. 3). Leucospis azteca and L. cayeniwnsis were most abundant at the 100 m site; L. bulbiventris and L. latifrons were most abundant at the 300 m site (Fig. 3). In ad- dition to C. bicornuta, C. nitida, and C. vit- tata, we also reared leucospids from C. an- alis F., undetermined species of Chalicodo- ma and Anthidium (Hymenoptera: Me- gachilidae), and at least two other undetermined species of megachilid bees, all from the trap nests described above. The experimental environment em- ployed in this study was artificial owing to the fact that the Ct'nfns-infested trap nests were highly accessible to leucospids by being clustered in baskets. The nests were thus at much higher densities than would be encountered in nature. In Costa Rican dry forests, Centris nests tend to be more widely dispersed, and the bees are not limited to nesting in holes in small sticks, which are maximally exposed for Leucospis oviposition. Therefore, the large number of leucospids recorded here is at Volume 8, Number 2, 1999 157 Frequency of Each Leucospis Species Reared azteca 21% bulbiventris 56% latifrons 11% cayennensis 11% Fig. 2. Relative proportions of Leucospis species reared from trap nesting bees. least in part a reflection of the experimen- tal manipulations. Observations on oviposition. — No female leucospids were observed around the wire baskets or on the sticks until nests were at least 3 weeks old (Fig. 4). Peak activity oc- curred towards the end of the third week, but since the experiment was terminated before all activity ceased, we lack data on how long nests remain attractive. Female leucospids fly up wind to the wire enclo- sures (100% N = 19), and ultimately land on one of the sticks containing bee nests. Generally, they walk the length of the stick slowly while alternately drumming the surface with their antennae. Prior to drilling, they stop to antennate the wood surface with both antennae held close to- gether, then move forward half a body length to drill the antennated spot with the ovipositor. Based on dissections of the 16 marked cells into which leucospids were ob- served ovipositing, female leucospids only attacked cells in which the mature larvae had at least begun to spin a co- coon. Leucospid eggs were always locat- ed inside the cocoons, either on the sur- face of a host larvae still finishing its co- coon (N = 1), a prepupa (N = 3), or a pupa (N = 12). Of the 18 other leucospid larvae or pupae recorded from randomly dissected nests, all were within the co- coon of a bee, indicating that either the bee is allowed to develop to a prepupa and spin a cocoon, or they are only par- asitized following cocoon formation. The failure to find leucospid larvae on youn- ger stages of bee larvae, and the delay in leucospid response to newly provisioned bee nests, suggests that these leucospids, at least, do not attack earlier stages of their hosts. Parasitized host larvae did not move, but it was not clear if they 158 Journal of Hymenoptera Research 100% 75% 50% 25% Proportions of Four Species of Leucospis Found at Different Elevations n L azteca UL. bulbiventris 0/.. cayennensis ■ L. latifrons _^^|_, 100 300 600 800 Elevation (m) Fig. 3. Proportions of four species of Leucofpiii found at different elevations. Total for each Leucospis species across all elevations adds to 100"Xi. Average Number of Leucospis Attacks on Bee Nests 7 6 5 4 - 3 - 2 1 0 3 day 1 day 2 day 3 WEEK 1 WEEK 2 1-4- day 4 WEEKS day 0 dJV '-■' -♦I day 1 WEEK 4 Fig. 4. Be at the begi of week 4 havioral observations. Number of Leucosfus attacks on provisioned sticks observed over time. Nests nning of week 1 were from 1 to 14 days old. Observations were not extended beyond tlie first day Volume 8, Number 2, 1999 159 were paralyzed since larvae at this stage of development are lethargic. There have been few prior studies in which more than one species of leucos- pid has been reared from a single host species. In addition to the five specific examples listed by Boucek (1974), RAW (unpublished) has collected both L. his- trio Maindron and L. niolei/rei Maindron from nests of a single species of Xi/locopm in Papua New Guinea. The data pre- sented here represent the first record, to our knowledge, of five leucospid species reared from a single host species. Fur- ther, we know of only one species of Leucospidae previously associated with Centris bees (Fritz and Genise 1980, De Santis 1983, Chandler et al. 1985). Parasitoid identifications. — Preliminary identifications of the Leucospis species were greatly facilitated by the excellent detail provided in the monograph by Boucek (1974). In our attempt to confirm the identifications of these species, how- ever, we uncovered several problems as- sociated with the primary types. The ho- lotypes of bulbiveutris and dubiosa Cres- son and lectotype of azteca should be in ANSP. They were all examined by Bou- cek, and the types returned as indicated by correspondence at ANSP. Despite considerable effort by D. Azuma, how- ever, no leucospid primary types could be found, though several others should also be at ANSP. The key presented here to the leucos- pids attacking Centris bees in dry forests of Costa Rica is adapted largely from this work, as is the terminology. The key has been greatly simplified for ease in identi- fication of leucospids attacking Centris in dry forests of Costa Rica, and should only be used in this context (or for comparison with Centris parasitoids from dry forests elsewhere). There are several other species of Leucospis known from Mesoamerica, and Boucek's work should therefore be consulted for any species not reared from Centris, and for rigorous confirmation of suspect individuals. KEY TO SPECIES OF LEUCOSPIS ASSOCIATED WITH CENTRIS BEES IN COSTA RICAN DRY FOREST HABITATS I- Pronotum with a transverse, premarginal cross carina and /or narrow yellow stripe near posterior margin. Mandible with triang^ular indentation on mesal chewing edge (Fig. 5) . . 2 - Pronotum without transverse, premarginal cross carina or narrow yellow stripe. Mandible with semicircular indentation on mesal, chewing edge (Fig. 6) . . . L. cayennensis Westwood 2. Scutellum at least partly yellow 3 - Scutellum completely without yellow coloration L. bulbiveutris Cresson 3. Hind femur with fewer than 10 small teeth on ventral margin. Hind coxa with a slender tooth (spine) on its dorsal posterior edge (as in Figs. 8, 9) 4 - Hind femur with more than 10 small teeth on ventral margin. Hind coxa with a thin, partially translucent lobe on dorsal-posterior edge (as in Fig. 7), never with a spine-like tooth L. egaia Walker 4. Setae on hind coxa converge towards center of depression (Fig. 8). Ovipositorial furrow on first metasomal tergum in the form of a simple, smooth, convex ridge down the midline (Fig. 11). Yellow coloration on scutellum restricted to posterior half, at least anterior half of scutellum black L. latifrons Schletterer - Setae on hind coxa nearly all pointing in the same direction, not converging towards the center (Fig. 9). Ovipositorial furrow on first metasomal tergum with a smooth, shiny, con- vex ridge down the midline, and concave slopes on either side of the ridge (Fig. 12). Nearly entire scutellum yellow or yellowish, with only anterior edge black L. azteca Cresson 160 Journal of Hymenoptera Research Figs. 5-9. Mandibles and hind legs of Leucospis species. 5. Mandible of L. latifrons, female. 6. Mandible of L. cayennensis, female. 7. Hind leg of L. bulbiventris, female. 8. Hind leg of L. latifrons, female. 9. Hind leg of L azteca, female. Leucospis azteca Cresson (Figs. 9, 12) Leucospis azteca, previously known only from the three specimens of the original type series, lacks obvious diagnostic fea- tures. It can be identified by the following combination of characters: mandible with triangular indentation (Fig. 5); pronotum with premarginal carina (a weak, trans- verse ridge within a yellow band that runs parallel and slightly anterior to the pos- terior margin of the pronotum), median lobe of metanotum (= dorsellum) coarsely sculptured, but without distinctly carinate lateral margins; hind tibia apically trun- cate; hind coxa regularly punctate and se- tose (as in Fig. 9); and ovipositorial furrow as in Fig. 12. Boucek (1974) noted that L. azteca shared several features in common with both L. latifrons and L. affinis Say, and that the species had been variously con- fused in previous studies. Leucospis latif- rons is readily identified by the pattern of dense pubescence on the hind coxa (com- pare Figs. 8 and 9), but one of the subspe- cies of L. affinis treated by Boucek, L. a. dubiosa Cresson, is particularly problem- atic as it is very similar to L. azteca. The apparent loss of the holotype of L. dubiosa and lectotype of L. azteca makes it even more difficult to separate L. affinis from L. azteca. Fortunately, the two other members of the original type series of L. azteca were located in ANSP. Structurally, both of Volume 8, Number 2, 1999 161 Figs. 10-12. Ovipositorial furrows on the first nietasomal terga of Leucospis species. 10. L. I'ulbhvntiis, female. 11. L, latifrons, female. 12. L. azteca, female. these agree very closely with our material, but there are some differences in color. This is particularly noticeable on the scu- tellum. In our material, the yellow color- ation covers at least the posterior half of the scutellum, providing a readily observ- able field characteristic for separating this species from the otherwise similar L. latif- rons. In both of the L. azteca paratypes, however, the scutellum is much less ex- tensively yellow, and similar to our latif- rons in this regard. Given this variation in color, we concur with Boucek (1974) that the differences in the ovipositorial furrow on the first metasomal tergum are impor- tant for distinguishing L. affinis from L. az- teca, and it is on this basis that we have determined our material as L. azteca. In L. azteca, the furrow is generally not as deep as in affinis (a feature that is difficult to assess without side by side comparison), and scattered setae occur on the polished median ridge. In the specimens of L. affinis available to us from California and south- em Texas, the median, polished ridge lacks setae (though these are abundant along the edge of the furrow). We have found that density of punctation on the hind leg (a feature used by Boucek 1974) is too variable to be used for separation of L. affinis, L. azteca, and L. latifrons unless side by side comparison is possible with a good series of specimens representing all three species (which we were fortunate to have at our disposal). The biology of this species was previ- ously unknown. Our data suggest that it is a generalist, capable of attacking several different species. Of the 58 individuals for which we had host data, 41 came from C. bicornuta, 1 from C. nitida, 3 from C. vittata, 4 from Anthidium, 5 from Chalicodoma, and 4 from other undetermined species of Me- gachilidae. Unlike L. latifrons, L. azteca was confined to lower elevations. Of the 53 specimens for which we had altitudinal data, 98%, were reared from nests at 100 m, and only 2% from 300 m (Fig. 3). Of the 63 individuals we reared, only 17% were male, showing a strong female bias (Fig. 13). Leucospis affinis has been reared from a wide variety of megachilid bees, and could conceivably be found on the same hosts as L. azteca where their ranges overlap (e.g. in Mexico). As these species 162 Journal of Hymenoptera Research are very difficult to separate, caution must be exercised when identifying them. Leucospis bulbwentris Cresson (Figs. 7, 10) Leucospis bulbiveiitris is readily identifi- able by the complete absence of yellow coloration on the scutellum, the shape of the ovipositorial furrow, which is excep- tionally broad anteriorly (Fig. 10), the deep, triangular incision of the mandible (much deeper than in Fig. 5), and the shape of the hind coxa, which has a thin, partially translucent lobe on its dorsal posterior edge (Fig. 7). It is a large but slender species with a distinctly petiolate abdomen. Leucospis bulbiveiitris was previously known only from the male holotype col- lected in Mexico. A second nominal spe- cies, L. maimica Roman, described from Brazil, has heretofore been known only from five females that are similar in many respects to the holotype of L. bulbiventris (Boucek 1974). We reared a good series of males and females, in several cases from the same host nest, enabling us to associ- ate the sexes with certainty. Boucek (1974) was the first to suggest that perhaps L. bul- biveiitris and L. maimica represent different sexes of the same species. We confirm that the differences in setation and overall shape between L. maiiaica and L. bulbiveii- tris noted by Boucek represent sexual di- morphism. Based on our rearings, we therefore strongly suspect that L. iiiaiiaica and L. bulbiventris are the same; but be- cause we do not have males from South America, and, more importantly, cannot locate the holotype of L. bulbiventris, we must unfortunately leave this problem un- resolved. The name bulbiventris has prior- ity over manaica, and since males from our material fit the description of L. bulbiven- tris provided by Boucek (1974), we have therefore used this name for our species. No biological information has previous- ly been published for either L. bulbiventris or L. manaica, nor have either of these been recorded before from Costa Rica. Of the 116 individuals for which we have host data, the majority (108) came from nests of C. bicornuta. The others were reared from C. aiialis (1), C. nitida (1), and C. vit- tata (6). Of 150 reared individuals for which we have elevation data, 21% were from 100 m, 73% from 300 m, and only 7% from nests at 600 m. None were found at 800 m (Fig. 3). Of 165 individuals reared, only 19% were males, showing a strong female bias (Fig. 13). Leucospis cayennensis Westwood (Fig. 6) This is a widespread Neotropical spe- cies recorded from Mexico to Argentina as well as the Caribbean (Fidalgo 1980, De Santis 1983). It is readily identified by the semicircular indentation of the mandible (Fig. 6), the complete lack of a transverse premarginal carina on the pronotum, and the relatively smooth hind coxa (with dor- sal two-thirds of the depression smooth, shiny, bare and impunctate). Fritz and Genise (1980) were the first to record Centris tarsata Smith as a host of L. cayennensis, and this is the only specific host recorded to date (Chandler et al. 1985, De Santis 1983). In Fritz and Genise's (1980) study, L. cayennensis was reared from 14% of the C. tarsata cells in old, abandoned Sceliphron asiaticum (L.) nests. Our data show that L. cayennensis attacks at least four other species of Centris bees. Of the 30 individuals for which we have host data, 22 came from nests of C. bicor- nuta, 4 from C. nitida, 3 from C. analis, and 1 from C. vittata. Leucospis cayennensis was found at all four elevations sampled dur- ing this study, but seemed to have a pref- erence for lower elevations. Approximate- ly 61% of the individuals for which we have elevational data were reared from nests at 100 m, 21% from 300 m, 14% from 600 m, and 4% from 800 m (Fig. 3). Of 33 individuals collected, one third were males, showing a female sex bias (Fig. 13). Volume 8, Number 2, 1999 163 100% 75% 50% 25% - 0% Sex Ratios for Reared Leucospis Species (N) 83% 81% Dmale 90% D female 17% 19% 67% 33% 10% azteca (63) bulbiventris (165) cayennensis (33) Fig. 13. Sex ratios of Lcucospiis species reared from trap-nesting hosts. latifrons (31) Leucospis egaia Walker This species is similar in general ap- pearance and coloration to L. aztecn and L. latifrons, but has a carinately margined dorsellum, a more rounded, translucent lobe rather than a spinose tooth on the hind coxa, and an oblique rather than truncate margin to the hind tibia. No host records existed previously for L. egaia (Boucek 1974). We reared a male from C. biconinta at 100 m, a male from a megachi- lid bee at 600 m, and a female from a me- gachilid bee at 600 m. Since only 3 indi- viduals of this species were found, we cannot make general inferences about its biology. Leucospis latifrons Schletterer (Figs. 5, 8, 11) This is another widespread Neotropical species, occurring from Mexico to Bolivia. It is readily identified by the arrangement of the dense patch of setae in the central depression of the hind coxa (Fig. 8). The color pattern on the scutellum was suffi- ciently stable in our material to use for separation of L. azteca from L. latifrons in the field. As noted above, however, these color patterns may vary from one locality to the next, and should be used cautiously for identification purposes. Hosts were previously unknown for L. latifrons. Of the 29 individuals for which we have rearing data, 17 came from C. hicornitta, 6 from megachilid bees, 4 from C. nitida, I from C. vittata, and 1 from Chalicodoma. Of the 32 individuals with altitudinal data, 6% came from nests at 100 m, 52% came from nests at 300 m, 19% from nests at 600 m, and 23% from nests at 800 m (Fig. 3). This species seems to be more of a generalist, able to adapt to a variety of hosts and el- evations, with an apparent preference for habitats at 300 m. Although L. latifrojis and L. azteca are extremely similar morpholog- ically, the elevation data suggest a biolog- ical difference supporting Boucek's (1974) finding that they are two distinct species. Of 31 individuals reared, only 10% were male, showing a strong female bias, as in all the other species in our samples (Fig. 13). 164 Journal of Hymenoptera Research ACKNOWLEDGMENTS We are most grateful to D. Azuma, Z. Boucek, and E. Grissell for their assistance in our attempts to lo- cate the type material. The Friends of Lomas Barbu- dal offered logistic support for this research. The Cal- ifornia Agricultural Experiment Station provided fi- nancial support for most of the field work. We also thank J. Oswald who offered the use of his scanner and computer during the final phase of preparing the illustrations, and two reviewers (J. Noyes and J. LaSalle) for suggesting improvements in the text. LITERATURE CITED Boucek, Z. 1974. A revision of the Leucospidae (Hy- menoptera: Chalcidoidea) of the world. BiiUctm of the British Miisctini (Natural History) Entomolo- gy, Supplement 23:1-241. Burgis, H. 1995. Leucospis gigas (Chalcidoidea: Leu- cospidae) as parasite of the mortal bee Megachilc sicula (Apoidea: Megachilidae). Bembix 5:27-32. Chandler, L., J. A. F. Barrigossi, and E. B. S. Diaz. 1985. The first definitive host record for Leucospis cai/ennensis Westwood (Hymenoptera: Leucospi- dae). Revistn Ceres 32:170-174. Clausen, C. P. 1940. Entomophagous Insects. McGraw- Hill, New York. 688 pp. De Santis, L. 1981. Catalogo de los Himenopteros Cal- cidoideos de America al Sur de los Estados Un- idos — Primer Suplemento. Revista Peruana ilc En- tomologia 24:1-38. Fidalgo, A. P. 1980. Nuevas citas de calcidoideos para Argentina, Bolivia y Peru (Hymenoptera). Neo- tropica 26:193-196. Frankie, G. W., L. E. Newstrom, and S. B. Vinson. 1993. Nesting-habitat preferences of selected Cen- tris bee species in Costa Rican dry forest. Biotro- pica 25:322-333. Frankie, G. W., S. B. Vinson, L. E. Newstrom, and J. F. Barthell. 1988. Nest site and habitat preferenc- es of Centris bees in the Costa Rican dry forest. Biotropica 20:301-310. Frankie, G. W., S. V. Vinson, M. A. Rizzardi, T. L. Griswold, 5. O'Keefe, and R. R. Snelling. 1997. Diversity and abundance of bees visiting a mass flowering tree species in disturbed seasonal dry forest, Costa Rica, journal of the Kansas Entomo- logical Societ}/ 70:281-296. Fritz, M. A. and J. A. Genise. 1980. Notas sobre nido de barro de Sphecidae (Hymenoptera) construc- tores, inquilinos, parasitoides, cleptoparasitos y detritivoros. Revista de la Sociedad Entomohigica Argentina 39:67-81. Habu, A. 1962. Chalcididae and Leucospidae and Po- dagrionidae (Insecta: Hymenoptera). Fauna ]a- ponica 1-232. Snelling, R. R. 1984. Studies on the taxonomy and distribution of American centridine bees (Hy- menoptera: Anfhophoridae). Contributions in Sci- ence (Los Angeles). 347:1-69. Vinson, S. B., G. W. Frankie, and H. J. Williams. 1996. Chemical ecology of bees of the genus Ccutris (Hymenoptera: Apidae). Florida Entonu^logist 79: 109-129. ]. HYM. RES. Vol. 8(2), 1999, pp. 165-196 The Taxonomy of Recent and Fossil Honey Bees (Hymenoptera: Apidae; Apis) Michael S. Engel Department of Entomology, American Museum of Natural History, Central Park West at 79th Street, New York, New York 10024-5192, USA Abstract. — An attempt is made to clarify the complicated and error fraught taxonomic history of the honey bees (genus Apis Linnaeus) by cataloguing the numerous names proposed for Apis. The 178 species- and 10 genus-group names (including those proposed herein) are brought for the first time into accord with the modern classification of the honey bees. The results of this synthesis reveal a number of names in current usage to be taxonomically confused. The authorship of the species Apis koschevnikovi is established as Enderlein and not Buttel-Reepen. The correct names for four subspecies of the Western honey bee, A. mtilifera, are established as A. mellifcra remipes Gerstacker (not A. mellifera anuenincn Skorikov), A. mellifera sicilianci Grassi (not A. mellifera Simla Montagano), A. mellifera jemenitica Ruttner (not A. mellifera yemenitica Ruttner), and A. mel- lifera Caucasia Pollmann (not A. mellifera caiicasica Gorbachev). The correct authorship of the sub- species A. mellifcra iberica is established as Ruttner and not Goetze, of A. mellifera syriaca as Sko- rikov and not Buttel-Reepen, and of A. mellifera intermissa as Maa and not Buttel-Reepen. Three replacement names are proposed for preoccupied subspecific taxa: A. mellifera artemisia Engel (new name for A. mellifera acervorum Skorikov, nee Linnaeus), A. mellifera iberiensis Engel (new name for the aforementioned A. mellifera iberica Ruttner, nee Skorikov), and A. mellifera sossimai Engel (new name for A. cerifera Gerstacker, nee Scopoli). The previously confused Himalayan honey bee (infrequently associated with A. cerana skorikovi Maa, nomen nudum or A. cerana "Him- alaya race") is proposed and validated as A. cerana skorikovi Engel (new subspecies). Similarly the large black race of China (Aba race) is proposed as A. cerana heimifeng Engel (new subspecies). A lectotype and 11 paralectotypes are designated for A. mellifera ruttneri (by Sheppard et «/.), a lectotype is designated for A. koschevnikovi, and a neotype is designated for A. mellifera iberica Ruttner. Apis cuenoti Theobald is newly synonymized under A. henshaun Cockerell, A. mellifera rodopica Petrov is synonymized with A. mellifera macedonica Ruttner, while Hauffapis scheeri Arm- bruster, H. scharmanni Armbruster, H. scheeri gallauni Armbruster, H. scheeri rahdei Armbruster, H. scheuthlei seemanni Armbruster, and H. scheuthlei zeuneri Armbruster are all synonymized under A. armbrusteri Zeuner (new synonymies). The name A. vetustus Engel for a fossil from the Oli- gocene of Europe is emended to A. vetiista Engel. All genus-group taxa are characterized on the basis of adult and larval morphology and ethology. Two new subgenera are proposed to accom- modate fossil honey bee species: Cascapis Engel new subgenus (type: A. armbrusteri Zeuner) and Priorapis Engel new subgenus (type: A. vetusta Engel). The honey bees (genus Apis Linnaeus) pict early humans collecting honey combs, are the most famous of all insects owing Similarly, the honey bees figure promi- to their import for the pollination of crops, nently in human cultures, mythologies, their social organization, and the honey and religions (e.g., see Ransome 1937). they produce. There has been a long as- Numerous world revisions have been at- sociation between Apis and man as is at- tempted for the genus (Gerstacker 1862, tested by the Paleolithic Arafia Cave 1863, Smith 1865, Ashmead 1904, Buttel- paintings near Valencia, Spain which de- Reepen 1906, Enderlein 1906, Skorikov 166 Journal of Hymenoptera Research Table 1. Classificatory structure of major honey bee revisions since Gerstacker (1862); living species only. Subspecies and varieties omitted so as to save space. Species are in alphabetical order and do not necessarily correspond in rows from one column to the next. 1 have replaced uses of wcWficn by some of these authors with mellifera. Although Maa (1953) observes that Skorikov (1929b) recognized 14 species, the later author left one unnamed (the "Egyptian bee") and 1 have therefore listed only 13 here. G = genus, sg = subgenus. Gerstacker 1862 Smith 1865 Ashmt'.id 1')II4 Butlcl-Reepen Endt'rlfin SkonkiH' W24b \Ki,i I'resent author G. Aph G. Apis G. Apis G. Apis G. Apis G. Apis G. Apis G. Apis dorsata adansoiiii cerana dorsata dorsata sg. Apis sg. Apis sg. Apis florea dorsata iitdica florea florea adansoiiii adansonii cerana indica florea mellifera mellifera indica cerana intennissa koschevnikovi mellifera indiea nigritarum mellifera cypria lamarckii mellifera mellifera nigrocincta indica meda nigrocincta nigrocincta unicolor japonica mellifera sg. Megapis sinensis G. Megapis johni remipes dorsata zonata dorsata meda unicolor sg. Micrapis zonata mellifera sg. Sigtnatapis andreniformis G. Micrapis remipes cerana florea florea syriaca indica 1871 unicolor javana added: sg. Megapis johni laboriosa dorsata sg. Micrapis florea koschcviiikovi licftiiicki nigrocincta peroni philippina samarensis vecliti G. Megapis bingliami breviligula dorsata laboriosa G. Micrapis andreniformis florea 1929b, Maa 1953) as well as an unpub- lished faunal revision from Thailand (Ma- laipan 1972). Since the first detailed study by Gerstacker (1862), anywhere from 3 to 24 species have been recognized while the four subgenera have at times been treated as distinct genera (Table 1). Despite the consistent and concentrated effort on Apis taxonomy for well over a century, a clear picture of the species and the numerous names associated with each has yet to be achieved. It is hoped that this paper (al- though mostly an exercise in "bookkeep- ing") will make a small step towards achieving these goals. The primary objective of this paper has been to pull together the extensive litera- ture pertaining to the taxonomy of honey bees and to associate the long lists of ju- nior synonyms with their valid counter- parts. The last catalogue of Apis names was undertaken by Maa (1953); however, his list of names contains a number of er- rors and his rather extreme classification is not easily reconcilable with current us- age. Moreover, Maa did not treat the nu- merous fossil honey bees and at least 60 names have either been discovered or were proposed since his study (approxi- mately one-third of the total number of names). Below I have compiled a listing of all names applied to honey bees, both liv- Volume 8, Number 2, 1999 167 ing and extinct. This compilation contains 178 species- and 10 genus-group names. Although Maa (op. cit.) and Ruttner (1988) noted that over 600 species-group names have been proposed for Apis, this does not mean that 600 names have been proposed for honey bees. Many of these names were proposed under a Linnean concept of the genus which included all bees. Thus, most of the 600-1- names pertain to species now placed in other genera and families (e.g., Anthopliora, Bonihiis, Megachile, Trigoiin, &c.). Interestingly most of the names ap- plicable to honey bees have been pro- posed during this century. A breakdown of the list reveals eight names appearing between 1758 and 1799, 17 from 1800 through 1849, 26 between 1850 and 1899, 66 in the period 1900 to 1949, and 61 names between the years 1950 and 1999. Of these names 146 are applicable to the living species (90 associated with Apis itiel- lifera alone!), 25 for the fossil species, and seven names are of dubious taxonomic status. Included with this catalogue I have pro- vided descriptions of all genus-group taxa now included in Apis, thereby adding to this work a subgeneric revision of the ge- nus. Table 2 summarizes the hierarchical classification of Apis as it is proposed herein. Phylogenetic studies on the genus have recently been undertaken by Alexander (1991 a,b), Engel (1998a), and Engel and Schultz (1997: see also works cited there- in). The current hypothesis of relation- ships among the subgenera and species is depicted in Figure 1 (modified from Engel 1998a). General works on honey bee anat- omy, biology, behavioral ecology, and di- versity have been presented by Snodgrass (1956), Winston (1987), Seeley (1985, 1995), and Smith (1991a) respectively. The distri- bution of the various species and subspe- cies has been thoroughly treated by Rutt- ner (1988, 1992), Otis (1996), and Hepburn and Radloff (1998). The position of the Ap- ini among the other tribes of corbiculate Table 2. Current hierarchical classification of the honey bees; tribe Apini Latreille (excluding infraspe- cific taxa). Daggers (t) indicate extinct taxa. Genus APIS Linnaeus subgenus Apiii Linnaeus cerana Fabricius koschevnikovi Enderlein meltifera Linnaeus nigrocincta Smith t subgenus Cnscapiis Engel t armhrusteri Zeuner subgenus Megnpis Ashmoad dorsata Fabricius subgenus Micrapns Ashmoad andreniformis Smith florea Fabricius + subgenus Priorapis Engel t vetusta Engel t subgenus Symipis Cockerell t ht'nslui-d'i Cockerell t longtibia Zhang t mioccnkci Hong + pctrcfactn (Rilia) bees has been considered most recently by Chavarria and Carpenter (1994), Engel (1998b, unpubl. data), Noll (1998, in prep.), and Schultz et al. (1999). FORMAT Standard formats for taxonomic histo- ries are used. Comments provided for each taxon give information on the assign- ment of authorship to certain names, sub- specific classifications, studies of the rec- ognition of difficult taxa, and occasionally historical information on the biogeogra- phy or biology of the species. Several hon- ey bee names were used in publications as iiomina iiudn and in those cases where no description was later provided they are in- dicated as such. However, whenever a name was later made available by an as- sociated description I have only listed the publication making the name available under the rules of the I.C.Z.N. (1985) and not the original paper in which it ap- peared as a nomen nudum. For those species in which subspecies are currently recognized (A. cerana, A. dor- 168 Journal of Hymenoptera Research £ vetusta henshawi petrefacta longtibia miocenica armbrusteri florea andreniformis dorsata - mellifera nigrocincta cerana koschevnikovi I Priorapis Synapis Cascapis Micrapis Megapis Apis Fig. 1. Phylogeny of the honey bees (genus A\m Linnaeus) with subgenera indicated (modified from Engel 1998a). sata, and A. ineUifera) I have associated each junior synonym with its correspond- ing subspecies. Names in brackets at the end of each entry indicate which subspe- cies the name corresponds to. In some cas- es (e.g., nomina iiiida) this decision cannot always be made with certainty and so are accordingly noted with an interrogative mark. Descriptions are provided for genus- group taxa and are based on information from all adult castes, the mature larva (of workers), and ethological information. Too little information exists at this time on Apis pupae to know whether additional character data can be gleaned from this life-stage. Characters given for workers also apply to drones and queens (except when indicated), although characters giv- en for drones and queens do not necessar- ily occur in the worker caste. The abbre- viations Tl, T2, . . . , T8 are used for the corresponding metasomal terga (S for ster- num). Since the literature on honey bees is vast, some names may have been inadver- tently overlooked. Researchers discover- ing names (nomina niida, nova, &c.) omit- ted here or of earlier usage than the dates cited are urged to contact me and, if pos- sible, to send a copy of the relevant pa- per(s). All in all, however, this list should prove a reasonable starting point for fu- ture research. RECENT HONEY BEES Genus Apis Linnaeus Diagnosis. — Modified and updated from Michener (1990: as Apinae). WORKER: Labral apex gently and often weakly con- cave; labrum three to four times broader than long (median width). Malar space as long as, or longer than basal width of mandible. Mandible without dentition Volume 8, Number 2, 1999 169 Figs. 2-3. Honey bee structural characteristics rep- resented by Apis {Apis) cerana Fabricius. 2, Worker head (scale bar = 500 M-m). 3, Inner surface of meta- tibia-metabasitarsus junction showing pollen press (auricle), rastellum, and absence of tibial spurs (scale bar = 200 (jim). (Fig. 2) (not so for queens or males). Hairs of compound eye long and erect (Fig. 2). Vertex extremely short, much less than ocellar diameter. Scutellum strongly con- vex and bulging, obscuring metanotum and basal area of propodeum. Basal area of propodeum extremely short and decli- vious. Mesocoxae nearly meeting medial- ly. Strigilis with prong on anterior margin. Corbicula, rastellum, and auricle present (workers only). Metatibial spurs absent (Fig. 3); metatibia without penicillum; in- ner surface of metabasitarsus with setal comb rows (worker only: referred to as brush combs in Michener 1990). Claws cleft; arolia present. Distal wing venation strong and complete; Ir-m and 2r-m strongly an- gled respective to M (Fig. 3). Marginal cell long (Fig. 3) and bluntly rounded at apex, not gently tapering over its length. Jugal lobe present. Sting straight (worker only). DRONE: Mandible usually bidentate, fre- quently only weakly. Holoptic (compound eyes meeting at top of head). S8 reduced to transverse bar, without spiculum. Spa- tha and volsella absent. QUEEN: Mandi- ble bidentate. Inner surface of metabasi- tarsus unmodified. Sting curved. MA- TURE LARVA: Without small dorsal tu- bercles on segments 1^. Mandible weakly sclerotized, bluntly pointed, without con- cavity on inner surface. ETHOLOGY: Nest a vertical comb of cells made of wax. Dance language communication system present. Highly eusocial (with morpholog- ically distinct castes). New colonies found- ed by fission and include the old queen. Brood and storage cells similar. Imma- tures progressively provisioned. Biogeography. — Honey bees are predom- inantly a tropical group and arose in the Indo-Malayan region during the early Ol- igocene (Engel 1998a). The genus is native to Europe, Africa, Madagascar, Arabia, the Near East, as well as Central and Southern Asia inclusive of most Southeast Asian Is- lands. Honey bees are not native to the Australian or American continents having been moved to these regions by humans. Species radiated from southern Asia per- haps being limited only by temperature extremes. The clade of living species (sub- genera Apis, Megapis, and Micrapis: Fig. 1) arose sometime in the latter half of the Miocene (Engel 1998a). The development of cavity-nesting behavior in the earliest Pliocene (perhaps in the latest Miocene) enabled at least two of the species, A. cer- ana and A. melUfera, to extend their ranges 170 Journal of Hymenoptera Research into more temperate areas. In the case of A. mellifera this was into Europe and northwestern Asia while for A. cerana this was into northeastern China, and the east- ern regions of the former Soviet Union. The fossil species were presumably open- nesting (refer to Engel 1998a for a cladistic reconstruction of this behavior) and the slightly warmer temperatures in the Mio- cene would have partly allowed species such as A. armbusteri to exist in Europe even in the absence of well developed thermoregulatory capabilities. As the tem- peratures continued to cool (e.g., by the Pliocene) open-nesting species would have been forced from these areas and the cavity-nesting species would be free to colonize. Subgenus Apis Linnaeus Apis Linnaeus 1758: 343, 574. Type species: Apis mcUifica Linnaeus 1761 (= Apis mellifera Lin- naeus 1758), designation of Latreille (1810). Michener 1944: 292. Maa 1953: 557. Michener 1990; 140. Apicnla Rafinesque 1814: 29. Unjustified re- placement name for Apis Linnaeus 1758. Apiarus Rafinesque 1815: 123. Unjustified re- placement name for Apis Linnaeus 1758. Apis (Siginntapis) Maa 1953: 556. Type species: Apis cerana Fabricius 1793, original designa- tion. Diagnosis. — WORKER: Forewing length 7-10 mm. Basal vein frequently gently curved, strongly distad cu-a (Fig. 4). An- gle of posteroapical margin of first sub- marginal cell less than 45°. Distal abscissa of vein M in hind wing variable (present [Fig. 5] or absent). DRONE: Hind basitar- sus without thumb-like process. Vertical arm of T8 longer than horizontal arm; S7 and SB fused mesally. Gonobase absent. Ventral gonocoxa membranous; dorsal gonocoxa reduced. Ventral cornua of en- dophallus recurved ventrally. MATURE LARVA: Labrum with peg-like setae re- stricted to apical surfaces and tubercles. Galea smaller than maxillary palpus. La- bial palpus not spiculate. Epipharynx with or without setae. Atrial inner walls smooth. ETHOLOGY: Nest constructed in a cavity. Dance language performed on vertical surface; wagging metasoma posi- tioned parallel to dance surface; recruits positioned next to dancer's metasoma (within a near field sound range). Drone cell cap variable (present or absent). Apis (Apis) cerana Fabricius The Eastern or Asian Honey Bee Apis cerana Fabricius 1793: 327. [cerana Fabri- cius] Apis indica Fabricius 1798: 274. [indica Fabricius] Apis socialis Latreille 1804a: 390. [indica Fabri- cius] Apis peroni Latreille 1804b: 173. [? indica Fabri- cius: see treatment of javana subspecies] Apis gronovii Guillou 1841: 323. [? indica Fabri- cius: see treatment of javana subspecies] Apis perrolletii Guerin-Meneville 1844: 461. [indica Fabricius] Apis delessertii Guerin-Meneville 1844: 461. [indica Fabricius] Aprs sinensis Smith 1865: 380. [cerana Fabricius] Apis mellifica variety japonica Radoszkowski 1887: 436. [japonica Radoszkowski] Apis delesserti Buttel-Reepen 1906: 168. Unjusti- fied emendation, [indica Fabricius] Apis indica variety javana Enderlein 1906: 337. [javana Enderlein] Apis indrca Baldensperger 1928: 173. Lapsus cal- ami, [indica Fabricius] Apis johni Skorikov 1929b: 251. [johni Skorikov] Apis indica philippina Skorikov 1929b: 252. [indica Fabricius] Apis indica skorikovi Maa 1944: 4. Nomen nudum, [skorikovi Engel] Apis mellifera gandhiann Muttoo 1951: 153. No- men nudum, [indica Fabricius] Apis {Sigmatnpis) lieftincki Maa 1953: 572. [johni Skorikov] Apis (Sigmatapis) samarensis Maa 1953: 580. [indica Fabricius] Apis indica sinensis ussuriensis Goetze 1964: 26. Nomen nudum. Refer to comments under A. mellifera. [cerana Fabricius] Apis cerana himalaya Smith 1991b: 154. Nomen nudum (see below), [skorikovi Engel] Apis ccrcna Willis, Winston, and Honda 1992: 169. Lapsus calami, [cerana Fabricius] Volume 8, Number 2, 1999 171 cu-a Figs. 4-6. Honey bee wing venation. 