J ouirnsLi or Hymenoptera esearch August 1995 ISSN #1070-9428 CONTENTS BANKS, D. Male nest defense in the digger wasp Cerceris binodis (Hymenoptera: Spheci- dae) 77 CARVER, M. Euryischomyia Girault (Hymenoptera: Chalcidoidea: Aphelinidae: Eriaporinae: EuryischiLni) 64 FINNAMORE, A. T. Revision of the world genera of the subtribe Stigmina (Hymenoptera: Apoidea: Sphecidae: Pemphredoninae), part 1 204 GESS, F. W. Descriptions of the male of Riekia nocatunga Richards, the male and two strik- ingly distinct sympatric colour forms of Riekia confluens (Snelling) and the male of Rolandia angulata (Richards) (Hymenoptera: Vespidae: Masarinae) from Aus- tralia 33 GESS, F. W., S. K. GESS and R. W. GESS. An Australian masarine, Rolandia angulata (Rich- ards) (Hymenoptera: Vespidae): nesting and evaluation of association with Good- enia (Goodeniaceae) 25 GOKHMAN, V. E. and D. L. J. QUICKE. The last twenty years of parasitic Hymenoptera karyology: an update and phylogenetic implications 41 HANSSON, C. Revised key to the Nearctic species of Chrysocharis Forster (Hymenoptera: Eulophidae), including three new species 80 HEYDON, S. L. A review of the North American species of Thinodytes Graham and Mauleus Graham (Hymenoptera: Pteromalidae) 1 LATTKE, J. E. Revision of the ant genus Gnamptogenys in the New World (Hymenoptera: Formicidae) 137 PULAWSKI, W. J. The wasp genus Tachytella Brauns, 1906 (Hymenoptera: Sphecidae) . . . 121 (Continued on back cover) INTERNATIONAL SOCIETY OF HYMENOPTERISTS Organized 1982; Incorporated 1991 OFFICERS FOR 1995 Donald L. J. Quicke, President James M. Carpenter, President-Elect Michael E. Schauff, Secretary John T. Huber, Treasurer Paul M. Marsh, Editor Subject Editors Eric Grissell, Robert Matthews, Wojciech Pulawski, David Rosen, Mark Shaw All correspondence concerning Society business should be mailed to the appropriate officer at the following addresses: President, Department of Biology, Imperial College at Silwood Park, Ascot, Berks SL5 7PY, England; President-Elect, Department of Entomology, American Museum of Nat- ural History, Central Park West at 79th Street, New York, New York 10024; Secretary, Systematic Entomology Laboratory, USDA, 7v National Museum of Natural History, NHB 168, Washington, D.C. 20560; Treasurer, Biological Resources Division, CLBRR, Agriculture Canada, K.W. Neatby Building, Ottawa, Ontario, Canada KIA 0C6; Editor, P. O. Box 384, North Newton, Kansas 67117. Membership. Members shall be persons who have demonstrated interest in the science of ento- mology. Annual dues for members are $25.00 (U.S. currency) per year, payable to The Interna- tional Society of Hymenopterists. Requests for membership should be sent to the Treasurer (ad- dress above). Journal. The Journal is published once a year by the International Society of Hymenopterists, % Department of Entomology NHB 168, Smithsonian Institution, Washington, D.C. 20560, U.S.A. Members in good standing receive the Journal of Hymenoptera Research. Nonmember subscriptions are $50.00 (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: Once a year (currently). Location of Office of Publication, Business Office of Publisher and Owner: International Society of Hymenopterists, % Department of Entomology, NHB 168, Smithsonian Institution, Washington, D.C. 20560, U.S.A. Editor: Paul M. Marsh, P. O. Box 384, North Newton, Kansas 67117. Managing Editor and Known Bondholders or other Security Holders: none This issue was mailed 25 September 1995 J. HYM. RES. Vol. 4, 1995, pp. 1-24 A Review of the North American Species of Thinodytes Graham and Mauleus Graham (Hymenoptera: Pteromalidae) Steven L. Heydon Bohart Museum, Department of Entomology, University of California, Davis, CA 95616-8584 USA Abstract. — The Halticoptera-group is defined as containing those pteromaline pteromalid genera with a reticulate body, acarinate pronotal collar, weakly developed notauli, weakly delimited frenum, propodeum with the median carina and plicae connected posteriorly by a W-shaped carina, petiole with a basal flange, and the hind margin of the first gastral tergite sinuous laterally and usually emarginate medially. Genera included in this group are Halticoptera Spinola, Halti- copterina Erdos, Andersena Boucek, Thinodytes Graham, Syntomopiis Walker, Mauleus Graham, and Ploskana Boucek. Thinodytes and Mauleus are revised for the Nearctic region and keys to the world's described species are given. New species include T. caroticus n. sp., T. cyzicopsis n. sp., T. petiolatus n. sp., M. cultratus n. sp., M. iligneus n. sp., and M. venetus n. sp. Polycystus nigritus Howard is transferred to Mauleus as M. nigritus n. comb, and Gastrancistrus cephalon Walker is transferred to Thinodytes as T. cephalon n. comb. Bubekia fallax Gahan n. syn. is synonymized with T. cephalon Walker. INTRODUCTION are Notoglyptus Masi, Sphegigaster Spino- la, and Schimitschekia Boucek. The Halhcoptera-group is herein de- ^ ^^^ Thinodytes and Mauleus, the fined as containing those pteromaline North American species of all the genera pteromalid genera with: body reticulately ^^ ^^e Halticoptera-group occurring in the sculptured, pronotal collar acarmate (Figs Nearctic region were recently reviewed 21, 22), notauli weakly developed (Figs ^^ ^^e currently being studied. Andersena 21, 22), frenum weakly delimited (Figs 21, includes only one species (Boucek 1993). 22), propodeum with a sharp median ca- Anderson (1990) reviewed Halticopterina rina and the plicae connected posteriorly f^^ North America and is currently revis- by a W-shaped carina (Figs 17-20), peti- i^g Halticoptera. The world species of No- ole braced basally by an anteriorly di- togh/ptus and the Nearctic species of Si/n- rected lateral and ventral flange (Figs 17- tomopus and Sphegigaster were reviewed 20), and hind margin of first gastral ter- by Heydon (1988),' Heydon (1993), and gite sinuous laterally and usually emar- Heydon and LaBerge (1988), respectively, ginate medially (Figs 21, 22). Genera of This paper reviews the Holarctic species this group include Halticopitera Spinola, of Thinodytes and Mauleus. [Ploskana was Halticopterina Erdos, Andersena Boucek, reviewed by Boucek (1976).] Research by Thinodytes Graham, Syntomopus Walker, the author has revealed the presence of Ploskana Boucek, and Mauleus Graham, numerous species of both these genera in Three other genera that show many of the Central and South America, characteristics of the Halticoptera-group Thinodytes previously contained two and which probably are more or less described species, the Palearctic species closely related to the Halticoptera-^roup T. cyzicus (Walker) 1839, and the Carib- Journal of Hymenoptera Research Figs. 1-10. 1, Thinodytes caroticus n. sp., female hind leg; 2, Thinodytes cephalon (Walker), 2, female hind leg; 3, 4, Thinodytes cyzicopsis n. sp., female head dorsal view, 4, male antenna; 5, 6, Thinodytes petiolatus n. sp., 5, female head dorsal view, 6, male antenna; 7, Thinodytes cyzicus n. sp., female head dorsal view; 8, Mauleus cultratus n. sp., female habitus; 9, Mauleus iligneus n. sp., female antenna; 10, Mauleus venetus n. sp., female antenna. bean species T. clypeatus (Girault) 1918. To these I add the New World species T. cephalon (Walker) 1843, n. comb., and de- scribe four new Nearctic species — T. car- oticus n. sp., T. cyzicopsis n. sp., T. petiol- atus n. sp., and T. santerna n. sp. Mauleus was created for the species M. maderensis Graham, 1981 from Madeira. However, this species is associated with plants na- tive to Mexico and is suspected to be na- tive to the New World (Boucek and Ras- plus 1991). The New World origin of this Volume 4, 1995 Figs. 11-16. 11, Thinodytes petiolatus n. sp., female clypeus; 12, Thinodyies cephahn (Walker), female clypeus; 13, Maiik'iis iligueus n. sp., male clypeus; 14, Synlomopiis amerkanus Ashmead, female head dorsal view; 15, Maidens iligneiis n. sp., male head posterior aspect; 16, Halticoptera sp., male head posterior aspect. species is given more credence by the new Nearctic species — M. cultratus n. sp., presence of a described Neotropical Man- M. iligneus n. sp., and M. venetus n. sp. /('HS species, M. ///^^r/fj/s (Howard), 1897 n. These three new species are described comb, and the author's discovery of three herein. Journal of Hymenoptera Research Figs. 17-22. 17, Tliinodijtes ceplmlon (Walker), female propodeum and petiole; 18, Tlunodi/tes ci/zicopsis n. sp., male propodeum and petiole; 19, Tlii)iodi/tes petiolatiis n. sp., female propodeum and petiole; 20, 21, Maidens digneus n. sp., 20, male propodeum and petiole, 21, female habitus; 22, Syntomopus arpedes Heydon, female habitus. METHODS ^\^ jg used instead of clava. In addition, Terminology generally follows that of the gastral tergites are numbered T1-T7 be- Graham (1969), except that genal concav- ginning with the first tergite after the pet- ity is used instead of genal hollow and iole. The following abbreviations are used: Volume 4, 1995 the median ocellar diameter is MOD, the ocellar-ocular distance is OOL, the poste- rior ocellar distance is POL, the lateral ocellar distance is LOL, the multiporous plate sensilla are MPP sensilla, the lower ocular line is LOcL, and the antennal ha- nicular segments are Fl through F6. The measurements given in the descriptions can be converted to millimeters by multi- plying by 0.02. The acronyms for the mu- seums from which material was borrowed are listed in the acknowledgments section. TJiinodytes Graham Dicydus Thomson, 1876:221, 253. Type species: Miscogaster cyzicus Walker, 1839:200; by monotypy (examined). New name needed because of Dicydus Walker, 1833:371, 455. Thinodytes Graham, 1956:261. Type species: Miscogaster cyzicus Walker, 1839:200; by orig- inal designation (examined). Peck, Boucek, and Hoffer, 1964:41. Graham, 1969:150, 167. Hedqvist, 1975:167. Dzhanokmen, 1978:82. Boucek and Rasplus, 1991:32. Description. — Body color varying from almost wholly black to metallic green; scape metallic or nonmetallic. Head and mesosoma mostly alveolate except sculp- turing sometimes smooth on frenum (T. cyzicopsis, T. petiolatiis, and T. santenm) and median panels of propodeum (T. san- terna); petiole alveolate in species with quadrate to elongate petiole (Figs 18-20), smooth in species with transverse petiole (Fig. 17); gastral tergites nearly smooth. Clypeus variable, with either three asym- metrically arranged denticles (Fig. 11) (T. cyzicopsis, T. cyzicus, T. petiolatiis), three symmetrically arranged denticles (T. san- terna and T. clypeatus) or with single broad asymmetrically placed denticle (Fig. 12) (T. cephalon and T. caroticus). Head with short genal concavity often present; anten- nal torulus above LOcL. Antenna with for- mula 1:1:2:6:3; scape cylindrical, slender; length of flagellum plus pedicel of female less than head width (subequal in length in T. cyzicopsis), equal to or slightly greater than head width in male; funicular seg- ments cylindrical; MPP sensilla usually in single row, their length about equal to length of funicular segment (Figs 4, 6); fe- male club simple apically and with small patch of micropilosity on ventral side of terminal segment, except T. cephalon with terminal spine and large patch of micro- pilosity. Male maxilla with stipites unen- larged; palps slender (Fig. 15). Mesosoma arched dorsally; pronotum with collar very short medially, anterior edge round- ed; mesoscutum with notaulus incomplete or extending to hind margin of mesoscu- tum as impressed line; scutellum as long as wide, lacking anterior median sulcus, with two or three pairs of lateral setae (ex- cept T. santenm with many pairs), frenum not set off by sulcus; dorsellum a short transverse ridge; propodeum with width of median panels about 1.6X their length (Figs 17-19), plicae and median carina dis- tinct and connected posteriorly by W- shaped carina (Figs 17-19), basal fovea sometimes bordered mesally by short straight carina (Fig. 17) (T. clypeatus, T. cy- zicopsis, and T. cephalon), or by long sinu- ous carina (Figs 18, 19) (T. cyzicopsis, T. cyzicus, and T. petiolatiis); spiracles ovate. Fore wing with postmarginal vein about as long as marginal vein but postmarginal vein sometimes distinctly shorter (T. car- oticus and T. cephalon); stigma small, but its height half the distance between stigma and anterior wing margin in T. clypeatus; costal cell with one complete and two or three partial distal setal rows; basal cell bare except sometimes a few setae distally (T. clypK^atus and T. cyzicus); basal vein se- tose except in T. cephalon; speculum de- veloped and open posteriorly. Petiole braced basally by lateral and ventral an- teriorly directed lamellate flange (Figs 17- 19); petiole variable, sometimes trans- verse, unsclerotized ventrally, and with- out lateral setae (Fig. 17) (T. caroticus, T. cephalon, and T. santcrna), or quadrate to elongate and sclerotized ventrally (7. cly- peatus, T. cyzicopsis, T. cyzicus, and T. pe- tiolatiis); lateral setae sometimes present Journal of Hymenoptera Research (Fig. 19) (T. cyzicopsis and T. petiolatus); weak median carina present in T. clypea- tiis. Gaster of female ovate, acuminate api- cally, 1.4-1.8X as long as wide; hypopy- gium extending V2 to % gastral length; hind margin of Tl sinuous laterally and emarginate or straight medially. Discussion. — Giving a few characters to separate Thinodytes from the other genera of the Halticoptera-group, particularly Hal- ticoptera, Maiileus, and Si/ntomopiis, is im- possible because Thinodytes is what is left when the more distinct genera of the Hal- ticoptera-group are characterized. Apomor- phic characters among related genera com- mon to all Thinodytes species such as the reticulate body, weakly developed notauli, poorly delimited frenum, propodeum with sharp median carina and plicae connected posteriorly by W-shaped carina, petiole with a basal bracing consisting of an ante- riorly directed lateral and ventral flange, and the hind margin of the first gastral ter- gite being sinuous laterally and usually emarginate medially are the same charac- ters defining the Hnlticoptera-group itself. So presently, Thinodytes can only be de- fined negatively. There are many South American species, some described and some not, that fit within the present defi- nition of Thinodytes. Once these are inves- tigated, it may be possible to divide Thi- nodytes into monophyletic generic units. Halticoptera is distinguished from Thinod- ytes by a number of apomorphic character states including a bidentate clypeus, the antennal torulus located at or below the LOcL, the scape usually nonmetallic, the male maxilla with lamellately expanded palps and usually with another lobe on the stipites, and a median longitudinal carina on the petiole. Thinodytes has the clypeus with variable numbers and arrangements of clypeal denticles, but no known Thinod- ytes species has a bidentate clypeus. The antennal torulus in Thinodytes is located distinctly above the LOcL (except in T. pe- tiolatus), and the scape usually has metallic coloration. The male maxilla of Thinodytes lacks any expansion of the palps or any lobes on the stipites. Thinodytes clypeatus is the only species of Thinodytes that has a median carina on the petiole. Halticoptera species are commonly bright metallic green; those of Thinodytes are usually dark. Mauleus is distinguished from Thinodytes by the bidentate clypeus (Fig. 13), the dor- sum of the mesosoma as high as the vertex (Figs 8, 21), the median panels of the pro- podeum short (2X as wide as long) (Fig. 20), and the lateral flanges of the petiole enlarged and thickened (Fig. 20). In Thi- nodytes, the clypeus has various numbers and arrangements of clypeal teeth, but is never bidentate, the dorsum of the meso- soma is distinctly lower than the vertex, the median panels of the propodeum are lon- ger (Figs. 17-19), and the basal flanges of the petiole are lamellate (Figs. 17-19). Syntomopus is distinguished from Thi- nodytes by having an elongate pronotum (length about V3 its width) (Fig. 22), three broad symmetrically arranged clypeal denticles (Fig. 14), and usually a flattened mesosoma (Fig. 22). The pronotum of Thi- nodytes is much shorter, those Thinodytes species having symmetrically arranged denticles have them fingerlike rather than broad (Fig. 11), and the mesosoma is nev- er so flattened as it usually is in Syntomo- pus species. Biology. — The known hosts of Thinodytes species are all small Diptera living in plants as leaf or stem miners. KEY TO HOLARCTIC SPECIES OF THINODYTES GRAHAM 1. Petiole transverse, smooth (Fig. 17) 2 - Petiole as long as wide or longer than wide, reticulate (Figs 18-20) 4 2. Scape and legs beyond coxae pale, nonmetallic. Frenum and median panels of propodeum smooth. Plicae rounded and smoothly convergent posteriorly santema n. sp. Volume 4, 1995 7 - Scape mostly or completely metallic. Legs beyond coxae with metallic coloration or dark bands on femora and tibiae. Frenum and median panels of propodeum weakly to strongly alveolate. Plicae with distinct angle between parallel basal part and convergent posterior part 3 3. Hind margin of Tl emarginate medially. Dark bands on tibiae with sharp borders (Fig. 1). Female club with apex simple caroticus n. sp. - Hind margin of Tl entire medially. Dark bands on tibiae with diffuse borders (Fig. 2). Female club with terminal spine at apex cephalon (Walker) 4. Clypeal denticles symmetrically arranged. Petiole with weak median carina. Ovipositor exserted for length equal to that of T7 clypeattis (Girault) - Clypeal denticles asymmetrically arranged, the median denticle displaced to the left (Fig. 11). Petiole without a median carina (Figs 18-19). Ovipositor sheaths hardly exserted .... 5 5. Costal cell with three rows of setae distally. Eye length <4X as long as the temple length (Fig. 5). Body more or less all dark cyzicus (Walker) - Costal cell with two rows of setae distally. Eye length >4X as long as the temple (Figs 3, 7). Body dark with diffuse metallic patches on head and mesosoma 6 6. Male with terminal segment of funicle appearing as wide or wider than long (Fig. 4). Petiole usually less than 1.7X as long as wide (Fig. 18). Female with petiole usually less than 1.5X as long as wide. Both sexes with hind margins of Tl and T2 as long medially as laterally, (eastern United States and Canada) cyzicopsis n. sp. - Males with terminal segment of funicle appearing longer than wide (Fig. 6). Petiole usually more than 1.7x as long as wide. Female with petiole usually more than 1.5 x as long as wide (Fig. 19). Both sexes with hind margins of Tl and T2 usually longer laterally than medially, (far western United States and Canada) petiolatus n. sp. Thinodytes caroticus Heydon, new species (Fig. 1) Holotype, female. — Color: Body dark blue-green except flagellum, fore tarsus, wing veins brown; pretarsi black; knees, apical K; of tibiae, middle and hind tarsi white, border between the light ends and dark median band of tibiae sharp (Fig. 1). Sculpture: Clypeus smooth; head and mesoscutum delicately and regularly al- veolate; scutellum, frenum finely alveo- late; median panels of propodeum ob- scurely alveolate. Structure: Body length 1.2 mm. Head width 1.4X height (23:17), 2.3X length (23: 10); anterior margin of clypeus with single broad tooth; malar distance 5x length of genal concavity; eye height 1.2X length (10:8); 2.0X malar distance (10:5), eye length 4.0 X temple length (8:2); ratio of MOD, OOL, POL, LOL as 2.0:3.0:6.5:3.0; vertex rounding regularly into occiput; to- rulus Vi own diameter above LOcL. An- tenna with length of pedicel plus flagel- lum 0.87X head width (20:23); relative lengths of scape, pedicel, annelli, Fl-6, club as 6.5:2.5:0.5:2.0:2.0:2.0:2.0:2.0:2.0:5.0; widths of Fl, F6, club as 2:3:3; apical club segment simple apically, with micropilos- ity ventrally. Mesosoma arched dorsally, length 1.3X width (25:19); dorsellum short smooth band; propodeum with basal fo- vea part of continuous groove across an- terior margin of median panels, with short weak longitudinal carina crossing groove halfway between plica and median carina, with spiracles on anterior margin of pro- podeum, with nucha lunate strip and car- inate anteriorly. Fore wing with ratio of lengths of submarginal, marginal, post- marginal, stigmal veins as 19.0:11.5:9.0:5.0; stigma small; basal cell bare; basal vein with row of 4 setae. Petiole conical, trans- verse, smooth, with median carina. Gaster fusiform, length 1.8X width (30.0:16.5); hind margin of Tl emarginate medially; ovipositor sheaths hardly exserted; hypo- pygium extending Vi gastral length. Journal of Hymenoptera Research Allotype, male. — Color pattern similar to holotype except basic body color dark blue; fore tarsi pale brown; pale portions of legs pale yellow-brown instead of white. Body length 1.2 mm. Antenna with length of pedicel plus flagellum 0.98 X head width (21.0:21.5); relative lengths of scape, pedicel, annelli, Fl-6, and club as 6.0:2.5:0.5:2.0:2.0:2.0:2.0:2.0:2.0:6.0; widths of Fl, F6, club as 2.5:3.0:3.0; funicular setae sparse, reclinate. Gaster ovate, length 1.5X width (23:15). Variation. — The body length of females examined varied between 1.2 and 1.8 mm and males between 1.0 and 1.4 mm. The color of the dorsum of the mesosoma var- ies from dark blue-green to dark blue. The dark bands on the legs are always distinct, but the intensity of their metallic colora- tion is variable. Discussion. — Thinodi/tes caroticus most closely resembles T. cephalon because both species are dark in color; have a single, broad, asymmetrically placed clypeal den- ticle; and have smooth, transverse peti- oles. Thinodytes caroticus differs from T. ce- phalon in the following: 1. The hind mar- gin of Tl is emarginate in T. caroticus, but straight in T. cephalon. 2. The basal vein of the fore wing is setose in T. caroticus, but bare in T. cephalon. 3. The median panels of the propodeum are weakly sculptured in T. caroticus, but distinctly alveolate in T. cephalon. 4. The hypopygium extends about Vi the length of the gaster in T. car- oticus, but about % the gastral length in T. cephalon. 5. The dark bands on the tibiae are distinct with sharp borders in T. caro- ticus, but are less distinct and have diffuse borders in T. cephalon. These distinctive tibial color bands are unique to T. caroticus and will identify the species at a glance (Fig. 1). Etymology. — The species name comes from the Greek karotikos, meaning stupe- fying or soporific, and refers to the general nondescript appearance of this species. Type Material. — The holotype, allotype (both UCDC) and one male paratype were all reared by the author from the leaf mines of Calycomyza promissa (Frick) (Dip- tera: Agromyzidae) collected 30 June 1985 on the South Farms of the University of Illinois, near Champaign, Illinois. Fifty- nine additional paratypes seen were col- lected as follows (CNCI, INHS, SEMC, UCDC, USNM): Bermuda. DEVONSHIRE PARISH: Devonshire Marsh, 27.VI.1988, 1 male. FAGOT PARISH: Berry Hill Road, 29.VI.1988, 2 females, 3 males; Botanical Garden, 27. VI. 1988, 1 female; Camdon Marsh, 29.VI.1988, 1 female. SMITHS PARISH: Spittal Pond, 27.VI.1988, 1 fe- male. Canada. ONTARIO: Chatham, 1952 (mass reared from Hessian fly straw), 1 male. United States. CALIFORNIA: Haw- thorne, IX. 1940 (ex Aster blotch), 3 fe- males, 1 male; Jepson Prairie Preserve (13 km s. Dixon), 20.V.1983, 1 female; Lake Hennessy (11 km ese. St. Helena), 28.x. 1990, 1 female; Los Angeles County (bred from dipterous leaf miner), 1 female; Sacramento, 16. IX. 1924 [ex Agromyza pus- ilia {prob.= Liriomyza pusilla)], 1 female; So- quel, 26.VIII.1948 (ex Agromyza sp.), 2 fe- males, 1 male; 11 km e. St. Helena (Lake Hennessey), 28. X. 1990 (on Baccharis), 1 fe- male, 7.IX.1991 (on Heraculeum), 1 male; 6 km e. Suisun City, 11.VIII.1990, 2 males; 1000 Palms, 29.III.1977, 1 female; West- wood Hills (Los Angeles County), 5.XI.1940 (ex serpentine leaf miner in Zin- nia), 4 females, 1 male. FLORIDA: Braden- ton, 27.XI.1946 [ex Liriomyza pissilla {proh.= Liriomyza pusilla)], 1 male; Jackson- ville, 2 females. GEORGIA: Savannah, 5. VI. 1943 (parasite of goldenrod leafmin- er), 3 females. ILLINOIS: South Farms of the University of Illinois, 19.V.1985, 1 fe- male; White Heath, 24.IX.1939, 1 female. INDIANA: 4 miles s. New Harmony (Har- mony State Park), 28. VI. 1983, 2 females; 2 miles s. New Lisbon, 14. VII. 1981, 1 male. IOWA: Sioux City (reared from leaf miner on sunflower leaf), 1 female, 5. VIII. 1921 (reared from mine of leaf on sunflower), 1 female. KANSAS: Lawrence, 14.V.1955, 1 female. MINNESOTA: Albert Lea, Volume 4, 1995 26.VII.1960, 1 female. NEW MEXICO: Ci- marron, 1909, 1 female; Springer, 1909, 1 female, 3 males. TEXAS: Clarendon, 19. IX. 1905 (on Grindelia squarrosa (Pursk) Dunal (Compositae), 1 female; Ennis, 27.IX.1905 [ex dipterous leaf miner of Ma- chaeranthera annua (Rydb.) Shinner (Com- positae)], 2 females; Roma, 26. III. 1948 (ex pupa of dipterous leaf miner), 3 females, 3 males. Biology. — This species has been reared from leaf-mining Agromyzidae, mostly on Compositae. Known hosts include Ca- lycomyza promissa (Prick) and Liriomyza pusilla (Meigen). Because L. pusilla is a Pa- learctic agromyzid species (Spencer 1976) and is not known from the Nearctic region (Spencer and Steyskal 1986), it is likely that this host record is in error. There is one record from Chatham, Ontario from Hessian fly straw. Thinodytes caroticus was also reared from leaf miners on Machaer- anthera annua, Zinnia, goldenrod, sunflow- er, and "Aster". It has been reared from both linear-mining and blotch-making leaf miners. Other plant associations of a more uncertain nature include the com- posites Baccharis sp., Heracleum sp., and Grindelia squarrosa. Thinodytes clypeatus (Girault) Polycystus clypeatus Girault, 1918:128. Holotype, female (USNM); Hym. Type No. 20682; (ex- amined). Thinodytes clypeatus (Girault): Heydon, 1989: 193. Redescription. — Holotype, female. Color: Mesosoma, petiole black, with blue tints on propodeum and pleural regions; gaster dark brown; legs with basal % of femora brown, remainder of legs white. Sculpture: Dorsum of mesosoma regu- larly and delicately reticulate; median panels of propodeum weakly alveolate; petiole alveolate. Structure: Head with anterior margin of clypeus with three symmetrically ar- ranged fingerlike denticles, torulus above LOcL. Antenna with lengths and widths of segments as follows (in iJim): scape 19.5X3.6: pedicel 6.0x4.5: annelli 2.6x3.6: Fl 5X6: F2 4.6X6.0: F3 5.2x6.2: F4 4.6x6.5: P5 4.6X6.8: F6 missing: club 13.4X6.8; club simple apically. Mesosoma with dorsum rather flat, length 1.4X width (33:24); pro- notum with collar width 11 X length (22: 2), sides converging posteriorly; notauli shallow posteriorly; scutellum with two pairs of lateral setae, frenum almost indis- tinguishable from remainder of scutellum; propodeum with length of median panels 0.45 X width, with row of foveae separated by carinae along anterior margin, spiracle on anterior margin of propodeum, nucha a raised smooth crescent. Fore wing with relative lengths of submarginal, marginal, postmarginal, stigmal veins as 21:12:11:6; stigma large, height Vi distance between stigma and anterior margin of wing; basal cell with one seta; basal vein with three setae. Petiole length 1.2X width (6:5); with weak median carina; lacking lateral setae. Gaster fusiform, length 1.7X width (31.0: 18.5); Tl emarginate medially; hypopy- gium extending % length of gaster; ovi- positor sheaths exserted for length equal to that of T7. Discussion. — The holotype has the head removed and crushed on a slide. Thinod- ytes clypeatus differs from all other de- scribed Thinodytes species because it has an enlarged stigma, exserted ovipositor sheaths, and a median carina on the peti- ole. Thinodytes clypeatus has three sym- metrically arranged denticles like Synto- mopus species, but T. clypeatus differs from Syntomopus species in several ways. The denticles of T. clypeatus are fingerlike lobes like those of the other Thinodytes species with three denticles, whereas the denticles of Syntomopus species are rather broadly triangular (Fig. 14). The pronotal collar of Thinodytes clypeatus is short, eleven times as wide as long; the pronotal collar in Syn- tomopus is only about three times as wide as long (Fig. 22). Most species of Halticop- tera have a median carina on the petiole, but there are no other characters to indi- 10 Journal of Hymenoptera Research cate a particularly close relationship be- tween T. clypeatus and Halticoptera. Distribution and Biology. — The type spec- imen was reared from a leaf miner on corn on 2 May 1916, on St. Vincent, British West Indies by F. Watts. De Santis (1979) reports this species from Barbados also, and gives the name of its host as Agromyza parvicornis Loew (Diptera: Agromyzidae). Thinodytes cephnlon (Walker), new combination Figs 1, 17 ?Pteromnlus Rhseo Walker, Walker 1839b:88-89. Lectotype, female (BMNH); Hym. Type No. 5.772 (examined). Gastmncistrus cephalon Walker 1843:30. Lecto- type, male (BMNH); Hym. Type No. 5.661 (examined). Bubekia fallax Gahan, 1933:114-116. Holotype, female (USNM); Type No. 44841 (examined). Allen and Painter, 1937:225. Nikol'skaya, 1937:25. Peck, 1951:538. Thompson, 1958:587. Peck, 1963:610-611. Morrill and Kieckhefer, 1971:1130. Allen and Pienkowski, 1973:616, 617. Burks, 1979:789. Hendrickson, 1979:300, 302. n. syn. Discussion. — Thinodytes cephalon is easily distinguished by its single broad clypeal denticle, apical spine on the female club, notauli traceable to the hind margin of the mesoscutum as impressed lines, bare dor- sal vein, smooth and short petiole (Fig. 17), and first gastral tergite having the hind margin entire. A possible senior synonym of T. cephal- on is Pteromalus rhaeo described from Chil- oe Island, Chile. The lectotype female (designated herein) of P. rhaeo is unfortu- nately missing the head. It appears similar to specimens of T. cephalon examined, of which many specimens in my collection and the CNC collection were from Chiloe Island, but it is significantly larger than any other specimen examined. The medi- an panels of the propodeum in P. rhaeo are entirely rugose and the plicae are poste- riorly divergent. In other T. cephalon ex- amined, the median panels have few ru- gae which are located mostly posteriorly, and the plicae are parallel. The hind tibia in P. rhaeo is almost uniform in color; in T. cephalon, the hind tibia are distinctly dark over the middle half or more. Pter- omalus rhaeo may be within the range of variation of what is recognized as T. ce- phalon, but I am reluctant to formally syn- onymize the name until more is known about the range of variation of T. cephalon in South America. Distribution. — Thinodytes cephalon is one of the most commonly collected New World pteromalids. Its distribution ex- tends throughout most of North and South America (from Chile to Canada and also Bermuda). Biology. — Thinodytes cephalon is a para- sitoid of dipterous larvae that mine leaves or grass stems. Thinodytes cephalon was originally described from puparia of the Hessian fly, Phytophaga destructor (Say) (Diptera: Cecidomyiidae)(Gahan 1933). It has also been reared from the wheat stem maggot, Meromyza americana Fitch (Dip- tera: Chloropidae), in a number of studies (Gahan 1933; Allen and Painter 1937; and Morrill and Kieckhefer 1971). Allan and Painter report that T. cephalon probably oviposits into the larval stage of this host. One additional host added herein is Lir- iomyza trifoliearum Spencer (Diptera: Agro- myzidae) on alfalfa. Thinodytes cyzicopsis Heydon, new species Figs. 3, 4, 18 Holotype, female. — Color: Head, mesoso- ma black with metallic green patches as follows: entire dorsellum, pairs of spots alongside median ocellus, lateral hind cor- ners of pronotum and upper epimeron; metallic blue patches as follows: on inner orbits, anterior part of lateral lobe of scu- tum and propodeum. Antenna with scape, pedicel dark blue; flagellum black. Petiole black. Caster brownish black with metallic blue reflections dorsally, green ventrally. Legs with coxae black with weak blue re- Volume 4, 1995 11 flections; trochanters, femora brownish black with weak green reflections, knees yellow, tibiae with diffuse-edged dark brown band extending % their length; fore tarsi brown; middle and hind tarsi with pretarsi black, remainder yellow. Sculpture: Head, mesosoma, regularly and delicately alveolate; frenum, propo- deum shallowly and indistinctly alveolate; petiole finely alveolate. Structure: Body length 1.6 mm. Head width 1.2X height (27:22), 2.2 X length (27: 12); anterior margin of clypeus with three minute asymmetrically arranged apical denticles; genal concavity absent; eye height 1.4X length (13:9), 1.9X malar dis- tance (13:7), length 4.5 X temple length (9: 2) (Fig. 3); ratio of MOD, OOL, POL, LOL as 2.5:4.0:7.0:3.0; vertex rounding smooth- ly into occiput; torulus V2 own diameter above LOcL. Antenna with length of ped- icel plus flagellum l.Ox head width (27: 27); relative lengths of scape, pedicel, an- nelli, Fl-6, club as 11.0:3.0:1.0:2.5:3.0:2.75: 2.75:2.5:2.5:7.0; widths of Fl, F6, club as 3.0:3.5:3.5, Fl^ appearing quadrate, F5-6 slightly transverse; club simple apically, with minute patch of micropilosity on ter- minal segment. Mesosoma arched dorsal- ly, length 1.6X width (36:22); pronotum with humeral angles squared; mesoscu- tum with notauli shallow; dorsellum bandlike, short, weakly alveolate; propo- deum with basal fovea bordered mesally by sinuous carina; nucha a raised band, weakly carinate anteriorly; spiracle < 0.5 X own diameter from anterior margin of propodeum. Fore wing with ratio of lengths of submarginal, marginal, post- marginal, stigmal veins as 25:13:14:8; stig- ma small, maximum width only slightly more than stigmal vein width; basal cell bare; basal vein setose. Petiole length 1.2X width (6:5); without median carina; with one pair of lateral setae. Gaster length 1.4X width (31:22); hypopygium extend- ing to about % length of gaster; ovipositor sheaths hardly exserted. Allotype, male. — Color similar to holo- type except frons blue, mesoscutum with a pair of large diffuse green spots; axilla green. Body length 1.3 mm. Antenna (Fig. 4) with length of pedicel plus flagellum l.Ox head width (24:23); relative lengths of scape, pedicel, annelli, Fl-6, club as 7.0: 3.0:1.0:2.0:2.5:2.5:2.5:2.5:2:6.5; widths of Fl, F6, club as 2.0:2.5:2.5; setae fine, reclinate. Petiole length 1.2X width (5:4)(Fig. 18). Gaster ovate, length 1.2X width (20.0: 16.5). Variation. — The body color varies from mostly dark, like the holotype, to the pat- tern shown by the allotype male. The in- tensity, size, number, and hue of the me- tallic spots on the body are highly vari- able. The color of the metallic patches var- ies from coppery green to green to blue- green. The body length ranges between 1.4 and 1.8 mm in females and 1.2 and 1.7 mm in males. The length of the marginal vein varies from 0.9 to 1.2 times the length of the postmarginal vein but averages about equal its length [x = 1.02± (S.E. = )0.02 (n = 12)]. The petiole averages 1.39±0.045 (n=7, range 1.2-1.5) times as long as wide in the females and 1.57 ±0.061 (n = 7, range 1.2-1.9) times in males (Fig. 18). The pair of lateral setae on the petiole are sometimes difficult to see because they tend to lie near the anterior flanges of the petiole. Discussion. -Thinodytes cyzicopsis resem- bles the Palearctic species T. cyzicus (Thomson), but differs from that species in the following: 1. Thinodytes cyzicopsis has a series of metallic patches on the body, which are lacking in T. cyzicus. 2. The ratio of the eye length to temple length aver- ages 4.25±0.12 (n=6, range 3.8^.5) in fe- male T. cyzicofisis (Fig. 3) but measured only 2.3 and 3.6 in two female T. cyzicus specimens examined (Fig. 5). 3. The costal cell of the fore wing of T. cyzicopsis has two rows of setae distally, whereas there are three rows distally in the costal cell of T. cyzicus. 4. The petiole has a pair of lat- eral setae in T. cyzicopsis, which are lack- ing from T. cyzicus. 12 Journal of Hymenoptera Research Thinodytes cyzicopsis and T. petiolatus are very similar species and specimens cannot always be confidently separated. These two species are distinct from other Thinod- ytes in their common possession of a sim- ilar pattern of metallic patches on the head and mesosoma. Individuals of T. cyzicopsis differ from those of T. petiolatus in the fol- lowing: 1. The ratio of eye height to malar distance in T. cyzicopsis averages 1.74±0.035 (n=8, range 1.5-1.8) in females and 2.04 ±0.032 (n=9, range 1.9 to 2.0) in males; in T. petiolatus the ratio is 1.48 ±0.23 (n=10, range 1.4 to 1.6) in females and 1.77±0.038 (n=10, range 1.6 to 2.0) in males. 2. All funicular segments of the male antenna are transverse to quadrate in T. cyzicopsis; all funicular segments are usually longer than wide in male T. pe- tiolatus. 3. The ratio of petiole length to width in T. cyzicopsis averages 1.39 ±0.045 (n=7, range 1.2 to 1.5) in females and 1.57±0.061 (n = 7, range 1.2 to 1.9) in males; in T. petiolatus it averages 1.93±0.062 (n=10, range 1.5 to 2.0) in fe- males (Fig. 19) and 1.95±0.071 (n=10, range 1.6 to 2.4) in males. 4. The first gas- tral tergite in T. cyzicopsis is about as long medially as laterally, whereas Tl is often shorter medially than laterally in T. petiol- atus. 5. The hind margin of T2 is straight to just noticeably concave in T. cyzicopsis, whereas it is usually distinctly concave in T. petiolatus. 6. The terminal segment of the male funicle appears quadrate to slightly transverse in T. cyzicopsis, whereas it appears elongate in T. petiolatus. In ad- dition to these structural characters, T. cy- zicopsis is found east of the Rocky Moun- tains, and T. petiolatus occurs from the Rockies west. Type Material. — The holotype (CNCI) is from Kouchibouquac National Park, New Brunswick, and was collected 9 August 1977 by S. J. Miller. The allotype (USNM) is from Ithaca, New York, and was col- lected on vernal alfalfa on 22 June 1968 by A. G. Wheeler. Sixteen paratypes were col- lected as follows (CNCI, UCDC, USNM): Canada. ALBERTA: Elkwater Lake, 21.VII.1956, 1 female; Lethbridge, 1 fe- male. NEW BRUNSWICK: Kouchibou- quac National Park, 20.IX.1977, 1 female. NOVA SCOTIA: Alton, IX. 1964 [ex Phy- tobia {Calycomyza) solidaginis on Solidago], 1 male; Crosby, 31 .VII. 1952 (on apple), 1 male. ONTARIO: Ottawa, 22.VI.1972 (swept from Salix hlanda), 1 male, 29.VI.1972 (swept from Salix blanda), 3 males, 24. VII. 1972 (swept from Salix hlan- da), 1 male. QUEBEC: Lac Brule, 21.VII.1947, 1 female, 25.VII.1947 (swept from Rosa rugosa), 1 female, 9.VIII.1945, 1 female. United States. ILLINOIS: 2 miles e. Shumway, 7.VII.1980, 1 female; South Farms of the University of Illinois, nr. Champaign, 23.VI.1981, 1 male. MASSA- CHUSETTS: Hopkinton, 9.VIII.1951 (ex Ilex leaf miner), 1 female. NEW MEXICO: Mesilla, 4. V. 1909, 1 female. WEST VIR- GINIA: Winchester, 16.VI.1964, 1 male, 17.VI.1964, 1 male. Etymology. — The species name is de- rived from the species name of Thinodytes cyzicus and the Greek suffix -opsis, mean- ing like or similar in appearance, and re- fers to the morphological similarity be- tween T. cyzicopsis and T. cyzicus. Biology. — Known hosts of T. cyzicopsis include Phytobia {Calycomyza) solidaginis (Agromyzidae)[on Solidago sp. (Composi- tae)] and an Ilex leafminer. The species has also been taken on Salix blanda (Salica- ceae), Rosa rugosa (Rosaceae), and alfalfa. Salix and Rosa are probably primarily nec- tar sources because neither has many leaf- mining agromyzids, but these plants are important sources of nectar and /or hon- eydew for many parasitic Hymenoptera. Thinodytes cyzicus (Walker) Miscogaster cyzicus Walker, 1839a:200. Lecto- type, female (BMNH); Hym. Type No. 5.2570 (examined). Syntomopus cyzicus (Walker): Walker, 1846:28. Schmiedeknecht, 1909:376. Dicyclus circulus Thomson, 1876:253. Lectotype, female (LUND), not seen. Volume 4, 1995 13 Thinodytes cyzicus (Walker): Graham, 1956:261. Graham, 1969:167. Askew, 1970:380. Hedqv- ist, 1975:180. Boucek, 1977:56. Kamijo, 1978: 457. Takada and Kamijo, 1979:21, 22, 23, 25. Hedqvist, 1983:167. Boucek and Rasplus, 1991:32. Discussion. — The synonymy of Dicyclus circiilus with Thinodytes cyzicus is accepted on the authority of Graham (1969). Thi- nodytes cyzicus resembles T. cyzicopsis and T. petiolatus, because all three species have three small, sharp, asymmetrically ar- ranged clypeal denticles (Fig. 11) and a long, reticulate petiole. Besides the geo- graphic separation of their ranges, T. cy- zicus can be distinguished from the two Nearctic species by its body color. Thino- dytes cyzicus is uniformly dark and lacks the distinctive diffuse metallic patches on the head and mesosoma present in T. cy- zicopsis and T. petiolatus. Thinodytes cyzicus also has three rows of setae distally in the costal cell, whereas T. cyzicopsis and T. pe- tiolatus have only two rows. The eye length in T. cyzicus is 4X or less the length of the temple (Fig. 5), whereas T. cyzicopsis and T. petiolatus have the eye length more than 4X the length of the temple (Figs 3, 7). Distribution. — Thinodytes cyzicus occurs throughout the Palearctic region, from Britain to Japan (Graham 1969; Boucek 1970; Kamijo 1978). Biology. — The hosts of T. cyzicus are all Agromyzidae. Askew (1970) recorded this species from an agromyzid, probably Phy- tomyza atricornis Meigen, on Senecio jaco- baea Linnaeus. Kamijo (1978) recorded T. cyzicus from Chromatomyia horticola (Gou- reau)(Diptera: Agromyzidae) on pea and from an agromyzid on Lathy rus maritimus. Takada and Kamijo (1979) recorded T. cy- zicus as emerging from the puparium of Chromatomyia horticola and speculated that T. cyzicus may parasitize the larval stage of its host. Thinodytes petiolatus Heydon, new species Figs 6, 7, 11, 19 Holotype, female. — Color: Body black but frenum, metanotum coppery; frons, ver- tex, gena, lateral region of pronotum, lat- eral lobe of mesoscutum, middle lobe of mesoscutum with pair of large diffuse me- tallic patches, anterior lateral corner of ax- illa dark metallic green; gaster with green- ish reflections. Antenna with scape dark green; remainder brown, pedicel with weak green reflections. Legs with coxae black with greenish reflections; femora brown with greenish reflections; remain- der of legs mostly yellow-brown, tibiae with broad diffuse brown band mesally, apical two tarsi brown. Wing veins pale reddish brown. Sculpture: Clypeus weakly alveolate; face alveolate, cells elongate in radiating fashion from clypeus; remainder of head alveolate; mesoscutum, scutellum deli- cately and regularly alveolate except fre- num smooth; median panels of propo- deum alveolate; petiole finely alveolate. Structure: Body length 1.8 mm. Head width 1.4X height (28:21), 2.6 X length (28.5:11.0); clypeus with three small asym- metrically arranged clypeal denticles (Fig. 11); weak genal concavity extending Vs malar distance; eye height 1.3X length (12.0:9.5), 1.5X malar distance (12:8), length 4.8X temple length (9.5:2.0)(Fig. 7) ratio of MOD, OOL, POL, LOL as 2:4:7:3 vertex rounding regularly into occiput antennal torulus just above LOcL. Anten- na with length of pedicel plus flagellum 0.91 X head width (26:28.5); ratio of lengths of scape, pedicel, annelli, Fl-6, club as 11.5:3.5:1.0:2.5:2.5:2.5:2.5:2.5:2.5:6.0; widths of Fl, F6, club as 2:3:3; club simple apically, with small patch of micropilosity ventrally on apical segment. Mesosoma arched dorsally, length 1.7X width (37:22); notauli shallow posteriorly; propodeum (Fig. 19) with basal fovea margined mesal- ly by long sinuous carina; nucha a weakly 14 Journal of Hymenoptera Research sculptured band, carinate anteriorly; spi- arate these two species are given in the racle <0.5X own diameter from anterior discussion section for T. cyzicopsis. The margin of propodeum. Fore wing with rel- hind margin of T2 in many specimens of ative lengths of submarginal, marginal, T. petiolatus is distinctly concave, but the postmarginal, stigmal veins as 27.0:14.5: visibility of this character depends on how 15.0:9.0; stigma small; basal vein with row the specimen has dried. This character is of setae along length. Petiole (Fig. 19) more distinct in the males than in the fe- length 2.0 X width (8:4); without median males. carina; with two pairs of lateral setae. Gas- Type Material: — The holotype (USNM) ter ovate-acuminate, length 1.4X width was collected at Roseworth, Idaho, on 19 (33:24); hind margin of Tl strongly sinu- June 1931, on Salsola pestifer by D. E. Fox ous laterally, emarginate mesally; ovipos- and the allotype (USNM) was collected at itor sheaths hardly exserted; hypopygium Twin Falls, Idaho, on 5 August 1920 by R. extending % gastral length. H. Smith. Sixty paratypes were collected Allotype, male.— Body color similar to as follows (CNCI, INHS, UCDC, USNM): holotype except metallic spots more dif- Canada. ALBERTA: Banff (Sunshine fuse and extensive; frenum, metanotum Lodge, 7500'), 24. VII. 1962, 1 female; Elk- blue-green; dark bands on tibiae very water, 9.VI.1956, 1 male; Elkwater Lake, weak. Body length 1.3 mm. Antenna (Fig. 21.VII.1956, 1 male; Lethbridge, 7.VI.1956 6) with length of pedicel plus flagellum (swept from barley), 1 female, 5.VIII.1956 1.1 X head width (28:25); ratio of lengths (swept from barley), 1 male; nr. Leth- of scape, pedicel, annelli, Fl-6, club as 9.0: bridge, 1924.1925, 1 female. BRITISH CO- 3.5:1.0:2.5:3.0:3.0:3.0:3.0:3.0:8.0; widths of LUMBIA: Bowser, 28.V.1955, 1 female; Fl, F6, club as 2:2:3; setae of flagellomeres McQueen Lake, 10 miles n. Kamloops, fine, reclinate. Petiole length 2.0 X width 18.VI.1973, 1 male. United States. ARI- (8:4). Gaster length 1.2X width (20:16); ZONA: Mesa, 3.VI.1958 (swept from alfal- truncate apically. fa), 1 male. CALIFORNIA: Albany, Variation. — The body color varies from 29. IV. 1958, 1 male; Alpine Lake, VI. 1971, like that of the holotype to very dark, with 1 female; Apple Valley, 8.V.1955, 2 males; the green areas on the head and the me- 15 miles w. Baker, 6.V.1977, 2 females, 1 sosoma reduced and obscure, the frenum male; Boca, 22. VII. 1970, 1 female; Bolinas, and metanotum green, and the tibiae dark 5.X.1975, 2 males; Camino, 21.VII.1948 (ex metallic green. The body length of females Phytomyza aqiiilegiana), 10 females; Cerro examined varied between 1.3 and 1.8 mm Noroeste (sw. corner of Kern Co.), and males varied between 1.0 and 1.6 mm. 15.VII.1965, 1 female, Cuyler Harbor (San The sinuous carina on the propodeum Miguel Island), 11. VII. 1970, 1 male; Dar- sometimes extends only Va the length of win Falls (nr. Panamint Springs), the propodeum. The row of setae on the 29. III. 1984 (on Encelia), 1 male; Emeryville, basal vein sometimes curls proximally, 28. V. 1958, 1 male; Eureka Dunes (Inyo running a short distance along the cubital Co.), 15.V.1979, 1 male; Lake Tenaya, vein. 23.VII.1949 (host Recurraria milleri), 1 fe- Disciission. — Thinodytes petiolatus and T. male; Lake Tahoe, 29. VI. 1927, 1 female; cyzicopsis are very similar species and are Lily Pond (alpine lake), VI. 1971, 1 female; distinct from other Thinodytes species in Los Angeles {Agromyza playptera Thom.), 1 the similar pattern of metallic patches on female; Los Angeles Co., 1 male; McClure an otherwise dark body. Individuals of T. Beach (Marin Co.), 18.VII.1970, 1 male; Mt. petiolatus and T. cyzicopsis, especially fe- Ingalls (Placer Co.), 11. VII. 1964, 1 male; males, cannot always be distinguished near Nicasio, 15.11.1991 (sweeping Salix), 2 with absolute certainty. Characters to sep- males; 22 miles w. Panamint Springs, Volume 4, 1995 15 7.V.1961, 2 males; Placer Co., Vlll, 1 male; (5), 1 male; Moscow, 6.VIII.1926, 1 male; Sagehen Creek (near Hobart Mills), 24. Murtaugh, 29.V.1930 (3 & 5), 1 male; Oak- VI.1970, 1 female, 21-25.VI.1982 (sweep- ley, 7.VIII.1929 (1, 3 & 5), 1 female; Rupert, ing Primus), 1 male, 12.VII.1972, 1 male, 29.V.1930, 1 female; Tuttle, 22.V.1931 (4), 11-15.VII.1982, 1 female, 23.VII.1968, 1 fe- 1 female; Twin Falls, 7.VI.1930, 1 male; male, 1.VIII.1970, 1 female; San Bernardi- Wendell, 22.V.1931 (5), 3 males. MON- no Co., San Gorgonio Wilderness, TANA: Big Butte, 8.VIII.195?, 1 male. OR- 19. VII. 1982, 1 male; Santa Cruz, EGON: Corvallis, 15.VI.1981, 1 male, 22.V11I.1948 (ex PJnjtonn/za sp. B), 1 female; 26.VI.1985, 1 female, 3 males; Near Cor- Santa Rosa, 25.V.1990, 2 males; Sheppard vallis (St. Mary's Peak), 15.VIII.1984 (road- Pass Trail (Inyo Co.), 16. VII. 1985 (on side vegetation), 1 female; Eugene, 6- Sphenosciadium capitellatum),! male; Shive- 12.VI.1984 (Malaise trap), 1 female, 11 ly, 19. VI. 1959, 1 female; Sonora Pass males; McKinzie Pass (Mt. Washington (Mono Co., 9624'), 20.V1II.1960, 1 female; Wilderness), 17.VIII.1984, 2 females, Soquel, 26.VIII.1960 (ex Agromijza sp.), 2 20.VIII.1984, 5 females, 2 males; 1 mile w. females; Titus Canyon (Death Valley Na- McKinzie Pass, 1 female, 1 male. UTAH: tional Monument), 24.III.1984 (on Cownia Logan Canyon, 21. VII. 1976, 1 female; mexicana), 1 male; Tomales Bay State Park Monte Cristo, 6.VII.1976, 1 male; My ton, 3 (Marin Co.), 14.IV.1961, 1 male; Trinidad, males; Wellsville, 13.V.1964, 1 male; 1.5 24.IX.1977, 1 female; Truckee, 20.VII.1970, miles w. Wild Horse Butte (Wild Horse 1 female; White Water, Snow Creek Creek), 31.VII.1982, 1 female. WASHING- (1500'), 29.III.1955, 1 male; 6 miles w. Bas- TON: San Juan Island (Barney's Place), setts (Yuba Pass), 9.VII.1970, 2 females, 2 23.VII.1944, 1 male. Vancouver, 15.VI. males. COLORADO: Chambers Lake (Lar- 1911, 1 female, 1 male, 16.VI.1911, 1 male, imer Co.), 16.VIII.1966, 1 male; 16 km n. 20.VI.1911, 1 female. WYOMING: Kem- Colorado Springs, 25. VII. 1991, 1 male; Ft. merer, 12. VII. 1985 {Eriogomim), 2 females, Collins, 20. VIII. 1895 (on boxelder foliage), 13.VII.1985 [Artemesia spinosa {?=spines- 1 female, 1 male; Glacier Basin, Rocky cens D.C. Eaton], 1 male; Snowy Range, Mountain National Park, 24.VII.1977, 1 23.VIII.1951, 2 females, 1 male, male; Echo Lake (Mt. Evans, 10,500'), Etymologxj. — The species name refers to 4. VIII. 1961, 1 female. IDAHO: Boise, the long petiole, characteristic of this spe- 28.V.1984 (Malaise trap), 2 males; Buhl, cies. 27.V.1929 (1 & 50, 2 males; Burley, B/o/ogi/.— The host(s) of T. petiolatus is 14.VI.1930 (3), 1 female, 9. VII. 1931, 1 fe- unknown but it has been taken in associ- male; Eden, 11. VIII. 1930 (2), 1 female; ation with a number of plants, such as Ar- Hobbs Butte, 22.V.1931, 1 male, 6. VI. 1931, temesia sp., Descurainia sophia, Eriogomim, 1 male; Hollister, 16.V.1931 (3), 1 male, Salsola pestifer, and Sisymbrium altissimum 20.V.1931, 2 males, 2.VI.1931, 1 female, 1 in Idaho and Encelia in California. It has male, 5. VI. 1931 (3), 1 female, 7. VI. 1931 (3 also been collected in association with & 5), 1 female, 13.VI.1931, 1 female; Hubbs crop plants, such as beets in Idaho and Butte, 22.V.1931, 1 male, 6.VI.1931, 1 male; barley in Alberta. Jerome, 11.VII1.1930 (2), 1 male; Kimberly, 27.VI.1931, 2 females; Milner, 29.V.1930 Thinodytes santerna Heydon, new species Holotype, female. — Color: Body black with blue-green reflections on vertex, side ' The host plants for the specimens from Idaho are j^^^^g ^f mesoscutum, SCUtellum, gaster; numbered as follows: 1= A.? rosae. 2= Beta viiharis n ■ <- i i i L. 3= Soplua sophia {=Descuraima sophia (L.) V.B. coppery reflections on frenum and dorsel- Webb). 4= Salsola pestifer A. Nelson. 5= N.(ortn ) al- lum; yellow-green reflechons on middle tissiiniim or S. altissiina { = Sisymbri\nu altissimum L.). lobe of SCUtellum and propodeum. Anten- 16 Journal of Hymenoptera Research na with scape yellow, flagellum brown dorsally, brownish white ventrally. Legs yellow beyond coxae, tarsi white, pretarsi brown. Fore wing veins pale brown. Sculpture: Clypeus alveolate; frenum, dorsellum, propodeum smooth; gaster with T5-7 coriaceous, remainder smooth. Structure: Body length 1.6 mm. Head width 1.4X height (28:20), 2.3 X length (28: 12); clypeus with three small symmetri- cally arranged clypeal denticles; weak genal concavity extending Va malar dis- tance; eye height 1.3 X length (11:10), 2.2 X malar distance (11:6), length 5.0 X temple length (10:2); ratio of MOD, OOL, POL, LOL as 2:3:6:3; vertex rounding regularly into occiput; antennal torulus just above LOcL. Antenna with length of pedicel plus flagellum 0.86 X head width (24:28); ratio of lengths of scape, pedicel, annelli, Fl-6, club as 10:3:1:2:2:2:2:2:2:6; widths of Fl, F6, club as 2:2:2; club simple apically, with small patch of micropilosity ventrally on apical segment. Mesosoma arched dorsal- ly, length 1.5 X width (32:22); notauli shal- low posteriorly; propodeum with basal fo- vea obscure, nucha lunate strip and cari- nate anteriorly, spiracles on anterior mar- gin of propodeum. Fore wing with relative lengths of submarginal, marginal, postmarginal, stigmal veins as 24:12:12:6; stigma small; basal vein with row of setae along length. Petiole transverse, smooth. Gaster ovate acuminate, length 1.6X width (36:23); hind margin of Tl nearly straight, slightly convex mesally; hypo- pygium extending Vi gastral length. Discussion. — This species is distinct from all other Thinodytes species by the characters listed in the key: scape and legs beyond coxae pale, nonmetallic; frenum and median panels of propodeum smooth; plicae rounded and smoothly convergent posteriorly. In these characters, T. santerna is phenetically similar to species of Noto- glyptiis. Type Material— The holotype (USNM) was collected 3 April 1984, at the Eagle Borax Works in Death Valley National Monument, Inyo County, California (USA) by E. E. Grissell on Distichlis in a brackish marsh. Etymology. — The specific epithet of this species is from the Latin noun santerna, meaning borax, and refers to the locality where the type specimen was collected. Biology. — Nothing is know of the host(s) of T. santerna. Mauleus Graham, 1981 Type Species: Mauleus maderensis Gra- ham, 1981 (examined); original designa- tion. Description. — Body very dark green or blue; scape brownish yellow, nonmetallic. Head, pronotum, mesoscutum, scutellum (including frenum), dorsellum, median panels of propodeum, petiole alveolate; gastral tergites nearly smooth. Head with clypeus bidentate (left tooth compound) (Fig. 13), lateral part of mouth margin with short shallow genal concavity; anten- nal torulus IX own diameter above LOcL. Antenna with scape cylindrical, >6X as long as wide; flagellum length less than head width in females, about equal to head width in males; funicular segments cylindrical; MPP sensilla in single row; fe- male club simple apically and with small patch of micropilosity on apical segment. Male maxilla with palps slender, stipites unenlarged. Mesosoma (Figs 8, 21) arched dorsally; pronotum with collar short (Figs 23-25), nearly level with vertex dorsally, anterior edge rounded; mesoscutum with notauli shallow, impressed lines at most; scutellum as long as wide, lacking anterior median groove, frenum indistinguishable from remainder of scutellum; dorsellum short, length about equal to length of ridge across anterior margin of propo- deum; propodeum (Fig. 20) with median panels short (width about 2X median length), plicae and median carina well de- veloped and connected posteriorly by W- shaped carina; spiracles strongly ovate al- most linear. Fore wing with relative lengths of veins as follows: submarginal > Volume 4, 1995 17 Figs. 23-25. 23, Maidens maderensis Graham, female pronotum and mesonotum; 24, Maidens iligneus n. sp., female pronotum and mesonotum; 25, Maidens venetiis n. sp., female pronotum and mesonotum. marginal > postmarginal > stigmal; stig- ma small, width about 2X width of stig- mal vein; costal cell with complete row of setae and sometimes a partial second row; basal cell bare; basal vein setose; specu- lum present, open posteriorly. Petiole (Fig. 20) longer than wide, with basal flange thickened laterally, without median cari- na, without lateral setae. Gaster of females lanceolate, length 1.6 or more times width; hypopygium extending V3 length of gaster or more; in both sexes, hind margin of Tl sinuous laterally, emarginate medially (Figs 8, 21). Discussion. — This genus is placed in the Halticoptera-group as defined in this paper by: the rounded pronotum, shallow no- tauli, undifferentiated frenum, propo- deum with a median carina and plicae connected by a W-shaped carina, reticu- late petiole with a complete basal flange, and Tl that is sinuous laterally and emar- ginate medially. Maidens and Halticopitera exhibit considerable phenetic similarity. The clypeus of Mauleus is secondarily bi- dentate (Fig. 13); the left hand clypeal den- ticle is divided by a sulcus formed as a result of the near fusion of the two ap- proximated left hand denticles. These den- ticles are shown in their plesiomorphic separated state in Thinodytes (Fig. 11). The same compound condition of the left den- ticle is found in Halticoptera Spinola (Hey- don, unpublished data). [In contrast, in the unrelated genus Sphegigaster Spinola, the left tooth in the bidentate clypeus is a single unit (Heydon, unpublished data)]. The compound clypeal denticular struc- ture and the nonmetallic scape are possi- ble evidence of a close phylogenetic rela- tionship between Halticoptera and Mauleus. Halticoptera is well-defined cladistically relative to Mauleus by the very low inser- tion of the antennae (at or below the LOcL) and by the male maxilla, which has the terminal two segments of the palps flattened, expanded, and nearly always yellow and often the stipites also expand- ed. Two apomorphic characters readily define Mauleus relative to Halticoptera and related genera: 1. The propodeum is short- ened in Mauleus (median panels over 2X as wide as long in Mauleus species exam- ined). 