5to3 "S.fijj! Journal of Hymenoptera Research Volume 16, Number 2 N^ / /RPMafcS^ October 2007 ISSN #1070-9428 CONTENTS GESS, F. W. The genus Quartinia Ed. Andre, 1884 (Hymenoptera: Vespidae: Masarinae) in southern Africa. Part I. Descriptions of new species with complete venation 211 GRISSELL, E. E. Torymidae (Hymenoptera: Chalcidoidea) associated with bees (Apoidea), with a list of chalcidoid bee parasitoids 234 NEFF, J. L. and A. W. HOOK. Multivoltinism and usage of multiple nest substrates in a west Texas sand dune population of Psendomasaris phaceliae Rohwer (Hymenoptera: Vespi- dae: Masarinae) 266 PACKER, L. Mydrosoma micheneri Packer, new species, a new diphaglossine bee from Brazil (Hymenoptera: Colletidae) 277 PACKER, L., A.-I. D. GRAVEL, and G. LEBUHN. Phenology and social organization of Halictus (Seladonia) tripartitus (Hymenoptera: Halictidae) 281 PULAWSKI, W. J. The status of Liris magnificus Kohl, 1884, and Trachogorytes costaricae R. Bohart, 2000 (Hymenoptera: Crabronidae: Crabroninae, Bembicinae) 293 PUNZO, F. Interspecific variation in hunting behavior of Pepsis grossa (Fabricius) and Pepsis thisbe Lucas (Hymenoptera: Pompilidae): a field study 297 SHIMIZU, A. and R. WAHIS. Systematic studies on the Pompilidae occurring in Japan: genus Irenangelus Schulz (Hymenoptera: Pompilidae: Ceropalinae) 311 WENG, J. L. and BARRANTES, G. Natural history and larval behavior of the parasitoid Zaty- pota petronae (Hymenoptera: Ichneumonidae) 326 INTERNATIONAL SOCIETY OF HYMENOPTERISTS Organized 1982; Incorporated 1991 OFFICERS FOR 2007 Michael E. Schauff, President James Woolley, President-Elect Michael W. Gates, Secretary Justin O. Schmidt, Treasurer Gavin R. Broad, Editor Subject Editors Symphyta and Parasitica Aculeata Biology: Mark Shaw Biology: Jack Neff Systematics: Andrew Deans Systematics: Wojciech Pulawski All correspondence concerning Society business should be mailed to the appropriate officer at the following addresses: President, Plant Sciences Institute, Bldg. 003, Rm. 231 BARC-West, Beltsville, MD 20705, USA; Secretary, Southwestern Biological Institute, 1961 W. Brichta Dr., Tucson, AZ 85745, USA; Treasurer, PO Box 37012, c/o Smithsonian Institution, MNMH, MRC168, Washington, DC 20013-7012, USA; Editor, Dept. of Entomology, The Natural History Museum, Cromwell Road, Lon- don SW7 5BD, UK. Membership. Members shall be persons who have demonstrated interest in the science of entomol- ogy Annual dues for members are US$45.00 per year (US$40.00 if paid before 1 February), payable to The International Society of Hymenopterists. Requests for membership should be sent to the Trea- surer (address above). Information on membership and other details of the Society may be found on the World Wide Web at http://hymenoptera.tamn.edu/ish/. Journal. The Journal of Hymenoptera Research is published twice a year by the International Society of Hymenopterists, % Department of Entomology, Smithsonian Institution, Washington, D.C. 20560- 0168, U.S.A. Members in good standing receive the Journal. Nonmember subscriptions are $60.00 (U.S. currency) per year. The Society does not exchange its publications for those of other societies. Please see inside back cover of this issue for information regarding preparation of manuscripts. Statement of Ownership Title of Publication: Journal of Hymenoptera Research. Frequency of Issue: Twice a year. Location of Office of Publication, Business Office of Publisher and Owner: International Society of Hymenopterists, 0 Department of Entomology, Smithsonian Institution, 10th and Constitution NW, Washington, D.C. 20560-0168, U.S.A. Editor: Gavin R. Broad, Department of Entomology, The Natural History Museum, Cromwell Road, London SW7 5BD, UK. Managing Editor and Known Bondholders or other Security Holders: none. This issue was mailed 24 September 2007 J. HYM. RES. Vol. 16(2), 2007, pp. 211-233 The Genus Quartinia Ed. Andre, 1884 (Hymenoptera: Vespidae: Masarinae) in Southern Africa. Part I. Descriptions of New Species with Complete Venation Friedrich W. Gess Albany Museum, Grahamstown, 6140 South Africa; email: f.gess@ru.ac.za Abstract. — In this publication, the first of a projected series revising the Afrotropical (essentially southern African) species of the genus Quartinia Ed. Andre, 1884 (Hymenoptera: Vespidae, Masarinae), eleven new species are described. Of these, seven occurring variously in the southern Namib Desert and in its southward extension down the western coast of South Africa, and one occurring on the southern coast of South Africa, have been found nesting in sand-filled snail shells. They are: australis, bonaespei, conchicola, namaqua, namaquettsis, obibensis, and refugicoln. To these species is added vexilhita which is presumed to have the same nesting habits. A key to distinguish these species is given. The other three newly described species, all from Namibia, are: femorata, geigeriae and lamellata. Following van der Vecht and Carpenter (1990) Quartinia Ed. Andre, 1884 is here understood to include, as junior subjective synonyms, Quartiniella Schulthess, 1929 and Quartinioides Richards, 1962. As has been pointed out by Carpenter (2001), Quartiniella and Quartinioides were primarily based on the partitioning of a trend in the reduction of wing venation, Quartiniella being defined on the basis of the loss of 3rs-m and 2m-cu and Quarti- nioides because it has 2m-eu present but attenuate and interrupted, whereas Quarti- nia has it complete. In Quartiniella in particular and to some extent also in Quartinioides reduction of wing venation is a correlate of overall size reduction As formal taxonomic partitioning of essential- ly continuous variation is an unacceptable practice, Carpenter synonymized Quarti- niella and Quartinioides with Quartinia, a view with which the present author is in full agreement. Nevertheless, in view of the large num- ber of species in Quartinia, adoption of the above venational characters to divide the genus into smaller, more manageable but totally informal, non-natural units is found to be useful. Thus the present paper deals with species with complete venation - that is species which in the past would have been placed in Quartinia sensu stricto. In his revision Richards (1962) dealt with a total of 61 southern African species, 18 being placed in Quartinia, 38 in Quarti- nioides and five in Quartiniella. Of these, 11, 26 and two respectively were described as new. One additional species, placed in Quartinioides was added (Richards 1982). Available to Richards in 1962 were just over one thousand specimens - 140 Quar- tinia, 727 Quartinioides and 148 Quartiniella. Ten species were known from only one specimen, 30 species from only one sex. It is clear that Richards suffered from a pau- city of material. Particularly the lack of large samples from individual populations spread over the distributional area pre- vented him from appreciating factors such as intraspecific variation and geographical clines. In some instances the associations of sexes is of doubtful validity, especially where males and females are from widely separated localities. 212 Journal of Hymenoptera Research The present study is based on over 6000 specimens, most of which were purpose- fully collected. A large proportion of the specimens have associated biological data - mostly flower visiting records but also, for some species, nesting data. Desirable as it might be to undertake a complete revision of the genus, this is at present not practicable. Rather than to get bogged down in a study which might never be completed and published, it is intended to publish a series of papers describing new species as well as review- ing some known species. It is envisioned that a new key to species will complete the series. Quartinia species range in length from a little over 2 mm to 7 mm. In comparison with the great majority of species of other genera of Masarinae even the largest Quartinia are relatively small. In view of the considerable range in size shown by species of Quartinia and in order to express relative size, categories based on length have been established for species of the genus. These are: minute (1.5-2.5 mm); small (2.5-3.5 mm); medium (3.5-4.5 mm); large (4.5-5.5 mm); very large (5.5- 6.5 mm); and gigantic (6.5-7.5 mm). The notation used for expressing geo- graphic co-ordinates is as in the gazetteer of The Times Atlas of the World (1981). The figures before the stop are degrees, those after the stop are minutes; the stop is not a decimal point. For purposes of plotting distributions, co-ordinates have been given in square brackets in the text for those localities for which none are given on the data labels. On a few data labels from collections other than that of the Albany Museum the collecting locality is followed by degree latitude and degree longitude and by half- and quarter-degree reference letters ac- cording to the Degree Reference System of Leistner and Morris (1976). As this system is not universally understood an attempt has been made here to find on a map the localities concerned and to add in square brackets the co-ordinates ex- pressed in the manner adopted in this paper. In listing the material examined, the localities have been arranged, as far as practicable, in north to south order within countries or, in the case of South Africa, within provinces. Acronyms for institutions in which material is housed are: AMG = Albany Museum, Grahamstown, South Africa; CAS = California Academy of Sciences, San Francisco, United States of America; FSCA = Florida State Collection of Arthro- pods, Gainesville, United States of Amer- ica; NCP = National Collection of Insects, Pretoria, South Africa; NNIC = Namibian National Insect Collection, Windhoek, Namibia. DESCRIPTION OF SPECIES AND COLLECTION DATA A) Species nesting in sand-filled snail shells or (vexillata) presumed to do so. Quartinia australis Gess, new species Diagnosis. — Large (5.0-5.4 mm). Fore wing with Cula and 2m-cu complete and as thick as the other veins. Tegula with posterior inner corner inwardly produced. Both sexes predominantly black with limit- ed white markings; male with clypeal disc and underside of scape and pedicel white. Description. — Female: Black. The follow- ing are white: narrow anterior margin of pronotum (in most specimens) and extreme postero-dorsal angle of same; tegula anteri- orly and posteriorly; lateral lamella of scutellum; posterior bands medially on terga I-V (that on V in some specimens reduced to a postero-medial spot); distal end of fore femur; streaks on fore and middle tibiae; proximal and distal ends of hind tibia. Brown are: rest of legs; underside of flagellum. Wings lightly browned. Length 5.0-5.4 (average of 5:5.3 mm); length of fore wing 3.4-3.6 mm (average of 4:3.53 mm); hamuli 6. Volume 16, Number 2, 2007 213 Head in front view 1.31 X as wide as long, finely microreticulate, matt; clypeus apunctate; frons and vertex with shallow punctures separated by about their width (punctures barely perceivable on lower regions of frons, clearer in ocular sinuses and upper part of frons and particularly on vertex. POL:OOL = 1:0.6. Clypeus 1.5 X as wide as long; anterior margin shallowly and widely emarginate; antero-lateral an- gles rounded. Mesosoma microreticulate, moderately shiny, with punctures larger and more obvious than on head. Gaster microreticulate but shiny; punc- tures finer and shallower than on head and much more so than those on mesosoma, becoming progressively finer posteriorly. Male: Black. White markings as in fe- male, with in addition: labrum; disc of clypeus (i.e. not sides nor areas adjacent to antennal sockets); underside of scape and pedicel; posterior band on tergum VI; distal end of middle and hind femora. Underside of flagellum white suffused with reddish-brown. Length 5.0-5.4 (average of 4:5.1 mm); length of front wing 3.4-3.6 mm (average of 4:3.42 mm); hamuli 6. Head in front view 1.5 X as wide as long; POL:OOL = 1:0.6. Clypeus 1.5 X as wide as long; anterior margin shallowly and widely emarginate; antero-lateral angles rounded. Microsculpture and punctuatation of head and body similar to that of female. Tergum VII (Fig. 6) with surface notice- ably depressed and with hindmargin with a short median slit. Sterna I-VI unmodi- fied; sternum VII trilobed, basally marked- ly concave between swollen and poster- iorly produced lateral lobes and with median lobe flat and densely setose. Etymology. — The name australis draws attention to the southern provenance of the species. Material examined.— Holotype: J, SOUTH AFRICA: WESTERN CAPE: Witsand (34.23S 20.52E), 14.viii.2002 (F. W. and S. K. Gess) (ex nest in shell of Tlieba pisana (Mull.), Helicidae) [AMG]. Paratypes: SOUTH AFRICA: WEST- ERN CAPE: same data as holotype, 7 99, 4 Jg (ex nests in shells of Tlieba pisana (Mull.), Helicidae) [AMG]. Geographic distribution. — Known only from the type locality, Witsand, near Port Beaufort at the mouth of the Bree River, a little to the west of the southernmost point of Africa. Floral associations. — Unknown. Nesting.— Found nesting in sand-filled shells of the exotic Tlieba pisana (Mull.) (Mollusca: Gasteropoda: Pulmonata: Heli- cidae) collected from the surface of the sand below bushes growing on supra- littoral dunes. Quartinia bonaespei Gess, new species Diagnosis.— Very large to gigantic (6.3- 7.0 mm). Fore wing with Cula and 2m~cu complete and as thick as other veins. Tegula short, laterally rounded, with pos- terior inner corner inwardly produced. Both sexes black with white-marked pro- notum, tegula and scutellar lamella and with wide, bright reddish-orange posterior bands on all but last two terga. Male with greatly enlarged and modified fore leg, with somewhat modified middle and hind legs, with tergum VII apico-medially drawn out into a robust, dorsally flattened and apically rounded process, and with sterna medially depressed. Description. — Female: Black. The follow- ing are white: medially interrupted trans- verse band on dorsum of pronotum and minute dot at postero-dorsal angle of same; anterior and posterior thirds of tegula (median third black); medially broadly interrupted band on lamellate margin of scutellum. Bright reddish-orange are: mandibles distally; posterior markings dorsally (i. e. not extending down sides) on terga I-IV (that of tergum I wide, covering entire dorsal surface, that of II slightly narrower, that of III wide medially but narrowed laterally, that oi IV a median 214 Journal of Hymenoptera Research transverse spot). Labrum brown. Under- side of antennae, to various degrees, pale. Coxa, trochanter, femur and tibia of all legs black with exception of yellowish streak on antero-dorsal surface of fore tarsus and same colour on extreme base of middle and hind tibiae; tarsomeres dark brown. Wings brown; veins dark brown to black. Length 6.3-7.0 mm (average of 5:6.7 mm); length of fore wing 4.3-4.5 mm (average of 5:4.4 mm); hamuli 6. Head in front view 1.29 X as wide as long, microreticulate, matt, with small, shallow punctures (sparse on clypeus, well separated on lower part of frons but progressively closer on upper part of frons and on vertex). POL:OOL = 1:0.75. Clyp- eus 1.3 X as wide as long; anterior margin shallowly emarginate; antero-lateral angles rounded. Mesosoma microreticulate, matt, with punctures slightly larger and deeper than on head (moderately well separated on mesoscutum and scutellum, closer on pro- notum and upper part of mesopleuron where sculpture almost reticulate-punctate). Gaster microreticulate but shiny; punc- tures finer and shallower than on head and mesosoma, becoming progressively finer posteriorly. Male: Black. White markings as in fe- male. Bright reddish-orange markings on gaster similar to those of female but present also on tergum V where transverse as on anterior terga. Underside of flagello- meres, antero-distal spot on fore femur, dorsal and anterior surfaces of fore tibia, fore tarsus, yellowish-orange. Length 6.3 mm; length of fore wing 4.6 mm; hamuli 6. Head in front view 1.33 X as wide as long, much more finely microreticulate and much more finely punctate than in female, moderately shiny. POL:OOL - 1:0.7. Clyp- eus shorter than that of female, 1.46 X as wide as long. Mesosoma much more finely microreti- culate and much more finely punctate than in female, moderately shiny. Fore leg much modified; coxa and trochanter enlarged; femur (Fig. 1) greatly swollen, postero-basally with pointed tu- bercle, its posterior surface depressed, smooth and very shiny and forming an angle with ventral surface; tibia greatly enlarged, ventrally with its swollen basal section fitting into opposing disto-ventral emargination of femur (best seen in ante- rior view); tarsomeres robust, noticeably setose. Middle and hind legs more robust than those of female; femora of both these legs swollen beneath but longitudinally grooved over distal half to accommodate tibia when opposed; tarsomeres II— IV of middle leg noticeably wider than those of hind leg. Gaster very finely microreticulate, shiny. Tergum VII (Fig. 7) baso-laterally with a pronounced rounded tubercle, apico- medially drawn out into a robust, dorsally flattened and apically rounded process raised above depressed surface on either side of it; process dorsally with a slight median longitudinal carina and laterally on each side with a smooth low carina (carried forward some distance onto the tergal disk) at angle formed by its dorsal and lateral surfaces; hind margin of tergum in lateral view forming a low smooth curve from basal tubercle to tip of apical process. Sternum II— VI depressed medially; ster- num II markedly so; III— VI progressively less so. Etymology. — The name bonaespei, a Latin noun in the genitive, refers to the Cape of Good Hope and draws attention to the provenance of the species, especially to the type locality which is within sight of Table Mountain. Material examined. — Holotype: o, SOUTH AFRICA: WESTERN CAPE: on coast 4 km north of Bloubergstrand (33.46S 18.27E), 12- 13.viii.2002 (F. W. and S. K. Gess) (on ground) [AMG]. Paratypes: SOUTH AFRICA: WEST- ERN CAPE: Donkinsbaai, 10 km S of Door- nbaai, low vegetated dunes and slacks behind beach (31.54S 18.17E), 9.ix.2005 (F. W. and S. K. Gess), 8 99 (4 99 from sand-filled Trigonephrus Volume 16, Number 2, 2007 215 Figs. 1-5. Left fore femur of male (posterior view). 1. Quartinia bonaespei, 2. Quartinia conchicola, 3. Quartinia namaquensis, 4. Quartinia vexillata, 5. Quartinia fenwrata. shells; 2 99 reared from mature larvae ex Quartinia nests ex sand-filled Trigouephrus shells, emerged in lab. first week of June 2006; 2 99 visiting white centred, pink flowers of Drosantheinuin sp., Aizoaceae: Mesembryan- thema) [AMG]; Lamberts Bay, dunes behind beach (32.05S 18.19E), 28.ix.2005 (F. W. and S. K. Gess), 1 9 (from Trigonepihrus shell) [AMG]; Lamberts Bay, sandy southern bank of lagoon (32.05S 18 19E), 28.ix.2005 (F. W. and S. K. Gess), 3 99 (visiting yellow flowers of Conicosia, Aizoaceae: Mesembryanthema) [AMG]; Elands Bay, low vegetated dunes behind beach (32.19S 18.20E), 30.ix.2005 (F. W. and S. K. Gess), 3 99, 1 J (1 9 from sand filled Trigouephrus shell; 2 99 visiting pink flowers of Drosanthemum, Aizoa- ceae: Mesembryanthema; 1 ,_J reared ex Quarti- nia nest in sand-filled Theba pisana (Mull.) shell, emerged in lab. 6.viii.2006) [AMG]; Roscherpan Nature Reserve (32.36S 18.18E), 24.iii.2001 (Feuerer & Thell), 4 99, 1 J (from shells of Trigouephrus porphyrostoma (Melvill & Pon- sonby) [Zool. Mus Berlin]; Yzerfontein (33.20S 18.10E), 15.X.2006 (D. W., G. T. and G. M. Gess), 1 9 (ex Theba pisana shell) [AMG]; S of Yzerfontein (33.22S 18.1 IE), 15.X.2006 (D. W., G. T. and G. M. Gess), 1 9 (on sand) [AMG]; Melkbosstrand (33.42S 18.26E), lO.x.2005 (F.W. and S. K. Gess), 2 99 (1 9 on sand beneath flowering Tracln/andra divaricata (Jacq.) Kunth., Asphodelaceae; 1 9 reared from mature larva ex Quartinia nest ex sand-filled Theba pisana shell) [AMG]; on coast 4 km north of Bloubergstrand (33.46S 18.27E), 12-13.viii.2002 (F. W. and S. K. Gess), 11 99, (6 99, 3 99 visiting white flowers of Tracln/andra divaricata; 1 9 visiting purplish pink flowers of Aizoaceae: Mesembryanthema; 1 9 ex nest in sand-filled Trigouephrus shell) [AMG]; same locality, 5.X.2005 (F. W. and S. K. Gess), 4 99, 1 6* (1 9 from sand filled Trigouephrus shell; 1 9 visiting white flowers of Tracln/andra divar- icata; 2 99 on sand beneath flowering Tracln/au- 216 Journal of Hymenoptera Research 6. 8. 10. 11 12. Figs. 6-13. Tergum VII of male (postero-dorsal view). 6. Quartinia australis, 7. Quartinia bonaespei, 8. Qumiinia conchkola, 9. Quartinia namaquensis, 10. Quartinia obibensis, 11. Quartinia refugicola, 12. Quartinia vcxillata, 13. Quartinia femorata. [TergumVII of Quartinia namaqua is very similar to that of Quartinia obibensis (Fig. 10)]. dm divaricata; 1 j reared ex Quartinia nest in sand-filled Theba pisana shell, emerged in lab. 8.viii.2006) [AMG]. Geographic distribution. — Known only from the supra-littoral dunes of the Atlan- tic seaboard of the Western Cape, from Donkinsbaai, circa 220 km north of Cape Town to Bloubergstrand at the northern extremity of Table Bay (the type locality). At Yzerfontein it has been found together with Q. namaqua and Q. obibensis. Floral associations. — Asphodelaceae (Tra- chyandra) and Aizoaceae: Mesem- bryanthema (including Conicosia and Dro- santhemum). Nesting. — The collection at all the listed localities of adult females from sand-filled snail shells, the discovery of an adult female at Bloubergstrand in a shell contain- ing also an open cell provisioned with a mixture of pollens including that of Trachyandra divaricata, and the rearing in the lab of adults from mature larvae extracted from cells found in shells from four of the localities, demonstrates that this species, like others occurring in sandy areas, utilizes sand-filled snail shells as a nesting niche. Shells of the indigenous desert snail, Trigonephrus species (Mol- lusca: Gasteropoda: Pulmonata: Dorcasii- Volume 16, Number 2, 2007 217 dae) are the original ones utilized and appear to be preferred; where these are in short supply, the smaller, thinner and therefore less opaque shells of the exotic Theba pisana (Mull.) (Mollusca: Gastero- poda: Pulmonata: Helicidae) are used. Quartinia conchicola Gess, new species Quartinia sp. (larger sp.) (Gess and Gess 1999, nesting) Diagnosis. — Very large (5.6-6.3 mm). Fore wing with Cula and 2m-cu complete and as thick as other veins. Both sexes with vertex behind posterior ocelli depressed, somewhat concave; with fore coxa not swollen basally nor anteriorly produced but evenly curved. Male with fore femur enlarged, excavated beneath and undulate postero-ventrally; tibia robust, markedly swollen, appreciably shorter than femur and, when opposed to femur, fitting into ventral excavation of same. Description. — Female: Black. The follow- ing are yellowish-white: short (almost medially interrupted) and laterally widen- ing transverse band on dorsum of prono- tum and minute spot at postero-dorsal angle of same; humeral streak of varying length; anterior and posterior thirds of tegula (median third clear, testaceous); medially interrupted band on lamellate margin of scutellum (specimens from Hondeklip Bay only). (The specimen from Knersvlakte lacks the humeral streak as do those from between Alexander Bay and Port Nolloth which in addition have the other markings on the thorax reduced and reddish-brown. Those from W of Wallek- raal are without thoracic markings.) The following are various shades of light reddish brown: mandible (other than base); labrum; lower aspect of pedicel and flagel- lum; posterior bands (in some specimens widened medially and usually not attain- ing lateral margins) on terga I-IV (or V); apices of all femora; most of tibia and tarsus of all legs. Venation light brown at base of wings, otherwise very dark brown. Wing membrane very slightly browned. Length 5.6-6.3 mm (average of 8 = 6.1 mm); length of fore wing 3.7-4.2 mm (average of 8 = 4.1 mm); hamuli 6. Head in front view 1.3-1.34 X as wide as long; POL:OOL = 1:0.65 (average of 5). Vertex behind posterior ocelli depressed, somewhat concave. In general facies similar to male (de- scribed below) but with legs and last tergum simple. Male: Head and mesosoma black, gaster and greater part of femora of all legs very dark brown to almost black. The following are yellowish-white: pair of small spots on frons immediately above frontoclypeal su- ture (in specimens from north of Vanrhyns- dorp only); short (almost medially inter- rupted) and laterally widening transverse band on dorsum of pronotum and minute spot at postero-dorsal angle of same; hu- meral streak of varying length; anterior and posterior thirds of tegula (median third clear, testaceous); medially interrupted band on lamellate margin of scutellum. (In a specimen from between Alexander Bay and Port Nolloth the humeral markings are absent and the other markings on the thorax are reddish-brown.) The following are various shades of light reddish brown: mandible (other than base); labrum; lower aspect of scape, pedicel and flagellum; posterior bands (slightly widened medially and laterally but not attaining lateral mar- gins) on terga I— VI; apices of all femora; most of fore tibia; middle tibia and hind tibia to variable extent and tarsus of all legs. Venation light brown at base of wings, otherwise very dark brown. Wing mem- brane very slightly browned. Length circa 5.8-6 mm; length of fore wing circa 4^4.5 mm. Head, mesosoma and terga I— VII very finely microsculptured (shagreened) but nevertheless shiny. Head in front view 1.4-1.45 X as wide as long; POL:OOL = 1:0.65. ). Vertex behind posterior ocelli depressed, somewhat con- cave. 218 Journal of Hymenoptera Research Tegula with posterior inner corner in- wardly produced. Wing venation with Cul and 2m-cu complete and as thick as other veins. Fore leg with coxa unmodified; femur (Fig. 2) enlarged, excavated beneath and undulate postero-ventrally; tibia robust, markedly swollen, appreciably shorter than femur and when opposed to femur fitting into ventral excavation of same. Middle and hind femora robust but otherwise not markedly modified. Sternum I postero-medially very slightly bi-tuberculate; sternum II somewhat raised on either side of median area. Tergum VII (Fig. 8) in posterior half with dorsal surface raised laterally and delimited by low carinae, produced apically and with a deep, narrow, slightly sub-parallel median slit. Etymology. — The name conchicola is a compound word formed from the Latin words concha - ae - the shell of a mollusc, and cola - a dweller. It serves to draw attention to the species' association, albeit not unique, with sand-filled shells of the Desert Snail, Trigonephrus, in which its nests are sheltered from prevailing winds. Material examined.