c 3 Journal of Hymenoptera Research . .IV 6«** Volume 15, Number 2 October 2006 ISSN #1070-9428 CONTENTS BELOKOBYLSKIJ, S. A. and K. MAETO. A new species of the genus Parachremylus Granger (Hymenoptera: Braconidae), a parasitoid of Conopomorpha lychee pests (Lepidoptera: Gracillariidae) in Thailand 181 GIBSON, G. A. P., M. W. GATES, and G. D. BUNTIN. Parasitoids (Hymenoptera: Chalcidoidea) of the cabbage seedpod weevil (Coleoptera: Curculionidae) in Georgia, USA 187 GILES, V. and J. S. ASCHER. A survey of the bees of the Black Rock Forest Preserve, New York (Hymenoptera: Apoidea) 208 GUMOVSKY, A. V. The biology and morphology of Entedon sylvestris (Hymenoptera: Eulophidae), a larval endoparasitoid of Ceutorhynchus sisymbrii (Coleoptera: Curculionidae) 232 KULA, R. R., G. ZOLNEROWICH, and C. J. FERGUSON. Phylogenetic analysis of Chaenusa sensu lato (Hymenoptera: Braconidae) using mitochondrial NADH 1 dehydrogenase gene sequences 251 QUINTERO A., D. and R. A. CAMBRA T The genus Allotilla Schuster (Hymenoptera: Mutilli- dae): phylogenetic analysis of its relationships, first description of the female and new distribution records 270 RIZZO, M. C. and B. MASSA. Parasitism and sex ratio of the bedeguar gall wasp Diplolqjis rosae (L.) (Hymenoptera: Cynipidae) in Sicily (Italy) 277 VILHELMSEN, L. and L. KROGMANN. Skeletal anatomy of the mesosoma of Palaeomymar anomalum (Blood & Kryger, 1922) (Hymenoptera: Mymarommatidae) 290 WHARTON, R. A. The species of Stenmulopius Fischer (Hymenoptera: Braconidae, Opiinae) and the braconid sternaulus 316 (Continued on back cover) INTERNATIONAL SOCIETY OF HYMENOPTERISTS Organized 1982; Incorporated 1991 OFFICERS FOR 2006 Michael E. Schauff, President James Woolley, President-Elect Michael W. Gates, Secretary Justin O. Schmidt, Treasurer Gavin R. Broad, 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, Plant Sciences institute, Bldg. 003, Rm. 231 BARC-West, Beltsville, MD 20705, USA; Secretary, Southwestern Biological Institute, 1961 W. Brichta Dr., Tucson, AZ 85745, USA; Treasurer, PO Box 37012, c/o Smithsonian Institution, MNMH, MRC168, Washington, DC 20013-7012, USA; Editor, Centre for Ecology & Hydrology, Monks Wood, Abbots Ripton, Hunting- don, Peterborough PE28 2LS, UK. Membership. Members shall be persons who have demonstrated interest in the science of entomol- ogy. Annual dues for members are US$45.00 per year (US$40.00 if paid before 1 February), payable to The International Society of Hymenopterists. Requests for membership should be sent to the Trea- surer (address above). Information on membership and other details of the Society may be found on the World Wide Web at http://hymenoptera.tamn.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, 0 Department of Entomology, Smithsonian Institution, 10th and Constitution NW, Washington, D.C. 20560-0168, U.S.A. Editor: Gavin R. Broad, Centre for Ecology & Hydrology, Monks Wood, Abbots Ripton, Hunting- don, Peterborough PE28 2LS, UK. Managing Editor and Known Bondholders or other Security Holders: none. This issue was mailed 20 October 2006 J. HYM. RES. Vol. 15(2), 2006, pp. 181-186 A New Species of the Genus Parachremylns Granger (Hymenoptera: Braconidae), a Parasitoid of Conopomorpha Lychee Pests (Lepidoptera: Gracillariidae) in Thailand Sergey A. Belokobylskij and Kaoru Maeto (SAB) Zoological Institute Russian Academy of Sciences, St. Petersburg, 199034, Russia, and Museum and Institute of Zoology Polish Academy of Sciences, Wilcza 64, Warszawa 00-679, Poland; email: sb@zin.ru (KM) Laboratory of Insect Science, Faculty of Agriculture, Kobe University, Rokkodai-machi 1-1, Nada-ku, Kobe 657-8501, Japan; email: maeto@kobe-u.ac.jp Abstract. — Parachremylns litchii Belokobylskij & Maeto, new species, from Thailand is described as a parasitoid of larvae of Conopomorpha sinensis Bradley and C. litchiella Bradley, the major pests of lychee and longan in South-East Asia. The taxonomic position of Parachremylns and the range of the hosts of related genera of parasitoids are discussed. Several insect pests are seriously threat- The genus Parachremylns with type spe- ening lychee (Litchi chinensis Sonn.) and cies P. seyrigi Granger was originally de- longan (Dimocarpus longan Lour.) (Sapin- scribed from Madagascar (Granger 1949); daceae) growers. They are the fruit borer this genus occurrs also in continental {Conopomorpha sinensis Bradley), leaf miner Africa — Nigeria and Niger (Wharton (Conopomorpha litchiella Bradley), longan 1993). Two additional species of this genus sucking bug (Tessaratoma papillosa Drury), have already been recorded from the fruit piercing moth (Othreis fnllonia Oriental region. Parachremylns oblongus (Clerck)), and twig borer (Zenzera coffeae (Papp) was described from India in the Nietner) (Menzel 2002). genus Avga Nixon (Papp 1990, 1997), and Conopomorpma sinensis, the lychee stem- P. temporalis Belokobylskij from Brunei end borer and the lychee fruit borer in (Belokobylskij 1999). A fourth species of China, Thailand and India, is the major this genus, similar to P. temporalis, is pest of lychee and longan in these coun- described below from Thailand. The sys- tries. Conopomorpha sinensis and the related tematic position of this genus is disputable. C. litchiella both attack lychee and longan, Parachremylns is included in the subfamily the latter preferring to mine leaves and Exothecinae (tribe Avgini: Belokobylskij shoots (Bradley 1986). There have been 1993), or conventionally in subfamily Hor- only tentative reports on braconid para- miinae (Wharton 1993). In spite of the sitoids of the pest Conopomorpiha borers: different understanding of the contents of Phanerotoma sp., Pholetesor (Apanteles) sp., subfamilies, the position of this genus close and Colastes sp. (Menzel 2002, Anupunt to Avga Nixon is suggested by both and Sukhvibul 2005), but possibly infor- authors. Belokobylskij (1993) discussed mation about Colastes is due to misdeter- the relationships of these genera with mination. Here we report a new braconid Parahormins Nixon, Pseudohormius Tobias of the genus Parachremylns Granger as & Alexeev and Allobracon Gahan (= Lcur- a larval parasitoid of C. sinensis and C. inion Muesebeck), which share the loss of litchiella. the prepectal (epicnemial) carina on the 182 Journal of Hymenoptera Research mesosoma. Wharton (1993) provisionally placed Avga near Parahormius and Pseudo- hormius and showed the possible relation- ship of Avga and Parachremylus (shared granulate mesonotal sculpture and the poorly developed propleural flange). How- ever, in his opinion, Allobracon does not appear to be closely related to Parachremy- lus in spite of it sharing a number of features with Avga and Parahormius. The host of Parachremylus has not been known till now. The new species described below as P. litchii sp. nov. was reared from larvae of Conopomorpha sinensis and C. litchiella (Gracillariidae), both important pests of lychee and longan trees in South- East Asia. The members of related genera of the tribe Avgini {Parahormius, Avga, Allobracon) are also recorded as parasitoids of the leaf-rollers or leaf-miners of the families Tortricidae, Gracillariidae, Lyone- tiidae, Cosmopterigidae, Coleophoridae, and Gelechiidae (Belokobylskij 1993, Whar- ton 1993) as well as rarely (recorded for Allobracon) of leaf-mining Coleoptera (Wharton 1993). The terms of wing venation are used as defined by Belokobylskij and Tobias (1998). The following abbreviations are used: POL — postocellar line; OOL — ocular-ocellar line; Od — maximum diameter of lateral ocellus; NIAES— National Institute of Agro-Environmental Sciences (Tsukuba, Japan); ZISP— Zoological Institute, Rus- sian Academy of Sciences (St. Petersburg, Russia). Parachremylus litchii Belokobylskij & Maeto, new species (Figs 1-11) Holotype female.— "Horticultural Re- search Center, Chiang Rai, Thailand, viii, 1996, Supatra Dolsopon", "Host: Conopo- morpha litchiella larvae on Lychee or Lon- gan" (NIAES). Paratypes. 2 females, 1 male, with the same labels as holotype (NIAES, ZISP); 5 females, "Horticultural Research Center, Chiang Rai, Thailand, 6. vi, 1997, Supatra Dolsopon", "Host: Con- opomorpha sinensis larvae" (NIAES, ZISP). Description. — Female. Body length 2.6- 2.8 mm; fore wing length 2.5-2.6 mm. Antennae: thickened, almost filiform, 29- 30-segmented, 1.1-1.2 times longer than body. Scapus 1.7-2.0 times longer than wide. First flagellar segment 2.5-2.8 times longer than its apical width, 1.1-1.2 times longer than second segment. Penultimate segment 2.0-2.3 times longer than wide, 0.6-0.7 times as long as first flagellar segment, 0.7-0.75 times as long as apical segment; the latter with distinct spine apically. Head: width 1.8-2.0 times its median length, 1.25-1.4 times width of mesoscutum. Temple very strongly and almost linearly narrowed behind eye (dor- sal view). Transverse diameter of eye (dorsal view) 5.5-7.0 times longer than temple length (7.0-7.7 times if measured on straight line). Ocelli small, in triangle with base 1.1-1.15 times its sides. POL 0.7-1.0 times Od, 0.3-0.5 times OOL. Vertex with narrow median longitudinal furrow. Oc- cipital carina dorsally distinctly curved towards ocelli, rather widely interrupted medially; not fused with hypostomal cari- na ventrally being obliterated for a short distance. Eye large, sub-round, glabrous, 1.1-1.2 times as high as broad. Malar space 0.25-0.3 times height of eye, 0.8-0.9 times basal width of mandible. Face width 0.9 times height of eye and 1.2-1.25 times height of face and clypeus combined. Malar suture absent. Clypeal suture rather distinct and complete. Clypeus weakly convex. Hypoclypeal depression sub- round, its width 0.8-0.9 times distance from edge of depression to eye, 0.35 times width of face. Head below eyes (front view) strongly and almost linearly nar- rowed. Mesosoma: length 1.5-1.55 times its height. Mesoscutum highly and almost perpendicularly raised above pronotum (lateral view), with rather fine longitudinal medioposterior keel (dorsal view). Notauli rather narrow, shallow anteriorly on verti- cal surface and very shallow to almost Volume 15, Number 2, 2006 183 Figs 1-11. Parachremylus litchii sp. nov. 1, Head, frontal view. 2, Head, dorsal view. 3, Propodeum. 4, Six basal segments of antenna. 5, Mesosoma, lateral view. 6, Hind tibia and two basal segments of hind tarsus. 7, Metasoma, dorsal view. 8, Hind coxa. 9, Hind femur. 10, Fore wing. 11, Hind wing. 184 Journal of Hymenoptera Research absent on dorsal surface, finely sculptured. Prescutellar depression short, shallow, finely crenulate-granulate, 0.15-0.2 times as long as scutellum. Scutellum almost flat. Metanotum medially with small and ob- tuse tubercle. Subalar depression rather shallow, wide, densely and curvedly stri- ate with fine granulation anteriorly. Ster- nauli shallow, rather wide, weakly curved, entirely smooth. Wings: Length of fore wing 2.8-3.0 times its maximum width. Radial cell not shortened, metacarpus 1.3 times longer than pterostigma. Pterostigma rather wide, 3.1-3.7 times longer than wide. Radial vein arising a little or rather distinctly before middle of pterostigma. Second radial abscissa 1.5-2.3 times longer than first abscissa, 0.25-0.3 times as long as the straight third abscissa, 1.15-1.25 times longer than the weakly curved first radio- medial vein. Second radiomedial cell short, weakly narrowed towards apex, its length 1.5-1.8 times maximum width, 0.9-1.1 times length of brachial cell. First medial abscissa rather distinctly S-shaped. Recur- rent vein 0.9-1.0 times as long as second abscissa of medial vein. Discoidal cell 1.55- 1.65 times longer than wide. Nervulus strongly postfurcal, distance from nervulus to basal vein nearly twice nervulus length. Parallel vein arising a little behind middle of distal margin of brachial cell. Hind wing 4.5-4.7 times longer than maximum width. First abscissa of costal vein 0.85-0.9 times as long as second abscissa. First abscissa of mediocubital vein 1.15-1.2 times longer than second abscissa. Recurrent vein short, unsclerotized, interstitial, curved toward base of wing. Legs: Hind coxa large, 1.5-1.6 times longer than wide, 0.7-0.75 times as long as hind femur. Hind femur wide, 3.1- 3.2 times longer than wide. Hind tibia thickened towards apex. Hind tarsus 1.1 times longer than hind tibia; hind basitar- sus 0.6-0.65 times combined length of second-fifth segments (without pretarsus). Second tarsal segment 0.4-0.45 times as long as basitarsus, 1,2-1.3 times longer than fifth segment (without pretarsus). First-fourth segments of hind tarsus ven- trally with wide and transparent flanges, which are pointed on the tops of each segment. Metnsomn: 1.7-2.0 times longer than its maximum width, 0.9-1.1 times as long as head and mesosoma combined. First tergite strongly, uniformly and line- arly widened from base to apex; with small spiracular tubercles before its middle; laterally with distinct high and rather wide carinae; fine dorsal carinae fused in basal 0.3 and then extending to apex as a single, elevated, median carina; dorsope absent. Apical width of first tergite 2.4-2.7 times its basal width; its length 0.6-0.65 times apical width. Second suture rather distinct and convex. Second and third tergites with rather distinct and fine longitudinal medi- an carina. Median length of second tergite about half its basal width, equal to or 1.1- 1.2 times length of third tergite. Combined median length of second and third tergites nearly equal to basal width of second tergite, 0.7-0.75 times maximum width of tergites. Ovipositor sheath (visible part in lateral view) 1.1-1.3 times longer than first tergite, 1.0-1.2 times longer than hind basitarsus, 0.25-0.4 times as long as meso- soma, 0.15-0.17 times as long as fore wing. Sculpture mid pubescence: Head very densely and minutely granulate, face additionally with rather fine and irregular striation. Mesoscutum very densely and distinctly granulate, with rather narrow and long rugulosity in medioposterior half. Scutel- lum finely and densely granulate. Meso- pleuron almost smooth in lower half. Metapleuron coarsely, regularly and curv- edly striate for the most part, with fine granulation between striae and anteriorly. Propodeum almost entirely coarsely and rather sparsely striate, striae in areola more or less transverse and partly undulate or rugulose, with fine granulation partly; with distinctly delineated basolateral areas; areola wide, its length 1.0-1.2 times maximum width; dorsal carina 0.8-1.0 times as long as areola fork. Hind coxa smooth; hind femur finely punctulate with Volume 15, Number 2, 2006 185 very fine granulation dorsally, smooth ventrally. Metasoma entirely densely granulate, granulation becoming finer to- wards apex of metasoma. Mesoscutum entirely shortly and very densely setose. Hind tibia dorsally with rather short, dense and semi-erect setae, its length 0.35-0.55 times maximum width of tibia. Colour: Head and anterior half of meso- soma (including mesoscutum) yellow, posterior part of mesosoma and metasoma pale yellow, metasoma additionally often with greenish tint. Antenna reddish brown or brown, scapus mostly yellow. Palpi pale yellow. Legs yellow, all tarsi (especially posterior ones) more or less brown. Ovipositor sheath brown in basal half and black in apical half. Fore wing faintly infuscate. Pterostigma brownish yellow. Male. — Body length 2.0 mm; fore wing length 2.4 mm. Head width 2.1 times its median length. Transverse diameter of eye (dorsal view) 8.8 times longer than temple length if measured on straight line. Anten- na 28-segmented. Otherwise similar to female. Diagnosis. — The new species is very similar to P. temporalis Belokobylskij from Brunei (Belokobylskij 1999) and differs in having the recurrent vein as long as second abscissa of medial vein, the nervulus strongly postfurcal, the pterostigma rather wide, the hind femur wide, the second tergite short, the face rather finely striate, and the propodeum almost entirely coarse- ly rugose-striate. Host. — Conopomorpha sinensis Bradley and C. litchiella Bradley (Gracillariidae). Distribution. — Thailand. Etymology. — This species is named after the name of the fruit tree — lychee {Litchi chinensis Sonn.) — on which their hosts develop. KEY TO SPECIES OF THE GENUS PARACHREMYLUS GRANGER 1. Temple longer; transverse diameter of eye (dorsal view) 4.0-5.0 times as long as temple length. Malar space larger than basal width of mandible. Mesopleuron smooth in upper half, striation partly present in subalar depression only. 1st— 4th segments of hind tarsus with narrow and partly indistinct flanges 2 - Temple shorter; transverse diameter of eye (dorsal view) 5.5-7.0 times as long as temple length. Malar space less than basal width of mandible. Mesopleuron distinctly curvedly striate in upper 0.4-0.5. lst-4* segments of hind tarsus with wide flanges 3 2. Notauli complete, rather deep posteriorly. Metacarpus 1.2-1.3 times as long as pterostigma. T1 flagellar segment 3.5-3.7 times as long as apical width. Median length of 2nd and 3rd metasomal tergites combined a little larger than basal width of 2nd tergite. Propodeum mostly coarsely and sparsely striate. Body length 2.2 mm. — Madagascar P. seyrigi Granger - Notauli incomplete, almost absent posteriorly. Metacarpus 1.5 times as long as pterostigma. 1st flagellar segment 3.0 times as long as apical width. Median length of 2nd and 3rd metasomal tergites combined 1.3 times basal width of 2nd tergite. Propodeum mostlv smooth. Body length 2.0 mm. — India P. oblougus (Papp) 3. Pterostigma narrow, 5.0 times as long as maximum width. Recurrent vein of fore wing about twice as long as second abscissa of medial vein. Nervulus not strongly postfurcal, distance from nervulus to basal vein 0.7 times nervulus length. Hind femur 3.5 times as long as wide. 2nd tergite 0.6 times as long as its basal width. Face almost entirely distinctly transversely striate. Propodeum within background areolation sparsely striate, mostly smooth. Body length 2.3 mm. — Brunei P. temporalis Belokobylskij Journal of Hymenoptera Research Pterostigma rather wide, 3.1-3.7 times as long as maximum width. Recurrent vein of fore wing almost as long as second abscissa of medial vein. Nervulus strongly postfurcal, distance from nervulus to basal vein nearly twice nervulus length. Hind femur 3.1-3.2 times as long as wide. 2nd tergite about 0.5 times as long as its basal width. Face finely and partly indistinctly transversely striate and with dense fine granulation. Propodeum within background areolation almost entirely coarsely rugose-striate with fine granula- tion partly. Body length 2.0-2.8 mm. — Thailand P. litchii sp. nov. ACKNOWLEDGMENTS We wish to thank Dr. S. Moriya (Tsukuba, Japan) and Dr. S. Dolsopon (Bangkok, Thailand) for giving us a chance to examine the reared material, and two reviewers for valuable comments of our manuscript. The present work was supported by the grant of Japan Society for the Promotion of Science, Invitation Fellowship Program for Research in Japan (No. L05564) and grant 2 P04C 001 28 of the Ministry of Science and Information Society Technologies (Poland) for the first author, and a Grant-in-Aid for Scientific Research (A) (No. 15208007) from the Japan Society for the Promotion of Science for the second author. LITERATURE CITED Anupunt, P. and N. Sukhvibul. 2005. Lychee and longan production in Thailand. Acta Horticultural' (ISHS) 665: 53-60. Beiokobylskij, S. A. 1993. [On the classification and phylogeny of the braconid wasps of subfamilies Doryctinae and Exothecinae (Hymenoptera, Bra- conidae). Part I. On the classification, 2.] Entomo- logicheskoe Obozrenie 72: 143-164. [In Russian.] Beiokobylskij, S. A. 1999. [New taxa of the braconid subfamily Exothecinae (Hymenoptera, Braconi- dae) from tropical and subtropical regions of the Old World. I.] Entomologicheskoe Obozrenie 78: 674-693. [In Russian.] Beiokobylskij, S. A. and V. I. Tobias. 1998. Fam. Braconidae. Introduction. Pp. 8-26. in: Lehr, P. A. ed. OpredeliteV nasekomykh Dal'nego Vostoka Rossii [Keys to the Insects of the Russian Far East]. 4(3). Dal'nauka, Vladivostok. [In Russian.] Bradley, J. D. 1986. Identity of the South-East Asian cocoa moth, Conopomorpha cramerella (Snellen) (Lepidoptera: Gracillariidae), with descriptions of three allied new species. Bulletin of Entomolog- ical Research 76: 41-51. Granger, C. 1949. Braconides de Madagascar. Mem- oires de llnstitut Scientifique de Madagascar. Ser. A, Biologic Animate 2: 1-428. Menzel, C. 2002. The lychee crop in Asia and the Pacific. Food and Agriculture Organization of the United Nations, Bangkok. 120 pp. Papp, J. 1990. New braconid wasps (Hymenoptera, Braconidae) in the Hungarian Natural History Museum, 1. Annates historico-naturales Musei nationalis hungarici 82: 175-190. Papp, J. 1997. New braconid wasps (Hymenoptera, Braconidae) in the Hungarian Natural History Museum, 5. Annates historico-naturales Musei nationalis lunigarici 89: 157-175. Shenefelt, R. D. 1975. Braconidae 8. Exothecinae, Rogadinae. Hymoiopterorum Catalogus (nova editio) 12: 1115-1262. Wharton, R. A. 1993. Review of the Hormiini (Hymenoptera: Braconidae) with description of new taxa. Journal of Natural History 27: 107-171. J. HYM. RES. Vol. 15(2), 2006, pp. 187-207 Parasitoids (Hymenoptera: Chalcidoidea) of the Cabbage Seedpod Weevil (Coleoptera: Curculionidae) in Georgia, USA Gary A. P. Gibson, Michael W. Gates and G. David Buntin (GAP) Agriculture and Agri-Food Canada, Biodiversity and Integrated Pest Management, K. W. Neatby Bldg., 960 Carling Avenue, Ottawa, ON, Canada, K1A 0C6; email: gibsong@agr.gc.ca (MWG) Systematic Entomology Laboratory, PSI, Agricultural Research Service, U.S. Department of Agriculture, c/o National Museum of Natural History, Smithsonian Institution, Washington, DC 20560-0168, USA (GDB) Department of Entomology, University of Georgia, Georgia Station, Griffin, GA 30223, USA Abstract. — Five families and 13 species of Chalcidoidea (Hymenoptera) were obtained from mass-reared seedpods of Brassica napus L. (Brassicaceae) as putative parasitoids of the cabbage seedpod weevil, Ceutorhynchus obstrictus (Marsham) (Coleoptera: Curculionidae), in Georgia, USA. The species are Conura torvina (Cresson) (Chalcididae), Euderus glaucus Yoshimoto and Necremnus tidius (Walker) (Eulophidae), Brasema allynii (French) n. comb, (from Eupelmus Dalman) and Eiipclimis cyaniceps Ashmead (Eupelmidae), Eurytoma tylodermatis Ashmead (Eurytomidae), and Lyrcus incertus (Ashmead), L. maculatus (Gahan), L. perdubius (Girault), Mesopolobus moryoides Gibson, Neocatolaccus tylodermae (Ashmead), Pteromalus cerealellac (Ashmead) and Pteromalus sp. (Pteromalidae). An illustrated key is provided to differentiate the taxa. Lyrcus maculatus constituted about 96% of all reared Pteromalidae and 86% of the total parasitoid fauna. The associations of B. allynii, E. glaucus, E. cyaniceps, E. tylodermatis, L. incertus, N. tylodermae, Pteromalus sp. and P. cerealellac with C. obstrictus are new, but some of these species likely are hyperparasitoids or emerged from insect contaminants of the mass-reared seedpods. The only previous report of a parasitoid of C. obstrictus in eastern North America, Trichomalus perfectus (Walker) (Pteromalidae), is a misidentification. The parasitoid fauna of C. obstrictus in Georgia is discussed relative to that known for western North America. The cabbage seedpod weevil, Ceuto- rhynchus obstrictus (Marsham) (Coleoptera: Curculionidae), was introduced from Eur- ope to western North America about 70 years ago. Since then it has become the most important insect pest of canola and rape, Brassica napus L. and B. rapa L. (Brassicaceae), in most areas of the conti- nent where these crops are grown (Car- camo et al. 2001, Kuhlmann et al. 2002). It was first reported from eastern North America in North Carolina, USA (USDA 1960), and is now known to extend from Georgia to Quebec and Ontario, Canada (Brodeur et al. 2001, Mason et al. 2004). There have been several surveys of the introduced and native chalcid (Hymenop- tera: Chalcidoidea) parasitoids of the cab- bage seedpod weevil in western North America, including Breakey et al. (1944), Doucette (1944, 1948), Hanson et al. (1948), Carlson et al. (1951), McLeod (1953), Walz (1957), and Dosdall et al. (in press). Murchie and Williams (1998) listed 7 identified and 4 unidentified species in 9 genera and 5 families of Chalcidoidea as parasitoids of C. obstrictus in North Amer- ica, but almost all of the species names either represent misidentifications or are now recognized as junior synonyms of older names (Gibson et al. 2005). Dosdall et al. (in press) reported another six chalcid species as reared from B. napus and B. rapa seedpods in Alberta. Consequently, the 188 Journal of Hymenoptera Research Table 1. Chalcid parasitoids associated with the cabbage seedpod weevil in North America, including for Georgia the number of specimens and percentage (in parenthesis) of total parasitoids reared by Buntin (1998). Western North Taxon America Georgia Chalcididae Conura albifrons (Walsh) + - ? Conura side (Walker)1 + - Conura torvina (Cresson) + 9 (0.8) Eulophidae Eudcrus albitarsis + - (Zetterstedt) Euderus glaucus — 2 (0.2) Yoshimoto Necremnus tidius (Walker) + 6 (0.5) Eupelmidae Brasema allynii (French) - 5 (0.5) Eupelmus cyaniceps — 4 (0.4) Ashmead Eupelmus vesicularis + - (Retzius) Eurytomidae Eun/toina ti/lodennatis + 25 (2.2) Ashmead Pteromalidae Chlorocytus sp. + - Lyrcus incertus (Ashmead) - 6 (0.5) Lyrcus maculatus (Gahan) + 967 (86.0) Lyrcus perdubius (Girault) + 60 (5.3) Mesopolobus bruchophagi + - (Gahan) Mesopolobus mayetiolae + - (Gahan) Mesopolobus moryoides + 2 (0.2) Gibson Neocatolaccus tylodermae - 33 (2.9) (Ashmead) Pteromalus eerealellae - 1 (0.1) (Ashmead) Pteromalus spp.2 + 4 (0.4) Trichomalus lucidus + - (Walker) ' Single record, likely a misidentification of C. torvina (see text). 2 Females in the two regions represent different species (see text). chalcid fauna purportedly parasitizing C. obstrictus in western North America in- cludes at least 14 species (Table 1). In contrast, there is only a single published report of parasitoids of C. obstrictus in eastern North America. Buntin (1998) stated that greater than 96% of the para- sitoids recovered from seedpods of B. napus in Georgia were Trichomalus perfectus (Walker) (Pteromalidae). This species is the most common biological control agent of C. obstrictus in Europe (Murchie and Williams 1998) and was long thought to have been introduced to North America along with the seedpod weevil. However, Gibson et al. (2005) snowed that all previous reports of T. perfectus in western North America were misidentifications of Trichomalus lucidus (Walker), another European species. Accurate identification of parasitoid spe- cies is a prerequisite for successful classical biological control and integrated pest man- agement. The senior author examined the parasitoids reared by Buntin (1998) as part of a larger study to document the diversity and identity of the chalcid parasitoids of C. obstrictus in North America. The primary purpose of Buntin (1998) had been to examine the effect of trap cropping on the number of seedpod weevils and its para- sitoids in canola crops in Georgia. The species identities of the parasitoids had therefore never been thoroughly investi- gated. The purpose of this paper is to provide the first comprehensive informa- tion on the diversity of the chalcid para- sitoids reared from canola seedpods in southeastern USA in order to facilitate future studies of the parasitoid fauna associated with C. obstrictus throughout North America. MATERIALS AND METHODS The chalcid parasitoids identified in this study were obtained from mass-reared seedpods of B. napus collected from the Bledsoe Research Farm (33 10.635'N 84°24.354'W) located near Griffin, Georgia, from 1994-1996, as per "Material and methods" in Buntin (1998). Although not stated, the pods were screened for insect contaminants prior to rearing. Contami- nants mainly included aphids (Hemiptera: Volume 15, Number 2, 2006 189 Aphididae) and larvae and pupae of the using Adobe Photoshop™ to enhance cla- diamondback moth, Plutella xylostella (L.) rity. (Lepidoptera: Plutellidae). The reared parasitoids had been stored in ethanol, RESULTS but were critical-point dried, point- A total of 1,127 specimens of Chalcidoi- mounted, and identified to genus by the dea were sufficiently intact that they could senior author using the relevant family be identified accurately. Of these, there keys in Gibson et al. (1997). The senior was a single male Pachyneuron aphidis author is responsible for all species identi- (Bouche) and a female and male Asnphes fications except Eurytoma tylodennatis Ash- suspensus (Nees) (Pteromalidae). Members mead (Eurytomidae), which was identified of Pachyneuron and Asaphes are obligate by MWG. Information concerning the hyperparasitoids of aphids (Gibson et al. method of species identification within 1997) and are not dealt with further. The each genus is provided under the relevant remaining 1,124 specimens included 5 species discussion. Voucher specimens are families, 10 genera, and 13 species of deposited in the Canadian National Col- Chalcidoidea that are possible parasitoids lection of Insects and Arachnids (CNC), of C. obstrictus. These taxa are keyed below Ottawa, Ontario, the University of Georgia and subsequently treated by family in Museum of Natural History (UGCA), alphabetical order. The key also segregates Athens, Georgia, and the United States Trichoinalus Thomson and Chlorocytus Gra- National Museum of Natural History ham, the only two genera reared from C. (USNM), Washington, District of Colum- obstrictus in western North America (Dos- bia. Terms used for parasitoid structure dall et al. in press) that were not recovered follow Gibson (1997). Photographs are by Buntin (1998) (Table 1). The two taxa composite serial images that were com- are included in the key because species of bined using Auto-Montage™. These both genera occur in eastern North Amer- images and the scanning electron micro- ica and may eventually be reared as part of photographs were digitally retouched the regional C. obstrictus parasitoid fauna. KEY TO CHALCIDOIDEA PUTATIVELY PARASITIZING C. OBSTRICTUS IN GEORGIA 1 Hind leg with elongate coxa of similar length to conspicuously swollen and ventrally toothed femur, and with curved tibia (Fig. 1) . . . Conura torvina (Cresson) (Chalcididae) Hind leg with comparatively short coxa, slender femur, and straight tibia (Figs 3-8) ... 2 2(1) Tarsi 4-segmented; flagellum with 3 or 4 funicular segments, the segments sometimes branched (Figs 9, 10) (Eulophidae) 3 Tarsi 5-segmented; flagellum with 5 or 6 unbranched funicular segments (Figs 11-18) ... 4 3(2) Meso- and metatarsi with basal 4 segments white; forewing membrane bare dorsally behind marginal vein, but with clearly visible row of long admarginal setae (Fig. 42, ams) on ventral surface near marginal vein; flagellum of both sexes with 4 unmodified funicular segments Euderus gJaucus Yoshimoto Meso- and metatarsi with basal 1 or 2 segments white; forewing membrane uniformly setose dorsally behind marginal vein, the setae largely obscuring admarginal setae on ventral surface (Fig. 41); flagellum branched in male (Fig. 10) and with only 3 funicular segments in female (Fig. 9) Necrcmmis tidius (Walker) 4(2) Head and mesosoma with coarse piliferous punctures and non-metallic, dark brown to black (Figs 3, 4); pronotal collar quadrangular in dorsal view, only slightly shorter than mesoscutum; male with elongate petiole and flagellar segments having whorls of conspicuously long setae (Fig. 4) 190 Journal of Hymenoptera Research Eiin/toma tylodermatis Ashmead (Eurytomidae) Head and mesosoma with finer mesh-like sculpture and often with metallic green to bluish luster; pronotum strongly transverse in dorsal view, conspicuously shorter than mesoscutum; male with short petiole and flagellar segments having short, inconspicuous setae 5 5(4) Mesopleuron elongate, convex or cushion-like, and uniformly finely sculptured (Figs 5, 7); middle leg with strong black spines at apex of tibia and on ventral surface of tarsal segments (Figs 5, 7, sp), the colour of spines contrasting distinctly with mostly yellowish leg (Eupelmidae: Eupelminae female) 6 Mesopleuron about as high as long, usually with a smooth region dorsally below base of wings and often with an oblique femoral depression or groove, but at least not convex or uniformly sculptured (Figs 23-25); middle leg with slender spines at apex of tibia and on ventral surface of tarsal segments, the colour of spines not contrasting with leg 7 6(5) Ovipositor sheaths projecting only slightly beyond apex of gaster and uniformly coloured (Fig. 5); forewing completely setose behind parastigma and marginal vein (Fig. 5) Brasema allynii (French) Ovipositor sheaths projecting beyond apex of gaster by about one-third its length and medially whitish between darker basal and apical bands (Fig. 7); forewing with slender, oblique bare band (Fig. 7, bb) below parastigma and base of marginal vein Eupelmus cyaniceps Ashmead 7(5) Flagellum with only basal segment conspicuously differentiated as strongly transverse ring segment and with 7 or 8 distinct funicular segments; head in frontal view with inner margin of eyes distinctly divergent over about ventral half; mesotibial spur much longer and thicker than metatibial spurs, as long as basal tarsal segment and about one-third length of tarsus (Eupelmidae: Eupelminae male) 8 Flagellum with 2 or 3 basal segments conspicuously differentiated as ring segments and then with 6 or 5 tubular funicular segments, respectively (Figs 11-18); head in frontal view with inner margin of eyes subparallel or slightly but uniformly incurved; mesotibial spur somewhat longer than, but otherwise similar to, metatibial spurs, the spur obviously shorter than basal tarsal segment and only about one-quarter length of tarsus (Pteromalidae) 9 8(7) Forewing completely setose behind parastigma and base of marginal vein (Fig. 6); hind leg with femur yellowish-white and tibia usually more or less distinctly brown (Fig. 6); flagellum clavate, the segments widening distinctly to clava and apical funicular segments transverse (Fig. 6) Brasema allynii (French) Forewing with large, oblique bare region (Fig. 8, bb) behind parastigma and base of marginal vein; hind leg with femur and tibia dark (Fig. 8); flagellum robust- filiform, the segments all about the same width and apical funicular segments quadrate (Fig. 8) Eupelmus cyaniceps Ashmead 9(7) Flagellum with 5 tubular funicular segments and 3 strongly transverse ring segments (Figs 11, 13, 14) 10 Flagellum with 6 tubular funicular segments and 2 ring segments (Figs 12, 15-18) ... 14 10(9) Female only: head and mesosoma dark with conspicuous, white, lanceolate setae (Figs 19, 20, 24, 25); costal cell ventrally with setae only within about apical half of cell (Figs 43, 46) 11 Female or male: head and mesosoma variable in colour, but with inconspicuous hairlike setae (Figs 21-23); costal cell ventrally with line of setae extending almost entire length of cell or at least setae present both basally and apically if line more or less interrupted medially (Figs 44, 45) 13 11(10) Forewing dorsally setose behind marginal vein over about apical half of vein, the setae partly obscuring at least 3 rows of ventral admarginal setae apically (Fig. 43); propodeum with transverse ridge or carina within anterior half dividing it into anterior and posterior sections on either side of median carina (Fig. 30); Volume 15, Number 2, 2006 191 metapleuron completely sculptured and with anterior margin on same plane as and abutting mesopleuron (Fig. 25, am) Neocatolaccus tylodertnae (Ashmead) Forewing dorsally bare behind marginal vein to level at least equal with middle of stigmal vein, completely exposing 1 or sometimes 2 partial rows of ventral admarginal setae (Fig. 46, ams); propodeum sometimes with transverse furrow near middle, but without transverse ridge (Figs 27, 29); metapleuron partly smooth anteriorly and with anterior margin either curved outward (Figs 24, 29) or extending anteriorly above and over posterior margin of mesopleuron (Figs 27, 28) 12 12(11) Propodeum with nucha (Fig. 29, nuc) delineated laterally by longitudinal carina within furrow along posterior margin; metapleuron with anterior margin (Figs 24, 29, am) curved outwards, extending as thin brown flange almost at right angle to posterior margin of mesopleuron; lower face without evident malar depression, evenly convex along oral margin between malar sulcus and clypeus Lyrcus perdubius (Girault) Propodeum with nucha not delineated laterally by carina, the furrow along posterior margin of callus continued uninterrupted mesally and anteriorly so as to delineate, more or less conspicuously, anterior limit of nucha (Fig. 27); metapleuron with anterior margin (Figs 27, 28, am) raised above and extending over posterior margin of mesopleuron; lower face with short but distinct, concave malar depression (Fig. 20, md) between malar sulcus and clypeus Lyrcus incertus (Ashmead) 13(10) Both sexes: mesonotum usually dark with conspicuous pattern of bluish-green spots, the spots usually most distinct on mesoscutum paramedially behind pronotum and laterally on lateral lobe adjacent to notaulus, though small specimens sometimes brown. Female: gaster lanceolate (Fig. 22); forewing dorsally bare behind marginal vein to level at least equal with middle of stigmal vein (Fig. 45). Male: flagellum brown with first funicular segment oblong and much longer than combined length of the 3 ring segments (Fig. 14); marginal vein strong, but only as thick as width of stigma and with posterior margin straight, parallel with anterior margin Lyrcus maculatus (Gahan) Both sexes: mesonotum metallic green. Female: gaster subcircular (Fig. 21); forewing dorsally bare behind marginal vein, but apically the setae extending to base of stigmal vein (Fig. 44). Male: flagellum yellowish with first funicular segment quadrate to slightly wider than long and at most as long as combined length of the 3 ring segments (Gibson et al. 2005, fig. 8); marginal vein conspicuously thickened relative to slender stigma and with posterior margin slightly convex (Gibson et al. 2005, fig. 31) Mesopiolobus moryoides Gibson 14(9) Male only: forewing with bare band behind marginal vein extending to level about equal with middle of stigmal vein, and with 1 or at most 2 partial rows of admarginal setae (Fig. 46, ams) that are obviously longer than setae on dorsal surface of disc; metapleuron partly smooth and with anterior margin (Figs 24, 28, am) curved outward or raised above mesopleuron 15 Male or female: forewing with bare region behind marginal vein less extensive, the discal setae extending to or almost to base of stigmal vein, and with more than 2 rows of admarginal setae of about same length as setae on dorsal surface of disc (Figs 43, 47-50); metapleuron completely sculptured and with anterior margin (Fig. 25, am) on same plane as and abutting mesopleuron 16 15(14) Propodeum with nucha (Fig. 29, nuc) delineated laterally by longitudinal carina within furrow along posterior margin; flagellum with most funicular segments only slightly longer than wide, the first segment subquadrate and shorter than pedicel (Fig. 15); metapleuron with anterior margin (Figs 24, 29, am) curved outwards, extending as thin brown flange almost at right angle to posterior margin of mesopleuron Lyrcus perdubius (Girault) Journal of Hymenoptera Research Propodeum with nucha not delineated laterally by carina, the furrow along posterior margin of callus extending uninterrupted mesally and anteriorly so as to delineate, more or less conspicuously, anterior limit of nucha (Fig. 27); flagellum with all funicular segments clearly oblong, the first segment as long as pedicel (Fig. 16); metapleuron with anterior margin (Figs 27, 28, am) raised above and extending anteriorly over posterior margin of mesopleuron Lyrcus incertus (Ashmead) 16(14) Male only: head and body dark with conspicuous, long, slightly lanceolate white setae (Fig. 25); propodeum with transverse ridge or carina within anterior half dividing it into anterior and posterior sections on either side of median carina (Fig. 30) Neocatolaccus tylodermae (Ashmead) Male or female: head and body metallic green with inconspicuous hairlike setae (Fig. 26); propodeum with or without median carina but without transverse ridge (Figs 35-38) 17 17(16) Pronotum anteriorly with collar rounded into neck, the reticulations extending uninterrupted from dorsal to inclined surface (cf. Figs 23, 24); forewing with marginal vein comparatively short, less than 1.5 times as long as stigmal vein (Figs 49, 50); propodeum with convex, reticulate nucha (Figs 31, 32, nuc), reticulate panels (Figs 31, 32, pnl), and often without distinct median carina 18 Pronotum anteriorly with shiny, transverse carina differentiating collar from neck (Fig. 26); forewing with marginal vein obviously (at least 1.5 times) longer than stigmal vein (Figs 47, 48); propodeum with flat or slightly concave, lunate or triangular adpetiolar strip (Fig. 35, aps) delineated by inverted Y-shaped median carina anterior to petiolar foramen or, if with reticulate nucha (Fig. 37, nuc), then with panels (Fig. 37, pnl) partly strigose (having oblique, irregular, fine carinae or striae) 19 18(17) Female: costal cell with line of setae on ventral surface interrupted medially (Fig. 49); scutellum anterior to frenum with reticulations distinctly smaller medially than laterally (Fig. 33); propodeum with plical carina (Fig. 31, pc) directed obliquely toward inner margin of spiracle Pteromalus cereahilae (Ashmead) Female: costal cell with entire line of setae ventrally (Fig. 50); scutellum anterior to frenum with almost uniform meshlike reticulations (Fig. 34); propodeum with plical carina (Fig. 32, pc) less strongly angled, directed distinctly mesal of inner margin of spiracle toward outer margin of basal fovea (Fig. 32, bf) ... Pteromalus sp. 19(17) Metacoxa setose dorsally only over about apical half; forewing of female without setae on basal fold (Fig. 48); propodeum with inverted Y-shaped median carina delimiting adpetiolar strip (Fig. 35, aps); propodeum in male without complete plical carina (Fig. 36, pc) and in female largely bare posterior to spiracle, setose only from callus to postspiracular sulcus (Fig. 35, pss) Chlorocytus Graham3 Metacoxa setose dorsally over at least apical two-thirds and often completely setose to base; forewing of female with at least a couple of setae on basal fold (Fig. 47, bf) differentiating basal cell from speculum; propodeum with or without median carina but with convex, reticulate nucha (Fig. 37, nuc); propodeum in male with complete plical carina (Fig. 38, pc) and in female extensively setose posterior to spiracle, from callus to complete, strong plical carina (Fig. 37) Trichomalus Thomson3 Genus not yet reported parasitizing C. obstrictus in eastern North America. Volume 15, Number 2, 2006 193 SPECIES NOTES, ARRANGED BY FAMILY Chalcididae One species of Chalcididae was reared — Conura torvina (Cresson), which comprised nine specimens (7 99, 2 33) or 0.8% of the parasitoid fauna. Delvare (1992) keyed the species-groups of Conura and differentiat- ed C. torvina as one of eight species of the side-group in a key to the "common species" of that group in North America north of Mexico. Prior to Delvare (1992), C. torvina was consistently misidentified as Conura side (Walker). Carlson et al. (1951) reported that a specimen of C. side was reared from C. obstrictus in California. We were unable to locate this specimen to confirm the identification, but it is possible that it is conspecific with the Georgia species because C. torvina is transcontinen- tal in North America (Delvare 1992, Noyes 2002). Because of the confusion in names prior to Delvare (1992), the list of published distribution and host records given for C. side by Noyes (2002) certainly contains many records that actually refer to C. torvina. Based on previous name usage, of those Curculionidae listed as hosts of C. side by Noyes (2002), the record of the cotton boll weevil, Anthonomus grandis (Boheman), probably does refer to C. side, whereas the records of Rhynchaenus palli- cornis (Say) and Hypera spp. likely refer to C. torvina. Because of variability in the colour pattern features given by Delvare (1992), females of C. torvina can be easily misidentified as Conura albifrons (Walsh), another transcontinental species that Dos- dall et al. (in press) reported parasitizing C. obstrictus in Alberta. Females of both species have paramedial yellow marks on the first gastral tergum, but in female C. torvina the distance between the marks is, at most, only about equal to the length of a mark (Fig. 2). In female C. albifrons the separation between the marks is at least similar to the width of a mark, if not conspicuously greater. Males of the two species are more easily differentiated. Males of C. torvina have the interantennal region and lower face yellow, whereas males of C. albifrons have the clypeus dark so that they have a conspicuous, angulate (A-like), yellow band extending dorsally between the antennal scrobes. Eulophidae Two genera and species of Eulophidae were reared — Euderus glaucus Yoshimoto (2 99; 0.2% of the parasitoid fauna) and Necremnus tidius (Walker) (1 9, 5 33; 0.5% of the parasitoid fauna). Although E. glaucus was known from Florida and Texas (Noyes 2002), its association with C. ob- strictus in Georgia represents a new state distribution record and a possible new host record. The only other reported host for E. glaucus is Epiblema obfuscana (Dyar) (Lepi- doptera: Tortricidae) (Yoshimoto 1971). Dosdall et al. (in press) reported a second Euderus species, E. albitarsis (Zetterstedt), as an incidental parasitoid of C. obstrictus in Alberta, but this association was also obtained by mass-rearing seedpods. Eu- derus glaucus and E. albitarsis are differen- tiated in Yoshimoto (1971), though prob- lems remain in species recognition within the genus. Necremnus tidius is a comparatively com- mon parasitoid of C. obstrictus in western North America, but it was misidentified as N. duplicatus Gahan prior to Gibson et al. (2005), who differentiated and illustrated the species. The specimens from Georgia represent the first record of the species in eastern North America. Eupelmidae Two genera and species of Eupelmidae were reared — Eupelmus (Eupelmus) cyani- ceps Ashmead (2 99, 2 33; 0.4% of the parasitoid fauna) and Brasema allynii (French) n. comb, (from Eupelmus Dalman) (1 9, 4 Jo; 0.5% of the parasitoid fauna). Brasema Cameron is unrevised for the region, but there are about 25 described species in North America north of Mexico. 194 Journal of Hymenoptera Research Most of the species are currently misclassi- fied in Eupelmus (Gibson 1995). Gahan (1933) described and partly illustrated both sexes of B. allynii as a parasitoid of the Hessian fly, Mayetiola destructor (Say) (Dip- tera: Cecidomyiidae). Phillips and Poos (1921) also provided both a dorsal and lateral habitus of the female, and for both sexes illustrated the colour pattern of the legs, important species-recognition fea- tures, when they described the immature stages of B. allynii as a parasitoid of the wheat jointworm, Tetramesia tritici (Fitch) (Hymenoptera: Eurytomidae). The sexes of Eupelminae are strongly dimorphic (Gib- son 1995), but the more important di- agnostic features of B. allynii females in- clude: head and mesosoma variably brown or dark with metallic green luster, scrobal depression finely coriaceous and quite shiny, lower face with relatively sparse and only inconspicuously lanceolate white setae, mesonotum finely coriaceous, and middle legs entirely or largely yellow beyond coxae (mesofemur and tibia often with some light brown infusion but meso- femur not extensively dark). Males of B. allynii are in part diagnosed within Brasema by a clavate flagellum with very short and inconspicuous setae (Fig. 6), head and mesosoma metallic green, head with only very slightly lanceolate and comparatively sparse white setae, and legs with all femora yellow (Fig. 6). Brasema allynii is trans- continental in North America and a po- lyphagous primary or facultative hyper- parasitoid of hosts in concealed situations. Noyes (2002) listed 58 host species in 22 families of Coleoptera, Diptera, Hemiptera, Hymenoptera, and Lepidoptera, though the putative host record of C. obstrictus is the first for Curculionidae. Gibson (1995) recognized three subge- nera in Eupelmus, including £. (Episolinde- lia) Girault and E. (Macroneura) Walker in addition to the nominate subgenus. Noyes (2002) listed 45 valid species of Eupelmus in the Nearctic region, but this includes all three subgenera and several species in- correctly classified to genus. Eupelmus is unrevised for the region, but there are about 15 described species of E. (Eupelmus) in North America north of Mexico. Eu- pelmus cyaniceps belongs to the urozonus species-group sensu Gibson (1995). Hunter and Pierce (1912, pi. XVIII, f) provided a dorsal habitus drawing of the female when they recorded the species as a para- sitoid of A. grandis. A species revision of E. (Eupelmus) that includes evaluation of size- correlated and host-induced variation is necessary to confidently characterize spe- cies limits within the subgenus. However, females of £. cyaniceps are differentiated from most other regional species of the subgenus by the following combination of features: macropterous, the forewing hya- line and with a linea calva (Fig. 7, bb), scape dark, mesosoma dark with metallic green luster, and ovipositor sheaths ex- tending for a distance at least equal to two- thirds length of the metatibia and with a medial white band (Fig. 7). Because of extreme sexual dimorphism (cf. Figs 7, 8), species recognition in Eupelmus is based almost entirely on females; males are not characterized for most species. The fea- tures provided in the key for males of E. cyaniceps are of family and generic level. Eupelmus cyaniceps is a primary or faculta- tive hyperparasitoid of hosts in concealed situations. The rearing from C. obstrictus re- presents a new putative host record, but Noyes (2002) listed 17 other species in 11 different genera of Curculionidae as part of 65 host records in 20 families of Coleoptera, Diptera, Hymenoptera and Lepidoptera. The only eupelmid previously reported as a parasitoid of C. obstrictus is Eupelmus (Macroneura) vesicularis (Retzius) from Washington state (Hanson et al. 1948) and British Columbia (McLeod 1953). This species likely represents one of the earliest accidental introductions from Europe to North America; only females are known in North America and they are brachypterous (Gibson 1990). Volume 15, Number 2, 2006 195 Eurytomidae sitoids or as hyperparasitoids. It is beyond One species of Eurytomidae was reared the scoPe of this study to assess the —Eurytoma tylodermatis Ashmead, which monophyly of species-groups or species comprised 25 specimens (8 99, 17 f$) or limits in Eurytoma, but at least the four 2.2% of the parasitoid fauna. The Georgia Eurytoma species listed above have the rearing represents a new host record for E. propodeum densely setose lateral to the tylodermatis, but a previously unidentified propodeal foramen, and the petiole species of Eurytoma was also reared in most (Fi§- 40' Pt}) has one dorsomedial and surveys of the parasitoid complex of C. two anterolateral processes. Furthermore, obstrictus in western North America (Douc- the anterior margin of the first gastral ette 1948, Hanson et al. 1948, McLeod 1953, tergum (Fig. 40, Gtl) is emarginate and Dosdall et al. in press). Examination of depressed medially, and the tergum is voucher and additional reared material deeply depressed anterolateral^, to ac- from Alberta, British Columbia, Idaho, commodate the processes of the petiole Oregon, and Washington indicates the when the gaster is raised. Bugbee's (1967) western species is also £. tylodermatis. keY to species differentiates E. tylodermatis, Noyes (2002) listed 19 other curculionid in part, by stating the sculpturing of the species in 11 genera as part of 46 reported fourth gastral tergum extends over the host species in 14 families of Coleoptera, dorsal surface at least narrowly along the Diptera, Hymenoptera and Lepidoptera. anterior margin (couplet 20). The species Over 90 nominal species of Eurytoma are description, however, states that the sculp- known from the Nearctic region (Noyes ture of the fourth tergum is heavy ventro- 2002). Bugbee (1967) revised the North laterally, continues dorsally for about one- American species, but species recognition half to two-thirds of the surface, and then remains extremely difficult because of fades out so that the dorsal surface is variability of the morphological features smooth and shiny. The extent of sculptur- he used to differentiate species, and be- ing on the fourth gastral tergum appears to cause sexual dimorphism (cf. Figs 3, 4) be variable in species of Eurytoma, and the presents difficulties in recognizing conspe- appearance is partly affected by telescop- cific sexes. Rearing is necessary to make ing of the terga. Features that can be used the association, and one sex of several in combination to differentiate £. tyloder- species of Eurytoma remains undescribed. matis from similar species include the Bugbee (1967) examined over 4,000 speci- malar space lacking an alveolate boss (a mens originally identified as E. tylodermatis slightly raised area), the ventrolateral in the USNM and stated that "even this margin of the scrobes (Fig. 39, vis) being number was not enough to give an produced anteriorly and reflexed poster- adequate picture of the geographical dis- iorly, and the median channel of the tribution, or the range of variation of propodeum being distinct and defined several species in the complex" (Bugbee laterally by carinae formed by longitudi- 1967, p. 492). He keyed E. tylodermatis as nally aligned crenulae (Fig. 40). one of 48 species of his "tylodermatis A single species of Eurytoma, E. curculio- complex" and considered the species to num Mayr, has also been reported as reared be most closely related to £. bolteri Riley, E. from C. obstrictus in Europe (Dmoch 1975). diastrophi Walsh, and E. pini Bugbee. He Individuals of E. curculionum have a meso- also stated that the four species were coxal lamella according to Claridge and probably associated with the larvae of Askew (1960, fig. 2), which is absent from weevils and small moths that live in stems the North American specimens identified of various plants, either as primary para- as E. tylodermatis. 