4, Apii (/l;'/s) cerana Fabricius, worker, forewing. 5, A. {A.) ceraiia, worker, hind wing. 6, A. {Synapns) hcnshawi Cockerel], worker, forewing reconstruction. In the forewing M indicates an abscissa of vein M that is called the basal vein, while in the hind wing a distal abscissa is indicated which is referred to as the indica vein by some authors; cu-a is the cubital-anal crossvein, also called cu-v by some authors. Apis nulucnsis Tingek, Koeniger, and Koeniger 1996 [1997]: 116. [mduensis Tingek et ai] Apis ccrnim hcimifcng Engel 1999: see below. [heimifeng Engel] Apis cerana skorikovi Engel 1999: see below, [sko- rikovi Engel] Comments. — The name gmnihiaim Mut- too (1951) was also used by Muttoo (1956) for an invalid variety of A. cerana. The name uuluensis was synonymized by my- self in Schultz et al. (1999) and therefore appears here for the first time as a sub- specific entity of A. cerana (see treatment of subspecies below). Distribution. — Refer to treatment of in- dividual subspecies below. Apis (Apis) koschevnikovi Enderlein, corrected authorship The Sundaland Honey Bee Apis tncllifica imiica variety koschcv)\ikovi Buttel- Reepen 1906: 192. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, and 50c). 172 Journal of Hymenoitera Research Apis indica variety koscheimikovi Enderlein 1906: 335. First available usage. Apis (Sigmatnpis) vechti Maa 1953: 572. Apis (Sigmatapis) vechti linda Maa 1953: 574. Apis melUfica adansonii koschevnikowi Goetze 1964: 25. Unjustified emendation. Comments. — The authorship of this name has almost universally been given to Buttel-Reepen who first proposed it in 1906. However, Buttel-Reepen used the name infrasubspecifically and thus ac- cording to the I.C.Z.N. (1985: Art. 10c) does not become available until it is first used as a species or subspecies, and then the authorship is assigned to the author who used the name in such a sense (Arts. 23j and 50c). Enderlein (1906) was the first to use the name koscheimikovi subspecifi- cally and he must therefore be considered as the author. This species was only recently reinstat- ed although under the junior synonym A. vechti (Koeniger et al. 1988, Tingek et al. 1988). Ruttner et al. (1989) recognized that the correct name for this species was A. koschevnikovi. There are presently no sub- specific forms recognized. Distribution. — Apis koschevnikovi occurs on the Malay Peninsula, Borneo, Brunei, Java, Sabah, Sarawak, and Sumatra. Lectotype.—V^ orkev; BORNEO: Kinaba- lugebirge, John Worterstradt, L. [leg.]; now with an additional label reading, "Lectotype, Apis koschevnikovi Enderlein, desig. M. S. Engel" (deposited in the Mu- seum fiir Naturkunde, Berlin). The origi- nal series of A. koschevnikovi used by But- tel-Reepen is located in the Museum fiir Naturkunde, Berlin. The type series con- sists of one worker from northern Borneo and eight labeled as from Cameroon. No cerana-MVe bees are natively known from the African continent, yet the specimens from "Cameroon" are clearly conspecific with the one from Borneo, and all corre- spond to the species widely recognized as A. koschevnikovi today. The specimens from Camaroon are very likely misla- beled; something even Buttel-Reepen (1906) suspected. Although Maa (1953) chose Cameroon as the type locality, this cannot be considered a type fixation since there are eight specimens from this local- ity and none are labeled as the type (in fact, Maa admits having never seen any specimens of A. koschevnikovi). I therefore have selected the single specimen from Borneo to serve as the lectotype. Apis (Apis) mellifera Linnaeus The Western Honey Bee Apis mellifera Linnaeus 1758: 576. [mellifera Lin- naeus] Apis mellifica Linnaeus 1761: 421. Unjustified emendation, [mellifera Linnaeus] ^;);s gregaria Geoffroy 1762: 407. [mellifera Lin- naeus] Apis cerifera Scopoli 1770: 16. [mellifera Linnae- us] Apis imicolor Latreille 1804b: 168. [iiuicolor La- treille] Apis fasciata Latreille 1804b: 171. Preoccupied {nee Linnaeus 1767, Scopoli 1770). [lamarckii Cockerell] Apis adansonii Latreille 1804b: 172. [adansonii La- treille] Apis ligustica Spinola 1806: 35. [ligustica Spinola] Apis capensis Eschscholtz 1822: 97. [capensis Eschscholtz] Apis caffra Lepeletier de Saint Fargeau 1836: 402. Preoccupied {nee Linnaeus 1767). [scutel- lata Lepeletier de Saint Fargeau] Apis scutellata Lepeletier de Saint Fargeau 1836: 404. [scutellata Lepeletier de Saint Fargeau] Apis nigritarum Lepeletier de Saint Fargeau 1836: 406. [adansonii Latreille] Apis daurica Fischer von Waldheim 1843: 1. [mellifera Linnaeus] Apis mellifica variety cecropia Kiesenwetter 1860: 315. [cecropia Kiesenwetter] Apis australis Kiesenwetter 1860: 317. Unjusti- fied replacement name for Apis ligustica Spi- nola 1806. [ligustica Spmola] Apis cerifera Gerstacker 1862: 60. Preoccupied (nee Scopoli 1770). [sossimai Engel] Apis mellifica variety remipes Gerstacker 1862: 61. [remipes Gerstacker] Apis liguna Smith 1861b: 14. Nomoi nudum. [? ligustica Spinola] Volume 8, Number 2, 1999 173 Apis mellifica germanica Pollmann 1879: 1. [mel- lifera Linnaeus] Apis mellifica caniica Pollmann 1879: 45. [carfiicn Pollmann] Apis mellifica hymettea Pollmann 1879: 50. Un- justified replacement name for Apis mellifica carnica Pollmann 1879. [carnica Pollmann] Apis mellifica cypria Pollmann 1879: 52. [cypria Pollmann] Apis siciliana Grassi 1881: 1. [siciliana Grassi] Apis mellifica variety nigrita Lucas 1882: 62. [mel- lifera Linnaeus] Apis mcllifida Pollmann 1889: 90. Lapsus calami, [mellifera Linnaeus] Apns mellifida [sic] variety caiicasia Pollmann 1889: 90. [caiicasia Pollmann] Apis ligiirica Dalla Torre 1896: 602. Nomen nu- dum (uec Tegetmeier 1859, 1860: see below). Corrected authorship, [ligustica Spinola] Apis mellifera carniolica Koschevnikov 1900: 1. Unjustified emendation, [carnica Pollmann] Apis mellifera cypriaca Koschevnikov 1900: 1. Unjustified emendation, [cypria Pollmann] Apis mellifica mellifica variety siziliana Buttel- Reepen 1906: 168. Unjustified emendation. [siciliana Grassi] Apis mellifica unicolor variety syriaca Buttel-Ree- pen 1906: 175. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, 50c). [syriaca Skorikov] Apis mellifica mellifica variety lelizeni Buttel-Ree- pen 1906: 184. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, 50c). [mellifera Linnaeus] Apis mellifica unicolor variety intermissa Buttel- Reepen 1906: 187. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, 50c). [intermissa Maa] Apis mellifica unicolor v ariety friesei Buttel-Ree- pen 1906: 188. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, 50c). [adansonii Latreille] Apis mellifera lamarckii Cockerell 1906: 166. Re- placement name for Apis fasciata Latreille 1804b. [lamarckii Cockerell] Apis mcllefica Enderlein 1906: 331. Lapsus calami, [mellifera Linnaeus] Apis mellifica unicolor variety frisei Enderlein 1906: 335. Lapsus calami, [adansonii Latreille] Apis mellifera sicula Montagano 1911: 26. [sicili- ana Grassi] Apis ndamsoni Meunier 1915: 210. Lapsus calami, [adansonii Latreille] Apis fiiscata Meunier 1915: 210. Lapsus calami, [lamarckii Cockerell] Apis mellifica kaffra Jack 1916: 397. Lapsus calami, [scutellata Lepeletier de Saint Fargeau] Apis mellifera caucasica Gorbachev 1916: 39. Un- justified emendation, [caucasia Pollmann] Apis mellifica variety banatica Grozdanic 1926: 57. [carnica Pollmann] Apis fascrata Baldensperger 1928: 173. Lapsus calami, [lamarckii Cockerell] Apis eurasiatica Skorikov 1929a: 14. Unjustified replacement name for Apis mellifica variety remipes Gerstacker 1862. [remipcs Gerstacker] Apis mellifera mellifera natio tesquorum Skorikov 1929a; 29. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, 50c). [artemisia Engel] Apis mellifera remipes natio absuatna Skorikov 1929a: 32. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, 50c). [remipes Gerstacker] Apis mellifera remipes natio siganica Skorikov 1929a: 32. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, 50c). [remipes Gerstacker] Apis mellifera remipes natio georgica Skorikov 1929a: 32. Proposed as new again in Skorikov (1929b). Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, 50c). [remipes Gerstacker] Apis meda Skorikov 1929b: 253. [meda Skorikov] Apis mellifera natio acervoriim Skorikov 1929b: 253. Preoccupied {nee Linnaeus 1758). [arte- misia Engel] Apis remipes transcaucasica Skorikov 1929b: 254. [remipes Gerstacker] Apis remipes transcaucasica natio absuana Skori- kov 1929b: 254. Lapsus calami, [remipes Ger- stacker] Apis remipes transcaucasica natio iberica Skorikov 1929b: 254. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, 50c). [remipes Gerstacker] Apis remipes armeniaca Skorikov 1929b: 254. [re- mipes Gerstacker] Apis mellifera banata Skorikov 1929b: 263. Un- justified emendation, [carnica Pollmann] Apis (Apis) syriaca Skorikov 1929b: pi. 1. First available usage, [syriaca Skorikov] Apis mellifica variety sahariensis Baldensperger 1932: 829. [sahariensis Baldensperger] Apis niger Baldensperger 1932: 830. Nomen nu- dum. [? mellifera Linnaeus] Apis mellifera mingrelica Lavrezhin 1935: 656. Nomen nudum, [remipes Gerstacker] Apis mellifera taurica Alpatov 1935: 665. Pro- posed as new again in Alpatov (1938). [taur- ica Alpatov] Apis (Apis) intermissa Maa 1953: 591. First avail- able usage [intermissa Maa]. Apis (Apis) mellifera anatoliaca Maa 1953: 599. [anatoliaca Maa] 174 Journal of Hymenoptera Research Apis meUifera siriaca Kerr and Amaral 1960: 12. Lapsus ailnnii. [si/rincn Skorikov] Apis mellifcra monticola Smith 1961a: 258. Pro- posed as new again in Smith {1961b). [imvi- ticola Smith] Apis mellifera litorea Smith 1961a: 259. Proposed as new again in Smith (1961b). [litorea Smith] Apis mellifera lamarchii Smith 1961b: 148. Lapsus calami, [lamarckii Cockerell] Apis millifica Goetze 1964; 9. Lapsus calami, [mel- lifera Linnaeus] Apis mellifica intermissa tellica Goetze 1964: 25. Nomen nudum. See also below, [intermissa Maa] Apis mellifica mellifica silvarum Goetze 1964: 26. Nomen nudum. See also below, [mellifera Lin- naeus] Apis mellifica lamarcki Ruttner 1968; 41. Unjus- tified emendation, [lamarckii Cockerell] Apis mellifica adami Ruttner 1975: 271. Proposed as new again in Ruttner (1980) with the spe- cies called mellifera. [adami Ruttner] Apis mellifera major Ruttner 1976a; 354. [inter- missa Maa] Apis mellifera nubica Ruttner 1976a: 359. [jcmen- itica Ruttner] Apis mellifera littorea Ruttner 1976a; 361. Lapsus calami, [litorea Smith] Apis mellifcra jemenitica Ruttner 1976a: 366. [je- menitica Ruttner] Aptis mellifera carpatica Barac 1977: 270. [cnrnica PoUmann] Apis mellifera anatolia Winston 1987; 12. Lapsus calami, [anatoliaca Maa] Apis mellifera yemenitica Ruttner 1988: 212. Un- justified emendation, [jemenitica Ruttner] Apis mellifera sudanensis Ruttner 1988; 214. No- men nudum (see below), [jemenitica Ruttner] Apis mellifera iberica Ruttner 1988: 236. Preoc- cupied {nee Skorikov 1929b, Goetze 1964: see treatment of A. mellifera subspecies below); corrected authorship, [iberiensis Engel] Apis mellifera macedonica Ruttner 1988: 249. [ma- cedonica Ruttner] Apis mellifera logustica Kugler 1988: 268. Lapsus calami, [ligustica Spinola] Apis mellifica rodopica Petrov 1991: 17. Proposed as new again in Petrov (1996). [macedonica Ruttner] Apis millifern Willis, Winston, and Honda 1992; 169. Lapsus calami, [mellifera Linnaeus] Apis mellitera Petrov 1993: 36. Lapsus calami. [mellifera Linnaeus] Apis mellifica capratica Petrov 1996; 58. Lapsus calami, [carnica Pollmann] Apis mellifera bandasii Radloff and Hepburn 1997; 57. Nomen nudum (see below), [jemeni- tica Ruttner] Apis mellifera ruttneri Sheppard, Arias, Grech, and Meixner 1997 [1998]; 290. [ruttneri Shep- pard et (?/.] Apis mellifera artemisia Engel 1999: see below. Replacement name for Apis mellifera acervo- runi Skorikov 1929b {nee Linnaeus 1758). [ar- temisia Engel] Apis mellifera iberiensis Engel 1999; see below. Replacement name for Apis mellifera iberica Ruttner 1988 {nee Skorikov 1929b). [iberiensis Engel] Apis mellifera sossimai Engel 1999: see below. Re- placement name for Apis cerifera Gerstacker 1862 (nee Scopoli 1770). [sossimai Engel] Comments. — The term "natio" as used by Skorikov (1929a, b) appears to be syn- onymous witli "variety" as used by other authors. Both Maa (1953) and Goetze (1964) in- clude the name A. domestica Ray (1710) in their taxonomic treatment of the v^estern honey bee. This is a pre-Linnean name and it is therefore excluded from zoolog- ical nomenclature (LC.Z.N. 1985: Art. 3) and my treatment above. It might be interpreted from my treat- ment of Pollmann's names above that A. mellifida as used by this author was not a lapsus calami but in actuality an unjustified emendation due to the fact that it appears to have been used twice (under mellifida and mellifida caucasia). This is not the case. The name tnellifida was only used once in Pollmann's (1889) work in the proposal of the variety caucasia. It appears twice in my list since two taxonomic entities are in- volved in this one action by Pollmann; one as the lapsus calami mellifida and the second for the proposal of the variety (now sub- species) caucasia. Maa (1953) and Dalla Torre (1896) both cite the name A. li;^urica with one author attributing it to Tegetmeier (1859) and the other to that paper as well as to Teget- meier (1860). Interestingly enough neither Volume 8, Number 2, 1999 175 of Tegetmeier's contributions use this name. Tegetmeier's 1859 paper, the pur- ported first usage of the epithet, is merely an untitled note about an exhibition of Apis specimens at a meeting of the Ento- mological Society of London and simply reads, "Mr. Tegetmeier exhibited speci- mens of Apis ligustica, of which he had lately received living examples of the queen and workers from the Continent, where it is considered a more profitable species to the owner than the common honey bee; he hoped, during the next sea- son, to test the correctness of this opin- ion." Tegetmeier (1860), a similarly unti- tled note from a societal meeting, is mere- ly a follow up to his 1859 notice and uses the name A. ligustica as well (not A. lig- urica). A perusal of other issues of the Pro- ceedings of the Entomological Society of Lon- don failed to find any usage of A. ligurica, and it would therefore appear that the first appearance of such a name was in Dalla Torre's catalog (1896). The name mingrelica was used by La- vrezhin (1935) and attributed to Skorikov (1929a); however, Skorikov did not pro- pose such a specific epithet. Skorikov (1929a: 32, 41, 44) used the name MHHipejT&cKaa only in Russian and as part of a vernacu- lar name (it even has a Russian adjectival ending which changes in the text as the case of the noun it modifies changes), nev- er once proposing it as a latinized taxo- nomic entity with a description. The same can be said for several of the nomiua niida listed by Goetze (1964); these being silva- rum attributed to Alpatov (1935), tellica at- tributed to Buttel-Reepen (1906), and us- suriensis attributed to Lawrjochin (1960). These names, as used by Goetze (op. cit.), are also excluded from nomenclature ow- ing to the fact that they are infrasubspe- cific names proposed after 1960 (I.C.Z.N. 1985: Art. lb and 45e-g). Radloff and Hepburn (1997) employ the subspecific names bandasii (attributed to Mogga) and sudanensis (attributed to Ras- had) both of which are nomina nuda. I failed to find in publications of Rashad available to me any official proposal of a subspecies with the latinized name suda- nensis that would satisfy the rules of no- menclature, although many of his papers mention the Sudanese bees as variable and distinctive (e.g., Rashad and El-Sarrag 1978). Ruttner (1988) also employs the name sudanensis (attributing it to an un- published manuscript by Rashad) and thus it is a nomen nudum with Ruttner ap- parently being the correct author. The name baiuiasii, however, was proposed by Mogga (1988) and would meet the appro- priate criteria except that this name was only treated in his unpublished Master's thesis, is therefore excluded from zoolog- ical nomenclature [I.C.Z.N., 1985: Art. 9(11)], and means that its use by Radloff and Hepburn (op. cit.) is a nomen nudum. Both of these names were once again used by Hepburn and Radloff (1998) although they were only mentioned as synonyms of jemenitica in their consideration of African races, a synonymy considered amply jus- tified by both Ruttner (1988) and myself (above). This seems an appropriate venue to comment on the apparent confusion of the valid name for this species. The persistent use of the unjustified replacement name A. mellifica by some modern authors is dis- tressing and either demonstrates a general ignorance of taxonomic rules or a flagrant dismissal of the International Code of Zoological Nomenclature. Ignorance of the Code is understandable as not all bi- ologists are intimately familiar, or even have an express need to be so familiar, with its rules and nuances. Dismissal of the Code is, however, a more serious sit- uation. The criticism that the name melli- fica is more descriptive than mellifera is be- side the point and, if the argument of Ben- ton (1904) is followed, vastly incorrect anyway. Suffice to say, the use of A. mel- lifera versus A. mellifica is not a matter of 176 Journal of Hymenoptera Research personal choice. Apis melUfera is the only valid name for this species and authors should abide by this. Distribution. —Ihe western honey bee presumably arose in the Near East or Pon- tocaspian regions at some point during the early Pliocene. This species subse- quently dispersed throughout Africa and Madagascar (entering via the Arabian Peninsula), into Europe, and into North- western Asia as far as the Russian Steppes. Localized adaptation gave rise to the nu- merous subspecies which are recognized today. Spread of this species elsewhere in the world has been through the action of human interference. Refer to treatment of individual subspecies for further details. Apis mellifera was introduced into the New World by European settlers during their colonization of the western hemi- sphere. Although some early authors felt that that North American honey bees were specifically distinct from the European honey bees (e.g., Belknap 1792) it was quickly recognized by naturalists of the time that Apis was not native to the New World (e.g.. Barton 1793), a belief con- firmed by their discussions with the Na- tive American populace. In fact. President Thomas Jefferson commented that the Na- tive Americans referred to the honey bees as "the white man's fly" (Jefferson 1787) and signaled to them the advance of col- onists. Early reports that honey bees oc- curred in the New World (Hernandez 1648) were based on misidentifications with species of Melipona (Gerstacker 1863). Most early authors correctly referred to the transport of honey bee colonies from Europe by colonists (e.g., Josselyn 1674). Despite the arguments raging in the late 1700's over the "distinctiveness" and na- tivity of A. mellifera in the Americas, no names were proposed for them. Apis (Apis) nigrocittcta Smith The Sulawesian Honey Bee Apis nigrocincta Smith 1861a: 93. Apis mdlifica indica variety picea Buttel-Reepen 1906: 193. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, 50c). Apis {Signmtapis} nigrocincta marginella Maa 1953: 575. Co»;?«e»fs.— Although 1 have previously been hesitant to recognize this species as distinct from A. cerana (e.g., Engel 1998a), I here correct this following the excellent work of Hadisoesilo et al. (1995), Hadisoe- silo and Otis (1996, 1998), and Hadisoesilo (1997). The absence of a pore in the drone cell cap is a plesiomorphic feature of Apis in general and for the cerana-c\ade in par- ticular (Fig. 1). Apis nigrocincta is the only member of the cerana-dade to lack this fea- ture. The apomorphic presence of such a pore unites A. cerana and A. koschevnikovi. No subspecies are recognized in A. nigro- cincta. Distribution.— Ihis species only occurs on the island of Sulawesi except in the ex- treme southwest corner in a lower eleva- hon band from a bit northwest of Ujung Padang around to Bulukumba and be- yond (G. W. Otis pers. comm.). Previous records of this species on Mindanao (Otis 1996, Damns and Otis 1997) are probably A. cerana as revealed by the development of the drone cell which had a hard cap with a pore (G. W. Otis pers. comm.), this being a feature of A. cerana and not A. ni- grocincta (Hadisoesilo and Ohs 1998). Pop- ulations of Mindanao are certainly distinc- tive from typical A. cerana and need to be investigated further. Subgenus Megapis Ashmead Megapis Ashmead 1904: 120. Type species: Apis dorsata Fabricius 1793, original designaHon. Maa 1953: 552. Diflgttosis.— WORKER: Forewing length 12-15' mm. Basal vein frequently gently curved, strongly distad cu-a (similar to Fig. 4). Angle of posteroapical margin of first submarginal cell less than 45°. Distal abscissa of vein M in hind wing present (similar to Fig. 4). DRONE: Hind basitar- sus without thumb-like process. Vertical Volume 8, Numbek 2, 1999 177 arm of T8 as long as horizontal arm; S7 and S8 not fused mesally. Gonobase ab- sent. Ventral gonocoxa membranous; dor- sal gonocoxa reduced. Ventral cornua of endophallus not recurved ventrally. MA- TURE LARVA: Labrum with peg-like se- tae scattered over surface. Galea larger than maxillary palpus. Labial palpus spic- ulate. Epipharynx without setae. Atrial in- ner walls ridged. ETHOLOGY: Nest con- structed in the open. Dance language per- formed on vertical surface; wagging me- tasoma held above dance surface; recruits potentially far from dancer's metasoma (potentially outside a near field sound range). Drone cell cap without pore. Apis (Megapis) dorsata Fabricius The Giant Honey Bee Apis dorsata Fabricius 1793: 328. [dorsata Fabri- cius] Apis nigripennis Latreille 1804b: 170. [dorsata Fa- bricius] Apis bicolor Klug 1807: 264. Preoccupied (nee Fa- bricius 1781, Villers 1789). [dorsata Fabricius] Apis testaeea Smith 1858: 49. [dorsata Fabricius] Apis zoiiata Smith 1859: 8. Preoccupied (jit'c Lin- naeus 1758). [binghajiii Cockerell] Apis laborwsa Smith m Moore et al. 1871; 249. [laboriosa Smith] Apis testaeea Smith 1871: 396. Lapsis ealami. [dor- sata Fabricius] Megapis zonata (Smith); Ashmead 1904: 121. [binghami Cockerell] Megapis dorsata (Smith); Ashmead 1904: 121. [dorsata Fabricius] Apis dorsata binghaiiii Cockerell 1906: 166. Re- placement name for Apis zonata Smith 1859. [binghami Cockerell] Apis binghami sladcni Cockerell 1914: 13. [labor- iosa Smith] Apis darsata Baldensperger 1928: 173. Lapsus eal- ami. [dorsata Fabricius] Apis himnlai/ana Maa 1944: 4. Nomen nudum, [la- boriosa Smith] Megapis breviligula Maa 1953: 563. [brevdigula (Maa)] Megapis binghann (Cockerell); Maa 1953: 564. [binghami Cockerell] Megapis laboriosa (Smith); Maa 1953: 570. [labor- iosa Smith] Apis dorsatao Ruttner 1988: 118. Lapsus calami. [dorsata Fabricius] Apis labortiosa Willis, Winston, and Honda 1992: 169. Lapsus calami, [laboriosa Smith] Distribution. — Refer to treatment of in- dividual subspecies below. Subgenus Micrapis Ashmead Micrapis Ashmead 1904: 122. Type species: Apis florea Fabricius 1787, monobasic and original designation. Maa 1953: 557. Diagnosis. — WORKER: Forewing length 6-7 mm. Basal vein frequently gently curved, strongly distad cu-a (similar to Fig. 4). Angle of posteroapical margin of first submarginal cell less than 45°. Distal abscissa of vein M in hind wing absent. DRONE: Hind basitarsus with thumb-like process. Vertical arm of T8 as long as hor- izontal arm; S7 and S8 not fused mesally. Gonobase represented by isolated frag- ments. Ventral gonocoxa sclerotized as transverse bar; dorsal gonocoxa not re- duced. Ventral cornua of endophallus not recurved ventrally. MATURE LARVA: La- brum with peg-like setae scattered over surface. Galea larger than maxillary pal- pus. Labial palpus spiculate. Epipharynx with setae. Atrial inner walls ridged. ETHOLOGY: Nest constructed in the open. Dance language performed on hor- izontal surface; wagging metasoma held above dance surface; recruits potentially far from dancer's metasoma (potentially outside a near field sound range). Drone cell cap without pore. Apis (Micrapis) andreniformis Smith The Black Dwarf Honey Bee Apis andreniformis Smith 1858: 49. Apis florea andreniformis variety sumatrana En- derlein 1906: 339. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, 50c). Micrapis andreniformis (Smith); Maa 1953: 601. Comments. — This species was reinstated just over a decade ago by Wu and Kuang (1986, 1987) with additional evicience for this decision provided by Wongsiri et al. 178 Journal of Hymenoptera Research (1989). It's common name comes from the nearly black Tl-2 while in A. florea these structures are reddish brown. No subspe- cies are recognized. Distribution. — Apis andreiiifonnis occurs as far north as southern-most Yunnan Province in China, south into Malaysia, eastward on the mainland to the coast of Vietnam, and westward as far as north- western India along the eastern border of Nepal. Although specimens are not re- corded from Bhutan, Cambodia, Myan- mar, or Nepal the records of A. andreiiifor- iiiis in the Bengal, Sikkim, and Assam re- gions of India suggests that it will even- tually be found in these countries (Otis 1996) unless it has since become locally ex- tinct. Apis (Micrapis) florea Fabricius The Red Dwarf Honey Bee Apis florea Fabricius 1787: 305. Ayis seminifa Hoffmannsegg 1818: 60. Apis lobatn Smith 1854: 416. Apis floralis Home and Smith 1870: 181. Lapsus calami. Apis testacea Bingham 1898: 129. Preoccupied (nee Smith 1858). Micrapis florea (Fabricius); Ashmead 1904: 122. [see Comments below] Apis florea variety rufiventris Friese in Buttel- Reepen 1906: 167, 170. Apis florea florea variety fnscata Enderlein 1906: 338. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23j, 50c). Apis nursci Cockerel! 1911a: 319. Replacement name for Apis testacea Bingham 1898. Apis florea nasicana Cockerell 1911b: 241. Comments. — Although when proposing the genus Micrapis Ashmead (1904) did not explicitly transfer the type species, A. florea, into this genus (he shll referred to the species as A. florea and not Micrapis florea) he is to be consider as having made the combination as the proposal of Mi- crapis was done prior to 1961 [I.C.Z.N. 1985: Art. 51c(ii)]. There are presently no subspecies recognized in A. florea. Distribution. — This species is distributed from the eastern regions of the Arabian Peninsula, through southern Iran and Iraq to Pakistan and India as far north as the Himalayan uplift. The species extends eastward as far as Vietnam, north to Yun- nan Province, China, and south into In- donesia. Refer to Otis (1996) for more pre- cise locality records. Apis florea was intro- duced into Sudan a little over a decade ago from Pakistan and has become estab- lished (Lord and Nagi 1987, Mogga and Ruttner 1988). SUBSPECIES Owing to both the high visibility and dramatic geographic variation of the hon- ey bees, the more common species have been heavily divided infraspecifically. Be- low I have attempted to outline the pres- ently recognized subspecies. There are no valid subspecies of A. andreniformis, A. flo- rea, A. koschevnikovi, or A. >iigrocincta and 1 have not recognized any subspecies among the fossil forms. A few of the names treated below are used incorrectly under the rules of zoological nomencla- ture (I.C.Z.N. 1985) and are therefore treated in further detail, otherwise I have not given detailed taxonomic histories for each subspecies. I have tried to provide common names for these morphs as well as historical, geographical, or important biological comments for some taxa. As far as I have been able to ascertain, holotypes or neotypes have not been des- ignated for certain subspecies. This will be necessary to assure stability in nomencla- ture. The general problems with the rec- ognition of subspecies have been dis- cussed by several authors, the best cri- tique being that by Wilson and Brown (1953). My recommendation is a complete cessation of proposals of new subspecies in Apis (primarily A. mellifera) until the taxonomy has been more thoroughly ex- amined (something which 1 hope this pa- per will facilitate). Volume 8, Number 2, 1999 179 Apis {Apis) cerana Fabricius The populations of A. cerana in the Phil- ippines are somewhat distinctive (e.g., Ruttner 1988, Smith 1991b, Smith and Ha- gen 1996, Damus and Otis 1997) and could possibly be elevated to subspecific status. If this is done the name of this race would become A. cerana philippina Skorikov with A. samareusis Maa as a junior synonym. 1. Apis cerana cerana Fabricius: The Eastern Honey Bee Distribution. — This subspecies occurs as far east as Afghanistan and south along the central deserts and mountain ranges. Along the Himalayan uplift in Tibet it is replaced by A. cerana skorikovi and further south in India it is replaced by A. cerana indica. The subspecies then ranges across most of central and southern China al- though not in the high regions of central China where the black bee (A. cerana heiin- ifeng) occurs. Apis cerana cerana then rang- es along the eastern edge of Asia up to Korea and the southern-most areas of the former Soviet Union (Ussuria) and south to the northern reaches of Vietnam. 2. Apis cerana heimifeng Engel, new subspecies: The Black Chinese Honey Bee Apis cerana, Aba race, Peng et al. 1989: 15. Diagnosis. — Distinguishable from typi- cal A. cerana cerana in China by the dark brown to black scutellum and T3-4 (these are yellow in A. cerana cerana). In this re- spect it resembles the Tibetan Honey Bee, A. cerana skorikovi, from which it differs in the larger body size (body length 12.5-13.7 mm; forewing length 8.8-9.3 mm) and low tendency to swarm (very strong in A. cer- ana skorikoz'i). This is the largest A. cerana race in China and is a relatively dark bee (hence its common name) with a dark me- tasoma, scutellum, and legs. This dark col- oration is partly owing to the very narrow to completely absent bands of orange on T3-5. Peng et al. (1989) provide several morphometric measures for this bee (as "Aba race"). Distribution. — This is a distinctive race in central China and occurs in relatively high elevation areas (although certainly not at the highest points for this region) mostly near river valleys. Its general dis- tribution is in northern Sichuan Province, southwestern Gansu Province, and eastern Qinghai Province. Holoti/pe.— Worker; CHINA: [Northern Sichuan Province], Nanping, 21 April 1920 (deposited in the American Museum of Natural History, New York). Etyniologi/. — The specific epithet is de- rived from the Mandarin words hei (mean- ing "dark") and mifeng (meaning "honey bee") and refers to the dark coloration of this race. The name is a noun in apposi- tion. 3. Apis cerana indica Fabricius: The Indian Honey Bee Distribution. — This race occurs through most of India except in the higher regions towards the north where it is replaced by A. cerana cerana or in Tibet by A. cerana skorikovi. 4. Apis cerana japonica Radoszkowski: The Japanese Honey Bee Distribution. — This race occurs on the is- lands of Japan. 5. Apis cerana javana Enderlein: The Javanese Honey Bee Comments. — The taxa described as A. peroni by Latreille (1804b) and A. gronovii by Guillou (1841) may be the same as En- derlein's A. cerana javana. Owing to this uncertainty 1 have for now retained A. cer- aim javana as the name of this taxon and considered A. peroni and A. gronovii to be A. cerana indica. Researchers, however, should be aware of this uncertainty and if the consubspecific nature of these taxa is confirmed in the future, then the name of the subspecies would have to be changed to A. cerana peroni with gronovii and javana 180 Journal of Hymenoptera Research as junior synonyms. This situation was also noted by Maa (1953: 580). If this un- fortunately turns out to be the case, then it would probably be advisable to petition the I.C.Z.N. to suppress A. peroni and A. gronovii in favor of A. cerana javana al- though such a case would not be a very strong one owing to the fact that A. peroni has been in use within the last 50 years (e.g., Maa 1953). Distribution. — This subspecies is pres- ently known from Java and as far east as Timor. 6. Apis cerana johni Skorikov: The Sumatran Honey Bee Distribution. — This subspecies only oc- curs on the island of Sumatra. 7. Apis cerana mihiensis Tingek, Koeniger, and Koeniger: The Malaysian Mountain Honey Bee Distribution. — This morph occurs in the mountains of Sabah, Malaysia. Otis (1996) provides details of various collection lo- cales in Malaysia. 8. Apis cerana skorikovi Engel, new subspecies: The Himalayan Honey Bee Apis cerana skorikovi Maa 1944: 4. Nonien nudum. Apis cerana, Himalaya race, Ruttner 1988: 121, 158. Apis cerana, Xizang race, Peng et al. 1989: 15. Apis cerana himalaya Smith 1991b: 154. Nomen nudum. Diagnosis. — This race is fairly gray in overall body coloration, some areas being slightly more yellow while others are clos- er to black (e.g., T4 is entirely black) and with paired yellow spots on S3. There are distinctive white hairs on the terga which contribute to the overall grayish appear- ance. Apis cerana skorikovi, like A. cera)m heimifeng, also differs from A. cerana cerana (as well as A. cerana indica and other sub- species) in the dark scutellum and slightly larger size (body length around 12.2 mm; forewing length around 8.75 mm). Addi- tional morphometric values are provided by Peng et al. (1989). Comments. — The Himalayan race of A. cerana (sometimes called the Tibetan race) was referred to as A. cerana skorikovi by Maa (1944); however he never provided a description for this subspecies. Many au- thors have subsequently recognized this race as distinctive but never associated a name with the numerous descriptions simply referring to this taxon as the Ti- betan honey bee (among other common names). Peng et al. (1989) have done the best to characterize and describe this tax- on referring to it as the Xizang race (Xi- zang is Chinese for Tibet, sometimes spelled Sitsang). I here validate the name A. cerana skorikovi for this taxon, referring readers to the diagnosis provided above as well as the descriptive comments given by Peng et al. (1989: as "Xizang race") and Ruttner (1988: as A. cerana "Himalaya race"). Smith (1991b) referred to this race as A. cerana hinialai/a making a Latin name out of Ruttner's "A. cerana Himalaya race". Ruttner's name was not used as, nor intended as, a scientific name and A. cerana himalaya as used by Smith (1991b) is a nomen nudum. Distribution. — This subspecies occurs along the Himalayan uplift at elevations of 1900-4000 meters. Ho/ofype.