— Holotype: J, SOUTH AFRICA: WESTERN CAPE: 12 km N of Vanr- hynsdorp (31.31S 18.43E), 27.ix.2005 (F. W. and S. K. Gess) (dead, ex nest in sand-filled Trigonephrus shell) [AMG]. Paratypes: SOUTH AFRICA: NORTHERN CAPE: Richtersveld Na- tional Park, 1.5 km from Helskloof Gate (28.18S 16.57E), 8.ix.l996 (F. W., S. K. and R. W. Gess), 1 9 (on white flowers of Pelargonium klinghardtense Knuth, Geraniaceae) [AMG]; Richtersveld, W of Brandkaros (28.29S 16.40E), 15.ix.1996 (F. W., S. K. and R. W. Gess), 1 S (dead) and fragments of 2 further <$<$ (ex nests in sand-filled Trigonephrus shells) [AMG]; between Alexander Bay and turnoff to Oranjemund (28.35S 16.30E) 13.ix.1996 (F. W., S. K. and R. W. Gess), 1 9 (dead, ex nest in sand-filled Trigonephrus shell) [AMG]; 24 km S of Alexander Bay on road to Port Nolloth [= 60 km N of Port Nolloth on road to Alexander Bay] (28.47S 16.38E), 27.ix.1997 (F. W. and S. K. Gess), 5 99 (4 ex nests in sand-filled Trigonephrus shells; 1 on ground) [AMG]; same locality, ll.x.2000 (F. W. and S. K. Gess), 1 9 (visiting pink flowers of Drosanthemum sp.) [AMG]; 60 km S of Alexan- der Bay on road to Port Nolloth (28.51 S 16.40E), 19.ix.1996 (F. W., S. K. and R. W. Gess), 1 J (dead, ex nest in sand-filled Trigonephrus shell) [AMG]; Hondeklip Bay (30.19S 17.17E), 12.X.1994 (F. W. and S. K. Gess), 3 99 (visiting yellow flowers of Conicosia sp., Aizoaceae: Mesembryanthema) [AMG]; W of Wallekraal (30.21S 17.26E), 8.X.1997 (F. W. and S. K. Gess), 2 99 (live) and fragments of 1 c? (ex nests in sand- filled Trigonephrus shells) [AMG]. WESTERN CAPE: Knersvlakte, 48 km N of Vanrhynsdorp (31.14S 18.32E), 20.ix.1996 (F. W., S. K. and R. W. Gess), 1 9 [AMG]; 12 km N of Vanrhynsdorp (31.31S 18.43E), 27.ix.2005 (F. W. and S. K. Gess), 2 99, 2 S3 (dead and incomplete, ex nests in sand-filled Trigonephrus shells) [AMG]; SE of Lutzville on road to Vredendal (31.36S 18.23E), 29.ix.2005 (F. W. and S. K. Gess), 399 (reared from mature larvae ex Quartinia nests ex sand- filled Trigonephrus shells, emerged in lab at the end of April 2007) [AMG]. Geographic distribution. — The species is known from South Africa from the western part of the Northern Cape, mainly along the seaboard from the Orange River south- wards, and from the northwestern Western Cape where it extends inland to a distance of about 50 km. In occurs variously togeth- er with Q. namaqua, Q. namaquensis, Q. obibensis, Q. rufigicola and Q. vexillata. Floral associations. — Aizoaceae: Mesem- bryanthema (Conicosia, Drosanthemum), Geraniaceae (Pelargonium). Nesting. — Throughout its presently known distributional area found nesting in sand-filled shells of the desert snail Trigonephrus sp. (Mollusca: Gasteropoda: Pulmonata: Dorcasiidae). For further de- tails see Gess and Gess (1999). Discussion. — Q. conchicola and Q. vexillata appear to be closely allied and at least in the north-western Richtersveld (S of Alex- ander Bay) overlap in their distribution. Whereas the males are readily distinguish- able on the basis of secondary sexual characters - notably the differently modi- fied fore legs - the females are deceptively similar and at first sight are very difficult to separate. They may, however, be distin- Volume 16, Number 2, 2007 219 guished by characters which they share (progressively darkened) of all legs; in two with their respective males: Q. conchicola by of the southern specimens apex of femur, the depressed, somewhat concave vertex base and apex of tibia and base of first and by the unmodified fore coxae and Q. tarsomere only. Wings slightly darkened; vexillata by the evenly convex vertex and veins brown. by the basally swollen and anteriorly pro- Length 6.2 mm; length of fore wing duced fore coxae. 3.9 mm; hamuli 6. Head in front view 1.3 X as wide as long; Quartinia namaqua Gess, new species clypeus 1.1 x as wide as long (length measured to bottom of emargination); Diagnosis.-Very large (5.8-6.2 mm). pOL:OOL 1:09 Q dosel Fore wing with Cula and 2m-cu complete microsculptured/ with bareiy discernable and as thick as other veins. Tegula with shaUow tureS/ dull; frons and vertex posterior inner corner markedly inwardly similarly micr0sculptured but somewhat produced, reddish brown. Both sexes with mQre obviously punctured, moderately head and thorax black (except, in most shiny; mesosoma micr0sculptured with specimens, a small reddish-brown marking obvious shallow pictures; interstices of medially on anterior margin of pronotum); puncture width or less; parapsidal furrows gaster black with a variable number of very obvious; gaster finely and closely reddish brown posterior bands which do punctured shinv not attain sides of terga. Parapsidal fur- Mak. Black The foUowing are reddish- rows very noticeable. Male with clypeus brown: mandibles distally; scape apically, evenly convex (not medially depressed), pedical/ upper and iower side 0f flagello- closely and finely sculptured; with ster- meres (except dista] part 0f dub); trans- num I postero-medially raised into a pro- verse marking on anterior margin of nounced tubercle; tubercle viewed from pronotum; tegula; posterior band not at- behind with widely rounded (almost sub- taining sides on tergum I and mere in- truncate) apex, viewed from the side dication of band on tergum II; apex of sloping steeply anteriorly and falling steep- femur/ most of tibia, tarsomeres (progres- ly posteriorly to hind margin of sternum. siveiy darkened) of all legs. Tergum VII with distinct dorsal and lateral Length 5.8 mm; length of fore wing surfaces; apex drawn out into a pair of 35 mm; hamuli 6. parallel processes flanking narrow and slit- Head in front view 1.4 x as wide as long; like emargination; emargination produced clypeUs 1.1 X as wide as long (length anteriorly as a median impression. measured to bottom of emargination); Description.— Female: Black. The follow- POL:OOL = 1:0.8. Clypeus evenly convex, ing are reddish-brown: mandibles distally; closely microsculptured, with barely dis- underside of pedicel and flagellum; in two cernable shallow punctures, only moder- of the northern specimens a mere indica- ately shiny; frons and vertex similarly tion of a transverse marking on anterior microsculptured but somewhat more obvi- margin of pronotum; tegula; posterior ously punctured, moderately shiny; meso- bands not attaining sides on terga 1-IV soma microsculptured with obvious shal- (in southern specimens on terga I — III only); low punctures; interstices of puncture that on I of even width and covering about width or less; parapsidal furrows very half of tergum; those of terga II and III obvious; gaster finely and closely punc- narrower but medially expanded; that of tured, shiny. Sternum I postero-medially IV short or (in two specimens) barely raised into a pronounced tubercle; tubercle indicated; in northern specimens apex of viewed from behind with widely rounded femur, entire or most of tibia, tarsomeres (almost subtruncate) apex, viewed from 220 Journal of Hymenoptera Research the side sloping steeply anteriorly and falling steeply posteriorly to hind margin of sternum. Tergum VII with distinct dorsal and lateral surfaces; apex drawn out into a pair of parallel processes flank- ing narrow and slit-like emargination; emargination produced anteriorly as a me- dian impression. Etymology.— The name, namaqua, a noun in apposition to the generic name, is derived from the Namaqua people of Namaqualand and refers to the provenance of the species. Material examined. ■ Holotype, 6\ SOUTH AFRICA: NORTHERN CAPE: Inland of Hon- deklip Bay (30.19S 17.17E), 25.ix.2005 (F W and S K Gess) (ex nest in sand-filled Trigonephrus shell) [AMG]. Paratypes: SOUTH AFRICA: NORTHERN CAPE: same data as holotype, 1 9 [AMG]; between Hondeklip Bay and Wallek- raal (30.22S 17.28E), 25.ix.2005 (F W and S K Gess), 1 9 (ex nest in sand-filled Trigonephrus shell) [AMG]. WESTERN CAPE: SE of Lutzville on road to Vredendal (31.36S 18.23E), 29.ix.2005 (F W Gess and S K Gess), 1 9 (visiting yellow flowers of Conicosia spv Aizoaceae: Mesem- bryanthema) [AMG]; Yzerfontein (33.20S 18.10E), 15.X.2006 (D. W., G. T. and G. M. Gess), 3 99 (ex Theba pisana shells) [AMG]. Geographic distribution. — The species is known from South Africa from the south- western Northern Cape, from the north- western Western Cape, and from Yzerfon- tein in the southwestern Western Cape, and therefore will probably be found to occur all along the coastal sandveld be- tween the above areas. In occurs variously together with Q. bonaespei, conchicola, na- maquensis and obibensis. Floral associations. — Aizoaceae: Mesem- bryanthema (Conicosia). Nesting. — At two localities found nesting in sand-filled shells of the desert snail Trigonephrus sp. (Mollusca: Gasteropoda: Pulmonata: Dorcasiidae) and at another obtained from shells of the exotic Tlieba pisana (Mull.) (Mollusca: Gasteropoda: Pulmonata: Helicidae). Discussion. — Q. namaqua is superficially very similar to Q. obibensis, most notably in the male in the possession of a raised tubercle postero-medially on sternum I. It may be distinguished in both sexes by the more distinct and somewhat less close puncturation of the mesoscutum and scu- tellum, by the broader and much more noticeable parapsidal furrows, and by the interocellar distance only slightly exceed- ing the ocellar-ocular distance [POL:OOL = 1:0.9 (9) and 1:0.8 (S) as against 1:0.7 (both sexes)]. The male may be distin- guished by the evenly convex, closely microsculptured and only moderately shiny clypeus in contrast to the medially depressed, non-microsculptured but sparsely punctured and shiny clypeus of Q. obibensis. In colour pattern the species differs in that the reddish-brown markings are reduced, most notably in that the posterior bands on the gaster do not attain the lateral margins of the terga. Quartinia namaquensis Gess, new species Diagnosis. — Very large (5.8-6.0 mm). Fore wing with Cula and 2m-cu complete and as thick as other veins. Tegula short, laterally rounded, with posterior inner corner inwardly produced. Male black with white-marked labrum, clypeus, frons, pronotum, tegula, scutellar lamella, and terga I-VI. Fore leg greatly enlarged and modified; middle and hind legs somewhat modified. Tergum VII drawn out apico- medially into a robust, pointed, dorsally flattened and apically narrowly rounded process. Description. — Male: Black. The following are white: labrum; disc of clypeus; parao- cular streak from mandibular insertion to level of top of antennal socket (specimen from Wallekraal only); supra-clypeal mark- ing (more or less quadrate and bilobed dorsally in specimens from Leliefontein but in specimen from Wallekraal expanded on each side with lobe directed laterally towards ocular sinus and another directed dorsally); underside of scape, pedical and proximal flagellomeres; continuous anteri- Volume 16, Number 2, 2007 221 or band on pronotum (narrowly and pointedly extended a little along dorso- lateral margin and broadly continuous onto humerus and beyond) and minute spot on postero-dorsal angle of same; small spot at top of mesopleuron (specimen from Wallekraal only); tegula (except for median testaceaous area); scutellar lamella (other than medially); lower two thirds of meta- notum (specimen from Wallekraal only); minute dots dorsally on propodeum (one specimen from Leliefontain only) or small streak unilaterally on angle of propodeum (specimen from Wallekraal only); narrow posterior bands, almost reaching sides, on terga I-VI. The following are light reddish yellow: mandible (except base and apex): labrum (if not white); posterior bands, slightly medially expanded, on sterna; underside of trochanter of all legs; entire anterior surface of fore femur as well as posterior surface of basal lamelliform angle of same; underside of basal half of mid femur (most specimens); apices of femora and entire tibiae, tarsi and claws of all legs. Wing membrane sub-hyaline; veins brown. Length 5.8-6.0 mm.; length of fore wing 3.8-3.9 mm.; hamuli circa 6. Head in front view 1.33 X as wide as long, microreticulate, moderately shiny, with shallow punctures (small and close on frons, slightly larger and more widely spaced on vertex). POL:OOL = 1:0.59. Clypeus 1.5 X as wide as long in midline; anterior margin widely and shallowly emarginate. Mesosoma microreticulate, moderately shiny, with punctures larger than those on head (moderately well separated on pronotum, mesoscutum and scutellum). Tegula short, laterally rounded, with posterior inner corner inwardly produced. Fore leg much modified; coxa and trochanter enlarged; femur (Fig. 3) greatly swollen, its posterior surface in proximal half markedly concavely excavate, smooth and very shiny, its baso-ventral region angulate and sublamellate; first tarsomere swollen, excavate and setose below; second tarsomere in posterior view curved, wide at base but otherwise narrow, with long , backwardly curved setae; middle and hind legs beneath with trochanters flattened and with femora angulate, flattened in proxi- mal half and longitudinally grooved in distal half. Metasoma moderately shiny, with punc- tures finer than those on head. Tergum VII (Fig. 9) drawn out apico-medially into a robust, pointed, dorsally flattened and apically narrowly rounded process. Female: Unknown, none of the specimens of several species from the relevant local- ities being assignable with any degree of confidence to this species. Etymology. — The name, namaquensis, an adjective, is derived from the Namaqua people of Namaqualand and refers to the provenance of the species. Material examined. — Holotype, 3, SOUTH AFRICA: NORTHERN CAPE: Leliefontein (30.23S 18.16E), 31.vii.2003 (C. Mayer), 1 3 (yellow trap) [AMG]. Paratypes: SOUTH AFRICA: NORTHERN CAPE: same data as holotype but date 15.ix.2003, 1 3 [AMG]; same data as holotype but date 22.viii.2004, and trap white, 1 3 [AMGJ; W of Wallekraal (30.21S 17.26E), 8.X.1997 (F. W. and S. K. Gess), 1 3 (ex nest in sand-filled Trigonepmrus shell) [AMG]. Geographic distribution. — Known only from two localities in Namaqualand, one in the coastal sandveld, the other in the Kamiesberg. In the former locality it occurs together with Q. conchicola and Q. namaqua. Floral associations. — Unknown. Nesting. — One specimen, freshly eclosed and with wings not yet fully hardened, was extracted from a cell of a nest in a sand- filled shell of the desert snail Trigonephrus sp. (Mollusca: Gasteropoda: Pulmonata: Dorcasiidae). Quartinia obibensis Gess, new species Diagnosis. — Large to very large (5.2- 5.7 mm). Fore wing with Cula and 2m-cu complete and as thick as the other veins. Tegula with posterior inner corner markedly inwardly produced, reddish brown. Both 222 Journal of Hymenoptera Research sexes with head and thorax predominantly Gaster finely microreticulate but shiny; black with limited reddish-brown markings; punctures finer than those on mesosoma, gaster black with well developed reddish- becoming progressively finer distally. brown posterior bands attaining or almost Tegula with posterior inner corner mark- attaining side of terga. Male with clypeus edly inwardly produced, medially depressed, sparsely punctured and Male: Black. The reddish-brown mark- shiny; with sternum I postero-medially ings as in the female, with in addition: raised into a small tubercle; tubercle viewed labrum (to varying degree); in some speci- from behind transversely subtriangular with mens small antero-lateral spots (occasion- a narrowly rounded apex, viewed from the ally joined ) on clypeus. side sloping gradually anteriorly and falling Length 5.2 mm; length of fore wing steeply posteriorly to hind margin of ster- 3.5 mm. num. Tergum VII with distinct dorsal and POL:OOL = 1:0.7 lateral surfaces; apex drawn out into a pair of Clypeus medially depressed, non-micro- parallel processes flanking narrow and slit- sculptured but sparsely punctured and like emargination; emargination produced shiny. Sternum I postero-medially raised anteriorly as a median impression. into a small tubercle; tubercle, viewed from Description. — Female: Black. The follow- behind, transversely subtriangular with ing are reddish-brown: mandibles (except a narrowly rounded apex, anteriorly grad- base); underside of pedicel and flagello- ually sloping, posteriorly falling steeply to meres; anterior margin of pronotum and hind margin of sternum. Tergum VII postero-dorsal angle of same; tegula; cres- (Fig. 10) with distinct dorsal and lateral cent (in some specimens broken up into surfaces; apex drawn out into a pair of spots) posteriorly and laterally on disk of parallel processes flanking narrow and slit- scutellum; scutellar lamella; in some speci- like emargination; emargination produced mens lower half of metanotum; posterior anteriorly as a median impression, bands attaining or almost attaining sides Etymology. — The name, obibensis, an ad- on terga I-V; that on I of even width and jective, is derived from the Obib Mountains covering about half of tergum; those of in the Sperrgebiet of south-western Nami- terga I I-V progressively narrower, undu- bia, the site from which the largest number late, expanded medially and laterally and of specimens was obtained. attaining or almost attaining sides of terga); • i * «. ™ t\7 „„,w n( c° I Material examined.— Holotype: $, NAMIBIA: apical spot on tergum IV apex of femur, „ _, , yr "' „,„„_. *\ *\ , l i t.\ £ ii Sperrgebiet, Obib camp site 28.00S 16.39E ), entire tibia, tarsomeres (except last) of all ,\ . ®n ' TA7 * v _ v . . , , . 14.ix.2003 (F. W. and S. K. Gess) (ex nest in legs. Last tarsomere and claws brown. sand.filled Trigonephrus shell/ emerged in the Wings slightly darkened; veins brown. lab 15_22.x.2003) [AMG]. Paratypes: NAMIBIA: Length 5.2-5.7 mm (average of 4:5.4 Sperrgebiet, W of Klinghardtberge (27.17S mm); length of fore wing 3.6-3.7 mm 15.36E), 20.ix.2003 (F. W. and S. K. Gess), 4 99 (average of 4:3.7 mm); hamuli 5. (ex nests in sand-filled Trigonephrus shells) Head in front view 1.35 X as wide as long, [AMG]; Sperrgebiet, Klinghardtberge (27.19S microreticulate, matt, with inconspicuous, 15.46E), 10.ix.2005 (F. W. and S. K. Gess), 1 9 very shallow, fine punctures. POL:OOL = (reared ex Quartinia nest in sand-filled Trigone- 1:0.7. Clypeus 1.33 X as wide as long (length Phrus shell) tAMGl; Sperrgebiet, Klinghardt- _ J\. u „ c • « 1 o n/ berge (27.30S, 15.44E), 10.ix.2005 (F. W. and S. measured to bottom of emargination; 1.2 X T, % \ « ' ' « . . ., j 1 1 r t , 1 K. Gess), 1 9, 1(5 (reared ex Quart una nests in it measured to level of antero-lateral an- , c.u , ^ . , , ,, . taa*/-! a sand-filled Trigonephrus shells) [AMG]; Aus - gles); anterior margin smooth, shiny, shal- Rosh pinah (27 44S 16 42E) 25 ix 2003 (F. w. lowly and evenly emarginate. and s K Gess)^ 1 s (visiting white centered, Mesosoma microreticulate with close, purplish-pink rayed Drosanthemum spv Aizoa- shallow, fine punctures, slightly shiny, ceae: Mesembrianthema) [AMG; ]; Sperrgebiet, Volume 16, Number 2, 2007 223 between Aurusberg and Scorpion Mine (27.45S 16.32E), 15.ix.2003 (F. W. and S. K. Gess), 1 9, 2 SS (ex nests in sand-filled Trigonephrus shells) [AMG]; Sperrgebiet, Scorpion Mine (27.49S 16.35E), 15.ix.2003 (F. W. and S,. K. Gess), 4 99, 1 j (ex nests in sand-filled Trigonephrus shells) [AMG]; Sperrgebiet, Obib camp site (28.00S 16.39E), 14.ix.2003 (F. W. and S. K. Gess), 37 99, 13 $<$ (ex nests in sand-filled Trigonephrus shells; 26 99, 2 J J emerged in the lab. 15- 22.X.2003; 7 99, 2 S3 emerged in lab. at a later date ) [AMG]; 12.8 km S Rosh Pinah (28.03S 16.51E) ll.ix.1996 (F. W., S. K. and R. W. Gess), 3 99/ 3 0*6* (ex nests in sand-filled Trigonephrus shells; 2 99, 3 $$ emerged in lab.) [AMG]. SOUTH AFRICA: NORTHERN CAPE: Richters- veld, W of Brandkaros (28.29S 16.40 E), 15.ix.1996 (F. W., S. K. and R. W. Gess), 3 99, 1 6* (ex nests in sand-filled Trigonephrus shells) [AMG]. WESTERN CAPE: Yzerfontein (33.20S 18.10E), 15.X.2006 (D. W., G. T. and G. M. Gess), 2 $$ (ex Theba pisana shells) [AMG]. Geographic distribution. — Q. obibensis is known from Namibia, from a limited area in the southern half of the Desert and Succulent Steppe (Winter Rainfall Area) of Giess (1971), from South Africa from a nearby locality in the Richtersveld and from a coastal site in the Western Cape. In the north of its range it occurs together with Q. conchicola, Q. rufigicola and Q. vexillata and in the south with Q. bonaespei and Q. namaqua. Floral associations. — Aizoaceae: Mesem- bryanthema (Drosanthemum). Nesting. — Throughout its presently known distributional area found nesting most commonly in sand-filled shells of the desert snail Trigonephrus sp. (Mollusca: Gasteropoda: Pulmonata: Dorcasiidae). At one coastal locality in the Western Cape obtained from shells of the exotic Theba pisana (Mull.) (Mollusca: Gasteropoda: Pulmonata: Helicidae). For further details see Gess and Gess (1999). Discussion. — See under Q. namaqua. Quartinia refugicola Gess, new species Quartinia sp. (smaller sp.) (Gess and Gess 1999, nesting; Greathead 1999, 2006, bombyliid parasite). Diagnosis. — Medium sized to large (4.1- 5.2 mm long). Fore wing with Cula and Im-cu complete and as thick as other veins. Tegula with posterior inner corner inward- ly produced. Posterior bands on terga reaching lateral margins. Description. — Female: Black. The follow- ing are yellow or yellow merging into brownish yellow: underside of flagello- meres; short, narrow, transverse band (in some specimens reduced to pair of small marks, in others totally absent) medially on pronotum and in some specimens a minute dot on postero-dorsal angle of same; tegula (except for testaceous medial spot); nar- row, medially interrupted, lamellate mar- gin of scutellum; in some specimens median part of metanotum; narrow poste- rior bands reaching lateral margins on terga I-V (that of tergum I widest, others progressively narrower); in some speci- mens a diffuse posterior band on sternum II; extreme apex of femur, entire tibia (except for elongate dark mark on posterior surface) and tarsomeres of all legs (except in some specimens brown terminal tar- someres of middle and hind legs). Mandi- ble with distal half bright ferruginous; labrum brown. Wings subhyaline; veins brown. Length 4.5-5.2 mm (average of 5:4.8 mm); length of fore wing 3.0-3.4 mm (average of 5:3.2 mm); hamuli 5-6. Head in front view 1.3 X as wide as long; clypeus 1.5 X as wide as long (length measured to bottom of emargination); POL:OOL = 1:0.6. Clypeus very closely microsculptured, with barely discernable shallow punctures; Irons and vertex simi- larly microsculptured but more obviously punctured (especially in region of ocelli); mesosoma microsculptured with obvious shallow punctures slightly larger than those on vertex and with interstices of puncture width or less; gaster closely and finely punctured. Male: Black. Pale markings as in female but with the addition of: in some speci- mens sub-basal spot on mandible between 224 Journal of Hymenoptera Research black base and ferruginous distal half; in some specimens part of the clypeus (ranging in extent from pair of antero- lateral spots, to uninterrupted anterior margin, to most of disc with exception of region below antennal sockets); in all specimens narrow posterior band on ter- gum VI and in most specimens apices of tergum VII. Length 4.1-4.3 mm (average of 5:4.2 mm); length of fore wing 2.8-2.9 mm (average of 5:2.8 mm); hamuli 4. Head in front view 1.37 X as wide as long; clypeus convex, 1.5 X as wide as long; POL:OOL = 1:0.6. Microsculpture and puncturation as in female. Tergum VII (Fig. 11) dorsally slightly depressed (flat- tened) and its apical margin with a narrow V-shaped median emargination flanked by narrowly rounded projections. Etymology. — The name refugicola is a com- pound word formed from the Latin words refugium - ii (n) - a place of refuge, and cola - a dweller. It serves to draw attention to the species' association with sand-filled cavities in which its nests are sheltered from prevailing winds. Material examined. — Holotype: 6\ NAMIBIA: 12.8 km S of Rosh Pinah (28.03 S 16.51E), ll.ix.1996 (F. W., S. K. and R. W. Gess) (on ground) [AMG]. Paratypes: NAMIBIA: Aus (26.39S 16.15E), 25.viii.2002 (F. W. and S. K. Gess), 1 9 (visiting yellow flowers of Leysera tenella DC, Asteraceae) [AMG]; Sperrgebiet, Kaukausib Spring - Grillental (26.58S 15.31E), 5.ix.2002 (F. W. and S. K. Gess), 4 99 (visiting white flowers of Zygaphyllum clavatum Schltr. & Diels, Zygophyllaceae) [AMG]; Sperrgebiet, near Grillental (26.59S 15.23E), 5.ix.2002 (F. W. and S. K.Gess), 1 6* (visiting yellow flowers of Foveolina albida (DC.) Kallersjo, Asteraceae) [AMG]; Sperrgebiet, Grillental (27.00S 15.21E), 8.ix.2005 (F. W. and S. K. Gess), 3 99, 1 $ (visiting white flowers of Zygophyllum sp.) [AMG]; Sperrgebiet, Klinghardtberge, Tsabiams Camp (27.10 S 15.42E), 4.ix.2002 (F. W. and S. K. Gess), 1 9 (visiting yellow flowers of Dimor- photheca polyptera DC, Asteraceae) [AMG]; Sperrgebiet, Klinghardtberge (27.14S 15.43E), l-3.ix.2002 (F. W. and S. K. Gess), 5 99 (1 9 visiting flowers of ? Cotula sp., Asteraceae; 1 9 visiting yellow flowers of Pteronia sp., Aster- aceae; 3 99 visiting yellow flowers of Zygophyl- lum simplex L., Zygophyllaceae) [AMG]; Sperr- gebiet, Klinghardtberge (27.14S 15.44E), 2.ix.2002 (F. W. and S. K. Gess), 6 99, 1 J (3 99 visiting yellow flowers of Pteronia sp., Aster- aceae; 3 99 visiting apricot coloured flowers of Phyllobolus occulatus (N.E.Br.) Gerbaulet, Aizoa- ceae: Mesembryanthema; 1 6" visiting pink flowers of Sarcocaulon sp., Geraniaceae) [AMG]; Sperrgebiet, W of Klinghardtberge (27.17S 15.36E), 20.ix.2003 (F. W. and S. K. Gess), 5 99, 3 S3 (ex nests in sand-filled Trigonephrus shells) [AMG]; Sperrgebiet, Klinghardtberge (27.18S 15.54E), 2.ix.2002) (F. W. and S. K. Gess), 1 6" [AMG]; Sperrgebiet, Klinghardtberge (27.19S 