196 Pteromalidae Three species of Lyrcus Walker, one species of Mesopolobus Westwood, one species of Neocatolaccus Ashmead, and what likely are two species of Pteromalus Swederus comprised about 96% of the reared parasitoids (Table 1). Lyrcus is restricted to the New World. The genus is unrevised for the Nearctic, but Noyes (2002) listed 16 species from the region. Species identifications in this study are based on examination of type material of the North American species contained in the USNM, which excludes the four oldest names assigned to Lyrcus from the Nearc- tic. Walker (1847) described four species collected in Florida that are now classified in Lyrcus (Noyes 2002) and type material of these species is in The Natural History Museum, London. Although Burks (1975) examined the types, the names have yet to be placed adequately within a species concept of Lyrcus. Until this is done within a comprehensive taxonomic revision, it is possible that one or more of the four Walker names represents a senior syno- nym of a name used in this paper. Lyrcus maculatus (Gahan) was the most commonly reared species of all the para- sitoids, comprising 967 specimens (515 99, 452 (J1) or 5.3% of the parasitoid fauna. Georgia represents a new state distribution record for the species. Dosdall et al. (in press) first reared L. perdubius from canola seedpods in Alberta, putatively as a parasitoid of C. obstrictus, and Noyes (2002) listed Antho- nomus grandis, A. rutilus (Boheman), A. signatus (Say), Lixus musculus Say, and Smicronyx tychoides Le Conte (Coleoptera: Curculionidae) as other hosts. Six specimens (3 99, 3 SS) 0.5% of the parasitoid fauna) of Lyrcus incertus (Ash- mead) were also reared. This species is widespread throughout southern and east- ern USA. Although C. obstrictus represents a new host association, Noyes (2002) listed several other genera and species of Curcu- lionidae as hosts, including A. grandis and a single report of a Ceutorhynchus sp. (Pierce et al. 1912). Only one female and male of Mesopolobus moryoides Gibson were reared, which rep- Volume 15, Number 2, 2006 197 resent 0.2% of the parasitoid fauna and the first distribution record for Georgia and eastern North America. This species is a common parasitoid of C. obstrictus, its only known host, in western USA (Gibson et al. 2005). Two other species of Mesopo- lobus have also been reported as putative parasites of C. obstrictus in western North America, M. mayetiolae (Gahan) in Califor- nia (Carlson et al. 1951) and M. bruchophagi (Gahan) in Alberta (Dosdall et al. in press). Mesopolobus moryoid.es was misidentified as Mesopolobus morys (Walker) in North America until Gibson et al. (2005) correctly identified it and provided features to differentiate the two species from each other and from other regional species. Mesopolobus is yet another unrevised, spe- ciose genus in North America, with Noyes (2002) listing 20 valid species for the region. A total of 33 Neocatolaccus tylodermae (Ashmead) (13 99, 20 SS) were reared, which represent 2.9% of the parasitoid fauna. Although Georgia is a new state distribution record, the species was known from Florida and is transcontinental in the USA (Noyes 2002). Ceutorhynchus obstrictus also represents a new putative host associ- ation, though Noyes (2002) listed 15 other curculionid species in 10 genera as hosts. Pierce (1909) reared it along with E. cyaniceps from Lixus musculus, Wilson and Andres (1986) reared it along with L. maculatus from Rhinocyllus conicus, and there is a single published association with Anthonomus grandis (Pierce 1909). Boucek (1993) provided a key to the three de- scribed North American species of Neoca- tolaccus. He differentiated N. tylodermae from N. moneilemae Gahan on the basis of a rounded rather than medially carinate pronotum and truncate rather than medi- ally narrowly emarginate clypeus. Fore- wing setal differences also help to differ- entiate the species. In N. tylodermae the ventral surface of the costal cell has setae only over about its apical half and dorsally the forewing is bare behind about the basal half of the marginal vein so that three or four rows of ventral admarginal setae are visible within a distinct speculum (Fig. 43), whereas individuals of N. moneilemae have a line of setae along the length of the costal cell and the forewing is setose behind the marginal vein more or less to its base so that a distinct speculum is lacking and the admarginal setae are covered by dorsal setae. Five individuals (2 99, 3 ^ct; 0.5% of the parasitoid fauna) of Pteromalus were reared. One female was identified as Pteromalus cerealellae (Ashmead) based on examination of type material in the USNM, but the other specimens remain unidenti- fied to species (see below). Girault (1917) provided a key to several species of Pteromalus (as Habrocytus Thomson) in North America, but there is no modern revision and Noyes (2002) listed 46 valid species names in the Nearctic region. In western North America, unidentified spe- cies of Pteromalus have been reported from surveys in Idaho (Walz 1957), Washington (Hanson et al. 1948), British Columbia (McLeod 1953), and Alberta (Dosdall et al. in press). Examination of voucher speci- mens from these studies by the senior author revealed at least one unidentified species common to the four areas as well as a single rearing of Pteromalus puparum (L.) from the surveys reported by McLeod (1953) in British Columbia (Gibson et al. 2006). The unidentified species from western North America and P. cerealellae belong to a comparatively small group of Nearctic species whose females have the line of setae on the ventral surface of the costal cell interrupted medially and the bare band behind the marginal vein ex- tending the length of the vein (sometimes with one or two setae apically within an otherwise distinct bare band, Fig. 49). Females of the two species differ from each other most conspicuously in propo- deal features. In P. cerealellae the setae on the callus extend mesal of the postspira- cular sulcus posteriorly, almost to the 198 Journal of Hymenoptera Research plical carina (Fig. 31), whereas females of the species from western North America have the region between the postspiracular sulcus and plical carina bare. The propo- deal structure of P. cerealellae is very similar to that of the European species Pteromalus semotus (Walker) (Graham 1969, fig. 385), which was reported as reared from C. obstrictus in England and Poland (Murchie and Williams 1998). Boucek (1977) once considered the two names conspecific, but subsequently (Boucek 1988) re-established P. cerealellae. Among other features, P. semotus has an entire costal setal line (cf. Fig. 50). Pteromalus cerealellae was de- scribed from, and until recently was thought to be a monophagous parasitoid of, the Angoumois grain moth, Sitotroga cerealella (Olivier) (Noble 1932). Flanders (1932) stated that it would also oviposit into the tuberworm moth, Phthorimaea operculella (Zeller) (Lepidoptera: Gelechii- dae). However, Brower (1991) showed that it effectively parasitized 12 different spe- cies in 4 families of Coleoptera (including 3 species of Curculionidae) that are pests of stored products. He concluded that al- though the species probably prefers S. cerealella, it is more habitat specific than host specific. The second Pteromalus female reared and that of P. cerealellae are similar in having the forewing dorsally bare behind the entire length of the marginal vein (Figs 49, 50), but differ in those features given in the key. A comprehensive generic revision is necessary to establish the correct species name of the unidentified female. The three unidentified males may represent the opposite sex of the unnamed female, based on the presence of a continuous line of setae on the costal cell, but species char- acteristics of male Pteromalus remain large- ly unknown and the males are not in- cluded in the key. DISCUSSION The parasitoid fauna reared from B. napus seedpods in Georgia, 1994 through 1996, revealed the same five chalcid fam- ilies that have been reported as reared from C. obstrictus in western North America, including six species apparently shared in common (Table 1). Of the shared species, L. maculatus was by far the most commonly reared parasitoid in Georgia, comprising about 86% of the fauna. This contrasts to western North America where it appears to be only an incidental parasitoid of C. obstrictus. Furthermore, two common para- sitoids of C. obstrictus in at least some parts of western North America, N. tidius and M. moryoid.es, were reared as only incidental parasitoids in Georgia. The latter rearings represent the first distribution records of the respective species in eastern North America. If C. obstrictus was introduced to Georgia from western North America, the two parasitoid species may have been introduced accidentally at the same time. The most common parasitoid of C. obstric- tus throughout most of western North America, T. lucidus, was not reared in Georgia despite the statement of Buntin (1998) that most of the reared specimens consisted of T. perfectus (a misidentification of T. lucidus prior to Gibson et al. 2005). The second and third most commonly reared species in Georgia were L. perdubius and N. tylodermae, respectively. Because of their relative abundance and because neither has been reported from hosts other than Curculionidae (Noyes 2002), both species very likely are parasitoids of C. obstrictus. However, it remains to be determined whether they are primary or hyperparasi- toids. At least some of the other incidental species, such as C. torvina, B. allynii and E. cyaniceps, likely are hyperparasitoids rather than primary parasitoids. The rearing of £. glaucus, B. allynii, E. cyaniceps, L. incertus, P. cerealellae and the unidentified species of Pteromalus from B. napus seedpods in Georgia represent new rearing records, but these are at most incidental parasitoids, if C. obstrictus was the actual host for all the species. Buntin (1998) obtained the para- sitoids from mass-reared seedpods. The Volume 15, Number 2, 2006 199 very few Asaphes and Pachyneuron that were reared, along with several Aphidiinae (Braconidae) also preserved with the ma- terial, show that some aphid mummies contaminated the seedpods even though an attempt was made to remove these prior to rearing. Likewise, one or more of the uncommon parasitoid taxa may have emerged from other undetected insects within or on the pods. For example, the only other host record for E. glaucus is a lepidopteran. Individual rearing of para- sitoids dissected from seedpods is neces- sary to definitively prove the host associa- tions listed in Table 1, which at present are only inferred. Both L. incertus and L. perdubius have been reported previously as parasitoids of the cotton boll weevil, as has also E. cyaniceps, E. tylodermatis, and N. tylodermae , though not the most commonly reared parasitoid of C. obstrictus in Georgia, L. maculatus. These results suggest that the chalcid parasitoid fauna acquired by C. obstrictus in any area where it is introduced is partly influenced by what other curcu- lionid species occur in the region. If so, the parasitoid fauna from eastern Canada and the southeastern USA might be expected to differ as substantially as between eastern and western North America. ACKNOWLEDGMENTS The senior author gratefully acknowledges Eric Grissell (USNM) for access to the USNM chalcid collection and loan of type and other material, without which most species identifications would not have been possible. We thank Jennifer Read (CNC) for preparing the plates of illustrations used to clarify species differentiation, and John Huber and Peter Mason (CNC) as well as two anonymous reviewers for helpful suggestions regarding improving this manuscript. LITERATURE CITED Bigger, J. H. 1933. Parasites of the sunflower weevil, Desmoris fulvus Lee, during 1931 and 1932. Journal of Economic Entomology 26: 652. Boucek, Z. 1977. A faunistic review of the Yugoslavian Chalcidoidea (parasitic Hymenoptera). Acta En- tomologica Jugoslavia! 13 (suppl.): 3-145. Boucek, Z. 1988. Australasian Chalcidoidea (Hyme- noptera). A Biosystematic Revision of Genera of Fourteen Families, with a Reclassification of Species. C.A.B International Wallingford. 832 pp. Boucek, Z. 1993. New taxa of North American Pteromalidae and Tetracampidae (Hymenoptera), with notes. Jour)ial of Natural History 27: 1239-1313. Breakey, E. P., R. L. Webster, and E. C. Carlson. 1944. The cabbage seed pod weevil, Ceutorhynchus assimilis, in western Washington. Bulletin of the Washington Agricultural Experiment Station 455: 118-119. Brodeur, J., L.-A. Leclerc, M. Fournier, and M. Roy. 2001. Cabbage seedpod weevil (Coleoptera: Cur- culionidae): new pest of canola in northeastern North America. Fhe Canadian Entomologist 133: 709-711. Brower, J. H. 1991. Potential host range and perfor- mance of a reportedly monophagous parasitoid, Pteromalus cerealellae (Hymenoptera: Pteromali- dae). Entomological News 102: 231-235. Bugbee, R. E. 1967. Revision of chalcid wasps of genus Eurytoma in America north of Mexico. Proceedi)igs of the United States National Museum 118: 433-552. Buntin, G. D. 1998. Cabbage seedpod weevil (Ceutor- hynchus assimilis, Paykull) management by trap cropping and its effect on parasitism by Tricho- malus perfectus (Walker) in oilseed rape. Crop Protection 17: 299-305. Burks, B. D. 1975. The species of Chalcidoidea described from North America north of Mexico by Francis Walker (Hymenoptera). Bulletin of the British Museum (Natural History), Entomology 32: 139-170. Carcamo, H. A., L. Dosdall, M. Dolinski, O. Olfert, and J. R. Byers. 2001. The cabbage seedpod weevil, Ceutorhynchus obstrictus (Coleoptera: Curculioni- dae) — a review. Journal of the Entomological Society of British Columbia 98: 201-210. Carlson, E. C, W. H. Lange Jr, and R. H. Sclaroni. 1951. Distribution and control of the cabbage seedpod weevil in California. Journal of Economic Entomology 44: 958-966. Claridge, M. F. and R. R. Askew. 1960. Sibling species in the Eurytoma rosae group (Hym., Eurytomidae). Entomophaga 5: 141-153. Delvare, G. 1992. A reclassification of the Chalcidini with a checklist of the New World species. Memoirs of the American Entomological Institute 53: 119-441. Dmoch, J. 1975. [Study on the parasites of the cabbage seed weevil (Ceuthorrhyuchus assimilis Payk.) (Co- leoptera, Curculionidae). I. Species composition and economic importance of the larval ectopar- asites.] Roczniki Nauk Rolmczych (E) 5: 99-112 (Text in Polish.) Dosdall, L. M., B. J. Ulmer, G. A. P. Gibson, and H. A. Carcamo. In press. The spatio-temporal distribu- ZOO Journal of Hymenoptera Research tion dynamics of the cabbage seedpod weevil, Ceutorhynchus obstrictus (Marsham) (Coleoptera: Curculionidae), and its larval ectoparasitoids in canola in western Canada. Biocontrol Science and Technology. Doucette, C. F. 1944. The cabbage seedpod weevil, Ceutorhynchus assimilis (Payk). Bulletin of the Washington Agricultural Experiment Station 455: 123-125. Doucette, C. F. 1948. Field parasitizarion and larval mortality of the cabbage seedpod weevil. Journal of Economic Entomology 41: 765-767. Flanders, S. E. 1932. Recent developments in Tricho- gramma production (abstract). The review of Applied Entomology. Series A: Agricultural 19: 95-97. Gahan, A. B. 1914. Descriptions of new genera and species, with notes on parasitic Hymenoptera. Proceeding of the United States National Museum 48: 155-168. Gahan, A. B. 1919. Some chalcid-wasps reared from cecidomyid galls. Annals of the Entomological Society of America 12: 159-170. Gahan, A. B. 1933. The serphoid and chalcid parasites of the Hessian fly. United States Department of Agriculture Miscellaneous Publication 174: 1-147. Gibson, G. A. P. 1990. Revision of the genus Macro- neura Walker in America north of Mexico (Hymenoptera: Eupelmidae). The Canadian Ento- mologist 122: 837-873. Gibson, G. A. P. 1995. Parasitic wasps of the subfamily Eupelminae: classification and revision of world genera (Hymenoptera: Chalcidoidea, Eupelmi- dae). Memoirs on Entomology, International 5: 1-421. Gibson, G. A. P. 1997. Chapter 2. Morphology and terminology. Pp. 16-44 in: Gibson, G. A. P., J. T. Huber, and J. B. Woolley, eds. Annotated Keys to the Genera of Nearctic Chalcidoidea (Hymenoptera). NRC Research Press, Ottawa. Gibson, G. A. P., H. Baur, B. Ulmer, L. Dosdall, and F. Muller. 2005. On the misidentification of chalcid (Hymenoptera: Chalcidoidea) parasitoids of the cabbage seedpod weevil (Coleoptera: Curculioni- dae) in North America. The Canadian Entomologist 137: 381-403. Gibson, G. A. P., D. R. Gillespie, and L. Dosdall. 2006. The species of Chalcidoidea (Hymenoptera) in- troduced to North America for biological control of the cabbage seedpod weevil, and the first re- covery of Stenomalina gracilis (Chalcidoidea: Pter- omalidae). The Canadian Entomologist 138: 285-291. Gibson, G. A. P., J. T. Huber, and J. B. Woolley eds. 1997. Annotated Keys to the Genera of Nearctic Chalcidoidea (Hymenoptera). NRC Research Press, Ottawa. 794 pp. Girault, A. A. 1917. The North American species of Habrocytus (Chalcid-flies). The Canadian Entomol- ogist 49: 178-181. Graham, M. W. R. de V. 1969. The Pteromalidae of north-western Europe (Hymenoptera: Chalcidoi- dea). Bulletin of the British Museum (Natural Histon/) (Entomology), supplement 16: 1-908. Hanson, A. J., E. C. Carlson, E. P. Breakey, and R. L. Webster. 1948. Biology of the cabbage seedpod weevil in northwestern Washington. State College of Washington Agricultural Experimental Station Bulletin 498: 1-15. Hunter, W. D. and W. D. Pierce. 1912. The Mexican cotton-boll weevil: A summary of the results of the investigation of this insect up to December 31, 1911. United States Department of Agriculture, Bureau of Entomology Bulletin 114: 1-188. Kuhlmann, U., L. M. Dosdall, and P. G. Mason. 2002. 11. Ceutorhynchus obstrictus (Marsham), cabbage seedpod weevil (Coleoptera: Curculionidae). Pp. 52-58 in: P. G. Mason, and J. T. Huber, eds. Biological Control Programmes in Canada, 1981- 2000. CABI Publishing, Wallingford. Mason, P. G., T. Baute, O. Olfert, and M. Roy. 2004 (2003). Cabbage seedpod weevil, Ceutorhynchus obstrictus (Marsham) (Coleoptera: Curculionidae) in Ontario and Quebec. Journal of the Entomological Society of Ontario 134: 107-113. McLeod, J. H. 1953. Notes on the cabbage seedpod weevil, Ceutorhynchus assimilis (Payk.) (Coleop- tera: Curculionidae), and its parasites. Proceedings of the Entomological Society of British Columbia 49: 11-18. Murchie, A. K. and I. H. Williams. 1998. A bibliogra- phy of the parasitoids of the cabbage seed weevil (Ceutorhynchus assimilis Payk.). Bulletin. Section Regionale Quest Palaearctiquc, Organisation Inter- nationale de Lutte Biologique 21: 163-169. Noble, N. S. 1932. Studies of Habrocytus cerealellae (Ashmead), a pteromalid parasite of the Angou- mois grain moth, Sitotroga cerealella (Olivier). University of California Publications in Entomology 5: 311-354. Noyes, J. S. 2002. Interactive catalogue of world Chalcidoidea. 2nd ed. [CD-ROM]. Taxapad, Van- couver, British Columbia, and The Natural History Museum, London. Phillips, W. J. and F. W. Poos. 1921. Life-history studies of three joint worm parasites. Journal of Agricultural Research 21: 405-^26 + 6 pis. Pierce, W. D. 1909. Notes on the biology of certain weevils related to the cotton boll weevil. United States Department of Agriculture, Bureau of Ento- mology Bulletin 63: 39^14. Pierce, W. D., R. A. Cushman, and C. E. Hood. 1912. The insect enemies of the cotton boll weevil. United States Department of Agriculture, Bureau of Entomology Bulletin 100: 1-99. Urbahns, T. D. 1919. Life-history observations on four recently described parasites of Bruchophagus funebris. Journal of Agricultural Research 16: 165-174 + 2 pis. Volume 15, Number 2, 2006 201 USDA. 1960. Cooperative Economic Insect Report 10: 414. Walker, F. 1847. Characters of undescribed Chalcidites collected in North America by E. Doubleday, Esq., and now in the British Museum. Annals and Magazine of Natural History 19: 392-398. Walz, A. J. 1957. Observations on the biologies of some hymenopterous parasites of the cabbage seedpod weevil in northern Idaho. Annals Entomological Society of America 50: 219-220. Wilson, R. C. and L. A. Andres. 1986. Larval and pupal parasites of Rhinocyllus conicus (Coleoptera: Curculionidae) in Carduus nutans in north- ern California. Pan-Pacific Entomologist 62: 329- 332. Yoshimoto, C. M. 1971. Revision of the genus Euderus of America north of Mexico (Hymenoptera: Eulophidae). The Canadian Entomologist 103: 541-578. Yunus, C. M. and C. A. Johansen, C. A. 1967. Bionomics of the clover seed weevil, Miccotrogus picirostris (Fabricius) in southeastern Washington and adjacent Idaho. Technical Bulletin, Washington Agricultural Experiment Station, Washington State University 53: 1-16. 202 Journal of Hymenoptera Research Figs 1-8. 1 and 2, Conura torvina, female: 1, lateral habitus; 2, metasoma, dorsal. 3 and 4, Eurytoma tylodermatis, lateral habitus: 3, female; 4, male. 5 and 6, Brasema allynii, lateral habitus: 5, female; 6, male. 7 and 8, Eupelmus cyaniceps, lateral habitus: 7, female (insert: mesotarsus and apex of mesotibia showing spines); 8, male. (Abbreviations: bb = bare band, sp = spines.) Volume 15, Number 2, 2006 203 ilus cerealellae 9 Figs 9-18. Antenna, male and female. (Abbreviations: fu = funicular segment.) 204 Journal of Hymenoptera Research Figs 19-26. 19, Lyrcus perdubius, head. 20, L. incertus, malar space. 21 and 22, dorsal habitus, female: 21, Mesopolobus moryoides, 22, L. maculatus. 23-25, mesosoma, lateral: 23, L. maculatus; 24, L. perdubius; 25, Neocatolaccus tylodermae. 26, Trichomalus perfectus, pronotum and mesoscutum. (Abbreviations: am = anterior margin of metapleuron, md = malar depression.) Volume 15, Number 2, 2006 205 Figs 27-34. 27 and 28, Lyrcus incertus: 27, propodeum; 28, metapleuron. 29-32, propodeum: 29, L. perdubins; 30, Neocatolaccus tylodermae; 31, Pteromalus cerealellae; 32, Pteromalus sp. 33 and 34, scutellar-axillar complex: 33, P. cerealellae; 34, Pteromalus sp. (Abbreviations: am = anterior margin of metapleuron, bf = basal fovea, nuc = nucha, pc = plical carina, pnl = propodeal panel.) 206 Journal of Hymenoptera Research WJffMfl''' "'"f'M/ifibfiiiP'**' Figs 35^42. 35 and 36, Chlorocytus sp., propodeum: 35, female, posterior; 36 male; posterolateral. 37 and 38, Trichomalus lucidus, propodeum: 37, female, posterior; 38, male, posterolateral. 39 and 40, Eurytoma tylodermatis female: 39, head, frontolateral; 40, propodeum and base of metasoma, dorsolateral. 41 and 42, forewing: 41, Necremnus tidius; 42, Euderus glaucus. (Abbreviations: ams = admarginal setae, aps = adpetiolar strip, Gtl = first gastral tergite, nuc = nucha, pc = plical carina, pnl = propodeal panel, pss = postspiracular sulcus, ptl = petiole, vis = ventrolateral margin of scrobes.) Volume 15, Number 2, 2006 207 Nebcatolaccus tylodermae (43 ■Sr\\< Mesopolobus moryoides v*3 J: Lvrcus maculatus (40-' ^-r* bf ^ Trichoma l us lucidus 4J m&c-'T, -^ s. »* v Chlorocytus sp. (48 J? ■-.•v>" . "~ "* ■» - > N Pteromalus cerealellae (49. Figs 43-50. Forewing, female. (Abbreviations: ams = admarginal setae, bf = basal fold, cc = costal cell, mv marginal vein, sv = stigmal vein.) J. HYM. RES. Vol. 15(2), 2006, pp. 208-231 A Survey of the Bees of the Black Rock Forest Preserve, New York (Hymenoptera: Apoidea) Valerie Giles and John S. Ascher* Division of Invertebrate Zoology American Museum of Natural History, Central Park West at 79,h Street, New York, NY 10024, USA ^Address for correspondence: ascher@amnh.org Abstract— We present the results of a survey of the bee fauna of Black Rock Forest, Orange County, New York, USA. The survey focused on bees, with more limited data gathered for other incidentally collected groups such as apoid and vespid wasps. Surveys in 2003 with nets and bowls recorded 144 bee species (26 genera), 22 vespid species (9 genera) and 23 crabronid species (12 genera). Noteworthy records are detailed. A preliminary checklist of the bee fauna of the BRF is presented and discussed in relation to that of New York State, selected sites within the state, and of the northeastern USA as a whole. The cleptoparasitic species Sphecodes fattigi Mitchell, Sphecodes johnsonii Lovell, and Lasioglossum (Dialictus) michiganense (Mitchell), and the oligolectic species Osmia (Melanosmia) inermis (Zetterstedt) are newly recorded from New York State. Ecological patterns pertaining to sociality, nest type, pollen specialization, parasitism, and phenology, are summarized and discussed, as are the efficacies of different collecting methods. The net collected sample was richer than the bowl trapped sample in total bee species (117 vs. 113) and in the number of unique species (29, 20.4% vs. 25, 17.6%). Key words. — native bees, Bombus, Andrena, Apidae, invertebrate survey, invertebrate biodiversity, bowl trap, pan trap, trap nest, Black Rock Forest, pollination Bees (Hymenoptera: Apoidea) are the T. Griswold, unpublished; cf. Hurd 1979). single most important animal pollinators of Published data on the distribution of bees both native and cultivated vegetation in in New York State (NY) is limited (e.g., most habitats worldwide (Williams et al. Leonard 1928), but Ascher (unpublished 2001, Michener 2000). The mutualist re- information) has compiled a list of 423 lationship between bees and plants forms species known from New York, of which a key process in the maintenance of both 405 are native to North America, local biological diversity and agricultural Single-site inventories of poorly known productivity. As primary pollinators, bees invertebrate groups have the potential to provide a vital ecosystem service, affecting establish useful quantitative baseline esti- the integrity of ecological communities as mates of local biodiversity, as well as to a whole, including the health of humans help illuminate large-scale distributional (Williams et al. 2001, Nilsson 2000, Cane patterns within those groups. Such esti- and Tepedino 2001). Despite this ecological mates can prove useful in testing hypoth- importance, our understanding of some eses arising from practices as diverse as basic aspects of bee biology, including theoretical biogeography and conservation species level distributional patterns, re- planning. In addition, geo-referenced spec- mains incomplete. There are about 20,000 imen data are amenable to re-analysis and species of bees worldwide (Michener 2000) comparison with related data sets in the and approximately 3500 described species context of regional studies of biodiversity occur in America north of Mexico (JSA and across diverse groups. Finally, such in- Volume 15, Number 2, 2006 209 ventories help to address the need for natural history information that is crucial for understanding community-level eco- logical patterns (e.g. phenological patterns, host associations, habitat use, etc). Black Rock Forest (BRF) harbors a variety of distinct habitat types, many of which are typical of the larger Hudson Valley Region, and is managed in part as a long-term research preserve. Because the landscape matrix surrounding BRF is under increas- ing pressure from land conversion and habitat degradation, a survey of the bee communities of BRF while surrounding habitats are still relatively intact should provide a valuable basis for future com- parison with a variety of other sites across a range of spatial and temporal scales. The effects of environmental change on bee communities remain insufficiently under- stood. Many relevant studies have been published (see, e.g., Matheson et al. 1996), but few of these are from eastern North America. Cane (2005) notes that bees "possess a unique combination of salient foraging and nesting traits that together set them apart from other taxa studied in the context of habitat fragmentation". Many bees utilize open areas for foraging and nesting, and may benefit from forest fragmentation, unlike forest-dwelling songbirds. However, bees are still poten- tially vulnerable to habitat change, partic- ularly the loss of their host plants. RESEARCH OBJECTIVES This survey was undertaken with two primary goals; first, to assemble a faunal list of the bees (and selected aculeate wasps) of BRF to serve as a baseline inventory of use to both ecologists and conservation biologists, and second, to compare BRF data with other bee samples in order to shed light on larger-scale (regional) patterns of bee distributions and diversity. Second order objectives in- cluded enhanced representation of the regional bee fauna in the collection of the American Museum of Natural History (AMNH) and creation of a synoptic (taxo- nomic reference) collection to be housed at BRF. Ecological data were also gathered, such as abundance of bees across the season and on selected host plants, and the efficacies of various collecting methods (net collecting vs. trapping with bowls of three different colors vs. trap nesting) were tested. STUDY SITE AND METHODS The Black Rock Forest (BRF) is a 1520 hectare preserve and research facility lo- cated in Orange County, New York (Lat. 41.42267, Long. 74.03039), ca. 50 miles north of New York City (NYC). The BRF is situated within the highest portion of the Hudson Highlands. The terrain comprising the preserve ranges in elevation from about 135 m to 446 m. A network of closed canopy dirt roads permits access to within 1 kilometer of any point within the BRF. The landscape, both within the forest pre- serve, and across several large adjacent tracts (including West Point Military Acad- emy), is mostly forested (upland hardwood forests dominated by Quercus spp., Barrin- ger and Clemants 2003). Other local habitat types include successional hardwood stands, hemlock coves, chestnut-oak woods, red maple swamps, ponds, reser- voirs, and marshes. Important habitats for bees at BRF include small meadows, exposed road edges and reservoir edges, dams, and marshes, where flowering shrubs (such as Viburnum spp., Ilex verti- cillata (L.) A. Gray, Kalmia latifolia L., Clethra alnifolia L., Spiraea spp., Rhododen- dron spp.), and herbs (such as Veronica, Polygonum spp., Gnaphalium, Solidago spp., and Viola spp.) provide seasonal sources of pollen and nectar to bees. In addition, during early spring (April) prior to leaf- out, forest habitats hosted bees attracted to flowering trees such as Acer rubrum L., Salix spp. including S. discolor Muhl. and Prunus spp. In May, other flowering trees such as Craetaegus macrosperma Ashe, and other Prunus spp. were important re- 210 Journal of Hymenoptera Research sources for many Andrena and other bee one of three florescent colors: yellow, white species. The early spring flower Erythro- or blue. A total of 150 traps were deployed nium americanum Ker-Gawler flowered on each of 17 survey visits and arrayed in sparingly during our survey season and ten transects on each visit. Each transect was visited by relatively few bees. Vacci- consisted of 15 traps (five of each color), nium species, especially highbush blueber- arrayed in alternating colors. Traps con- ry Vaccinium corymbosum L. and lowbush tained a solution of Dawn brand blue blueberry V. august ifloiwn Ait. dominated dishwashing liquid (1 table spoon to 1 large areas of the forest understory at BRF, gallon tap water) and were placed in ten including dry hillsides, damp forest areas, sites on the ground along transect lines, open forest gaps created by fires, and wet Traps were deployed over a period of marsh edges in association with herba- approximately 1 hour beginning at 0730 hr ceous communities. Beginning in May, and and were in place before 0900 hr on survey continuing into late June, Vaccinium stands visits during which they were used. In- composed of several species were visited dividual traps were placed at approximate- by large concentrations of nectaring and ly one meter apart. Transect sites were "buzz" pollen-collecting bees. Vaccinum chosen opportunistically and included: stamineum L. (Deerberry) was moderately open fields, roadsides, reservoir edges, common in hillside forests. Patches of dams, forest floors and stone outcroppings Lysimachia were noted. throughout the BRF property. At the close The survey season during the spring and of each survey visit the traps were re- summer of 2003 was generally wet and trieved during a two-hour period begin- cool in southern New York as confirmed ning at approx. 1600 hr. The contents were by weather data collected at BRF. Above poured through sieves and the recovered average rainfall and below average tern- specimens were transferred to plastic peratures would be expected to depress whirl-packs containing 75% ethyl alcohol, bee numbers and collecting success. Locality data and bowl trap color labels Sampling schedule. — We conducted bee were recorded, surveys at BRF during 24 days between 31 Hand netting of bees was conducted March 2003 and 16 October 2003. Each between 09:00 and 16:00 during 23 survey survey day began at approximately 0730 hr visits. Collecting by hand-net was under- and was completed generally between taken opportunistically at sites where bees 1800 hr and 1900 hr. Most fieldwork was were thought to be concentrated. Hand conducted on days with predominantly netting was pursued most intensively in sunny skies and warm temperatures. Col- exposed sunny habitats such as fields, road lection sites visited per sampling day and edges, reservoir and marsh edges, where the time spent at each site varied. In many shrubs and herbaceous perennials addition, individual collecting sites were bloom and where bees were most likely to chosen throughout the BRF property op- occur. When bees were captured they were portunistically in response to the presence transferred to cyanide killing jars before of bees or abundance of flowering plants, being stored in vials. Vials were labeled UTM coordinates were recorded for all and placed in a cooler for transfer to the sites where bees were collected. laboratory. Sampling methods.— We collected bees at Twenty wooden 'Binderboard' brand BRF using 3 principal methods: colored trap-nests were deployed for the duration plastic pan (or bowl) traps, hand-held of the survey beginning on 27 May 2003. insect nets, and wooden trap nests. Bowl Ten trap-nests consisted of a wooden block traps were made from 6 oz. plastic Solo bearing 39 holes, each measuring 5.5 mm brand bowls that were spray-painted with diameter, and a depth of 10 cm. The Volume 15, Number 2, 2006 211 remaining 10 trap-nests were similar in other respects, but each bore 21 holes measuring 5 mm in diameter, drilled to a depth of 16 cm. Each hole was lined with a kraft paper tube to facilitate recovery of specimens. Trap-nests were mounted in sets of two, at 10 sites dispersed across the BRF property. Each nest was hung from a tree limb approximately 1.5 meters above the ground with the holes oriented to face south. Trap nests were checked periodical- ly to determine if any Hymenoptera had inhabited the holes and to ensure that they were intact and undisturbed. Trap nests were retrieved from BRF on 20 March 2004, and each trap-nest was examined in the laboratory for evidence of occupation by Hymenoptera. Specimens were sorted, mounted, and identified to species by the authors (initial- ly sorted by VG; species determinations then made or confirmed by JSA) except the more difficult metallic Lasioglossum (Dialic- tus) females, determined by S. Droege, Vespidae, determined by J. M. Carpenter, and the more difficult Crabronidae, de- termined by P. Gambino. S. Droege made, confirmed, and revised identifications for numerous Nomada, and L. Day made and confirmed identifications of Bombus sauder- soni and B. vagnns. Vouchers are deposited in the collection of the American Museum of Natural History (AMNH). A synoptic collection is housed at the BRF research facility. Duplicate specimens were dis- persed to various bee specialists. Comparative data on the North Ameri- can bee fauna as a whole, and on the fauna of New York State (NY), and of particular areas within NY, were compiled by JSA based on study of relevant taxonomic and faunistic literature and study of historical insect collections, especially those housed at: AMNH; Cornell University (CUIC); New York State Museum; National Muse- um of Natural History; University of Connecticut Insect Collections, Storrs; and Parker Gambino's personal collection (af- filiated with the AMNH). Recent collec- tions from across NY and from nearby states such as Connecticut were available, including material collected by the authors, P. Gambino, S. W. T. Batra, K. N. Mag- nacca, B. N. Danforth, D. L. Wagner, R. G. Goelet and their associates. All discussion of the past and present status and life histories of bee species found at BRF is based, in part, on these historical and recent collections and the literature in addition to the sample obtained during the survey of BRF. Totals cited for "south- ern New York" are for the area encom- passing New York City (NYC), Long Island, and all counties adjacent to Orange County (i.e. the southeast portion of the state north to Sullivan, Ulster, and Dutch- ess Counties). RESULTS AND DISCUSSION The survey collected and individually databased 6,542 bee specimens represent- ing 144 species, of which 138 are native and six are exotic (Appendix 1), 26 genera, and five families (Table 1). Of these, several records detailed below represent notable range extensions, the only recent known collection of a species in NY, or otherwise fill gaps in the known distributions of New York State bees. Other aculeate specimens incidentally sampled included 22 vespid species (9 genera), 24 crabronid species (13 genera), and 2 species of Isodontia (Spheci- dae sensu stricto) (Appendix 2). Only a sin- gle bee, an Osmia cornifrons female, em- erged from our trap nests. These were occupied primarily by eumenine (Vespi- dae) and Trxfpoxylon {Trypoxylon) (Crabro- nidae) wasps and were not examined in detail due to the lack of bees. Species totals. — Of the 144 bee species found at BRF we regard 138 as native to North America (Appendix 1). These are discussed first followed by the six species known or suspected to have been intro- duced deliberately or accidentally to North America from Europe or East Asia. Native bee species. — 212 Journal of Hymenoptera Research Table 1. Summary of the number of described bee species for each genus known from New York State, with totals for the Eastern USA (sensu Mitchell 1960, 1962), New York State (NY), southern NY as defined in the text (SNY), Black Rock Forest (BRF), New York City (NYC; i.e. the five boroughs), and Ithaca (within city limits; many additional species are known from the Ithaca vicinity in Tompkins Co.). The number of species not native to North America (i.e. adventive and introduced species) is given in parentheses following the total number of species. *No recent records. Superfamily Apoidea: Clade Anthophila (Bees) EUSA NYS SNY BRF NYC Ithaca Family Colletidae: Subfamily Colletinae Colletes Subfamily Hylaeinae: Tribe Hylaeini Hi/laeits Family Halictidae Subfamily Rophitinae: Tribe Rophitini Dufourea Subfamily Halictinae Tribe Augochlorini Augochlorella Augochlora Augochloropsis Tribe Caenohalictini Agapostemon Tribe Halictini Subtribe Sphecodina Sphecodes Subtribe Halictina Halictus Lasioglossum Family Andrenidae Subfamily Andreninae: Tribe Andrenini Andrcna Subfamily Panurginae Tribe Calliopsini Calliopsis Tribe Protandrenini Pseudopanurgus Tribe Panurgini Subtribe Panurginina Panurginus Subtribe Perditina Perdita Family Melittidae Subfamily Melittinae Tribe Macropidini Mncropis Tribe Melittini Melitta Family Megachilidae Subfamily Megachilinae Tribe Anthidiini Anthidiellum Anthidium Paranthidium 35 24(3) 34 5 114(2) 125(1) 4 3 17 14(2) 25 8(2) 15 0 0 8(3) 10(2) 3 1 1 1 1 1 1 1 1 1 1 1 3 1 1 1 1 1 8 3* 0 13 4 3 3 3 3 67(2) 39(1) 29(1) 25(1) 44(2) 87(1) 70(1) 40(1) 58(1) 66(1) 3 1 1 1 1 1 15 4 2 1 1 1 3 1 0 0 0 0 27 7 1 0 3 2 2 1 2 1 1 0 0 1 4(2) 2(2) 2(2) KD 2(2) 2(2) 1 1 1 0 0 1 Volume 15, Number 2, 2006 213 Table 1. Continued. Superfamily Apoidea: Clade Anthophila (Bees) EUSA NYS SNY BRF NYC Ithaca Stelis 15 6 2 2 1 4 Tribe Osmiini Chelostoma 3(2) 3(2) 2(1) 0 1 3(2) Heriades 3 3 1 1 2 1 Hoplitis 8(1) 6(1) 3 2 1 4 Osmia 30(4) 20(3) 12(3) 10(2) 5(3) 13(2) Tribe Megachilini Megachile 43(5) 22(4) 17(2) 7(2) 16(4) 13(2) Coelioxys 25 12 7 2 5 9 Family Apidae Subfamily Xylocopinae Tribe Xylocopini Xylocopa 2 1 1 1 1 1 Tribe Ceratinini Ceratina 4 3 3 2 3 3 Subfamily Nomadinae Tribe Nomadini Nomada 76 47 25 18 19 26 Tribe Ammobatoidini Holcapasites 3 2 1 0 0 2 Tribe Epeolini Epeolus 21 7 4 0 1 3 Triepeolus 26 7 1 0 1 2 Subfamily Apinae Tribe Osirini Epeoloides 1 1* 1* 0 1* 0 Tribe Emphorini Ptilothrix 1 1 0 0 1 0 Tribe Eucerini Eucera 7 1 1 0 0 0 Melissodes 27 12 6 2 9 8 Pcpounpis 1 1 1 0 1 1 Svastra 5 1 0 0 1* 0 Tribe Anthophorini Anthophora 6(1) 4 3 1 3 3 Habropoda 1 1 0 0 0 0 Tribe Bombini Bombus 21 18 15 8 12 15 Tribe Apini Apis 1(1) KD 1(1) KD KD KD TOTALS: 743(22) 423(18) 269(13) 144(8) 210(15) 274(15) Colletidae We collected only two species of Colletes, C. compactus and C. simulans, neither of which was numerous. This scarcity of individuals and species (vs. the nine known from southern NY; see Table 1) may reflect the low frequency with which Colletes is captured in bowls (S. Droege, T. Griswold, pers. comm.) and perhaps a scar- city of appropriate sandy nest banks in the vicinity of the sampling sites. Absence of Colletes inaequalis Say in net-collected sam- ples from early spring is surprising, as this is a conspicuous and locally abundant species across much of the northeastern USA and is often encountered as it collects pollen from maples (Acer spp.)/ which are numerous at BRF. 214 Journal of Hymenoptera Research Our sample of Hylaeus, including small below), both widely distributed across NY, series of only two ubiquitous Hylaeus spp., and by two species, L. acuminatum and L. H. mesillae and H. modestus, is also im- fuscipenne, restricted to eastern NY (e.g., poverished. At least four additional species absent from the Fingerlakes Region; see are abundant in nearby Putnam County range maps in McGinley 1986). (JSA and P. Gambino, unpublished) and Two widely distributed, pollen-general- should occur at BRF. ist species of carinate Lasioglossum (Evy- laeus) (sensu Michener 2000) were collected Halictidae (L cinctipes and L. quebecense) in addition to All three augochlorine species known the more localized L. (Evylaeus) nelumbonis. from NY are numerous at BRF. The The latter seems to be strongly associated abundance of Augochlora pura in our with aquatic emergent flowers. In our samples probably reflects the local avail- study, numerous L. nelumbonis were col- ability of rotting logs in which this species lected in pan traps placed along a causeway excavates its nests. The most abundant bee bisecting Jim's Pond, in which grew abun- in our sample, Augochlorella aurata (1,222 dant Nymphaeaceae (Nymphaea odorata). individuals collected) is a eusocial, Museum label data suggest that L. nelum- ground-nesting species that is numerous bonis may be a pollen-specialist of Nym- across most of eastern North America, phaeaceae and /or Nelumbonaceae, but Populations of this species from northeast- direct observations of pollen collecting ern USA and southern Canada were behavior by this species have not yet been known as A. striata (Provancher) prior to made due to the difficulty of observing and recent synonymy with A. aurata in Coelho's collecting bees on aquatic vegetation. (2004) revision of Augochlorella. Lasioglossum (Dialictus) individuals were, The two most common species of Aga- as expected, particularly abundant in our postemon in NY (A. sericeus and A. virescens) bowl samples. These were found to belong were collected, but two species present to 22 identified species (additional, poorly more locally in southern NY [A. texanus known species may be included among our Cresson and A. splendens (Lepeletier)] were undetermined metallic Dialictus; most not found. Absence of A. splendens is not males of this subgenus were not deter- surprising, as this species seems to be mined) including two socially parasitic associated with sandy nesting substrates. species (Paralictus sensu Mitchell 1960) The cleptoparasitic genus Sphecodes was and two black, non-metallic Dialictus spe- represented by S. galerus, S. levis, S. fattigi, cies ( = noncarinate Evylaeus; see Michener and S. johnsonii, the last two recorded for 2000). Among the identified species of the first time in NY (JSA has also collected metallic, pollen-collecting Dialictus collect- S. johnsonii in Fairfield County, Connecti- ed in BRF (i.e., Dialictus sensu Mitchell cut, new state record). Two additional 1960) the wood-nesting species L. coeru- Sphecodes species, S. atlantis Mitchell and leum, L. cressonii, and L. oblongum were each S. dichrous Smith, not found at BRF were numerous. Other notable metallic, pollen- collected elsewhere in Orange County in collecting Dialictus species include two 1962 (Tuxedo Park vicinity; AMNH). species typical of northern forests (L. Three Halictus species ubiquitous in the nigroviride and L. versans), a distinctive eastern United States were found in good species often found in sand pits (L. hctero- numbers, but the more precinctive H. gnathum), and an infrequently recorded parallelus Say was not collected. species (due in part to identification diffi- Lasioglossum sensu stricto was represent- culties) previously known in NY from a few ed by L. coriaceum and L. leucozonium specimens collected in or near the lower (regarded for the first time as exotic, see Hudson River Valley (L. cattellae). Single Volume 15, Number 2, 2006 215 females of the two socially parasitic Dia- The long malar space of A. rufosignata, in lictus species were collected, one of which, comparison to its likely sister species L. michiganense, has previously been re- Andrena mandibularis (LaBerge 1980), may corded in the literature solely from Mitch- be an adaptation to collecting nectar from ell's (1960) unique holotype female, col- the bell-shaped corollas of Vaccinium spe- lected in Wayne County, Michigan, in 1940. cies. Another oligolectic Andrena collected, Our single female specimen and another A. cornelli, is now thought to be a Rhodo- female collected 30 June 2004 at the inlet to dendron specialist based on field observa- Lake Myosotis, Edmund Niles Huyck Pre- tions by JSA in Virginia, label data for serve, Rensslaerville, Albany County, NY, newly identified material in museum col- by JSA and C. J. Daley are the first records lections, and the widely spaced scopal outside of Michigan. Despite a lack of hairs of this species that can be considered published records, this species is probably an adaptation that holds Rhododendron widely distributed across the northeastern pollen connected by viscin threads USA. It was recently found in Maryland (S. (Ascher, unpublished; cf. LaBerge 1980). Droege, pers. comm.; new state record) and Andrena violae, an oligolege of Viola, pos- southern Ontario, Canada (L. Packer, pers. sesses elongate maxillary palpi used to comm; new Canadian record). The male of extract nectar from its host. Viola is other- this species remains unknown. The female wise most often visited by long-tongued of L. michiganense possesses a conspicuous, bees such as Osmia that are able to reach its inner, subapical mandibular tooth, where- concealed nectaries. Andrena violae is nu- as the mandibles of other parasitic female merous across much of the eastern United L. (Dialictus) are simple (i.e., lack an inner States, excluding the colder areas of the tooth) with elongate slender tips. The other northeast, but was previously known in socially parasitic Dialictns found at BRF, L. NY solely from a single male collected at cephalotes, has recently been found in NYC Van Natta's Dam, Six Mile Creek, Ithaca, in Central Park, Prospect Park (JSA, new Tompkins County, 2 May 1936 (specimen records), and the Bronx (collected by P. examined, CUIC). This species was not Gambino). represented among collections made on Viola at this site and elsewhere in the Andrenidae Fingerlakes Region by JSA during 1997- Our BRF sample included 40 species of 2002, so evidence of its persistence in NY at Andrena but is still far from complete, as an a new station of occurrence is welcome, additional 32 species known from southern Other oligolectic Andrena at BRF include A. NY were not recorded. Our sample was krigiana, a specialist of Krigia (dwarf dan- rich in vernal species characteristic of delion), and A. fragilis, a specialist of northeastern forests such as A. imitatrix Cornus (Svida). Three Andrena specialists and A. nivalis. Species associated with of Solidago and Aster (tribe Astereae) were blueberry were particularly well represent- found, A. hirticincta, A. nubecula, and A. ed including the Vaccininm oligoleges simplex (but not its sister species, A. placata (pollen specialists) A. hradleyi and A. Mitchell, which has been collected recently Carolina, and the polylectic A. carlini and in Putnam Co., NY), as was the panurgine A. rufosignata. The last species is abundant Astereae specialist Pseudopanurgus andre- (but under-collected; cf. LaBerge 1980) in noides [we recognize genus Pseudopanurgus northern blueberry bogs, and evidently in the broad sense of Mitchell, 1960, reaches the southern limits of its range at including Protandrena {Heterosarus) and P. or near Black Rock Forest, as it is unknown (Pterosarus) of Michener, 2000]. Andrena from New York City, Long Island, and arabis is a specialist of Brassicaceae that elsewhere along the mid-Atlantic coast, may actually benefit from spread of in- 216 Journal of Hymenoptera Research vasive Garlic Mustard Alliaria petiolata (Bieb.) Cavara & Grande. Many species that regularly collect pollen from rosaceous trees and shrubs, and are known or suspected to be important pollinators of apples, were collected in good numbers, including A. miserabilis, A. (Melandrena) spp., and A. (Trachandrena) species. An- drena (Trachandrena) nitda was numerous at BRF, which is near the northern edge of its range in NY (see map in LaBerge 1973). Melittidae Although a deliberate effort was ex- pended to locate and collect from Vaccinum stamirieum, the host plant of Mclitta eick- worti Snelling and Stage (1995), this re- cently described species was not recorded during our survey. However, it has been collected nearby in Putnam County by P. Gambino, as has O. virga Sandhouse, another poorly known oligolege of Erica- ceae (see Cane et al. 1985; they recorded O. virga, as O. "felti", collecting "surprisingly pure" loads of Deerberry pollen; this species also uses other ericaceous hosts, M. Arduser pers. comm.). No Macropis were collected in this study although their host plant Lysimachia was present. Megachilidae Native megachilid species collected at BRF included the cleptoparasites Stelis (Dolichostelis) louisae (one female) and Stelis (Stelis) nitida (one female). The former is a colorfully marked parasite of native resin bees in subgenus Megachile (Chelosto- moides), including M. (C.) campamdae (the likely host in NY and New England), which reaches its northern distributional limits in southern New York. Stelis nitida was described in 1878 from specimens collected in Canada and NY, but there have been few subsequent collections from eastern North America. It is most likely a northern and montane species that parasitizes Osmia, or possibly large Hoplitis species. Our sample of eight native Osmia species includes