2. The basal flanges of the petiole are exceptionally large and thick, giving the petiole a connate appearance. The bas- al flanges in related genera are generally free-standing lamella and the petiole is more or less cylindrical. Biology. — Mauleus iligneus has been reared from pupae of the native holly leaf miner, Phytouiyza ilicicola Loew (Diptera: Agromyzidae). Like many other genera of the Miscogastrinae, it is likely that Mau- leus species are parasitoids of the pupal stages of leaf-mining or stem-mining Dip- tera. 18 Journal of Hymenoptera Research KEY TO DESCRIBED SPECIES OF MAULEUS GRAHAM 1. Pronotal collar with sides parallel in dorsal view (Fig. 23) 2 - Pronotal collar with sides widest near anterior margin and converging posteriorly in dorsal view (Figs 24, 25) 3 2. Vertex and mesoscutum with conspicuous pale setae. Propodeum with reticulations much less coarse than on scutellum, median panel each with broad shallow groove along anterior margin nigritus (Howard) - Vertex and mesoscutum with indistinct dark setae. Propodeum with reticulations as coarse as on scutellum, median panel each with a pair of elongate sublateral depressions along anterior margin maderensis Graham 3. Propodeum with anterior depression between basal foveae; median panels with extensive area of weak, almost smooth sculpture. Pronotum with humeral angles acute, coming to blunt points in dorsal view (Fig. 25). MPP sensilla in two or more rows on funicular segments venetus Heydon - Propodeum without distinct anterior depression between basal foveae; median panels al- most entirely alveolate, with at most a small central patch of weak sculpturing. Pronotum with humeral angles either squared or slightly and smoothly convergent posteriorly (Fig. 24). MPP sensilla in single row on funicular segments (male of M. cultratus unknown) ... 4 4. Eye height 2.5-2.7 X genal distance. Female with combined length of head and mesosoma longer than gaster; gaster less than twice as long as wide; hypopygium extending around % gastral length (United States) iligneus Heydon - Eye height 2. 1-2.4 X genal distance. Female with combined length of head and mesosoma less than or equal to length of gaster (Fig. 8); gaster more than twice as long as wide; hypopygium extending to near tip of gaster (southern Mexico to Argentina) cultratus Heydon Maidens cultratus Heydon, new species Structure: Body length (excluding ovi- Fig. 8 positor sheaths) 1.6 mm. Head width 1.2X height (26:21), 2.2 X length (26:12); genal Holotype. female. -Color: Head, pleural concavity extending Vs malar distance; eye regions, propodeum dark blue; pehole j^^-g^t 1.4X length (13.5:9.5), 2.2X malar bluish black; collar, dorsum of mesosoma distance (13.5:6.0), length 4.8X temple greenish black; gaster dark brown, Tl length; ratio of MOD, POL, OOL, LOL as with dark blue reflections. Antenna with 2:4:6:3; torulus IX own diameter above scape brownish yellow with weak metallic lqcL. Antenna with length of pedicel plus reflections; pedicel, flagellum brown. Legs flagellum 0.81 X head width (21:26); ratio with coxae, trochanters, femora dark blue; of lengths of scape, pedicel, annelli, Fl-6, tibiae brown except basal and apical tips dub as 9.0:3.5:1.0:1.5:2.0:2.0:2.0:2.0:2.0:5.0; brownish yellow; fore tarsi brown, middle widths of Fl, F6, club as 2:3:3; MPP sparse, and hind tarsi yellow-brown with pretarsi only one or two visible per segment from brown. Wing veins pale brown. single view. Mesosoma length 1.4X width Sculpture: Clypeus, median portion of (33:23); pronotum with sides converging face finely alveolate; remainder of head, posteriorly; notauli incomplete; propo- mesonotum, scutellum, frenum, median deum with width of median panels 2.2 X panels of propodeum alveolate; petiole length (11:5); basal fovea a shallow, tri- finely alveolate; gaster smooth except T5- angular depression; plicae fading out in 7 weakly coriaceous. anterior Vz; spiracles 0.5 X own diameter Volume 4, 1995 19 from anterior margin of propodeum. Wing with relative lengths of submargin- al, marginal, postmarginal, stigmal veins as 27:16:14:7; costal cell with single com- plete row of setae; basal vein setose with one seta posteriorly on cubital vein. Peti- ole length 1.5X width (9:6); sides narrow- ing posteriorly. Gaster lanceolate (Fig. 8), length 2.6 X width (42:16), length l.OX that of head and gaster (42:42); hypopygium reaching to apex of T7; ovipositor sheaths exserted for distance equal to half length of hind tibia. Variation. — The color of the head, pleu- ral regions, and petiole varies from dark blue, as in the holotype, to dark green and nearly concolorous with the dorsum of the mesosoma. The body length of specimens from the type locality is about 1.9 mm, whereas the length of the specimen from Ixtapan is 2.3 mm, the one from Morelia is 2.4 mm, and the female from Panama is 3.1 mm in length. The ovipositor sheaths are exserted for a distance equal to ¥i to Vi the length of the hind tibia. Discussion. — Maidens cultratus can be distinguished from M. iligneus by the char- acters given in the discussion section for that species. Eti/mologx/. — The species name comes from the Latin word cultratus, meaning knife-shaped, and refers to the shape to the female gaster. Type Material. — The holotype female (CNCI) and one paratype female were col- lected from San Cristobal de las Casas, Mexico, on 29 June 1969. Eight other para- type females were collected as follows (CASC, CDAE, CNCI, SEMC, USNM): Ar- gentina. SALTA: Rosario de Lerma, 4- 8.XI.1983 (Malaise trap). Mexico. CHIA- PAS: San Cristobal de las Casas, 1- 12.V.1969; MICHOACAN: Morelia, 6.IX.- 1938; MEXICO: Ixtapan, 9.VII.1954; TA- MAULIPAS: 6 miles n. Ciudad Victoria. Panama. Chiriqui, XII. 1946. Biology. — The host(s) of this species are unknown. Mauleus iligneus Heydon, new species Figs. 13, 20, 21, 24 Holotype, female. — Color: Like that of M. cultrata except pedicel not metallic, and middle and hind tarsi white. Sculpture. Clypeus and immediate vi- cinity finely alveolate (Fig. 13); remainder of head, mesoscutum, scutellum, frenum, median panels of propodeum alveolate; petiole finely alveolate; gaster smooth ex- cept T6 and T7 coriaceous. Structure. — Body length 1.8 mm. Head width 1.3X height (29:23), 2.2X length (29: 13); eye height 1.4X length (15.0:10.5), 2.5 X malar distance (15:6), length 5.2 X temple length (10.5:2.0); ratio of MOD, OOL, POL, LOL as 2.5:4.0:7.0:3.0; torulus located IX own diameter above LOcL. An- tenna with length of pedicel plus flagel- lum 0.88 X head width (25.5:29.0); ratio of lengths of scape, pedicel, annelli, Fl-6, club as 11.0:3.0:1.0:2.5:2.5:2.5:2.5:2.5:2.5:7.0; widths of Fl, F6, club as 2.0:3.0:3.5. Me- sosoma length 1.5X width (36.0:24.5); pro- notum with humeral angles squared but sides convergent posteriorly (Fig. 24); no- tauli extending to hind margin of mesos- cutum as impressed lines; propodeum (Fig. 20) with plicae fading out in anterior half, basal fovea extending halfway down median panels and bordered mesally by carina, nuchal area raised but acarinate anteriorly, spiracles almost on anterior margin of propodeum. Fore wing with rel- ative lengths of submarginal, marginal, postmarginal, stigmal veins as 25:15:11:6; costal cell with one complete setal row and one partial row distally; basal vein with row of three setae. Petiole (Fig. 20) length 1.3X width (8:6); narrowing poste- riorly; basal flanges large. Gaster fusiform (Fig. 21), length 0.87X length of head and mesosoma (39:45); length 1.8X width (39: 22); hypopygium extending % length of gaster; ovipc^sitor sheath exserted for a distance equal to/ length of hind tibia. Allotype. — Male. Color similar to holo- type except frons, callus green; flagellum 20 Journal of Hymenoptera Research paler ventrally; tibiae brownish yellow, allotype (USNM), and an additional nine Body length 1.8 mm. Head with ocelli rel- female and seventeen male paratypes atively larger, ratio of MOD, OOL, POL, (UCDC, USNM) were reared from Phijto- LOL as 3.0:3.0:7.5:3.5. Antenna with mi/za ilicicola on Ilex opaca Solander at Lex- lengths of pedicel plus flagellum l.Ox ington, Kentucky by D. A. Potter collected head width (31:31); relative lengths of 16 May 1984. An additional 43 paratypes scape, pedicel, annelli, Fl-6, club as 10.0: were collected as follows (CNCI, UCDC): 3.5:1.0:3.0:3.5:3.0:3.0:3.0:3.0:8.0; widths of ILLINOIS: Cave-in-Rock State Park, near Fl, F6, club as 2.5:3.0:3.0; setae reclinate. Cave-in-Rock, 4.VI.1981, 1 female. KAN- Gaster ovate, length 1.8X width (42: SAS: Oswego, 17.V.1976, 1 female. MARY- 23)(Specimen critical-point dried so gaster LAND: College Park, V.1954 (ex holly leaf in air-dried specimen will be shorter). miner) 1 male, VII-VIII.1937 (ex Phytomyza Variation. — Length of female specimens ilicis), 4 females, 8 males; Laurel, varies between 1.5 and 2.1 mm and males 14.V.1965, 1 female. NEW JERSEY: New between 1.1 and 1.8 mm. Other than the Brunswick, 26.V.1947 (ex Phytomyza ilici- rather large variation in size, this is a mor- cola), 1 female. TEXAS: Houston, 8.XII.- phologically uniform species. 1929 (ex leaf miner on Ilex vomitoria Solan- Discussion. — Mauleus iligneus differs der), 1 female, 1 male; Sanderson, 9.V.- from M. cultratus in the following: 1. The 1912, 1 female. VIRGINIA: Norfolk, V.- ratio of the eye height to the genal dis- 1929 (ex Phytomyza ilicis), 2 females, 6 tance is more in M. iligneus [x= males; Richmond, V.1938 (ex P. ilicis), 10 2.56±(S.E.)0.022 (n = 12); range 2.5-2.7] females, 4 males. than in M. cultratus [x = 2.24 ±0.038 (n = 7); Biology. — This species has been reared range 2.1-2.4]. 2. Mauleus iligneus has a numerous times from the complex of Phy- partial second row of setae in the costal tomyza mining the leaves of holly [Potter cell of the fore wing, which is lacking from and Gordon 1985 (as Sphegigaster sp.), Un- M. cultratus. 3. The ratio of the length of derhill 1943 (as Sphegigastrinae, new ge- the head and mesosoma to the length of nus, new species), and Langford and Cory the gaster is relatively greater in M. ilig- 1936 (as Sphegigaster sp.)]. It may also be neus [x = 1.23 ±0.03 (n = 10); range 1.2- the species called Halticoptera sp. by Kulp 1.4] (Fig. 21) than in M. cultratus (1968), but voucher specimens from Kulp [x=0.918±0.035 (n=6); range 0.8-1.0](Fig. have not been seen. Host records prior to 8). 4. The ratio of the gastral length divid- 1968 are questionable because the Phy to- ed by its width is less for M. iligneus myza complex on holly was not studied in [x=1.76±0.04 (n=10); range 1.7-1.9] than detail until then (Kulp 1968). for M. cultratus [x=2.56±0.11 (n=6); range Potter and Gordon (1985) reported that 2.2-3.1)]. 5. The hypopygium extends to Mauleus iligneus was a primary parasite about % the length of the gaster in M. //- when its agromyzid host, the native holly /^news (Fig. 21), but is nearly even with the leafminer, Phytomyza ilicicola, was unpar- tip of T7 in M. cultratus (Fig. 8). 6. The asitized, but was a facultative hyperpar- ovipositor sheaths are exserted for a asite on Opius striativentris Gahan (Hy- length equal to Va the length of the hind menoptera: Braconidae) when that species tibia in M. iligneus (Fig. 21), but for about had already parasitized the agromyzid V3 to Vi the length of the hind tibia in M. maggot. Potter and Gordon reared no cultratus (Fig. 8). Mauleus iligneus from puparia collected Etymology. — The specific epithet for just a few days prior to the emergence of Mauleus iligneus is an adjective based on Phytomyza ilicicola and speculate that Mau- the Latin word ilex, meaning holly. leus iligneus parasitizes the fly late in its Type Material. — The holotype (USNM), pupal stage. Volume 4, 1995 21 Mauleus maderensis Graham Fig. 23 Mauleus maderensis Graham, 1981:8. Holotype, female (BMNH) Hym. Type No. 5.3454 (examined). Boucek and Rasplus, 1991:41. Diagnosis. — Mauleus maderensis differs from the three newly described species in that it has the sides of the pronotum in dorsal view as for most other Pteromali- dae — more or less parallel, but weakly convex, with the broadest point being about halfway between its anterior and posterior edge (Fig. 23). The three new species of Mauleus described in this paper have the pronotum in dorsal view broad- est near its anterior margin and distinctly convergent posteriorly. Mauleus luaderensis is similar in propodeal structure to the Ca- ribbean species M. niritus. However, these two species are easily distinguished by the characters given in the key. Distribution. — Madeira (Pico das Arru- das, near Sao Martinho), and possibly Mexico (Boucek and Rasplus, 1991) be- cause the species is associated with plants of Mexican origin (Boucek, pers. comm.). Biology. — The insect host(s) of M. mad- erensis remain unknown. Mauleus nigritus (Howard), new combination Polycystus nigritus Howard, 1897:142. Holotype, female (BMNH); Hym. Type No. 5.876 (examined). Heydon, 1989:193. Diagnosis. — The type of Mauleus nigritus is fragmented with most of the antennae gone and the gaster and petiole mounted on the card separately from the remainder of the body. Enough is remaining how- ever, to confirm that this species belongs in Mauleus. Mauleus nigritus differs from the three newly described species in that it has the sides of the pronotum in dorsal view more or less parallel, with the broad- est point being about halfway between its anterior and posterior edge. The three new species of Mauleus described in this paper have the pronotum in dorsal view broad- est near its anterior margin and distinctly convergent posteriorly. Mauleus nigritus is similar in pronotal structure to M. mader- ensis; however, these two species are eas- ily distinguished by the characters given in the key. Distribution. — Mauleus nigitus is known to me only from the holotype collected at Baltazar, on the windward side of Grena- da (West Indies) by H. H. Smith. Biology. — The insect hosts of M. nigritus remain unknown. Mauleus venetus Heydon, new species Fig. 25 Holotype, female. — Color: Body black ex- cept anterior aspect of head, collar, me- soscutum steel blue and lateral portions of metanotum, propodeum, Tl blue. Anten- na with scape, ventral side of pedicel and Fl brownish yellow, remainder of pedicel and flagellum brown. Legs with coxae dark blue; femora brown with metallic blue reflections, except basal and apical tips brownish yellow; tibiae brownish yel- low with weak brown bands mesally; tarsi white with pretarsi brown. Fore wing clear with veins brownish yellow, paras- tigma darker. Sculpture. — Clypeus and immediate vi- cinity finely alveolate; remainder of head, mesoscutum, scutellum, frenum, median panels of propodeum alveolate; petiole finely alveolate; gaster smooth except T6 and T7 coriaceous. Structure. — Body length 2.2 mm. Head width 1.4X height (42:30), 2.5 X length (42.0:16.5); eye height 1.5X length (20:13), 2.0X malar distance (20:10), length 7.5X temple length (13:2); torulus 2X own di- ameter above LOcL; ratio of MOD, OOL, POL, LOL as 3.0:6.5:9.0:4.0. Antenna with length of pedicel plus flagellum 0.90 X head width (38:42); relative lengths of scape, pedicel, annelli, Fl-6, club as 15.0: 4.0:1.0:4.5:4.5:4.5:4.0:4.0:3.5:8.0; relative widths of Fl, F6, club as 3.5:4.0:4.0; MPP sensilla in two rows on each flagellar seg- 22 Journal of Hymenoptera Research merit; terminal club segment with small ventral patch of micropilosity. Mesosoma length 1.5 X width (26.5:18.0); pronotum with humeral angles acute (Fig. 25); dor- sellum short, anterior and posterior edges parallel; propodeum with region between basal foveae depressed (this depression shorter mesally than laterally and bound- ed posteriorly by weak carina), plicae fad- ing out before reaching anterior margin of propodeum, nuchal region raised, mar- gined anteriorly by carina, spiracles al- most on anterior margin of propodeum. Fore wing with relative lengths of sub- marginal, marginal, postmarginal, stigmal veins as 37:23:15:8; costal cell with 1 com- plete and 1 distal partial row of setae; bas- al cell bare; basal vein with row of setae; speculum open posteriorly. Petiole length 1.1 X width (9:8). Gaster length 1.4X width (43:31), 0.66 X combined length of head and mesosoma (43:65); hypopygium ex- tending about Vi gastral length; ovipositor sheaths hardly extending beyond hind margin of T7. Allotype, male. — Similar to female ex- cept: body length 1.8 mm. Head with ratio of MOD, OOL, POL, LOL as 3.5:4.0:7.0:3.0. Antenna with length of pedicel plus fla- gellum 1.1 X head width (37:34); relative lengths of scape, pedicel, annelli, Fl-6, club as 11.0:3.0:1.0:4.5:4.5:4.0:4.0:4.0:4.0:9.0; relative widths of Fl, F6, club as 3.0:3.5: 3.0; MPP sensilla numerous and distrib- uted over funicular segments in many rows; setae reclinate, nearly absent. Gaster length 1.1 X width (29:26). Diagnosis. — Mauleus venetus is distin- guished from the other two Nearctic Mau- leus species by the acute humeral angles of the pronotum, the broad depression along the anterior margin of the propo- deum, and the lack of sculpture over most of the median panels of the propodeum. Mauleus venetus most closely resembles M. iligneus — the coloration of the two species is almost identical except the propodeum is brighter blue in M. venetus. In addition, the MPP sensilla of M. venetus are ar- ranged in two or more rows on each fu- nicular segment in both sexes, whereas they are arranged in only one row in M. iligneus. Etymology. — The species name comes from the Latin word venetus, meaning blue or sea-blue, and refers to the distinc- tive blue propodeum in this species. Type Material. — The holotype (IRCW) is a female, collected in Grant Co. (T6N, R6W, SI 7), Wisconsin (USA), in a malaise trap exposed 7-14.VL1976. The allotype (UCDC) was collected by the author on the South Farms of the University of Illi- nois, near Champaign, Illinois (USA) on 26.V.1985. Six paratypes were collected as follows (CNCI, IRCW, UCDC, USNM): Canada. QUEBEC: La Trappe, 12.VII.1942, 1 female. United States. ILLINOIS: Uni- versity of Illinois South Farms, near Champaign, 26.V.1985, 1 male. MICHI- GAN: Midland Co., 2.VII.1943, 1 male. VIRGINIA: Winchester, 16.VI.1964, 1 male. WISCONSIN: Grant, T6N, R6W, S17, 3-8.VI.1976 (gypsy moth Malaise trap), 1 female, 14-21. VI.1976 (gypsy moth Malaise trap), 1 female. Biology. — The host(s) of M. venetus are unknown. ACKNOWLEDGMENTS I thank Melissa Bennett and two anonymous re- views for their careful reading of this manuscript. I also thank the following people for the loan of ma- terial used in this study: Dr. J. S. Noyes, The Natural History Museum, London, ENGLAND (BMNH); Dr. W. J. Pulawski, California Academy of Sciences, San Francisco, CA (CASC); Dr. F. G. Andrews, California State Collection of Arthropods, Sacramento, CA (CDAE); Dr. G. A. P. Gibson, Canadian National Col- lection, Ottawa, ON (CNCI); Dr. W. E. LaBerge, Illi- nois Natural History Survey, Champaign, IL (INHS); S. Krauth, University of Wisconsin, Madison, WI (IRCW); Dr. R. W. Brooks, Snow Entomological Museum, University of Kansas, Lawrence, KS (SEMC); Dr. E. E. Grissell, United States National Mu- seum, Washington, D.C. (USNM). The acronym for the collection of the Bohart Museum at the University of California, Davis, CA is UCDC. Volume 4, 1995 23 LITERATURE CITED Allen, M. W. and R. H. Painter. 1937. Observations on the biology of the wheat maggot in Kansas. Journal of Agricultural Research 55:215-238. Allen, W. A. and R. L. Pienkowski. 1973. Parasites reared from puparia of the frit fly, Oscinella frit, in Virginia. Environmental Entomology 2:615-617. Anderson, H. 1990. Three new species of Halticopter- ina Erdos (Hymenoptera: Pteromalidae). Pan-Pa- cific Entomologist 66:131-139. Askew, R. R. 1970. Observations on the hosts and host food plants of some Pteromalidae [Hym., Chalcidoidea]. Entomophaga 15:379-385. Boucek, Z. 1970. Contribution to the knowledge of Italian Chalcidoidea, based mainly on a study at the Institute of Entomology in Turin, with de- scriptions of some new European species. Me- morie della Societa Entomologica Italiana 49:35-102. Boucek, Z. 1976. African Pteromalidae (Hymenop- tera): new taxa, synonymies and combinations. Journal of the Entomological Societi/ of Southern Af- rica 39:9-31. Boucek, Z. 1977. A faunistic review of the Yugosla- vian Chalcidoidea (parasitic Hymenoptera). Acta Entomologica Jugoslavica 13(Suppl.): 3-145. Boucek, Z. 1993. New taxa of North American Pter- omalidae and Tetracampidae (Hymenoptera), with notes. Journal of Natural History 27:1239- 1313. Boucek, Z. and J.-Y. Rasplus. 1991. Illustrated Key to West-Pa lea rctic Genera of Pteromalidae (Hy- menoptera: Chalcidoidea). Institut National de la Recherche Agronomique. Paris. Burks, B. D. 1979. Family Pteromalidae. pp.768-835. In, K. V. Krombein, P. D. Hurd, D. R. Smith, and B. D. Burks (eds.). Catalog of Hi/menoptera in America North of Mexico. Smithsonian Institution Press. Washington, D. C. De Santis, L. 1979. Catdlogo de los Himenopteros Cal- cidoideos de America al sur de los Estados Unidos. Comision de Investigaciones Cienti'ficas de la Provincia de Buenos Aires. La Plata. Dzanokmen, K. A. 1978. Identification of the insects of the European part of the USSR. Vol. 3. Hy- menoptera. Second Part. Pteromalidae. Opredeli- teti Faune SSSR 120:57-228. Gahan, A. B. 1933. The serphoid and chalcidoid par- asites of the Hessian fly. U.S. Department of Ag- riculture, Miscellaneous Publications 174:1-147. Girault, A. A. 1918. New and old West Indian and North American chalcid-flies (Hym.). Entomolog- ical News 29:125-130. Graham, M. W. R. de V. 1956. A revision of the Walker types of Pteromalidae (Hym., Chalcidoi- dea). Part 2 (including descriptions of new gen- era and species). Entomologist's Monthly Magazine 92:246-263. Graham, M. W. R. de V. 1969. The Pteromalidae of northwestern Europe (Hymenoptera, Chalcidoi- dea). Bulletin of the British Museum (Natural His- tory) Entomology Supplement 16:1-908. Graham, M. V. R. de V. 1981. A survey of Madeiran Chalcidoidea (Insecta: Hymenoptera) with addi- tions and descriptions of new taxa. Bocagiana 58: 1-20. Hedqvist, K.-J. 1975. Notes on Chalcidoidea 7. A key to Swedish species of the genus Halticoptera Spin. and three related genera (Hymenoptera: Pter- omalidae). Entomologica Scandinavica 6:167-181. Hedqvist, K.-J. 1983. Malmsteklar pa Olands Stora alvar. Entomolgisk Tidskrift 104:165-168. Hendrickson, Jr., R. M. 1979. Field studies and par- asites of Liriomyza trifoliearum (Diptera: Agro- myzidae) in northern USA. Journal of the Neiv York Entomological Society 87:299-303. Heydon, S, L. 1988. A review of the world species of NotoglypHus Masi (Hymenoptera: Pteromali- dae). Proceedings of the Entomological Society of Washington 9l:U2-\23. Heydon, S. L. 1989. Relationships among Holarctic genera in the Cyrtogaster-group with a review of the species of North America north of Mexico. (Hymenoptera: Pteromalidae). Journal of the New York Entomological Society 97:192-217. Heydon, S. L. 1993. Syntomopus Walker: The nearctic species with a review of known host associations (Hymenoptera: Pteromalidae). Journal of Hyme- noptera Research 2:107-116. Heydon, S. L. and W. E. LaBerge 1988. A review of North American species of Sphegigaster north of Mexico and the biology of their hosts (Hyme- noptera: Pteromalidae). Journal of the Kansas En- tomological Society 61: 258-277. Howard, L.O. 1897. On the Chalcididae of the island of Grenada, B.W.I. Journal of the Linnean Society of London. Zoology 26:129-178. Kamijo, K. 1978. Chalcidoid parasites (Hymenop- tera) of Agromyzidae in Japan, with description of a new species. Kontyu 46:455—469. Kulp, L. A. 1968. The taxonomic status of dipterous holly leaf miners (Diptera: Agromyzidae). Uni- versity of Maryland Agricultural Experiment Station Bulletin A-155:l-42. Langford, G. S. and E. N. Cory. 1936. The holly leaf- miner and its control. Proceedings of the National Shade Tree Conference 13:109-112. Morrill, W. L. and R. W. Kieckhefer. 1971. Parasitism of the wheat stem maggot in South Dakota. Jour- nal of Economic Entomology 64:1129-1131. Nikol'skaya, M. N. 1937. The chalcidoid parasites (Hymenoptera) of some injurious flies of grain crops. Entomological Review 27:1-27. Peck, O. 1951. Superfamily Chalcidoidea. pp. 410- 594. In, C. F. W. Muesbeck, K. V. Krombein, and H. K. Townes. Hymenoptera of America North of 24 Journal of Hymenoptera Research Mexico, Synoptic catalog. U.S. Department of Ag- riculture, Agriculture Monograph 2:1-1420. Peck, O. 1963. A catalogue of the Nearctic Chalci- doidea (Insecta: Hymenoptera). Canadian Ento- mologist Supplement 30:1-1092. Peck, O., Z. Boucek, and A. Hoffer. 1964. Keys to the Chalcidoidea of Czechoslovakia. Memoirs of the Entomological Society of Canada 34:1-120. Potter, D. A. and F. C. Gordon. 1985. Parasites as- sociated with the native holly leafminer, Phyto- myza ilicicola Loew (Diptera: Agromyzidae), on American holly in Kentucky. Journal of the Kansas Entomological Society 58:727-730. Schmiedeknecht, O. 1909. Hymenoptera family Chalcididae. Genera Insectorum 97:1-550. Spencer, K. A. 1976. The Agromyzidae (Diptera) of Fennoscandia and Denmark. Fauna Entomologica Scandinavica 5:1-606. Spencer, K. A. and G. C. Steyskal. 1986. Manual of the Agromyzidae (Diptera) of the United States. U.S. Department of Agriculture, Agriculture Hand- book 638:1^78. Takada, H. and K. Kamijo. 1979. Parasite complex of the garden pea leaf-miner, Phytomyza horticola Goureau in Japan. Kontyu 47:18-37. Thomson, C. G. 1876. Hymenoptera Scandinaviae. 4. Pteromalus (Svederus). Lund. Thomson, W. R. 1958. Catalogue of parasites and pred- ators of Insect pests. Section 2. Host parasite cata- logue. Part 5. Hosts of the Hymenoptera (Mis- cogasteridae to Trigonalidae), Lepidoptera and Strepsiptera. p. 562-698. Commonwealth Insti- tute of Biological Control. Ottawa, Ontario. Underhill, G. W. 1943. Some insect pests of orna- mental plants. Virginia Agricultural Experiment Station Bulletin 349:27-38. Walker, F. 1833. Monographia Chalciditum. Ento- mologist's Magazine 1:367-384, 455^66. Walker, F. 1839a. Monographia Chalciditum, 1. Lon- don. Walker, F. 1839b. Monographia Chalciditum, 2. Lon- don. Walker, F. 1843. Descriptions of Chalcidites discov- ered near Conception, in South America, by C. Darwin, Esq. Annals and Magazine of Natural His- tory 12:30-32. Walker, F. 1846. List of the specimens of Hymenopterous insects in the collection of the British Museum. Part I-Chalcidites. London. J. HYM. RES. Vol. 4, 1995, pp. 25-32 An Australian Masarine, Rolandia angulata (Richards) (Hymenoptera: Vespidae): Nesting and Evaluation of Association with Goodenia (Goodeniaceae) Friedrich W. Gess, Sarah K. Gess and Robert W. Gess Albany Museum, Grahamstown, South Africa Abstract. — An account is given of some aspects of the nesting of Rolandia angulata (Richards) (Masarinae). This wasp nests in multicellular burrows in compacted sandy soil. The sloping en- trance to the burrow, not surmounted by a turret, is concealed beneath a pebble, a plantlet or a suitable item of debris. The main thrust of the shaft is vertically downwards, however, at intervals it curves outwards to end in a horizontal cell so that each cell, except the last excavated and therefore deepest one, appears to be accessed by a lateral shaft. The cells are unlined. The archi- tecture of the nest and the method of its construction are discussed. The association between Rolandia angulata and Goodenia (Goodeniaceae) flowers, the source of nectar and pollen, is evalu- ated. It is shown that the association with Goodenia pinnatifida Schldl., at least, is mutually bene- ficial. Indeed it is suggested that in some areas, at some times, R. angulata may be the most important potential pollinator of this plant. The genus Rolandia Richards, 1962 is (28.03S, 148.30E), and 85 km E of St listed by van der Vecht and Carpenter George (28.03S, 148.30E) (27-29.X.1993, (1990) as a junior synonym of Metaparagia F.W., S.K. and R.W. Gess). Meade-Waldo, 1911. As the analysis on There is only one published observation which this is based has not yet been pub- on the nesting of Rolandia — that of R. ma- lished the generic name Rolandia is used culata entering a turretless, simple, in the present paper. Rolandia is restricted oblique, blindly ending burrow in sandy to Australia. Four species are known, two, ground (Houston, 1984). However, Hous- R. maculata (Meade-Waldo) and R. hous- ^^^ is currently investigating further the toni Snelling, from Western Australia, a riesting of this species and has made some third, R. borreriae Snelling, from Northern ^^^^i^^ available for comparison. Territory, and the fourth, R. angulata ^^ investigations concerning R. angii- (Richards), the subject of this paper, from ^"^" published m the present paper were Queensland and New South Wales. undertaken by the authors during the r, .^ J- 1 I u u J J course of a fieldtrip to Australia in Octo- Rolandia angulata has been recorded , ^^^^ w , £ .u .. r-i 1 J J ber 1993. Voucher specimens have been from southwestern Queensland and , . , . , »„ ,, ^ - , o .1 T»r , r deposited in the Albany Museum (R. an- northwestern New South Wales, from [,.,<- ^ . i .i i n u no o/ic 1/ir /.niri /o- u j gulata, the forage plants, and the bee visi- Cunnamulla [28.04S, 145.40E (Richards ; «. i.u r i x. x .^u a .. i- .^^„ T^. , . , , ^ ,. . tors to the forage plants), the Australian 1968 as Riekia auQulata and SneWine, 1986) xt .• it .. V- n ..• /- u /o , , '^ t> / National Insect Collection, Canberra [R. m the north to 90 km W of Cobar, Barnato ,,,^,,,,^,,1 and the Australian National Tanks [31.38S, 144.59E] (about 400 km Herbarium, Canberra (the forage plants). south of Cunnamulla) (Richards 1968) in the south, and from three sites to the east NESTING of Cunnamulla, 80 km E of Cunnamulla Description of the Nesting Areas. — Two (28.04S, 145.40E), 27 km W of St George nesting areas of Rolandia angulata were lo- 26 Journal of Hymenoptera Research Figs. 1-4. 1, Eucalyptus woodland between St George and Cunnamulla, southwestern Queensland; 2, Acacia scrub 80 km east of Cunnamulla, southwestern Queensland; 3, a nesting site of Rolandia ai^gulata 27 km west of St George, southwestern Queensland; 4, a nest entrance of Rolandia angulata beneath a piece of plant debris (approximately X 3.6). cated between St George and Cunnamulla in southern Queensland, one 27 km west of St George on 28 October and the other 80 km east of Cunnamulla on 29 October, the former in dry open woodland domi- nated by Eucalyptus (Fig. 1) and the latter in dry open scrub dominated by Acacia (Fig. 2). Both areas had recently received rain in the form of localized thundershow- ers which had resulted in a growth of an- nuals which were in flower on the road verges and in the low lying areas. The soil in both areas was sandy, compact but fri- able. It was increasingly moist at least to the depth of the deepest nest cells, 38 cm. Flozvers Visited. — The only previous rec- ord of flower visiting by R. angulata is that of Richards (1968) for females collected on Goodenia cycloptera R.Br, in C.Sturt (Good- eniaceae) at Barnato Tanks. In the present study females and males were collected and observed visiting the yellow flowers of a prostrate herb also identified as G. cy- cloptera (Fig. 11) and an erect herb Good- enia pinnatifida Schldl. (Fig. 6) in the two areas where nests were discovered and Volume 4, 1995 27 also in the area 85 km east of St George. At the most easterly site, that is 85 km east of St George, both species of Goodenia were common. Travelling westwards to Cunnamulla it was observed that G. cy- cloptera rapidly became uncommon whereas G. puinntifida became increasingly common and at the site 80 km east of Cun- namulla was the most common roadside herb. The abundance of R. angiilata fol- lowed a similar pattern to that of G. pin- natifida. Visits to the flowers by female R. angii- lata were abundant by lOhOO and contin- ued through the heat of the day and the afternoon. Visits became fewer in the late afternoon and ceased after IThOO. Males were observed to patrol the flowers and only later in the day to visit them for nec- tar. All plants flowering together with the two Goodenia species, most notably several Asteraceae and a Wahlenbergia species (Campanulaceae), were sampled for flow- er visitors. None was being visited by R. angulata. Provision. — Provision in the form of a firm, white pollen loaf was obtained from each of three cells. Pollen from the loaves was examined microscopically and found to match that obtained from the Goodenia flowers. Mate location. — During the morning males patrolled the flowers where they sought the females. They did not alight on the flowers but rested on a neighbouring plant, for example on a grass stem, or on the ground. Numerous attempted copu- lations at flowers were observed and a sin- gle attempt to copulate with a nest-exca- vating female was noted. Several instances of a male and female grappling on the ground next to plants and of "hot pur- suit" were noted. Description of the Nest. — The nest (n = 8) consists of a subterranean burrow (Fig. 5) excavated in horizontal sandy soil in a clearing (Fig. 3). The entrance is concealed beneath a pebble, a plantlet or a suitable item of debris (Fig. 4). It is a simple hole 3-4 mm in diameter, not surmounted by a superstructure. For approximately the first 10 mm the shaft slopes gently down- wards. Thereafter its main thrust is verti- cally downwards, however, at intervals it curves outwards to end in a horizontal cell so that each cell, except the last excavated and therefore deepest one, appears to be accessed by a lateral shaft. These "lateral shafts" radiate out through 360° each deeper than that preceding it. The diameter of the shaft narrows at a depth of approximately 80 mm and then continues constant. The first cell in the nests investigated was at a depth of be- tween 180 and 370 mm, the "lateral shafts" were 30 mm in length, and the cells 13 mm long and 4 mm in diameter at the widest point. A lateral shaft is packed with sand after the cell which ter- minates it has been provisioned. Up to seven cells per nest were recorded. The cells are unlined. Method of Construction of the Nest and Provisioning. — Water is not required for the excavation of the nest as the sandy soil, though compact, is friable. The sand extracted from the burrow is carried out of the shaft held between the head and the prosternum, the genae being fringed with ammochaetae. Whilst excavation is in progress, the female, when leaving the nest, backs out. During the initial stages of burrow excavation the extracted sand is dropped in flight in a more or less con- stant area to one side of the entrance and about 120 mm from it. As the burrow be- comes deeper the extracted sand is dropped further from the nest in a con- stant arc about 250 mm from the entrance. The sand being scattered, there is no de- tectable accumulation and so no tumulus develops. The presence of a recently hatched larva in a sealed, fully provisioned cell indicat- ed that mass provisioning is practised. The sand used for packing a shaft leading to a cell is almost certainly obtained with- 28 Journal of Hymenoptera Research in the nest during the excavation of the shaft leading to the next cell. The fully fed larva spins a white cocoon which completely fills the cell. Like the sealed cell it is therefore rounded at the inner end and truncate at the outer end. Discussion of Nesting. — The nest of R. an- gulata is essentially similar to that of R. maculata, which is described by Houston (pers. comm.) as a vertical burrow in sandy soil, about 300 mm deep and un- lined, with unlined cells at the lower end. Thus the basic nest type for these two spe- cies of Rolnndia can be defined as a mul- ticellular sub-vertical burrow in horizontal ground excavated by the nester, without an entrance turret and with excavated cells not containing constructed cells. When compared with the seven basic nest types recognized for the Masarinae as a whole by Gess and Gess (1992) this ap- pears to fit Nest type 1 except for the lack of an entrance turret. When the method of construction is compared with that of the species listed for this nest type, a basic dif- ference is apparent. Water is used in the excavation and construction of Nests of type 1, like nests of types 2 and 3, whereas the nests of the two Rolandia species are excavated without the use of water. This is possible due to the friable nature of the soil in which they are sited. As noted in Gess and Gess (1992) nesting in friable soil in the Vespidae is probably derived rather than primitive as in the Pompilidae and Sphecidae. The nest type of the two Rolan- dia species is therefore seen as a sub-type which can be derived from Nest type 1. Both species of Rolandia carry sand, ex- tracted from the shaft, out of the burrow held between the head and the proster- num, the genae being fringed with am- Fig. 5. Plan of vertical section of a nest of Rola7tdia attgulata. Cells 2 — 5 followed cell I, radiating out in sequence through 360° and at successively greater depths. 50mm 29 Figs. 6-10. 6, Goodenia pinnatifida, an erect herb; 7, a flower of Goodenia piiiiiatifida, indusium concealed by the bases of the adaxial petals; 8, Rolandia angulata entering a flower of Goodenia pinnatifida, indusium of the flower exposed (X 2.8); 9, Rolandia angulata in nectar drinking position in a flower of Goodenia pinnatifida, indusium of the flower fitting snugly over the wasp's mesosoma (X 2.8); 10, Rolandia angulata withdrawing from a flower of Goodenia pinnatifida, showing dusting of pollen on the head, dorsum of the prothorax, and anterior part of the mesoscutum (X 2.8). mochaetae. The possesion of ammochae- tae fringing the genae is a generic char- acter (Snelling 1986) and it is therefore ex- pected that all members of the genus excavate their nests in a similar manner. The only other vespid genus recorded as having a psammophore for the removal of the spoils of excavation is Pterochihis (Eu- meninae) (Bohart 1940) for which nesting in vertical burrows in friable soil by two species has been described (Isely 1914 and Evans 1956). EVALUATION OF ASSOCIATION WITH FLOWERS VISITED Both R. angulata males and females ob- tain nectar and possibly pollen for their own nourishment and females collect nec- tar and pollen for provisioning their nest- cells apparently solely from Goodenia flow- ers. They are therefore probably depend- ant on Goodenia flowers. To determine whether or not the association between the wasp and the flowers is mutually ben- eficial or not, that is whether or not the wasp in addition pollinates the flowers, requires a consideration of the functional morphology of the flowers, the behaviour of the wasps in the flowers, and wasp/ flower fit. Goodenia flowers are distinctly two lipped. The lower wings of the two ad- axial petals are differentiated in their low- er parts into auricles which envelop the 30 Journal of Hymenoptera Research Figs. 11-13. 11, Goodenia cycloptera, a prostrate herb; 12, three flowers of Goodenia cycloptera, indusium con- cealed by the bases of the adaxial petals; 13, Rolandia angulata in nectar drinking position in a flower of Goodenia cycloptera, indusium of the flower pressed down on the wasp's folded wings (X 3.3). indusium in the flower (Figs 7 and 12) so that it is only exposed when an insect vis- itor pushes the corolla lobes apart as it seeks the nectar at the base of the flower (Figs 8 and 13). The indusium is a cup at the top of the style and the surrounding stigmatic initial that collects and retains pollen from the stamens, which dehisce in the bud, and presents it thus to the pollen vector (Carolin, Rajput and Morrison 1992). Later the stigmatic initials mature and grow out of the indusium and collect pollen from pollen vectors. R. angulata when visiting the flowers for nectar always alights on the lower lip and then pushes its way in beneath the indu- sium so that, if the flower is in the pollen presenting phase, the wasp receives pollen on its head. When it is in the nectar drink- ing position in a flower of G. pinnatifida, the mesosoma (Fig. 9) is snugly capped by the indusium but, when it is in a flower of G. cycloptera, which is deeper, it is the folded wings which are pressed beneath the indusium (Fig. 13). On emerging from a flower of G. pinnatifida it can be clearly Volume 4, 1995 31 seen to have been well dusted with pollen not only on the head but also on the dor- sum of the prothorax and the anterior part of the mesonotum (Fig. 10). When coming thus laden with pollen to a flower with receptive stigmas the wasp would be ide- ally suited to pollinate it. It is not clear, however, whether R. angulata would effec- tively pollinate the deeper flowers of G. cycloptera. It is possible that pollen may be transferred from its head to a receptive stigma but the pollen received on the wasp's wings is unlikely to be successfully transferred. It is concluded that, within its limited distribution, R. angulata is a potential pol- linator, at least, of the widespread species G. pinnatifida with which it therefore has a mutually beneficial association. However, at none of the sites was R. angulata the sole visitor to the Goodenia flowers. At the most easterly site a second ma- sarine, Riekia nocatunga Richards, which was also recorded from Goodenia fascicu- laris F.Muell. & Tate at Kondar to the southeast, was a relatively abundant vis- itor to G. pinnatifida. Furthermore the flowers of both species of Goodenia, par- ticularly those of G. cycloptera, were vis- ited abundantly by bees. The most com- mon species was Leioproctus {Chri/socolle- tes) moretonianus (Cockerell) (Colletidae) which was also recorded from Goodeni- aceae at several sites in Queensland by Michener (1965). Less common visitors were a second, but slightly smaller, spe- cies of Leioproctus {Chrysocolletes) and a species of Megachile (Megachilidae). Rel- atively uncommon visitors were six fur- ther species of Megachile, an additional coUetid, a halictid, a few anthophorids and the honey-bee. Travelling westwards the bees were uncommon at the site 27 km west of St George and at the site 80 km east of Cunnamulla all but the second species of Leioproctus (Chrysocolletes) were absent. At the latter site a third much smaller species of Leioproctus was also re- corded. However, R. angulata was the only abundant visitor suggesting that in some areas, at some times, it may be the most important potential pollinator of G. pinnatifida. ACKNOWLEDGMENTS The following are thanked with appreciation: Terry Houston of the Western Australian Museum for per- mission to quote his unpublished observations con- cerning the nesting of Rolandia maciilata; Jo Cardale of the Australian National Insect Collection, Canberra for facilitating the identification of forage plants; Kir- sten Cowley, Australian National Herbarium, Can- berra for the identifications of the forage plants; Alan Weaving of the Albany Museum for producing black and white negatives from the authors' colour trans- parencies for Figs 1-4 and 6-13; and the South Afri- can Foundation for Research Development for a roll- ing support grant to F.W.Gess, which made possible the fieldwork in Australia. The manuscript was re- viewed by Karl V. Krombein and an anonymous re- viewer, who are thanked for their constructive com- ments. LITERATURE CITED Bohart, R.M. 1940. A revision of the North American species of Pterocheilus and notes on related gen- era (Hymenoptera, Vespidae). Annals of the En- tomological Society of America 33(1): 162-208. Carolin, R.C., M.T.M. Rajput, and D. Morrison. 1992. Goodeniaceae. In: George, A.S. ed., Flora of Aus- tralia, volume 35: Brunoniaceae, Goodeniaceae. Can- berra: Australian Government Publishing Ser- vice, pp. 4-300. Evans, H.E. 1956. Notes on the biology of four spe- cies of ground-nesting Vespidae (Hymenoptera). Proceedings of the Entomological Society of Washing- ton 58(5): 265-270. Gess, F.W. and S.K. Gess. 1992. Ethology of three southern African ground nesting Masarinae, two Celonites species and a silk-spinning Quartinia species, with a discussion of nesting by the sub- family as a whole (Hymenoptera: Vespidae). Journal of Hymenoptera Research 1(1): 145-155. Houston, T.F. 1984. Bionomics of a pollen-collecting wasp, Paragia tricolor (Hymenoptera: Vespidae: Masarinae), in Western Australia. Records of the Western Australian Museum 11(2): 141-151. Isely, D. 1914. The biology of some Kansas Eumen- idae. Kansas University Scietice Bulletin (2)8(7): 233-309. Meade-Waldo. G. 1911. Notes on the family Masar- idae (Hymenoptera), with descriptions of a new genus and three new species. Annals and Maga- zine of Natural History (8)8: 747-750. Michener, CD. 1965. A classification of the bees of the Australian and South I'acific regions. Bulletin 32 Journal of Hymenoptera Research of the American Museum of Natural History 130: 1- 362. Richards, O.W. 1962. A revisional study of the Masarid wasps {Hymenoptera, Vespoidea). London: British Museum (natural History). Richards, O.W. 1968. New records and new species of Australian Masaridae (Hymenoptera: Vespo- idea). Journal of the Australian Entomological Soci- ety 7: 101-104.' Snelling, R.R. 1986. The taxonomy and nomenclature of some Australian paragiine wasps (Hymenop- tera: Masaridae). Contributions in Science 378: 1- 19. van der Vecht, J. and J.M. Carpenter. 1990. A cata- logue of the genera of the Vespidae (Hymenop- tera). Zoologische Verhandelingen 26: 1-62. J. HYM. RES. Vol. 4, 1995, pp. 33-10 Descriptions of the Male of Riekia nocatiinga Richards, the Male and Two Strikingly Distinct Sympatric Colour Forms of Riekia confluens (Snelling) and the Male of Rolandia angiilata (Richards) (Hymenoptera: Vespidae: Masarinae) from Australia Friedrich W. Gess Albany Museum, Grahamstown, 6140, South Africa Abstract. — The male of Riekia nocatunga Richards from southern Queensland, the male and two strikingly distinct sympatric colour forms of Riekia confluens (Snelling), comb, nov., from Western Australia, and the male of Rolandia angnlata (Richards) from southern Queensland are described. The characters distinguishing R. nocatunga and R. confluens are discussed. The Australian Masarinae and their nat- ural history are relatively poorly known. During a recent collecting trip to Australia by the author, S.K.Gess and R.W.Gess, material collected included males of three species, Riekia nocatunga Richards, Riekia confluens (Snelling), comb, nov., and Rolan- dia angnlata (Richards), hitherto described only from the female sex. Two strikingly distinct sympatric colour forms of Riekia confluens were found to occur. One of these was previously unknown and helps to elucidate the identity of a single male Riekia collected by T.F.Houston and dis- cussed by Snelling (1986) who was unable to place it in a species. Riekia Richards, 1962 and Rolandia Rich- ards, 1962 have been sunk as junior sub- jective synonyms of Metaparagia Meade- Waldo, 1911 by van der Vecht and Car- penter (1990) on the basis of an as yet un- published cladistic analysis by Carpenter. Until such time as this analysis is pub- lished and may be studied, the present au- thor considers it best to continue to accept Riekia and Rolandia as genera in their own right. Institutions in which the material stud- ied is deposited are: Albany Museum, Grahamstown, South Africa (AMG); Aus- tralian National Insect Collection, Canber- ra (ANIC); Western Australian Museum, Perth (WAM). Riekia nocatunga Richards Riekia nocatunga Richards, 1962: 55-57, female. This species was described from 7 miles N of Nocatunga [on maps as Nockatunga] [27.40S, 142.40E], Queensland (not New South Wales as in Richards 1962; correct in Richards 1968) (holotype female and 4 female paratypes) and Bourke [30.09S, 145.59E], New South Wales (1 female paratype). Subsequently an additional five females from 90 km W of Cobar, Barnato Tanks [31.38S, 144.59E], New South Wales, were attributed to this species by Richards (1968). He noted that they differed from his original description with regard to the pale markings of which he gave some de- tails. Two of these females, in the collec- tion of the Natural History Museum, Lon- don, have been examined by the present author. They are confirmed as being R. no- catunga. They are less melanistic than the specimens originally described. The dark areas are less extensive and some, in par- 34 Journal of Hymenoptera Research ticular those on the pronotum and tergites each side of mesoscutum where its fur- 1 and 2, are brownish rather than black, rows meet pronotum (usually but not al- The light areas are more extensive and ways present in females) absent in both their colour is a strong yellow rather than males examined, pale lemon-yellow. Antennal scape (with radicle) very Snelling (1986) figured and briefly dis- slightly longer than interantennal dis- cussed a male Riekia from 43 km ENE Lan- tance; first flagellomere less than half (0.4) dor Homestead (25.08S, 116.54E) in West- as long as scape (with radicle), one and a ern Australia. Whereas it agreed generally third times as long as wide at its distal with the original description of R. nocatun- end, and one and one fifth as long as the ga, Snelling stated that there was no cer- second flagellomere. Last three flagellom- tainty that it was the opposite sex of that eres flattened and slightly concave be- species and that he suspected that it was neath; ultimate flagellomere narrowing not. He listed some discrepancies, allow- apically and distinctly curved to form a ance being made for sexual differences, hook. and stated that more material would have Tergite 7 evenly rounded apically. to be available before the specific status of Genitalia (Figs 3 and 4); parameral spine the male could be determined. broad and thick, evenly curved over its Recently collected Riekia material, of rel- length and not hook-like apically; ventral evance to the clarification of the above tax- process of each paramere relatively broad onomic question, consists of associated fe- and short, subtriangular, without a nar- males and males from both Queensland row elongation; digitus short, and Western Australia. The specimens Length 7.5-7.8 mm; length of forewing from Queensland are certainly R. nocatun- 5.8 mm, hamuli 14. ga, as established by the comparison by New Material Examined. — Queensland: the author of several of the females with Lundavra State School, Kondar (28. OSS, the holotype in the Australian National In- 149.51E), 25.X.1993 (F.W.,S.K.& R.W.Gess) sect Collection, Canberra. The specimens 11 females, 1 male (in yellow flowers of from Western Australia on the other hand Goodenia fascicularis F.Muell. & Tate, represent two strikingly distinct colour Goodeniaceae); Southwood Road, western forms of Riekia confluens. One is superfi- boundary of Southwood National Park cially similar looking to Riekia nocatunga (27.56S, 149.30E), 26.x. 1993 (F.W.,S.K.& but morphologically distinct. The males of R.W.Gess) 2 females (at water); 85 km E this form show the characters noted by of St George (28.03S, 148.30E), 27.X.1993 Snelling for his male. (F.W.,S.K.& R.W.Gess) 8 females, 1 male Male. — (Figs 3 and 4). Colour pattern (in yellow flowers of Goodenia pinnatifida very similar to that of female but differing Schldl., Goodeniaceae). Two females and in that the following parts are pale lemon- 1 male in ANIC; 2 females in WAM; rest yellow: small streak on scapes distally, en- of material in AMG. tire clypeus other than for narrow lateral and anterior margins, single large frontal Rekia confluens (Snelling) comb. nov. spot between and above antennal sockets, .1 . . 1 , .. Paravia confluens Snelling, 1986: 14 and Fie. 28, narrow streak marginmg lower orbits, c \ & ' elongate spot in lower half of each ocular p. , • ^ n- mo^ ^ o j t- mm ^ , ^ . . ^ ^ Riekia sp. Snelling, 1986: 6, 8 and Figs 10, 19- sinus (not rising above upper limit of fron- 21 male tal spot and separated from it by about width of antennal socket), uninterrupted Snelling (1986) erroneously described but posteromedially narrowly emarginate confluens as a species of Paragia Shuck- pronotal band connecting humeri. Spot on ard, 1837 despite the fact that, in his own Volume 4, 1995 35 Figs. 1-4. Riekia nocatunga. 1, dorsofrontal view of vertex and dorsal view of anterior third of thorax of female (X17.6); 2, dorsal view of posterior third of thorax of female (X17.6); 3-4, ventral and ventrolateral views of genitalia of male (X57). Figs. 5-8. Riekia confluens. 5, dorsofrontal view of vertex and dorsal view of anterior third of thorax of female (X17.6); 6, dorsal view of posterior third of thorax of female (X17.6); 7-8, ventral and ventrolateral views of genitalia of male (X63); [all lemon-yellow colour form] generic key, the species with its uncon- stricted gastral tergum 2 runs down to Riekia Richards, 1962. Further, with the exception of the number of spine-Hke se- tae at the apex of the mid- and hindtibiae (a specific character) the characters given by him in his diagnosis of the species are all common to both coiiflnciii^ and nocn- tungn and may therefore be considered to be generic characters pertaining to Riekia. Similarly the confluence of the postocular and preoccipital carinae (leading to the name confluens), or alter- natively the absence of the postocular ca- 36 Journal of Hymenoptera Research rina, is a character shared with nocatiin- ga. Conversely, the same characters set confluens (and nocatiinga) generically apart from Paragia. Material collected at a single site near Carnarvon, Western Australia, at distanc- es of 165 km and 260 km in a westerly and west-south-westerly direction respectively from the type localities of confluens (16 km WSW Lyons River Homestead, 24.38S, 115.20E — paratype female, and 36 km ESE Minnie Creek Homestead, 24.00S, 115.42E — holotype female) consists of 53 females and 4 males. The material is di- visible into two very distinct groups, one black and reddish-brown and the other black and lemon-yellow. Whereas intra- group variability is negligible, inter-group differences with respect to colour gener- ally and to colour pattern on the metaso- ma in particular are striking. The distinctness of the two sympatric groups and the notable absence of any in- termediate forms initially led to the belief that two species were represented. How- ever, no morphological characters could be found that supported this view and it is concluded that the apparent difference in overall facies is a product of the differ- ences in colour and colour pattern. The reddish-brown colour form, consistent with the description of Snelling's two fe- males (particularly with that of the para- type), is represented in the present mate- rial by females only, the lemon-yellow col- our form by both sexes. To facilitate intra- specific comparison, descriptions limited to colour pattern are given of both female forms whereas for the purpose of inter- specific comparison with nocatiinga the de- scription of the male is given in greater detail. In passing it is noted that in Snelling's description it is incorrectly stated that the probasitarsus is slightly more than twice wider than long. This is clearly an inad- vertent error as it is in fact slightly less than half as wide as long. Reddish-brown (RB) Colour Form Female. — Black. The following reddish- brown: mandibles other than for teeth and extreme base, upper half of clypeal disc, a variously developed transverse band (seemingly formed of a pair of medially fused spots) between and above antennal sockets and narrowly separated ventrally (where sometimes angularly emarginate) from clypeal marking, variously sized elongate spot in upper half of each ocular sinus (rising to level of anterior ocellus and exceeding upper margin of frontal marking) [median frontal transverse band and lateral spots sometimes broadly fused above], a large oval spot behind eyes dor- sally, entire upper surface of prothorax other than for narrow streak bordering posterior margin, tegulae except for clear testaceous central spot and narrow mar- gin, scutellar disc other than for its ante- rior margin, a large spot on raised part of axillae and a minute spot between it and tegula, a diffuse spot on metanotum me- dially, a large spot on angles of propo- deum, large spot on mesopleuron below tegula, distal end of mid- and hindcoxae, part of foretrochanter and whole of mid- and hindtrochanters, femur, tibia and tar- someres of all legs, gastral tergite 1 other than for anterior declivity, tergite 2 other than for a variously developed anterior black band which may be triangularly produced in the middle and for a pair of widely separated diffuse dark markings on posterior margin, an anteriorly point- ing triangular spot on posterior margin of tergite 3, tergite 4 other than for a narrow anterior black band, tergite 5, diffuse transverse posterior band on gastral ster- nites 2-5 (those of sternites 2 and 3 and sometimes 4 strongly and widely anteri- orly produced in the middle). Lemon-yellow (LY) Colour Form Female. — (Figs 5 and 6). Black. The fol- lowing reddish-brown: tip of mandibles except actual teeth, flagellomeres 1-10 be- Volume 4, 1995 37 neath, trochanters of mid- and hindlegs, eral and anterior margins, single large proximal half to three quarters of outer as- frontal spot between and above antennal pect of femora (where moderately to sockets, lower aspects of mid- and hind- heavily black suffused) and inner and coxae, entire outer aspect of midfemora, lower aspect of same, inner and lower as- entire outer aspect of all tibiae (including pect and distal end of tibiae, all tarsomer- distal ends), narrow longitudinal streak on es, posterolateral portions of dark mark- all basitarsi, transverse posterior band on ings on sternites 2-4 and sternite 6 apical- gastral sternites 2-5 (all with median lobe ly (all black-suffused). The following lem- strongly and widely anteriorly produced, on-yellow: proximal half to two-thirds of together giving the effect of a wide me- mandibles, subapical spots on scapes, dian longitudinal band), clypeus other than for narrow lateral and Antennal scape (with radicle) very anterior margins and in some specimens slightly longer than interantennal dis- arcuate black lines on distal half, a pair of tance; first flagellomere slightly more than irregularly shaped frontal spots between half (0.53) as long as scape (with radicle), and above antennae, a narrow streak on one and three quarter times as long as lower inner orbits flanking clypeus, a wide at its distal end, and twice as long as broad mark in ocular sinus rising to level the second flagellomere. Last three flagel- of anterior ocellus, an oval spot behind lomeres flattened and slightly concave be- eyes dorsally, uninterrupted but postero- neath; ultimate flagellomere narrowing medially narrowly emarginate pronotal apically and distinctly curved to form a band connecting humeri, posterior corner hook. of pronotum, tegulae except for clear tes- Tergite 7 truncate and narrowly trans- taceous central spot and narrow margin, a verse apically. large transverse suboval shield-like spot Genitalia (Figs 7 and 8); similar to those on scutellum (not quite reaching anterior of R. nocatunga but differing in detail par- margin), a small spot on raised part of ax- ticularly with respect to the form of the illae, a large spot on angles of propodeum, inwardly directed lobe of the volsella be- large spot on mesopleuron below tegula, low the digitus. a small spot on mesosternum anterior to Length 7.8-8.3 mm; length of forewing coxal cavities, a small lateral spot on mid- 5.5-5.7 mm; hamuli 12-13. and hindcoxae, distal half of outer aspect Material Examined. — Western Australia: of forefemora, distal spot on outer aspect 8 km NE of Carnarvon (24.51S, 113.45E) of mid- and hindfemora, outer aspect of on road to Bibbawarra Bore, 3-5. x. 1993 all tibiae (except distal ends), wide ante- (F.W.,S.K.