15.46E), 10.ix.2005 (F. W. and S. K. Gess), 2 99, 1 S (reared from larvae ex Quartinia nests ex sand-filled Trigonephrus shells) [AMG]; Sperrge- biet, Klinghardtberge, Nomitsas (27.27S 15.52E), 31.viii.20Q2 (F. W. and S. K. Gess), 1 9, 1 o (ex sand-filled Trigonephrus shells) [AMG]; Sperrge- biet, Uguchab River, NW of Aurus Mountains (27.31S 16.12E), 17.ix.2003 (F. W. and S. K. Gess), 31 99, 20 (J6* (17 99, 10 26* ex sand-filled Trigonephrus shells; 14 99, 10 $<$ ex nests in sand-filled Trigonephrus shells) [AMG]; 12.8 km S of Rosh Pinah (28.03 S 16.51E), ll.ix.1996 (F. W., S. K. and R. W. Gess), 29 99, 32 6*6" (21 99, 26 SS on ground; 1 9 visiting yellow flowers of Cotula sp., Asteraceae; 1 9 on blue rayed Filicia sp., Asteraceae; 1 9, 1 j visiting yellow flowers of Hirpicium sp., Asteraceae; 1 9, 1 6* visiting yellow flowers of Osteospermum sp., Asteraceae; 1 9 visiting yellow flowers of Grielum sp., Neurada- ceae; 3 99, 4 ^J ex nests in sand-filled Trigone- phrus shells; 2 SS ex sand-filled cavities in calcrete) [AMG]; Sperrgebiet, W of Obib Moun- tains (28.08S 16.42E), 15.ix.2003 (F. W. and S. K. Gess), 1 9 (ex nest in sand-filled Trigonephrus shell) [AMG]; E of Oranjemund, 28 km from checkpoint on road to Sendelingsdrif (28.26S 16.42E), 25.ix.1997 (F. W. and S. K. Gess), 1n-cu complete and as thick as other veins. Both sexes with vertex behind posterior ocelli evenly convex; with fore coxa swollen basally and anteriorly produced, very markedly so in male, less so in female where swelling, however, forms a rounded right angle. Male with fore femur greatly enlarged, robust, proximally produced ventrally to form a sturdy, subquadrate flange, distally markedly downcurved; flange with its posteriorly facing surface markedly concave with pronounced distal angles and its anteriorly facing surface convex with a pronounced submedian distal tubercle; tibia robust with dense setae on lower surface. Description. — Female: In general facies similar to male (described below) but with legs and last tergum simple. Head without any pale markings. Specimens from between Alexander Bay and Port Nolloth have the mesosoma and gaster with both yellowish white and reddish brown markings very similar to those of males from the same population; speci- mens from SSE of Grillental and from Obib have the markings on the mesosoma tending to reddish brown. Microsculpture (shagreening) of head and mesosoma somewhat coarser and these parts, in particular mesoscutum, semi-matt rather than shiny. Length 5.8-6.5 mm (average of 7 = 6.1 mm); length of fore wing 3.9-4.4 mm (average of 7 = 4.1 mm); hamuli 6. Head broad, 1.33 X as wide as long; POL:OOL = 1:0.5. Vertex behind posterior ocelli evenly convex. Fore coxa enlarged, basally markedly and roundly anteriorly produced. Male: Head and mesosoma black, gaster and greater part of femur of all legs very dark brown to almost black. The following are yellowish white: lower aspect of scape (excluding radicle) and pedicel; labrum (in one specimen testaceous); clypeus (other than for irregular area below antennal socket); a small transverse spot situated on either side of midline of frons immedi- ately above frontoclypeal suture (in one specimen only); narrow paraocular streak from mandibular insertion to level of top of antennal socket; short (in one specimen almost medially interrupted) and laterally widening transverse band on dorsum of pronotum and minute spot at postero- dorsal angle of same; humeral streak of varying length; anterior and posterior thirds of tegula (median third clear, testa- ceous); medially interrupted band on la- mellate margin of scutellum; distal portion of flange on fore femur. The following are various shades of light reddish-brown: mandible (other than base); flagellomeres (other than for dark suffusion on upper 226 Journal of Hymenoptera Research surface); posterior bands (slightly widened medially, narrowed laterally, and not quite attaining lateral margins of terga) on terga I-VI; streak on anteriorly protruding por- tion of fore coxa (in one specimen) and lower surface of middle and hind coxae; all trochanters; distal portion (up to almost half the length) of anterior aspect of fore femur; basal flattened lower surface of middle femur; apex of middle and hind femora; tibia and tarsus of all legs. Venation light brown at base of wings, otherwise very dark brown. Wing membrane very slightly browned, a little darker on fore wing in and beyond marginal cell. Length 5.2-6.4 mm; length of fore wing 3.6-3.8 mm; hamuli 6. Head, mesosoma and terga I-VI very finely microsculptured (shagreened) but nevertheless shiny, with moderately sized punctures; punctures on head and terga somewhat shallow and undefined with interspaces generally less than puncture diameter, those on mesosoma deeper and well defined with interspaces at least on mesoscutum often greater than puncture diameter. Tergum VII without microsculp- ture; punctures more pronounced than those on other terga, irregularly spaced, some separated by wide interspaces and others coalescing. Setation on head and particularly on body sparse and short throughout, more noticeable on tibiae and tarsi and strikingly developed on underside of front tibia where dense and long. Head broad, 1.45 X as wide as long; POL:OOL = 1:0.5. Vertex behind posterior ocelli evenly convex; Tegula with posterior inner corner in- wardly produced. Wing venation with Cula and 2m-cu complete and as thick as other veins. Fore leg uniquely and greatly modified; coxa enlarged, basally markedly and roundly anteriorly produced; femur (Fig. 4) greatly enlarged, robust, proximal- ly produced ventrally to form a sturdy, subquadrate flange, distally markedly downcurved; flange with its posteriorly facing surface markedly concave with pro- nounced distal angles and its anteriorly facing surface convex with a pronounced submedian distal tubercle; tibia robust with dense setae on lower surface. Middle and hind femora robust, mark- edly angled below and with lower surface both proximal and distal to angle distinctly flattened (more so on middle than on hind femur). Tergum VII (Fig. 12) in posterior half with dorsal surface raised laterally and slightly concave medially, produced api- cally and with a deep, narrow, slightly sub- parallel-sided median slit. Etymology. — The name vexillata is formed from the Latin noun vexillum meaning a flag or standard. It refers to the conspic- uously modified front femur of the male which may possibly have a communicatory role in courtship behaviour. Material examined.— Holotype: & SOUTH AFRICA: NORTHERN CAPE: 23 km S of Alexander Bay (28.46S 16.37E), 11.x. 2000 (F. W and S. K. Gess) (on ground) [AMG]. Para- types: NAMIBIA: Sperrgebiet, main north/ south road, 47 km SSE of Grillental (27.23S 15.32E), 6.ix.2002 (F. W. and S. K. Gess), 2 99 (on ground next to Drosanthemum sp., Aizoaceae: Mesembryanthema) [AMG]; Sperrgebiet, Obib camp site (28.00S 16.39E), 14.ix.2003 (F. W. and S. K. Gess), 9 99 (7 99 visiting yellow flowers of Cephalophyllum sp., Aizoaceae: Mesem- bryanthema; 2 99 visiting yellow flowers of Othoiuia cylindrica (Lam.) DC, Asteraceae) [AMG]. SOUTH AFRICA: NORTHERN CAPE: 23 km S of Alexander Bay (28.46S 16.37E), 11.x. 2000 (F. W and S. K. Gess), 5 99 (4 99 visiting pink flowers of Drosanthemum sp., Aizoaceae: Mesembryanthema; 1 9 visiting purple-centred white flowers, Aizoaceae: Me- sembryanthema) [AMG]; 60 km N of Port Nolloth (28.47S 16.38E), 27. ix. 1997) F. W. and S. K. Gess), 6 99, 1 $ (2 99, £ visiting pale pink flowers of Drosanthemum sp.; 4 99 on ground) [AMG]. Geographic distribution. — The species is known from Namibia from the Desert Volume 16, Number 2, 2007 227 and Succulent Steppe (Winter rainfall area) Floral associations. — Aizoaceae: Mesem- of Giess (1971) and from South Africa from bryanthema {Cephalophyllum, Drosanthe- the adjoining northern Strandveld of the mum); Asteraceae (Othonna). West Coast of Acocks (1953). It occurs Nesting. — Unknown; probably utilizing variously together with Q. conchicola, Q. sand-filled snail shells as a nesting niche. obibcnsis and Q. refugicola. Discussion. — See under Q. conchicola. KEY TO SPECIES NESTING IN SAND-FILLED SNAIL SHELLS OR (VEXILLATA) PRESUMED TO DO SO Males 4. 5. 6. Fore legs not modified 2 Fore legs markedly modified 5 Sternum I not modified 3 Sternum I posteriorly raised into a tubercle 4 Large (5.0-5.4 mm long); black with white markings; tegula with white anterior and posterior markings contrasting markedly with dark brown to black median part; pale posterior bands on terga not extending onto sides; clypeus and labrum white australis Gess n. sp. Medium (4.1-4.3 mm long); black with yellow to brownish-yellow markings; tegula with pale anterior and posterior markings not contrasting markedly with testaceous median part; pale posterior bands on terga reaching lateral margins refugicola Gess n. sp. Clypeus convex namaqua Gess n. sp. Clypeus depressed to concave obibensis Gess n. sp. Posterior bands on terga white; not contrasting in colour with markings on head and mesosoma; fore femur (Fig. 3) greatly swollen, its posterior surface in proximal half markedly concavely excavate, smooth and very shiny, its baso-ventral region angulate and sublamellate namaquensis Gess n. sp. Posterior bands on terga reddish-brown or bright reddish-orange, in most specimens contrasting in colour with markings on head and mesosoma; fore femur differently formed 6 Tibiae and tarsi of all legs predominantly black; fore femur (Fig. 1) greatly swollen, postero-basally with a pointed tubercle, its posterior surface depressed, smooth and very shiny and forming an angle with ventral surface bonaespei Gess sp. n. Tibiae and tarsi of all legs predominantly light reddish-brown 7 Vertex behind posterior ocelli evenly convex; fore coxa swollen basally and markedly anteriorly produced; fore femur (Fig. 4) greatly enlarged, robust, proximally produced ventrally to form a sturdy subquadrate flange, distally markedly downcurved; flange with its posterior facing surface markedly concave with pronounced distal angles and its anterior facing surface convex with a pronounced submedian distal tubercle vexillata Gess n. sp. Vertex behind posterior ocelli depressed, somewhat concave; fore coxae unmodified; fore femur (Fig. 2) enlarged, excavated beneath and undulate postero-ventmlly conchicola Gess n. sp. Females Species not included: namaquensis Gess n. sp. (9 not known) 1. Markings on mesosoma and gaster generally concolorous 2 228 Journal of Hymenoptera Research Markings on mesosoma and gaster not of same colour; posterior bands on terga reddish-brown or bright reddish-orange, generally contrasting with pale markings on mesosoma 5 2. Black with white markings; tegula with anterior and posterior markings contrasting markedly with dark brown to black median part; posterior bands on terga not extending onto sides anstralis Gess sp. n. Black with yellow, brownish-yellow or reddish-brown markings; tegula with anterior and posterior markings not contrasting markedly with median part; posterior bands on terga variously developed 3 3. Posterior bands on terga reaching lateral margins; scutellar disk black and scutellar lamella yellow to brownish yellow refugicola Gess sp.n. Without this combination of characters 4 4. Mesoscutum and scutellum with interstices between punctures not obviously microreticulate (shagreened); scutellar disk and scutellar lamella black; scutellar lamella at most slightly flattened postero-medially; metanotum not transversly impressed, black throughout namaqua Gess n. sp. Mesoscutum and scutellum with interstices between punctures very obviously microreticulate (shagreened); scutellar disk laterally and medially with reddish- brown markings and scutellar lamella of same colour; scutellar lamella slightly emarginate postero-medially; metanotum transversely impressed with lower section reddish brown and contrasting with almost black upper section obibensis Gess n. sp. 5. Tibiae and tarsi of all legs predominantly black bonaespei Gess sp. n. Tibiae and tarsi of all legs predominantly light reddish-brown 6 6. Vertex behind posterior ocelli evenly convex; fore coxae swollen basally and markedly anteriorly produced vcxillata Gess n. sp. Vertex behind posterior ocelli depressed, somewhat concave; fore coxae unmodi- fied conchicola Gess n. sp. B) Other species Quartinia femorata Gess, new species Diagnosis. — Very large to gigantic (5.8- 6.7 mm long). Fore wing with Cula and 2m-cu complete and as thick as the other veins. Tegula with posterior inner corner inwardly produced. Both sexes predomi- nantly yellow. Male fore femur robust, notched ventrally in basal third and with a distally directed, apically rounded, la- mellate process. Description. — Female: Predominantly yel- low. Black greatly reduced leaving only: occiput; irregular median band on vertex (posteriorly wide along occipital carina but anteriorly narrowed and closely encom- passing ocelli) and on frons (on upper half of similar width to part encompassing ocelli but on lower half trifid with middle arm and outcurved lateral arms reaching clypeal suture and antennal sockets re- spectively); propleuron (in greater part) and presternum; median and parapsidal bands on mesoscutum (median band wide at anterior margin, narrowing posteriorly; parapsidal bands not reaching anterior margin and of even width throughout); small antero-median mark on scutellum; anterior half of propodeal dorsum and small spot on each side at bottom of propodeal declivity; transverse marking (either continuous or broken up into three) on declivity of tergum I; abbreviated anterior transverse bands (only visible if metasoma is downwardly flexed) on terga II and III. The following are various shades of light reddish-brown: mandibular teeth; antennal club (apex of last flagellomere dark brown); last one or two tarsomeres Volume 16, Number 2, 2007 229 (arolia dark brown); usually concealed anterior third of terga II— VI and poorly defined laterally abbreviated and medially interrupted pre-apical transverse bands on terga II- V. Tegulae yellowish-white except for unpigmented translucent central area and outer margin. Wing membrane hya- line; costa, subcosta, media, thickening at junction of Rs & M, parastigma and stigma light brown, rest of venation contrastingly dark brown. Length 5.8-6.7 mm (average of 6:6.3 mm; length of front wing 3.8-4.3 mm (average of 6:4.1 mm); hamuli 7. Head, thorax and gaster sparsely cov- ered with short, semi-erect pale pilosity, slightly longer and most noticeable on head, declivity of propodeum, declivity of tergum I, and sternum VI. Head in front view 1.25 X as wide as long, microreticulate, with close, fine, shallow punctures on vertex. POL:OOL = 1:0.6. Clypeus 1.2 X as wide as long. Mandible simple, apically strongly biden- tate. Thorax microreticulate; mesoscutum and scutellum with only scattered, inconspicu- ous, very shallow, small punctures; prono- tum and mesopleuron with conspicuous, moderate-sized, shallow punctures. Tegula 1.5 X as long as wide, the posterior inner corner distinctly inwardly produced. Pro- podeal angles evenly rounded. Gaster microreticulate and with fine punctures. Male: Coloration as in female. Parameres light reddish-brown. Length 5.9-6.3 mm; length of fore wing 3.6-4.3; hamuli 7. Structurally similar to female but differ- ing in the following respects: fore femur (Fig. 5) considerably more robust, notched ventrally in basal third and with distally directed, apically rounded, lamellate pro- cess; tergum VII (Fig. 13) with surface flattened medially, with hind margin widely rounded and medially deeply and narrowly emarginate; sternum VII with surface convex medially, concave laterally, with apical margin widely trilobed, lateral lobes ventrally curved. Genitalia very large (1.5 mm long; i.e. half the length of the gaster); outer ramus of parameres broad in dorsal view, apically obliquely truncate and densely covered with fine, long setae; inner ramus proximally of varying width and distally progressively narrowing and mark- edly and evenly downcurved to form a sharp, well sclerotized hook attaining level of lateral posterior angle of outer ramus. Etymology. — The name femorata serves to draw attention to the uniquely modified front femur of the male. Material examined. — Holotype: j, NAMIBIA: 1 1 km S of Swakopmund on inland side of road B2 to Walvis Bay (22.46S 14.32E), 7.iv.2002 (F. W. and S. K. Gess) [AMG]. Paratypes: NAMI- BIA: same data as holotype, 6 99, 13 $ $ [AMG]; same data as holotype but date 14. iv. 2002, 1 9, 2 $3 [AMG]; same data as holotype but date 20.iv.2002, 1 9, 4 S3 [AMG]; same data as holotype but date 30.iii.2004, 1 9, 2 J 3 [AMG]; same data as holotype but date 31 .iii.2004 [AMG], 1 9 [AMG]; Walvis Bay, 22.ii.1990 (W. J. Pulawski), 6 99, 2 3S) [CAS]. (All specimens collected by F. W. and S. K. Gess were visiting the pink flowers of Trianthema hereroensis Schinz (Aizoaceae: non-Mesembryanthema) or were on the sand immediately next to these plants where resting or mating.) Geographic distribution. — Q. femorata is known only from Namibia, from a single locality on the seaward side of the coastal dunes at the northern extremity of the Southern Namib of Giess (1971). Floral associations. — Q. femorata has con- sistently been found to be associated solely with Trianthema hereroensis Schinz (Aizoa- ceae: non-Mesembryanthema). Nesting. — Unknown; probably in the sand beneath the hummock forming Tri- anthema bushes. Quartinia geigeriae Gess, new species Diagnosis. — Medium sized to large (3.8- 5.0 mm). Fore wing with Cula and 2m-cu complete and as thick as other veins. Tegula short, laterally rounded, with pos- 230 Journal of Hymenoptera Research terior inner corner a near right angle. Both sexes with angles of propodeum very markedly posteriorly produced, lamellate and subhyaline. Female with head and mesosoma black, tegulae and gaster red- dish brown. Male with head, mesosoma and gaster black with yellowish-white markings. Description. — Female: Black. The follow- ing are various shades of reddish brown: labrum; distal two thirds of mandibles; tegula; scutellar lamella; median section of metanotum; in some specimens a narrow streak dorsally on outer aspect of lamellate propodeal angle (rest of lamella subhya- line); terga I-IV or V (narrow posterior bands lighter in colour than rest of terga). Underside of antenna, distal quarter of femur, entire tibia and all tarsomeres of all legs light reddish yellow. Wings hyaline; veins brown. Length 4.6-5.0 mm (average of 6:4.8 mm); length of fore wing 2.7-3.0 (average of 6:2.9 mm); hamuli 5; length of extended tongue 3.1-3.2 mm. Head in front view 1.23 X as wide as long, microreticulate but shiny, with sepa- rated, moderate sized punctures. PO- L:OOL - 1:0.85. Clypeus 1.6 X as wide as long (length measured to bottom of emar- gination; 1.36 X if measured to level of antero-lateral angles), markedly raised an- teriorly and laterally, a little flattened medially; anterior margin deeply and evenly emarginate; antero-lateral angles narrowly rounded, lamellate, subhyaline. Mesosoma microreticulate but shiny; mesonotum and scutellum with punctures slightly larger and sparser than on head; pronotum with punctures similar to those on head; mesopleuron with punctures close together, reticulate-punctate ventral- ly. Propodeum dorso-laterally markedly raised, dorso-medially depressed to expose metanotum, posteriorly with upper three quarters flat, closely reticulate-punctate and lower quarter unpunctured and shiny, laterally with a smooth, shiny depression and arising from it a very pronounced posteriorly directed lamella; lamella flat, very thin, subhyaline, basally slightly rugose but elsewhere smooth, marginally widely and evenly rounded. Gaster microreticulate but shiny; punc- tures finer and shallower than on head and mesosoma, becoming progressively finer posteriorly. Vestiture generally very short and sparse, longer and more noticeable on labrum, posterior flat surface of propo- deum and declivous anterior face of tergum I. Male: Black. The following are yellowish- white: base of labrum (in some specimens only); clypeal disk and adjoining it a large medial marking on frons together forming an hour-glass-like figure); scape, pedicel and proximal flagellomeres; anterior mar- gin of pronotum (transverse band in some specimens medially interrupted and re- duced to two spots); tegula (except for pale testaceous discal spot); in some specimens a narrow streak dorsally on outer aspect of lamellate propodeal angle (rest of lamella subhyaline); narrow posterior bands on terga I-VI (very narrowly anteriorly wid- ened medially on II— VI; immediate vicinity of emargination of tergum VII; distal quar- ter of femur, entire tibia and all tarsomeres of all legs. Varyingly reddish brown are: mandible distally; concave declivous ante- rior surface of tergum I. Underside of antennal club light reddish, upper side brown. Wings hyaline; veins brown. Length 3.8-4.5 mm (average of 6:4.1 mm); length of front wing 2.4-2.8 mm (average of 5:2.6 mm); hamuli 4-5. Structurally very similar to female but puncturation on gaster noticeably coarser. Tergum VII reticulate punctate, postero- medially with a shallow V-shaped emar- gination. Parameres postero-laterally smoothly curved to apex; apex not hooked and inner edge of parameres not toothed. Labrum shiny, non-carinate. Antenna with poorly defined, elongate club. Etymology. — The name geigeriae, genitive singular, is formed from the generic name Volume 16, Number 2, 2007 231 of the plants, Geigeria spp. (Asteraceae), on the capitula of which the wasp was found foraging for nectar or nectar and pollen. Material examined. — Holotype: 9, NAMIBIA: Solitaire (23.52S 16.00E), 30.iv.2002 F. W. and S. K. Gess) (visiting yellow flowers of Geigeria ornativa O. Hoffrn., Asteraceae) [AMG]. Para- types: NAMIBIA: same data as holotype, 2 99, 6 o'JlAMG]; between Solitaire and Nomtsas (24.15S 16.33E), l.v.2002 (F. W. and S. K. Gess), 10 99, 2 33 (9 99, 1 S visiting yellow flowers of Geigeria ornativa; 1 9, 1 o visiting yellow flowers of Geigeria pectidea (DC.) Harv.) [AMG]; 1 km N of Mariental (24.37S 17.58E), 2.V.2002 (F. W. and S. K. Gess), 30 99, 4 33 (22 99, 2 33 visiting yellow flowers of Geigeria ornativa; 8 99, 2 33 visiting yellow flowers of Geigeria pectidea) [AMG]; between Mariental and Keetmanshoop (24.54S 17.55E), 2.V.2002 (F. W. and S. K. Gess), 1 3 (visiting yellow flowers of Geigeria pectidea) [AMG]; 18 km from Ariamsvlei on road to Aroab [28.00S 19.43E], 14.V.1973 (C. F. Jacot-Guillar- mod), 5 99, 1 3 [AMG]; SOUTH AFRICA: NORTHERN CAPE: Langvlei, 103 km WNW of Upington [28.10S 20.16E], 14.V.1973 (C. F. Jacot- Guillarmod), 21 99, 2 33 [AMG]. Geographic distribution. — Q. geigeriae is known from Namibia, from a limited area in the Semi-desert and Savanna Transition (Escarpment Zone) and the adjoining Dwarf Shrub Savanna of Giess (1971), and from a closely adjoining locality in the Northern Cape. Floral associations. — Known only in asso- ciation with two species of Geigeria, Aster- aceae). Nesting. — Unknown. Discussion. — Q. geigeriae shares with Q. arteritis Richards, Q. breyeri Richards and the below described Q. lameilata the pos- session of markedly backwardly produced propodeal lamellae. Q.geigeriae together with breyeri and lameilata is readily distin- guished from artemis in having the poste- rior inner corner of the tegula rounded or a near right angle, not markedly produced inwards; it is distinguished from both breyeri and lameilata in having the epicne- mium rounded, not defined by a low carina. Quartinia lameilata Gess, new species Diagnosis. — Large to very large (5.0- 6.2 mm). Fore wing with Cula and 2m-cu complete and as thick as other veins. Clypeus raised and protruding with, espe- cially in female, marked disto-lateral lobes. Labrum large, very noticeable, in female setose. Epicnemium defined by a low carina. Tegula rounded posteriorly, with posterior inner corner a near right angle. Angles of propodeum markedly back- wardly produced, lamellate. Description. — Female: Black. The follow- ing are yellowish-white: in some speci- mens a small spot on disto-lateral lobe of clypeus; transversely oval or bilobed me- dial marking (in some specimens reduced to two round spots) distally on frons immediately above clypeus; in a single specimen a small round spot in ocular sinus; broad streak behind top of eye; scape (distally), pedicel, intermediate flagello- meres, and underside of antennal club; pair of spots on dorsum of pronotum; large mark on humeral angle (in some speci- mens remote from spots on dorsum, in others fused with them to form a continu- ous band); variously developed streak on postero-dorsal angle of pronotum; in some specimens a small spot on mesopleuron; tegula (except for testaceous median area); in some specimens a small streak laterally (flanking tegula) on mesonotum; curved posterior band on disk of scutellum; angles of propodeum; posterior bands, reaching sides and generally slightly expanded medially and laterally, on terga I-V; apical half of tergum VI; postero-lateral corners of sterna II— V and apical half or more of sternum VI; distal half or less of femur, entire tibia and tarsus of all legs. Mandibles, labrum and suffusion on upper surface of antennal club reddish- brown. Wing membrane hyaline; veins brown. Length 6.0-6.2 mm (average of 3:6.06 mm); length of fore wing 3.9-4.08 mm (average of 3:4.0 mm); hamuli 4. 