— Worker; NEPAL: Ko Pasi near Panauti, 19 March 1982, B. A. Under- wood (deposited in the Cornell University Insect Collection, Ithaca). Paratypes. — Three workers, same collec- tion data as holotype, deposited in the same collection as the holotype; two work- ers, same collection data as holotype, de- posited in the American Museum of Nat- ural History, New York. Etymology. — I have retained the specific epithet used by Maa (1944) for this race. The epithet is a patronymic honoring Dr. A. S. Skorikov who did much work on the classification of bees; honey bees and bumble bees in particular. Volume 8, Number 2, 1999 181 Apis (Apis) mellifera Linnaeus The western honey bee, A. mellifera, is remarkable for its geographic variation and, for this reason, has been broken into a plethora of taxonomic entities over the past two centuries. In earlier studies these forms were often treated as distinct spe- cies; however, today they are recognized as subspecies of a single variable species. 1. Apis mellifera adami Ruttner: The Cretan Honey Bee Comments. — No type has been designat- ed for this race. Ruttner's specimens are among the collection of the Institut fiir Bi- enenkunde in Oberursel, Germany. A sin- gle worker from this material should be selected as the lectotype and the others designated as paralectotypes. Distribution. — This race occurs on the is- land of Crete in the Mediterranean Sea. 2. Apis mellifera adansonii Latreille: The West African Honey Bee Distribution. — This subspecies has a wide distribution in western Africa rang- ing from Niger in the north, east to Sene- gal, and as far south as Zaire. 3. Apis mellifera anatoliaca Maa: The Anatolian Honey Bee Distribution. — This race occurs through- out Turkey, bordering with A. mellifera ma- cedonica in the European portion of that country. In the east, A. mellifera anatoliaca borders with A. tnellifera caucasia and A. mellifera remipes in the area of Azerbaijan, Armenia, and Georgia and with A. melli- fera meda in the southeastern-most regions of Turkey. 4. Apis mellifera artemisia Engel, new name: The Russian Steppe Honey Bee Apis mellifera incUifera natio tesquorum Skorikov 1929a: 29. Unavailable (I.C.Z.N. 1985: Arts. 10c, 23), 50c). Apis mellifera acervorum Skorikov 1929b: 253. Preoccupied (nee Linnaeus 1758). Etymology. — Named for Artemis (Greek; Diana in Roman mythology), the goddess for whom the honey bee was a symbol and whose temple at Ephesus, the Artemi- sium, was listed by Callimachus of Cyrene and Philo of Byzantium as one of the sev- en wonders of the world. Distribution. — This race occurs in the central Russian Steppes. 5. Apis mellifera capensis Eschscholtz: The Cape Honey Bee Comments. — This subspecies is biologi- cally distinctive for its ability to undergo parthenogenetic reproduction when de- prived of a queen (Jack 1916, Anderson 1963, Ruttner 1977). Distribution. — As is indicated by the name of this race, it occurs in the Cape region of South Africa. 6. Apis mellifera caniica Pollmann: The Carnolian Honey Bee Distribution. — This European race oc- curs south of the Alps, east into northern Italy, and west into Yugoslavia and Ro- mania. 7. Apis mellifera caucasia Pollmann, reinstated name: The Caucasian Honey Bee Apis mellifida [sic] variety caucasia Pollmann 1889: 90. Apis mellifera caucasica Gorbachev 1916: 39. Un- justified emendation. Distribution. — This race occurs in the Caucasus Mountains. It is sometimes re- ferred to as the "grey" Caucasian honey bee. 8. Apis mellifera cecropia Kiesenwetter: The Greek Honey Bee Distribution. — This race occurs through- out most of Greece and surrounding Ae- gean islands. It borders A. mellifera mace- donica in the northeast regions of Greece and A. mellifera carnica in the northwest corner of the country. 182 Journal of Hymenoptera Research 9. Apis mellifera cypria PoUmann: The Cyprian Honey Bee Distribution. — This race is presently known only from the island of Cyprus. 10. Apis mellifera iberiensis Engel, new name: The Iberian Honey Bee Apis nicllifica intermissa ibericn Goetze 1964: 25. Infrasubspecific name proposed after 1960 and therefore unavailable (I.C.Z.N. 1985: Art. lb); preoccupied (nee Skorikov 1929b). Apis mellifera iberica Ruttner 1988: 236. Correct- ed authorship (first available usage); preoc- cupied {nee Skorikov 1929b). Comments. — Goetze (1964) proposed this form as a variety of the subspecies in- termissa. Since this is a varietal name (i.e., infrasubspecific) proposed after 1960 it is unavailable and does not enter into zoo- logical nomenclature (I.C.Z.N. 1985: Arts, lb and 45e-g). The first correct application of this name as a subspecies was by Rutt- ner (1988) which thereby validates the name to that author. Unfortunately, a fur- ther complication arises as the name is a primary junior homonym of an available name proposed by Skorikov (1929b) for another type of honey bee from Azerbai- jan. Since no type exists for Ruttner's sub- species I here designate a neotype for A. mellifera iberica, the name of which will be replaced by A. mellifera iberiensis. Neotype. — Worker; SPAIN: Cordoba, Andalusia, September 1986, D. Smith and R. Hagen; now with additional labels reading "NEOTYPE, Apis mellifera iberica Ruttner, desig. M. S. Engel, 1999" and "Apis mellifera iberiensis Engel" (deposited in the Snow Entomological Collection, Natural History Museum, University of Kansas, Lawrence). Three additional workers, identical to the neotype, have also been deposited with the University of Kansas and the American Museum of Natural History, New York. Etymology. — The new subspecific epi- thet refers to the distribution of this sub- species on the Iberian Peninsula. Distribution. — This race natively occurs on the Iberian Peninsula in Europe. Colo- nies were transferred to the South Amer- ican tropics several decades ago, prior to the introduction of A. mellifera scutellata, but populations never became established. Two ecotypes of A. mellifera iberiensis oc- cur in Spain as shown by Santiago et al. (1986) as well as two mitochondrial types as demonstrated by Smith et al. (1991) and Smith and Glenn (1995). The neotype se- lected above comes from the Cordoba populations sampled by Smith and Glenn (1995: their colony 1-1). 11. Apis mellifera intermissa Maa, corrected authorship: The Tellian Honey Bee Comments. — As was noted in the taxo- nomic history of A. mellifera, the name in- termissa as used by Buttel-Reepen (1906) is unavailable for the same reasons as pre- sented for Buttel-Reepen's name koschev- nikovi (see Comments for A. koschevnikovi). The name of this subspecies was first made available by Maa (1953). Distribution. — This race has a tight dis- tribution along the northern coast of Af- rica as far west as Morocco, into Tunisia in the east, but bordered by the Atlas range in the south. 12. Apis mellifera jemenitica Ruttner, reinstated name: The Arabian or Nubian Honey Bee Apis mellifern nnbien Ruttner 1976a: 359. Apis mellifera jemenitica Ruttner 1976a: 366. Apis mellifera yemenitica Ruttner 1988: 212. Un- justified emendation. Apis mellifera sudanensis Ruttner 1988: 214. No- men nudum. Corrected authorship. Apis mellifera bandasii Radkiff and Hepburn 1997: 57. Nomeii nudum. Corrected author- ship. Comments. — Ruttner (1988) synony- mized A. mellifera jemenitica with A. melli- fera nubica and acting as first reviser (I.C.Z.N. 1985: Art. 24) gave jemenitica pri- ority over nubica. In addition, the name/V- Volume 8, Number 2, 1999 183 menitica was unnecessarily emended from its original spelling. The correct name of this taxon should be A. mellifera jemenitka, not yemenitica. No type has been designated for this race. See comments under A. mellifera ada- mi for location of Ruttner's original mate- rial. Distribution. — This subspecies of small bees occurs in hot arid zones of eastern Africa and the Arabian Peninsula. Coun- tries in which it occurs include Chad, Oman, Saudi Arabia, Somalia, Sudan, and Yemen. 13. Apis mellifera lamarckii Cockerel!: The Egyptian Honey Bee Distribution. — This honey bee race oc- curs in a narrow range along the Egyptian Nile Valley. 14. Apis mellifera ligustica Spinola: The Italian Honey Bee Distribution. — This subspecies occurs along the Italian Peninsula although it has been commercially transported through- out the world. The bees currently distrib- uted in Italy have hybridized much with A. mellifera mellifera and A. mellifera carnica in the north. In fact, untainted populations of A. mellifera ligustica appear to be con- fined to Kangaroo Island, Australia where they are being maintained as an unhybri- dized strain (Ruttner 1976b). I recently (January 1999) had the opportunity to visit Kangaroo Island and to see one of these colonies. 15. Apis mellifera litorea Smith: The East African Honey Bee Comments. — This name was originally proposed by Smith (1961a) but was also proposed as new by Smith (1961b). No type appears to have been designated for this subspecies. Distribution. — This subspecies is distrib- uted along the eastern coast of tropical Af- rica occurring from Kenya (perhaps even the southern-most portions of Somalia) south to Mozambique. 16. Apis mellifera macedonica Ruttner: The Macedonian Honey Bee A}ns mellifera macedonica Ruttner 1988: 249. Apis mellifica rodopica Petrov 1991: 17. New syn- onymy. Comments. — No type has been designat- ed for this race. See comments under A. mellifera adami for location of Ruttner's original material. The name rodopica was proposed again as new by the same au- thor five years later (Petrov 1996). There appears to have been no type designated for Petrov's race either. Distribution. — This subspecies occurs as far north as southern Romania, east to Yu- goslavia, and south to northern Greece where it borders A. mellifera cecropia. In the Carpathian Mountains and in Yugoslavia it borders A. mellifera carnica. 17. Apis mellifera meda Skorikov: The Median Honey Bee Distribution. — This race is most common in Iran and Iraq but does range into south- eastern Turkey and northern Syria. 18. Apis mellifera mellifera Linnaeus: The Western or European Honey Bee Distribution. — This subspecies originally ranged throughout central Europe north of the Alps, as far south as southern France in the west, southern Sweden in the north, central Russia in the east, and on the British Isles. In the Ukraine there is a transition over the steppe region to A. mellifera sossimai. 19. Apis mellifera monticola Smith: The East African Mountain Honey Bee Comments. — The name was first pro- posed by Smith (1961a) but was designat- ed as new a second time in Smith (1961b). No type appears to have been designated for this subspecies. Distribution. — This race occurs within the mountains of eastern Africa (e.g., in 184 Journal of Hymenoptera Research Kenya and Tanzania). The occurrence of A. mellifera motiticola in the mountains of Cameroon should be checked carefully. 20. Apis mellifera remipes Gerstacker, reinstated name: The Yellow Armenian Honey Bee Apis remipes Gerstacker 1862: 61. Apis mellifera armeniaca Skorikov 1929b: 254. Comments. — The name armeniaca was subjectively chosen by Ruttner (1988: 192) over remipes. This decision is invalid be- cause remipes has priority and the senior synonym must be recognized as the name for the subspecies. Distribution. — This race occurs in Ar- menia and may be the same as A. mellifera anatoliaca, in which case Gerstacker's name has priority for the subspecies. 21. Apis mellifera ruttneri Sheppard, Arias, Grech, and Meixner: The Maltese Honey Bee Comments. — No type was originally des- ignated for the subspecies but this is here corrected by the original authors thereby stabilizing the name of this taxon. Lectotype. — Worker; MALTA: St. Julians, 10 August 1995, W. S. Sheppard; desig- nation of W. S. Sheppard, M. C. Arias, A. Grech, and M. D. Meixner. The lectotype now bears a label indicating it as such and is deposited in the M. T. James Entomo- logical Museum, Washington State Uni- versity. Paralectotypes. — Nine workers; same col- lection data as lectotype; deposited in the same collection as the lectotype. A further two paralectotype workers are deposited in the American Museum of Natural His- tory, New York. All paralectotypes desig- nated by W. S. Sheppard, M. C. Arias, A. Grech, and M. D. Meixner. Distribution. — This race is only distrib- uted on the island of Malta in the Medi- terranean Sea. 22. Apis mellifera sahariensis Baldensperger: The Saharan Honey Bee Comments. — Ruttner (1988) attributed this name to Baldensperger (1923). He has, however, unfortunately confused recog- nition of the subspecific entity with taxo- nomic proposal of the subspecies. Al- though Baldensperger (1923) notes the distinctive character of the Saharan honey bees he does not provide a latinized name for them (in fact, nowhere in his work does he even mention the name Apis, let alone sahariensis). Instead Baldensperger (1923) only uses the vernacular name 'T'abeille saharienne". Thus, as far as I have been able to determine, the name is first made taxonomically available by Bal- densperger in his later paper (1932) where he uses the latinized name and provides a diagnosis. This is one of the races for which I know that no type was ever designated; how- ever I have not been able to confirm whether any material survives of Balden- sperger's original colony which he brought back to France. I have therefore hesitated to designate a neotype. This race is markedly lighter in coloration, particu- larly on Tl-3, than the more common A. mellifera intermissa (which has a noticeably shiny, dark brown to black integument with more sparse pubescence). See Rutt- ner (1988) for further descriptive details. It is possible that this race is synonymous with A. mellifera lamarckii in which case the name lamarckii has priority for the subspe- cies. This possibility should be closely ex- amined in the future. Distribution. — This race, like A. mellifera intermissa, has a tight range in northwest- ern Africa. It occurs along the southern side of the Atlas range. 23. Apis mellifera scutellata Lepeletier de Saint Fargeau: The African Honey Bee Comments. — This is the race introduced into Brazil in the 1950's which has capti- Vol UME 8, Number 2, 1999 185 vated the apicultural world as well the imagination of the public; popularly known as the "Africanized Honey Bee" or "Killer Bee". An account of the transport of African bees to Brazil and their acciden- tal release is given by Kerr (1957, 1967: see also Michener 1975, Spivak et al. 1991, Taylor 1977, 1985 for information on the introduction, spread, and effect of this race in the western hemisphere and for di- rection to other literature sources). Distribution. — This subspecies also has a large distribution in Africa and ranges from South Africa northward along the eastern half of the continent to about So- malia, it apparently does not occur along the eastern coastal plain where A. nielli fern litorea is found. 24. Apis meUifera siciliana Grassi, reinstated name: The Sicilian Honey Bee Apis siciliana Grassi 1881: 1. Apis mellifica mellificn variety siziliana Buttel- Reepen 1906: 168. Unjustified emendation. Apis sicula Montagano 1911: 26. Comments. — Ruttner's (1988) arbitrary choice of the epithet sicula over siciliana for the Sicilian race of honey bees does not meet the criterion of priority and thus the name siciliana must be reinstated for this morph. Distribution. — This subspecies occurs on the island of Sicily in the Mediterranean Sea. 25. Apis mellifera sossimai Engel, new name: The Ukrainian Honey Bee Apis cerifera Gerstacker 1862; 60. Preoccupied {nee Scopoli 1770). Etymologi/. — The new specific epithet is derived from St. Sossima, patron saint of beekeeping in the Ukraine. St. Sossima may be a Christian version of the bee-god Zosim of some early pagan tribes of Rus- sia. Distribution. — This race occurs along the eastern borders of A. nwllifera carnica's range. It occurs mostly in the Ukraine and easterly over to the northern regions of the Caucasus Mountains where it borders A. mellifera caucasia. South in the Crimea it is replaced by A. mellifera taurica. 26. Apis mellifera syriaca Skorikov, corrected authorship: The Syrian Honey Bee Comments. — As was noted in the taxo- nomic history of A. mellifera, the name si/r- iaca as used by Buttel-Reepen (1906) is un- available for the same reasons as present- ed for Buttel-Reepen's name koschevnikovi (see Comments for A. koschevnikovi). The name of this subspecies was first made available by Skorikov (1929b). Distribution. — Apis mellifera syriaca oc- curs along the eastern shores of the Med- iterranean Sea; north from Syria to the Ne- gev Desert in the south. It is sometimes known as the Palestine honey bee. 27. Apis mellifera taurica Alpatov: The Crimean Honey Bee Distribution. — This race occurs along the north-central shores of the Black Sea; in the Crimea. 28. Apis mellifera tmicolor Latreille: The Malagasy Honey Bee Distribution. — The distribution of this race in Madagascar is discussed by Brooks and Michener (1988). Apis (Megapis) dorsata Fabricius Some authors prefer to recognize one or more of the A. dorsata races as distinct spe- cies. This is most often done with A. dor- sata laboriosa followed by A. dorsata brevi- li^ula and to a much lesser degree with A. dorsata biiighami. 1. Apis dorsata binghami Cockerell: The Giant Sulavv'esi Honey Bee Distribution. — This race of giant honey bees occurs on the island of Sulawesi. 186 Journal of Hymenoptera Research 2. Apis dorsata breviligula (Maa): The Giant Philippine Honey Bee Comments. — Since Maa (1953) originally proposed the name breviligula in the genus Megapis his name should be written in pa- rentheses (I.C.Z.N. 1985: Art. 51c). Distribution. — This morph of giant hon- ey bees occurs in the Philippines. The spe- cies status of this race is oft argued based on nest site characteristics as reported by Morse and Laigo (1969: as A. dorsata) and Starr et al. (1987). 3. Apis dorsata dorsata Fabricius: The Common Giant Honey Bee Distribution. — This subspecies has the largest distribution of the group; ranging from India, east to the coast of Vietnam, and into the southeast Asian islands where it is sometimes replaced by other subspecies (see accounts for binghami and breviligula). 4. Apis dorsata laboriosa Smith: The Giant Himalayan Honey Bee Comments. — I currently do not recognize the subspecific form laboriosa as a separate species (see also Engel 1998a; but see Sak- agami et al. 1980, Roubik et al. 1985, McEvoy and Underwood 1988, Under- wood 1990a, b). Growing evidence, how- ever, suggests that the distinction may be valid and my decision to place laboriosa within dorsata should be examined more closely. Sakagami et al. (1980) provided a detailed account of the morphology of la- boriosa versus typical dorsata. Many of the characters they use to justify specific sta- tus are, however, quite variable (e.g., pro- tuberance of the ocelli, general pubescence patterns, &c.) and across a large range of specimens blend naturally into one anoth- er except some measures of body size which I presently feel are more indicative of a subspecies rather than of a separate species. One character which appears to me to be somewhat reliable is the minute, broad medioapical extension of the grad- ulus on S3 in dorsata while this is com- pletely absent in laboriosa. Distribution. — This subspecies is distrib- uted at high aWtudes (1,200 to 4,000 m) on the slopes of the mountains from northernmost India to the northernmost boundary of Laos. Further details of its distribution are provided by Otis (1996). FOSSIL HONEY BEES More paleontological work has focused on the honey bees than on any other group of bees. Recently I have treated some of these extinct species and attempt- ed to place them into a phylogenetic framework with the living species (Engel 1998a: Fig. 1). Petrov (1992, 1997) also dis- cussed Apis evolution with reference to the fossil record, in particular focusing on the origin of the Bulgarian honey bee (A. mellifera rodopica = A. mellifera macedonica in my system). Hong (1984) described a compression fossil from the Early Creta- ceous of China as the oldest fossil bee and as a relative of Apis. Hong's Palaeapis bei- boziensis is in actuality a sphecid wasp (Michener 1997, Engel 1998a) and has no bearing on apine evolution. Lastly, Nel et al. (1999) presented a small review of fossil Apis specimens but overlooked the species presented by myself, recent treatments of Apini (e.g., Michener 1990), as well as the numerous synonymies within the group, instead simply regurgitating the classifi- cation of Zeuner and Manning (1976) which is fraught with errors (e.g., refer to Engel 1998a). Nel et al. (op. cit.) have also followed Zeuner and Manning (1976) and other authors in over splitting "species" of honey bees based on minor morphometric differences in wing venation and size (e.g., their unnamed species A-J: Nel et al. 1999); characters which distinguish mere subspe- cies in Apis at best. These authors conclud- ed (p. 31) that the fossil honey bees could not be studied from a phylogenetic per- spective; however, such an attempt had al- ready been successfully undertaken at least three times previously (e.g., Buttel- VdiUME 8, Number 2, 1999 187 Reepen 1906, Statz 1931, Engel 1998a). I am presently involved in a monographic study of the fossil bees, particularly those from Baltic amber, and cladistic analyses have been successfully completed for the living and fossil corbiculate bees, includ- ing Apini (Engel 1998b, in prep.). In order to maintain the monophyly of the generally recognized subgenera (re- cent and fossil) it seems appropriate to propose two new subgenera thereby breaking up the paraphyletic subgenus Si/iiapis. One of these new groups was called Hnuffnpis by Armbruster (1938), but he failed to designate a type species for the genus, thereby making this name un- available according to the I. C.Z.N. (1985: Art. 13c) (see also discussions in Michener 1990, 1997). None of the fossil species be- longs to the living subgenera Apis, Megnp- is, or Micrapis. Subgenus Cascapis Engel, new subgenus Hauffapis Armbruster 1938: 37. Unavailable as no type species was originally designated (I.C.Z.N. 1985: Art. 13c). See also Michener (1990, 1997). Ti/pe species. — Apis arnibrusteri Zeuner 1931. Diagnosis. — WORKER: Basal vein gently curved, strongly distad cu-a (similar to Fig. 4). Angle of posteroapical margin of first submarginal cell greater than 45°. Dis- tal abscissa of vein M in hind wing present (similar to Fig. 5). Drone, mature larva, queen, and ethology all unknown. Etymology. — The new genus-group name is a combination of cascus (L. old) and apis (L. bee). Comments. — The subsequent designa- tion of Hauffapis scheiithlei by Zeuner and Manning (1976) as the type species was in the synonymy of Hauffapis with Apis and is therefore invalid (I.C.Z.N. 1985: Art. lie) (see also Michener 1990). Apis (Cascapis) armbnisteri Zeuner Armbruster's Honey Bee Ayis armbnisteri Zeuner 1931: 292. Hnuffnpis sclieiithlci Armbruster 1938: 43. Hnuffapis scheeri Armbruster 1938: 43. New syn- onymy. Hnuffnpis schnnuniiiii Armbruster 1938: 44. New synonymy. Hnuffnpis scheeri variety gaUnuni Armbruster 1938: 45. New synonymy. Hnuffnpis scheeri variety rnhdei Armbruster 1938: 45. New synonymy. Hnuffnpis scheuthlei variety seeiiminii Armbruster 1938: 45. New synonymy. Hnuffnpis scheuthlei variety zeuiieri Armbruster 1938: 45. New synonymy. Apis nnnbruiteri chcuthlei Hong and Miao 1992: 2. Lnpsus cnlnmi. Comments. — The species and subspecies of Hauffapis proposed by Armbruster (op. cit.) were all described from the Miocene of Randecker Maar and were based on mi- nor wing variations. None of these varia- tions justifies subspecific status even un- der the extreme criteria used for splitting morphs of extant species (e.g., A. mellifera). The species is presently known from the Miocene of Germany. Subgenus Synapis Cockerell Apis (Synnpis) Cockerell 1907: 229. Type species: Apis {Syunpis) henshmvi Cockerell 1907, monobasic. Zeuner and Manning 1976: 240. Diagnosis. — WORKER: Basal vein gently curved, only slightly distad cu-a (Fig. 6). Angle of posteroapical margin of first sub- marginal cell greater than 45°. Distal ab- scissa of vein M in hind wing present (as in Fig. 5). Drone, mature larva, queen, and ethology all unknown. Apis (Synapis) henshaivi Cockerell Henshaw's Honey Bee Apis (Synnpis) henshnwi Cockerell 1907: 229. Apis donnitnus Cockerell 1907: 228. Preoccupied (uec Heyden 1862). Apis oligoceuicn Meunier 1915: 210. Syuapis dorinitniis (Cockerell); Statz 1931: 45. Synapis henshnzei (Cockerell); Statz 1931: 45. 188 Journal of Hymenoitera Research Synapis kaschkei Statz 1931: 50. Apis cuenoti Theobald 1937: 401. New synony- my. Apis oligocaenica Goetze 1964: 9. Lapsus calami. Apis heiischnuii Goetze 1964: 9. Lnpsiis calami. Apis knschki Goetze 1964: 9. Lapsus calami. Apis henshawi dormiens Zeuner and Manning 1976: 241. Replacement name for Apis dor- mitans Cockerell 1907. Apis henshawi kaschkei (Statz); Zeuner and Man- ning 1976: 243. Apis aquisextana Nel, Martinez-Delclos, Arillo, and Pefialver in Arillo et al. 1996: 60. Pro- posed as new again in Nel et al. (1999). Comments. — This species is perhaps the most famous of the fossil honey bees. The type was redescribed and new synony- mies presented by Engel (1998a). This spe- cies is from the Oligocene of Europe (in Germany, France, and Spain). Apis (.Synapis) lotigtibia Zhang The Long-legged Honey Bee Apis longtibia Zhang 1990: 85. Comments. — Engel (1998a) gave some characters to support the recognition of this species, although far much work re- mains to be done on the fossil honey bees of Asia. This fossil is from the Miocene of Shandong Province, China. Apis (Synapis) miocenica Hong The Chinese Miocene Honey Bee Apis miocenica Hong 1983: 10. Apisfota Zhang 1989: 323. Apis shandongica Zhang 1989: 325. Comments. — This species was briefly treated by Engel (1998a) and reasons for the synonymy of A.fota and A. shandongica were given. The species is from the Mio- cene of Shandong Province, China. Apis (Synapis) petrefacta (Riha) The Petrified Honey Bee Sipiapis petrefacta Riha 1973: 217. Apis petrefacta (Riha); Engel 1998a: 275. strata of the Ceske Stredhori Mountains of the Czech Republic. Subgenus Priorapis Engel, new subgenus Type species. — Apis vetiista Engel 1998a. Diagnosis.— WORKER: Basal vein straight, confluent with cu-a. Angle of posteroapical margin of first submarginal cell greater than 45°. Distal abscissa of vein M in hind wing present. Drone, ma- ture larva, queen, and ethology all un- known. Etymologij. — The new subgeneric name is derived from prior (L. earlier) and apis (L. bee). Apis (Priorapis) vetusta Engel, emended name The Aged Honey Bee Apis vetustus Engel 1998a: 271. Comments. — Through a lapse on my part I failed to adjust the gender of the specific epithet to match that of the genus (which is feminine). I here correct this er- ror. The species occurred in the Oligocene of Germany. NOMINA DUBIA AND UNASSOCIATED NOMINA NUDA Apis aenigmatica Rayment, nomen dubiiim Apis aenigmaticus Rayment 1925: 67. Comments. — As pointed out by Cardale (1993), this name was based on the draw- ing of a comb and anecdotal reports of the insect in the absence of any specimens. Rayment (1935) gives another lively ac- count of this elusive "species" which still had not been found at that time (nor has it been found since). See also discussion in Michener (1965: 232). Apis catanensis Roussy, nomen dubiutn Apis catanensis Roussy 1960: 8. Apis catanensis avolii Roussy 1960: 8. Comments. — A fossil form from Miocene Comments. — ^Judging from the original Volume 8, Number 2, 1999 189 description, which is exceedingly incom- plete, this may not be a honey bee at all. The subspecific name avolii is unnecessary and invalid since no subspecific forms were recognized; besides, any subspecies based on the type specimen would have to take the nominate specific epithet and would therefore become A. catanensis ca- tanensis, with A. cataiiensis avolii being an invalid synonym. This is reportedly a fos- sil species from Miocene Sicilian amber. Apis melistiga Zeuner and Manning, nomen dtibittm "Apidae" melisuga Handlirsch 1907: 893. Un- available (I.C.Z.N. 1985: Art. llh[iii]). Apis melisuga Zeuner and Manning 1976: 248. Comments. — The original proposal of this epithet by the great paleoentomolo- gist Anton Handlirsch (1907) was not in accord with the principles of zoological nomenclature and is therefore unavailable under Article llh (iii) (I.C.Z.N. 1985). Zeu- ner and Manning (1976) made the name available through their monograph by publishing this name is combination with a genus-group name and a diagnosis; al- though these authors attributed the name to Handlirsch. Neither Prof. Zeuner nor Dr. Manning had seen the type and their diagnosis of the species was a mere re- gurgitation of Handlirsch's original de- scriptive comments that the specimen re- sembled A. mellifera. In the absence of the type (which is presumably lost) and any real character information it is impossible to confidently place this as a species of Apis. The specimen was reportedly from the Miocene of Italy. Refer to Engel (1998a) for a more thorough treatment of this name. Apis postadamitica Buttel-Reepen, nomen nudum Apis postadamitica Buttel-Reepen 1906: 163. Comments. — Buttel-Reepen (op. cit.) con- sidered this as the hypothetical direct an- cestor of A. mellifera and postulated its oc- currence in the Pliocene. No specimen for this species exists, and it was a mere hy- pothesis of Buttel-Reepen's that this spe- cies existed and would someday be dis- covered. The fact that this name is for a hypothetical taxon means that it is also ex- cluded from zoological nomenclature (I.C.Z.N. 1985: Art. lb). Apis styriaca Pongracz, nomen nudum Apis styriaca Pongracz 1931: 105. Comments. — A supposed fossil species from the Miocene of Germany represented by a wing fragment. No description or fig- ure was provided and the whereabouts of this material is unknown. Apis trigona Rayment, nomen nudum Apis trigona Rayment 1925: 69. Preoccupied (nee Schrank 1798). Comments. — The decision to place this as a nomen nudum was made by Cardale (1993) and is followed here. HONEY BEE VERSUS HONEYBEE In closing I should like to make a brief comment on common names for this group of bees. The great arthropod mor- phologist Robert E. Snodgrass presented a short discussion on the common name for honey bees in the preface to his work con- cerning the anatomy of A. mellifera (Snod- grass 1956). His brief discussion advocat- ing the use of the two word common name {honey bee) over a single word (bo)2- eybee) does not appear to have been wide- ly accepted since apiculturists still com- monly use "honeybee." Although there are no absolute rules for the use of this and other common names, I agree with Snodgrass' preference for a two word name and the logic by which he justified this position. Since I cannot hope to word his position more eloquently, I quote here his short argument: "Regardless of dictio- naries, we have in entomology a rule for insect common names that can be fol- lowed. It says: If the insect is what the 190 Journal of Hymenoptera Research name implies, write the two words sepa- rately; otherwise run them together. Thus we have such names as house fli/, blow fly, and robber fly contrasted with dragonfly, caddicefly, and butterfly, because the later are not flies, just as an aphislion is not a lion and a silverfish is not a fish. The honey bee is an insect and is preeminently a bee; 'honeybee' is equivalent to 'Johnsmith.'" ACKNOWLEDGMENTS I am sincerely grateful to the librarians of Cornell University's Comstock Memorial Library and the Kroch Rare and Manuscript Collection and of the li- brary of the American Museum of Natural History for their help in locating references examined in the course of this study. Kumar and Valerie Krishna helped locate several older references and provided delightful company during these searches. Numerous bee biologists and paleontologists throughout the world graciously supplied copies of their work and shared discussions on Apis classification; to each of them I extend my thanks. David A. Grimaldi, E. Eric Grissell, Charles D. Michener, Card W. Otis, Wojciech Pulawski, Molly G. Rightmyer, and Jerome G. Rozen, Jr., kindly read versions of the manuscript and made valuable corrections and criticisms. Their assistance greatly improved the presentation of this material. Not all agreed with my classification or conclusions and any errors or idiosyncrasies which remain are, of course, my own. 1 am particularly thankful to Mich- ener for discussions on this material. I owe additional thanks to Jeffrey G. Engel for assistance during the preparation of the manuscript and to Molly G. Right- myer and Zhiwei Liu for sharing with me their knowledge of Mandarin in the construction of the ep- ithet heimifeng. I am indebted to numerous institu- tions and individuals for hosting me during my trav- els and /or for the loan or donation of material; in this regard 1 should particularlv recognize Robert W. Brooks, Gabriela Chavarria, E. Richard Hoebeke, Card W. Otis, W. Steve Sheppard, and Deborah R. Smith. W. Steve Sheppard and his coauthors kindly allowed me to include their lectotype designation for A. mellifera ruttneri herein and I am grateful to them for this honor. Donald B. Baker was the first to rec- ognize the correct authorship of A. koschcviiikm'i (pers. comm. to C. D. Michener); I am grateful for his advice and assistance. This work is dedicated in memory of Prof. Dr. Friedrich Ruttner (1914-1998), leading authoritv on Apis classification who did much to clarify honey bee systematics (as well as contributing enormously to many other fields of apiculture). A lovely account cel- ebrating Prof. Ruttner's life was given by Koeniger (1998). LITERATURE CITED Alexander, B. A. 1991a. A cladistic analysis of the ge- nus Apis. Pages 1-28, in D. R. Smith (ed.) Diivr- siti/ in the Genus Apis. Westview Press, Boulder, Colorado, United States, xiv-l-265 pp. Alexander, B. A. 1991b. A phylogenetic analysis of the genus Apis (Hymenoptera: Apidae). Annals of the Entomological Society of America 84: 137-149. Alpatov, V. V. 1935. Contribution to the study of var- iation in the honey bee. 3. The cubital cell on the wings of different forms of the genus Apis and its taxonomical and evolutionary significance. Zoohgischeskiy Zhnrnal 14: 664-673. [In Russian with English summary] Alpatov, V. V. 1938. Contribution to the study of var- iation in the honey bee. 4. Carnolian and Cri- mean bees and their place among other forms of Apis mellifera L. Zootogisctieskiy Zhurnal 17: 473- 481. [In Russian with English summary] Anderson, R. H. 1963. The laying worker in the Cape honeybee. Apis mellifera capensis. lounial of Api- cultiiral Research 2: 85-92. Arillo, A., A. Nel, and V. M. Ortuno. 1996. Two fossil bees from the Oligocene of Izarra (Alava, Spain) (Hymenoptera, Apoidea). Bulletin tie la Societe en- tomologique de France 101: 59-64. Armbruster, L. 1938. Versteinerte Horugbienen aus dem obermiocanen Randecker Maar. Archiv fiir Bienenkumte 19: l-t8, 97-133. Ashmead, W. H. 1904. Remarks on honey bees. Pro- ceciiings of the Entomological Society of Washington 6: 120-122. Baldensperger, P. J. 1923. Sur I'abeille saharienne. 