& R.W.Gess) 16 RB females, 29 riorly trilobed posterior band on gastral LY females and 4 LY males (in yellow tergites 1-5, tergite 6 (other than for vari- flowers of Lechenaiiltia sp., Goodeniaceae), ously developed posteriorly pointing V- 4 RB females, 1 LY female (at water), 2 RB shaped black mark leaving lateral and me- females, 1 LY female (without biological dian spots or almost eliminating these), data). Two RB females, 2 LY females and transverse posterior band on gastral ster- 1 LY male in both ANIC and WAM; rest nites 2-5 (those of sternites 2 and 3 tri- of material in AMG. lobed with median lobe strongly and Discussion. — R. confliieiis may be distin- widely anteriorly produced). guished from R. nocatunga in both sexes Male. — (Figs 7 and 8). Colour pattern by the colour pattern and the below indi- very similar to that of female but differing cated morphological characters, in that the following parts are pale lemon- In R. confluens the ratio of POL (distance yellow: entire anterior aspect of scapes, between posterior ocelli):OOL (distance entire clypeus other than for narrow lat- between a posterior ocellus and nearest 38 Journal of Hymenoptera Research eye margin) is 1:1.3-1.4, whereas in R. no- the species differ in the male sex in the catunga it is 1:1.0-1.1. relative proportions of the first two flagel- In R. confliiens the pilosity of the pro- lomeres, in the form of gastral tergite 7 notum is only moderately dense, the in- and in slight differences in the genitalia, dividual hairs being slender and only Present data would indicate that the slightly curved apically whereas in R. no- distributions of the species are widely sep- catunga the pilosity is much denser and arated, R. nocatungci occurring in New the individual hairs are more robust and South Wales and Queensland and R. con- markedly and evenly curved. fluens in Western Australia. In R. confluens the scutellum (Fig. 6) has its disc posteriorly widely and evenly to Rolandia angulata (Richards) bluntly rounded and abruptly declivitous ^.^^^.^ ^,^^^^^^^^ j^.^j^^^^^^ ^^^^. ^q^_^q2, female, to the metanotum which it slightly over- ^^,^„^^,.^ ^^^^^^^^^ (Richards), Snelling, 1986: 8. hangs; the posterolateral free edge of each lateral wing is almost straight. The meta- This species was described from Cun- notum in its median third is not depressed namulla [28.04S, 145. 40E], Queensland middorsally and is therefore of greatest (holotype female and 1 female paratype) depth in the middle; it is vertical, not vis- and from 90 km W of Cobar, Barnato ible from above. In R. nocatunga the scu- Tanks [31.38S, 144.59E] (about 400 km tellum (Fig. 2) has its disc posteriorly nar- south of Cunnamulla), New South Wales rowly rounded and gradually descendant (12 female paratypes). Snelling (1986) re- to metanotum which it does not overhang; corded an additional 8 females from Cun- the posterolateral free edge of each lateral namulla. Hitherto the male appears to wing is widely and shallowly concave, have been unknown. One of the present The metanotum in its median third is females was compared with the holotype somewhat depressed middorsally and is in the Australian National Collection of therefore of even depth; it is sloping, vis- Insects, Canberra, and found to be identi- ible from above. cal. In R. confluens gastral tergite 1 is two Male. — (Figs 9 and 10). Colour pattern and two-fifths to two and a half times very similar to that of female, differing wider than long whereas in R. nocatunga chiefly in that the yellow markings are it is two and a quarter times wider than lighter in hue and brighter and contrast long. more strongly with the light reddish In R. confluens the number of spine-like brown areas. The bright yellow mandi- setae on the apex of the mesotibia is 1 bles, clypeus and transverse lower frontal whereas in R. nocatunga it is 2-4, most marking, and the distinct median and lat- commonly 3; the number of such setae on eral spots on the gaster facilitate recogni- the apex of the metatibia is 1-2, most com- tion of the male even in the field, monly 1, and 2-3 respectively. Antennal scape (with radicle) shorter The species differ in the female sex in than interantennal distance; first flagello- the form of the pronotum as seen from mere three times as long as broad at its above. In R. confluens (Fig. 5) the anterior distal end, about two-thirds as long as and lateral margins describe a semicircle scape (with radicle) [three-quarters as long whereas in R. nocatunga (Fig. 1) the as scape without radicle], and a little more "shoulders" are anterolaterally produced than twice as long as second flagellomere; so that the anterior and lateral margins do flagellomeres 3-9 subequal in length and not describe a semicircle but the anterior with the exception of 3 wider than long; margin appears subtransverse. 3-7 increasing in width; 7 and 8 of same As may be seen from the descriptions, width; 9 as wide as 6; 10 slightly longer Volume 4, 1995 39 Figs. 9-10. Rolandia angulata. 9, posterior view of right fore leg of male (X44.5); 10, ventral view of genitalia of male (X44.5). than preceding flagellomeres, narrowing and apically rounded; distal flagellomeres not modifled beneath. Foreleg (Fig. 9); process of foretrochan- ter large, compressed, broadly digitiform in posterior view and slightly forwardly curved, its apex attaining level of angulate base of forefemur; forefemur in posterior view broad and ventrally strongly angu- late at base and widely but shallowly in- curved from there to about midlength, with ventral margin angulate along its length but becoming lamellate at basal an- gle, with ventral surface basally (i.e. ad- jacent to lamellate ventral angle) excavat- ed. Genitalia (Fig. 10); parameral spines apically sharply pointed and ventrally re- curved; ventral process of each paramere tapering towards apex in lateral view but in ventral view more or less of even width over most of its length and apically nar- rowly rounded; digitus and inwardly di- rected lobe of volsella as figured. Length 8.7-9.2 mm; length of forewing 6.3-6.5 mm, hamuli 16-18. Material Examined. — Queensland: 85 km E of St George (28.03S, 148.30E), 27.X.1993 (F.W.,S.K.& R.W.Gess) 6 females (in yel- low flowers of Goodenia cycloptera R.Br.in C.Sturt, Goodeniaceae), 7 females (in yel- low flowers of Goodenia pinnatifida Schldl., Goodeniaceae); 27 km W of St George (28.03S, 148.30E), 28.X.1993 (F.W.,S.K.& R.W.Gess) 31 females, 10 males (in yellow flowers of Goodenia cycloptera R.Br.in C.Sturt, Goodeniaceae), 1 female (in yel- low flowers of Goodenia pinnatifida Schldl., Goodeniaceae), 2 females (nesting in sandy soil), 2 females, 3 males (without bi- ological data); 80 km E of Cunnamulla (28.04S, 145.40E), 29.X.1993 (F.W.,S.K.& R.W.Gess) 49 females, 31 males (in yellow flowers of Goodenia pinnatifida Schldl., Goodeniaceae), 4 females (nesting in sandy soil), 9 females, 2 males (without bi- ological data). Three females and 3 males in ANIC; 3 females and 3 males in WAM; rest of material in AMG. Discussion. — Snelling (1986) was correct in his assumption that the males of R. an- gulata would differ from those of the very similar R. houstoni in the same characters as do the females. Of particular note is the scutellum which in both sexes of R. an- gulata slopes evenly to the metanotum and 40 Journal of Hymenoptera Research lacks a definite posterior face. The males run down satisfactorily in Snelling's key, allowance being made for the longer than broad second flagellomere. ACKNOWLEDGEMENTS The author wishes to thank Ian Naumann and Jo Cardale of the Australian National Insect Collection, Canberra for hospitality in their department and for facilitating the comparison of some of the present specimens with type material in their care, Kirsten Cowley of the Australian National Herbarium, Can- berra for identifying the forage plants, Suzanne Lewis of the Natural History Museum, London for lending material under her care, and Robin Cross of the Elec- tron Microscopy Unit, Rhodes University, Grahams- town for producing the scanning electron micro- graphs reproduced as Figs 1-10. An earlier version of the manuscript was reviewed by Karl V. Krombein of the Smithsonian Institution, Washington and James M. Carpenter of the American Museum of Nat- ural History, New York. The latter, in particular, is thanked for his constructive critique and for drawing attention to a potential synonymy arising from the fact that R. confluens had been described by Roy Snell- ing as a species of Paragia, a genus to which the wasp in question quite clearly does not belong. Gratitude to the South African Foundation for Research Devel- opment is expressed for a rolling support grant which made the field work in Australia possible. LITERATURE CITED Meade-Waldo, G. 1911. Description of new species of African Diploptera in the collection of the Brit- ish Museum. Annals and Magazine of Natural His- tory (8)8: 448-457. Richards, O.W. 1962. A revisional study of the Masarid ivasps (Hymenoptera, Vespoidea). London: British Museum (natural History). Richards, O.W. 1968. New records and new species of Australian Masaridae (Hymenoptera: Vespo- idea). Journal of the Australian Entomological Soci- ety 7: 101-104." Shuckard, W.E. 1837. Descriptions of new exotic acu- leate Hymenoptera. Transactions of the Entomolog- ical Society of London 2: 68-82. Snelling, R.R. 1986. The taxonomy and nomenclature of some Australian paragiine wasps (Hymenop- tera: Masaridae). Contributions in Science 378: 1- 19. van der Vecht, J. and Carpenter, J.M. 1990. A cata- logue of the genera of the Vespidae (Hymenop- tera). Zoologische Verhandelingen 26: 1-62. J. HYM. RES. Vol. 4, 1995, pp. 41-63 The Last Twenty Years of Parasitic Hymenoptera Karyology: An Update and Phylogenetic Implications Vladimir E. Gokhman and Donald L. J. Quicke (VEG) Botanical Garden, Moscow State University, Moscow 119899, Russia; (DLJQ) Department of Biology, Imperial College at Silwood Park, Ascot, Berks SL5 7PY, U.K. Abstract. — A survey of karyological data in the parasitic Hymenoptera shows a considerable uniformity in chromosome numbers within higher taxonomic groupings, with most chalcidoids having n = 5-6 and most cynipoids, ichneumonoids and those very few species of Diapriidae and Scelionidae examined to date having n = 10-12. The Encyrtidae and Eurytomidae differ from other chalcidoids examined to date in having higher n values (8-11). The braconid subfamilies Aphidiinae {ii - 4-7), Charmontiinae (n = 5) and Exothecinae (» = 6) are shown to have lower values than do other braconids. New data are presented for 32 species, including the first records of chromosome numbers for the Eupelmidae, Bethylidae, Diapriidae, Ormyridae, the ichneumonid subfamilies Banchinae, Pimplinae, and Tryphoninae, and the braconid subfamilies Aphidiinae, Charmontinae, Doryctinae, Exothecinae, Macrocentrinae, Meteorinae, and Miracinae. Chromo- some number data are discussed in relation to current views on Hymenoptera phylogeny. INTRODUCTION There has long been a tendency for tax- onomists to ignore many potential sources of systematic evidence in favour of more traditional morphological, and nowadays molecular, data. Whilst this lack has not been ubiquitous amongst taxonomists, where it has occurred it may have been due to a number of reasons including lack of understanding of novel characters, lack of facilities for investigating them and lack of availability of information about such characters. In this respect, it is not sur- prising that chromosome number and oth- er karyological features have played little role in considerations of hymenopteran phylogeny, and this is especially true for the parasitic families. Reasons for this are manifold. Hymenopteran chromosomes have traditionally been studied in imma- ture stages in which dividing cells are common, but larvae of parasitic wasps are perhaps especially difficult to identify and obtain except when species are in culture. Recent developments in the study of hy- menopteran chromosomes (Imai et al. 1988, Baldanza et al. 1991b), have opened the possibility of greatly expanding our knowledge of parasitic wasp karyology. Further, it is now possible to obtain good results working with adult tissues such as ovaries (Gokhman 1985, 1990a), and such protocols have allowed, for the first time, an examination of chromosome number and structure in natural populations. Karyological studies have, nevertheless, played an important role in the systemat- ics and species level taxonomy of various groups of the non-parasitic Hymenoptera including, sawflies, ants, bees and social wasps (principally Polistinae) and these have been reviewed in detail elsewhere (see for example, Kerr 1972, Kerr and da Silveira 1972, Naito 1982, Imai et al. 1977, Moritz 1986, Pompolo and Takahashi 1987, 1990, Hoshiba, Matsuura and Imai 1989, Costa et al. 1993). Wide variation in n is well known among the ants (Formic- idae) and ranges from 1 to 47 (Imai and Taylor 1989, Imai et al. 1990), but is modal at 11 (Imai et al. 1988). Indeed, the greater 42 Journal of Hymenoptera Research part of this range can even be found with- in the single ant genus Myrmecia. How- ever, within the parasitic Hymenoptera there appears to be be rather more consis- tency, and for most famiUes the observed range of variation is far narrower. Outside of the Hymenoptera, both chro- mosome numbers, structure and size have all been found useful in phylogenetic re- construction, usually but not exclusively in combination with other morphochar- acters, and this has been particularly true of plant systematics. For example, George & Geethamma (1992) have recently pro- posed a phylogeny of jasmines based on chromosome numbers and assumed poly- ploidy events. Among the insects, chro- mosome numbers have been used exten- sively in the systematic treatment and phylogeny of the Hemiptera (Blackman 1980, Emeljanov & Kirillova 1989, 1991). Kuznetsova (1985), for example, has con- cluded that the homopteran subfamily Or- geriinae is monophyletic based on an au- tosomal fusion giving 2m =27 (in males) compared with the plesiomorphic value of 2n = 29 (in males), and similarly that the tribe Almanini of the Orgeriinae is mono- phyletic on the basis of an autosome-sex chromosome fusion (giving 2n=26). In this paper we review the current state of knowledge of the karyology of parasitic wasps and, in addition, provide new data for approximately 32 taxa including the first records of chromosome numbers for the Eupelmidae, Bethylidae, Diapriidae, Ormyridae, the ichneumonid subfamilies Banchinae, Pimplinae, and Tryphoninae, and the braconid subfamilies Aphidiinae, Charmontinae, Doryctinae, Exothecinae, Macrocentrinae, Meteorinae, and Miraci- nae. MATERIALS AND METHODS New karyotypic data were obtained us- ing ovarian tissues of adult wasps which were either collected from the wild at Sil- wood Park, Berkshire, U.K., during early September 1995, or obtained from labora- tory cultures. Chromosome preparations were obtained according to the schedule described in Appendix I. Voucher speci- mens for Ichneumoninae are deposited in the collection of Moscow State University, those for other taxa are in the Natural His- tory Museum, London. RESULTS AND DISCUSSION The last review of chromosome number in the parasitic Hymenoptera was by Cro- zier (1975) at which time values were known for only about twenty species (with data published before 1930 not in- cluded for the reason of insufficient reli- ability); his data are summarized in Figure 1. Since then, chromosomes of nearly 140 additional species have been studied (Gokhman 1994), and for this review we have investigated 32 more in order to help fill in a number of gaps and to confirm some previous findings. These data are presented in Table 1 and summarized in Figure 2. The great majority of published chromosome numbers come from studies on the ichneumonid subfamily Ichneu- moninae by Gokhman (1985, 1987, 1989, 1990a, 1990b, 1991a, 1991b, 1993a), though they also include a substantial body of ev- idence especially relating to various fam- ilies of Chalcidoidea. Within the better studied parasitoid families (i.e. Ichneu- monidae, Cynipidae, Torymidae) chro- mosome numbers were found to be rela- tively stable, and were generally uniform at the generic level. This contrasts mark- edly with observations for many aculeate Hymenoptera which often show striking chromosomal variability even within gen- era (Imai and Taylor 1989, Imai et al. 1990). The apparent bimodality in haploid number of parasitic Hymenoptera appar- ent in Crozier's small sample (Fig. 1) now appears to be well founded (Fig. 2). How- ever, the exact modal numbers are slightly different (n = 5 and 11 compared with n = 5 and 10 of Crozier), probably due to the strong bias in favour of the ichneumo- Volume 4, 1995 14 -T- 12 - 10 43 6 - 2 -- + + + + 4- 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Haploid chromosome number Fig. 1. Histogram of haploid chromosome numbers for parasitic Hymenoptera, data modified after Crozier (1975) to take into account subsequent taxonomic changes. nid subfamily Ichneumoninae in the pres- ent enlarged data set. As will be apparent from the figures, the peak in haploid val- ues around 5 largely comprises members of the Chalcidoidea and there is only slight overlap between these and the other peak centred around 11. The same is true, though less pronounced, when the data are plotted for genera rather than species (Figure 3: each n value in a genus being represented only once in the histogram). Members of three of the chalcidoid fami- lies investigated to date, the Eurytomidae (Eiiri/toma), Encyrtidae (6 species in three genera) and Aphelinidae (one species, Pteroptrix (= Archenomiis) orientalis Silves- tri, out of 5 investigated; see below) differ from other chalcidoids in having haploid numbers ranging between 8 and 11, whilst members of the braconid subfamilies Aphidiinae {Aphidius, Diaretiella, Ephcdriis, Praon), Charmontinae [Charmon) and Ex- othecinae [Rhysipolis) are atypical in hav- ing haploid numbers of 4 to 7, 6 and 5 respectively, compared with other ichneu- monoids whose haploid n values range from 8 to 17. GENOMIC SIZE AND ITS IMPLICATIONS Rasch et nl. (1975, 1977) calculated the haploid genomic size of the braconid wasps, Habrobracon juglandis and H. seri- nopae both to be 0.154).16 x 10 ^^g dNA and that of the pteromalid, Nnsonin (as Mormoniclla) vitripennis to be 0.33-0.34 X 10 '-^g. These values translate (using an average molecular weight of a base pair as 660) to base pair numbers of approxi- mately 1.4 X 10** and 3.0 X 10^ base pairs for the ichneumonoid and the chalcidoid respectively. More recently. Bigot et al. (1991) using DNA reassociation kinetics calculated the haploid genomic sizes of the ichneumonid Diadroniiis piilchclliis and the chalcidoid, Eiipclmus viiillcti, as 1-2 X 10'^ base pairs and 10 X 10"^ base pairs re- spectively. Unfortunately, no data for the numbers of chromosomes are available for Eupelmus vuiUeti but we have investigated 44 Journal of Hymenoptera Research Table 1. Chromosome number in parasitic wasps Taxon 2/!* Reterence(h)t Diaprioidea Diapriidae Beli/ta depressa Thomson Cinetus lanceolatus Thomson Scelionoidea Scelionidae Telejtomus fariai Lima Chalcidoidea Aphelinidae Apheliniis mali Haldeman Aphytis mytilaspidus (Le Baron) Archenonnis orientalis Silvestri Encarsia berlesei Howard Encarsia pergandiella Howard Chalcididae Bradiyineria intermedia Nees B. lasus Burks B. ovata Say Dirhiniis himalayanus Westwood Encyrtidae Ageniaspis fiiscicoUis Dalman Copidosoma buyssoni Mayr C. gelechiae Howard C. gelechiae Howard C. truncatelliim (Dalman) (= Ifloridanum Ashmead) C. floridanum Ashmead C. floridanum Ashmead Apoanagyrus lopezi (DeSantis) Eulophidae Cirrospilus diallus Walker Colpoclypeiis florus Walker Melittobia chalybii Ashmead Tetrastichus gigas Burks T. megachilidis Burks Eupelmidae Anastatus catalonicus Bolivar Macroneura vesicularis (Retzius) Eurytomidae Eudecatoma bigidtata (Swederus) Eurytoma californica Ashmead Leucospidae Leucospis affinis Say Ormyridae Ormynis sp. Pteromalidae Anisoptcromalus calandrae (Howard) Coelopisthia extenta (Walker) Dibrachys sp. Lariophagus distinguendus Foerster Muscidifurax zaraptor Legner Nasonia vitripennis (Walker) 8 10 10 6 6 5 6 6 5 5 9 10 ?6 7 5 5 5 5 5 16 present papertt 20 present paper 20 Dreyfus & Breuer 1944 5 10 Viggiani 1967 5 10 Rossler & De Bach 1973 11 22 Baldanza et al. 1991a 5 10 Baldanza et al. 1991b 6 12 Hunter et al. 1993 3 6 Hung 1986 5 10 Hung 1986 5 10 Hung 1986 5 10 Amalin et al. 1988 c. 10 C.20 Silvestri 1908; Martin 1914 12 24 Silvestri 1914 11 22 Hegner 1915 10 20 Patterson 1921 10 20 Hunter & Bartlett 1975 8 16 Leiby 1922; Patterson 1917, 1921 Patterson & Porter 1917 11 22 Strand & Ode 1990 10 20 Dijken 1991 12 present paper 12 Dijkstra 1986 10 Schmieder 1938 12 Goodpasture 1974 12 Goodpasture 1974 10 present paper 10 present paper 18 present paper 20 Goodpashire 1974 12 Goodpasture 1974 ?12 present paper 14 present paper 10 present paper 10 Goodpashjre 1974 10 Gershenzon 1968 10 Goodpashare 1974 10 Gershenzon 1946, 1968; Pennypacker 1958; Whiting 1960, 1968; Wahr- man & Zhu 1993 Volume 4, 1995 45 Table 1. Continued Taxon Reference(s)t Nasonia vitripennis (Walker) 5 + 0-1 IB 10 Nasonia vitripiennis (Walker) 6 12 Ptewmalus puparum L. 5 10 P. venustus Walker 5 10 Torymidae Monodontoinerus dementi Grissell 6 12 M. moniivagus Ashmead 6 J2 M. obscurus (Westwood) 4 8 M. obscurus (Westwood) 6 12 M. saltuosus Grissel 5 10 Tory mils baccharidis Huber 6 12 T. californicus (Ashmead) 6 12 T. capillaceus Huber 6 12 T. koebelci Huber 5 10 T. occidentalis Huber 6 12 T. tubicola Osten-Sacken 6 12 T. umbilicatus Gahan 5 10 T. vesiculi Moser 6 12 T. warreni (Cockerell) 6 12 Trichogrammatidae Trichogmmma cliilonis Ischii 5 10 T. dcicvi Pinto & Oatman 5 10 T. dendrolimi Matsumura 5 10 T. evanescens Westwood 5 10 T. nubialale Ertle & Davis 5 10 T. pretiosum Riley 5 10 T. spp. (7 strains) Cynipoidea Cynipidae Andricus curvator Hartig A. feciindator Hartig A. kollari Hartig A. quercuscalicis Burgsdorf Aulacidea hiemcii Bouche Biorrhiza pallida Olivier Callirhi/tis palmiformis Ashmead Cynips divisa Hartig Diastrophus nebulosus Osten-Sacken Diplolepis elganteriae Hartig D. nervosum Curtis D. rosae L. D. rosae L. D. spinosissimae Girault Dryocosmus kuripliilus Yasumatsu Dryophanta erinacea Mayr Neurotcrus laeviusculus Schenck N. numismalis Fourcroy N. quercusbaccarum L. Trigo)iaspis megaptera Panzer Xestopluvics potcntillac Retzius 10 Nur et al. 1988; Werren 1991 Goodpasture 1974 Guhl & Dozortseva 1934, Dozortseva 1936 McDonald & Krunic 1971 Grissell 1973b; Goodpasture 1975a Goodpasture 1975a Goodpasture 1975a McDonald & Krunic 1971 Grissell 1973b; Goodpasture 1975a Goodpasture & Grissell 1975 Goodpasture & Grissell 1975 Goodpasture & Grissell 1975 Goodpasture & Grissell 1975 Goodpasture & Grissell 1975 Goodpasture & Grissell 1975 Goodpasture & Grissell 1975 Goodpasture & Grissell 1975 Goodpasture & Grissell 1975 Hung 1982 Stouthamer & Kazmer 1994 Liu & Xiong 1988 Hung 1982 Hung 1982 Hung 1982; Stouthamer & Kazmer 1994 Fukada & Takemura 1943 (cited by Hung 1982) 10 20 Dodds 1938; Sanderson 1988 10 20 Dodds 1938 10 20 Hogben 1920; Sanderson 1988 10 ?20 Sanderson 1988 10 20 Dodds 1938 10 20 Dodds 1938 w 20 Goodpasture 1979b 10 20 Sanderson 1988 10 20 Hegner 1915' 9 27(3n) Sanderson 1988 9 18 Sanderson 1988 9 18 Henking 1892; Stille & Davring 1980; Sanderson 1988 ?12 ?24 Schleip 1909 9 18 Sanderson 1988 10 20 Abe 1994 c. 12 24 Wieman 1915 10 20 Sanderson 1988 10 20 Dodds 1938 10 20 Doncaster 1910, 1911, 1916; Dodds 1938 10 20 Dodds 1938 10 20 Dodds 1938 46 Journal of Hymenoptera Research Table 1. Continued Taxon ;;' 2n* Reference(s)t Eucoilidae Pseudeucoila bochei Weld 10 20 Jungen cited in Crozier 1975 Ichneumonoidea Braconidae Aphidius rhopalosiphi De Stefani Perez 7 14 present paper Biosteres carbonarius (Nees) U 28 present paper Charmon cruentatiis Haliday 5 10 present paper Cotesia gloineratiis (L.) (as Apanteles) 12 24 Hegner 1915i Dacnusa sp. 17 34 present paper Diaeretiella rapae Mcintosh 6 12 present paper Ephedrus sp. 17 ?14 present paper Habwbracon hebetor Say 10 20 Torvik-Greb 1935 H. jiiglandis Ashmead^ 10 c. 20 Torvik-Greb 1935; Speicher 1937; Rasch et al. 1977 H. jiiglandis Ashmead^ 11 22 Whiting 1927 H. pectinophorae Watanabe 10 20 Inaba cited in Makino 1959 H. serinopae Ramakrishna 10 20 Rasch et al. 1977 Heterospihis prosopidis Viereck 17 34 present paper Macrocentrus thomcicus (Nees) 13 26 present paper Meteorus versicolor Wesmael 8 16 present paper Meteorus gyrator Thunberg 10 20 present paper Meteorus pallipes Wesmael 10 20 present paper Mirax sp. 10 20 present paper Phaenocarpa persimilis Papp 17 34 Prince & Stace (cited by Crozier 1977) Praon abjectum Haliday 4 8 present paper Rln/sipolis decorator (Haliday) 6 12 present paper Ichneumonidae Aethercerus discolor Wesmael 11 22 Gokhman 1985 Ae. dispar Wesmael 12 24 Gokhman 1991 Ae. nitidus Wesmael 11 22 Gokhman 1990a Ae. raniui Gokhman 11 22 Gokhman 1991 Agrothereutes extrematus (Cresson) 10 20 Koonz 1939 Aophis pulchricornis (Gravenhorst) 13 26 Gokhman 1990b Aptesis puucticollis (Thomson) 8 16 Gokhman 1990a Baeosemus dent if er Gokhman 11 22 Gokhman unpubl. obs. Baranisobas ridibundus (Gravenhorst) 11 22 Gokhman 1990b Chasmias niotatorius (Fabricius) 17 34 Gokhman 1985 Coelichneiimon cyaniventris (Gravenhorst) 13 26 Gokhman 1990a C. sugillatorius (L.) 13 26 Gokhman 1990a Colpognathus celerator (Gravenhorst) 11 22 Gokhman unpubl. obs. Cratichneumon viator (Scopoli) 14 28 Gokhman 1985 Diadroinus prosopius Holmgren 11 22 Gokhman 1990a D. pulchellus Wesmael 11 22 Hedderwick et al. 1985 D. subtilicornis (Gravenhorst) 11 22 Gokhman 1990b D. troglodytes (Gravenhorst) 11 22 Gokhman 1990a D. varicolor Wesmael 11 22 Gokhman pers. obs. Dicaelotus sp. nr. parvulus (Gravenhorst) 11 22 Gokhman 1990a D. piimilis (Gravenhorst) 11 22 present paper Diphyus latebricola (Wesmael) 12 24 Gokhman 1990a D. raptorius (L.) 12 24 Gokhman 1990a Dirophanes callopus (Wesmael) 9 18 Gokhman 1987 D. fidvitarsis (Wesmael) 10 20 Gokhman 1990a,b D. invisor (Thunberg) 10 20 Gokhman 1987 Dyspetes arrogator Heinrich 10 20 present paper Volume 4, 1995 47 Table 1. Continued Taxon Ret'erence(s)+ Ephialtes manifestator (L.) 15 30 Eurylahiis tonms Wesmael 10 20 Gelis sp. 13 26 Gli/pta lapponica Holmgren 9 18 Heterischnus nigricollis (Wesmael) 11 22 H. tnincator (Fabricius) 11 22 Homothenis locutor (Thunberg) 11 22 Ichneimion albiger Wesmael 12 24 /. auip}iibolus Kriechbaumer 12 24 /. bucciilcntus Wesmael 12 24 /. confiisor Gravenhorst 12 24 /. crassifemiir Thomson 12 24 /. croceipes Wesmael 12 24 /. extensorius L. 12,13 24, 26^ /. cf. extensorius L. 11 22 /. formosus Gravenhorst 11 22 /. gracilentus Wesmael 12 24 (25) /. gracilicomis Wesmael 11 22 /. ingratiis Hellen 12 24 /. inquinatus Wesmael 13 26 /. iiisidiosus Wesmael 12 24 /. liigens Gravenhorst 12 24 /. melanotis Holmgren 12 24 /. mmutorius Desvignes 12 24 /. molitorius L. 12 24 /. nereni Thomson 11 22 /. sarcitorius L. 11 22 /. stramentarius Gravenhorst 12 24 I. submarginatus Gravenhorst 10 20 /. subquadmtiis Thomson 12 24 /. suspiciosus Wesmael 12 24 (26) /. validicornis Holmgren 22 24 Lissonota sp. 11 22 Mastnis smithii (Packard) 13 26 Oroiwtus binotatus (Gravenhorst) 11 22 Orthocentrus sp. 14 28 Orthopelma mediator (Thunberg) 11 22 Patrodoides chah/beatiis (Gravenhorst) 8 16 Phaeogencs melaiiogonos (Gmelin) 11 22 Ph. nigridens Wesmael 11 22 Ph. semivulpimis (Gravenhorst) 9 18 Ph. spiuiger (Gravenhorst) 11 22 Polysphmcta tiiberosa Gravenhorst 9 18 Pseudoambhjteles homocerus (Wesmael) 9 18 Stenichneumon cidpator (Schrank) 14 28 Sycaonia sicaria (Gravenhorst) 11 22 Syspasis alboguttata (Gravenhorst) 11 22 S. scutellator (Gravenhorst) 11 22 Tycherus australogemimis Gokhman 11 22 T. bellicontis (Wesmael) 10, U 20(21) T. dilleri Ranin n 22 T. fuscicornis (Wesmael) n 22 T. ischiomelinus (Gravenhorst) 9 18 present paper Gokhman 1987 present paper present paper Gokhman 1990b Gokhman unpubl. obs. Gokhman 1990b Gokhman 1990a Gokhman 1990b Gokhman 1993a Gokhman 1985 Gokhman 1985 Gokhman 1990b Gokhman 1993a Gokhman 1990a Gokhman 1990a Gokhman 1993a Gokhman 1990a Gokhman 1990b Gokhman 1993a Gokhman 1990a Gokhman unpubl. obs. Gokhman 1990a Gokhman 1987 Gokhman 1990a Gokhman 1990a Gokhman 1990a Gokhman 1987 Gokhman 1990a Gokhman 1990b Gokhman 1993a Gokhman unpubl. obs. present paper Koonz 1936 Gokhman 1987 Gokhman 1990a Hogben 1920 Gokhman 1993a Gokhman 1990b Gokhman 1990b Gokhman unpubl. obs. Gokhman 1990a present paper Gokhman unpubl. obs. Gokham 1985 Gokham 19990b Gokhman 1985 Gokhman unpubl. obs. Gokhman 1991 Gokhman 1989 Gokhman 1989 Gokhman 1990b Gokhman 1991 48 Journal of Hymenoptera Research Table 1. Continued Taxon n* 2n* Reference(s)+ T. ophthalmicus (Wesmael) 22 Gokhman 1990a T. osculator (Thunberg) 22 Gokhman 1989 T. suspicax (Wesmael) 22 Gokhman 1987 Ventiiria canescens (Gravenhorst) 22 Speicher 1937 Virgichneumon digrammus (Gravenhorst) 34 Gokhman 1990a V. f annus (Gravenhorst) 22 Gokhman 1990a Vulgichneumon snturatorius (L.) 9 18 Gokhman 1987 Chrysidoidea Bethylidae Epyris niger Westwood 14 28 present paper Laelius utilis Cockerell 10 20 present paper * In papers which only quote n or 2)i, the other value has been surmised and is given in italics, t Data appearing in works before 1930 should be considered with great caution as most resulted from his- tological rather than cytological protocols, involving sectioned material rather than squash preparations and also often involving fixation techniques not well suited for the study of chromosomes, though some of these earlier findings are clearly correct including the oldest one (Henking 1892). tt For some new data we were not able to obtain an unambiguous chromosome number but our best ap- proximation is presented (data indicated in table with a "?") as in some cases these still provide potentially valuable information. ' Hegner (1915) did not provide a definitive statement on chromosome number and the data here come from his rather stylized figures; such data need therefore to be considered with extreme caution. ^ Some workers consider H. juglandis to be a junior synonym of H. Iwbetor, however, this is not yet absolutely confirmed and therefore we prefer to keep these records separate. ^ A single, probably aneuploid, female specimen with 25 chromosomes has also been found. 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Haploid chromosome number Fig. 2. Histogram of haploid chromosome numbers of parasitic Hymenoptera, data from Table 1 based on n values of each species. Volume 4, 1995 30 T 25 - 49 20 - 15 - 10 7 8 9 10 11 12 13 Haploid chromosome number 14 15 16 17 Fig. 3. Histogram of haploid ciiromosome numbers of parasitic Hymenoptera, data from Table 1, each n value occurring in a genus being represented only once. two other eupelmid species in the genera Macroneura and Anastatiis, which were both found to have haploid numbers of 5 in common with Nasonia. Although the re- sults of Rasch et al. (1975, 1977) and of Big- ot et al. (1991) don't quite agree in the or- der of magnitude of base pairs they esti- mate, probably due to differences in pro- cedure, they did both show the chalcidoid to have a larger genome than the ichneu- monid. Thus, if these values are roughly representative of other members of their superfamilies, then the chromosomes of most chalcidoids would be expected to have between four and twenty times as much DNA on average than chromosomes of ichneumonids. CHROMOSOME NUMBER IN RELATION TO HYMENOPTERA PHYLOGENY In karyological studies it is common practice to interpret modal chromosome numbers as representing the initial (ances- tral) number, though to many cladists this would be interpreted as the application of the much decried commonality principle. The 'common equals primitive' associa- tion is of course probabilistic rather than deterministic (Watrous and Wheeler 1981, Frohlich 1987, Quicke 1993). Some further insight into whether modal chromosome number is likely to reflect the ancestral number can come from the comparison of modal and median numbers. If these numbers coincide, and the whole distri- bution may be approximated to a normal one, then the data provide no evidence of a directed change, though it must be re- membered that there is always the possi- bility that an evolutionary change in chro- mosome number early in the evolution of the group could lead to the same distri- bution. In the Ichneumonoidea and the Cynipoidea both the median and modal chromosome numbers are the same 11 and 10 respectively; in contrast, for example, ants (Formicidae) have a modal number of 11 but the median is 15. However, cladis- tic analyses based on independent char- 50 Journal of Hymenoptera Research acter systems, via outgroup comparisons, provide the most reliable means of deter- mining plesiomorphic chromosome num- bers and, where possible, this is the ra- tional that we have employed. According to currently accepted views of Hymenoptera phylogeny, the sawflies ('Symphyta') form a paraphyletic group with respect to the Apocrita with the latter being the sister group of the Orussidae, and the Apocrita + Orussidae in turn most probably being the sister group of the Xiphydriidae (Konigsmann 1977, Rasnit- syn 1980, 1988, Gibson 1985). Unfortunate- ly, chromosome numbers are not known either for the Orussidae or for the Xiphy- driidae, although they are known for members of two other sawfly families with claims for a close relationship with the Apocrita, viz the Siricidae and Cephi- dae (Konigsmann 1977, Basibuyuk & Quicke 1994, 1995). In the Cephidae n ranges from 9 to 22-26 (Mackay 1955, Cro- zier and Taschenberg 1972), whilst in the Siricidae, according to Sanderson (1932, 1970), the haploid number varies between 8 and 18. As regards other, less derived sawflies, haploid chromosome number ranges from 5 to 22 in the Tenthredino- idea, with three quarters of species having an n value ranging between 7 and 10 (Nai- to 1982). Taking the Siricidae and Cephi- dae as the two sawfly families closest to the ancestral lineage of the Apocrita (i.e. putative sister groups of the Apocrita + Orussidae) for which chromosome num- bers are available, it seems reasonable to conclude that the plesiomorphic haploid chromosome number in the latter was at least 8 and possibly rather higher. Within the Apocrita there is a picture emerging from independent investiga- tions of phylogenetic relationships (Ras- nitsyn 1988, Johnson 1988, Gibson 1985, Mason 1983, Quicke et al. 1993, 1994, Her- aty et al. 1994) that the group divided rel- atively early in its history into a lineage giving rise to the Ichneumono- idea+Aculeata and a second comprising the bulk of the taxa currently regarded as 'Microhymenoptera' including Chalcidoi- dea, Cynipoidea, Scelionoidea, Diaprioi- dea and Proctotrupoidea s.s. (Fig. 4). Our data show that the modal n value in the Ichneumonoidea, the probable sister group of the Aculeata, is 11. Further, our limited data for the less derived aculeates of the family Bethylidae {Epyris and Lae- liiis; Fig. 10), whilst demonstrating some degree of variation in haploid number be- tween 10 and 14, when considered togeth- er with data for other aculeates suggest the ancestral aculeate may have had a haploid number around 11, as was also concluded by Hoshiba, Matsuura and Imai (1989). Similarly, available values for three other parasitoid superfamilies, the Diaprioidea (Fig. 5), Scelionoidea and Cy- nipoidea, are similar. According to Ras- nitsyn's (1988) phylogenetic hypothesis (see Fig. 4), the Scelionoidea are putatively the sister group of the Chalcidoidea s.l., and thus taking the former as the out- group, the plesiomorphic haploid chro- mosome number for the Chalcidoidea may be postulated as being approximately 10. Therefore the data collectively support the hypothesis that the small values of n (from 3 to 7) found in the majority of Chal- cidoidea are likely to be apomorphous. Unfortunately, there are no well founded views of relationships within the Chalci- doidea (Trjapitzin 1978, LaSalle 1987, Bou- cek 1988a, Woolley 1988, Noyes 1990, Gib- son 1990), largely perhaps because of the considerable plasticity in adult morphol- ogy displayed by many of the families, which may result because of the undoubt- edly polyphyletic natures of some family level taxa. Trjapitzin (1978) made few pro- posals about higher level relationships, and only suggested two possible group- ings, his 'pteromaloid' group comprising Pteromalidae, Tanaostigmatidae, Eupel- midae and Encyrtidae, and a 'tetracam- poid' group comprising Tetracampidae, Eulophidae, Elasmidae and Aphelinidae. A relationship between the Aphelinidae Volume 4, 1995 51 Chalckjidae Apheiinidae Ormyridae Pteromalidae Torymidae Trichogrammatidae Encyrtidae Cynipoidea EvantOKJea CeraphronoKiaa Trigonatyoidea Megatyroidea Stephanoidea Diaprioidea Ichneumonidae Braconidae Chrysidoidea (lO-l'^ Vespoidea + Apoidea Cephoidea Tenthredinoide Megaiodontoidea Xyeloidea Fig. 4. Chromosome number and evolution of the parasitic Hymenoptera, with haploid values overlain on a cladogram in which superfamilial relationships are those proposed by Rasnitsyn (1988) but with relation- ships within the Chalcidoidea based on current karyological evidence for clarity. Ranges and (in parentheses) modal values. and the Encyrtidae has also been pro- posed by a number of workers, but Gib- son (1986) and WooUey (1988) considered that the supposed synapomorphies might be better regarded as resulting from con- vergence. LaSalle (1987) upheld Trjapit- zin's view that the Tanaostigmatidae, Eu- pelmidae and Encyrtidae form a mono- phyletic group, citing two putative syna- pomorphies. Boucek (1988b) suggested that four families, the Chalcididae, Eury- tomidae, Torymidae and [some] Pterom- alidae were relatively 'ancient', partly be- cause of their 5 segmented tarsi, compar- 52 Journal of Hymenoptera Research © © ® »<^\ ^'^Z *v> © ® I ¥ r Fig. 5-10. Photomicrographs of chromosomes of adult female Hymenoptera revealed by Giemsa staining of ovarian tissue. 5, Behjia depressa (Diapriidae), meiosis, diakinesis in mature egg {2n = 16); 6, 7, Anastatus catalonicus (Eupelmidae), 6, meiosis, diplotene figures in developing oocyte, 7, metaphase mitosis (2« = 10); 8, Diaretiella rapae (Braconidae, Aphidiinae), metaphase mitosis {2n = 12); 9, Meteorus versicolor (Braconidae, Meteorinae), mitosis, one cell at pro-metaphase and one at metaphase {2n = 16); 10, Laelius utilis (Bethylidae), metaphase mitosis {2« = 20). Volume 4, 1995 53 atively large size and thoracic structure, they do not agree well with the hypothesis and he agreed with Trjapitzin and LaSalle of a close relationship between Eupelmi- that at least the Eupelmidae and Encyrti- dae and Encyrtidae (Trjapitzin 1978, La- dae might form a closely related group. Salle 1987, Boucek 1988a, Woolley 1988, However, this view was not supported by Noyes 1990) which have haploid numbers Gibson (1989, 1990) who considered both of 5 (Figs 6, 7) and 8-11 respectively, be- that the Eupelmidae might not be mono- cause if a modal haploid chromosome phyletic and that the characters used to number of 5 represents a synapomorphy, unite them with the Encyrtidae and Tan- then this relationship would require either aostigmatidae are ". . . either primitive two separate reductions (from c. 10) or a features or apparently were derived in- reversal. However, a close relationship be- dependently several times". tween Eupelmidae and Encyrtidae is not The present data could suggest that universally accepted (Gibson 1989, 1990), there has been an approximate halving of and chromosome number should be in- chromosome numbers within the Chalci- corporated in future cladistic analyses of doidea, from around a modal number of the superfamilies as an independent char- 10 as shown by the Eurytomidae and En- acter. cyrtidae to 5 or 6 in Aphelinidae (but see Potentially significant variation occurs below), Chalcididae, Eulophidae, Leucos- within the Braconidae and the Ichneu- pidae, Ormyridae, Pteromalidae, Torymi- monidae. In the former family, several dae and Trichogrammatidae. It should be subfamilies have haploid numbers be- noted that within the Aphelinidae, whilst tween 8 and 11, for example, the Bracon- most taxa examined have haploid values inae (Habrobrncon), Meteorinae {Meteonis; of 5, Pteropterix orientalis has an n value of Fig. 9), and Miracinae {Mirnx). However, 11 (referred to as Archenomus orientalis by rather higher numbers from 12 to 17 are Baldanza et al. 1991a). However, P. orien- found in the Doryctinae (Heterospilus), talis is an highly derived species within Alysiinae {Phaenocarpa, Dacnusa) and Opi- the Coccophaginae, being either a sister inae (Biosteres), whereas substantially low- group of Coccophagoides or representing a er numbers {n = 4 to 7) are found in the derived branch within Encarsia itself (A. Aphidiinae {Aphidius, Diareticlla (Fig. 8), Polaszek personal communication). In ei- Ephedrus, Praon), in the exothecine genus ther case, the high haploid chromosome Rln/sipolis (n = 6), and in the unrelated number in this taxon is clearly derived Charmontinae {Charmon; n = 5). The pos- with respect to the values of 5 (and 6) dis- session of low values within the four gen- played by the other aphelinids studied era of Aphidiinae examined provides a (collectively representing both the Cocco- potential synapomorphy for the group, phaginae and Aphelininae), and these which otherwise seems an heterogeneous lower values may therefore be taken as assemblage in which members are united representing the ancestral range of values by few characters other than biology, and for the family as a whole. The n value of more taxa will need to be examined before 11 in Pteropterix orientalis, being nearly this can be confirmed. The low n values twice that found in the other aphelinids found in Rln/sipolis might be an autapo- investigated, is further suggestive that this morphy, and it would be interesting to taxon could have originated through a know something about chromosome num- polyploidy event. bers in the apparently closely related Cli- The karyological data summarized here nocentrini and other Rogadinae s.s.. The lend some support to Boucek's view that haploid number of 5 found in Charmon the Eurytomidae are a relatively ancient was especially surprising as the subfamily and underived family of Chalcidoidea but seems to be close to the Macrocentrinae VARIATION AND IMPLICATIONS FOR HYMENOPTERA SYSTEMATICS 54 Journal of Hymenoptera Research (Quicke and Achterberg 1990, Quicke et any great extent in studies of Hymenop- al. 1994) for which we obtained a value of tera systematics. ^^Within the Ichneumonidae, most data ,, . J^J^l ^^D INTEFSPECIFIC available up until now were for the sub- family Ichneumoninae but a few chromo- some numbers for the Campopleginae Only with the advent of techniques for (Venturia), Cryptinae {Agr other eutes, Mas- examining chromosome number and mor- triis), Orthocentrinae (Orthocentrus), and phology in adult Hymenoptera has it be- Orthopelmatinae (Orthopelma) had also come possible to study intraspecific vari- been published. The overwhelming major- ation in nature and thus even to reveal the ity of the species had haploid chromo- presence of hitherto unsuspected cryptic some numbers modal at 11. We extended species or species complexes. As will be this data set by making chromosomal apparent from Table 1, chromosome num- preparations also for the Banchinae {Glyp- ber is often relatively consistent within a ta, Lissonotn), Pimplinae {Ephialtes and Po- single genus. For example, in the Ichneu- lysphincta), and Tryphoninae (Dyspetes). monidae, the 5 species of Diadromus for These additional data generally support which chromosome numbers are available the earlier findings, although the ephial- all have a haploid value of 11. Even for tine genus Ephialtes had an n value of 15, the large genus Ichneumon, 18 of the 25 rather higher than appears typical for the species examined have n = 12. family, while the polysphinctine genus Po- Although chromosome numbers are lysphincta had the more typical, even low, usually considered as differentiating char- haploid value of 9. acters, serving to help distinguish between In addition to chromosome number, closely related forms, they may also be chromosome size and structure have been used in an integrative fashion, providing used extensively for cytotaxonomic pur- evidence for uniting related forms if the poses in other groups of organisms. Dis- chromosome number represents a syna- cussion of size and centromere position in pomorphy. For example, the discovery the parasitic Hymenoptera is currently se- that all members of the cynipid genus Di- verely hampered by the relative paucity of plolepis have n=9, whereas the haploid data; in fact, in many illustrations, and es- values for all members of the other cyni- pecially among the earlier ones, centro- pid genera investigated to date for which meres are hardly (if at all) discernable. reliable figures are available is 10, pro- However, if we look at the karyotypes of vides additional evidence for the mono- the best studied groups (i.e. Ichneumoni- phyly of Diplolepis. dae, Torymidae, Cynipidae), bi-armed Chromosome numbers are fixed in the (metacentric in the broad sense) chromo- great majority of species of parasitic somes predominate in most cases (Figs 7- wasps whose populations have been stud- 10). Some Hymenotera, for example sev- ied in detail in the field. Two possible ex- eral Diplolepis species (Sanderson 1988), ceptions, however, are the ichneumonines may also have numerous acrocentric chro- Icheumon extensorius and I. suspiciosus, mosomes. Even less can be said about both of which were revealed by Gokhman chromosome size, though in general in the (1993a) to comprise individuals with two parasitic Hymenoptera it is inversely re- different diploid numbers, 24 and 26. lated to chromosome number. Much more Since specimens with the intermediate by way of comparative and quantitative chromosome number were not found data will have to be assembled before it (with one possible exception in /. extenso- will be possible to use these features to rius) the possibility that these represent Volume 4, 1995 55 Table 2. Sibling species in the parasitic Hymenoptera detected by karyological features Family Species Chromosomal characteristics Reference Ichneumonidae Ichneumonidae Ichneumonidae Ichneumonidae Encyrtidae Pteromalidae Torymidae Torymidae Aethecerus dispar Wesmael Aethecerus ranuii Gokhman Tycherus australogemiiius Gokhman Tycherus ischionieliniis (Gravenhorst) Ichneumon extensorius L. Ichneumon suspiciosus Wes- mael Copidosoma "tnincatellum (Dalman)'" Copidosoma floridanum (Ash- mead) C. floridamim (Ashmead) Nasonia vitripennis (Walker) Nasonia vitripennis (Walker) Torymus californicus (Ash- mead) Totrymus warreni (Cocker- ell)^ Monodontomerus obscurus (Westwood)'* Monodontomerus obscurus (Westwood)^ 2n = 24 2» = 22 2n = 22 2« = 18 2n = 24, 26' 2n = 24, 26 n = 10, 2n = 20 n = 11, 2n = 22 n = 8,2n = 16 n = 5 + - IB, 2n = 10 n = 6, 2« = 12 n = 6(6M), 2n = 12 n = 6(5M + lA), 2n = 12 n = 6, 2n = 12 Gokhman 1991a Gokhman 1991a Gokhman 1993a Gokhamn 1993a Hunter & Bartlett 1975 Strand & Ode 1970 Leiby 1922; Patterson 1917, 1921; Patterson & Porter 1917 Nur ct al. 1988 and others Goodpasture 1974 Goodpasture & Grissell 1975 Goodpasture & Grissell 1975 McDonald & Krunic 1971 Abbreviations: A = acrocentric; M = metacentric (in a broad sense). ' A single, probably aneuploid, female specimen with 25 chromosomes has also been found. ^Apparently belongs to C. floridamim. ■ Apart from T. californicus, the second chromosome pair in T. ivarreni has secondary constrictions. ■* Apparently belongs to M. laticornis Grissell & Zerova. two sibling species rather than chromo- somal races, seems much more likely, though the reverse cannot at present be discounted. Despite intensive effort, in neither case was it possible to detect any external morphological criteria to permit delineation of these putative taxa in the absense of karyological evidence. Further, in one instance, a chromosomal polymor- phism in terms of C-banding pattern has been detected in the ichneumonine spe- cies, Dirophanes invisor (Thunberg). The C- banding patterns of the two homologous chromosomes of the second pair of meta- centrics do not differ from one another in some individuals, but in others, this pair is obviously heteromorphic, one member of the pair being substantially longer than the other, and its segment of pericentric heterochromatin is also much more devel- oped (Gokhman 1993b). Several other recently discovered ex- amples of apparent interspecific variation within other parasitic Hymenoptera are summarized in Table 2 and discussed be- low in more detail. The torymid chalci- doids Torymus californicus and T. ivarreni were considered to form a single species by Grissell (1973a). However, karyological analysis has shown that despite the fact that these two wasps have the same chro- mosome number (2n = 12), T. ivarreni has 56 Journal of Hymenoptera Research a pair of acrocentric chromosomes and has secondary constrictions on the second largest pair of submetacentrics, whilst in T. californicus all the chromosomes are bi- armed and the second pair has no con- strictions (Goodpasture and Grissell 1975). Thus these two may well be best inter- preted as sibling species. Other cases of possible karyologically-detected sibling species in the Chalcidoidea are more prob- lematical. Goodpasture (1975a) and Mc- Donald and Krunic (1971) reported n = 4 and n = 6 respectively for apparently the same torymid species, Monodontomerus ob- scurns. The most likely explanation for this difference is that one of these works in- volved a misidentified species with the specimens examined by McDonald and Krunic actually belonging to the very sim- ilar species, M. laticornis Grissell & Zerova, described 14 years later (Zerova and Gris- sell 1985). Two new ichneumonid species of the subfamily Ichneumoninae, Tycherus australogeminus and Aetheceriis ranini, were originally detected on the basis of karyo- logical evidence, but in each case reliable morphological differences were also found (Gokhman 1991a; see also Table 2). Application of karyology may also be important in laboratory cultures as avail- able evidence suggests that, at least in some instances, strains that were believed to belong to a single species may in fact represent more than one, with different laboratories working on different entities. For example. Hunter and Bartlett (1975) working with what they referred to as Copidosoma truncatellum, reported it as having a haploid number of 10. C. trun- catellum was subsequently partly synony- mized with C. floridanum, but Strand & Ode (1990) reported n = 11 for apparently the same species. Several earlier workers had reported the haploid number for C. floridanum (as Paracopidosomopsis floridan- us) to be 8 (Leiby 1922, Patterson 1917, 1921, Patterson and Porter 1917), but their findings have to be treated with consid- erable caution as the techniques for fixa- tion and preparation used in pre-1930 studies are often unreliable. Also in the genus Copidosoma, Hegner (1915) reported n = 11 and Patterson (1921) reported n = 10 for C. gelechiae. Again these data may not be fully reliable for technical reasons. However, it is harder to interpret the ap- parent conflict in reported numbers for the widely studied pteromalid, Nasonia vitri- pennis. Many workers (e.g. Gershenzon 1946, 1968, Pennypacker 1958, Whiting 1960, 1968) have reported an n value of 5, but Goodpasture (1974) working on the University of California at Davis culture found n = 6. As with the case of the ich- neumonines discussed below, intensive morphological investigation of these and other strains by Goodpasture failed to re- veal any differences. Therefore, the possi- bility that the Davis culture had devel- oped as a unique chromosomal race must be considered. Such variants are not un- common in cultures of other organisms. However, it should be noted that Darling and Werren (1990) recently discovered two cryptic species of Nasonia in North America, and the karyological results could also reflect a sibling species com- plex. The discovery of sibling species that can only reliably be separated by karyotype may pose a considerable nomenclatural problem, since the current Code of the Zoological Nomenclature requires new taxa to be differentiated from existing ones. As it is not normally possible to ob- tain karyological data from the type spec- imens of the species that have already been described, it would not possible to give scientific names to both of them (Gokhman 1993a) unless it were possible (for instance through geographic distri- bution) to infer the karyotype of the de- scribed taxon. Intraspecific karyotypic variation may also be of interest for population cytoge- netics. This type of variation is favoured in Hymenoptera by some characteristics of their genetic system, which allows the sur- Volume 4, 1995 57 vival and comparatively high viability of aneuploids (Imai et al. 1984). The data ob- tained, for example, for the ichneumonid Tycherus heUicornis suggest the long-term persistence of a chromosomal population polymorphism, probably induced by a translocation and subsequent non-disjunc- tion of chromosomes (Gokhman 1989). Occasional aneuploid specimens were also found in Ichneumon extensor ius and /. gra- cilentiis. Perhaps the most interesting case of numerical chromosomal polymorphism is described by Nur et al. (1988) and Wer- ren (1991) and occurs in the pteromalid, Nasonia vitripennis. This type of polymor- phism implies the existence of a particular B chromosome. Being transmitted pater- nally into the diploid zygotes, this chro- mosome eliminates all other chromosomes of the paternal set from the zygote, thus converting the originally diploid zygote into a haploid one. Therefore this B chro- mosome may be considered as the most selfish genetic element ever known. FUNCTIONAL IMPLICATIONS Reduced chromosome numbers will generally be associated with reduced lev- els of recombination (Vorontsov 1966, White 1973). Under many circumstances, a reduction in recombination will be mal- adaptive. However, situations that favour high levels of inbreeding, for example, ei- ther parasitisation of aggregated hosts or gregarious parasitism combined with mat- ing near the emergence site, may lead to selection in favour of low intrinsic levels of genetic diversity. Such situations may, for example, favour parthenogenesis or in the case of the parasitic Hymenoptera, thelytoky. Similarly, under such circum- stances, high levels of recombination, or large numbers of separate linkage groups, will not be favoured, and consequently, mutations leading to reduced recombina- tion will not necessarily be so deleterious. Such circumstances will permit a reduc- tion in chromosome number more readily than will situations favouring high levels of genetic diversity. Consideration of the data currently available for parasitic Hy- menoptera may be relevant in this respect. For example, in the Aphidiinae the chro- mosome number is markedly lower than in the rest of the Braconidae. Aphidiines parasitise aphids which in turn often form clonal patches. In chalcidoids, many spe- cies are also gregarious or attack clustered hosts, perhaps in part as a result of their small size and dispersal capabilities. It would be difficult to quantify this, but the possibility exists that inbreeding may be more common in the Chalcidoidea as a whole than in the Ichneumonoidea or Cy- nipoidea (Askew 1968). Information on re- combination levels in parasitic Hymenop- tera is extremely limited to date (Crozier, 1975), being based on linkage data {Habro- bracon hebetor) and chiasmata per bivalent arm {Aphytis mytilaspidiis). Further studies taking into account cross-over and levels of heterozygosity in parasitic wasps with different biologies and different chromo- some numbers and morphologies might provide additional evidence in this re- spect. FURTHER PERSPECTIVES Though we have managed for the first time to obtain cytogenetic information for several major groups of parasitic Hyme- noptera, the overwhelming majority of these insects still remains untouched by karyological investigation, and this in- cludes some entire superfamilies such as the Stephanoidea, Megalyroidea, Trigon- alyoidea, Evanioidea and Ceraphronoidea (Fig. 4). The data available at present, however, suggest that the new chromo- somal evidence may be a substantial help in future phylogenetic and taxonomic studies. As regards the higher level phy- logeny of the parasitic Hymenoptera, new karyological evidence is especially needed for various groups of Proctotrupoidea sen- su lato and of Chalcidoidea, especially of such apparently underived taxa as the Mymaridae, Tetracampidae and Rotoiti- 58 Journal of Hymenoptera Research dae (Noyes 1990) and the pteromalid sub- family Cleonyminae (Boucek personal communication). Recent investigations (e.g. Gokhman 1991b, 1994, Costa et al. 1993, Odierna et al. 1993) also show that other karyological data, especially those obtained using dif- ferential chromosome staining (e.g. C- banding), can still provide useful infor- mation even in the absence of differences in chromosome number. More detailed morphological investigations are therefore particularly likely to be of use in future species level work. ACKNOWLEDGEMENTS We would like to thank Efi Kazantzidou (Horti- cultural Research International, Wellesbourne), Prof. Helmut van Emden (University of Reading) and Robert Belshaw, Peter Mayhew and Patricia Reader (Imperial College) for providing living material for original investigation, Lynn Kimsey for enabling us to see C. Goodpasture's Ph.D. thesis, Andy Polaszek for helpful discussions pertaining to the Aphelini- dae, and Zdenek Boucek, John LaSalle, and Mark Shaw for identifying various specimens. This work was partly supported by grants from the British Council and International Science Foundation to VEG; DLJQ is supported by the NERC Initiative in Taxonomy. LITERATURE CITED Abe, Y. 1994. The karyotype of the chestnut gall wasp Dn/ocosmus kiiriphihis (Hymenoptera, Cy- nipidae). Applied Entomology and Zoology 29:299- 300. Amalin, D. M., L. M. Rueda and A. A. Barrion. 