232 Journal of Hymenoptera Research Head in front view 1.21 X as wide as long. POL:OOL = 1:0.83 Clypeus raised and protruding, medially depressed, distally widely and deeply emarginate and with marked disto-lateral lobes. Labrum large, longer than wide, apically pointed, setose. Clypeus and frons moderately shiny, with close, fairly coarse punctures and finely micro- sculptured interstices; pronotum, me- soscutum and scutellum with larger, much more sparsely arranged punctures and extremely finely microsculptured inter- stices; terga uniformly finely punctured. Epicnemium defined by a low carina. Tegula rounded posteriorly. Angles of propodeum markedly backwardly pro- duced, at mid-height forming a rounded projection and below that translucently lamellate. Male: Black. The following are yellow: clypeus (other than for, in some specimens including holotype, a variously sized me- dian longitudinal marking and in all speci- mens areas immediately adjacent to anten- nal sockets); large transverse marking distally on frons immediately above clyp- eus; broad streak behind top of eye; scape (distally), pedicel, intermediate flagello- meres , and underside of antennal club; most or almost entire dorsal surface of pronotum (except in all specimens small postero-lateral area flanking tegula); spot on mesopleuron; tegula (except for testa- ceous median area); in all specimens a marking (ranging from a minute spot to a small streak) flanking tegula on mesono- tum; curved posterior band on disk of scutellum; scutellar lamella; angles of propodeum; posterior bands (anteriorly ill-defined and grading into reddish- brown), reaching sides on terga I-VI and, to a varying degree, apical half of tergum VII; ill-defined posterior bands on sterna II— VI; most of sternum VII; distal half or less of femur, entire tibia and tarsus of all legs. Mandibles (wholly or in part), labrum and suffusion on upper surface of antennal club, terga and sterna anterior to posterior bands reddish-brown. Wing membrane hyaline; veins brown. Length 5.0-5.8 (average of 3:5.2 mm; holotype 5.0 mm); length of fore wing 3.0 mm. Head in front view 1.24 X as wide as long Structurally very similar to female but puncturation on head and mesosoma markedly coarser. Tergum VII with hind margin shallowly emarginate and postero- lateral lobes rounded. Etymology. — The name lamellata is in- tended to draw attention to the markedly backwardly produced, lamellate angles of the propodeum. Material examined.— Holotype: 3, NAMIBIA, Rooibank [23.11S 14.39E], 19.xii.1978 (H. Empey) [AMG]. Paratypes: NAMIBIA: same data as holotype but date 28.xii.1978, 2 $$ [AMG]; Kaokoland [Dist.], Otjinungwa (SE 1712 Ab) [17.17S 12.27E], 19-22.viii.1973 (? collector), 1 9 [NNIC]; Kaokoland [Dist.], Khowarib R. (SE 1914 Ac) [locality not pin- pointed], 17-19.V.1978 (S. Louw, M.-L. Pen- rith), 1 ct [NNIC]; Namib Naukluft Park, Vogelfederberg (23.03S 15.00E), 21. ii. 1988 (G. D. Butler), 1 9 [NCP]; same locality, 24.L1988 (R. Miller and L. Stange), 1 9 [FSCA]; Luderitz [Dist.], Sossusvlei (SE 2415 Cd) [24.43S 15.20E], 12-19.ix.1971 (? collector), 3 99 [NNIC]; Luderitz [Dist.], Kanaan 104 (SE 2516 Cc) [25.50S 16.09E], 6-7.X.1972 (? collector), 10 99 [NNIC]; Geographic distribution. — Quartinia lamel- lata is widespread in the western parts of Namibia, collection localities spanning eight degrees of latitude and falling in the Mopane Savanna, Central Namib and Southern Namib /Semi-desert and Savanna Transition (Escarpment Zone) of Giess (1971). Floral associations. — Unknown. Nesting. — Unknown. Discussion. — See discussion under geiger- iae. On the basis of the characters there listed, lamellata is closest to breyeri but may readily be distinguished from that species by its larger size, differently developed clypeus and labrum, differences in punc- turation and in colour pattern. Volume 16, Number 2, 2007 233 ACKNOWLEDGMENTS The following individuals are thanked for much appreciated assistance as specified: Dr Sarah Gess of the Albany Museum, Grahamstown, co-collector of most of the Albany Museum's Quartinia material, for over thirty years of happy, productive and synergistic fieldwork, for valuable discussion and encourage- ment; Mr Robert W Gess for field assistance in southern Namibia and the Northern Cape in 1996; Mr David W Gess, Ms Gaby T Gess and Miss Gaby Maria Gess for field assistance at Melkbosstrand in 2005 and at Yzerfontein in 2006; Coleen Mannheimer of the National Herbarium of Namibia, Windhoek for her invitation to join the Herbarium party on their expeditions to the Sperrgebiet in 2002, 2003 and 2005 and also for her determination of voucher specimens of Namibian plants visited for pollen and nectar by masarines; Eugene Marais of the Namibian National Insect Collection, Windhoek, Connal Eardley of the National Collection of Insects, Pretoria, Wojciech Pulawski of the California Academy of Sciences, San Francisco, and Lionel Stange and Jim Wiley of the Florida State Collection of Arthropods, Gainesville for the loan of specimens from their respective collec- tions; Caroline Mayer of BIOTA-Southern Africa, Hamburg University for the gift of specimens collect- ed by herself in Namaqualand. Grateful thanks are expressed to all those bodies which issued permits for the collection of insects and plant samples, namely: the Namibian Ministry of Environment and Tourism; the Namibian Ministry of Mines and Energy as also NAMDEB (Pty) Ltd (for the Sperrgebiet - Diamond Area No 1); the Department of Nature and Environmental Conservation, Northern Cape; CapeNature (Western Cape Nature Conserva- tion Board); Department of Economic Affairs, Envi- ronment and Tourism, Eastern Cape (Western Re- gion); and the Nature Conservation Division, City of Cape Town (for the Blaauwberg Conservation Area). Debi Brody of the Graphics Services Unit of Rhodes University, Grahamstown is thanked for help with the production of the figures. The National Research Foundation (NRF) is thanked for running expenses grants for fieldwork during the course of which much of the present material was collected. The Board of Trustees of the Albany Museum is thanked for Research Contracts granted to the author and Dr Sarah Gess since 2003, which have given them continued use of the mu- seum's facilities since their retirements. LITERATURE CITED Andre, Ed. 1884 Species des Hymenopteres d'Europe et Algerie. Vol. 2. Beaune, Andre and Andre. Carpenter, J. M. 2001. Checklist of species of the subfamily Masarinae (Hymenoptera: Vespidae). American Museum Novitates 3325: 1-39. Gess, F. W. and S. K. Gess. 1999. The use by wasps, bees and spiders of shells of Trigonephrus Pilsb. (Mollusca: Gasteropoda: Dorcasiidae) in desertic winter-rainfall areas in southern Africa, journal of Arid Environments 43: 143-153. Giess, W. 1971. A preliminary vegetation map of South West Africa. Dinteria 4: 1-114. Greathead, D. J. 1999. Apolysis sp. (Diptera: Bombylii- dae) reared from Quartinia sp. (Hymenoptera: Vespidae: Masarinae). journal of Arid Environ- ments 43: 155-157. . 2006. New records of Namibian Bombyliidae (Diptera), with notes on some genera and descriptions of new species. Zootaxa 1149: 1-88. Leistner, O. A. and J. W. Morris. 1976. Southern African Place Names. Annals of the Cape Provincial Museums (Natural History) 12: i-iv, 1-565. Richards, O. W. 1962. A revisional study of the masarid wasps (Hymenoptera, Masaridae). London: British Museum (Natural History). . 1982. A new species of Quartinioides Richards (Hymenoptera, Masaridae. Bollettino del Museo civica di storia naturale di Venezia 32: 199-200. Schulthess, A. von. 1929. Contribution to the knowledge of African Masaridae (Vespoidea). Annals and Magazine of Natural History (10)3: 498-511. Vecht, J. van der. and J. M. Carpenter. 1990. A catalogue of the genera of the Vespidae (Hyme- noptera). Zoologische Verhandelingen 260: 1-62. J. HYM. RES. Vol. 16(2), 2007, pp. 234-265 Torymidae (Hymenoptera: Chalcidoidea) Associated with Bees (Apoidea), with a List of Chalcidoid Bee Parasitoids E. E. Grissell Systematic Entomology Laboratory, PSI, Agricultural Research Service, U. S. Department of Agriculture, c/o National Museum of Natural History, Smithsonian Institution, P.O. Box 37012, MRC 168, Washington, D. C. 20013-7012 Abstract. — Thirty-one species of Torymidae (Hymenoptera: Chalcidoidea) are associated with bees. In this review each is keyed and discussed, and geographic ranges and hosts are given. Most species are illustrated. Torymids represent about one-fourth of the 135 species of Chalcidoidea associated with bees. Two summary lists are presented for all chalcidoids, including Torymidae, and the 216 bee species with which they are associated. One is arranged as a bee/parasitoid list and the other as a parasitoid/bee list. Considering that 22,000 species of Chal- cidoidea (Noyes 2003) and 16,000-17,000 bee species (Michener 2000) have been described, the number of chalcidoids re- ported associated with bees is surprisingly small. At most 135 different chalcidoids have been reared from, or associated with, 216 bee species (see Appendix, de- rived from Noyes 2003). Of these, the families Torymidae and Leucospidae have the highest percentage of the bee parasitoids (each 22-23%), followed closely by Pteromalidae (18%). The other families associated with bees are: En- cyrtidae (13%), Eulophidae (13%), Chalci- didae (5%), Eurytomidae (5%), Eupelmi- dae (3%), Mymaridae (0.6%), and Perilam- pidae (0.6%) (Appendix: based on Noyes 2003). Although Torymidae and Leucospidae have the highest number of bee parasitoids among Chalcidoidea, this figure is some- what misleading. Of approximately 1,000 torymid species, only 31 are known (or suspected) to attack bees (Grissell 1995, 2000, 2005; Noyes 2003), so a predilection for bee hosts is not especially pronounced in the family. The host range of this family is extremely broad, but nearly 80% of the known hosts are shared equally between the Hymenoptera and Diptera, most of which are gall-forming cynipids and ceci- domyiids (Grissell 1995). Conversely, the entire family Leucospidae, consisting of 135 species, has been presumed to parasit- ize aculeate Hymenoptera — solitary bees, and less frequently, solitary wasps. In reality, however, hosts are known only for about 30 leucospid species (Boucek 1974, Noyes 2003), so the true relationship of the family to bees is largely unknown. Recently a species of leucospid was re- ported as an ectoparasitoid of an ichneu- monid attacking a cerambycid in limbs of apricot in Iran (Hesami et al. 2005). This finding casts doubt on our concept of host specificity in Leucospidae. In this paper I present a summary of torymid species reported to attack bees, including a review of published informa- tion for each species and a key. I also include a world bee/chalcidoid and chal- cidoid/bee list for all Chalcidoidea re- portedly associated with bees (Appendix). Volume 16, Number 2, 2007 235 As with many chalcidoid records, the true host is not always indicated by the host record given (Noyes 1994). Many bee host records are simply nest rearings and may have been contaminated by other true parasitoids, cleptoparasitoids, inquilines, and simple space usurpers of all sorts, many not even hymenopteran. Similarly, a mud wasp's nest may be usurped by a nesting bee, thus causing confusion as to the true host (Rust 1974). Bee nests, as well as almost any other ecological niche, offer complex arrays of hosts, many of which are not even suspected at the time of rearing. For example, Glypkomerus stigma (Fabri- cius) was reported from Melitoma taurea (Say) (Apidae), but this is likely to be an error because all other records for species of Glyphomerus are gall-forming cynipids or rarely eurytomids (Grissell 1995). With respect to bee parasitoids, therefore, all records should be considered tentative until established by dissection and obser- vation. Within the Torymidae listed in this paper, I point out that several are likely not to be true bee parasitoids. In those few cases where the biologies of torymids are known they are generally solitary, idiobio- tic larval ectoparasitoids, but in several genera (e.g., Monodontomerus, Microdonto- merus) larvae are known to be gregarious (Grissell 2000, 2005). In examining host records presented in the Appendix several reviewers suggested that it might be informative to summarize parasitoid data with respect to bee biology as there appeared to be a bias towards twig and cavity nesting bees, with ground- nesters being under-represented. I solicited the input of two recognized bee authori- ties: Frank Parker, who specializes in twig- nesters, and Jerry Rozen, who specializes in ground-nesters, and both agreed that the data suggested cavity nesters were the predominant host representatives. These are primarily twig nesters, bees that nest in pre-existing crevices or cavities, and bees that re-use old bee nests. Some of these nests may be external, for example resin nests attached to objects such as twigs and rocks. According to Rozen most of the records are indicative of shallow nesting bees, and he suggested that ground nesting bees in general would be less likely to harbour parasitoids because they might have a more difficult time entering nests and crawling down the "... long, main tunnels" to find their host. He also pointed out that old bee nests and shallow cavities are frequently re-used several times, thus encouraging the build-up of large parasit- oid populations. Parker suggested that twig-nesting bees are more likely to be sampled because they readily come to artificial traps set out by the collector. They are also easier to extricate and study in these nests. Conversely, ground nesting bees must be actively hunted by the collector, are less easily found, and require painstaking excavation to reveal nest de- tails. In general, then, records summarized in the Appendix indicate that host data are biased towards parasitoids attacking cavity nesting bees and that multiple causes contribute to this bias. Whatever cursory glimpses the bee/parasitoid host list may reveal, and considering the numerical size of the chalcidoid and apoid groups, it appears that much remains to be discov- ered. Within existing literature, relatively little is devoted to parasitization and then primarily only to a few solitary bee species (e.g., the alfalfa leafcutting bee, Stolbov et al. 1986), whereas with few exceptions (e.g., Zerova and Romasenko 1986) there is scarcely any comprehensive literature pertaining to solitary bee parasitoids. METHODS In the following discussions host names are given without authors. Complete authors' names may be found in the Appendix. Within discussions, hosts are listed alphabetically by family, but in the host listing all hosts are alphabetic regard- less of family. 236 Journal of Hymenoptera Research KEY TO TORYMIDAE ASSOCIATED WITH SOLITARY BEES 1 Anterior edge of metapleuron straight, not projecting forward as lobe into mesepimeron (Fig. 2) 4 — Anterior edge of metapleuron (usually its upper half) projecting forward as lobe into mesepimeron (Fig. 1), which is subdivided into upper and lower sections, lower section delimited by anterior groove Torymus Dalman 2 2 Hind coxa dorsally covered with short setae (Fig. 3), coarsely reticulate; propodeum areolate-rugose, heavily carina te (Fig. 5); frenal area less than 1/5 length of scutellum (Fig. 7) 3 Hind coxa dorsally bare (a few long setae may be present; Fig. 4), smooth and polished; propodeum essentially smooth (Fig. 6); frenal area 1/3 to almost 1/2 length of scutellum (Fig. 8) (Palearctic, Australasian [?introduced]) Torymus armatus Boheman 3(2) Head dorsum, mesosoma, and hind coxa coppery with greenish tints; at least part of hind femur orange, concolorus with tibia (Palearctic) . . . Torymus cupreus (Spinola) Head dorsum, mesosoma, and hind coxa metallic green or blue; entire hind femur metallic green or blue, contrasting with orange tibia (Nearctic) Torymus zabriskii (Cresson) 4(1) Fore wing with marginal and stigmal veins conspicuously thickened relative to submarginal vein, postmarginal vein not projecting beyond tip of stigmal vein (Figs 9, 12, 13), and with marginal vein slightly removed from margin of wing (Fig. 13; may be somewhat difficult to see); malar distance longer than intermalar distance (Figs 14, 15); mandibles reduced, scarcely visible, tips not meeting medially when closed, apically without teeth Echthrodape Burks 5 Fore wing with marginal and stigmal veins not conspicuously thickened relative to submarginal vein, with postmarginal vein longer than stigmal vein (Fig. 10), and with marginal vein at edge of wing margin; malar distance subequal to or shorter than intermalar distance (Fig. 11); mandibles visible, tips meeting medially when closed, apically with teeth 6 5(4) Postmarginal vein developed, longer than stigmal vein, which is slender and petiolate (Fig. 12); genae straight, not concave (Fig. 14) [Papua New Guinea, Austra- lia] Echthrodape papuana Boucek Postmarginal vein reduced, subequal to stigmal vein, which is thick and sessile (Fig. 13); genae concave (Fig. 15) [Kenya] Echthrodape africana Burks 6(4) Occipital carina absent (Fig. 16), weakly or questionably developed, or if apparent, then medially arched and midway between hind ocelli and occipital foramen and not reaching hypostomal carina (Fig. 17) (head usually vertical with dorsoposterior aspect slightly concave and the carina, if present, easily seen); hind femur slender, apicoventrally either without tooth (Fig. 34), angulate, or vaguely serrate; metasomal terga with or without apicomedian emarginations, often weakly sclerotized 26 Occipital carina well developed, dorsal margin not greatly arched but nearly horizontal (Fig. 18), closer to occipital foramen than to hind ocelli and reaching hypostomal carina (head usually tilted forward with dorsoposterior aspect conspicuously concave and occipital carina easily seen, but head must be removed to see hypostomal carina); hind femur apicoventrally with abrupt tooth (Figs 35, 37, 38), or greatly swollen and angulate (Fig. 36); metasomal terga heavily sclerotized, without apicomedian emarginations [Holarctic, Neotropical, Oriental] Monodontomerus Westwood 7 Volume 16, Number 2, 2007 237 7(6) First 2 flagellar segments reduced in length, ring-like (Fig. 19); hind femur swollen with distal subapical angle but without distinct tooth (Fig. 36) [Nearc- tic] Monodontomerus thorpi Grissell At most, first flagellar segment reduced in length (Fig. 20); hind femur relatively narrow with distinct subapical tooth (Figs 37, 38) 8 8(7) Female, face transverse, intermalar distance 3.5 to 5x length of malar distance; male, face grotesquely modified, entirely sunken medially (as if entirely consisting of scrobal basin) (Fig. 22), with sharp edge mesad of eye (Fig. 21) [Palearctic, Nearctic (introduced)] Monodontomerus osmiae Kamijo Both sexes, face at most slightly transverse, intermalar distance from 1 to 3X length of malar distance; male with face not medially sunken, scrobal basin normal, though areas on either side of scrobe may be slightly depressed 9 9(8) Clypeus greatly elongate (Figs 23, 24) [Palearctic] . . Monodontomerus anthidiorum Lucas Clypeus either barely reaching to or beyond line drawn across lateral corners of oral fossa (Figs 25-28) 10 10(9) Upper mesepimeral area with anterior half reticulately sculptured and anterodorsal corner diagonally striate extending nearly to transepimeral sulcus (Fig. 53) .... 11 Upper mesepimeral area nearly entirely polished with striae scarcely extending half way to transepimeral sulcus (Figs 54, 55) 12 11(10) Discal setae of fore wing not extending into basal area (as in Fig. 31); female with ovipositor sheaths shorter than metasoma; male with clypeus recessed (not extending beyond line drawn across lateral corners of oral fossa), malar sulcus absent or obscure, malar distance subequal to intermalar distance; scape with ventral surface slightly keeled vertically (i.e., not flat), no pores visible [Palearctic] Monodontomerus laticornis Grissell and Zerova Discal setae of fore wing extending into basal area (as in Fig. 29); female with ovipositor sheaths as long as or longer than entire body; male with clypeus extending beyond line drawn across lateral corners of oral fossa, malar sulcus present; malar distance about 1.5X intermalar distance; scape with ventral surface flat, covered with pores visible at 100 X [Nearctic] Monodontomerus dementi Grissell 12(10) Metasomal tergum 2 dorsally with reticulate to strigate sculpture in distal half ... 13 — Metasomal tergum 2 dorsally smooth, polished in distal half 14 13(12) Distal portion of postmarginal vein equal in length to proximal portion (as in Fig. 30); rim of scutellum apically widened, somewhat projecting; females, metasomal tergum 6 acute in profile (as in Fig. 32); male, fore leg unmodified (i.e., normal) (as in Fig. 39), tibia equal in length to femur and not ventrobasally concave, tarsomeres elongate (claw length equal to or shorter than tarsomere 4) [Nearctic] Monodontomerus dianthidii Gahan Distal portion of postmarginal vein 0.33 X length of proximal portion (as in Fig. 31); rim of scutellum apically even in width, not projecting; female, metasomal tergum 6obtuse in profile (as in Fig. 33); male, fore leg modified (Fig. 40), tibia shorter in length than femur and ventrobasally concave, tarsomeres shortened (claw length equal to tarsomeres 3 and 4) [Nearctic] Monodontomerus breincrus Grissell 14(12) Malar sulcus absent (Fig. 28), or if weakly apparent, greatly curved backward from lower margin of eye then curving downward to join edge of malar opening (Fig. 27); lower face protuberant in profile (Fig. 28) [Nearctic] Monodontomerus bakeri Gahan Malar sulcus well developed, straight (Figs 25-26), or slightly curved from lower margin of eye to lateral edge of malar opening; lower face flat (not bulging) in profile (Fig. 26) 15 15(14) Frenal area medially highly polished, appearing glabrous, faint coriaceous sculpture may be seen with difficulty at some angles of view (questionable species will run through either couplet of key) 16 238 Journal of Hymenoptera Research Frenal area medially sculptured, may be uniformly similar overall or relatively less prominent than laterally, never glabrous, sculpture easily visible at any angle of view (questionable species will run through either couplet) 19 16(15) Costal cell above with apical setal row incomplete, confined to distal 1 /2 or less of cell (as in Fig. 31); female, metasomal tergum 6 strongly concave in profile (as in Fig. 32) [Nearctic] Monodontomerus torchioi Grissell (most specimens) — Costal cell above with apical setal row complete (as in Fig. 29); female, metasomal tergum 6 weakly concave in profile (as in Fig. 33) 17 17(16) Frenal area apicomedially intruding into rim, punctures of rim reduced in size at point of intrusion (Fig. 42); stigma and uncus relatively short, postmarginal vein with proximal and distal section subequal in length (Fig. 30) [Holarctic] Monodontomerus aeneus (Fonscolombe) — Frenal area with apical rim not interrupted posteriorly at median margin, punctures of rim as large or larger at apex as on sides (as in Fig. 41); stigma and uncus elongated, postmarginal vein with proximal section longer than distal (as in Fig. 31) 18 18(17) Female ovipositor sheaths swelling distally (i.e., not parallel-sided); male hind femur broad, widening apically, about 2.5 X as long as wide (Fig. 38) [Palearctic] .... Monodontomerus rugulosus Thomson Female ovipositor sheaths same width throughout (i.e., parallel-sided); male hind femur narrow, dorsal and ventral margins essentially parallel (Fig. 37), about 3.5 X as long as wide [Neotropical] Monodontomerus argentinus Brethes 19(15) Costal cell above with apical setal row complete (Fig. 29) 20 Costal cell above with apical setal row incomplete, confined to distal 1/3 to 1/2 of cell (Fig. 31) or appearing absent (2 or 3 setae may be present at apex as in Figs 43, 44) ... 22 20(19) Scape about 4x longer than wide, greater in length (about 1.3X) than distance from venter of torulus to apical margin of clypeus [Nearctic, Neotropical] Monodontomerus mexicanus Gahan Scape about 3X longer than wide, subequal in length to distance from venter of torulus to apical margin of clypeus 21 21(19) Stigma rectangular, proximally elongated towards base of wing (Figs 51, 52); postmarginal vein with distal length less than proximal length (Figs 51, 52); male face with depression laterad of scrobal basin [Nearctic] Monodontomerus acrostigmus Grissell Stigma squarish, neither stigma nor proximal angle elongated (as in Figs 10, 44); postmarginal vein with distal length subequal to basal length (as in Fig. 44); male face convex laterad of scrobal basin [Holarctic] Monodontomerus obscurus Westwood 22(19) Admarginal setae reaching bases of marginal vein and parastigma (Fig. 43); intermalar distance subequal to 3x malar distance (Fig. 48); both mandibles with single apical tooth and small secondary tooth on dorsal margin (Fig. 48) [Nearctic] Monodontomerus mandibularis Gahan Admarginal setae either not reaching base of marginal vein or apex of parastigma (Fig. 44); intermalar distance less than 2.5 X malar distance (Fig. 47); both mandibles with 2 apical teeth, and small third tooth on dorsal margin (Figs 47) 23 23(22) Transepimeral sulcus incomplete (Figs 53, 55); upper anterior margin of costal cell with setal row in apical 1/4 to 1/3 (as in Fig. 31); male, scape in side view slightly curved in profile (Fig. 56), area beneath torulus flat, sculptured, and setose .... 24 Transepimeral sulcus complete (Fig. 54), appearing as a sculptured groove; upper anterior margin of costal cell with 1 to 3 setae at apex (Figs 43, 44); male, scape in lateral view strongly C-shaped (Figs 57, 58), area beneath torulus slightly swollen, polished, and asetose 25 24(23) Frenal area medially with reticulate sculpture readily apparent, area may be shiny, but sculpture visible at any angle of view; male, ventral surface of scape without pores visible at 100X [Nearctic] Monodontomerus montivagus Ashmead Volume 16, Number 2, 2007 239 Frenal area medially with reticulate sculpture visible only at some angles of view and seen only with difficulty, area shiny and appearing polished; male, ventral surface of scape evenly covered with pores easily visible at 100X [a few atypical specimens run here, but most to couplet 16 based on the polished frenal area] [Nearctic] Monodontomerus torchioi Grissell 25(23) Females, ovipositor subequal to metasoma (ca. 1-1. 