6"' Congrcs International d' Apiculture, Marseille 1922: 61-64. Baldensperger, P. J. 1928. Bees in their natural state and with beekeepers. Bee World 9: 173-174. Baldensperger, P. J. 1932. Varietes d'abeilles en Af- rique du Nord. 5"' Congres International d'Entomologie, Pans 1932: 829-839. Barac, I. 1977. Preserving the genetic pool of Apis mel- lifera carpatica. Proceedings of the International Bee- keeping Congress 26: 270-274. Barton, B. S. 1793. An inquiry into the question whether the Apis nwUifica, or true Honey-bee, is a native of America. Transactions of the Anurican Philosophical Society, Philadelphia 3: 251-261. Belknap, J. 1792. A Discourse Intended to Commemorate the Discovery of America by Christopher Columbus; Delivered at the Request of the Historical Society in Massachusetts, on the 23d Day of October, 1792, be- ing the Completion of the Third Century siiue that Me\norable Event. To zvhich are added Four Disser- tations, connected with various Parts of the Dis- course. Apollo, Belknap & Hall, Boston, Massa- chusetts, United States, 132 pp. Benton, F. 1904. The specific name of the common Volume 8, Number 2, 1999 191 honey-bee. Proceedings of Ihe Entoniolcigical Societi/ of Wasliinghm 6: 71-73. Bingham, C. T. 1898. On some new species of Indian Hymenoptera. journal of the Bombay Natural His- tory Society 12: 115-130. Broolc— •►d- Characters 7-8 g a— •-b— ►c— ►d Characters 17-18 g ' t t a— ► b— •► c— *- d— »► e f J Characters 21-24 f t a— *-b— ►c— •-d Characters 31-32 t a— ►b— »-c \ Characters 34-36 f i .i t a^4-b- -►c- \ \ e h Characters 42-44 a— *>b— #>c \ d Characters 48-49 Fig. 1. Diagrams of reticulate transformation series. -b— ►►c Characters 68-69 208 Journal of Hymenoptera Research Table ]. Matrix of Alciodcs species and character states. See Appendix for complete species names and text for explanation of characters. 01234567 ANCESTOR 0000000111000000000000100010000010010000001110000000010000001000000100000 ABDOMINA 021102031101000013 0002100011101132110001002110000021121000101000221100000 ACICULAT 0111010311000000010001000111000121010000001210020021011000001000210100000 AESTUOSU 021112031101000213 0012100010001212110001002110001121023000001000211100100 AGILIS 0111120311000002130013100110001211110001002110001121123000001000110000100 ALBITIBI 0111011311000000110002110011000111110000001110000011011010001000210100000 ALBOANNU 011112030101000003 0002100111701131110111002110001021121000101000210101000 ALUTACEU 011112031100000003 0001200111000111110010101110000021010000000000110100000 ALIGARHE 0111001211000000010001100011000111010000001110100011011000001000220100000 ALTERNAT 0111010311000000010001000111000121110000001210010021011000001000221100000 ANGUSTAT 0121011211100000010001000111000121110000001210010071010000001000210101000 ANNULATU 0101010211000?00030001100111000111010000001110000021010000000100?10100000 ANTENNAT 0111120311000010130002100010001111110111003110000021101000001000210100000 APICALIS 0211020311010000130012100011101132110001002110000021121000101000221100000 ARCTICUS 0111010311000000010001100011000111100000001310030021011010002000210100000 ARIZONIE 0211120311100010030002100010011111210010003110000021111000002000221100000 ARMATUS 0111011211000000010001100011001111110000001210000021010000001000210100000 ARNOLDI 0111120311010000130012100111701131110001002110000021022000001000210101000 ARSENJEV 0111120311010010030002100010011111110111103110000020001000001000710100000 ATRICEPS 0211020311010000130002100011101132110001002110000021122000001000221100000 ATRICORN 0111120311010000030012100111001141110001002110002021112000001000210101000 AUTOGRAP 0111011211000000010001100011000111110000001110000011011000011000210100000 BAKERI 1111011211010000020001100111000111110000001110000071011000001000210100000 BICOLOR 0111010311000000010001000111000121100000001310010021011010001000210100000 BOREALIS 0111011211000000010001100011000111110000001110000011011000011000210100000 BRETHESI 0211020311010000130002100011101132110001002110000021131000001000221100000 BREVIPEN 01110103110017000301013 00111000110100000111110000020000000000100110100000 BREVIRAD 0111020311010000030001000111001121100000001210000021001700001000710100000 BUOCULUS 0111011211000010010001100011000111110000001110000011011000001001210100000 BURRUS 0111120311010000130013100111001131100001002110001021122000001000210101000 CAMERONI 0211120301010010030012100010011111210000003110000021011000002000221100000 CANTHARI 0111011211000000010001100011000111110000001110000011017000011001710100000 CARINATU 0111120311010000130013100111001131110001002110001021112000001000221101000 CARINIVE 0211120311010000130013100011001131110001002120000021122010001000221100000 CAUCASIC 01111213110000001300121000110111311000010021100000210010000010002 7 7100000 CAZIERI 0211120311110010030002100010001131210000002110000021111000002000221100000 CHLOROTI 0111111301000010010002100110070111210000003110000021110000001001110100000 CIRCUMSC 0111011211000000010001100011000111110000001110000011011000011000210100000 COMPRESS 0221001211100000010001100011000111110000001110100011011000001000220100010 CONFORMI 0111120311010000130013100111001131110001002110001021122000001000210101000 CONVEXUS 0211020311010000130002100011101212110001002110000021121010001000221100001 COXALIS 0111010311000000010001000111000121100000001310010021011010001000210100000 COXATOR 0211010311010000010001000111000121100000001210010021011000001000221100000 CRASSIPE 0101011311170000010001100111000111110000001110010071011000001000710100000 CRUENTIS 021112031101000013 0013100110001311120001002120001021122000001000221101100 CURTISPI 0111010311000000010001200111000111110000001110000021011100000000 710100000 DENDROLI 0111011211000010010001100010000111110000001110000011011010001000210100000 DESERTUS 0111120311000002130012100110001211110001002110001171723000001001121000100 DISPAR 011102031100100003 01013 00111001111110010111110000021000000000100110100000 DISSECTO 021112031101000013 0012100010001311120001002120001021122010001000221101100 DIVERSIC 0111010311071000030001200111000121010010101110000021101000000100110100000 DIVERSUS 0211120311010000130012100010 701211110001002120001021022011001000221101100 DUCTOR 0211020311010000130002100011101132110001002110000021121000101000221100000 EARINOS 0111121311000010030012100110011131210000003110000021101000002000220100000 ECUADORI 0111120311010000030014100111001131110011002110000021102000001000210101000 EURINUS 0111120311010000130013100111001131110001002110001021122000101000210101000 EXCAVATU 0111010311001000030101300111001121110000111211030020000000001100110100000 FAHRINGE 0111020311010002130012100010001212110001002110002021123000001000210001100 FEMORATU 1111010211000000030001100111000111010000001110000011010000001000210100000 FERRUGIL 0111020311000000130012100111001141110001002110000021121000001000210001000 FLAVIDUS 0111121301170010030002100020001131210000003110000021111000001000211100000 FLAVISTI 0211020321010001130012100000001101130001002110002021132110001000221101100 Volume 8, Number 2, 1999 209 Table 1. Continuod. 01234567 FLAVITAR 011102031101000003 0002100011101112110001002100000021111000001000221100000 FORTIPES 0111121311070000030012100011011131110000003110000021011000002000210100000 FORTIS 0111120311010000130013100111001131110001002110001021122000001000210101000 FUSCIPEN 01110213110000000 10002110011000111 1100000011100000 71021000001000710100000 GASTERAT 0211120311010000130013100111001131110001002110001021122000101000211101000 GASTRITO 0111011211000100010001100011000111110000001110000011011000001001110100000 GEOMETRA 0211121301000010030012100010011111210000003110000021011000002000221100000 GLABER 0121720311010012130012100110001211120001002100001121023000002000110000100 GOSSYPII 0111011211000000010001100011000111110000001110000011011000001000210100000 GRANDIS 021112031101000003 0011000011101121100001002110000021111010001000221100000 GRANULAT 0111010311001000010001100111000111010000101210020011010000001100210100000 GRAPHICU 0111110311000010111002100001000101110000101110000021021000001000210100000 GRASSATO 011112031101000013 0013100111001131110001002110000021121000001070210101000 GRESSITT 0111001011000000010011100011000111111000001110000011012000001000210100000 HARRIMAN 0111020311010000130012100110001131110001002100001021121000101000210101000 HELLENIC 0111011311000000010001000111001121110000001217020021010000000000710100000 HIRTUS 0211020311010000130012100111001131110001002110001021121000101000210101000 INCERTOI 0111010311000000010001000111000121100000001210020071011000001000210100000 INCERTUS 0111010311000000010001000111000121110000001210010021111000001000210000000 INDESCRE 0211011311000010010001100011000111110000001110000011011000001000221100000 JAKOWLEW 011101031110000001000100011100012110001000131003 0021111000011000710100000 JAROSLAW 021112031101000013 0012100011700131110001002110000020121000101000221101000 KRULIKOW 0111020311010002130013100110001131120001002120001021103000101000210101100 KUZTLIZK 1111011211000000010001100111000112110000001210000011010000011000210100000 LAPHYGMA 0111011211000000010001100011000111110000001110000011011000001001210100000 LONGICOR 7111010311170000030101200111000111010010001110000021110000000100770100000 LONGIPEN 0111010311101000030101300111000120100000101211000020000000000100110100000 LUCIDUS 0211020321010001130012100000001101140001002100002031232110001000221201100 LYMANTRI 0211011211000010010001100010001111210000001110000011011000011000221100000 MALACOSO 0111011211000000010001100011000111110000001110000011011000001001110100000 MANDIBUL 0211120311010002130012100110001211110001002110002021123000001000221101100 MEDIANUS 0221020311010000030002100111001111110000002110000071021000001000221101000 MELANOPT 0211020321010001130012100000001101140001002100002021232110001000221201100 MEXICANU 021102031101000113 001210000000110113 0001002110002021132000001000221101100 MICROCUL 011112031101000003 00031001117 70310120001102110000021101010001000221101100 MINIATUS 012112031101000113 0013101111001311120000002120002021123000001000210001100 MODESTUS 0111010311000000010012100110000111120000002120001021022000001000210101000 MOLDAVIC 0101010311000000010001000111001121100000001310020021011000001000710100000 MOLESTUS 021102031101000013 0002100111101132110001002110000021121000001000221100000 MORIO 0111120311010000130012100111001311110001002110000021121010001000221101000 NARANGAE 0111010211001700010001000011000111110000001210020011000000000100110100000 NEGATIVU 0211010311000000010001000110001121110001101210000021110000001000221101000 NEOTROPI 7111011311170000010001100011000111110000001110000021011000001070710100000 NIGRIBAS 1111011211100010020001100111000111110000001110000071110000001000210100000 NIGRICOR 0111011211000000010001100011000111110000001110000011011000011000110100000 NIGRISTE 1111010211000000020001100111000111010000001110000011010000001000210100000 NOCTURNU 0121001211000010010011100010000111110000001110000071011000001001210100000 NOLOPHAN 0111011211000000010001100011000111110000001110000011011000001001110100000 NOTOZOPH 0211121301000010030002100010011131210000003110000021011000002000221100000 ^a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ournal of Hymenoitera Research Table 1. Continued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ll 0111010311000000010001000111000121100000001310010021011010001000210100000 NEW SP12 0111011211101010010001100110000111110000001210010011011000001000210100000 NEW SP13 0211010311000000010001000111000121100000001310010021011010001000210100000 NEW SP14 011101031100000001000110001100011110000010131003 0021011010001000210100000 NEW SP15 0111001211000100010001100011000111110000001210000011011000001011210100000 NEW SP16 0221001011000000010001100011000111111000001010000011011000001000111110000 NEW SP17 0111010311000000010001100011000111100000101310030021011010001000210100000 NEW SP. .1 NEW SP, ,2 NEW SP. .3 NEW SP. .4 NEW SP. .5 NEW SP. .6 Volume 8, Number 2, 1999 Table 1. Continued. 211 NEW NEW NEW NEW NEW NEW NEW NEW NEW NEW NEW NEW NEW NEW NEW NEW NEW NEW NEW NEW NEW SP18 SP19 SP20 SP21 SP22 SP23 SP24 SP25 SP26 SP27 SP28 SP29 SP30 SP31 SP32 SP33 SP34 SP35 SP36 SP37 SP38 02110112 01110012 01111203 01110203 02111203 02111203 01110112 01110203 02110103 01110103 01111203 01111104 11110112 01111203 01110112 11110112 02110103 01110012 01010103 02110103 01110112 1 11000 11000 11010 11000 11010 11010 11100 11010 12000 11001 11010 11000 11000 11010 11000 11000 11000 11000 11000 11000 10100 01001 00001 00013 00213 00213 00013 00001 00013 00001 00001 00003 00011 00003 00013 00001 00002 00001 00001 00001 01001 01001 00011 00011 00131 00121 00121 00121 00011 00121 00010 00010 00141 10021 00011 00121 00011 00011 00010 00011 00010 00011 00011 00010 00011 00111 00010 00100 00110 00011 00110 00111 00111 00111 00001 00111 00111 00011 00111 00111 00011 00111 00111 00011 00011 00011 00113 00121 00131 00121 00011 00113 00012 00012 00013 00010 00011 00113 00011 00011 00012 00011 00012 00111 00111 1110000 1110000 1110001 1110001 1110001 1110001 1010000 1110001 1110000 1100000 1110011 1110000 1010000 1110001 1110000 1010000 1100000 1110000 1100000 1110000 1110000 001110 001110 002110 002100 002110 002110 000110 002110 101210 001310 002110 101110 001110 002110 001110 001110 001210 001110 001210 001110 001110 00001 00001 00102 00112 00102 00102 11001 00102 01002 01002 00002 00002 00001 00102 00101 0000? 01002 00001 00002 0000? 00001 10110 10110 11220 10230 11330 11230 10110 11210 10110 10110 11120 10210 10100 11220 10110 10100 11110 10110 10110 10110 10110 000020 000010 000010 000010 000010 000010 000010 001010 000010 000010 000010 000010 000010 000010 000010 000010 000010 000010 000010 000010 000110 00221 11210 00210 00210 00221 00221 00220 00210 00221 00210 00210 00210 00210 00210 01210 00210 00221 11210 00210 00221 00210 100000 100000 101000 000100 101100 001100 100010 101000 100000 100000 101000 100000 100000 101000 101000 100000 100000 100000 100000 100000 100000 the analysis exceeded 128 taxa, a compre- hensive matrix (Table 1) was maintained from which 'sample matrices' were drawn for analysis of newly analyzed species within the context of a representative sam- pling of species analyzed to date. 'Species-group' definition refinement. — It was found that "species-groups" consis- tently clustered on cladograms derived from Nelson consensuses throughout the course of the study (see Results). Refine- ment of these species-group definitions was aided by use of sample matrices as follows. For a given species-group, all spe- cies previously found to be in the group were included in a sample matrix as de- scribed above, as well as most or all spe- cies in contiguous species-groups. Manual branch swapping was applied to the Nel- son consensus tree after running the ma- trix on Hennig86 (m*; bb*; xs w; cc) in or- der to further refine the definition of the species-group by further reducing the length of the tree. After definitions of all characters, char- acter states, transformation series, and species-groups had been finalized, a final matrix with an approximately equal per- centage of species from each species- group (about 61.5%) was analyzed using the mhennig*; option of the Hennig86 pro- gram, in order to find the most parsimo- nious tree (Farris 1988). This percentage was found to meet the requirement of ob- taining the tree with 128 taxa, with taxa most evenly distributed across it with re- spect to species-groups. Host association analysis. — After a clado- gram had been generated as described above, host associations were mapped onto the cladogram. Based on available lit- erature (Shenefelt 1975, Achterberg 1985, Marsh 1979, Shaw 1983, 1994; Shaw 1995, Shaw et al. 1997), and museum specimens with host information attached were ex- amined (Rocky Mountain Systematic En- tomology Museum, Laramie, WY and Ca- nadian National Collection, Ottawa, ON), host associations were assigned. CHARACTER SUMMARY Following is a list describing characters and evolutionary transformation series of those characters in this analysis, along with explanations of how polarities were hypothesized for each transformation se- ries. The number assigned to each char- acter identifies it on the matrix and on the 212 Journal of Hymenoitera Research cladogram illustrated by CLADOS. The numbers in brackets indicate character states within the transformation series for a given character, and also appear under- neath the numbers assigned to the respec- tive characters in the matrix, and on the CLADOS-generated cladograms for each species-group. Squares on cladogram branches below designated character numbers represent character state trans- formations which help define the node or species following the squares. If the square is darkened, the square represents an increase in character state number; if white, a reversal is indicated. Unless oth- erwise stated, the [0] state was that hy- pothesized to be ancestral based on Cli- noceutrus, Rogas and Stiropius specimens examined. I. Head Characters 4. Shape of 15th flagellomere from base, fe- males: [0] width/length less than 0.83; [1] compact: width /length greater than 0.83. State [1] was found only in some Aleiodes species. Thus, state [0] was hypothesized to be the ancestral state. 6. Occipital cari)ia: [0] strong, complete medially; [1] weak or effaced medially. State [1] was found in most Stiropius and no Cliuoceutriis or Rogas examined, and was hypothesized to be ancestral. This hy- pothesis was tested by comparing relative parsimony with respect to character po- larization. The same matrix was run on Hennig86 as described above for two runs except that polarization was switched for each. Since the alternate polarity was most parsimonious (strong, complete occipital carina ancestral), this alternate hypothesis was accepted; the RogaslClinocentrus con- dition was hypothesized to be ancestral. 13. Flagellum, female: [0] > 33 flagello- meres; [1] s 33 flagellomeres. All Stiropius species examined expressed state [1], and no Clinocentrus or Rogas species did. The more parsimonious character polarization was found by running each of the two possibilities on Hennig86. Based on par- simony, the reduced number of flagello- meres was hypothesized to be derived where it occurred in Aleiodes. 14. Ocellar diameter/distance between lat- eral ocellus and compound eye: [0] < 2.5; [1] > 2.5 (Shaw et al. 1997, Figs. 1-5). State [1] was observed in all Rogas species exam- ined, and no Stiropius. A comparison of parsimony in which the polarity of this character was switched between two oth- erwise identical matrices showed that the polarization used here was most parsi- monious. Thus, state [0] was hypothesized to be the ancestral state. 15. Ratio of clypeo-antennal space to width of oral space: [0] greater than 0.82; [1] 0.69- 0.82; [2] less than 0.69. Since states [1] and [2] were not found in non-Aleiodes species examined, state [0] was hypothesized to be ancestral. 20. Occipital carina at hypostomal carina: [0] complete or nearly so; [1] incomplete, effaced well before reaching hypostomal carina. Determining polarity was prob- lematic, since of four Rogas species exam- ined, two expressed state [0], and two ex- pressed [1]. The Clinocentrus species ex- amined (four unnamed morphospecies) consistently exhibited state [0]. Two Sti- ropius species examined consistently pos- sessed state [1]. Both codings were tried for the ancestor, with no change in the consistency index (c.i.). The above polari- zation represents the most parsimonious solution. Thus, state [0] was hypothesized to be the ancestral state. 27. Ratio of malar space to mandibular bas- al width: [0] less than 1; [1] greater than or equal to 1. State [0] was found in all Rogas and Cli)iocentrus species examined. State [1] was observed in all Stiropius examined. A comparison of parsimony between al- ternate polarizations revealed that the po- larization presented here is the simplest explanation. 31. Vertex sculpturing (path a-d): [0] coarsely granulate; [1] smooth, granulate, or faintly rugulose (Shaw ef al. 1997, Fig. 6); [2] smooth with shallow, dense punc- Volume 8, Number 2, 1999 213 tation; [3] shiny, coarsely and densely punctate. 32. Vertex sculpturing {path e-h): [0] nit- id; [1] one of conditions in #31; [2] with strong laterally running ridges, or rugose; [3] with fine laterally running ridges (gen- eralized advanced condition; Shaw et al. 1997, Fig. 5); [4] finely rugulose. The retic- ulate character described by the above 2 paths is hypothesized as follows: a = 31[0]; b = 31[1]; c = 31[2]; d = 31[3]; e = 32[0]; f = 32[2]; g = 32[3]; h = 32[04]. a b c d path n 0 1 2 1 a-c (#48) n + 1 0 0 0 1 b-d (#49) a b c d e f g h path n 0 1 2 3 1 1 1 1 a-d (#31) n + 1 1 1 1 1 0 2 3 4 e-h (#32) The above phylogenetic hypothesis is il- lustrated in Fig. Id. Stiropius species ob- served all expressed state (a), and Rogns species all expressed state (e). A compar- ison of parsimony between alternate po- larizations revealed that the polarization presented here is the simplest explanation. 33. Medial carina extending down frons: [0] absent; [1] present, extending less than 0.55 of distance from line between bases of scapes to clypeus; [2] present, extending 0.55 or more of distance from line between bases of scapes to clypeus. States [1] and [2] were observed in no non-Aleiodes spe- cies examined. Thus, state [0] was hypoth- esized to be the ancestral state. 38. Lightnesf of midsection of antennae: [0] not lighter than both basal and apical sec- tions; [1] lighter than both basal and apical sections (Shaw et al. 1997, Fig. 23). State [1] was not found in non-Aleiodes species examined. Thus, state [0] was hypothe- sized to be the ancestral state. 48. Clypeal carina (path a~-c): [0] absent; [1] present, not shelflike; [2] shelflike. 49. Clypeal carina (path b-d): [0] clypeus either not elongate, or with a carina; [1] elongate clypeus without carina. The retic- ulate character described by the above 2 paths is hypothesized as follows: a = 49[0]; b = 49[1]; c = 49[2]; d = 50[1] The above phylogenetic hypothesis is il- lustrated in Fig. Ig. States [1] and [2] were observed in no non-Aleiodes examined. Thus, state [0] was hypothesized to be the ancestral state. 51. Length of first flagellomere: [0] s scape length; [1] < scape length. State [1] was not observed in non-Aleiodes species examined. Thus, state [0] was hypothe- sized to be the ancestral state. 54. Cli/peus height divided by icidth: [0] > 0.65; [1] < 0.65 > 0.42; [2] <\42 > .30; [3] ^ 0.30. Height was defined as distance be- tween clypeo-labral suture and dorsal edge of clypeus. Width was defined as distance between inner edges of tentorial pits. All Stiropius and Rogas examined ex- pressed state [0]. Clinocentrus expressed state [1]; However, given that Stiropius and Rogas are phylogenetically closer to Aleiodes than Clinocentrus (Whitfield 1992), the elongated clypeal condition in Clino- centrus was hypothesized to be indepen- dently derived with respect to the Aleiodes condition, and state [0] was hypothesized to be ancestral. 55. Third maxillary palpomere: [0] not swollen; [1] swollen. State [1] was not ob- served in non-Aleiodes species examined. Thus, state [0] was hypothesized to be the ancestral state. 69. Clypeus shape: [0] not abruptly edged, not flat ventrad (ancestor); [1] abruptly edged, flat ventrad. State [1] was not observed in non-Aleiodes species ex- amined. Thus, state [0] was hypothesized to be the ancestral state. 70. Wide, flat flange on lower side of man- dible: [0] not strongly present; [1] strongly present. State [1] was not observed in non- Aleiodes species examined. Thus, state [0] was hypothesized to be the ancestral state. 214 Jouy 0.28 of pronotal length. State [1] was not found in non-Al- eiodes species examined. Thus, state [0] was hypothesized to be the ancestral state. 50. Postero-dorsal surface of scutum: [0] mostly nitid, sometimes weakly rugulose; [1] weakly rugulose, not shiny; [2] at least some strong costae; costae smooth; [3] smooth, not nitid. State [0] was observed in Clinocentrus, Rogas, and Stiropius species examined, but in no Aleiodes, and was hy- pothesized to be the ancestral condition. 52. Notauli at mid-dorsal surface of scu- tum: [0] without carinae or foveae or finely foveate, non-carinate; [1] coarsely foveate, and /or usually carinate posteriad; [2] smooth, not nitid. States [1] and [2] were not found in any Rogas or Stiropius exam- ined. However, the coarsely foveate con- dition was ubiquitous in Clinocentrus. Giv- en the above, it was hypothesized that state [0] was ancestral, and that the Cli- nocentrus condition was derived in parallel with the state's appearance in Aleiodes. 53. Scutal sculpturing excluding postero- dorsal surface and notauli: [0] rugulose sculpturing over a nitid surface; [1] gran- ulate; [2] punctate; [3] nitid or smooth, not nitid. States [1], [2], and [3] were not ex- pressed by non-Aleiodes species examined. Stiropius expressed the coarsely granulate condition, Rogas expressed rugulose sculpturing over a nitid surface, and Cli- nocoitrus expressed a nitid-punctate scutal surface. Since Rogas is considered the sis- ter group of Aleiodes (Whitfield 1992), state [0] was hypothesized to be ancestral. 61. Propodeum dorsal profile: [0] rounded; [1] flattened. State [1] was not observed in non-i4/t';o(ft's species examined. State [0] was hypothesized to be ancestral. 63. Ratio of pronotal medial length to head length: [0] >'0.30; [1] < 0.30. Head length was defined as distance from the occipital carina at the vertex to the line between posterior edges of antennal sutures. State [1] was observed in Stiropius and some Al- eiodes, but not in Rogas or Clinocentrus. Given the very small ocelli of Stiropius, such that the ocellocular distance /ocellar diameter ratio was consistently much greater than one, the vertex broad in com- parison with the three other genera, and the antennal flagellomeres over twice as long as wide in contrast with the three other genera, it was hypothesized that the Stiropius condition was independently de- rived. Thus, it was hypothesized that state [0] was the ancestral condition. III. Metasomal Characters 2. Median carina of tergite II: [0] undif- ferentiated from other coarse carinae (Achterberg 1991, Fig. 124); [1] differenti- ated from other carinae (Achterberg 1991, Fig. 377); [2] undifferentiated from other fine carinae. State [0] is the Stiropius/Rogas condition. Clinocentrus shared [1] with some Aleiodes. Since Clinocentrus is consid- ered most distantly related to Aleiodes of the three outgroup genera (Whitfield 216 Journal of Hymenoptera Research 1992), it was hypothesized that state [1] was independently derived in Clinocentrus and Aleiodes. 5. Median triangle of tergite U: [0] absent; [1] small and without well developed an- terior carinae (Achterberg 1991, Fig. 290); [2] large and associated with anterior ca- rinae which run laterally nearly to mar- gins of tergite (Achterberg 1991, Fig. 300). The presence of a median triangle differ- entiates most Aleiodes with respect to Ro- gas and Stiropiiis. Thus, state [0] was hy- pothesized to be the ancestral state. 10. Length of metasomal tergite I, female: [0] tergite apical width /tergite length greater than or equal to 0.87 (not elon- gate); [1] tergite apical width/tergite length less than 0.87 (elongate). An elon- gate first metasomal tergite was not pre- sent in Stiropius, and only occasionally present in Clinocentrus and Rogas. Thus, its occasional presence in Aleiodes, Clinocen- trus, and Rogas was hypothesized to be in- dependently derived. 28. Setal mat on metasomal tergites 4-6 of males: [0] absent; [1] present (Shaw et al. 1998, Fig. 1). State [1] was expressed by no non-Aleiodes species examined. Thus, state [0] was hypothesized to be the ancestral state. 29. Medial pits on tergites 4-7, males: [0] absent; [1] present (Shaw et al. 1997, Figs. 13, 15, 17) State [1] was expressed by no non-Aleiodes species examined. Thus, state [0] was hypothesized to be the ancestral state. 34. Sculpturing of metasomal tergite I (path a-c): [0] weakly or strongly costate, costae relatively widely spaced, often over weak granulation; [1] rugulocostate or rugocostate; or costate, costae narrowly spaced, no underlying sculpturing; [2] finely rugulose to finely rugulocostate (Shaw et al. 1997, Figs. 11, 12). 35. Sculpturing of metasomal tergite I (path e-fl: [0] rugose or rugulose; [1] one of conditions of #34; [2] smoothly rugose; [3] strongly costate, costae narrowly spaced, no underlying sculpturing; [4] weakly costate, costae narrowly spaced, no underlying sculpturing, or nitid. 36. Scidpturing of metasomal tergite I (path b-h): [0] not smoothly, finely granu- late; [1] smoothly, finely granulate. The re- ticulate character described by the above two paths is hypothesized as follows: a = 34[0]; b = 34[1]; c = 34[2]; d = 35[3]; e = 35[0]; f = 35[4]; g = 35[2]; h = 36[1]. abcde fgh path n 0 12 1 1 1 1 1 a-c (#34) n -f- 1 1 1 1 3 0 4 2 1 e-f(#35) n+2 0 0 0 0 0 0 0 1 b-h (#36) The above phylogenetic hypothesis is il- lustrated in Fig. le. State (a) was observed in all Clinocentrus and Rogas examined, and in some Aleiodes species. The coarsely granulate Stiropius condition was thus hy- pothesized to be a derived synapomorphy for that group, and (a) was hypothesized to be ancestral. 42. Metasomal tergite III sculpturing (path a-d): [0] completely, finely smooth-acicu- late; [1] weakly to strongly rugulocostate or weakly costate over granulate back- ground; costae not smooth, tergite not shiny; [2] smooth-aciculate or smoothly rugulose anteriad, nitid or smooth-punc- tate posteriad, or entirely nitid or smooth punctate; [3] shallowly rugulose or rugu- lopunctate anteriad, finely punctate pos- teriad or completely finely punctate. 43. Metasomal tergite III sculpturing (path ^^S^' [0] smooth, faintly granulate, shiny; [1] one of states of #42; [2] entirely cari- nate; [3] strongly rugose. 44. Metasomal tergite III sculpturing (path h-i): [0] completely smooth, shiny, no trace of granulation; [1] one of conditions of #42; [2] densely punctate or rugulo- punctate anteriad; shiny posteriad, or en- tirely densely punctate. The reticulate character described by the above 3 paths is hypothesized as follows: a = 42[0]; b = 42[1]; c = 42[2] d = 42[3]; e = 43[0] f = 43[2]; g = 43[3]; h = 44[0]; i = 44[2]. Volume 8, Number 2, 1999 217 a b c d e f g h i path n 0 12 3 11 1 2 2 a-d(#42) n+ 1 1 1 1 1 0 2 3 1 1 e-g(#43) n + 2 1 1 1 1 1 1 1 0 2 h-i (#44) The ancestral state is (b). The above phylogenetic hypothesis is illustrated in Fig. If. Since state (a) was ubiquitous in all Rogns species examined and in many Aleiodes, it was hypothesized to be ances- tral. 45. Excavated medial areas on metasomal tergites II and III: [0] absent; [1] present (Achterberg 1985, Fig. 12). State [1] was not found in non-Aleiodes species exam- ined. Thus, state [0] was hypothesized to be the ancestral state. 46. Apex of abdomen, females: [0] not compressed; [1] compressed. State [1] was not found in non- Aleiodes species exam- ined. Thus, state [0] was hypothesized to be the ancestral state. 47. Fourth metasomal tergite (MT IV): [0] MT IV unsculptured; [1]'mT IV slightly sculptured; [2] MT IV mostly sculptured, not heavily rugulose; [3] MT IV heavily rugulose, but not a complete carapace over apicad tergites; [4] MT IV a heavily rugulose, complete carapace over apicad tergites. States [I] through [4] were found in no non-Aleiodes species examined. Thus, state [0] was hypothesized to be the an- cestral state. 56. Metasomal dorsum color: [0] not en- tirely black; [1] entirely black. State [1] was observed in some unicolored Stiropius and Clinocentrus species, and in some Al- eiodes species. Since all Rogas species ob- served were uniformly yellow, and many Aleiodes expressing state [1] had heads and /or mesosoma that were not black, it was hypothesized that the Stiropius and Clinocentrus conditions were independent- ly derived from the condition in Aleiodes, and that state [0] was ancestral. 57. Metapleural pit: [0] not slightly larger than propodeal spiracle, deep and round; [1] slightly larger than propodeal spiracle, deep and round. State [1] was not ob- served in non-Aleiodes species examined. Thus, state [0] was hypothesized to be the ancestral state. 58. Head/mesosoma/apical metasomal col- oration (metasomal tergite III and posteriad; females): [0] not all black with bicolored metasoma; [1] all black with bicolored me- tasoma. State [1] was not observed in non- Aleiodes species examined. Thus, state [0] was hypothesized to be the ancestral state. 59. Metasomal tergite II coloration: [0] not consistently black laterally, yellow medi- ally; [1] consistently black laterally, yellow medially. State [1] was not observed in non-Aleiodes species examined. Thus, state [0] was hypothesized to be the ancestral state. 62. Ovipositor length: [0] less than Vi length of metafemur; [1] greater than or equal to V2 length of metafemur. All Sti- ropius species examined had short ovipos- itors, and most Rogas species and all Cli- nocentrus species examined had long ovi- positors. Since most Aleiodes species ex- amined had short ovipositors, short ovipositor length was hypothesized to be the ancestral condition. 71 . Lateral edges of metasomal tergite I: [0] not parallel; [1] parallel. State [1] was not observed in non-Aleiodes species exam- ined. Thus, state [0] was hypothesized to be the ancestral state. 72. Metasomal tergite III: [0] not cara- pace-like; [1] carapace-like, extending over apicad tergites. State [1] was not observed in non-Aleiodes species examined. Thus, state [0] was hypothesized to be the an- cestral state. IV. Wing Characters 9. Second suhmarginal cell: [0] long, nar- row, vein 2RS/vein 2-M less than 0.4; 2RS parallel or nearly so with r-m; [1] trape- zoidal or rectangular, defining veins often arched, 2RS/2-M greater than 0.45 and less than 0.71; [2] quadrate; 2RS parallel or nearly so with r-m, and 2RS/2-M greater 218 Journal of Hymenoptera Research than 0.71. Since the second submarginal cells of Cliiwcentru:^, Rogas, and Stiropius species examined all expressed state [1], and since the ordering described above is most logical, it was hypothesized that this is a bipolar character 17. RS vein of hind wing (path a-ci): [0] marginal cell widest basally, RS vein not recurved apically (Achterberg 1991, Fig. 54); [1] marginal cell narrowest in middle; RS vein moderately to slightly recurved apically, never nearly touching wing mar- gin (Achterberg 1991, Fig. 278); [2] RS vein parallel with costal wing margin in basal 1/6-%, abruptly angling posteriad apically; [3] marginal cell not narrowest in middle; RS vein not recurved or angling; straight or nearly (Achterberg 1991, Fig. 291). 18. RS vein of hind wing (path b-e): [0] not extremely narrow in middle; vein not nearly touching wing margin; [1] extreme- ly narrow in middle, vein nearly touching wing margin (Achterberg 1993a, plate 31). The reticulate character described by the above two paths is hypothesized as fol- lows: a = 17[0]; b = 17[1]; c = 17[2]; d = 17[3]; e = 18[1]. a b c d e path n 0 1 2 3 1 a-d (#17) n -H 1 0 0 0 0 1 b-e (#18) State (a) was observed in Clinocentnis, Rogas, and Stiropius species examined, but not in any Aleiodes. Thus, it was hypothe- sized to be the ancestral state. State (b) was hypothesized to be the most basal state for this character within Aleiodes, since it represents only one modification from the ancestral state: the recurvation of the apical section of the RS vein. States (c) and (d) were hypothesized to represent a progressive straightening of the RS vein from the original curved condition. State (e) was hypothesized to represent a bifur- cation in the evolution of this character, such that the (b) condition evolved in some Aleiodes lineages into the straight- ened (c) and (d) conditions, and on the other hand into the extremely recurved (e) condition in another. The above phyloge- netic hypothesis is illustrated in Fig. lb. 26. Wing coloration: [0] fumate; [1] hya- line; [2] patterned. State [1] was universal for all Clinocentnis, Rogas, and Stiropius species examined, and most Aleiodes spe- cies. States [0] and [2] were not observed in any non-Aleiodes species examined. Giv- en the ubiquity of state [1], were non-hy- aline wings to occur in non-Aleiodes spe- cies, they would almost certainly represent a parallel derived condition. Thus, state [1] was hypothesized to be the ancestral state. 30. Hindwing vein lA: [0] not extending past vein cu-a; [1] extending past vein cu- a. State [1] was found in no non-Aleiodes species examined. Thus, state [0] was hy- pothesized to be the ancestral state. 