1988. Cytology of a parasitic wasp, Dirhinus himalay- anus Westwood (Chalcididae: Hymenoptera). Philippine Entomologist 7:272-274. Askew, R. R. 1968. Considerations on speciation in Chalcidoidea (Hymenoptera). Evolution 22:642- 645. Baldanza, F., L. Gaudio and G. Viggiani. 1991a. Ri- cerche cariologiche suW Archenomus orientalis Sil- vestri (Hymenoptera: Aphelinidae), parassitoide di Pseudaulacapis pentagona (Targioni Tozzeti) (Homoptera: Diaspididae). Atti XVI Congresso na- zionale italiano di Entomologia Bari-Martina Franca (Ta) 23/28 settembre 1991:457^61. Baldanza, P., G. Odierna, and G. Viggiani. 1991b. A new method for studying chromosomes of par- asitic Hymenoptera, used on Encarsia berlesei (Howard) (Hymenoptera: Aphelinidae). Bolletti- no del Laboratorio di Entomologia Agraria "Filippo Silvestri" 48:29-34. Basibuyuk, H. H. and D. L. J. Quicke. 1994. Evolu- tion of antennal cleaner structure in the Hyme- noptera (Insecta). Norwegian Journal of Agricultur- al Science, Supplement 16:199-206. Basibuyuk, H.H. and D. L. J. Quicke. 1995. Mor- phology of the antenna cleaner in the Hymenop- tera with particular reference to non-aculeate families (Insecta). Zoologica Scripta 24:157-177. Bigot, Y., M. H. Hamelin and G. Periquet. 1991. Mo- lecular analysis of the genomic organization of the Hymenoptera Diadromus pulchellus and Eu- pelmus iniilleti. Journal of Evolutionary Biology 4: 541-556. Blackman, R. L. 1980. Chromosome numbers in the Aphididae and their taxonomic significance. Sys- tematic Entomology 5: 7-25. Boucek, Z. 1988a. An overview of the higher classi- fication of the Chalcidoidea (Parasitic Hymenop- tera). In, V. K. Gupta (Ed.). Advances in Parasitic Hymenoptera Research. E. J. Brill: Leiden, pp. 11- 23. Boucek, Z. 1988b. Australasian Chalcidoidea (Hymenop- tera). A biosystematic revision of genera of four- teen families, with a reclassification of species. CAB International: Wallingford. 832 pp. Costa, M. A., G. A. R. de Melo, S. das Gracas Pom- polo, and L. A. O. Campos. 1993. Karyotype and heterochromatin distribution (C-band patterns) in three species of Microstigmus wasps (Hyme- noptera, Sphecidae, Pemphredoninae). Revista Brasileira de Genetica 16: 923-926. Crozier, R. H. 1975. Animal Cytogenetics. Vol. 3, Part 7. Gebriider Borntraeger: Berlin, pp. 1-95. Crozier, R. H. 1977. Evolutionary genetics of the Hy- menoptera. Annual Review of Entomology 22: 263- 288. Crozier, R. H. and E. F. Taschenberg. 1972. Chro- mosome number polymorphism in the sawfly Ja- nus integer (Hymenoptera: Cephidae). Psyche 79: 111-119. Darling, D. C. and J. H. Werren. 1990. Biosystematics of Nasonia (Hymenoptera: Pteromalidae): Two new species reared from birds' nests in North America. Annals of the Entomological Society of America 83: 352-370. Dessart, M. P. 1987. Quelques cas de la determina- tion du sexe chez les hymenopteres. Bulletin et Annales de la Societe Royale Entomologique de Bel- gique 123:55-58. Dijken, M. J. van, 1991. A cytological method to de- termine primary sex ratio in the solitary parasit- oid Epidinocarsis lopezi. Entomologia Experimentalis et Applicata 60:301-304. Dijkstra, L. J. 1986. Optimal selection and exploita- tion of hosts in the parasitic wasp Colpoclypeus Volume 4, 1995 59 floras (Hym., Eulophidae). Netherlands Journal of Zoology 36:177-301. Dodds, K. S. 1938. Chromosome numbers and sper- matogenesis in some species of the hymenopter- ous family Cynipidae. Genetica 21:177-190. Doncaster, L. 1910. Gametogenesis in the gall-fly, Neurolenis lenticularis (Spntliegaster baccaruni). Part I. Proceedings of the Royal Society, Series B 82: 88-112. Doncaster, L. 1911. Gametogenesis in the gall-fly, Neuroteriis lenticularis. Part II. Proceedings of the Royal Society, Series B 83:476-489. Doncaster, L. 1916. Gametogenesis and sex-deter- mination in the gall-fly, Neuroterus lenticularis (Spathegaster baccaruni). Part III. Proceedings of the Royal Society, Series B 89:183-200. Dozortseva R. L., 1936. Chromosome morphology in the parasitic wasp Pteromalus puparum. Bulletin of the Academy of Sciemces of the USSR. 6:1220-1231. (In Russian with English summary) Dreyfus, A. and M. E. Breuer. 1944. Chromosomes and sex determination in the parasitic hymenop- teran Telenonnis fariai Lima. Genetics 29:75-82. Emeljanov, A. F. and V. I. Kirillova. 1989. Trends and modes of karyotype evolution in the Cicadina (Homoptera). I. Entomologicheskoye Obozreniye 68: 587-603. Emeljanov, A. F. and V. 1. Kirillova. 1991. Trends and modes of karyotype evolution in the Cicadina (Homoptera). II. Entomologicheskoye Obozreniye 70: 796-817. Frohlich, M. W. 1987. Common-is-primitive: A par- tial validation by tree counting. Systematic Botany 12:217-237. Fukada, H. and M. Takemura. 1943. Genetical stud- ies of Trichogramma. Japanese Journal of Genetics 19:275-281.' Gauld, I. D. and B. Bolton. 1988. The Hymenoptera. British Museum (Natural History) /Oxford Uni- versity Press:Oxford. 332 pp. George, K. and S. Geethamma. 1992. Cytology and evolution of jasmines. Cytologia, Tokyo 57: 27-32. Gershenzon, S. M. 1946. The genetic structure of the natural populations of Mormoniella vitripennis Wlk. (Chalcididae: Hymenoptera). Zhurnal Ob- shchei Biologii 7:165-173. (in Russian). Gershenzon, S. M. 1968. The chromosomes and sex determination in the parasitic wasp, Mormoniella vitripennis Wlk. Tsitologiya i Genetika 2:3-13 (in Russian). 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. Canadian Entomologist 117: 1395- 1443. Gibson, G. A. P. 1986. Evidence for monophyly and relationships of Chalcidoidea, Mymaridae, and Mymarommatidae (Hymenoptera: Terebrantes). Canadian Entomologist 118:205-240. Gibson, G. A. P. 1989. Phylogeny and classification of Eupelmidae, with a revision of the world gen- era of Calosotinae and Metapelmatinae (Hyme- noptera: Chalcidoidea). Memoirs of the Entomolog- ical Society of Canada 149: 1-121. Gibson, G. A. P. 1990. A word on chalcidoid classi- fication. Chalcid Forum 13: 7-9. Gokhman, V. E. 1985. Chromosome sets in some Ich- neumoninae (Hymenoptera: Ichneumonidae). Zoologichesky Zhurnal 64: 1409-1413. (In Russian) Gokhman, V. E. 1987. Chromosomes in the Ichneu- moninae (Hymenoptera: Ichneumonidae). Zoolo- gichesky Zhurnal 66:543-548. (In Russian) Gokhman, V. E. 1989. Karyotypes of ichneumon flies of the Tycherus osculator group (Hymenoptera: Ichneumonidae). Entomologicheskoye Obozreniye 68: 710-714. (In Russian) Gokhman, V. E. 1990a. Karyology and systematics of the subfamily Ichneumoninae (Hymenoptera: Ichneumonidae). Summary of the PhD thesis. Moscow State University: Moscow. 17 pp. (In Russian) Gokhman, V. E. 1990b. Main trends of the karyotype evolution in Ichneumoninae (Hymenoptera: Ich- neumonidae). Zoologichesky Zhurnal 69:70-80. (In Russian) Gokhman, V. E. 1991a. New species of Phaeogenini (Hymenoptera: Ichneumonidae) from the Euro- pean part of the USSR. Zoologichesky Zhurnal 70: 73-80. (In Russian) Gokhman, V. E. 1991b. Taxonomic aspects of kar- yology of the Ichneumoninae (Hymenoptera: Ichneumonidae). / AlTRiissian Conference on Insect Genetics. Abstracts. Moscow, p. 32. (In Russian) Gokhman, V. E. 1993a. New data on the karyology of Ichneumonina (Hymenoptera: Ichneumoni- dae). Zoologichesky Zhurnal 72:85-91. (In Russian) Gokhman, V. E. 1993b. Sibling species and chromo- somal polymorphism in natural populations of Ichneumoninae (Hymenoptera, Ichneumonidae). Karyosystematics of the invertebrate animals. II. S. Petersburg, pp. 25-27. (In Russian) Gokhman, V. E. 1994. Cytogenetics of the parasitic Hymenoptera: karyotype diversity and taxonom- ic implications. 5th European Congress of Ento- mology. Uni\'ersity of York, UK. 29 August-2 September 1994. Abstracts, p. 87. Goodpasture, C. 1974. Cytological data and classifi- cation of the Hymenoptera. Unpublished Ph.D. thesis. University of California: Davis. 178 pp. Goodpasture, C. 1975a. Comparative courtship be- haviour and karyology in Monodontomcrus (Hy- menoptera: Torymidae). Annals of the Entomolog- ical Society of America 68:391-397. Goodpasture, C. 1975b. The karyotype of the cynipid 60 Journal of Hymenoptera Research Callirhytis palmiformis (Ashmead). Annals of the Entomological Society of America 68:801-802. Goodpasture, C. and E. E. Grissell. 1975. A karyo- logical study of nine species of Torymus (Hyme- noptera: Torymidae). Canadian journal of Genetics and Cytology 17:413-432. Grissell, E. E. 1973a. Revision of western Nearctic species of Torx/mus Dalman. Unpublished Ph.D. thesis. University of California: Davis. Grissell, E. E. 1973b. New species of North American Torymidae (Hymenoptera). Pan-Pacific Entomolo- gist 49:232-239. Guhl, A. P. and R. L. Dozortseva, 1934. A contribu- tion to the knowledge of sex determination in Hymenoptera. Compiles Rendiis Academic Sciences Russie. 3:524-526. (In Russian with English sum- mary) Hedderwick, M. P., M. el Agoze, P. Garaud and G. Periquet. 1985. Mise en evidence de males het- erozygotes chez I'hymenoptere Diadromus pul- chellus (Ichneumonide). Genetique Selection et Evo- lution 17:303-310. Hegner, R. W. 1915. Studies of germ cells. IV. Pro- toplasmic differentiation in the oocytes of certain Hymenoptera. Journal of Morphology 26:495-561. Henking, H. 1892. Untersuchen iiber die ersten En- twicklungsvorgange in der Eiern der Insekten. III. Specielles und AUgemeines. Zeitschrift fiir Wissenschaftliche Zoologie 54:1-274. Heraty, J. M., J. B. WooUey, and C. Darling. 1994. Phylogenetic implications of the mesofurca and mesopostnotum in Hymenoptera. Journal of Hy- menoptera Research 3: 241-277. Hogben, L. T. 1920. Studies on Synapsis. I. Oogenesis in the Hymenoptera. Proceedings of the Royal So- ciety, Series B 91:268-293. Hoshiba, H., M. Matsuura and H. T. Imai. 1989. Karyotype evolution in the social wasps (Hy- menoptera, Vespidae). Japanese Journal of Genetics 64:209-222. Hung, A. C. F. 1982. Chromosome and isozyme studies in Trichogramma (Hymenoptera: Tricho- grammatidae). Proceedings of the Entomological So- ciety of Washington 84:791-796. Hung, A. C. F. 1986. Chromosomes of three Brachy- meria species (Hymenoptera: Chalcidoidea). Ex- perientia 42:579-580. Hung, A. C. F., Day, W. H. and Hedlung, R. C. 1988. Genetic variability in arrhenotokous and thelyt- okous forms of Mesochonis nigripes Ratzeburg. (Hym.: Ichneumonidae). Entomophaga 33:7-15. Hunter, M. S. 1993. Sex allocation in a field popu- lation of an autoparasitoid. Oecologia 93:421^28. Hunter, K. W., Jr., and A. C. Bartlett, 1975. Chro- mosome number of the parasitic encyrtid Copi- dosoma truncatellum (Dalman). Annals of the En- tomological Society of America 68:61-62. Hunter, M. S., U. Nur and J. H. Werren. 1993. Origin of males by genome loss in an autoparasitoid wasp. Heredity 70:162-171. Imai, H. T., R. H. Crozier and R. W. Taylor. 1977. Karyotype evolution in Australian ants. Chro- mosoma 59:341-393. Imai, H. T. and R. W. Taylor. 1989. Chromosomal polymorphism involving telomere fusion, cen- tromeric inactivation and centromere shift in the ant Myrmecia (pilosula) n = 1. Chromosoma 98:456- 460. Imai, H. T., R. W. Taylor, M. W. J. Crosland and R. H. Crozier. 1988. Modes of spontaneous chro- mosomal mutation and karyotype evolution in ants with reference to the minimum interaction hypothesis. Japanese Journal of Genetics 63:159- 185. Imai, H. T., R. W. Taylor, M. Kubota, K. Ogata, and M. Y. Wada. 1990. Notes on the remarkable kar- yology of the primitive ant Nothomyrmecia ma- crops, and of the related genus Myrmecia (Hy- menoptera: Formicidae). Psyche 97:133-140. Johnson, N. F. 1988. Midcoxal articulations and the phylogeny of the order Hymenoptera. Annals of the Entomological Society of America 81:870-881. Kerr, W. E. 1972. Numbers of chromosomes in some species of bees. Jouriial of the Kansas Entomological Society 45: 111-122. Kerr, W. E. and Z. V. da Silveira. 1972. Karyotypic evolution of bees and corresponding taxonomic impications. Evolution 26:197-202. Konigsmann, E. 1977. Das phylogenetische System der Hymenoptera. Teil 2: Symphyta. Deutsche Entomologische Zeitschrift 24:1-40. Koonz, K. H. 1936. Some unusual cytological phe- nomena in the spermatogenesis of haploid par- thenogenetic Hymenopteran, Aenoplex smithii (Packard). Biological Bulletin 71: 375-385. Koonz, K. H. 1939. Spermatogenesis of a haploid parthenogenetic hymenopteran, Spilocryptus ex- trematus (Cresson). Transactions of the American Microscopical Society 58: 292-303. Kuznetsova, V. G. 1985. Phylogenetic analysis of the chromosome variability and karyosystematics of the leaf-hoppers of the family Dictyopharidae (Homoptera, Auchenorryncha). Entonwlogiches- koye Obozreniye 64: 539-553. (in Russian) LaSalle, J. 1987. New World Tanaostigmatidae (Hy- menoptera, Chalcidoidea). Contributions of the American Entomological Institute 23: 1-181. Leiby, R. W. 1922. The polyembryonic development of Copidosoma gelechiae with notes on its biology. Journal of Morphology 37:195-285. Liu, W. and P. Xiong, 1988. Karyotype study in Trichogramma dendrolimi. Journal of the Wuhan University 2: 105-108. (In Chinese) McDonald, M. D. and M. D. Krunic. 1971. Chromo- some numbers of Monodontomerus obscurus and Volume 4, 1995 61 Pteromahis venustus, chalcid parasites of Megachi- le rotiindata. Arhiv Bioloskih Naiika 23: 9pp. Mackay, M. R. 1955. Cytology and parthenogenesis of the wheatstem sawfly, Cephus ductus Nort. (Hymenoptera; Cephidae). Cauadiau journal of Zoology 33: 161-174. Makino, S. 1951. An Atlas of Chromosome Numbers iu Animals. Iowa State College Press: Ames. 290pp. Martin, F. 1914. Zur Entwicklungsgeschichte der po- lyembryonalen chalcidiers Ageuiaspis (Eucyrtus) fuscicollis Dalm. Zeitschrift fur Wissenschaftliche Zoologie 110: 419^79. Mason, W. R. M. 1983. The phylogeny of the Apo- crita (Unpublished lecture notes cited in Gauld & Bolton, 1988). Moritz, R. F. A. 1986. Genetics of bees other than Apis mellifera. pp. 121-154. In T. E. Rinderer (ed.) Bee Genetics and Breeding. Academic Press: Orlan- do. Naito, T. 1982. Chromosome number differentiation in sawflies and its systematic implication (Hy- menoptera, Tenthredinidae). Kontyu, Tokyo 50: 569-587. Noyes, J. S. 1990. A word on chalcidoid classifica- tion. Chalcid Forum 13: 6, 7. Nur, U., J. H. Werren, D. G. Eickbush, W. D. Burke, and T. H. Eickbush. 1988. A 'selfish' B chromo- some that enhances its transmission by eliminat- ing the paternal genome. Science 240: 512-514. Odierna, G., F. Baldanza, G. Aprea, and E. Olmi. 1993. Occurrence of G-banding in metaphase chromosomes of Encarsia berlesei (Hymenoptera: Aphelinidae). Genome 36: 662-667. Patterson, J. T. 1917. Studies on the biology of Para- copidosomopsis. III. Maturation and fertilization. Biological Bulletin of the Marine Biological Labora- tory Woods Hole, Mass. 33: 38-50. Patterson, J. T. (1921) 1922. The development of Par- acopidosomopsis. Journal of Morphology 36:1-70. Patterson, J. T. and L. T. Porter. 1917. Studies on the biology of Paracopidosomof)sis. II. Spermatogene- sis of males reared from unfertilized eggs. Biolog- ical Bulletin of the Marine Biological Laboratory Woods Hole, Mass. 33:28. Peacock, A. D. and R. A. R. Gresson. 1932. Male hap- loidy and female diploidy in Sirex cyaneus F. (Hy- men.). Proceedings of the Royal Society of Edinburgh 57: 97-103. Pennypacker, M. I. 1958. The chromosomes of the parasitic wasp Mormoniella vitripennis. Part I. In spermatogenesis of haploid and diploid males. Archives de Biologic, Liege 69:483-495. Pompolo, S. G. and C. S. Takahashi. 1987. Cytoge- netics of Brazilian Polybiini wasps (Hymenop- tera, Vespidae, Polistinae). Revista Brasileira de Genetica 10: 483-497. Pompolo, S. G. and C. S. Takahashi. 1990. Karyotype of ten species of social wasps (Hymenoptera, Pol- istinae, Polybiini). Revista Brasileira de Genetica 13: 469-477. Quicke, D. L. J. 1993. Prijiciples ami Techniques of Con- temporary Taxonomy. Blackie Academic and Pro- fessional: Glasgow. 311 pp. Quicke, D. L. J. 1994. Phylogenetic and biological transitions in the Braconidae (Hymenoptera: Ich- neumonoidea). Nonvegian journal of Agricultural Sciences. Supplement 16: 155-162. Quicke, D. L. J. and C. van. Achterberg. 1990. Phy- logeny of the subfamilies of the family Braconi- dae (Hymenoptera: Ichneumonoidea). Zoologis- che Verhandelingen, Leiden, 258:1-95. Quicke, D. L. ]., M. G. Fitton and S. Ingram. 1992. Phylogenetic implications of the structure and distribution of ovipositor valvilli in the Hyme- noptera (Insecta). journal of Natural History 26: 587-608. Quicke, D. L. J., M. G. Fitton, J, R. Tunstead, S. In- gram and P. V. Gaitens. 1994. Ovipositor struc- ture and relationships within the Hymenoptera, with special reference to the Ichneumonoidea. Journal of Natural History 28:635-682. Rasch, E. M., J. D. Cassidy and R. C. King. 1975. Es- timates of genome size in haploid-diploid species of parasitoid wasps. Journal of Histochemistry and Cytochemistry. 23: 317. Rasch, E. M., J. D. Cassidy and R. C. King. 1977. Ev- idence for dosage compensation in parthenoge- netic Hymenoptera. Chromosoma 59:323-340. Rasnitsyn, A. P. 1980. Origin and evolution of Hy- menoptera. Trudy Palcontologicheskogo Instituta 174:1-190. (in Russian) Rasnitsyn, A. P. 1988. An outline of evolution of the hymenopterous insects (Order Vespida). Oriental Insects 22:115-145. Rossler, Y. and P. De Bach. 1973. Genetic variability in a thelytokous form of Aphytis mytilaspidus (Le Baron) (Hymenoptera: Aphelinidae). Hilgardia 42:149-175. Sanderson, A. R. 1932. The cytology of parthenogen- esis in Tenthredinidae. Genetica (The Hague) 14: 321^94. Sanderson, A. R. 1970. Further studies on the cytol- ogy of sawflies. Proceedings of the Royal Society of Edinburgh B61:29-40. Sanderson, A. R. 1988. Cytological investigations of parthenogenesis in gall wasps (Cynipidae, Hy- menoptera). Genetica (The Hague) 77:189-216. Schleip, W. (1909) 1910. Die Reifung des Eies von Rhodites rosae L. und einige allgemeine Bemer- kungen uber die Chromosomen bei parthenoge- netischer Fortpflanzung. Zoologischer Anzeigcr 35: 203-213. Schmieder, R.G., 1938. The sex ratio in Melittobia chah/bii Ashmead, gametogenesis and cleavage in females and in haploid males (Hymenoptera: Chalcidoidea). The Biological Bulletin. 74:256-266. 62 Journal of Hymenoptera Research Silvestri, F. 1908. Contribuzioni alia conoscenza biol- ogica degli Imenotteri Parassiti. II. (1) Sviluppo dell' Ageniaspis fuscicollis (Dalm.) e note biografi- che. Bolletino del Laboratorio di Zoologia Generale e Agraria della R. Scuola Superiore d'Agricoltum in Portici 3:29-53. Silvestri, F. 1914. Prime fast de sviluppo del Copido- soma bin/ssoni (Mayr), imenottere calcidide. Aua- tomische Anzeiger 47:45-56. Speicher, B. R. 1937. Oogenesis in a thelytokous wasp, Nemeritis canescens (Grav.). Journal of Mor- phologij 61:453-467. Speicher, K. G. and B. R. Speicher. 1938. Diploids from unfertilized eggs in Habrobracon. Biological Bulletin 74:247-252. Speicher, B. R. and K. G. Speicher. 1940. The occur- rence of diploid males in Habrobracon brevicornis. American Naturalist 74:379-382. Stille, B. and L. Davring. 1980. Meiosis and repro- ductive strategy in the parthenogenetic gall wasp Diplolepis rosae (L.) (Hymenoptera, Cynipidae). Hereditas 92:353-362. Stouthamer, R. and D. J. Kazmer. 1994. Cytogenetics of microbe-associated parthenogenesis and its consequences for gene flow in Trichogramma wasps. Heredity 73: 317-327. Strand, M. R. and P. J. Ode. 1990. Chromosome num- ber of the polyembryonic parasitoid Copidosoma floridanum (Hymenoptera: Encyrtidae). Annals of the Entomological Society of America 83:834-837. Torvik-Greb, M. 1935. The chromosomes of Habro- bracon. Biological Bulletin 68:25-34. Trjapitzin, V. A. 1978. Superfamily Chalcidoidea. In G. S. Medvedev (ed.) Keys to the Identification of Insects of the European Part of the U.S.S.R. Nauka Publishers: Leningrad, pp. 39-58. Viggiani, G. 1967. Richerche sugli Hymenoptera Chalcidoidea. XV. Osservazioni caryolgische pre- liminari sulli Aphelinus mali. Bollettino del Labor- atorio di Entomologia Agraria "Filippo Silvestri" 25: 326-330. Vorontsov, N. N. 1966. The karyotype evolution. In: A Manual on Cytology. Vol. 2. Nauka Publishers, Moscow /Leningrad, pp. 59-389. [in Russian] Wahrman, M. Z. and S. Zhu. 1993. Haploid and dip- loid cell cultures from a haplo-diploid insect. In- vertebrate Reproduction and Development 24: 79-86. Watrous, L. E. and Q. D. Wheeler. 1981. The out- group comparison method of character analysis. Systematic Zoology 30:1-11. Werren, J. H. 1991. The patemal-sex-ratio chromo- some of Nasonia. American Naturalist 137:392^02. White, M. J. D. 1973. Animal Cytology and Evolution. Cambridge University Press: Cambridge. 961 pp. Whiting, A. R. 1927. Genetic evidence for diploid males in Habrobracon. Biological Bulletin 53:438- 449. Whiting, P. W. 1960. Polyploidy in Morinoniella. Ge- netics 45:949-970. Whiting, P. W. 1968. The chromosomes of Mormon- iella. Journal of Heredity 59:19-22. Wieman, H. L. 1915. Observations on the spermato- genesis of the gall-fly, Dryophanta erinacei (Mayr). Biological Bulletin of the Marine Biological Labora- tory VJoods Hole, Mass. 28:34-46. Woolley, J. B. 1988. Phylogeny and classification of the Signiphoridae (Hymenoptera: Chalcidoidea). Systematic Entomologi/ 13: 465-501. Zerova, M.D. and E. E. Grissell. 1985. A new species of the genus Monodontomerus Westw. (Hymenop- tera, Torymidae) — Parasite of leafcutting bees. Entomologicheskoye Obozreniye 64:203-206. (in Russian) APPENDIX I. METHODOLOGY FOR STAINING CHROMOSOMES IN ADULT PARASITIC WASPS General comments. Chromosomes can be stained in many adult parasitic wasp females especially in recently eclosed individuals using ovarian tissue. As always in insects, care should be taken to count mul- tiple cells so that the occasional polyploid cell can be recognised and discounted. In common with Cro- zier (1975) we recommend counting at least ten in- dividual metaphase plates though occasionally this may not be possible. If fewer plates are available, one may gain extra confidence in the results if it is possible to identify through size and morphology, particular pairs of chromosomes (see for example. Fig. 9). Usually, mitotic divisions are most evident, but in some individuals and taxa, meiosis may also be observed, sometimes with very clear spreads. The number of plates may also be increased if the wasp is fed on honey water containing colchicine for a few hours before preparation, though beyond about 5 hours one stands the risk of increasing the propor- tion of polyploid cells. The general method de- scribed below is modified after that of Imai et al. (1988). Method. Metasomas of adult female wasps are dissected in hypotonic sodium citrate solution with colchicine (Solution A). Ovaries are incubated in this solution at room temperature (c. 25°C) for 20 min- utes (optimal times for different taxa may vary slightly). Following incubation, ovaries are trans- ferred to a thoroughly cleaned microscope slide, ex- cess citrate solution pipetted off and the slide is gently flushed with Fixative I taking care not to wash off the ovaries. Whilst still moist with Fixative I, the ovaries are disrupted (e.g. using fine mounted needles) and their cells spread gently over the mid- dle part of the slide. One or two drops of Fixative II are then applied to the centre of the area of spread cells and the more aqueous phase which is displaced to the edge of the slide is blotted off. The same pro- Volume 4, 1995 63 cedure is then performed with Fixative III. The slide is then air dried before staining for at least 20 min- utes. Excess stain should be washed off with dis- tilled water and the slide examined dry or under emersion oil (do not apply mounting media). Slides can be restained if the initial result was insufficient- ly intense, or they can be destained with alcohol in the reverse situation. Solutions (A) Hypotonic sodium citrate with colchicine Ig Na citrate.2H,0 5mg colchicine 100 ml distilled water (B) Fixative I 3 parts Ethanol 3 parts Glacial acetic acid 4 parts distilled water (C) Fixative II 1 part Ethanol 1 part Glacial acetic acid (C) Fixative III Glacial acetic acid (D) Stain 2 ml Giemsa solution 50 ml 0.089M Na.HPO, 50 ml 0.066M KH>0, J. HYM. RES. Vol. 4, 1995, pp. 64-76 Euryischomyia Girault (Hymenoptera: Chalcidoidea: Aphelinidae: Eriaporinae: Euryischiini) Mary Carver Division of Entomology, CSIRO, G.P.O. Box 1700, Canberra, A.C.T. 2601, Australia Abstract. — Types of some Australian Euryischiini described by A. A. Girault have been exam- ined, and compared with other material identified by Girault, together with other reared and collected material of Euryischomyia Girault. Euryischomyia saintpverrei Girault, 1915, syn. nov.; Myiocnema marmorativentris Girault, 1915, syn. nov.; E.fasciata Girault, 1916, syn. nov.; and £. setosa Girault, 1929, syn. nov.; are placed in synonymy with £. flaznthorax Girault & Dodd, 1915; and £. alami Shafee, 1975, syn. nov., (described fom India), is placed in synonymy with E. washingtoni Girault, 1914. A key to the 2 remaining species of Euryischomyia is provided. £. washingtoni and £. flavithorax are redescribed. £. zuasJiingtoni is known as a parasite of pseudococcids, and £. flavithorax as a hyperparasite of psyllids, aphids, coccids and pseudococcids. INTRODUCTION compressed hind coxae, large, bristle-like np, , ill. J i.- setae on the lees and elsewhere, postaxil- The placement, status and composition , ,, of the subfamily Eriaporinae (Chalcidoi- ^^^^^ ^^'^^ scutellum and a large propo- dea: Aphelinidae) have long been a matter of uncertainty. The subfamily was erected '^^^ nomenclatural history of the Eu- in the family Aphelinidae by Ghesquiere ^yischiini is complicated by the fact that (1955). Members have been variously ^^^^^^ ^^^^^^ ^^^ ^^'"l^^^ ^^^^^^^ ^^^ f^^" placed in the Aphelinidae (e.g. Ashmead ^^y Euryischidae {sic) to contam Euryischia 1900; Compere 1947; Ferriere 1965) and ^^^' subsequently, he included Myiocnema Elasmidae (e.g. Girault 1914; Smith and ^^^ Euryischomyia in the family also (Sha- Compere 1928; Nikol'skaya 1963). Cur- ^^^ 1^75) but, later, Shafee and Rizvi rently, the Eriaporinae are included in the (^^^O) retained only Euryischia in the fam- Aphelinidae (e.g. Hayat 1994), albeit with i^Y- transferring the other 2 genera to the some reservation (Goulet and Ruber Aphelinidae. The elevation of these genera 1993). Contained taxa of Eriaporinae pos- ^ family status, and their separation, are sess an unbroken, posteriorly expanded "^o* accepted herein. parastigma in the fore wing, which is usu- Although Euryischia is well character- ally continued posterobasally and bears 1- ised, the limits of the other genera of Eu- 3 coarse, bristle-like setae. ryischiini have not been understood and Hayat and Verma (1980) created 2 tribes their history is also confusing. Essentially, within the Eriaporinae, namely, the Eria- Girault (1914) described the new genus porini to contain the genera Promuscidea and species Euryischomyia washingtoni and, Girault and Eunotiscus Compere; and the in 1917, erected the genus Paramyiocnema Euryischini (sic) for 4 genera: EuryiscJiia for Myiocnema marmorativentris Girault but Riley, 1889; Myiocnema Ashmead, 1900; did not describe or diagnose the genus. Euryischomyia Girault, 1914; and Para- Girault (1930) placed Euryischomyia in syn- myiocnema Girault, 1917. Euryischiini are onymy with Myiocnema. Hayat and Verma characterised by possession of enlarged, (1980) transferred Euryischomyia flavithorax Vt^LUME 4, 1995 65 Girault & Dodd to Parnmyiocnema. They also opined that E. saintpierrei Girault, E. fasciata Girault and E. setosa Girault might also belong in Paramyiocnema. Shafee et al. (1985) placed Paramyiocnema in synonymy with Myiocnema and Hayat (1994) placed it in synonymy with Eiiryischomx/ia. The types of Eiiryischomyia washingtoni, E. flavithorax, E. saintpierrei, Myiocnema marmorativentris, E. fasciata, and £. setosa have been examined by the author, to- gether with other material identified by Girault, and more recently reared and col- lected material of EuryiscJioirn/ia. As a re- sult, Eiiryischomyia saintpierrei, Myiocnema marmorativentris, E. fasciata, and £. setosa are placed in synonymy with £. flavithorax. In addition, £. alami Shafee, 1975, de- scribed from India, is proposed as a junior synonym of £. washingtoni, thus reducing Eiiryischomyia to 2 species, namely, £. washingtoni and £. flavithorax. METHODS All museum material was examined as provided, as dry-mounts, or as uncleared or cleared slide-mounts. Girault's material was recognizable from information pro- vided by Dahms (1984, 1986), and by rec- ognition of Girault's or Mrs E.J. Girault's handwriting on the labels. Whenever feasible, reared material was obtained by caging live, field-collected material to allow maximal mummification of already-parasitized hosts. Aphid mum- mies (mummified, parchment-like, some- times silk-lined, aphid skins containing immature parasites) were removed and placed individually in gelatin capsules. Except when slide-mounted, each emer- gent parasite was preserved in association with its mummy. Reared material was dry-mounted, or preserved in 80% etha- nol, or cleared and slide-mounted using 10% potassium hydroxide, chloral-phenol, and gum-chloral mounting media. Micrometer eyepieces were used for measurements, all of which are of maxi- mal values; those of the scape do not in- clude the radicle; those of the metasoma include the ovipositor. Length of mesoso- ma + metasoma is given rather than body length because the head of most of Gi- rault's specimens is separated from the trunk and slide-mounted fronto-occipital- ly and, in those instances of two or more mountings per coverslip, head and trunk could not be matched with confidence. Girault customarily enumerated the number of specimens he used in a descrip- tion of a new species, and then implicitly designated only 1 or 2 types from that ma- terial. Dahms (1983) considered as types all specimens stated by Girault to have been used by him in a description. Dahms' practice is followed herein. Dahms (1984, 1986) and Hayat and Ver- ma (1980) catalogued Girault's material of Eiiryischomyia in detail, obviating the ne- cessity to do so herein. The information contained in the lists of specimens exam- ined has therefore been simplified. Where appropriate, label data are enclosed in quotation marks; added or interpreted in- formation is enclosed in square brackets. Latitudes and longitudes have in most cases been provided by the author. The terminology adopted is that of Nau- mann (1991), except for the following: An- tennal formula: representation of the num- ber of segments comprising, respectively, the scape, pedicel, anellus, funiculus, club. Postaxilla: a small triangular sclerite pos- terior to, and apparently a division of, the mesonotal axilla; a term proposed by Sha- fee (1975). Parastigmal spur: a posterobas- ally orientated extension of the parastig- mal vein, bearing 1-3 stout, bristles. Abbreviations used include the follow- ing: ANIC, Australian National Insect Col- lection, Canberra; BMNH, British Museum (Natural History), London; QM, Queens- land Museum, Brisbane; SAM, South Aus- tralian Museum, Adelaide; USNM, United States National Museum of Natural His- tory, Washington DC; WADA, Western Australian Department of Agriculture, 66 Journal of Hymenoptera Research South Perth; ZMA, Zoological Museum, Aligarh. Euryischomyia Girault Euryischomyia Girault, 1914: 285. — Girault 1915a: 178; Girault 1930: 1; Shafee 1975: 27; Hayat & Verma 1980: 35; Hayat 1983: 87; Dahms 1986: 689; Viggiani 1988: 101; Shafee & Rizvi 1990: 103. Euryisdiomyiia [sic] Girault, 1929: 331. Paramyiocnema Girault, 1917: 93. — Hayat 1994: 85. Type species Euryischomyia washingtoni Gi- rault, by monotypy and original designation. The following characters, in combina- tion, are diagnostic for Euryischomyia: Head with 7 pairs of long, stiff setae near ocelli and inner orbital margins; max- illary and labial palps 2-segmented; man- dibles tridentate (or, depending on inter- pretation, bidentate, the wide dorsal tooth being apically concave and truncate); se- tose. Antennae: scape unexpanded in both sexes; 1-2 anelli; funicle 3-segmented; club 3-segmented. fronotum with row of long, stiff setae, 3 on each side, posterosublaterally. Me- soscutum between notauli with 2-3 rows of shorter, finer setae anterior to row of 4 and row of 2 longer, stiffer setae; scutel- lum wider than long, with 2 pairs of long, stout setae, otherwise bare. Postaxillae present, bare. Phragma evenlv tapered, apically incised. Metanotum not produced posteromedially. Propodeum wide, not excavated posterolaterally. Fore wings hyaline or medially infus- cated; parastigmal spur little- or well- developed, with 1-2, rarely 3, long, bris- tle-like setae of similar size to those on submarginal vein; 0-2 bristle-like setae posterobasal to parastigma, or area pos- terior to submarginal vein and proximal of parastigmal spur with 6-13 shorter, stiff setae in 2-3 irregular rows; this area of disc otherwise bare; discal setation proximally regularly or irregularly mar- gined. Legs: hind coxa enlarged, elongate- ovate, bilaterally compressed ventrally; hind femur elongate, broad, laterally com- pressed, laminose and arcuate in section; fore and mid femora also broad and lat- erally compressed. Only one hind tibial spur. Long, bristle-like setae on legs most commonly disposed in differently sized pairs (as subapically on coxae) or in lon- gitudinal rows (as on mid and hind tibi- ae). Petiole wide, boomerang-shaped. Bases of cerci low, tuberculate. Notes. — Euryischomyia is most closely re- lated to (the monotypic) Myiocnema, which differs in possessing 2 hind tibial spurs, 3- segmented maxillary palps, 3 anelli, and a distinctive fore wing chaetotaxy; the costal cell bears numerous short, stiff setae, and the basal discal area bears both a very long bristle-like seta posterior to the par- astigmal spur and several irregular rows of shorter, stiff setae in the angle between the submarginal vein and the parastigmal spur. The chaetotactic differences may be of specific value onlv. Studies in progress by J.B. Woolley and M. Hayat show that Myiocnema differs significantly from Eu- ryischomyia in not possessing a looped me- sofurca (Hayat 1994). Euryischia is very distinctive in possess- ing almost circular, disc-like hind coxae, a posterolaterally excavated propodeum to accommodate the coxae, and a postero- medial metanotal process ('elasmid' char- acteristics). Euryischia is further distin- guishable from Euryischomyia in possess- ing 2 hind tibial spurs, 4-segmented max- illary palps, and cylindrical cereal bases. The scape of the male of some species of Euryischia is greatly enlarged, and the chaetotaxy of the mesonotum and fore wing of Euryischia is interspecifically vari- able. Hayat and Verma (1980) provided keys to the genera. Taxa of Euryischiini clearly demonstrate different degrees of development of eria- porine and euryischiine features, namely, development of the parastigmal spur; spi- Volume 4, 1995 67 nosity of the legs, fore wing and body; en- thorax, Myiocnema comperei Ashmead and largement and compression of the hind Euryischia spp. are successively more en- legs and concomitant modifications of the dowed in these respects, mesosoma. Euryischia washingtoni, E. flavi- KEY TO SPECIES OF EURYISCHOMYIA GIRAULT 1. Parastigmal spur well-developed, bearing 2, rarely 3, long, bristle-like setae. Area imme- diately posterior to submarginal vein and proximal to parastigmal spur with 2-3 irregular rows of 6-13 relatively small, stiff setae. Subapical margin of costal cell with 4-7 short, stiff setae anterior to cluster of fine, ventral setae; otherwise bare. Fore wings hyaline, not narrow. Mesoscutum with 11-21 setae in 2-3 irregular rows anterior to row of 4 and row of 2 bristle-like setae. Body coloration yellow, orange, brown and black. Distribution: Australia £. flavithorax Girault - Parastigmal spur scarcely developed, bearing 1-2 bristle-like setae. Area posterobasal to parastigma bearing 0-2 large, bristle-like setae. Costal cell ventro-subapically with some pale, fine setae; otherwise bare. Forewings narrow, 3.5 times longer than width; medially with broad band of infuscation. Mesoscutum with only 8 setae anterior to row of 4 and row of 2 setae. Body coloration predominantly black, legs predominantly lemon-yellow. Distribution: Australia and India E. washingtoni Girault Euryischomyia washingtoni Girault Euryischomyia washingtoni Girault, 1914: 285. — Girault 1915a: 178; Hayat & Verma 1980: 37; Dahms 1986: 649. Euryischomyia alami Shafee, 1975: 27. — Hayat & Verma 1980: 37. Syn. nov. Type Material Examined. — Euryischomyia washingtoni Girault: Lectotype 9, paralec- totype 6 (both on 1 slide, also parts of ho- lotype of Pleurotropomyia seditiosus Gi- rault; all separate); paralectotype 9, (on tag, minus head). (QM 4060; [Hy]2742). QUEENSLAND: 20°32'S 145°24'E Cape- ville (Pentland), by sweeping in forest along the banks of Cape River, 8. i. 1913 (Girault 1914). Other Specimen Examined; Identified by A. A. Girault. — "Euryiscliomyiia [sic] wash- ingtoni" 9 (on tag and on slide; slide also contains a paratype of Amicromelus gran- daevus Girault). QUEENSLAND: 17°23'S 145°19'E Watsonville (QM). Other Records. — Euryischomyia washing- toni: QUEENSLAND: 19°15'S 146°48'E Townsville, sweeping in forest, 19. i. 1913, 9 (Girault 1915a). Euryischomyia alami: ex Nipaecoccus viri- dis (Newstead) on Mangifera indica, MY- SORE: Bangalore, Channapatna, 2.xii.l968, holotype 9, paratype 9, S.A. Shafee (Sha- fee 1975) (ZMA). Euryischomyia alami: ex Nipaecoccus viri- dis, TAMIL NADU: Shencottah, 2 9 9, M. Hayat (Hayat and Verma 1980) (ZMA). Redescription. — Coloration of 9 : Body shiny, blackish brown to black; mesono- tum obscurely dark green metallic. Body setae mostly pale. Legs whitish except for following: fore coxae dusky basally; mid and hind coxae black except apically; mid and hind femora black except apically and basally; tarsi brown apically; sockets of hind tibial bristles black. Fore wings each with broad band of pale to dark infusca- tion extending from anterior to posterior margins in region of marginal and post- marginal veins, otherwise hyaline. Morphology: Length mesosoma + me- tasoma: 0.90 mm (paralectotype 9); 1.05 mm (lectotype 9); 1.40 mm (9 t'.v Watson- ville); 0.84 mm (paralectotype 6). All bris- tle-like setae smaller and finer than cor- responding ones of £. flavithorax. 68 Journal of Hymenoptera Research Head: submedian frontal setae (of 6): (48|JLm); interocellar setae: (68) ixm. Max- illary and labial palps longer and more slender than in E. flavithorax. Dorsalmost tooth of mandibles pointed. Antennae of 9 ex Watsonville (and S): antennal formula: 1:1:1-2:3:3; relatively longer and more slender than in E. flavi- thorax. Scape: length 5.56 (4.44) X maximal width; 1.82 (1.73) X length of pedicel; 1.22 (1.11) X length of club. Pedicel: 2.5 (2.28) X maximal width. Anellus small, wedge- shaped; a 2nd anellus may be partly de- limited from it. Funicular segments mostly as long as or longer than wide; segment 1 (fl): length, 1.18 (1.0) X width; f2: length, 1.41 (0.94) X width; f3: length 1.04 (1.14) X width. Club: length: 2.56 (3.56) X width. Lengths of segments: scape, 150 (107) |xm; pedicel, 83 (62) fxm; fl: 30 (20) fxm; f2: 47 (21) |xm; f3: 41 (30)|jLm; club: 123 (96) jim. Mesosoma of lectotype 9 (and 6): Pro- notum: 6 bristle-like setae postero-subla- terally, 68 (75) fxm long. Mesoscutum be- tween notauli with row of 2, and row of 6, fine, bristle-like setae, 33 (26-29) |xm long; row of 4 (38) iJim long; and row of 2 thicker, bristle-like setae (51) ixm long. 2 midsublateral and 2 posterolateral scutel- lar setae, (45 ixm) and 90 (93) fxm long, respectively. Axilla with 2 setae; posterior seta (34) ixm long. Legs of lectotype 9 (and 6): Hind coxa, 0.21 (0.22) mm long, 1.87 (1.74) X width, slightly shorter than hind femur, which is shorter than hind tibia; subapical setae of hind coxa: 30 (36) and 48 (54) |xm long; hind femur, 0.23 (0.26) mm long, 2.53 (2.74) X width, subapical seta, 26 (30) |xm; hind tibia, 0.31 mm long, gradually broad- ening apicad, preaxial row of setae maxi- mally 50 |xm long, subapical setae 44 and 62 fxm long, spur 48 jxm long. Fore wings of lectotype 9 (and 6 ) nar- row, 3.5 X width. Parastigmal spur hardly developed, bearing 1-2 bristle-like setae, 45 |JLm long; 1 or 2 bristle-like setae pos- terobasal of parastigma, 31 (35) ixm and (18) fxm long, respectively. Subapical area of costal cell devoid of marginal setae. Discal setation with regular margin. Metasoma in dry-mounted 9 9 up- turned at an angle of approximately 60°, strongly tapered apicad. Ovipositor sheaths 0.18 mm long, 0.59 X hind tibial length, slightly protrusive beyond apex of metasoma, densely and evenly furnished with short, bristle-like setae. Notes. — Based on a knowledge of the chaetotactic variability within the tribe, £. alami is hereby placed in synonymy with E. washingtoni, despite the fact that the types of E. alami have not been seen by the pres- ent author. £ alami reportedly differs from £. zvashingtoni in not bearing discal setae posterior to the parastigma, and in possess- ing 3 setae anteriorly on the axilla (Shafee 1975; Hayat and Verma 1980). These differ- ences are too small to warrant separation at the species level. Hayat and Verma (1980), who studied the above-listed specimens from Tamil Nadu, are also of the opinion that the differences between the 2 taxa ap- pear to be of doubtful specific value. The above description of £. zuashingtoni supplements those of Girault (1914, 1915a) and Hayat and Verma (1980) of £. wash- ingtoni and that of Shafee (1975) of £. ala- mi. Because of the paucity and poor con- dition of the material available for study, a composite and comparative method of description is adopted. £. flavithorax, the standard of comparison, is conventionally described below in greater detail. The lem- on yellow coloration described by Girault (1914, 1915a) can be presumed to have faded with time to white. The antenna of the paralectotype 9 (on slide) is com- pressed; measurements and ratios of the 9 ex Watsonville are given instead. £. washingtoni has less developed eu- ryischiine features than the other taxa of Euryischiini, having a scarcely developed parastigmal spur, and shorter, finer bris- tle-like setae on the body, limbs and fore wings. Volume 4, 1995 69 ,.J^. I I'tnAniri Fig. 1. Euryischomyia flavithorax, male (X67). Cleared, slide-mounted specimen. Biology. — Parasite of unknown status of Nipaecoccus viridis, Pseudococcidae. Distribution.— AXJSTR ALIA: Queens- land; possibly also Beenleigh, Victoria (Viggiani 1988); INDIA: Mysore, Tamil Nadu. Euryischomyia flavithorax Girault & Dodd (Figure 1) Euryischomyia flavithorax Girault & Dodd, 1915 in Girault 1915a: 178.— Dahms 1984: 614. Euryischomyia saintpierrei Girault, 1915a: 178. — Hayat & Verma 1980: 37; Dahms 1986: 495. Syn. nov. Myiociicma iimnuorativentris Girault, 1915b: 64.— Dahms 1984: 795. Syn. nov. Euryiscliomyia fasciata Girault, 1916: 212. — Hay- at & Verma 1980: 37; Dahms 1984: 579. Syn. nov. Paramyiocnema marmorativentris (Girault). — Gi- rault 1917: 93; Hayat & Verma 1980: 35. Euryischomi/i[i]a sctosa Girault, 1929: 331.— Hay- at & Verma 1980: 37; Dahms 1986: 522. Syn. nov. Paramyiocnema flavithorax (Girault & Dodd). — Hayat & Verma 1980: 33. Type Material Exmuiued. — Eiin/isdioun/ia flavithorax Girault & Dodd: Lectotype 9 (on slide); paralectotype 9 (on tag; head and hind tibia on same slide as lectotype, herein called paralectotype 1); paralecto- type 9 (also on same slide as lectotype, herein called paralectotype 2). QUEENS- LAND: 17°05'S 145°47'E "Gordonvale (Cairns)" (QM 3903; Hy2744). Euryischomyia saintpierrei Girault: Holo- type 9 (torso and 1 fore wing on tag, fore wing separate, not listed by Dahms (1986); head, funicular segments 2 and 3 and club, both hind legs on slide) (QM 3905; Hy2743). QUEENSLAND: 27°28'S 153° 02'E Brisbane (Girault 1915a). Myiocnenia marmorativentris Girault: Ho- lotype 9 (on slide) QUEENSLAND: 17°05'S 145°47'E "Gordonvale (Cairns)" (QM 3907; Hy2959). Euryischomyia fasciata Girault: Holotype 70 Journal of Hymenoptera Research 6 (not ?, as described by Girault; on tag minus antennae, right wings and leg parts; antenna minus scape, fore wing, hind femur and hind tibia on slide; anten- na not listed by Dahms (1984)). WESTERN AUSTRALIA: 31°57'S 115°51'E "Perth, W. Austr., G. Compere, Collector, 837" (USNM 19679). Euryischomyiia [sic] setosa Girault: Holo- type 9 (on tag, minus head, with 1 fore wing on slide) SOUTH AUSTRALIA: 32°49'S 138°11'E "Melrose, Oct., A.M. Lea" (SAM). Other Specimens Examined: identified, as follows, by A.A. Girault. — "Myiocnema flav- ithorax Gir. [sic], brown scale [Coccus hes- peridum] parasites, [WESTERN AUSTRA- LIA:] 9, [2]6S, coll. L.J. Newman, 70". (On 1 slide. Host identified as Coccus hes- peridum in F. Wilson (I960)). (WADA). Euryischomyiia [sic] flavithorax Gir. [sic], 9 . Host Lecanidae [Coccidae] sp. Ent. Div. Dep. Ag. & Stk Qld." (QM Hy 10). "Myiocnema saintpierrei Gir., S , 9 . Host Saissetia sp., Sydney [33°53'S 151°12'E], N.S.W., June 19-31, S.E. Flanders. Univ. Cal. Citrus Exp. Sta. Ace. No. 27)." (On slide; only 2 of 6 specimens easily observ- able). (QM). "Myiocnema setosa Gir., 9, Myiocnema saintpierrei Gir., 6 , 9, on lime scale [Coc- cidae?], [QUEENSLAND: 26°45'S 150°- 38'E] Chinchilla, Aug. 14, 1931, S.E. Flan- ders". (On slide; only 29 9, 26 S easily observable) (QM). Other Specimens Examined. — QUEENS- LAND: 15°39'S 144°31'E Split Rock, 28.v.- 26.vi.1993, 9, P. Zborowski, I.D. Nau- mann (ANIC). WESTERN AUSTRALIA: 33°45'S 122° 32 'E Condingup, 55km E of Esperance, 30.xii.l986, 6, J.S. Noyes (BMNH). ex Psyllaephagus sp. / Cardiaspina albitex- tura Taylor on Eucalyptus blakelyi, VICTO- RIA: nr 36°21'S 146°19'E Wangaratta, ll.iLl955, 39 9,76 6, E. Lewis (ANIC). ex Cardiaspina sp. on Eucalyptus blakelyi, AUSTRALIAN CAPITAL TERRITORY: 35°18'S 149°08'E Canberra (site 15a), 3.ii.l954, 2 9 9, A. Magassy. NEW SOUTH WALES: 35°50'S 147°15'E Woomargana, 3.xii.l954, 9, E. Lewis (all in ANIC). ex Creiis sp., WESTERN AUSTRALIA: 31°59'S 115°52'E Western Australian Insti- tute of Technology, South Perth, 1982, 9, S.J. Curry (ANIC). ex Aphidius colemani Viereck / Br achy - caudus helichrysi (Kaltenbach) on safflow- er, NEW SOUTH WALES: 29°14'S 149°51'E 25km N. of Moree, 12.ix.l993, 3 9 9, Id, J. Stanley (ANIC). ex Aphidius colemani / Myzus persicae (Sulzer) on nectarine, AUSTRALIAN CAPITAL TERRITORY: 35°15'S 149°05'E Aranda, Canberra, 15.xi.l991, 29 9, P.J. Hart (ANIC). ex Trioxys complanatus Quilis / Therioa- phis trifolii (Monell) f. maculata on lucerne, VICTORIA: 36°25'S 145°56'E Goorambat, 1979, 9, P. Ridland; 36°33'S 145°59'E Ben- alla, 9.ii.l978, 9, 26 6, L.T. Woolcock; SOUTH AUSTRALIA: 34°40'S 138°34'E Virginia, 21.ii.l980, 6, C. Wilson; 34°56'S 138°36'E Northfield, Adelaide, 20.iii.l978, 8.ii.l979, 49 9, 36 6, L.T. Woolcock; 35°41'S 139°20'E Meningie, 27.ii.1979, 6, C. Wilson (all in ANIC). ex Aphelinus mali (Haldeman) / Eriosoma lanigerum (Hausmann) on apple, NEW SOUTH WALES: 30°31'S 151°40'E Armi- dale, December 1992, 6 9 9, S. Asante (ANIC). ex Pseudococcus sp., on Acacia melanoxy- lon, VICTORIA: 37°42'S 145°04'E LaTrobe University, Bundoora, Melbourne, l.xii.- 1980, 29 9,26 6, G. Farrell (BMNH). Other records.— QUEENSLAND: 15°28'S 145°15'E Cooktown, in forest, 2.iii.l914, 9, A.P. Dodd (Girault 1915a; Dahms 1984). WESTERN AUSTRALIA: 27°41'S 114°- 33 'E Kalbarri National Park, December 1986, 9, J.S. Noyes; 31°36'S 116°13'E Avon Valley National Park, 50km NE of Perth, 25.xii.1986, 9, J.S. Noyes; 31°43'S 116°04'E Walyunga National Park, 40km NE of Perth, 24.xii.1986, 6, J.S. Noyes (BMNH) (Identified by L.D. Coote). ex Ceroplastes rubens Maskell on Schef- Volume 4, 1995 71 flera actinophi/lla, QUEENSLAND: 27°30'S Morphology: Length mesosoma + me- 153°01'E Brisbane, May 1980, 46 6, J.S. tasoma, 0.7-L3 mm long (mean, LOl mm; Noyes (BMNH) (Recorded as Paramyioc- n = 2Q; lectotype, L06 mm). Integument nema sp. by Viggiani (1988)). variously reticulate-imbricate, except that Redescription of Female. — Coloration: of most of metanotum and propodeum. Most body setae dark. Head yellow, or- which is smooth; imbrication on tibiae ange or tawny except for following: finely setose at interstices. Large setae Ion- brown, linear macula suborbitally on gitudinally ridged. gena; small, brown macula surrounding Head: 2 pairs of large, stiff, bristle-like posterior half of median ocellus and an- setae on frons near margins of eyes; 1 sub- terior half of each lateral ocellus; exten- median pair between these (lOOixm long in sive, broad, brown to black, inverted U- lectotype); 1 pair lateral to median ocellus; shaped macula in region adjacent to pro- 1 pair near margin of eyes anterolateral to thorax; mandibles apically testaceous. An- lateral ocelli; 1 pair between lateral ocelli tennae yellow, scape may be dusky. Eyes (113 ixm); 1 pair posteromedial to lateral red; margin and, internally, medial half of ocelli. Dorsalmost tooth of mandible usu- eye socket darkly sclerotized (conspicuous ally slightly blunt, only in cleared specimens). Ocelli red. Antennae inserted not far distant from Mesosoma, orange or tawny except for clypeal margin, the distance subequal to following dark sclerotization: extensive, distance between insertions; setae short, brown to black macula on prothorax ad- with stouter, stiffer hairs concentrated on jacent to occipital region of head; notauli one side; antennal formula: 1:1:1-2:3:3. anteriorly, boundaries of anterior angles Scape: length, 2.9-3.4 X width (mean, 3.2); of axillae, anterior and posterior bound- 1.7-2.3 (2) x length of pedicel; 1-1.2 (1.1) aries of pronotum, scutellum and meta- X length of club. Pedicel: length, 1.3-2.5 notum mav be brown; 1-2 small, brown (1.8) X width; 0.6-1 (0.75) X length of to black maculae near base of fore wing; anellus(i) + funicle. Anellus 1 large sometimes 1 macula at base of posterolat- enough to be considered funicular, finely eral setae of scutellum and submarginally setose but non-sensoriated; a 2nd, very on metanotum; propodeum light to dark small, anellus sometimes variouslv devei- brown anteriorly and on each lateral third oped on half of apex of anellus 1, mav be except for spiracles, lateral margins may setose. Funicular segments all wider than be darker, or whole segment may be long, becoming progressivelv wider and brown and laterally black. Ventrally and longer apicad; segment 1 (fl) length, 0.58- sometimes pleurally, mesosoma light 0.87 (0.7) X width; f2: length, 0.47-0.82 brown to dark brown. (0.74) X width; f3: length, 0.64-0.9 (0.79) Legs yellow except for following: fore X width; length of anellus (i) + funicle, femur with linear macula postaxially; mid 0.6-0.9 (0.7) x club. Club 3-segmented; coxa may be slightly dusky basally; hind segments separable by segmentally ar- coxa brown to dark brown except usually ranged setal and sensorial patterns, not bv apically, sometimes also dark brown mar- sutures; length, 1.5-2.7 (2) X width; width ginally; hind femur sometimes streakily 1-1.8 X width of f3. Lengths of segments: dusky; sockets of hind tibial bristles black, scape, 99-135 fim (mean, 117 |xm; n = 20); Metasoma light brown to black; petiole pedicel, 45-68 (58) |xm; fl, 14-24 (17.5) usually brown, often black on lateral mar- |xm; f2, 12-37 (24) |xm; f3, 24^1 (31) |jim; gins; base of gaster medially and gastral club, 93-123 (107) |xm. tergites 2-4 usually darker than rest of Mesosoma: Pronotum with row of 3+3 gaster, and fasciate, especially under high long setae (115 ixm long in lectotype) pos- magnification. terosublaterally. Mesoscutum anteriorly. 