2X); scape orange to yellow without metallic infusion especially ventrally; male, scape greatly laterally compressed (ventral and dorsal surfaces essentially absent), outer surface flat, polished, asetose, and curving smoothly to inner surface without interruption (Fig. 57), no pores visible at 100x [Nearctic] Monodontomerus parkeri Grissell Females, ovipositor obviously longer than metasoma (ca. 1.5-1.8X); scape with metallic green infusion at least ventrally (may be complete or confined to area just beneath pedicel); male, scape dorsoventrally compressed, curved, with dorsal and ventral surfaces parallel and delimited by right-angled edge (Fig. 58), ventral surface polished and covered with pores visible at 100 x though difficult to see [Nearctic] Monodontomerus tepedinoi Grissell 26(6) Marginal vein long, 3-7X length of postmarginal vein and at least 6X length of stigmal vein; occipital carina present, its lateral edges extending at least in line with dorsum of hypostomal foramen [Oriental] Pseudotorymus indicus (Mani) Marginal vein short, 1-2.5 X length of postmarginal vein and 2-5 X length of stigmal vein (Figs 45, 46); occipital carina absent (Fig. 16) or, if present, its lateral edges not (or scarcely) extending in line with venter of occipital foramen (Fig. 17) 27 27(26) Occipital carina absent (Fig. 16) Microdontomerus Crawford 28 Occipital carina visible in dorsal view as finely polished line raised distinctly above surface sculpture (as in Fig. 17) [Palearctic] Adontomerus Nikol'skaya (A. gregalis (Steffan) and A. nesterovi Zerova) 28(27) Fore wing setation (Figs 45, 46) reduced; basal cell open behind, i.e., cubital vein basally at most with few isolated setae; basal vein at most with isolated setae; basal cell without distinct setal row paralleling submarginal vein 29 Fore wing setation (as in Figs 29, 31) not reduced (except in admarginal area of some species): basal cell closed behind, i.e., cubital vein essentially completely setose to base of wing; basal vein with distinct setal row and basal cell with setal row paralleling nearly entire submarginal vein 30 29(28) Postmarginal vein (Fig. 46) about 0.75 X as long as marginal vein; fore wing with admarginal area (Fig. 46) not well defined posteriorly by setal line, with admarginal setae nearly as uniform as central area of wing [Nearctic] Microdontomerus enigma Grissell Postmarginal vein (Fig. 45) about 0.5 X as long as marginal vein; fore wing with admarginal area (Fig. 45) well-defined posteriorly by setal line, with few sparse setae not as uniform as central area of wing [Nearctic] .... Microdontomerus parkeri Grissell 30(28) Eye height nearly 3x malar distance (Fig. 49); distance between eyes less than eye height (Fig. 49) [Nearctic] Microdontomerus anthidii (Ashmead) Eye height 2.5 X or less than malar distance (Fig. 50); distance between eyes equal to eye height (Fig. 50) [Nearctic] Microdontomerus apianus Grissell Adontomerus Nikol'skaya postmarginal vein, 2 to 5x the length of the stigmal vein, and marginal + postmarginal Recognition. — Adontomerus is recognized veins equal to 0.2 X the length of the wing; by the straight anterior edge of the meta- the occipital carina visible in dorsal view as pleuron (Fig. 2); the fore wing with mar- a finely polished line raised distinctly ginal vein 1 to 2.5 X the length of the above surface sculpture, medially arched 240 Journal of Hymenoptera Research and midway between hind ocelli and occipital foramen (Fig. 17); and the hind femur ventrally without a tooth (as in Fig. 34). Number of Species. — 8. Number Associated with Bees. — 2. Distribution. — Species of this genus are reported in the Palearctic Region including the former Soviet Union, Bulgaria, former Yugoslavia, Hungary, Italy, Sardinia, Spain, Jordan, and Algeria. Hosts of Genus. — Species of Adontomerus have been reared from cocoons of Lasio- campidae (Lepidoptera), galls of Cynipi- dae (Hymenoptera), and cocoons of Mega- chiiidae (Hymenoptera). In the National Museum of Natural History, Washington, DC, there are specimens reared from weevils in seed heads of Asteraceae. Discussion. — Records for the species listed below have been cited in the litera- ture under the genus Mellitotorymus , which was synonymized with Adontomerus by Grissell (1995). Adontomerus gregalis (Steffan) Distribution.— PALEARCTIC: Reported only from Sardinia (Steffan 1964). Host. — Reared from Pseudoanthidium (re- ported as Anthidium) lituratum (Megachili- dae). Discussion. — I believe that this species and Adontomerus nesterovi are synonyms, but I have not seen material of the latter to confirm this. Both share essentially similar descriptions as well as the same host. I treat them separately here to retain the known data, but there is no way to distinguish the species as far as I can tell. Adontomerus nesterovi Zerova Distribution.— PALEARCTIC: Reported from Turkmenistan (Zerova and Roma- senko 1986). Host. — Reared from cocoons of Pseu- doanthidium (as Paraantliidiellum) lituratum (Megachilidae). Discussion. — Zerova and Romasenko (1986) keyed and figured this species in a paper on the parasitoids of megachilid bees in the former Soviet Union. Echthrodape Burks Recognition. — Echthrodape is recognized by the straight anterior edge of the meta- pleuron (as in Fig. 2) and by the relatively short wing venation and the thickened marginal vein (Figs 9, 12, 13), with the postmarginal vein some distance from the distal edge of the wing (Fig. 9). Additional characters that help in recognition are the toothed hind femur (as in Fig. 38), the developed occipital carina that lies mid- way between the hind ocelli and occipital foramen, and the reduced mouth opening (Figs 14, 15, indicated, in part, by the long malar distance) with reduced mandibles (scarcely visible and obscured by other mouth parts). Number of Species. — 2. Number Associated with Bees. — 2. Distribution. — The genus is found in the Afrotropical Region in Kenya, and in the Australasian Region from Papua New Guinea. Host. — Hosts for both species belong to the genus Braunsapis (Apidae). Discussion. — The species of this genus are uncommonly encountered and are presently the only indigenous torymid bee parasitoids known from sub-saharan Africa and Australasia. The lack of records for these areas is probably the result of a paucity of collecting and rearing both bees and parasitoids. Echthrodape africana Burks Re- Distribution.— AFROTROPICAL: ported from Kenya (Burks 1969). Host. — Reared from nests of Braunsapis (as Allodapula) (Apidae) as reported by Burks (1969) and expounded upon by Michener (1969) who reported the follow- ing host records: Braunsapis simplicipes, B. rolini, and B. rufipes. Volume 16, Number 2, 2007 241 Biology. — Larvae of E. africana are exter- nal feeders on pupae of Braunsapis (Mich- ener 1969). One parasitoid was seen per host. The bee is a progressive feeder which uses burrows in the pith of dead Lantana stems. It moves its larvae and pupae about and does not distinguish between its own progeny and those of E. africana. Morphology. — Michener (1969) illustrated and described the peculiar larva of this species as well as the pupa. Discussion. — The two known species are relatively easily identified based on the distinctive heads (Figs 14, 15) and wing veins (Figs 12, 13) as well as their disjunct distributions. Eclitlirodape papnana Boucek Distribution.— AUSTRALASIAN: Known from Papua New Guinea (Boucek 1988) and Australia (R. Matthews, per. comm.). Host. — Reared from cells of Braunsapis unicolor (Apidae) nesting in bamboo inter- nodes (R. Matthews, per. comm.). Discussion. — A voucher specimen for the Australian record was kindly placed in the U. S. National Museum collection by Robert Matthews. Microdontomerus Crawford Recognition. — Microdontomerus is recog- nized by the straight anterior edge of the metapleuron (as in Fig. 2), the simple hind femur (as in Fig. 34), the absence of an occipital carina (Fig. 16), and the marginal vein short, 1 to 2.5 X the length of the postmarginal vein and 2 to 5X the length of the stigmal vein (Figs 45, 46). Number of Species. — 22. Number Associated with Bees. — 4. Distribution. — This genus is transconti- nental in the Nearctic, but limited in other regions of the world. In the Palearctic it is found in Spain, Italy, Algeria, and Libya, and in the Afrotropical Region it is found in Senegal. [The genus was reported in India (see Farooqi 1986, David et al. 1990), but this is probably a misidentification resulting from the confusion in names that existed at the time.] Hosts. — Species are reported from mega- chilid bees and cynipid gall-formers (Hy- menoptera), tephritids (Diptera), buprestid eggs and curculionids (Coleoptera), mantid eggs (Mantodea), and coleophorids, geli- chiids, lasiocampids, and tortricids (Lepi- doptera). At least one Nearctic species attacks saturniid eggs (Lepidoptera). Spe- cies have also been documented as facul- tative hyperparasitoids of braconids (Hy- menoptera) (Grissell 2005). Microdontomerus anthidii (Ashmead) Distribution. — NEARCTIC: This species has been collected in southern California, USA. Host. — Reared from Dianthidium pudicum consimile (as Anthidium consimile) (Mega- chilidae). Discussion. — Microdontomerus anthidii, M. enigma, and M. parkeri are difficult to distin- guish. Generally M. anthidii is smaller (2.3 mm or less) with a shorter ovipositor (less than 1.2 X hind tibia), whereas M. parkeri is larger (up to 3.0 mm) with a longer ovipositor (more than 2x hind tibia). Micro- dontomerus enigma is about the size of M. anthidii, but with the longer ovipositor of M. parkeri. Microdontomerus anthidii is fairly easily separated from the other two, however, based on discrete morphological differences in the fore wing: M. anthidii has a complete setal row along the upper anterior margin of the costal cell (absent in the other two species) and the basal cell is closed (open in the other two species). It appears that while all three species attack megachilid bees, M. anthidii is usually associated with species of the tribe Anthidiini that create nests of resin and sand grains, whereas M. parkeri and M. enigma are associated with Osminiini and Megachilini that make stem nests. Microdontomerus apianus Grissell Distribution. — NEARCTIC: Known from California, USA. 242 Journal of Hymenoptera Research Host. — Reared from Megachile montivaga (Megachilidae). Discussion. — In addition to characters given in the key, this species differs from M. anthidii in having the intermalar dis- tance about 1.7x the malar distance (about 2.5 X in M. anthidii), and in having the ovipositor sheaths subequal to the body length and 2.0-3.0 X as long as the hind tibia (in M. anthidii ovipositor sheaths subequal to metasoma and usually less than 1.5X as long as hind tibia). Microdontomerus enigma Grissell Distribution.— NEARCTIC: Known only from one locality in Nevada, USA. Hosts. — Reared from Hoplitis bullifacies (Megachilidae). Discussion. — This species is phenotypi- cally nearly identical to M. parkcri. Char- acters to separate the two are given in the key. Somewhat more difficult to assess is that in M. enigma the longest diameter of the lateral ocellus is less than the ocellocular distance, whereas it is subequal to or greater than the distance in M. parkeri. Microdontomerus parkeri Grissell Distribution.— NEARCTIC: Widespread in the western and southwestern United States. Hosts. — Reared from Megachilidae: Ash- meadiella bigeloviae, AshineadieUa cubiceps, Ashmeadiella gillettei, AshineadieUa rufipes, Hoplitis bullifacies, Hoplitis palmarum, Mega- chile brevis, and Osmia marginata. Biologi/. — Microdontomerus parkeri is a gre- garious parasitoid within individual bee cells. The number of individuals ranged from 2 to 33 per cell, with an average of about 8-9. For these rearings the total number of M. parkeri specimens was 229 females and 125 males for a sex ratio of 1.8 to 1. Ten of these rearings contained no males (Grissell 2005). Discussion. — This species has also been reared from Ancistrocerus sp. and Leptoclii- lus sp. (Vespidae: Eumeninae). It is the most common and widespread species of Microdontomerus attacking bees. Motiodontomerus Westwood Recognition. — Monodontomerus is recog- nized by the straight anterior edge of the metapleuron (as in Fig. 2), the presence of a frenal line on the scutellum (as in Fig. 8), the hind femur with a single, apicoventral tooth (Figs 35, 37; though in one species this tooth is poorly defined, Fig. 36), and by the well developed occipital carina which is nearly horizontal on its dorsal margin and closer to the occipital foramen than to the hind ocelli (Fig. 18). Number of Species. — 32. Number Associated with Bees. — 19. Distribution. — The species of this genus are widespread throughout the Holarctic, and somewhat less common in the Neo- tropical (Cuba, Mexico, Colombia, Argen- tina) and Oriental (Sri Lanka, India, Paki- stan) regions. Hosts. — Numerous hosts are known for this genus including families in Diptera, Hymenoptera, and Lepidoptera. The pri- mary hosts are solitary aculeate bees and wasps, sawflies, and moths (including their tachinid and ichneumonid parasi- toids). An authentic record of Monodonto- merus (undetermined species) attacking social vespids {Mischocyt tarns; Litte 1979) in Arizona occurs in the literature, but voucher specimens are now lost (Litte, /'// litt.). Unfortunately, some species of Mono- dontomerus are extremely difficult to tell apart and as a consequence there have been many misidentifications resulting in incorrect host records for some species. For example, Monodontomerus aereus Walker has been reported from Megachile muraria (now = M. pariet /;m)(Constantineanu et al. 1956), but this would not be considered a host based on the majority of records, which are from Lepidoptera (Grissell 2000, Noyes 2003). Monodontomerus vicicellae (Walker), a common parasitoid of larval Volume 16, Number 2, 2007 243 Lepidoptera and sawflies, was reported to be reared from an ichneumonid parasitoid in the nest of Megachile "ramulorum Rond." (Rondani 1877), which is a nomem nudum. There are no other records from bees for this species and the host record is consid- ered to be incorrect. Similarly, Monodonto- merus minor (Ratzeberg), also a parasitoid of Lepidoptera and sawflies, has been reported from several bees, but while these records appear in lists (e.g., Herting 1977) they apparently have no basis in the primary literature. Discussion. — In the following section, summary data are documented in Grissell (2000) unless otherwise specified. Identifi- cation is often more easily based on male characters. Although females predominate in reared series, species have gregarious larvae and some males are almost always present. Monodontomerus acrostigmus Grissell Distribution.— NEARCTIC: Eastern Tex- as, USA. Hosts. — Reared from pupa of Megachile sp. (Megachilidae) in a "mud-dauber nest". Discussion. — Monodontomerus acrostigmus is similar in appearance to M. obscurus, but differs from it (and all other known species) by having the stigma posteriorly appendiculate (Figs 51, 52). In addition, it differs from M. obscurus by having the distal portion of the postmarginal vein one half or less than the proximal portion (subequal in M. obscurus) and in males, which have the face lateral to the scrobal basin distinctly depressed (not depressed in M. obscurus). Monodontomerus aeneus (Fabricius) Distribution.— NEARCTIC: Widespread throughout the northern United States and southern Canada. PALEARCTIC: Re- portedly widespread in western Europe (Nikol'skaya and Zerova 1978) and often confused with M. obscurus, which has the same distribution and general host range. Hosts. — There are a great number of hosts listed for this species (as obsoletus) in the Old World (see Grissell 1995). Only bee hosts are listed here because these are certainly correct whereas all other hosts are suspect. Old World: Anthophora retusa, Ceratina callosa (Apidae); Anthidium floren- tinum, Hoplitis (as Osmia) adunca, Megachile parietina (as Chalicodoma muraria) (and Stelis nasuta, a cleptoparasite of this host), Mega- chile apicalis, Megachile centuncularis, Mega- chile (as Chalicodoma) sicula, Osmia (as Metallinella) brevicornis, Osmia coerulescens, Osmia rufa cornigera, Osmia cornuta, Osmia emarginata, Osmia fulviventris, Osmia latreil- lei, Osmia rufa, Osmia submicans, Osmia tricornis (all Megachilidae). New World: Verifiable records for this species include Megachile concinna, Megachile rotundata, and Osmia nigrifrons (Megachilidae). Biology.— Newport (1849, 1852, 1853) provided information and illustrations of the larvae, their digestive tract, and feeding habits. Johansen and Eves (1966) and Eves (1970) (and possibly Hobbs and Krunic 1971) published biological information on this species (as obscurus, reidentified by me, based upon Eves' specimens) as a parasit- oid of Megachile rotundata. Females ovipos- ited through the leaf-lined cell and/or cocoon of the host. Between 3 and 51 eggs were laid externally on the host. An average of 10 survived in one study (Johansen and Eves 1966), but Bonelli and Campadelli (1990) gave an average of 24 (range = 10 to 51 adults for 15 bee cells). All immature stages of the host are vulnerable to attack but parasitization of early instars is rarely successful. Larvae are non-cannibalistic. The life cycle can be completed in about 20 days. Goodpasture (1975) detailed the mating behavior of M. aeneus (reported as M. obscurus, but sub- sequently confirmed as M. aeneus in Gris- sell 2000). Tepedino (1988a) demonstrated that 7-12% of females mated before emer- gence from the host cocoon. He also 244 Journal of Hymenoptera Research showed (Tepedino 1988b) that females had an initial obligatory requirement for host cocoon and prepupal authenticity, but after 24 hours this would break down and females would oviposit into gelatin cap- sules holding bee prepupae or even agar replicates of bees. Females oviposited onto fresh host prepupae or prepupae that were up to 16 days old. Tepedino (1988c) showed that superparasitism occurs but that rates go down as resident parasitoids become older. In Spain, rates of parasitism for M. aenens (reported as M. obsoletus) on Osmia cornuta (Megachilidae) varied from 0.5% (Bosch 1994b) to 73% (Bosch 1994a). According to Bosch (1993) 53-76% of managed bee cocoons were parasitized when paper straws containing bee cells were extracted from their nesting blocks, but cells left in grooved boards were left untouched. In the Nearctic this parasitoid (as M. obscurus) reportedly replaced the native species M. montivagus in the mid- 1960's as the most important parasitoid of the alfalfa leafcutting bee in North Amer- ica, but then was itself replaced by a pter- omalid in the mid 1970's (Eves 1982). A paper on control of an unknown species of Monodontomerus in Utah by Brindley (1976) undoubtedly refers to this species. Morphology. — Goodpasture illustrated the karyotype of M. obsoletus (1975, re- ported as M. obscurus but confirmed as M. aeneus by Grissell 2000). The chromosomes number 4 in males, 8 in females. Good- pasture (1975) illustrated male scapes, and Walther (1983) illustrated antennal sensil- lae of this species. Discussion. — This species was introduced into the Nearctic in the 1930's (Johansen and Eves 1966), but it was misidentified as M. obscurus. Its correct identity as M. obsoletus was reported by Tepedino (1989) based upon my identification. The name has since been changed to M. aeneus by Graham (1992) who studied the type material of the species involved. Almost all previously published host records (e.g., Peck 1969) for M. obscurus are wrong and most should now refer to M. aeneus. Both M. aeneus and M. obscurus are common and widespread and are among the two most difficult species of the genus to distinguish from each other. This is disconcerting because they are economically important, have both been introduced into the New World along with the alfalfa leafcutter bee, and have been confused with each other since their introductions. Only the appar- ent absence of sculpture (though faint coriaceous sculpture may be apparent at some angles of view) on the median frenal area and the construction of the frenal apex offer reliable diagnostic information to separate these two species, but even this can be difficult to interpret on occasion. An additional character that may sometimes help to define these two taxa is found in the mesepimeron. In M. aeneus the entire mesepimeron is essentially smooth (po- lished) except for some slight reticulation (or carinae) above the ventral margin. In M. obscurus the ventral 1/5 of the mesepi- meron below the transepimeral sulcus is reticulate and the anterior 1/3 is alutac- eous to lightly reticulate. Monodontomerus anthidiorum (Lucas) Distribution.— PALEARCTIC: Found only in Algeria. Host. — Reared from Rhodanthidiuiu sticti- cum (Megachilidae). Biology. — This species was reared from the larva of its host. According to Lucas (1849) the bee nested in empty snail shells (Helix spp). The larvae were gregarious with 40-50 specimens of M. anthidiorum found in each shell. Discussion. — This species apparently has not been collected since its original de- scription. In both sexes this is one of the most distinct species of the genus based on the elongated clypeus (Figs 23, 24). Monodontomerus argentinus Brethes. Distribution.— NEOTROPICAL: Costa Rica, Panama, Colombia, and Argentina. Volume 16, Number 2, 2007 245 Hosts. — Reared from cells of Eufriesea nigrescens (as Euplusia longipennis) (Apidae) in Colombia. A species oi Megachile (Mega- chilidae) also serves as host. Biology. — Sakagami and Sturm (1965) reported that this species developed on the pupal stage. Discussion. — Monodontomerus argentinus is similar to M. mexicanus especially in proportions of the head and antenna and in details of the wing, hi both sexes of M. argentinus the median area of the frenum is highly polished, whereas in M. mexicanus the median frenal area is longitudinally sculptured similar to the lateral areas. Monodontomerus bakeri Gahan Distribution.— NEARCTIC: Colorado, Utah, Idaho, USA, and Alberta, Canada. Hosts. — Megachile pugnata, Megachile re- lativa, Megachile rotundata, Osmia coloraden- sis, and Osmia texana (Megachilidae). Discussion. — This species is relatively uncommon, but large numbers were trapped from Megachile rotundata blocks as a nuisance species at the USDA Bee Biology and Systematics Laboratory in Logan, Utah (pers. obs.). Monodofitomerus bakeri is unique among species of the genus in two ways. The absence of a malar sulcus (Fig. 28), or its expression as a greatly curving, indefinite line (Fig. 27 ), is atypical compared to the straight, well-defined sulcus found in most other species (e.g., Fig. 26). Also, the bulging lower face (Fig. 27) is not found in any other species, all of which have the area essentially flat (as in Fig. 26). Monodontomerus brevicrus Grissell Distribution.— NEARCTIC: California, USA. Hosts. — Reared from nests of Osmia ribifloris (Megachilidae). Discussion. — Monodontomerus brevicrus resembles M. dianthidii in having metaso- mal tergum 2 dorsally sculptured, but it is separated as follows: Both sexes of M. brevicrus have the distal portion of the postmarginal vein about one-third the length of the proximal portion (Fig. 31) (about equal in M. dianthidii) and the rim of the scutellum apically even in width and not projecting (apically widened and some- what projecting in M. dianthidii). In females of M. brevicrus metasomal tergum 6 is obtuse in profile (as in Fig. 33) (acute in M. dianthidii, as in Fig. 32) The males of M. brevicrus are unique among New World males in modifications found in the fore leg and in the sunken lower face. In males the fore leg is reduced (Fig. 40) with the tibia shorter in length than the femur and ventrobasally concave, and the tarsomeres shortened with the claw length equal to tarsomeres 3 and 4 (fore leg unmodified in other species, cf. Fig. 39). Monodontomerus dementi Grissell Distribution.— NEARCTIC: Wyoming and Colorado, USA. Hosts. — Dianthidiinn heterulkei (Megachi- lidae) [also reared from the factitious host Megachile rotundata (Megachilidae) in the laboratory]. Biologxj. — Clement (1976) found this spe- cies feeding on prepupae in cocoons of D. heterulkei. Goodpasture (1975) described the mating behavior, which is identical to that of Monodontomerus montivagus. Discussion. — Monodontomerus dementi and M. laticoruis are similar in having the anterior half of the upper mesepimeral area reticulately sculptured and the anterodor- sal corner with diagonal striations extend- ing nearly to transepimeral sulcus (Fig. 53). They differ in the characters outlined in couplet 11 of the key. Monodontomerus dianthidii Gahan. Distribution. — NEARCTIC: Eastern Cali- fornia and southwestern Oregon, USA. Hosts. — Dianthidiinn sp. (Megachilidae). Biology. — Reared from resin nests. Discussion. — Monodontomerus dianthidii is phenetically most similar to A4. brevicrus 246 Journal of Hymenoptera Research based upon the completely sculptured frenal area and metasomal tergum 2; the differences between these species are discussed in detail under M. brevicrus above. Monodontomerus laticornis Grissell and Zerova Distribution.— PALEARCTIC: Russia, Kazakhstan, Ukraine, and Moldavia. Hosts. — Reared from Megachile rotundata (Megachilidae); Megachile centuncularis and Anthidium florentinum (Zerova and Stolbov 1986) (Megachilidae); Anthidium septemspi- nosum (Zerova and Seryogina (2002). [A report of Apis mellifera as host (documented in Noyes 2003) seems unlikely]. Biology. — This is a gregarious parasitoid within cocoons of the hosts. Discussion. — Zerova and Romasenko (1986) key and figure this species in a paper on the parasitoids of megachilid bees in the Former Soviet Union. This species and M. clement i are similar in appearance, and characters to distinguish them are given under couplet 11 of the key. Monodo)ito- merus laticornis is a Palearctic species and M. dementi a Nearctic one, so they should not be readily confused. Monodontomerus mandibularis Gahan Distribution.