60. Fore wing width divided by length: [0] less than 0.29 (long, narrow); [i] 0.29-0.34; [2] greater than or equal to 0.35 (broad). States [0] and [2] were not observed in non-Aleiodes species examined. State [1] was hypothesized to be ancestral; thus, the character was hypothesized to be bipolar. V. Leg Characters 0. Inner apex of hind tibia: [0] with setae normal and unmodified; [1] with setae flattened coming to a point apically, form- ing an apical fringe (Achterberg 1991, Fig. 122). State [1] is the norm for Rogas, but does not occur in any Stiropius or Clino- centrus species so far as known (Shaw 1993). It was considered to be derived in Rogas and in any Aleiodes in which it oc- curred. Thus, state [0] was hypothesized to be the ancestral state. 1. Tarsal claiv: [0] either no tarsal teeth or if present, preapical lobe present (Ach- terberg 1991, Figs. 123, 322); [1] no preap- ical lobe, with basal tarsal teeth only; [2] no preapical lobe, tarsal teeth extending almost to apical tooth, or tarsi completely pectinate (Fig. 2c; Shaw et al. 1997, Figs. 20-22). Since the presence of tarsal teeth with no preapical lobe is an autapomor- Volume 8, Number 2, 1999 219 ^c S.A. melanoplenis ^^^^^Kf^mnMSJ^ [<' 1^' ^''Smator | Fig. 2. Tarsal claws, a, "new species 21;" b, A. nlbitihui; c, ,4. iiichiin'plcnis; d, A. ^tignmtor. phy for genus Aleiodes distinguishing it from Clinoccntnis, Rogas, and Stiropins, state [0] in part defines the hypothetical ancestor. Thus, state [0] was hypothesized to be the ancestral state for these four gen- era, and state [1] was hypothesized to be the ancestral state for Aleiodes. 11. Tarsal segment 4, female: [0] elongate: length > 1.6 times apical width; [1] com- pact: length < 1.6 times width. State [1] was not observed in non-Aleiodes species examined. Thus, state [0] was hypothe- sized to be the ancestral state. 37. Lightness of hind tarsus: [0] not con- siderably lighter than tibia; [1] consider- ably lighter than tibia (Shaw et al. 1997, Fig. 24). State [1] was not found in non- Aleiodes species examined. Thus, state [0] was hypothesized to be the ancestral state. 64. Number of tarsal ehnv teeth: [0] none; [1] 5 or less; [2] more than 5. State [0] was shared by some Rogas, all Clinocentrus and Stiropius, so far as known, and was hy- pothesized to be the ancestral state. 65. Shape of apical tarsal tooth: [0] not present; [1] incompletely toothed; [2] com- pletely toothed (Fig. 2c; Shaw et al. 1997, Figs. 19-22). The incomplete form was reminiscent of a thickened seta, with a hair-like, flexible apex. State [0] was hy- pothesized to be ancestral, for reasons stated for character 64. 66. Setae between apical tarsal tooth and clazo: [0] present (Fig. 2a, b); [1] absent (Fig. 2c, d). Several Aleiodes species which expressed incompletely toothed apical tar- sal teeth, also expressed basal tarsal teeth only. In acidition, they expressed a gra- dation of thickened, prominent setae which became progressively thinner with 220 Journal of Hymenoptera Research increasing apical position apicad of the apical tooth. Thus, it was hypothesized that state [1] represented the completion of a transformation of apical setae into teeth, and thus that state [0] was ancestral. 67. Shape of apical tarsal claw (path a-c): [0] nearly without a hook (Fig. 2a); [1] not abruptly hooked (Fig. 2b); [2] abruptly hooked with straight shank which angles at nearly 90 degrees (Fig. 2c). 68. Shape of apical tarsal claw (path b-d): [0] one of above states; [1] abruptly hooked with straight shank which angles at less than 90 degrees (Fig. 2d). The retic- ulate character described by the above two paths is hypothesized as follows: a = 67[0]; b = 67[1]; c = 67[2]; d = 68[1]. abed path n 0 1 2 1 a-c (#67) n + 1 0 0 0 1 b-d (#68) State (b) was universal in all non-Aleio- des species examined, and most Aleiodes species. Thus, it was hypothesized to be the ancestral state. The above phylogenet- ic hypothesis is illustrated in Fig. Ih. RESULTS Throughout the development of this analysis, species-groups emerged and de- fined themselves by consistently cluster- ing on the most parsimonious consensus tree. Once all species had been analyzed, 18 well-defined species-groups had emerged. Figs. 6 to 8 show the major spe- cies-groups within genus Aleiodes. Fig. 5 shows the overall phylogenetic pattern of the genus, with species-groups represent- ed by exemplar species. Initially assigning equal weight to all characters resulted in one most parsimo- nious tree using the mhennig* option. This tree had a length of 475 steps, a consisten- cy index (c.i.) of .25, and a retention index (r.i.) of .78. Extended branch swapping (bb* option) resulted in 755 equally par- simonious trees all with lengths of 475 steps. The final solution for successive itera- tions to character weighting applied to the trees before the Nelson consensus was run resulted in 755 most parsimonious trees, each with length of 624 steps, a c.i. of .56, and an r.i. of .91. The modified Nelson consensus tree of the final solution for the overall unweighted analysis, after manual branch swapping, is shown in Fig. 5. The Nelson consensus tree length of the un- weighted tree loaded into CLADOS was reduced by 140 steps by manual branch swapping. The genus can be divided into three ma- jor sections: basal, intermediate, and de- rived species. Basal species are defined as those in which 1) the occipital carina is usually weak or effaced medially, 2) hind- wing RS vein is recurved, 3) hind-wing vein lA usually does not extend past vein cu-a, 4) propodeal sculpturing is coarsely granular, or rugulocostate over a granu- late surface, 5) mesopleuron is granulate, 6) mesopleural pit posteriad to mesopleu- ron central disk is absent, and 7) median triangle of tergite II is small or absent and without well developed anterior carinae. In the overall cladogram (Fig. 3), groups 1-6 are basal groups. Intermediate species are defined as those with the above characteristics except 1) the occipital carina is complete medi- ally, 2) propodeal sculpturing is rugose, 3) the hind-wing RS vein may be straight, and 4) mesopleuron may be rugulose (groups 7-9). Derived species are defined as those with the characteristics for intermediate species except 1) mesopleuron sculpturing is smooth punctate or smooth, 2) the hind- wing RS vein is almost always straight, 3) third metasomal tergite is either smooth- aciculate or smoothly rugulose or rugulo- costate anteriad and either smooth-shiny or smoothly rugulose posteriad, or entire- ly smooth or entirely punctate 4) hind- wing vein lA extends past vein cu-a, and 5) the antero-medial triangle of metasomal tergite II is large and associated with an- Volume 8, Number 2, 1999 221 terior carinae which run laterally. There is an increasing tendency with derived po- sition on the cladogram for the mesopleur- al pit posteriad to mesopleuron central disk to be present (groups 10-18). Species-group definitions Some species in unresolved clusters, or which did not clearly fall within any clus- ter, were not placed within any species- group. These species either were in the basal section, or fell at the base of a well- defined species-group. Such species were A. armatus, A. arnoldi, A. cariniz'entris, A. ferrugileti, A. jawslawensis, A. medianus, A. modestus, A. pellucens, A. quadrum, A. ros- sicus, A. ruficeps, A. sibiricus, A. iingularis, and new species 7 and 9. Species included below which will be published later by Shaw et al. are desig- nated as new species (number). This text does not constitute a publication of any new species. 1. The gressitti species-group (Fig. 4a) consists of A. gressitti (Muesebeck) and new species 16. This is the basal-most group in the genus, and is defined by propodeum and metasomal tergite I finely, smooth-shiny granulate. 2. The compressor species-group (Fig. 4a) consists of known species A. ali- garhensi (Quadri), A. compressor (Her- rich-Schaffer), A. palmatus (Walley), and new species 24. The group is monophyletic, and defined by the fol- lowing synapomorphies: 1) apex of fe- male abdomen compressed, and 2) apical tarsal tooth complete. 3. The gastritor species-group (Fig. 4b) is a large monophyletic basal group de- fined by pronotal medial length :£ .30% of head length. It consists of A. biioculus Marsh, A. cantharius (Lyle), A. dendrolimi (Matsumura), A. gastritor (Thunberg), A. laphygmae (Viereck), A. malacosomatos (Mason), A. noctuniiis Telenga, A. iiolophatiae (Ashmead), A. pallescens Hellen, A. perplexus (Gahan), A. platypterygis (Ashmead), A. similli- miis (Ashmead), A. stigmator (Say), A. testaceus (Telenga), A. xanthus (Mar- shall), and new species 3, 15, 32, and 35. 4. The circumscriptus species-group (Fig. 6c) is a basal, monophyletic group de- fined by metasomal tergite II consis- tently yellow medially and black lat- erally. Species within the group are: A. aiitographae (Viereck), A. borealis Thomson, A. circumscriptus (Nees von Esenbeck), A. nigricornis Wesmael, A. pictiis (Herrich-Schaffer), A. similis (Curtis), A. scrutator (Say), and new species 38. 5. The pallidator species-group (Fig. 4d), also basal and monophyletic, is de- fined by eye diameter greater than 2.5X ocellocular space, and tarsal claws pectinate. The species in this group are A. indiscretus (Reardon), A. lymantriae (Watanabe), A. martini Marsh and Shaw, A. pallidator (Thun- berg) and new species 37. 6. The seriatus species-group (Marsh et al. 1998; Fig. 4e) is basal and mono- phyletic and is defined by hind tibial apex with flattened setae forming a fringe. The group consists of A. bakeri (Brues), A. femoratus Cresson, A. kus- litzkyi Tobias, A. nigribasis (Enderlein), A. nigristemmaticum (Enderlein), A. sanctivincentensis (Shenefelt), A. percur- rens (Lyle), A. scriptus (Enderlein), A. seriatus (Herrich-Schaffer), and new species 30 and 33. 7. The procerus species-group (Fig. 5a), which is monophyletic, consists of A. angustatus (Papp), A. crassipes Telenga, A. granulatus (De Gant), A. narangae (Rohwer), and A. procerus Wesmael, and new species 1. It is closely related to the A. dispar species-group. The group shares the following synapo- morphies: 1) metasomal tergite 4 slightly sculptured, and 2) metasomal tergite 1 elongate. All except A. cras- sipes and A. angustatus share a long, shelflike pronotum. 222 Journal of Hymenoptera Research ANCESTOR ^^^^f '6 A. gressilli species-eroiin PALMATUS N -ALIGARHE Ce H No K, — PELLUCEN — ROSSICUS D G L Ly N Py NEWSP 35 NEWSP 19 r- TESTACEU ">— XANTHUS — DENDROLI L CANTHARI G I_Hmalacoso l NEWSP 3 N STIGMATO G L Ly N GASTRITO C d G I No PALLESCE N No I— NEWSP 7 KUZLITZK PERCURRE N Nc Ur— SERIATUS A Uo N I L FEMORATU NEWSP 30 — SCRIPTUS — NIGRIBAS NIGRICOR G N PY NEWSP 36 G A. compix.ssor spetics-groiip "— 1-GAST LpALL A. };a.\thior species-tjroup _ A. seriatus spccics-giuiip -AUTOGRAP N -CIRCUMSC B G Ly N No Ps Py t -SIMILIS PALLIDA! G GE LY N PS T NEWSP 18 LYMANTRI Ly NEWSP 9 - ARMATUS G N PROCERUS CRASSIPE NEWSP 12 NEWSP 1 ANGUSTAT NEOTROPI PUNCTIPE Host Legend Arctiidae Bombycidae Catocallnae Choreutidae Drepanidae Gelechlidae Geometrldae Incurvarildae Lasiocampidae Limacodldae Lycaenidae Lymantrildae Noctuldae Notodontidae Nymphalidae Psychidae Pterophoridae Pyralidae Sphingidae Tortrlcidae Zygaenldae ■ A. circuniscrijiiiii spccics-gioup - A. iHillidaior spccics-gioui) L- A. procerus species-gioup CURTISPI - PLEURILI - ANNUUTU - TRANSVER -ROGEZENS -A. dispor species-tfoup I DISPAR 'l-TATIANAE r— RUGOSICO •— BREVIPEN A B Nc C D Ge G I L Li Lc Ly N No Ny Ps P Py S T Z Fig. 3. Cladogram showing overall phylogenetic pattern of Akwdes. Species-groups are represented by ex- emplar species. Hosts and respective codes are listed in legend. Host associations for Aleioiies species are represented by codes next to species. The coxalis species-group (Fig. 5b) is an intermediate monophyletic group defined by: 1) rugulose to rugose me- sopleuron, and 2) vertex with widely separated laterally running ridges. The following species fall in this group: A. aciculatus Cresson, A. alter- nator (Nees von Esenbeck), A. arcticus Thomson, A. bicolor (Spinola), A. cox- alis (Spinola), A. coxator Telenga, A. helleniciis Papp, A. incertoides Telenga, A. incertus Kokoujev, A. jakolezvi Ko- koujev, A. mohiavicus Tobias, A. iwga- tivus Tobias, A. nunbergi Noskiewicz, Volume 8, Number 2, 1999 223 -{l NEWSP 5 r— NEWSP 36 NEGATIVU N L Ly {' HELLENIC 1 TRISTIS L_ SATANAS "NEWSP 10 Lc ■ ALTERNAT A r- INCERTOI -NEWSP 2 Ly - NEWSP 13 ~\- BICOLOR O Lc Ly N Ny P Z COXALIS ■ MOLDAVIC — JAKOWLEW - TETRASPH - NEWSP U G -ARCTICUS - FUSCIPEN -PRAETOR S TEXANUS S GRAPHICU S UNGULARI MEDIANUS CAUCASIC p PULCHRIP L Ly N No I — CA2IERI ■ PEDALIS - ANTENNAT -VAUGHANI N CAMERONI No GEOMETRA G ROSSI RILEY! Ly N Py GRANDIS N FLAVITAR N P MOLESTUS N -ABDOMINA N ' — ATRICEPS CONVEXUS pPR> M L U I— SIBIRICU FERRUGIL r TERMINAL N I ^ HARRIMAN _ L UNIPUNCT GRASSATO r-TURKESTA EURINUS NEWSP 31 — PALLIDIS -I| ATRICORN N I NEWSP 28 N BURRUS N _ NEWSP 4 -c A . coxalls species-group ,) albillhia species-group .'). praetor species-group A. pulchripes species-group A. ductor species-group A schirjajewi species-group A . gaslerator species-group RUFIPES Fig. 3. Continued. 224 Journal of Hymenoitera Research UFEI N r- MORIO N r-RUGULOSU N ■DIVERSUS N DISSECTO N -■ A. ufei species-group ''\^ A. rugulo.sus species-group ^ ■ MICROCUL ^ SHESTAKO NEWSP 23 h: NEWSP 22 -FLAVISTI MELANOPT LUCIDUS MINIATUS I FAHRINGE c ^ — DESERTUS AGILIS AESTUOSU N NEWSP 21 GLABER Fig. 3. Continued. — A. mekmopterus species-group A. sanctihyacinthi (Provancher), A. sa- tauus Telenga, A. signatus (Nees von Esenbeck), subgenus Tetrasphaeropyx Ashmead; A. tristis Wesmael, and new species 2, 5, 10, 11, 13, 14, 17, 26, 27, 34, and 36. 9. The dispar species-group (Fig. 5c), a monophyletic group, consists of A. aleutacens Granger, A. annulatus Granger, A. hrevipeiidulatits Achter- berg, A. breviradialis Granger, A. cur- tispina Granger, A. dispar (Curtis), A. diversicornis Granger, A. excavatus (Te- lenga), A. longicornis Granger, A. lon- gipendulatus Achterberg, A. neotropical- is (Shenefelt), A. perinetensis Shenefelt, A. phmlineatus (Cameron), A. punctipes Thomson, A. rogezensis Granger, A. rii- gosicostalis Achterberg, A. iatianae (Te- lenga), and A. transversestriatus Grang- er. It is defined by the narrow fore- wing, which has a length /widest width ratio < 0.29. 10. The albitibia species-group (Shaw et al. 1998b; Fig. 6b), consisting of A. al- bitibia (Herrich-Schaffer) and A. fiisci- pennis (Szepligeti), is defined by nitid mesopleuron. This character state also occurs in A. pulchripes Wesmael, which falls well outside of this group. 11. The praetor species-group (Shaw ei al. 1998b; Fig. 6b) is monophyletic and consists of A. graphicus (Cresson), A. praetor (Reinhard), A. texauus Cresson, and new species 8 and 29. The group is defined by: 1) the strongly recurved hindwing radiellen vein which nearly touches the costal margin; 2) lateral ocellar diameter greater than 2.5 times ocellocular distance; and 3) short 15th flagellomere. 12. The apicalis species-group (Shaw et al. 1998a; Fig 6a) is a derived monophy- letic group. 13. The pulchripes species-group (Shaw et al. 1997; Fig. 6c) is a derived mono- phyletic group. 14. The gasterator species-group (Fig. 7a) is paraphyletic, and it is the basal spe- cies-group within the derived species- group cluster of the derived section. This species-group cluster is defined by clypeus abruptly edged, flat vcn- trad (69[1]), and is composed of the A. Volume 8, Number 2, 1999 225 r ANCESTOR 1 2 } 6 17 2121313)14 I I I I I I I I I I iOSl 54 646S I I I I I 11111111111112 1 64 NOLOPHAN 5064 gastritor species-group NEWSP 3 2 1 114) -♦-♦- NEWSP 15 I- NEWSP 19 SIMILLIM DENDROU 2 2027 *-t-0- NOCTURNU ■ TESTACEU — PURPLEXU 4869 NEWSP 32 CANTHARI '-, unipunctator \ species subgroup \ -HIRTUS r— NEWSP 25 '-O-i )e'4 HARRIMAN UNIPUNC T - FORTIS GRASSATO 1 58 ♦-♦- GASTERAT 5) I-O-CARINATU 1 52 NEWSP 6 0 "- NEWSP 4 .gasterator species subgroup Fig. 7. The base of the apical species-group cluster of the derived section. Volume 8, Number 2, 1999 229 r-»-»-lH)-RUFICEPS 110 0 1044SBT1 I I I I UFEI mm -fl-Hh 1 : 1 1111 10 1 6S66 I t I QUADRUM r-0-l-0-NEWSP23 0 ) 0 5, [-0- RUGULOSU ~H 1 (11670 ' Mhhhi- morio 1156 110 0 r-O-M-DIVERSUS 1 0 I r nigulosus species-group !52 0 0 pOlSSECTO 1 15)S6516 l-O-t-HHI- I 1 1 1 0 I n;7i«657o \ -(HH-(HH)- PERISCEL 0 0 10 10 \ ilS6 \ CRUENTIS ' 1 1S21J71011445!M „„^,„' '„°'. '„' I 10 110 0 10 1 0 1 0 Lo4Jn-sHESTAKO 10 2 1 « 3 2WlinoaeS4^7 \ I I I I I 0 I I 0- MINIATUS 111)02)0 \ 2S48 \ iHHhVENUSTUL 0 0 ^ '-0- melanoplerus species-group B 20 27M i-o-i- 0 ) p4-0-0-WAOAI 10 0 \ !nil2i)5B \ I I D I 0-t- KRULIKOW 11)01 \ 25)149 \ 0-t-HFAHRINGE 0 2 2 \ " 1 '1-+-AGIUS 1 1 1) AESTUOSU 2S4«52 i-O-Kh )S44f7 l-(H)-0- 4 6) 1 0 NEWSP 21 \^ 2u;')S6064 \ Mill 0-GLABER 2 112 2 1 Fig. 8. The apex of the apical species-group cluster. 230 Journal of Hymenoptera Research gasterator (Jurine), A. rugulosus (Nees von Esenbeck), and A. melanoptenis (Erichson) species-subgroups. The A. gasterator species-group is com- posed of A. arnoldi Tobias and new species 31 basally, and three species subgroups: the A. atricornis (Cresson), A. gasterator (Jurine), and A. unipunctator (Thunberg) species-subgroups. The group does not easily cluster based on any good synapo- morphy. The retention of tarsal claw teeth basal only in the A. gasterator species- group distinguishes it from species- groups apicad to it. Although three spe- cies clusters exist within the A. gasterator species-group, they were considered sub- groups, since they 1) clustered based on characters which reversed within the spe- cies-group or subgroup {A. atricornis spe- cies-subgroup); 2) were based on a color character, which are notoriously plastic across the genus {A. unipunctator species- subgroup), or 3) resulted in a paraphyletic group, such as the A. unipunctator species- subgroup, in which A. eurinus shares char- acter states which define both the A. uni- punctator and A. gasterator species-groups. 15. The ufei species-group (Fig. 8) is com- posed of one species, so far as known: A. ufei (Walley). The bicolored body is similar to the A. unipunctator species- subgroup, but the species differs in that the sides of the first metasomal tergite are parallel, and the female ab- dominal apex is compressed. Al- though this group is monospecific so far as known, it was given species- group status based on the unique me- tasomal features described above. 16. The rugulosus species-group (Fig. 8a) is monophyletic, and is defined by the uniquely shiny, coarsely and densely rugulopunctate vertex sculpturing and all black body. It consists of A. cruentus (Nees von Esenbeck), A. dis- sector (Nees von Esenbeck), A. diversus (Szepligeti), A. microculatus Watanabe, A. morio (Reinhard), A. periscelis (Rein- hard), A. rugulosus (Nees von Esen- beck), and A. shestakovi (Shenefelt). 17. The melanoptenis species-group (Fig. 8b) is monophyletic and is defined by the clypeo-antennal space /width of oral space ratio < 0.69, scutal sculp- turing excluding postero-medial dor- sal surface smooth or finely punctate, and clypeus height/ width less than 0.42. Members of the group are A. aes- tuosus (Reinhard), A. agilis (Telenga), A. desertus (Telenga), A. fahringeri (Te- lenga), A. flavistigma Shaw, A. glaber (Telenga), A. krulikoivskii Kokoujev, A. lucidus (Szepligeti), A. melanopterus (Erichson), A. mexicanus Cresson, A. tnandihularis (Cresson), A. niiniatus (Herrich-Schaffer), A. politiceps (Ga- han), A. ruficeps (Telenga), A. venustu- lus (Kokoujev), A. wadai (Watanabe), and new species 21, 22, and 23. 17a. The melanopterus species-subgroup (Shaw 1993, Fig 8b), a monophyletic subset of the above, consists of A. flavi- stigma Shaw, A. lucidus (Szepligeti), A. mandihularis (Cresson),. A. melanopte- rus (Erichson), A. mexicanus Cresson, A. politiceps (Gahan) and new species 23. The subgroup is defined by pecti- nate tarsal claws (Fig. 2c) and strongly protruding clypeal carina. DISCUSSION OF HOST ASSOCIATION The following data and that in Appen- dix 1 were drawn from Shenefelt (1975), Shaw (1983, 1994), and S. R. Shaw and P. M. Marsh (unpublished data). Of 208 Al- eiodes species analyzed, host records were available for 40.4% of Aleiodes species. Host records were available for 65.4% of basal species, 27.3% of intermediate spe- cies, and 39.4% of derived species (Appen- dix 1). Basal species for which host associations were known either attacked host species from more than one family, or attacked hosts other than non-catocaline Noctui- dae, Notodontidae, or Sphingidae. Excep- tions were A. gossi/pii, A. laphygmae, A. noc- Volume 8, Number 2, 1999 231 turniis, and A. nolophanae. Two basal spe- cies in the A. gastritor species group (A. pallesceus and A. stigmator) are gregarious. Within intermediate species, the A. cox- alis species-group showed a tendency to attack setose hosts within Arctiidae, Lasio- campidae, and Lymantriidae. Aleiodes cox- alis species-group species comprise 50.0% of Aleiodes species known to attack arctiids and lymantriids. Of the 15 intermediate species with host records, 40% attack ex- clusively non-catocaline noctuids so far as known. Within the derived section apicad of the A. pulchripes and A. apicalis species groups, only trifine Noctuidae are attacked so far as known. The A. praetor species-group at- tacks only sphingids. The A. albitihin spe- cies-group attacks only notodontids. With- in the A. pulchripes species-group, geome- trids, noctuids, and notodontids are at- tacked, so far as known. Of the seven of 11 A. apicalis species- group species with known hosts, four at- tack plusiine noctuids (A. brethesi, A. api- calis, A. molestus, and A. parasiticus). One is associated with an unidentified noctuid mummy {A. abdominalis), and three attack non-noctuids. Of all derived Aleiodes with host re- cords, 69.7% attack trifine noctuids. Of the above except the A. albitibia and A. praetor species groups, 84.6% attack trifine noc- tuids. Basal and intermediate species tend to have broad host ranges, and derived spe- cies tend to specialize on Noctuidae, No- todontidae, or Sphingidae. The only monophyletic groups which specialize on Notodontidae and Sphingidae lie at the base of the apical section. In the basal sec- tion, the A. pallidator species-group tends to specialize on Lymantriidae. The only Aleiodes species with host records for Ca- tocalinae are in the A. seriatus species- group in the basal section. Among inter- mediate species, the A. coxalis species- group tends to specialize on setose hosts, especially arctiids and lymantriids. According to the classification scheme of Nielsen and Common (1991), Noctuidae is the most derived lepidopteran family. The lepidopteran families Notodontidae, Lymantriidae, and Arctiidae are placed in superfamily Noctuoidea with Noctuidae. Notodontidae is the most basal noctuoid family (Nielsen and Common, 1991). Su- perfamily Sphingoidea, consisting of one family (Sphingidae), is immediately basal to Noctuoidea (Nielsen and Common, 1991). Based on order of placement in a list of 19 subfamilies of Noctuidae, subfamily Catocalinae is placed within the four bas- al-most subfamilies (Nielsen and Com- mon, 1991). Catocaline noctuids are uti- lized only by one Aleiodes species-group in the basal section, so far as known. Family Arctiidae is placed just basad to the noctuid subfamilies referred to in the study (Nielsen and Common 1991). The only basal Aleiodes species known to attack arctiids are within the A. seriatus species- group. Mitchell et al. have recently investigated the phylogeny of superfamily Noctuoidea by using characters derived from the nu- clear gene EF-la, which encodes elonga- tion factor- la protein (Mitchell et al. 1997). Based on their most parsimonious tree de- rived from all sites of EF-la, the authors concluded that 1) Notodontidae comprises a monophyletic group at the base of Noc- tuoidea; 2) "quadrifine" Noctuidae com- prise a paraphyletic group which consists of a monophyletic subgroup (Eutellinae, Nolinae, Sarrothripinae), and the basal- most subfamilies (Catocalinae, Hermini- inae) of a more derived monophyletic group. Arctiidae and Lymantriidae are sis- ter-groups at the apex of this latter mono- phyletic group; 3) the "trifine" noctuids, which comprise the remainder of Noctui- dae, form a monophyletic group at the apex of the cladogram (Mitchell et al. 1997). The findings of Mitchell et al. corrobo- rate Nielsen and Common with respect to 232 Journal of Hymenoptera Research Noctuoidea, except that Mitchell et al. find that Noctuidae is a paraphyletic group, with the quadrifine noctuids (including Catocalinae) more closely related to Arc- tiidae and Lymantriidae than to trifine noctuids (Mitchell et al. 1997). The latter finding is corroborated by Weller et al. (1994). Since older koinobiont parasitoid spe- cies are thought to have broader host ranges while newer species are thought to have narrower ranges (Shaw 1994), and since this study provides evidence that basal and intermediate Aleiodes species have broader host ranges while derived species have narrower host ranges, the overall pattern of phylogenetic inferences made from morphological data in this study is corroborated by host association data in light of the above. This study suggests that genus Aleiodes overall has co-evolved with its Lepidop- teran hosts. Basal host families appear more likely to be attacked by basal Aleio- des species while the most derived host family, trifine Noctuidae, is more likely to be attacked by derived Aleiodes species. Non-derived groups with well defined synapomorphies such as the basal A. ser- iatus and A. pallidator species groups, and the intermediate A. coxalis species-group, appear to have co-evolved with noctuoid groups basal to trifine Noctuidae such as catocalines, lymantriids and arctiids. The exclusive associations of Aleiodes species within the derived section with specific host groups, such as that of the A. praetor species-group with sphingids, of the A. al- bitibia species-group with notodontids, and of Aleiodes species apicad of the A. ap- icalis and A. pulchripes species-groups with Noctuidae add weight to this co-evolu- tionary hypothesis. ACKNOWLEDGMENTS The following collections provided specimens for this study. The North American collections cited in Shaw et al., 1997, The Koninklijk Museum voor Mid- den-Afrika, Belgium, The Museum of Natural His- tory, Warsaw, Poland, the insect collection of Lund University, Lund, Sweden, The Natural History Mu- seum, London, England, The Natural History Muse- um of Paris, France, The Natural History Museum of Vienna, Austria, The Royal Institute of Natural Sci- ences of Belgium, Brussels, The Natural History Mu- seum of Hungary, The Museum of Natural History, St. Petersburg, Russia. This research was supported by grant DEB-930-6314 from the National Science Foundation. Additional support was provided by a Canadian National Collection grant and by Univer- sity of Wyoming summer stipends. LITERATURE CITED Achterberg, C. van. 1982. Notes on some type-species described by Fabricius of the subfamilies Bracon- inae, Rogadinae, Microgastrinae, and Agathidi- nae (Hymenoptera: Braconidae). Eutonwliigishdw Benchtcn, 42: 133-139. Achterberg, C. van. 1985. Notes on Braconidae I-IV. Zoolcgische Mededclingen 59: 163-187. Achterberg, C. van. 1991. Revision of the genera of the Afrotropical and W. Palearctic Rogadinae Foerster (Hymenoptera: Braconidae). Zcologische Vfrhnmielingen 273: 3-102. Achterberg, C. van. 1993. Illustrated key to the sub- families of the Braconidae (Hymenoptera: Ich- neumonoidea). Zoohgische Verhaudclingen 283: 3- 189. Achterberg, C. van. 1993 Generic revision of the sub- family Betylobraconinae (Hymenoptera: Bracon- idae) and other groups with modified fore tarsus. Zoclogischc Verlmndelingen 298: 3-242. Achterberg, C. van and A. M. Penteado-Dias. 1995. Six new species of the Alcwda di-ipar group (Hy- menoptera: Braconidae: Rogadinae). Zoologhclie Mededeiingen Leiden 69: 1-18. Farris, J. S. 1970. Methods for computing Wagner trees. Systematic Zoology 19: 83-92. Farris, J. S. 1988. Hcnnig86 reference, version 1.5. Port Jefferson, New York. Harris, R. A. 1979. A glossary of surface sculpturing. Occasional Papers in Entomology 28: 1-33. Hennig, W. 1966. Phylogenetic Systematics. John Wiley and Sons. New York. Hennig, W. 1981. Insect Phytogeny. John Wiley and Sons. New York. Huber, J. and M. Sharkey. 1993. Chapter 3. Structure. In: Hymenoptera of the World: An Identification Guide to Families (ed. by H. Goulet and J. T. Huber). Agriculture Canada Publication 1894/ E. Ottawa. Lipscomb, D. L. 1992. Parsimony, Homology and the Analysis of Multistate Characters. Cladistics 8: 45-65. Marsh, P. 1979. The braconid (Hymenoptera) para- sites of the gypsy moth, Lymantria dispar (Lepi- Volume 8, Number 2, 1999 233 dopfera: Lymantriidae). Annals of the Entomolog- ical Society of America 72: 794-810. Marsh, P. M. and S. R. Shaw 1998. Revision of North American Aleiodes Wesmael (Part 3): the seriatus species-group in the New World (Hymenoptera; Braconidae, Rogadinae). Proceeiiings of the Ento- mological Societ}/ of Washington 100: 395—408. Mickevich, M. F. 1982. Transformation series analysis. Systematic Zoology 31: 461—178 Mickevich, M. F. and D. Lipscomb. 1991. Parsimony and the choice between different transformations for the same character set. Claiiistics 7: 111-139. Mickevich, M. F. and S. J. Weller. 1990. Evolutionary character analysis: tracing character change on a cladogram. Cladistics 6: 137-170. Mitchell, A., S. Cho, J. C. Regier, C. Mitter, R. W. Poole, and M. Matthews. 1997. Phylogenetic util- ity of elongation factor-la in Noctuoidea (Insec- ta: Lepidoptera): The limits of synonymnous sub- stitution. Molecular Biology and Evolution 14(4); 381-390. Mitter, C. and E. Silverfine. 1988. On the systematic position of Catocala Schrank (Lepidoptera: Noc- tuidae). Systematic Entomology 13: 67-84. Nielsen, E. S. and 1. F. B. Common. 1991. Chapter 41. Lepidoptera. In: The Insects of Australia. Cornell University Press, Ithaca, New York. Nixon, K. C. 1991. CLADOS Version 1.0. Cornell Uni- versity, Ithaca, New York. Shaw, M. 1983. On(e) the evolution of endoparasit- ism: the biology of some genera of Rogadinae (Braconidae). Contributions of the American Ento- mological Institute 20: 307-328. Shaw, M. 1994. Chapter 7. Parasitoid host ranges, in Parasitoid Community Ecology, (ed. by B.A. Haw- kins and W. Sheehan). Oxford University Press, Oxford, Great Britain. Shaw, M. and T. Huddleston. 1991. Classification and biology of braconid wasps. Handbooks for the Iden- tification of British Insects Vol. 7, Part II 7: 1-126. Shaw, S. R. 1993. Systematic status of Eucystomastax Brues and characterization of the Neotropical species, journal of Hymenoptera Research 2: 1-11. Shaw, S. R. 1995. Chapter 12.2. Braconidae. In: The Hymenoptera of Costa Rica (ed. by P. Hanson and I. D. Gauld). Oxford University Press, Oxford, Great Britain. Shaw, S. R., P. M. Marsh, and J. C. Fortier. 1997. Re- vision of North American Aleiodes Wesmael (Part 1): the flai'idus Cresson species-group in the New World (Hymenoptera: Braconidae, Rogadinae). journal of Hymenoptera Research 6: 10-35. Shaw, S. R., P.M. Marsh, and J. C. Fortier. 1998a. Revision of North American Aleiodes Wesmael (Part 2): the apicalis (Brulle) species-group in the New World (Hymenoptera: Braconidae, Roga- dinae). Journal of Hymenoptera Research 7: 62-73. Shaw, S. R. and P. M. Marsh. 1998b. Revision of North American Aleiodes Wesmael (Part 4): the praetor and atbitibia groups in the New World (Hymenoptera:Braconidae, Rogadinae). Proceed- ings of the Entomological Society of Washington 100: 553-565. Shenefelt, R. D. 1969. Notes on some rogadine genera (Hymenoptera: Braconidae). Proceedings of the En- tomological Society of Washington 71: 428-444. Shenefelt, R. D. 1975. Braconidae 8. Exothecinae, Ro- gadinae. Hymenopterorum Catalogiis (nova editio) 12: 1163-1262. Schuh, R. T. 1991. Phylogenetic, host and biogeo- graphic analyses of the Pilophorini (Heteroptera: Miridae: Phylinae). Cladistics 7: 157-189. Weller, S. J., D. P. Pashley, J. A. Martin, and J. L. Constable. 