series of the forest-associ- ated O. bucephala and O. pumila. We collected single specimens of three species that are scarce or absent in other recent collections from New York State, O. collin- siae, O. felti, and O. inermis. The last species, a probable oligolege of Ericaceae (M. S. Arduser, pers. comm.) previously unre- ported from NY, has also been identified among recent samples of bees from the Adirondacks (JSA and W. L. Romey, new record). Another Osmia species, O. dis- tincta, has been found elsewhere in NY (e.g., South Hill Swamp, Ithaca, Tompkins County; and along the Hudson River) and in Pennsylvania to visit Penstemon, includ- ing P. digitalis Nutt. The tuft of curved hairs on the ocellar region of this species would seem to be an adaptation for collecting pollen from Penstemon, although O. distincta is apparently not a strict oligo- lege of this genus (M. Arduser, pers. comm.). Our sample of native Megachile and associated Coelioxys cleptoparasites is im- poverished, perhaps reflecting the ineffi- ciency of bowl traps for capturing these strong-flying species (although Megachi- lini can be trapped in numbers in bowls of appropriate color, S. Droege, pers. comm.). The species captured are widely distribut- ed and numerous across New York, ex- cepting M. montivaga, which is known in the state from a few collections in southern NY (e.g., recently collected at Edmund Niles Huyck Preserve in Albany County). A report of this species from Ithaca (Leonard 1928) is based on a misidentified M. inermis. Apidae, excluding bumble bees Large and small carpenter bees were represented respectively by Xylocopa virgi- nica (locally very numerous at BRF, but most uncollected) and by two abundant sister species of Ceratina (Zadontomenis), C. calcaratn and C. dupla, that cannot be distinguished in the females. The related C. strenna Smith is also common in NY but was not collected. Volume 15, Number 2, 2006 217 All 18 identified species belonging to the size and long flight season (JSA has ruficornis group of Nomada (=Nomada s.str.) observed males flying as late as November in our samples are known or suspected to 10 in Ithaca, NY, a colder locality than BRF) be cleptoparasites of Andrena species, characteristic of this species. Other bumble Commonly encountered species of this bee species encountered include B. per- group at BRF and other forested areas of plexus and B. vagans, both generally nu- the northeastern USA and southeastern merous in New York forests and bogs, and Canada include the large, conspicuous the widely distributed B. bimaculatus and B. species N. luteolbides (a valid species griseocollis. We found few B. vagans, but the distinct from N. sulphurata Smith; see extremely similar (and thus infrequently Schwarz and Gusenleitner 2004) and N. identified) B. Sanderson i was found in maculata, both cleptoparasites of large surprisingly large numbers, including se- Andrcna belonging to the subgenus Mclan- ries of queens, males, and workers. Two drena (Milickzy and Osgood 1995), and N. Bombus ternarius were found. This is bella, a cleptoparasite of A. imitatrix. A new a species of northern affinities found probable host association between N. bella commonly south to the Catskills. Leonard and A. imitatrix was inferred by JSA (new (1928:1031-1032) regarded it as, "Essential- information) based on repeated co-occur- ly a Canadian and northern transition rence of these species at several sites species...", and stated that "the species is across several years. Females of N. bella not found near NYC. (Beq) [indicating J. have been identified (M. Schwarz, pers. Bequaert as the source]." Long-tongued comm.) but remain undescribed. Further bumble bee species belonging to subgenera study of Nomada with bidentate mandibles Fervidobombus [B. fervidus (Fabricius) and B. ( = Gnathias sensu Mitchell 1962) is needed pensylvanicus (Degeer)] and Subterraneo- to clarify separation of N. bella from N. bombus (B. borealis Kirby) that frequently ovata, N. lepida, and other similar species, visit clovers (especially Trifolium) were not We collected a single male Nomada austra- collected. Absence of B. fervidus is surpris- es, which is one of the three species ing, but B. pensylvanicus has been scarce in belonging to the erigeronis group ( = Cen- NY in recent years and is no longer, "An trias) known from NY. These are aestival abundant southern species, common as cleptoparasites of Agapostemon. far north as central NY..." (Leonard Anthophora was represented by the 1928:1032). Bombus borealis has always been wood-nesting species A. (Clisodon) termina- uncommon in New York State (Leonard lis, which is widely distributed and nu- 1928), and is generally absent from de- merous in northern and montane forests veloped areas (e.g., it is unknown from the from Siberia to eastern Canada [Davydova city of Ithaca, NY, but occurs in nearby and Pesenko 2002; these authors distin- countryside). guished the Holarctic A. terminalis from the Absence of Bombus (Bombus) affiuis in our Palearctic A. furcata (Panzer)]. sample of 1261+ bumble bee individuals is troubling because this species is well Bumble bees represented in historical collections from Black Rock Forest is a favorable habitat the northeastern United States, and is for bumble bees, and certain species were expected to be "...moderately abundant found in large numbers, especially Bombus in the eastern to southern parts of the [New (Pyrobombus) impatiens and its social para- York] State..." (Leonard 1928: 1031). How- site B. (Psithyrus) citrinus (also known to ever, this species has recently disappeared attack other Bombus species). Large num- from New York (e.g., from Ithaca and the bers of B. impatiens in our late season NYC area, JSA, unpublished) and else- samples reflect the unusually large colony where (L. Day, pers. comm.). The regional 218 Journal of Hymenoptera Research disappearance of B. affinis is coincident NYC) habitats in the eastern United States with an abrupt decline in B. (Bombus) to the point where it could be classified as terricola Kirby at Ithaca NY (Ascher, un- invasive. We collected 66 specimens from published), and elsewhere (L. Day, pers. on or around native vegetation and in comm.), as well as the extirpation of the bowls, and one female emerged from a trap closely related B. (Bombus) occidental is nest. Non-specificity to orchards should Greene from the San Francisco Bay Area not be surprising as Osmia (Osmia) species and elsewhere in western North America, such as O. cornifrons and the closely related and the precipitous decline of the endan- native species O. lignaria are polylectic, not gered B. (Bombus) franklini from its excep- specialists of fruit crops. In areas near tionally restricted range in southern Ore- where O. cornifrons were deliberately re- gon and northern California (Thorp 2005). leased (e.g., Patuxent National Wildlife Populations of B. affinis, and of all North Refuge, see Cane 2003), a very similar American species of subgenus Bombus, and Asian species, Osmia (Osmia) taurus Smith their obligate social parasites [e.g., B. has been found to be established. This (Psitin/rus) ashtoni; a queen of this species species has also been found in Huntingdon was collected at BRF on June 13 1988, by J. County in south-central Pennsylvania (VG, G. Rozen], should be carefully monitored, new data), but not yet in NY. as parasitism by Nosema and other para- We collected 10 Anthidium oblongatum, sites introduced and spread via the green- a species native to Europe and only re- house trade in Bombus colonies poses cently detected in North America (Hoebeke a potentially severe threat to their survival, and Wheeler 1999). This species is now Introduced bee species. — Our samples in- abundant in the mid-Atlantic States, New eluded numerous individuals of certain York, and southern New England, usually exotic bee species that have become estab- in association with favored host plants such lished and locally invasive in eastern North as Lotus corniculata, a weed generally America beginning in the 1990's. distributed in waste places such as road- Megacliile sculpturalis, a giant resin bee sides and abandoned lots, and Sedum. native to northeastern Asia, was first The halictine species Lasioglossum (L.) collected in New York State in 1997 leucozonium has long been present in North (Ascher 2001) and is now widely distrib- America and has therefore been generally uted and locally abundant in the Finger- regarded as native. However, its North lakes Region, and in southern NY, in- American range is restricted to northeast- eluding NYC. Outside of New York, M. ern USA and southeastern Canada and sculpturalis is now quite widely distributed does not include northwestern Canada or and has recently been found in additional Alaska (see maps in McGinley 1986). This northeastern states such as Massachusetts distributional pattern, and association of (Martha's Vinyard, P. Gambino pers. this species with introduced weeds such as comm.), Vermont, and New Hampshire Chicorium (Asteraceae), suggests that this (S. Droege, pers. comm.), as predicted by ground-nesting species is adventive from Hinojosa-Diaz et al. (2005). Europe, not native as has been assumed. The horn-faced mason bee Osmia comi- Molecular phylogenetic placement of L. frons, native to eastern Asia including leucozonium and L. zonulum (Smith) within Japan, was deliberately introduced by the otherwise exclusively Old World leuco- USDA scientists as a managed pollinator zonium species group, and lack of signifi- of apples. After wide distribution and cant genetic differences between Old and release, this species has recently estab- New World samples (see, e.g., Danforth lished large populations in natural and and Ji 2001), further support the idea that urban (e.g., Manhattan and Brooklyn, the occurrence of these species in North Volume 15, Number 2, 2006 219 America is adventive. It is possible that and 1,179 specimens (18.2%) belong to these species were introduced in soil parasitic species (Fig. la). Of the 144 bee carried in ships' ballast as has been species recorded in this study, 116 (80.5%) hypothesized for another ground-nesting are pollen-collecting species and 28 (19.4%) bee species native to Europe and found in are parasitic (Fig. lb). The abundance and our study, Andrena wilkella. Extensive diversity of parasites reflects a rich fauna sampling of variable molecular markers of vernal Nomada associated with Andrena such as COI is needed to test hypotheses of hosts. The preponderance of females (4321 native vs. adventive origin for bee species vs. 1977 males vs. 245 of unrecorded sex) with Holarctic distributions. Mcgachile cen- corresponds with the large number of tuncularis (L.) may be another early in- workers of eusocial species, including the traduction from Europe, as this species has two most numerous species at BRF. Of the not been recorded in Alaska as would be 6,543 bee specimens collected, 1,222 expected for a species with a naturally (18.7%) were Augochlorella aurata and 845 Holarctic range. (12.9%) were Bombus impatiens. The sample Workers of Apis mellifera (L.) were of 1,113 bumble bees collected was domi- abundant from mid-June and into October nated by B. impatiens (845, 75.9%) and its but were generally not collected. social parasite B. (Psithyrus) citrinus (154, Of the 144 bee species recorded in this 14.1%). study, six (4.2%) are exotic and 138 (95.8%) Seasonal patterns of occurrence can be are native. Of the 6,543 specimens collect- obtained from Appendix 1, which gives ed, 115 (1.7%) belong to exotic species, and extreme dates for BRF (by calendar date), 6,428 (98.2%) belong to native species. NY as a whole (by month), and the entire Wasps and other non-bees. — Our apoid North American range (by month) for each wasp samples include 23 crabronid species species. General patterns include an abun- (12 genera; Appendix 2). Some of these are dance and diversity of Andrena and their generally numerous in forest edge habitats Nomada parasites at forest floor sites prior in New York such as Ectemnius continuus, to leaf-out. At more open sites, seasonal which nests in holes in wood. Other turnover of the bee fauna was apparent, species collected such as Astata leitthostromi with notable peaks of abundance and and Bia/rtes quadrifasciata are ground-ne- species diversity corresponding with the sters that favor more open, often sandy bloom of favored plants such as Vacciniinn habitats. Our vespid wasp sample includes in late spring (visited by, e.g., Andrena and long series of the native paper wasp Polistes Osmia spp.) and Solidago in late summer fuscatus, both sexes of Vespula consobrina, (visited by, e.g., Colletes spp., Andrena a yellowjacket of northern (Canadian and simplex Pseiidopamirgus andrenoides, and Transition Zones) affinities, one individual the workers and males of the dominant of the rather scarce Zethus spinipes, and eusocial species Augochlorella aurata and a variety of eumenines including cavity- Bombus impatiens). Rather few oligolectic nesting species found in our trap nests. bee individuals were captured (292, 4.5% Ecological and behavioral patterns. — Eco- of the total) (Fig. 2a), but these represented logical information (summarized in Ap- a significant number of species in our pendix 1) was compiled for each of the 144 sample (19, 13.2%) (Fig. 2b). bee species from information found in Although soil nesting individuals and catalogs and revisions, primary literature, species predominated in our samples, hive and field observations, including those nesters, wood burrowers, and cavity- made during the BRF survey. nesters were also well represented (Fig. 3a, Of the 6,543 specimens collected, 5,364 b). Cavity-nesting species were numerous (82.0%) belong to pollen collecting species, relative to the number of individuals, as 220 Journal of Hymenoptera Research 1A Sociality of Individuals 1 B Sociality of Species Parasitic 18% Subsocial 1% Solitary 27% Parasitic 21% Subsocial Rusocial 21% Eusocial 54% Solitary 57% 2A Oligolectic vs. Polyectic Individuals 2B Oligolcctic vs. Polylectic Species Oligolcctic 5% Polylectic Solitary ^ 23% Other^^ „..-- 72% Oligolectic 13% | Polylectic Solitary 44% 3A Nest Substrate by Individuals 3B Nest Substrate by Species Other 18% Other 21% Wood Burrowers 7% Cavity 7% Wood Burrowers 3% Cavity 17% 4A Net and Bowl Catch by Individuals 4B Species Catch by Method Unknown 0% White Bowl 24% Net Only 21% * Blue Bowl 9% Yellow Bowl 33% Both 61% Bowl Only 18% Figs 1-4. Summary of ecological properties of Black Rock Forest (BRF) bees. 1, Sociality. The category "solitary" includes communal species. Some individual nests or local populations of species categorized as "eusocial" may be solitary: 1 A, percentage of individual bees belonging to each of the four recognized categories of sociality; IB, percentage of bee species belonging to each of the categories. 2, Pollen specialization; those classified as oliglolectic are specialists that usually collect pollen from only a single family of plants; the polylectic category includes polylectic solitary bees only; those in the "other" category include parasites and social bees, which are necessarily generalists (i.e. polylectic): 2A, percentage of bee individuals that are Volume 15, Number 2, 2006 221 several cavity-nesters were represented by singletons or doubletons. By contrast, wood burrowing species were relatively few (Fig. 3b), although some of these species were captured in large numbers (e.g., Augochlora pura). The large number of hive-nesting individuals relative to species likely reflects their eusociality (see above). A few species typical of more open and sandy areas were found at BRF (e.g., Lasioglossum heterognathum, Bicyrtes, As- tata), but sand specialists such as L. vierecki were not found. Efficacy and Biases of trapping methods. — The year 2003 was characterized by long periods of cold and cloudy weather and pans may have been particularly useful under these conditions as these allow catch during brief windows of sun on days when net-collecting would be unrewarding. Nearly twice as many individuals were bowl trapped than netted (4,322 vs. 2,221) (Fig. 4a), but the net sample was biased against certain of the most common and readily identified taxa (see above). Bowls were found to be particularly useful in forest and at the forest edge where flowers are few, dispersed, or in the case of trees and shrubs, difficult to reach. Where flowers are scarce, bowls may be particu- larly effective due to lack of competition from real flowers. Using bowls, we found certain inconspicuous forest-associated spe- cies rarely taken in nets such as Stelis nitida. Well known biases of bowl traps reinforced by our study include low catch rates for certain groups, especially fast and high- flying species of, e.g., Colletes, Megachile, and perhaps Melissodes, and high catch rates for slow and low-flying species of, e.g., small Lasioglossum, Andrena, Osmia, and Nomada. Our results generally support the currently accepted view that a combination of bowl trapping using multiple colors and netting is the best way to efficiently collect a plural- ity of species (S. Droege et al. protocol). Only 89 of the 144 bee species collected (61.8%) were collected by both nets and bowls, with 30 species (20.8%) unique to nets and 26 (17.6%) unique to bowls (Fig. 4b). The net collected sample was richer than the bowl trapped sample in total bee species (117 vs. 113) and in number of unique species (29, 20.4% vs. 25, 17.6%). Although wood and cavity-nesting bees were numerous in this survey, only one individual bee (the introduced Osmia cor- nifrons) used our trap-nests. The poor performance of trap-nests might possibly be explained by an abundance of natural nesting substrates (standing dead wood) at BRF. Alternatively, bees may have been out-competed for the trap-nests by eume- nines and Trypoxylon, or else the nests may have been placed in sites that ultimately proved to be too shady. Comparison to other bee faunas. — In com- parison to the bee fauna of NYS as a whole (423 species) and to the fauna of some well- sampled localities within the state such as Ithaca (274 species), the 144 species iden- tified in our BRF sample is relatively few (Table 1). However, several of these re- cords are of considerable biogeographic or ecological interest (see above). The NY bee fauna includes many species which are regionally rare and/or have highly special- ized ecological requirements, and are therefore unlikely to be found at BRF. oligolectic, polylectic and solitary, or other; 2B, percentage of bee species that are oligolectic, polylectic and solitary, or other. 3, Nest substrates: 3A, percentage of individuals belonging to each nesting category: soil, cavity, wood burrowers, hive, or other (primarily cleptoparasites that live in the nests of their hosts); 3B, percentage of bee species known or inferred to use the nest substrate indicated. 4, Collecting method: net vs. white bowl, vs. blue bowl, vs. yellow bowl: 4A, percentage of bee individuals caught by each method; 4B, percentage of bee species caught by net only vs. bowl only vs. both net and bowl. 222 Journal of Hymenoptera Research Nonetheless, it seems highly probable that at least 250 bee species could be present at BRF based on totals of 274 species recorded from Ithaca, Tompkins County, NY (Ascher, unpublished), in a colder climate than BRF, and ca. 300 species recorded from the vicinity of Carlinville in southern Illinois (Robertson 1929, Marlin and LaBerge 2001), in seemingly unre- markable farm country. The high number and proportion of singletons (28 spp., 19.4%), of doubletons (12 spp., 8.3%), of species known from a single sex (ca. 31 spp., ca. 21.0%) excluding Lasioglossum, and of rarely col- lected species (i.e., 3-10 individuals col- lected: 36 species, 25.3%), indicate that more prolonged and intensive surveying using the same methods would reveal many additional species, likely resulting in taxonomically and biogeographically significant specimens. Another indication of the incomplete- ness of sampling of the total BRF fauna is that only 57.8% of the 249 bee species known from southern New York excluding NYC and Long Island (i.e., Sullivan, Ulster, and Dutchess, Orange, Putnam, Rockland, and Westchester, Counties) were found. These might be considered to represent a regional pool of species from an area relevant to BRF. The total of 249 species known from an area relevant to BRF is only 59% of the species total for New York State as a whole (423), which in turn is only 57% of the 743 bee species known from the eastern USA. Twenty additional species recorded in NY only from coastal NYC and Long Island (e.g., the coastal dune special- ist Lasioglossum marinum) are less likely to occur at BRF. Most bee species at BRF are widely distributed in NY and have been recorded from other well-collected sites such as Ithaca (123 species shared with BRF, 85.4% of the BRF total) and NYC (103 species shared with BRF, 71.5% of the BRF total). Northern elements of the fauna at BRF can be defined as those species known from the northern and montane portions of NY (e.g., the Adirondack Mountains and in most cases Ithaca), but absent from NYC, Long Island, and other warmer and coastal areas. Examples of northern species occur- ring at or near their southern limits at BRF and unknown from NYC include Andrena rufosignata, A. algida, Stelis nitida, Osmia felti, O. inermis, Bombus ternarius, and possibly B. sandersoni (southern distribu- tional limits of this species remain un- certain due to identification difficulties versus B. vagans). Although these northern species are likely genuinely absent from NYC, many of the 42 species known from BRF, but not NYC may be found in the latter area when more thorough samples have been made of semi-natural habitats such as Pelham Bay Park. Southern ele- ments in the BRF fauna include the following species that are unknown from the very well collected Fingerlakes Region (which includes Ithaca): Lasioglossum bni- neri, Andrena nuda, A. eonfederata, A. hilaris, and Melissodes subillata. The apparent ab- sence of these species from Ithaca and elsewhere in central and northern New York is probably genuine and likely reflects a real faunal difference from BRF. Andrena violae is another species of southern affin- ities that is very rare in Ithaca (see above). The Sorensen index [Cs = 2a /(2a + b + c) where a is the number of species shared between two sites, b is the number of species found at only one site, and c is the number of species found only at the other site] was used to quantify similarity be- tween various sites. The total for BRF vs. Ithaca is 58.9% whereas the total for BRF vs. NYC is 58.2%. The similar Sorensen values for comparisons involving these two areas (despite BRF's much greater geographical proximity to NYC) reflect many shared widespread and northern species with Ithaca, and significant differ- ences between BRF and NYC due to the presence of northern forest elements (e.g., blueberry associates) only at BRF and of coastal /sand associates only in NYC. Volume 15, Number 2, 2006 223 ACKNOWLEDGMENTS We are grateful to the staff of Black Rock Forest for their invaluable advice and assistance with practical matters, and to the Steifel Foundation for an award from their Black Rock Forest Consortium Small Grants for Scientific Research and Education program awarded to J.G. Rozen and V. Giles. We thank: S. Droege, L. Day, and T. Griswold for identifying bees, and M. Arduser for sharing unpublished keys and other taxonomic information and commenting on the manuscript. P. Gambino provided unpublished re- cords of NY bees and identified many of our crabronid wasps. K. Magnacca provided bee records and other assistance. The following people kindly provided access to and information about historical collections of New York bees: B. N. Danforth, E. R. Hoebeke, J. K. Liebherr, T. McCabe, T. Schultz, D. Yanega. S. Kornbluth generously provided assistance with specimen labeling and data entry and manage- ment. We thank K. Russell for her helpful discussons of methods and data base management. C. Encarna- cion, and E. Burbano prepared our specimens. T. Nguyen lent his assistance with technical matters. J. G. Rozen, J. M. Carpenter, A. C. Vallely, and C. Dong made valuable comments on this manuscript. J. M. Carpenter identified our vespid wasps. We thank R. G. Goelet for his support of bee research at the AMNH. We are especially grateful to J.G. Rozen for his guidance and support. In addition to his help in this project, we both owe our interest in bees to his generosity and encouragement. LITERATURE CITED Ascher, J. S. 2001. Hylaeus hyalinatus Smith, a European bee new to North America, with notes on other adventive bees (Hymenoptera: Apoidea). Proceed- ings of the Entomological Society of Washington 103: 184-190. Barringer, K. and S. E. Clemants. 2003. The vascular flora of Black Rock Forest, Cornwall, New York. journal of the Torrey Botanical Society 130: 292-308. Cane, J. H. 2003. Exotic nonsocial bees (Hymenoptera: Apiformes) in North America: Ecological impli- cations. Pp. 113-126 in: Strickler, K. and J. H. Cane, eds. For Non-native Crops, Whence Pollinators of the Future. Thomas Say Publications in Ento- mology: Proceedings. Cane, J. H. 2005. Bees, pollination, and the challenges of sprawl. Pp. 109-124 in: E. A. Johnson and M. W. Klemens, eds. Nature in Fragments: the Legacy of Sprawl. Columbia University Press. Cane, J. H., G. C. Eickwort, F. R. Wesley, and J. Spielholz. 1985. Pollination ecology of Vaccinium stamineum (Ericaceae: Vaccinioideae). American journal of Botany 72: 135-142. Cane, J. H. and V. J. Tepedino. 2001. Causes and extent of declines among native North American polli- nators: detection, evidence, and consequences. Conservation Ecology 5: art. 1 [online]. Coelho, B. W. T. 2004. A review of the bee genus Augochlorella (Hymenoptera: Halictidae: Augo- chlorini). Systematic Entomology 29: 282-323(42). Danforth, B. N. and S. Ji. 2001. Australian Lasioglossum + Homalictus form a monophyletic group: Re- solving the "Australian enigma". Systematic Bi- ology 50: 268-283. Davydova, N. G. and Yu. A. Pesenko. 2002. Bees of the subgenus Clisodon (Anthophora, Hymenoptera, Apidae) in the fauna of Russia. Entomological Review 82: 621-624. [translated from Zoologicheskii Zhumal 81(11)] Hinojosa-Diaz, I. A., O. Yaiiez-Ordonez, G. Chen, A. T. Peterson, and M. S. Engel. 2005. The North American invasion of the giant resin bee (Hyme- noptera: Megachilidae). Journal of Hymenoptera Research 14: 69-77. Hoebeke, E. R. and A. G. Wheeler, Jr. 1999. Anthidium oblongatum (Illiger): an Old World bee (Hyme- noptera: Megachilidae) new to North AMerica, and new North American records for another adventive species, A. manicaium (L.). The Univer- sity of Kansas Natural History Museum Special Publication 24: 21-24. Hurd, P. D., Jr. 1979. Superfamily Apoidea. Pp. 1741- 2209 in: K. V. Krombein, P. D. Hurd, Jr., D. R. Smith, and B. D. Burks, eds. Catalog of Hymenop- tera of America north of Mexico, vol. 2. Smithsonian Institution Press, Washington, D. C. LaBerge. 1973. A revision of the bees of the genus Andrena of the Western Hemisphere, Part VI: Subgenus Trachandrena. Transactions of the Amer- ican Entomological Society 99: 235-371. LaBerge. 1980. A revision of the bees of the genus Andrena of the Western Hemisphere, Part X: Subgenus Andrena. Transactions of the American Entomological Society 106: 595-525. Leonard, M. D. 1928. A list of the insects of New York, with a list of the spiders and certain other allied groups. Cornell University Agricultural Experiment Station Memoir 101: 5-1121. Marlin, J. C. and W. E. LaBerge. 2001. The native bee fauna of Carlinville, Illinois, revisited after 75 years: a case for persistence. Conservation Ecology 5: art. 9. [online] Matheson, A., S. L. Buchmann, C. O'Toole, P. Westrich, and 1. Williams, eds. 1996. The Conserva- tion of Bees. Linnaean Society Symposium Series, Number 11. Academic Press, London. McGinley, R. J. 1986. Studies of Halictinae (Apoidea: Halictidae), I: Revision of New World Lasioglos- sum Curtis. Smithsonian Contributions to Zoology, no. 429: 294 pp. Michener, C. D. 2000. The Bees of the World. The Johns Hopkins University Press, Baltimore, Maryland. Miliczky, E. R. and E. A. Osgood. 1995. Bionomics of Andrena (Melandrena) vicina Smith in Maine and 224 Journal of Hymenoptera Research Washington, with new parasite records for A. (M.) regularis Malloch and a review of Melandrena biology, journal of the Kansas Entomological Society 68: 51-66. Mitchell, T. B. 1960. Bees of the eastern United States, Vol. 1. North Carolina Agricultural Experimental Station Technical Bulletin. Mitchell, T. B. 1962. Bees of the eastern United States, Vol. 1. North Carolina Agricultural Experimental Station Technical Bulletin. Nilsson, L. A. 2000. Critical resource levels for viable wild bee populations, www.cbm.slu.se/forskning/ naturvardskedjan/pdf/ wildbee.pdf Robertson. 1928. Flowers and Insects. Lists of visitors of four hundred and fifty-three flowers. Carlinville, Illinois: Published by the author. Schwarz, M. and F. Gusenleirner. 2004. Beitrag zur Klarung und Kenntnis parasitarer Bienen der Gattungen Coelioxys and Nomada (Hymeno- ptera, Apidae). Linzer biologische Beitrdge 62: 1413-1485. Snelling, R. R. and G. I. Stage. 1995. A revision of the Nearctic Melittidae: The subfamily Melittinae (Hymenoptera: Apoidea). Natural History Museum of Los Angeles County Contributions in Science No. 451: 1-17. Thorp, R. W. 2005. Species profile: Bombus franklini. In: Shepherd, M. D., D. M. Vaughan, and S. H. Black, eds. Red List of Pollinator Insects of North America. CD-ROM Version 1 (May 2005). The Xerces Society for Invertebrate Conservation, Portland, OR. Williams, N. M., R. L. Minckley, and F. A. Silviera. 2001. Variation in native bee faunas and implica- tions for detecting community changes. Conserva- tion Ecology 5: art. 7. [online] Volume 15, Number 2, 2006 225 O X; 01 _2 X "3 E o 3 u s en c "re OJ o E (J 01 01 x a. en v ^^ an re 3 3 G 73 C -5 as (0 c QJ 1/1 pb U ~- 0) to c _^ •; 01 X 1/1 en QJ 0) qj y 3> O) > a. in ^ QJ -a re e o « >■ Si 5 e_) QJ 01 1-7 en e P-. O .3 .3 3 3 .£ 3 >, OS Oh bC 3 ■ en £ £ .s £ QJ q a en 2 3 Oi jj. * 3 z -^ £ E XT1 o C/l =3 S 2 CO QJ O S 9 £ 5 o -5 re a. 3 «3 x, -a bD qj X "O ^ c >" w Z ^ o M •S .2 u. § 3 a X *. *- X KJ bD QJ - 2 (0 ej C 3 Sri G Ol ■ — * y X oi >, 73 to u £ aj en 3 01 IB c o C/j en -a i* o u 0) »*-H PL Cei CO OJ X 03 O >-~. ej >, 03 l. m O be 3 re J2 ~ .-H ^ i sj i1 QJ Oh en bC — 3 "3 — 1-73 Ol QJ 03 ^ oi e QJ en en .2 oi ■y Cj- ej ^ ~ O 'en ^ ' en ~ 03 S > C =3 .2 c .-5 '- 73 T5 "2 c rt O en ■r; re3 7-j -S 0 o 2 J 01 ch i- Q ^ e^ 2 fa C CQ Si T3 c E 3 en u '> ej 03 — ej n tfl 'u _, o £ 1/3 ■O en QJ QJ U T. 03 03 *- U « 03 C X _03 h a- S JS ■cl 03 — x; a. a. *-. ^ 2 "a U QJ •S "3 Ol 0) 3 60 if S * 5 o g c w -. SdxT en o w Oi 0 2- x C ej '3 C Oi a O en 0) 4= ffl en s § O1 7* O Ph O e-0 en _o a; bo 03 3 .y « T3 en 3 ^ en 01 y - 2 i— —- ^-> Xi — J5 U 7= '3 ^ 5 03 o re -r Oi £ •- 3 en QJ JZ ■M (J 03 03 — QJ en ^ ' O to -S en en _g r^ 03 Uh 2 01 CD (J Sh en 0) 0 ii / o u- _2 ^; o ■4^ ej o ^H ci 3 .tjj ej •"tr. O1 a Dh oj en S g =- S 73 ^ 3 OJ S x T3 ° 3 73 03 Ol a I 03 ej jy 3 en 3 X en T3 S o (C '-I en ■a o QJ 03 ■a .2 C QJ o y 2 5 03 _ 03 .3 ^ 3 s; bC i- QJ s £ O" O g 3 "3. c Oi 73 0) u _0) "o u 01 > 3 en Ol X oi 03 03 — , X 3 03 O ej 01 * b .2P-2 SI ffi E S £ £ T3 •G £ Ol 01 Cj-X C 3 3. 3 -3 Ol QJ 01 2 £ > o o o ^ Z Oi X > en 03 -r; X 03 i— "^ > ^ 3 ro g 3 y -T7 u s .y .y. S '■£ X 01 en a X en en en JS X "o w M O en 3 ■*-" . J c= « 22 x '0 < S a . .Jh en u 3 X 03 Oi ej 1- 73 03 II — w °2 3 .y g S 8 £ ^ B 2 3. • C c en 3 I > C en 01 Oi 'qj ~ •a en S en « ^ a, ej 8 >, X O 03 »*S 01 . en X £ 73 '3 .y en 73 "3 5 "2 en Oi 3 N 3 en 3 3 > Ji 73 2 o Oh ej O1 bO C .£ 73 ' OJ aij QJ M-. 73 ^^ 3 01 K, en Si O 01 > 2 3 o X en •4-> OS L_; X tj 3 3 en en OS en en 0) _o T3 u 0 . re 3 X Oi ~ -3 >- I 5 3- en en re U .5 QJ ■' 33 '— O OJ U > _>, 03 i-, m 03 QJ 3 03 bC y S T3 - £ QJ en 3 OS 3 5 o ■H 73 Q 3 9 en en 03 i- X ^ ebO Qj 3 >- ■ 3 >-. ±i Qj o '"3 >- .3 5 Oi ,a X x- c re c Oi < Us O c _o 3 .3 D y .3 en 73 3 a bo £ t •- 73 3 0> c u 3 01 Oh en 3 73 01 01 73 X 01 ,^- 73 O 0 u / u 23 >- en 1— i 01 01 >< X -5 (0 c 01 Oh 3 73 Oh '§ < 73 3 + Qj ' 1 £ c H O c c t y 01 x y c D 3 73 u 3 -C1 bCTJ 3 .2 u ej Qj u I a £ 3 £| 3°| u *^ .y >- 2 jg z I ■x y 3^ y £ -£ P > ™ 3 O Z C Oh en 2 £ S £ — re 01 3 ^3 OJ Oh 3 •— o bC^ > -p re y x .y qj 33 X ° OJ i 1- 3 •. "o £• Oh .y ^ -t-1 to "5 ."3 re T3 lH O 03 en ^-X •- y o re UJ •? 3 W >, > 3 „ti £ 3 re .y 3 5 o X .3 QJ re 3 •-^- 03 11 .y PL, 73 s 3 (- ■ — O 3 3 re 3 x £ ~ o "3 II 73 U 3 en QJ •-1 c 01 .3 O _3 en 3 O O) £ 3 c eu u 03 3l ^ .3 QJ ^ y 73 -x. QJ o 3 u £ o 3 c '03 Oh ^ £ X eu C3 ■£ t; bo 3 O O 3 u 3 Qj OS In en -- £ y u X 03 -3 QJ '73 X c I 2 5 o £ c >, 01 X X x o 03 "g 01 3 « o c ■ Sl OS QJ S II X .5 < 2 en i^ qj re 01 > X 73 O u lH »H 5 O Oh en X 3 en en 01 z 3 o X 3 X -5 ra x £ 73 £ 2 en I - ^ 01 : — ^ g_ "c 3 ■> 0 2 ^ E 1 ° -s I >,'en u .— 0 u en V 3 O. Qj en _-; C co 01 !s p- u O en en qj X 3 u 01 Oh ^_v en i-2 73 Oi >> -3 O x y bO X 3 .y S --! Oh Sj O ^- ^~. 01 E 3 re c 3 ^ JS u _0) g 3 ^ roz bC J •bb"£ 3 G G ^ -T Oi Qj >h ^ 73 £ .y Qj *4h o g 03 en '-I QJ U c r^ Q> -1- X X f- 4—1 y ° G £ "° -r- ° ^ O ^ ^ X -ri _ re en 75 ^ S -2 3? ,2 XS 1 £ o >, en t- Qj en en 03 G 73 ~ QJ m_ N G 'C >. y O u '" re 01 re u 73 3 > OS 03 X x 01 u x o OhX be c .£ < •—I ^N 7J ° 3 3 Q u Bjeqil A4ID I^OA M3N ■Miaypads ua[)oj -\4!lBpog a;ej4sqns isa^ 3140x3 sa aApefyj asuppunqe \^ pouad iqSiy 3tpJEa|sl poiisd mSiij ^n pouad jviSiq jaa s|pnpiAipui j>ig xxxx xxxxxx xxxx xxxxx G G Cl, Cl, ll.Ill.3,3.1 eXlLTlcrityj e^pJeXjeXje^OH i/i«UU pocn^c/jen 1 zzzz zzzzzz U _L 1 n t-h Tf it, ro^'^f'^iricc 00 000000 T— 't-^(JNC7N ,— ' l-H ^H ^H t— < r— I I I I I I I I I I I ^ ,-h ^ ^ ._ ^ ^ ej o ^ ^ u M \Z £X M n O 4 1 1 1 40 1 Q (^ 00 O T ^ T 111 -r *-r 1 w I 00 ^ N ^f o M ir, OfNrHlNtNrH^tnrtOrH ^ bObC^y y 'C >^bO>.bJ0 i-iH--i-ireSSp,re-ire-^ $<<:>iix;<,2<2< OvLnntNt^tNNN i-h tN (N (N 00 i-h M tN ^— , r<3 2 -*-< C3N r-H v s en 3 LTj I > .V. ej T— •* 'C t^ -^ X IN en e? 03 IX %* 3 '0 OJ X O es -- 0 ^^ ,U 03 X 00 7J 00 en _QJ 03 £ 3ee3 oc 03 LL, en 00 A T— 1 c 3 en en Ol 00 00 1— 1 Oi u C U I-H <-z en S 2 03 en .en Lf) 3JC 03 en '7! 5 eo en 3 ^ 3 c 0 en en en en en 73 en "r- O bo i CJ3 ^ Q — > 0 u Vi Oi Vh u en K en 3 eij .en -L. Q .en lC ■S b r- IH(J\UU001Jl^r lOS<^OONt^t^ONi— I ON i— < ON „ _L I I I I I I I I I I I I I I I I I I I I I I I I I I »^in^rO'tin<*inin'tvom^t'*in^|iriiri't^ioniir4fn't no in •* r-t ,>,>, >^ >, ■£ 3CrerererereCX ON •<* ON fN Cn| m CO 3 00 3 < I (N CX QJ en 1 -+ Cn| re C s 1 00 _^ _h ej o I ON a, ex a> 01 en en 1 1 on on in U en en 1 00 ^ ►^ ^ rN T tn O vo 1 1 00 00 — < CX Q-, <° ex re re re 00 _H 2 < a Ci. re re < 5 S rl ^ bC O.O < rX _L 1 -. °C 'C ■£ S ^ f^ at 2 < < CA) H 33 .-fcj O 01 en I Vi 3 -S -5 .<50 3 NO 00 ON >;cn o?m — 00 3 rH -C K -t-> — s i en y 00 ix m ^ m >.- 00 -rS 1— 1 t ■5 -3 5 & en ^^ O 00 ON 00 00 00 <— 1 c J 01 c re 01 X; O T3 — .O &u s 2 fc fc s — 1 w « o s kJ O lid lie, ia) n/1 J « S c 1 .V) to Selado Vi m vi ■^ ro ^u "« ■a -'- Vj V) tn Q 0 0 .u ,ej tj -C ^st ^^; ' — 1 t- — 1 > — 1 c: c ^ en e^ en * £ £ -2 -2 00 00 ■2 -Q U-j Vj « e -J -J V) V) V} Vj o o ■~ 5 2 "" 8 U K u '_ ►a CN ON ^__ ^1 =§ s en _£, o £ M C K if ri NO ,— . ON CO „ 00 ON o ^ 00 ON „> ^ °0 o ON tn UJ 0 0 en 0 ^ en JS 01 X) In 0) "re O J^ O es 0 Pi o =3 ^ il "H- "oJ c e. •£ K ^ s; fc £j Pi 60 3 DO ' — ' Q ■*- P^ -~ s a ^J -3 — ' 00 o NO ON 01 ej en 22 U •5 ~ ~ VJ ^ S en -*— ■*— n» ey eu ?U j;- t^- C -3 -3 en en rr ' r^, e~ es R a ►J g k| kl 3 3 -I I 'Tt Vi Vi Vh in iTi Vi V) ej ■ S -S -S -^ ."= "£ "£ "2 "5 "5 "5 "5 "5 "5 "S C 3 K 3 S 2 6 OO DO ■2 -2 en en e= c -J »-) en Vj en en •2 -2 00 cjq •2 -2 en en -J >-J £ £ S S£33S:33K «~~ = assSS£S « « « « Oi Oj » m K m B £ •£ IT, tSl ,j-j ,,> V) v> #, Zl 3 3 en en en Wj V) r^j -2-2^-2 0000 S 8 § I I I I g g 'I 'I 'I '§ 'I bo do .0 .0 en en -J -J Volume 15, Number 2, 2006 227 -c 3 C c o U X T3 C 01 a, < e:>t?mi X4Q ^JO^ M3(\l Aipypads uajjoj /Viijepog a)BJ4sqng jsa^j D140X3 SA 3AI4EN ajuepunqe XN pouad 4L|8i[J Ji43iEa[\i pouad 4M§!U AJM pouad 4M§!U JJ19 s|BnpiAipui jag a. X X X X X X X X X X X X X XXX XXXXXX X XX XXX xxxxx XXX XX xxxx 1 1 1 1 1 1 1 0OOM^^fcO00COt-t.Oc-Ot-t-CL.c. w w w u w w w inu*iw£inNO»lJii-ifl\B\BNvOi£N'fl nO I I I I I I I I I I I I I I I I I I I I I I I I I I i CM ON 1 — ' T — I rsl o 00 00 I CM CN On <_T) I y Si 41 T en en 4, I I ON CN cm 00 _ bJD M) >, >, 'C >. T^ Si in (N __ n3 n3 r^ CC CM 01 7 1 Ol tN C 3 ^ ►— , CS -4H ^$ . c xL 2L « ^2 A ** I— I <— > ,— I < S a. £ a, 3 1 1 VnO rf ON ] ^ 'C "^ « a, 0. S < < O I T— t ON 01 >, c « as cm *— i CD ON "* 00 CO O ^h NO 00 1— 1 <-h CM * — 1 1 — ' 1 — ' nnin^NNoninrMNrHooonwNr-N T— I < — ■ i— I 00 CM CO LTj LT) ON 00 ON 01 c o U £ b on co J r-. in 8£ ON o ON 00 ^H , — , c ° o 2 1/1 >- 0) o 8 S- go § « -~ LO 01 ,n £ emu o 00 C « h u t; - > 1-1 00 o ON Pi Ul Vl « -; i) 3 a t, « r^ *r* **- .r* t-» (^ £5 c ^ 5 .00 2 VI VO "J] VI Wl 'Ji j^ r1 r1 5 5 .2 ~ ?-, t. rt r; t> ^ .« s £ S £ -7-! K C 5 « %:s j<: '5 -5 ON 00 CM 00 01 _N- o u 1 -s •§ CO r-" ^ g 5 & » 1- j-j Xi o -C '? C x ■" H rv t: 0) X; o ON 00 ON ON 00 P4 B. 8 * P ■= 0 en 01 l-H u 1 — 1 -£ u (1> 11 1-1 01 > u CI 0 U B u - 'S3 N 3 3 'X, >~ K ~ E U <-> ^3 C « c C £ ss _c a» s ?> ■~ § ^2 -s e ^2 .a ~s VlViV)COUNV)VS,^r^f^. tK)t>oCx3(>otoGOb05u Jj .O .O .Q .O .O .O -S .2 -w ' s^-^^^^^^ 3 a c e S K E E 'ij "ij ^J *- ^ ^ « 13 13 SEE S -2 -2 ^ ^J H) 1j <3 Q E E C3 <3 E E (3 cs E E « « « « c: a 13 E E E E E ~~j '~j ~~j ~~j "— "— « « C « E E E E ; T3 >: 13 E E C E ^ "< -< ^ 228 Journal of Hymenoptera Research T3 CU p o u X c Oh ex < tOElfl] ^0 >tJ°A MaN Aipypads ua|[Oj A)i|Bpog aie-Usqiig isa|\[ 3140x3 sa aAi;i'N aouEpunqt? ^|nj pouad iqSiu djpibb^ pouad iqSijj ,\[nj pouad iqSi]j j>jg spmpiAipui j>jg xxxxxxxxxxxxxx xxxxxx xxxxxx X xxxxxxxxxxxxxxxxxxxx X X cuc^cuaH&HCLHOPH(xcxciHCiHeHPHCiHfxcuci.cuciHO(x 1 1 a. c cu cu cu (/it/i^wmt/i^i/it/ii/ii/iaiait/ic/i^aiaii/i^aimfcciHi/i^tfiwi/i (/)c/i(/)^a)c/i^(^^(/)^a)i/)^i/)ui(/)(/it/)(/iLn(/) 1 lUUUUU 2222222222wZZZZZZZZZZwZZZZ22Z DaC^O\O\O\O>O00ifl00HrHl»NO\N00KvO I I I I I I I I I I I I I I I I I I I I I I I I I I I I Ln^ir)^inin^int^^^toiLo^^^^NO'*NocoNOLXNONOLrjNo^^t| I 1 tx lx CN CN ^ >n (0 rS <* IN 00 ™ a, X^Xi. T I I CO LT) ON ^H (U (C (U 2 2 S o o tN 3 1 01 00 C 3 tN CN 1) p I I I ON LH tN TO 03 C 2 2 I LO CN II CO eN 00 00 SI N <" C i f NO 01 m NO >^ § 5P E^ >^'s ta re {^ ta re re 3 Q ^ 00 7 CO __ CO r~l CO ^ I 00 CN ojd NO CO x? ^ 00 NO CN < X A :>, E^ ^. bJD x ^ cu >, 1 H CO c X> 3 3 2P >, < 5, 1 1 On On >-, >^ rc ra 2 2 N^iBOigtOin^OONOOOrO^HaiNtO-f-l'rHOrHHU-iinNvOO to n ■* ifi lonocn colo cn t— 1 rj i— ico CO in CO ^=£ ON 00 '""' CO > in P CO 0 i-H t: 4 01 ■" -^ 2 c* tn NO ON 00 CO <"< ON „> 00 cu §•- t/j rc -tJ 1 — ( Cu « "§.Q cs -a IX o ON IX «-■ CO -* ^ CU ■* i-H C CU O EC cu X 0 C LX ON ON ,_l t — I ^ o & "^ .2 xi > o "3 u o i3 k -s ~ e LT; ON ^ ON -- 00 ~* I — I o in _r -I-' HH ir o 3 cu ei xi CO O iJ Pi LTj ON CO 00 (/) CU J X § ~ -S- 3 s^ -s k a. & _ ~" K K !Sj CU ~ i. V. S S K K <3 -2 -2 CU "S 55 CS cs S rS .Vj _(/! ^-^ nS <-n , tSg CO CJ 1= 13 CS Q cs cs C C K C Cu Cu Cu 2"S"2'^"«t3"C3"C3'T3 cs cs cti Cu -cs 13 Cj Cj £ £ Ju Cu ■? -5 cs cs k e Cu Cu ts "g s s "G CJ cs cs CO m 00 1 — I <^~l u~> X CO 4= X tjO ■? .a c ~ en g cu s: "S. £ 5 "^ sa. NO o CO cu 00 , CO CO Tf CO no ^r '-| CO 0050^ X -= I .- ? cu c £?-* C = cu (JO O O ^ * u 5^ Q 7 Cu Cu "cs -^ c '5 £ ■§ NO 00 0 QJ X cu £ u u c- u u T3 "e iS c CO 01 u cs .2 cs C -S cs *; x: a. ■ 2 § v 05 2 j> t^ ^ .c; .c ^ Si S £ £ fc CO «3 -cs s S 2 cu cu Cu -13 13 S 53 ^^^^^^^^^^^^^^^^^^^N. 0-, a _, p -is c k ~-^^3'< -3 c O) ex < E3BUHI ^D 1J°A MaN A^pypads uo||Oj A)i|t?pog 3;Ei)sqng isajsj 3140x3 sa aAjn'N aauepunqp \f^ pouad }q8tu 3tpJi'a|\j pouad iqgiy ^N pouad )q8iy d>jg S[t>npiAipui jyg XXX xxxxxxx xxxxxxxx xxxxxxx XXXXXXXXX XXXXX X X X X fcfctO^fcDHfcfct.L^t.OCH I I I I I I I I I I I I ai(/i(/i(/i(/)i/)aic/i«^an/)i/i^a)C.fcocDCQ c-c.c_o.c-o,a,&. UUUUUUUUUUUUUUUI 1 I U U 1 1 1 1 1 1 I Z Z Z Z Z 2 w Z w 2 2 2 Z Z Z Z Z Z Z Z ZZZZZZZZ Dti,eiDP-UUUU(JUDUiiHU,Uu.U On i— i On 00 i— ' 00 i— i i— 'On .1 _ .1 I I i I I I I I I I I I I I no -?ix^cnolxsOnc II I I 1 I I I I I I I I I 1 1 1 1 1 1 ill -tlnNomln''*''tcONOcNln^LnNOT^LnNOcocO'3, ■3,cO"*inLn^t-r--t- o 00 llllllllllllllllllll •tlnlnlnln•fi|^o^o^Oln^u,)^o^o^Ol£fO^'l'!), in ro -t in in I I I I I I I NO rN >, 3 -* 00 CN i— 1 >-. re X1 c 00 -* re c fN ON CN CN 01 CN CN NO 0) en 1 NO CX) I o> en >, 01 . re C 2, s 1 ,-H r-H ^ «N CO M " " r> a &, o, a, < < < en 3, >, >, Kg !>N NO (N 1 i— ' i-< CN IX bJD < qj 1 en 1 Oh CL, 0) CU en en 1 1 00 00 re ■* cn N ,i I-1 C/1 H rn t- 3 0> 3 ►X en ' o oJD J*£ ^ ^ ^ 01 re re c 22 = a, ** CN ^ a, if I IX ON CN re re ° ~ ~. K*^l C re re re cv 2 S 2 < on in in ■* m in in + , rinHiHHKoornmNnN no cn no no co 1—1 co m 00 - co ~ ■* 00 i- • c i~* v X. v c 1— 1 T— 1 4-< 0 (—1 ■a in c 7 0) >-, 0 in 01 in J S-4 O 2 in 0) u "5 0) -^ O 01 N in 0 u g IX o 0 o u CO o ON tX 00 CO CO in 00 =3 00 o> > o O M N in O T! re _c 01 £ -o. in 01 l^U 00 ^ ix c2 o . in ^ SP CN -* N S ^ 00 °2 00 —* o Ex 2 S ^ 00 ~- = -13 re 'TN m en rS in 01 p- *~ 5oU C . c IX in ro in 2 £ re ~— en 3 « ? QJ ^ 5c .2 .« .c .c .c .2 F £ S £ £ £ S R m m « in m in o o o o c o -^ K K K ~ c s: c -S -S -S -2 -2 -S £ cij ^j iu cu ^j *ij rr .- QJ a H X .0 .! » I £ in re " 01 .£ o ^ -s ■ -: « So -2 > o ^c Jkj to -r- "»- t; "J "«j "J g ^ 5 5 5 5 u u Ji_ ^l S S -2 -2 -r- i^ ^ ^ ^ i: ,S-> 5j ttj 4) o c X U U U 2 2 « 1= 2 2 c « a c: -r— -r- 1^- -r- 2 2^2 3 O 3 Q 2 2^2 230 Journal of Hymenoptera Research 73 QJ 3 C ■-3 C O u T3 C QJ a, EDeiHI b]D >F°A M3N /tipypads uanoj A)i|t'pog ojpjisqng jsa|\[ JI}OX3 SA 3Api?N aauepunqi' \f^ pouad iqSju aipjea[\[ pouad qqSiu ^N pouad H|%1 jyg sienpiAipui j>ig XXXXXXXXX X X X X X X X XX X X X X X XXX I I I I I I I XXXXXXXX X X OOp-iiiiiiii I ^DHCHCHt^fcttn OlOK/lfcWWWHJUBW UJ <-n^\XXXXXXX X ZZZZZ2ZZZ ZZZZZZZZZZZ w UUU^DftUB.B- tui-uUUUUcm't^|^<-t-t,-tirj m CN CN tN CN QJ (N June ly 26 0) QJ CN 2 41 2 3 2 c 3 2 QJ 2 3 3 3 1 3 t 3 H— * ■* ^ V ~- ** •*r 1 CN 1 00 ,— 1 1 00 CN i — i *T On m ON 1 — 1 in r-i ON ^ >^ 'E >, >^ 'C S>~> >N 'jh 'C >> as ns c^ to ns Cl ns as Cu O. ns ~- ~ < 2 S < s s < < s CN og ^O ,— t ^D 1 1 ON NO T 1 T— 1 CN NO CN no rH (N - SP- ■+- > T-H £ ept ■Oct 3 u 3 u o < o 1 1 QJ CO 'J o QJ CO g co ^ NO -"F -t> 1 1 1 I 00 1 I— 1 CN 1— 1 o ON <* i 00 '"' 00 in 1 — 1 cn CN cn >, >• >1 >, 'C 'C 'C QJ ■M QJ QJ X" « ns £X D. £X c O- c C £S 3 s < < < 3 QJ co 3 Z^ t <~nI 00 ^-* ^ CO NO o ■tmHNi»mtN(NiooNH\oinHotNiOi— < i— ' >— i 00 i— i QJ X (J s 3i m m ON 00 >— < CO v NO C 00 28 2 ° co <0 NO ig 00 °° cn c -9. o o Qj « >-, HJ c o in QJ u 0 U -S -is ■8^ ° r -2 ^J r- K B S « "a ts ts e « c S £ £ _p o o « 5 o a. C o ID u u g a s s "a. ? ^ ^i a. a. 00 o ON CO '-' ON „- 00 —> ^ Si c _£ O u « O QJ XI « O b «5 X ■r xj s ^ o ^ qj t: -y § "?^ 3 3 ^ S O O bO bb S 00 -; ^ ^ NO uT °° QJ " QJ bC QJ CD 03 QJ u QJ S 2 QJ ^ u m m h NO VO ON 00 00 <— ' ■a u C 2 n ""■ O 3 00 00 e CD u u bo « vi M ■~ "5 S O ns ."3 CD | a CD 00 in ■=- QJ R3 3 .5 ^ f-o oi in y-j m v-j r-n a ^ "^ q s -a -a K p a s C C ^j ?M P o ~ -a « P cr, Ci, a g O -a o o -a o s~, >, >, ^> D, G, Dh CX, ^ ^ -- Oh SX^^^^^^S^1^ g g S .« .» s ^ S II 5 5 ^ eg a. 51 -a-a-a-a-a-a-a-a aooooaoo^-. CQCQCQCQCQCQCQCQ^ Volume 15, Number 2, 2006 231 Appendix 2. List of wasp species collected incidentally at F3RF in 2003. Family Subfamily Specie! Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Crabronidae Sphecidae Sphecidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Vespidae Scoliidae Pompilidae Astatinae Bembecinae Bembecinae Crabroninae Crabroninae Crabroninae Crabroninae Crabroninae Crabroninae Crabroninae Crabroninae Crabroninae Crabroninae Crabroninae Crabroninae Pemphredoninae Pemphredoninae Pemphredoninae Pemphredoninae Pemphredoninae Philanthinae Philanthinae Philanthinae Philanthinae Sphecinae Sphecinae Eumeninae Eumeninae Eumeninae Eumeninae Eumeninae Eumeninae Eumeninae Eumeninae Eumeninae Eumeninae Eumeninae Eumeninae Eumeninae Polistinae Polistinae Vespinae Vespinae Vespinae Vespinae Vespinae Vespinae Vespinae Scoliinae Ceropalinae Astata leuthstromi Ashmead, 1897 Biq/rtes quadrifasciata (Say, 1824) Gorytes deceptor Krombein, 1958 Ectemnius (Clytochrysus) lapidarius (Panzer, 1804) Ectemnius (Ectemnius) atriceps (Cresson, 1865) Ectemnius (Ectemnius) borealis (Zetterstedt, 1838) Ectemnius (Ectemnius) dives (Lepeletier & Brulle, 1834) Ectemnius (Hypocrabro) continuus (Fabricius, 1804) Ectemnius (Hypocrabro) decemmaculatus (Say, 1823) Ectemnius (Hypocrabro) stirpicola (Packard, 1866) Liris (Leptolarra) argentata (Beauvois, 1811) Lyroda subita (Say, 1837) Tn/poxylon (Trypargilum) lactitarse Saussure, 1867 Trypoxylon (Trypoxylon) frigidum Smith, 1856 Tnjpoxylon (Tn/poxylon) pennsylvanicum Saussure, 1867 Pemphredon (Cemonus) inornata Say, 1824 Pemphredon (Cemonus) rugifera Dahlbom Mimumesa nigra (Packard, 1867) Psen erythropoda Rohwer, 1910 Pseneo shnplicicornis (Fox, 1898) Cerceris atramontensis Banks, 1913 Cerceris fumipennis Say, 1837 Cerceris halone Banks, 1912 Philanthus gibbosus (Fabricius, 1775) Isodontia (Isodontia) philadelphica (Lepeletier, 1845) Isodontia (Murrayella) mexicana (Saussure, 1867) Parancistrocerus pedestris (Saussure, 1855) Parancistrocerus pensylvanicus (Saussure, 1855) Parancistrocerus perennis (Saussure, 1857) Euodynerus foraminatus (Saussure, 1853) Euodynerus hidalgo (Saussure, 1857) Euodynerus leucomelas (Saussure, 1855) Ancistrocerus adiabatus (Saussure, 1852) Ancistrocents antilope (Panzer, 1798) Ancistrocerus campestris (Saussure, 1852) Ancistrocerus waldenii (Viereck, 1906) Symmorphus (Symmorphus) canadensis (Saussure, 1855) Eumenes (Eurnenes) fraternus Say, 1824 Zethus (Zethus) spinipcs Say, 1837 Polistes dominulus (Christ, 1791) Polistes fuscatus (Fabricius, 1793) Dolichovespula arenaria (Fabricius, 1775) Dolichovespula maculata (Linnaeus, 1758) Vespula consobrina (Saussure, 1864) Vespula flavopilosa Jacobson, 1978 Vespula germanica (Fabricius, 1793) Vespula maculifrons (Buysson, 1905) Vespula vidua (Saussure, 1854) Scolia (Discolia) bicincta Fabricius, 1775 Ceropales maculata (Fabricius, 1775) J. HYM. RES. Vol. 15(2), 2006, pp. 232-250 The Biology and Morphology of Entedon sylvestris (Hymenoptera: Eulophidae), a Larval Endoparasitoid of Ceutorhynchus sisymbrii (Coleoptera: Curculionidae) Alex V. Gumovsky Schmalhausen Institute of Zoology, 15 Bogdan Khmelnitsky St., 01601 Kiev-30, Ukraine; email: gumovsky@izan.kiev.ua Abstract. — The biology and morphology of preimaginal stages of Entedon sylvestris Szelenyi (Hymenoptera: Eulophidae), are described in detail for the first time. Entedon sylvestris is a larval endoparasitoid of the seed-feeding larvae of the weevil Ceutorhynchus sisymbrii Dieckmann on the small tumbleweed mustard, Sisymbrium loeselii L. (Brassicaceae). In the Ukraine, females of £. sylvestris begin ovipositing in late May and continue to lay eggs until the beginning of July. Females of E. sylvestris parasitize weevil larvae of various instars. The parasitoid larva remains within the body of the host weevil larva until the emergence of the latter from the dried host-plant pods. The morphology of each of the three larval instars is described in detail. The moult of the parasitoid larva into the final instar, as well as pupation, takes place underground. Adults of E. sylvestris must therefore penetrate a soil layer to emerge the following spring. Key words. — Entedoninae, larval endoparasitoids, parasitoid-host relationships, preimaginal morphology, Sisymbrium loeselii Parasitic wasps of the genus Entedon Dalman (Eulophidae, Entedoninae) are endoparasitoids of immature stages of beetles. Curculionidae (including Scolyti- nae), Brentidae (including Apioninae), An- obiidae, Chrysomelidae (including Bruchi- nae), Buprestidae, Cerambycidae, Mordel- lidae, and Nitidulidae are recorded as hosts (Boucek and Askew 1968, Graham 1971, Askew and Kopelke 1989, Rasplus 1991). Entedon ergias Walker has been imported from Europe into North America for the biological control of the smaller European elm bark beetle, Scolytus multi- striatus (Marsham) (Peck 1963). For some species, parasitism rates and/or general descriptions of the larval morphology are given (Ferriere 1939, Erdos 1944, Abedin and Quayum 1972, Tiwari 1976). However, these descriptions lack many morphologi- cal details, especially for the first instar larvae. Beaver (1966) and Fisher (1970) gave the most complete bioassays and reported egg-larval parasitism for Entedon ergias Walker, E. rumicis Graham, E. phar- nns Walker. Askew (1991) and Gumovsky (1997) provided some information on per- centage parasitism, the biology of the final instar larva and pupation procedure of E. cioni Thomson, E. cionobius Thomson and E. zanara Walker. In general, despite some thorough re- views (Parker 1924, Parker and Thomson 1925) and occasional detailed descriptions (e.g., Darling 1992, 1995) of the larvae of Chalcidoidea, our knowledge of morphol- ogy of preimaginal stages of chalcid wasps is incomplete. Most discussions on larval morphology and biology concern ectopar- asitoids, whereas endoparasitoid larvae traditionally attract less attention, mainly due to the difficulties with their prepara- tion and identification. The larvae of Eulophidae were classified by Parker (1924) in group II (the ectoparasitoid forms) and V (egg endoparasitoids), dif- fering mostly in having spiracles (group II) or being apneustic (group V). Later, when Volume 15, Number 2, 2006 233 discussing the morphological peculiarities ical control agents. It is therefore of vital of the first instar larvae of Anastatus sp. importance to determine the parasitoid- (Eupelmidae) and Miscogaster sp. (Ptero- host associations of Ceutorhynchinae and malidae), Parker and Thomson (1925) host plant preferences in various geo- stated that some endoparasitoid larvae graphic regions. represent a transitional type between the This paper reports results obtained from groups corresponding to groups V and VI and the methodological approaches used sensu Parker (1924). during the study of the parasitoid-host The larval morphology of Entedon spe- relationships between Entedon sylvestris cies is even more vague. The papers of Szelenyi and its host, the weevil Ceuto- Beaver (1966) and Fisher (1970) are the only rhynchus sysimbrii Dieckmann on Sisym- sources of comprehensive descriptions of brium loeselii, which is the first host record preimaginal stages, but these concern for the parasitoid. mainly the size, body proportions and number of spiracles, and are illustrated MATERIALS AND METHODS mostly by diagrammatic figures. Many of Adults of Entedon sylvestris were collect- the minute morphological structures (e.g. ed in the field on plants of Sisymbrium sensorial organs) remain obscure and un- loeselii during late June and early July in described. 