— NEARCTIC: Widespread throughout the eastern USA and Canada from Saskatchewan south to Louisiana. Hosts. — Anthophora abrupta, A. bomboides bomboides, Melitoma taurea (Apidae); Osmia cordata (Megachilidae) (Rau 1947). Biology. — Rau (1947) published some preliminary information on the life history of this species, which he concluded had one or two generations per year. He believed the wasp to be a primary, gregar- ious parasitoid of its host. Discussion. — Monodontomerus mandibu- laris is morphologically similar to M. montivagus but differs in both sexes (and from all other Monodontomerus species) by the mandibles having a single large, ventral tool and a smaller, subapical dorsal one (Fig. 48). Other species have two ventral teeth and a small subapical dorsal one (as in Fig. 47) or have the dorsal tooth so reduced as to be easily overlooked. The mandibles are not generally exposed, however, so that for practical purposes M. mandibularis is best distinguished from M. montivagus as follows: In females the intermalar distance is about 3x the malar distance (about 2X in M. montivagus; this is the result of the malar distance being relatively shorter in M. mandibularis and the face less produced ventrally below the eyes, cf. Figs 47, 48) and the posterior outline of metasomal tergum 6 is deeply concave (shallow in M. montivagus, cf. Figs 32, 33); in males the scape (Fig. 57) is laterally compressed and distinctly C- shaped in profile with dorsal and ventral arches asymmetrical (in M. montivagus the scape is dorsoventrally compressed and nearly symmetrically curved in profile, Fig. 56, sometimes greatly so). Monodontomerus mexicanus Gahan Distribution.— NEARCTIC/NEOTROPI- CAL: Spotty distribution in Arizona, north- central Mexico, and western Panama. Hosts. — Megachile peruviana (Megachili- dae) (Rau 1947); Ancyloscelis apiformis (as armata) (Torchio 1974) and Anthophora marginata (Apidae) (Herting 1977). Discussioji. — This species has also been reared from Trypoxylon mexicanum (Gahan 1941), T. monteverde, and Passaloecus ( = Polemistus) pusillus (Rau 1947) (all Crabro- nidae). It has been seen walking on the surface of Trypoxylon mud nests and drilling with its ovipositor through the mud walls (Brockmann in litt.). It is similar to M. argentinus and is discussed under that species. Monodontomerus montivagus Ashmead Distribution.— NEARCTIC: Widespread throughout southern Canada and USA. NEOTROPICAL: Southern Mexico (Guer- rero). Volume 16, Number 2, 2007 247 Hosts. — This species has been reared from the following bees. Apidae: Antho- phorn abrupta, Anthophora bomboides bom- boid.es, Anthophora bomboides neomexicana, Anthophora linsleyi, ? Anthophora occidentalis, lAnthophora vallorum, Bombns morrisoni, IMelissoides sp., Xylocopa tabaniformis orpi- fex. Megachilidae: Anthidium collectum, An- thidium emarginatum, Anthidiun Imormo- num, Anthidium nest, Ashmeadiella Califor- nia!, Dianthidium curvatum sayi, Dianthi- diinn pudicum pudicum, Dianthidium pudicum consimile, Hoplitis anthocopoides nest, Megachile centuncularis, Megachile re- lativa, Megachile rotundata, Osmia sp. cocoon (in Ttypoxylon politum nest [Crabronidae]), Osmia cordata, Osmia kincaidii, Osmia lati- sulcata, Osmia lignaria, Osmia ribifloris, Osmia sanrafaelae, Osmia texana, Stelis de- pressa. Biology. — This is a gregarious, external parasitoid of aculeate Hymenoptera. Al- though there are numerous references to this species in the literature (see Peck 1963), most of these are simply host records without biological data. A few papers cited by Peck are of interest and are cited below. Davidson (1893: 153) stated that females of M. montivagns deposited 10 to 20 eggs in each cell of Xylocopa tabaniformis orpifex and that some broods were all males while others were all females. Hicks (1926: 224) stated that M. montivagns was parasitic both on Anthophora occidentalis and its parasitoid Oryttus mirandus, thus acting as a primary and secondary parasitoid. Mickel (1928: 72-73) reared 415 specimens, of which 94% were females, from 21 cells of Anthophora occidentalis. He found no hyperparasitic relationship on the same bee host as reported by Hicks (1926). Linsley and MacSwain (1942: 409-411) also reported montivagns as both a primary and a hyperparasitoid on Anthophora linsleyi and its mutillid parasitoid Photopsis auraria (now = Sphaeropthalma itnicolor). These authors discussed the courtship behavior of montivagns and stated that its larvae fed on the prepupal stage of the bee. They stated that only one cell (of 9) had mixed sexes of this parasitoid, the others being either female (average 26 per cell) or male (average 40 per cell). In later rearings, however, MacSwain (1958: 395) found mixes of males and females in each of four cells of A. occidentalis. The sex ratio (males:females) varied from 1 to 12 to 1 to 30. Rau (1922) found a ratio of 1 to 6. Goodpasture (1975) described and illus- trated the courtship behavior of M. mon- tivagns. It is apparent from the literature and from reared specimens that M. mon- tivagns is parasitic on bees, wasps, and their nest associates. New and old nests of aculeate Hymenoptera are complex sites of diverse taxa, behaviorial types, and suc- cessional faunas. Therefore, our biological knowledge of M. montivagns is almost wholly inadequate. Morphology. — Goodpasture (1975) de- scribed and illustrated the male scapes and the haploid karyotype. This species has 6 chromosomes in males, 12 in females. Discussion. — Females of M. montivagus are morphologically similar to other spe- cies reared from bees (e.g., M. parkeri, M. tepedinoi, M. torchioi, M. mandibularis), but males differ notably in morphology of the scape. The differences between M. monti- vagns and the others mentioned are dis- cussed under each of these species. Monodontomerus obscurus Westwood Distribution.— NEARCTIC: Widespread from coast to coast in the United States and southeastern Canada. [Undoubtedly introduced into the Nearctic along with its host the alfalfa leafcutting bee.] PALEARC- TIC: Reportedly widespread in western Europe (Nikol'skaya and Zerova 1978) and probably often confused with M. aenens which appears to be sympatric. The species is also reported from the oriental Region (India). Hosts. — Hoplitis (as Osmia) adunca, Mega- chile argentata, Megachile centuncularis, Megachile cephalotes, Megachile flavipes, 248 Journal of Hymenoptera Research Megachile lanata, Megachile parietina (as Ckalicodoma muraria), Megachile rotundata, M. willughbiella, Osmia cordata, Osmia corni- frons, Osmia latreillei, O. lignaria, Osmia ribfloris, Osmia rufa rufa, Osmia rufa corni- gera, Osmia sanrafaelae, (Megachilidae); An- thophora plumipes, Xylocopa fenestrata (Api- dae). Biology. — In Spain, M. obscurus is consid- ered to be extremely destructive to the alfalfa leafcutting bee industry and chemi- cal methods of control have been devised (Asensio 1982). Krunic and Radovic (1973) reported that M. obscurus can go through a number of generations without diapause and that diapause could be interrupted after keeping them for a time at 5 C. Morphology. — Radu and Botoc (1968) illustrated female genitalia in detail. Mac- Donald and Krunic (1971) illustrated the somatic chromosomes for M. obscurus, which number 6 in males and 12 in females. (This differs from M. aeueus and thus strengthens the case for reproductive isolation between these two nearly identi- cal species.) Baker et al. (1985) described and illustrated the last instar larva and pupa of this species (adult identity con- firmed by examination of voucher speci- mens in North Carolina State University Insect Collection). Discussion. — Zerova and Romasenko (1986) key and figure this species in a paper on the parasitoids of megachilid bees in the former Soviet Union. This species is similar to M. aeueus and is often reared from the same species of host in the same locality. I discuss the two species more fully under M. aeueus, above. Monodontomerus osmiae Kamijo D/sfn7?Hfz'on.— PALEARCTIC: Known from Japan and the Russian Far East and introduced into the Nearctic (Grissell 2003). Hosts. — Osmia cornifrons, Osmia excavata, and Osmia taurus, (Megachilidae) (Kamijo 1963, 1965). Biology. — Iwata and Tachikawa (1966) reported a preponderance of females for rearings of this species from Osmia taurus. From 61 cocoons emerged 87 males and 726 females. The number of parasitoids per host (counted for 4 cocoons only) varied from 14 to 26. Discussion. — Zerova and Romasenko (1986) key and figure this species in a paper on the parasitoids of megachilid bees in the former Soviet Union. Grissell (2003) illus- trated the peculiar male head of this species based on specimens collected in Silver Spring, Montgomery County, Mary- land, USA. Monodontomerus japonicus Ash- mead was reported from Osmia taurus, but this record is undoubtedly a misidentifica- tion of M. osmiae (Grissell 1995). Males of this species are easily identified by the peculiar head (Figs 21, 22). Females have the discal area entirely setose. Monodontomerus parkeri Grissell Distribution. — NEARCTIC: Known from widespread localities in western North America stretching from Alberta, Canada to New Mexico, USA. Host. — Anthophora occidentalis (Apidae). Discussion. — Monodontomerus parkeri ap- pears most similar to M. tepedinoi and their separation is discussed under the latter species. This species is also easily confused with M. montivagus. Males of the two species may be readily distinguished by comparing scapes: in M. parkeri the scape is laterally compressed and asymmetrically bent with the apex enlarged and a polished, depressed area on its outer side that continues onto the ventral surface (some- what as in Fig. 57); in M. montivagus the scape is dorsoventrally compressed, sym- metrically bent, and has the polished area completely ventral (Fig. 56). Additionally, in both sexes of M. parkeri, the apex of the costal cell dorsally has few setae (0 to 3 as in Fig. 43) whereas in M. montivagus there is a dorsal row of setae in the apical half to quarter (as in Fig. 31), and the transepim- Volume 16, Number 2, 2007 249 eral sulcus is complete (Fig. 54), whereas in M. montivagus it is not (Fig. 55). There are several less obvious and more relative characters that are difficult to use without comparative material. In M. parkeri the frenal area appears medially polished un- der reflected light even though it is sculptured, whereas in M. montivagus this area is generally entirely sculptured. In M. parkeri the admarginal wing area contains a few, widespaced setae (Fig. 44), and relatively few setae (3 to 5) are directly adjacent to the marginal vein (so that there is no setal row parallel to the vein), but in M. montivagus this area is evenly setose to the marginal vein (as in Fig. 43); there are en- ough setae to form a row parallel to the vein. Monodontomerus rugulosus Thomson Distribution— PALEARCTIC: Wide- spread in western and central Europe (Zerova and Seryogina 2002). Hosts. — This species has been reared from Megachile rotundata (Megachilidae) (Zerova and Romasenko 1986). Biology. — This is a gregarious parasitoid in cocoons of its host. Discussion. — Monodontomerus rugulosus appears quite similar to M. argentinus, but the two species occur in different, widely spaced zoogeographic regions. They may be separated by characters given in the key. Monodontomerus tepedinoi Grissell Distribution. — NEARCTIC: Known from Oregon and Utah, USA. Hosts. — The species has been reared from Osmia lignaria (Megachilidae). Discussion. — Females of M. tepedinoi are easily confused with M. montivagus and M. parkeri. From M. montivagus it is most readily separated by the upper anterior margin of the costal cell with only 1 to 3 setae at the apex (as in Fig. 43), whereas in M. montiva- gus the upper anterior margin has a setal row in its apical 1/4 to 1/3 (as in Fig. 31). From M. parkeri it is separated by the longer ovipositor (ca. 1.5 to almost 2X the metaso- mal length; 1 to 1.2 X in M. parkeri) and by the scape, which has some metallic green color at least ventrally (all yellow to orange in M. parkeri). Males of M. tepedinoi are easier to distinguish than females based on the antenna as described in the key and compared in Figs 56, 57, 58). Monodonto- merus tepedinoi is so far associated only with Megachilidae and M. parkeri with Apidae. Monodontomerus thorpi Grissell Distribution. — NEARCTIC: Known from isolated localities in southern California, Arizona, and western Texas, USA. Hosts. — Reared from nests of Anthidium maculatum (Megachilidae). Discussion. — This species has been reared from twig nests in the eastern and western extremes of its distribution. It is one of the easiest species of the genus to identify in both sexes as it is the only species to have the first two flagellar segments reduced (i.e., ring-like, Fig. 19), whereas all other species have only the first segment reduced (Fig. 20). Additionally, the hind femur is enlarged with only a ventral angle (Fig. 36) as opposed to other species that have a distinct tooth (Figs 37, 38). Monodontomerus torchioi Grissell Distribution. — NEARCTIC: The species is known only from Utah, USA. Hosts. — Reared from nests of Osmia lignaria and O. sanrafaelae (Megachilidae). Discussion. — Monodotitomerus torchioi is easily confused with M. montivagus, M. tepedioni, and M. parkeri in females. The diagnostic characters used to separate these three taxa are given in key couplets 15 and 24-25 and under the discussion of the species mentioned. Pseudotorymus Masi Recognition. — Anterior margin of meta- pleuron straight (as in Fig. 2); occipital carina medially arched and midway be- 250 Journal of Hymenoptera Research tween the hind ocelli and occipital foramen (as in Fig. 17); hind femur ventrally with a slight indication of a tooth; marginal vein long, 3 to 7x length of postmarginal vein and at least 6X length of stigmal vein. Number of Species. — 43. Number Associated with Bees. — 1 (ques- tionably). Distribution. — The genus is most abun- dant in the Palearctic Region (30 species) where its species are widespread and extend into northern Africa. It is also known from the Afrotropical Region (7 species) from Madagascar, Mali, Mozam- bique, Nigeria, Rwanda (Republic of the Congo), Senegal, South Africa, and Sudan. There are 4 species known from India in the Oriental Region and a single, widespread species is known from the Nearctic (southern Canada and northern USA). Hosts. — Members have a broad host association including Curculionidae (Co- leoptera) in leguminous seed pods; Bruchi- dae (Coleoptera) from galls on Asteraceae, Combretaceae, Fabaceae, Orchidaceae, and Rubiaceae; Cecidomyiidae (Diptera) asso- ciated with Apiaceae, Cruciferae, Fabaceae, Lamiaceae, Rosaceae, Salicaceae, and Scro- phulariaceae; Eurytomidae (Hymenoptera) in grass stems (Poaceae); Cynipidae (Hy- menoptera) in pods of Papaveraceae; Ten- thredinidae (Hymenoptera); and Pyralidae (Lepidoptera). Discussion. — The inclusion of this genus in relation to bee hosts is highly question- able and is based upon the single record for P. indicus as indicated below. Among the other 42 known species of Pseudotor- ymus the use of bees is unknown so this record is likely to be incorrect. Pseudotorymus indicus (Marti) Distribution. — This species is known only from southern India (Uttar Pradesh, Tamil Nadu) (Mani 1989). Hosts. — The type series was reared from "flower bud galls" on Dalbergia sissoo (Fabaceae). Mani (1989) listed the host as a "leafcutting bee". Discussion. — In light of the original rear- ing and the entire host range given above, I am inclined to dismiss this record until it can be reconfirmed. Torytnus Walker Recognition. — This genus is easily recog- nized by the anterior edge of the meta- pleuron (usually its upper half) projecting forward as a lobe into the mesepimeron which is subdivided into upper and lower sections, the lower of which is delimited by an anterior groove (Fig. 1, compare with Fig. 2, arrows). Number of Species. — Approximately 375. Number Associated with Bees. — 3. Distribution. — All zoogeographic regions except Australia where it was apparently introduced (Grissell 1995). Hosts of Genus. — Members of this genus are mostly parasitoids of larvae of gall- forming Diptera and Hymenoptera. A few have been reared from bees, and a few are phytophagous in seeds. Discussion. — Until 1998 the species that parasitized bees were treated as the genus Diomorus Walker. Graham and Gijswijt (1998) synonymized Diomorus under Tor- ymus, dividing its members into several species groups of the latter. Ton/nuts armatus (Boheman) Distribution. — This species is widespread in the Palearctic, being reported from Europe (Graham and Gijswijt 1998) and Japan (Kamijo 1979). It was possibly in- troduced into Papua New Guinea (Boucek 1988). Hosts. — Kamijo (1979) reported T. arma- tus from Ceratina japojiica (Apidae) in Rubus twigs (Rosaceae) in Japan. Discussion. — This species has reportedly been reared from several genera of Crab- ronidae, including Rliopalum (Box 1920) and Crossocerus (Gijswijt 1974), and seems to be associated with wasps and bees that Volume 16, Number 2, 2007 251 nest in the stems of Rubus (Graham and Gijswijt 1998). It is the most distinct of the three Torymus species known from bees, having the hind coxa dorsally bare and smooth, and the propodeum without cari- nae. Toymus cupreus (Spinola) Distribution. — The species is widespread in the Palearctic (Nikol'skaya and Zerova 1978), mostly in the "southern parts and middle of Europe" and reaching into the Netherlands (Graham and Gijswijt 1998). It is reported from Burma in the Oriental Region (Mani and Kaul 1972). Host. — The original hosts given by Spi- nola included 7 species of cynipid galls, but these all probably housed aculeate bees or wasps. Mani and Kaul (1972) reported the species as "widely distributed as [a] parasitoid of Osmia sp. (Megachilidae) and Sphecidae." Biology. — Enslin (1922) illustrated and discussed the larval and pupal stages of this species (as Diomorus kollari). Discussion. — This Palearctic species and the following Nearctic species are geo- graphically separated but show no mor- phological differences. In coloration, how- ever, they are distinct as explained in key couplet 3. Torymus zabriskii (Cresson) Distribution. — The species is widespread in the United States. Hosts. — The only reported bee host is Ceratina dupla (Apidae) (Zabriskei 1890). Biology. — Krombein (1964) reported some short biological notes on this species (as Diomorus) as a parasitoid of Ectemnius paucimaculatus (Crabronidae). He sug- gested that T. zabriskii parasitized several cells in a succession of cells and that oviposition was probably through the wall of the plant stem {Hibiscus: Malvaceae) in which the wasp nested. Discussion. — Ceratina, the only reported bee host (Zabriskei 1890), has been listed in the secondary literature several times but has never been reconfirmed. A number of other hosts in the family Crabronidae have been reported for this species including Ectemnius, Crossocerus, and Rhopalum (sum- marized by Grissell 1995). ACKNOWLEDGMENTS 1 thank Robert Matthews, University of Georgia, Athens, for supplying unpublished collecting and rearing information for Echthrodape papuana from Australia and for donating a specimen to the U. S. National Museum of Natural History. I especially thank Terry Griswold, USDA Bee Biology and Systematica Laboratory, Logan, Utah, for checking the bee names used in this paper. His help has been most appreciated, but I remain responsible for errors in their subsequent use. For reading the host list and offering suggestions on host data I am grateful to Jerome Rozen, American Museum of Natural History, New York, and Frank Parker (retired), USDA Bee Biology and Systematics Laboratory. For additional help with information regarding hosts I thank John Brown, Systematic Entomology Laboratory, Washing- ton, DC. For reading the manuscript and offering positive criticism I thank Sam Droege, U. S. Geological Survey, Laurel, Maryland, and Norman Woodley, Thomas Henry, and David Nickle, Systematic Ento- mology Laboratory. Additionally, the editor of journal of Hymenoptera Research and several reviewers have provided a number of suggestions and corrections to the manuscript for which I am grateful. LITERATURE CITED Asensio, E. 1982. Leafcutter bee management in Spain: problems of parasitism. Pp. 71-79 in: G. H. Rank. Proceedings of the first international symposium on alfalfa leafcutting bee management. Saskatoon, Sas- katchewan, Canada: University of Saskatchewan Printing Services. 281 pp. Baker, J. R., S. B. Kuhn, and E. D. Bambara. 1985. Nests and immature stages of leafcutter bees (Hyme- noptera: Megachilidae). Journal of the Kansas Entomological Society 58: 290-313. Baur, H. and F. Amiet. 2000. The Leucospidae (Hymenoptera: Chalcidoidea) of Switzerland, with a key and data on the European species. Revue Suisse de Zoologie 107: 359-388. Bonelli, B. and G. Campadelli. 1990. Note biologiche su Chalicodoma parietina Geoffr. (Hymenoptera- Megachilidae). Bollettino dell'Istituto di Entomolo- gia "Guido Grandi" delta Universita degli Studi di Bologna 44: 1-9. Bosch, J. 1993. Parasitism in wild and managed populations of the almond pollinator Osmia cornuta Latr. (Hymenoptera: Megachilidae). Jour- 252 Journal of Hymenoptera Research nal of Agricultural Research 31: 77-82. (Journal dated 1992). — . 1994a. Osmia cornuta Latr. (Hym., Megachili- dae) as a potential pollinator in almond orchards: Releasing methods and nest-hole length. Journal of Applied Entomology 117: 151-157. 1994b. Improvement of field management of Osmia cornuta (Latreille) (Hymenoptera, Mega- chilidae) to pollinate almond. Apidologie 25: 71-83. Boucek, Z. 1974. A revision of the Leucospidae (Hymenoptera: Chalcidoidea) of the world. Bul- letin of the British Museum (Natural History), Entomology, Supplement 23: 1-241. . 1988. Australasian Chalcidoidea (Hymenoptera). Wallingford, UK: C. A. B. International. 832 pp. Box, L. A. 1920. [Note with exhibit of Hymenoptera]. Proceedings of the South London Entomological and Natural History Society 1919: 50. Brii.dley, W. A. 1976. Carbaryl control of chalcidoid parasitoids from alfalfa leafcutting bees. Journal of Economic Entomology 69: 225-228. Burks, B. D. 1969. New Perilampidae (Hymenoptera: Chalcidoidea). Proceedings of the Entomological Society of Washington 71: 73-81. Clement, S. L. 1976. The biology of Dianthidium heterulkei heterulkei Schwarz, with a description of the larva (Hymenoptera: Megachilidae). Was- mann Journal of Biology 34: 9-22. Constantineanu, M. I., I. Suciu, I. Andreiescu, V. Ciochia, and C. Pisica. 1956. Contributii la studiul Chalcidoidelor din R. P. R. Chalcidoide (Chalci- doidea Ashmead 1899) parazite in albilitia pru- nului (Aporia crataegi L.) din Moldova. Analele Stiintifice Universitatii lasi (New Series) 2: 113-126. Davidson, A. 1893. The nest and parasitoids of Xylocopa orpifex, Smith [sic]. Entomological News 4:151-153. David, P. M. M., A. M. Hanifa, and S. Natarajan. 1990. Biology and control of blossom midge Contarinia sp. (Diptera: Cecidomyiidae) on Jasminum sambac in Tamil Nadu. Entomon 15: 193-196. Enslin, E. 1922. Zur Biologie des Solonius rubicola Duf. et Perr, (larvatus Wesm.) und seiner Parasiten. Konowia 1: 1-15. Eves, J. D. 1970. Biology of Monodontomerus obscurus Westwood a parasitoid of the alfalfa leafcutting bee, Megachile rotundata (Fabricius) (Hymenop- tera: Torymidae; Megachilidae). Melanderia 4: 1-18. . 1982. Evolution of predators, nest destroyers, and parasitoids of the alfalfa leafcutting bee in the Pacific Northwest. Pp. 57-59 in: G. H. Rank. Proceedings of the first international symposium on alfalfa leafcutting bee management. Saskatoon, Sas- katchewan, Canada: University of Saskatchewan Printing Services. 281 pp. Farooqi, S. I. 1986. Family Torymidae. Pp. 259-277 in: B. R. Subba Rao, and M. Hayat, eds. The Chalcidoidea (Insecta, Hymenoptera) of India and the adjacent countries. Pt. II. A Catalogue. Oriental Insects 20: 1-430. Gahan, A. B. 1941. A revision of the chalcid-flies of the genus Monodontomerus in the United States National Museum. Proceedings of the United States National Museum 90: 461-482. Gijswijt, M. J. 1974. Faunistical and biological records of some palaerctic Chalcidoidea. Entomologische Berichten (Amsterdam) 34: 89-93. Goodpasture, C. E. 1975. Comparative courtship behavior and karyology in Monodontomerus (Hy- menoptera: Torymidae). Annals of the Entomolog- ical Society of America 68: 391-397. Graham, M. W. R. 1992. Diagnostic characters of Monodontomerus obscurus Westwood, 1833 and M. aeneus (Fonscolombe, 1832) (= obsoletus (Fabri- cius, 1798), preocc.) (Hym., Torymidae). The Entomologist's Monthly Magazine 128: 143-146. and M. J. Gijswijt. 1998. Revision of the European species of Torymus Dalman (s. lat.) (Hymenoptera: Torymidae). Zoologische Verhande- lingen Leiden 317: 1-202. Grissell, E. E. 1995. Toryminae (Hymenoptera: Chalci- doidea: Toryminae): a redefinition, generic clas- sification, and annotated world catalog of species. Memoirs on Entomology, International 2: 1-470. — . 2000. A Revision of New World Monodonto- merus Westwood (Hymenoptera: Chalcidoidea: Torymidae). Contributions of the American Entomo- logical Institute 32 (1): 1-90. . 2003. Discovery of Monodontomerus osmiae Kamijo (Hymenoptera: Torymidae) in the New World. Proceedings of the Entomological Society of Washington 105: 243-245. 2005. A review of North American species of Microdontomerus Crawford (Torymidae: Hyme- noptera). Journal of Hymenoptera Research 14: 22-65. Herting, B. 1977. Hymenoptera. A catalogue of parasites and predators of terrestrial arthropods. Section A. Host or Prey/Enemy 4: 176. Commonwealth Agri- cultural Bureaux, Institute of Biological Control. Hesami, S., M. A. Akrami, and H. Baur. 2005. Leucospis dorsigera Fabricius (Hymenoptera, Leucospidae) as a hyperparasitoid of Cerambycidae (Coleop- tera) through Xoridinae (Hymenoptera: Ichneu- monidae) in Iran. Journal of Hymenoptera Research 14: 66-68. Hicks, C. H. 1926. Nesting habits and parasitoids of certain bees of Boulder County, Colorado. The University of Colorado Studies (Series A) 15: 217-252. Hobbs, G. A. and M. D. Krunic. 1971. Comparative behavior of three chalcidoid parasitoids of the alfalfa leafcutter bee, Megachile rotundata, in the laboratory. The Canadian Entomologist 103: 674-685. Iwata, K. and T. Tachikawa. 1966. Biological observa- tions on 53 species of the superfamilies, Chalci- Volume 16, Number 2, 2007 253 doidea and Proctotrupoidea, from Japan (Hyme- noptera: Apocrita) [sic]. Transactions of the Shikoku Entomological Society 9: 1-29. Johansen, C. and J. Eves. 1966. parasitoids and nest destroyers of the alfalfa leafcutting bee. Washing- ton Agricultural Experiment Station, Circular 469: 1-12.' Kamijo, K. 1963. A revision of the species of the Monodontomerinae occurring in Japan (Hyme- noptera: Chalcidoidea) [Taxonomic studies on the Torymidae of Japan, II]. Insecta Matsumurana 26: 89-98. . 1965. A new host-record of Monodontomerus osimae Kamijo. Insecta Matsumurana 28: 78. 