72 Journal of Hymenoptera Research between notauli, with 11-21 fine, stiff se- tae (41 |jLm) in 2-3 irregular rows ; a mid- dle transverse row of 4 long, stout setae (98 fxm), and posterior row of 2 long, stout setae (90 |xm). Scutellum with 2 long setae anterosublateral of midpoint (87 ixm) and 2 very long, stout setae posterolaterally (168 |xm). Axillae each with long, stout seta (84 |xm) anteriorly and 2-3, rarely 4, shorter, differently sized setae (26-36 ixm) near posterior angle. Mesosomal setae of smaller specimens shorter and less stout. Legs bearing long, stiff setae, mostly in rows or in differently sized pairs; other se- tae mostly disposed dorsoapically on fe- mur and longitudinally on tibia; Fore coxa (of 1 Northfield specimen): pair of long, stiff setae subapically (39 and 57 ixm long); fore femur: subapically, 1 seta ventrally (29|jLm), 1 dorsally (45 lam); fore tibia ap- pearing fusiform, calcar plumose (57 ixm), 3 small, stiff setae apically (Slfxm), and row of 7 setae (maximally 31 ixm) preaxi- ally. Mid coxa: pair of setae subapically (42 and 57 lam); mid femur: very long pair subapically (65 (xm ventrally, 78 |xm dor- sally); mid tibia: spur long (87 |xm), plu- mose, 1 smaller seta near spur (18 |xm), 1 subapically (53 lam), preaxial row (maxi- mally 32 (xm) hardly differentiated from other setal rows. Hind coxa (0.22 mm long), shorter than hind femur (0.28 mm), which is slightly shorter than hind tibia (0.32 mm) and subequal to hind tarsus (0.27 mm); hind coxa large, elongate- ovate, length 1.44-1.69 X width, bilater- ally compressed ventrally, pair of setae subapically (44 and 66 ixm); hind femur broad, approximately parallel-sided, length 2.83-3.14 X width, 1 seta subapi- cally (58 |JLm); hind tibia, basally narrow, gradually and slightly widening apicad, spur almost smooth (54 |xm long), pair of very long, stout setae subapically (68 and 89 |jLm) and 1 longitudinal row of 8 very long, stout setae (maximally 75 iJim). Mid basitarsus longer than fore or mid basitar- sus, but not conspicuously long. Fore wings: 6-13 (mean, 10) moderately long and fine, stiff setae (30^0 |xm long in 1 Northfield specimen) present in 2-3 rows below submarginal vein and proxi- mal of parastigmal spur; 2-3 parastigmal setae (70-76 |xm); costal cell in distal third with 4-7 short, stiff, marginal setae (25 fxm) anterior to cluster of fine, stiff, ventral setae; margin of discal setation irregular. Metasoma not upturned or strongly ta- pered: Petiole boomerang-shaped except for truncated ends. Variously sized, stiff setae metamerically disposed laterad and apicad on dorsum of gaster. Ovipositor sheaths 0.10-0.11 mm long, 0.23-0.33 X hind tibial length, each fur- nished with about 10 mostly subapical se- tae. Description of Male. — Similar to female. Colour pattern similar but yellower be- cause of less extensive dark sclerotization. Linear macula usually present, anterome- dially, on vertex; suborbital, linear macula of gena absent and U-shaped macula of occipital region absent. Mesosoma later- ally pale. Gaster: dark sclerotization con- fined to narrow area adjacent to petiole and to medially coalesced fascia on gastral tergites 2-4, extending medially on gastral tergite 5 and apicolaterad. Antennae (1 specimen): scape: length, 111 |xm, 3.7 X width, 2.18 X length of ped- icel, 1.19 X length of club; pedicel: length, 51|xm, 1.7 X width, 0.71 X length of anel- lus + funicle; funicular segment 1 (fl): length, 12|jLm, 0.47 X width; f2: length, 21|jLm, 0.66 X width; f3: length, 27 |am, 0.66 X width; anellus + funicle: length, 0.77 X length of club; club: length, 0.93 |jLm, 1.32 X width. Hind coxa 0.24 mm long, twice as long as wide; hind femur 0.24 mm long, 3.2 X width; hind tibia 0.3 mm long. Mesosoma + metasoma 0.9 mm long. Anterior mesoscutal setae, 33^5 juim long; mid and posterior setae, 70 and 81 ixm long respectively. Anterior and posterior scutel- lar setae 64 and 121 \xm, respectively; only 1 posterior axillar seta, 24 |jLm long. Notes. — The type material of £. flavitlior- Volume 4, 1995 73 ax comprises the slide-mounted lectotype as possessing 2 anelli, which may have designated by Hayat in 1979 (Hayat and been the reason why he considered it a Verma 1980), and paralectotype 1, with species distinct from the other species of parts on a tag and parts on the same slide Euryischomyia, and why he identified spec- as the lectotype but mounted separately, imens from 'lime scale' as 2 species, M. inbetween the lectotype and paralectotype saintpierrei and M. setosa (see above list of 2 (see list of type material examined). Par- specimens identified by Girault). These alectotype 2 is the specimen stated by specimens, on 1 slide, have antennae each Hayat and Verma (1980) to be a species with either 1 or 2 anelli but are otherwise different from the lectotype, resembling similar to one another. Most of the speci- marmorativentris, and probably not part of mens of £. flavithorax examined in this the original type material. Consequent study have both antennae with 1 anellus; upon the above statements, the specimen others have 2; or 1 antenna may bear 1 was referred to by Dahms (1984) as an un- anellus and the other bear 2; or 1 small identified chalcidoid. Paralectotype 2 is, anellus may be partly delimited from a without doubt, Euri/ischoniyia flavithorax. larger one. Smaller and yellower than the lectotype, it Distribution. — AUSTRALIA: Queens- does indeed resemble the holotype Myioc- land. New South Wales, Australian Capi- nema marmorativentris, which was collect- tal Territory, Victoria, South Australia, ed in the same locality a week or so later. Western Australia (Figure 2). The female collected by A. P. Dodd at Biology. — E. flavithorax is not highly Cooktown has still not been located (see host-specific, its recorded hosts being chal- list of other records above). It may be part cidoid and ichneumonoid Hymenoptera, of Dodd's collection. and psylloid, aphidoid and coccoid He- £. flavithorax was originally described miptera Sternorrhyncha (Table 1). E. flav- from 3 swept specimens, and each of the ithorax was reared occasionally from im- synonymized species was described from mature Aphidiinae and Aphclimis mali, en- single specimens. The relevant parts of the doparasitic in aphidine and drepanosiphi- descriptions are based mainly on colora- ne, and pemphigine aphids, respectively, tion. Other characters described by Girault as part of a long-term study by the author are supra-specific or are not interspecifi- of aphids and their insect natural enemies; cally comparable. However, all of them fit and, also occasionally, from Psyllacphagus into the above conception of £. flavithorax. sp(p). parasitic in lerp-forming psyllid £. flavithorax is variable in coloration. In nymphs, as part of an ecological study by general, smaller females are yellower and L.R. Clark in 1950-74 of psyllids associat- less darkly sclerotized than larger ones, ed with eucalypts (Riek 1962). and males are smaller and yellower than £. flavithorax, however, is not a true hy- females. The differences are a reflection of perparasite of aphids if, like certain other differences in extent of dark sclerotization. chalcidoids and proctotrupc^ids {e.g. Pachy- The pattern of coloration, however, is cc:)n- neuron spp. (Pteromalidae) and Dcndroce- stant, and is best expressed in the lighter- rus spp. (Megaspilidae) (Takada 1973a,b)), coloured specimens. The lectotype of £. it does not parasitize the hymenopterous flavithorax, for instance, belongs near the host until after the latter has consumed all large, dark end of the scale of variability, of its aphid host except for the dead, dry whereas paralectotype 2 of £. flavithorax aphid skin (mummy); i.e. if it parasitizes and the holotype M. marmorativentris be- the hymenopterous contents of a capsule long near the small, pale end. The small, it is strictly a primary parasite. Such a pale holotype £. fasciata is a male. probability is supported by its broad host Girault (1913a) described £. saintpierrei spectrum; parasitizing live hosts of such 74 Journal of Hymenoptera Research ■ y c rt -eadon, and Eiicerceris (Evans and O'Neill 1988; O'Neill 1990). In this mating system, males apply volatile secretions originating from mandibular glands to a substrate with clypeal brushes, and defend these non-resource based territories from con- specific males. Philanthine females typi- cally nest in dispersed conditions, mate once at the beginning of the nesting period away from the nest, presumably are at- tracted to mandibular gland secretions, and enter territories to copulate (Evans and O'Neill 1988). The mating system of C. binodis departs from these conditions. Females nest in aggregations, and mate multiply at the nest. Resident C. binodis males stand on or directly adjacent to nests, exclude conspecific non-resident males from nests, and do not scent mark territories. Of the eight philanthine gen- era, male nest defense has been described only in Philanthiis, and presumably is de- rived within that genus (Evans and O'Neill 1988). Male nest defense is de- scribed here for the first time in the genus Cerceris, and also presumably represents a derived condition within this taxon. METHODS An aggregation of nesting C. binodis fe- males was discovered on Barbour Lathrop trail on Barro Colorado Island, Republic of Panama on July 8, 1994. A group of five nests that were close enough together to be observed simultaneously were marked with numbered tags. Male and female wasps were individually marked with enamel paint on the thorax, and head widths were measured to the nearest 0.05 mm with Manostat calipers. The nests were observed on two days after individ- uals were marked. Observations were made from 1 355-1 525h on the first day, and from 1515-1650h on the second day. The behaviors recorded were the amount of time that males spent standing on or adjacent tc^ nests, defensive interactions between males, and copulations. 78 Journal of Hymenoptera Research RESULTS When a female returned to a nest with prey, and a resident male was present, she would land on the nest tumulus, and cop- ulate. However, a female would fly di- rectly into her nest if no male was present. This behavior is also described in the sphecid genus Oxybelus (Crabroninae) (Peckham et. al. 1973). Females were pro- visioning nests in all cases where males defended nests and mated. Not only did males defend nests while females were on provisioning trips, but on two occassions, a male was observed to defend a nest while the female was inside the nest. These males were observed to bite at the females, and attempted to pull them out of the nest with their mandibles. One of these males mated with the female when she exited the nest. Nest defense duration was determined by the length of time a resident male was either in contact with a nest, directly ad- jacent to a nest, or defending a nest from non-resident males. On day one, males YB, RB and BY, spent a total of 66.7, 21.1, and 12.2 percent of the observation period defending nests, respectively. On day two, males WRB, YB, RY, BB and RB spent a total of 67.3, 41.1, 16.8, 1.2 and 0.8 percent of the observation period defending nests, respectively. Males YB and WRB obtained 71% (5/7) of the observed copulations, and there is a positive correlation between the duration of time which males defend nests and the number of copulations that males obtain; r = 0.86. At times, multiple males were simulta- neously defending different nests in the aggregation. Male-male interactions were considered a territorial defense when a resident male flew from his nest to chase a non-resident male. These chases rarely ended in grappling on the ground; other- wise, there was no physical contact. Four- ty eight aggressive interactions were re- corded between resident and non-resident males. Resident males won significantly more (95%) interactions than non-resident males (Chi-square = 40.1, df = 2, P< 0.001). In 11 of the 48 interactions, identity and size of both males were known; resi- dent males won 7 of these 11 interactions. Mean head widths of winner and looser males were 3.83mm and 3.76mm, respec- tively. The difference in mean headwidths between winner and looser males is not statistically significant (t-test = .82, df = 20, P > 0.4). DISCUSSION Size has been demonstrated to deter- mine dominance in territorial sphecids such as Philanthus (O'Neill 1983a, b), and in many other insect mating systems (Thornhill and Alcock 1983). In C. binodis, however, territory ownership, not size, ap- parently determines dominance (one should note the small sample size (n=ll) where identity of both males was known). This phenomenon has also been demon- strated in other insect mating systems (Da- vies 1978, Eickwort and Ginsberg 1980, Wang and Greenfield 1991). Data also in- dicate that as the total time invested in nest defense per male increases, the num- ber of copulations obtained per male also increases. This is not surprising, since fe- males mate with the resident male which is defending her nest. In many Cerceris species, females nest in aggregations, but males do not defend nests from conspecific males. The require- ment which is likely responsible for the difference between the mating systems of these species and C. binodis is that C. bin- odis females retain their receptivity after the initial mating. This might maintain male territoriality, particularly if there is last male sperm precedence (Hook and Matthews 1980, Brockmann and Grafen 1989). However, males in several species of Philanthus defend territories which con- tain nesting females that do not display continuous receptivity (Gwynne 1980, Evans and O'Neill 1988). Data collected from this study do not indicate that fe- Volume 4, 1995 79 males are continuously receptive through- out the nesting period, only that they mate multiply. In addition, last male sperm pre- cedence is not demonstrated in this spe- cies, but often occurs in insects (Thornhill and Alcock 1983). CONCLUSION This is the first description of male nest defense in Cerceris. This derived mating system shows striking convergence with the mating systems of both Oxyhehis (Peckham et. al 1973, Hook and Matthews 1980), and Philanthus (Evans and O'Neill 1988) species. Moreover, C. binodis males do not scent mark territories, and their clypeal brushes are reduced. Since clypeal brushes are a synapomorphy of Philanthi- nae, and are present in most species of Cerceris (Alexander 1992), reduction of the clypeal brushes in C. binodis represents a derived condition. Evans and O'Neill (1988) state that scent marking occurs in all territorial philanthines where the mat- ing systems have been described. There- fore, this mating system represents an ex- ception to their statement (see also Alcock and Gamboa 1975). ACKNOWLEDGEMENTS Helpful comments and suggestions were provided by B.A. Alexander, W. T. Wcislo, D. Windsor, R. L. Minckley, and four anonymous reviewers. This re- search was funded by a Smithsonian Tropical Re- search Institute 10 week graduate fellowship, and by a Sigma Xi research grant. LITERATURE CITED Alcock, J. 1975a. Male mating strategies of some phi- lanthine wasps(Hymenoptera: Sphecidae). Jour- nal of the Kansas Eiitoniological Society 48: 532-545. Alcock, J. 1975b. Territorial behavior by males of Phi- lanthus multimaculatus (Hymenoptera: Sphecidae) with a review of territoriality in male Sphecidae. Animal Behavior 23: 889-895. Alcock, J. and G. Gamboa. 1975. Home ranges of male Cerceris simplex macrosticta (Hymenoptera: Sphecidae). Psyche 81: 528-533. Alcock, J., C.E. Jones and S.L. Buchmann. 1977. Male mating strategies in the bee Centris pallida Fox (Hymenoptera: Athophoridae). American Natu- ralist 111: 145-155. Alexander, B.A. 1992. A cladistic analysis of the sub- family Philanthinae (Hymenoptera: Sphecidae). Systematic Entomolog]/ 17: 91-108. Bohart, R.M. and A.S Menke. 1976. Sphecid Wasps of the World: A Generic Revision. Berkely: Uni- versity of California Press. 695 pp. Brockman, H. J. and A. Grafen. 1989. Mate conflict and male behavior in a soitary wasp. Trypoxylon (Trypargilum) polifum (Hymenoptera: Sphecidae). Animal Behavior 37: 232-255. Davies, N.B. Territorial defense in the speckled wood butterfly (Pararge aegeria): the resident always wins. Animal Behavior 26: 137-147. Eickwort, G.C. and H. S. Ginsberg. 1980. Foraging and mating behavior in Apoidea. Annual Review in Entomology 25: 421-426. Evans, H.E. 1971. Obser\'ations on the nesting be- havior of wasps of the tribe Cercerini. journal of the Kansas Entomological Society 44: 500-523. Evans, H. E. and K. M. O'Neill. 1985. Male territorial behavior in four species of the tribe Cercerini (Sphecidae: Philanthinae). journal of the New York Entomological Society 93(3): 1033-1040. Evans, H. E. and K. M. O'Neill. 1988. The Natural History and Behavior of North American Bee- wolves. Comstock Publishing Associates. Hook, A. W. and R. W. Matthews. 1980. Nesting bi- ology of Oxybelus sericeus with a discussion of nest guarding by male sphecid wasps (Hyme- noptera). Psyche 87: 21-37. Gwynne, D.T. 1980. Female defense polygyny in the bumblebee wolf Pliilanthus hicinctus (Hymenop- tera: Sphecidae). Behavioral Ecolology and Socio- btology 7: 213-225. O'Neill, K.M. 1983a. Territoriality, body size, and spacing in males of the beewolf Phihvithus basi- laris (Hymenoptera: Sphecidae). Behavior 86: 295: 321. O'Neill, K.M. 1983b. The significance of body size in territorial interactions of male beewolves (Hy- menoptera: Sphecidae, Philanthus). Animal Behav- ior 3\A0i-AU. O'Neill, K.M. 1990. Female nesting behavior and male territoriality in Aphilanthops subfrigidus Dunning (Hymenoptera: Sphecidae). Pan-Pacific Entomologist 66: 19-23. Peckham, D.J., F.E. Kurczewski and D.B. Peckham. 1973. Nesting behavior of Nearctic species of Ox- ybelus (Hymenoptera: Sphecidae). Annals of the Entomological Society of America bb: 647-661. Peckham, D. J. 1977. Reduction of miltogrammine cleptoparasitism by male Oxybelus subulatus (Hy- menoptera: Sphecidae). Annals of the Entomologi- cal Society of America 70: 823-828. Thornhill, R.'and J. Alcock. 1983. The Evolution of Insect Mating Systems. Harvard Press. Wang, G. and M.D. Greenfield. 1991. Effects of ter- ritory ownership on dominanace in the desert clicker (Orthoptera: Acrididae). Animal Behavior 42: 579-587. J. HYM. RES. Vol. 4, 1995, pp. 80-98 Revised Key to the Nearctic Species of Chrysocharis Forster (Hymenoptera: Eulophidae), Including Three New Species Christer Hansson Department of Systematic Zoology, Lund University, Helgonavagen 3, S-223 62 Lund, Sweden Abstract. — The Nearctic species of Chrysocharis are keyed, altogether 63 species. The paper in- cludes the description of three new species: waJili, cornigera, sentenaca and two new combinations from Chrysocharis to Omphale Haliday: varia Hansson, gracilicornis Hansson. The Nearctic and Neotropical species of The page number after each species in Chrysocharis were revised by Hansson the revised key refers to Hansson (1987). (1987). Sixty one (61) species were treated The descriptions, diagnoses, distribution and 18 of the species were described as and hosts of Nearctic species of Chrysochar- new. From the Nearctic region 57 species is are found in Hansson (1985, 1986, 1987). were recognized. However, the kev for the identificatton of the species is difficult and RECOGNIZING CHRYSOCHARIS laborious to use, mainly because too many To facilitate the recognition of Chryso- characters are used under each item, and charis, following should be helpful. Anten- characters used in the two alternatives un- na with apical two segments fused, or der each item are not always exclusively with all flagellar segments free (exception: the same. To improve the key I have re- imhrasiis with three apical segments fused, arranged it so that the most obvious char- recognized as a Chrysocharis through the acters are used first, and under each item long postmarginal vein — 1.5X stigmal only 1-2 characters are used — only occa- vein — and antennal scrobes join below sionally are more than two characters frontal suture in female). Clypeus not de- used. The characters selected are those limited (exception: flaviclypeus, recognized easy to see and that vary as little as pos- as a Chrysocharis through female antennal sible. In some cases a species varies in scrobes that join below frontal suture), characters used, and this species is found Postmarginal vein longer than stigmal under both alternatives. vein (exception zvahli Sacornigera, recog- In the revised key I have also included nized as members of Chrysocharis through changes that has taken place since the the possession of the three abovemen- publication of the previous key: Entedon tioned characters). Without a sulcus sur- imbrasus Walker has been transferred to rounding ocellar triangle (this separates Chrysocharis from Neochrysocharis Kurdju- Chrysocharis from Derostemis). mov (Hansson in press); the species treat- ed as Zaommomyia Girault by Hansson (1986) have been transferred to Chryso- charis (Schauff 1991). The three species de- The terms used in the key and in the scribed as new in this paper are also in- text are indicated on Figs 1-8. eluded. Furthermore, two species regard- Abbreviations of collections: CNC= Ca- ed as Chrysocharis by Hansson (1987) are nadian National Collection of Insects and hereby transferred to Omphale Haliday: Arachnids, Ottawa; LUZM= Lund Uni- gracilicornis Hansson, varia Hansson, both versify Zoological Museum; TAMU= Tex- n. comb. as A&M University, College Station. TERMINOLOGY AND ABBREVIATIONS Volume 4, 1995 81 KEY TO NEARCTIC SPECIES OF CHRYSOCHARIS 1. Flagellum with all five segments free (e.g. Figs 35, 39) 2 - Flagellum with apical two or three segments fused, i.e. the constrictions between the basal funicular segments are narrower than the constrictions between the apical seg- ments (e.g. Figs 37, 38) 19 2. Forewing with a complete row of setae on underside of costal cell (Figs 48, 49) 3 - Costal cell without row of setae 4 3. Forewing speculum closed (Fig. 49); 5th flagellar segment (including the narrow tip) 1.3X as long as 4th segment (Fig. 35) C. chilo (Walker) (female, male) p. 30 - Forewing speculum open laterally (Fig. 48); 5th flagellar segment 0.9-l.Ox as long as 4th segment (Fig. 36) C. pilosa Delucchi (female, male unknown) p. 30 4. Pronotal collar with a transverse carina — at least on median pronotum — parts of pro- notum behind carina smooth and shiny (Fig. 87) 5 - Pronotum without transverse carina, hind margin of pronotum at most with a very narrow smooth strip (Fig. 86) 7 5. Petiole longer than wide (Fig. 62) C. acoris (Walker) (female, male) p. 64 - Petiole at most as long as wide (Figs 56, 59) 6 6. Scape and pedicel bright orange-yellow; clypeus pale yellow C. iUnstris Graham (male) p. 59 - Scape whitish or pale brown, pedicel brown; clypeus metallic bluish-green C. occidentalis (Girault) (female, male) p. 54 7. Propodeal callus with 2 setae 8 - Propodeal callus with at least 3 setae 14 8. Petiole pale; male flagellar segments with a single whorl of setae at the base of each segment (Fig. 42) C. beckeri Yoshimoto (female, male) - Petiole dark; male flagellar segments also with setae in middle and at apex 9 9. Propodeum with 2 complete and parallel submedian grooves (Fig. 84) C. sulcata (Hansson) (female, male) - Propodeum without such grooves 10 10. Forewing speculum open below (Fig. 47) C. vonones (Walker) (female, male) - Forewing speculum closed below 11 11. Scape bright orange-yellow; reticulation on thoracic dorsum fine and engraved C. fulviscapus Hansson (male, female unknown) p. 45 - Scape whitish or brown; reticulation on thoracic dorsum raised and strong 12 12. Occipital margin with a complete (reaching from eye to eye), high and sharp carina (Fig. 91); frontal suture raised C. liriomyzae Delucchi (male) p. 26 - Occipital margin without a complete carina; frontal suture not raised 13 13. All femora predominantly dark C. phijtomyzivora Hansson (male) p. 20 - Femora pale C. cerodonthae Hansson (male) p. 21 14. Petiole longer than length of median propodeum (Fig. 85) . . C. viridis (Nees) (male) p. 29 - Petiole at most as long as length of median propodeum 15 15. Postmarginal vein 3.0-3.5 x as long as stigmal vein 16 - Postmarginal vein 2.0-2.5X as long as stigmal vein 17 16. Petiole about as long as median propodeum; dorsellum usually excavated and divided in two parts by a median carina and hind part with a median incision (Fig. 69) C. entedonoides (Walker) (female, male) p. 28 - Petiole shorter than median propodeum; dorsellum convex to flat, without median in- cision and carina (Fig. 63) C. amyite (Walker) (female, male) p. 28 17. Hind coxa conspicuously long and slender, about 2.5 X as long as wide (Fig. 52); petiolar foramen triangular C. longicoxa Hansson (male) p. 26 - Hind coxa stouter, about 1.5X as long as wide; petiolar foramen semicircular to qua- drangular 18 82 Journal of Hymenoptera Research Figs. 1-8. Terminology. 1, beckeri, female: cc= costal cell; dm= dorsellum; m= marginal vein; ml= midlobe; msc= mesoscutum; om= occipital margin; OOL= ocell-ocular line; pc= propodeal callus; pet= petiolus; pm = postmarginal vein; POL= postocellar line; POO= distance between hind ocelli and occipital margin; prm= Volume 4, 1995 g3 18. Petiole as long as wide (as in Fig. 77); femora usually pale — in a few specimens pre- dominantly dark C. avia Hansson (female, male) p. 28 - Petiole transverse (Fig. 80); femora always predominantly dark C. phytomyzivora Hansson (male) p. 20 19. Flagellum with apical three segments fused (Fig. 41) C. imbrasus (Walker) (female, male unknown) - Flagellum with apical two segments fused 20 20. Forewing with a complete row of setae on underside of costal cell (as in Figs 48, 49) . . C robusta Yoshimoto (female, male) p. 62 - Costal cell without row of setae 21 21. Pronotal collar with transverse carina— at least on median pronotum (Fig. 87) 22 - Pronotum without transverse carina (Fig. 86) 55 22. Flagellum yellow; frontal suture missing (Fig. 26) C. lualleyi Yoshimoto (male) p. 60 - Flagellum brown; frontal suture present 23 23. Petiole at least 1.5x as long as wide 24 - Petiole at most as long as wide 26 24. Malar space very narrow, 1/15 the width of mouth opening (Fig. 92); frontal suture smoothly curved; with a procession between antennal toruli C. prodice (Walker) (female, male) p. 65 - Malar space not as narrow, 1/7 the width of mouth opening (Fig. 93); frontal suture more straight; without procession between toruli 25 25. Occipital margin with a low sharp carina; propodeal callus with 2 (3) setae; female frons above fork usually with rather strong reticulation; male frons smooth or with weak reticulation C. acoris (Walker) (female, male) p. 64 - Occipital margin rounded and smooth (in a few cases with a very weak carina); pro- podeal callus with 4-5 (3) setae; female frons above fork smooth or with very weak reticulation (male unknown) C. amasis (Walker) (female) p. 65 26. Propodeum with a strong and complete median carina (Fig. 58) C. walleyi Yoshimoto (female) p. 60 - Propodeum without a complete median carina 27 27. Posteromedian part of propodeum with two submedian, slightly curved carinae (Fig. 70); male pedicel bright orange-yellow C. gemma (Walker) (female, male) p. 59 - Posteromedian propodeum without carinae, or with 2-6 short and straight carinae (e.g. Fig. 73); male pedicel brown or whitish 28 28. Anteromedian part of propodeum with a wide and strong median carina — reaching half the length of propodeum and then dividing into 2 weaker carinae which diverge towards the hind edge of propodeum (Fig. 73) C. ilhistris Graham (female) p. 59 - Median carina on anteromedian part of propodeum weaker or missing 29 29. Clypeus partly or completely pale yellow 30 - Clypeus completely dark and metallic 31 30. Reticulation on thoracic dorsum very dense and strong, almost like punctulation; pro- podeal callus with 5 setae C. flaviclypeiis Hansson (female, male unknown) p. 63 propodeum; prn= pronotum; s= stigmal vein; scu= scutellum; sl= sidelobe; spe= speculum; sul= spiracular sulcus. 2, coUaris, female: pcl= pronotal collar. 3, Antenna, coinpresiiconiis, female: an= anelli; cla= clava; fla = flagellum; fun= funiculus; MPS= multiporous plate sensilla; pcl= pedicel; sca= scape. 4, Head, side view, phytoniyzh'om, female. 5, Head, front view, clarknc male: cly= clypeus; fs= frontal suture; he= height of eye; mo= width of mouth opening; ms= width of malar space; scr= antennal scrobes; tor= antennal toruli. 6, Mesothorax, side view, longicoxa, female: b= borderline between upper and lower mesepimeron. 7, Meso+metathorax, side view, tristis, female: b= as in Fig. 6; C2= mid coxa; C3= hind coxa; pc= propodeal callus. 8, iiicdiivia, female: pf= petiolar foramen. 84 Journal of Hymenoptera Research 10 11 Figs. 9-15. Chrysocharis n.spp. 9-12. cornigera. 9, Head, side view, female. 10, Head, front view, female. 11, Head, dorsal view, female. 12, Antenna, male. 13-15. wahli. 13, Head, side view, female. 14, Head, front view, female. 15, Antenna, male. Volume 4, 1995 85 Figs. 16-26. Head, front view. 16-17, ainslici, female. 18, Ditto, male. 19, oaciiiidi^, female. 20, Ditto, male. 21, chromatomyiae, female. 22, Ditto, male. 23, perditor, female. 24, Ditto, male. 25, amasis, female. 26, wallcyi, male. Figs. 27-40. 27, Head, front view, tristis, female. 28, Ditto male. 29, Head, lateral view, coptodiscae, male. 30, Ditto, pallidigaster, female. 31, Ditto, submiitica, male. 32, Head, dorsal view, occidentalis, female. 33, Ditto, coptodiscae, female. 34-40. Antennae. 34, occidentalis, female. 35, chilo, female. 36, Apical 2 flagellar segments, pilosa, female. 37, coptodiscae, female. 38, liriomyzae, female. 39, Ditto, male. 40, Flagellum, compressicornis, male. Figs. 41-56. 41-45. Antennae. 41, /»/^ras/- supposed species were found in close rorator that the senior author has seen in proximity to each other, i.e., north central that the two spines at the apex of meta- China and Mongolia, further supporting somal tergum 3 are not as sharp as those their conspecific identity. In their recent of S. irrorator. The variation in this char- revision of the Chinese species. He et al. acter in S. irrorator does not appear to be did not consider S. mongoliciis. It is the se- great, but series of specimens from the nior author's opinion that S. mongoliciis it- east Palaearctic must be checked. These self is likely synonymous with S. irrorator two species are therefore provisionally ac- because they apparently differ only in col- cepted. ^Q2 Journal of Hymenoptera Research PHYLOGENETIC CONSIDERATIONS is another character that was not included in the Achterberg and Austin data set but No hypothesis of the phylogenetic re- ^^^^^^ .^ ^^^ ^^^^^ ^^ ^^^ cladogram a lationships of Sigalphus species has been posteriori. An assessment of the distribu- published. Achterberg and Austin (1992) ^.^^ ^^ ^j^.^ character state shows that the analyzed the genera of the Sigalphinae. ^^^^^i^^j.^ of the non-Ethiopian species Despite this valiant first attempt at ap- ^^ Sigalphus are relatively more costate proximating relationships among the gen- ^^^^ ^^^^^ ^f the Ethiopian species which era, the monophyly of Sigalphus is ques- ^^^ mostly rugose, as in species of Min- tionable due to the inclusion of the Ethi- ^^^^ ^^^ Acampsis. Again, this character opian species (S. fiilvus, S. testaceus, and S. ^^^^^ ^^es not support the monophyly of neavei). Achterberg and Austin (1992) in sigalphus. their cladogram of sigalphine genera cite 3 -p^j-gt tergite with complete and 4 synapomorphies for the genus; they are g^^ong dorsal carina." Although this char- all suspect. These will be treated here in ^^^^^ ^^^^^ -g included in the data matrix, the order they appear on the cladogram. ^^^ cladogram does not present the most 1. "Third tergite with lobes or teeth parsimonious distribution. The data ma- ventro-posteriorly." This character state is ^^^^ j^g^g ^j^g following taxa as possessing not possessed by any of the three Ethio- ^^^ derived character state: Sigalphus, Min- pian species, conversely it is found in all ^^^^^ Acampsis, and Afrocampsis. This dis- members of Aposigalphus and Notosigal- tribution is correct, but on the cladogram phus in a condition almost identical to that Achterberg and Austin indicated that the of S. romeroi (Fig. 2a). Although the char- character state is an autapomorphy for Si- acter state appears on the cladogram of galphus and ignore the rest of the character Achterberg and Austin it is not included g^^te distribution. There are several parsi- in their data set. Species of Minanga also monious reconstructions of the character have teeth ventro-posteriorly on the third \y^^^ ^n indicate that the presence of strong tergum but since these do not emanate dorsal carina on the first tergite is plesiom- from the margin of the tergum as they do orphic with reference to Sigalphus. in the aforementioned groups their ho- 4 "Face with shallow semicircular de- mology is doubtful. The character state pression." This depression apparently re- has three possible origins given the phy- f^j-g to the two longitudinal depressions logeny of Achterberg and Austin, viz. a) a t^gt define the raised medial area of the synapomorphy for Sigalphus + (Aposigal- f^^e. Some specimens of S. irrorator appear phus + Notosigalphus) with a loss in the ^ have these depressions somewhat ex- Ethiopian species of Sigalphus, b) a syna- panded laterally but in other species of Si- pomorphy for the non-Ethiopian species galphus there is no discernable difference of Sigalphus and a synapomorphy (conver- between these depressions and those of gence) for Aposigalphus + Notosigalphus, or most other genera of Sigalphinae except c) a synapomorphy for Minanga + (Sigal- Afrocampsis which does not have them. phus + (Aposigalphus + Notosigalphus)) in summary, there is no evidence for the with a loss in the Ethiopian species of Si- monophyly of Sigalphus as presently de- galphus. This last hypothesis assumes that fined. That all four of the putative syna- the condition in Minanga and the other pomorphies proposed by Achterberg and genera is homologous. None of these char- Austin are incorrect suggests that their en- acter state distributions provides support tire analysis is suspect. We do not wish to for the concept of Sigalphus including the address generic concepts in this paper, Ethiopian species. only to draw attention to the fact that care 2. "Strongly costate metapleuron." This must be taken in interpreting their re- Volume 4, 1995 103 b Fig. 2. a, posterior aspect of metasoma of Sigalphus nmieroi showing ridges on ventral apex (d = denticles); b, dorsal aspect of metasomal tergum 2 of Sigalpliiis masoni; c, posteroventral aspect of metasoma of Sigalpluis nmsoiii showing denticles at ventral apex (d = denticles, o = ovipositor, v.f. = ventral foramen); d, dorsal aspect of metasoma of Sigalphus romcwi. 104 Journal of Hymenoptera Research ^V««Kvv,5*^<^VV*S^'^''"''^'' ,,<^x>'/''- .^yyy-^y-^yyy^ ,^^- m-cu U CU-a ^^^^^^VVVXVVXVSVVV,V.VNV .v>S-^" .vV-V---^ ^v^*- .v=N^^ *»»-^\\V«^^CSVV»^XVVNXVV^^ JJ.'tN"" "^"^^^--V-..ssvv.vs^-^-^ Fig. 3. a, fore wing of Sigalphiis romeroi; h, hind wing of Sigalphus romeroi; c, fore wing of Sigalphus inasoni. suits. A phylogenetic analysis of the spe- paper. The non— Ethiopian members of Si- des of Sigalphus must be preceded by a galpJius may be monophyletic, based on reexamination of the relationships of the the presence of denticles at the ventropos- genera and this is beyond the scope of this terior margin of tergum 3. KEY TO WORLD SPECIES OF SIGALPHUS 1. New World species 2 1.' Old World species 3 2. Metasoma orange; posterior flange of ventral foramen of metasoma smooth apically, ridg- es reduced to small crenulae not reaching apex of flange (south eastern Canada and east- ern U. S. A. south to Florida) S. bicolor (Cresson) Volume 4, 1995 105 2'. Metasoma black; posterior flange of ventral foramen of metasoma with ridges extending to apex (Fig. 2a) (known only from the type locality in northwestern Costa Rica, but the two species of noctuid hosts occur throughout the Mesoamerican Pacific dry deciduous forest and this parasitoid may be found there as well) S. romeroi n. sp. 3. Head and mesosoma entirely black 4 3.' Head and mesosoma not enhrely black, rather with pale coloration 9 4. Apex of ventral foramen of metasoma lacking denticles and two rounded lobes (southern Africa) S. neavei (Turner) 4'. Apex of ventral foramen of metasoma with two denticles or two rounded lobes (Fig. 2c) (Oriental, Palaearctic) 5 5. Apex of ventral foramen of metasoma with two sharp denticles 6 5'. Apex of ventral foramen of metasoma with two rounded lobes 7 6. Metasoma entirely black; stigma black (Palaearctic) S. irrorator (Fabricius) 6'. Metasoma orange and black; stigma yellow (Mongolia and northern China) S. tuougolicus Tobias 7. Tergum 2 with two parallel median longitudinal carinae with a median longitudmal groove between them 8 7'. Tergum 2 without two parallel median longitudinal carinae and without a median lon- gitudinal groove S. masoni n. sp. 8. Fore wing entirely melanic-infuscate; tergum 3 as wide as tergum 2 S. gyrodontus He and Chen 8'. Fore wing melanic-infuscate in distal half only; tergum 3 wider than tergum 2 S. rtifiabdominalis He and Chen 9. Vertex of head and apex of hind tibia black (southern Africa) S. fiilviis Brues. 9'. Head and hind tibia entirely reddish brown (Madagascar) S. testaceus Granger. Sigalphus romeroi Sharkey n. sp. (Figs 2a, d, 3a, b) MALE. Length. — 8.4 mm. Color. — Entirely black except fore tarsus brown, wings evenly infuscate. Head. — Antenna with 42-44 flagello- meres. Mesosoma. — Pronotum smooth except for weak crenulae along posterior border; subpronope present; notaulus deeply im- pressed and with weak punctures; scutel- lar furrow (sulcus) smooth except for me- dian longitudinal carina; posterior scutel- lar depression composed of two large pits; median depression of metanotum semicir- cular (rounded posteriorly); propodeum entirely areolate, with pair of strong lon- gitudinal carinae defining medial area; sternaulus complete to epicnemial carina; fore wing (Fig. 3a) with veins M and m- cu not distinctly converging anteriorly; crossvein a of fore wing not present; sec- ondary crossveins r and m-cu of hind wing present though weak and not tubu- lar (Fig. 3b). Metasoma (Figs 2a, d). — All terga rugose; first tergum with two precurrent longitu- dinal carinae; second tergum with median longitudinal carina, and without deep an- terolateral depressions (as in S. bicolor); tergum 3 lacking long dense setae; ventral margins of carapace crenulate; apex of ventral metasomal foramen with 6 riciges forming weak, sharp, denticles (Fig. 2a); ventral foramen of carapace not extending to apex of metasoma, i.e., tergum 3 curved around apex of metasoma forming poste- rior surface. Material Examined. — Holotype male: COSTA RICA, Guanacaste Province, Guanacaste Conservation Area, Sector Santa Rosa; 300 m. Janzen rearing data base voucher number 90-SRNP-828. Lar- val host Epithisanotia saiictijolunuia^i (Noc- tuidae) collected 24 June 1990, wasp 106 Journal of Hymenoptera Research eclosed 26 May 1991. Institute Nacional de Biodiversidad (INBio), Santo Domingo de Heredia, Costa Rica. Paratype male: COSTA RICA, Guana- caste Province, Guanacaste Conservation Area, Sector Santa Rosa, 300m. Janzen rearing data base voucher number 92- SRNP-2777. Larval host Erochia miimmia (species complex) (Noctuidae) collected 7 July 1992, wasp enclosed 10 May 1993. Ca- nadian National Collection, Ottawa, Can- ada. Remarks. — The female of this species is unknown. S. romeroi is easily distin- guished from the only other New World species, S. bicolor by the characters given in the key. Eti/mologi/. — The species is named in honor of Sra. Luz Maria Romero in rec- ognition of her outstanding drive and in- spiration in guiding and developing the Biological Education Program of the Guanacaste Conservation Area in north- western Costa Rica. Biogeographical Notes. — This is the first record of members of the genus Signlphiis in the Neotropics and only the second spe- cies to be discovered in the New World. The genus is unknown in South America. As suggested in the section on phyloge- ny, the non- Ethiopian members of Sigal- phus may be monophyletic based on the presence of denticles at the ventroposterior margin of tergum 3. If this is so, the distri- bution of the New World species is inter- esting from a biogeographic perspective. It is generally agreed that the closest phylogenetic relationships of most tem- perate North American taxa are found in the northern hemisphere of the Old World (Lavin and Luckow 1993). In contrast the usual hypothesis put forward to explain North American tropical fauna is dispers- al from South America. If the non-Ethiopian members of Signl- phus are monophyletic, it suggests a north- ern hemisphere origin of S. romeroi, the only Neotropical member of the genus. This distribution is consistent with what has been referred to as the boreotropics hypothesis, which suggests a historically more widespread tropical northern hemi- sphere biota. (Wolfe 1975; Tiffney 1985a, b; Lavin and Luckow 1993). Biology. — The larval hosts, Epithisanotia snnctijohannasi and Erochia miimmia (both: Noctuidae, Agaristinae), are common to exceedingly rare (depending on the year) and univoltine in the primary to early suc- cessional dry forests in eastern Sector Santa Rosa of the Guanacaste Conserva- tion Area in northwestern Costa Rica (200-300m elevation; about 35 km north of the town of Liberia, to the west of the In- teramerican Highway) (see Janzen 1987, 1988a, 1988b, 1993 for a general site de- scription of this forest and its caterpillars). E. sanctijohannasi and E. miimmia are most easily located when they are feeding on the large woody vine Tetracera vohibilis (Dilleniaceae, Dilleniales) (hundreds of re- cords), and £. sanctijohannasi has also been found feeding on Curatela americana (4 rec- ords). These larvae eat mature leaves and are non-gregarious feeders in the penulti- mate and ultimate instars (earlier instars unknown) on both host plants. Both spe- cies are univoltine (last instar caterpillars occurring in late June and early July, the second and third months of the rainy sea- son) and the pupae remain dormant through the remainder of the six month rainy season and all of the dry season, with the moths eclosing in May at the be- ginning of the rains (see Janzen 1987, 1988a, 1993 for other examples of the same phenomenon). Of 36 £. sanctijohannasi 31 £. miimmia last instar larvae collected and reared from T. voliibilis in 1992 and 1993, 84 and 93 percent respectively produced last instar larvae of Sigalphus romeroi. This extremely high percent of parasitization is not rep- resentative of caterpillar-parasitoid inter- actions in this forest, where 1-30% para- sitization by a given species of parasitoid is usually recorded (D. H. Janzen, unpub- lished rearing records). The very large Volume 4, 1995 107 number of Sigalphiis romeroi larvae did not generate a large paratype series because, as explained below, the larvae proved to be very incompetent at spinning cocoons and surviving subsequently under artifi- cial rearing conditions. The parasitized last instar caterpillars of £. sanctijohannnsi and E. uiiimmia appear to be perfectly normal in feeding, fleeing and walking behaviour. The prepupal caterpil- lar searches in the litter for a piece of rot- ting wood, tunnels into the wood, and makes its pupal chamber there. Within 1- 3 days of the caterpillar entering the wood the wasp larva eats nearly all of the con- tents of the prepupal caterpillar and exits the cadaver. The wasp larva is so large that there does not appear to be sufficient food for more than one per caterpillar of either host. The wasp larva then spins an ovoid white cocoon in the caterpillar's pu- pal chamber (late June to early July), where it remains dormant as a prepupal larva until the following May, when it pu- pates and exits through a round hole cut in the cocoon, and then burrows out through the wood chip-filled entrance tunnel made initially by the caterpillar. The biology described in the previous paragraphs is based on observation of all of this behaviour with captive caterpillars individually confined in large plastic bags and bottles in an open air laboratory a few meters from the forest where the caterpil- lars and wasps are found naturally. Dor- mancy of moth pupae and of wasp larvae in their cocoons occurred in the individ- uals confined under three conditions of captivity, i.e., under moist conditions throughout the wet season with a change to dry conditions during the dry season, entirely under dry conditions throught^ut the wet and dry seasons, and for those captured in the dry season they remained in a dry environment in captivity. Moths in the pupal stage and wasp larvae both use some environmental cue that is per- ceivable through the walls of a plastic bag to initiate further development in late April or early May, and this cue is most likely the abrupt drop in temperature that accompanies the first rains (see Janzen 1993). By eclosing at the same time as the moths, the adult wasps are presumably present at exactly the same time as are the first instar larvae of the next generation of caterpillars which, extrapolating from the known biology of S. bicolor, is the stage most likely to be attacked. The larva of Sigalphus romeroi is excep- tional, among the species of parasitoid wasps that the junior author has reared in this forest, in its inability to spin a cocoon except in the pupal chamber of its host. It was usually not able to spin a cocoon be- tween leaves in the litter, among crum- pled layers of dry or moist toilet tissue, or in tubular chambers of toilet tissue. Of ten larvae that did succeed in spinning co- coons among toilet tissue layers, all died of unknown causes over the following seven months. Malaise traps run for the equivalent of tens of years in this dry forest between 1984 and 1988 (e.g., Gauld 1988, Gaston and Gauld 1993) have captured no speci- mens of S. romeroi (I. D. Gauld and P. A. Mitchell, personal communication). This result is particularly striking given that the Malaise traps were located within sev- eral meters of the host plants and in all ages of forest succession. In contrast to S. romeroi being large, sol- itary, and univoltine, Signlplnis bicolor, the only North American member of the ge- nus, is small, gregarious and multivoltine. S. bicolor oviposits in the first instar larvae of a noctuid caterpillar, Acronicta clarescens (= Apatela clnrescetis), and the larv^ae come out of the prepupal caterpillar and spin their cocoons within the host cocoon (Cushman 1913, who described the wasp's natural history under the name Sphaeropyx bicolor). The only western Palaearctic spe- cies of the genus, S. irrorator has also been reared from Acronicta in Europe, but be- cause of its large size (about the same as S. romeroi) it is inferred to be solitary (S. 108 Journal of Hymenoptera Research Ward, personal communication). You, Zhou and Tong (1991) report S. irrorator (as S. anomis) as a parasitoid of Anomis flava (Noctuidae). The biology of all other spe- cies of Sigalphus is unknown. Sigalphus masoni Sharkey n. sp. (Figs. 2b, c, 3c) HOLOTYPE FEMALE. Length. — 10.9 mm. Color. — Mostly black with yellow on bases and ventral sides of all tibiae and on metasomal sterna, and orange on meta- somal terga 1, 2, and anterior % of tergum 3; fore wing entirely infuscate though somewhat paler basally; hind wing clear in basal Vs, infuscate in distal %. Head. — Antenna broken. Mesosoma. — Pronotum weakly rugose; subpronope present; notaulus deeply im- pressed and crenulate; scutellar furrow (sulcus) smooth except for three longitu- dinal carinae; posterior scutellar depres- sion composed of many small pits; medi- an depression of metanotum sharply tri- angular (acute posteriorly); propodeum weakly rugose laterally with pair of lon- gitudinal carinae defining medial area; sternaulus complete to epicnemial carina; fore wing with veins M and m-cu distinct- ly converging anteriorly; crossvein a of fore wing present and distinct; hind wing with secondary crossvein r present and secondary crossvein m-cu absent. Metasomn (Figs 2b, c). — All terga rugose; first tergum with two precurrent longitu- dinal carinae; second tergum with median longitudinal carina and an equally strong pair of lateral longitudinal carinae that converge towards the mid-line posterior- ly; second tergum with moderately deep anterolateral depressions (as in S. bicolor); tergum 3 with long dense setae; ventral margins of carapace crenulate; apex of ventral metasomal foramen with two rounded denticles (Fig. 2c); ventral fora- men of carapace (Fig. 2c) not extending to apex of metasoma. Material Examined. — Holotype 9: NE- PAL: Katmandu, Godavari, 5,000 ft. (1524 m.), 6. VIII. 1967, Canadian Nepal Expedi- tion, [Canadian National Collection, Otta- wa, Canada] Biology. — Unknown Remarks. — Members of S. masoni are similar to those of S. irrorator and S. mon- golicus in some respects, particularly in their general size and the long setae on metasomal tergum 3. In S. masoni, these setae are longer and denser than in the New World species of Sigalphus, though they are not as exaggerated as those of S. irrorator, nor are they gold colored. Mem- bers of the two species differ in color, those of S. masoni having the metasoma predominantly orange versus entirely black in S. irrorator. The species also differ in the shape of the denticles at the apex of the ventral foramen, i.e., those of S. masoni are wide and blunt whereas those of S. ir- rorator and S. mongolicus are relatively nar- row and acute. The most distinctive fea- ture distinguishing S. masoni from other species of Sigalphus is the sculpture of the second metasomal tergum. All other Pa- laearctic and Oriental species of Sigalphus have a median longitudinal groove bor- dered laterally by longitudinal carinae. These are not present in S. masoni (Fig. 2d). Etymology. — The species is named after Dr. W. R. M. Mason for his many contri- butions to Hymenopterology, and also, though it is not stated on the label, be- cause he is thought to be the collector of the unique specimen. ACKNOWLEDGMENTS This study was supported by NSF grant BSR 90- 24770 and DEB 94-00829 to D. H. Janzen. Junhua He is especially acknowledged for the loan of the type specimens of S. hunanus and S. anoniis. Sondra Ward and Ian Gauld are gratefully acknowledged for hav- ing expressed great excitement over the first speci- men reared. We thank Ian Gauld for having stimu- lated this dry forest parasitoid study in the first place. We thank R. W. Poole for identifying the moths. Hen- ri Goulet drew and Susan Rigby inked the habitus of S. bicolor which was first published in the "The Hy- menoptera Of The World" Goulet and Huber (1993). I Volume 4, 1995 109 Jason Hardis drew the fore wing S. masoni. All other drawings were done by Barry Flahey. We thank Da- vid Althoff, Paul Marsh, Lubomir Masner, Eric Rick- ey, Ales Smetana and two anonymous reviewers for valuable comments on early drafts. LITERATURE CITED Achterberg, C. van and A. D. Austin. 1992. Revision of the genera of the subfamily Sigalphinae (Hy- menoptera: Braconidae), including a revision of the Australian species. Zoologische Veiiuvniclingen 280:1-44. Cushman, R. A. 1913. Biological notes on a few rare or little known parasitic Hymenoptera. Proceed- ings of the Entomological Society of Washington 15: 153-161. Gaston, K. J. and I. D. Gauld 1993. How many spe- cies of pimplines (Hymenoptera: Ichneumoni- dae) are there in Costa Rica? Journal of Tropical Ecology 9:491^99. Gauld, I. D. 1988. A survey of the Ophioninae (Hy- menoptera: Ichneumonidae) of tropical Mesoam- erica with special reference to the fauna of Costa Rica. Bulletin of the British Museum (Natural His- tory), Entomology Series 57 :l-309. Goulet, H. and N.B. Huber. 1993. Hymenoptera of the world: an identification guide to families. Research Branch Agriculture Canada Publication 1894 /E. 668pp. He, J. and X. Chen. 1993. Descriptions of two species of the genus Sigalphus Latreille from China (Hy- menoptera: Braconidae: Sigalphinae). Acta Ento- mologica Sinica 34:225-229. (In Chinese with En- glish summary). He, J., Chen, X. and Ma H. 1994. Revision of Sigal- phus species from China with descriptions of two new species. Journal Zhejiang Agricultural Univer- sity 20:441-448. Janzen, D. H. 1987. How moths pass the dry season in a Costa Rican dry forest. Insect Science and its Application 8:489-500. Janzen, D. H. 1988a. Ecological characterization of a Costa Rican dry forest caterpillar fauna. Biotro- pica 20:120-135. Janzen, D. H. 1988b. Guanacaste National Park: Tropical ecological and biocultural restoration. In: J. Cairns, Jr., ed.: Rehabilitation of damaged eco- systems, Vol. II, CRC Press, Boca Raton, Florida, pp. 143-192. Janzen, D. H. 1993. Caterpillar seasonality in a Costa Rican dry forest, hi: N. E. Stamp and T. M. Casey eds.: Caterpillars. Ecological and evolutionary con- straints on foraging, Chapman and Hall, New York, pp. 448-477. Lavin, M. and M. Luckow. 1993. Origins and rela- tionships of tropical North America in the con- text of the boreotropics hypothesis. American Journal of Botany 80:1-14. Shenefelt, R.D. 1973. Braconidae 6 Cheloninae. Hy- menopterorum Catalogus (n.e.) 10:813-936. Tiffney, B. H. 1985a. Perspectives on the origin of the floristic similarity between eastern Asia and east- ern North America. Journal of the Arnold Arbore- tum 66: 73-94. Tiffney, B. H. 1985b. The Eocene North Atlantic land bridge: its importance in Tertiary and modern phytogeography of the northern hemisphere. Journal of the Arnold Arboretum 66: 243-273. Tobias, V.I. 1974. Contributions to the fauna of Bra- conidae (Hymenoptera) of Mongolia. [In Rus- sian]. The Insects Of Mongolia 2:261-274. Lenin- grad Publishing House. You, L., Zhou, Z. and X. Tong. 1991. Two new spe- cies of Sigalphus Latreille, 1802 from Hunan Province (Hymenoptera: Braconidae: Cheloni- nae). Acta Entomologica Sinca 34:225-229. You, L. and Z. Zhou. 1994. Revision of the genus Sigalphus Latreille, 1802 of the subfamily Sigal- phinae (Hymenoptera: Braconidae) in China. Journal of Hunan Agricultural College 20:363-365. Wolfe, J. A. 1975. Some aspects of plant geography of the northern hemisphere during the late Cre- taceous and Tertiary. Annals of the Missouri Bo- tanical Garden 62: 264-279. J. HYM. RES. Vol. 4, 1995, pp. 110-120 Ovipositor Steering Mechanisms in Braconid Wasps Donald L. J. Quicke, Mike Fitton, Jane Harris (DLJQ) Department of Biology, Imperial College at Silwood Park, Ascot, Berks SL5 7PY, U.K.; (MF, DLJQ) Department of Entomology, The Natural History Museum, London SW7 5BD, U.K.; (JH) Department of Animal and Plant Sciences, PO Box 601, University of Sheffield, Sheffield SIO 2UQ, U.K. Abstract. — Ovipositor features which allow the ovipositor tip to be manipulated and which have restricted distributions among the subfamilies of Braconidae are described and illustrated for the first time. Members of the Amicrocentrinae, Charmontinae, Helconinae and Macrocentrinae have the rhachis (the tongue-like part of the mechanism that interlocks the upper valve with the lower ones) with an increased density of scale-like sculpture and often also swollen pre-apically. This modification leads to increased friction and restricted movement between the upper and lower valves and is also associated with an ovipositor bending mechanism in which retraction of the lower ovipositor valves relative to the upper one causes the ovipositor tip to bend ventrally. Many members of the subfamilies Agathidinae and Orgilinae possess a pair of pre-apical boss-like pro- jections on the upper ovipositor valve (gonapophyses 9) and a corresponding structure on each lower valve (gonapophyses 8) that together enable the wasps to bend and manipulate the ovi- positor apex by retraction of the lower ovipositor valves relative to the upper one. Most members of the Doryctinae have the aulax (the groove-like part of the mechanism that interlocks the lower valve with the upper one) constricted opposite modified pre-apical, dorso-lateral tooth-like struc- tures. These constrictions restrict the relative movement between the upper and lower valves and thus operate as an ovipositor bending mechanism. The potential phylogenetic significance of these mechanisms is discussed, and preliminary parsimony analyses are described which suggest that the pre-apical boss-like projections of the Agathidinae and Orgilinae may represent a synapo- morphy uniting these two subfamilies. INTRODUCTION Oviposition behaviour and, as a conse- quence, the ovipositor have been of im- mense importance in the evolution of the Hymenoptera, and are central to the par- asitic way of life (Gauld & Bolton 1988). The commonly held vievv^ that the ovipos- itors of parasitic w^asps are fairly simple analogues of hypodermic needles is there- fore likely to be a considerable over-sim- plification as has in fact been shown by several previous investigations of func- tional morphology (Oeser 1961; Austin & Browning 1981; Quicke et al. 1992a,b, 1994). Several braconid genera have recently been shown to possess highly modified ovipositors whose tips are capable of be- ing independently actively manipulated by the wasp even though the ovipositor valves themselves are devoid of intrinsic musculature and their relative movements are controlled by muscles within the me- tasoma. Mechanisms, as found for exam- ple in the braconine genus Zagli/ptogastra Ashmead, were apparent because of very conspicuous features of the intact ovipos- itor (Quicke 1991; Quicke & Marsh 1992). However, an ability to manipulate or steer the ovipositor tip has been observed in a number of other parasitic Hymenoptera whose ovipositors appear more or less un- modified externally (Delanoue & Aram- bourg 1965; Compton & Nefdt 1988). Fur- ther, our own observations have also Volume 4, 1995 111 shown that an ability to manipulate the ovipositor tip is present in members of several subfamilies of Braconidae that possess moderately long ovipositors. We have therefore examined ovipositor tip structure in detail across a range of brac- onid taxa. Here we describe three different types of morphological specialisation found in the Braconidae that appear to un- derlie the manipulatory abilities observed. These modifications all utilise the same mechanical principle but they achieve this through modification of different struc- tures. In these cases, the structures in- volved can often only be revealed by ex- amination of separated ovipositor valves using scanning electron microscopy; how- ever, their existence supports the view that the ovipositors of parasitic wasps are highly specialised and complex structures. Two different ovipositor steering mech- anisms, one involving a sinuous apical re- gion of the ovipositor the other involving the development of an aulaciform rod be- tween the largely separated halves of the upper ovipositor valve have been report- ed previously (Quicke 1991; Quicke et al. 1994). The first of these is possessed by several unrelated taxa of Braconidae and Ichneumonidae, the second so far is known only from members of the Ichneu- monidae. During investigations of ovipos- itors belonging to a large number of other ichneumonid wasps collectively repre- senting most subfamilies, no steering spe- cialisations similar to any of the three de- scribed here for Braconidae have been de- tected. Functionally similar but morpho- logically different mechanisms in members of two other families of parasitic Hyme- noptera, Gasteruptiidae and Aulacidae, are being described elsewhere (Quicke & Fitton, in press). GENERAL MORPHOLOGY The ovipositors of Hymenoptera are comprised of components that are be- lieved to be derived primitively from two pairs of metasomal appendages (Smith 1968, 1970). The two gonapophyses of ab- dominal segment 9 are usually fused along their length and form the single up- per valve, while those of segment 8 re- main separate and form the two lower valves. The upper valve is interlocked with each lower valve by a k:)ngitudinal tongue and groove joint referred to as the olistheter. The tongue or rhachis situated ventro-laterally on each side of the upper valve is T'-shaped in transverse section and in life runs within the T'-shaped groove or aulax on the dorso-lateral face of the lower valve (Fig. 1). MATERIALS AND METHODS Ovipositors of 70 species of Braconidae, collectively representing members of 19 of the 40-or-so currently recognised subfam- ilies (Quicke & van Achterberg 1990; Shaw & Huddleston 1991) have been examined using scanning electron microscopy. For most of these, the upper and lower valves were carefully separated after soaking in- tact wasp metasomas or excised oviposi- tors in either water or dilute aqueous mild detergent. In the case of a few specimens in which one or both lower ovipositor valves were protruding beyond the apex of the dorsal valve, the intact ovipositor was examined. Limited observ^ations of ovipositor ma- nipulation were made on some hand-held living female wasps. In addition, experi- ments were carried out using excised ovi- positors from freshly killed wasps. Ovi- positors were submerged under water and the effects observed of pulling and push- ing the cut ends of the basal portions of the upper and lower valves differentially with fine forceps. The taxa investigated are listed below according to subfamily. Where more than one individual of a species was examined the number is given in parentheses. Agathidinae: Braunsia sp., Cameroun (3); Cnmptothlipsis sp., Kenya; Mcgngnthis sp.. Sierra Leone; Microdiis rufipcs Nees, UK. Alysiinae: Phacnocarpa conspurcator 112 Journal of Hymenoptera Research 2a. Macrocentrus type o < < 1. Stylized TS ovipositor 2b. Agathidine type upper valve aul lower valve c ■o •D n> < < (D O S < < 2c. Doryctine type upper valve retracted 2d. Doryctine type upper valve protracted lower valve upper valve lower valve rh aul Figs. 1,2. 