1994. Phylogeny of noctuoid moths and the utility of combining independent nuclear and mitochondrial genes Systematic Biology 43: 194-211. Whitfield, J. B. 1992. The polyphyletic origin of en- doparasitism in the cyclostome lineages of Bra- conidae (Hymenoptera). Systematic Entomology 17: 273-286. Wiley, E. O. 1981. Phylogenetics; the Theory and Practice of Phylogenetic Systematics. John Wiley and Sons, Inc. New York. 234 Journal of Hymenoptera Research Appendix I. List of Akiotlcf species examined. I losl t.imil\(ifs) abdominalis Cresson 1869 aciculatus Cresson 1869 aestuosiis (Reinhard) 1863 agilif (Telenga) 1941 albitWm (Herrich-Schaffer) 1838 alboannulalus Belokobylskyj alutaccus Granger 1949 aligharensi (Quadri) 1933 alternator (Nees von Esenbeck) 1934 angustatus (Papp) 1969 annulatiis Granger 1949 antennatus Belokobylskyj 1996 apicalis (BruUe) 1832 arctkus Thomson 1891 arizoniensis Marsh and Shaw 1997 armatus Wesmael 1838 arnoldii Tobias 1976 atricqjs Cresson 1869 atricornis (Cresson) 1872 autograpliae (Viereck) 1910 bakeri (Brues) 1912 bicolor (Spinola) 1808 borealis Thomson 1891 brethesi (Shenefelt) 1909 breinpendulatKS Achterberg and Penteado-Dias, 1995 brei'iradialis Granger 1949 buoculus Marsh 1989 burrus Cresson 1869 cameronii (Dalla Torre) 1898 cantherius (Lyle) 1919 carinatus (Ashmead) 1888 cariniventris (Enderlein) 1912 caucasicus Tobias 1976 cazieri Marsh and Shaw 1997 chloroticus Shestakov 1940 circuniscriptus (Nees von Esenbeck) 1834 compressor (Herrich-Schaffer) 1838 confonnis (Muesebeck) 1960 convexus Achterberg 1991 coxalis (Spinola) 1808 coxator Telenga 1941 crassipes Thomson 1891 cruentis (Nees von Esenbeck) 1834 curtispina Granger 1949 dendroliwi (Matsumura) 1925 desertiis (Telenga) 1941 dispar (Curtis) 1834 dissector (Nees von Esenbeck) 1834 diversicornis Granger 1949 dhvrsus (Szepligeti) 1903 Noctuidao Noctuidae Noctuidae (Noctuinae, Ophiderinae) unknown Notodontidae unknown unknown Gelechiidae, Noctuidae, Notodontidae Arctiidae, Lasiocampidae, Lymantriidae unknown unknown unknown Arctiidae, Geometridae, Lasiocampidae, Noctuidae (Hadeninae, Plusiinae) unknown unknown Geometridae, Noctuidae unknown unknown Noctuidae (Hadeninae) Noctuidae unknown Geometridae, Lycaenidae, Lymantriidae, Noctuidae, Nymphalidae, Pterophoridae, Zygaenidae unknown Noctuidae (Plusiinae) unknown unknown unknown Noctuidae (Acronictinae, Noctuinae) Notodontidae Geometridae unknown unknown unknown unknown unknown Bombycidae, Geometridae, Lymantriidae, Noctuidae, Notodontidae, Psychidae, Pyralidae, Tortricidae Gelechiidae, Geometridae, Limacodidae, Noctuidae, Notodontidae unknown unknown Hesperiidae, Satyridae unknown unknown Noctuidae (Hadeninae), Lymantriidae unknown Lasiocampidae unknown Noctuidae, Nvmphalidac Noctuidae (Acronictinae, Hadeninae) unknown Noctuidae (Acronictinae, Hadeninae) Volume 8, Number 2, 1999 235 Appendix 1. Continued. ,\/i'iit(irs specifs Host tamiK'des) eariiiof Marsh and Shaw 1997 ecuadi'ricima (Brues) 1926 eurinus (Telenga) 1941 excavatus (Telenga) 1941 fahringeri (Telenga) 1941 femoratui Cresson 1869 ferrugilcti (Shenefelt) 1975 flai'idiis (Cresson) 1865 flavisligma Shaw 1993 flaviUnsus Marsh & Shaw 1998 fortipcf. (Reinhard) 1863 forth (Muesebeck) 1960 fuscipennis (Szepligeti) 1904 gasterator (Jurine) 1807 gastritor (Thunberg) 1822 geomctrac (Ashmead) 1888 glabcr (Telenga) 1941 gossypii (Muesebeck) 1960 grmuiif Giraud 1857 graiiK/rtfus (De Gant) 1930 graphkus Cresson 1872 grassaior (Thunberg) 1822 gressitti (Muesebeck) 1964 harrimaiii (Ashmead) 1902 helleniai!. Papp 1985 hirtiia (Thomson) 1891 iiicertoides Telenga 1941 incerlus Kokoujev 1898 (>irf;sfri'fi(S (Reardon) jakolewi Kokoujev 1898 krulikou'skii Kokoujev 1898 kiisUtzk\/i Tobias 1976 lapln/gnme (Viereck) 1912 longiccrnis Granger 1949 longipemiiilatiif Achterberg and Penteado-Dias 1995 hichlus (Szepligeti) 1906 lymantrmc (Watanabe) 1937 mnhico^omcitos (Mason) 1979 mandibularis (Cresson) 1872 medianus (Thomson) 1896 mclanoptcrus (Erichson) 1848 mexicaniis Cresson 1869 microculntiii (Watanabe) 1937 }niiiiatiis (Herrich-Schaffer) 1838 nnidcftus (Reinhard) 1863 H;oMi!i'k»s Tobias 1986 »i()/i'sfi(s (Cresson) 1872 inoriti (Reinhard) 1863 narangac (Rohwer) 1934 negativus Tobias 1960 neotropiailis (Shenefelt) 1975 nigrilmsif (Enderlein) 1918 nigricorni^ Wesmael 1838 unknown unknown unknown unknown unknown unknown unknown unknown Unknown Noctuidae unknown unknown unknown Noctuidae (Plusiinae) Choreutidae, Drepanidae, Geometridae, Incurvari- idae, Notodontidae Geometridae unknown Noctuidae Noctuidae (Acronictinae) Noctuidae Sphingidae unknown unknown unknown unknown unknown unknown unknown Lvmantriidae unknown unknown unknown Noctuidae unknown unknown unknown Lvmantriidae Lasiocampidae unknown unknown unknown unknown unknown unknown Geometridae, Lasiocampidae unknown Noctuidae (Plusiinae, Noctuinae) Noctuidae unknown Noctuidae unknown unknown Noctuidae, Pvralidae, Geometridae 236 Journal of Hymenoptera Research Appendix I. Continued. ,Ut'(oi(fS species Host tamiiv(ies) nigristemmaticiim (Enderlein) 1918 nocturnus Tobias 1960 nohphanae (Ashmead) 1888 notozophus Marsh and Shaw 1997 luinbergi Noskiewicz 1956 pallcscens Hellen 1927 pallidator (Thunberg) 1822 pnllidistiginus (Telenga) 1941 palmatus (Walley) 1941 parasiticus Norton 1869 pedalis Cresson 1869 pellucens (Telenga) 1941 pcrcurrens (Lyle) 1921 perinetensis (Shenet'elt) 1975 perisCL'lis (Reinliard) 1863 perplexiis (Gahan) 1917 pictiis (Herrich-Schatfer) 1838 platypterygis (Ashmead) 1888 plurilineatus (Cameron) 1911 politicqjs (Gahan) 1917 praetor (Reinhard) 1863 procerus Wesmael 1838 pulchripcs Wesmael 1838 punctipes Thomson 1891 quadrum Tobias 1976 quebecensis (Provancher) 1880 rileyi Cresson 1869 rogezensis Granger 1949 rossi Marsh and Shaw 1997 rossicus Kokoiijev 1898 riificeps (Telenga) 1941 rutficoniis (Herrich-Schaffer) 1838 rufipes (Thomson) 1891 rugosicostalis Achterberg and Penteado-Dias, 1995 rugulosui (Nees von Esenbeck) 1811 sanctihyacinthi (Provancher) 1880 sanctivincentcusis (Shenefelt), 1975 satanas Telenga 1941 schirjajnoi (Kokoujev) 1899 scriptus (Enderlein) 1918 scrutator (Say) 1836 seriatus (Herrich-Schaffer) 1838 shestakovi (Shenefelt) 1975 sibiricus (Kokoujev) 1903 sigimtus (Nees von Esenbeck) 1911 similis (Curtis) 1834 simillinius (Ashmead) 1889 stigmator (Say) 1824 Noctuidae (Catocalinae) Noctuidae Noctuidae unknown Geometridae Noctuidae, Notodontidae (gregarious) Gelechiidae, Geometridae, Lymantriidae, Noctuidae, Psychidae, Tortricidae unknown Noctuidae Noctuidae unknown unknown Noctuidae (Catocalinae, Acronictinae) unknown unknown Geometridae, Noctuidae unknown Drepanidae unknown Noctuidae (Hadeninae, Noctuinae) Sphingidae unknown Lasiocampidae, Lymantriidae, Noctuidae (Acronicti- nae), Notodontidae unknown unknown Noctuidae (Acronictinae) Lymantriidae, Noctuidae (Acronictinae, Hadeninae), Pyralidae unknown unknown Drepanidae, Geometridae, Lasiocampidae, Lymantri- idae, Noctuidae, Pyralidae unknown Lasiocampidae, Lymantriidae, Noctuidae (Hadeninae, Noctuinae, Plusiinae) unknown unknown Noctuidae (Acronictinae) Arctiidae unknown unknown unknown unknown Choreutidae, Incurvariidae, Notodontidae Arctiidae, Lyonitiidae, Noctuidae unknown unknown unknown unknown Geometridae Geometridae, Lasiocampidae, Lymantriidae, Noctui- dae (gregarious) Volume 8, Number 2, 1999 237 Appendix I. Continued. Host tamiiv(ies) sudatorium Papp 1985 tatianae (Telenga) 1941 testaceus (Telenga) 1941 terminalh (Cresson) 1869 Tetrafphaeropyx Ashmead 1888 spp. texanui (Cresson) 1869 transversestriatuf Granger 1949 tristis Wesmael 1838 turkcitanicus (Telenga) 1941 ufei (Walley) 1941 ungulnris Thomson 1891 unicolor Wesmael 1838 unipundator (Thunberg) 1822 vaughani (Muesebeck) 1960 venustulus (Kokoujev) 1905 wadai (Watanabe) 1937 xanthus (Marshall) 1892 new species 1 new species 2 new species 3 new species 4 new species 5 new species 6 new species 7 new species 8 new species 9 new species 10 new species 11 new species 12 new species 13 new species 14 new species 15 new species 16 new species 27 new species 18 new species 19 new species 20 new species 21 new species 22 new species 23 new species 24 new species 25 new species 26 new species 27 new species 28 new species 29 new species 30 new species 31 new species 32 new species 33 new species 34 new species 35 new species 36 new species 37 new species 38 unknown unknown unknown Noctuidae (Acronictinae, Hadeninae, Noctuinae) Geometridae Sphingidae unknown unknown unknown Noctuidae (Noctuinae) unknown unknown Noctuidae (Acronictinae, Hadeninae) Noctuidae (Acronictinae) unknown unknown unknown imknown Lvmantriidae Noctuidae unknown unknown unknown unknown Sphingidae unknown Lycaenidae unknown unknown unknown Notodontidae unknown unknown unknown unknown unknown unknown unknown unknown unknown unknown Noctuidae (Cuculliinae) Arctiidae unknown Noctuidae (Hypeninae) Sphingidae unknown unknown Lycaenidae unknown unknown unknown unknown L\mantriidae Geometridae I. HYM. RES. Vol. 8(2), 1999, pp. 238-250 Does the Mating System of Trissolcus basalis (Wollaston) (Hymenoptera: Scelionidae) Allow Outbreeding? A. D. Loch and G. H. Walter (ADL and GHW) Department of Zoology and Entomology, The University of Queensland, Brisbane, Queensland, 4072, Australia; (ADL Present address: CSIRO Entomology, c/- Department of Conservation and Land Management, Brain St, Manjimup, Western Australia, 6258, Australia) Abstract. — The quasi-gregarious egg parasitoid Trissolcus basalis (Wollaston) is generally consid- ered to be an entirely inbreeding species because it is a sib-mating species that has female-biased sex ratios. Whether the species also outbreeds has not been previously investigated although several aspects of its mating behaviour suggest this might be possible. This question was inves- tigated indirectly in two ways by quantifying: (1) the inseminative capacity of T. basalis males in relation to the rate of female emergence, and (2) the effects of age and mating status on sexual receptivity of T. basalis. Trissolcus basalis females emerged over a period of several days, concen- trating their emergence in the morning hours. Males were able to inseminate many females (> 50) in rapid succession, apparently without sperm depletion. However, approximately 20% of females did not produce female offspring, probably because they did not mate. Although the mated females produced proportionately more male offspring with time, this outcome is not readily explained by sperm depletion of their mating partiiers and remains an unresolved issue. Male sexual receptivity appears to be unaffected by age and would be expected to be unaffected by mating status because males typically are polygynous. Although a previously successful mating encounter did not preclude females from mating again, female sexual receptivity decreased sig- nificantly after mating. Female sexual receptivity also decreased significantly with age. These results suggest that both T. basalis males and females have the ability to mate away from the natal site and that outbreeding is possible in this species. Whether males and females can locate one another away from their own natal site therefore warrants further investigation. Arrhenotokous parthenogenesis is the al 1993; Godfray 1994; but see Walter and usual means of reproduction in Hyme- Clarke 1992; Ode et al. 1997). Because fe- noptera. Female-biased sex ratios and sib- male Hymenoptera can control the fertil- mating are characteristic of many arrhen- isation of each egg they deposit, LMC the- otokous species, especially those v^hose ory predicts that single foundress broods males develop in the vicinity of their fe- of gregarious and quasi-gregarious spe- male siblings and emerge before them cies will contain only enough male off- (protandry). This occurs most frequently spring to mate all of their sisters in the in gregarious parasitoids, which deposit immediate vicinity (Hardy et al. 1998). many eggs per host, and quasi-gregarious Several recent studies have shown that ones (van den Assem et al. 1980), which strict local mating does not occur in some lay one egg per host into hosts that are species with female-biased sex ratios and invariably aggregated. Hamilton's (1967) which therefore should be inbreeding spe- local mate competition (LMC) theory is cies (e.g. Myint and Walter 1990; Nadel generally seen as the best explanation of and Luck 1992; Molbo and Parker 1996; female-biased sex ratios (e.g. Waage and Hardy et al. 1999). The term partial local Lane 1984; Waage and Ng 1984; Hardy et mate competition has been often used for Volume 8, Number 2, 1999 239 such cases. Another species whose mating behaviour appears not to be strictly local is Trissolciis ba:^alis (Wollaston) (Hymenop- tera: Scelionidae), a quasi-gregarious egg parasitoid of the green vegetable bug, Ne- zara viridula (L.) (Hemiptera: Pentatomi- dae). Trissolcus basalis is regarded as an in- breeding species that manifests local mate competition (LMC) (Hamilton 1967) be- cause it has female-biased sex ratios and sib-mating (Noble 1937; Anon. 1939; Smith 1945; Wilson 1961; Thomas 1972). How- ever, several aspects of its mating system are inconsistent with LMC and indicate that a proportion of each brood may out- breed if they are to transmit genes beyond the next generation. Field observations in- dicate that nearly 20% of newly-emerged females depart the egg mass unmated, ap- proximately 25% of mated females were mated more than once and often by mul- tiple males, virgin and mated females re- mained nearby the egg mass for up to sev- eral hours after emergence, and males dis- persed from the natal site (A. D. Loch and G. H. Walter unpublished data). Given that both males and females leave the na- tal site in the field, it is certainly possible that unrelated males and females meet and mate away from their natal site. The female mating pattern in T. basalis may be a consequence of the males mating many females in quick succession, and be- coming sperm depleted. Females mated by sperm depleted males may receive in- sufficient sperm to produce the usual pro- portion of daughters in a brood. They may therefore mate again, with outbreeding being more likely in such circumstances. The inseminative capacity of T. basalis males was therefore investigated in rela- tion to the rate and sequence of female emergence from the host patch. The ex- periment was designed to emulate typical rates of sibling emergence and mating in the field, rather than the unnaturally high rates used in some studies (e.g. Nadel and Luck 1985). Specifically, the number (and proportion) of females emerging from a single egg mass and inseminated by one male was quantified. The consequence of female emergence position for the amount of sperm received from the male was quantified by recording the number and sex ratio of progeny from every tenth fe- male to emerge. Whether T. basalis males become sperm depleted at mating rates typical in the field could thus be deter- mined. The potential for outbreeding in T. ba- salis was also evaluated by determining whether males and females are sexually receptive after mating and /or leaving the natal site. We therefore investigated the effects of age and mating status (virgin or mated) on T. basalis sexual receptivity in the laboratory, by exposing different aged virgin and once-mated females to newly- emerged virgin males. For completeness, the effect of age on male sexual receptivity was also studied by mating different aged virgin males with newly-emerged virgin females. MATERIALS AND METHODS Laboratory cultures. — Green vegetable bugs were reared at 28 ± TC, 65 ± 10% R.H., 16L:8D in mesh cages (0.45 m sides) on a diet of green bean pods (Phaseolus inilgaris L.), shelled peanuts (Arachis hy- pogea L.) and water. Cultures were aug- mented regularly with field-collected bugs. Green vegetable bug egg masses were collected daily from cages and were used to maintain cultures of green vege- table bug or T. basalis (see below). Laboratory cultures of T. basalis were es- tablished from parasitoids that emerged from green vegetable bug egg masses col- lected from mungbean, Vigna radiata (L.) Wilezek, and soybean. Glycine max (L.) Merr., during March-April 1997 and Jan- uary-April 1998 at Pittsworth (27" 43'S, 151° 38'E), Bongeen (27° 34'S, 151° 27'E) and Cecil Plains (27° 32'S, 151° 12'E) in south-eastern Queensland, Australia. All T. basalis individuals that emerged from a single egg mass were held together in a 240 Journal of Hymenoptera Research ventilated vial streaked with honey. Cul- tures of T. basalis were kept at 15 ± 1°C 65 ± 10% R.H. and 16L:8D. The identifi- cation of r. basalis was confirmed by Dr Norman Johnson (Ohio State University). Voucher specimens from the T. basalis cul- ture are deposited in The University of Queensland Insect Collection. In all experiments, virgin wasps of the F ,- Ft generation were used. Wasp virgin- ity was ensured by holding single wasp pupae in ventilated vials with honey, after breaking the host egg mass into individual eggs soon after parasitoid pupation. For each experimental replicate, wasps were derived from different field-collected egg masses to ensure siblings were not includ- ed as replicates. Inseminative capacity. — To determine male inseminative capacity, all of the T. basalis females that emerged from each of 10 parasitised green vegetable bug egg masses were tested for insemination (see below). These original egg masses con- tained 85 ± 5 eggs, the mean size for green vegetable bug egg masses in south- eastern Queensland. Each egg mass had been parasitised by a single, once-mated female T. basalis over two days in a 50 x 25 mm ventilated vial. Self-superparasit- ism is unlikely to arise under such condi- tions because females use a chemical marker to mark parasitised eggs (Wilson 1961; Ganesalingam 1966; Field et al. 1998). After 9-10 days, when the first males began to emerge, vials containing the parasitised egg masses were moni- tored frequently (every 5-10 minutes) dur- ing the 10 hours of artificial laboratory light each day. Before females began to emerge, all males were removed except for the dominant male occupying the egg mass. The dominant male was lightly marked on the thorax with fluorescent dust to distinguish him from males that emerged subsequently. These latter males were removed immediately they ap- peared. At each morutoring period any females that had emerged were removed and each was placed alone in a ventilated vial and provided with honey. Females were typi- cally found at the top of the vial. All fe- males, except those used to assess fecun- dity (see below), were provided 5-10 fresh green vegetable bug eggs to establish whether they produced female offspring, a certain indication they had been insem- inated (Wilson 1961). Females were al- lowed 24 hours to parasitise eggs before being removed. Lifetime fecundity was assessed for the first emerging female and for every tenth female that emerged from each egg mass. Each was provided with a frozen ( — 70°C) green vegetable bug egg mass each day for the first 12 days. The frozen eggs were < 1 month old and still viable for T. basalis (Powell and Shepard 1982, Kelly 1987). Earlier trials (n = 6) indicated that daily fecundity decreased rapidly and females were unlikely to produce offspring after 12 days. Large egg masses (85 ± 5 eggs each) were supplied on each of the first two days, half masses (40 ± 5 eggs) for each of the next four days, and small masses (20 ± 5 eggs) for each of the last six days, so that females had an excess of hosts at all times (see Results). Parasitised egg masses were placed singly in ventilat- ed vials and incubated until all offspring had emerged. Eggs that were obviously parasitised, but from which parasitoids failed to emerge, were dissected and the parasitoid removed for sexing. Counts of the numbers of male and female offspring produced per female per day were then made. Adult size. — Adults were measured to assess whether their size was affected by emergence sequence and whether fecun- dity was influenced by size. Two mea- surements were taken from all males and females to emerge from each egg mass: head width and right hind tibial length. Head width was measured as the distance between the outermost points of the eyes. Volume 8, Number 2, 1999 241 Measurements were made under a dis- secting microscope, accurate to 0.01 mm. Sexual receptivity. — Two experiments were conducted to investigate the effect of adult age and mating status (virgin or mated) on sexual receptivity. The first ex- amined whether age affected the male's readiness to mate. A single virgin male aged \, 5, 10, 15 or 20 days old was intro- duced into one end of a 50 X 12 mm ven- tilated vial containing a virgin female less than 24 hours old at the other end. The male and female were observed until mat- ing occurred or for 20 minutes, as virgin males and females would usually mate within 10 minutes with an average pre- mating time of ca 3 minutes. The number of contacts between the male and female before mating was re- corded, as were the pre-mating and mat- ing times. In addition, pre-mating and mating behaviours were observed for any differences between treatments. Once mat- ing had taken place, the male was re- moved and the female provided with ca 10 green vegetable bug eggs. The eggs were removed one day later and incubat- ed at 28 ± 1°C until offspring emerged. Because T. basalis is arrhenotokous (Wil- son 1961), a female was regarded as suc- cessfully inseminated if any female off- spring were produced. The other experiment examined wheth- er female age and mating status affected her readiness to mate. Virgin females were assigned to two groups. Those in one group were not mated, whereas the others were mated within 24 hours of emergence by a virgin male. All females were held, until needed, in a ventilated vial streaked with honey. Subgroups of females (virgin or once-mated) were exposed to virgin males less than 24 hours old, at ages 1, 5, 10, 15 or 20 days, one pair per 50 X 12 mm ventilated vial. Procedures and con- ditions were the same as for the first ex- periment. For both experiments 15 repli- cates of each treatment were conducted, all at 25 ± 2°C and 65 ± 10% R.H. Statistical analysis. — Logistic analyses were conducted to test whether emer- gence position influences the probability of a female being inseminated. A logistic regression was conducted for each of the 10 experimental replicates, in which the binary response variable, whether a fe- male was inseminated (assigned 1) or un- inseminated (assigned 0), was regressed against her position in the emergence se- quence. The fecundity of females in different emergence positions and their offspring's sex ratio were analysed by 1-way ANOVA after log(x + 0.5) and arcsine(Vp) trans- formations, respectively. The significance of any differences was assessed by Fish- er's protected least significant difference test. Linear regression was employed to assess the relationship between fecundity and female head width or hind tibial length. The effect of age of males, virgin fe- males and mated females on the number of pre-mating contacts, pre-mating time and mating time was tested by 1-way AN- OVA after data were log(x + 0.5) trans- Table 1. Summary statistics from logistic analyses testing whether emergence position influences the probability of a female being inseminated. A logistic regression was conducted for each of the 10 experi- mental replicates, in which the binary response var- iable, whether a female was inseminated (assigned 1) or uninseminated (assigned 0), was regressed against her position in the emergence sequence. Relationship refers to whether females later in the emergence se- quence tended to be uninseminated (negative) or in- seminated (positive). \.. KL-plll.UC- k'm.iles \ p.^,,lu. Ki-I.itionship 1 78 0.58 0.45 negative 2 40 1.52 0.22 positive 3 30 0.56 0.45 negative 4 73 0.30 0.58 positive 5 59 5.74 0.02 negative 6 55 0.80 0.37 positive 7 60 11.41 <0.01 negative 8 57 0.15 0.70 negative 9 61 2.78 0.10 positive 10 78 2.56 0.11 positive 242 Journal of Hymenoptera Research Day Fig. 1. Pattern in which Tnsso/cHS basnlis siblings emerged from parasitised green vegetable bug egg masses. Number of males and females that emerged each day from each of 10 egg masses (parasitised on day 0) in the laboratory at 28 ± rC, 65 ± 107o R.H. and 16L:8D. Error bars represent standard errors for the mean number of siblings (males and females) that emerged each day. formed. G-tests were employed to test the effect of age of males, virgin females and mated females on the number of females mated within 20 minutes. G-tests were also employed to test if the number of fe- males inseminated was related to age of males and virgin females. The effect of mated female age on the probability of be- ing inseminated was not analysed statis- tically because mated females were pre- sumed to have been successfully insemi- nated at their first mating. The effects of female age and mating status (virgin or once-mated) on the num- ber of pre-mating contacts, pre-mating time and mating time were tested by two- way ANOVA after data were log(x + 0.5) transformed. A log-linear analysis was employed to test the effects of female age and mating status on the number of fe- males mated within 20 minutes. RESULTS Emergence patterns. — Most wasps (> 95%) emerged under lighted conditions, with the majority emerging during the first 3-4 hours of morning light. A mean ± s.e. of 67.9 ± 4.6 wasps emerged from Table 2. Overall number of offspring and off- spring sex ratio (mean ± s.e.) produced by Trissokus basnlis females in different positions in the entire emergence sequence. Fecundity and sex ratio values derive only from those females that had been insem- inated (as indicated by their production of daugh- ters). See Fig. 2 and Table 1 for details. Emergence No. No nol position nitiled mated (t< m.i!e) 1 8 2 139 ± 13 0.23 ± 0.03 10 8 1 125 ± 17 0.26 ± 0.04 20 9 1 129 ± 16 0.29 ± 0.05 30 9 1 116 ± 11 0.36 ± 0.06 40 8 0 100 ± 12 0.30 ± 0.08 50 7 1 153 ± 23 0.30 ± 0.04 60 5 1 95 ± 15 0.22 ± 0.08 70 1 2 116 0.72 ' Column means for fecundity (F-^„ = 1. 12, p = 0.37) and sex ratio (F-„ = 1.61, p = 0.16) were not signif- icantly different. each of the 10 original egg masses, com- prising 59.1 ± 5.0 females and 8.8 ± 2.3 males. Males began emerging on day nine with emergence peaking on days 10 and 11 (Fig. 1). Few males, if any, emerged from egg masses later than day 14. Fe- males emerged on days 10-19 with emer- gence peaking on days 11-13. The largest number of females that emerged from any one egg mass in one day was 57 females on day 12. Inseminative capacity and fecundity. — The dominant males that were left alone on egg masses to mate their sisters insemi- nated a mean ± s.e. of 48.3 ± 3.9 females, with 68 females being the maximum num- ber inseminated by one male. The propor- tion (mean ± s.e.) of emerging females that was inseminated by the dominant males was 0.82 ± 0.02 across egg masses. Of the 10 replicates conducted, five showed a positive relationship between the probability of a female being insemi- nated and her emergence position and five showed a negative relationship (Table 1). Two of the negative relationships were significant at < 5% and two of the positive relationships were significant at < 11% (Table 1). Volume 8, Number 2, 1999 243 Table 3. Number and sex ratio (mean ± s.e.) of offspring produced by females that emerged as adults during a single day, but in different positions in the emergence sequence. Only females that emerged on the first day of female offspring emer- gence were included in the analysis because at this stage males would presumably have had a full sperm supply, and short term rates of sperm depletion could be assessed most accurately. Fecundity and sex ratio values are calculated only from mated females in each position. Emerge me position No. mated No. nol mated f-ecunditv Sev r.ilio (., male) 1 8 2 139 ± 13 0.23 ± 0.03 10 6 0 112 ± 19 0.27 ± 0.04 20 3 1 150 ± 4 0.27 ± 0.05 30 2 0 108 ± 4 0.46 ± 0.17 40 2 0 99 ± 45 0.21 ± 0.00 50 1 0 257 0.35 Column means for fecundity (F,,^ = 2.31, p = 0.09) and sex ratio (F^,„ = 1.47, p = 0.25) were not signif- icantly different. Fecundity was highly variable, and ranged from 42 to 257 (mean ± s.e. = 121 ± 5, n = 65) offspring per female. Progeny production peaked during the first 24 hours after emergence with about 40-50 offspring on average, and then decreased rapidly with time (Fig. 2). The number of offspring produced per day was always less than the number of hosts provided. The sex ratio (proportion male) of off- spring increased with time such that fe- males produced few or no female off- spring after 10 days, although by then few offspring were being produced (Fig. 2). Fecundity and brood sex ratio were not significantly affected by the position of parent females in the overall emergence sequence (Table 2), nor by the position of females in the emergence sequence on the first day of female emergence (Table 3). Adult size. — Head widths and hind tibial lengths for male and female T. basalts showed little variation within and across replicates. Females were significantly larg- er than males: mean ± s.e. head widths were 0.61 ± 0.001 mm and 0.58 ± 0.002 mm for females and males respectively (F,,„ = 410.3, p < 0.0001), and their re- spective hind tibial lengths were 0.41 ± 0.001 mm and 0.39 ± 0.001 mm (F,,,,, = 52.3, p < 0.0001). No trend between emer- gence position and head width or hind tibial length was apparent except that the last 1-5 wasps to emerge from an egg mass tended to have head widths and hind tibial lengths up to 0.05 mm smaller than previously emerged wasps. Fecundity increased significantly with increases in female head width and hind tibial length (Fig. 3). However, regressions of fecundity against each of the two size measurements fitted poorly (r- £ 0.10). Sexual receptivity — The age of males had no significant effect on the number of pre- mating contacts, mating time or the num- Table 4. Effect of virgin male age on their propensity to mate within 20 minutes of exposure to a virgin female (expressed as number of females mated). Also given is the number of females inseminated, number of pre-mating contacts, pre-mating time and mating time (last three values are mean ± s.e.). The number of males used to calculate each mean and s.e. is the number of males mated in 20 minutes (first row) from the 15 replicates. Male aj;e (days) 1 5 111 1^ 20 No. mated lS,i ISa 1 5a 15a 15a No. inseminated 12a 12a 15a 12a 12a No. contacts 2.6 ± 0.3a 1.9 ± 0.3a 1.8 ± 0.3a 2.7 ± 0.4a 2.2 ± 0.2a Pre-maling time (s) 173 ± 37a 92 ± 17b 88 ± 21b 181 ± 33a 176 i 24a Mating time (s) 12.0 ± 1.3a 10.8 ± 1.0a 12.6 ± 0.6a 12.4 ± 0.8a 14.5 ± 2.1a Row means tolknved h\ the same letter are not significantly ditterent (t;-test tor first 2 rows, l-way ANOVA for last 3 rows, P > 0.05). 244 Journal of Hymenoptera Research Table 5. Effect of virgin female age on their propensity to mate within 20 minutes of exposure to a virgin male (expressed as number of females mated). Also given is the number of mated females that was successfully inseminated, number of pre-mating contacts, pre-mating time and mating time (last three values are mean ± s.e.). The number of females used to calculate each mean and s.e. is the number of females that mated within 20 minutes (first row) from the 15 replicates. Virgin female age ( ;days) 1 5 10 15 20 No. mated 15a 15a 14ab lib 10b No. inseminated 12a 11a 10a 8ab 3b No. contacts 2.6 ± 0.3a 5.3 ± 0.8b 8.6 ± 1.9b 5.0 ± 1.1b 5.8 ± 0.8b Pre-mating time (s) 173 ± 37a 303 ± 43b 419 ± 113b 275 ± 93ab 370 ± 73b Mating time (s) 12.0 ± 1.3a 11.5 ± 0.7a 8.1 ± 0.8b 10.9 ± 0.7a 9.3 ± 1.5ab Row means followed by the same letter are not significantlv different (G-test for first 2 rows, 1-way ANOVA for last 3 rows, P > 0.05). ber of females mated or inseminated (Ta- ble 4). Pre-mating time was affected by male age with males aged 5 and 10 days old having a significantly shorter pre-mat- ing time than males aged 1, 15 and 20 days old. No differences in male pre-mat- ing or mating behaviour were observed between males of different ages. In contrast, virgin female pre-mating and mating behaviours were affected by their age (Table 5). Females aged 5-20 days old tended to resist the males' mat- ing attempts by moving away from them, aggressively chasing males away and /or refusing to allow males to copulate after mounting. The numbers of pre-mating contacts and pre-mating times were great- er for females aged 5-20 days than for 1 day old females (Table 5). Females aged 1, 5 and 15 days old mated for significantly longer than 10 day old females. The num- ber of females mated within 20 minutes and the number successfully inseminated decreased significantly with female age (Table 5). The age of mated females also affected their pre-mating and mating behaviours, with mated females aged 5-20 days gen- erally resisting mating attempts in the way described above for virgin females of different age. The mating propensity of mated 5-20 day old females was signifi- cantly less than that of one day old mated females (Table 6). The numbers of pre- mating contacts and pre-mating times in- creased significantly with female age, but mating time was not significantly affected by their age (Table 6). Two-way ANOVA examining the ef- fects of female age and sexual status (vir- Table 6. Effect of age of previously-mated females on their propensity to mate within 20 minutes of ex- posure to a virgin male (expressed as number of females mated). Also given is the number of pre-mating contacts, pre-mating time and mating time (values are mean ± s.e.). The number of females used to calculate each mean and s.e. is the number of females that mated within 20 minutes (first row) from the 15 replicates. Age of pre\'iouslv-mafed female ( da\s) 1 5 10 IS :ii No. mated 15a 6b 8b 9b 6b No. contacts 4.7 ± 1.0a 9.7 ± 2.9b 7.0 ± l.Oab 8.1 i 1.1b 9.3 ± 3.4b Pre-mating time (s) 235 ± 48a 549 ± 143b 373 ± 74ab 514 ± 97b 547 ± 205ab Mating time (s) 8.6 ± 0.7a 8.4 ± 1.1a 6.1 ± 1.1a 7.4 ± 1.5a 9.7 ± 2.2a Row means followed by the same letter are not significantly different (G-test for first row, 1-wav ANOVA for last 3 rows, P > 0.05). ' Volume 8, Number 2, 1999 245 Table 7, Summary of results from 2-way ANOVA (log (x + 0.5) transformed) testing whether the individual and interactional effects of the factors, female age and mating status (virgin or once-mated), affected the number of pre-mating contacts, pre-mating time and mating time. .ilin^ tinu- Milling linu' Age Mating status Age X mating status F^„ = 7.06, P < 0.0001 F,„ = 8.37, P = 0.005 F,« = 0.89, P = 0.47 F^^ = 4.50, P = 0.002 F,^ = 8.65, P = 0.004 F,« = 0.83, P = 0.51 3.31, p = 0.01 F,„ = 10.01, P F^'^ = 0.96, P = = 0.002 0.44 gin or once-mated) on mating propensity indicated that the interaction between fe- male age and mating status was not sig- nificant for all three measures of mating propensity (Table 7). As single factors, fe- male age and mating status significantly affected the number of pre-mating con- tacts, pre-mating time and mating time (Table 7). A log-linear analysis on the effect of fe- male age on the propensity of virgin and mated females to mate within 20 minutes revealed that the model could be de- scribed best by two interactions: female mating status and the number of females mated, and female age and the number of females mated (Maximum likelihood x" = 8.38, df = 8, p = 0.40). The propensity of females to mate within 20 minutes was significantly greater for virgin females and younger females. DISCUSSION The results from this study suggest that strict local mating does not occur in T. ba- salts and that outbreeding away from the natal site may commonly occur. Results that are inconsistent with LMC theory in- clude: (1) males do not achieve the maxi- mum rate of insemination expected de- spite showing no apparent signs of sperm depletion, (2) males remain sexually re- ceptive probably throughout their lifetime (Table 4), and (3) females can be mated multiple times, despite becoming decreas- ingly sexually receptive after mating and with age (Tables 6, 7). We discuss the im- plications that these results have on the mating system of T. hasalis and ask wheth- er this species is likely to outbreed. Emergence of T. basalis females is con- centrated during the early morning hours over several days (Wilson 1961; Fig. 1). In the field a dominant male usually guards the parasitised egg mass from which fe- males are emerging. These females are his sisters unless the egg mass has been su- perparasitised. In the laboratory, the sin- gle male left on the mass successfully mat- ed many females in succession, with sometimes up to 50 or more females emerging over several hours (Table 3). Fe- males in all positions in the emergence se- quence apparently received similar quan- tities of sperm because their offspring sex ratios were not affected by emergence po- sition (Table 2; Fig. 2), even if those fe- males all emerged and were mated on the same day by a single male (Table 3). Although the above results suggest that a single T. basalis male can fully insemi- nate each of his female siblings from the same egg mass, two observations indicate that the dominant male does not achieve the maximum rate of insemination that is possible. First, only ca 80% of females were inseminated by dominant males (Ta- ble 1, see also Wilson 1961). Wilson (1961) proposed that temporary sperm depletion in the dominant male may be the cause. But even when large numbers of females emerged in a day, the offspring sex ratio produced by inseminated females did not vary with their position in the mating se- quence (Table 3). In addition, uninsemi- nated females appeared throughout the emergence sequence (Tables 1-3). Field observations have shown that a similar percentage of emerging females is not 246 Journal of Hymenoptera Research mated by the dominant male guarding the egg mass (A.D. Loch and G.H. Walter un- published data). Also, ca 18% of mahngs between virgin males and females in the laboratory do not lead to successful in- semination (A.D. Loch and G.H. Walter unpublished data). In our experiments, we did not observe females to confirm they mated or to ascertain why they may not have mated. However, the high rate of un- inseminated females is likely to be partly the consequence of simultaneous female emergences, during which males become occupied with some emerging females, while others move unmated to the top of the vial, a behaviour that has parallels in the field (A.D. Loch and G.H. Walter un- published data). The second observation suggesting that maximum insemination rates are not achieved by the dominant male, is that even those females that were inseminated produced proportionately more male off- spring with age (Fig. 2). This trend has also been reported in other studies of T. basalis fecundity (Powell and Shepard 1982; Correa-Ferreira and Zamataro 1989; Awan et al. 1990) and in work on the con- familial Telenomus busseolae (Gahan) (Cha- bi Olaye et al. 1997). Females apparently do not receive sufficient sperm to fertilise all their eggs. However, temporary sperm depletion in males or insufficient sperm transfer by males are unlikely explana- tions. First, this trend was uniform for fe- males in all emergence positions (Fig. 2), indicating that sperm depletion in the male was not the cause. Second, females held with males throughout their lifetime, and therefore assumed to be mated mul- tiple times, also produce proportionately more male offspring with time (Powell and Shepard 1982; Awan et al. 1990), sug- gesting that this trend occurs irrespective of the number of times a female is mated. The reason for this trend is not clear, but a decrease in sperm viability over time is possible, or it could have a behavioural or physiological basis. Results from this study suggest that T. basalis males do not become sperm de- pleted at rates of mating that are typical for this species in nature. In this study, green vegetable bug egg masses of 85 ± 5 eggs were used, and represent the largest known host masses for T. basalis, in terms of the number of eggs. The test males were, therefore, exposed to a high number and frequency of matings. In any case, dominant males in control of egg masses in the field are unlikely to become sperm depleted because changeovers in male dominance occur frequently (A.D. Loch and G.H. Walter unpublished data), and female emergence continues over several days (Fig. 1). Sexual receptivity of T. basalis males ap- pears unaffected by age (Table 4) and mat- ing status, thus