1979. Four new species of Torymidae from Japan, with notes on two known species. Akitu (N. S.) 24: 1-11. Krombein, K. V. 1964. Natural History of Plummers Island, Maryland. XVIII. The hibiscus wasp, an abundant rarity, and its associates (Hymenoptera: Sphecidae). Proceedings of the Biological Society of Washington 77: 73-112. Krunic, M. D. and I. Radovic. 1973. Some aspects of cold-hardiness of the parasitoids Monodontomerus obscurus, Pteromalus venustus and Melittobia chaly- bii (Chalcidoidea, Hymenoptera) and their host Megachile rotundata (Megachilidae, Hymenop- tera). Arhiv Bioloskih Nauka, Beograd 25: 45-48. Linsley, E. G. and J. W. MacSwain. 1942. The parasitoids, predators, and inquiline associates of Anthophora linslei/i. American Midland Naturalist 27: 402-417. Litte, M. 1979. Mischocytarrus flavitarsis in Arizona: social and nesting biology of a polistine wasp. Zeitschrift fiir Rierpsychologie 50: 282-312. Lucas, H. 1849. Histoire naturelle des animaux articules. Insectes. Exploration Scientifique de l'AT gerie: Sciences Physiques Zoologie. Vol. 3. Paris: Bertrand. 403 p. MacDonald, M. D. and M. D. Krunic. 1971. Chromo- some numbers of Monodontomerus obscurus and Pteromalus venestus, chalcid parasitoids of Mega- chile rotundata. Arhiv Bioloskih Nauka, Beograd 23: 93. MacSwain, J. W. 1958. Taxonomic and biological observation on the genus Hornia (Coleoptera: Meloidae). Annals of the Entomological Society of America 51: 391-396. Mani, M. S. 1989. The fauna of India and adjacent countries. Chalcidoidea (Hymenoptera). Pt. I and II. Madras, India: Zoological Survey of India. 1633 p. and B. K. Kaul. 1972. Some Torymidae (Hymenoptera: Chalcidoidea) from India. Orien- tal Insects 6: 313-331. Michener, C. D. 1969. Immature stages of a chalcidoid parasitoid tended by allodapine bees (Hymenop- tera: Perilampidae and Apidae). journal of the Kansas Entomological Society 42: 247-250. Michener, D. C. 2000. The bees of the world. The Johns Hopkins University Press, Baltimore Mickel, C. E. 1928. The biotic factors in the environ- mental resistance of Anthophora occidentalis Cress. (Hym.: Apidae; Dip., Coleop.). Entomological News 39: 69-78. Newport, G. 1849. On the anatomy and development of certain Chalcididae and Ichneumonidae, com- pared with their special economy and instincts; with descriptions of a new genus and species of bee parasitoid. Part I. Proceedings of the Linnean Society of Eondon 2: 23-26. [November]. . 1852. The anatomy and development of certain Chalcididae and Ichneumonidae, com- pared with their special oeconomy of instincts; and descriptions of a new genus and species of bee parasitoids. The Transactions of the Linnean Society of London 21: 61-77. [May or June]. . 1853. Further observations on the habits of Monodontomerus; with some account of a new Acarus (Heteropus ivntricosus), a parasitoid in the nests of Anthophora retusa. The Transactions of the Linnean Society of London 21: 95-102. [June]. Nikol'skaya, M. N. and M. D. Zerova. 1978. Family Torymidae (Callimomidae) (Torymids). Pp. 651-685 in: G. S. Medvedev, ed. Keys to the Insects of the European part of the USSR, Vol. 3, Part 2. New Delhi [Translated from the Russian by Amerind Published Co., 1987]. 1341 pp. Noyes, J. S. 1994. The reliability of published host- parasitoid records: a taxonomist's view. Nonve- gian journal of Agricultural Sciences 16: 59-69. . 2003. Universal Chalcidoidea Database. World Wide Web electronic publication, www.nhm.ac. uk/entomology/chalcidoids/ index.html [ac- cessed 10-Sep-2003]. Peck, O. 1963. A catalogue of the Nearctic Chalcidoi- dea. The Canadian Entomologist Supplement 30: 1-1092. . 1969. Chalcidoid parasitoids of the alfalfa leaf- cutter bee, Megachile rotundata, in Canada. The Canadian Entomologist 101: 418-422. Radu, V. G. and M. Botoc. 1968. Contributii la studiul constitutiei morfofunctionale a armaturii genitale femele la calcidoide. Babes-Bolyai, Cluj (Ser. Biol.) 13: 71-76. Rau, P. 1922. Ecological and behavior notes on Missouri insects. Transactions of the Academy of Science of St. Louis 24: 1-71 pis. 5-8. . 1947. Bionomics of Monodontomerus mandibu- laris Gahan, with notes on other chalcids of the same genus. Antials of the Entomological Society of America 40: 221-6. Rondani, C. 1877. Vesparia parasita non vel minus cognita. Bollettino della Societa Entomologica Italiana 9: 166-213. Rust, R. W. 1974. The systematics and biology of the genus Osmia, subgenera Osmia, Chalcosmia, and 254 Journal of Hymenoptera Research Cephalosmia (Hymenoptera: Megachilidae). Was- mann Journal of Biology 32: 1-93. Sakagami, S. F. and H. Sturm. 1965. Euplusia long- ipeiuiis (Friese) und ihre merkwurdigen brutzel- len aus Kolumbien. Insecta Matsumurana 28: 83-92. Steffan, J. R. 1964. Deux nouveaux genres de Torymi- dae et de Cleonymidae Palearctiques. Entomo- phaga 9: 101-106. Stolbov, N. M., S. M. Palevych, M. D. Zerova, O. F. Grobov, T. V. Konovalov, and O. P. Shatskaya. 1986. [Recommendations for the study of the parasitic Chalcidoidea of Megachile rotundata.] [In Russian.] Agricultural Academy of the USSR, Moscow Tepedino, V. J. 1988a. Incidence of pre-emergence sib- mating in Monodontomerus obsoletus, Pteromalus venustus, and Tetrastichus megachilidis, three chal- cid parasitoids of the alfalfa leafcutting bee, Megachile rotundata (Hymenoptera: Chalcididae). Pan-Pacific Entomologist 64: 63-66. . 1988b. Aspects of host acceptance by Pteroma- lus venustus Walker and Monodontomerus obsoletus Fabricius, parasitoids of Megachile rotundata (Fabricius), the alfalfa leafcutting bee (Hymenop- tera: Chalcididae). Pan-Pacific Entomologist 64: 67-71. . 1988c. Host discrimination in Monodontomerus obsoletus Fabricius (Hymenoptera: Torymidae), a parasitoid of the alfalfa leafcutting bee Megachile rotundata (Fabricius) (Hymenoptera: Megachili- dae). journal of the New York Entomological Society 96: 113-118. . 1989. Repulsion of three chalcid parasitoids of the alfalfa leafcutting bee, Megachile rotundata (F.) by the fungicides CAPTAN and DPX. Pan-Pacific Entomologist 65: 403-409. Torchio, P. F. 1974. Notes on the biology of Ancyloscelis annata Smith and comparisons with other antho- phorine bees (Hymenoptera: Apidae). Journal of the Kansas Entomological Society 47: 54-62. Walther, J. R. 1983. Antennal patterns of sensilla of the Hymenoptera — a complex character of phy- logenetic reconstruction. Verhandlungen der Ver- eins fiir Naturzvissenschaftliche, Hamburg 26: 373-392. Zabriskei, J. L. 1890. Society notes. Entomologica Americana 6: 100. Zerova, M. D. and L. P. Romasenko. 1986. [Chalci- doids (Hymenoptera, Chalcidoidea), parasitoids of leafcutting bees (Hymenoptera, Megachilidae), in the fauna of the USSR]. [In Russian]. Zoolo- gicheskii Zhurnal 65: 1631-1643. - and L. Ya. Seryogina. 2002. A revision of Old World Monodontomerus (Hymenoptera: Chalci- doidea: Torymidae). National Academy of Sciences of Ukraine, I. I. Schmalhausen Institute of Zoology, Separate Issue, Kiev. 1-74. - and N. M. Stolbov. 1986. New data on the trophic relation of Monodontomerus laticornis Griss. et Zer. (Hymenoptera, Torymidae). Vestnik Zoologii, Kiev 1986 (5): 86-87. APPENDIX Two lists are presented: bee host/chalci- doid and chalcidoid/bee host. The bee host list presents names as they currently are applied in the literature (i.e., valid names), not as they were originally published. The lists are derived from Boucek (1974), Noyes (2003), and Grissell (1995, 2000, 2005). Authors' names are given for bee host in the first list; chalcidoid authors are given in the subsequent list. The placement of bee genera in families is based on an electronic version (http://faculty.ucr.edu/~heraty/ beepage.html) of Michener (2000). Bee Host/Chalcidoid Apidae Allodape exoloma Strand: Xylencyrtus tridens Allodape mucronata Smith: Xylencyrtus tridens Allodape panurgoid.es Smith: Xylencyrtus tridens Allodape rufogastra Lepeletier and Serville: Xy- lencyrtus tridens Allodapula grandiceps (Friese): Xylencyrtus tridens Allodapula melanopus (Cameron): Xylencyrtus mumifex Ancyloscelis apiformis (F.): Monodontomerus mex- i can us Anthophora abrupta Say: Melittobia acasta, Melit- tobia megachilis, Pediobius williamsoni, Mono- dontomerus mandibular is, Monodontomerus montivagus Anthophora bomboides bomboides Kirby: Leucospis gigas, Monodontomerus mandibularis, Monodon- tomerus montivagus Anthophora bomboides neomexicana Cockerell: Monodontomerus montivagus Anthophora liusleyi Timberlake: Monodontomerus montivagus Anthophora marginata Smith: Monodontomerus mexicanus Anthophora occidentalis Cresson: Monodontomerus montivagus, Monodontomerus parkeri Anthophora plumipes (Pallas): Monodontomerus obscurus Volume 16, Number 2, 2007 255 Anthophora retusa (L.): Melittobia acasta, Melitto- bia pelopoei, Monodontomerus aeneus 1 Anthophora vallorum (Cockerell): Monodonto- merus montivagus Apis cerana (¥.): Antrocephalus sp. Apis mellifera L.: Dibrachys boanniac, Melittobia aeasta, Monodontomerus laticornis, Nasonia vi- tripennis, Pteromalus apum, Tetrastichus ho- wardi Bombus agrorum ¥.: Pteromalus conopidarum Bombus amerieanorum ¥:. Pediobius williamsoni Bombus atratus Franklin: Pediobius williamsoni Bombus fervidus F.: Melittobia chalybii Bombus hortorum (L.): Melittobia aeasta Bombus lapidarius (L.): Pteromalus conopidarum Bombus morrisoni Cresson: Monodontomerus mon- tivagus Bombus ruderatus (¥.): Melittobia aeasta Bombus sp.: Dibrachys cavus, Melittobia haioaiien- sis, Pachyerepoideus vindemmiae Bombus terrestris (L.): Melittobia aeasta Braunsapis leptozonia (Vachal): Xylencyrtus tri- dens Braunsapis rolini (Vachal): Echthrodape africaua Braunsapis rufipes (Friese): Echthrodape africaua Braunsapis simplicipes Michener: Echthrodape af- ricaua Braunsapis unicolor Smith: Echthrodape papuana Centris analis ¥.: Leucospis cayennensis Centris bicornuta Mocsary: Leucospis cayennensis Centris nitida Smith: Leucospis cayennensis Centris tarsata (Smith): Leucospis cayennensis Centris vittata Lepeletier: Leucospis cayennensis Ceratina calcarata Robertson: Axima zabriskiei Ceratina callosa (¥.): Eurytoma nodularis, Mono- dontomerus aeneus Ceratina dallatorreana Friese: Eurytoma apiculae Ceratina dupla Say: Axima zabriskiei, Baryscapus amerieanus, Habritys latrus, Torymus zabriskii Ceratina flavipes Smith: Neochalcis breviceps Ceratina ignara Cresson: Baryscapus amerieanus Ceratina japonica Cockerell: Cleonymus ceratinae, Torymus armatus Ceratina nanula Cockerell: Baryscapus amerieanus, Eurytoma apiculae Ceratina punctigena Cockerell: Eurytoma apiculae Ceratina sequoiac Michener: Baryscapus ameri- eanus Ceratina sp.: Chciloneurus leptulus, Epistenia coeruleata, Melittobia megachilis, Micrapion da- lyi, Micrapion nasutum, Micrapion richardsi Ceratina truncata Friese: Micrapion steffaui ICtenoplectra chalybea Smith: Leucospis histrio Diadasina distincta (Holmberg): Leucospis genalis Eufriesea nigrescens (Friese): Monodontomerus argentinus Euglossa annectans Dressier: Melittobia sp. Euglossa ignita Smith: Polistomorpha fasciata Euglossa sp.: Polistomorpha couura, Polistomorpha fasciata Eulaema meriana (Oliver): Leucospis pinna ?Melissodes sp.: Monodontomerus montivagus Melitoma taurea (Say): Monodontomerus mandibu- laris Trigona sp.: Brachymeria discreta Xylocopa brasilianorum (L.): Leucospis klugii Xylocopa caerulea (¥.): Coelopencyrtus pallidiceps Xylocopa caffra (L.): Coelopencyrtus callainus, Coelopencyrtus taylori Xylocopa divisa Klug: Coelopencyrtus callainus, Coelopencyrtus taylori Xylocopa fenestrata (¥.): Monodontomerus obscurus Xylocopa flavicollis (De Geer): Coelopencyrtus callainus Xylocopa flavorufa (De Geer): Coelopencyrtus callainus, Coelopencyrtus taylori, Coelopencyrtus watmoughi Xylocopa frontalis (Oliver): Coelopencyrtus gar- gar is Xylocopa inconstans Smith: Coelopencyrtus callai- nus Xylocopa nogueirai Hurd and Moure: Leucospis xylocopae Xylocopa pubescens Spinola: Coelopencyrtus sp. Xylocopa scioensis Gribodo: Coelopencyrtus cy- prius Xylocopa sp.: Leucospis reversa Xylocopa submordax Cockerell: Leucospis anthi- dioides Xylocopa sulcatipes Maa: Coelopencyrtus sp. Xylocopa tabaniformis orpifex Smith: Monodonto- merus montivagus Xylocopa tenuiscapa Westwood: Coelopencyrtus krishnamurtii Xylocopa tranquebarorum (Swederus): Melittobia sosui Xylocopa turanica Morawitz: Melittobia acasta Xylocopa watmoughi Eardly: Coelopencyrtus sp. Colletidae Hylaeus communis Nylander: Coelopencyrtus are- narius, Coelopencyrtus callidii Hylaeus cressoni Cockerell: Coelopencyrtus hylaei Hylaeus ellipticus (Kirbv): Coelopencyrtus hylaeol- eter Hylaeus fuscipennis (Smith): Coelopencyrtus kaalae 256 Journal of Hymenoptera Research Hylaeus heraldicus (Smith): Coelopencyrtus nothy- laei Hylaeus koae (Perkins): Coelopencyrtus kaalae Hylaeus nigritus (¥.): Coelopencyrtus arenarius Hylaeus polifolii Cockerell: Eurytoma stigmi Hylaeus pubescens (Perkins): Coelopencyrtus kaa- lae, Coelopencyrtus sexramosus Hylaeus sp.: Eurytoma nodularis, Melittobia acasta, Melittobia hawaiiensis Hylaeus varifrons Cresson: Pteromalus analis Halictidae Halictus africanus Friese: Aperilampus varians ILasioglossum pruinosum (Robertson): Eupelmus ashmeadi, Eupelmus rhizophelus Nomia melauderi Cockerell: Mesopjolobus brucho- phagi Megachilidae Anthidiellum perplexum Smith: Leucospis affinis Anthidiellum sp.: Leucospis slossonae Anthidiellum strigatum (Panzer): Leucospis bifas- ciata, Leucospis dorsigera Anthidium collectum Huard: Monodontomerus montivagus Anthidium diadema Latreille: Leucospis dorsigera Anthidium emarginatum (Say): Leucospis affinis, Leucospis dorsigera, Monodontomerus montiva- gus Anthidium florentinum (F.): Melittobia acasta, Monodontomerus aeneus, Monodontomerus lati- coruis Anthidium maculatum Smith: Monodontomerus thorpi Anthidium maculosum Cresson: Leucospis affinis Anthidium Imormonum Cresson: Monodonto- merus montivagus Anthidium septemspi}iosum Lepeletier: Monodon- tomerus laticornis Ashmeadiella aridula astragali Michener: Leucospis affinis Ashmeadiella bigeloviae (Cockerell): Microdonto- merus parkeri Ashmeadiella californica (Ashmead): Monodonto- merus montivagus Ashmeadiella cubiceps (Cresson): Microdontomerus parkeri Ashmeadiella gillettei Titus: Microdontomerus par- keri Ashmeadiella meliloti Cockerell: Leucospis affinis Ashmeadiella rufipes Titus: Microdontomerus par- keri Coelioxys octodentata Say: Aprostocetus sp., Mer- isus sp., Tetrastichus coelioxydis ICoelioxys quadridentatus (L.): Leucospis gigas Dianthidium curvatum sayi Cockerell: Monodon- tomerus montivagus Dianthidium heterulkei Schwarz: Monodontomerus dementi Dianthidium pudicum pudicum (Cresson): Leucos- pis affinis, Monodontomerus montivagus Dianthidium pudicum consimile (Ashmead): Mi- crodontomerus anthidii, Monodontomerus mon- tivagus Dianthidium sp.: Monodontomerus dianthidii Heriades crenulatus Nylander: Eurytoma heriadi, Melittobia acasta ? Heriades sp.: Leucospis dorsigera Heriades truncorum (L.): Melittobia acasta Hoplitis acuticornis (Dufour and Perris): Leucospis biguetina Hoplitis adunca (Panzer): Eurytoma nodularis, Leucospis dorsigera, Melittobia acasta, Mono- dontomerus aeneus, Monodontomerus ob- scurus Hoplitis anthocopoides (Schenck) (nest): Monodon- tomerus montivagus Hoplitis bullifacies Michener: Microdontomerus enigma, Microdontomerus parkeri Hopilitis palmarum (Cockerell): Microdontomerus parkeri Hoplitis producta (Cresson): Cleonymus amabilis, Eurytoma amplicoxa, Eurytoma stigmi, Leucospis affinis Hoplitis tridentata (Dufour and Perris): Leucospis biguetina, Neochalcis osmicida Hoplosmia ligurica (Morawitz): Leucospis dorsigera Lithurgus capensis Friese: Leucospis ornata, Leu- cospis varicollis Megachile aetheria Mitchell: Melittobia hawaiiensis Megachile albitarsis Cresson: Ablaxia cupraeus Megachile apicalis Spinola: Monodontomerus ae- neus Megachile argentata (F.): Dibrachys cavus, Melitto- bia acasta, Monodontomerus obscurus Megachile bombycina Radoszkowski: Melittobia acasta Megachile brevis Say: Aprostocetus sp., Cricellius megachilis, Leucospis affinis, Melittobia chalybii, Merisus sp., Microdontomerus parkeri, Tetrasti- chus coelioxydis Megachile centuncularis (L.): Ablaxia cupraeus, Anagrus putnamii, Aprostocetus pygmaeus, Bar- yscapus megachilidis, Dibrachys sp., Melittobia acasta, Melittobia chalybii, Melittobia megachilis, 1 ■ ' Volume 16, Number 2, 2007 257 Monodontomerus aeneus, Monodontomerus lati- cornis, Monodontomerus montivagus, Monodon- tomerus obscurus, Pteromalus apum, Pteromalus macronychivorus Megachile cephalotes Smith: Monodontomerus ob- scurus Megachile concinna Smith: Baryscapus megachili- dis, Melittobia australica, Monodontomerus ae- neus Megachile disjunctiformis Cockerell: Leucospis japonica Megachile ericetorum Lepeletier: Leucospis dorsi- gera Megachile flavipes Spinola: Monodontomerus ob- scurus Megachile gentilis Cresson: Baryscapus megachili- dis, Leucospis affinis Megachile gomphrenae Holmberg: Melittobia ha- waiiensis Megachile gratiosa Cameron: Melittobia sp. Megachile hungarica Gerstaecker: Leucospis gigas Megachile biennis Provancher: Leucospis affinis, Melittobia chah/bii Megachile lanata (¥.): Melittobia australica, Mono- dontomerus obscurus Megachile mendica Cresson: Leucospis affinis Megachile montivaga Cresson: Leucospis affinis, Microdontomerus apianus Megachile nipponica Cockerell: Leucospis japonica Megachile pallefacta Vachal: Melittobia hawaiiensis Megachile palmarum Perkins: Melittobia hawaiien- sis Megachile parietina (Geoffrey): Leucospis gigas, Melittobia acasta, Monodontomerus aeneus, Monodontomerus obscurus Megachile peruviana Smith: Monodontomerus mex- icanus Megachile poeyi Guerin-Meneville: Leucospis poeyi Megachile pugnata Say: Dibrachys sp., Leucospis affinis, Melittobia sp., Monodontomerus bakeri Megachile pyrenaica Lepeletier: Leucospis gigas, Pteromalus apum Megachile rancaguensis Friese: Leucospis hopei Megachile rangii Cheesman: Leucospis amino Megachile relativa Cresson: Dibrachys relativus, Leucospis affinis, Melittobia acasta, Melittobia chalybii, Monodontomerus bakeri, Monodonto- merus montivagus, Pteromalus apum Megachile rotundata (F.): Baryscapus daira, Bar- yscapus megachilidis, Dibrachys confusus, Dibra- chys maculipenuis, Melittobia acasta, Melittobia australica, Melittobia chalybii, Melittobia ha- waiiensis, Monodontomerus aeneus, Monodonto- merus bakeri, Monodontomerus dementi (in lab), Monodontomerus laticornis, Monodontomerus montivagus, Monodontomerus obscurus, Mono- dontomerus rugulosus, Pteromalus apum, Pter- omalus conopidarum, Pteromalus veneris, Tetra- stichus sp. Megachile sculpturalis Smith: Leucospis japonica Megachile sicula Rossi: Leucospis gigas, Monodon- tomerus aeneus Megachile sp.: Brachymeria paraguayensis, Calosota fumipennis, Horismenus albipes, Kocourekia cla- vigera, Leucospis histrio, Leucospis intermedia, Melittobia pelopoei, Monodontomerus acrostig- mus, Monodontomerus argentinus Megachile spissula (Cockerell): Lariophagus obtu- sus, Melittobia acasta Megachile ustulatum (Smith): Leucospis histrio Megachile willozomorensis Brauns: Leucospis ornata Megachile willughbiella (Kirby): Melittobia acasta, Monodontomerus obscurus, Pteromalus apum Megachile xylocopoides Smith: Baryscapus mega- chilidis Megachile zaptlana Cresson: Melittobia australica Microthurge corumbae (Cockerell): Leucospis sp. Osmia atriventris Cresson: Leucospis affinis Osmia bicolor (Schrank): Eulophus osmiarum Osmia bicornis (Schrank): Leucospis dorsigera, Leucospis gigas Osmia brevicomis (F.): Monodontomerus aeneus Osmia californica Cresson: Leucospis affinis Osmia "coerulea" [?lapsus for O. coerulescens, Baur and Amiet 2000]: Leucospis gigas Osmia coendescens (L.): Aprostocetus pygmaeus, Eurytoma nodularis, Leucospis gigas, Monodon- tomerus aeneus Osmia coloradensis Cresson: Monodontomerus bakeri Osmia cordata Robertson: Monodontomerus man- dibularis, Monodontomerus nuvitivagus, Mono- dontomerus obscurus Osmia cornifrons Radoszkowski: Monodonto- merus obscurus, Monodontomerus osmiae Osmia cornuta (Latreille): Leucospis dorsigera, Monodontomerus aeneus Osmia emarginata Lepeletier: Leucospis interme- dia, Monodontomerus aeneus Osmia excavata Alfken: Leucospis japonica, Mono- dontomerus osmaie Osmia fedtschenkoi (Morawitz): Leucospis dorsi- gera Osmia fulviventris (Panzer): Leucospis dorsigera, Monodontomerus aeneus Osmia globicola (Stadelmann): Leucospis osmiae 258 Journal of Hymenoptera Research Osmia kincaidii Cockerell: Leucospis affinis, Mono- dontomerus montivagus Osmia latisulcata Michener: Monodontomerus montivagus Osmia latreillei (Spinola): Calosota vernalis, Mono- dontomerus aeneus, Monodontomerus obscurus Osmia leueomelana (Kirby): Eurytoma sp., Melit- tobia acasta Osmia lignaria Say: Monodontomerus montivagus, Monodontomerus obscurus, Monodontomerus te- pedinoi, Monodontomerus torchioi Osmia marginata Michener: Microdontomerus parkeri Osmia nigrifrons Cresson: Dibrachys pelos, Mono- dontomerus aeneus Osmia niveata (F.): Leucospis dorsigera Osmia parietina Curtis: Leucospis dorsigera Osmia parvula Dufour and Perris: Eurytoma nodularis Osmia pumila Cresson: Leucospis affinis Osmia ribifloris Cockerell: Monodontomerus brevi- crus, Monodontomerus montivagus, Monodonto- merus obscurus Osmia rostrata Sandhouse: Leucospis affinis Osmia rufa comigera (Rossi): Monodontomerus aeneus, Monodontomerus obscurus Osmia rufa rufa (L): Leucospis dorsigera, Leucospis gigas, Melittobia acasta, Monodontomerus ae- neus, Monodontomerus obscurus Osmia sanrafaelae Parker: Monodontomerus mon- tivagus, Monodontomerus obscurus, Monodonto- merus torchioi Osmia simiUima Smith: Leucospis affinis Osmia sp.: Epistenia coeruleata, Monodontomerus montivagus (cocoon in Trypargilum politum nest), Torymus cupreus Osmia submicans Morawitz: Monodontomerus aeneus Osmia taurus Smith: Leucospis japonica, Mono- dontomerus osmiae Osmia texana Cresson: Monodontomerus bakeri, Monodontomerus montivagus Osmia tricornis Latreille: Leucospis dorsigera, Monodontomerus aeneus Pachyanthidium cordatum (Smith): Leucospis tri- color Pachyanthidium truncataum (Smith): Leucospis tricolor Pseudoantliidium lituratum (Panzer): Adontomerus gregalis, Adontomerus nesterovi, Neochalcis fer- toni Khodanthidium sticticum (F.): Monodontomerus anthidiorum Serapista denticulata (Smith): Leucospis africana, Leucospis tricolor Stelis depressa Timberlake: Monodontomerus mon- tivagus Stelis nasuta Latreille: Melittobia acasta, Mono- dontomerus aeneus Stelis sexmaculata Ashmead: Cleonymus amabilis, Leucospis affinis Chalcidoid/Bee Host Chalcididae Antrocephalus sp.: Apis ceraua Brachymeria discreta Gahan: Trigona sp. Brachymeria paraguayensis Girault: Megachile sp. Neochalcis breviceps (Masi): Ceratina fiavipes Neochalcis fertoni (Kieffer): Pseudoanthidium litur- atum Neochalcis osmicida (Saunders): Hoplitis tridentata Encyrtidae Cheiloneurus leptulus Annecke and Prinsloo: Ceratina sp. Coelopencyrtus arenarius (Erdos): Hylaeus commu- nis, Hylaeus uigritus Coelopencyrtus callainus Annecke: Xylocopa caffra, Xylocopa divisa, Xylocopa flavicollis, Xylocopa flavorufa, Xylocopa inconstans Coelopencyrtus callidii (Jansson): Hylaeus commu- nis Coelopencyrtus cyprius Annecke: Xylocopa scioen- sis Coelopencyrtus gargaris (Walker): Xylocopa fronta- lis Coelopencyrtus hylaei Burks: Hylaeus cressoni Coelopencyrtus hylaeoleter Burks: Hylaeus ellipti- cus Coelopencyrtus kaalae (Ashmead): Hylaeus fusci- pennis, Hylaeus koae, Hylaeus pubescens Coelopencyrtus krishnamurtii (Mahdihassan): Xy- locopa tenuiscapa Coelopencyrtus nothylaei Annecke: Hylaeus heral- dicus Coelopencyrtus pallidiceps (Girault): Xylocopa caer- ulea Coelopencyrtus sexramosus Timberlake: Hylaeus pubescens Coelopencyrtus sp.: Xylocopa pubescens, Xylocopa sulcatipes, Xylocopa watmouglii Coelopencyrtus tai/lori Annecke and Doutt: Xylo- copa caffra, Xylocopa divisa, Xylocopa flavorufa Volume 16, Number 2, 2007 259 Coelopencyrtus watmoughi Annecke: Xylocopa flavorufa Eulophidae Aprostocetus pygmaeus Zetterstedt: Megachile centuncularis, Osmia coerulescens Aprostocetus sp.: Coelioxys octodeutata, Megachile brevis Baryscapus americanus (Ashmead): Ceratina du- plet, Ceratina ignara, Ceratina nanula, Ceratina sequoiae Baryscapus daira (Walker): Megachile rotumiata Baryscapus megachilidis (Burks): Megachile cen- tuncularis, Megachile concinna, Megachile genti- lis, Megachile rotundata, Megachile xylocopoides Eulophus osmiarum Robineau-Desvoidy: Osmia bicolor Horismenus albipes (Schrottky): Megachile sp. Kocourekia clavigera Boucek: Megachile sp. Melittobia acasta (Walker): Anthidiumflorentinuiu, Anthophora abrupta, Anthophora retusa, Apis mellifera, Bombus hortorwn, Bombus ruderatus, Bombus terrestris, Heriades crenulatus, Heriades truncorum, Hoplitis adunca, Hylaeus sp., Mega- chile argentata, Megachile bombycina, Megachile centuncularis, Megachile parietina, Megachile relativa, Megachile rotundata, Megachile spis- sula, Megachile willughbiella, Osmia leucome- lana, Osmia rufa, Stelis nasuta, Xylocopa tur- anica Melittobia australica Girault: Megachile concinna, Megachile lanata, Megachile rotundata, Mega- chile zaptlana Melittobia chalybii Ashmead: Bombus fervidus, Megachile brevis, Megachile centuncularis, Megachile inermis, Megachile relativa, Megachile rotundata Melittobia hawaiiensis Perkins: Bombus sp., Hy- laeus sp., Megachile aetheria, Megachile gom- phrenae, Megachile pallefacta, Megachile pal- marum, Megachile rotundata Melittobia megachilis (Packard): Anthophora abrupta, Ceratina sp., Megachile centuncularis Melittobia pelopoei [unavailable name]: Antho- phora retusa, Megachile sp. Melittobia sosui Dahms: Xylocopa tranquebarorum Melittobia sp.: Euglossa annectans, Megachile gratiosa, Megachile pugnata Pediobius williamsoni Girault: Anthophora abrupta, Bombus americanorum, Bombus atratus Tetrastichus coelioxydis (Burks): Coelioxys octodeu- tata, Megachile brevis Tetrastichus howardi (Olliff): Apis mellifera Tetrastichus sp.: Megachile rotundata Torymus armatus (Boheman): Ceratina japonica Torymus cupreus (Spinola): Osmia sp. Torymus zabriskii (Cresson): Ceratina dupla Xylencyrtus mumifex Annecke: Allodapula mela- nopus Xylencyrtus tridens Annecke: Allodape exoloma, Allodape mucronata, Allodape panurgoides, Allo- dape rufogastra, Allodapula grandiceps, Braunsa- pis leptozonia Eupelmidae Calosota fumipennis Curtis: Megachile sp. Calosota vernalis Curtis: Osmia latreillei Eupelmus ashmeadi Melander and Brues: ILasio- glossum