1. Diagramatic transverse section (TS) through mid-region of braconid ovipositor to show olistheter mechanism: (ol) olistheter, (rh) rhachis, (aul) aulax. 2. Diagramatic representations of how the three ovipositor bending mechanisms are believed to operate, a, Macrocentrinae, Amicrocentrinae, and Charmontinae, lateral views showing preapical swelling of rhachis; b, Agathidinae and Orgilinae, lateral views showing pre-apical bosses; c, d, Doryctinae, dorsal showing in the postulated increased grip against the rhachis when the upper valve is protracted. (Haliday), Netherlands. Amicrocentrinae: Amicrocentrum sp., Tanzania. Blacinae: Ar- tiocus spinarius Achterberg, Brazil (2); Bla- cus hastatus Haliday, Sweden; B. longipen- nis (Gravenhorst), UK; B. maculipes Wes- mael, Austria; B. nigricornis Haeselbarth, UK; B. paganus Haliday, UK; B. pallipes Haliday, UK; B. ruficornis Nees, Hungary, Russia (2); Dyscoletes lancifer (Haliday), UK, Germany (2). Braconinae: Digonogas- tra sp., Trinidad; Eiiurohracon yokahamae Dalla Torre, Japan; Glyptomorpha sp., Gam- bia; Nesaiilax sp., Papua New Guinea; Ode- sia sp., Kenya; Virgulihracon endoxylapha- gus Quicke & Ingram, Australia. Cardi- ochilinae: Cardiochiles sp., Cameroun. Gen- Volume 4, 1995 113 ocoeliinae: Capitonius spp., Brazil (2). Charmontinae: Charmon cruentatus Haliday, UK (2). Doryctinae: Acanthodoryctes tomen- tosa (Szepligeti), Australia; Acropjliasmiis fernigineus (Marsh), USA; Allorhogas pyr- alophagus Marsh, Trinidad; Binarea sp., Ar- gentina; Doryctes erythromeles (Brulle), USA; Liobracon sp., Brazil; Megaloproctiis platyantennus Marsh, Brazil; Monarea sp., Brazil; Rhaconotus sp., USA; Rutheia sp., Bolivia; Syngaster sp., Australia; Spathiiis sp., Kenya; Zombrus sp., Kenya. Helconi- nae: Austrohelcou sp., Australia; Aspicolpus carinator Nees, Hungary; Aspidocolpus sp., Nigeria; Baecis abietis (Ratzeburg), Poland; Briilleia latiannulatus (Cameron), New Guinea; Calohelcon obsciiripennis Turner, Australia; Diospilus capita (Nees), Hungary (2); Diospilus morosus Reinhard, Austria; Eubazus pallipes Nees, Romania; £. tibialis Haliday, Bulgaria; £. (Brachistes) spp., Ar- gentina, UK (2); Helcon uunciator Fabricius, Hungary; Helconidea tentator (Fabricius), Belgium; Polydegmon sinuatus Foerster, Hungary; Taphaeus hiator Thunberg, Hun- gary; Wroiightonia sp., Canada; unde- scribed genus A, Australia; undescribed genus B, Australia. Homolobinae: Homo- lobiis truncator (Say), Bulgaria (2); H. {Oiil- ophiis) sp., USA. Macrocentrinae: Austroz- ele uniformis Provancher, USA; Macrocen- trus linearis Nees, Hungary (2); Macrocen- triis sp., Poland. Microgastrinae: Sathon falcatus (Nees), Germany. Microtypinae: Microtypus sp., Hungary. Opiinae: Doryc- tobracon sp., Trinidad. Orgilinae: Orgilus leptocephnlus (Hartig), Germany; Orgilus spp., Cameroun, France (3); Orgilonia sp., Cameroun; Stantonia sp., Cameroun. Pam- bolinae: Monitoriella elongatn Hedqvist, USA. Pselaphaninae: Pselaphanus trogoides Szepligeti, Surinam. Rhysipolinae: Rhysi- polis sp., UK RESULTS Upper Valve Rhachis Scaling and Expansions In all Hymenoptera both the rhachis and the margins of the aulax are furnished with scale-like sculpture which, in the case of the parasitic Apocrita and the Orussi- dae, are generally rather widely spaced (Figs. 16, 17). In the Amicrocentrinae, Bla- cinae, Charmontinae and Helconinae the density of scaling was found to be consid- erably increased over a short region pre- apically (Figs. 3-11). Further, in the Char- montinae and Macrocentrinae this is ac- companied by a distinct swelling of the rhachis (Figs. 3-6; arrowed) and slight widening of the corresponding length of the aulax. Observation of living Macrocentrus fe- males held in the hand showed that they are capable of bending the apex of their ovipositors ventrally through nearly 90°. It seems likely that this is achieved in an analogous way to that proposed for the Agathidinae and Orgilinae (see below), with the swollen and more coarsely sculp- tured region of the rhachis providing in- creased resistance to relative movement of the ovipositor valves at the apical part of the ovipositor. Thus, if the dorsal valve is protracted relative to the ventral ones, they will not be able to slide past one an- other freely at the apex, and so the ovi- positor will be forced to curve ventrally in order to accommodate the relative exten- sion of the upper valve in a manner anal- ogous to the bending of a bimetal strip in a thermostat, as illustrated diagrammati- cally in Fig. 2a. The plausibility of this was conclusion was substantiated by manipu- lation of the cut ends of an excised Macro- cent rus ovipositor which showed that pro- traction of the upper valve relative to the lower one causes a ventral curving of the ovipositor tip. Upper and Lower Valve Bosses In all members of the Orgilinae and most Agathidinae examined, the pre-api- cal lateral margins of both the upper (Figs. 12, 13, 16, 17) and lower (Figs. 14, 15) ovi- positor valves possess an abrupt escarp- ment-like boss in lateral aspect. These structures are located lateral tt) the olis- 114 Journal of Hymenoptera Research Figs. 3-6. Scanning electron micrographs of apical parts of isolated upper ovipositor valves showing preap- ical expansion of rhachis (r) with denser scaling (arrows). 3, 4 (detail), Clmrniou cruoitntus. 5, Macrocentrus linearis. 6, Macrocentrus sp. Scale line: 3, 5: 50|jim; 4, 6: 25(xm. Volume 4, 1995 115 Figs. 7-11. Scanning electron micrographs of apical parts of isolated upper ovipositor valves showing weak preapical expansion of rhachis (r) with denser scaling (arrows). 7, 8, Amicrocenirum sp. 9, Hckonidea lentator. 10, Wwiightonia sp. 11, BniUeia latmwnilatiis. Scale line: 7: 100|xm; 8: ZOjim; 9, 50n.m; 10, 11: 40^,01. theter. In the case of the dorsal valve, the steep (scarp) face of the stop is at the distal end of the modified region whereas it is at the proximal end in the lower valve. The positioning of these modifications is such that when the ovipositor is in its rest- ing position with neither upper nor lower valves apically protruding beyonci the other, the scarp faces of both processes are closely apposed. The presence of these stops limits the relative longitudinal movement of the upper and lower valves such that the upper valve cannot be pro- tracted apically, relative to the lower valves, beyond the resting position though the upper valve can be retracted basally. The functional significance of this modifi- cation was revealed when living agathi- dines of the genus Agathis were held in the hand and the apical part of the ovipositor was observed to change its degree of ven- tral curvature as the wasp attempted to 116 Journal of Hymenoptera Research I Figs. 12-17. Scanning electron micrographs of apical parts of isolated upper (12, 13, 16, 17) and lower ovi- positor valves (14, 15) showing (arrows) corresponding preapical, bosses lateral to the rhachis (r) and aulax (a) of the upper and lower valves respectively. 12, 13, 15, Braunsia sp. 14, 16, Megagathis sp. 17, Orgilonia sp. Scale line: 12: 50|jLm; 13, 16: 20|xm; 14, 15: 30|xm; 17: 35|xm. free itself. Such bending appeared to be posing scarp faces of their bosses. Because achieved by the wasp protracting the up- the valves are interlocked along their per valve posteriorly relative to the lower whole length by the olistheter mechanism, valves, thereby forcing together the op- the ovipositor is forced to curve ventrally. Volume 4, 1995 117 as illustrated in Figure 2b. The plausibility of this was conclusion was substantiated by manipulation of the cut ends of an ex- cised Agnthis ovipositor which showed that protraction of the upper valve relative to the lower one causes a ventral curving of the ovipositor tip. No other braconids examined possessed this modification and since neither the Agathidinae nor the Orgilinae are likely to be the most basal of extant braconids (Quicke & van Achterberg 1990; Wharton et al. 1992; van Achterberg & Quicke 1992) (for example, because of their koinobiont endoparasitoid life history), it seems likely that the presence of a scarp-like stop mechanism is apomorphous. The potential implications of this are dealt with in the discussion. Pre-apical Constrictions of the Aulax Unlike other braconid subfamilies, most members of the subfamily Doryctinae have the apex of the ovipositor much darker (piceous or black) and much more strongly sclerotized than the remainder (Quicke et al. 1992a) such that it is usually extremely difficult to separate the valves for study even after softening in aqueous potassium hydroxide; though with persis- tent manipulation, the three parts could usually be separated. Scanning electron microscopy shows that in many genera (e.g. Binnrea, Liobracon, Monarea, Syngasier, Zotnbrus) the apical part of the lower valves, as well as having ventro-lateral serrations, also possess one or two appar- ently distinct dorso-lateral tooth-like structures (Figs. 18-21; arrows) (Quicke et al. 1992a). The significance of these in terms of ovipositor steering has been re- vealed by examination of the adjacent part of the aulax. The aulax is an apprc^ximate- ly parallel-sided groove along most of its length, in isolated ventral valves it is strongly constricted opposite these addi- tional teeth by an infolding of the lateral margin (Figs. 22-24). Together with the in- creased sclerotization of the ovipositor tip and our observation that it is very difficult to separate the dorsal and ventral valves at the tip region, we conclude that the nar- rowed region of the aulax acts to grip the rhachis tightly, thereby reducing the free- dom of movement of the valves near to the ovipositor tip. The asymmetric ar- rangement of the constrictions of the aulax are such that retraction of the upper valves is likely to afford less resistance than protraction (Figs. 2c cf d) which will cause the lateral wall of the aulax to roll inwards thus increasing the friction against the rhachis. This prediction was borne out by observations of living Hcca- holiis females which were found \.o be able to bend their ovipositors ventrally but not dorsally. DISCUSSION In all the cases in which we have found to possess probable ovipositor steering mechanisms, their hosts are concealed in- sect larvae. Orgilines mostly attack weakly concealed hosts such as leaf miners, and they probe into their hosts' tunnels often through frass holes rather than by boring directly through the plant material (Flan- ders & Oatman 1982; Shaw & Huddleston 1991). Agathidines, charmontines and ma- crocentrines similarly attack weakly con- cealed hosts such as leaf rollers, leaf tiers, shoot borers, and, particularly in the case of the Agathidinae, leaf miners /tunnellers. The steering mechanisms may therefore enable the wasps to locate their hosts better within their concealed feeding places. It may also enable the wasps to place their eggs accurately on or within their hosts. In the case of agathidines, for example, it is known that eggs are usually placed very precisely within or next to the host's cen- tral nervous system (references in Shaw & Huddleston 1991). Given that their hosts are concealed and often relatively small, such accurate egg placement would be fa- cilitated by the existence of some sort of directional control of the ovipositor tip. In contrast to the above mentioned sub- 118 Journal of Hymenoptera Research Figs. 18-24. Scanning electron micrographs of apical parts of intact ovipositors (18, 19) and isolated lower ovipositor valves (20-24) of doryctine braconids showing dorso-lateral tooth like structures (arrow) and in 24, the corresponding narrowing of the aulax (a). 18, Megalopwctus plaiyantenniis. 19, 24, Binarea sp. 20, Acroplias- mus ferrugineus. 21, Spathius sp. 22, Syngaster sp. 23, Zombrus sp. Scale line: 18: 130(xm; 19: 70|xm; 20, 21: 60|xm; 22, 23: 30|xm; 24: 10|xm. families, many members of the Doryctinae attack more strongly concealed insects such as wood-boring beetle larvae, and since they are idiobiont ectoparasitoids, the exact placement of the paralysing sting or of the egg is usually likely to be of little importance. The process of boring through bark or other hard substrates may not be very accurate and is also quite time consuming, which in turn probably places the wasp under an increased risk of pre- dation. Locating a host tunnel might therefore be a good first step towards reaching a host if the wasp can subse- quently manipulate its ovipositor tip. Steering mechanisms might be seen in these cases as an adaptation which over- comes some of the potential problems of reaching a concealed host directly through a thick layer of substrate. Of some surprise was that we failed to find any morphological modifications that Volume 4, 1995 119 could be involved in ovipositor steering in a number of taxa with long to very long ovipositors such as members of the bra- conine genera Archibracon, Euurobracou, Glyptomorpha, Nesaidax, Odesia and Virgii- libracon. Nor were any steering mecha- nisms identified in the Alysiinae, Cardi- ochilinae, or Cenocoeliinae examined. Ob- servations of living females of these taxa would be valuable as it is quite possible that they have some other, perhaps less conspicuous, mechanisms that enable them to manipulate their ovipositors or ovipositor tips. Certainly for many of the braconines with long ovipositors, ovipo- sition involves boring through hard wood (van Achterberg 1986) and this may pose limitations on the bending mechanisms they might possess. Alternatively, it may be that most braconines that attack deeply concealed, wood-boring hosts are simply not adapted to hit the host's tunnel first and then to use the ovipositor to follow the tunnel to the host, but rather reach their hosts by drilling directly through the substrate to them. If this is generally the case, then the strong, possibly defensive, smell of most braconines (Quicke 1988; Quicke & Wharton in preparation) could be an adaptation to protect them during protracted periods of drillng for hosts, during which time they may be particu- larly vulnerable to predation. Interestingly, the Agathidinae and Or- gilinae did not appear together in the phy- logenetic analyses of Quicke & van Ach- terberg (1990) and Wharton et al. (1992) though their relationships were rather la- bile. These two subfamilies have a similar general appearance and perhaps more im- portantly, have two other putative syna- pomorphies, a relatively narrow forewing marginal cell with an approximately straight vein 3-SR, and in those species with a complete 2nd submarginal cell, this is small and usually triangular or at least has vein 2-SR very short. The discovery of a new apomorphous character present in both, further suggests the possibility that they could be sister groups, though the possibility also exists that the ovipositor bending mechanisms possessed by mem- bers of the two subfamilies reflect conver- gence as a result of the similarities in their hosts' biologies. In a preliminary investi- gation, we have added the presence or ab- sence of an ovipositor boss as an addition- al character to a data matrix of braconid subfamilies based on that of Quicke & van Achterberg (1990) modified in accordance with the suggestions of Wharton et al. (1992) and of van Achterberg & Quicke (1992). Whereas the Agathidinae and Or- gilinae did not appear as sister groups in the most parsimonious trees (length 391), trees with them so arranged were only one step longer than the most parsimoni- ous ones obtained. The absence of this character in the Pselaphaninae might be taken to indicate that they are not the sis- ter group of the Agathidinae (van Achter- berg 1985), however, it should be noted that in Pselaphanus, the ovipositor is rather robust and is probably unsuited to "steer- ing" and that such a mechanism could have been secondarily lost. In this context, it should also be noted that members of some genera of Agathidinae that have shorter and more robust ovipositors were also found to lack the bosses, and there- fore this feature may be readily lost in spe- cies whose ovipositors evolve to become less flexible. We therefore conducted a sec- ond phylogenetic analysis on the data ma- trix described above but this time scoring the Pselaphaninae as unknown for the new ovipositor character, thus allowing the possibility of a reversal in that group without adding to tree length. In this case the most parsimonious trees included a subset containing the clade (Orgilinae -I- (Agathidinae + Pselaphaninae)). We therefore suggest that the newly discov- ered ovipositor features should be incor- porated in future phylogenetic analyses of the Braconidae. ACKNOWLEDGMENTS We would like to th.ink Rachel Kruft and Laraine Ficken for their assistance with scanning electron mi- 120 Journal of Hymenoptera Research croscopy, and Kees van Achterberg, Paul Marsh and Jeno Papp for providing some of the material exam- ined. Our thanks also to Mark Shaw and Robert Bel- shaw who provided many helpful comments on an earlier version of the manuscript. The work was part- ly supported by a NERC research grant (GR3/8060) to Donald Quicke and Mike Fitton and by the NERC Initiative in Taxonomy. LITERATURE CITED Achterberg, C. van. 1985. Notes on Braconidae. V. The systematic position of the genera Ecnomios Mason and Pselaphanus Szepligeti (Hymenop- tera: Braconidae). Zoologische Verhandeliiigen, Lei- den 59: 341-348. Achterberg, C. van. 1986. The oviposition behaviour of parasitic wasps with very long ovipositors (Ichneumonoidea: Braconidae). Eutomologische Berichten 46: 113-115. Achterberg, C. van and D. L. J. Quicke. 1992. Phy- logeny of the subfamilies of the family Braconi- dae: A reassessment assessed. Cladistics 8: 237- 264. Austin, A. D. and T. O. Browning 1981. A mecha- nism for movement of eggs along insect ovipos- itors. International Journal of Insect Morphologx/ and Emhryologi/ 10: 93-108. Compton, S. and R. Nefdt. 1988. Extra-long ovipos- itors in chalcid wasps; some examples and ob- servations. Antenna 12: 102-105. Delanoue, P. and Y. Arambourg. 1965. Contribution a I'etude en laboratoire d'Eupelmus urozonus Dalm. (Hym. Chalcidoidea Eupelmidae). Annales de la Societe Entomologique de France (N.S.) 1: 817- 842. Flanders, R. V. and E. R. Oatman. 1982. Laboratory studies on the biology of Orgiliis jenniae (Hyme- noptera: Braconidae), a parasitoid of the potato tuberworm, Phthorimaea operculella (Lepidoptera; gelechiidae). Hilgardia 50: 1-33. Gauld, I. D. & B. Bolton (eds). 1988. The Hymenoptera. British Museum (Natural History/Oxford Uni- versity Press, Oxford. 332pp. Oeser, R. 1961. Vergleichend-MorphologischeUnter- suchungen iiber den Ovipositor der Hymenop- teren. Mitteihmgen aiis dem Zoologische Museum in Berlin 37: 1-119. Quicke, D. L. J. 1988. Host relationships in the Bra- coninae (Hymenoptera: Braconidae) — How little we know! Entomological Society of Queensland News Bulletin 16: 85-92. Quicke, D. L. J. 1991. Ovipositor mechanics of the braconine wasp genus Zaglyptogastra and the ich- neumonid genus Pristomerus. Journal of Natural History 25: 971-977. Quicke, D. L. J. and C. van Achterberg. 1990: Phy- togeny of the subfamilies of the family Braconi- dae (Hymenoptera). Zoologische Verhandelingen 258: 1-95. Quicke, D. L. J., L. C. Ficken, and M. G. Fitton. 1992a. New diagnostic ovipositor characters for doryc- tine wasps (Hymenoptera, Braconidae). Journal of Natural History 26: 1035-1046. Quicke, D. L. J., L. C. Ficken, and M. G. Fitton. 1992a. New diagnostic ovipositor characters for doryc- tine wasps (Hymenoptera, Braconidae). Journal of Natural History 26: 1035-1046. Quicke, D. L. J. and M. G. Fitton. In press. Ovipositor steering mechanisms in parasitic wasps of the families Gasteruptiidae and Aulacidae (Hyme- noptera). Preceedings of the Royal Society. Quicke, D. L. J., M. G. Fitton, J. R. Tunstead, S. N. Ingram, and P. V. Gaitens. 1994. Ovipositor structure and relationships within the Hymenop- tera, with special reference to the Ichneumono- idea. Journal of Natural Histoty 28: 635-682. Quicke, D. L. J. and P. M. Marsh. 1992. Two new species of Neotropical parasitic wasps with high- ly modified ovipositors (Hymenoptera, Braconi- dae: Braconinae and Doryctinae). Proceedings of the Entomological Society of Washington 94: 559- 567. Quicke, D. L. J. and R. A. Wharton. In prep. Antero- lateral scent glands of braconine wasps (Hyme- noptera: Braconidae). Shaw, M. R. and T. Huddleston. 1991. Classification and biology of braconid wasps. Handbooks for the Identification of British Insects. Vol. 7, Part 11. Roy- al Entomological Society, London. Smith, E. L. 1968. Biosystematics and morphology of Symphyta. I. Stem-galling Euura of the California region, and a new female genitalic nomenclature (Hymenoptera: Tenthredinidae). Annals of the En- tomological Society of America 61: 1389-1407. Smith, E. L. 1970. Evolutionary morphology of the external insect genitalia. 2. Hymenoptera. Annals of the Entomological Society of America 63: 1-27. Wharton, R. A., S. R. Shaw, M. J. Sharkey, D. B. Wahl, J. B. Woolley, J. B. Whitfield, P. M. Marsh and J. W. Johnson 1992. Phylogeny of the subfamilies of the family Braconidae (Hymenoptera: Ichneu- monoidea): a reassessment. Cladistics 8: 199-235. J. HYM. RES. Vol. 4, 1995, pp. 121-136 The Wasp Genus Tachytella Brauns, 1906 (Hymenoptera: Sphecidae) WOJCIECH J. PULAWSKI California Academy of Sciences, Golden Gate Park, San Francisco, California 94118, USA Abstract.— The southern African genus Tachytella Brauns includes aurcopilosa Brauns, 1906 (west- ern South Africa) as well as two new species: heliophila (western Namibia) and nama (southern Namibia, northwestern South Africa). A lectotype is designated for Tachytella aureopilosa var. nana Arnold, 1936, the latter name being synonymized with Tachytella aureopilosa Brauns, 1906. The genus is redescribed using previously known and newly discovered characters, a key to the spe- cies is provided, and phylogenetic relationships between the species are discussed. INTRODUCTION General. — Tachytella Brauns, a little known genus from southern Africa, was thought to be monotypic until now. Spec- imens are rarely collected (Bohart and Menke 1976, for example, knew of only six). A decade ago, I discovered an unde- scribed species in materials from the State Museum Windhoek that Ole Lomholdt, then at the Zoological Museum, Copen- hagen, kindly had transshipped to me. During our expedition to Namibia in 1990, Maximilian Schwarz and I collected spec- imens that represented another new spe- cies. These findings prompted me to re- vise the genus. I follow the format used for my revisions of Kohliella Brauns, Holo- tacfn/spliex de Beaumont, and Gastrosericus Spinola (Pulawski 1991, 1992, 1995). Technical Terms. — Most morphological terms follow Bohart and Menke (1976), but the mandibular terms are taken from Michener and Fraser (1978). The following terms are here redefined for convenience: Clypeal lobe: the projecting medioventral portion of the clypeus. Mandible: — adductor ridge: extends distad from the adductor swellings (on the inner side) and gradually becomes visible from the outside, constituting the distal part of the mandibular posterior margin; the two portions differ in size, the distal one being higher than the basal one; — condylar ridge: arises from the condyle, extends distad, and forms the basal por- tion of the posterior mandibular mar- gin; it is angulate distally in many Lar- rinae; — condyle: mandibular articulation on the occipital side of the head capsule; — notch: an emargination on the posterior margin, delimited basally by the con- dylar ridge and distally by the expand- ed portion of the adductor ridge; — posterior margin: extends between the condyle and mandibular apex; called externoventral margin by Bohart and Menke (1976) and lower margin by Michener and Fraser (1978); it actually consists of two components: the con- dylar ridge basally and the adductor ridge distally (the term posterior is pre- ferred because the head is hypognath- ous, and this edge is thus oriented pos- terad); Sternum, tergum: shortened terms gastral sternum, gastral tergum. Vertex: for — length: the distance between a hind- ocellus and an imaginary line connecting eye hindcorners (i.e., the point where the 122 Journal of Hymenoptera Research inner and the posterior portions of the or- bit meet); — width: the shortest interocular distance in the ocellar region. specimens that have Origin of Material. — The 29 examined apparently are all been collected so far. Institutional or per- sonal collections in which the material is deposited are abbreviated in the text as follows (names of contact persons are giv- en in parentheses): AMG: Albany Museum, Grahams- town, South Africa (Friedrich W. Cess). BMNH: British Museum (Natural His- tory), London, current nonstat- utory name: The Natural Histo- ry Museum (Loraine Ficken). CAS: California Academy of Sciences, San Francisco, California (Woj- ciech J. Pulawski). FSCA: Florida State Collection of Ar- thropods, Gainesville, Florida (Lionel A. Stange, James R. Wi- ley). MS: Maximilian Schwarz, Ansfelden bei Linz, Austria (personal col- lection). SAM: South African Museum, Cape Town, South Africa, including G. Arnold collection previously housed in Bulawayo, Zimbabwe (Margie A. Cochrane, Vincent B. Whitehead). FMP: Transvaal Museum, Pretoria, South Africa (Ilema Fourie). USNM: Smithsonian Institution, Na- tional Museum of Natural His- tory, Washington, D.C. (Arnold S. Menke). Genus Tachytella Tachytella Brauns, 1906:56. Type species: Tach- ytella aureopilosa Brauns, 1906, by monotypy. Diagnosis. — Tachytella is a member of Larrini (Bohart and Menke 1976) and thus has the hindocelli modified to flat, elon- gate scars (Figs. 8, 12, 20), the only rem- nant of a lens being a narrow, translucent band along the scar's outer margin (the band is broadly interrupted on the scar's posterolateral or lateral side, depending on its orientation). Within the tribe, Tach- ytella can be recognized by the following four characters: 1. The frons has a median, slightly raised, flat area that is delimited by a lat- eral sulcus (Figs. 6, 10, 18, 22, 28, 31), the sulci being subparallel and close to the or- bits. Similar sulci are found in some An- cistromma W. Fox, e.g., capax W. Fox, and in less specialized Parapiagetia Kohl such as genicularis (F. Morawitz) and odontosto- ma (Kohl), in which however, the frons is not raised. 2. The pronotal collar is rounded and reaches the scutum level, a condition also found in some Gastrosericiis, e.g. baobabiciis Pulawski and eurypus Pulawski. Tachytel- la, however, has three submarginal cells, whereas Gastrosericiis has two. 3. The submarginal cell III in the fore- wing is essentially rhomboidal (Fig. 1), a previously unused character, with the an- terior and posterior margins equal in length or nearly so; the proximal and dis- tal margins are weaklv sinuous, almost straight, nearlv parallel to each other. 4. The hindwing jugal lobe is short and ends well before crossvem cu-a; conse- quently, the jugal and anal excisions are widely separated (Fig. 1). A similar con- dition is found in some specialized Gastro- sericiis such as braunsi Arnold, herero Pu- lawski, and pulchellus R. Turner (Pulawski 1995). Description. — Bohart and Menke (1976) regarded Tachytella as monotypic and, consequently, considered as generic some structures that are found only in aureopi- losa but not in the other two species (e.g., such male structures as the edentate inner mandibular margin, dentate clypeal free margin, and sterna VI-VIII with dense, erect setae). They thought that the prono- tal collar not closely appressed to scutum Volume 4, 1995 123 0.5 mm 0.25 mm Figs. 1-3. Some of the generic characters of Tachytella. 1, fore and hindwing of aureopilosa (drawn from the holotype, the model Bohart and Menke 1976, used for their illustration); arrows indicate third submarginal cell, jugal excision, and anal excision. 2, anterior part of thorax of aureopilosa (arrow indicates pronotum). 3, foretarsus of female nama. 124 Journal of Hymenoptera Research also was a generic character, but in fact it is only individual. The following rede- scription takes into account the three spe- cies now recognized and also some struc- tures not considered previously. Included are those features that vary within other Larrini but are universal in Tachytella. Labrum flat, not emarginate apically, not or barely protruding beyond clypeal free margin; prementum convex but not compressed laterally; stipes nearly flat. Mouthparts not elongate. Paramandibular process broadly separated from back side of clypeus (mandibular socket open). Pos- terior mandibular margin notched; notch delimited proximally by angulate apex of condylar ridge, and distally by marked, roundly expanded distal portion of ad- ductor ridge. Clypeus produced into mes- al lobe, free margin of lateral section con- cave. Antennal socket narrowly separated from frontoclypeal suture. Inner orbits convergent above. Frons flat, with no gla- brous swelling above each antennal sock- et. Hindocellar scars elongate, broadly di- verging anterolaterad (their long axes forming an angle of about 120-140°), shorter than distance that separates each one from midocellus. Occipital carina ef- faced just short of hypostomal carina. Pro- notum anterodorsally with transverse groove; collar swollen, reaching level of scutum. Postspiracular carina evanescent, about as long as midocellar diameter. Ep- isternal sulcus originating near middle of subalar fossa, ending well before reaching anteroventral mesopleural margin; scrobal sulcus absent; subalar fossa not margined below; metapleural flange not expanded. Axilla the usual shape, i.e., not expanded or carinate. Propodeum short, distance be- tween spiracle and metanotum less than spiracle's length; dorsum setose through- out. No additional sclerites between me- tasternal apex and propodeum (no "pro- podeal sternum"). Forewing with three submarginal cells, none petiolate; recur- rent veins separate or (some aureopilosa) forming short petiole. Hindwing crossvein cu-a vertical (not inclined). Forecoxal apex not expanded into process. Hindcoxal dor- sum: inner margin not carinate. Hindtibia not margined. Forebasitarsus and apical tarsomeres without ventral spines. Claws in each pair equal in size. Hindtarsomere II long (0.6-0.7 X hindtarsomere I). Base of tergum I variable: with or without short, oblique carina that extends from each anterolateral corner. Lateral carina of tergum I complete, reaching tergal hind- margin. Tergum II not carinate laterally. Body without erect setae (except setae on apical sterna in male of aureopilosa); those on propodeal dorsum inclined obliquely toward head. Female. Inner mandibular margin with tooth. Forecoxa not carinate anteriorly, venter slightly convex; outer surface of foretibia without preapical spines. Foretar- sus with rake that consists of stiff, flat- tened, widely spaced spines (four on for- ebasitarsus, two on foretarsomere II); tar- someres IV: length more than apical width, apicoventral margin emarginate; tarsomeres V: apicoventral margin slightly arcuate. Claws without teeth. Tergum VI fairly convex: angle between lateral mar- gin of tergum and lateral margin of pygid- ial plate, in side view about 30-40°. Pygid- ial plate well defined (margined laterally), rounded apically, sparsely punctate and with thin, inconspicuous setae (Figs. 9, 21, 30). Sting sheaths somewhat flattened dor- soventrally. Male. Foretrochanter and forefemur not emarginate; forebasitarsus 1 with one rake spine (in basal half); outer, apical spine of foretarsomere II shorter than foretarso- mere III; sterna without velvety patches; tergum VII carinate apicolaterally, not de- pressed apically; sterna III- VI impunctate at least mesally; sternum VIII rounded apically or (some aureopilosa) minimally emarginate. Venter of penis valve finely denticulate subapically (Figs. 17, 27, 35). Character Polarization. — The tribe Larri- ni, of which Tachytella is a member, con- sists of two lineages, the subtribes Larri- Volume 4, 1995 125 heliophila aureopilosa nama aureopilosa heliophila nama Figs. 4a and b. Cladograms showing hypothetical cladistic relationships of the three species of Tachytella; 2: modified cleft of inner mandibular margin, 8: penis valve with medioventral tooth. na and Tachytina (Bohart and Menke 1976). The autapomorphy of Larrina is a unique frons: the midocellus is placed in a depression, and the transverse sweUing delimiting the depression anteriorly joins a linear swelling along the inner orbit. The hindocelli are oriented along a trans- verse line, and the transparent lens rem- nant is horseshoe-like, with the two branches parallel and close to each other, apparently the plesiomorphic condition within the tribe. In Tachytina, the frons is the usual plesiomorphic condition, with no depression or swelling, but the hin- docelli are oriented in a V-shaped pattern and the lens remnant is clearly derived: the anterior branch is longer than the pos- terior or (in Parapiagetia) the branches are roughly circular (open on the lateral side). Because of the shape of the lens remnant, Tacln/tella is a member of Tach- ytina, but its relationships within the sub- tribe are unclear. Bohart and Menke (1976) noted that Tacln/telln resembled Ancistrotnma and Larropsis Patton in the shape of the mandible, ocellar scar, fe- male foretarsal rake, and male sternum VIII. None of these similarities, however, appears to be a synapomorphy. I have found no single derived character linking Tacln/telln with any of the other nine gen- era of the subtribe. For the purpose of this analysis, I regard the Larrina and the re- maining Tachytina as the outgroup. The characters found in both Tacln/telln and the outgroup are considered plesiom- orphic, unless there is strong evidence that they have been acquired indepen- dently. The characters present only in some Tncln/tella but nc^t the outgroup are regarded as apomorphic. The character codings are: 0. ancestral, 1 and 2. derived. The following characters have been con- sidered: 1. Female mandible inner margin: sub- basal tooth: 0. well defined, 1. inconspic- uous, obtuse (as in Tnclnjtelln nureopilosa). The tooth is present in most members of the outgroup, but absent in Lnrrn Fabri- cius, the subgenus Motes Kohl of Liris Fa- bricius, in Parapiagetia, several specialized Gastrosericus, and in two Tachysphex Kohl {nefarius Pulawski and ramses Pulawski). Because of their many specializations, none of these are likely to share a unique common ancestor with Tacln/telln, and consequently I regard the presence of a well defined mandibular tooth in Tacliy- tella as ancestral. 2. Female mandible inner margin: cleft: 0. narrow, 1. shallow, obtusely angulate (as in Tacln/tella aureopilosa , 2. deep, broad (as in Tacln/tella nama). A cleft is present and narrow in most Larrinae (including Tacln/tella heliophila), but absent in Larra, the subgenus Motes of Liris, Dalara Ritse- 126 Journal of Hymenoptera Research 9 aureopilosa Q heliophiia nama ( Fig. 5. Collecting localities of Tachytella. ma, Paraliris Kohl, Holotachysphex, many specialized Gastrosericus, Parnpiagetia geni- cularis (F. Morawitz), Tachysphex nefarius and ramses. None of these is likely to share a unique common ancestor with Tachytel- la, and I therefore accept that a narrow cleft is ancestral in the latter genus. The broad, deep cleft is unique within the tribe. 3. Male clypeus: 0. without longitudinal swellings, 1. swellings present. The clyp- eal swellings of aureopilosa are unique within Larrinae and thus apomorphic. 4. Propodeum: 0. without V-shaped im- pression, 1. with V-shaped impression. The impression is unique to Tachytella au- reopilosa and thus apomorphic. 5. Male foretarsus: 0. tarsomere II not expanded, 1. tarsomere II expanded. The expanded tarsomere of Tachytella aureopi- losa is unique within the tribe and thus apomorphic. 6. Tergum I: 0. with a pair of short, bas- al carinae, 1. carinae absent. Within Lar- rini, the carinae are absent in Gastrosericus, Holotachysphex, Kohliella, Parapiagetia, Tach- Volume 4, 1995 127 0.25 mm 0.5 mm 8 f-mm 0.25 mm Figs. 6-9. Tachytella aiireopilosa, female. t>, head frontally. 7, clypeus. S, head dorsally. 9, pygidial plate. i/sphex, manv Tachvtes Panzer, and Larrop- sis chilopsidis (Cockerell and Fox). They are present m all other Larrini and other Sphecidae (those with a petiolate gaster are obvious exceptions). 1 therefore regard the absence of carinae in Tachytella aiireo- pilosa as derived. 7. Male gaster: 0. sternal setae all short, 1. setae of sterna VI-VIII long, erect. Ster- nal setae are short, appressed in most Lar- rinae, but long, erect on apical sterna in many Liris and in Tachytella aiireopilosa. These species are not likely to share a sin- gle common ancestor. In addition, the se- tal configuration c^f T. aiireopilosa is unique and I regard it to be derived. 8. Penis valve: 0. without medioventral tooth, 1. with medioventral tooth. This tooth, found m Tachuteila heliophila and nama, is unique within Larrinae and thus apomorpnic. The following 5 characters cannot be polarized because both states are com- monly found in the outgroup: 9. Gena: a. thin in dorsal view, b. thick in dorsal view. 10. Mesopleural vestiture: a. not con- cealing integument, b. concealing integu- ment. 11. Hindtarsomere IV: a. length about 1.0 X apical width, b. length about 1.2-1.3 X apical width. 12. Pygidial plate of female: a. apex with a row of setigerous punctures. 128 Journal of Hymenoptera Research 0.5 mm 0.5 mm 12 0.25 mm 11 0.25 mm 13 Figs. 10-13. Tachytella aureopilosa, male. 10, head frontally. 11, clypeus. 12, head dorsally with outline of vertex showing variation. 13, scape. b. apex without a row of setigerous punc- tures. 13. Male mandible: a. inner margin an- gulate, b. inner margin simple. Cladistic Analysis. — Distribution of the 8 polarized characters is given below: Number: ancestor: aureopilosa: heliophila: nama: 12345678 00000000 11111110 00000001 02000001 Only two derived characters are found in more than one species: character 2 (a modified cleft on inner mandibular mar- gin) is shared by aureopilosa and nama, and character 8 (penis valve with medioventral tooth) by heliophila and nama. The clado- gram resulting from character 2 (Fig. 4a) is heliophila + {aureopilosa + nama), the cladogram based on character 8 (Fig. 4b) is aureopilosa + {heliophila + nama). I see no reason to prefer one over the other. Life History. — Unknown. The presence, in the female, of a foretarsal rake and a pygidial plate suggests ground nesting. Geographic Distribution (Fig. 5). — Namib- ia and western half of South Africa. Volume 4, 1995 129 KEY TO THE SPECIES 1. Propodeal dorsum with V-shaped impression that begins behind each spiracle and ends at hindmargin (impression inconspicuous in smallest individuals); tergum I basally without pair of short, oblique carinae. Female: clypeus (except basally) with finely, closely punctate median impression that separates two unsculptured, longitudinal areas (Fig. 6); imier man- dibular margin with inconspicuous subbasal tooth (Fig. 6). Male: clypeus with pair of glabrous, longitudinal swellings, free margin with small, median tooth (Fig. 10); sterna VI- VIII with dense rows or tufts of erect setae (Fig. 15); penis valve without ventral tooth (Fig. 17) aurcopilosa Brauns - Propodeal dorsum without such impression; tergum I basally with pair of short, oblique carinae. Female: clypeal bevel without median row of small punctures; inner mandibular margin with well-defined subbasal tooth (Figs. 18, 28). Male: clypeus with no longitudinal swellings, free margin without small, median tooth; sterna without dense rows or tufts of erect setae; penis valve with ventral tooth (Figs. 27, 35) 2 2. Mesopleural vestiture conspicuous, fully concealing integument. Length of hindtarsomere IV 1.2-1.3 X apical width. Female: free margin of clypeal lobe weakly arcuate, corner well defined but not prominent (Fig. 18); inner mandibular margin with usual, narrow cleft (Fig. 18). Male: clypeal lobe broad, nonprominent (Fig. 22); inner mandibular margin sim- ple (Fig. 22) heliophila sp. n. - Mesopleural vestiture inconspicuous, integument easily visible. Length of hindtarsomere IV about 1.0 X apical width. Female: free margin of clypeal lobe sinuate, with prominent corner (Fig. 28); inner mandibular margin with broadly expanded cleft (Fig. 28). Male: clypeal lobe prominent, markedly narrowing anterad (Fig. 31); trimmal margin obtusely dentate (Fig. 31) nama sp. n. DESCRIPTIONS OF SPECIES (TMP), present designation, examined. New synonym. — As Tachytella aiireopilosa nana: Diagnostic characters for each species ^^^^^^ ^^^ ^^^^^ ^^7^^253 (illustration of are given m the key and are not repeated ^^^^ ^ead), 256 (new status, listed), individually in the descriptions below. Synonymy. — Arnold (1936) described a Tachytella aiireopilosa Brauns yar. nana for specimens that had red gas- Figures 6-17 tral segments I-III, shallower V-shaped „ , ^ ,, ,1, inn^ err i • imprcssion on thc propodcal dorsum, and Tnan/tella aiireo-vilosa Brauns, 1906:57, male, in- '^ ,,,,,, • . . : , , . .. u 1 . ^ truncate rather than shallowly emarginate correct original hyphenation. Holotype: -^ " male. South Africa: Cape Province: Willow- male sternum VIII. Because these charac- more (TMP), examined.— Arnold 1923:218 ters vary individually, I regard var. nana (revision), 1930:5 (listed); Bohart and Menke as a synonym of aiireopilosa. 1976:231 (illustration of ocelli), 253 (illustra- Description. — Gena thin in dorsal view tion of male head), 255 (illustration of (Figs. 8, 12). Vertex with triangular, gla- wings), 257 (listed), 258 (illustrations of pro- brous, impunctate, dull area that extends notum and scutum), 268 (illustration of fe- f^.^^ Q^bit toward midline. Head, thorax, male foretarsus), 279 (illustration of male ^^^^^^^ ^^^ femora minutely punctate, sternum VIII), 280 (illustration of male geni- , i.u i ^- I^ .*. ,^ ^^^..f talia); Gess 1981:19 (South Africa: nefting punctures no more than 1 diameter apart, probably in friable soils). hence integument mat except for shiny Tachytella aureo-pilosa race nana Arnold, 1936:29, clypeal swellings in male. Propodeal dor- female, male. Lectotype: male. South Africa: sum, in most specimens studied, with Cape Province: Resolution in Albany District pair of impressions that meet postero- 130 Journal of Hymenoptera Research 0.25 mm 0.25 mm 15 0.1 mm Figs. 14-17. Tachytella aiireopilosa, male. 14, foretarsomere. 15, gastral apex in profile. 16, volsella. 17, penis valve. mesally and delimit triangular area re- throughout. Length of hindtarsomere IV sembling propodeal enclosure of most about 1.0 X apical width. Gastral seg- Sphecidae (impressions barely visible in ment I without basolateral, oblique cari- smallest specimens, apparently a result of na. allometric growth). Propodeal hindface Mesopleural vestiture concealing integ- not ridged. Femora and tibiae setose ument. i Volume 4, 1995 131 0.5 mm 20 0.25 mm 19 0.1 mm Figs. 18-21. Tacln/tt'lla heliophila, female. 18, head frontally. 19, clypeus. 20, head dorsally. 21, pygidial plate. Head and thorax black but the follow- ing are reddish: mandible (apical third dark), scape, tegula, and humeral plate. Gastral segment I to I-III red, remainder black. Mid- and hindfemora black, forefe- mur (except dorsally), tibiae and tarsi red. Wings nearly hyaline. Terga I-V silvery fasciate apically. Female. — Mandible: inner margin di- vided by an obtuse angle into basal and distal portions, with shallow, obtusely an- gulate cleft. Clypeus with median, finely, densely punctate impression that sepa- rates two unsculptured, longitudinal areas (Figs. 6, 7); lobe free margin arcuate except markedly concave near corner, which is prominent (Figs. 6, 7). Length of flagello- mere I 2.0 X apical width. Vertex width 1.4 X length. Tergum V punctate and se- tose throughout. Pygidial plate with preapical row of setigerous punctures (Fig. 9). Length 7.3 mm Male. — Mandible: inner margin simple, not angulate (Fig. 10). Clypeus (Figs. 10, 11): lobe free margin with median tooth, corner well defined but not prominent; distance between corners about 0.8 X dis- tance between corner and orbit; middle section with a pair of longitudinal swell- ings that are either microsculptured, dull, or unsculptured, shiny. Scape elongate, swollen apicoventrally (Fig. 13). Length of flagellomere I 1.7-1.8 X apical width. Ver- tex width 1.6 X length. Foretarsomere II enlarged (Fig. 14). Tergum VII: punctures averaging about 1 diameter apart, some of them about 3 diameters apart. Apex of sternum VI, sternum VII, and VIII with 132 Journal of Hymenoptera Research 0.5 mm 24 0.25 mm 23 0.25 mm 25 Figs. 22-25. Tachytella heliophila, male. 22, head frontally. 23, clypeus. 24, head dorsally. 25, scape. erect setae; longest setae equal to basal mandibular width. Length: 6.1-8.5 mm. Volsella: Fig. 16. Penis valve without ven- tral tooth (Fig. 17). Geographic Distribution. — Western South Africa. Records.— SOUTH AFRICA: Cape Prov- ince: Colesberg (1 female, SAM), Gra- hamstown: Hilton, sand pit (1 male, AMG), NE Ouberg Pass 31 road km NE Montagu at 33°39'S 20°18'E (1 male, AMG), Resolution in Albany District (1 fe- male, 1 male, AMG; 1 male, SAM; 1 head- less female, 1 male lectotype of nana, TMP), 10 km W Steinkopf (1 female, FSCA), Willowmore (1 male, TMP, holo- type of aureopilosa), Worcester (3 males, BMNH; 1 male, USNM). Tachytella heliophila Pulawski, new species Figures 18-27 Name Derivation. — Heliophila, a Neolatin feminine adjective derived from two Greek words: helios, sun; and philos, a friend; with reference to the open, sunny habitat where the species was found. Description. — Gena thin in dorsal view (Figs. 20, 24). Vertex without impunctate, triangular area. Head, thorax, coxae, fem- ora and tibiae minutely punctate, punc- tures no more than 1 diameter apart, hence integument mat except for the un- sculptured, shiny clypeal bevel. Propodeal hindface not ridged. Femora and tibiae se- tose throughout. Length of hindtarsomere IV 1.2-1.3 X apical width. Gastral segment I basolaterally with short, oblique carina. Volume 4, 1995 133 mTTT^. 0.1 mm Figs. 26 and 27. Tachytella heliophila. 26, volsella. 27, penis valve. Mesopleural vestiture concealing integ- ument. Head and thorax black but the follow- ing are reddish: mandible (apical third black), clypeal bevel anteriorly, scapal venter apically, pronotal lobe, tegula, and humeral plate. Gaster red. Coxae and fem- ora black (femora red apically), tibiae and tarsi red. Wings hyaline. Terga I-V silvery fasciate apically. Female. — Mandible: inner margin with small subbasal tooth and narrow cleft (Fig. 18). Clypeus (Figs. 18, 19): free margin of lobe arcuate, corner angulate but not prominent; apicomedian portion of mid- dle section shiny, practically unsculptured (with only a few microscopic punctures). Length of flagellomere I 2.75 X apical width. Vertex width 1.7 X length. Tergum V punctate and setose throughout. Pygid- ial plate with preapical row of setigerous punctures (Fig. 21). Foretibia without preapical spines. Length 6.6-7.5 mm. Male. — Mandible: inner margin simple. not angulate (Fig. 22). Clypeus (Figs. 22, 23): lobe free margin arcuate, corner ob- tusely angulate; surface finely punctate and setose throughout except for narrow apical lip; distance between corners about 0.9 X distance between corner and orbit. Scape swollen ventrally (Fig. 25). Length of flagellomere I 2.1 X apical width. Ver- tex width 3.0 X length. Tergum VII: punc- tures averaging less than one diameter apart. Length 6.4 mm. Volsella: Fig. 26. Pe- nis valve with ventral tooth near mid- length (Fig. 27). Geographic Distribution. — Known from a single area in Namib desert, western Na- mibia. Records. — Holotype: female, Namibia: Karibib District: 65 km SW Usakos, 24 Feb. 1990, W.J. Pulawski (CAS). Paratypes: NAMIBIA: Karibib District: 55 km SW Usakos, 1 Mar. 1990, W.j. Pu- lawski (1 female, CAS); 65 km SW Usakos, M. Schwarz, 24 Feb. 1990 (1 female, MS), 1 Mar. 1990 (1 female, MS; 1 male, CAS). 134 Journal of Hymenoptera Research 0.5 mm 29 0.25 mm Figs. 28-30. Tachytella nama, female. 28, head frontally. 29, head dorsally. 30, pygidial plate. Tachytella nama Pulawski, new species Figures 28-35 Name Derivation. — Nama, a Hottentot tribe that immigrated from southern Af- rica into the central Namib; a noun in ap- position to the generic name. Description. — Gena thick in dorsal view (Figs. 29, 32). Vertex without impunctate, triangular area. Scutal punctures varying from mostly about 1 diameter apart (fe- male from Goegap) to many diameters apart (most specimens). Mesothoracic ven- ter with minute punctures that are several diameters apart. Propodeal hindface mi- croscopically, densely ridged and with ev- anescent punctures. Fore- and midfemo.ral venter with a few, scattered punctures, as- etose. Outer side of foretibia as well as mid- and hindtibial dorsum impunctate, asetose. Length of hindtarsomere IV about 1.0 X apical width. Gastral segment I ba- solaterally with short, oblique carina. Mesopleural vestiture inconspicuous, not concealing integument. Head and thorax black except mandible yellowish basally and reddish distally, middle clypeal lobe of male reddish, male flagellum yellowish brown ventrally. Gas- tral segments I-III red, remainder black. Wings almost hyaline. Terga I-III silvery fasciate apically, but fasciae evanescent in most specimens. Female. — Mandible: inner margin with subbasal tooth and broadly expanded cleft (Fig. 28). Clypeus (Fig. 28): lobe free margin sinuous, corner prominent; sur- face all or largely unsculptured (at most punctate basally). Length of flagellomere I 2.0 X apical width. Vertex width 2.4 X length. Tergum V with impunctate, gla- Volume 4, 1995 135 0.5 mm 32 0.5 mm 0.1 mm Figs. 31-35. Tachytella nama, male. 31, head frontally. 32, head dorsally. 33, scape. 34, volsella. 35, penis valve. brous apical depression. Pygidial plate with no preapical row of setigerous punc- tures (Fig. 30). Foretibia with two spines between inner surface and dorsum: one at midlength, the other at two thirds of length. Length 7.5 mm. Male. — Mandible: inner margin obtuse- ly angulate (Fig. 31). Clypeus (Fig. 31): middle section markedly convex; lobe prominent, markedly narrowing anterad, with unsculptured apical part and shal- lowly emarginate anterior margin. Scape minimally swollen ventrally (Fig. 33). Length of flagellomere I 1.5 X apical width. Vertex width 2.4 X length. Tergum VII: punctures averaging more than one 136 Journal of Hymenoptera Research diameter apart. Length 7.2 mm. Volsella: Fig. 34. Penis valve with ventral tooth (Fig. 35) near midlength. Geographic Distribution. — Southern Na- mibia and northwestern South Africa. Records. — Holotype: female. South Afri- ca, Cape Province: Hester Malan [now Goegap] Nature Reserve near Springbok, 15-21 Oct. 1987, F.W. and S.K. Gess (AMG). Paratypes: NAMIBIA: Liideritz District: Namuskluft 88, 2716 Dd [= between 27°45' and 28°00'S and 16°45' and 17°00'E], collector unknown, 12-15 Sept. 1973 (1 female, SMW), 7-15 Oct. 1970 (1 male, SMW). SOUTH AFRICA: Cape Province: Anenous, 29°14'30"S, 17°34'45"E, 11-13 Oct. 1988, D.W. Gess (2 females, AMG); Dikbome, Merweville Koup, [South Afri- can] Museum Expedition, Oct. 1952 (1 male, SAM); Hester Malan [now Goegap] Nature Reserve near Springbok, 15-21 Oct. 1987, F.W. and S.K. Gess (1 female, CAS); same data but 10-12 Oct. 1988 (1 male, AMG); between Kamieskroon and Springbok, [South African] Museum Staff, Oct. 1952 (2 females, CAS, SAM). ACKNOWLEDGMENTS Illustrations were prepared by Mary Ann Tenorio on a Macintosh computer using the Adobe Illustrator program. Fieldwork in Namibia in 1989, where I col- lected specimens, was supported by the National Sci- ence Foundation (grant number BSR-8722030) and museum work in South Africa the same year by the California Academy of Sciences through its In-House Research Fund. I sincerely thank persons who sent specimens for study. I am indebted to Arnold S. Men- ke and Friedrich W. Gess for their critical reviews of the manuscript. I sincerely thank my friend and trav- el companion to Namibia, Herr Maximilian Schwarz, who generously donated the unique male of Tachy- tella heliophila to the California Academy of Sciences. LITERATURE CITED Arnold, G. 1923. The Sphegidae of South Africa. Part III. Annals of the Transvaal Museum 9:191-253. Arnold, G. 1930. A check-list of the Sphegidae of the Ethiopian Region. University Press, Cambridge, England. 21 pp. Arnold, G. 1936. New African Hymenoptera No. 3. Occasional Papers of the Rhodesian Museum 5:1-38, pi. I. Bohart, R.M., and A.S. Menke. 1976. Sphecid zoasps of the world. A generic revision. University of Cali- fornia Press, Berkeley, Los Angeles, London. 1 color plate, IX + 695 pp. Brauns, H. 1906. Zur Kenntnis der siidafrikanischen Hymenopteren. Verhandlungen der Zoologisch-Bo- tanischen Gesellschaft in Wien 56:43-59. Gess, F. 1981. Some aspects of an ethological study of the aculeate wasps and the bees of a karroid area in the vicinity of Grahamstown, South Af- rica. Annals of the Cape Provincial Museums (Nat- ural History) 14:1-80. Michener, CD. and A. Fraser. 