1995, 1997, 2001, 2002, 2004 in Kiev Weevils of the subfamily Ceutorhynchi- (50 28'N; 30°32'E) and Kherson oblast of nae have a wide range of host plants, but Ukraine (v. Lazurnoe) (46 04'N; 3229'E). many are restricted to Brassicaceae. Some Collections were made by both sweeping Ceutorhynchinae species have gained spe- and capture of individual adults in tubes, cial attention as pests of economically Females were kept in Petri dishes and fed important plants (e.g., Ceutorhynchus napi with diluted honey. Gyllenhal, C. pallidactylus (Marsham) and Mature pods of S. loeselii infested by the C. obstrictus (Marsham) damaging cabbage larvae of Ceutorhynchus sysimbrii (living and oilseed rape in North America). Other inside the pods and feeding on seeds) were species (e.g. C. merkli Korotyaev, C. cardar- exposed to females of E. sylvestris kept in iae Korotyaev, C. alliariae Brisout, C. roberti various reservoirs (Petri dishes, sealed Gyllenhal) are under investigation as po- plastic bags or boxes). Infestation by the tential biological control agents against weevil was indicated by the presence of some introduced weeds (i.e. whitetops a distinct hole in the pod made by the Cardaria spp., garlic mustard Alliaria petio- female's rostrum before oviposition. laria: Hinz et al. 2004, Hinz and Gerber, The sites where the parasitoid's oviposi- 1998). The small tumbleweed mustard, tion took place were marked with black Sisymbrium loeselii L., is a plant of European ink. To study the morphology of the differ- origin that was accidentally introduced into ent larval instars of E. sylvestris, weevil the New World, and is now recorded in 31 larvae were removed from the pods at states of the USA, and is regarded as an regular intervals and dissected. Endopar- invasive weed (Stubbendieck et al. 1994). asitoid larvae found were fixed in Bouin's Some parasitoid species (e.g. Tersilochus fixing solution (15 cm3 picric acid (saturat- spp., Microctonus spp.) have been released ed), 5 cm3 formaldehyde solution, 1 cm3 in North America to control Ceutorhynchus acetic acid) to keep their original shape, spp. (e.g. C. obstrictus) that are pests of and further washed out in 96-98% economic plants (e.g. Brassica spp.). How- ethanol. ever, parasitoids — unless sufficiently host- Mature weevil larvae, leaving the pods specific — could hamper the effectiveness of of their host plant, were put into plastic the Ceutorhynchus spp. released as biolog- tubes (50 X 16 mm) filled 3/4^1/5 with Journal of Hymenoptera Research soil taken from field sites of S. loeselii. The behavior of the mature weevil larvae was observed through the transparent walls of the tubes. In September/October, when the pupae of E. sylvestris were expected to be completely formed, all tubes were carefully searched for earthen cells or dead weevil larvae. Dead weevil larvae found, were put into 5-10% solution of lactic acid to regain their original shape and soften. The soft- ened larvae were then dissected in order to find the parasitoid larvae, in particular, the final instar larvae. These were transferred into lactic acid solution of higher concen- tration (40-70%) to obtain maximum swell- ing. The parasitoid larvae were then put into Bouin's fixing solution in order to fix the regained shape and then washed out in 96-98% ethanol. All fixed larvae were kept in 100% ethanol for one day and then in 100% molecular sieved ethanol for maximal de- hydration. After absolute ethanol the spe- cimens were Critical Point Dried. The minute parasitoid larvae were put into pipette tips of various diameters sealed on the sides with cotton wool plugs, to avoid their loss during drying. The dried para- sitoid larvae were transferred to SEM stubs on metallic pins, using static electrical charges to avoid damaging their extremely soft integuments. Finally the specimens were coated with gold and observed using a Scanning Electronic Microscope LEO 1530VP in the Max-Planck Institute for Metal Research, Stuttgart (MPI). Field and laboratory video recordings were made using either 8 mm VP-A800 Pal Samsung Video Camera or with digital imaging of Leica IC A Videomodule in- tegrated in the Leica MZ 125 stereomicro- scope, using the video grabbing option conducted in Adobe Photoshop 6.0 pro- gramme through the use of the Falcon \ Ea- gle Frame Grabber. The alignment of the photos corresponding to different layers in focus was conducted using the Combine Z programme Version 3.9 (designed by Alan Hadley, http://www. hadleyweb.pwp. blueyonder.co.uk/CombineZ/CombineZ3. zip). RESULTS Taxonomy Entedon sylvestris Szelenyi, 1981 Entedon sylvestris Szelenyi, 1981: 277; Entedon sylvestris Szelenyi; Askew, 1992: 119; Entedon (Entedon) sylvestris Szelenyi; Gumovsky, 1999: 142; Entedon sylvestris; Gumovsky, Boyadz- hiev, 2003: 23. Material examined.— Types: Holotype female, paratypes 10 females, 6 males, Hungary, Horto- bagy National Park (Szelenyi) (Hungarian Mu- seum of Natural History, Budapest); more than 2,000 specimens from Kiev (June 1995, 1997, 2003, 2002, 2004) and v. Lazurnoe (Kherson oblast), Ukraine, collected on plant shoots and fruits of S. loeselii; 7 females, 4 males, Kiev vicinity, Velyka Oleksandrivka, swept from Berteroa incana and Capsella bursa-pastoris, 31.V.2004, (Schmalhausen Institute of Zoology, Kiev); 54 females, 1 male, Kyiv, Trukhaniv Island, "Park Druzhby Narodiv", 24.VI.1995, swept from Sisymbrium loeselii (Natural History Museum, London); 1 female, 1 male, Giefien, ex Capsella seeds V.1995 (Weiffenbach) (Zoolo- gische Staatssammlung Munchen). Recent literature. — Entedon sylvestris has gained little attention since its description (Szelenyi, 1981). This species belongs to the assemblage of the cyanellus and costalis species groups in having the anterior margin of the clypeus produced (Figs 1A, C, E, cly; 2C). Askew (1992) reported it for Great Britain, presented a short corrected di- agnosis and reported the possession of traces of pale strips on the fore tibiae. He also proposed to locate this species in the cyanellus species group of Entedon based on the presence of these hardly discern- ible foretibial strips and the 3-segmented funicle of the male. Gumovsky (1999), when revising the cyanellus group of the genus, argued against Askew's placement of E. sylvestris based on that only about 20% of specimens in a population possess Volume 15, Number 2, 2006 235 EHT * 5.00 kV Deleclor = SE1 Ma9 ■ 3 59 K X 2pm WO = 16 mm \—\ Fig. 1. Entedon sylvestris, female, details of morphology: A, C, head in frontal view; B, antenna; D, pedicel and 1st funicular segment; E, lower face; F, anelli; cly, clypeus. the dim foretibial stripes, whereas the majority of the specimens have fore tibiae largely darkened, and proposed instead to accommodate E. sylvestris in the costalis species group (decreasing the value of the 3-segmented funicle of male as a species-group character). Gumovsky and Boyadzhiev (2003) reported this species for Bulgaria and also provided its comparative diagnoses in their key to species. Comparative notes. — Within the European fauna, E. sylvestris can easily be confused with £. cyanellus and E. fufius. From the former species, E. sylvestris is easily distin- guishable as female has a shorter 1st 236 Journal of Hymenoptera Research Fig. 2. Entedon sylvestris: A, B, female: A, posterior mesosoma; B, metasoma; C-F, male: C, head in frontal view; D, antenna; E, posterior mesosoma and anterior metasoma (petiole); F, body. funicular segment (1.5-1.6 times as long as broad, Figs IB, D, whereas in E. cyanellus — 1.8-5.0 times) and by the narrower anten- nal scape of males (3.8 times as long as broad, Fig. 2D, about 3.0 times in E. cyanellus). Also, the foretibial stripes (when present) are much narrower and not terminated by a pale area on the distal end of the tibia (the tibia is always with distal pale band in E. cyanellus). The distal pale bands on the tibiae also distinguish £. sylvestris from £. fufius in which all tibiae are completely dark. Also, the 1st funicular segment of females is about 2.2 times as long as broad and 1.4 times as long as the 2nd segment in £. fufius, whereas in £. Volume 15, Number 2, 2006 237 sylvestris the segment is 1.5-1.6 times as long as broad and just slightly (about 1.1 times) longer than the 2nd (Figs IB, D). Moreover, all flagellar segments are free in males of E. fufius (the funicle is 4-segment- ed), whereas in E. sylvestris the funicle is 3- segmented and two last funicular segments are closely attached (forming the 2-seg- mented clava, Fig. 2D). Distribution. — Hungary (Szelenyi 1981), Britain (Askew 1992), Bulgaria (Gumovsky and Boyadzhiev 2003), Ukraine (Gu- movsky 1999). Ecology The host, Ceutorhynchus sisymbrii Dieck- mann. — Ceutorhynchus sisymbrii is similar to C. pulvinatus Gyllenhal, 1837 in having dense coverage of dorsum and black rostrum, but differs from that species in that only the tibiae are red, not the entire legs (L. Behne, pers. comm.). Des- pite C. sisymbrii (Dieckmann 1966) being described nearly fifty years ago, its bio- logy remained unknown. Dieckmann (1972) mentioned that studies on the life cycle of this species are of special value. Below I propose a brief synopsis of the field and laboratory observations on this species. Adults of Ceutorhynchus sisymbrii (Figs 3A, B) feed on shoots, flowers and fresh seeds of Sisymbrium loeselii. Oviposi- tion takes place in May-July. The female of the weevil lays its eggs into the pods of the host plant when the seeds are fully grown, but still soft and green (mature seeds are yellow to orange and hard). At first, the female makes an opening in the pod with her rostrum. It deeply penetrates the rostrum into the pod (up to the base of the rostrum) and eats the seed below the hole. Then the female turns back and presses its caudal end into the prepared opening, and begins oviposition. This behaviour is discernable by the rhythmic pulsing of her gaster. Occasionally, the female fails to aim into the prepared hole and lays an egg directly onto the surface of the plant pod. She then eats the egg, as in other weevil species (Kozlowski 2003). The larvae feed on the seeds and remain in pods until they are fully grown (Figs 3C, D, E; Figs HA, B). The pods then split and the mature larvae fall to the ground (Fig. 11D). Larvae leave the host plant's pods at the end of June - beginning of July. They quickly bury themselves into the soil, where they prepare an earthen cell, in which they pupate (Fig. HE). I have found adult beetles in the dissected earthen cells in October. However, I assume that adult weevils are already present before, but leave their soil/earthen cocoons only the following spring. Host searching and oviposition of the parasitoid, E. sylvestris. — The females of E. sylvestris can be found in the field from late May until early July (Figs 4A-G). The parasitoid female searches along the pods of S. loeselii for weevil larvae, by drumming the pod surface with her antennae (Fig. 4A). Once she has located a host, she walks back and forth several times before starting to oviposit. She bends her gaster downwards and briefly hooks the ovipos- itor saw into the plant tissues. She then releases the gaster so that it strengthens in a position perpendicular to the ovipositor (Figs 4B, C). Thereafter she penetrates the pod wall with her ovipositor, and carries out rhythmic, twisting movements to find the weevil larva, as in other parasitoid species. Oviposition lasts about 30-90 seconds. Quite often the ovipositor pene- tration causes the host's haemolymph to exude, and then females of E. sylvestris feed on these excretions. Immature stages of E. sylvestris Egg. — Elongate, white to transparent, without discernible sculpture. No stalks or terminal bulbous projections, reported by Beaver (1966) for £. ergias, were found. Size, about 260-300 urn long and about 80- 100 urn wide. 238 Journal of Hymenoptera Research Fig. 3. A-E, Ceutorhynchus sisymbrii. A, B, female, C-E, mature larva. 1st instar larva. — Habitus. The first instar different forms, which were found during larva of E. sylvestris is hymenopteriform, dissection of the host larvae. One is "slim" pale (nearly transparent), has 13 body (Figs 5A, 6A), about 300 jim long and segments and a cranium. There are some about 104 |im wide (max.), and another Volume 15, Number 2, 2006 239 Fig. 4. A-G, the female of Entedon sylvestris oviposits into the larva of Ceutorhynchus sisymbrii in pods of Sisymbrium loeselii: A-C, in laboratory; D-G, in the field; hole, oviposition hole of weevil female. Scale bars: A-C, 2 mm; D-G, 5 mm. form is "swollen" (Figs 6E, F), about 450 |im long and about 206 urn wide. Occasionally, specimens of an intermediate form were found (Fig. 6G), with a length of about 260 urn and a maximum width of about 108 fim. Despite the differences in size, I regard all three forms as belonging to the same larval instar, because of the possession of the peculiar shape of the head capsule (cranium) and body surface. The different body proportions are proba- bly caused by different nutritional condi- tions, and fixation and drying circum- stances. Body segments. The last, XHIth, segment bears sharp triangular tubercles along its margin, arranged in two or three rows in the shape of a crown (Figs 6C, D). These tubercles are very distinct in freshly emerged larvae, but occasionally are coat- ed with secretions (Fig. 6H). These rows of tubercles are also distinct in the "swollen" larvae (Fig. 6H), however, they are not so clearly distinguishable in these larvae due 240 Journal of Hymenoptera Research Fig. 5. £. sylvestris, first-instar larva: A, habitus; B, E, head close-up; C, D, cauda close-up. to the thickness of the preceding segment which consist of small curved teeth. Seg- (Figs 6E, F, H). Segments IV-XII bear ments I— III bear no distinguishable teeth or distinct dorsal semicircular serrations serration, which reflects the subdivision (Fig. 6C, se) along their anterior margins, into thoracic (I — III) and abdominal (IV- Volume 15, Number 2, 2006 241 Fig. 6. E. sylvestris, first-instar larva: A, E-G, habitus; B, head; C, D, H, cauda; se, serration. XIII) segments. No spiracles were found on trally, with a characteristic "beak"-shaped the body of the larvae. end, which is formed by the protruding Head capsule (cranium). The head cap- palpi of the labrum. In light microscopy, sule is weakly sclerotized, narrowing ven- the labrum adopts an active backward and 242 Journal of Hymenoptera Research Photo No. = 74 Date :3 Aug 2004 EHT= 5.00 kV Detector = SE1 Mag= 3.53 KX 2um WD= 20 mm H H MPI Metallforschung Metallographie ant Photo No. = 78 Date :3 Aug 2004 EHT= 5.00 kV Detector = SE1 Mag= 2.56 KX 10(jm WD= 20 mm H — MPI Metallforschung Metallographie Fig. 7. E. sylvestris, first-instar larva, details of the head morphology (see text for abbreviations). forward motion. Antennae are absent, in- developed and arranged in a relatively dicated only as small swellings on the fixed position (Figs 7, 8). The lateral area of upper part of the head capsule. The the cranium bears three pairs of cranial sensorial structures of the head are well- palpi: the upper (cpl), the lower (cp2) and Volume 15, Number 2, 2006 243 Fig. 8. £. sylvestris, first-instar larva, details of the head morphology (see text for abbreviations). the posterior (cp3) palpi. There is a pair of enlarged pleurostomal palpi (pip) just above the labrum. The labrum bears 3 pairs of labral palpi (grouped by one from each side): the upper lateral labral palpi (ulrp), the lower lateral labral palpi (llrp) and the inner labral palpi (ilrp). A large palpus is located near each maxilla (mxp) 244 Journal of Hymenoptera Research o No. = 113 EHT = 5.00 kV Mag = 332 X JOOprr .5 Aug 2004 Detector = SE1 WD- 18 1 MPI Metalirorschung Metallographie Mag = 496 X 20um WD = 18 mm MPI Metallforschung PholoNo. = 119 EHT= 5.00W Mag = 293X 100pm Date :5 Aug 2004 Detector = SE1 WD ■ 18 mm I — — Fig. 9. £. sylvestris, second-instar larva, details of morphology: A, E, habitus; B, head in frontal view; C, head close-up; D, mouth area, close-up; F, mouth area, lateral view; pip, pleurostomal palpi; lb, labrum, md, mandible. and behind them there is a pair of smaller labial palpi (lbp). A pair of comparatively- long pharyngeal setae (phs) is situated behind the latter palpi. 2nd instar larva. — The second instar larva of E. si/lvestris is also hymenopteriform and pale, with similar body segmentation (13 body segments and the head capsule), 500- 590 (im long. However, the proportions and shape of the segments are different: the larva is more robust (Figs 9A, E) and none of the segments has serrations. The last segment bears no "caudal crown". The head is more spherical and the sensorial structures of the head are less clear (Figs 9C, D). However, it is still possible Volume 15, Number 2, 2006 245 to discern the pleurostomal palpi, the delimited labrum and short mandibles (about 20 jam long). No spiracles are recognizable. Final instar larva. — Dissections of the dead host larvae revealed two forms of the parasitoid larvae, which differ from each other by their size and body shape. The smaller form fits the size range and morphological peculiarities of the second instar larva, described above. The larger form found within the body of the dead, buried host larvae, were 0.96-1.02 mm long (Fig. 10A), which is nearly twice as long as the second instar, but still nearly half of the average length of the pupa (1.8 mm). It has large (46 urn long) and heavily sclerotized (dark brown in color) mandibles and discernible hypostome (Figs 10D, F). I regard these peculiarities to be specific to the final instar. The sensory structures of the head are hardly distinguishable, apart from large antennae situated on broad swellings (Figs 10B, C). Spiracles were difficult to distinguish because the skin of the larva was too shriveled. The large antennae are peculiar to the final instar larvae of Chalcidoidea (Parker 1924). Fisher (1970) also draws the large round areas (equivalent to the large antennae) on the upper part of the head of the final instar larva of E. rumicis, but does not mention them in the text. The posses- sion of large antennae is one of the characters supporting the assumption that this is the final instar larva of E. sylvestris. The smaller body size of the larva is pro- bably artificial and resulted from incom- plete swelling of the dried, dead larvae. Pupa. — Generally, pupation takes place in the host's earthen cell (Fig. 11F). Some- times, pupae could be found directly in soil samples, which suggest that parasitoid pupation can also take place without the successful creation of an earthen cell by the host. In the laboratory, the host's earthen cells and the "free" pupae were found 2- 3 centimeters from the upper soil level (the overall height of the soil level in tubes was about 3 cm), and the host larvae can probably pupate even deeper. The pupa is black, obtect, with distinct outlines of head, mesosoma, metasoma, wings, legs and antennae. The average length is 1.8-1.9 mm, the width of the head is about 0.6-0.7 mm, of mesosoma - 0.8 mm, of metasoma - 0.9 mm. The last larval skin, covered by soil particles, is often attached to the caudal end of the pupa. Parasitoid-host relations. — Eggs of E. sylvestris were found singly in the host larvae. When the host larvae were dis- sected, the parasitoid eggs and first instar larvae were found free-floating anywhere in the host body cavity. No attempts to parasitize the host eggs were recorded. Parasitoid larvae hatch about one day after being laid. Females of E. sylvestris oviposited into weevil larvae of various instars, but only second instar larvae of the parasitoid were found within the mature host larvae. Only one parasitoid larva per host larva was found in all studied samples. Parasitoid larvae could some- times be observed while inside the host. The larvae of £. sylvestris never emerged from the host's body within the pods of S. loeselli. Behaviour and pupation of parasit- ized host weevil larvae did not differ from unparasitized. DISCUSSION Life history of Entedon sylvestris Females of Entedon sylvestris attack their hosts, the weevils Ceutorhynchus sisymbrii, at their larval stage. The larvae of C. sisymbrii feed on the seeds of the small tumbleweed mustard, Sisymbrium loeselii L., and leave the host plant's pod when mature. Unlike most other parasitoids of seed-feeding weevils, which finish their ontogenesis and kill the hosts inside the host plant pods (e.g. Trichomalus spp., Mesopolobus spp., Necremnus spp.), the larva of E. sylvestris is in its second instar when the the host larva leaves the pod. The 246 Journal of Hymenoptera Research Fig. 10. A-E. E. sylvestris, final instar larva (isolated from the dead host larva): A, habitus, B, head, lateral view; C, antenna enlarged; D, head, frontal view; E, mandibles; ant, antenna; md, mandible. Volume 15, Number 2, 2006 247 Fig. 11. A, B, the mature larva of Ceutorhynchus sisymbrii; C, first-instar larva of E. sylvestris isolated from the host's body; D, the dead weevil larvae isolated from the soil sample and treated by the lactic acid solution; E, the mature larva of C. sisymbrii creates an earthen cell (the larva is visible inside); F, the pupa of £. sylvestris in the host's earthen cell; par, parasitoid larva. Scale bars: A, 3 mm; B, 0.3 mm; C, 0.3 mm; D, E, F, 1 mm. 248 Journal of Hymenoptera Research moult of the parasitoid larva into the final The anterior margin of the clypeus is less instar and final consumption of the host produced in these species than in E. take place within the host's body, un- sylvestris and E. cyanellus. Further observa- derground. Pupation takes place outside tions on the emergence procedure may of the remnants of the host's body, also reveal other behavioral peculiarities and underground, at a depth of at least of 2- functional devices for the species of En- 3 centimeters. The underground pupation tedon having an "underground span" in means that adult parasitoids must pene- their life cycles. trate at least 2 cm of soil to emerge the , „ Larval morphology following spring. r o; Underground pupation is rather rare Beaver (1966) mentioned that the first- within Chalcidoidea, perhaps due to the instar larva of E. ergias has 12 segments, but small body size of these insects and the last segment bears a dorsal plate with corresponding problems with emergence a series of sharp spines on its edge. Fisher from soil. Apart from Entedon sylvestris, (1970), when describing the first instar the pupation and consequent emergence larva of E. rumicis, mentioned that it has from the earthen cells of its host, is the head capsule and 13 body segments, recorded for E. cyanellus Dalman, a parasit- with "the last abdominal segment ventrally oid of Tichius quinquepunctatus (L.) (Gu- with a sclerotized plate with irregularly movsky 1997). However, no detailed bi- spinous edges". The same author also ological information is available for E. mentioned that the first-instar larvae of E. cyanellus, and records of its biology refer pharnus and E. philiscus have a similar body just to labels of museum specimens. So, E. shape. Since the first-instar larvae of other sylvestris is the only chalcid species of species have 13 body segments, Beaver's known biology, which begins its ontogen- statement of the possession of only 12 body esis above the ground and ends it un- segments by the first instar larva of E. derground. ergias is likely erroneous. Both known species for which unde- The morphological peculiarities of the ground pupation is known or suspected (E. first-instar larva of E. sylvestris include the sylvestris, E. cyanellus), have the produced notable indentation on the XIII (last) body anterior margin of the clypeus (Fig. 1, cly). segment (Figs 6C, D, 8C, D). This indenta- The produced clypeus is diagnostic for the tion is likely equivalent to the plates with cyanellus and costalis groups of Entedon spinous edges reported for the first-instar (Gumovsky 1997), and occasionally present larva of E. ergias (Beaver 1966) and E. in species of some other groups (E. kerteszi rumicis (Fisher 1970), and also may be Erdos, E. occidentalis Girault, E. diabolus assumed for the larvae of E. pharnus and Rasplus). The function of the shape of the E. philiscus (mentioned as similar to the clypeus is unknown. The adult chalcids larva of E. rumicis by Fisher 1970). The feed by sucking liquids through the use of tubercles of the indentation (the "caudal their labio-maxillary complex. Neither the crown") is likely homologous to the small clypeus nor the mandibles are actively curved teeth of the dorsal semicircular involved in this process. One of the serrations (Fig. 6C) along the anterior possible functions of the produced clypeus margins of the segments IV-XII. The may be associated with facilitating emer- function of the "caudal crown" for these gence from the ground. Entedon rumicis, E. larvae remains obscure, but there is no pharnus and E. philiscus Walker also have evidence of its usage for movement within the anterior margin of the clypeus pro- the body cavity, since the parasitoid larvae duced, but their adults emerge from the observed within the host's body are gen- host plant's stems, not soil (Fisher 1970). erally passive. Volume 15, Number 2, 2006 249 The first-instar larva of E. sylvestris is, to some extent, similar to the larva of Mis- cogaster sp. described by Parker and Thomson (1925). Both larvae are apneustic, have the cuticular spines arranged in encircling lines (striations) on the body segments (IV-XII) and peculiar indentation of the last body segment (a semi-circular or bilobed "caudal crown"). However, the tubercles of the "caudal crown" are sub- equal in length in the larva of E. sylvestris (two inner spines are stouter than the surrounding spines of the "caudal crown" in Miscogaster sp.) and the sensorial struc- tures of the head have a somewhat differ- ent arrangement. Although data is available for only a few species, the "caudal crown" is likely in- herent to all first instar larvae of Entedon. This character is quite remarkable and rare in Chalcidoidea (Parker and Thomson 1925), and thus may be used later to support monophyly of Entedon, if found in other species of the genus. Also, the combination of a nearly bare body and the "caudal crown" may facilitate separating the larvae of this genus from other en- doparasitoid larvae. Furthermore, the proper affiliation of the first-instar larva of E. sylvestris to a peculiar larval type requires a revision of the current classifi- cation (Parker 1924) of these types in Chalcidoidea. ACKNOWLEDGMENTS This work was in part supported by the grants of the following agencies and organizations: SRSF (State Fund of Fundamental Research, Ukraine, grant No. 05.07/00078), SCOPES Institutional Partnership project "Emphasising Classical and Conservation Biological Control in Research and Teaching" (7 IP 65648), DAAD (Deutcher Akademischer Austausch- dienst - German Academic Exchange Service, grant 322 - A/04/15867) and the SYNTHESYS Project (GB- TAF-535). The help of Drs. Yakiv Didukh and Antonina U'ins'ka (Kholodny Institute of Botany, NUAS, Kiev) in identification of S. loeselii and other Brassicaceae is also much appreciated, as well as the assistance of Dr. Stefan Schmidt during the author's visit to Munchen. My sincere appreciation is to Dr Lutz Behne (Deutsches Entomologisches Institut (DEI) im Zentrum fur Agrarlandschafts- und Nutzungs- forschung (ZALF)) who kindly confirmed my identi- fication of C. sisymbrii. I am also grateful to Drs. Stanislav Gorb, Ulrike Wegst, Sabinne Kuhnemann and Jiirgen Siewert (MPI) for their help in conducting the SEM work. I would like also to acknowledge Dr Hariet Hinz (CABI-Bioscience, European Station, Delemont, Switzerland) for the suggestions on the biology of Ceutorhynchus weevils. Special thanks to Drs Hariet Hinz, Andrew Polaszek (BMNH) and Gavin Broad for careful reading of the manuscript and suggested corrections. LITERATURE CITED Abedin, J. and M. A. Quayum. 1972. Some studies on the Entedon sp. (Eulophidae: Chalcidoidea), an important parasite of the jute weevil, Apion corchori Marsh. Pakistan Journal of Science, Lahore 23: 135-138. Askew, R. R. 1991. Review of species of Entedon Dalman having a complete frontal fork with redefinition of the species-group cioni Thomson (Hymenoptera: Eulophidae). Entomologica Scandi- navian 22: 219-229. Askew, R. R. 1992. Additions to the British list of Entedon Dalman (Hym., Eulophidae) with de- scriptions of three new species. Entomologist's Monthly Magazine 128: 119-128. Askew, R. R. and J. -P. Kopelke. 1989. Entedoninae associated with Pontania and allied sawflies (Hymenoptera: Eulophidae and Tenthredinidae). Entomologica Scandinavica 19: 431-434. Beaver, R. A. 1966. The biology and immature stages of Entedon leucogramma (Ratzeburg) (Hymenop- tera: Eulophidae), a parasite of bark beetles. Proceedings of the Royal Entomological Society of London (A) 41: 37-41. Boucek, Z. and R. R. Askew. 1968. Palaearctic Eulophidae sine Tetrastichinae. Index of Entomoph- agous Insects 3: 260 pp. Delucchi, V., Remaudiere, G. Le Francois, Paris. Dieckmann, L. 1966. Die Mitteleuropaischen Arten der Gattung Neosirocalus Ner. et Wagn. (mit Beschrei- bung von drei neuen Arten). Entomologische Blatter 62: 82-110. Dieckmann, L. 1972. Beitrage zur Insektenfauna der DDR: Coleoptera - Curculionidae: Ceutorhynch- inae. Beitrage zur Entomologie 22: 1-128. Darling, D. C. 1992. The life history and larval morphology of Aperilampus (Hymenoptera: Chal- cidoidea) with a discussion of the phylogenetic affinities of the Philomidinae. Systematic Entomol- ogy 17: 331-339. Darling, D. C. 1995. New species of Krombeinius (Hymenoptera: Chalcidoidea: Perliampidae) from Indonesia, and the first description of first-instar larva for the genus. Zoologische Mededelingen 69: 209-229. 250 Journal of Hymenoptera Research Ferriere, C. 1939. Chalcid flies attacking noxious beetles in India and New Guinea. Bulletin of Entomological Research 30: 163-168. Fisher, J. P. 1970. The biology and taxonomy of some chalcidoid parasites (Hymenoptera) of stem- living larvae of Apion (Col., Curculionidae). Transactions of the Royal Entomological Society of London 122: 293-322. " Graham, M. W. R. de V. 1971. Revision of British Entedon (Hymenoptera: Chalcidoidea) with de- scriptions of four new species. Transactions of the Royal Entomological Society of London 123: 313-358. Gumovsky, A. V. 1997. Review of Cionus weevils (Coleoptera, Curculionidae) with notes on their chalcidoid parasitoids (Hymenoptera, Chalcidoi- dea). Journal of the Ukrainian Entomological Society 3: 49-72. Gumovsky, A. 1999. Review of the genus Entedon Dalman, 1820 (Hymenoptera, Eulophidae, Ente- doninae). 5. Revision of the cyanellus species group. Annates historico-naturales Musei nationalis hungarici 1999 (91): 141-176. Gumovsky, A. and P. Boyadzhiev. 2003. Review of the Bulgarian Entedon Dalman, 1820 (Hymenoptera: Eulophidae, Entedoninae). Acta Entomologica Bul- garia 55 (3): 3-32. Hinz, H. L. and E. Gerber. 1998. Investigations on potential biological control agents of garlic mustard, Alliaria petiolata (Bieb.) Cavara & Grande. Annual Report 1998. CABI Bioscience Centre Switzerland, Delemont, Switzerland. Hinz, H. L., M. Cripps, B. J. L. Renteria, and A. Winspurdy. 2004. Biological control of whitetops, Lepidium draba and L. appelianum. Annual Report 2003. CABI Bioscience Centre Switzerland, Dele- mont, Switzerland. Kozlowski, M. W. 2003. Consumption of own eggs by curculionid females (Coleoptera: Curculionidae: Curculioninae, Ceitorhynchinae). Weevil News 10: 4, http://www.curci.de/WeevNews/KOZLOW_ A.htm. Parker, H. L. 1924. Recherches sur les formes postembryonaires de chalcidiens. Annates de la Societe Entomologique de France 93: 261-379. Parker, H. L. and W. R. Thompson. 1925. Notes on the larvae of the Chalcidoidea. Annals of the Entomo- logical Society of America 18: 384-395. Peck, O. 1963. A catalogue of the Nearctic Chalcidoi- dea (Insecta; Hymenoptera). Canadian Entomolo- gist (Supplement) 30: 1-1092. Rasplus, J.-Y. 1990. Nouvelles especes afrotropicales du genre Entedon Dalman et notes sur leur biologic Bulletin de la Societe Entomologique de France 94: 223-245. Stubbendieck, J., G. Y. Friisoe, and M. R. Bolick. 1994. Weeds of Nebraska and the Great Plains. Nebraska Department of Agriculture, Bureau of Plant Industry, Lincoln, Nebraska. 589 pp. Szelenyi, G. 1981. On the chalcidoid flies of the Hortobagy, I. Eulophidae (Hymenoptera). The Fauna of the Hortobagy National Park 1: 275-284. Mahunka, S., ed. ISBN 963-05-2519-4. Tiwari, N. K. 1976. Observations on Entedon sp. (Hymenoptera: Eulophidae) parasitising Frauen- feldiella jamboolii Grover. Indian journal of Ento- mology 37: 197-198. J. HYM. RES. Vol. 15(2), 2006, pp. 251-265 Phylogenetic Analysis of Chaenusa sensu lato (Hymenoptera: Braconidae) using Mitochondrial NADH 1 Dehydrogenase Gene Sequences Robert R. Kula,* Gregory Zolnerowich and Carolyn J. Ferguson (RRK, GZ) Department of Entomology, Kansas State University, Manhattan, KS 66506, USA (CJF) Division of Biology, Kansas State University, Manhattan, KS 66506, USA fCurrent address and address for correspondence: Systematic Entomology Laboratory, USDA, c/o U.S. National Museum of Natural History, MRC-168, 10th & Constitution Ave, N.W., Washington, DC 20560-0168, USA. Abstract. — Alysiinae currently contains over 1,500 described species and is divided into the tribes Alysiini and Dacnusini. There is disagreement on how species should be grouped within Dacnusini, and Chaenusa Haliday is a prime example. Chaenusa sensu lato is defined by the presence of setae on the compound eyes (Griffiths 1964). Alternatively, Riegel (1950, 1982) treated Chaenusa s.I. as three genera, Chaenusa sensu stricto, Chorebidea Viereck, and Chorebidella Riegel, and differentiated the genera primarily using forewing venation and shape of the forewing stigma. Phylogenetic analyses using molecular data have not been undertaken. Therefore, we assessed the monophyly and interspecific relationships of Chaenusa s.I., Chaenusa s.s., Chorebidea, and Chorebidella through maximum parsimony, maximum likelihood, and Bayesian analyses using mitochondrial NADH 1 dehydrogenase gene sequences. Chaenusa s.I. and Chorebidea were not monophyletic in any of the analyses, but four of five species of Chorebidea always formed a clade. Further, Chaenusa s.s. and Chorebidella were monophyletic in all analyses and were always sister taxa. The results of this study largely support Riegel's (1950, 1982) treatment of Chaenusa s.I. as Chaenusa s.s., Chorebidea, and Chorebidella. However, we suggest that Chaenusa s.I. be retained until additional phylogenetic analyses have been undertaken to confirm the relationships inferred in this study. In addition to the phylogenetic analyses, we discuss the morphological features relevant to Griffiths' definition of Chaenusa s.I. and Riegel's definition of Chaenusa s.s., Chorebidea, and Chorebidella. Alysiinae currently contains over 1,500 described species, and estimates of global richness range from 2,900 to 5,300 species (Dolphin and Quicke 2001). The mono- phyly of Alysiinae is firmly established based on the possession of exodont mand- ibles and the complete loss of the occipital carina (Griffiths 1964, Shaw and Huddle- ston 1991, Wharton 1997). Host records suggest that all alysiines are koinobiont endoparasitoids of cyclorrhaphous Diptera (Shaw and Huddleston 1991, Wharton and Austin 1991, Wharton 1997). Two tribes are currently recognized in Alysiinae: Alysiini and Dacnusini. Alysiini is probably nonmonophyletic as it is de- fined by the presence of forewing vein r-m (a plesiomorphy). Dacnusini is considered monophyletic based on the absence of forewing vein r-m (an apomorphy) (Grif- fiths 1964, Shaw and Huddleston 1991, Wharton 1994) and has consistently been recognized, although at different hierar- chal levels, since Forster (1862). There is widespread disagreement on how species should be grouped within Dacnusini, and Chaenusa Haliday is a prime example. Nixon (1943) divided dacnusines with setiferous compound eyes (Fig. 1) into two genera, Chaenusa and Chorebidea Vier- eck, and differentiated the genera using forewing venation and shape of the fore- wing stigma. Riegel (1950) established Chorebidella Riegel, a third genus contain- 252 Journal of Hymenoptera Research - Figs 1-4. Chaemisa sensu lato, Chaemtsa sensu stricto, Chorebidea, and Chorebidella. 1, Chorebidea americana, setiferous compound eyes. 2, Cha. quadriceps, 1st subdiscal cell closed. 3, Chorebidea saxicola, 1st sub- discal cell open, RS+M partially present, and stigma "long". 4, Chaemisa sp. 3, 1st subdiscal cell open, RS+M absent, and stigma "short, wide", a = 1st subdiscal cell, b = RS+M, and c = stigma. ing dacnusines with setiferous eyes. Like Nixon (1943) Riegel (1950) differentiated the genera primarily using forewing vena- tion and shape of the forewing stigma. Riegel (1950) regarded all dacnusines with setiferous eyes and a closed 1st subdiscal cell as Chaemisa (Fig. 2); he segregated dacnusines with setiferous eyes and an open 1st subdiscal cell into Cliorebidea or Chorebidclla. Species with forewing vein RS+M at least partially present and a "long" stigma were considered Chorebi- dea (Fig. 3); species with RS+M absent and a "short, wide" stigma were considered Chorebidella (Fig. 4). Griffiths (1964) hy- pothesized that all dacnusines with setifer- ous eyes form a monophyletic group and synonymized Chaemtsa sensu stricto, Clior- ebidea, and Chorebidella (i.e., Chaemtsa sensu lato). However, Riegel (1982) disagreed with Griffiths' synonymies and continued to treat Chaemisa sensu Griffiths (1964) as three genera. Riegel (1982), the only com- prehensive treatment of North American species of Chaemisa sJ., included several new species in Chaemisa s.s. and Chorebidea, but Wharton (1997) followed Chaemisa sensu Griffiths (1964) rather than Chaemisa sensu Riegel (1982). With 29 described species worldwide, Chaemisa s.l. is small relative to other dacnusine genera (e.g., over 240 species of Chorebus Haliday). Nearly all species are Nearctic or Palaearctic, but three species are known from Australia, and one species each is known from Madagascar and Argentina. As far as is known, flies in the ephydrid genus Hydrellia Robineau-Des- voidy are exclusively utilized as hosts (Griffiths 1964, Shaw and Huddleston 1991, Wharton and Austin 1991, Wharton 1997). Hydrellia is an important group for classical biological control of aquatic weeds. For example, Hydrellia pakistanae Deonier and Hydrellia balciunasi Bock have been imported and released for control of Hydrilla verticillata (L.f.) Royle in the United States. However, Hydrellia also contains species that are rice pests, such as Hydrellia griseola (Fallen) and Hydrellia philipvina Ferino. Species of Chaemisa s.l. may hinder classical biological control programs as contaminants in the quaran- tine phase (Wharton 1997) or through parasitism (by endemics) of introduced natural enemies. Conversely, species of Chaemisa s.l. may be important natural enemies of pest flies (Natarajan and Mathur 1980). Volume 15, Number 2, 2006 253 Table 1. Species analyzed in this study and their respective taxonomic placements, locality data, source repositories or collectors, and GenBank accession numbers. CNG = Cimarron National Grassland, KPBS = Konza Prairie Biological Station, and SFF = Santuario de Fauna y Flora. Species Taxonomic Placement Locality Source Accession No. Chaenusa n. sp. 1 Chaenusa s.s. Chile: Isla Chiloe, Vilupulli UCDC DQ9 17269 Chaenusa quadriceps Chaenusa s.s. Canada: ON: Ottawa TAMU DQ917272 Chaenusa n. sp. 2 Chorebidea U.S.A.: GA: Clarke Co., Athens TAMU DQ9 17270 Chaenusa n. sp. 3 Chorebidea Colombia: Boyaca, SFF Iguaque IAVH/HIC DQ917271 Chorebidea americana Chorebidea U.S.A.: FL: Putnam Co., Rodman Reservoir TAMU DQ917276 Chaenusa sp. 1 Chorebidea Canada: SK: —35 km W. Rosthem MJY DQ9 17273 Chaenusa sp. 2 Chorebidea U.S.A.: SC: Lexington Co., Lexington CNC DQ917274 Chaenusa bergi Chorebidella U.S.A.: KS: Riley Co., KPBS RRK/GZ DQ917268 Chaenusa sp. 3 Chorebidella India: Karnataka, Bangalore, Kumbalgodu TAMU DQ917275 Chorebus sp. 1 affinis group U.S.A.: AZ: Santa Cruz Co., Peha Blanca Lake RRK DQ9 17277 Chorebus sp. 2 affinis group U.S.A.: AZ: Santa Cruz Co., Pena Blanca Lake RRK DQ9 17278 Coelinius ferruginea Coelinius s.l. (Lt 'pton) U.S.A.: KS: Riley Co., KPBS RRK/GZ DQ9 17279 Coelinius hopkinsii Coeli)iius sd. {Li 'pton) U.S.A.: KS: Riley Co., KPBS RRK/GZ DQ917280 Dapsilarthra sp. 1 balteata group U.S.A.: TX: Brazos Co., Lick Creek Park RRK DQ917281 Opius sp. 1 Opiinae U.S.A.: KS: Morton Co., CNG RRK/GZ DQ917282 The taxonomic history discussed above illustrates that the limits of Chaenusa are uncertain. The monophyly of Chaenusa s.l., Chaenusa s.s., Chorebidea, and Chorebidella has never been assessed through phyloge- netic analysis, and character systems other than morphology (e.g., DNA sequence data) have not been utilized. Resolving the taxonomic limits of Chaenusa and understanding the evolutionary relation- ships among species in the genus are important factors for predicting their po- tential as biological control antagonists or agents. Additionally, increased taxonomic stability facilitates revisionary work on a group. Smith et al. (1999) and Michel- Salzat and Whitfield (2004) demons- trated the utility of mitochondrial NADH 1 dehydrogenase (ND1) gene sequences for resolving evolutionary relationships among aphidiine and microgastrine braco- nids, respectively. Thus, the objective of this study was to assess the monophyly and interspecific relationships of Chaenusa s.l, Chaenusa s.s., Chorebidea, and Chorebi- della using ND1 gene sequences. MATERIALS AND METHODS Terminologi/. — Terminology for mandibu- lar teeth and external male genitalia follows Wharton (1977). Terminology for all other anatomical features, including wing cells and veins, follows Sharkey and Wharton (1997). Abbreviations for repositories are as in Evenhuis and Samuelson (2005). Taxon sampling. — Species analyzed in this study and their respective taxonomic placements, locality data, source reposito- ries or collectors, and GenBank accession numbers (DQ917268-DQ917282) are listed in Table 1. Specimens used for DNA isolations were acquired from repositories as indicated in Table 1 or were collected by RRK, GZ, and Matthew J. Yoder (MJY, Texas A&M University) using yellow pan traps, sweep nets, and Malaise traps. Voucher specimens for each species are deposited in the Ambrose Morell Collec- tion for Molecular and Microbial Research at the American Museum of Natural History. With the exception of Chaenusa pallidinervis (Brethes), holotypes were ex- amined for all described alysiines dis- 254 Journal of Hymenoptera Research cussed in this paper. The holotype of Gyrocampa pallidinervis Brethes is housed in the Museo Argentina de Ciencias Nat- urales (MACN). The first author made multiple requests, but the MACN did not loan the holotype. The ingroup was composed of either 13 species of Dacnusini or 13 species of Dacnusini and one species of Alysiini depending on the analysis. Nine species of Chaenusa s.l. were included, with Chae- nusa s.s., Chorebidea, and Chorebidella repre- sented by two, five, and two species, respectively. Undescribed species were considered Chaenusa s.s., Chorebidea, or Chorebidella based on forewing configura- tion. Chaenusa n. sp. 1-3 will be described in a taxonomic revision of New World Chaenusa s.l. (Kula in preparation). Chae- nusa sp. 1 and 2 appear to be undescribed species but are only known from one and two individuals, respectively. Thus, RRK awaits the discovery of additional speci- mens before describing them. Evaluation of the literature for Old World Chaenusa s.l. suggests that Chaenusa sp. 3 is also unde- scribed. Two species each from Chorebus and Coelinius Nees were also treated as ingroup taxa to test the monophyly of Chaenusa s.l. Species of Chorebus and Coelinius possess morphological features (i.e., eye setation, number and position of mandibular teeth, metapleural setation, metasomal compres- sion) that suggest the potential for a close relationship with certain species of Chae- nusa s.l. (Kula personal observation). Both species of Chorebus fit in the affinis group (Griffiths 1968), and both species of Coeli- nius fit the concept of Lepton Zetterstedt ( = Coelinidea Viereck) in Griffiths (1964) (as a subgenus) and Riegel (1982) (as a genus). A species of either Opius Wesmael or Dapsilarthra Forster was specified as the outgroup to root trees depending on the analysis. Previous phylogenetic analyses support a sister group relationship be- tween Alysiinae and Opiinae (Quicke and van Achterberg 1990, Wharton et al. 1992, Quicke 1994, Belshaw et al. 1998, Dowton et al. 1998, Shi et al. 2005). Griffiths (1964) suggested that species of Dapsilarthra (Aly- siini) and Dacnusini might be closely related based on parasitism of leaf-mining agromyzids. Species of Dapsilarthra almost exclusively attack leaf-mining agromyzids (Wharton 1984, 1997), and dacnusines that Griffiths (1964) considered morphological- ly plesiomorphic are parasitoids of leaf- mining agromyzids. In analyses with Opius sp. 1 used to root trees, Dapsilarthra sp. 1 was included in the ingroup to explore the monophyly of Dacnusini. Dapsilarthra sp. 1 was used to root trees in analyses that excluded Opius sp. 1. DNA isolation, amplification, sequencing, and alignment. — Genomic DNA was iso- lated from individual wasps using a DNea- sy® Tissue Kit (Qiagen) according to the manufacturer's protocol for insects. Most specimens were ethanol-preserved, but several were dried, pinned specimens up to 14 years old. Polymerase chain reaction (PCR) amplifications and sequencing reac- tions were performed using an MJ Re- search PTC-200 thermal cycler. A portion of the ND1 gene was amplified using PCR set up in 25 liI volume. Oligonucleotide primers (ND1F: 5'-GATAAATCAAAW- GGKGT-3', ND1R: 5'-CAACCTTTTAGT- GATGC-3') and the PCR program were as in Smith et al. (1999) except the annealing temperature was optimized at 47 C. PCR products were purified using a Qiaquick® PCR Purification Kit (Qiagen) according to the manufacturer's protocol. Both strands of all purified PCR products were se- quenced using the PCR primers as se- quencing primers. Sequencing reactions were performed in 10 jal volume using an ABI Prism® BigDye™ Terminator v3.0 Ready Reaction Cycle Sequencing Kit (Applied Biosystems) according to the manufacturer's protocol. Sequencing reac- tion products were purified using spin columns filled with Sephadex® (Amer- sham Biosciences), dehydrated by vacuum centrifuge, and sent to the DNA Sequenc- Volume 15, Number 2, 2006 255 ing & Synthesis Facility at Iowa State University for gel runs on an ABI Prism® 3700 DNA Analyzer (Applied Biosystems). Sequences generated from the forward and reverse primers were aligned and edited in Sequencher™ 4.1.2 (Gene Codes Corpora- tion) to acquire a consensus sequence for each species. Consensus sequences were manually aligned in SeqPup 0.6 (Gilbert 1996) to produce a DNA sequence data matrix. The DNA data matrix was trans- lated to construct an amino acid (AA) sequence data matrix using the Drosophila Fallen mtDNA genetic code in MacClade 4.06 (Maddison and Maddison 2003). DNA and AA sequence characteristics and phylogenetic analysis. — The number of con- stant, variable parsimony uninformative, and parsimony informative characters were determined using PAUP* 4.0M0 (Swofford 2002), as were mean base fre- quencies. PAUP* 4.0M0 was also used to test for significant heterogeneity of base frequencies across taxa; base frequencies were considered significantly heteroge- neous if P < 0.05. Maximum parsimony (MP) and maxi- mum likelihood (ML) analyses were per- formed using PAUP* 4.0bl0. Maximum parsimony analyses were conducted for the DNA and AA data matrices using the branch and bound algorithm. Modeltest 3.06 (Posada and Crandall 1998) was used to determine the model of molecular evolution that best fit the data, and sub- sequently, ML analyses were conducted for the DNA data matrix using the heuristic search option with stepwise addition, 100 random addition sequence replicates, and tree bisection-reconnection (TBR) branch swapping. If the Hierarchical Likelihood Ratio Test (hLRT) and the Akaike Informa- tion Criterion (AIC) in Modeltest selected different models, ML analyses were per- formed using each model. Support for individual clades was assessed via boot- strap analyses. For MP 1,000 pseudorepli- cates with the branch and bound algorithm were used. For ML 100 pseudoreplicates using the heuristic search option with stepwise addition, 50 random addition sequence replicates, and TBR branch swap- ping were used. Bayesian analyses were performed using MrBayes 3.0b4 (Huelsenbeck and Ronquist 2001). Modeltest 3.06 was used to deter- mine the model of molecular evolution that best fit the data, and subsequently, Bayes- ian analyses were performed for the DNA data matrix. The data matrix was parti- tioned by codon position (i.e., 1st, 2nd, 3rd), and among-site rate variation was set (as a prior) to allow variable rates across partitions. The model of nucleotide sub- stitution and among-site rate variation was set as determined using Modeltest. The following model parameters were un- linked across the partitions: substitution rates of GTR model, character state fre- quencies, gamma shape parameter, and proportion of invariable sites. Each run consisted of 1,000,000 generations with a random starting tree and sample fre- quency of every 100 generations. The burnin was determined by constructing an XY scatter plot (i.e., generation X log likelihood value) using Microsoft® Excel to determine the number of generations until log likelihood values stabilized. Trees sampled prior to the generation at which log likelihood values stabilized were not included in the consensus tree. A 50% majority-rule consensus of the retained trees, showing the frequency of all ob- served bipartitions (i.e., posterior probabil- ities), was constructed using PAUP* 4.0bl0. Maximum parsimony analyses with Chaenusa s.l. constrained as monophyletic were also performed for the DNA data matrix. The search parameters were the same as for unconstrained MP analyses as discussed above. Most parsimonious trees (MPTs) from unconstrained and con- strained analyses were compared statisti- cally using the "Compare-2" permutation test (Faith 1991) in PAUP* 4.0M0. Under MP each of 10,000 random matrices (with 256 Journal of Hymenoptera Research Table 2. Number of constant (C), variable parsimony uninformative (VPU), and parsimony informative (PI) characters for all nucleotide (nuc) and amino acid (A A) sites and positions (Pos) 1, 2, and 3 with Opius sp. 1 excluded (excl) and included (incl). Dataset C vru P] All nuc sites (Opius sp. 1 excl) 241 66 123 All nuc sites (Opius sp. 1 incl) 234 60 136 Pos 1 (Opius sp. 1 excl) 83 26 34 Pos 1 (Opius sp. 1 incl) 81 25 37 Pos 2 (Opius sp. 1 excl) 119 7 17 Pos 2 (Opius sp. 1 incl) 116 8 19 Pos 3 (Opius sp. 1 excl) 39 33 72 Pos 3 (Opius sp. 1 incl) 37 27 80 All AA sites (Opius sp. 1 excl) 79 24 40 All AA sites (Opius sp. 1 incl) 74 26 43 all taxa randomized) were analyzed using the heuristic search option with stepwise addition, 500 random addition sequence replicates, and TBR branch swapping. The length difference between two trees (i.e., alternative hypotheses of relationships) was considered significant if P < 0.05. RESULTS DNA and AA sequence characteristics. — After sequence editing the aligned DNA data matrix was 430 bp and included no gaps. The DNA data matrix translated to an AA data matrix of 143 AAs. The number of constant, variable parsimony uninfor- mative, and parsimony informative char- acters for all sites and positions 1, 2, and 3 with Opius sp. 1 excluded and included are reported in Table 2. Evaluation of the mean base frequencies revealed a high A+T nucleotide bias, particularly in the first and third positions (Table 3). However, significant heteroge- neity of base frequencies across taxa was detected only for position 3 when Opius sp. 1 was included (Table 4). High A+T nucle- otide bias and less constrained nucleotide change relative to positions 1 and 2 may cause a high level of homoplasy in position 3 of insect mitochondrial protein-coding genes. Therefore, MP and bootstrap analy- ses were performed, as described above, with Opius sp. 1 included and position 3 excluded. In Modeltest the hLRT selected the TIM model with a proportion of invariable sites and gamma distributed rate variation among sites; the AIC selected the TrN model with a proportion of invariable sites and gamma distributed rate variation among sites. Maximum likelihood analyses using each model resulted in trees with identical topologies, and the results of analyses using the TrN model are pre- sented below. Phylogenetic analysis. — Maximum parsi- mony analysis of the DNA data matrix with Opius sp. 1 excluded resulted in two MPTs (tree length = 387 steps, consistency index excluding uninformative characters (CI) = 0.5609, retention index (RI) = 0.5959) (Fig. 5). The trees differed only in the placement of Chaenusa n. sp. 3 as either sister to Chorebus sp. 1 or sister to the rest of the ingroup. Chaenusa s.l. was not mono- Table 3. Mean base frequencies for all sites and positions (Pos) 1, 2, and 3 with Opius sp. 1 excluded (excl) and included (incl). Dataset A c G T All sites (Opius sp. 1 excl) 0.35626 0.10279 0.07695 0.46401 All sites (Opius sp. 1 incl) 0.35709 0.10381 0.07671 0.46239 Pos 1 (Opius sp. 1 excl) 0.35317 0.07908 0.09507 0.47267 Pos 1 (Opius sp. 1 incl) 0.35514 0.07901 0.09441 0.47144 Pos 2 (Opius sp. 1 excl) 0.20829 0.18746 0.11588 0.48836 Pos 2 (Opius sp. 1 incl) 0.20812 0.18726 0.11619 0.48843 Pos 3 (Opius sp. 1 excl) 0.50735 0.04165 0.02022 0.43078 Pos 3 (Opius sp. 1 incl) 0.50803 0.04496 0.01983 0.42719 Volume 15, Number 2, 2006 257 Table 4. Results of tests for significant hetero- geneity of base frequencies across taxa for all sites and positions (Pos) 1, 2, and 3 with Opius sp. 1 excluded (excl) and included (incl). Dataset P All sites {Opius sp. 1 excl) 17.509269 0.99882241 All sites (Opius sp. 1 incl) 19.096701 0.99908575 Pos 1 {Opius sp. 1 excl) 12.619144 0.99998081 Pos 1 (Opius sp. 1 incl) 13.202722 0.99999385 Pos 2 (Opius sp. 1 excl) 2.302238 1.00000000 Pos 2 (Opius sp. 1 incl) 2.338632 1.00000000 Pos 3 (Opius sp. 1 excl) 53.386385 0.06219822 Pos 3 (Opius sp. 1 incl) 60.168328 0.03417160 phyletic. In both trees the Coelinius clade was sister to the clade formed by four of the five species of Chorebidea included in the analysis. Further, Chaenusa n. sp. 3 either formed a clade with Chorebus sp. 1 and Chorebus sp. 2 or was sister to the rest of the ingroup. Chorebidea was not mono- phyletic, although four of five species of Chorebidea included in the analysis formed a clade with 94% bootstrap support. Chor- ebus was monophyletic in one tree, but bootstrap support was <50%. Chaenusa s.s., Chorebidella, and Coelinius were monophy- letic with 97%, 100%, and 75% bootstrap support, respectively. Bootstrap support for the relationships among these clades was <50% except for the sister group relationship between Chaenusa s.s. and Chorebidella (99% bootstrap support). 