enabling males to mate probably throughout their lifetime. The only aspect of male sexual receptivity that was affected by age was pre-mating time, which was significantly shorter for 5-10 day old males than for 1, 15 and 20 day old males. They may be more receptive at 5-10 days because they emerge up to sev- eral days before females (Anon. 1939; No- ble 1937; Smith 1945; Wilson 1961; Thom- as 1972) and would therefore not normally need to mate immediately upon emer- gence. In contrast, female sexual receptivity decreased rapidly after mating and with age (Tables 5-7). Such decreases are con- Fig. 2. Number (mean ± s.e.) of progeny produced each day after emergence by inseminated Tnssc/CKS basalis females. Data are presented separately for each group of parent females according to their position in the emergence sequence (i.e. (a) 1st, (b) 10th, (c) 20th, (d) .3()th, (e) 40th, (f) 50th, (g) 60th and (h) 70th), and thus the sequence in which they were inseminated by the dominant male on their host egg mass. Numbers above error bars indicate the number of females still alive at that time. Volume 8, Number 2, 1999 (a) Female 1 70 247 ■ Females Q Males 12 3 4 5 6 r^r^- 9 10 11 12 (e) Female 40 70 - 60 - (f) Female 50 9 10 11 12 (g) Female 60 E z r^r— ir^ 9 10 11 12 1 2 3 4 5 6 7 6 aiU1112 70 • 60 ■ 50 I. HmnH^ 1 1 1 Female age (days) 123456789 10 1112 Female age (days) 248 Journal of H-imenoptera Research sistent with LMC theory because mating is assumed to occur only at the natal site among siblings (Hamilton 1967). Howev- er, females can be mated multiple times, which is inconsistent with LMC theory. This inconsistency coupled with others such as male dispersal from the natal site (A. D. Loch and G. H. Walter unpublished data), and males remaining sexually re- ceptive probably throughout their lifetime (Table 4) suggest that T. basalis of both sexes may mate away from the natal site and therefore outbreed. Trissolciis basalis may outbreed if males and females can locate and / or attract each other once they have left the natal site. Males may be able to locate unrelated, newly-emerged (and thus sexually recep- tive) females directly, or they could do so indirectly by locating hosts parasitised by T. basalis and then competing with emerg- ing males for mating access to females. Similarly, newly-emerged females may be able to locate males directly, or indirectly by searching for parasitised hosts with males in occupation. Currently, no evi- dence is available on whether males or fe- males can locate potential mating partners away from the natal site. LMC models as- suming strict local mating (Hamilton 1967) suggest that T. basalis males and fe- males will not be able to locate each other in the field. However, in other hymenop- terous species with female-biased sex ra- tios, such as Spalangia cameroui (Perkins) (Myint and Walter 1990) and Pacln/crepo- ideus vindemiae (Rondani) (Nadel and Luck 1992), males are able to locate hosts and thus potential mating partners, a feature likely to be found in other species (Hardy 1994). Males may not only be able to outbreed with newly-emerged virgin females but also with newly-emerged mated females because a previous mating encounter did not preclude females from mating again (Table 6). This result is not likely to be an artefact of laboratory conditions or pro- cedures because females have been ob- Head width (mm) Fig. 3. Fecundity of Trissolcus basalis females in re- lation to head width (y = 1711x - 928, r^ = 0.10, n = 65, p = 0.01). The trend for fecundity versus hind tibial length (y = 1294x - 407, r- = 0.07, n = 65, p = 0.04) is not shown because it was similar to the displayed trend. Both trends were determined irre- spective of female position in the emergence se- quence (see Table 2). served to be mated multiple times and by multiple males in the field (A. D. Loch and G. H. Walter unpublished data). Whether T. basalis females are truly polyandrous has yet to be established, for matings after the first successful mating may not lead to successful insemination. For instance, mating plugs may be used by males to en- sure additional matings do not result in insemination. This study also made a number of find- ings pertaining to the fecundity of T. ba- salis. The mean fecundity recorded in this study is higher than fecundities recorded by Noble (1937), Ganesalingam (1966) and Thomas (1972), but similar to values re- corded by Powell and Shepard (1982) and Correa-Ferreira and Moscardi (1994), and lower than fecundities recorded by Cor- rea-Ferreira and Zamataro (1989) and Awan et al. (1990) for the same species. These differences are likely to be the result of differences in laboratory procedures and conditions, although differences in adult female size may have contributed because fecundity is greater for larger fe- males (Fig. 3). The trend whereby fecun- dity peaked on the first day after female emergence and decreased rapidly over Volume 8, Number 2, 1999 249 time, differs somewhat from the results of Ganesalingam (1966) and Powell and Shepard (1982), who showed that fecun- dity peaked on day 2. These differences are less readily attributable to different laboratory procedures and conditions, and their significance is unclear. The claim by Field et al. (1998) that T. basalis is a syno- vigenic species was supported by our re- sults because females laid eggs for 10-12 days with progressively fewer eggs each day (Fig. 2) despite sufficient hosts being available during the first few days for them to have deposited their lifetime com- plement of eggs then. In conclusion, although uninseminated females leaving the egg mass may well be mated by the other males (also likely to be their siblings) that wait around the egg mass, the possibility that these females could mate unrelated males near or away from the natal site may not be low. Al- though we have no direct evidence of T. basalis outbreeding in nature, the results from this study and other related studies (A.D. Loch and G.H. Walter unpublished data), suggest that it may be more fre- quent than anticipated by LMC theory. In addition, outbreeding is likely to occur when > 1 female oviposits in an egg mass. Further research investigating the mating system of T. basalis is required before the question of the species' outbreeding can be resolved. Specific issues that need to be addressed include the questions of wheth- er T. basalis haa a means of mate-attrac- tion, and whether T. basalis females are truly polyandrous. ACKNOWLEDGMENTS We thank Tony Clarke for discussing aspects of this manuscript, and Norman Johnson (Ohio State University) for confirming the identification of Tris- solcus basalis. We also thank Ian Hardy and Peter Mayhew for their comments and suggestions on ear- lier versions of the manuscript. The senior author was supported bv a Grains Research and Development Corporation lunior Research Fellowship. LITERATURE CITED Anon. 1939. The egg parasite of the green vegetable bug. The Agricultural Gazette of Neu' Simf/i Wales 50; 277-278. Awan, M. S., Wilson, L. T. and Hoffman, M. P. 1990. Comparative biology of three geographic popu- lations of Trissolcus basalis (Hymenoptera; Sce- lionidae). Environmental Entomology 19; 387-392. Chabi Olaye, A., Schulthess, F., Shanower, T. G. and Bosque-Perez, N. A. 1997. Factors influencing the developmental rates and reproductive potentials of Teteiumius busseolae (Gahan) (Hym.: Scelioni- dae), an egg parasitoid of Sesamia calamistis Hampson (Lep.: Noctuidae). Biological Control 8; 15-21. Correa-Ferreira, B. S. and Moscardi, F. 1994. Temper- ature effect on the biology and reproductive per- formance of the egg parasitoid Trissolcus basalis (WoU.). Anais da Sociedade Entomologica do Brasil 23: 399-108. Correa-Ferreira, B. S. and Zamataro, C. E. O. 1989. Reproductive capability and longevity of the egg parasitoids Trissolcus basalis (WoUaston) and Trissolcus mitsukurii Ashmead (Hymenoptera; Scelionidae). Revista Brasileira de Biologui 49; 621- 626. Field, S. A., Keller, M. A. and Austin, A. D. 1998. Field ecology and behaviour of the egg parasit- oid Trissolcus basalis (WoUaston) (Hymenoptera; Scelionidae). Transactions of the Royal Society of South Australia 122: 65-71. Ganesalingam, V. K. 1966. Some environmental fac- tors influencing parasitisation of the eggs of Ne- zara viridula L. (Pentatomidae) by Telenornus ba- salis WoUaston (Hymenoptera: Scelionidae). Cey- lon Journal of Science (Biological Sciences) 6: 1-14. Godfray, H. C. ]. 1994. Parasitoids: behavioural and evo- lutionary ecolog}/. Princeton University Press, Princeton, 473 pp. Hamilton, W. D. 1967. Extraordinary sex ratios. Sci- ence 156; 477-188. Hardy, I. C. W. 1994. Sex ratio and mating structure in the parasitoid Hymenoptera. Oikos 69: 3-20. Hardy, I. C. W., Dijkstra, L. J., Gillis, J. E. M. and Luft, P. A. 1998. Patterns of sex ratio, virginity and developmental mortality in gregarious parasit- oids. Biological journal of the Linnean Society 64: 239-270. Hardy, 1. C. W., Ode, P. J. and Strand, M. R. 1993. Factors influencing brood sex ratios in polyem- bryonic Hymenoptera. Oecologia 93: 343-348. Hardy, I. C. W., Pedersen, J. B., Sejr, M. K. and Lin- deroth, U. H. 1999. Local mating, dispersal and sex ratio in a gregarious parasitoid wasp. Ethol- ogy 105: 57-72. Kelly, G. L. 1987. Factors affecting the success of Tns- solcus basalis (Hvmenoptera: Scelionidae) as a bi- ological control agent of the green vegetable bug. 250 Journal of Hymenoptera Research Nezara viriciula (Hemiptera: Pentatomidae). Un- published PhD thesis. University of Sydney, Syd- ney, 329 pp. Molbo, D. and Parker Jr, E. D. 19%. Mating structure and sex ratio variation in a natural population of Nnsi'ii/n I'itripennis. Proceedings of the Roi/nl Societt/ cf Lciidoii. Series B. Biological Sciences 263: 1703- 1709. Myint, W. W. and Walter, G. H. 1990. Behaviour of Si'alangia cameroni males and sex ratio theory. Oi- kos 59;' 163-174. Nadel, H. and Luck, R. F. 1985. Span of female emer- gence and male sperm depletion in the female- biased, quasi-gregarious parasitoid, Piicln/crepo- ideus vindcmiae (Hymenoptera: Pteromalidae). Annals of the Entomological Society of America 78: 410-414. Nadel, H. and Luck, R. F. 1992. Dispersal and mating structure of a parasitoid with a female-biased sex ratio: implications for theory. Evolutionary Ecolo- gy 6: 270-278. Noble, N. S. 1937. An egg parasite of the green veg- etable bug. Agricultural Gazette of Neiu South Wn/cs48: 337-341. Ode, P. J., Antolin, M. F. and Strand, M. R. 1997. Con- strained oviposition and female-biased sex allo- cation in a parasitic wasp. Oecologia 109: 547-555. Powell, J. E. and Shepard, M. 1982. Biology of Aus- tralian and United States strains of Trissolcus ba- salis, a parasitoid of the green vegetable bug, Ne- zara viridula. Australian Journal of Ecologi/ 7: 181- 186. Smith, J. H. 1945. Useful parasitic insects. Queensland Agricultural journal 60: 340-351. Thomas, J. W. Jr. 1972. Evaluation of Trissolcus basalts (Wollaston) as an egg parasite of Nezara viridula (Linnaeus). Unpublished MSc thesis, Louisiana State University, Louisiana, 99 pp. van den Assem, J., Gijswijt, M. J. and Niibel, B. K. 1980. Observations on courtship - and mating strategies in a few species of parasitic wasps (Chalcidoidea). Netherlands Journal of Zoology 30: 208-227. Waage, J. K. and Lane, J. A. 1984. The reproductive strategy of a parasitic wasp. II. Sex allocation and local mate competition in Trichogramma evanes- cens. journal of Animal Ecolog]/ 53: 417^26. Waage, J. K. and Ng, S. M. 1984. The reproductive strategy of a parasitic wasp. I. Optimal progeny and sex allocation in Trichogramma evanescens. journal of Animal Ecology 53: 401^15. Walter, G. H. and Clarke, A. R. 1992. Unisexual broods and sex ratios in a polyembr\'onic encyr- tid parasitoid (Copndosonia sp.: Hymenoptera). Oecologia 89: 147-149. Wilson, F. 1961. Adult reproductive behaviour in Asolcus basalts (Hymenoptera: Scelionidae). Aus- tralian journal of Zoology 9: 739-751. J. HYM. RES. Vol. 8(2), 1999, pp. 251-267 The Nearctic Species of Protarchiis Foerster (Hymenoptera: Ichneumonidae: Ctenopelmatinae) Luc Leblanc Eastern Cereal and Oilseed Research Centre, Agriculture and Agri-Food Canada, Ottawa, Ontario, KIA 0C6, Canada Abstract. — The Nearctic species of the Holarctic genus Protnrchus Foerster (Ichneumonidae, Cten- opelmatinae, Mesoleiini) are reviewed. A key to the Nearctic species is provided. Seven species are recognized. P. testntorius (Thunberg) and P. sarin (Ratzeburg) are Holarctic, P. bolbogaster Leblanc, n. sp. and P. mellipes (Provancher) are transcontinental Nearctic, P. magnus (Davis) and P. pallidiconiis (Walley) are eastern Nearctic, and P. atrofacies Leblanc, n. sp. is found in Alaska. The name P. longipes (Cushman) is synonymized with P. sorbi, new synonymy. The Holarctic genus Protarchus Foerster belongs to the ichneumonid subfamily Ctenopelmatinae (Scolobatinae sensu Townes 1970) and tribe Mesoleiini. The tribe is the most speciose and taxonomi- cally difficult in the subfamily. Only a few genera have been studied in detail (Kaur 1989, Leblanc 1989, Viitasaari 1979). The Palearctic species of Protarchus were re- vised by Viitasaari (1979), who recognized four species and discussed the biology of the species that parasitize sawflies of the genus Trichiosoina (Cimbicidae) in bogs, based on extensive collecting and rearing. The Nearctic Protarchus were studied by Walley (1938), who recognized four spe- cies based on specimens in the Canadian National Collection. Ichneumonids of the genus Protarchus are large sized, the nervellus in the hind wing is intercepted above its middle, the areolet is present, the clypeus is small and the medial dorsal carinae on the first me- tasomal tergite extend beyond the spira- cles. Both Nearctic and Palearctic species parasitize Trichiosoma spp. Study of the Nearctic and Palearctic specimens shows that the Nearctic P. lon- gipes (Cushman) and the Palearctic P. sorbi (Ratzeburg) represent a single Holarctic species. In addition, P. testatorius (Thun- berg) is Holarctic but had not been re- ported for the Nearctic region by previous authors. The discovery of Holarctic distri- bution patterns and of two undescribed Nearctic species encouraged me to revise the Nearctic species of Protarchus. MATERIALS, METHODS AND TERMINOLOGY Material studied. — A total of 87 Nearctic specimens were examined from 5 collec- tions as follows (collections acronyms are from Amett and Samuelson (1986)): AEIC: American Entomological Institute, Gaines- ville, Florida, D.B. Wahl, 31 specimens; ANSP: Academy of Natural Sciences, Phil- adelphia, Pennsylvania, D. Azuma, 2 spec- imens; CNCI: Canadian National Collec- tion of Insects, Ottawa, Ontario, J.R. Bar- ron, 43 specimens; LLIC: Luc Leblanc pri- vate collection, Montreal, Canada, 1 specimen; USNM: National Museum of Natural History, Washington, D.C., B. Danforth, 10 specimens. Morphological terms. — The morphologi- cal terms used in the descriptions are from Gauld (1984) except that mesosoma is used instead of thorax, and metasoma is used instead of gaster. Terms used to 252 Journal of Hymenoptera Research characterize microsculpture are from Al- len and Ball (1980). In characterizing me- tasomal punctures and sculpture, only the second tergum was utilized as sculpture is best defined on that segment. Color descriptions. — Areas used to de- scribe color patterns are as follows. Anten- nae are divided into dorsal and ventral surfaces and legs are divided into anterior, posterior, dorsal and ventral surfaces if imagined as stretched out horizontally at right angles from the body. Areas in bilat- erally symmetrical parts (face, clypeus, epicnemium, dorsal and ventral surfaces of mesosoma and metasomal terga) are identified as median /lateral and basal/ apical for face, clypeus and epicnemium, or median /apical and anterior /posterior for other parts. Areas in asymmetrical parts are identified as anterior /posterior and upper /lower. Measurements. — A total of 37 females and 20 males were measured with an oc- ular micrometer. Length of the forewing was measured from the junction of costal vein with humeral plate to the most dis- tant tip between the ends of veins Rs and M. Height of face is the distance between a line connecting the lower margins of an- tennal sockets and a line connecting the tentorial pits. Width of face is the distance between the inner margins of the com- pound eyes at the level of the middle of face height. Length of the first metasomal tergum was measured in lateral view from the base to the tip of its dorsal surface. Width of the first metasomal tergum was measured in dorsal view at its widest part near the apex. The length of the hind fe- mur was measured on the anterior surface from the base to the apex. The width of the hind femur was measured at its mid length. The widths of the hind tibia and hind tarsomere 1 were measured at their widest part in lateral view. Genus PROTARCHUS Foerster Protnrchus Foerster 1869:201. Type species: Try- pluvi rufus Gravenhorst (= testatorius Thun- berg). Designated by Woldstedt 1877:460. Lectotype not examined. Zacalles Foerster 1869:204. Type species: Zacalles mngnus Davis. Designated by Davis 1898:283. Synonymized by Cushman 1924:8. Holotype examined. Protarchoidcs Cushman 1922:25. Type species: Protarchoidcs loiigipes Cushman. Original des- ignation. Synonymized by Townes 1945:505. Holotype examined. Diagnostic characters. — Large (forewing 9.8 to 18.3 mm long). Clypeus (Fig. 3) small, transversly convex near base (near middle in P. sorbi), apically almost flat, and with apical margin truncate. Mandi- ble short and broad, its upper tooth a little wider and longer than lower tooth (Fig. 4) or much wider than lower tooth (Fig. 5). Forewing with areolet usually present, large (Fig. 28) or very small (Fig. 29). Ves- tige of vein 1/Rs + M (= ramulus) often present (Fig. 29) (always absent in other genera of Mesoleiini). Vein cu-a separated from vein Rs-I-M by 0.2 to 0.5 of its length. Hind wing with vein 1/Cu longer than vein cu-a ("nervellus intercepted above middle"). Tibial spurs of middle and hind legs unequal, the longest spur about 0.3 to 0.4 as long as first tarsomere. First tergum moderately stout, with median dorsal ca- rinae well defined and strong (Figs. 17,19), or reduced to a median furrow (Fig. 21) but always extending beyond spiracle. Terga 2 to 4 in some species each with two large sublateral swellings where punc- tures are sparser (Fig. 23). Hairs on female hypopygium directed backward. Biology and biogeography. — Hosts of Pro- tarchus are almost invariably larvae of Tri- chiosoma (Hymenoptera: Cimbicidae). Published host records for the Palearctic region are: T. nanae Vikberg & Viitasaari, parasitized by P. testatorius in Finland (Vi- itasaari 1979, Vikberg & Viitasaari 1991); T. ?lucoruni L., parasitized by P. testatorius and P. sorbi in Finland (Viitasaari 1976, 1979). Records of Palaeociiubex femorata, parasitized by P. testatorius (Townes et al. 1965) and Cindu'x parasitized by P. Iieros Volume 8, Number 2, 1999 253 (Holmgren 1876) would require confir- mation. Label data included with Nearctic specimens indicate that Trichiosoma trian- guliim is attacked by P. testatorius, P. sorbi and P. mellipes. Bogs are the usual habitat of Protarchus spp. (Viitasaari 1979). P. tes- tatorius, P. sorbi and P. bolbogaster release a strong odor when picked up, as indicat- ed by label data. The species P. sorbi and P. testatorius are present in the boreal zone of North America (Figs. 31-32) as well as across the Palearctic region. The new spe- cies P. atrofascies, from Alaska, may also be Holarctic. The remaining three species, on the other hand, are apparently restrict- ed to north-eastern North America. KEY TO NEARCTIC SPECIES OF PROTARCHUS 1. Hind tibiae entirely light colored, brown or yellowish brown; ocelli enlarged and hind ocelli separated by about their diameter (Fig. 1) 2 - Hind tibiae apically to entirely dark, reddish black to black; ocelli smaller and hind ocelli separated by more than their diameter (Fig. 2) 4 2. Epomia clearly defined and prominent (Fig. 12); metasoma black or brownish black beyond tergum 2; size forewing 12.3-16.0 mm long 3 - Epomia indistinct; metasoma generally mostly brown, but in some specimens black beyond tergum 2; forewing 16.5-18.3 mm long magniis (Davis) 3. Face black, or sometimes black with a yellowish brown median spot; mesosoma entirely black except light tegula; metasoma entirely black mellipes (Provancher) - Face brown or yellowish brown; mesosoma brownish black with extensive yellowish or- ange and brown markings; metasoma brownish black with tergum 1 and part of tergum 2 brown pallidicontis (Walley) 4. Metasomal terga black, with extensive orange on terga 2 to 4; areolet usually present and very small (Fig. 29) testatorius (Thunberg) - Metasomal terga entirely black; areolet large (Fig. 28) or absent 5 5. Metasomal terga 2 to 4 each with two large sublateral swellings (Fig. 22); upper mandibular tooth much wider than lower tooth (Fig. 5); hind tibia entirely black bolbogaster Leblanc, n. sp. - Metasomal terga 2 to 4 without sublateral swellings (Fig. 23); upper mandibular tooth subequal to lower tooth (Fig. 4); hind tibia basally yellow to orange and apically black ... 6 6. Notaulus weak and reduced to shallow impressions (Figs. 6,7); hind tibia of male black in apical 0.6; face of male yellow sorbi (Ratzeburg) - Notaulus anteriorly strong and sharply defined (as Figs. 8,9); hind tibia of male black m apical 0.5; face of male black except faint brown median spot (female unknown) atrofacies Leblanc, n. sp. Protarchus atrofacies Leblanc, new species (Fig. 32) Diagnostic combination. — Face 1.2 X as wide as high (wider in other species), ar- eolet absent, face of male black except faint light spot (female unknown), meso- soma and metasoma predominantly black. Description. — Structure: Antenna with >20 flagellomeres (broken). Ocelli en- larged and sitting on a swelling, hind ocel- li separated by about their diameter. An- tennal sockets in lateral view forming a moderately strong angle with vertical axis of compound eye. Frons not strongly de- pressed. Face 1.2 X as wide as high. Me- dian swelling of face moderate. Upper mandibular tooth subequal in size and shape to lower tooth. Notaulus strong and sharply defined only at anterior end of mesoscutum. Epomia weak. Mesopleuron separated by 0.5-2.0 of their diameter and 254 Journal of Hymenoptera Research 3 ■^^ ^ v^^^i Figs. 1-12. 1-2, head in dorsal view: 1, P. mngniif, 2, I', tcftatoniis. 3, face ot i'. -n/i'i 4-S, mandibular teeth; 4, P. testiitoriiis; 5, P. bolbof:af,ter. 6-11, niesoscutum and notauli, dorsal and lateral: 6-7, P. scrbi; 8-9, /'. ningniif, 10-11, P. (I'sfnfpriKS. 12, pronotum of P. iiwUipcf, with epomia (EPM). microsculpture with meshes well outlined and sculpticells convex; well defined on posterior 0.3 of mesopleuron below spec- ulum and gradually fading towards other parts of mesepisternum. Carinae of pro- podeum strongly defined. Forewing 12.3 mm long, areolet absent. Hind femur 3.7 mm long and 6.2 X as long as wide. Hind tarsomere 1 weakly compressed laterally and 0.55 X as wide as hind tibia near apex. First metasomal tergum 2.1 X as long as wide, with median dorsal carinae reduced to a median furrow; in lateral view weakly and regularly curved at mid- length. Sublateral swellings on terga ab- sent. Metasomal tergum 2 with punctures almost coalescent with outlines almost po- lygonal when close, and microsculpture Volume 8, Number 2, 1999 255 Figs. 13-24. 13-14, mesopleuron: 13, P. sor/'i; 14, P. l'olbo\;ii^U-i. 15-16, propiidt-um: 15, P. forbi; 16, P. Ic~^l(itnriiis. 17-21, metasomal tergum 1, dorsal and lateral: 17-18, P. holhogasfer. 19-20, P. forbi. 21, P. tefnfdmis. 22-24, metasomal tergum 2, dorsal: 22, P. maginis; 23, P. bolbogaster; 24, P. sorbi. well defined with sculpticells convex along posterior and lateral borders of ter- gum, at most suggested at centre of ter- gum and almost entirely faded anteriorly. Coloration: Antenna with scape, pedicel and flagellum black except traces of yel- lowish brown on flagellomeres 9 to 13. Head black except faint brown median spot on face below tubercle, brown clyp- eus and brown anterior surfaces and api- cal third of lateral surfaces of mandibles. Mesosoma black. Wings with light yellow infuscation. Legs with all coxae and tro- chanters black. Femur, tibia and tarsus of all legs orange except apical half of hind tibia and whole hind tarsus black. Meta- soma black. Etymology. — From the Latin aier (black) 256 Journal of Hymenoptera Research and fades (face), referring to the dark col- ored face. Distribution. — Known only from the type locality (Fig. 32). Specimen examined. — Holotype male: "Unalakleet, Alaska 8.viii.l961 R. Madge". Condition of type: missing left antenna be- yond flagellomere 17 and right antenna beyond flagellomere 20. [CNCI]. Protarchtts bolbogaster Leblanc, new species (Figs. 5,14,17,18,23,30) Diagnostic combination. — Upper mandib- ular tooth much wider than lower tooth (Fig. 5) (subequal in other species), subla- teral swellings on metasomal tergites 2 to 4 (Fig. 22) (absent in other species), me- sosoma and metasoma predominantly black. Description. — Structure: Antenna with 48-53 flagellomeres. Ocelli of moderate size and not sitting on a swelling, hind ocelli separated by more than their diam- eter. Antennal sockets in lateral view forming a moderately strong angle with vertical axis of compound eye. Frons not strongly depressed. Face 1.7-1.8 (female) and 1.5-1.7 (male) X as wide as high. Me- dian swelling of face moderate. Upper mandibular tooth much wider than lower tooth (Fig. 5). Notaulus strong and sharply defined to middle of mesoscutum. Epomia indistinct. Mesopleuron (Fig. 14) with punctures separated by 0.5-2.0 of their di- ameter and microsculpture with meshes lightly impressed and sculpticells slightly convex, but more convex anterior to spec- ulum. Carinae of propodeum strongly de- fined. Forewing 13.6-15.2 (female) and 13.9-14.8 (male) mm long, areolet present and large. Hind femur 4.2-4.6 mm long and 5.2-6.2 X as long as wide. Hind tar- somere 1 weakly compressed laterally and 0.50 X as wide as hind tibia near apex. First metasomal tergum 1.8-2.1 X as long as wide, with median dorsal carinae well defined and strong (Fig. 17); in lateral view decurved with a strong angle before midlength (Fig. 18). Terga 2 to 4 each with two large sublateral swellings with spars- er punctures (Fig. 23). Metasomal tergum 2 (Fig. 23) with punctures almost coales- cent with outlines polygonal, but sparser and faded on sublateral swellings, and mi- crosculpture well defined along posterior and lateral borders, but with meshes much smaller and flat along anterior bor- der. Dorsal notch on ovipositor with an- terior margin without a strong angle, gradually sloping. Coloration: Female. — Antenna with scape and pedicel reddish black, and flagellum with flagellomeres dorsally reddish brown and ventrally brownish orange gradually turning red- dish brown near apex. Head black except anterior surfaces and apical half of lateral surfaces of mandibles brown. Mesosoma black except tegulae yellowish brown. Wings with yellow infuscation. Legs or- ange except hind tibia and tarsus reddish black to black. Metasoma black. Male. — Antenna with scape and pedicel brown, and flagellum with flagellomeres dorsally reddish brown and ventrally brown. Head black except usually brown median lon- gitudinal spot on face below tubercle, brown to reddish black clypeus and brown anterior surfaces and apical half of lateral surfaces of mandibles. Mesosoma black except tegulae yellowish brown. Wings with yellow infuscation. Legs or- ange except hind tibia and tarsus reddish black to black. Metasoma black. Etymology. — From the Greek bolbos (swelling) and gaster (belly), referring to the characteristic sublateral swellings on metasomal terga 2 to 4. Distribution. — Transcontinental in cold temperate and boreal regions (Fig. 30). Specimens examined. — 3 females and 6 males. Holotype male, "Estes Pk. Colo.[rado] 7500 ft 7 mi. E. 8-18-48 Evans"; "Protarchus Det. W.R.M. Mason '48". Condition of type: intact. [CNCI]. Paratypes: CANADA. BRITISH COLUM- BIA:' Jesmond, 14. ix. 1938, J.K. Jacob (1F,CNCI); Racing River, 2400', Volume 8, Number 2, 1999 257 24.viii.1973, H.&M. Townes (1F,AEIC). ONTARIO: Orrville, 21.vii.l958, L.L. Pe- chuman (1M,AEIC). QUEBEC: Lac Rol- land, R.I.F. 41, specimen no. 12150-B (1M,CNCI). YUKON TERRITORY: 14 mi.E.Dawson, 1300', 30.vi.l962, R.E. Leech (1F,CNCI) (used for SEM). UNITED STATES. MAINE: Dryden, 30.viii.l959 (1M,AEIC); Roque Bluff, 10.viii.l907, J.A. Cushman, "Pwtarclwides maiidibulnris" Al- lotype # 25975, USNM (1M,USNM). MICHIGAN: Huron Mts, 25.viii.1959, H. Townes, "strong Pimpla odor" (1M,AEIC). Remarks. — This species is closely related to the Palearctic P. heros (Holmgren), both species sharing as synapomorphies the unique mandible shape and the large sub- lateral swellings on terga 2 to 4. P. holbo- gaster differs from P. hews by its predom- inantly orange legs. Cushman's original allotype of P. mamiibularis belongs to P. bolbogaster. Protarchus magnus (Davis) (Figs. 1,8,9,22,26,30) Znccnles mngnus Davis 1898 (1897):283. Diagnostic combination. — Whole body predominantly light colored (brown), hind tibiae entirely light colored, ocelli en- larged and hind ocelli separated by about their diameter, dorsal notch on ovipositor with anterior margin with strong angle, abruptly sloping (Fig. 26). Description. — Structure: Antenna with 47^8 flagellomeres. Ocelli enlarged and sitting on a swelling, hind ocelli separated by about their diameter (Fig. 1). Antennal sockets in lateral view forming a moder- ately strong angle with vertical axis of compound eye. Frons not strongly de- pressed (Fig. 1). Face 1.5-1.7 X as wide as high. Median swelling of face very weak. Upper mandibular tooth subequal in size and shape to lower tooth. Notaulus strong and sharply defined only at anterior end of mesoscutum (Fig. 8,9). Epomia weak. Mesopleuron with punctures separated by 0.5-2.0 of their diameter and microsculp- ture with sculpticells very convex, making surface mat; meshes well outlined on pos- terior half below speculum, but gradually fading anteriorly. Carinae of propodeum strongly defined. Forewing 16.5-18.3 mm long, areolet present and large. Hind fe- mur 4.6-5.2 mm long and 6.8-8.0 X as long as wide. Hind tarsomere 1 weakly compressed laterally and 0.50-0.55 X as wide as hind tibia near apex. First meta- somal tergum 2.1-2.6 x as long as wide, with median dorsal carinae reduced to a median furrow; in lateral view weakly and regularly curved at midlength. Sub- lateral swellings on terga absent. Metaso- mal tergum 2 (Fig. 22) with punctures round and not coalescent, and microsculp- ture uniform over all tergum, with sculp- ticells convex. Dorsal notch on ovipositor with anterior margin with a strong angle, apruptly sloping (Fig. 26). Coloration: Fe- male.— Body uniformly brown except the following. Yellowish brown parts are an- tennae, clypeus, mandibles, hind corner lobe of pronotum, tegula, subalar promi- nence, mesepimeron and legs beyond cox- ae. Black parts are base of first antennal flagellomere and narrow bands at apices of metasomal terga 3 to 8. Wings with yel- low infuscation. Color variation: Some fe- males have extensive black markings: frons, vertex and genae, propleuron, all pronotum except upper and posterior margins and hind corner lobe, dorsal sur- face of forecoxa, anteromedian band on mesoscutum to half of mesoscutum length and two bands on mesoscutum lateral to notauli, upper fourth of epicnemium, up- per fourth to third of mesopleuron except anterior and posterior margins, subtegular ridge, speculum, mesosternum except an- terolateral corners, large central spot on metasomal tergum 2 and whole metasoma beyond tergum 2. Distribution. — Northeastern United States and Southeastern Canada (Fig. 30). Specimens examined. — 10 females. Holo- type female, 4 labels, "N!N"; "HoloTYPE 4340"; "Zaccales magnus Davis"; "Collec- 258 Journal of Hymenoptera Research Figs. 25-27. 25, P. testatoriiis, metasonial tergum 2, dorsal. 26-27, ovipositor; 26, P. magnum; 27, P. sorb/. tion of THE ACADEMY OF NATURAL SCIENCES of Philadelphia. ANSP". Con- dition of type: missing left flagellum be- yond flagellomere 11, right antenna, left fore wing; right fore and hind wings pinned below specimen. [ANSP]. Other specimens: CANADA. QUEBEC: Joliette Co., Ste-Beatrix, 24.VIII.1978 (1F,LLIC) (used for SEM). UNITED STATES. MICH- IGAN: Midland, vii.1950 (1F,AEIC). NEW YORK: Allegany St.Pk., 30.vii.l938 (1F,USNM), 31.vii.l938 (1F,USNM), A.R. Shaddle; Essex Co., Keene Valley, 17.ix.l917, H. Nortman (1F,ANSP); Six Miles Creek, Ithaca, 17.vii.l947, J.G. Fran- clemont (3F,AEIC). PENNSYLVANIA: Glenside, "10.12.1929", G.G. Sleesman (1F,USNM). Protarchus mellipes (Provancher) (Figs. 12,31) Coelocentrus mellipes Provancher 1886:113. Protarchoides mellipes: Walley 1938:231. Protarchoides pnllipes Cushman 1927:15. Synon- ymy by Walley 1938:231. Diagnostic combination. — Epomia clearly defined and prominent (Fig. 12), ocelli en- larged and hind ocelli separated by about their diameter, face entirely black, or at most with yellowish brown median spot, mesosoma and metasoma almost entirely black, hind tibiae entirely light colored. Description. — Structure: Antennal flagel- lum with 43^6 (female) and 41^3 (male) flagellomeres. Ocelli enlarged and sitting on a swelling, hind ocelli separated by about their diameter. Antennal sockets in lateral view forming a moderately strong angle with vertical axis of compound eye. Frons not strongly depressed. Face 1.5-1.7 X as wide as high. Median swelling of face moderate. Upper mandibular tooth subequal in size and shape to lower tooth. Notaulus strong and sharply defined only at anterior end of mesoscutum. Epomia clearly defined and prominent (Fig. 12). Mesopleuron with punctures separateed by 0.5-2.0 of their diameter and micro- sculpture with sculpticells convex, making surface slightly mat; meshes well outlined on posterior half below speculum, but gradually fading anteriorly. Carinae of propodeum well defined and strong. Fore- wing 13.8-16.0 (female) and 12.3-13.2 (male) mm long. Areolet present and large. Hind femur 3.9-4.9 mm long and 6.0-7.0 X as long as wide. Hind tarsomere 1 weakly compressed laterally and 0.50- 0.55 X as wide as hind tibia near apex. First metasomal tergum 2.3 x as long as Volume 8, Number 2, 1999 259 wide with median dorsal carinae reduced to a median furrow and; in lateral view weakly and regularly curved at mid- length. Sublateral swellings on terga ab- sent. Metasomal tergum 2 with punctures round and not coalescent, and microsculp- ture uniform over all tergum with sculp- ticells convex. Dorsal notch on ovipositor with anterior margin without a strong an- gle, gradually sloping. Coloration: Fe- male.— Antenna with scape and pedicel black or black with ventral surfaces yel- lowish orange, and flagellum with basal section yellowish orange, except basal half of first flagellomere and dorsal surfaces of first few flagellomeres generally black, and apical section black to brown. Head black except clypeus reddish brown to yellowish brown, lower margin of face along clypeus sometimes yellowish brown and mandibles black to yellowish brown. Mesosoma black except yellowish brown to reddish black tegula. Wings with yel- low infuscation. Legs yellowish brown ex- cept fore, middle and hind coxae and usu- ally fore, middle and hind trochanters 1 black, usually a narrow brownish orange longitudinal line on black dorsal surface of hind coxa and brownish orange fore and middle tarsomeres 5. Metasoma black. Male. — Antenna with scape and pedicel black, and flagellum with basal section yellowish orange, except dorsal surfaces of first few flagellomeres black and with apical section dark brown. Head black except clypeus yellowish brown and, in one specimen, faint traces of yel- lowish brown on face. Mesosoma black except tegula yellowish brown. Wings with yellow infuscation. Legs yellowish brown except fore, middle and hind coxae and trochanters 1 reddish black. Metaso- ma black. Distribution. — Transcontinental in bore- al region (Fig. 31). Specimens examined. — 5 females and 2 males. Type material: Lectotype of Coelocen- trus mellipes Provancher, designated by Barron 1975:508; male, 6 labels: "Holotype male Coleocentrus mellipes Provancher No. 4235"; "Coleocentrus n. spec!"; "G 484"; "This must be the type of Coleocentrus mel- lipes Prov. which g.