1978. A comparative anatomical study of mandibular structures in bees. The University of Kansas Science Bulletin 51: 463^82. Pulawski, W.J. 1991. A revision of the wasp genus Kohliella (Hymenoptera: Sphecidae). Proceedings of the California Academy of Sciences 47:289-302. Pulawski, W.J. 1992. World species of the wasp ge- nus Holotachysphex de Beaumont (Hymenoptera: Sphecidae). Proceedings of the Etitomological Society of Washington 94:223-242. Pulawski, W.J. 1995. The wasp genus Gastrosericus Spinola, 1839 (Hymenoptera: Sphecidae). Mem- oirs of the California Academy of Sciences (in press). J. HYM. RES. Vol. 4, 1995, pp. 137-193 Revision of the Ant Genus Gnantptogenys in the New World (Hymenoptera: Formicidae) John E. Lattke Institute de Zoologia Agricola, Universidad Central de Venezuela, Apartado 4579, Maracay, Venezuela 2101-A Abstract. — Ants of the genus Gnmnptogenys Roger in the New World are revised. The study is based mostly on worker morphology; 73 species are recognized, fifteen of which are new: G. andina, G. hoUviensis, G. brunnea, G ciineiforma, G. ejuncida, G. extra, G. gentri/i, G. latkephala, G. ilimaui, G. nigrivitrea, G. pilosa, G. siapensis, G. stellae, G. transversa and G. volcano. G. ericae is revalidated. In the present work sixteen names have been synonymized, and a morphological synopisis of the genus is given. The species range from southern U.S. to northern Argentina, but most are found in Central America and tropical South American forests. The evolutionary history of the genus is studied using phylogenetic systematics criteria. Six species groups are recognized: striatula group (22 species); niiniita group (9 species); rastrata group (14 species); sulcata group (8 species); nwrdax group (17 species), and the concinna group (3 species). A key for the identification of workers is presented. Illustrations of morphological features supplement the key and descrip- tions. INTRODUCTION The genus Gnamptogem/s Roger is a di- verse group of ponerines with a distribu- tion that covers the Neotropics, Indoma- laysia and parts of the Nearctic and Aus- tralia. The genus as treated here is mostly the result of extensive synonymy by Brown (1958) in his revision of the Tribe Ectatommini. In this treatment, however, the minuta group species (= Alfarin Forel group of previous authors), are excluded and treated elsewhere (Brandao and Latt- ke 1990; Lattke 1992). Since Brown (1959), research on Gnamptogem/s has consisted mostly of the description of new forms and scattered bits of ecological informa- tion. The only known fossils of the genus are twc^ species reported from Dominican Amber by Baroni Urbani (1980). Lattke (1990) reviews the Venezuelan species and their natural history. Due to the growing number of specimens in collections, and the additional information published c^r accumulated in notes, a revision of the New World species seemed desirable. MATERIALS AND METHODS The revision is based primarily on worker morphology and the species have been defined conservatively because of considerable interpopulation variation. The criteria for judging allopatric popula- tions as conspecific are as in Ward (1984: 131): if they are at least as dissimilar as closely related, sympatric species and are not connected by known intermediate forms. For a phylogenetic analysis, the ge- nus Ectatomma Fr. Smith is used as an out- group in order to help determine the char- acter state polarities. This genus is consid- ered very close to Gnamptogemjs and pre- sents a greater number of primitive character states than either Gnamptogemjs or Rhytidoponera Mayr (Lattke 1994). Spe- cies groups were determined and then compared using a character matrix (Table 1) and Farris's Hennig '86 v. 1.5 phyloge- netic inference program. The "ie-" option was used to generate the most parsimo- nious tree (Fig. 19) and a nelsen consensus 138 Journal of Hymenoptera Research Table 1. Character state matrix 123456789 0123456789 012345 Ectatomma 000000000 0000000000 000000 LEMQ G. haenschi 110001101 0010000000 012110 G. concinna 100011111 1110000001 000001 G. schmitti 100001021 1010000011 001010 strigata sbgp. 010010020 0000011201 000100 porcata sbgp. 000010020 0000011201 000100 LNKD striatuln sbgp. 000010020 0000011200 000100 moniax group 111111121 111-011102 000011 MCZC sulcata group 110111121 1111011122 001011 rastrata group 110011120 0010111202 000010 miuiita group 100011000 0010010001 110110 MHNG MIZA tree calculated from the 13 most parsi- moniuos trees calculated by the "ie" op- tion. Due to logistical difficulties, it was not possible to examine some types, this is especially true for European specimens. Thus this revision depends, in such cases, upon Brown's concept of the species in- volved, either through examining species determined by him during the course of his studies, reading the notes he took while visiting European collections in 1963 or from his 1958 revision. When type material was examined it is followed by a "T" in the species list. Specimens were examined from or de- posited in the following collections: BMNH The Natural History Museum, London, U.K. CMLB Coleccion Martha Lucia Baena, Cali, Colombia CASC California Academy of Sciences, San Francisco, California, U.S.A. CFFC Coleccion Fernando Fernandez, Bogota, Colombia CKWJ Coleccion Klaus Werner Jaffe, Universidad Simon Bolivar, Ca- racas, Venezuela CNIC Cornell University Insect Col- lection, Ithaca, New York, U.S.A. CUNB Coleccion Universidad Nacion- al, Bogota, Colombia JTLC John T. Longino Collection, Ev- ergreen College, Olympia, Wash- ington, U.S.A. LACM Los Angeles County Museum of Natural History, Los Ange- les, California, U.S.A. Lyman Entomological Muse- um, McGill University, Ste. Anne de Bellevue, Quebec, Canada. Landessammlungen fiAr Natur- kunde, Karlsruhe, Germany Museum of Comparative Zool- ogy, Harvard University, Cam- bridge, Massachusetts, U.S.A. Museum d'Histoire naturelle, Geneva, Switzerland. Museo del Instituto de Zoologia Agricola, Universidad Central de Venezuela, Maracay, Vene- zuela MUSP Museu Universidade de Sao Paulo, Sao Paulo, Brazil NHMW Naturhistorisches Museum, Vi- enna, Austria. PSWC Philip S. Ward Collection, Uni- versity of California, Davis, California, U.S.A. USNM United States National Muse- um, Washington, D.C., U.S.A. WPMC William P. MacKay Collection, University of Texas, El Paso, Texas, U.S.A. Measurements and Indices Morphological measurements were made at magnifications of up to 100 X with an ocular micrometer on a Wild M5 ster- eoscope. Since ratios derived from the measurements are conveniently scaled in- dices by themselves, the orthodox proce- dure of multiplying them by 100 was not followed. The measurements follow those of Lattke (1990). HL Head length: midline length of head proper, measured in full-face (dorsal) view, from the anterior clypeal margin to the midpoint of a line drawn across the vertexal margin. ML Mandibular length: length of Volume 4, 1995 139 closed mandibles, from the anteri- or clypeal margin to the apices of the mandibles, measured in the same plane as HL. SL Scape length: length of the first an- tennal segment, excluding the neck and basal condyle. ED Eye Diameter: diameter of eye, measured along its long axis in lat- eral view HW Head width: maximum width of head, measured in full-face (dor- sal = frontal) view, excluding the eyes. WL Weber's length of the mesosoma (alitrunk): diagonal length mea- sured in lateral view, from the an- terior margin of the pronotum (ex- cluding collar) to the posterior ex- tremity of the metapleural lobe. CI Cephalic index: HW/HL MI Mandibular index: ML/HW SI Scape index: SL/HW OI Ocular index: ED/HW SSC Scape setal count: the number of standing hairs (not pubescence), i.e. those forming an angle of 45° or more with the cuticular surface, visible in outline on the upper sur- face of the scape, with the line of view at right angles to the plane of funicular inflexion. New World Gnamptogenys Roger Generic Synopsis (Worker). — Sculpture consisting mostly of even parallel costae, costulae or strigae, occasionally rugose; distinct dorsomedian cephalic carinae usually absent, at most never extending more than half-way between clypeus and vertex; eyes at midlength of head or pos- terad, surrounded by a fine groove; lobes of frontal carinae broadly convex and par- tially raised, incompletely covering anten- nal condyles (except miniitn group); fun- niculus filiform or incrassate, never with a distinct club; anterior clypeal border with narrow lamella of variable length; mese- pisternum anteroventrally bordered by narrow lamella; pronotum unarmed, without protuberances; propodeal spira- cles round or slightly ovoid, never slit- like; transverse sutures and grooves on mesosomal dorsum present or absent; mesonotum never prominently convex and bulging; anterior prostemal process bidentate; anterior mesepisternal process produced as thin triangular lobe with pointed or bluntly pointed apex; metepis- ternum with deep posterior cleft for peti- olar insertion and continuous with open metacoxal fossae; metepisternal process located anterad of cleft and of variable de- velopment; metacoxal dorsum usually with denticle, lobe or tubercle, absent in some species; helcium protruding medi- anly on anterior postpetiolar face; tergite of helcium much larger than sternite; for- tibial apex lacking stout moveable setae; outer border of foretarsal comb (opposite calcar) usually with single prominent seta; meso- and metatibial spurs one or two (weakly developed), barbulate or simple; empoida lacking. In most species the pre- tarsal claws are bidentate on all legs, and the median tooth may vary in its position among species, and the claws may not necessarily be alike on all legs. In small species the claws may be hard to observe. Malpighian tubule number six (Brown 1988). The genus is so diverse that char- acterizing it can be difficult. The following characters can be considered synapomor- phies of the genus which distinguish it from its closest relatives: Ectatomnia and Rhi/tidoponera: the single stout moveable seta on the foretibial apex and a spine or tubercle on the metacoxal dorsal surface. This former trait is lacking in some spe- cies. The following combination of char- acters are synapomorhic for the genus if one excludes the problem species dis- cussed further on in the text: inconspicu- ous metanotum, disappearance of the me- dian cephalic carina, lack of row of stout setae on foretarsal base opposite the strig- il, leaving only one seta. Characters. — The choice of characters for 140 Journal of Hymenoptera Research 1 8 Figs. 1-8. Lateral view of mandibles and anterior section of head of Ectatomma and Gnmnpiogemjs. Transverse section of mandible depicted in black with chewing border to the left and frontal surface below. 1, E. niidum; 2, G. porcata; 3, G. moellerius; 4, G. liaenschi; 5, G. triangularis; 6, G. mordax; 7, G. tortuolosa; 8, G. concinna. use is complicated by convergence in sev- eral instances. Some character states con- sidered apomorphic have apparently been reached independently by one or two spe- cies in different lineages. Examples are the smooth and shining scapes of G. rastrata and of G. mordax, or the conspicuous me- tanotal groove of G. bisulca and G. brimnea. There are also some apparent reversals, as well as retention of plesiomorphies pres- ent in the outgroup. But these cases are few, isolated and do not represent the ma- jor trends of their respective species groups. Therefore, they were not taken into account for defining group character states. Mandibular shape is an apparent useful indicator of phylogenetic trends, with the general direction being from ple- siomorphic triangular mandibles to apo- morphic forms as subtriangular or subfal- cate. Subtriangular mandibles have been independently derived in the haenschi, sul- cata, and mordax groups, as indicated by other mandibular traits such as the config- uration of the chewing border, sculpture and cross section. In the last two groups (Figs. 6-8), the mandibular cross section reveals thick, robust mandibles as op- posed to the slim cuneiform section of G. haenschi, which is the plesiomorphic state Volume 4, 1995 141 11 Figs. 9-14. Lateral view of propodeum of Ectatomma and Gnamptogenys. 9, E. lugeus; 10, £. quadridens; 11, E. tuberculatum; 12, G. bispinosa; 13, G. perspicax; 14, G. triangularis. exhibited by the outgroup and the striatula and rastrata groups (Figs. 1-5). The sub- petiolar process in its primitive state is tri- angular, lobe-like and anteriorly protrud- ing. Modification has been towards a sub- quadrate shape, but some species, such as G. striatula, exhibit both conditions, and in the derived mordax subgroup the triangu- lar subpetiolar porocess is probably a re- versal. Some characters are shared by the out- group and certain groups or subgroups of Gnamptogenys. Two of these are particu- larly important: the propodeal denticles and the setae on the foretarsal base. The 16 Figs. 15-16. Outer lateral view of apex of tibia and base of tarsus of foreleg. 15, typical configuration in Ectatomma with single calcaria on tarsal apex and row of setae in foretarsal concavity; 16, same in most Gnamptogenys with only one stout seta on foretarsal concavity. propodeal denticles of Ectatomma may be absent in some species and when present, they are quite modest, usually lobe-like and directed dorsally. The declivitous pro- podeal face runs evenly and directly to the denticular apex (Figs. 9-11). The denticles in the rastrata group are stout and inclined posterad, and their base forms a more abrupt separation from the propodeal de- clivity (Figs. 12-14). With the exception of a few highly derived species, propodeal denticles are lacking in Rhytidoponcra and the striatula group of Gnamptogenys. This suggests that the presence of propodeal denticles is probably independently de- rived in Ectatomma and Gnamptogenys and consequently was lacking in the ancestral stock of this group of genera. For this rea- son this state has been given value in the character matrix for Ectatomma. The base of the foretarsus, in the concavity oppos- ing the strigil, has an outer row of stout setae in Ectatomma, but only one promi- nent seta in most Gnamptogenys (Figs. 15- 16). The exceptions are all the minuta- group species besides three other species, and each of these exceptions are unique within the genus as will be discussed later on. The brief anterior peduncle of the pet- iole, typical of Ectatomma is considered plesiomorphic for the porcata subgroup. 142 Journal of Hymenoptera Research 0^ 17 18 Figs. 17-18. Ventral view of the anterior postpetiolar process. 17: a, G. triajjgularis; b, G. haenschi; c, G. striatiila. 18, G. sulcata. but autapomorphic for the mordax sub- group. The following characters were taken into account for all the species groups or individual species in the matrix, but have been restricted to the minuta sub- group of the minuta group (Lattke 1991). Characters 8, 17, 18 were considered non- additive. The following character states were taken into account: 1. Head shape in frontal view: 0= wider posterad than anterad, l=subquadrate/ other. 2. Funiculus shape: = filiform to subfi- liform, l = incrassate. 3. Scape length: 0= surpassing vertexal border (Figs. 39, 41), l = does not surpass, at most reaching the vertexal border (Fig. 43). 4. Scape sculpture: 0= scapes with ap- preciable sculpture, 1 = smooth and shin- ing. The aforementioned sculpture usually consists of punctulae, punctae or striae. 5. Development of the clypeal lamella: 0= poorly developed, very short, l=well developed. This character refers to how far anterad the lamella extends. In poorly developed cases the longitudinal cephalic costulation is not clearly delimited from the lamella. 6. Shape of anterior clypeal margin: 0= convex clypeus (Figs. 33, 37, 39, 41), 1 = straight/concave (Figs. 34-36, 62, 64). 7. Shape of lateral extremities of ante- rior clypeal lamella: O=softly curves (Figs. 33, 37, 39, 41), 1= angular /pointed (Figs. 34-35). 8. Presence of median clypeal carinae: 0= present, 1= variable, 2 = absent or indis- tinct. 9. Mandibular shape: = triangular (Figs. 39, 41), l=subtriangular/subfalcate (Figs. 43, 54, 69, 62). 10. Shape of cross section at apical one third of mandible: = thin and cuneiform (Figs. 2-5), l=wide and robust (Figs. 6, 7). 11. Sculpture of mandibular dorsum: 0=striae/rugulae (Fig. 54), l=smooth and shining (Fig. 43). 12. Development of promesonotal su- ture: = deeply impressed, cleaving sculp- ture (Figs. 46, 49, 52), 1= weakly im- pressed, effaced (Fig. 55). 13. Development of metanotal groove: 0= deeply impressed (Fig. 55), l=weak to effaced (Figs. 46, 49, 52). 14. Development of propodeal denti- cles: 0= absent (Figs. 46, 49, 52), 1= present (Figs. 12-14). 15. Position of spiracle on propodeal side, in lateral view: = anterad, well sep- arated from declivitous propodeal face, 1= posterad, close to declivitous propo- deal face. 16. Presence of setae on foretarsal base opposite the strigil: O = row of stout setae (Fig. 15), l=one prominent seta only (Fig. 16). 17. Development of dorsal metacoxal Volume 4, 1995 143 process: = absent, 1 = variable, 2= well de- veloped (Figs. 46, 49). 18. Shape of the metasternal process: = stout denticles or tubercles, l=acicular, straight denticles, 2=acicular, arched den- ticles. 19. Shape of petiolar node: 0=erect and high node (Figs. 46, 49, 52), l=high and posterad, 2 = subcylindrical/other (Figs. 61, 63, 65). 20. Petiolar spiracles: = directed obliquely; 1= directed ventrally. 21. Petiolar spiracles: = -at the same level as the surrounding integument; l=in a distinct depression. 22. Shape of postpetiolar process in lat- eral view: O=broad (Fig. 17), l=bluntly pointed (Fig. 18). 23. A brief median carina on the dor- sum of the postpetiolar process: 0= absent, 1= present. This is a low carinae that starts at the anterior margin of the process and briefly extends posterad. 24. Abdominal segment IV: 0=with a ventral stridulitrum, 1= without a ventral stridulitrum. Usually recognizable as a longitudinal band that refracts light into rainbow colors. 25. Shape of the fourth abdominal seg- ment: = ventrally reflexed (Fig. 55), l=straight (Fig. 44). Species Groups. — Within the New World Gnamptogenys six species groups are de- fined here: striatula, sulcata, concinna, min- uta, mordax and rastrata. The internal clas- sification of the genus is depicted in Table 2. The characters used in defining the fol- lowing informal groupings are diverse, thus they may or may not be synapomor- phies. A question mark following the spe- cies name means that its position within the group is not clear. \. striatula group: mandibles triangular with rugulae or striae on frontal surface; convex anterior clypeal margin; scapes usually surpassing vertexal margin, sculp- turing punctate or vermiculate, never smooth and shining; eyes slightly behind cephalic midlength; head wider posterad Table 2. Internal classification of Gnamptogenys (except minuta group) striatula group porcata subgroup: acuta, brunnea, ejuncida, extra, gentryi, gracilis, nigrivitrea, pilosa, pleurodon, porcata striatula subgroup: ammophila, moelleri, striatula strigata subgroup strigata complex: andina, bisulca, ilimani, pitti- eri, strigata haytiana complex: haytiana, inina, reichenspergi, rclicta rastrata group rastrata subgroup: hispinosa complex: bispinosa, perspicax rastrata complex: cuneiforma, ingeborgae, lanei, lineolata, lucaris, mecotyle, meiiozzi, rastrata, triangu- laris banksi subgroup: banksi complex: laticephala, mediatrix semiferox complex: semiferox sulcata group sulcata subgroup: fernandezi complex: fernandezi sulcata complex: acuminata, curvoclypeata, tor- tuolosa ericae subgroup: ericae complex: ericae lucaris complex: lucaris siapensis volcano complex: volcano mordax group regularis subgroup: hartmaui complex: bruchi, hartmani, transversa regularis complex: horni, nana, pristina, regularis, rimulosa ainndata complex: annulata, kempfi mordax subgroup: continua, boliviensis, interrup- ta, mordax, stellae alfaroi subgroup: alfaroi levinates subgroup: levitjates concinna group concinna, haenschi, schmitti than anterad; propodeal spiracle close to declivity; anterolateral propodeal declivity without spines or tubercles; anterior pros- ternal process broadly concave medianly; metacoxal dorsum always with denticle or lobe; high petiolar node; relatively wide anteroventral postpetiolar process; second 144 Journal of Hymenoptera Research gastric segment ventrally arched. Most view; anterior clypeal margin usually seem to be generalist predators, but noth- straight; mandibular front usually striate ing is yet known about the diet of the stri- or rugulose, sometimes smooth; scapes gata subgroup species. usually surpassing vertex, sometimes with Three subgroups are recognizable: 1) longitudinal rugulae, vermiculate or striatula subgroup: petiolar node erect and smooth and shining; promesonotal suture sessile; subpetiolar process variably feebly impressed to absent, never totally shaped, either subquadrate or triangular dividing sculpture; metanotal suture well and projecting anterad in lateral view. 2) impressed; propodeum usually armed porcata subgroup: petiolar node posterior- with denticles or spines; petiolar node ly inclined, with short anterior peduncle; low; subpetiolar process shape variable, subpetiolar process lobe-like, projecting usually projecting anterad but sometimes anterad in lateral view. 3) strigata sub- subquadrate; metacoxal teeth always pres- group: petiolar node variably shaped; sub- ent, usually acicular; second gastric seg- petiolar process subquadrate with a cu- ment ventrally arched. The ants of this neiform ventral edge (Fig. 30) as opposed group are specialized millipede predators to the uniformly parallel sides of the other and can be subdivided into two sub- striatula group species (Fig. 29). Many of groups. 1) rastrata subgroup: mandibles the species have their propodeal spiracles triangular. Within the subgroup two com- elevated on small prominences. Most spe- plexes are defined: a) bispinosa complex: cies are smaller in size than those of other large species, anterior clypeal margin con- subgroups and all are cryptobiotic leaflit- vex, propodeal teeth above level of spira- ter inhabitats. Two species complexes can cles; anterior lobe of subpetiolar process be pointed out in this subgroup: a) strigata prominent, b) rastrata complex: smaller complex: cephalic vertex sculptured; no species; anterior clypeal margin straight; distinct propodeal lobes; petiolar node not propodeal teeth at same level as spiracles; parallel-sided in lateral view, anterior subpetiolar process subquadrate. 2) banksi margin usually convex (Figs. 40, 57). All subgroup: mandibles subfalcate to falcate, species inhabitat cloud forests, mostly of Two species complexes are defined, a) the Andes, with the exception of one banksi complex: metacoxae and propo- found in the Cordillera de la Costa of Ven- deum with denticles, b) semiferox complex: ezuela. b) haytiana complex: cephalic ver- metacoxae and propodeum lacking denti- tex mostly smooth and shining; small pro- cles. In a study of the sting apparatus of podeal lobes present; petiolar node several species of Gnamptogenys Kugler strongly compressed with subparallel an- (1991) grouped triangularis and bispinosa, terior and posterior faces (Fig. 21); some the only members of the rastrata group as species have modified sculpturing from defined above, into one group, the usual costulate patterns of the genus. IIL mordax group: antennal scapes usu- Most species inhabitat the Amazon-Ori- ally do not reach vertex and are com- noco river basin, except for one species pressed, smooth and shining; mandibles found on Hispaniola Island. Kugler (1991) subtriangular to subfalcate with front obtained similar results studying the sting smooth and shining; head subquadrate to apparatus of brunnea, nr. strigata, porcata, elongate; mesosoma mostly parallel sided; moelleri and gracilis. They were the only small propodeal lobes frequently present, members of the striatula group included, denticles absent (except one species); me- as defined above, and were grouped to- tacoxal denticles or lobes frequently lack- gether. ing; second gastric segment straight, with- II. rastrata group: head subquadrate or out ventral arching. Considerable inter- wider anterad than posterad in frontal population variability in the average size Volume 4, 1995 145 is observed in species of this group. The group can be divided into 4 subgroups: 1) regulciris subgroup: mandibles subtrian- gular; anterior clypeal border straight and usually with laterally acutely pointed la- mella; no transverse dorsal mesosomal su- tures; petiolar node lacking anterior pe- duncle; subpetiolar process subquadrate. Within this subgroup three species com- plexes are recognizable, a) hartfiiaiii com- plex: mandibles triangular to subtriangu- lar, frontal surface partially sculptured; vertical sculpturing on declivitous propo- deal face; b) regiilaris complex: mandibles subtriangular, their frontal sides smooth and shining; vertical sculpturing on de- clivitous propodeal face; c) annulata com- plex: mandibles subtriangular, front smooth and shining; horizontal sculptur- ing on propodeal declivity. 2) mordax sub- group: mandibles subfalcate, frontal sur- face smooth and shining; anterior clypeal margin concave with rounded lamellar sides; well impressed metanotal groove; petiolar node slightly pedunculate; sub- petiolar process projects anterad. 3) alfaroi subgroup: mandibles triangular, head prominently elongate; denticles on pro- podeum. The elongate head and armed propodeum makes this subgroup quite distinct from the others. 4) levinates sub- group: subtriangular mandibles, frontal surface sculptured; scapes surpassing ver- tex; subpetiolar process subquadrate. Its clypeus projects anterad more than in any extant species of the group. IV. sulcata group: mandibles subtrian- gular, front smooth and shining; head su- bquadrate; anterior clypeal border usually straight; lamella usually laterally angular; scapes smooth and shining; metacoxal denticles wanting in some species; usually no transverse mesosomal sutures; petiolar node low, lacking peduncle; metasternal process acicular and arched; second gas- tric segment relatively straight; anteroven- tral postpetiolar process relatively narrow and bluntly pointed, in contrast with wid- er process of other species. Two sub- groups can be pointed out. 1) sulcata sub- group: posterior face of petiolar node with horizontal costulae. Two species complex- es are recognizable: a) fernandezi complex: mandibles triangular; clypeal lamella con- vex, b) tornata complex: mandibles sub- triangular; clypeal lamella straight. 2) sul- cata subgroup: posterior node face with longitudinal costulae. This subgroup is made up of three complexes, a) ericae com- plex: clypeal lamella laterally acutely an- gulate; mandibles subtriangular. b) lucaris complex: clypeal lamella laterally round- ed; mandibles subtriangular. c) volcano complex: clypeal lamella laterally obtusely angulate; mandibles triangular. V. concinna group: head subquadrate; scapes surpassing vertex in frontal view; median clypeal carina frequently present; body sculpture densely striate; small pit frequently situated medianly on weakly impressed promesonotal suture; metaster- nal process slender to acicular; row of stout setae on base of foretarsus opposite strigil present; petiolar node with no an- terior peduncle; metacoxae and propo- deum unarmed. All three members of the group are large species. This group is nev- ertheless heterogeneous and its three members were considered individually in the matrix. VI. minuta group: head subquadrate; frontal carinae broadly expanded laterad; row of stout setae on base of foretarsus opposite strigil present; petiolar spiracles facing directly ventrad and sunken within a pit. Phylogenetics. — The possible evolution- ary history of the six species groups is de- picted in Fig. 19. Fig. 19 was computed by the "ie-" option of Hennig '86. Figure 20 is a nelsen consensus tree calculated from 13 trees generated by the "ie" option of Henning '86. The striatula group subgroups are con- sidered a sister group to the rest of the Gnamptogem/s species considered. The sul- cata and mordax groups are the most de- rived and are considered sister groups. 146 Journal of Hymenoptera Research 19 Ectatomma strigata subgroup porcata subgroup striatula subgroup minuta group rastrata group G. schmitti G. haenschi G. concinna mordax group sulcata group 20 Ectatomma strigata subgroup porcata subgroup striatula subgroup G. haenschi G. concinna G. schmitti rastrata group minuta group mordax group sulcata group Figs. 19-20. Cladograms generated from the data in Table 1 by Hennig '86. 19, most parsimonious tree found by the "ie-" option, length 51, ci 58. 20, Nelsen consensus tree generated from the 13 most parsimonious trees found by the "ie" option, length 51, ci 58. These results closely parallel Kugler's (1991) findings based upon the sting ap- paratus of several Gnamptogeni/s species. Within each of these groups we find spe- cies that apparently represent the transi- tion from striatida-\ike ancestors and thus give additional support to the program re- sults: the hartmani complexes of the mordax group, and the volcano and fernandezi com- plexes of the sulcata group. The rastrata group is probably closer to the striatula group than to the mordax or sulcata groups on account of characters such as triangular mandibles, long and usually sculptured scapes, the convex clypeal lamella of the bispinosa complex and the well developed metacoxal tooth. The consensus tree rec- ognizes two major groupings within the genus: striatula, and the rest. The "rest" make up a six-branched polytomy. Clearly further studies will be needed in the fu- ture in order to gain a better picture of the phylogenetic history of Gnamptogenys. The concinna group is probably not monophyletic. It is diverse in mandibular shape, and in other details such as the de- velopment of the clypeal lamella or me- dian clypeal carina. These traits as well as vestigial promesonotal suture are unusual remnants of plesiomorphies when com- pared with most other members of the ge- nus. G. haenschi retains more plesiomor- Volume 4, 1995 147 phies than the other two species and could be considered closer to the rootstock of the group. All are apparently specialized, ei- ther in feeding or nesting habits. This group opens a number of questions which can not be answered to satisfaction at present. Are the unarmed metacoxae ple- siomorphic or apomorphic for this group? Could they represent relicts of an other- wise extinct lineage (or lineages) in Gnamptogenys history? Interesting paral- lels can be found in the shallow areolate ground sculpture and well-developed ver- texal lobes of mimita subgroup members and those of several Old World species groups. A comparative study with Old World species could shed new light on the phylogeny of the genus. The striate sculp- ture of the concinna group can also be found, in a lesser degree, in the suniilans complex. The origins of Ectatommini can be traced to the late Cretaceous to early Ter- tiary, at least after the separation of Africa from Gondwana (80-90 Ma), but before the separation of South America from Pa- leoantarctica (36-41 Ma) (Lattke 1994). Given the presence of Gnamptogenys in the Indomalaysian Region and their absence in Africa, the origin of the genus also ex- tends at least to this time period and geo- graphically could be related to southern South America -Palaeoantarctica. The spread of the ancestral fauna from South America to Central America probably took place during the mid Tertiary (30-38 Ma) as the Panama land bridge devel- oped. Chances for an earlier exchange through possible connections between parts of the Greater Antilles and South America could have existed during the late Cretaceous (Donelley 1988:26), but the geological history of the Caribbean region is still fraught with controversy (Rull 1989). The mordax group fossils from Do- minican Amber show that by the late Oli- gocene to early Miocene (15-25 Ma), di- versification of the American Gnamptogen- ys fauna was well underway. Potential ter- restrial connections between Central America and the Greater Antilles could have existed during the Cenozoic via the presently submerged Nicaraguan rise which lays between Honduras and Jamai- ca (Donnelly 1988:28). It offered possibili- ties for dispersal until its subsidence dur- ing the middle Cenozoic. The presence of the Hispaniolan endemic species G. hayti- ana, G. schmitti, and G. semiferox lend sup- port to such a connection and to the di- versification of Gnamptogenys lineages some 26-34 Ma. Ecology. — The ecology of Gnamptogenys found in Venezuela is treated in Lattke (1990) and should be consulted for more information. With the exception of a few additional tidbits of information little is added here. Generally most species are dwellers of mesic, forested habitats from Texas and Louisiana to northern Argenti- na. Nests are generally constructed in de- composing wood on the ground and tend to be small, rarely exceeding 500 adults. A few species exhibit polygyny. The striatula group species are mostly generalist predators, but other species groups have developed varying degrees of specialization. This is taken to an extreme by the rastrata group species which prey only on certain diplopods. Beetles are the favorite of many members of other groups, and some species such as G. hart- manni and G. horni have also become fond of other ants as prey. New World Gnamptogenys: A Synonymic List When type material of a determined species was personally studied a "T" fol- lows the distribution, plus the acronym of the collection where the specimen is de- posited. Type material, in this case, in- cludes specimens labelled either holotype, paratype, cotype, type series, or syntype. acuniiiiaia Emery 1869. Tropical S. America. acuta (Brown 1957a). Colombia-Bolivia. T:MCZC alfaroi Emery 1894. Costa Rica-Ecuador aitimopliila Lattke 1990. se Venezuela. T:MIZA 148 Journal of Hymenoptera Research andina Lattke new sp. sw Colombia-Ecuador. annulata Mayr 1887. Costa Rica-tropical S. America. hanksi (Wheeler 1930). Panama-Ecuador. T:MCZC bispinosa (Emery 1890). Costa Rica-Colombia. bisiilca Kempf y Brown 1968. Costa Rica-Ecuador. T: MCZC bufonis (Mann 1926). s Mexico-Nicaragua. boliviensis Lattke, new species, n Bolivia. briichi (Santschi 1922). Argentina. T:MCZC brunnea Lattke, new species, nw Colombia. caelata Kempf 1967. Colombia-se Brazil. = soror Kempf and Brown 1968 concinna (F. Smith 1858). Mexico-tropical S. America. T:BMNH = romani Wheeler 1923. = semicircularis Borgmeier 1929. = conica Borgmeier 1929. continiia Mayr 1887 Mexico-tropical S. America. = panamensis Santschi 1931 = exarata Emery, 1901. new synonymy. cuneiforma Lattke, new species. Panama. curvoclypeata Lattke 1990. Colombia-Venezuela. T: MCZC ejuncida Lattke, new species, s Colombia. ericae Forel 1912, revalidated, n S. America. = bufonum Weber 1938, new synonymy. extra Lattke, new species. Colombia-Ecuador. falcifera Kempf 1967. n S. America. fernandezi Lattke 1990. Colombia-Ecuador. T:MIZA fieldi Lattke 1990. n central Venezuela. T:MIZA gentryi Lattke, new species, sw Colombia. gracilis (Santschi 1929). Guiana Shield. T:MCZC laticephala Lattke, new species. Ecuador. haenschi Emery 1902. Costa Rica-tropical S. America. hartmanni (Wheeler 1915). s United States-n S. Amer- ica. T:MCZC = nigrifrons Borgmeier, 1949. new synonymy. T: MUSP = turmalis Kempf y Brown, 1968. new synonymy. T:MCZC haytiana (Wheeler and Mann 1914). Hispaniola. T: MCZC horni Santschi 1929. Panama-Bolivia. T:MCZC ilimani Lattke, new species. Bolivian Andes. ingeborgae Brown 1992. e Colombia. T:MCZC interrupta Mayr 1887. Mexico-tropical S. America, Ja- maica. kempfi Lenko 1964. Brazil-Peru. lanei Kempf 1960. Brazilian Amazonas. levinates Baroni Urbani 1980. Dominican Amber. lineolata Brown 1992. Hispaniola. T:MCZC lucaris Kempf 1968. s Brazil. mecotyle Brown 1958. Panama-tropical S. America. T: MCZC mediatrix Brown 1958. Amazon drainage. T:MCZC menozzi Borgmeier 1948. s Brazil. TiMUSP = schubarti (Borgmeier, 1948) new synonymy. T: MUSP mina (Brown 1957). w S. America. T:MCZC minuta (Emery 1896). Belize-Brazil. = bufonis (Mann 1922) = emeryi (Forel 1901) = scabrosus (Mann 1922) = mus (Santschi 1931) = panamensis (Weber 1940) = carinata (Weber 1940) = pneodonax Kempf 1968 moelleri (Forel 1912). Tropical S. America. = splendens (Santschi 1929) = concinna (Santschi 1929) = teffensis (Santschi 1929) new synonymy. T:MCZC niordax (F. Smith 1858). Mexico-tropical S. America. T:BMNH = nodosa (Latreille 1802) = purensis Forel 1912 = sebastiani Borgmeier 1937 nana Kempf 1960. Brazil. nigrivitrea Lattke, new species, sw Colombia. perspicax Kempf and Brown 1970. Colombia-Ecuador. T:MCZC petiscapa Lattke 1990. ne Venezuela. T:MIZA pilosa Lattke, new species, sw Colombia. pittieri Lattke 1990. n Central Venezuela. T:MIZA pleurodon (Emery 1896) Tropical S. America. = emeryi (Santschi 1929). = vidua (Santschi 1929). porcata (Emery 1896). Honduras-Bolivia. = magnifica (Santschi 1921). new synonymy. pristina Baroni Urbani 1980. Dominican Amber. rastrata (Mayr 1866). s Brazil. = trigona Emery 1905. new synonymy. regularis Mayr 1870. Mexico-Paraguay. = splendida Pergande 1895. T:CASC = fiebrigi Forel 1909. = arcuata (Santschi 1929). reichenspergi (Santschi 1929). Amazon-Orinoco drain- age. relicta (Mann 1916). Amazon-Orinoco drainage. T: USNM rimulosa (Roger 1861). s Brazil. schmitti (Forel 1901). Hispaniola. = minor (Wheeler 1936). T:MCZC semiferox Brown 1958. Hispaniola. T:MCZC siapensis Lattke, new species, s. Venezuela. simulans (Emery 1896). Costa Rica. T:USNM stellae Lattke, new species. Costa Rica. striatula Mayr 1883. Mexico-Argentina, Caribbean. T: BMNH = curtula (Emery 1896) new synonymy. = stolli (Forel 1899) new synonymy. = brasiliensis (Emery, 1902). = angustiloba (Forel 1908). = paulina (Forel 1908). = simplicoides (Forel 1908) new synonymy. = pernambucana (Santschi 1929). = calcarata (Santschi 1929). Volume 4, 1995 149 = antillana (Santschi 1929). T:MCZC = hybrida (Santschi 1929). = mayri (Santschi 1929). = isthmica (Santschi 1929). = recta (Santschi 1929). T:USNM = rustica (Santschi 1929) new synonymy. = wheeleri (Santschi 1929) new synonymy. T:MCZC = arcuata (Santschi 1929) new synonymy. = wasmanni (Santschi 1929) new synonymy. striolata (Borgmeier 1957). se Brazil. T:MUSP strigata (Norton 1871). Mexico-Colombia. = simplex (Emery 1896). sulcata (F. Smith 1858). Mexico-tropical S. America. T: BMNH = tornata (Roger 1861). new synonymy. = lineata Mayr 1870. = cearensis Forel 1912. = nitens Mann 1916. TiUSNM = ypirangensis Borgmeier 1928. tortuolosa (F. Smith 1858). Amazon-Orinoco drainage. T:BMNH = quitensis Forel 1920. transversa Lattke, new species. Panama. triangularis Mayr 1887. Panama-Argentina; se United States. = richteri (Forel 1920) = aculeaticoxae (Santschi 1921) new synonymy. volcano Lattke, new species. Costa Rica. vriesi Brandao & Lattke, 1990. Ecuador. KEY TO THE WORKERS OF THE NEW WORLD SPECIES OF GNAMPTOGENYS. Note: In some species the difference between the promesonotal suture and the metanotal groove may be hard to establish when only one of the two is present, which is the usual case. In a dorsal view the lateral margins of the pronotum converge posterad and at the end of the convergence, when the margins more or less straighten out, the promesonotal suture is usually slightly arched anteriorly, while the metanotal groove is usually straight. 1. Petiolar spiracle directed ventrally and within a depression .... mn;Hfrt group (see Lattke 1991) - Petiolar spiracle directed anteroventrally and not depressed; Pronotum separated from mesonotum by very distinct suture which completely cuts dorsal sculpture 2 - Petiolar spiracle directed ventrolaterally and not depressed; Promesonotal suture absent to well impressed, but never completely cuts sculpture 23 2. Petiolar in lateral view with subparallel anterior and posterior margins sharply set off at right angles to relatively brief dorsal margin (Fig. 21); vertex usually smooth and shining, occasionally with weak transverse triae 3 - Petiole not as above, in lateral view with anterior margin either broadly curving into dorsal margin or separated by broad angle (Figs. 22-4, 46, 49, 52); vertex strongly sculp- tured 6 3. Mesosomal dorsum with extensive smooth and shining areas 4 - Mesosomal dorsum totally sculptured, usually rugulose or costulate 5 4. Metanotal groove well impressed, breaking sculpture; metacoxae armed; anterolateral lobes on declivitous propodeal face lacking (Amazon-Orinoco drainage) relicta - Metanotal groove absent; metacoxae unarmed; small lobes present on anterolateral de- clivitous propodeal face (Amazon-Orinoco drainage) reichenspergi 5. Postpetiolar dorsum transversely rugulose; body color yellowish (Hispaniola) .... haytiana - Postpetiolar dorsum longitudinally costulate; black to dark brown (w South America) .... mina 6. Mesosomal dorsum with well developed metanotal groove 7 - Metanotal groove absent 9 7. Dorsum of gastric tergites I and II smooth and shining (sw Colombia) gentryi n. sp. - Dorsum of gaster with longitudinal costulae 8 8. Subpetiolar process subquadrate, with posterior angle (Fig. 21; Costa Rica to Ecuador, Andes) bisulca - Subpetiolar process shaped as an anteriorly projecting lobe with no posterior angle Fig. 42) (nw Colombia) brunnea n. sp. 9. Petiolar node in lateral view relatively erect, posterodorsal and anterodorsal angles not differing notably (Fig. 22) 10 - Petiole in lateral view posteriorly inclined, anterior margin joins dorsal margin through broad convexity that contrasts with sharp angle between posterior and dorsal margins (Figs. 23, 24, 46, 49, 52) 12 150 Journal of Hymenoptera Research 10. SL < 1.08 mm 11 - SL > 1.08 mm (tropical S. America) moelleri 11. Declivitous propodeal face with 5-11 longitudinal costulae between spiracles and 1-3 transverse costulae between each spiracle and longitudinal costulae (Fig. 26); decumbent pubescence on scapes sparse, < six decumbent and suberect hairs (Mexico to Argentina, Caribbean) striatula - Declivitous face with 13-16 longitudinal costulae between spiracles and no transverse costulae between each spiracle and longitudinal costulae (Fig. 25); scapes with dense and uniform decumbent pubescence, 1-3 subdecumbent hairs (se Venezuela) ammophila 12. Posterodorsal angle of petiolar node forms a sharp to bluntly pointed apex that overhangs posterior margin (Figs. 23, 24, 46, 49, 52); subpetiolar process in ventral view with uni- formly narrow ventral edge, posteriorly widening only at junction with node (Fig. 29) 13 - Petiolar node not, or slightly overhanging posterior margin (Figs. 40, 57); subpetiolar process in ventral view with cuneiform ventral edge: anteriorly thin, wider posterad with bifurcate apex that forms posterior angle of subquadrate process (Fig. 30) 20 13. Petiolar node with a blunt posterodorsal apex (Figs. 24, 46, 49, 52) 14 - Petiolar node with acute posterodorsal point (Fig. 23; Colombia to Bolivia) acuta 14. Postpetiolar sternum totally costulate or striate; vertex usually with one (occasionally up to 3) transverse costulae next to vertexal carinae 15 - Postpetiolar sternum with a basal median smooth and shining area; vertex with 4-5 trans- verse costulae next to vertexal carinae (Guiana Shield) gracilis 15. SSC < 10 (Fig. 27) 18 - SSC > 10 (Fig. 28) 16 16. Body and legs brown, sculpture costulate to striate; body with abundant decumbent pu- bescence; subpetiolar process lobiform and without acute projection 17 - Body piceous and legs ferruginous; body with abundant pubescence; subpetiolar process usually with an acute anteroventral projection, sometimes blunt, especially in specimens from Central America (Honduras to Bolivia) porcata 17. Femora with abundant erect to suberect hairs; ventral surfaces of meso- and metacoxae with erect hairs; gastric sculpture striate (sw Colombia) pilosa, n. sp. - Femora with decumbent to subdecumbent hairs and few or no erect to suberect hairs; ventral surfaces of meso- and metacoaxe wirthout erect hairs; gastric sculpture costulate (s Colombia) ejuncida, n. sp. 18. Five or more standing hairs on scapes and tibiae 19 - One or no suberect and subdecumbent hairs on scapes and tibiae (Colombia-Ecuador) extra, n. sp. 19. Declivitous propodeal face with transverse costulae; petiolar node with anterior peduncle (Fig. 24; tropical S. America) pleurodon - Declivitous propodeal face with longitudinal costulae; node without peduncle (Fig. 59; sw Colombia) nigrivitrea, n. sp. 20. Postpetiolar sternum with well defined costulae or striae; dorsal propodeal face not no- tably depressed below mesonotum, dorsal and declivitous faces confluent 21 - Postpetiolar sternum with weak irregular rugae; in lateral view dorsal propodeal face separated from mesonotum by notable depression, dorsal and declivitous faces sharply separated by a ridge (n Central Venezuela) pittieri 21. Scapes with moderate to weak pubescence, five or more standing hairs present 22 - Scapes with dense, white subdecumbent pubescence, 0-2 standing hairs present (Bolivian Andes) ilimani n. sp. 22. HW > 0.84, WL > 0.35; Ol < 0.16 (sw Colombia-Ecuador) andina, n. sp. - HW < 0.84, WL < 0.35, OI > 0.16 (Mexico-Colombia) strigata 23. With head in frontal view antennal scapes barely or do not reach the vertexal margin, never beyond, when laid back as straight as possible from insertions 24 - Scapes clearly reach beyond vertexal margin 38 Volume 4, 1995 151 24. Mandibular dorsum with continuous striae or costulae over the basal one-fourth or more of length 25 - Mandibular dorsum entirely smooth and shining, sometimes with scattered punctures 27 25. Distinct median carinae visible on clypeus; eyes small, OI < 0.1; mandibular dorsum entirely and densely striate (Costa Rica-tropical S. America) haenschi - No distinct median carinae on clypeus; eyes larger, OI> 0.2; mandibular dorsum without dense striae 26 26. Mandibular dorsum with costulae on basal one fourth or less, rest smooth and shining; propodeal declivity with longitudinal costulae (s United States-n South America) .... hartmani (in part) - Mandibular dorsum mostly rugulose; propodeal declivity with transverse costulae . . . transversa n. sp. 27. Propodeal dorsum mostly transversely striate 28 - Propodeal dorsum mostly longitudinally costulate or striate, continuous with rest of notal sculpturing 30 28. Head subquadrate, CI > 0.78; body brown to dark brown 29 - Head notably rectangular and elongate, CI< 0.78; body black (Costa Rica-Ecuador) . . . .alfaroi 29. Costulae on mesosomal and petiolar dorsum subopaque; propodeum with small teeth; metacoxal teeth (Brazil and Peru) kempfi - Very fine striae on mesosoma and node; no teeth on propodeum nor on metacoxae (Costa Rica to tropical S. America) attnulata 30. Clypeal lamella medianly straight and laterally acutely angulose; subpetioloar process subquadrate, usually with acute posteroventral tooth 36 - Clypeal lamella usually concave and laterally rounded, never acutely angular (rarely with slight median projection and laterally bluntly angular); subpetiolar process usually pro- jecting anteriorly as subtriangular lobe, sometimes subquadrate, but never with acute posteroventral tooth 31 32. Mandibles subtriangular to subfalcate; metanotal groove well-impressed 32 - Mandibles triangular; metanotal groove absent (Brazil) nana 32. Metacoxa with dorsobasal lobe or tooth 34 - Metacoxa lacking tooth or lobe, at most small swelling or low tubercle present 33 33. Body sculpture striate; gastric tergum 2 smooth and shining; clypeal lamella laterally bluntly angulose, slightly projecting anterad medianly and with small concavity in middle of projection (Fig. 60; Costa Rica) stellae n. sp. - Body sculpture costulate; gastric tergum 2 longitudinally costulate; clypeal lamella lat- erally rounded and medianly concave (Fig. 43; Mexico to tropical S. America, Jamaica) interrupta 34. Metacoxal dorsum with low triangular lobe; HW < 1.12, WL < 1.81 mm 35 - Metacoxal dorsum with high, approximately parallel-sided lobe or tooth; HW usually > 1.12, WL > 1.81 mm (Mexico to tropical S. America) mordax 35. Cephalic dorsum striate; anterolateral lobes of declivitous propodeal face weakly devel- oped (n Bolivia) holiviensis, n. sp. - Cephalic dorsum costulate; anterolateral propodeal lobes well developed (Mexico to trop- ical S. America) continua 36. Metanotal groove weakly impressed, visible only with limited angles of view; subpetiolar process with acute posteroventral tooth; anterolateral angles of declivitous propodeal face bordered by raised carinae that cross over costulate sculpture 37 - Metanotal groove well-impressed, visible with any angle of view; subpetiolar process without acute posterior tooth; declivitous propodeal face lacking carinae different from costulate sculpture (s. Brazil) rimiilosa 37. Declivitous propodeal face with longitudinal costulae that converge posterad; anterior nodal face with transverse costulae; meso-metapleural suture absent or very weakly im- pressed (Panama to Bolivia) homi 152 Journal of Hymenoptera Research - Declivitous propodeal face with longitudinal parallel costulae; anterior nodal face mostly longitudinally costulate, basal transverse costulae frequently present and rarely occupy all; meso-metapleural suture distinctly impressed (Mexico to Paraguay) regularis 38. Mandibles very elongate: falcate or subfalcate, only their apices touching or crossing when closed (Figs. 35, 54) 39 - Mandibles triangular or subtriangular, apical and basal margins separated by rounded or angular basal angle (Figs. 33-34, 36-37) 43 39. Declivitous propodeal face with spines or denticles; metacoxal tooth present 40 - Declivitous propodeal face rounded, no trace of spines or denticles; coxal tooth absent 42 40. Inner basal mandibular border with short triangular tooth that represents basal angle (Fig. 54) 41 - Inner basal mandibular border with broadly convex lobe or flange (Amazon drainage) mediatrix 41. Clypeus with two lobes betwen antennal fossae and anterior lamella, lobe protrude over lamella in front view; mesothoracic spiracles not set in deep depressions (Ecuador) . . laticephala, n. sp. - Clypeus without lobes between antennal fossae and anterior lamella; spiracles of meso- thorax set in deep depressions (Panama to Ecuador) banksi 42. Mandibles long and slender (Fig. 36); dorsal propodeal face evenly and transversely cos- tulate (Hispaniola) schmitti - Mandibles shorter and wider; dorsal propodeal face subvermiculately and longitudinally costulate (Hispaniola) semiferox 43. Mandibles mostly smooth and shining, with scattered punctures, subtriangular to trian- gular, sometimes striae or costulae present on basal one-third or less; usually without transverse sutures on mesosomal dorsum (except concinna) 44 - Mandibles with longitudinal rugulae or striae (sometimes weakly impressed), always triangular, never subtriangular; metanotal suture distinctly impressed 55 44. HW > 1.43; WL > 2.27 mm 45 - HW < 1.43; WL < 2.27 mm 46 45. Sculpture costulate; metacoxae with dorsal tooth; petiolar node ends in blunt angle; black (Amazon-Orinoco drainage) tortuolosa - Sculpture striate; metacoxae without dorsal tooth, with low tubercle at most; petiole ends in acute point or beak; ferruginous (Mexico to tropical S. America) concinna 46. Sculpture striate; anterolateral corners of declivitous propodeal face borderd by carinae that cross costulae; antennal scape when laid back does not surpass vertexal margin by more than one apical width 47 - Sculpture costate; declivitous propodeal face without carinae distinct from costae; scapes surpass vertexal margin by more than an apical width 48 47. Posterior face of petiolar node longitudinally costulate (s United States to n S. America) hartmani (in part) - Posterior nodal face transversely costulate (Argentina) bruchi 48. Posterior face of petiolar node transversely costulate 49 - Posterior face of node longitudinally costulate 51 49. Anterior clypeal lamella laterally acutely angular; mandibles subtriangular, basal and apical margins joined by very broad continuous convexity (Fig. 34) 50 - Clypeal lamella very evenly convex laterad; mandibles triangular, basal and apical mar- gins separated by angle (Fig. 33; Venezuela to Ecuador) femandezi 50. Petiolar node with posterior pointed projection; declivitous propodeal face with longi- tudinal costulae; body color uniformly brown (tropical S. America) acuminata - Node sometimes with slight posterior projection, but never pointed; declivitous propodeal face with transverse costulae, rarely longitudinal; color variable, usually with brown and ferruginous areas, rarely totally brown (Mexico to tropical S. America) sulcata 51. Anterior clypeal lamella laterally acutely angulate 52 Volume 4, 1995 153 - Clypeal lamella laterally rounded or obtusely angulate 53 52. HW < 1.11, WL < 1.72 mm; anterior clypeal lamella straight (Fig. 31; n S. America) . . . ericae reval. sp. - HW > 1.11, WL > 1.72 mm; anterior clypeal lamella medially convex and laterally con- cave (Fig. 32; Venezuela and Colombia) ciirvoclypeata 53. Declivitous propodeal face with longitudinal costae; clypeal lamella straight anterad and laterally rounded; mandibles subtriangular, basal and apical margins meeting through broad convexity (Fig. 62) 54 - Declivitous propodeal face with transverse costae; clypeal lamella excised mesad and with two lateral convexities; mandibles roughly triangular, apical and basal margins separated by notch (Fig. 64; Costa Rica) volcano, n. sp. 54. Petiolar node with dorsal margin more or less evenly convex in lateral view; metacoxal tooth denticle-like; subpetiolar process with acutely pointed angles (s Brazil) lucaris - Petiolar node with dorsal surface flat; metacoxal tooth broadly triangular; subpetiolar process bluntly angulate (s Venezuela) siapensis, n. sp. 55. Antennal scapes usually smooth and shining 56 - Scapes with longitudinal striae or rugulae (sometimes partially effaced) 60 56. Scapes surpass vertexal margin by two or more apical widths; HW > 1.07, WL > 1.64 mm 59 - Scapes when laid back do not surpass vertexal margin by more than one apical width; HW < 1.07, WL < 1.64 mm 57 57. Body sculpture striate; no propodeal denticles; ventral petiolar process in lateral view projecting anterad as a lobe, lacking a posterior angle; Hispaniola Island lineolata - Sculpturing costulate; propodeal denticles present; petiolar ventral process with a pointed posterior corner; South America 58 58. Costulae on dorsum of metanotum and propodeum well defined; petiole in dorsal view with concentric costulae (s Brazil) rastrata - Costulae on dorsum of metanotum and propodeum mostly effaced; petiole in dorsal view transversely costulate (e Colombia) ingeborgae 59. Propodeal teeth longer than diameter of propodeal spiracles; petiolar node longer than wide and subpetiolar process projects anterad as triangular lobe (Brazilian Amazon drain- age) land - Propodeal teeth shorter than diameter of propodeal spiracles; petiolar node wider than long; subpetiolar process subquadrate (s Brazil) menozzi 60. Anterior pronotal margins with at least five transverse costulae or rugulae; eyes promi- nent and bulging (Fig. 37) 61 - Anterior pronotal margin with longitudinal costulae that extend caudad across rest of mesosomal dorsum; eyes not bulging and prominent, relatively flat (Fig. 38; Florida, Pan- ama to Argentina) triangularis 61. Vertex with 4-5 transverse costulae; anterior one-half of postpetiolar tergum rugose ... 62 - Vertex with longitudinal costulae; anterior one-half of postpetiolar tergum longitudinally costulate 63 62. Dorsum of metanotum and all of propodeum with transverse costulae (Panama) .... cuneiforma n. sp. - Dorsum of metanotum and propodeum longitudinally costulate; declivitous propodeal face with weak longitudinal costulae, smooth and shining mesad (Panama to tropical S. America) mecotyle 63. Propodeal spines at least as long as distance between inner sides of their base (Fig. 12); ferruginous species (Costa Rica to Colombia) bispinosa - Propodeal spines shorter (Fig. 13); black species (Colombia to Ecuador) perspicax 154 Journal of Hymenoptera Research New World species of Gnamptogenys In this section the species are arranged in alphabetical order. In the synonymy listings, when more than one citation fol- lows a junior synonym, the first author, after the author of the species, is respon- sible for that synonymy. The diagnosis is meant to complement the results obtained in the key. In order to avoid repeating in- formation already published the reader should consult Lattke (1990) for additional data concerning ecology and localities of species previously known from Venezue- la. For the sake of brevity the dates and collector's names have been omitted. Gnamptogenys acuminata Emery Gnamptogenys acuminata Emery 1896:50; Brown 1958:228,^236, 299; Kempf 1961:491; Kempf 1972:111; Kempf 1976:51; Lattke 1990:9. Diagnosis. — Mandibles subtriangular; longitudinal costulae on mesosoma and declivitous propodeal face; transverse su- tures weak, visible only with certain an- gles of light; body dark brown to black; coxae and basal two-thirds of femora tes- taceous to brown-testaceous; tibiae, apex of femora, antenna and mandibles brown. Ecology. — Found in humid forests, nest- ing in rotting wood. Comments. — The acute posterior tooth on the petiolar node, combined with lon- gitudinal costulae on the declivitous pro- podeal face, seem to provide the most re- liable distinction from tornata. Color in ac- uminata is not so variable but light-pig- mented forms can occasionally be found. The mesometepisternal suture is usually absent. Specimens Examined. — BRAZIL, Amazo- nas: Ig. Marianil, Rio Branco Rd, 24 km NE Manuas. COLOMBIA, Meta: Cano EL Buque, S Villavicenicio, 480 m. PERU, Lo- reto: 15 km WSW Yurimagua, 5°59'S 76°13'W, 220 m; Panguana: 9°37'N 74°56'W, 220 m. VENEZUELA, Bolivar: Rio Cuyuni, 66 km SSE El Dorado, 6° 09' N 61° 30'W, 250 m; 10 km E Icabaru, 700 m. Amazonas: Rio Baria, 0°50'N 66°10'W, 140 m. Gnamptogenys acuta Brown Holcoponera acuta Brown 1957:491. Gnamptogenys acuta (Brown); Brown 1958:300; Kempf 1972:111. Diagnosis. — Similar to striatula; four to six transverse costulae on anterior prono- tal face, rest of mesosoma with longitudi- nal costulae; longitudinal costulae on an- terior face of postpetiole; body dark brown to black, legs brown. Figs. 21-24. Lateral view of petiole. 