75/88 1 1.0(1 1 — <50/<50 100/97 | 1.00 I 0.75 98/72 1.00 94/64 1.00 <50/<50 0.76 97/78 | "0.76 1 <50/<50 99/100 1.00 0.68 loo/iod uhT] ■ ■ <50/84 1 <50/* ' r~ 0.96 1 1 Coelinius f err ugine a Coelinius hopkinsii Coelinius Chaenusa sp. 2 Chaenusa n. sp. 2 Chaenusa sp. 1 Chorebidea americana Chaenusa n. sp. 1 Chaenusa quadriceps Chaenusa bergi Chorebidea Chaenusa s.s. Chorebidella Chaenusa sp. 3 Chaenusa n. sp. 3 Chorebus sp. 1 Chorebus sp. 2 Chorebus Chaenusa n. sp. 3j| Chorebidea Dapsilarthra sp. 1 1 Alysiini (Outgroup) Fig. 5. Composite cladogram of two most parsimonious trees resulting from maximum parsimony analysis of the DNA data matrix with Opius sp. 1 excluded. Dashed line indicates alternative placement of Chaenusa n. sp. 3. Maximum parsimony bootstrap values are above branches and left of slashes. Where clades were recovered in maximum likelihood (ML) analysis with Opius sp. 1 excluded, bootstrap values are above branches and right of slashes. Where clades were recovered in Bayesian analysis with Opius sp. 1 excluded, posterior probabilities are below branches. Asterisks above and below branches indicate clades not recovered in MI and Bayesian analysis, respectively. 258 Journal of Hymenoptera Research Table 5. Bootstrap support and posterior pro- babilities for groups within Dacnusini recovered through maximum parsimony (MP) and Bayesian analyses with Opius sp. 1 included. Maximum parsimony analyses were conducted with position (Pos) 3 included (incl) and excluded (excl). nr = groups not recovered. MP MP Group (Pos 3 incl) (Pos 3 excl) Bayesian Chaenusa s.s. 97 100 0.77 Chorebidella 100 99 1.00 4 of 5 Chorebidea 98 85 1.00 Chaenusa s.s. + 99 81 1.00 Chorebidella Chorebus nr nr 0.88 Coelinius 67 67 1.00 Chorebus sp. 1 + Chaenusa 59 <50 nr n. sp. 3 Chorebus + Chae nusa n. 62 <50 0.84 sp. 3 Maximum likelihood and Bayesian anal- yses of the DNA data matrix with Opius sp. 1 excluded resulted in a most likely tree (-Ln likelihood value = 2246.66073) and a 50% majority-rule consensus tree, re- spectively, with topologies identical to the MP tree with Chaenusa n. sp. 3 sister to the rest of the ingroup (Fig. 5). For the Bayes- ian consensus tree, the burnin was the first 50 trees. Bootstrap support for ML and posterior probabilities for Bayesian are reported in Fig. 5. Maximum parsimony analysis with Opius sp. 1 included resulted in a single MPT (tree length = 435 steps, CI = 0.5474, RI = 0.5729) (tree not shown). Dacnusini was monophyletic with 78% bootstrap support. The relationships among dacnu- sines were identical to the MP tree with Opius sp. 1 excluded and Chaenusa n. sp. 3 sister to Chorebus sp. 1 (Fig. 5). Analysis with position 3 excluded resulted in a single MPT (tree length = 177, CI = 0.5755, RI == 0.6118) (tree not shown) with a topology identical to the tree with position 3 included. Dacnusini was mono- phyletic, but bootstrap support was <50%. Bootstrap support for groups within Dac- nusini for analyses with position 3 in- cluded and excluded are presented in Table 5. Bayesian analysis of the DNA data matrix with Opius sp. 1 included resulted in a 50% majority-rule consensus tree (tree not shown) with a topology nearly identi- cal to the MP tree with Opius sp. 1 excluded and Chaenusa n. sp. 3 sister to Chorebus sp. 1 (Fig. 5). The burnin was the first 70 trees. In terms of the relationships among dacnusines, the only differences between the trees were (1) Chorebus was monophyletic with Chaenusa n. sp. 3 sister to the Chorebus clade and (2) the clade containing all dacnusines except Chaenusa n. sp. 3, Chorebus sp. 1, and Chorebus sp. 2 was not recovered. Dacnusini was mono- phyletic with a posterior probability of 0.99. Posterior probabilities for groups within Dacnusini are presented in Table 5. Maximum likelihood analysis of the DNA data matrix with Opius sp. 1 included resulted in a most likely tree (-Ln likeli- hood value == 2437.12877) with a topology considerably different than trees from all other analyses (Fig. 6). Dacnusini was not monophyletic. Rather, Dapsilarthra sp. 1 was sister to Chaenusa n. sp. 3, but bootstrap support for this relationship was <50%. Chaenusa s.l. was not mono- phyletic. Chorebidea was not monophyletic, although four of five species of Chorebidea included in the analysis formed a clade with 67% bootstrap support. Chaenusa s.s., Chorebidella, Chorebus, and Coelinius were monophyletic with 75%, 99%, 80%, and 71% bootstrap support, respectively. Boot- strap support for the relationships among these clades was <50% except for the sister group relationship between Chaenusa s.s. and Chorebidella (98% bootstrap support). Maximum parsimony analysis with Opius sp. 1 excluded and Chaenusa s.l. constrained as monophyletic resulted in two MPTs (tree length = 393 steps, CI = 0.5503, RI = 0.5782) (trees not shown) six steps longer than the MPTs from the unconstrained analysis. The "Compare-2" test revealed that the two MPTs from the Volume 15, Number 2, 2006 259 71 Coelinius ferruginea Coelinius hopkinsii Coelinius 80 • <50 <50 <50 Chorebus <50 Chorebus sp. 1 ^^ Chorebus sp. 2 ^™ Chaenusa n. sp. 3 1 Chorebidea ^ — Dapsilarthra sp. 1 1 Alysiini 75 98 99 67 79 98 I Chaenusa n. sp. 1 Chaenusa quadriceps Chaenusa hergi Chaenusa sp. 3 Chaenusa sp. 2 Chaenusa n. sp. 2 Chaenusa s.s. L Chorebidella ( hacnusa sp. 1 Chorebidea americana Chorebidea 0.05 substitutions/site 0/;///.v sp. 1 1 Opiinae (Outgroup) Fig. 6. Phylogram resulting from maximum likelihood analysis of the DNA data matrix with Opius sp. 1 included. Bootstrap values are above branches. unconstrained analysis are not significant- ly shorter than either of the two MPTs from the constrained analysis (P = 0.125100, 0.127600, 0.162700, 0.165300). Maximum parsimony analysis with Opius sp. 1 in- cluded and Chaenusa s.l. constrained as monophyletic resulted in one MPT (tree length = 444 steps, CI = 0.5344, RI 0.5499) (tree not shown) nine steps longer than the MPT from the unconstrained analysis. The "Compare-2" test revealed that the MPT from the unconstrained analysis is significantly shorter than the MPT from the constrained analysis (P = 0.031300). Maximum parsimony analysis of the AA data matrix with Opius sp. 1 included resulted in two MPTs (tree length = 161 steps, CI = 0.7087, RI = 0.7176). Dacnusini was not monophyletic in the strict consen- sus of the two MPTs (Fig. 7). Rather, Dapsilarthra sp. 1 was sister to Chaenusa n. sp. 3, but bootstrap support for this re- lationship was <50%. Chaenusa s.l. was not monophyletic. Chorebidea was not mono- phyletic, although four of five species of 260 Journal of Hymenoptera Research <50/53 Coelinins 63/69 <50/* <50/<5<] Chorebus 98/96 94/95 <50/<50 Chaenusa s.s. 100/100 Chorebidella <50/na 96/96 93/94 Coelinius ferruginea Coelinins hopkinsii Chorebus sp. 1 Chorebus sp. 2 Chaenusa n. sp. 1 Chaenusa quadriceps Chaenusa bergi Chaenusa sp. 3 Chaenusa n. sp. 3 1 Chorebidea Dapsilarthra sp. 1 1 Alysiini Chaenusa sp. 2 Chaenusa sp. 1 Chaenusa n. sp. 2 ( Chorebidea americana Opius sp. 1 1 Opiinac (Outgroup) Chorebidea Fig. 7. Strict consensus of two most parsimonious trees resulting from maximum parsimony (MP) analysis of the amino acid data matrix with Opius sp. 1 included. Maximum parsimony bootstrap values are above branches and left of slashes. Where clades were recovered in MP analysis with Opius sp. 1 excluded, bootstrap values are above branches and right of slashes. The asterisk indicates a clade not recovered in MP analysis with Opius sp. 1 excluded, na = not applicable. Chorebidea included in the analysis formed a clade with 93% bootstrap support. Chae- nusa s.s., Chorebidella, and Chorebus were monophyletic with bootstrap support of 98%, 100%, and 63%, respectively. Coelinins was monophyletic, but bootstrap support was <50%. Bootstrap support for the relationships among these clades was <50% except for the sister group relation- ship between Chaenusa s.s. and Chorebidella (94% bootstrap support). Maximum parsimony analysis of the AA data matrix with Opius sp. 1 excluded resulted in six MPTs (tree length = 142 steps, CI = 0.7273, RI = 0.7479). In terms of the relationships among dacnusines, the strict consensus of the six MPTs (tree not shown) was identical to the strict consen- sus tree in Fig. 7 except the monophyly of Chaenusa s.l. was unresolved (see asterisk in Fig. 7). Bootstrap support is reported in Fig. 7. DISCUSSION DNA sequence characteristics. — Character- istics of the ND1 DNA sequences in this study are consistent with ND1 DNA sequences of other braconids (e.g., Smith and Kambhampati 1999, Smith et al. 1999, Michel-Salzat and Whitfield 2004). As in the aforementioned studies, the sequenced fragments in this study (including Opius sp. 1) are biased towards adenine (35.7%) and thymine (46.2%), particularly in the first (82.7%) and third (93.5%) positions. Significant heterogeneity of base frequen- Volume 15, Number 2, 2006 261 cies was detected for position 3 when Opins sp. 1 was included. As mentioned in Michel-Salzat and Whitfield (2004), the A+T nucleotide bias observed for insect mitochondrial DNA could influence the level of homoplasy, particularly in the first and third positions. However, 27.2% and 58.8% of the parsimony informative char- acters in the DNA data matrix with Opius sp. 1 included are in the first and third positions, respectively. Therefore, for most analyses all positions were considered and were not differentially weighted. We per- formed MP and bootstrap analyses with Opius sp. 1 included and position 3 excluded to examine the influence of position 3 on tree topology and branch support. The exclusion of position 3 had no influence on tree topology but resulted in lower bootstrap support for several clades. Conversely, there was a slight increase in CI and RI values when position 3 was excluded. This suggests that position 3 contains phylogenetic information that supports several clades but also increases the level of homoplasy in the data matrix. Tribe Dacnusini. — Griffiths (1964) and Wharton (1994) suggested that Dacnusini is monophyletic based on the absence of forewing vein r-m. Further, Dacnusini is homogeneous in terms of host utilization; the tribe exclusively contains parasitoids of plant-mining flies, particularly parasitoids of leaf- and stem-mining agromyzids, chloropids, and ephydrids (Wharton 1997). Maximum parsimony, ML, and Bayesian analyses were conducted with Dapsilarthra sp. 1 included in the ingroup to explore the monophyly of Dacnusini. In MP and Bayesian analyses of the DNA data matrix, Dacnusini was monophyletic with 78% bootstrap support and a posterior probability of 0.99, respectively. However, neither ML analysis of the DNA data matrix nor MP analysis of the AA data matrix recovered Dacnusini. Rather, Dapsi- larthra sp. 1 was always sister to Chaenusa n. sp. 3, but bootstrap support for this relationship was <50%. In MP analysis of the AA data matrix with Chaenusa n. sp. 3 excluded, Dacnusini was monophyletic in two of six MPTs, but bootstrap support was <50% (results not presented). Dacnu- sini was not monophyletic in ML analysis of the DNA data matrix with Chaenusa n. sp. 3 excluded (results not presented). Thus, ND1 DNA sequences and the ab- sence of forewing vein r-m largely, but not conclusively, support the monophyly of Dacnusini. Exclusive utilization of plant- mining flies as hosts, particularly leaf- and stem-mining agromyzids, chloropids, and ephydrids (i.e., biological homogeneity), provides further indication that Dacnusini is monophyletic. However, more extensive taxon sampling and the use of additional markers more conserved than ND1 are needed to confirm the monophyly of Dacnusini and resolve the more ancient divergences within the tribe. Genus Chaenusa sensu lato. — Chaenusa s.l. was not monophyletic in any of the analyses. Rather, the results indicate that certain species of Chaenusa s.l. are more closely related to species of Chorebus and Coel in ius than they are to other species of Chaenusa s.l. This result is not surprising for several reasons. Certain species of Chaenusa s.l. possess morphological fea- tures that suggest the potential for a close relationship with species of Chorebus and Coelinius. As is observed for species of the Chorebus affinis group (Griffiths 1968), several species of Chaenusa s.l. have four- toothed mandibles with three major teeth and one small tooth along the ventral margin of elongate tooth 2. In this study Chaenusa n. sp. 3, Chaenusa n. sp. 1, and Cha. quadriceps (Ashmead) exhibit this condition, as do four described (i.e., Chae- nusa anticostae Riegel, Chaenusa californica Riegel, Chaenusa illinae Riegel, Chaenusa rossi Riegel) and two undescribed Nearctic species of Chaenusa s.l. not included in this study (Kula unpublished). Further, the metapleural setation of Chaenusa n. sp. 3 is nearly oriented in a rosette surrounding a raised swelling, a character state used to 262 Journal of Hymenoptera Research define Chorebus. Chaenusa n. sp. 3 forms a clade with Chorebus in certain MP and Bayesian analyses, and it is possible that Chaenusa n. sp. 3 is a species of Cliorebus with setiferous eyes. A character state in females of Coelinius is lateral compression of the metasoma. Females of Chorebidea americana Riegel, Chorebidea bessae Riegel, Chorebidea mcclurei Riegel, Cha. rossi, Chorebidea saxicola Riegel, and one undescribed Nearctic species of Chaenusa sd. have a laterally compressed metasoma (Kula unpublished). In this study only Chorebidea americana clearly exhibits this condition. Further, Coelinius is partially defined on the possession of four-toothed mandibles with three major teeth and one small tooth between tooth 1 and 2. In this study Chaenusa sp. 2, Chaenusa n. sp. 2, and Chaenusa sp. 1 exhibit this condition, and it also occurs in an undescribed Nearctic species of Chaenusa sd. not included in this study (Kula unpublished). Griffiths (1964) proposed that among dacnusines setiferous eyes is unique to species of Chaenusa sd. and is a synapo- morphy that defines Chaenusa sd. How- ever, dacnusines in genera other than Chaenusa sd. have setiferous eyes. New World species of Cliorebus (47 morphospe- cies), Coelinius (19 morphospecies), Colo- neura Forster (two morphospecies), Dac- nusa Haliday (18 morphospecies), Epimicta Forster (two morphospecies), Exotela For- ster (14 morphospecies), Laotris Nixon (six specimens), and Synelix Forster (one mor- phospecies) all have setiferous eyes. Only New World species of Symphya Forster (13 morphospecies) have glabrous eyes (Kula unpublished). Character states other than setiferous eyes clearly place the aforemen- tioned species in their respective genera. In most cases the setae are straight and are so minute that they could easily escape de- tection using a stereomicroscope at 120X magnification (i.e., usually snorter than a facet width). For species of Chaenusa sd., at least some setae on the eyes are conspicuously longer than a facet width and are curved. However, 8.5% of the Chorebus and 5.3% of the Coelinius mor- phospecies examined have curved setae on the eyes longer than a facet width. Thus, the mere presence of setae on the eyes cannot be regarded as a synapomorphy that defines Chaenusa sd. Genus Chaenusa sensu stricto. — Chaenusa s.s. was monophyletic in all analyses, and branch support was moderate to strong. Chaenusa s.s. should be more extensively sampled in future phylogenetic analyses to provide a more robust assessment of monophyly. Six of the 11 described New World species of Chaenusa sd. fit in Chae- nusa s.s. (i.e., Cha. anticostae, Cha. californica, Cha. illinae, Cha. pallidinervis, Cha. quadri- ceps, Cha. rossi). However, all except Cha. quadriceps are only known from the holo- type. Thus, a very small number of New World specimens of Chaenusa s.s. are avail- able for DNA sequencing. Extensive col- lecting will be needed to increase the representation of New World Chaenusa s.s. in future phylogenetic analyses. The most successful methods for collecting speci- mens of Chaenusa sd. are yellow pan traps placed along the shore of permanent lakes, ponds, and streams and sweeping within and along the edge of aquatic habitats. Riegel (1950, 1982) defined Chaenusa s.s. using the following features: (1) 1st sub- discal cell closed, (2) stigma "short, wide", and (3) labial palpi four-segmented. Both species of Chaenusa s.s. included in this study have the 1st subdiscal cell closed, a relatively broad stigma, and three- or four-segmented labial palpi. The length of the distal palpomere in specimens with three-segmented labial palpi is approxi- mately the combined length of palpomeres 3 and 4 in specimens with four-segmented labial palpi. Further, examination with a scanning electron microscope revealed that the distinction between palpomeres 3 and 4 is extremely weak in some specimens of Chaenusa n. sp. 1, Cha. quadriceps, and an Volume 15, Number 2, 2006 263 undescribed Nearctic species that fits Chae- nusa s.s. Thus, it appears that three-seg- mented labial palpi in Chaenusa n. sp. 1 and Clin, quadriceps resulted from the fusion of palpomeres 3 and 4 or the division of palpomere 3 into two palpomeres. Genus Chorebidea. — Chorebidea was not monophyletic in any of the analyses. However, four of five species of Chorebidea included in this study formed a clade in all analyses, and branch support was weak to strong. Riegel (1950, 1982) defined Chorebi- dea using the following features: (1) 1st subdiscal cell open, (2) forewing vein RS+M at least partially present, (3) stigma "long", (4) labial palpi three-segmented, and (5) gonoforceps "stocking-shaped in lateral view". All species of Chorebidea included in this study have an open 1st subdiscal cell through the partial or com- plete absence of forewing veins 2-1A and 2cu-a, and forewing vein RS+M is at least partially present. Both features exhibit some degree of intraspecific variation. The 1st subdiscal cell is rarely (3.1%, one of 32 specimens examined) closed in Chaenusa n. sp. 3, and although forewing vein RS+M is present for all species, it may vary from complete and tubular to minute- ly present posteriorly. Riegel (1950, 1982) included a "long" stigma in his concept of Chorebidea, but Chorebidea americana and Chorebidea bessae have a relatively broad stigma. The stigma is relatively long for Chaenusa sp. 2, Chaenusa n. sp. 2, Chaenusa sp. 1, and Chaenusa n. sp. 3 but is relatively broad for Chorebidea americana. Chaenusa sp. 2, Chaenusa n. sp. 2, Chaenusa sp. 1, and Chorebidea americana have three-segmented labial palpi, but the labial palpi are four- segmented for Chaenusa n. sp. 3. Lastly, Chorebidea americana has "stocking- shaped" gonoforceps, but Chaenusa sp. 2, Chaenusa n. sp. 2, and Chaenusa sp. 1 have gonoforceps that gradually narrow proxi- mally to distally and are roughly triangu- lar-shaped. Chaenusa n. sp. 3 has roughly rectangular-shaped gonoforceps that are truncate distally. Genus Chorebidella. — Chorebidella was monophyletic in all analyses, and branch support was strong. Chorebidella should be more extensively sampled in future phylo- genetic analyses to provide a more robust assessment of monophyly. Only one of the 11 described New World species of Chae- nusa s.l. fits in Chorebidella (i.e., Chaenusa bergi (Riegel)). We acquired two Old World species in addition to Cha. bergi but only had permission to use one for DNA sequencing. As for Chaenusa s.s. extensive collecting will be needed to increase the representation of New World Chorebidella in future phylogenetic analyses. Riegel (1950, 1982) defined Chorebidella using the following features: (1) 1st sub- discal cell open, (2) forewing vein RS+M absent, (3) stigma "short, wide", (4) labial palpi three-segmented, and (5) gonofor- ceps "not stocking-shaped in lateral view". Both species of Chorebidella included in this study have the 1st subdiscal cell open through the partial or complete absence of forewing veins 2-1 A and 2cu-a, forewing vein RS+M absent, a relatively broad stigma, and gonoforceps that gradually narrow proximally to distally and are roughly triangular-shaped. Chaenusa bergi has three-segmented labial palpi, but Chae- nusa sp. 3 has two-segmented labial palpi. Two-segmented labial palpi have not been recorded for any species of Chaenusa s.l. CONCLUSIONS The results of this study indicate that Chaenusa s.l. is not monophyletic, but Chaenusa s.s. and Chorebidella are mono- phyletic groups with moderate to strong support. Chorebidea was not monophyletic in any of the analyses, but four of five species of Chorebidea included in this study formed a clade in all analyses. The species of Chorebidea that did not form a clade with the other species of Chorebidea (i.e., Chae- nusa n. sp. 3) exhibits morphological character states observed for species of Chorebus. Further, Chaenusa n. sp. 3 forms a clade with Chorebus in certain MP and 264 Journal of Hymenoptera Research Bayesian analyses, and this suggests that Chaenusa n. sp. 3 may actually be a species of Chorebus with long curved setae on the eyes. Phylogenetic analyses using ND1 gene sequences largely support Riegel's (1950, 1982) treatment of Chaenusa s.l. as Chaenusa s.s., Chorebidea, and Chorebidella. However, we suggest that Chaenusa s.l. be retained until phylogenetic analyses with nuclear markers, morphology, and greater taxon sampling have been undertaken to confirm the relationships inferred in this study. ACKNOWLEDGMENTS We thank the following persons and repositories for providing specimens used in this study: David B. Wahl (AEIC), Henri Goulet (CNC), Lubomir Masner (CNC), E. Richard Hoebeke (CUIC), Michael J. Sharkey (HIC), IAVH, Colin Favret (INHS), KSUC, Stefan Cover (MCZ), Robert A. Wharton (TAMU), Matthew J. Yoder (TAMU), Steven L. Heydon (UCDC), Mark F. O'Brien (UMMZ), and David R. Smith (USNM). Srinivas Kambhampati provided valuable technical advice and suggestions on the manuscript. We are grateful to Suzanne C. Strakosh for providing technical support throughout the course of this research. We thank Matthew L. Buffington for consultation on the use of MrBayes. Kent Hampton (Scanning Electron Microscope Laboratory, KSU) captured the original scanning electron micrograph used to illustrate compound eye setation. This article is Contribution No. 05-271 -J from the Kansas Agricul- tural Experiment Station and was supported by KAES Hatch Project No. 583, Insect Systematics, and Na- tional Science Foundation grant DEB-0089656 awarded to CJF. LITERATURE CITED Belshaw, R., M. Fitton, E. Herniou, C. Gimeno, and D. L. J. Quicke. 1998. A phylogenetic reconstruction of the Ichneumonoidea (Hymenoptera) based on the D2 variable region of 28 S ribosomal RNA. Systematic Entomology 23: 109-123. Dolphin, K. and D. L. ]. Quicke. 2001. Estimating the global species richness of an incompletely de- scribed taxon: an example using parasitoid wasps (Hymenoptera: Braconidae). Biological Journal of the Unman Society 73: 279-286. Dowton, M., A. D. Austin, and M. F. Antolin. 1998. Evolutionary relationships among the Braconidae (Hymenoptera: Ichneumonoidea) inferred from partial 16 S rDNA gene sequences. Insect Molec- ular Biology 7: 129-150. Evenhuis, N. L. and G. A. Samuelson. 2005. The Insect and Spider Collections of the World Website. Bishop Museum, Honolulu, Hawaii, http://hbs. bishopmuseum.org/codens/codens-r-us.html [accessed 25 February 2005]. Faith, D. P. 1991. Cladistic permutation tests for monophyly and nonmonophyly. Systematic Zool- ogy 40: 366-375. Forster, A. 1862. Synopsis der Familien und Gattun- gen der Braconen. Verhandlungen des Naturhistor- ischen Vereines preussischen Rheinlande und West- phalens 19: 225-288. Gilbert, D. G. 1996. SeqPup, Version 0.6. Indiana University, Bloomington, Indiana, http://iubio. bio.indiana.edu/soft/molbio/seqpup/ [released July 1996]. Griffiths, G. C. D. 1964. The Alysiinae (Hym. Braconidae) parasites of the Agromyzidae (Dip- tera). I. General questions of taxonomy, biology and evolution. Beitrdge zur Entonwlogie 14: 823-914. Griffiths, G. C. D. 1968. The Alysiinae (Hym. Braconidae) parasites of the Agromyzidae (Dip- tera). VI. The parasites of Cerodontha Rondani s. 1. Beitriige zur Entonwlogie 18: 63-152. Huelsenbeck, J. P. and F. Ronquist. 2001. MRBAYES: Bayesian inference of phylogeny. Bioinformatics 17:754-755. Maddison, D. R. and W. P. Maddison. 2003. MacClade 4. Analysis of Phylogeny and Character Evolution. Version 4.06. Sinauer Associates, Sunderland, Massachusetts. Michel-Salzat, A. and J. B. Whitfield. 2004. Preliminary evolutionary relationships within the parasitoid wasp genus Cotesia (Hymenoptera: Braconidae: Microgastrinae): combined analysis of four genes. Systematic Entomology 29: 371-382. Natarajan, K. and K. C. Mathur. 1980. New record of Chaenusa sp. (Braconidae-Hymenoptera) as a par- asite of the rice whorl maggot. Science and Culture 46: 337-338. Nixon, G. E. J. 1943. A revision of the European Dacnusini (Hym., Braconidae, Dacnusinae). The Eiitomologist's Monthly Magazine 79: 20-34, 159-168. Posada, D. and K. A. Crandall. 1998. Modeltest: testing the model of DNA substitution. Bioinfor- matics 14: 817-818. Quicke, D. L. J. 1994. Phylogenetics and biological transitions in the Braconidae (Hymenoptera: Ichneumonoidea). Norwegian journal of Agricul- tural Sciences 16: 155-162. Quicke, D. L. J. and C. van Achterberg. 1990. Phylogeny of the subfamilies of the family Braconidae (Hymenoptera: Ichneumonoidea). Zo- ologische Verhandelingen 258: 1-95. Riegel, G. T. 1950. A new genus and species of Dacnusini (Hym.: Braconidae). Entomological News 61: 125-129. Volume 15, Number 2, 2006 265 Riegel, G. T. 1982. The American species of Dac- nusinae, excluding certain Dacnusini (Hyme- noptera: Braconidae). Novitates Arthropropodae 1: 1-185. Sharkey, M. J. and R. A. Wharton. 1997. Morphology and terminology. Pp. 19-38 in: Wharton, R. A., P. M. Marsh, and M. J. Sharkey, eds. Manual of the New World Genera of the Family Braconidae (Hyme- noptera). Special Publication of the International Society of Hymenopterists, Number 1. Shaw, M. R. and T. Huddleston. 1991. Classification and biology of braconid wasps (Hymenoptera: Braconidae). Handbooks for the Identification of British Insects 7 (11): 1-126. Shi, M., X. X. Chen, and C. van Achterberg. 2005. Phylogenetic relationships among the Braconi- dae (Hymenoptera: Ichneumonoidea) inferred from partial 16S rDNA, 28S rDNA D2, 18S rDNA gene sequences and morphological char- acters. Molecular Phylogenetics and Evolution 37: 104-116. Smith, P. T. and S. Kambhampati. 1999. Status of the Cotesia flavipes species complex (Braconidae: Microgastrinae) based on mitochondrial 16 S rRNA and NADH 1 dehydrogenase gene se- quence. Journal of the Kansas Entomological Society 72: 306-314. Smith, P. T., S. Kambhampati, W. Volkl, and M. Mackauer. 1999. A phylogeny of aphid parasi- toids (Hymenoptera: Braconidae: Aphidiinae) inferred from mitochondrial NADH 1 dehydro- genase gene sequence. Molecular Phylogenetics and Evolution 11: 236-245. Swofford, D. L. 2002. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other Methods). Version 4.0bI0. Sinauer Associates, Sunderland, Massa- chusetts. Wharton, R. A. 1977. New World Aphaereta species (Hymenoptera: Braconidae: Alysiinae), with a dis- cussion of terminology used in the tribe Alysiini. Annals of the Entomological Society of America 70: 782-803. Wharton, R. A. 1984. Biology of the Alysiini (Hyme- noptera: Braconidae), parasitoids of cyclorrha- phous Diptera. Texas Agricultural Experiment Station Technical Monograph 11: 1-39. Wharton, R. A. 1994. New genera, species, and records of New World Alysiinae (Hymenoptera: Braconi- dae). Proceedings of the Entomological Society of Washington 96: 630-664. Wharton, R. A. 1997. Subfamily Alysiinae. Pp. 84-1 17, 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 Hymenop- terists, Number 1. Wharton, R. A. and A. D. Austin. 1991. Revision of Australian Dacnusini (Hymenoptera: Braconidae: Alysiinae), parasitoids of cyclorrhaphous Diptera. Journal of the Australian Entomological Society 30: 193-206. Wharton, R. A., S. R. Shaw, M. J. Sharkey, D. B. Wahl, J. B. Woolley, J. B. Whitfield, P. M. Marsh, and W. Johnson. 1992. Phylogeny of the subfamilies of the family Braconidae (Hymenoptera: Ichneumo- noidea): a reassessment. Cladistics 8: 199-235. I. HYM. RES. Vol. 15(2), 2006, pp. 266-269 NOTE Observations on Flower Association and Mating Behaviour of the Pollen Wasp Species Celonites abbreviates (Villers, 1789) in Greece (Hymenoptera: Vespidae, Masarinae) VOLKER MAUSS Staatliches Museum fur Naturkunde, Abt. Entomologie, Rosenstein 1, D-70191 Stuttgart, Germany, email: volker.mauss@stechimmenschutz.de Celonites abbreviatus (Villers) ranges from Satureja thymbra L. (Lamiaceae) with neigh- Morocco across the entire North Mediter- bouring Phrygana vegetation (Fig. 1) (Ordo ranean Area to Turkey. In the North it Cisto-Micromeretalia Oberdorfer 54; cf. reaches the southern parts of Central Horvat et al 1974). The ground was hard, Europe (Gusenleitner 1997). Most informa- stony and only sparsely covered with tion on the bionomics stem from the vegetation. An open source of water was Central European range of the species present only in one of the localities. (Schremmer 1959, Bluthgen 1961, Bellmann Females of C. abbreviatus were frequently 1984, 1995, Miiller 1996, Amiet and Mauss observed to visit flowers of Satureja thym- 2003) whereas bionomical records from bra. During the flower visits the proboscis other parts of its distribution are rare (cf. was protruded into the corolla tube, in- Lichtenstein 1869, Ferton 1901, Fahringer dicating nectar uptake (Fig. 4). Simulta- 1922). Moreover, until now the mating neously, the facial part of the head was behaviour of the species has been com- rubbed over the nototribic anthers of the pletely unknown. flower, and pollen grains accumulated on On a field trip to the Maleas Peninsula of the frons of the female (Fig. 4), where the Peloponnese (Greece, Laconias) C. modified setae form a pollen-collecting abbreviatus was observed between the 5th apparatus (Schremmer 1959, Miiller 1996). and 10th of June 2005 at four localities in the At regular intervals the females interrupted vicinity of Profitis Ilias near Neapoli (I. nectar and pollen uptake and transferred 36 26.687' N 23 06.926' E, 55 m a.s.L; II. pollen from the head to the mouthparts by 36 26.719' N 23 08.482' E, 77 m a.s.L; III. alternating grooming movements of the 36 26.122' N 23 08.303' E, 10 m a.s.L; IV. forelegs and ingested the pollen (Fig. 5). 36' 26.252' N 23 07.012' E, 20 m a.s.L). All On two occasions females of C. abbreviatus sites were situated in old, open fallow collected pollen from Thymus capitatus (L.) areas characterised by large patches of Ho. and Li. (Lamiaceae) at localities where Figs 1-5. Habitat and behaviour of Celonites abbreviatus in vicinity of Profitis Ilias (Peloponnese, Greece). 1, Fallow area at locality II with large patches of Satureja thymbra (middle and right foreground), at which male and female pollen wasps were observed; Phn/gana vegetation in background. 2, Female resting on sun-exposed stone. 3, Insertion phase of copulation; male clasping to the back of the female. 4, Female visiting flower of S. thymbra; her proboscis is protruded into the corolla tube, and the pollen collecting apparatus on the frons is Volume 15, Number 2, 2006 267 rubbed over the nototribic anthers. 5, Female transferring pollen from the fore-tarsal pollen comb of her left foreleg to her mouthparts, after grooming the pollen collecting hairs on the frons with the foreleg. 268 Journal of Hymenoptera Research flowers of S. thymbra were nearly or followed by the male which finally completely withered. During visits to T. pounced on her. The pair fell to the capitatus females stood with their middle ground, where a short period of grappling and hindlegs on the lower lip of the flower, occurred of about one second, after which raised the anterior part of the body and the male was positioned on the back of the rubbed the facial area of the head over the female with his middle and hindlegs nototribic anthers, while the proboscis wrapped around the female's metasoma remained retracted. A single flower visit and his genitalia inserted into the genital of a male was recorded at S. thymbra. The chamber of the female (Fig. 3). The pair observed flower visiting behaviour of C. remained motionless in this position for abbreviatus on the Maleas Peninsula adds about 10 seconds. Then the partners sepa- further evidence that C. abbreviatus is rated and flew away. The second copula- specialised with regard to its pollen source tion differed from the first in that the to flowers of Lamiaceae since in Central patrolling male pounced on a female vis- Europe, Italy and Croatia pollen collecting iting a flower of S. thymbra. During the of C. abbreviatus is also restricted to various following insertion phase the pair re- flowers of Lamiaceae (e.g. Acinos arvensis mained on the flower. The male released (Lam.) Dandy, Ballota nigra L., Ballota the female's body and fell over backwards pseudodictamnus Bentham, Salvia officinalis but his genitalia remained in the female's L., Stacln/s cretica L., Teucrium montanum L. genital chamber. After less than 10 sec- and Thymus spec; Bellmann 1984, 1995, onds, the pair separated and both partners Schremmer 1959, Miiller 1996). Although flew off. The female alighted on a nearby imagines have been recorded also from stone and cleaned her head with her flowers of other plants families such as forelegs, while the male flew off and Boraginaceae, Crassulaceae and Gerania- disappeared. This is the first record of ceae (Schremmer 1959, Schmiedeknecht mating behaviour in C. abbreviatus and also 1930, Bliithgen 1961), these visits were the first description of the copulation of probably for nectar uptake only (Schrem- a species of the genus Celonites, in general, mer 1959). However, it has been recorded that males Between flower visits the females often of Afrotropical species of Celonites also alighted briefly on the ground or on small search for females in the vicinity of forage stones close to the forage plants (Fig. 2). On plants (Gess 1996: 59) indicating that re- a single occasion a female defecated after source based mating systems may be more alighting on a stone. Similarly, females of widespread in this monophylum. The un- C. abbreviatus from Central Europe were usual position of the male in the second frequently observed to alight on sun- copulation looked similar to the hanging exposed stones or on the ground in the position in the first phase of the copulation vicinity of forage plants (Bluthgen 1961). A of vespine wasps (cf. Schulz-Langner comparable behaviour is shown by females 1954). However, it remains to be shown of the Afrotropical Celonites clypeatus whether this is a regular alternative mating Brauns (Gess 1993). position in C. abbreviatus or if the male Males repeatedly patrolled flowers of S. accidentally lost his hold on the female's thymbra flying at the level of the inflor- metasoma. escences. Copulations were observed in Imagines of C. abbreviatus were not two instances. The first was initiated by observed at water, which is in agreement a patrolling male that rapidly approached with the behaviour of C. abbreviatus in a female flying towards a plant of S. Central Europe (cf. Bellmann 1984, 1995) thymbra. The female responded, in that and of Afrotropical species of Celonites she flew about 0.2 m back from the plant, (Gess 1996: 107). Volume 15, Number 2, 2006 269 ACKNOWLEDGMENTS John Plant (Vienna) kindly read the manuscript and improved my English. LITERATURE CITED Amiet, F. and V. Mauss. 2003. First report of male sleeping aggregations in the pollen wasp Celonites abbreviatus (Villers, 1789) (Hymenoptera: Vespi- dae: Masarinae). Journal of Hymenoptera Research 12: 355-357. Bellmann, H. 1984. Beobachtungen zum Brutverhalten von Celonites abbreviatus Villers (Hymenoptera, Masaridae). Zoologischer Anzeiger 212: 321-328. Bellmann, H. 1995. Bienen, Wespen, Ameisen: die Hautfliigler Mitteleuropas. Kosmos Naturfuhrer. Franckh-Kosmos, Stuttgart. Bliithgen, P. 1961. Die Faltenwespen Mitteleuropas (Hymenoptera, Diploptera). Abhandlungen der Deutschen Akademie der Wissenschaften zu Berlin Klasse fur Chemie, Geologie und Biologie 1961 (2): 1-249. Fahringer, J. 1922. Hymenopterologische Ergebnisse einer wissenschaftlichen Studienreise nach der Tiirkei und Kleinasien (mit Ausschluss des Amanusgebirges). Archiv fur Naturgeschichte 88: 149-222, 9 figs. Ferton, C. 1901. Notes detachees sur l'instinct des Hymenopteres melliferes et ravisseurs avec la description de quelques especes. Annalcs de la Societe Entomologique de France 70: 83-148. Gess, S. K. 1993. The Karoo Violet and a masarid wasp - a mutually beneficial association. The Naturalist 37: 32-36. Gess, S. K. 1996. The pollen wasps — Ecologi/ and natural history of the Masarinae. Harvard University Press, Cambridge, Massachusetts. Gusenleitner, J. 1997. Die europaischen Arten der Gattung Celonites Latreille 1802 (Hymeno- ptera, Masaridae). Linzer Biologische Beitrdge 29: 109-115. Horvat, I., V. Glavac, and H. Ellenberg. 1974. Vegetation Siidosteuropas. Geobotanica selecta 4. Gustav Fischer, Stuttgart. 768 pp. Lichtenstein, J. 1869. No title. Annalcs de la Societe Entomologique de France (4) 9: XXIX. Miiller, A. 1996. Convergent evolution of morpholog- ical specializations in Central European bee and honey wasp species as an adaptation to the uptake of pollen from nototribic flowers (Hyme- noptera, Apoidea and Masaridae). Biological Journal of the Linnean Society 57: 235-252. Schmiedeknecht, O. 1930. Handbuch der Hymenopteren Mitteleuropas. 2. Aufl. Gustav Fische, Jena. Schremmer, F. 1959. Der bisher unbekannte Pollen- sammelapparat der Honigwespe Celonites abbre- viatus Vill. (Vespidae, Masarinae). Zeitschrift fiir Morphologic und Okologie der Tiere 48: 424- 438. Schulz-Langner, E. 1954. Beobachtungen zur Kopula- tion gefangen gehaltener Wespen. Zoologischer Anzeiger 152: 39^42. J. HYM. RES. Vol. 15(2), 2006, pp. 270-276 The Genus AUotilla Schuster (Hymenoptera: Mutillidae): Phylogenetic Analysis of its Relationships, First Description of the Female and New Distribution Records DlOMEDES QUINTERO A. AND ROBERTO A. CAMBRA T. Museo de Invertebrados G.B. Fairchild, Estafeta Universitaria, Universidad de Panama, Panama, Panama; email: dquinter@ancon.up.ac.pa rcambra@ancon.up.ac.pa Abstract. — The morphological characteristics of the monotypic genus AUotilla Schuster are discussed; new distribution records and a morphology-based phylogenetic analysis are presented, clarifying possible relationships with other sphaeropthalmine genera. Females of the genus are described and illustrated for the first time; male genitalia are illustrated. Until now, AUotilla gibbosa Schuster, 1949 was known only from the holotype male from Cordoba, Argentina. In a list of Mutillidae from Argentina, Fritz (1998) included Buenos Aires as an additional collection locality for this species, without providing any additional information such as the number of specimens examined, their sex, or the depository. We must consider the Buenos Aires record not to be valid because it is incomplete and because we have been unable to locate any of Fritz's AUotilla specimens in his reference collections at the AMNH, New York (to which they were sold by his widow after his death) or at the Museo de Ciencias Naturales "B. Rivadavia," Buenos Aires (Roig-Alsina, pers. comm.). The female of the Neotropical, mono- typic genus AUotilla is described and illustrated for the first time; new distribu- tion records and a morphology-based phylogenetic analysis are presented. MATERIALS AND METHODS We follow Brothers (1999) for the sub- family classification of the Mutillidae. For SEM examination we used a JEOL model JSM 5600. The AUotilla gibbosa specimens examined are deposited in the follow- ing institutions: American Museum of Natural History, New York (AMNH); Museum of Comparative Zoology, Har- vard University , Cambridge (MCZ); Mu- seo de Invertebrados G. B. Fairchild, Uni- versidad de Panama (MIUP); D. J. Broth- ers's personal collection, University of KwaZulu-Natal, South Africa (DJBC); and Institut Royal des Sciencies Naturelles de Belgique (ISNB). RESULTS AUotilla Schuster AUotilla Schuster 1949: 89-93. Type species: AUotilla gibbosa Schuster, 1949, by original designation and monotypy. Generic diameters of female. — Body cov- ered with simple setae only. Head: almost round, narrower than mesosoma; eye small, nearly circular and flat (Fig. 3); genal carina absent; scrobal carina absent; antenna 12-segmented; antennal tubercle strongly projecting, with low lamellate ridge on anteromedial surface (Fig. 3); mandible slender, with preapical tooth nearly obsolete or totally absent (Fig. 3); mandible evenly arcuate on ventral mar- gin, without tooth or constriction; probos- cidal fossa with carina nearly reaching inner mandibular base; hypostoma without tooth or tubercle; maxillary palp 6-seg- Volume 15, Number 2, 2006 271 merited; labial palp 4-segmented, the sec- ond segment almost as long as wide. Mesosoma: broader than long, pyriform (Fig. 4); dorsum strongly convex, without transverse carina along posterior margin of pronotum or lateral carina; dorso-lateral margins without spines; scutellar scale absent; mesopleuron swollen; leg with apex of tarsus simple, not produced above claws. Metasoma: first metasomal tergum not constricted posteriorly, sessile with tergum 2 (Fig. 5); tergum II evenly convex, without rows of longitudinal carinae; ter- gum II with felt lines; sternum II without felt lines; tergum VI with surface totally sculptured and evenly merging with rest of tergum, pygidial area poorly defined by postero-lateral carina, only visible under high magnification (>30x) (Fig. 6). Allotilla gibbosa Schuster, 1949 (Figs 3-8) Allotilla gibbosa Schuster, 1949: 93-95, Holotype male, Argentina: Cordoba, col. W. M. Davis, Harvard University, MCZ, type 30516, exam- ined. Description of female. — Integumental col- or: head, mesosoma, all metasomal sterna, terga I, III and IV, reddish brown; antennae and legs yellow-red; mandible reddish brown except apical third blackish; tergum II reddish brown except posteriorly with two lateral, transverse, black spots inter- rupted medially; terga V and VI black. Head: vertex and gena with sparse, medi- um-sized punctures one or more diameters apart (Fig. 3); punctures of frons denser, less than one diameter apart; vertex and frons with long, sparse, erect and semi- erect, dark setae; gena, clypeus laterally and hypostoma with long, sparse, pale setae. Mesosoma: pronotum and mesono- tum with punctures as on vertex (Fig. 4); metanotum with transverse reticulate band (Fig. 4); dorsum of propodeum mostly densely micropunctate-rugose (Fig. 4) ex- cept a narrow impunctate area near meta- notum; mesopleuron with dense, medium- sized punctures, except for mostly im- punctate anterior area near lateral area of pronotum; metapleura and lateral area of propodeum impunctate, smooth; setation of pronotum, mesonotum and metanotum similar to that of vertex; mesopleura and dorsum of propodeum with pale setae; metapleura and lateral area of propodeum glabrous. Metasoma: terga I and II with small, dense punctures, sparser in apical areas (Fig. 5); terga III and IV with small, sparse punctures; tergum V mostly smooth, except basal lateral area with a few punctures; tergum VI with scale-like surface sculpture basally, scales diminish- ing in size toward apex, gradually turning into granules (Fig. 6); sterna I and VI smooth; sterna II and III with small, somewhat sparse punctures; sterna IV and V mostly smooth, except apex with small, dense punctures; tergum I mostly with pale setae, a few dark setae at apex; tergum II with dark setae, except lateral area and apex with pale setae; terga III and IV mostly with pale setae; tergum V mostly glabrous, with a few pale and dark setae laterally; tergum VI glabrous; sterna I to V with pale setae; sternum VI glabrous. Additional male characters. — The external male genitalia and the penis valve (Figs 7- 8) are illustrated here for the first time (paramere, cuspis and digitus were de- scribed previously but not illustrated). The volsella has distinctive long setae on the cuspis apex (Fig. 7). The penis valve has a subapical tooth, more distant from the apical tooth than in males of Protophotopsis (see Figs 11-14, Cambra and Quintero 1997). Material examined (56 males, 5 females). — All males were captured with Malaise traps (B. Garcete coll.) and females with pitfall traps (T. Delsinne coll.). PARAGUAY: Boqueron Depart- ment: Parque Nacional Teniente Enciso, Admin- istracion, 239 m (21 12' S, 61 39' W) 16-19 Sep 2003, 20 males [MIUP, DJBC]; same loc, 20-24 Mar 2004, 6 males [MIUP]; same loc, 23-26 Sep 2004, 3 males [MIUP]; Siracua, 275 m (21 02' S, 61 45' W) 20-22 Sep 2003, 21 males [MIUP, MCZ, AMNH, ISNB, DJBC]; Estancia Maria 272 Journal of Hymenoptera Research Vicenta, 235-244 m (20 55' S, 61 23' W) 26-30 Sep 2004, 5 males [MIUP]; TransChaco, Mister Long, (20°35' S, 62 02' W), 17 Sep 2003, 1 female [MIUP]; Parque Nacional Teniente Enciso, TransChaco, 23-25 Sep 2004, 3 females [MIUP, ISBN, DTBC]; same data but 24-25 Sep 2003, 1 female [MCZ]. Presidente Hayes Department: Reserva Tinfunke, La Verde, 146 m (23 56' S, 69°29' W) 29 Nov-1 Dec 2003, 1 male [MIUP]. Variations. — Female frons dark reddish brown to black; tergum IV varying from totally reddish brown to black or the lateral areas black with reddish brown in the middle. Males from Paraguay are identical to the holotype, except that the propodeal lateral area is rugose on the holotype, but punctate with smooth areas to rugose- punctate or totally rugose in specimens from Paraguay. We consider this variation to be size-related: male rugosity increases with body length. In addition, larger males have the notauli nearly obsolete (same as the holotype), but notauli are absent in smaller males. Total length, females: 3.5- 5 mm; males: 4-7 mm. Distribution. — Paraguay and Argentina. AUotilla gibbosa was previously known only from the holotype from Cordoba, Argen- tina. PHYLOGENETIC ANALYSIS T/7A77. — To test the subtribal position of AUotilla (currently included in the subtribe Pseudomethocina, Brothers 1975), and to recognize its phylogenetic affinities, we selected as the outgroup the following two genera: Timulla (Mutillini) and Dasy- labris (Dasylabrini, genus not present in America); as ingroup taxa, we selected the following 17 Sphaeropthalmini genera, with mainly South American distributions and fully winged males: ten Pseudometho- cina (Euspinolia, Tallium, Atillmu, Calomu- tilla, Horcomutilla, Pseudomethoca, Hoplo- crates, Pappognatha, Hoplomutilla and AUo- tilla), females with head transversely sub- quadrate, broader than the mesosoma, genal carina present (except Euspinolia and Tallium), first metasomal segment sessile, evenly merging with second; and seven Sphaeropthalmina (Nanotopsis, Pro- tophotopsis, Reedomutilla, Scaptodactyla, Li- maytilla, Suareztilla and Limaytilla), females with head nearly round, narrower than the mesosoma, and genal carina absent. Characters. — Twenty-three binary and multistate characters of adult male (M) and female (F) external morphology and male genitalia were coded for analysis; all were treated as unweighted and unor- dered. No autapomorphies were used. The character matrix used is given in Table 1. The following characters were employed for cladistic analysis: Head: 2. 3. 4. 5. 7. 8. Head shape (F): 0 — small, almost round, not broader than mesosoma; 1 — trans- versely subquadrate, large, distinctly broader than mesosoma. Head (M, F): 0 — without large conical projection ventrally; 1 — with large conical projection ventrally. Scrobal carina (F): 0 — present; 1 — absent. Genal carina (F): — absent; 1 — present. Mandible basal ventral margin (M, F): 0 - with constriction; 1 — with broad lamellate projection; 2 — almost straight. Antennal tubercle (F): 0 - without lamellate projection; 1 — with lamellate projection on anteromedial surface. Antenna (F): 0-12-segmented; 1-13-seg- mented. Ocelli (M): 0 — small (diurnally active); 1 — large (nocturnally active). Mesosoma: 3. 4. 5. Dorsum of mesosoma (F): 0 — longer than broad, sometimes as broad as or slightly broader than long; 1 — distinctly broader than long. Shape of mesosoma (F): 0 — subrectangular; 1 — violin- shaped, strongly constricted at the propodeal spiracles; 2 — pyriform. Axilla of mesonotum (M): 0 — not expand- ed; 1 — expanded posterolateral^ as a rect- angular or acute protruberance. Scutellar scale (F): 0 — present; 1 — absent. Notauli (M): 0 — present; 1 — absent. Volume 15, Number 2, 2006 273 Table 1. Data matrix for the 23 characters used in the phylogenetic analysis Characters Taxon 0 1 9 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Timulla* 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ( ) 0 0 0 0 0 0 0 Dasylabris* 0 0 1 0 0 0 0 0 0 2 1 0 1 2 0 ( ) 1 2 0 0 0 0 0 Euspinolia 1 0 1 0 2 1 0 0 0 0 0 0 2 0 1 0 1 0 0 1 0 0 Tallium 1 0 1 0 - - 0 0 0 0 0 0 0 2 0 I 0 1 0 0 1 - 0 Allotilla 0 0 1 0 2 1 0 0 1 2 0 - 0 0 I 0 1 0 0 1 0 0 Nanotopsis 0 0 1 0 2 0 0 0 0 0 0 1 0 0 I 0 1 1 1 1 0 0 Protophotopsis 0 0 1 0 2 0 0 0 0 0 0 1 0 0 I 0 1 1 1 1 0 0 Scaptodactyla 0 0 1 0 2 1 0 1 1 2 0 0 0 0 I 0 2 0 0 0 1 0 Atillum 1 1 0 1 - 0 1 0 0 1 0 0 0 I 0 0 0 0 0 0 0 Hoplocrates 1 1 0 1 1 0 1 0 0 1 0 0 1 ' [ 1 1 0 0 0 0 0 Hoplomutilla 1 0 0 1 2 0 0 0 0 1 0 1 1 [ 1 0 - 0 0 0 0 Pappognatha 1 0 0 1 1 0 0 0 0 2 0 1 1 [ 1 1 1 1 0 0 0 Pseudomethoca* 1 0 0 1 2 0 0 0 0 1 0 1 o : I 0 0 0 0 0 0 0 Calomutilla 1 0 0 1 2 0 0 0 0 1 0 1 0 ' I 0 0 1 1 0 0 0 Horcomutilla 1 0 - 1 2 0 0 0 0 1 0 1 o : I 0 0 0 0 0 0 0 Xystromutilla 0 0 1 0 2 0 0 0 0 2 0 0 0 0 ' [ 1 2 1 1 1 1 0 Reedomutilla 0 0 1 0 2 0 0 0 0 2 1 1 0 0 ] I 1 2 0 0 0 0 1 Suareztilla 0 0 0 0 2 0 0 0 0 2 1 1 2 0 ] I 1 2 0 0 0 0 1 Limaytilla 0 0 1 0 0 1 0 1 1 2 0 0 0 0 1 0 2 0 0 1 1 0 (- = not applicable or with intragroup variation) *Only the type species examined: Timulla dubitata (Smith), Pseudomethoca frigida (Smith) and Dasylabris {Dasylabris) maura (Pallas). 6. Mesosternum posterior margin (F): 0 — with short triangular or spiniform process; 1 — with large truncate laminate process between and over inner margins of poste- rior coxae; 2 — obsolete spiniform or tri- angular process. 7. Tarsomere 5 (M, F): 0 — apex not produced above base of claws; 1 — apex produced above base of claws, forming lamellate plate. 8. Pterostigma (M): 0 — slightly sclerotized, membranous or absent; 1 — heavily sclero- tized. Metasoma: Tergum I (F): 0 — broad, sessile with second apically; 1 — nodose, dis- ciform or petiolate, not sessile with second. Tergum I (M): 0 — broad, sessile with second apically; 1 — subsessile-campanu- late at apex; 2 — nodose or petiolate with distinct lateral constrictions apically. Tergum VI (F): 0 — defined by lateral carinae; 1 — not defined by lateral carinae. Tergum VI (F): 0 — mostly flattened, broad and generally totally sculptured (punctate, 2. 3. 4. rugose, granulate or striate); 1 — more convex, not broad, with a medial smooth area or entirely smooth. 5. Felt line on sternum II, lateral margin (M): 0 — absent; 1 — present. 6. Setae, 40 X (M, F): 0— only simple; 1— both simple and plumose. 7. Parameres (M): 0 — scarcely recurved api- cally or almost straight; 1 — apex sharply and strongly recurved apically. Characters of male genitalia not exam- ined in all genera but considered poten- tially useful in a future phylogenetic analysis of the Sphaeropthalminae are: shape of cuspis, digitus and penis valve. The following characters were examined but not used because these characters need a more detailed study to determine if intragroup variation modifies them: length of antenna; eye size; sculpture of head; number of teeth on apex of mandibles; propodeum teeth; length of pterostigma and second submarginal cell. The following characters were examined but not used because they probably repre- 274 Journal of Hymenoptera Research sent autapomorphies in the present analy- sis: posterolateral tubercles on vertex, present in some females of Dasymutilla, Traumatomutilla and Cephalomutilla; mand- ibles covered with dense, short pubes- cence, an autapomorphy of Pappognatha; humeral angle of pronotum produced as a hook-like tooth, an autapomorphy of Gurisita females (males unknown); posteri- or coxa with tooth apically on inner margin, an autapomorphy of Vianatilla females; tibia 2 with one apical spur, an autapomorphy of Acanthophotopsis males; males apterous or brachypterous, Morsyma (apterous), Myrmilloides and Stethophotopsis (brachypterous), Dasymutilla (rarely bra- chypterous); tergum II with arcuate trans- verse band of dense, curled setae and slight integumental ridge at anterior margin of band, in most Dimorphomu- tilla females; female with felt line on sternum II, autapomorphy of Patquiatilla; sternum II with anteromedian seta-filled pit, in some males of Dasymutilla and Traumatomutilla; tegula elongated to or beyond the level of transscutal arti- culation, autapomorphy for Timulla males; eye inner margin deeply and abruptly notched, autapomorphy for Timulla males. A heuristic search of trees derived from parsimony analysis was carried out using NONA version 2.0 using WinClada version 1.00.08 (Nixon 2002), resulting in four cladograms. We preferred two of these minimal-length cladograms (Figs 1-2), see Results and Discussion. The following options were used: maximum trees to keep = 1000; number of replications (mult*N) = 1000; starting trees per rep (hold/) = 100; random seed = 1000; unconstrained search; search strategy of multiple TBR + TBR (mult* max*). RESULTS AND DISCUSSION The female of Allolilla gibbosa was recog- nized based on morphological similarities to the known male, mainly the distinctive and peculiar inflation and broadening of HI H^ Atlllum Hoploc rates Pseudomethoca Horcomulilla Calomutilla Hoplomutilla Pappognatha Eusplnolla — Talhum Nanotopsls Protophotopsis Allotilla Us-[T Scaptodactyla Llmaylllla - Xystromutilla - Reedomutilla - Dasylabrls - Suareztilla — Atlllum I Psoudomcthoca ' Horcomutllla i — Calomutilla Figs 1-2. Preferred minimum-length cladograms of four trees derived from heuristic analysis of the character data presented in Table 1 using ratchet (WINCLADA) and mult* (NONA). Tree length = 62; consistency index = 0.43; retention index = 0.66. Synapomorphies are show by the black circles. the mesosoma, and because these were the only female and male Sphaeropthalminae collected from the same geographic area, Paraguay's Chaco, that we recognized as belonging to a genus whose female was not known. The relationships between American mutillid genera (none is known to be present outside of America, except for Timulla) and those of Africa and Australia are almost totally unknown, although different but closely related genera are involved. Vicariant biogeographical hy- potheses and their common ancestral geo- Volume 15, Number 2, 2006 275 Figs 3-4. Allot ilia gibbosa, female. 