[ahan] & R.[ohwer] could not locate. It was evidently returned to geddes by Prov. & the label [# 2] in red ink is in Geddes hand (note by G.S. Wal- ley april/20/37)"; "Lectotype Coleocentrus mellipes Provancher Comeau '40"; "LEC- TOTYPE Coleocentrus mellipes Provancher G 484 Barron '71". Specimen from Rocky Mountains according to Walley, 1938. Condition of type: missing right fore and middle legs beyond coxae, left middle leg beyond tibia and both hind tarsi but tar- someres 1-2 of one hind leg glued on first label. [CNCI]. Holotype of Protarchoides pallipes Cushman. Female. 3 labels: "Ed- monton, Alberta 23.viii.1916 G. Salt"; "Type No. 40444 U.S.N.M."; Protarchoides pallipes Type. Cush.". Condition of type: missing entire left antenna and right hind tarsus; left hind tarsus broken and re- glued between tarsomeres 2 and 3. [USNM]. Other specimens. CANADA. AL- BERTA: 15 mi.E.Morley, 14.viii.l962, K.C. Herrmann (1F,CNCI) (used for SEM); Ed- monton, 23.Vm.1926, G. Salt, paratype # 40444, USNM (1F,USNM). BRITISH CO- LUMBIA: Houston, 17.vi.l959, Forest In- sect Survey specimen no. 58-1792-0119, ex. Trichiosoma triangulum (1M,CNCI); Rob- son, 13.ix.l949, H.R. Foxlee (1F,CNCI). ONTARIO: Smoky Falls, near Kapuska- sing, 9.viii.l937, R.V. Whelan (1F,CNCI). Remarks. — The specimen used by Pro- vancher, collected by G. Geddes in the Rocky Mountains (Provancher 1886), could not be located by Rohwer in his 1915 visit to the Provancher collection (Gahan and Rohwer 1917, Cushman and Rohwer 1920). Walley (1938) discovered a specimen in the Geddes collection, donat- ed to CNCI, labelled "Coleocentrus n sp" and agreeing with Provancher's descrip- tion. He accepted this specimen as the one originally used by Provancher. Barron (1975) designated the specimen as lecto- type. 260 Journal of Hymenoptera Research Figs. 28-29. Forewing, showing areolet: 28, P. sorhi; 29, P. testatcriKS. Protarchus pallidicomis (Walley) (Fig. 30) Protarchoides pnllidicoruis Walley, 1938:231. Diagnostic combination. — Epomia clearly defined and prominent (Fig. 12), ocelli en- larged and hind ocelli separated by about their diameter, face entirely brown or yel- lowish brown, mesosoma and metasoma predominantly brownish black with ex- tensive yellowish orange and brown markings, hind tibiae entirely light col- ored. Description. — Structure: Antenna with 45 (female) flagellomeres. Ocelli enlarged and sitting on a swelling, hind ocelli sep- arated by about their diameter. Antennal sockets in lateral view forming a moder- ately strong angle with vertical axis of compound eye. Frons not strongly de- pressed. Face 1.5 X as wide as high. Me- dian swelling of face moderate. Upper mandibular tooth subequal in size and shape to lower tooth. Notaulus strong and sharply defined only at anterior end of mesoscutum. Epomia clearly defined and prominent. Mesopleuron with punctures separated by 0.5-2.0 of their diameter apart and microsculpture with sculpticells convex, making surface slightly mat and meshes well outlined on posterior half be- low speculum, but gradually fading ante- riorly. Carinae of propodeum strongly de- fined. Forewing 14.5 (female) and 12.8 (male) mm long, areolet present and large. Hind femur 4.3 (female) and 4.7 (male) mm long and 6.9 (female) and 7.8 (male) X as long as wide. Hind tarsomere 1 weakly compressed laterally and 0.50 x as wide as hind tibia near apex. First meta- somal tergum 2.4 x as long as wide, with median dorsal carinae reduced to a me- dian furrow; in lateral view weakly and regularly curved at midlength. Sublateral swellings on terga absent. Metasomal ter- gum 2 with punctures round and not co- alescent, and microsculpture uniform over all tergum with sculpticells convex. Dorsal notch on ovipositor with anterior margin without a strong angle, gradually sloping. Coloration: Female. — Antenna brownish orange except dorsal half of scape brown- ish black. Head brown except frons, area around ocelli and vertex behind ocelli brownish black. Mesosoma brownish black except the following. Yellowish or- ange parts are hind corner lobe of prono- tum, tegula, subalar prominence, anterior margin of mesopleuron, mesepimeron and upper division of metapleuron. Brown parts are sublateral longitudinal bands on mesoscutum along notauli from anterior margin to three quarters of scutum length and lateral bands along mesoscutum mar- gin from wing tegula to end of mesoscu- tum, lower half of epicnemium, anterolat- eral corner of mesosternum, lower margin of mesopleuron, speculum, scutellum in- cluding axillae, postscutellum, upper an- terior corner of metapleuron, anterior half of area lateralis of pronotum and along median furrow of propodeum. Wings Volume 8, Number 2, 1999 261 • P A P ■ P bolbogai^ter magnus pallidicoirnis Fig. 30. Distribution of P. I'ltlbogaster (circles), P. magnus (triangles) and P. pallidicornis (squares). with yellow infuscation. Legs yellowish brown except faint traces of brownish black on dorsal surface of forecoxa and apical half of dorsal surfaces of mid and hind coxae. Metasoma brownish black ex- cept first tergum brown and faint traces of brown on second tergum. Male. — Antenna with scape and pedicel dorsally brownish black and ventrally yellowish brown, and flagellomere 1 yellowish brown (rest of antennae missing). Head brownish black except face, clypeus, malar space and mandibles yellowish brown. Mesosoma brownish black except the following. Yel- lowish orange parts are hind corner lobe of pronotum, tegula, subalar prominence and mesepimeron. Brown parts are two sublateral longitudinal bands on mesoscu- tum along notauli from anterior margin to three quarters of scutum length and two lateral bands along margins of scutum from base of sublateral band to three quar- ters of scutum length, lower two thirds of epicnemium, anterolateral corner of me- sosternum, lower margin of mesopleuron, speculum, scutellum including axillae, postscutellum and entire propodeum ex- cept faint brownish black on metapleuron. Wings with yellow infuscation. Legs yel- lowish brown. Metasoma brownish black except first tergum and basal half of sec- ond tergum brown and faint brown traces on anterolateral corners of third tergum. Distribution. — Ontario, Maine (Fig. 30). Specimens examined. — 1 female and 1 male. Holotype, examined, female, 2 labels: "Holotype female Protarchoides pallidicor- nis Walley No. 4410"; "Smokey Falls, Ont[ario] (near Kapuskasing) Aug 4, 1937 R.V. Whelan". Condition of type: missing hind tarsomeres 4-5. [CNCI]. Other speci- men. UNITED STATES. MAINE: Glen- burn, 5.vii.l928, Gypsy moth Lab, 10088 NIG, ex. Tenthredinid (!) (1M,USNM). Remarks. — This species is closely related to P. mellipes. The prominent epomia (Fig. 12) is a synapomorphy. Protarchtis sorbi (Ratzeburg) (Figs. 3,6,7, 1 3, 1 5, 19,20,24,27,28,31 ) Trypho)! sorbi Ratzeburg 1844:126. Holotype lost. PsUosargc (!) longipes Ashmead, in Slosson 1902: 321 {noincn nmiuw). 262 Journal of Hymenoptera Research Protarchoides longipes Cushman 1922:26. Protar- chus longipes: Townes, 1945:505. NEW SYN- ONYM. Protnrchoides mainiibidnris Cushman 1924:9. Des- ignated synonym to P. longipes by Townes 1945:505. NEW SYNONYM. Diagnostic combination. — Hind tarsomere 1 strongly compressed laterally and 0.55- 0.70 as wide as hind tibia near apex, me- tasomal tergum 1 decurved with a strong angle (Fig. 20), face of male yellow, me- sosoma and metasoma predominantly black. Description. — Structure: Antenna with 36-45 flagellomeres. Ocelli of moderate size and not sitting on a swelling, hind ocelli separated by more than their diam- eter. Antennal sockets in lateral view forming a moderately strong angle with vertical axis of compound eye. Frons not strongly depressed. Face (Fig. 3) 1.6-1.9 X as wide as high. Median swelling of face almost absent. Upper mandibular tooth subequal in size and shape to lower tooth. Notaulus weak and reduced to shallow impressions (Figs. 6,7). Epomia indistinct or very weak. Mesopleuron (Fig. 13) with punctures separated by less than 0.1 of their diameter and microsculpture vari- able in different specimens from absent over all mesopleuron to almost absent with meshes absent to lightly convex be- low speculum to irregularly striated through fusion of sculpticells on posterior 0.5 of mesopleuron below speculum. Ca- rinae of propodeum strongly defined (Fig. 15). Forewing 9.9-13.9 (female) and 9.8- 12.5 (male) mm long, areolet present and large (Fig. 28). Hind femur 2.8-3.9 mm long and 5.2-6.5 X as long as wide. Hind tarsomere 1 strongly compressed laterally and 0.55-0.70 x as wide as hind tibia near apex. First metasomal tergum 1.4-1.9 X as long as wide, with median dorsal carinae well defined and strong (Fig. 19) or re- duced to a median furrow; in lateral view decurved with a strong angle before mid- length (Fig. 20). Sublateral swellings on terga absent. Metasomal tergum 2 (Fig. 24) with punctures very dense, polygonal, crater-like but not coalescent, and micro- sculpture with sculpticells convex along posterior border of tergum, otherwise completely faded on the rest of tergum. Dorsal notch on ovipositor with anterior margin without a strong angle, gradually sloping (Fig. 27). Coloration: Female. — An- tenna with scape and pedicel black, and flagellum with flagellomeres dorsally brownish black and ventrally brownish orange and gradually turning brownish black near apex, or sometimes entire fla- gellum brownish black. Head black except anterior surface of mandible to entire mandible yellowish orange, and clypeus brownish orange to reddish black or black. Mesosoma black except hind corner lobe of pronotum and tegula reddish brown. Wings hyaline. Legs orange except sometimes traces of reddish black anteri- orly at base of each coxae, and apical two thirds to nine tenth of hind tibia and whole hind tarsus reddish black or black. Metasoma black. Male. — Antenna with scape and pedicel reddish black except usually yellow ventral surface of scape and sometimes yellow ventral surface of pedicel, and flagellum with flagellomeres dorsally brownish black and ventrally brownish black to brownish orange. Head black except orbits between compound eyes and antennal sockets yellow, and face, clypeus and mandibles yellow. Me- sosoma black except tegula yellow. Wings hyaline. Legs orange except apical six tenths of hind tibia and whole hind tarsus reddish black or black and sometimes yel- low fore and middle coxae, fore, middle and hind trochanters, foretibia, foretarsus and basal half of middle tibia. Metasoma black. Distribution. — Palearctic (Sweden, Fin- land, Russia) and transcontinental in North America (Fig. 31). Specimens examined. — 17 females and 14 males. Type material. Holotype of Protar- choides longipes Cushman. Female. 5 labels: "MT.WASH'[ingto]N."; "39"; "Type No. Volume 8, Number 2, 1999 263 • P. sorbi ( A P. mellipesi Fig. 31. Distribution of P. sinbi (circles) and P. melliftef (triangles). <^^t^t:> /" p. testaioriu^ P. atrofacies Fig. 32. Distribution of P. tc^tatoriin (circles) and P. u/ra/flot's (triangles). 264 Journal of Hymenoptera Research 25030 U.S.N.M."; "Psilosage longipes .Type Ashm[ead]"; "Protarchoides longipes Type. Cush.". [New Hampshire, A.T. Slosson, 1902]. Condition of type: missing left hind tarsus; both antennae broken and glued on first label. [USNM]. Holotype of Protar- choides mandibiilaris Cushman. Female. 4 labels: "Wellington B.C. 28.V1I.[19]04"; "1260"; "Type No 25975 U.S.N.M."; "Pro- tarchoides mandihularis Type. Cush.". Con- dition of type: missing: left antenna be- yond flagellomere 7, right antenna beyond flagellomere 37, right foreleg tarsus be- yond tarsomere 1 and right hind leg tarsus beyond tarsomere 1; right antenna section containing flagellomeres 4 to 37 and right hind leg beyond coxa broken and glued on locality label. [USNM]. Other specimens: CANADA. ALBERTA: Edmonton, 28.vii.1926, E.H. Strickland (1M,AEIC); Entranse, emerged 21.iii.l950, Forest In- sect Survey specimen no. A113313, ex. Tri- chiosoma sp. (1M,CNCI). BRITISH CO- LUMBIA: Summit Lake, Mile 392, Alaska Hwy, 4600', 16.vii.l959 (1F,CNCI), 4700', 15.vii.l959 (1F,CNC1), 5000', 6.vii.l959 (1M,CNCI); Stone Mtn Nat. Pk., "3800', 12.vii.l975, 13.vii.l973, 18.vii.l973, 20.vii.l973 (7M,AEIC). MANITOBA: Christopher Lake, emerged 4.iii.l963, For- est Insect Survey specimen no. 1914(02), ex. Trichiosorna triangulum (1F,CNCI); Tur- tle Mtn., 21.vii.l953, Brooks-Kelton (1F,CNCI); Wanless, 29.vi.1961, H.E. Mil- liron (1F,CNCI). PRINCE EDWARD IS- LAND: Alberton, 15.vii.l940, G.S. Walley (1F,CNCI). QUEBEC: Hemmingford, l.vii.l928, G.H. Hammond (1F,CNCI); Aylmer, 20.V.1920, G.H. Hammond, ex. Trichiosorna sp. (1M,CNCI). SASKATCH- EWAN. Buffalo Narrov^s, emerged 4.iii.l964, Forest Insect Survey specimen no. 63W-2241(03), ex. Trichiosorna triangu- laris (!) (1F,CNCI); Candle Lake, emerged 10. ii. 1951, Forest Insect Survey specimen no. W50-42786, ex. Trichiosorna triangulum (1M,CNC1); Parr Hill, emerged 17.ii.l961, Forest Insect Survey specimen no. W61- 2392(04), ex. Trichiosorna triangulum (1F,CNCI); Waskesiu, 21. vi. 1938, J.G. Rempel (1F,AEIC). UNKNOWN PROV- INCE: Crimson Lake, emerged before 12.vii.l961, Forest Insect Survey specimen no. 60A1409-03, ex. Trichiosoma sp. (1F,CNCI) (used for SEM). UNITED STATES. ALASKA: Seaward, <300', 25.vii.1951, W.J. Brown (1F,CNCI). OREGON: Mt. Hood, 3500', 19.vii.l978, 24.vii.1978, H.&M. Townes (2F,AEIC). WASHINGTON: Mt. Rainier, 4200', 15.vii.l940, H.&M. Townes, "odor like Pirnpla" (1M,AEIC). UNKNOWN LOCAL- ITY. Reared specimen labelled "99 88281" (1M,CNCI). Protarchus testatorius (Thunberg) (Figs. 2,4,10,11,16,21,25,29,32) Ichneiinion testatorius Thunberg 1822:276. Lec- totype (designated by Roman, 1912) not ex- amined. Tryphon ruftis Gravenhorst 1829:200. Holotype lost. Mesoleius (Protarchus) welanurus Thomson 1895: 2020. Leptotype designation and synonymy by Viitasaari 1979:34. Lectotype not exam- ined. Diagnostic combination. — Areolet very small or sometimes absent (Fig. 29), me- tasoma black with extensive orange mark- ings on tergites 2 to 4. Description. — Structure: Antenna with 40-44 flagellomeres. Ocelli of moderate size and not sitting on a swelling, hind ocelli separated by more than their diam- eter (Fig. 2). Antennal sockets in lateral view forming a very strong angle with vertical axis of compound eye. Frons strongly depressed (Fig. 2). Face 1.3-1.4 X as wide as high. Median swelling of face moderate. Upper mandibular tooth sub- equal in size and shape to lower tooth (Fig. 4). Notauli strong and sharply de- fined to middle of mesoscutum length (Figs. 10,11). Epomia indistinct. Mesopleu- ron with punctures separated by 0.5-2.0 of their diameter and microsculpture gener- ally absent, at most expressed as slightly convex sculpticells below speculum. Ca- Volume 8, Number 2, 1999 265 rinae of propodeum weakly defined to al- most absent (Fig. 16). Forewing 11.2-13.3 mm long. Areolet in forewing very small (Fig. 29) or sometimes absent. Hind femur 3.6-4.4 mm long and 6.0-6.8 x as long as wide. Hind tarsomere 1 weakly com- pressed laterally and 0.45-0.50 X as wide as hind tibia near apex. First metasomal tergum 1.8-2.2 X as long as wide, with median dorsal carinae reduced to a me- dian furrow (Fig. 21); in lateral view weakly and regularly curved at mid- length. Sublateral swellings on terga ab- sent. Metasomal tergum 2 (Fig. 25) with punctures very dense, slightly polygonal but not coalescent, but density decreasing towards posterior margin, and with mi- crosculpture well defined on apical 0.3, flat in middle 0.3 and completely faded basally. Dorsal notch on ovipositor with anterior margin without a strong angle, gradually sloping. Coloration: Female. — Antenna with scape and pedicel reddish black, and flagellum brownish orange ex- cept basal portion of flagellomere 1 red- dish black and dorsal surfaces of first few flagellomeres reddish black gradually turning brownish orange to yellowish or- ange near apex. Head black except the fol- lowing. Yellow parts are orbits between eyes and antennal sockets, clypeus, man- dibles and face except narrow median red- dish black line below tubercle. Brownish orange parts are genal orbits from poste- rior ocelli down to two thirds to a quarter of eye height. Mesosoma black except the following. Yellow parts are tegula, scutel- lum, postscutellum and often four small spots on anterior margin of mesoscutum on the sides of notauli. Wings with light yellow infuscation. Legs with coxae black except anteromedian line on or whole an- terior surface of forecoxa yellow and an- teroapical spot on middle coxa yellow. Trochanters 1 black except anterior sur- face of foretrochanter 1 and apex of dorsal surface of middle trochanter 1 yellow. Fore and middle legs beyond trochanters 1 yellowish orange except ventral basal quarter to half of middle femur reddish black. Hind trochanter 2 yellowish orange. Hind femur black except yellowish orange at base and apex. Hind tibia with basal 0.6 yellowish orange and apical 0.4 reddish black. Hind tarsi yellowish orange to brownish orange. Metasoma black with the following brownish orange: apical 0.25 of tergum 1, apical 0.33 to 0.5 of tergum 2 to entire tergum 2 except lateral black markings on basal 0.5, entire tergum 3 ex- cept sometimes a large median black spot, basal 0.15 of tergum 4 with a small sub- apical spot to entire tergum 4 except api- cal 0.1. Color variatio?is: Two females have brownish orange markings on propodeum anterior to postpectal carina and lateral to median furrow and extending laterally be- low lateral carinae. These markings are present in most Palearctic specimens of P. testatorius. Distribution. — Palearctic (Sweden, Fin- land, Germany, Russia) and transconti- nental in North America (Fig. 32). Remarks. — The lectotypes from Europe have not been examined because a good series of Palearctic material (AEIC, USNM) was studied and found to be con- specific with the Nearctic material. Specimens examined. — 28 females. AL- BERTA: Banff, 18.viii.l946, E.H. Strickland (1F,AEIC); Banff, Black's Camp Ground, emerged 28.vi.1958, Forest Insect Survey specimen no. 58A108-03, ex. Trichiosoma sp. on alder (1F,CNCI); Cameron Lake, 9.vii.l949, C.P. Alexander (1F,AE1C); Can- yon Creek, emerged 9.iii.l953, Forest In- sect Survey specimen no. A2182A, ex. Tri- chiosoma sp. (1F,CNCI); Eisenhower Junc- tion, Banff National Park, 9.viii.l962, K.C. Hermann (1F,CNCI); Fawcett, emerged 9.vii.l956, Forest Insect Survey specimen no. 55A1438-12, ex. Trichiosoma sp. (1F,CNCI); Kanamaskis, emerged 23.ii.1949, Forest Insect Survey specimen no. A-455-K, ex. Trichiosoma sp. (1F,CNCI); Lake Louise, 5600', 26.vii.1938, G.S. VVal- ley (1F,CNCI); Nordegg, 21.vii.l926, E.H. Strickland (1F,CNCI); 18 mi.W. Strachan, 266 Journal of Hymenoptera Research emerged 24.iii.1959, Forest Insect Survey specimen no. 58A1821-02, ex. Trichiosoma sp. (1F,CNCI). BRITISH COLUMBIA: Hope Mts, 20.viii.l931, A.N. Gartrell (1F,CNCI) (used for SEM); Saanich, emerged 23.V.1958, Forest Insect Survey specimen no. 58-3-01-A, ex. Trichiosoma triaugulum (1F,CNCI). ONTARIO: Sud- bury, 1891 (1F,CNCI). QUEBEC: Lac St- Jean, 20.viii.l939, P.L. Mercier (1F,AEIC); Kazabazua, 28.viii.1928, G.H. Fisk (1F,CNCI); Ste-Agathe des Monts, 7.viii.l937, G.S. Walley (1F,CNCI). YU- KON TERRITORY: Whitehorse, 7.viii.l948, W.R. Mason (1F,CNCI). UN- KNOWN PROVINCE: Hot Springs Road, 2.viii.l902, N.B. Sanson (1F,CNCI). FRANCE. Miquelon Island, 15.viii.l990, D. Abraham, collected in bog (1F,CNCI). UNITED STATES. ALASKA: Thompson Pass, 14.viii.l973, H.&M. Townes (1F,AEIC); Tsaina River, 17.viii.l973, 18.viii.l973, H.&M. Townes (2F,AEIC). COLORADO: Gould, 5.viii.l974, 6.viii.l974, H.&M. Townes (3F,AEIC). NEW HAMPSHIRE: Franconia (1F,USNM). WASHINGTON: Mt. Rainier, 22.vii.1940, H.&M. Townes, "strong Pim- pla odor" (2F,AEIC). ACKNOWLEDGMENTS The museum curators listed in the paper are sin- cerely acknowledged for their cooperation in the loan of specimens. For reviewing the manuscript, 1 thank Henri Goulet, from the Biological Resources Program, ECORC, Agriculture and Agri-Food Canada, Ottawa, as well as the three external reviewers. Dr. Goulet's contribution in accurately describing microsculpture is highly appreciated. Scanning Electron Micrographs were taken by K. Bolte, from ECORC. I want to ded- icate this study to Dr. John Barron (1932-1997), who has greatly helped me to improve this paper and who has always encouraged me to study Ctenopelmati- nae. LITTERATURE CITED Allen, R. T. and G. E. Ball. 1980. Synopsis of Mexican taxa of the Lo.\nth1ruf series (Coleoptera: Carabi- dae: Pterostichini). Transactions of tlw American Entomological Socicti/. 105:481-576. Arnett, R. H. and G. A. Samuelson. 1986. The insect and spider collections of llie world. E.J. Brill/ Flora and Fauna Publications. Gainesville, Florida. 220 PP- Barron, J. R. 1975. Provancher's collections of insects, particularely those of Hymenoptera, and a study of the types of his species of Ichneumonidae. Na- tiiraliste Camidien. 102:387-591. Cushman, R. A. 1922. On the Ashmead manuscript species of Ichneumonidae of Mrs. Slosson's Mount Washington lists. Proceedings of the United States National Museum. 61(8);l-30. Cushman, R. A. 1924. On the genera of the Ichneu- mon-flies of the tribe Paniscini, Ashmead, with discussions and descriptions and discussion of related genera and species. Proceedings of the United States National Museum. 64(20):l-48. Cushman, R. A. 1927. Miscellaneous notes and de- scriptions of Ichneumon-flies. Proceedings of the United States National Museum. 72(13):l-22. Cushman, R. A. and S. A. Rohwer. 1920. The North American Ichneumon-flies of the tribe Acoeniti- ni. Proceedings of the United States National Muse- um. 57:503-523. Davis, G. C. 1898(1897). A review of the Ichneumonid subfamily Tryphoninae. Transactions of the Amer- ican Entomological Society. 24:193-348. Foerster, A. 1869. Synopsis fiir Familien und Gattun- gen der Ichneumonen. Verli. Naturli. Ver. Rliein- lande. 25:135-227. Gahan, A. B. and S. A. Rohwer. 1917. Lectotypes of the species of Hymenoptera (except Apoidea) de- scribed by Abbe Provancher. Canadian Entomolo- gist. 49:391^00. Gauld, I. D. 1984. An introduction to the Ichneumonidae of Australia. British Museum (Natural History). London. 413 pp. Gravenhorst, J. L. C. 1829. Ichneumonologia Europaea, Vratislaviae, 1. 989 pp. Holmgren, A. E. 1876. Dispositio synoptica meso- leiorum scandinaviae. Koiigl. Svenska Vetenskap- sakad. Handl. (N.F.) 13:3. Kaur, R. 1989. A revision of the mesoleiinae genus Dentimachus Heinrich (Hymenoptera: Ichneu- monidae). Oriental Insects. 23:291-305. Leblanc, L. 1989. A taxonomic revision of the Nearctic species of Himerta (Hymenoptera: Ichneumoni- dae). Contributions of the American Entomological Institute. 25(3):l-76. Provancher, L. 1886. Additions et corrections nu volume 11 de la faiine entomologique du Canada. C. Darveau. Quebec. 475 pp. Fam IV. Ichneumonidae pp. 29- 121. Ratzeburg, J. T. C. 1844. Die Ichneumonen der Forstin- sekten in forstliclier und entomologisclwr Beziehung. Ein Anhang zur Abbildung und Beschreibung der Forstinsekten. 1, VlII:I-224. Roman, A. 1912. Die Ichneumonidentvpen C.P. Thun- bergs. Zool. Bidrag Upp'^ala 1:229-293. Slosson, A. T. 1902. Addilional list of insects taken in Volume 8, Number 1, 1999 267 Alpine Region of Mt. Washington. Eiiloniological News. 13:319-321. Thomson, C. G. 1895. Opiaai EntoiHotci\;ica. Fascicle 19: pp. 1971-2137. Thunberg, C. P. 1822. Ichneumonidea, Insecta Hy- menoptera illustrata. Memoires de I'Acadernie Iniperiale des Sciences de St.Petershourg. 8:249- 281. Townes, H. K. 1945. A catalogue and reclassification of the Nearctic Ichneumonidae (Hvmenoptera), Memoirs of the American Eiitoinolo^;iail Institute. No 11 (2):480-925. Townes, H. K. 1970. The genera of Ichneumonidae, part 3. Memoirs of the American Entomological In- stitute. 13:1-307. Townes, H. K., S. Momoi and M. Townes. 1965. A catalogue and reclassification of the Eastern Pa- learctic Ichneumonidae. Memoirs of the American Entomological Institute. 5:1-661. Viitasaari, M. 1976. Protarchus sorbi (Ratz.) new to the Finish fauna. Annates Entomologici Fennici 42:146. Viitasaari, M. 1979. A studv of the Palearctic species of the genus Protarchus Forster (Hvmenoptera, Ichneumonidae). Notiilae Ent. 59:33-39. Vikberg, V. and M. Viitasaari. 1991. Tncluosoma naiuie sp. n., a monophag on Betula nana from Finland (Hvmenoptera, Cimbicidae). Entomologica Fcnni- ca.' 2:67-77. VValley, G. S. 1938. Notes on the genus Protarchoides (Hymen., Ichneumonidae. Canadian Entomologist. 70:230-232. Woldstedt, F. W. 1877. Beitrag zur Kenntnis den um St. Petersbourg vorkommenden Ichneumoniden. Bulletin de I'Acadernie Imperiale des Sciences de St. Petersbourg. 23:432-460. J. HYM. RES. Vol. 8(2), 1999, pp. 268-269 BOOK REVIEW Geologische und biologische Entomookologie der rezenten Seidenbiene CoUetes. Volume I. Detlef Mader. Logabook, Koln, 1999. xliii + 807. Price: Germany DM 98.00 or 50.00 Euro (hardcover). ISBN 3-87361-263-1. In our concept-driven time dominated by molecular biology, few authors dare to publish lengthy accounts on pure natural history, summarizing raw facts of organ- ismal life or the idiosynchrasies of a par- ticular species. I happen to enjoy pulling those books off library shelves and im- mersing myself in organismal trivia. I fre- quently find a library note slipped into those books saying that they have not been circulating for a decade and that, un- less soon used, they will be permanently impounded in compact storage. So I reg- ularly check out these books from the li- brary, often to return them immediately. I like to think that this fools the librarians, such that they will grant these books an- other decade of accessible existence on a prime shelf. And I imagine that some fu- ture scientist will stumble, like me, over these books and keep alive some natural history minutia that otherwise may be buried forever. Detlef Mader's book on the Geological and Biological Entomo-Ecology of CoUetes Bees is one of those natural history ac- counts. It is thick in detail, particularly the nesting idiosynchrasies of the species Col- letes daviesanus. It is a specialist's book, therefore, and its audience may be rather limited. But it has enourmous depth in natural history, inspired by a Humbold- tian appetite for careful and complete doc- umentation. Leafing through pages and pages of flower records, or the geological details of the substrate used by C. davie- snniis for nest construction, I came to ad- mire Mader for taking the time to sum- marize his decades of work on the nesting habits of this species. Few of us ever make the time to do that for the species that are close to our heart. Bees in the basal bee genus CoUetes are best known for their unique nest architec- ture. The typical coUetid bee is solitary, constructs an underground nest (some species use twigs or rotting wood), and lines the nest tunnel with a cellophanelike tapestry derived from glandular secre- tions. A series of cells are constructed in- side this tunnel, separated by partitions of additional tapestry. The tapestry linings are tranlucent, giving the appearance that these bees keep their brood in a series of plastic bags. Hence also the vernacular name of the plastic-bag bee. Most colletids prefer to nest in level ground, but C. daviesnnus, the most com- mon species in central Europe, is an ex- ception. It prefers to nest in vertical em- bankments or cliffs, and there only in sub- strate of particular kinds, such as loose sandstone. Much of Maders' book is de- voted to documenting these substrate preferences for populations in central Ger- many, with comparative references to oth- er populations throughout Europe. This treatment has a rather geological flavor, and I admit that the geological aspects of this work were rather lost on me, as I as- sume they will be lost on anyone unfa- miliar with the geology and stratigraphy of central Europe. Other parts of Mader's book are more accessible, summarizing for example the literature on flowers visited by CoUetes, or profiling the communities of other soil-dwelling species that secondar- ily use tunnels of abandoned C. daviesaniis nests. Mader's book actually comes in two volumes, only the first of which is pub- lished at this point. The second volume is scheduled for publication next year, but a Volume 8, Number 2, 1999 269 table of contents is already available. While both volumes are organized around aspects of the nesting biology of C. dav- ieaniis and other Colletes species, the first volume seems to focus more on biological aspects (e.g., presenting information on Colletes parasites, flower visitation, etc.), while the second volume seems more spe- cialized and appears to concentrate more on geological aspects. Who would benefit from these vol- umes? Clearly the readership is rather lim- ited, yet it certainly should not be missing from libraries specializing on hymenop- teran literature or general natural history. But apart from Colletes afficionados inter- ested in knowing everything about Colle- tes, including the geological trivia of nest substrate choice of a single species in cen- tral Europe, I do not think that these vol- umes will find much use among bee bi- ologists. This is not a reflection of the quality of the work, but more a reflection of the currently perceived value of natural history accounts. So I imagine that the true value of these volumes will not be appar- ent until C. dai'iesaniis, though abundant at this point, may become extinct some time in the future. The same is certainly true for many other species for which time did not permit a comprehensive study of all life-history aspects, as for example the nesting habits of the passenger pigeon. Ulrich G. Mueller, Integrative Biology, Patterson Labs, The University of Texas at Austin, Austin, TX, 78712, USA. INSTRUCTIONS FOR AUTHORS General Policy. The Journal of H\/nwiwptera Racarch in\'ites papers of high scientific quality reporting comprehensive research on all aspects of Hymenoptera, including biology, behavior, ecology, systematics, taxonomy, genetics, and morphology. Taxonomic papers describing single species are unlikely to be accepted unless a strong case is evident, such as importance in economic entomology or with concurrent biology or ecology. Manuscript length generally should not exceed 50 typed pages; however, no upper limit on length has been set for papers of exceptional quality and importance, including taxonomic monographs at generic or higher level. All papers will be reviewed by at least two referees. The referees will be chosen by the appropriate sub- ject editor. However, it would be helpful if authors would submit the names of two persons who are compe- tent to review the manuscript. The language of publication is English. Summaries in other languages are acceptable. The deadline for receipt of manuscripts is 1 October (for the April issue) and 1 April (for the October issue). Format and Preparation. Three copies of each manuscript, including copies of illustrations, should be submitted on letter size or A4 paper, double spaced, with at least 25 mm margins on all sides. On the upper left of the title page give name, address and telephone and fax numbers of the author to whom all corre- spondence is to be sent. The paper should have a concise and informative title, followed by the names and addresses of all authors. The sequence of material should be: title, author(s), abstract, text, acknowledgments, literature cited, appen- dix, figure legends, figure copies (each numbered and identified), tables (each numbered and with heading). Each of the following should start a new page: (1) title page, (2) abstract, (3) text, (4) literature cited, (5) figure legends, (6) footnotes. Following acceptance of the manuscript, the author should provide the editor with one copy of the manuscript accompanied by a copy on diskette using DD, double sided computer diskettes — IBM compati- ble MS DOS 5.25 inch or IBM and Macintosh 3.5 inch diskettes. (Authors who do not have access to a com- puter should submit three copies of the manuscript.) The paper may be submitted in most PC and Mac word processor programs such as Microsoft Word, FullWrite Professional, WordPerfect, WriteNow, Nisus, Mac- Write, or MacWrite 11. If possible, all words that must be italicized should be done so, not underscored. Tables may be formatted in a spread sheet program such as MS Works or MS Excel. Text should be double spaced typing, with 25 mm left and right margins. Tables should be put in a separate file. Diskettes should be accompanied by the name of the software program used (e.g., WordPerfect, Microsoft Word). Authors should keep backup copies of all material sent to the Editor. The Society cannot be responsible for diskettes or text mislaid or destroyed in transit or during editing. Illustrations should be planned for reduction to the dimension of the printed page (14 X 20.5 cm, column width 6.7 cm) and allow room for legends at the top and bottom. Do not make plates larger than 14 x 18 in. (35.5 X 46 cm). Individual figures should be mounted on a suitable drawing board or similar heavy stock. Photographs should be trimmed, grouped together and abutted when mounted. Figure numbers should be on the plate, and it is strongly recommended that names be included after the numbers (e.g.. Fig. 2, tcxaiius). Include title, author(s) and address(es), and illustration numbers on back of each plate. Original figures need not be sent until requested by the editor, usually after the manuscript has been accepted. Reference to figures/tables in the text should be in the style "(Fig. 1)" "(Table 1)". Measurements should be in the metric system. All papers must conform to the International Code of Zoological Nomenclature. The first mention of a plant or animal should include the full scientific name including the authority'. Genus names should not be abbre- viated at the beginning of a sentence. In taxonomic papers type specimens must be clearly designated, type depositories must be clearly indicated, and new taxa must be clearly differentiated from existing taxa by means of keys or differential diagnoses. Authors are required to deposit all type material in intemahonally recognized institutions (not private collections). Voucher specimens should be designated for specimens used in behavioral or autecological studies, and they should be deposited similarly. Acceptance of taxonomic papers will not require use of cladistic methods; however, authors using them will be expected to specify the phylogenetic program used (if any), including discussion of program options used. A data matrix should be provided if the subject is complex. Cladograms must be hung with characters and these should include descriptors (not numbers alone) when feasible. The number of parsimonious cladograms generated should be stated and reasons given for the one adopted. Lengths and consistency indices should be provided. Adequate discussions should be given for characters, plesiomorphic conditions, and distributions of characters among outgroups when problematical. References in the text should be (Smith 1999), without a comma, or Smith (1999). Two articles by a single author should be (Smith 1999a, 1999b) or Smith (1999a, 1999b). For multiple authors, use the word "and," not the symbol "&" (Smith and Jones 1999). For papers in press, use "in press," not the expected publication date. The Literature Cited section should include all papers referred to in the paper. Journal names should be spelled out completely and in italics. Charges. Publication charges are $10.00 per printed page. At least one author of the paper must be a member of the International Society of Hymenopterists. Reprints are charged to the author and must be ordered when returning the proofs; there are no free reprints. Author's corrections cind changes in proof are also charged to the author Color plates will be billed at full cost to the author. All manuscripts and correspondence should be sent to: Dr E. Eric Grissell Systematic Entomology Laboratory, USDA % National Museum of Natural History Washington, DC 20560-0168 Phone: (202) 382-1781 Fax: (202) 786-9422 E-mail: egrissel@sel.barc.usda.gov Heckman B I N I) 1- K ^, I N t , [Viurul T