21, G. relicta; 22, G. inoelleri; 23, G. acuta; 24, G. pleurodon. Volume 4, 1995 155 26 Figs. 25-26. View of declivitous propodeal face. 25, G. ammophila; 26, G. striatiila. Figs. 27-28. Frontal view of scapes, showing hairs and pilosity. 27, G. pleurodon; 28, G. porcata. Ecology. — Found in low to premontane jungles. Comments. — Not frequently collected, knowledge of its range has expanded slowly. Specimens Examined. — ECUADOR, Napo: 2-8 miles N Puyo, Pastaza, 935 m; original locality unknown: Quarantine Station, San Pedro, California. PERU, Ju- nin: Colonia Perene, Rio Perene, 18 km NE La Merced; Loreto: Boqueron, 500 m; CO- LOMBIA, Meta, Mesetas, 780 m. Gnamptogenys alfaroi Emery Gnamptogenys alfaroi Emery 1894:145; Brown 1958:301;'Kempf 1972:111. Diagnosis. — Mandibular dorsum smooth and shining with some punctulae; meta- coxal spine present; subpetiolar process anteriorly projected; gastric tergum 2 with basal one-half longitudinally costulate, the rest smooth and shining; disk of pospetio- lar sternum smooth and shining; body black; legs and antennae brown. Ecology. — Known from montane rain forest. Comments. — The Ecuador record ex- tends its distribution from the only pre- viously known localities in Costa Rica. This species is a striking member of the mordax group due to its elongated, almost cylindrical body. Its relationship with oth- er members of the mordax group is not clear. It is rarely collected. Specimens Examined. — COSTA RICA, Turrialba. ECUADOR, Guayas: 3 km SW Bucay (MCZC). Gnamptogenys ammophila Lattke Gnamptogenys ammop>hila Lattke 1990:9. Diagnosis. — Black, finely costulate with brown scapes and mandibles, light brown legs. Subpetiolar process subquadrate, not as projecting anterad as in striatula. Me- sosomal dorsum with longitudinal costu- lae and erect hairs up to 0.30 mm in length. Ecology. — This species is unique within the genus due to its savanna nest sites as opposed to the forest preferences of most species. Its range is apparently restricted to high (>1200m), cool savannas of the Upper Rio Caroni watershed (La Gran Sa- bana). It has not been found in lower sa- vannas, despite the fact that another ant, Pogonomyrmex naegeli, is widespread in sa- vannas isolated in the midst of great ex- panses of forest. The extent of savannas in the Guiana Shield was much broader dur- ing the last glaciation (75,000-11,000 B.P.), as were generally cooler temperatures and a drier climate (Schubert 1988; Clapperton 1993). Such conditions could concievably have favored the origin of this species from its apparent sister species G. striatula during the course of the last glaciation. Comments. — See Lattke 1990. Specimens Examined. — VENEZUELA, Bo- livar: via El Doradc-)-Santa Elena, km 141, Rio Apanwao, 5°52'N 61°26'W, 1500 m. 156 Journal of Hymenoptera Research ^: iT 31 29 30 32 Figs. 29-30. Ventral view of petiole showing shape of subpetiolar process (outlined in black). 29, G. gracilis; 30, G. andina. Figs. 31-32. Frontal view of clypeal lamella. 31, G. sulcata; 32, G. curvoclypeata. Gnamptogenys andina Lattke, new species (Figs. 39-40) Type Specimens Examined. — Holotype worker. COLOMBIA, Valle: Parque Far- allones de Cali, El Topacio, 20 km NW Cali, 76 37'N 2 30'W, 1550 m, 31-XII-1981, J. Lattke No. 220. Deposited in MIZA. Paratypes (all from Colombia, Valle): (1). 22 workers from tye same nest series as the holotype. (2) 21 workers, same data as holotype except collection numbers 220, 214, 218. (3). 18 workers, Cali-Buenaven- tura road, km 21, 15 km NE Cali, 1300- 200 m, l-X-1975, J. Lattke, leg. (4). 6 work- ers, CVC station near Pance, 15 km W Cali, 1700 m, 12-XII-1975, J. Lattke, leg. Additional Specimens (not paratypes). — ECUADOR, Bolivar, 20 km de Palzabam- ba, 1500-1800 m, 18-X1-87, N. Zavala, leg. Six paratypes deposited in MCZC and three in each of the following : LACM, BMNH, CASC, MUSP. The remainder of paratypes and other specimens deposited in MIZA. Worker. — Holotype (Paratypes) mea- surements: HL 1.04 (1.02-1.08); ML 0.40 (0.26-0.36); HW 0.88 (0.82-0.94); SL 0.86 (0.86-0.92); ED 0.12 (0.10-0.14); WL 1.36 (1.36-1.60) mm; CI 0.85 (0.83-0.90); SI 0.98 (0.93-1.10); OI 0.14 (0.11-0.17) n = 7. Typical striatida group member: Rough- ly costulate sculpture. Vertexal margin concave in frontal view; clypeal costulae extend slightly onto anterior lamella; man- dibles triangular and mostly striate except for smooth and shining lamellate basal corner and margin of chewing border; compound eyes relatively small; in lateral view dorsal propodeal margin not notably depressed below rest of mesosomal dor- sum; propodeal spiracle slightly elevated above rest of sculpture and at edge of de- clivitous surface; petiolar node posteriorly inclined; subpetiolar process subquadrate, typical of strigata subgroup. Longitudinal costulae present on declivitous postpetio- lar face. Weak transverse costulae on post- petiolar ventrum. Fairly abundant standing hairs on body, but very sparse appressed pubescence. Legs smooth and shining, except for nu- merous piligerous punctures. Metacoxal dorsum with a well-developed denticle. Body dark brown to black with yellowish legs. Male. — Longitudinal costulae on cephal- ic dorsum; mandibles triangular and cos- tulate; pronotum smooth and shining with piligerous punctures; pronotum with pili- gerous punctures; mesonotum mostly smooth and shining except for narrow an- terior band of transverse striae; declivitous propodeal face with longitudinal rugulae that do not reach anterior dorsal face; gas- ter smooth and shining. Queen. — Unknown. Discussion. -This species is near strigata and though the differences are not striking, they seem consistent enough to define a Volume 4, 1995 157 35 Figs. 33-38. Frontal view of head. 33, G. fernandezi; 34, G. sulcata; 35, G. haenschi; 36, G. schmitti; 37, G. bispinosa; 38, G. triangularis. species. G. strigata is smaller (HW 0.74- 0.80); WL 1.20-1.34 mm), with relatively larger eyes (OI 0.17-0.20) and shorter scapes (0.70-0.78 mm) and dorsal propo- deal face in lateral view is usually de- pressed below the rest of the mesosomal dorsum. It is probably sympatric with nn- dina in part of its range, having been found only 51 km away from the andina type lo- cality. The male of strigata has a median area of rugosity on the mesonotum and to- tally rugulose propodeum. Individuals of the Ecuadorean sample are slightly larger than the Colombian specimens. Etymology. — The name alludes to the Andes, a portion of which is inhabited by this species. Gnamptogenys annulata Mayr Ectntomma {Gnamptogenys) rimulosa var. annulata May 1887:543. Ectatomma {Gnamptogenys) annulatutn Mayr; Mann 1916:408; Luederwalt 1926:23. Gnamptogenys annulatum Mayr; Mann 1922:3; Brown 1958:300; Kempf 1960:390; Kempf 1961:491; Kempf 1969:275; Kempf 1972:112; Kempf 1976:51; Lattke 1990:11. Diagnosis. — Finely striate species; longi- tudinal striae on head, promesonotum and gastric segments 2 and 3; petiolar dor- sum with transversely arched striae; man- dibles subtriangular; metacoxal spine ab- sent; body reddish brown; coxae ferrugi- nous yellow; tibiae and femora yellow to testaceous but apically ferruginous yellow to ferruginous. Ecology. — J. Longino (pers. comm.) re- ports from a nest midden chitinous frag- ments of Coleoptera, aradid bug nymphs, an unidentified male ant, and a Cauipotio- tus worker. An apparently recently killed 158 Journal of Hymenoptera Research 42 Figs. 39-42. Frontal view of head, and lateral view of mesosoma and petiole. 39-40, G. andina; 41-41, G. brunnea. Pheidole (cephalica?) queen was also found in a nest chamber. Comments. — The gauge of costulation can vary from striate to finely costulate, but coloration is fairly stable. SPECIMENS EXAMINED: BOLIVIA, Caranavi: vie. Radio, 800 m. BRAZIL, Amazonas: 80 km NNE Manaus, 2°25'S 59°46'W, 80 m. COLOMBIA, Valle: Bajo Calima; Cauca: Isla La Gorgona. COSTA RICA, vie. Guapiles, Rio Toro Amarillo; Prto. Limon; Limon: Tortuguero, < 100m, 10°35'N 83°31'W; Heredia: 10°20'N 84°84'W, 500 m; 3 km S Puerto Viejo, 10°26'N 84°00'N, 50 m; Pq. Nac. Braulio Carrillo, Carrillo, 500 m; Peninsula de Osa, Sirena, 8°28'N 83°35'W; Alajeula: Rio Peii- as Blancas, 10°19'N 84°43'W, 800 m. EC- UADOR, Pichincha: 6 km W Sto. Domingo de los Colorados, 953 m; Maquipucuna, 5 km ESE Nanegal, 0°70'N 78°38'W, 1250m; Sucua. GUIANA, Kartabo. PERU, Pangu- ana: 9°37'S 74°56'W, 220 m; Tingo Maria: Valle de Monson; Madre de Dios: Res. Rio Tambopata, 30 km SW Prto. Maldonado, 290 m; Loreto: Boqueron, 500m; San Mar- tin: Davidcillo, 30 km NNE Tarapoto, 6°15'S 76°15'W. VENEZUELA, Tachira: Uribante-Caparo, Las Cuevas, 7°47'N 71°46'W, 500 m; Barinas: Ticoporo, 8°04'N 70°48'W, 250 m; Amazonas: Alto Mavaca, 2°01'N 65°07'W, 200 m; Alto Siapa, 1°42'N 64°38'W, 500 m. Gnamptogenys banks! (Wheeler) Emeryella banksi Wheeler 1930:10. Gnamptogenys banksi (Wheeler); Brown 1958: 227,233; Kempf 1972:112. Volume 4, 1995 159 Diagnosis. — Finely costulate; head broad from the first sighting to the prey was 6 with longitudinally striated scapes; meso- m. I followed a returning worker as far as metanotal suture not deeply impressed, I could into a tree fall tangle, 10 m from but distinct. It lacks the long, fine mandib- prey." The second attack involved 46 ular sulcus present in laticepJmla. workers tackling a 6 cm long millipede Ecology. — Found in rotten logs in low- within a rc^tten log. Once the millipede land rain forest, this species is a specialist was subdued they carried it back to the millipede predator. nest. Comments. — It is rarely collected. A Two millipedes taken from the nest by specimen from Barro Colorado has two Longino were determined by R.L. Hoff- deep fossae on the posterolateral margin man as Trichomorpha sp. (Polydesmida: of the mesonotum. The type was exam- Chelodesmidae), and Epinannolene sp. ined. (Spirostreptida: Epinannolenidae). Specimens Examined. — ECUADOR, Comments. — This unique ant was previ- Guayas: 3 km S Bucay. PANAMA, Isla ously only known from a restricted area Barro Colorado (MCZC). in Central America. Along with perspicax, it represents a highly specialized form in Gnamptogenys bispinosa (Emery) ^^^ lineage of millipede hunters. The an- Ectatomma {Poneracantha) bispinosum Emery terior clypeal setae are more developed 1897:547. than in any other of the rastrata group spe- Ectatomma (Holcoponera) bispinosum Emery; Em- cies and an interesting autapomorphy is a ery 1890:40. brief anteromedian longitudinal sulcus on Holcoponera bispinosa (Emery); Wheeler 1952: \\^q clypeus. 132-133. Specimens Examined. — COLOMBIA, Val- Gnamptogenys bispinosa (Emery); Brown 1958: ^^. ^QSTA RICA- Heredia- 10°20'N ?g79 i^r^^ ''''^ ^'°'^'' ^^^^''^^^'' ^'""^^ 84°04'W, 500m; Peninsula de Osa, Pq. Nac. Corcovado, Sirena, < 100m, 8°28'N Diagnosis. — Eyes subglobulose, scapes 83°35'W; Puntarenas: 2 km S San Vito, Fin- long and surpassing vertexal margin; ca Las Cruces, 1219 m; Rio Reventazon, 4 mandibles edentate. Promesonotal suture km E Turrialba (MCZC). breaks sculpture dorsally but not laterally; mesometanotal suture deep and broad. Gnamptogenys bisulca Kempf and Head, mesosoma and postpetiole rugose; Brown „ ^ ■ . . . . Gnamptogenys bisulca Kempf and Brown 1968: £co/o^/.-This wet forest species is a 92-3; ^ Wheeler and Wheeler 1971:1202; millipede specialist. J. Longino (pers. Kempf 1972112 comm.) observed the following two raids from the same colony, both at 11:00 AM: Diagnosis. — Longitudinal costulae on " A column of 20 workers was moving propodeal declivity, metacoxal spines along a liana on the forest floor. They left hooked and triangular, not acicular; well the liana and moved very slowly across defined transverse costulae on postpetio- the leaf litter, frequently bunching up. lar sternum; propodeal spiracle as low as They all went under a leaf and then agi- surrounding sculpture, not raised, tated ants could be seen coming out from Ecology. — Occurs in mesic forested hab- under the leaf, 2 or 3 ants at a time, clean- itats. The localities suggest preference for ing their antennae and mandibles. I re- relatively cool sites as the lowest altitudes moved the leaf to find the ants attacking are found at higher latitudes (10°N) and a millipede. Some were stinging and some as latitude approaches the equator the lo- were pulling on the legs. The distance calities ascend in altitude. 160 Journal of Hymenoptera Research Comments. — Within the strigata complex this species is unique due to the develop- ment of its metanotal groove, and within the striatula group two other species also show this character state. Recent collec- tions show a wider range than previously known. Specimens Examined. — COLOMBIA, Choco: 10 km SW San Jose del Palmar, Rio Torito, Finca Los Guaduales, 800 m, 1/4- VI-78; Valle: vie. Lago Calima, 1600 m; vie. Saladito, 1900-2100 m. COSTA RICA, He- redia: 17 km S. Puerto Viejo, 10°18'N 84°02'W, 550 m; 17 km N Volcan Barba, 10°17'N 84°05'W, 880-1020 m. ECUA- DOR, Pichincha: 25 km ENE Alluriquin, via Chiribaga, 1400-1800 m; Tinalandia, 16 km SE Sto. Domingo de los Colorados, 9-VI-76; 3 km E Tandapi, 1300 m. Gnamptogenys boliviensis Lattke, new species (Figs. 43, 44) Type Material. — Holotype worker. BO- LIVIA, Tumupasa, W.M. Mann, leg. De- posited in the USNM. Paratypes: Seven workers deposited in the USNM; One worker and one male in each of the fol- lowing: BMNH, MIZA, LACM, MCZC. One worker in MUSP. All from same lo- cality and probably from the same nest. One specimen bears an additional label: "Mulford Biological Expedition, 1921- 1922." Consultation of maps locates Tu- mupasa at 14°09'S 67°55'W in lowland for- ests of the upper Rio Beni watershed. Worker. — Holotype (Paratypes) mea- surements: HL 0.88 (0.83-0.88); ML 0.36 (0.39-0.41); HW 0.70 (0.69-0.70); SL 0.49 (0.48-0.50); ED 0.10 (0.10-0.12); WL 1.23 (1.06-1.15) mm; CI 0.70 (0.80-0.84); SI 0.70 (0.69-0.71); OI 0.15 (0.15-0.18). n=4. Head in frontal view elongate, lateral margins fairly parallel, vertexal margin concave; anterior lamella of clypeus later- ally rounded and medianly convex; eyes situated at mid-length; antennae smooth and shining, dorsoventrally compressed widest apicad; mandibles smooth and shining dorsally, with sparse punctae, su- bfalcate, laterobasally with small area of longitudinal rugae, including small sulcus never longer tha one-fourth of the man- dible; gula with longitudinal costulae, front of head with longitudinal striae. Occasional transverse costulae on pron- otal collar; mesosomal dorsum with lon- gitudinal costulae, slightly thicker on metanotum and propdeum than on pro- mesonotum; pronotal suture softly im- pressed, visible only in certain angles of view; metanotal suture deeply impressed but doesn't break longitudinal sculpture; declivitous propodeal face with oblique to transverse costulae and two superolateral ridges; pronotum laterally with transverse costulae, smooth and shining along ven- tral margin; anepisternum elongate and obliquely costulate; katepisternum and metepisternum with transverse costulae; costulae on metepisternum have same di- rection as on katepisternum, but costulae on lateral propodeal face are more hori- zontal; propodeal spiracle round and not prominently higher than surrounding sculpture; petiole slighty pedunculate, dorsally with longitudinal costulae, ante- rior face with transverse costula or rugae, sometimes partially effaced, laterally with longitudinal costulae with slight oblique tendency; node in lateral view with softly convex dorsal margin, anteriorly concave; anterior and posterior faces fairly sharply separated from dorsal face; posterior face with 4-5 convex costulae; subpetiolar pro- cess varies from subquadrate anteriorly and posteriorly concave. Anterior postpetiolar face smooth and shining; gastric terga I and II with longi- tudinal striae, sternum I laterally costula- te, but discal area smooth and shining as is rest of gaster; procoxa anteriorly and anterolaterally smooth and shining, pos- terolaterally costulate; mesocoxae with transverse costulae that tend to be effaced; dorsum of metacoxa with transverse cos- tulae and basal, low triangular lobe; tibia and femora smooth and shining. Body fer- Volume 4, 1995 161 Figs. 43-44. Micrographs of G. boliviciiisis. Scale bars = 0.5 mm. 43, head; 44, thorax and gaster. 162 Journal of Hymenoptera Research ruginous brown; legs and antennae testa- ceous. Queen. — Unknown. Male. — Mandibles triangular, apical edge serrate, dorsally smooth and shining; cephalic dorsum with longitudinal striae that tend to be effaced before reaching mid-ocellus, faint transverse rugae above the posterior ocelli but rest of head smooth and shining; propodeum rugose; petiole with lateral rugulae, dorsum shin- ing, with slight roughened aspect and me- dian longitudinal costulae; gastric sterna and terga smooth and shining; vestigial arolea present. Discussion. — Several specimens were found with determination labels, probably written by Mann, indicating that he rec- ognized this species as new. G. boliviensis is near to G. continua and the possibility exists that the collection and study of more material may show them to be con- specific. In G. continua the sculpture is coarser, costate-costulate, and the mandi- bles are more robust, without a concave inner edge. Etymology. — The name is derived from the name of the type locality country, Bo- livia. Gnamptogenys bruchi (Santschi) Ectatomma (Parectatomma) bruchi Santschi 1922: 241. Gnamptogenys bruchi (Santschi); Brown 1958: 227; Kempf and Brown 1968:92; Kempf 1969: 275; Kempf 1972:112. Diagnosis. — Striate body sculpture with weak transverse impressions; metacoxal process is shaped as a low, triangular lobe; clypeal lamella broadly convex anterad. Ecology. — Apparently a predator of Tra- chymyrmex and other leaf cutting ants (Kempf and Brown, 1968:92). Comments. — Kempf (1969) notes bruchi as close to hartmani and that a synonymy could be possible. He also mentions the variability of the extent of impression of the metanotal groove and discards it as an indicator of species separation. This could very well be only a variant of hartmanni but due to differences of sculpture on the posterior petiole face, postpetiolar sternal process and clypeal lamella, I have chosen to conserve this name until more material can be studied. The sculpture on the dor- sum of the petiole can vary from longitu- dinal to whorled. Specimens Examined. — ARGENTINA, Cordoba: Alta Gracia (cotype:MCZC). BRAZIL, Sao Paulo: Fazenda B. Rico. Gnamptogenys brunnea Lattke, new species (Figs. 41,42) Type Material. — Holotype worker. CO- LOMBIA, Choco, 10 km SW San Jose del Palmar, Finca Los Guaduales, 760 m, 2-VI- 1978, C. Kugler, leg. Deposited in MIZA. Para types: One queen and 18 nidotype workers. Queen and 6 workers deposited in MIZA. Two workers deposited in each of the following: MCZC, LACM, BMNH, MUSP, USNM. Worker. — Holotype (Paratypes) mea- surements: HL 0.90 (0.88-0.92); ML 0.30 (0.30-0.32); HW 0.74 (0.68-0.80); SL 0.74 (0.80-0.84); WL 1.20 (1.10-1.28) mm; CI 0.82 (0.77-0.87); SI 1.00 (0.95-1.09); OI 0.18 (0.16-0.18) n=5. Cephalic dorsum longitudinally costu- late; in frontal view vertexal margin con- cave and eyes slightly behind mid-length; mandibles triangular, with dorsal costu- lae, basal and apical margins with narrow smooth and shining strip that is sharply separated and lower than the rest of man- dibular surface; anterior pronotal margin with 6-8 transverse costulae which arch around and become longitudinal on dor- sum; mesometanotal suture well-im- pressed; mesonotum narrow and trans- verse; in lateral view dorsal propodeal face slightly depressed below rest of me- sosomal dorsal margin; declivitous pro- podeal face distinctly separated from rest of mesosoma by low lateral ridges; mostly transverse costulae with longitudinal cos- tulae from dorsum extending only to an- Volume 4, 1995 163 terior one-fourth of declivitous face; peti- Diagnosis. — A large finely striate species olar node in lateral view transversely cos- with transverse striae on propodeum, de- tulate and posteriorly inclined: a broad clivity medianly smooth; promesonotal anterior convexity with sharp posterior suture weakly impressed or absent; me- drop and concave posterior margin; sub- sometanotal suture deep and wide; meta- petiolar process shaped as an anteriorly coxal tooth absent, low crest or tubercle at projecting lobe; anterior postpetiolar face most. and sternum with transverse costulae. Ion- Ecology. — Consistently collected and ob- gitudinal on rest; costulae on the gastric served on trees, this species is an arboreal tergum II finer than on preceding seg- nester. Longino found a nest inside a ment; metacoxal spine present. large, deep knothole and records prey Body with sparse decumbent pubes- items of the following: Heteroptera: Pen- cence and erect to suberect hairs; pubes- tatomidae, Aradidae; Coleoptera: Ceram- cence denser on legs; scapes with abun- bycidae, Histeridae, Platypodidae, Chrys- dant decumbent pubescence, no erect or omelidae, Tenebrionidae, and Passalidae. suberect hairs. Body reddish-brown and One observed foraging group was made legs brownish-yellow. up of 9 workers and a dealate queen. Ob- Queen. —HL 0.98; HW 0.86; ML 0.32; SL servations of group foraging of up to 40 0.84; ED 0.18; WL 1.44 mm; CI 0.88; SI single-file workers, plus the following 0.97; Ol 0.21. Like workers with exception Longino observation, indicate trail and /or of usual caste differences. recruiting pheromone communication in Male. — Unknown. this species. "The foragers walked with Discussion. — This species is close to ni- their gasters curled and touching the sub- grivitrea on account of size, posteriorly tilt- strate. When they were together near the ed petiolar node, sleek appearance, sub- prey I could often make out a tiny white petiolar process shape, sparse pilosity and area (gland?) exposed at the very tip of the lack of any trace of anterior petiolar pe- gaster, which other workers would inves- duncle. But nigrivitrea is different in its pi- tigate". ceous body coloration with dark brown Comments. — Color can vary from ferru- legs, lack of metanotal groove, slightly ginous to light-brown, and the striae on more elevated propodeal spiracles, longer the posterior nodal face may be effaced, pilosity, and a sleeker petiolar node with Specimens Examined. — BRAZIL, Amazo- a more prominent apical point. nas: Ilha de Curari, (varzea); Amapa: Villa Etymolog}/. — The name comes from the Amazonas. COLOMBIA, Meta: Reserva Latin term for brown, brunneus. La Macarena, Cafio La Curia, 580 m. COS- TA RICA, Peninsula de Osa: Pq. Nac. Cor- Gnamptogenys concinna (Smith) covado, Llorona; Heredia: Finca La Selva. PANAMA, Isla Barro Colorado. PERU, Ectatommn concinnna F. Smith 1858:103. Madre de Dios: 15 km E Prto. Maldonado, Ectakvnma{Gnamptogenys)concinmim{¥.Smm; 200 m; Amazonas, Panguana, 9°37'S Mann 1916:406; Wheeler 1922:2. ^a^^c^'wi nnn ^ x- \a ^ \7 n a r. ,^ 74 56 W, 220 m; Tmgo Maria: Valle de tctatomma (Gnamptogenys) concinnum var. ro- ^. . -roiMTMr-v a r^ a • \i w i r-^ ■ AAfu 1 in-^oo D ir^ron-^r, Monson. TRININDAD, Arima Valley, 152 mnn? Wheeler 1923:2; Brown 1958:227. a/cxtc^ttct a a ai/ r,- Lcia\omma{Gnampto^em,s)concmnumv,.r.comca "^- VENEZUELA, Amazonas: Alto Rio Borgme.er 1928:196; Brown 1958:227. Mavaca, 2 02'N 65°06'W, 200m. Ectatommn {GnarnptoQenys) concimmni var. senii- ^ ^ .• »,r , D . ir.or.ir^r r, irvr-r, Gnamptogenys contiHua Mavr circularis Borgmeier 1929:195; Brown 1958: r & j / 227. Ectatommn [Gnnmptogenys) conti)U(uiii Mayr Grwmptogcnys co}tci)uia (Smith); Brown 1958: 1887:544. 227; Kempf 1972:112; Lattke 1990:11. Ectatommn {Giinmptogcin/^) cxaratum Emery 164 Journal of Hymenoptera Research 1901:50; Brown 1958:303; Kempf 1972:112; Lattke 1990:13. NEW SYNONYMY. Ectaomma {Gnamptogenys) continuum var. pana- mensis Santschi 1931:265; Brown 1958:228, 304. Gnamptogenys continua (Mayr); Brown 1958:228, 304; Kempf 1972:112; Lattke 1990:12. Diagnosis. — Mandibles with linear ten- dency; mesometanotal suture impressed; declivituous propodeal face with small su- perolateral lobes; upper half with longi- tudinal costulae and inferior half with transverse costulae; sternal disc of post- petiole smooth; anterior nodal face rugose with some inferior transverse costulae. Ecology. — Collected in leaflitter samples from humid forsets. Comments. — The synonymy of exarata was inevitable as more material accumu- lated and it became impossible to separate the two forms. The gap between OD and apical scape width closed. There are changes in the gauge of the costulation, degree of anterior clypeal convexity, ef- facement of the sculpture, shape of the subpetiolar process, extent of the petiolar peduncle, and size (HW 0.56-0.96; HL 0.64-1.16; WL 0.88-1.62). The occiput has a median dent that is most noticeable when the occiput is smooth and shining as opposed to costulate forms. Deposited in the MCZC is a specimen from Vera Cruz that was compared with the type by Brown in 1954. Specimens Examined. — BELIZE, Caves Branch, 4 km S Belmopan. BRAZIL, Rio de Janeiro: Itatiara; Sao Paulo: Fazenda Con- quista, Rio Preto; Guquitiba, BRE km 76. HONDURAS, La Lima. ECUADORA, Pi- chicincha: Tinalandia, 16 km SE Sto. Do- mingo del los Colorados, 680 m. COLOM- BIA, Meta: Quebrada Susumuko, 23 km NW Villavicencio, 1000 m; Magdalena: Parque Tayrona, Pueblito, 360 m. MEXI- CO, Vera Cruz, Tuxtla, Las Hamacas, 17 km N Santiago. PERU, Panguana, 9°37'S 74°56'W, 220 m; Huanuco: 16 km SW Las Palmas, 1000 m; Pasco: Pozuzo, 1000, SU- RINAM, Kartabo, Bartica Dam. VENE- ZUELA, Barinas: 17 km SSW Ciudad Bo- livia, 8°02'N 70°46'W, 240 m; Guarico: 24 km N Altagracia de Orituco, 300 m. Gnamptogenys cuneiforma Lattke, new species Type Material. — Holotype worker: PAN- AMA, Bocas del Toro, 8°47'N 82°11'W, 500m; 16/18-VII-87, D.M. Olson, leg. De- posited in MCZC. Paratypes: One worker deposited in MCZC and two workers de- posited in MIZA, all with same locality data as the holotype. Worker. — Holotype (Paratype) measure- ments: HL 0.90 (0.84-0.88); ML 0.62 (0.60- 0.64); HW 0.84 (0.82-0.84); SL 0.82 (0.80- 0.82); ED 0.20 (0.21-0.22); WL 1.40 (1.38- 1.40) mm; CI 1.07 (1.02-1.05); SI 0.98 (0.98); OI 0.24 (0.26-0.28), n=4. Head subquadrate in frontal view, sides very broadly convex, slightly wider ante- riorly than posteriorly, posterior margin broadly concave, longitudinally costulate. Anterior clypeal margin with a small la- mella, medianly straight and laterally rounded. A brief longitudinal carinae, thicker than surrounding sculpture, pres- ent between frontal lobes. Mandibles tri- angular and elongate, masticatory border concave, dorsal surface of blade convex and with longitudinal rugulae. Scapes shining and with low rugosities. Scapes without pilosity, only decumbent hairs. Anterior pronotal face with 4-5 trans- verse costulae, sometimes quite rough and resembling rugae, longitudinal costulae on the dorsum posterad including the mesonotum. Metanotum and propodeum with transverse costae. Promesonotal suture well to poorly im- pressed, metanotal groove well im- pressed. In lateral view the mesosoma has a very broadly convex dorsal margin, pro- podeal denticles brief and triangular, de- clivity concave. Sides of pronotum with regular longitudinal, slightly oblique cos- tae. Anepisternum fulcrum shaped. Kat- episternum with a well-developed flange. Mesopleura, propodeum and dorsal two- Volume 4, 1995 165 thirds of metapleura with transverse/ oblique costae; metapleura ventrad of spi- racle with longitudinal costae. Petiole in lateral view with a low and broadly convex dorsal margin, ventral process projects anterad. Dorsally with transverse costae, posterior margin ru- gose. Gastric dorsum with longitudinal costulation; ventrum of postpetiole with transverse rugae along posterior half; ir- regular rugulae anterad to the process. Base of foretarsus with a single stout se- tae opposite the strigil. Forecoxae with transverse costulae on lateral face, as well as dorsal faces of meso- and metacoxae. Legs shining and with abundant shallow punctulae. Very sparse pilosity on femora and tibiae. Color ferruginous to dark brown; mandibles, antennae and legs fer- ruginous. Queen, Male. — Unknown. Discussion. — This species may be con- fused with two other small rastrata group species: mstrata and mecoti/le. It is separa- ble from the former on account of the lon- ger scapes (surpassing the vertexal mar- gin), transverse costulae on the propo- deum as opposed to longitudinal, and a well-defined metapleura, separated from the propodeum laterally by an impressed line, lacking in rnstratn. The costulae of ine- cotyle have a coarse aspect in contrast with the smooth regularity in cuneiforma, tneco- tyle also has the striae on the scapes and dorsal mandibuar surface better defined and the masticatory border lacks the con- cavity of cuneiforma. Etymology. — The name alludes to the shape of the anepisternum. Gnamptogenys curvoclypeata Lattke Gnamptogenys curvoclypeata Lattke 1990: 12. Diagnosis. — Piceous species with ferru- ginous antenna, legs and coxae; prome- sonotal suture very weakly impressed, visible only with certain angles of light; meso- and metacoxae smooth and shining on median and basal lateral faces; meta- coxae with weakly developed triangular dorsal lobe. Ecology.— See Lattke (1990). Comments. — See Lattke (1990). Specimens Examined. — COLOMBIA, Amazonas: 7 km N Leticia (MCZC). Gnamptogenys ejuncida, Lattke, new species (Figs. 48, 50) Type Material. — Holotype worker. CO- LOMBIA, Putumayo, Mocoa, 610 m, 4-1- 77, C. Kugler, leg. Deposited in MIZA. Worker. — Holotype measurements: HL 1.10; ML 0.26; HW 0.88; SL 1.02; ED 0.18; WL 1.54 mm, CI 0.80; SI 1.56; OI 0.21. With head in frontal view: vertexal mar- gins slightly convex, lateral margins broadly convex and anterior clypeal mar- gin convex; mandibles triangular and with rugose dorsum; anterior pronotal face with six transverse costulae, dorsum with longitudinal costulae; mesonotum with anteriorly convex costulae, transverse at metanotum and posteriorly convex on dorsal propodeal face; anepisternum with three transverse costulae that descend from mesonotum; rest of mesosomal side obliquely costulate, some curve around propodeal spiracles; declivitous propodeal face with longitudinal costulae; in lateral view dorsal mesosomal outline evenly convex, with no abrupt breaks or depres- sions; petiolar node posteriorly inclined: in lateral view transversely costulate with convex anterior margin, concave posterior margin and overhanging apex; subpetiolar process triangular and projecting anterad. Anterior face to dorsal two-thirds of post- petiolar dorsum with transverse costulae; rest of gastric sculpture longitudinal; post- petiolar sternum anteriorly transversely costulate, posteriorly divergent; transverse costulae on procoxae and anterior faces of meso- and metacoxae; metacoxal tooth well developed. Abundant suberect hairs and pubescence on body and extremities. Queen, Male. — Unknown. Ecology. — Unknown. 166 Journal of Hymenoptera Research Figs. 45-53. Lateral view of posterior leg, mesosoma and petiole, and frontal view of head. 45-47, G. extra; 48-50, G. ejuncida; 51-53, G. pilosa. Discussion. — On account of size, sculp- ture, pilosity and posteriorly inclined pet- iole node this species seems to be most closely related to pilosa, a very similar spe- cies found in an Andean canyon NW of Cali, Colombia. The sculpture of pilosa is much finer and its pilosity is a bit denser, including abundant erect hairs, practically lacking in ejuncida. The presence of erect to suberect hairs in pilosa vs. subdecum- bent to decumbent hairs in ejuncida is es- pecially notable on the femora. In lateral view the angle formed by the junction of the dorsal and declivitous propodeal faces is greater in ejuncida than in pilosa, giving it a more slender appearance. Etymology. — This sleek species inspired the use of the Latin word for slender, ejun- cida. Gnamptogenys ericae Forel, revalidated Gnamptogenys tornata var. ericae Forel 1912:33; Brown 1958:229. Grtamptogenys sulcata (F. Smith); Brown 1958: 329; Lattke 1989:24. Gnamptogenys sulcatum bufonum Weber 1938: 208. NEW SYNONYMY. Diagnosis. — Mandibles subtriangular; posterior mesosomal dorsum and propo- deum with longitudinal costulae, trans- verse sutures barely visible under certain angles of lighting; small metacoxal tooth. Body black; antennae, legs and mandibles brown to dark brown. Ecology. — Can be found nesting in dis- turbed forests and coffee plantations as well as in undisturbed forests. Comments. — Much confusion has accu- Volume 4, 1995 167 mulated regarding the identity of this spe- cies and the similar sulcata. Many previous references to sulcata are actually ericae, but the confusion makes it quite difficult to distinguish between each reference. Dr. Ivan Lobl of the MHNG kindly examined the ericae type and confirmed the longitu- dinal direction of the costulae on the pos- terior face of the node of the specimen. A specimen in the MCZC from Guiana with a determination label as bufonum in N. Weber's handwriting and another label stating "in Bufo marinus stomach" coin- cides with the information in Weber's de- scription. Even though it has no label des- ignating it as a type, it seems reasonable to conclude that Weber's bufonum is a syn- onym of ericae. For a broader discussion on the separation of the forms known as tornata and sulcata the reader should turn to the "Discussion" for G. sulcata. Color in this species is relatively constant but oc- casional ferruginous, brown or mottled specimens will turn up. Size and shape of the petiolar node does not vary as much as in sulcata. Workers seem to always have longitudinal costulae on the propodeal de- clivity, though queens may have trans- verse costulae. Specimens Examined. — BOLIVIA, Rosa- rio. COLOMBIA, Magdalena: Minca, 610 m; Guajira: Serrania de Macuira, 7 km S Nazareth, 70-200m; Meta: San Juan de Ar- ama, 914m; La Macarena, Rio Guayabero, 260m; 65 km E Puerto Lopez. GUIANA, Forest Settlement, R. Mazaruni. TRINI- DAD, Port of Spain, Federation Park. VENEZUELA, Merida: Santa Cruz de Mora, 600m; Guarico: Hato Masaguaral; Bolivar: Campamento Rio Grande, 8°07'N 6r48'W, 280 m; Anzoategui: 20 km S El Tigre. Gnamptogenys extra Lattke, new species (Figs. 45-47) T\/pe Material. — Holotype worker. EC- UADOR, Pichincha: Tinalandia, 16 km SE Santo Domingo de Los Colorados, VI- 1976, S. & J. Peck, leg. Deposited in MIZA. Paratypes: (1) One worker with the same data as the holotype. Deposited in MCZC. (2) Two workers from COLOMBIA, Cho- c6: Lloro, Vereda Pefialoza, 26-III-1988, V. Jaimes, leg. One worker in BMNH and LACM, respectively. Worker. — Holotype (Paratypes) mea- surements: HL 1.20 (1.10-1.12); HW 1.00 (0.94-1.00); ML 0.34 (0.30-0.34); ED 0.16 (0.14-0.18); SL 1.26 (1.00-1.20); WL 1.68 (1.44-1.50) mm; CI 0.83 (0.85-0.98); SI 1.26 (1.06-1.20); lO 0.16 (0.15-0.19) n=4. With head in frontal view: vertexal mar- gin straight, laterally convex with eyes slightly behind midlength, and anterior clypeal margin convex; mandibles trian- gular with dorsal striae; scapes relatively thick in appearance and surpassing ver- texal margin; vertex with 2-3 transverse costulae; anterior pronotal face with 5-7 transverse costulae, dorsally with anteri- orly arching costulae; costulae on meso- notal concentric or longitudinal; costulae on metanotum and dorsal propodeal face transverse to longitudinal, sometimes arched; declivitous propt^deal face with longitudinal costulae that diverge poste- riorly, basally with 2-3 transverse costu- lae. Petiolar node in lateral view subquad- rate, with broad anterodorsal convexity and sharp posterior drop with overhang- ing apex; subpetiolar apex is an anteriorlv projecting lobe; anterior postpetiolar face with transverse costulae that arch back and bect")me longitudinal posteriorlv; rest of dorsal gastric cc^stulae longitudinal, di- vergent on posterior postpetiolar sternum; ventral stridulitrum present on fourth ab- dominal segment. Scapes with abdundant decumbent pilosity and no suberect or subdecumbent hairs; decumbent pilosity also present on coxae and legs, with no standing hairs; sparse suberect hairs on rest of body. Body dark brown, legs and antennae slightly lighter. Male, Female. — Unknown. Discussion. — This species has a fairly pointed petiolar node apex, but it doesn't 168 Journal of Hymenoptera Research approach the extremely pointed condition of G. acuta. The node is more robust and subquadrate (lateral view) than in other porcata subgroup species and is reminis- cent of the condition in G. tortiiolosa. This species could be closely related to G. stria- tula on account of the node shape (erect) and the lacking anterior peduncle. Gnamptogenys fernandezi Lattke Gnamptogenys fernandezi Latike 1990: 14. Diagnosis. — Triangular mandibles with a smooth and shining dorsum; anterior la- mella of clypeus convex; promesonotal su- ture lightly impressed and mesometanotal suture absent; low, elongate petiole, pos- terodorsally bluntly pointed; metacoxal teeth absent. Ecologif. — A rarely collected, but fairly widespread species in n South America. In Venezuela known only from 2 forested lo- calities in the Cordillera de la Costa be- tween 650-1100 m. Comments. — See Lattke (1990). Specimens Examined. — COLOMBIA: Val- le: 50 m (no other data). ECUADOR. Mo- rora: Santiago, Los Tayos. PERU, Madre de Dios: 5 km SE Prto. Maldonado, Rio Tambopata, 260m; Cuzco Amazonico, 15 km NE Prto. Maldonado, 200m. VENE- ZUELA, Carabobo: Hacienda Palmichal, Canoabo 900 m. Gnamptogenys gentryi Lattke, new species Type Material. — Holotype worker. CO- LOMBIA, Valle, 2250 m, Insp. Pance, Re- serva Natural Hato Viejo, J.B. Hillaire, leg. Deposited in MIZA. Thank you to Fernan- do Fernandez for making the specimen available. Worker. — Holotype measurements: HL 1.22; ML 0.32; HW 0.92; ED 0.22; SL 1.04; WL 1.64 mm; CI 0.75; SI 1.13; Ol 0.24. Head in frontal view elongate, anterior clypeal margin convex, posterior margin concave, eyes relatively flat, finely costu- late throughout, though diverging at ver- tex, with a band of three transverse cos- tulae next to the vertexal carinae; mandi- bles triangular, frontal surface rugulose. Scapes mostly smooth and shining except for sparse punctulae. Pronotal dorsum with whorled costulae almost enclosing 3 brief longitudinal cos- tulae, rest of mesosomal dorsum with lon- gitudinal costulae, metanotal groove deeper than pronotal suture. In lateral view the mesosoma has a convex pronotal profile that becomes an inclined and more or less straight slope that bends sharply at the propodeal declivity, which has a broadly concave profile. Node in lateral view tilted backwards, with oblique to longitudinal rugulae laterally and trans- verse costulae on the anterior and poste- rior faces, ventral process subquadrate, with anterior corner projecting more than posterior corner. Gastric tergite I mostly smooth and shining except for anterior face which has a small area of transverse costulae and the sternite with transverse striae; the rest of the gaster is smooth and shining. Color black. Queen, Male. — Unknown. Discussion. — This striatula group species is quite easy to separate from all others just considering the presence of a metan- otal groove and the mostly smooth and shining gaster. Etymology. — Named in honor of the late botanist Alwyn Gentry, he died in the line of duty in the Andes of Ecuador. Gnamptogenys gracilis Santschi Holcoponera gracilis Santschi 1929:468. Gnamptogenys gracilis (Santschi); Brown 1958: 228, 232; kempt 1972:112; Lattke 1990:17. Diagnosis. — Finely costulae; pronotal costulae concentric to arched; mesonotum frequently with longitudinal costulae sur- rounded by concentric costulae; metano- tum and propodeal dorsum with trans- verse costulae, and longitudinal on decliv- itous propodeal face; subpetiolar process anteriorly triangular and posteriorly an- gular. Volume 4, 1995 169 Ecology. — This is an epigaeic forager known only from primary and secondary wet forests of the Guiana Shield and its remnants in the Venezuelan Amazonas, mostly between altitudes of 850-1600 m. Within the range of 1000-1200 it can be locally abundant. One record from 550 is next to Angel Falls. This suggests that proximity with the higher forested talus slopes of Auyan Tepuy could explain its ocurrence at such a low elevation. This species obviously prefers cool, mesic hab- itats. This distribution plus climatic evi- dence (Schubert 1988) suggests that G. gracilis may have been more widely dis- tributed in this area during the last glacial period. Comments. — On account of size, similar sculpture, posteriorly inclined node and fair amount of pilosity, its nearest relative appears to be pilosa, also an apparent en- demic, of the Colombian Cordillera N and NNW of Cali. An occasional specimen may have a second, smaller seta just be- low the prominent seta of the foretarsal base. Specimens Examined. — VENEZUELA, Bolivar: Auyan Tepuy, Sector Aonda, 6°02'N 62°37'W, 1600 m; Auyan Tepui, Cafion del Diablo, Isla Raton, 550 m; San Ignacio de Yuruani, 5°00'N 61°08'W, 800 m (MIZA). Gnamptogenys haenschi Emery Gnamptogcmis haenschi Emery 1902:27; Brown 1958:302;"Kempf 1972:112. Diagnosis. — Vertexal margin concave in frontal view; mandibles with longitudinal costulae; declivitous propodeal face with longitudinal striae superiorly and basally transverse rugae; petiolar node transverse; postpetiolar sternum mostly transversely strigulate; no metacoxal tooth. Ecology. — This species inhabitats forest and apparently nests in rotten wood as well as in soil. It has also been taken in oil palm plantations. Some specimens were collected from carrion traps baited with iguana meat. Longino (pers. comm.) re- ports finding a nest beneath a rotten log; the ants burrowed into the soil on expo- sure. Comments. — This species is possibly the only survivor of an otherwise extinct lin- eage in the genus. See discussion on phy- logeny of the different species groups. Specimens. — BOLIVIA, Lower Rio Mad- idi; El Montero, 70 km N Sta. Cruz. CO- LOMBIA, Cauca: Isla La Gorgona; Guaji- ra: Quebrada Guacoche, vie. Don Diego, 10 m; Magdalena: 4 km N San Pablo, 10°57'N 74°03'W, 550 m. COSTA RICA, Peninsula de Osa: Corcovado, 8°28'N 83°35'W; Heredia: La Selva; 3 km S Prto. Viejo, 10°26'N 84°00'W, 50m. ECUADOR, Napo: Prto. Misahualli, 350 m. PERU, Tin- go Maria: Valle de Monson; Puerto Mal- donado, 260m; Loreto: Ramon Castillo, 5 km NW Leticia. PANAMA, Darien: Rio Tacaruna, 580m. VENEZUELA, Alto Rio Siapa, 1°40'N 64°35'W, 530m. Gnamptogenys hartmani (Wheeler) Ectatoniuw {Parectatomma) hartmani Wheeler 1915:390. Gnamptogenys txartmani (Wheeler); Brown 1958: 228, 302; Echols, 1964:137; Kempf 1972:112; MacKay 1988:127. Gnamptogenys nigrifnvis Borgmeier 1948; Brown 1958: 228, 236; Kempf 1972:114. NEW SYN- ONYMY. Gnamptogenys turmalis Kempf and Brown 1968: 93; Kempf 1972:116. NEW SYNONYMY. Diagnosis. — Superolateral corners of de- clivitous propodeal face with small lobes or carinae; mandibles triangular to sub- triangular; metanotal groove vaguely im- pressed and posterior nodal face has lon- gitudinal costulae; metacoxal dorsum with lobes; subpetiolar process subquad- rate. Ecology. — One record from soil in a ba- nana farm and another from soil in a de- stroyed Trachymyrmex nest. These ants have on several occasions been reported as predators of Trachymrmex ants (Echols, 1964:137; Kempf and Brown, 1968:94). J. 170 Journal of Hymenoptera Research Longino (pers. comm.) reports from Costa Rican field notes of Dana Myer: "a nest was found found in leaf litter amidst the remains of a Trachymrmex nest and many cut up workers and a queen of the attines were also found along with many wound- ed Gnamptogenys workers." Longino has observed this species carrying its brood in a loose 3 m column, fleeing from raiding Eciton. One specimen was taken from the stomach contents of Dendrobates lecomelns. Comments. — Kempf and Brown de- scribed G. turmalis as being close to G. ni- grifrons but slightly larger and more ro- bust, with finer sculpturing and a uniform ferruginous color. These authors realized that G. hartmani, G. nigrifrons and G. tur- malis were so close as to constitute possi- ble synonyms, and they were right. The study of specimens at hand show colors vary enough to make it an unreliable char- acter for separating species and the same is true for irregularities in the sculpture. Specimens from more southern localities have finer costulation. Also found to vary continously was the length vs. width of the petiolar node, as well as other size in- dicators. Specimens Examined. — BRAZIL, Bahia, Ilheus-Itab., km 22, (CEPEC). COSTA RICA, Peninsula de Osa: Sirena, 8°28'N 83°35'W, 50m. GUYANA, Kartabo; HON- DURAS, La Lima, Zapote farm. MEXICO, Tamaulipas: 10 km W EL Encino, 23°N 08'W 99°10'W; PANAMA, Isla Barro Col- orado. PERU, Valle de Chanchamayo, 800 m. USA, Louisiana: Lucky; Texas: Bent- son-Rio Grande State Park, 30. VENE- ZUELA, Amazonas, vie. Cerro Ya-Pakana. Gnamptogenys haytiana (Wheeler and Mann) Spaniopone haytiana Wheeler and Mann 1911:11. Gnamptogenys hai/tiana (Wheeler and Mann); Brown 1958:228, 316; Kempf 1972:113. Diagnosis. — Petiole node disciform, with anterior and posterior faces more or less parallel to one another, dorsal and lateral nodal faces with transverse rugulae; me- tanotal groove barely impressed; propo- deal spiracles mounted on turrets at mid- height of the lateral edge of declivitous propodeal face. Ecology. — Taken from forest and coffee plantation leaf litter. The dramatic defor- estation of Hispaniola has considerably re- duced the range of this species. Comments. — Endemic to Hispaniola Is- land. On account of the yellowish color, more pronounced curvature of the second gastric segment and different sculpture, this species is outstanding amongst the other members of its species complex. It is possible that it represents an independent development from the strigata complex. The holotype was examined in the MCZC. Specimens Examined. — DOMINICAN REPUBLIC, Barahana: 2 km N Polo, 1000m. Gnamptogenys horni Santschi Gnamptogenys regularis horni Santschi 1929:475. Gnamptogenys horni Santschi; Brown 1958:228, 235, 316;' Kempf 1961:491; Kempf 1972:113; Lattke 1990:17. Diagnosis. — Subtriangular mandibles; dorsum of petiolar node with posterome- dian longitudinal costulae, transversely arched anteriorly; episternal costulae curve on to declivitous propodeal face. Body dark brown; legs and antennae fer- ruginous. Ecology. — A series was taken from the stomach of Colostethus nubicola. See Lattke 1990. Comments. — This common species has a wide range in the neotropics. There is fair size variation and smaller specimens usu- ally have more vermiculate costulation. The position of the propodeal spiracle can vary from mid-heigth to below mid-heigth and the metacoxal tooth from triangular to denticulate. Nests of small forms can sometimes be found in the same locality as larger forms. Intermediates were found in other localities. A specimen labelled as a "cotype" was examined in the MCZC. Volume 4, 1995 171 Specimens Examined. — BRAZIL, Amazo- nas: Rio Taruma; Benjamin Constant; Re- serva Ducke; via Manaus-Itacoatiara km 50; Manaus; Rondonia: Vilhema; Para: Ca- rajas; Utinga Tract vie. Belem. BOLIVIA, Caranavi, vie. radio, 800m. COLOMBIA, Choco, Rio Napipi. ECUADOR, Pichincha: Estacion Rio Palenque; Sucua; Morona: Santiago, Los Tayos; Esmeraldas: 48 km S Atacames. GUYANA, Kamakusa. PANA- MA, Isla Barro Colorado. PERU, Pangu- ana, 9°37'S 74°56'W, 220m; Tingo Maria: Valle del Monson. SURINAM, Tambah- redjo; Dirkshoop; La Poolle; Maripa- hueuei, Vank. VENEZUELA, Amazonas: Alto Rio Mavaca, 2°02'N 65°06'W, 200m; Alto Rio Siapa, 1°40'N 64°35'W, 600m. Gnamptogenys ilimani Lattke, new species Type Material. — Holotype worker. BO- LIVIA, 22 km N Caranavi, Vivero Ilimani, 1700 m, 22-VI-81, C. Kugler, leg. Primary forest clearing with Cinchona, nest in rot- ten wood. Deposited in the MIZA. Para- types: paranidotypic workers from the same nest as the holotype. Deposited in each of the following: BMNH, LACM, MUSP, MCZC, MIZA. Worker. — Holotypes (Paratypes) mea- surements: HL 0.90 (0.84-0.90); HW 0.78 (0.72-0.78); ML 0.30 (0.26-0.32); ED 0.14 (0.14-0.18); SL 0.70 (0.70-0.72); WL 1.20 (1.12-1.20) mm; CI 0.87 (0.84-0.88); SI 0.90 (0.92-0.97); OI 0.19 (0.18-0.27) n=5. Coarsely costulate strigata group spe- cies. With head in frontal view vertexal margin convex; anterior pronotal face with 3-4 costulae that sharply bend back laterally to become longitudinal; declivi- tous propodeal face with longitudinal cos- tulae; node anteroposteriorly compressed, its posterior face sharply marginate later- ally with a flat to slightly concave surface and effaced sculpture; anterior postpetio- lar face with 2-3 transverse costulae, rest longitudinal; postpetiolar sternum with longitudinal costulae; lateral mesocoxal face with oblique, rough costulae and that of metacoxae with rough rugulae; meta- coxal tooth low and triangular. Body black with yellow-brown legs and scapes; abun- dant appressed pilosity on legs and scapes, but no standing hairs. Qi/mj.— Measurements: HL 0.88 (0.88); HW 0.72 (0.74); ED 0.20 (0.18); SL 0.72 (0.72); WL 1.28 (1.20) n = 2. Differences from workers are the usual; caste differ- ences, though the node is more disciform. Male. — Unknown. Discussion. — This species is nearest to G. strigata and G. pittieri but they are sepa- rable on several on several accounts: the posterior petiolar node face is not sharply set off and has raised costulae; in dorsal view their node is anteriorly convex, not straight; standing hairs on the body and especially the scapes are abundant, scape pilosity is longer. In G. pittieri the anterior transverse costulae of the pronotum curve around gently at the sides. Etymology. — The species name alludes to the type locality. Gnamptogenys ingeborgae Brown Gnamptogenys ingeborgae Brown 1992:279. Diagnosis. — Mandibular masticatory margin denticulate, promesonotal suture present, at least partially, and the longi- tudinal costulae on the propodeal dorsum and postpetiolar tergite is frequently ill- defined to effaced. Ecology. — A specialized millipede feeder from Colombia. See Brown 1992. Comments. — See Brown 1992. Specimens Examined. — COLOMBIA, Cundinamarca, Bogota-Villavicencio road, km 79, 1100 m (type series: MCZC). Gnamptogenys interrupta Mayr Ectatomuia {Gnamptogenys) interruptum Mayr 1887:543. Gnamptogenys interruptum (Mayr); Mann 1922:3; Brown 1958:228, 303; Kempf 1968:377; Kempf 1972: 113; Lattke 1990:18. Diagnosis. — Sublinear mandibles. Ce- phalic dorsum, mesosoma, and gastric ter- 172 Journal of Hymenoptera Research gum I with longitudinal costulae; gastric tergum II smooth; pleura also with smooth patches; metacoxal tooth absent. Body reddish brown; legs and antennae ferruginous. Ecology. — Found in humid forests of lowlands and mountains (cloud forest). Taken from leaf litter samples and beneath bark of rotting logs on ground. Comments. — Lateral mesosomal costulae can be effaced to a variable degree and the second gastric segment can ocassionally have weak longitudinal costulae, medi- anly effaced. Declivitous propodeal face with longitudinal costulae and weakly de- veloped anterolateral lobes. Specimens Examined. — COLOMBIA, Magdalena: 3 km SE Minca, 11°08'N 74° 06'W, 1050m; Valle: Puerto Merizalde, 10 m. COSTA RICA, Puntarenas: Montever- de, 10°18'N 84°48'W, 1500m. HONDU- RAS, Lombardia; JAMAICA, Mandeville. Gnamptogenys kempfi Lenko Gnamptogenys kempfi Lenko 1964:257; Kempf 1972:114. Diagnosis. — Mandibles subtriangular, dorsally smooth and shining, transverse costulae on mesosomal dorsum and node, no apparent transverse sutures on meso- somal dorsum. Propodeal spiracle large and metacoxal tooth present. Ecology. — Apparently from lowland for- est, one nest was found in rotten wood on the ground. Comments. — A widepread but uncom- mon species. Specimens Examined. — COLOMBIA, Amazonas: 7 km N Leticia. PERU, Pan- guana, 9°37'S 74°56'W, 220m (MCZC). Gnamptogenys lanei Kempf Gnamptogenys lanei Kempf 1960:388-90; Kempf 1968:377; Kempf 1970: 325; Kempf 1972:114. Diagnosis. — Relatively small eyes; clyp- eal lamella medianly concave; promeson- otal suture vestigial and small denticles on the propodeum; node elongate, with transverse costulae. Ecology. — Its morphology puts it in the rastrata group of millipede hunters. Comments. — Kempf 1968:377 reports a series with transverse costulae on the an- terior pronotal face, differing from the lon- gitudinal sculpture of the type series. MflfenVi/.— BRAZIL, Para: Belem, Rio Guama (MUSP). Gnamptogenys laticephala Lattke, new species (Figs. 54, 55) Type Material. — Holotype worker. EC- UADOR, Guayas: 3 km S Bucay, 24-VII- 73, W.L. Brown, leg. Deposited in MCZC. Worker. — Holotype measurements: HL 1.71; ML 1.01; HW 1.94; SL 1.53; ED 0.34; WL 2.32 mm; CI 1.13; SI 0.79; OI 0.18. Head in frontal view broad, vertexal margin fairly straight, lateral margins slightly convex and anteriorly diverging; eyes large and situated at cephalic mid- length; anterior clypeal lamella broadly concave with lateral triangular teeth near mandibular insertions; clypeus with two small anteriorly projecting lobes between anterior border of antennal fossae and an- terior lamella; cephalic dorsum with lon- gitudinal rugae that diverge slightly pos- terad on to vertex; rugae between frontal lobes and eyes more irregular than rugae between and behind frontal lobes; abun- dant piligerous punctures present. Anten- nal scapes smooth and shining, surpassing vertexal margin; gula with median longi- tudinal carinae and transverse costae at each side; mandibles falcate, longitudinal- ly costulate with interior glabrous sulcus that ends at apical tooth, basal tooth marks the end of basal margin and is fol- lowed by two more pre-apical teeth; an- terior pronotal face opaque, not as shiny as rest of pronotum and with faint trans- verse rugulae; mesosomal dorsm with parallel longitudinal costulae, promeson- otal suture well impressed but does not break longitudinal sculpture, mesometa- Volume 4, 1995 173 Fig. 54-55. Micrographs of G. laticcplmla. Scale bars =1.0 mm. 54, head; 55, body. 174 Journal of Hymenoptera Research notal suture deep and interrupting sculp- ture except for some ridges that join a few costulae on either side of suture; lateral pronotal face with slightly oblique parallel costulae; pronotal-mesopleural suture very broad and deep; meso- and metepis- terna with parallel, longitudinal costulae; mesometepisternal suture well impressed and metepisternal-propodeal suture dis- tinct; metepisternal lobe well developed; declivitous propodeal face with longitu- dinal costulae; petiolar node in lateral view low, anterodorsal margin convex and posterodorsal margin sharply convex; transversely costulate; subpetiolar process anteriorly triangular, with slighty round- ed posterior lobe. Postpetiolar costulation longitudinal, weakly roughened and becoming rugulo- se laterally, ventrally with transverse cos- tulae that tend to fade medially; gastric tergum II strongly vaulted anteroventrally with dorsal and lateral sculpture as on preceding segment; apical gastric seg- ments with faint transverse strigulae, tending to smooth and shining; anterior face of procoxae smooth and shining lat- erally with oblique costulae; meso- and metacoxae laterally smooth and shining, dorsally transversely costulate; metacoxal teeth well developed; body with sparse suberect and decumbent hairs, longest on gastric apex and clypeus, no appressed pi- losity on scapes, just suberect and decum- bent hairs. Mesosoma dark reddish brown; head, node and gaster darker; legs and antennae brown; mandibles brown to yellow brown. Queen, Male. — Uknown. Discussion. — The two lobes between the antennal fossae and clypeal lamella distin- guish this species from all other New World members of the group, which lack such lobes, including its closest relative, banksi. Other characters in which banksi differs are the following: areolate sculp- ture on the cephalic dorsum, occiput and anterior pronotal face with transverse ru- gae, deep lateral and mesonotal pits for the spiracles (such pits lacking in frauda- trix), pronotal suture interrupts sculpture, and the inner mandibular mandibular sul- cus only reaches the basal angle. In gen- eral the sculpture of banksi is rougher, mostly of gaster I, which is reticuate ru- gose and gaster II has rough, widely spaced, dorsal longitudinal rugae, becom- ing more irregular laterally. The form of the pronotal-mesopleural suture and the lobes formed by the pronotum and anepi- sternum at the base of the suture are rem- iniscent of some minuta group species. The possibility exists that this could be an ab- errant specimen of banksi (it was found de- termined as such), but giv