3. Head, dorsal view. 4. Mesosoma, postero-dorsal view. Abrevia- tions: LP = lamellate projection of antennal tubercle, M = metanorum, P = pronotum, Pr = propodeum, S = scape. .,-: ■••■ Wm— ■■ — > — * Figs 5-6. Allotilla gibbosa, female metasoma, dorsal view. 5. Tergum 1 and basal part of tergum 2. 6. Tergum 6. Figs 7-8. Allotilla gibbosa, male genitalia. 7. Para- meres dorsal view. 8. Penis valve, lateral view. Abbreviations: P = paramere; ST = subapical tooth; VS = volsellar setae. graphic areas suggest close evolutionary relationships with South American genera. Some genera widely distributed in the Americas, such as Sphaeropthalma and Traumatomutilla, cannot be included in the present cladistic analysis because they are as yet poorly defined. The heuristic analysis resulted in four cladograms, the two preferred minimal- length cladograms (Figs 1-2) postulate a sister relationship of Allotilla with Scapto- dactyla and Li nun/ 1 ilia. The other two cladograms postulate a less parsimonious, complex relationship for the selected South Amerian taxa and, following Occam's razor, were not accepted: Allotilla + [(Scap- todach/la + Limaytilla) + (((Xystromutilla + ((Reedomutilla + (Dasylabris + Suareztilla)))]. These preliminary cladograms suggest than Allotilla is more closely related to genera in the subtribe Sphaeropthalmina than to those in the Pseudomethocina. At present, we do not know of any unique morphological characters to separate these two subtribes. The biology of Allotilla gibbosa is un- known, but the black integument, moder- ately sized eyes, and very small ocelli of the males suggest that they are diurnal. The following morphological characters of 276 Journal of Hymenoptera Research AUotilla females indicate that they spend most of their lives underground: small, flattened eyes, relatively short legs with a fore tarsal rake (used to excavate soil), and a mostly reddish-brown integument. Extensive visual samplings carried out during daylight hours in Teniente Enciso National Park did not yield any AUotilla females from the ground or on the sparse vegetation; females were collected only with pitfall traps. These capture data lend support to the postulated underground, burrowing habits of the females. Probably they parasitize small, underground-nesting aculeates. Females of the here-recognized closely related sphaeropthalmine genera, Scaptodactyla and Limaytilla, sister genera to AUotilla, have a similar morphological habitus, suggesting that they have similar hypogeal lives and burrowing activities. A morphology-based phylogenetic anal- ysis of 18 mostly South American mutillid genera and one from outside America (Dasylabris) permits us to construct a hypo- thetical scenario of biogeographic diver- gences. The vicariant event that divided the population of the common ancestor of the taxa presented in Table 1 was the uplifting of the Andean high mountain range. This uplifting event was followed by ecological divergence of the two An- dean regions: the elongated West costal region, draining into the Pacific Ocean, isolated from a more extensive and ecolog- ically diverse Eastern region. The ancestor population of the present-day Euspinolia, an almost exclusively Chilean and Peru- vian taxon, was isolated on the Pacific Andes slopes. The common ancestor of both Tallium and Limaytilla + Scaptodactyla + AUotilla was relegated to the region east of the Andes. The climatic and vegetational changes brought about by the Andean uplift, particularly the desertification of the Chaco region, was the driving selective force that caused the diversification and evolutionary split between Tallium and the sister genera of AUotilla + Scaptodactyla + Limaytilla. Spichiger et al. (2004), investigating the geographical zonation in South America of 32 common tree species encountered in Paraguay, found that the xeromorphic forests of the Chaco area act as an edaphic barrier to many species that are centered in northern Argentina. The genus AUotilla is reported only from northern Argentina and Paraguay. If additional sphaeropthalmine genera eventually are included in the data matrix presented here, the larger data matrix might provide a better resolution in the phylogenetic analysis. ACKNOWLEDGMENTS Our sincere thanks to Thibaut Delsinne, ISBN, for sorting and sending us all the mutillids from his pitfall traps from Paraguay. We are grateful to Philip D. Perkins, MCZ, for the loan of the holo- type of A. gibbosa; Bolivar R. Garcete B., Museo Nacional Historia Natural de Paraguay for col- lecting mutillid specimens; Annette Aiello, Smithsonian Tropical Research Institute, and Denis J. Brothers, University of KwaZulu-Natal, South Africa for reading the manuscript and for provid- ing us with very helpful suggestions to improve its contents; Alcides Muhoz, University of Panama, for SEM work. LITERATURE CITED Brothers, D. J. 1999. Phytogeny and evolution of wasps, ants and bees (Hymenoptera, Chrysidoi- dea, Vespoidea and Apoidea). Zoologica Scripta 28: 233-249. Cambra. T. R. A. and D. Quintero, A. 1997. A revi- sion of Protopmotopsis Schuster (Hymenoptera: Mutillidae). journal of Hymenoptera Research 6: 263-272. Fritz, M. A. 1998. Mutillidae. Pp. 445-451 in: Morrone, J. J. and S. Coscaron, eds. Biodiversidad de Artropodos Argentinos, Ediciones Sur. Nixon, K. C. 2002. Winclada version 1.00.08. Published by the author, Ithaca, NY. Schuster, R. M. 1949. Contributions toward a Mono- graph of the Mutillidae of the Neotropical Region. III. A Key to Subfamilies Represented and the descriptions of several New Genera (Hymenoptera). Entomologica Americana (n.s.) 29: 59-140. Spichiger, R., C Calenge, and B. Bise. 2004. Geo- graphical zonation in the Neotropics of tree species characteristic of the Paraguay-Parana Basin. Journal of Biogeograpln/ 31: 1489-1501. J. HYM. RES. Vol. 15(2), 2006, pp. 277-285 Parasitism and Sex Ratio of the Bedeguar Gall Wasp Diplolepis rosae (L.) (Hymenoptera: Cynipidae) in Sicily (Italy) Maria Concetta Rizzo and Bruno Massa Dipartimento SENFIMIZO, Universita di Palermo, viale delle Scienze 13, 1-90128 Palermo Italy; email: zoolappl@unipa.it Abstract. — The Diplolepis rosae gall community is analysed in Sicily (Italy), based on collections totalling 82 galls from 12 sites from which 1,026 adult insects were obtained. The gall wasp exits in March-June from galls induced the previous year. On average 5.6 D. rosae individuals per gall were obtained, corresponding to 44.8% of all the emerged insects. We obtained 4.3% of D. rosae males overall, the highest figure found till now for the cynipid overall in Europe, where male D. rosae are visually rarer, and the first record of them for the Mediterranean area. No inquilines were found, and, consequently, none of their specific parasitoids. However, Eurytoma rosae, generally considered as a specific parasitoid of the inquiline Periclistus brandtii, was obtained, together with the polyphagous Exeristes roborator. In all, seven parasitoid species emerged from the galls: four of them, Orthopelma mediator, Torymus bedeguaris, Exeristes roborator, and Eupelmus urozonus, start to emerge together with D. rosae, while Glyphomerus stigma, Pteromalus bedeguaris, and E. rosae, have their maximum peaks later in the year. All the cited species, except for E. roborator, showed a second peak of emergence in September, when D. rosae is absent. Parasitization ranged from 12.5 to 100%, reaching more than 70% in 66.6% of the samples, but it was rather low (30.5%) when males were present, even though there was no overall correlation between parasitization and D. rosae sex ratio. Statistical analysis showed however that all the parasitoid species (except for males of T. bedeguaris and P. bedeguaris) are longer (which we take to signify larger) than D. rosae males, and neither the size nor the sex ratio of parasitoids differed statistically depending on the presence of male D. rosae. Diplolepis rosae (Linnaeus, 1758) is a pa- laearctic species, introduced in the nearctic region (Shorthouse and Ritchie 1984). It is widespread in peninsular Italy (Pagliano 1995), from where it has been known since at least 1600 (Tuscany: Pagliano et al. 1997), and in Sicily (De Stefani Perez 1887, as Rhodites rosae L.). Its multilocular galls have been recorded on 18 species of Rosa in Italy (De Stefani Perez 1887, 1902, 1905, Pelliz- zari Scaltriti 1988, Pagliano et al. 1997), and on 32 species in Europe and the Mediter- ranean basin (Houard 1908, Nieves-Aldrey 2001), though the gall wasp seems to prefer those species that are taxonomically allied in the Section Caninae (Randolph 2005), and particularly R. canina L. (Schroder 1967). Although D. rosae reproduces itself by amphitokous parthenogenesis (Beauvi- sage 1883, Callan 1940), males are repro- ductively inactive (Stille and Davring 1980) and of spasmodic occurrence probably due to female infection with Wolbacliia bacteria, which is also common in other Diplolepis species (Plantard et al. 1998, 1999). Larvae grow inside the galls and overwinter in diapause as prepupae; they pupate the following spring and emerge soon after (Schroder 1967). Diplolepis rosae galls host a large insect community that has been extensively studied in many countries (Randolph 2005, and references therein). Considering only Hymenoptera parasi- toids, Noyes (2003) lists 29 species of Chalcidoidea, and numbers increase when parasitoids belonging to other superfami- lies are included (Fulmek 1968). However, several of the names listed by Fulmek (1968) are invalid or misidentifications or probably erroneously associated with D. 278 Journal of Hymenoptera Research Table 1. Differences in the number of males and in the percentage of parasitization in different groups of gall samples. Total number of Average number Sex ratio (No. % of D. rosae Diplolepis rosae of D. rosae emerging males/No. males as total of Average emerging per gall females) emerged insects parasitization Sicilian galls collected 394 5.18 (min.: 0; 0 0 57.6% (min.: 12.5%; from 1992 to 2005 max.: 92) max.: 100%) (n = 76) Sicilian galls collected in 66 11 0.43 21 30.5% 1965 (n = 6) Hungarian galls collected 42 21 0.31 15.6 34.3% in 2001 (n = 2) rosae (Askew pers. comm.). Despite this deep knowledge of biology and ecology of the bedeguar gall wasp all over Europe, little or nothing is known about it in Italy (cf. Randolph 2005). The aim of this study is to rectify that situation, with special emphasis on aspects needing further in- vestigation (D. rosae sex-ratio and parasit- ism), as suggested by Randolph (2005). MATERIALS AND METHODS A total of 82 galls were collected in the following localities in Sicily (Italy) (in brackets the number of galls per sample): Madonie Mts. (Palermo), 13.IV.65 (6); Ma- donie, Piano Porno 2. 11.96 (1); Madonie, Piano Zucchi 18.11.96 (1); Madonie, Piano Cervi 30.V.96 (1), 31.V.01 (1); Madonie, Fosso Carina 6.VI.99 (1); Madonie, Castel- buono 13.X.96 (1); Madonie, loc. Vicaretto 27.X.96 (10); Ficuzza (Palermo) 16.IH.97 (5), 7.IX.97 (9), 7.II.99 (3), 28.111.99 (1), 27.VI.99 (2), 19.111.00 (8), 6.II.05 (2); Contessa En- tellina, loc. S. Maria del Bosco (Palermo) XI.92 (1); Bivona (Agrigento) 29.X.96 (16); Nebrodi Mts. (Messina), Biviere di Cesaro 2.XI.96 (1), 17.X.99 (2); Pergusa (Enna), 13.XI.04 (10). One additional sample was collected in Hungary (Ferto-Hansag Na- tional Park, Koszeg, 16.V.01, 2 galls). Galls of D. rosae examined during this study were collected on Rosa canina L., Rosa sempervirens L. and Rosa sp. Galls were placed separately in single small cages, at room temperature in the Palermo labora- tory. They were kept for at least one year after the collecting date. In this way different emergence dates could be re- corded for both gall inducers and para- sitoids. Each emerged individual was counted, mounted and labelled. Parasitoid identification was carried out by M.C. Rizzo, if not differently reported in the Acknowledgements. Total body length, from the apex of the head to the tip of the abdomen, of all D. rosae adults and 30 specimens (males and females) of each parasitoid species was measured with the aid of a binocular microscope. Parasitoid length does not include the exserted ovipositor sheaths. Statistics were per- formed in STATISTIC A (StatSoft 2003). RESULTS AND DISCUSSION Sex ratio and phenology of Diplolepis rosae. — A total of 460 individuals of D. rosae was obtained from the 82 galls examined, corresponding on average to 5.6 cynipids per gall and to 44.8% of all the emerged insects. However, while no males of D. rosae emerged from the 76 galls of 1992-2005, 21% males emerged from those collected in 1965 (Table 1). Also, in the latter sample, the average number of emerged D. rosae per gall was relatively higher (11 versus 5.18 of the other sam- ples), while the percentage parasitization was somewhat lower than in the other samples on average (30.5 versus 57.6%) (Table 1). Table 1 also includes an occa- sional sample from Hungary, which also yielded a considerable number of males 419 6 1.5 6007 46 0.8 104 2 1.96 601 1 0.17 1264 51 4.2 3425 137 4 671 7 1 127 1 0.8 700 0 0 2684 0 0 2000 4 <0.2 388 6 1.5 249 0 0 49 0 0 460 20 4.3 66 20 30.3 42 10 23.8 Volume 15, Number 2, 2006 279 Table 2. Diplolepis rosae sex ratio known until now (from Randolph 2005, except for the last five sets of data). Total number of D. rosae Country and Author individuals emerging No. of males % of males USA: Kinsey (1920) ENGLAND: Callan (1940) ENGLAND: Blair (1943) ENGLAND: Niblett (1949) NORTHERN ENGLAND: Askew (1960) DENMARK: Hoffmeyer 0925) GERMANY: Adler & Straton (1894) GERMANY: Weidner (1956) FRANCE: Hardouin (1943) FRANCE, SPAIN, SWITZERLAND, AUSTRIA, SOUTH GERMANY: Schroder (1967) ENGLAND, ITALY, DENMARK, SWEDEN: Picard (1926) SOUTHERN SWEDEN: Nordlander (1973) SPAIN: Nieves-Aldrey (1981) SPAIN: Pujade Villar (1983) SICILY (ITALY): this study (the whole Sicilian data set) SICILY (ITALY): this study (Sicilian sample of 1965) HUNGARY: this study (one sample, 2001) (15.6% of the total insects); the number of Our results, however, lead us to reconsider cynipids emerged per gall (21) and per- the existence of a latitudinal gradient; centage parasitization (34.3%) reflect those alternatively, we suppose that emergence of the Sicilian sample of 1965 (cf. Table 1). of D. rosae males could depend on the Also considering Sicilian samples on the combined effects of environmental and whole, D. rosae male percentage obtained biological factors (such as female infection during this study is among the highest with Wolbachia bacteria and its spatial and reported in the literature (4.3%), and it is temporal diffusion), as other authors have very much higher than the value given by already hypothesized (Shorthouse and Picard (1926) for Italy and other countries Ritchie 1984, and references therein, Plan- cumulatively (0.2%) (Table 2). High male tard et al. 1998, 1999). figures for single samples, similar to those The cynipid phenology recorded during recorded by us (Sicilian sample of 1965: this study in Sicily matches what is already 30.3% of the total number of cynipids; known for the species in North and central Hungarian sample of 2001: 23.8%; cf. Europe (Callan 1940, Schroder 1967, Stille Table 2) are however reported by other and Davring 1980, Randolph 2005), even if authors (38%: Walsh 1924, 13%: Askew's its cycle starts earlier, probably because of unpublished data for a sample from the warmer weather: the majority of D. Yorkshire (northern England), in Randolph rosae adults exited galls in March (Fig. 1), 2005), even if never in the Mediterranean with a few individuals continuing to area. Askew (1960) suggested the presence appear till the end of the spring as of a decreasing latitudinal gradient recorded also by Schroder (1967) in central (North>South) to explain the distribution Europe. Male D. rosae of the 1965 Sicilian of males of D. rosae in northern Europe, sample were obtained in May and June, at and data previously collected in southern the same time as females, in contrast to the Europe (reporting the complete absence of finding of Callan (1940), who recorded males) seemed to confirm his hypothesis male appearance distinctly earlier than that (Nieves-Aldrey 1981, Pujade Villar 1983). oi females. During this study neither the 280 Journal of Hymenoptera Research N o 400 □ emerged males ■ emerged females v vi vn vm ix x xi xn Months Fig. 1. Seasonal appearance of adults of D. rosae in Sicily. inquiline Periclistus brandtii (Ratzeburg) nor its parasitoids were found. Altough sometimes a common inhabitant of D. rosae galls, this species has a variable distribu- tion in central and northern Europe and seems rarer in Mediterranean countries (Nieves-Aldrey 1981, Pujade Villar 1983, Randolph 2005). In Italy it is known only from the north (Pagliano 1995). No other inquiline species has been reared during this study. Parasitoid complex. — Although a large number of parasitic wasps is associated with the bedeguar gall wasp, a typical list of parasitoids comprises about a dozen species (Schroder 1967, Askew 1984, Ran- dolph 2005). During this study a total of 566 Hymenoptera, mainly belonging to the most common D. rosae parasitoid species, was obtained from the Sicilian galls as follows: 374 Orthopelma mediator (Thun- berg) (= O. luteolator Gravenhorst) (Ich- neumonidae), 98 Torymus bedeguaris (Lin- naeus) (Torymidae), 52 Glyphomerus stigma (Fabricius) (Torymidae), 23 Pteromalus be- deguaris (Thomson) (= Habrocytus bede- guaris Thomson) (Pteromalidae), 12 Exer- istes roborator (Fabricius) (Ichneumonidae), 5 Eurytoma rosae Nees (Eurytomidae), and 2 Eupelmus urozonus Dalman (Eupelmidae). They represented 55.2% of all insects and their relative abundance is reported in Fig. 2. Orthopelma mediator is an endophagous species, largely known as a widespread parasitoid of D. rosae (Askew 1960, Nieves- Aldrey 1981, Redfern and Askew 1992, Randolph 2005, and references therein), previously unrecorded from Sicily (Scara- mozzino 1995, Noyes 2003). It was known as a parasitoid of D. rosae in Italy as early as 1600, since an unidentified Ichneumonidae, corresponding to it, is portrayed by Redi in an unpublished plate on D. rosae galls collected in Tuscany (Pagliano et al. 1997). In this study it was the most common (88.9% of the samples) and abundant parasitoid (66.1% of all parasitoids) (Fig. 2). O. mediator emerged from March onwards (Fig. 3), so that its phenology matches that of the gall inducer, as Nieves-Aldrey (1981) already observed in Spain, except in autumn when there was a late peak of emergences. Torymus bedeguaris is a holarctic species (Grissell 1995,' Noyes 2003), already re- corded in Italy (De Stefani Perez 1905, Pagliano 1995, Pagliano and Navone 1995, Noyes 2003), mainly known as an ectopar- asitoid of cynipid gall wasps belonging to the genus Diplolepis or of their inquilines (De Stefani Perez 1905, Askew 1960, Nieves-Aldrey 1981, Noyes 2003); it is also occasionally reported attacking O. mediator (Askew 1960, Schroder 1967). In Sicily T. bedeguaris represented 17.3% of all emerged parasitoids and its phenology overlaps that of D. rosae and O. mediator (cf. Figs 2 and 3). This trend is similar to that recorded by Nieves-Aldrey (1981) in Spain. Volume 15, Number 2, 2006 281 Diplolepis rosae 44.8% ' Torymus bedeguaris 7.3% Orthopelma mediator 66.1% Pteromalus bedeguaris 4.1% Exeristes roborator 2.1% Eupelmus Eurytoma urozonus 0.3% rosae 0.9% Fig. 2. Relative abundance of parasitoids of D. rosae in Sicily. Glyphomerus stigma is another holarctic species typically associated with the genus Diplolepis (Noyes 2003), mainly known as an ectoparasitoid of D. rosae and its inquilines (Blair 1943, Askew 1960, Nieves-Aldrey 1981), and occasionally as a hyperparasitoid of E. rosae (Redfern and Askew 1992). This is the first record for Sicily, although the species was previously known from peninsular Italy (Pagliano 1995, Pagliano and Navone 1995, Noyes 2003). Nieves-Aldrey (1981) reported it as the second most abundant parasitoid in Spain; in Sicily it reached only 9.2% of all parasitoids (Fig. 2). It appeared later in the year than the inducers and the previous parasitoid species (Fig. 3), which agrees with the findings of Nieves-Aldrey (1981) in Spain. Pteromalus bedeguaris is one of the two species of Pteromalidae typically associat- ed with the galls of D. rosae. It is a holarctic species, hitherto unknown in Italy (Pa- gliano 1995, Pagliano and Navone 1995, Noyes 2003), although De Stefani Perez (1905) listed many other Pteromalidae as N° 250 200 150 100 v1 vn vra K \ln ii l lis Orthopelma mediator Torymus bedeguaris "EZZEZZZZQ Glypitomerus stigma J~ | Pteromalus bedeguaris Exeristes roborator Eurytoma rosae Eupelmus urozonus » xn Fig. 3. Seasonal appearance of adults of parasitoids of D. rosae in Sicily. 2g2 Journal of Hymenoptera Research parasitoids of D. rosae in Sicily. Pteromalus C. inflexa were found during this study, bedeguaris is considered a parasitoid of and the five individuals of E. rosae, which several species of Diplolepis, but it is also emerged in September (Figs 2 and 3), may known to attack O. mediator, T. bedeguaris have developed at the expense of D. rosae, and G. stigma (Blair 1943, Redfern and in agreement with what Nieves-Aldrey Askew 1992, Noyes 2003), and occasionally (1981) recorded in Spain. The typical peak parasitizes Periclistus brandtii (Nordlander of emergences in late summer has been 1973, Noyes 2003). The phenology of P. considered by many authors (Blair 1945, bedeguaris in Sicily shows two emergence Niblett 1951, Claridge and Askew 1960) as periods, the first in late spring and the a precocious emergence from galls of the second in autumn (Fig. 3), similar to what same year. Nordlander (1973) in Sweden and Nieves- Finally, Eupelmus urozonus, a cosmopoli- Aldrey (1981) in Spain recorded. In the tan generalist (Pagliano 1995, Pagliano and Sicilian samples, P. bedeguaris amounted to Navone 1995, Noyes 2003) and occasional 4.1% of all emerged parasitoids (Fig. 2), parasitoid of D. rosae (Schroder 1967, while 22 individuals of this species (100% Noyes 2003) has been obtained during this of all parasitoids) emerged from the Hun- study: two individuals of this bivoltine garian sample in June. This species is species (Askew 1961) emerged in March recorded as the second most common and September. parasitoid in many countries of central Parasitism and parasitoid relationships with Europe (Randolph 2005, and references the gall inducer. — Parasitization ranged therein). The palaearctic Caenacis inflexa from 12.5 to 100%, reaching more than (Ratzeburg) (Pteromalidae) was not ob- 70% in 66.6% of the samples, in line with tained from any sample, being a specific other authors who found the bedeguar gall parasitoid of the inquiline P. brandtii wasp heavily parasitized (Schroder 1967, (Callan 1944), which itself was absent from Nordlander 1973, Stille 1984). In Sicily, the our samples. D. rosae parasitoid complex comprises two Exeristes roborator (Fabricius) is a pa- groups of species; namely, a first group, laearctic species, already known in Italy including O. mediator, T. bedeguaris, and E. (Scaramozzino 1995). It is a polyphagous roborator, whose phenology overlaps and ectophagous parasitoid of Lepidoptera, strictly follows that of the gall inducer, and Coleoptera and Hymenoptera larvae, a second group, with the prevailing species known only as a parasitoid of Biorhiza G. stigma and P. bedeguaris, which appears pallida (Olivier) among Cynipidae (Fulmek later in the year (Fig. 3). All the recorded 1968, Constantineanu and Pisica 1970). We parasitoid species, except for E. roborator, cannot definitely affirm here that it is showed a late peak of emergences in a primary parasitoid of D. rosae, even if autumn, when D. rosae is absent (Fig. 3). no inquiline species emerged from our This peak has often been reported by other samples. Twelve individuals of this species authors (see accounts above of single (2.1% of all parasitoids) exited from the parasitoid species) and explained, depend- galls in March (Figs 2 and 3). ing on species, as a precocious emergence Eurytoma rosae is another palaearctic (i.e. for E. rosae) or as a bivoltine cycle (i.e. species, already known in Italy (Pagliano E. urozonus). Our present data do not allow and Navone 1995, Noyes 2003), associated us to decide which of the two explanations with galls induced by Diplolepis and con- applies to O. mediator, T. bedeguaris, G. sidered as usually a specific parasitoid of stigma and P. bedeguaris. P. brandtii, less commonly of D. rosae (Blair Moreover, it seems that there is a nega- 1945, Claridge and Askew 1960). However, tive association between the presence of D. neither P. brandtii nor its specific parasitoid rosae males and parasitization (cf. Table 1), Volume 15, Number 2, 2006 283 Table 3. Summarized results for the one way ANOVA analysis (FD = 273; F = 49.67; p < 0.001) performed on the total length of male and female D. rosae and its parasitoids (small samples were excluded from analysis). D. rosae males are statistically shorter than all the parasitoid species, except for T. bedeguaris and P. bedeguaris males (in bold). D. rosae females D. rosae males O. mediator both sexes longer* T. bedeguaris males T. bedeguaris females G. stigma females P. bedeguaris males P. bedeguaris females shorter longer not different longer* longer not different longer shorter not different not different longer * Even if statistically longer, O. mediator and T. bedeguaris females overlap the maximum size of D. rosae females (Fig. 4), only O. mediator males being clearly longer. Moreover the two parasitoid species are much more slender than the massive D. rosae females, the latter probably being yet a suitable food source for them. the tested parasitoid species, except for T. bedeguaris and P. bedeguaris males (FD = 273; F = 49.67; p < 0.001) (Table 3 and Fig. 4). Given similar shapes of the species concerned, we take "shorter" to be a good indication of "smaller". From the ANOVA analysis parasitoid size did not differ between samples with or without males of D. rosae (cf. Fig. 4), and neither did their sex ratio differ (Wilcoxon Test; Z = 0.36; p = n.s.; fd = 4). However, our data are insufficient to ascertain whether D. rosae males represent an inadequate food re- source for parasitoid development and /or whether galls which contain them are avoided by parasitoids. It would be in- teresting to compare parasitization per- centages in other parts of Europe where good numbers of males of D. rosae are recorded. even if no correlation came out between parasitization and D. rosae sex ratio (r = -0.3; fd = 16, p = 0.21). Comparing total lengths of D. rosae adults (males and females) with total lengths of their com- monest parasitoids through a one way ANOVA analysis, we found that D. rosae males are statistically much shorter than all ACKNOWLEDGMENTS We thank very much R.R. Askew who revised the manuscript and an anonymous referee for his useful suggestions. We are also very grateful to Gavin Broad and Mark Shaw, whose advice greatly improved the last draft. For their kind collaboration in the identifi- cation of specimens we thank: Jose Luis Nieves Aldrey (Cynipidae), Klaus Horstmann (Ichneumoni- 5,5 5,0 4,5 4,0 3,5 3,0 2,5 2,0 1,5 'i i —i r™ ca It XTlj I o co O a o D LL. O Q CO O or q I [■$#'] | 1 | •!• ! X i hi T^~ 0 Mean I I ± Std Err ~1~ ± Std. Dev o <0 O and the features we have stud. antecostal suture (Fig. 26) is also observed ied need to be surveyed for a , le in Tnchogramma (Krogmann 2005), but not of the family According to Gibson et al. outside Chalcidoidea. A more comprehen- (1999)/ the diversity of Mymarommatidae sive survey of the occurrence of this trait is not sufficiently reflected by the current within Chalcidoidea is required to estab- classification that has all extant species lish its possible phylogenetic significance, lumped in Palaeomymar, a situation that The presence of well developed apodemes will probably change once the family has internally of the spiracle indicates that it is been comprehensively revised. Given the the only functional spiracle in the meso- small size of all Mymarommatidae and soma of P. anomalum. that it possibly occupies a key phylogenetic The slender, two-segmented petiole is position within Apocrita, detailed morpho- only observed in the extinct taxon Serphi- logical studies of additional species are tidae apart from in Mymarommatidae; recommended to test our inferences, based on this, it has been hypothesized Many of the putative autapomorphies of that they form a monophylum (cf. Gibson Mymarommatidae are reductional and 1986). The first petiolar segment of P. probably correlated with miniaturization anomalum essentially is a narrow, sclero- (e.g., the absence of the anterior thoracic tised tube with little indication of bound- spiracle and the metafurca). Furthermore, aries between individual sclerites. This being morphologically highly derived and condition occurs in many Apocrita, in- isolated confounds attempts to homologize eluding some Chalcidoidea, and is difficult features with those in other Apocrita (e.g., to interpret. The ventral longitudinal carina tne extensive fusion in the propectus, the extending along most of the first petiolar difficulties in establishing the identity and segment in P. anomalum is perhaps formed configuration of the prepectus). The sis- by the fusion of the lateral margins of the tergroup relationship to the Chalcidoidea second abdominal tergum (T2), indicating (also a highly autapomorphic taxon) seems that T2 is expanded ventrally and the the best supported hypothesis at present, sternum of this segment is reduced or but may Prove difficult to corroborate absent (Gibson in litt.). However, the setae further from morphological evidence situated anteroventrally just posterior to alone' Mymarommatidae has not been the anterior margin of the first petiolar represented in any phylogenetic analyses segment (Figs 28, 29) are probably homol- of Hymenoptera including molecular data, ogous with the sensillar patches situated not even the most recent comprehensive sublaterally on S2 in Apocrita that have °Anes (CamPbf et aL 2000' Dowto» and this sclerite separate from T2. This indi- Austin 2001). Sequence data might confirm cates that at least the anteriormost part of °r chaUen^ I™"™* Cu°nCePtS °f the P^" Ct ■ , r> 1 -ri , • logenetic affinity of these intriguing ani- S2 is present in P. anomalum. The continu- , & b . , , mals. ous transverse carina anteriorly on the first petiolar segment (Fig. 29) is also observed ACKNOWLEDGMENTS in many Proctotrupoidea, Cynipoidea, Pla- tygastroidea, Ceraphronoidea as well as in Jens-Willu-lm Janzen, Hamburg, Germany, and >-,, , . , . , . , ,. , , , Kiel] Arne Johansson, Naturhistoriska Riksmuseet, some Chalcidoidea (unpublished observa- Stockholm, Sweden/ kindly provided access to mate. ti°ns)- rial employed in this study. John Huber, CNCI, 306 Journal of Hymenoptera Research Ottawa, Canada provided useful comments on earlier versions of the paper. Gary Gibson, CNCI, Ottawa, and an anonymous referee commented extensively on the submitted version. LITERATURE CITED Bucher, G. E. 1948. The anatomy of Monodontomerus dentipes Boh., an entomophagous chalcid. Canadi- an Journal of Research Section D 26: 230-281. Campbell, B., J. M. Heraty, J.-Y. Rasplus, K. Chan, J. Steffen-Campbell, and C. Babcock. 2000. Molecu- lar systematics of the Chalcidoidea using 28S-D2 rDNA. Pp. 59-73 in: Austin, A. D. and M. Dowton, eds. Hymenoptera — evolution, biodiversity and biological control. CSIRO publishing, Colling- wood. Debauche, H. R. 1948. Etude sur les Mymarommidae et les Mymaridae de la Belgique (Hymenoptera, Chalcidoidea). Memoires du Musee Royal d'Histoire Naturelle Belgique 108: 1-248. Dowton, M. and A. D. Austin. 2001. Simultaneous analysis of 16S, 28S, COl and morphology in the Hymenoptera: Apocrita — evolutionary transi- tions among parasitic wasps. Biological Journal of the Linnean Society 74: 87-111. Gibson, G. A. P. 1985. Some pro- and mesothoracic structures important for phylogenetic analysis of Hymenoptera, with a review of terms used for the structures. The Canadian Entomologist 117: 1395-1443. Gibson, G. A. P. 1986. Evidence for monophyly and relationships of Chalcidoidea, Mymaridae, and Mymarommatidae (Hymenoptera: Terebrantes). The Canadian Entomologist 118: 205-240. Gibson, G. A. P. 1999. Sister-group relationships of the Platygastroidea and Chalcidoidea (Hymenop- tera)— an alternate hypothesis to Rasnitsyn (1988). Zoologica Scripta 28: 125-138. Gibson, G. A. P., J. M. Heraty, and ]. B. Woolley. 1999. Phylogenetics and classification of Chalcidoidea and Mymarommatoidea — a review of current concepts (Hymenoptera, Apocrita). Zoologica Scripta 28: 87-124. Gibson, G. A. P. and J. T. Huber. 2000. Review of the family Rotoitidae (Hymenoptera: Chalcidoidea), with description of a new genus from Chile. Journal of Natural History 34: 2293-2314. Grimaldi, D. and M. S. Engel. 2005. Evolution of the insects. Cambridge University Press, New York. Heraty, J. M., J. B. Woolley, and D. C. Darling. 1994. Phylogenetic implications of the mesofurca and mesopostnotum in Hymenoptera. Journal of Hy- menoptera Research 3: 241-277. Heraty, J. M., J. B. Woolley, and D. C. Darling. 1997. Phylogenetic implications of the mesofurca in Chalcidoidea (Hymenoptera), with emphasis on Aphelinidae. Systematic Entomology 22: 45-65. Krogmann, L. 2005. Molekulargenetische und mor- phologische Untersuchungen zur systematischen Stellung der Pteromalidae innerhalb der Chalci- doidea (Hymenoptera: Apocrita). Ph.D.-thesis, University of Hamburg, 204 pp. http://www. sub.uni-hamburg.de/opus/volltexte/2005/ 2652/ Ronquist, F., A. P. Rasnitsyn, A. Roy, K. Eriksson, and M. Lindgren. 1999. Phylogeny of Hymenoptera: A cladistic reanalysis of Rasnitsyn's (1988) data. Zoologica Scripta 28: 13-50. Vilhelmsen, L. 2000a. Cervical and prothoracic ske- leto-musculature in the basal Hymenoptera (In- secta): comparative anatomy and phylogenetic implications. Zoologischer Anzeiger 239: 105-138. Vilhelmsen, L. 2000b. Before the wasp-waist: compar- ative anatomy and phylogenetic implications of the skeleto-musculature of the thoraco-abdominal boundary region in basal Hymenoptera (Insecta). Zoomorphology 119: 185-221. J. HYM. RES. Vol. 15(2), 2006, pp. 307-316 Nesting Biology of a Tropical Myrmicine Ant, Myrmicaria arachnoides (Formicidae), in West Java, Indonesia Bakhtiar Effendi Yahya and Seiki Yamane (BEY, SKY) Department of Earth and Environmental Sciences, Faculty of Science, Kagoshima University, Korimoto-1, Kagoshima, 890-0065 Japan, BEY email: bakhtiareffendi@yahoo.co.uk, SKY email: sky@earth.sci.kagoshima-u.ac.jp (BEY) Institute for Tropical Biology and Conservation, Universiti Malaysia Sabah, Kota Kinabalu, Sabah, Malaysia Abstract. — Nesting biology of a myrmicine ant, Myrmicaria arachnoides, was studied on Java island, Indonesia. The colonies were polydomous and polygynous. Nests of a colony were located close to each other on different leaves of a tree. The number of dealated queens was positively correlated with the numbers of nests, adult workers and immatures, but not with the numbers of males and new queens present. Adaptive aspects of nesting site, polygyny and polydomy are discussed. The genus Myrmicaria is widely distrib- uted in the Old World tropics, i.e., South- east Asia, South Asia and tropical Africa. Emery (1922) sorted Myrmicaria species into two principal groups based on mor- phology, i.e., the M. brunnea group and M. arachnoides group, and briefly mentioned that species of the former construct huge nests underground while those of the latter construct carton nests on trees. Our ob- servations have confirmed this for most of the Southeast Asian forms (also see Kar- avaiev 1935). However, there are few detailed studies on the nesting habits of Myrmicaria. In Cameroon, Africa, one species, M. opaciventris Emery, belonging to the M. brunnea group, has been in- tensively studied for its ecology/bionomics including nesting habits (Kenne and De- jean 1999, Kenne et al. 2000, 2001). This species has polydomous and polygynous colonies in the soil. Interconnected nests are built and trenches and tunnels are constructed as underground trails connect- ing nests. Species of the M. arachnoides group also often construct polydomous and polygynous colonies but on vegetation (Karawajew 1935, this study, Bakhtiar and Yamane unpubl.). This may allow them to develop relatively large colonies similar to Polyrhachis species nesting in similar situa- tions (R. Kohout pers. com.). Research on colony growth in eusocial insects has focused on the relationship between colony size and productivity. At each growth period, colonies make invest- ment decisions about whether to produce more workers and grow larger, or rather to invest that energy in reproductive output (Billick 2001). Because colony size is often the most important factor determining reproductive output (Odum and Pontin 1961, Michener 1964, Fowler 1986, Tschin- kel 1993, Savolainen et al. 1996), maximiz- ing the long-term size of the colony is an important component of colony fitness (Oster and Wilson 1978). Colony size is known to be related with queen number and also to affect caste/sex composition in ant nests. In this respect nesting behaviour of the M. arachnoides group is oi special interest. 308 Journal of Hymenoptera Research Little has been studied on the colony structure of the Southeast Asian Myrmi- caria arachnoides F. Smith (but see Kara- vaiev 1935). During our study on the taxonomy of Oriental Myrmicaria we ob- tained relatively good samples of M. arachnoides colonies on Java island, Indo- nesia. We report here the nesting sites, colony size, and reproductive output of these colonies. MATERIALS AND METHODS The taxonomy of Myrmicaria is still unresolved, particularly the status of in- fraspecific forms of some species. Myrmi- caria arachnoides was originally described from Borneo and consists of several 'sub- species' (Bolton 1995), of which at least some would be good species. The form studied here is in coloration most similar to 'M. arachnoides arachnoides'. Samples were collected from three dis- turbed sites in West Java in September 2004: foot of G. Salak - Site 1 (6 39' S, 106 46' E, 560 m) (BOG3, 10 & 18), Salak- Halimun Corridor - Site 2 ( 6 45'S, 106 37'E, 710 m) (BOG24, 25 & 26), and Bogor Botanical Garden - Site 3 (6 36'S, 106 48'E, 220 m) (BOG 38). The distance between each plant where these colonies were found was approx. 5-50 m in Site 1, and 5-10 m in Site 2. These habitats have been infringed by plantation or agricultur- al activities or surrounded by residential areas. Nests constructed on a same plant were thought to constitute a single colony as the ants use same foraging trails. Nests of each colony from different plants were collected intact and put into plastic bags separately. In total, seven colonies of different sizes were collected. Nests were measured for their maximum width and length, and then dissected carefully. Workers, reproductives (dealat- ed queens, young winged queens and males), immatures (eggs, larvae and pu- pae) were counted and preserved in 80% alcohol. Pupae were sorted into sexes and castes as much as possible. RESULTS Nesting site and structure. — In Salak and Salak-Halimun Corridor, colonies were found in plantation areas with sparse trees and bushes, while in the botanical garden they were found in a forested area with relatively high trees. Nests were located on the underside of leaves of various plant species at 0.5-1.5 m above the ground in the former two sites (Fig 1 a,b), but in the botanical garden they were positioned higher at around 3-4 m above the ground. In the case of polydomous colonies com- ponent nests were generally constructed separately on different leaves of the same plant, but in one case, two nests were built narrowly connected on one and the same leaf (BOG25-2 & -4) (Appendix 1). Nests were made of carton-like material (probably chewed plant tissues), flattened domes in shape, and greyish brown in colour. Various sizes of nests were built on the underside of various sizes of leaves (Appendix 1). Within the nest, there were numerous chambers for adults and imma- tures; some of these would be used as galleries for the movements of workers (Fig. 1 b,c). The number of nests per colony varied from 1 to as many as 12; nests of a colony were usually built close to each other on one plant (approx. 15-30 cm apart). Colony composition. — Dealated queens (Q) were considered to be foundresses or those that have joined later (simply called 'queens' hereafter) and were found in all colonies and all nests except in BOG18-3 and BOG25-11. The two queenless nests had immatures, suggesting the transportation of them from other nests where a laying queen(s) existed. The number of queens differed from colony to colony and nest to nest (Fig. 2). Number of queens is highly correlated with the number of nests (Fig. 3a). The highest mean number of queens per nest, 1 1 .67, was found in a 3- nest colony (BOG18) (range: 0-22) (Fig. 2). However, BOG38 which also had 3 nests had the smallest number of queens in each of the Volume 15, Number 2, 2006 309 Fig. la-d. a, Nests constructed on underside of leaves, b, Nest on the underside a leaf, c, Structure of a nest; top of envelop removed to show the interior, d, Compartments within nest; outer walls removed. nests (mean=1.33 with a range of 1 to 2). BOG24 (1-nest colony) and BOG3 (2-nest colony) contained relatively small numbers (1-3) of queens per nest. In larger colonies, BOG10 (4-nest colony) and BOG25 (1 2-nest colony), the mean number of queens per nest was larger and relatively stable (4.00 ± 2.16 & 4.00 ± 3.34 respectively). Within each colony, workers represent the greatest number among inhabitants (approx. 40%) except in BOG 25 and BOG26 (approx. 30.3%), followed by eggs (approx. 20-40%) and larvae (approx. 10- 30%). Worker pupae and male adults constitute relatively lower ratios, while other life forms (stages) were much fewer. Most of the nests had queen(s), workers and immatures, and there was no striking specialization for a certain nest(s) in a sin- gle colony with respect to worker /repro- ductive production. However, BOG18-3, BOG38-1, BOG26-3 and BOG 10-2 con- tained higher percentages of workers com- pared with other nests composing these colonies (Appendix 1). In the single 1-nest colony (BOG24), all the adults were workers (approx. 40% of all the inhabitants) except for three queens 310 Journal of Hymenoptera Research BOG10 Nests/colony Fig. 2. Number of queens in each nest and colony. that were possibly egg layers. There were numerous eggs (approx. 45%), while larvae and pupae were relatively few, being approx. 10% and 5%, respectively. Further- more, all the pupae were workers, all this suggesting that the colony was in its ergonomic stage (reproduction and dis- persal are not its immediate concern) (cf. Oster and Wilson 1978). In the single 2-nest colony (BOG3) all life stages were present except the pupae of new queens (Appendix 1). Males had started to be produced, representing ap- proximately 5% of the total adults for the colony (Appendix 1). Male pupae also existed in BOG3-1 (Appendix 1). Adult workers, eggs and larvae each had similar percentages for the whole colony and also for each component nest. In the three 3-nest and 4-nest colonies (BOG18, 26 and 38), all life stages were present except the pupae of new queens (in all colonies), and winged adult queens in colony BOG26 (Appendix 1). As in smaller colonies mentioned earlier, workers again constitute between 30 and 40% of all inhabitants. Following the workers, eggs and larvae also occupied large proportions except in BOG18-3, just as in the smaller colonies. In BOG18-3, more adult males were observed. Worker pupae were ob- served to constitute approximately 5% in the colonies BOG38 and BOG26. In the 12-nest colony (BOG25; Appen- dix 1), winged queens were seen only in two nests in small numbers, while males were distributed more evenly; the pupae of new queens were absent in this colony. For the whole colony, workers, eggs and larvae had approximately the same numbers (around 30%) while approximately 10% were worker pupae. Reproductive output. — There were strong relationships between the number of Volume 15, Number 2, 2006 311 12000 - ♦ v = 160.81x- 567.79 y R2 = 0.8057 /~ + 2000 /& 0 i ^— y 1 1 — 1 1- 20 30 Queen number 30 Queen number g 10000 E ♦ y = 334 79x - 1658.3 R2 = 0 7845 y^ / */ 1 1 ♦ 1 1 1 ♦ y = 164 89x - 394.22 R2 = 0.8491 / / // ♦ — h ♦ 1 1 1 1 20 30 40 Queen number 20 30 40 Queen number Fig. 3a-d. a, Correlation between queen and nest numbers, b, Correlation between queen and egg numbers, c, Correlation between queen and immature numbers, d, Correlation between queen and worker numbers. queens and those of workers and imma- rures (Figs 3 b-d) (but the correlation was often strengthened by the values for the largest colony BOG25). These relationships sometimes hold also among the nests of single colonies but usually with a smaller R2 as seen in BOG10 and BOG25 (R2 = 0.0577 and 0.5549 respectively). In BOG38, however, the relationships were negative. Egg:worker ratio was relatively constant with a mean of 0.86 ± 0.5 and 0.88 ± 0.39 for each nest and colony respective- ly (Table 1). Number of queens was not significantly correlated with the numbers of new queens and males found in every colony (R2 = 0.0919 and 0.0254 respectively). DISCUSSION Nesting site, nest structure and adaptation to arboreal nesting. — We found nests of Myrmicaria araclinoides in small remnant patches of secondary forests surrounded by human settlements and plantations, and colony/nest density was relatively high. However, not all such patches harboured this ant. Generally, nests are built of carton on the under-surface of leaves with 1-2 entrances in the outer envelope. This nesting habit may protect the nests from rain and visual predators; see Holldobler and Wilson (1990) for other adaptive features of arboreal nests. Several colonies were found to be located close to each other but on different trees (approx. 5- 10 m). Closer observations, both behavioral and genetic, are needed to clarify if they are completely independent from each other or form a kind of supercolony. Poli/gyin/ and polydomy. — Myrmicaria ara- clinoides colonies observed in the present study clearly showed a polydomous and polygynous condition. At present it is not 312 Journal of Hymenoptera Research Table 1. Egg:worker (E:W) ratio. E:W ratio for E:W ratio for E:W ratio for E:W ratio for Colony Nest nest colony Colony Nest nest colony BOG24 BOG24 1.08 1.08 BOG25 BOG25(l) 1.00 0.99 BOG3 BOG3(l) 0.47 0.50 BOG25(2) 1.15 BOG3(2) 0.53 BOG25(3) 0.88 BOG 18 BOG18(l) 0.65 0.65 BOG25(4) 1.28 BOG18(2) 0.66 BOG25(5) 0.86 BOG18(3) 0.00 BOG25(6) 0.85 BOG26 BOG26(l) 1.98 1.59 BOG25(7) 1.02 BOG26(2) 2.00 BOG25(8) 1.38 BOG26(3) 0.37 BOG25(9) 0.22 BOG38 BOG38(l) 0.22 0.50 BOG25(10) 0.73 BOG38(2) 1.37 BOG25(ll) 0.71 BOG38(3) 0.69 BOG25(12) 0.82 BOG10 BOG10(1) BOG10(2) BOG10(3) BOG 10(4) MEAN 1.73 0.23 0.61 0.69 0.86 ± 0.5 0.82 0.88 ± 0.39 known whether polygyny in the present case was the result of pleometrosis (colony foundation by multiple queens) or later joining of inseminated daughters or alien females, and whether all the queens are egg layers or not. But the distribution of queens and immatures in colonies strongly suggests the existence of multiple egg layers. Ecological factors have been invoked to explain the emergence of polygyny, in particular, high dispersal risks, high prob- abilities of colonies losing their queens, short queen lifespan compared to colony survivorship and low success of individual queens conducting independent colony founding (Keller 1995, Elias et al. 2005). Sudd and Franks (1987) also reasoned that queens may come together at the colony- founding stage to pool their resources during the first and most vulnerable stages in colony growth (for the advantage of pleometrosis, see also Holldobler and Wilson, 1990). The presence of multiple queens in nests and colonies in most of the samples studied here can be discussed in relation to nesting site and competition with other species. As mentioned above, nesting of this species (and probably of its related forms) is found in rather restricted patches of suitable habitat, but usually in high densities in the patches where they have established themselves. Successful estab- lishment of immigrant queens may be not common, but accomplished by rapid de- velopment of colonies helped by pleome- trosis and polygyny, and surviving popu- lations will be maintained by polydomy and adoption of additional queens (for Camponotus, see Hansen and Klotz 2005). Thus, trees of a certain area can be dominated by this species as observed in Salak-Halimun Corridor. Although the ant fauna associated with M. aracluwides is not yet known, the above reasoning does not contradict the "ant mosaic" mentioned by Majer (1993), Djieto-Lordon and Dejean (1999), and others. In polygynous colonies of Myrmicaria araclmoides, there is a possibility of colony reproduction due to budding (cf. Elias et al. 2005). In this case there should be a poly- ethism among new queens: some disperse to other patches after mating flights, and others prefer to move to nearby branches Volume 15, Number 2, 2006 313 or trees with a group of workers. However, we do not have any evidence supporting this view. Coloni/ size and reproductive production — Smaller colonies (eg. BOG24) obviously had lower numbers of inhabitants (Appen- dix 1). These may have been at stages just after the colony foundation (cf. Sudd and Franks 1987). Data for other colonies show that as a colony grows in terms of the number of nests per colony, the number of colony members increases dramatically (Appendix 1), as suggested by Oster and Wilson (1978). In this subsequent stage, profits are mainly re-invested in workers and infrastructure such as the nest (Sudd and Franks 1987). At some critical size, a colony begins to produce sexual offspring in order to realize its inclusive fitness. In our case even the 2- nest colony had already produced some males. But there was no relation between the colony size and number of males present. For example, in the 12-nest colony (BOG25), the number of males was rela- tively small as compared with BOG38 and BOG10 with fewer nests. Several factors may be responsible for this. Some males might have already left the nest when it was collected. Furthermore we do not know whether males are produced throughout the year or during restricted seasons. The increase in queen number may increase the size of the colony, and finally the number of reproductives. Although in this study queen number positively affect- ed the number of workers and immatures, the relationship between queen number and new queen (also male) number was not positive. This shows that other factors may have operated in determining the development of queens and males as reported for many other genera (for the Argentine ant, see Aron et al. 2001). In M. araclinoides maintaining a large worker force on one tree itself may be important under certain conditions (e.g., presence of competitors) at the cost of producing more reproductives. Furthermore we cannot know how many reproductives have been produced in a colony at the time of collection because dispersed individuals do not leave any indication of their pre- vious presence in the colony unlike the case for social vespids where reproductive production can be measured rather pre- cisely by observing pupal remnants. ACKNOWLEDGMENTS We would like to express our appreciation to the Indonesian Institute of Sciences (LIPI) for granting us the permission to carry out this survey in West Java. We also would like to express our great appreciation to our research counterparts in Indonesia, particularly Mr. Rosichon Uhaidillah and his staff from Museum Zoologicum Bogoriense (Cibinong); and also Mr. Akhmad Rizali for giving us an immense amount of help during the survey. Our sincere thanks are extended to Dr. Rudy Kohout of the Queensland Museum for giving us valuable information on Polyrhachis nesting habits. LITERATURE CITED Aron, S., L. Keller, and L. Passera. 2001. Role of resource availability on sex, caste and reproduc- tive allocation ratios in the Argentine ant Eine- pithema humile. journal of Animal Ecology 70: 831-839. Billick, I. 2001. Density dependence and colony growth in the ant species Formica neorufibarbis. journal of Animal Ecology 70: 895-905. Bolton, B. 1995. A new general catalogue of the ants of the zvorld. Harvard University Press, Cambridge and London. Djieto-Lordon, C. and A. Dejean. 1999. Tropical arboreal ant mosaics: innate attraction and im- printing determine nest site selection in dominant ants. Behavioral Ecologi/ Sociobiology 45: 219-225. Elias, M., R. Rogengren, and L. Sundstrom. 2005. Seasonal polydomy and unicoloniality in a polyg- ynous population of the red wood ant Formica truncorum. Behavioral Ecology Sociobiology 57: 339-349. Emery, C. 1922. Hymenoptera, Fam. Formicidae, subfam. Myrmicinae. Pp. 95-206 in: Wytsman, P., Genera insectorum. Fasc. 174B. Hansen, L. D. and J. H. Klotz. 2005. Carpenter ants of the United States and Canada. Comstock Publishing Associates, Ithaca and London. Holldobler, B. and E. O. Wilson. 1990. The ants. Harvard University Press, Cambridge. 314 Journal of Hymenoptera Research Karavaiev, W. 1935. Neue Ameisen aus dem Indo- Australischen Gebiet, nebst Revision einiger For- men. Treubia 15: 57-117. Keller, L. 1995. Social life: the paradox of multiple- queen colonies. Trends in Ecologx/ and Evolution 10: 353-360. Kenne, M. and A. Dejean. 1999. Spatial distribution, size and density of nests of Myrmicaria opaciven- tris Emery (Formicidae, Myrmicinae). Insectes Sociaux 46: 179-185. Kenne, M., B. Schatz, R. Feneron, and J. L. Durrand. 2000. Changes in worker polymorphism in Myrmicaria opaciventris Emery (Formicidae, Myr- micinae). Insectes Sociaux 47: 50-55. Kenne, M., B. Schatz, R. Feneron, and A. Dejean. 2001. Hunting efficacy of workers from incipient colonies in the myrmicine ant Myrmicaria opaci- ventris (Formicidae: Myrmicinae). Sociobiology 37: 121-134. Majer, J. 1993. Comparison of the arboreal ant mosaic in Ghana, brasil, papua New Guinea and Aus- tralia - its structure and influence on arthropod diversity. Pp. 115-141 in: LaSalle, J. and I. D. Gauld eds. Hymenoptera and Biodiversity. CAB International, Wallingford. Oster, G. F. and E. O. Wilson. 1978. Caste and ecology in the social insects. Princeton University Press, Princeton. Savolainen, R., K. Vepsalainen, and R. J. Deslippe. 1996. Reproductive strategy of the slave ant Formica podzolica relative to raiding efficiency of enslaver species. Insectes Sociaux 43: 201-210. Sudd, J. H. and N. R. Franks. 1987. The behavioral ecology of ants. Chapman and Hall, New York. Tschinkel, W. R. 1993. Sociomerry and sociogenesis of colonies of the fire ants Solenopsis invicta during one annual cycle. Ecological Monographs 63: 425-457. Volume 15, Number 2, 2006 315 Appendix 1. Composition of life forms (stages) in colonies and nests. Nest Nest & Leaf Measure w Q WQ M Egg Larva WorkerP MaleP QueenP Total BOG24 Nest 1 N:5.0x7.5 cm L:7.7X24.5 cm 880 3 0 0 954 209 81 0 0 2127 BOG3 Nest 1 N:5.0xll.0 cm L:5.5xl6.9cm 549 1 2 38 257 390 7 4 0 1248 Nest 2 N:7.0xl2.0cm L:10xl7.0cm 607 3 2 58 324 359 1 0 0 1354 HHHH BOG 18 Nest 1 N:14.5xl6.0cm L:22.0X28.0 cm 1780 22 38 69 1156 1039 1 0 4105 Nest 2 N:10.0X14.5 cm L:26.0X27.5 cm 1413 13 14 86 929 657 39 5 0 3156 Nest 3 N:NR L:NR 15 0 0 2 0 0 0 0 0 17 Subtotal 3208 157 2085 1696 ■Bw BOG26 Nest 1 N5.5X9.5 cm L:8.2X23.5 cm 415 7 0 6 820 214 91 1 0 1554 Nest 2 N:4.5X8.8 cm L:6.0xl7.0cm 436 10 0 1 871 187 32 1 0 1538 Nest 3 N:3.8X6.0 cm L:5.8X18 cm 278 1 0 3 104 43 1 0 0 430 Subtotal 1129 1795 444 ■24 3522 BOG38 Nest 1 N:5.0X9.5 cm L:10.3X29.2 cm 335 2 3 36 75 61 17 12 0 541 Nest 2 N:3.2X6.3 cm L5.7X17.5 cm 62 1 0 6 85 105 9 2 0 270 Nest 3 N5.5X7.5 cm L:7.5Xl9.0cm 212 1 0 20 147 274 44 6 0 704 609 ■n ■m 62 307 440 ■70 20 1515 BOG 10 Nest 1 N:7.8Xll.0cm L:NR 417 2 3 64 719 295 69 1 0 1570 Nest 2 N:6.0X9.0 cm L:NR 256 3 6 11 58 31 0 0 0 365 Nest 3 N:6.0xl2.0cm L:3.5xl6.0cm 489 4 3 108 299 186 48 3 0 1140 Nest 4 N:8.0xll.5cm L:6.5xl6.0cm 956 7 11 139 661 312 14 0 0 2100 HBHHi BOG25 Nest 1 N:5.5xl4.0cm L:7.5X36.5 cm 655 3 0 6 658 746 186 1 0 2255 Nest 2 N:8.0X20.0 cm 6.0X16.5 cm L:9.0X48.0 cm 1380 7 0 12 1586 1260 339 3 0 4587 Nest 3 N:6.0X26.0 cm L:9.5X34.5 cm 1259 6 1 7 1107 1100 363 3 0 3846 Nest 4 N:6.5X20.0 cm L:9.5x31.cm N:5.0xl2.5cm L:7.0xl8.0cm 1437 9 0 6 1843 1545 321 7 0 5168 Nest 5 N:6.5xl6.0cm L:9.5X44.0 cm 534 4 1 34 461 214 63 0 0 1311 Nest 6 N:6.0xl7.0cm L:l 2.0X30.5 cm 714 2 0 1 610 503 157 0 0 1987 316 Journal of Hymenoptera Research Appendix 1. Continued. Nest Nest & Leaf Me, asure w Q WQ M Egg Larva WorkerP MaleP QueenP Total Nest 7 N5.3X18.5 L:9.5X40.5 cm cm 617 4 0 0 632 224 120 2 0 1599 Nest 8 N:7.0X19.0 L:9.4x32.0 cm cm 1205 12 0 14 1658 1613 138 5 0 4645 Nest 9 N5.5X14.5 L:8.5X27.0 cm cm 413 2 0 1 91 528 63 0 0 1098 Nest 10 N:6.5X12.5 L:7.8X20.5 cm cm 662 4 0 2 481 393 45 0 0 1587 Nest 11 N:6.0X16.0 L:9.3X36.0 cm cm 722 0 0 1 516 455 121 2 0 1817 Nest 12 N:7.0X19.0 L:9.8x28.5 cm cm 968 3 0 3 794 460 194 2 0 2424 ■^K Subtotal 10566 10437 9041 2110 25 WKBM 32324 W=worker, Q=queen, WQ=winged queen, M=male, WorkerP=worker pupa, MaleP=male pupa, Queen- P=queen pupa, NR=not recorded, N=nest, L=leaf. J. HYM. RES. Vol. 15(2), 2006, pp. 317-347 The Species of Sternanlopius Fischer (Hymenoptera: Braconidae, Opiinae) and the Braconid Sternaulus Robert A. Wharton Department of Entomology, Texas A&M University, College Station, TX 77843, USA email: rawbaw2@tamu.edu Abstract. — The opiine braconid genus Sternaulopius Fischer is recognized as valid, redefined, and one new species, Sternaulopius duplicatus, from Madagascar, is described. The type species, Sternaulopius bisternaulicus Fischer, is fully redescribed, with new host and distribution records. The only other previously included species, Sternaulopius beieri Fischer, is placed as a junior subjective synonym of Opius bajulus Haliday, new synonym. Opius bajulus is also redescribed and the genus group name Biophthora Foerster is resurrected for this species. Opius rossicus Szepligeti, is transferred to the genus Biophthora (new combination) and Opius castaneus Granger, type species of Frekius Fischer, is transferred to Utetes Foerster (new combination). Thus, Frekius is a junior subjective synonym of Utetes, new synonym, but is retained as a valid subgenus. Biophthora and Sternaulopius are compared to Xynobius Foerster (where Opius bajulus has frequently been placed), and Xynobius is redefined and treated as a subgenus of Eurytenes Foerster. Stigmatopoea Fischer is also recognized as a valid subgenus of Eurytenes. Characters used to define these genus-group taxa are discussed in detail, with emphasis on venation, placement of metasomal spiracles, and sculptural details of the body. Use of the term sternaulus for a longitudinal groove on the ventral mesopleuron in Ichneumonoidea is reviewed, and it is shown that the sternaulus in cryptine and mesochorine Ichneumonidae is not homologous to the precoxal sulcus in Braconidae based on dissections of associated thoracic musculature. A true sternaulus, defined internally as the ridge supporting the origin of the mesopleural-basalare muscle, is rarely present in Braconidae. The genus Sternaulopius Fischer, 1965 Opiinae is still unsettled. Wharton (1988) was described to accomodate a single observed that, except for the double ster- species from the Democratic Republic of naulus, the two described species more the Congo. Subsequently, Fischer (1968) closely resembled other species within described a second species from Germany. Opius Wesmael s. /. than they did each No additional species have been described, other. On this relatively limited evidence, and only five specimens have been re- Wharton (1988) transferred both species to corded (Fischer 1965, 1972, Quicke et al. Opius, thus treating Sternaulopius as a syn- 1997). The two species that have been onym of Opius while at the same time included in Sternaulopius (Figs 1-20) have noting that Opius s. I. was not demonstra- two distinct grooves (sternauli) on each bly monophyletic. Quicke et al. (1997) side of the mesopleuron but in nearly all recognized six genera that were formerly other braconid wasps only a single groove treated by Fischer (1972, 1977, 1987) as is present, or the groove is completely lost subgenera of Opius, and reported that the (exceptions include Trigastrotheca laikipien- venom apparatus of a specimen of Sternau- sis Quicke and some species of Pambolus lopius beieri Fischer resembled that of some, Haliday). These two species of Sternnulo- but not all of the species that they included pius are thus distinctive, though their in one of these genera, Xynobius Foerster. placement in the classification of the Nothing else has been published on Ster- 318 Journal of Hymenoptera Research Figs 1-4. Sternaulopius bisternaulicus reared from Strombosia fruits: 1, face. 2, lateral view of head. 3, dorsal view of head. 4, right mandible, with oblique views of clypeus and labrum. naulopius, undoubtedly due to the paucity of individuals available for study. Van Achterberg (2004) recently re-characterized Xynobius, but made no mention of Sternau- lopius. The term sternaulus has long been used by students of Ichneumonidae to describe a longitudinal groove on the lower part of the mesopleuron extending posteriorly from the ventral-lateral region of the epicnemial (= prepectal) carina towards the coxifer (or pleural coxal process) (Viereck 1916, Richards 1956, Townes 1969). Several workers (e.g. Granger 1949, Fischer 1958, Mason 1964, Marsh 1971, Wharton et al. 1997) have also applied this term to a similar groove on the mesopleuron of Braconidae, while other braconid specialists have used the name precoxal suture/sulcus (notably van Achterberg 1975, 1993, Shaw and Hud- dleston 1991) or simply longitudinal fur- row of mesopleuron (Muesebeck 1970). The term precoxal sulcus has often been applied more widely in the Hymenoptera, though not necessary consistently (com- pare Richards 1956 with Bohart and Menke 1976). Gibson et al. (1998) treat the sternaulus as a synonym of the transepisternal line/sulcus (though their focus is on Chalcidoidea). Fischer (1972) noted that it was probably incorrect to equate the sternaulus with the precoxal sulcus in Braconidae, and van Achterberg Volume 15, Number 2, 2006 319 Figs 5-10. Stemaulopius bisternaulicus reared from Strombosia fruits: 5, left side of mesosoma, left arrow = propleural flange, upper right arrow = precoxal sulcus, lower right arrow = sternaulus. 6, ventral-lateral view of mesopleuron, upper arrow = precoxal sulcus, lower arrow = sternaulus. 7, dorsal view of mesosoma. 8, dorsal view of pronotum, upper arrow = oblique groove on propleuron, lower arrow = left notaulus. 9, dorsal view of metanotum and propodeum. 10, dorsal view of petiole. (1993) illustrated the two as separate be observed in some Opiinae (undoubt- structures for braconids. In a newsletter, edly referring to Stemaulopius), and Pam- van Achterberg (1994) stated more specif- bolinae where "the sternaulus is clearly ically that the difference between the bent downwards anteriorly and situated sternaulus and the precoxal sulcus can more ventrad." 320 Journal of Hymenoptera Research Figs 11-12. Sternaulopius bisternaulicus reared from Strombosia fruits: 11, lateral view of metasoma. 12, same showing dorsope (left arrow) and spiracle on second tergum (right arrow). There are no published host records for Sternaulopius (with the exception of in- formation I have recently included on a website: hymenoptera.tamu.edu/parof- fit). However, examination of material reared from fruit in both Cameroon and Kenya (Steck et al. 1983, Copeland et al. 2002), indicates that tephritid fruit flies (Diptera) are the hosts of at least some of the species of Sternaulopius. This reared material forms the basis for the present treatment, including a preliminary exami- nation of the nature of the sternaulus in Ichneumonoidea. MATERIALS AND METHODS Specimens of Sternaulopius (sensu Fischer 1972, 1987) were either reared from fruits infested with Tephritidae or bor- rowed from museums. Rearing methods, localities, and methods of identification of flies and plants are described in Steck et al. (1983) and Copeland et al. (2002). Material was borrowed from the Koninklijk Muse- um voor Midden-Afrika, Tervuren, Bel- gium (MRAC), Museum fur Naturkunde der Humboldt-Universitaet, Berlin, Ger- many (ZMHB), Hungarian Natural History Volume 15, Number 2, 2006 321 Fig. 13. Sternaulopius bisternaulicus reared from Strombosia fruits, ovipositor and ovipositor sheaths. Museum, Budapest (HNHM), Museum National d'Histoire Naturelle, Paris (MNHN), and the U. S. National Museum of Natural History (USNM). Additional specimens examined are in the Naturhis- torisches Museum Wien, Austria (NHMVV) and the National Museum of Ireland, Dublin (NMID). Fly puparia were individually isolated in Kenya by R. Copeland from one heavily- infested sample of Strombosia scheffleri Engl. (Olacaceae) fruit, collected in Kakamega Forest, Kenya on 30th April 2000. Associ- ation of wasps with hosts was made by the author, based on characteristics of the puparia. All other host records listed below are based on wasps that emerged from tephritid puparia that were not individu- ally isolated. Numerous specimens of Braconidae and Ichneumonidae from material housed at Texas A&M University were dissected to examine the internal musculature of the mesothorax. All dissected material was initially stored in 70-80% ethanol, then either air dried following transfer through 95% ethanol and 99% amyl acetate or dissected while still in alcohol. Specimens of Apis mellifera L. and several other Apocrita collected in central Texas with Malaise traps were also dissected to ascer- tain the pattern of general thoracic muscu- lature, using the works of Daly (1964) and Gibson (1985) for orientation and identifi- cation of major muscles. All identifica- tions were made by the author; voucher specimens are deposited as Texas A&M University voucher number 656 in the Texas A&M University Insect Collection (TAMU). Measurements are as in Wharton (1977, 1986); terminology for descriptions of 322 Journal of Hymenoptera Research V 14 15 Figs 14-16. 14, 15, Fore and hind wings of Sternaulopius bisternaulicus. 16, Hind wing of Biophthora bajulus, arrow = 2-1A. venation and external features of the body precoxal sulcus (following van Achterberg follows Sharkey and Wharton (1997), ex- 1993). Details of the mesothorax associated cept that in the present paper a distinction with the musculature are provided in the is made between the sternaulus and the results and discussion where additional Volume 15, Number 2, 2006 323 Figs 17-20. Sternaulopius beieri holotype female (= B. bajulus): 17, lateral habitus. 18, face. 19, dorsal-posterior view of posterior part of mesosoma showing mesonotal midpit and sculpture of scutellum and propodeum. 20, lateral oblique view of mesosoma, upper arrow = precoxal sulcus, lower arrow = sternaulus. terms, when used, are specifically defined. Thoracic muscles relevant to discussions of the sternaulus are noted in Figs 21-32. Of particular focus are various axillary mus- cles and especially the mesopleural-basa- lare muscle (sensu Gibson 1985). The latter originates on the wall of the mesopleuron and inserts on the basalare sclerite near the base of the wing, at least in all the Ichneumonoidea (22 genera) and Aculeata (4 genera) examined (Table 1). RESULTS AND DISCUSSION Musculature and associated external features The observations made here refer specif- ically to the Ichneumonoidea (Ichneumo- nidae + Braconidae). However, they will also apply to many other taxa both within and outside apocritan Hymenoptera, as evidenced by the detailed studies of Daly (1964) and Gibson (1985, 1986a, b, 1993). The major muscles discussed below for the Ichneumonoidea were roughly similar in their points of origin and insertion in the few aculeates and non-aculeate apocritans that were dissected (namely A. mellifera, Khopalosoma nearcticum Brues, and unde- termined species of Gasteruption Latreille and Polities Latreille). Many of the details of thoracic musculature found in such excellent works as those of Gibson (1985, 1993) have been omitted here since the major interest is to provide a better char- acterization of the sternaulus. On the mesopleuron, the sternaulus and various pits, depressions, or other changes 324 Journal of Hymenoptera Research " '% X 4 )26B| I^^B Jjj Figs 21-26. Sections through mesosomas of Ichneumonidae, showing thoracic musculature. 21-24, Limonethe sp., a typical Ichneumoninae lacking a sternaulus, left side of body in ethanol: 21, dorsolongitudinal indirect flight muscle (arrow); 22, dorsolongitudinal muscle and associated phragma removed to expose dorsoventral indirect flight muscle (arrow); 23, dorsoventral muscle removed (arrows = fragments left to show origin of this muscle mass) to expose mesopleural-basalare muscle originating in trough of mesopleuron immediately laterad dorsoventral muscle; 24, mesopleurahbasalare muscle removed to expose 3 sets of muscles (arrows) originating on lateral wall of mesopleuron. 25, Barichneumon sp., Ichneumoninae, dorsal view of inside of trough of Volume 15, Number 2, 2006 325 in elevation are, to a greater or lesser recent taxonomic treatments), rather than extent, the external representation of the parapsidal furrows. attachment of the thoracic muscles. In- Immediately laterad the dorsoventral ternally, the mesothorax, at least in ichneu- indirect flight muscle is the smaller monoids and aculeates, is dominated by (though still fairly large) mesopleural- the massive dorsolongitudinal (Fig. 21) basalare muscle (Figs 23, 25-29, 31). This and dorsoventral (Fig. 22) indirect flight is a distinctive band of muscle fibers muscles. The former originates largely readily identified by the dorsal, sclerotized from the median mesoscutal lobe (and cap that is strongly constricted to form partly from the phragma along the anterior a tendon-like attachment connected to the margin of the lobe) and extends posteriorly basalare (Figs 27, 29, 31). In Ichneumonoi- through the middle of the mesothorax. On dea, the mesopleural-basalare originates either side of this median muscle mass is ventrally in the trough of the mesothorax a broad band of dorsoventral muscle fibers just laterad the dorsoventral muscle or it (Fig. 22), which insert on the lateral lobes may arise somewhat higher on the vertical of the mesoscutum and originate ventrally wall of the mesepisternum. The subtegular on the mesothorax. The floor of the (= subalar) ridge, immediately ventrad the mesothorax, where the dorsoventral mus- anterior subalar depression of van Achter- cles originate, is a trough interrupted berg (1988), is the external representation medially by a phragma representing the of the internal pocket in which resides the midventral invagination of the mesoster- sclerotized cap of the mesopleural-basalare num (partially shown in Figs 25 and 29). muscle. The epicnemial carina, when pres- The lateral and ventral sides of the trough ent, delineates the anterior border of this are formed on each side of the body by the muscle mass ventrally, and dorsally it is relatively sharp vertical to horizontal tran- delineated by the pronotal-mesopleural sition in the mesepisternal wall ventrally as suture. it curves towards the ventral midline. The mesopleural-basalare muscle origi- Notauli are perhaps best defined as nates on the sternaulus in those members external grooves on the mesoscutum in- of the Ichneumonidae that possess a ster- dicating the position of the internal ridge naulus (Figs 25, 28), but does not originate or phragma that separates the attachment on the structure that has frequently (e.g. point of the dorsolongitudinal muscle mass Sharkey and Wharton 1997) been called the from the attachment points of the right and sternaulus in Braconidae (Figs 29-30). In left dorsoventral muscle bundles (see the the ichneumonid taxa Cryptini and Meso- excellent discussion in Gibson (1985) for chorinae, both of which have a well-de- terminology of mesonotal grooves). In veloped sternaulus ventrolateral^ on the Ichneumonoidea, the position of these mesepisternum, the mesopleural-basalare grooves, when present, relative to the muscle originates on the internal ledge attachment points of these two indirect formed by the anterior portion of the flight muscles confirms that they should be sternaulus as the latter extends posteriorly called notauli (as they are in nearly all from the epicnemial carina. In those mem- mesopleuron, dried specimen, left arrow showing mesopleural-basalare muscle, right arrow showing internal ledge of short sternaulus on which anterior end of mesopleural-basalare originates. 26, Enicospilus sp., Ophioninae, right side of body in ethanol, showing origin of elongate mesopleural-basalare in trough of mesopleuron. 326 Journal of Hymenoptera Research Figs 27-32. Longitudinal sections through mesosomas of Ichneumonoidea, showing muscles originating on mesopleuron: 27-28, Ichneumonidae, Cryptinae, right side of body: 27, in ethanol showing mesopleural-basalare muscle (arrow) and its dorsal cap; 28, same specimen, dried (resulting in gaps in muscle bundle), showing origin of mesopleural-basalare muscle on internal ledge formed by sternaulus (arrow). 29-30, Fopius vandenboschi (Braconidae, Opiinae), dried: 29, ventral and left portions of body, bottom arrow = midventral phragma (partially broken in dissection), top arrow = mesopleural-basalare muscle; 30, right side of body, mesopleural- basalare removed, bottom arrow = region where mesopleural-basalare originates, middle arrow == precoxal sulcus visible through cuticle, top two arrows = muscles inserting on axillary sclerites. 31, Wroughtonia ferruginea Volume 15, Number 2, 2006 327 bers of the subfamily Ichneumoninae in inside, the ledge along the dorsal side of which the sternaulus is distinct as a short this braconid "sternaulus" serves as a gen- anterior crease (e.g. Fig. 25), the meso- eral attachment point for three sets of pleural-basalare also originates there, pro- muscles located laterad the mesopleural- viding evidence that the sternaulus in these basalare muscle (the ledge best seen in three subfamilies is homologous. In the Fig. 30 where this "sternaulus" is visible pimplines, ichneumonines, and campople- through the semi-transparent cuticle). All gines that lack a visible sternaulus, the three of these muscles are fan-shaped, with mesopleural-basalare muscle originates at a broad origin on the wall of the meso- the lateral margin of the ventral curve of pleuron. They are strongly tapered dorsal- the mesepisternum just laterad the origin ly, where the muscle bundles in each fan of the dorsoventral indirect flight muscle unite to form tendons (Figs 24, 30, 32). Two (Figs 21-23, 26). This is also where the of these muscles appear to correspond to mesopleural-basalare originates in all of the muscles numbered 8 and 9 in Gibson the braconids that were dissected. The (1986b) and are clearly axillary muscles, sternaulus is long in nearly all cryptines inserting on the third axillary sclerite at the and tends to curve dorsally as it extends base of the fore wing. The third is similar in posteriorly. However, in the species dis- position to Gibson's (1986b) muscle num- sected, the mesopleural-basalare is at- ber 4, and inserts on the mesoscutum near tached to the anterior end of the sternaulus, the posterior notal wing process. This same which is more ventrally displaced. In those pattern of three muscles was found in all braconids with a well-developed "sternau- ichneumonoids examined. However, the lus," such as members of the Agathidinae, position of the muscles and in particular Doryctinae, Helconinae, and Rogadinae, their orientation (nearly longitudinal vs. the mesopleural-basalare passes over this dorsoventral) varies with the shape of "sternaulus" from its more ventral origin thorax and, in braconids, the angle of the on the floor of the mesopleuron. In both "sternaulus." When the braconid "sternau- braconids and ichneumonids, the muscle lus" is strongly oblique, for example, the varies in length and width (compare dorsal-most of the axillary muscles may be Fig. 29 with Fig. 31), as might be expected nearly longitudinal rather than dorsoven- given the differences among species in the tral in orientation. shape of the mesothorax. Although the origin of a muscle may be In the opiine braconid Sternaulopius, less stable than the insertion for establish- which has a second "sternaulus" below ing homologies (see especially Daly 1963, the first, the mesopleural-basalare muscle 1964), the above observations support the originates on the internal ledge formed by conclusions of van Achterberg (1993, 1994) the ventral-most sternaulus, which, not that the ventral sternaulus in Sternaulopius coincidentally, is in the same position as is homologous to the sternaulus of ichneu- the attachment point of this muscle in other monids. Similarly, the dorsal "sternaulus" braconids. The dorsal-most "sternaulus" in of Sternaulopius is homologous to the single Sternaulopius is homologous to the "ster- groove found in other braconids, which naulus" of other braconids, and on the van Achterberg has termed the precoxal (Braconidae, Helconinae), left side of body in ethanol, showing broad mesopleural-basalare muscle, arrow = dorsal cap of muscle. 32, Alabagrus sp. (Braconidae, Agathidinae), right side of body, dried, mesopleural- basalare muscle removed, arrows = muscles inserted on axillary sclerite. 328 Journal of Hymenoptera Research Table 1. List of taxa dissected for examination of mesopleural-basalare and other mesothoracic muscles. BRACONIDAE Agathidinae Alabagrus sp., College Station, TX, USA Braconinae Atanycolus simplex (Cresson) and Digonogastra sp., College Station, TX, USA Cenocoeliinae Cenocoelius saperdae (Ashmead), College Station, TX, USA Doryctinae Heterospilus sp., Gainesville, FL, USA Helconinae Helconidea ferruginea (Brues), College Station & Angelina National Forest, TX, USA Homolobinae Homolobus truncator (Say), Conroe, TX, USA Macrocentrinae Austrozele sp., Gainesville, FL, USA Opiinae Fopius arisanus (Sonan), Honolulu, HI, USA Fopius vandenboschi (Fullaway), Tangerang, W. Java, Indonesia Psyttalia incisi (Silvestri), Honolulu, HI, USA Sternaulopius bisternaulicus Fischer, coastal Kenya Rogadinae Aleiodes aciculdtus Cresson, College Station, TX, USA ICHNEUMONIDAE Anomaloninae Anomalon sp., Conroe, TX, USA Banchinae Syzeuctus sp., Conroe, TX, USA Cryptinae Cryptini spp., College Station, TX, USA and Springfield, Dominica Ichneumoninae Barichneumon sp., Bastrop, TX, USA Cratichneumon sp., Huntsville, TX, USA Limonethe sp., Gainesville, FL, USA Mesochorinae Mesochorus sp., College Station, TX, USA Ophioninae Enicospilus sp., Gainesville, FL, USA Pimplinae Pimpla sp., College Station, TX, USA Rhyssinae Megarhyssa macrurus (L.), College Station, TX, USA APIDAE ' Apis mellifera L., College Station, TX, USA GASTERUPTIIDAE Gasteruption sp. (Rhydinofoenus sensu Townes), College Station, TX, USA RHOPALOSOMATIDAE Rhopalosoma nearcticuni Brues, College Station, TX, USA VESPIDAE Polistes sp., College Station, TX, USA sulcus. The sternaulus of ichneumonids likely to be homologous. It therefore seems has the same function as the "sternaulus" appropriate to use the term precoxal sulcus (= precoxal sulcus) of most braconids, but for the more dorsally displaced groove the muscles whose origin they support are found in most braconids, despite the different and thus the structures are un- probability that this term is used for non- Volume 15, Number 2, 2006 329 homologus features in other Apocrita. conceal ventral corner of pronotum above Coining another term would simply add base of fore coxa; oblique carina and to the confusion, and seems unnecessary associated, shallow groove (Fig. 8) weakly since precoxal sulcus has become fairly to distinctly developed, separating flange well established at least in the European from rest of propleuron. Pronotum dorsal- braconid literature. ly with transverse, crenulate groove along posterior margin; pronope absent. Notauli Taxonomy deep anteriorly, shallow to barely indicat- Sternaulopius Fischer, 1965 ed posteriorly, weakly curving into dis- tinct, rounded midpit; unsculptured pos- Stemaulopius Fischer, 1965: 311. teriorl Mesoscutum carinateiy margined Type species: Stemaulopius bisternauhcus . ,, , , , i . -n £. ■ -.nsr- *„ i • j i j laterally from base of notaulus to axillarv Fischer, 1965. Monobasic and original desig- _ J tT^. _. _ , ... flange (Fig. 5). Iransscutal articulation nation. ? , j represented by a distinct sulcus lmmedi- Head (Figs 1-4). Occipital carina widely ately anteriorad crenulate scutoscutellar separated from hypostomal carina at base sulcus. Scutellum (Fig. 7) polished, un- of mandible, extending dorsally more than sculptured to weakly punctate (except half height of head, not or only weakly along posterior margin). Precoxal sulcus curved medially at dorsal end, completely (Fig. 6) broad, crenulate anteriorly, absent absent dorsal-medially; hypostomal carina or apparently so posteriorly; distinct from protruding as a flange below base of the more ventrally placed sternaulus; ster- mandible. Gena (including temple behind naulus well-developed as a broad, crenu- eye) and vertex smooth, polished. Frons late groove extending more than half bare, polished and impunctate but usually length of mesopleuron; both precoxal with weak, irregular sculpture along mid- sulcus and sternaulus straight over poste- line. Mandible (Fig. 4) evenly curved along rior half, never curving ventromedially. dorsal margin; ventral margin distinctly Hind margin of mesopleuron crenulate carinate, without basal tooth or lobe; above (though sometimes weakly) and bidentate, dorsal tooth longer than ventral below mesopleural fovea (Figs 5, 6). Ven- tooth; outer surface of mandible weakly to tral midline crenulate, with carinations stongly curved from base to apex, species distinct posteriorly but not developed as with weaker curvature have a completely a postpectal carina. Metapleural/propo- concealed labrum. Labrum varying from deal junction either obscured by sculpture exposed to completely concealed; clypeus or represented by a weak impression. (Figs 1, 2, 4) with ventral margin truncate, Propodeum rugose to carinate-rugose. varying from sharp (when labrum ex- Hind tibia lacking basal carina typical of posed) to slightly thicker (when labrum Utetes Foerster, 1862. concealed), in profile clypeus evenly con- Wings (Figs 14, 15). Stigma thickened, vex, moderately protruding, distinctly wedge-shaped; r arising from basal 0.3-0.4; overhanging labrum when the latter ex- second submarginal cell long (2RS<3RSa); posed. Epistomal sulcus generally more m-cu arising distad base of 2RS, (RS+M)b deeply impressed laterally than dorsally. thus absent; (RS+M)a arising distinctly Malar sulcus a distinctly impressed line, below parastigma, 1RS>0.3X1M; 1-1 A Maxillary palp about as long as height of curved towards posterior margin but re- head, 6 segmented; labial palp 4 segment- maining separate from margin by at least ed. First flagellomere longer than second; width of 1-1A; 1st subdiscal cell closed, apical flagellomere spinose at tip. Hind wing RS nearly absent, represented Mesosoma (Figs 5-9). Propleural flange by a weak, unpigmented crease; m-cu large, strongly curved laterally to partially extending at least half distance to wing 330 Journal of Hymenoptera Research margin as a curved, distinctly impressed, 1837 (new synonym, see below) affords an weakly pigmented line. opportunity to compare Sternaulopius to Metasoma (Figs 10-12). Petiole with spi- Xynobius Foerster, 1862 as well as to racle located near middle of segment; pit- discuss the placement of Sternaulopius like dorsope present basally at junction of within the Opiinae. dorsal and lateral carinae; laterope present Although O. bajulus has been placed in at extreme base. Second and following Xynobius in all recent treatments of Opiinae visible terga neither sculptured nor un- (e. g. Fischer 1972, Tobias and Jakamavi- usually shortened. Second metasomal ter- cius 1986, van Achterberg 1997, 2004; gite with spiracle placed dorsally, just Quicke et al. 1997), there are distinct mesad the rounded lateral fold. Hypopy- differences between bajulus and the type gium weakly protruding at midline, never species of Xynobius {viz, X. pallipes Foerster, strongly attenuate nor extending beyond 1862, for which the current valid name is metasomal tergites. Ovipositor lacking Eurytenes caclatus (Haliday, 1837), new sharp subapical notch. combination, see below). One of the more Biology. — Reared from tephritids infest- important of these differences is the ar- ing fruits belonging to several plant fam- rangement of metasomal spiracles. In £. ilies. Known only from Subsaharan Africa caelatus (as well as in Eurytenes abnormis and Madagascar. Included species: Ster- (Wesmael, 1835), the type species of Eur- naulopius bistemaulicus Fischer, 1965 and ytenes Foerster, 1862), the spiracle of the the new species from Madagascar de- second metasomal tergum is situated lat- scribed below. erally, well below the crease that separates Excluded species. — Sternaulopius beieri the dorsal portion of the tergite from the Fischer, 1968 (= Opius bajulus Haliday, lateral portion of the tergite. In both 1837). bistemaulicus and beieri ( = bajulus), the Diagnosis. — Sternaulopius is distin- spiracle (Fig. 12) is located dorsally in guished from nearly all other supraspecific a position typical of most other opiines: taxa in the Opiinae by the presence of both either adjacent to the crease (when a dis- a sculptured precoxal sulcus and a true tinct crease is present) or situated more sternaulus (Fig. 6), and from Biophthora by medially. Xynobius was originally de- the absence of rugose sculpture on the scribed as a cyclostome braconid (Foerster scutellum dorsomedially. Sternaulopius is 1862), but Foerster obviously had difficulty further recognizable by the presence of placing this taxon since in the same paper a dorsope, a distinct median pit on the he described two other genera (Biophthora mesoscutum, and a five-sided second sub- Foerster, 1862 and Aclisis Foerster, 1862) in marginal cell, as well as the absence of both two different family-group taxa that are a basal tooth or lobe on the mandible and now (Fischer 1972, van Achterberg 2004) a basal carina on the hind tibia. included in Xynobius. Muesebeck and Remarks. — The type species of Stemaulo- Walkley (1951) were apparently the first plus was described on the basis of one to associate the name Xynobius with Opii- female and two males from the Democratic nae when they placed all three of these Republic of the Congo (Fischer 1965). A Foerster genera as synonyms of Opius second species, Sternaulopius beieri Fischer, Wesmael, 1835. Fischer (1972) established was subsequently described from Ger- a comprehensive classification for the many on the basis of a single female Opiinae that included Xynobius as one of (Fischer 1968). Only one additional speci- 36 subgenera of Opius. Fischer (1972) men has been reported to date (Quicke et presented the first clear delineation of al. 1997). The discovery that S. beieri is Xynobius, which he characterized based a junior synonym of Opius bajulus Haliday, on its exposed labrum and sculptural Volume 15, Number 2, 2006 331 features of the mesothorax. He noted separated throughout. Despite these differ- especially the heavily sculptured scutellum ences, removal of bajulus from Xynobius is (Fig. 19), an unusual feature in the Opiinae, complicated by the fact that both bajulus and this is shared by both bajulus and and caelatus can be characterized by the caelatus. Both species have been character- shared possession of strongly elevated and ized as having an exposed labrum and also heavily sculptured scutellum (Fischer a deep, pit-like dorsope on the petiole, 1972), an unusual feature found in rela- features emphasized by Tobias (1977) and tively few Opiinae. In one of the other van Achterberg (2004) respectively. Thus, species with a similarly sculptured scutel- both of these species have been included in lum, Opius severini Fischer, 1964, the Xynobius since 1972 regardless of the scutellum, fore wing venation, clypeus differences in how Xynobius has been and mandible are as in caelatus, but the characterized by Fischer (1972) and sub- spiracle of the second metasomal segment sequent authors, and regardless of whether is positioned as in bajulus. One is therefore Xynobius has been accorded subgeneric left with the choice of a very broad (Fischer 1972, Tobias 1977) or generic (van definition for Xynobius, such as that offered Achterberg 1997, 2004) rank. Thus, if by Tobias (1977), Tobias and Jakimavicius bajulus is retained in Sternaulopius, follow- (1986) and, apparently, van Achterberg ing the discovery here of the synonymy (2004), which could accomodate bajulus, between bajulus and beieri, then Sternaulo- or a much more restricted one (e.g. Fischer phis becomes a synonym of either Xynobius 1972). The relative merits of each are or Biophthora (bajulus is the type species of outlined below. Biophthora). Throughout much of the 1900s, most The difference between bajulus and cac- Opiinae were placed in the genus Opius, latus in placement of the spiracle of the which eventually encompassed over 1000 second metasomal tergum argues against species. Fischer (1972) facilitated work on retention of bajulus in Xynobius. Differences the Opiinae by presenting a classification in the details of fore wing venation and in which several distinctive genera were shape of the clypeus, mandibles, and recognized and an extensive subgeneric petiole provide further evidence that baju- classification was proposed for Opius. lus and caelatus may not be closely related, Opius has been subsequently reduced by and bring into question the homology of removal of putatively monophyletic taxa some of the character states previously such as Utetes and Psyttalia Walker, 1860 used to place these two species together in (Wharton 1987, 1988), and by an attempt to Xynobius. The second submarginal cell is restrict the definition of Opius to species parallel-sided in caelatus, m-cu is antefur- with a basal tooth on the mandible (van cal, the stigma is more nearly parallel- Achterberg and Salvo 1997). Further deli- sided than wedge-shaped, the clypeus is mitation of monophyletic groups is essen- short and strongly protruding, the mandi- tial for progress in understanding the ble is more massive basally and strongly relationships and evolutionary biology of tapered distally, and the dorsal carinae of this large and important group of dipteran the petiole are strongly convergent basally. endoparasitoids. Xynobius in the sense In bajulus (Figs 16-20), the second sub- used by Tobias (1977) and van Achterberg marginal cell is narrowed distally, m-cu is (2004) is fairly well delineated, but difficult postfurcal, the clypeus is truncate and to define as monophyletic because all of the nearly flat (and the labrum consequently characters given for recognition of the more weakly exposed), the mandible is genus can be considered plesiomorphic. slenderer and nearly parallel-sided, and As correctly noted by van Achterberg the dorsal carinae of the petiole are widely (2004), for example, Xynobius has a number 332 Journal of Hymenoptera Research of features suggestive of a basal placement argument for the monophyly of Xynobius within the Opiinae. Features such as using characters emphasized by previous a distinct dorsope, distinct hypoclypeal workers, I prefer to retain Xynobius as depression (labrum exposed by a gap a subgenus at this time rather than elevate between the clypeus and mandibles), ab- it to generic rank. If Opius is defined in sence of a transverse carina on the pro- a restricted sense to include only species pleuron and absence of a true sternaulus with a basal tooth on the mandible (van may be useful for recognition of Xynobius Achterberg and Salvo 1997), then Xynobius within the Opiinae (van Achterberg 2004), is better placed in Eurytenes. but they are also indicative of the basal Placement of Xynobius in Eurytenes by no placement of this group since these fea- means solves the problem of monophyly of tures also characterize the Exothecini, the Xynobius, but does bring together several putative outgroup taxon and are thus groups of opiines that share characters plesiomorphic in this context. Some suggestive of a close relationship (though exothecines also have a short (relative to it is admittedly tricky to define relation- 2RS) cross-vein r in the fore wing, the only ships among those putatively basal taxa other recognition feature noted by van that retain both a deep dorsope and an Achterberg (2004). The latter is significant exposed labrum). Eurytenes has long been because it serves to separate Xynobius characterized by the extremely basal origin sensu van Achterberg (2004) from Eury- of r on the stigma, and only two species tenes. have been described with this characteris- The sculptural patterns noted by Fischer tic. As pointed out by Wharton (1988), (1972) in his more restricted definition of however, the shape of the pterostigma is Xynobius provide a seemingly better op- also unusual, and the shape of both the portunity for defining this taxon as mono- stigma and the petiole suggest a relation- phyletic, but there are problems of homo- ship between the type species of Eurytenes, logy. The species included in Xynobius by £. abnormis, and species such as Opius Fischer all have a sculptured scutellum, but macrocerus Thomson, 1895 (type species of the pattern differs markedly among the Stigmatopoea Fischer, 1986). Critical exam- species. For example, Opius rudis Wesmael, ination of the fore wing of macrocerus 1835, included in Xynobius by Fischer demonstrates that the Eurytenes condition (1972), has a flattened scutellum with has been achieved by increasing the angle coriaceous sculpture lacking rugosities. In between r and 3RSa and increasing the species such as caelatus, bajulus, and sever- length and curvature of r so that the latter ini, the scutellum is both elevated and extends to the base of the stigma. The close rugose, providing a putative synapomorph relationship between macrocerus and abnor- in terms of a complex character. However, mis is further supported by other shared this same complex character state is found features of the fore wing (shape of second in some of the species of Biosteres Foerster, submarginal cell and placement of m-cu 1862. Further, as detailed above, these and 2CUb) as well as the fact that both three species (caelatus, bajulus, and severini) have the spiracle of the second metasomal differ markedly in wing venation, shape of segment displaced laterally. Fischer (1998) the clypeus and mandibles, and placement accepted the relationship between macro- of the metasomal spiracles. The available cents and Eurytenes based on fore wing evidence suggests, then, that this scutellar venation, and in particular the shape of the feature has either developed independent- stigma, and tranferred several additional ly in at least a few opiine lineages or is species to Eurytenes. Van Achterberg a primitive feature retained in these (2004), on the other hand, placed Stigmato- lineages. As I am unable to provide a strong poea as a synonym of Xynobius, but did not Volume 15, Number 2, 2006 333 include or discuss Eurytenes despite the fact that Quicke et al. (1997) included Eurytenes as a subgenus of Xynobius, citing van Achterberg and van Zuijlen (in press). In either case, macrocerus is nicely posi- tioned as morphologically intermediate between caelatus and abnormis, with suffi- cient similarities in the shape of the stigma, the fore wing venation, and the placement of T2 spiracles to justify their grouping as a single genus. Considering only these three taxa, the hypothesized relationships are Xynobius + (Stigmatopoea + Eurytenes s. s.). If the shape of the elongate distal portion of the stigma is used to support the monophyly of Eurytenes s. /., then the narrowed basal portion in conjunction with a relatively long vein r can be used to support a Stigmatopoea + Eurytenes s. s. relationship. However, this would be an oversimplified picture of relationships since Eurytenes, as defined here, embodies a fair number of species, including most of Fischer's (1998) Eurytenes and van Achter- berg's (2004) Xynobius. Here, I propose restricting the name Xynobius to those species with the spiracle of the second metasomal segment located more laterally, though this will need to be re-evaluated in a more comprehensive treatment of this group. Therefore, I reject the inclusion of bajulus in Xynobius, though an argument could be made that bajulus is simply a highly derived species of Xynobius. Additional differences in the stigma, vena- tion, clypeus, and mandibles all suggest that bajulus should be excluded not only from Xynobius but also from Eurytenes s. I. Similarly, rudis is also excluded from both Eurytenes and Xynobius since it lacks both the ventrally displaced T2 spiracle and a dorsope. Rank is subjective, and thus an argument could be made for retaining both Eurytenes and Xynobius as separate genera. The shared placement of the spiracle laterally on the second metasomal segment, which differs from that in Exothecini, provides an argument against this course of action and for monophyly of the included species. Additionally, based on the above hypoth- esized relationships, recognition of Eury- tenes separate from Xynobius would render the latter paraphyletic if Stigmatopoea is treated as a synonym of Xynobius (as in van Achterberg 2004). Eurytenes and Xynobius also compete for priority since both were described in the same publication (Foerster 1862). Eurytenes has been consistently recognized as a genus since it's initial description; hence my preference for use of this name over Xynobius. This still leaves unresolved the question of the most ap- propriate genus group name(s) for bajulus and bisternaulicus. If Opius is temporarily maintained as a paraphyletic assemblage from which monophyletic groups are extracted as they become recognized and delineated over time, following Wharton (1988), then both bajulus and bisternaulicus can be retained in Opius until their relationships are better understood. A restricted definition for Opius is ultimately desirable, however, and since van Achterberg and Salvo (1997) have provided one, this leaves bajulus and bisternaulicus excluded from Opius, but with Biophthora and Sternaulo- pius, respectively, as available genus-group names. Aulonotus Ashmead, 1900 is also available for species with the second metasomal spiracle placed dorsally and possessing a distinct dorsope, a sculptured precoxal sulcus, and a mesonotal midpit. Fischer (1972) and Fischer and Koponen (1999) treated Aulonotus as a subgenus of Opius, Fischer (1998) elevated it to generic rank, and van Achterberg (2004) listed it as a synonym of Xynobius. Aulonotus is a con- venient place to put species formerly placed in either Xynobius or Eurytenes, but which do not fit the restricted definition presented above. Unfortunately, the name Biophthora, which has remained in obscu- rity since originally proposed by Foerster (1862) and has been treated as a synonym since the works of Marshall (1891) and Dalla Torre (1898), has priority over the 334 Journal of Hymenoptera Research more widely used name Aulonotus. Apodes- mia and Utctes also date from Foerster (1862), and also have been recently treated as genera, but their definitions would have to be expanded considerably to accomo- date bajulus and/or bisternaulicus since the type species of both Apodesmia and Utetes lack a dorsope. Biophthom Foerster, new status, is thus the most suitable genus group name for bajulus and Sternaulopius is the most suitable name for bisternaulicus. Following the discovery of a species from Madagascar (described below) with clypeus and mandibles resembling bajulus but otherwise more closely resembling bisternaulicus, it has become much more challenging to find morphological differ- ences useful for retaining Sternaulopius and Biophthora as separate genera. In addition to similar sculptural features already not- ed, the venation of fore and hind wing is nearly identical (and similar to many other opiines). Nevertheless, I retain them as separate largely because Sternaulopius is an Afrotropical group of tephritid parasitoids and bajulus is a temperate species with the general habitus and short ovipositor sug- gestive of it being a leaf-miner parasitoid. In addition to differences in the ovipositor and shape of the metasoma, bajulus has a rugose, elevated scutellum. Other, seem- ingly more trivial, differences are noted in the descriptions below. Sternaulopius bisternaulicus Fischer, 1965 (Figs 1-15) Sternaulopius bisternaulicus Fischer, 1965: 312- 314, holotype female in MRAC; Fischer 1968: 105, key. Fischer 1971a: 125, catalog; Fischer 1987: 562-564, redescription, figures; van Achterberg 1993: 121, figures; Yu and van Achterberg 2005, electronic catalog. Opius bisternaulicus: Wharton 1988: 355-357. Redescription. — Head (Figs 1-4) in dorsal view 1.8-2.0 X broader than long, 1.40-1.55 X broader than length in lateral view, eyes distinctly bulging, width at eyes 1.05-1.10 X width at temples; face 1.45-1.65 X wider than high; eye in lateral view large, 2.9-3.9 X longer than temple. Frons, vertex, and occiput highly polished, face appearing less polished due to punctation; low, median ridge present from epistomal sul- cus to level of antennal bases, replaced on frons by shallow, median, crenulate line extending to or nearly to median ocellus; area between and immediately posteriorad antenna often weakly rugulose; face dis- tinctly punctate throughout, punctures separated by about their own diameter; frons with 3-5 setae near eye margin, vertex and frons otherwise bare; occipital margin with a single row of setae dorsally. Width of ocellar field 1.1-1.5 X distance from lateral ocellus to eye; width of ocellar field 2.2-2.8 X width of lateral ocellus. Hypostomal carina protruding as a short but distinct flange beneath mandible when mandible closed; occipital carina widely separated from hypostomal carina ventral- ly (separation about equal to basal width of mandible), sharp and distinctly elevated throughout, extending dorsally just below top of eye in lateral view, not reflected medially at dorsal terminus; very shallow- ly curved in lateral view. Malar space distinct, 0.15-0.25 X eye height, a little shorter than basal width of mandible; punctate (similar to face) between distinct malar sulcus and margin of clypeus, smooth and polished laterad sulcus. Clyp- eus about twice wider than high; truncate ventrally in frontal view, slightly protrud- ing ventrally in profile; nearly triangular in outline, the epistomal sulcus only weakly curved; uniformly punctate, pattern of punctation as on face; epistomal sulcus distinctly impressed laterally, more weakly so medially; anterior tentorial pits small, round. Mandibles abruptly narrowing over basal half, more gradually and evenly narrowing from midpoint to apex; outer surface strongly convex; mandibles strong- ly deflected ventrally, broadly exposing labrum. Antenna with 23-30 flagellomeres (27 in holotype), clearly varying with body length; first flagellomere slightly longer Volume 15, Number 2, 2006 335 than second, second subequal to third; first posteromedially, transverse carina varying flagellomere roughly 3 X longer than wide, from distinct to indistinct; propodeal spi- tenth flagellomere roughly 2 X longer than racle minute, situated about midway be- wide; apical flagellomere sharply pointed, tween anterior and posterior margins; but the tip not attenuate. Maxillary palps propodeum separated from metapleuron about equal in length to height of head. by a shallow to deep groove. Metapleuron Mesosoma (Figs 5-9) 1.25-1.35 X longer broadly impressed and carinately rugose than high, 1.65-1.80 X longer than wide, around margins; median plate varying Pronotum dorsally without median pit, from polished, punctate, and largely un- crenulate along posterior margin, other- sculptured to uniformly rugulose. Hind wise polished, unsculptured; pronotum margin of mesopleuron crenulate through- laterally with dorsal, crenulate line extend- out, though sometimes more weakly so ing as a broad, median groove to ventral dorsally, the crenulate impression forming corner, the groove crenulate for most of its nearly a straight line. Precoxal sulcus length, though often more weakly so incomplete posteriorly, but extending most medially, posterior margin crenulate at of way to mid coxa; gradually widening least over ventral 0.5, area between poste- anteriorly, roughly twice as broad anteri- rior crenulate margin and median crenu- orly as posteriorly; crenulate throughout; late groove varying from weakly to mod- precoxal sulcus anteriorly joining broadly erately rugulose, pronotum laterally other- crenulate groove along anterior margin of wise smooth and polished dorsoposteriorly mesopleuron and extending dorsally then and anteriorad median groove. Propleural posteriorly ventrad the setose subtegular flange large, distinct, sharply bent poster- ridge. Sternaulus crenulate throughout, oventrally; separated from remainder of nearly parallel to but distinctly separated propleuron by oblique, weakly to strongly from more dorsally positioned precoxal sculptured groove. Anterior declivity of sulcus, slightly longer than the latter and of mesoscutum densely covered with decum- approximately uniform width throughout, bent setae, setae extending onto base of Fore wing (Fig. 14) with stigma broad, disc, in 2-3 rows along notauli, and as wedge-shaped: widest at origin of r, a scattered patch of somewhat more erect tapered into metacarpus distally; r arising setae between midpit and scutellar sulcus; from about basal 0.4, length of r 0.75-0.90 notauli deeply impressed and sculptured X width of stigma; second submarginal on basal 0.4-0.5 of mesoscutum, abruptly cell large, weakly converging distally, 5- to gradually transforming to very shallow, sided with m-cu distinctly postfurcal, 2RS unsculptured depressions extending pos- strongly reclivous, r-m slightly inclivous, teriorly and sometimes almost impercepti- completely depigmented and descler- bly to oval or tear-drop shaped midpit; otized; 3RSa 1.35-1.70 X longer than 2RS; midpit covering apical 0.2 of disc, well 3RSb 1.45-1.60 X longer than 3RSa; 3RSb separated from distinct transscutal articu- extending to wing margin very close to lation; carinate lateral margin of disc wing apex; RS+M weakly sinuate, nearly crenulate, deeply impressed between te- straight, arising low on almost evenly gula and rugose base of notaulus. Scutellar bowed 1M (the curvature slightly stronger sulcus about 4 X wider than length along posteriorly), IRS 0.35-0.45 X length of 1M; midline; smooth, with 6-8 distinct ridges, large median bulla covering all of 2 Ma, Scutellum evenly convex, not strongly posterior extremities of RS+M and 2RS, elevated; polished throughout, with scat- and anterior portion of m-cu; 3M tubular tered setae. Metanotum with small, low and distinctly pigmented for more than median tubercle. Propodeum densely ru- half its length; lcu-a inclivous, separated gose, the sculpture often somewhat weaker from 1M by nearly its own length; 1st 336 Journal of Hymenoptera Research subdiscal cell closed, 2CUa strongly incli- vous, nearly twice length of tubular 2cu-a; 1-1A weakly bowed towards wing margin, separated near mid-length from the latter by about twice its width. Hind wing (Fig. 15) with RS represented by unpig- mented, spectral crease; 2M and m-cu nearly always weakly but distinctly pig- mented, m-cu strongly bowed with poste- rior end directed towards wing base; 2-1A absent. Metasoma (Figs 10-13) with gaster in dorsal view broadly oval, distinctly taper- ing anteriorly and posteriorly. Petiole 1.0- 1.3 X longer than apical width, apex 1.8- 2.2 X wider than base; largely smooth, polished basal-medially, striate to strigose laterally and over entire apical half, dorsal carinae converging near mid-length but not quite touching, replaced by a low, median ridge over apical half; dorsope distinct, deep. Hypopygium weakly scler- otized and folded along midline; short, with posterior margin moderately protrud- ing medially but apex not extending to tip of metasoma. Ovipositor protruding dis- tinctly beyond metasoma, 1.4-1.5 X longer than mesosoma, upper valve with distinct subapical node, valve slightly narrowed immediately basad node; ovipositor sheath 0.85-0.95 X length of mesosoma, with tuft of long setae over apical third, 2 ventral and 1 lateral row of more widely spaced setae basally. Color mostly dark brown; legs yellow, fifth tarsomere brown, apical 0.15-0.20 of posterior face of hind tibia and hind tarsi dorsally light brown in specimens from Rutshuru and those reared from Psychotria fruits in Kenya, brown infumation on tibia absent or very weakly indicated in speci- mens from Cameroon and those reared from Strombosia fruits in Kenya; flagellum brown to dark brown; scape, occasionally pedicel, most of mandible, labrum, palps, tegula, and pleural areas adjacent tegula yellow; apical margin of clypeus, posterior traces of notauli, extreme base of metaso- mal tergum 2, scutellum posteriorly, and often parts of metanotum reddish brown; metasomal terga and sterna otherwise dark brown in specimens from Rutshuru and Psychotria fruits in Kenya, the laterotergites lighter brown with pale margins, terga yellow brown and sterna pale in specimens from Cameroon, terga 2+3 with large, dark brown spot medially, remainder of meta- soma dorsally and ventrally yellow; wings hyaline. Length of body (exclusive of antenna and ovipositor) 2.5-3.6 mm; of wing 2.6- 3.6 mm; of antenna about 3.1-4.3 mm. Male as in female except as follows: antenna with 23-30 flagellomeres; petiole narrower apically, thus appearing more parallel-sided, 1.25-1.40 X longer than apical width, apex 1.5-1.8 X wider than base; body and wing length more variable, 1.8-3.4 and 2.0-3.3 mm respectively; speci- mens from Cameroon with antenna entire- ly yellow and metasomal terga 2+3 yellow brown with remaining terga brown; speci- mens reared from Strombosia fruits in Kenya also with antenna entirely yellow, but metasomal terga entirely dark brown. Fischer (1965) noted that in males, r was half the length of 2RS in the male paratypes and 2/3rds the length of 2RS in the female holotype; in the longer series examined here, most males had a slightly shorter r cross-vein than females, but the difference was smaller than in Fischer's material. Biology. — Reared from larvae of the following Tephritidae (Diptera) infesting fruit of Moraceae (from Ceratitis MacLeay spp. infesting Antiaris toxicaria (Pers.) Lesch. in Kakamega forest, Kenya), Olaca- ceae (from Ceratitis anonae Graham and /or C. fasciventris (Bezzi) infesting Strombosia scheffleri Engl, in Kakamega), Sapotaceae (from C. anonae infesting Engleropmytum oblanceolatum (S. Moore) T. D. Penn. in Kakamega) and Rubiaceae (from Trirhi- thrum Bezzi and/or Ceratitis infesting Coffea spp. for all Cameroon collections; from Trirhithrum meladiscum Munro infest- ing Psychotria fractinervata Petit in Gata- mayu forest, Kenya and P. mahonii C. Volume 15, Number 2, 2006 337 Wright in Mt. Kenya forest; and from T. senex Munro infesting P. alsophila K. Schum. in Taita Hills, Kenya). See material examined section for details of localities. Bactrocera amplexa (Munro) was also pres- ent in the samples of Strombosia, but the isolated puparia from which wasps emerged were those of Ceratitis. Emer- gence, as in all known opiines, was from the host puparium. Host stage attacked is unknown. Material examined. — Holotype female, DEMO- CRATIC REPUBLIC OF THE CONGO Rut- shuru, 29.V.1936 (L. Lippens) (MRAC). Additional specimens: BURUNDI 1 male(?), Bururi, Urundi, 2200 m, 4.ix.l948 (Francois) (MRAC). CAMEROON 2 males, Akonolinga, viii.1982 (G. Steck), TAMU Quarantine # 82-31, TAMU Voucher # 82; 2 females, 3 males, Nkolbisson, x.1982 (G. Steck), TAMU Quaran- tine #s 82-39 to 82-41, TAMU Voucher #s 87 to 89; 1 male, same locality, 1980 (Perkins) (TAMU). DEMOCRATIC REPUBLIC OF THE CONGO 2 females, 2 males, same locality as holotype, ix-x.1936 & 7.iv.l937 (MRAC). KENYA 4 females, 4 males, Central Province, Gatamayu forest, 0:58.45'S, 36°41.83'E, 17.iv.2001 & 19.X.2001 (R. Copeland); 1 male, Coast Province, Ngangao forest, Taita Hills, 3 21.255'S, 38 20.579'E, 12.viii.2002 (R. Cope- land & R. Wharton); 2 females, 2 males, Eastern Province, Mt. Kenya forest, 0 12.568'S, 37 30.218'E, 6.xii.200l'(R. Copeland); Western Province, Kakamega forest, 4 females, 3 males, 0 14.16'N, 34 51.82'E, reared from Ceratitis infesting fruits of Strombosia scheffleri, 30.iv.2000 (R. Copeland); 1 male, 0 13.14'N, 34 53.76'E, reared from Trirhitlirum senex infest- ing fruits of Englerophytum oblanceolatum, 7.vi.2001 (R. Copeland);l female, 1 male, 0 14.16'N, 34 51.87'E, reared from Ceratitis infesting fruits of Antiaris toxicaria, 3.iii.2000 (R. Copeland); 1 male, same but 0 13.66'N, 34°53.12'E, 19.ii.2001 (TAMU, National Mu- seums of Kenya). Diagnosis. — This species is similar to Sternaulopius duplicatus new species, Biophthora rossicus (Szepligeti) new combi- nation, and B. bajulus new status, in the possession of both a sculptured sternaulus and a sculptured precoxal sulcus. It differs from these species by its more broadly exposed labrum and long ovipositor with distinct subapical node. Remarks. — In overall appearance (body shape, size, and fore wing venation), S. bisternaulicus resembles species of Utetes, but has a distinct dorsope and lacks the sharp carina on the base of the inner side of the hind tibia that characterizes members of the genus Utetes. Since species of both genera have occasionally been reared from tephritids infesting the same host fruits, and both lack the attenuate hypopygium of many of the other parasitoids of fruit- infesting tephritids, careful attention needs to be paid to these distinguishing charac- teristics in order to avoid misidentifica- tions. The most obvious difference between bisternaulicus and the two other species treated in detail here (duplicatus and bajulus) is the more broadly exposed labrum formed by the ventrally deflected mandibles. Compared to bajulus, both bisternaulicus and duplicatus have an evenly and weakly convex, dorsally unsculptured scutellum, a more broadly impressed an- terior margin of the metapleuron and a distinctly longer fore wing cross-vein r. Minor differences between these two and bajulus include a straighter carinate groove on the posterior margin of the meso- pleuron and somewhat coarser propodeal and mesopleural sculpture. There is considerable morphological variation amongst the material available for examination, but I am unable to de- lineate more than a single species. Though not entirely satisfactory, I consider all of the specimens in the material examined section as representing a single, variable species because the series reared from Trirhithrum meladiscum infesting Psi/chotria fruits in central Kenya encompasses most of the range of variation observed in the rest. More specifically, the larger speci- mens reared from Strombosia fruits in western Kenya have slightly larger ocelli, and consequently a broader ocellar field. 338 Journal of Hymenoptera Research Individuals reared from Psychotria fruits in measuring this structure on intact speci- various Kenyan localities, as well as indi- mens and the slightly longer, narrower viduals from Cameroon, have somewhat petiole of the Rutshuru specimens. Perhaps smaller ocelli as in the specimens from the most distinctive difference among the Rutshuru. Specimens from Cameroon tend material examined is in coloration, with to have more extensive, better-developed specimens from Rutshuru and from Psy- sculpture on the pronotum laterally and chotria and Engleraphytum fruits in Kenya a more consistently deep and heavily darker than specimens from Cameroon sculptured oblique groove on the pro- and from Strombosia and Antiaris fruits in pleuron than the larger individuals reared Kenya. Thus, both lighter and darker forms from Strombosia fruits in western Kenya, can be found in Kakamega forest in Individuals reared from Psychotria fruits in western Kenya. central Kenya are highly variable in this The species is distributed across equato- regard, covering the extent of variation rial Africa from coastal Kenya to Camer- seen in all other populations. The setae on oon. The type material of bister naulicus is the median mesoscutal lobe extend further from Rutshuru, in the eastern portion of posteriorly in specimens from Cameroon present day Democratic Republic of the than in the others, but in these the Congo, near the Uganda border, mesoscutum is still distinctly less setose The medially desclerotized hypopygium than in the new species described below. is reminiscent of some Cardiochilinae and The propodeal sculpture is much weaker Microgastrinae and similar in this respect posteriorly in most of the specimens from to the equally short hypopygium found in Cameroon and the Kenyan material reared fruit-infesting tephritid parasitoids of the from Strombosia fruits. The transverse genus Utetes. The functional significance is carina is also distinct and readily visible unknown, in these specimens. The material from Rutshuru and most of the specimens Sternaulopius duplicatus Wharton, reared from Psychotria fruits in Kenya are new species densely sculptured throughout, and as (Figs 33-35) a consequence, the transverse carina is usually indistinct. However, there is some Head in dorsal view 1.85-2.00 X broader variation within localities and in particular than long, 1.50-1.55 X broader than length the large series from Psychotria fruits in lateral view, eyes weakly bulging be- contains some individuals with weaker yond temples; face 1.30-1.45 X wider than propodeal sculpture and a couple of the high; eye in lateral view large, 3.35 (female) males from Cameroon have denser sculp- and 2.90 (male) X longer than temple, ture posteriorly. The sternaulus is a little Frons, vertex, and occiput as in bisternau- broader throughout in Cameroon and licus except median sculpture on frons Strombosia material than in Rutshuru and confined to area immediately posteriad Psychotria material, but is of the same antennal bases; face (Fig. 33) distinctly length and very well developed in both punctate on either side of polished, narrow groups. The second submarginal cell is midridge about as in bisternaulicus but with similarly shaped in all populations, but 1-2 distinct vertical rows of more closely- extreme within-population variation in set punctures on either side of midridge in length of 2RS resulted in the broad range addition to the irregular pattern of punc- for the 3RSa/2RS ratio noted in the above tures. Ocelli smaller than in most individ- description. Variation in the shape of the uals of bisternaulicus; width of ocellar field petiole, on the other hand, is partly 1.05-1.10 X distance from lateral ocellus to a function of the difficulty in accurately eye; width of ocellar field 2.2-2.5 X width Volume 15, Number 2, 2006 339 Figs 33-35. Sternatdopius duplicatus, n. sp., holotype female: 33, face. 34, anterior-dorsal view of mesonotum. 35, lateral habitus, arrow = precoxal sulcus. 340 Journal of Hymenoptera Research of lateral ocellus. Hypostomal and occipital carinae as in bisternaulicus. Malar space short but distinct, 0.15-0.20 X eye height, a little shorter than basal width of mandi- ble; region of malar space sculptured as in bisternaulicus. Clypeus (Fig. 33) a little tal- ler and more oval in male than female, appearing truncate below in both depend- ing on angle of view; nearly flat in profile; uniformly setose and punctate, distinctly but more sparsely punctate than in bister- naulicus; epistomal sulcus impressed later- ally, barely indicated medially; anterior tentorial pits round, a bit larger than in bisternaulicus. Mandibles gradually and evenly narrowing from base to apex; about 3 X longer than median width; outer surface very weakly convex, nearly flat; mandibles barely deflected ventrally, com- pletely concealing labrum when closed (Fig. 33). Antennae broken in both speci- mens, male with 25 flagellomeres remain- ing, female with 18; first flagellomere slightly longer than second, second slightly longer than third; first flagellomere rough- ly 3.5 X longer than wide, tenth flagello- mere more than twice longer than wide and longer in male than female. Maxillary palps curled and length difficult to esti- mate but appear slightly shorter than in bisternaulicus. Mesosoma (Figs 34, 35) 1.35 X longer than high, 1.75 X longer than wide. Pronotum not visible dorsally; laterally as in bisternaulicus. Propleural flange distinct, bent posteroventrally; separated from re- mainder of propleuron by oblique, weakly to strongly sculptured groove. Anterior declivity of mesoscutum densely covered with decumbent setae, these extending onto disc, covering entire median mesos- cutal lobe and anterior and median por- tions of lateral lobes (Fig. 34), though not as densely as on anterior declivity; notauli weaker than in bisternaulicus, impressed and weakly sculptured on basal 0.2-0.3 of mesoscutum, faintly indicated medially, imperceptible posteriorly; midpit small, round, deep, separated from distinct trans- scutal articulation; carinate lateral margin of disc crenulate, deeply impressed be- tween tegula and rugose base of notaulus. Scutellar sulcus a bit narrower medially than in bisternaulicus, about 5.6 X wider than length along midline; smooth, with 6- 8 distinct ridges. Scutellum and metano- tum as in bisternaulicus. Propodeum dense- ly rugose; propodeal spiracle minute, situated about midway between anterior and posterior margins; propodeum sepa- rated from metapleuron by a shallow groove. Metapleuron broadly impressed and carinately rugose around margins; median plate polished and largely un- sculptured. Hind margin of mesopleuron crenulate throughout, the crenulate im- pression forming nearly a straight line. Precoxal sulcus incomplete posteriorly, but extending most of distance to mid coxa; crenulate and of uniform width through- out; precoxal sulcus anteriorly as in bister- naulicus. Sternaulus crenulate throughout, nearly parallel to but distinctly separated from more dorsally positioned precoxal sulcus (the two grooves slightly converg- ing posteriorly), slightly longer than pre- coxal sulcus and of approximately uniform width throughout. Fore wing stigma about as in bisternau- licus but a little more gradually tapered distally; r appearing to arise more proxi- mally, from basal 0.30-0.35; second sub- marginal cell, m-cu, r-m, 1M, IRS, 2RS, 3M, 1-1A, and median bulla as in bisternaulicus; 3RSa about 1.70 X longer than 2RS; 3RSb 1.55-1.60 X longer than 3RSa; 3RSb ex- tending to wing margin very close to wing apex; RS+M straight to nearly so; lcu- a vertical to nearly vertical, separated from 1M by half its length or slightly less; 1st subdiscal cell about as in bisternaulicus but 2cu-a longer, distinctly more than half length of 2CUa. Hind wing about as in bisternaulicus but m-cu shorter, straighter, and more weakly pigmented than in most individuals of the latter; 2-1A absent. Metasoma with gaster in dorsal view broadly oval, distinctly tapering anteriorly Volume 15, Number 2, 2006 341 and posteriorly. Petiole 1.35 (female) and densely setose mesoscutum, and concealed 1.60 (male) X longer than apical width, labrum. apex 1.6 X wider than base in male, base Remarks. — The concealed labrum of this not completely visible in female; striate to species makes it more challenging to argue strigose except in basal depression, dorsal for the exclusion of bajulus from Sternaulo- carinae extending to apex though weaker pius since the differences in appearance of posteriorly, weakly converging throughout the clypeus and mandibles between dupli- length; dorsope as in bisternaulicus. Hypo- catus and bajulus are relatively minor. As pygium short, not extending to tip of noted above in the general discussion metasoma; shape about as in bisternaulicus under Sternaulopius, recognition of Bio- but difficult to discern because of deflected phthora as a separate genus is based ovipositor. Ovipositor about 0.75 X length primarily on the nature of the scutellum of mesosoma, upper valve without distinct and in this duplicatus and bisternaulicus are subapical node; ovipositor sheath about markedly different from bajulus. Other 0.35 X length of mesosoma, with tuft of essential features that assist in the place- long setae apically, and more widely ment of duplicatus in Sternaulopius rather spaced setae basally. than Biophthora are found in the details of Color mostly dark brown to black; scape, sculpturing of the malar region, meso- clypeus, palps, tegula, base of metasomal pleuron, and propodeum as well as the tergum 2 and legs yellow, the coxae and punctation and shape of the clypeus. See trochanters more pale in male; mesosoma further remarks under bisternaulicus. dark brown to black dorsally and ventrally, The length of the ovipositor and sheath with small reddish patches laterally in given in the above description are esti- female, male more extensively reddish to mates since the holotype was not dissected, reddish-yellow; petiole reddish brown, However, since the base of the ovipositor is terga 2+3 yellow-brown to reddish brown, partially exposed in the holotype, the remaining metasomal terga dark brown; estimate is fairly accurate. The ovipositor wings hyaline or nearly so. of this species is thus considerably shorter Length of body (exclusive of antenna than in bisternaulicus, and at rest probably and ovipositor) about 2.3 mm; of wing 2.6- protrudes only slightly beyond the apex of 2.7 mm. the metasoma. As in bajulus, there is no . , , , , A discernible subapical node on the dorsal Material examined. — Holotype female MADA- , , ,. , ._. ., , ,. GASCAR. Label with 4 lines as follows: "COLL. Valve in du?™tus ^ggeshng that duphca- MUS. CONGO", "Madagascar: Ankaratra", tus maV ovlPoslt mto an earher host sta8e "IV-1944" and "A. Seyrig" (deposited in than bisternaulicus. MRAC). ^ rne presence of a sternaulus (in Paratype male MADAGASCAR same label addition to the precoxal sulcus) is over- data as holotype (MRAC). looked, duplicatus runs to Opius (Firkins) Diagnosis. — This species is characterized castaneus Granger in the keys to Opiinae by the presence of both a sculptured published by Granger (1949) and Fischer sternaulus and sculptured precoxal sulcus, (1972, 1987). Opius castaneus is the type a combination known only in Sternaulopius species of Frekius Fischer, 1971b. The and Biophthora among the Opiinae. Ster- holotype of castaneus lacks a sternaulus naulopius duplicatus lacks the elevated, but surprisingly the hind tibia is carinate dorsally sculptured scutellum of the spe- basally as in Utetes. 1 therefore transfer cies placed in Biophthora below, and is most Granger's species to Utetes where the new readily differentiated from the only other combination is Utetes [Frekius) castaneus valid species of Sternaulopius, S. bisternau- (Granger). I retain, at least temporarily, the licus, by the shorter ovipositor, more name Frekius as a valid subgenus pending 342 Journal of Hymenoptera Research a badly needed revision of the genus Utetes. Typical Utetes have a broadly ex- posed labrum and thus the inclusion of Frekius in Utetes is not without problems since castaneus does not have an exposed labrum. The parallels between Utetes and Sternaulopius in the unusually variable nature of the labral exposure are also deserving of closer attention. Sternaulopius duplicatus is known from only two specimens; the species name refers to the two parallel grooves on the mesepisternum (precoxal sulcus and ster- naulus). Biophthora bajulus (Haliday, 1837) (Figs 16-20) Opius bajulus Haliday, 1837: 214; Marshall 1891: 43, redescription, key, English; Marshall 1894: 327-328, redescription, key, French; Dalla Torre 1898: 59, catalog; Szepligeti 1904: 164, checklist; Fischer 1958: 58-60 redescription, keys; Fischer 1971a: 46, catalog; Papp 1981b: 35, key; Tobias and Jakimavicius 1986: 96-97, key, figures; Belokobylskij et al. 2003: 393, checklist, clarification of date of original description in this and other 19th century publications relevant to Opiinae. Biophthora bajulus: Foerster 1862: 260. Sternaulopius beieri Fischer, 1968: 103-105, new synonym; Fischer 1971a: 125, catalog; Fischer 1972: 478-479, redescription; Papp 1981a: 273-274, distribution, diagnosis; Papp 1981b: 135, redescription, key; Quicke et al. 1997: 25- 36, venom glands, relationships; Tobias 1998: 559-561, key; Belokobylskij et al. 2003: 396, checklist; Yu and van Achterberg 2005, electronic catalog. Opius (Xynobius) bajulus: Fischer 1972: 88-90, redescription, key, figures; Tobias and Jaki- mavicius 1986: 29, key, figures. Xynobius bajulus: van Achterberg, 1997: 18, original description listed as published in 1836, following Horn and Schenkling (1928); van Achterberg 2004: 315. Redescription. — Head in dorsal view 1.8- 1.9 X broader than long, 1.3-1.4 X broader than length in lateral view, as wide at temples as at eyes; face (Fig. 18) 1.7-1.8 X wider than high; eye in lateral view 1.2-1.3 X longer than temple. Face (medially), vertex, frons and occiput polished; low median ridge present from epistomal sul- cus to level of antennal bases, replaced by shallow groove on frons; face sparsely and finely punctate medially, punctures sepa- rated by at least their own diameter, weakly rugulose and dull (unpolished) adjacent eye margin; frons with 3-5 setae along eye margin, otherwise bare and unsculptured; vertex sparsely setose. Ocelli and ocellar field small, width of ocellar field 0.70-0.75 X distance between lateral ocellus and eye; width of ocellar field 3.0- 3.4 X width of lateral ocellus. Hypostomal carina protruding as a short but distinct flange beneath mandible when mandible closed; occipital carina very widely sepa- rated from hypostomal carina ventrally (by a distance greater than basal width of mandible), relatively short, extending dor- sally just above middle of eye in lateral view, not reflected medially at dorsal terminus, broadly curved in lateral view due to prominantly swollen gena. Malar space broad, 0.3-0.4 X eye height, about equal to basal width of mandible; rugu- lose-strigose (Fig. 18), with sculpture ex- tending from malar sulcus to ventral- lateral margin of clypeus; malar sulcus complete from eye to subgenal margin but barely distinguishable from adjacent sculp- ture. Clypeus broadly hemi-elliptical, trun- cate and sharpy margined ventrally, punc- tate and setose along dorsal margin, other- wise polished and bare; epistomal sulcus distinctly impressed throughout, though slightly less so medially; anterior tentorial pits narrow, slit-like. Mandible gradually and evenly narrowing from base to apex; slightly more than twice longer than basal width but less than 3 X longer than median width; weakly deflected ventrally, expos- ing labrum in a narrow but distinct gap between clypeus and mandibles when the latter are closed (Fig. 18). Antenna with 24- 26 flagellomeres; first flagellomere slightly longer than second; apical flagellomere sharply pointed, but tip not attenuate. Volume 15, Number 2, 2006 343 Maxillary palps shorter than height of sculptured. Hind margin of mesopleuron head. finely crenulate throughout, the crenulate Mesosoma (Figs 17, 19, 20) 1.2-1.3 X impression emarginate near middle. Pre- longer than high, 1.8-1.9 X longer than coxal sulcus (Figs 17, 19) incomplete pos- wide. Pronotum dorsally without median teriorly, extending slightly more than half pit, crenulate along posterior margin, oth- distance from anterior margin to mid coxa; erwise polished, unsculptured; pronotum finely crenulate throughout, weakly ta- laterally finely sculptured throughout ex- pered posteriorly; precoxal sulcus weakly cept largely smooth anteriorad median separated from crenulate groove along groove. Propleural flange large, distinct, anterior margin of mesopleuron or con- sharply bent posteroventrally; separated nected to groove only as a faint trace; from remainder of propleuron by oblique, anterior margin of mesopleuron finely strongly sculptured groove. Anterior de- crenulate, the sculpture extending poster- clivity of mesoscutum densely covered iorly ventrad subtegular ridge, with a few with decumbent setae; disc bare except weak striae extending ventrally towards for scattered row of setae along each middle of mesopleuron from subtegular notaulus and on either side of midpit; ridge. Sternaulus (Figs 17, 19) finely cren- notauli extending onto anterior 0.4 of disc ulate throughout, parallel to but distinctly as crenulate grooves, abruptly transform- separated from more dorsally positioned ing to shallow, unsculptured depressions precoxal sulcus; of uniform width through- extending posteriorly to narrow, slit-like out. Midventral groove crenulate; postpec- midpit, largely to completely obscured at tal carina completely absent, base by rugose sculpture; midpit (Fig. 19) Fore wing stigma broad, wedge-shaped: weakly sculptured, covering apical 0.25 of widest at origin of r, tapered into meta- disc, extending to barely perceptible trans- carpus distally; r arising from basal 0.35, r scutal articulation; lateral margin of me- short, at most half as long as width of soscutum crenulate, deeply impressed be- stigma; size and shape of second sub- tween tegula and rugose base of notaulus. marginal cell as in bisternaulicus; m-cu Scutellar sulcus about 4-5 X wider than distinctly postfurcal, r-m and 2Ma slightly length along midline, with 6-8 ridges more distinct than in most specimens of partially obscured by rugulose sculpture, bisternaulicus, due to trace of pigmentation; Scutellum (Fig. 19) arising vertically from 3RSa 1.65-1.75 X longer than 2RS; 3RSb posterior margin of scutellar sulcus, dis- 1.2-1.3 X longer than 3RSa; 3RSb extend- tinctly elevated above plane of mesoscu- ing to wing margin near tip but not as close turn; dorsal surface flattened, densely to apex as in bisternaulicus and duplicatus; rugose throughout, with sculpture extend- RS+M weakly but distinctly sinuate, aris- ing to metanotum. Metanotum with small, ing low on almost evenly bowed 1M (the low median tubercle. Propodeum (Fig. 19) curvature slightly stronger posteriorly), densely and finely rugose, the sculpture IRS 0.30-0.35 X length of 1M; 3M tubular somewhat weaker posterior-medially, and distinctly pigmented for slightly less transverse carina weak, barely discernible; than half its length; lcu-a vertical to very propodeal spiracle minute, situated about weakly inclivous, separated from 1M by midway between anterior and posterior about 0.5-0.7 X its length; 1st subdiscal cell margins; propodeum separated from me- closed, 2CUa strongly inclivous, distinctly tapleuron by a shallow groove margined longer than tubular 2cu-a; 1-1 A weakly medially by a low carina posteriorly, bowed towards wing margin, separated Metapleuron narrowly impressed and cari- near mid-length from the latter by nearly 3 nately rugose around margins; median X its width. Hind wing (Fig. 16) as in plate polished, punctate, and largely un- bisternaulicus except m-cu not so strongly 344 Journal of Hymenoptera Research curved, nearly reaching wing margin and 2-1A short but distinct. Metasoma (Fig. 17) with gaster in dorsal view nearly parallel-sided, gradually ta- pering posteriorly. Petiole short, length 0.85-0.95 X apical width; 2.35-2.45 X wider at apex than at base; sparsely striate to strigose (Fig. 19), dorsal carinae distinct basally, difficult to distinguish from sur- rounding sculpture apically, apparently extending to apex, not converging; dorsope small, round, deep. Spiracle of second metasoma] tergum positioned as in bister- naulicus, dorsad and nearly adjacent lateral crease of median tergite. Hypopygium short, apex not extending to tip of meta- soma. Ovipositor very short, not extending beyond apex of metasoma, upper valve without subapical node; ovipositor sheath about 0.25 X length of mesosoma, with tuft of long setae apically. Head and mesosoma dark brown to black; tegula, metasoma and coxae dark reddish brown, fore coxa sometimes more yellowish; tarsomeres 5 and sometimes 4 dark brown, remainder of legs dark yel- low; clypeus and mandibles yellow; palps light brown; wings hyaline or nearly so. Length of body (exclusive of antenna and ovipositor) about 2.6-2.7 mm; of wing 2.8-2.9 mm; of antenna 3.0-3.1 mm. Material examined. — Holotype female of baju- lus, ENGLAND (Walker) (NMID). Van Achter- berg (1997) noted that the holotype of bajulus is a female but was listed in the original de- scription as a male probably because the genitalia are not readily visible. Holotype female of beieri, GERMANY "Sach- sen Altenbach bei Wurzen Coll. Dr. R. Krieger 31.5.93" (ZMHB). Additional specimens: 1?, AUSLRIA, Thurin- gen (NHMW); CZECH REPUBLIC, 1?, Hradek (NHMW); 1?, Moravia, Brno (NHMW); 1 female, IRELAND, County Kildare, R. Canal, 18.vi.1944 (Stelfox) (USNM). Distribution. — Previously recorded from Czech Republic, Denmark, Germany, Hungary, England, Ireland, and Turkey (Fischer 1971a, Papp 1981a, Quicke et al. 1997). The specimen noted above from Thiiringen in NHMW is assumed to be from Austria but this new record needs verification. The holotype female of beieri has the date hand-written sideways on the label, "Sachen" and "Dr. R. Krieger" printed, and the specific locale hand- written and difficult to read. Fischer (1968) interpreted the specific locality as Altenberg, which was a well-known local- ity in Saxony of that time period. However, there was also a small village called Altenbach just east of Leipzig near Wur- zen, and I interpret the label to read Altenbach rather than Altenberg. Data on the holotype of bajulus are provided by Marshall (1891) and van Achterberg (1997). Diagnosis. — Biophthora is readily recog- nized by the presence of both a sculptured sternaulus and a sculptured precoxal sul- cus in addition to the rugose dorsal surface of the elevated scutellum and a deep dorsope. The scutellar feature separates bajulus from members of the genus Ster- naulopius, but not from Biophthora rossicus (Szepligeti, 1901), new combination, which is nearly identical to bajulus (including the presence of a true sternaulus). The holo- type of rossicus (HNHM) has the notauli better developed than in specimens I have seen of bajulus (including beieri). Remarks I exclude bajulus from Sternaulopius, despite the presence in bajulus of both a crenuate sternaulus and a sculptured precoxal sulcus; bajulus differs from the species of Sternaulopius by the more rugose malar region, the more parallel-sided me- tasoma, and in particular the sculpture and shape of the scutellum. The presence of a true sternaulus in bajulus is hypothesized as a case of remarkable convergence with the condition in Sternaulopius. The sternau- lus of bajulus, though distinct when not obscured by glue or the position of the legs (as is often the case), is decidedly finer and weaker than in bisternaulicus and duplicatus. See remarks section under the genus Ster- Volume 15, Number 2, 2006 345 naulopius for rationale regarding recogni- tion of Biophthora and placement of bajulus therein. The synonymy between beieri and bajulus has undoubtedly been overlooked because few specimens have been available for study and the sternaulus in bajulus is difficult to see relative to the better de- veloped sternaulus in bisternaulicus. Thus, both Fischer (1972) and van Achterberg (2004) placed bajulus in Xynobius on the basis of more readily visible characters such as the sculptured scutellum and the dorsope. Compared to other species placed in Xynobius, bajulus has a more completely concealed labrum, with the clypeus broader, flatter, truncate and not as sharply margined ventrally and the mandibles are flatter and less strongly narrowed from base to apex. The fore wing venation is very different from that of typical Xynobius due to the shape of the stigma, the postfurcal m-cu, the relatively short 1M, the strongly reclivous 2RS and less strong- ly bowed 1-1A. Except for the notably shorter cross-vein r, the fore wing of bajulus is identical in all important respects to that of bisternaulicus and duplicatus. The present study focuses on the mor- phological differences and similarities amongst Biophthora, Eurytenes s. /., and Sternaulopius, but does little to resolve their phylogenetic relationships, either relative to one another or within the Opiinae as a whole. Quicke et al. (1997), using venom gland morphology to assess relationships amongst a wide variety of opiine and alysiine braconids, suggested a relationship between beieri and various species of Xynobius based on venom gland apparatus of a specimen from Denmark. The utility of the tripartite reservoir and notably small glands will need to be reassessed now that the classification has been considerably altered, especially since the glands of caelatus were not examined. Nevertheless, the hypothesis of a close relationship between Biophthora and Eurytenes s. I. cannot be rejected and is deserving of more detailed study as is the possible relationship between Sternaulopius and Utetes. The differences noted by Wharton (1988) between bisternaulicus and bajulus relative to hind wing RS and m-cu have not held up upon examination of more material, though another potentially useful character (presence vs. absence of 2-1A) has been discovered. The ovipositor and sheath of bajulus are distinctly shorter than in bister- naulicus, and also appear to be shorter than in duplicatus. Although the ovipositor is very short in bajulus, the exact length could not be measured accurately on the avail- able material. Papp (1981a) has provided some addi- tional features for separation of bajulus (as beieri) from rossicus. Since the holotype and only known specimen of rossicus is male and bajulus is rare, the separation of the two species will eventually need to be examined in more detail. ACKNOWLEDGEMENTS 1 am particularly grateful to J. O'Connor (NMID) for examining Haliday's type of bajulus in Dublin, as well as the following curators and their assistants for extended loans and /or access to material in their care: E. De Coninck, S. Hanot, and M. De Meyer (MRAC), H. Zettel and M. Fischer (NHMW), F. Koch (ZMHB), C. Villemant (MNHN), J. Papp (HNHM), and P. Marsh and D. Smith (USNM). Both R. Copeland and G. Steck were instrumental in collection of fruits that produced the reared Sternaulopius, and special thanks to R. Copeland for isolating the puparia from Strombosia. This work has also benefitted from discussions with C. van Achterberg about the nature of the braconid sternaulus and G. Gibson regarding thoracic musculature (though all interpretations of homology and any associated errors are my own). Special thanks to Jim Ehrman at the Digital Micros- copy Facility, Mount Allison University, Sackville, NB, Canada for doing all of the SEM work and to Matt Yoder and Karl Roeder who provided much-needed assistance in acquiring and assembling the other figures. This project was supported primarily by NSF/PEET grant no. DEB 0328922 and in part by Initiative for Future Agriculture and Food Systems grant no. 00-52103-9651 from the USDA/ CSREFS, USAID grant no. PCE-G-00-98-0048-00 (the latter two in collaboration with ICIPE in Kenya), and an NSF/ REU supplement #0616851 to the PEET grant. 346 Journal of Hymenoptera Research LITERATURE CITED Ashmead, W. H. 1900. Some changes in generic names in the Hymenoptera. The Canadian Entomologist 32: 368. Belokobylskij, S. A., A. Taeger, C. van Achterberg, E. Haeselbarth, and M. Riedel. 2003. Checklist of the Braconidae of Germany (Hymenoptera). Beitrdge zur Entomologie 53: 341^35. Bohart, R. M. and A. S. Menke. 1976. Sphecid wasps of the world, a generic revision. University of Califor- nia Press, Berkeley. 695 pp. Copeland, R. S., R. A. Wharton, Q. Luke, and M. De Meyer. 2002. Indigenous hosts of Ceratitis capitata (Diptera: Tephritidae) in Kenya. Annals of the Entomological Society of America 95: 672-694. Dalla Torre, C. G. 1898. 5. Subfam. Opiinae. Pp. 54-70. in: Catalogus Hymenopterorum hucusque descrip- torum systematicus et synonymicus IV: Braconidae. G. Engelmann, Leipzig. 323 pp. Daly, H. V. 1963. Close-packed and fibrillar muscles of the Hymenoptera. Annals of the Entomological Society of America 56: 295-306. Daly, H. V. 1964. Skeleto-muscular morphogenesis of the thorax and wings of the honey bee Apis mellifera (Hymenoptera: Apidae). University of California Publications in Entomology 39: 1-77. Fischer, M. 1958. Die europaeischen Arten der Gattung Opius Wesm. Tiel la. (Hymenoptera, Braconidae). Annali del Museo Civico di Storia Naturale di Genova 70: 33-70. Fischer, M. 1965. Zwei neue Opiinen-Gattungen aus dem Kongo (1. Beitrag ueber die Opiinae des Musee Royal de l'Afrique Centrale) (Hymenop- tera, Braconidae). Revue de Zoologie et de Botanique Africaines 71: 309-323. Fischer, M. 1968. Sternaulopius in Deutschland (Hy- menoptera, Braconidae, Opiinae). Aivialen des Naturhistorischen Museum in Wien 72: 103-106. Fischer, M. 1971a. Index of Entomophagous Insects. World Opiinae. Le Francois, Paris. 189 pp. Fischer, M. 1971b. Die aethiopischen Opius-Arten der Sektion A, aufgeteilt auf die Untergattungen (Hymenoptera, Braconidae, Opiinae). Annalen des Naturhistorischen Museum in Wien 75: 387-433. Fischer, M. 1972. Hymenoptera Braconidae (Opiinae I). Das Tierreich 91: 1-620. Fischer, M. 1977. Hymenoptera Braconidae (Opiinae II-Amerika). Das Tierreich 96: 1-1001. Fischer, M. 1986. Neue Bestimmungsschlussel fuer palaearktische Opiinae, neue Subgenera, Rede- skriptionen und eine neue Art (Hymenoptera, Braconidae). Annalen des Naturhistorischen Muse- um in Wien 88/89: 607-662. Fischer, M. 1987. Hymenoptera Braconidae Opiinae III - aethiopische, orientalische, australische und oceanische Region. Das Tierreich 104: 1-734. Fischer, M. 1998. Neue taxonomische Untersuchungen ueber Madenwespen der Alten Welt mit beson- derer Beruecksichtigung der Gattungen Eurytenes Foerster, Aulonotus Ashmead, Biosteres Foerster und der Untergattung Gastrosema Fischer (Hyme- noptera, Braconidae: Opiinae). Linzer Biologische Beitriige 30: 21-51. Fischer, M. and M. Koponen. 1999. A survey of Opiinae (Hymenoptera, Braconidae) of Finland, part 1. Entomologica Fennica 10: 65-93. Foerster, A. 1862. Synopsis der Familien und Gattun- gen der Braconen. Verhandlungen des Naturhistor- ischen Vereines Preussischen der Rheinlande und Westphalens 19: 225-288. Gibson, G. A. P. 1985. Some pro- and mesothoracic structures important for phylogenetic analysis of Hymenoptera, with a review of terms used for the structures. The Canadian Entomologist 117: 1395-1443. Gibson, G. A. P. 1986a. Evidence for monophyly and relationships of Chalcidoidea, Mymaridae and Mymarommatidae (Hymenoptera: Terebrantes). The Canadian Entomologist 118: 205-240. Gibson, G. A. P. 1986b. Mesothoracic skeletomuscu- lature and mechanics of flight and jumping in Eupelminae (Hymenoptera, Chalcidoidea: Eupel- midae). The Canadian Entomologist 118: 691-728. Gibson, G. A. P. 1993. Groundplan structure and homology of the pleuron in Hymenoptera based on a comparison of the skeletomusculature of Xyelidae (Hymenoptera) and Raphidiidae (Neu- roptera). Memoirs of the Entomological Society of Canada 165: 165-187. Gibson, G. A. P., J. D. Read, and R. Fairchild. 1998. Chalcid wasps (Chalcidoidea): illustrated glossa- ry of positional and morphological terms. URL - http://res2.agr.ca/ecorc/apss/chglintr.htm. June 15, 1998. Granger, C. 1949. Braconides de Madagascar. Mem- oircs de L'Institut de Recherche Scientifique de Madagascar Serie A. Biologic Animate 2: 1-428. Haliday, A. H. 1837. Essay on parasitic Hymenoptera. Entomologist's Monthly Magazine 4: 203-221. Horn, W. and S. Schenkling. 1928-1929. Index Litter- aturae Entomologicae. Serie 1. Die Welt-Literatur ueber die gesamte Entomologie bis inklusive 1863. Berlin, Dahlem. 1426 pp. Marsh, P. M. 1971. Keys to genera of the families Braconidae, Aphidiidae, and Hybrizontidae (Hy- menoptera). Annals of the Entomological Society of America 64: 841-850. Marshall, T. 1891. A monograph of British Braconidae. Part IV. Transactions of the Entomological Society of London 1891: 7-61. Marshall, T. 1894. Les braconides (suite), 10 Tribu - Opiidae. Pp. 280-359 in: E. Andre, ed. Species des Hymenopteres d'Europe et d'Algerie 5. Bouffaut freres, Gray, 635 pp. Mason, W. R. M. 1964. The genus Chrysopophthorus Goidanich (Hymenoptera: Braconidae). The Cana- dian Entomologist 96: 1005-1017. Volume 15, Number 2, 2006 347 Muesebeck, C. F. W. 1970. The Nearctic species of Orgilus Haliday (Hymenoptera: Braconidae). Smithsonian Contributions to Zoology 30: 1-104. Muesebeck, C. F. W. and L. Walkley. 1951. Braconidae. Pp. 90-184 in: C. F. W. Muesebeck, K. V. Krombein, and H. K. Townes, eds. Hymenoptera of America north of Mexico. Synoptic catalog. Mono- graph No. 2. U.S. Government Printing Office, Washington D. C. 1420 pp. Papp, J. 1981a. A survey of the braconid fauna of the Hortobagy National Park (Hymenoptera, Braco- nidae), I. The Fauna of the Hortobdgx/ National Park, 1981: 255-274. Papp, J. 1981b. Fuerkeszdarazs-Alkatuak IX/A. - Ich- neumonoidea IX/A. Fauna Hungariae 144: 1-161+. Quicke, D. L. J., K. van Achterberg, and H. C. J. Godfray. 1997. Comparative morphology of the venom gland and reservoir in opiine and alysiine braconid wasps (Insecta, Hymenoptera, Braconi- dae). Zoologica Scripta 26: 23-50. Richards, O. W. 1956. Hymenoptera. Introduction and keys to families. Handbooks for the Identification of British Insects 6 (1): 1-94. Sharkey, M. J. and R. A. Wharton. 1997. Morphology and terminology. Pp. 19-37 in: R. A. Wharton, P. M. Marsh, and M. I. Sharkey, eds. Manual of the Neiv World Genera of the Family Braconidae (Hyme- noptera). The International Society of Hymenop- terists, Washington, D. C. 439 pp. Shaw, M. R. and T. Huddleston. 1991. Classification and biology of braconid wasps. Handbooks for the identification of British Insects 7 (11): 1-126. 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. Entomophaga 31: 59-67. Szepligeti, G. 1901. Hymenoptera. Pp. 121-169 in: Zoologische Ergebnisse der dritten Asiatischen For- schungsreise des Grafen Eugen Zichy. Vol. 2. Szepligeti, G. 1904. Hymenoptera. Braconidae. 24. Subfamily Opiinae. Genera Insectorum 22: 158-167. Thomson, C. G. 1895. LII. Bidrag till Braconidernas kannedom. Opuscula Entomologica (Lund) 20: 2141-2339. Tobias, V. I. 1977. The genus Opius Wesm. (Hyme- noptera, Braconidae) as parasites of fruit flies (Diptera, Tephritidae). Entomologicheskoe Obozre- nie 56: 420-430. Tobias, V. I. 1998. Subfamily Opiinae. Pp. 558-655 in: A. P. Ler, ed. Key to the insects of Russian Far East. 4. Neuropteroidea, Mecoptera, Hymenoptera. 3. Hy- menoptera (part). Dal'nauka, Vladivostok. 706 pp. Tobias, V. I. and A. B. Jakimavicius. 1986. Subfamily Opiinae. Pp. 7-100 in: G. S. Medvedev, ed. Keys to the insects of the European part of the USSR. 3. Hymenoptera. 5. Braconidae. Academia Nauk, Leningrad. Townes, H. 1969. The genera of Ichneumonidae, part 1. Memoirs of the American Entomological Institute 11: 1-300. Van Achterberg, C. 1975. A revision of the tribus Blacini (Hymenoptera, Braconidae, Helconinae). Tijdschrift voor Entomologie 118: 159-322. Van Achterberg, C. 1988. Revision of the subfamilv Blacinae Foerster (Hymenoptera, Braconidae). Zoologische Verhandelingen 249: 1-324. Van Achterberg, C. 1993. Illustrated key to the sub- families of the Braconidae (Hymenoptera: Ichneu- monoidea). Zoologische Verhandelingen 283: 1-189. Van Achterberg, C. 1994. New morphological terms. Ichnews 14: 5. Van Achterberg, C. 1997. Revision of the Haliday collection of Braconidae (Hymenoptera). Zoolo- gische Verhandelingen 314: 1-115. Van Achterberg, C. 2004. New Indo-Australian sub- genera and species of the genera Xynobius Foerster and Ademoneuron Fischer (Hymenoptera: Braconidae: Opiinae). Zoologische Mededelingen 78: 313-329. Van Achterberg, C. and A. Salvo. 1997. Reared Opiinae (Hymenoptera: Braconidae) from Argen- tina. Zoologische Mededelingen 71: 189-214. Viereck, H. L. 1916. Guide to the insects of Connecti- cut part III. The Hymenoptera, or wasp-like insects, of Connecticut. Connecticut State Geology and Natural History Survey Bulletin 22: 1-824. Walker, F. 1860. Characters of some apparently undescribed Ceylon insects. Annals and Magazine of Natural History 3 (5): 304-311. Wesmael, C. 1835. Monographic des Braconides de Belgique. Nouveaux Memoires de V Academic Royale des Sciences el Belles-lettres de Bruxelles 9: 1-252. Wharton, R. A. 1977. New World Aphaereta species (Hymenoptera: Braconidae: Alysiinae), with dis- cussion of terminology used in the tribe Alysiini. Annals of the Entomological Society of America 70: 782-803. Wharton, R. A. 1986. The braconid genus Alysia (Hymenoptera): a description of the subgenera and a revision of the subgenus Alysia. Systematic Entomology 11: 453-504. Wharton, R. A. 1987. Changes in nomenclaure and classification of some opiine Braconidae (Hyme- noptera). Proceedings of the Entomological Society of Washington 89: 61-73. Wharton, R. A. 1988. Classification of the braconid subfamily Opiinae (Hymenoptera). The Canadian Entomologist 120: 333-360. Wharton, R. A., P. M. Marsh, and M. J. Sharkey, eds. 1997 Manual of the New World Genera of the Family Braconidae (Hymenoptera). The International Soci- ety of Hymenopterists, Washington, D C. 434 pp. Yu, D. S. and C. van Achterberg. 2005. Interactive catalogue of World Ichneumonoidea. Compact Disc. Taxapad, Vancouver. INSTRUCTIONS FOR AUTHORS General Policy. The Journal of Hymenoptera 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 acceptable if the species has economic importance or provides new data on the biology or evolution of the genus or higher taxon. 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 September (for the April issue) and 1 March (for the October issue). Format and Preparation. Authors are strongly encouraged to submit manuscripts electronically to the editor at the email address below, and in the format specified below. On the upper left of the title page give name, address, telephone and fax numbers, and email address of the author to whom all correspondence 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. Upon final acceptance of a manuscript, the author should provide the editor with an emailed IBM formatted electronic version. CD-ROMs or 3.5 inch floppy disks are acceptable. Because symbols and tables are not always correctly translated it is best to also send a printed copy of the manuscript. Preferred word processing programs are Microsoft Word and WordPerfect. 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. CDs and 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 mm) and allow room for legends to 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. 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.l)" "(Table 1)". Measurements should be in the metric system. Electronic plates may be submitted on disc, via email or uploaded to an ftp site (instructions will be given). They must be fully composited, labeled, and sized to fit the proportions of the journal page. Line art should be scanned at 1200 dpi (minimum input resolution is 600 dpi). Color or grayscale (halftone) images should have a dpi of 300-350. Color files should be in CMYK and not RGB. Graphics should be submitted as TIFF, Adobe Illustrator or EPS files. No PowerPoint or Word /Word Perfect files with images embedded in them are acceptable. All papers must conform to the International Code of Zoological Nomenclature. The first mention of a plant or animal name should include the full scientific name including the authority. Genus names should not be abbreviated 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 Recognized institutions (not private collections). Voucher specimens should be designated for specimens used in behavioral or autecological studies, and they should be deposited similarly. DNA sequences must be deposited in GenBank/EMBL/DNA Databank of Japan. Acceptance of taxonomic papers will not require use of cladistic methods; however, authors using them will be expected to specify the phylogenetic program used, including discussion of program options used. A data matrix should be provided for morphological characters. 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 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 addressed to: Dr Gavin Broad Centre for Ecology & Hydrology Monks Wood, Abbots Ripton Huntingdon PE28 2LS, UK Phone: +44(0)1487 772406; Fax: +44(0)1487 773467; Email: gabro@ceh.ac.uk SNirTHSONIAN INSTrTUTION L»"*"J" 3 9088 01283 7498 CONTENTS (Continued from front cover) YAHYA, B. E. and S. YAMANE. Nesting biology of a tropical myrmicine ant, Myrmicaria arachnoides (Formicidae), in West Java, Indonesia 307 NOTES: MAUSS, V. Observations on flower association and mating behaviour of the pollen wasp species Celonites abbreviatus (Villers, 1789) in Greece (Hymenoptera: Vespidae, Mas- arinae) 266 ROIG-ALSINA, A. Hylaeus punctatus (Brulle) (Colletidae), a Palaearctic bee long estab- lished in South America 286