pruinosum Eupelmus rhizophelus Brues: ILasioglossum prui- nosum Eurytomidae Axima zabriskiei Howard: Ceratina calcarata, Ceratina dupla Eurytoma amplicoxa Bugbee: Hoplitis producta Eurytoma apiculae Bugbee: Ceratina dallatorreana, Ceratina nanula, Ceratina punctigena Eurytoma heriadi Zerova: Heriades crenulatus Eun/toma nodularis Boheman: Ceratina callosa, Hylaeus sp., Osmia adunca, Osmia coerulescens, Osmia parvula, Eun/toma sp.: Osmia leucomelana Eurytoma stigmi Ashmead: Hoplitis producta, Hylaeus polifolii Leucospidae Eeucospis affinis Say: Anthidiellum perplexum, Anthidium emarginatum, Anthidium maculo- sum, Ashmeadiella aridula astragli, Ashmeadiella meliloti, Dianthidium pudicum, Hoplitis pro- ducta, Megachile brevis, Megachile geutilis, Megachile inermis, Megachile mendica, Mega- chile montivaga, Megachile pugnata, Megachile relativa, Osmia atriventris, Osmia californica, Osmia kincaidii, Osmia pumila, Osmia rostrata, Osmia simillima, Stelis sexmaculata. Eeucospis africana Cameron: Serapista denticulata Eeucospis anthidioides Westwood: Xylocopa sub- mordax Eeucospis aruina Walker: Megachile rangii Eeucospis bifasciata Klug: Anthidiellum striga- tum Eeucospis biguetina J urine: Hoplitis acuticornis, Hoplitis tridentata 260 Journal of Hymenoptera Research Leucospis cayennensis Westwood: Centris analis, Centris bicornuta, Centris nitida, Centris tarsata, Centris vittata Leucospis dorsigera ¥.: Anthidiellum strigatum, Anthidium diadema, Anthidium emarginatum, IHeriades sp., Hoplitis adunca, Hoplosmia ligur- ica, Megachile ericetorum, Osmia bicornis, Osmia cornuta, Osmia fedtschenkoi, Osmia fulviventris, Osmia niveata, Osmia parietina, Osmia rufa rufa, Osmia tricornis Leucospis genalis Boucek: Diadasina distincta Leucospis gigas F.: Anthophora bomboides bom- boides, ICoelioxys quadridentatus, Megachile hungarica, Megachile parietina, Megachile pyr- enaica, Megachile sicula, Osmia bicornis, Osmia Icoerulescens, Osmia rufa rufa Leucospis histrio Maindron: ICtenoplectra chaly- bea, Megachile ustulatum Leucospis hopei Westwood: Megachile rancaguen- sis Leucospis intermedia Illiger: Megachile sp., Osmia emarginata Leucospis japonica Walker: Megachile disjunctifor- mis, Megachile nipponica, Megachile sculpturalis, Osmia excavata, Osmia taunts Leucospis klugii Westwood: Xylocopa brasilia- norum Leucospis ornata Westwood: Lithurgus capensis, Megachile willowmorensis Leucospis osmaie Boucek: Osmia globicola Leucospis pinna Grissell and Cameron: Eulaema meriana Leucospis poeyi Guerin-Meneville: Megachile poeyi Leucospis reversa Boucek: Xylocopa sp. Leucospis slossonae Weld: Anthidiellum sp. Leucospis sp.: Microthurge corumbae Leucospis tricolor Kirby: P achy anthidium corda- tum, V achy anthidium truncatum, Serapista den- ticulata Leucospis varicollis Cameron: Lithurgus capensis Leucospis xylocopae Burks: Xylocopa nogueirai Micrapion dalyi Boucek: Ceratina sp. Micrapion nasutum Boucek: Ceratina sp. Micrapion richardsi Boucek: Ceratina sp. Micrapion steffani Boucek: Ceratina truucata Polistomorpha conura Boucek: Euglossa sp. Polistomorpha fasciata Westwood: Euglossa ignita, Euglossa sp. Mymaridae Anagrus putnamii Packard: Megachile centuncu- lari Perilampidae Aperilampus varians Strand: Halictus africauus Pteromalidae Ablaxia cupraeus (Provancher): Megachile albitar- sis, Megachile centuncularis Cleonymus amabilis Cockerell: Hoplitis producta, Stelis sexmaculata Cleonymus ceratinae Kamijo: Ceratina japonica Cricellius megachilis Ashmead: Megachile brevis Dibrachys boarmiae (Walker): Apis mellifera Dibrachys cavus (Walker): Bombus sp., Megachile argent at a Dibrachys confusus (Girault): Megachile rotundata Dibrachys maculipennis Szelenyi: Megachile rotun- data Dibrachys pelos Grissell: Osmia nigrifrons Dibrachys relativus Doganlar: Megachile relativus Dibrachys sp.: Megachile centuncularis, Megachile pugnata Epistenia coeruleata Westwood: Ceratina sp., Osnua sp. Habritys latrus Wallace: Ceratina dupla Lariophagus obtusus Kamijo: Megachile spissula Merisus sp.: Coelioxys octodentata, Megachile brevis Mesopolobus bruchophagi (Gahan): Nomia melan- deri Nasonia vitripennis (Walker): Apis mellifera Pachycrepoideus vindemmiae Rondani: Bombus sp. Pteromalus analis Ashmead: Hylaeus varifrons Pteromalus apum (Retzius): Apis mellifera, Mega- chile centuncularis, Megachile pyrenaica, Mega- chile relativa, Megachile rotundata, Megachile willughbiella Pteromalus conopidarum (Boucek): Bombus agrorum, Bombus lapidarius, Megachile rotun- data Pteromalus macronychivorus Perez: Megachile cen- tuncularis Pteromalus veneris Dalla Torre: Megachile rotun- data Torymidae Adoiitomerus gregalis (Steffan): Pseudoanthidium lituratum Adontomerus nesterovi Zerova: Pseudoanthidium lituratum Echthrodape africana Burks: Braunsapis rolini, Braunsapis rufipes, Braunsapis simplicipes Echthrodape papuana Boucek: Braunsapis unicolor Volume 16, Number 2, 2007 261 Microdontomerus anthidii (Ashmead): Dianthi- dium pudicum consimile Microdontomerus apianus Grissell: Megachile mon- tivaga Microdontomerus enigma Grissell: Hoplitis bullifa- cies Microdontomerus parkeri Grissell: Ashmeadiella bigeloviae, Ashmeadiella cubiceps, Ashmeadiella gillettei, Ashmeadiella rufipes, Hoplitis bullifa- cies, Hoplitis palmarum, Megachile brevis, Osmia marginata Monodontomerus acrostigmus Grissell: Megachile sP. Monodontomerus aeneus (F.): Anthidium florenti- num, Anthophora retusa, Ceratina callosa, Hopli- tis adunca, Megachile apicalis, Megachile cen- tuncularis, Megachile concinna, Megachile par- ietina, Megachile rotundata, Megachile sicula, Osmia brevicornis, Osmia coerulescens, Osmia cornigera, Osmia cornuta, Osmia emarginata, Osmia fulviventris, Osmia latreillei, Osmia rufa, Osmia nigrifrons, Osmia submicans, Osmia tricornis, Stelis nasuta Monodontomerus anthidiorum (Lucas): Rhodanthi- dium sticticum Monodontomerus argentinus Brethes: Eufriesea nigrescens, Megachile sp. Monodontomerus bakeri Gahan: Megachile pug- nata, Megachile relativa, Megachile rotundata, Osmia coloradensis, Osmia texana Monodontomerus brevicrus Grissell: Osmia ribi- floris Monodontomerus dementi Grissell: Dianthidium heterulkei, Megachile rotundata Monodontomerus dianthidii Gahan: Dianthidium sP. Monodontomerus laticornis Grissell and Zerova: Anthidium florentinum, Anthidium septemspino- sum, Apis mellifera, Megachile centuncularis, Megachile rotundata Monodontomerus mandibularis Gahan: Anthophora abrupta, Anthophora bomboides bomboides, Meli- totna taurea, Osmia cordata Monodontomerus mexicanus Gahan: Ancyloscelis apiformis, Anthophora marginata, Megachile peruviana Monodontomerus montivagus Ashmead: Anthidium collectum, Anthidium emarginatum, Anthidiun Imormonum, Anthophora abrupta, Anthophora bomboides bomboides, Anthophora bomboides neo- mexicana, Anthophora linsleyi, lAnthophora occi- dentalis, lAnthophora vallorum, Ashmeadiella cali- fornica, Bombus morrisoni, Dianthidium curvatum sayi, Dianthidium pudicum consimile, Dianthi- dium pudicum, Hoplitis anthocopoides (nest), Megachile centuncularis, Megachile relativa, Mega- chile rotundata, IMelissodes sp., Osmia cordata, Osmia kincaidii, Osmia latisulcata, Osmia lignaria, Osmia ribifloris, Osmia sanrafaelae, Osmia texana, Stelis depressa, Xylocopa tabaniformis orpifex Monodontomerus obscurus Westwood: Anthophora plumipes, Hoplitis adunca, Megachile argentata, Megachile centuncularis, Megachile cephalotes, Megachile flavipes, Megachile lanata, Megachile parietina, Megachile rotundata, Megachile will- ughbiclla, Osmia cordata, Osmia cornifrons, Osmia latreillei, Osmia lignaria, Osmia ribfloris, Osmia rufa rufa, Osmia rufa cornigera, Osmia sanrafaelae, Xylocopa fenestrata Monodontomerus osmiae Kamijo: Osmia cornifrons, Osmia excavata, Osmia taurus Monodontomerus parkeri Grissell: Anthophora occidentalis Monodontomerus rugulosus Thomson: Megachile rotundata Monodontomerus tepedinoi Grissell: Osmia lignaria Monodontomerus thorpi Grissell: Anthidium ma- culatum Monodontomerus torchioi Grissell: Osmia lignaria, Osmia sanrafaelae 262 Journal of Hymenoptera Research frenal line upper mesepimeron lower mesepimeron metapleuron frenum transepimeral sulcus epimeron metapleuron frenum malar distance 11 intermalar distance submarginal vein marginal vein postmarginal vein 10 , stiqma stigmal y parastigma vein Figs. 1-11. Torymidae. 1-2, Mesosoma, side (arrow indicates anterior margin of metapleuron). 3-4, Metacoxa, side. 5-6, Propodeum, dorsal. 7-8, Mesosoma, dorsal (arrow indicates frenum). 9, Fore wing, dorsal (Echthrodape africana). 10, Fore wing, dorsal, showing venation terminology. 11, Head, anterior, showing measurements. Volume 16, Number 2, 2007 263 14 15 16 occipital carina hypostomal carina 17 18 Figs. 12-28. Torymidae. 12-13, Fore wing venation, dorsal. 14-15, Head, anterior, lines indicate malar and intermalar distances. 12, 14 Echthrodape papuana. 13, 15 Echthrodape africana. 16-18, Head, posterior (showing carinae). 16, Microdontomerus. 17, Pseudotorytnus. 18, Monodontomerus. 19-20, Antenna, side. 19, Monodontomerus thorpi. 20, Monodontomerus spp. 21-28, Head. 21-22, Monodontomerus osmiae (from Kamijo 1963). 23-24, Monodontomerus anthidiorum. 25-26, Monodontomerus mexicanus. 27-28, Monodontomerus bakeri. 264 Journal of Hymenoptera Research basal cell costal cell admarginal area postmarginal vein ■*■ i* /*"/* /■y'gT^v - Figs. 29-42. Torymidae, Manodontomerus spp. (except 34, Pseudotorymus). 29-31, Fore wing. 29, M. aeneus. 30, M. sp. 31, M. dementi. 32-33, Metasomal tergum 6. 32, M. argentinus. 33, M. rugulosus. 34-35, Hind femur and tibia, side. 34, P. sp. 35, M. aeneus. 36-38, Hind femur. 36, M. thorpi. 37, M. argentinus. 38, M. rugulosus. 39^0, Fore leg (left side view, right ventral view). 39, M. aeneus. 40, M. brevierus. 41-42, Frenum (apex of scutellum). 41, M. acrostigmus. 42, M. aeneus. Volume 16, Number 2, 2007 265 . - Ti i •> » &^scz<'<*s% 43 \^> 51 52 \\V^ e ■ t '.- ■:'." I transepimeral sulcus Figs. 43-58. Torymidae, Motwdontomerus (Mo.) and Microdontomerus (Mi.). 43-46, Fore wing, part. 43, Mo. mandibularis. 44, Mo. parked. 45, Mi. parkeri. 46, Mi. enigma. 47-50, Heads. 47, Mo. montivagus. 48, Mo. mandibularis. 49, Mi. anthidii. 50, M/'. apiamis. 51-52, M. acrostigmus (variation in stigma). 53-55, Mesopleuron. 53, Mo. dementi. 54, Mo. parkeri. 55, Mo. montivagus. 56-58, Scape, male. 56, Mo. montivagus. 57, Mo. mandibularis. 58, Mo. tepedinoi. J. HYM. RES. Vol. 16(2), 2007, pp. 266-276 Multivoltinism and Usage of Multiple Nest Substrates in a West Texas Sand Dune Population of Pseudomasaris phaceliae Rohwer (Hymenoptera: Vespidae: Masarinae) John L. Neff and Allan W. Hook (JLN) Central Texas Melittological Institute, 7307 Running Rope, Austin, Texas 78731, USA; email: jlnatctmi@yahoo.com (AWH) Department of Biology, St. Edward's University, Austin, Texas 78704-6489, USA; email: allanh@stedwards.edu Abstract. — A west Texas population of Pseudomasaris phaceliae was found to be multivoltine and active from April to September. Unlike previous reports of nests constructed only on stones, nests were commonly constructed on the stems and infructescences of its host plant, Phacelia integrifolia. Emergence data indicated the primary sex ratio is strongly female biased (53:14). Nest parasitism was rare, but predation was common. Data on nest architecture, nest construction, and foraging behavior are presented. Pseudomasaris is the only North Ameri- phaceliae, like most other Pseudomasaris can genus of the Masarinae, a relatively species, constructs its nests on stones small (300+ spp.) but widespread clade of (Parker 1967, Torchio 1970), we found that pollen and nectar provisioning vespid at Monahans, P. phaceliae was commonly wasps (Carpenter 1982, 2001, Gess 1996). attaching its nests to plant stems, and only Torchio (1970) reported in considerable rarely to stones. In addition, P. phaceliae, detail on the biology of Pseudomasaris like most other temperate masarine spe- edwardsii (Cresson) but little is known of cies, had previously been assumed to be the biology of the remaining 14 species, univoltine (Parker 1967, Gess 1996), but Brief reports on nest structure and /or nest wasps emerged in early July from a nest sites have been published for eight of these collected during June suggesting multi- species, including Pseudomasaris phaceliae voltinism. This prompted a series of visits Rohwer, but only floral records are avail- to this site to gather additional information able for four other species, and nothing at on the nests and behavior. Follow-up trips all has been published on the biology of the were made in July, August, and September remaining two. 2005 and April, May, June, and August During a brief visit to Monahans Sand- 2006. hills State Park in June 2005, we encoun- Habitat. — Monahans Sandhills State Park tered a large population of Pseudomasaris (32.128'N, 103.953 W) is located in Ward phaceliae visiting flowers of Phacelia integri- County, Texas on the southern edge of folia Torr. (Boraginaceae). Pseudomasaris a large dune field of quartz-rich Quaterna- phaceliae is an infrequently collected but ry sands that stretch northward into sometimes locally abundant species of the southeastern New Mexico (Machenberg arid American Southwest (Arizona, New 1984, Muhs 2001). While much of the dune Mexico, west Texas, and adjacent Mexico) field is partially stabilized by shin oak (Richards 1966, and pers. obs.). Although (Quercus havardii Rydb.) and other peren- previous reports had indicated that P. nials, large moving dunes are common in Volume 16, Number 2, 2007 267 Figs. 1-2. 1. Pseudomasaris phaceliae nest (indicated by arrow) on Phacelia integrifolia stem. 2. Incomplete P. phaceliae nest with one completed cell and one cell under construction. the Park. Soils in the sandhills consist almost exclusively of loose sands. Occa- sionally, winds expose the underlying caliche layer, but generally there are no rocks or stones in the dunes beyond those brought in for the caliche service roads. Like most of Texas, summers are warm with July maximum temperatures averag- ing 35 C. Rainfall is low, averaging 33.6 cm per year, with 75% of precipitation occur- ring during a six-month May to October summer/fall period. Despite the relative aridity, the water table is quite shallow in the sand hills due to an underlying impermeable caliche layer. Water-loving plants like Salix nigra Marshall (Salicaceae) and Baccharis salicina Torrey & A. Gray (Asteraceae) are found in some of the deeper depressions among the dunes where temporary ponds may form follow- ing unusually heavy rains (Machenberg 1984). Nests. — Although a few nests were found on miscellaneous plant stems (such as a sapling of Prosopis glandulosa Torr. (Fabaceae) or stems of the erect herb Mentzelia strictissima (Wooton & Standi.) J. Dark (Loasaceae), the vast majority of the nests we found were on stems or infruc- tescences of larger (over 40 cm high) individuals of Phacelia integrifolia (Fig. 1.). Nests were located at heights of 20.3- 71.1 cm above the ground (n = 50, mean = 41.0 ± 9.9 cm). Typically there was only one nest per plant although a few plants had two, and one Phacelia plant had four nests, two older nests from which emer- gence was complete, and two newer nests under construction. Like those of other Pseinloniasaris species, nests consisted of one or more cylindrical cells attached lengthways one to another (Torchio 1970). Cells of nests on plants were always positioned with the long axis 268 Journal of Hymenoptera Research oriented vertically, regardless of the orien- tation of the stem or infructescence (Fig. 2). Nests on plants averaged 5.4 ± 3.8 cells per nest, (1-14, n = 53), a value that probably underestimates the true number of cells per nest since some of the nests may not have been complete when censused. Nests on stems were initiated by laying down a strip of the sand-soil mix along a stem or infructescence. The female then con- structed the hemispherical inner end of the cell at the upper end of the strip, and completed the cell wall by adding irregular strips or scale-like patches of the moist sand-soil mix. Upon returning with a sand- soil load, the female inserted her head in the nest and curled her metasoma so its tip was opposite her head. The moist sand-soil mix was added to the cell wall by simultaneously working the mix from within with her mandibles and tapping it from the outside with the flattened, hirsute surface of metasomal sternite 6. Time on the nest depositing cemented sand was typically brief (0.6-2.9 min). The first cell is a true cylinder as the walls are complete, the plant substrate not being used as part of the cell wall (Fig. 2). Additional cells are attached to the walls of the initial cell so the nest grows in an asymmetrical manner away from the stem rather than around it (Fig. 3). Cell walls are rather thin, only 0.4-0.5 mm thick, but quite strong. Nests within 100 meters of the caliche road were usually constructed with a mix of relatively coarse sand particles (0.1-0.4 mm in diameter), and much finer (0.02-0.04 mm) particles, pre- sumably caliche dust. Nests further from the road are constructed of sand alone. The particles are bound together with regurgi- tated nectar, and perhaps, glandular prod- ucts. As there was no free water in the dunes, nectar (or perhaps honey-dew) was the only likely source for the bulk of the liquid used to moisten the sand-soil mix. Nest fragments placed in water softened but were still intact after being immersed for 48 hours, suggesting substances be- yond just sugar may be holding the walls together. These moistened nests quickly sprouted fungal hyphae, the fungi pre- sumably growing on the nectar sugars. In contrast, the true mud nest walls and partitions of nests constructed by Trypargi- lum politum (Say) or Osmia lignaria (Say) dissolve almost instantly when placed in water (pers. obs.). Individual cells were 15.2 ± 1.2 mm (12.8-17.6, n = 19) mm long with an average diameter of 4.2 ± 0.2 mm (3.8- 4.5, n = 28). The distal end of the cell was hemispherical while the cell opening was simple and truncate. After provisioning, each cell was closed with a cemented sand plug 0.5-0.8 mm thick medially and 1.0- 1.3 mm thick at the sides. Additional sand and fine particles were added to the nest exterior as construction proceeded, filling the spaces between the cells (Fig. 9). In one unfinished nest in which the second cell was only half finished, soil had been added along the juncture of the two cells, strengthening their connection, although additional soil had not been added else- where to the outer surface of the first cell. Upon completion, the nest has flattened, relatively smooth walls concealing the outlines of individual cells. We found no indication of empty spaces between the cells. Completed nests on stems usually lacked obvious ornamentation but a few (4 of 50) had conical projections on the margins of the nest (Fig. 4). Females were commonly observed col- lecting fine soil particles along the caliche road (Fig. 6) and, less frequently, in the dunes. Females would hover 10-15 cm above the soil surface and repeatedly drop to collect sand or soil. We did not obtain a complete picture of soil collection but it was clear that during most bouts on the soil surface, the wasps did not add nectar to the soil surface prior to collection. Rather, they used their mandibles and foretarsi to add soil or sand to a moist soil bolus held behind the mandibles by their modified labial palpi. Regurgitated nectar Volume 16, Number 2, 2007 269 Figs. 3-5. 3. Completed Pseudomasaris phaceliae nest showing asymmetrical position on Phacelia integrifolia stem. 4. Ornamented P. phaceliae nest (projection indicated by arrow). 5. P. phaceliae nest opened laterally, probably by birds. is apparently added to this bolus while the wasps are hovering. As in Pseudomasaris edwardsii, females constructing cells or adding sand to the nest exterior typically made long trips (20-30 min duration) to gather nectar at flowers as well as sand, interspersed between series of 3^4 relative- ly short (0.9-3.1 min) trips. The latter trips were presumably for sand only since they were too brief to allow for both sand and nectar collection. While the majority of nests were con- structed on plant stems, some nests were constructed on small stones brought in for a caliche service road through the dunes. These were structurally similar to the nests 270 Journal of Hymenoptera Research - Figs. 6-10. 6. Pseudomasaris phaceliae female collecting sand (sand bolus indicated by arrow). 7. Completed P. phaceliae nest on stone. 8. Female extending walls of new P. phaceliae nest. 9. Female adding sand (indicated by arrow) to exterior of P. phaceliae nest. 10. Emergence holes from six- celled P. phaceliae nest. constructed on stems, but had fewer cells, never more than three, averaging only 1.3 ± 0.5 cells (1-3, n = 28). While all the stem nests were constructed during the year in which they were discovered, many of the rock nests were worn and /or damaged and may have been constructed in prior years. Unlike the situation in stem nests, the first cells constructed on stones were usually not true cylinders as the stone was used as part of the cell wall. The first cells constructed typically were attached to the stone along their entire length (Fig. 8), although in a few cases, only the basal half of the cell was attached to the substrate, with the outermost portion arching away. Completed nests on stones often appeared to be triangular in cross-section, due to the smoothing of the sides and the addition of soil, which greatly widened the base of the nest (Fig. 7). Nests on rocks occurred both on the sides and upper surfaces without any consistent orientation. Completed nests on stones were quite cryptic as the outer nest covering matched the color of the pale stone on which they were con- structed. Nests on Phacelia plants were more conspicuous since the pale nests contrasted with the green stems or young infructescences, but the nests were still somewhat cryptic since the infructes- cences, leaves and /or leaf tips all com- monly turn brown with age. Nests are provisioned with a dense mass of regurgitated pollen-nectar pellets. The pellets had short projections and were oriented so the projections, rather than the main mass of the pellets, contacted the cell walls. Pellets averaged 0.6-0.7 mm across with projections from 0.2 to 0.6 mm long. Unlike the provisions of Pseudomasaris edwardsii, the innermost face of the provision mass of P. phaceliae was not smoothed so the individual pellets were easily distinguished. The outer face of the provision mass (the surface facing Volume 16, Number 2, 2007 271 the cell closure) was smooth and convex. We found several cells being provisioned, and others with feeding larvae but re- covered only a single cell with a completed, intact, provision mass. This mass was 15.2 mm long, occupying nearly all the cell, leaving a small open space of 0.8 mm between the mass and the roof of the hemispherical inner end, and 1.6 mm between the mass and the cell closure. Host Plant and Foraging. — Pseudomasaris phaceliae is believed to restrict its foraging to flowers of Pliacelia spp. (Torchio 1970). At Monahans dunes, individuals of P. phaceliae foraged only at flowers of Pliacelia integrifolia, the only Pliacelia species flower- ing. Pliacelia integrifolia is a widespread annual of the American Southwest, occur- ring on a variety of rocky or sandy substrates, particularly gypsum or lime- stone (Correll and Johnston 1970). The plants bear scorpioid cymes of small, pale purple flowers. In west Texas, flowering occurs primarily from March through May (Correll and Johnston 1970), although herbarium records from the Plant Re- sources Center of the University of Texas, Austin, Texas, show flowering as late as the end of July in the Monahans Sandhills area, and a non-technical guide indicates flowering through September (Ajilvsgi 2003). Pliacelia integrifolia was common and in flower during our first 2005 visit to Monahans dunes on 16 April, although cold, wet weather prohibited any observa- tions of floral visitors. When we revisited Monahans Dunes on 18 June 2005, we found P. integrifolia flowering was still widespread although most abundant in certain depressions between the dunes. When the site was revisited a month later in July, most P. integrifolia plants on the dunes were brown and dead, and flower- ing was restricted to the large plants occurring in the depressions between the dunes. Unlike the typical, erect plants on the dunes and most other habitats (Fig. 1), these plants were sprawling and multi- branched, almost shrub-like in aspect, with greatly enlarged stems. These depression plants were still strongly flowering when the site was revisited in August. On the final 2005 visit on 16-18 September low levels of flowering were still occurring among the large depression plants al- though most plants were dead, and flower- ing was limited to a few inflorescences. On 21 April 2006, P. integrifolia was flowering and locally abundant on the dunes, but surprisingly, was absent in the depressions where the large, long-flowering individu- als were found in 2005. Flowering contin- ued, with a gradual decline in overall abundance through May and June 2006. By August 2006, only secondary inflores- cences of the few surviving individuals were still flowering. Individual flowers of Pliacelia integrifolia begin opening around 0830 CDT (approx. 2 hrs after sunrise), and flowers continue opening through the day. Individual flow- ers last approximately two days. Anthers dehisce fully shortly after flower opening, and are usually stripped of pollen within two hours of opening. Some larger bees, such as species of Habropoda, Osmia, or Martinapis began foraging on P. integrifolia at or before 0800 early in the season, but Pseudomasaris phaceliae was always a late starter. Females were rarely observed before 0830 CDT and they were generally not active until after 0930 CDT with air temperatures above 30 C. Foraging contin- ued until sundown, although females were sometimes observed resting on stems or branches during periods of peak heat in late afternoon (air temperatures above 39 C). Foraging rates recorded on 13 July 2005 were quite rapid, perhaps a reflection of the high temperatures (33-35 C) and re- source depletion due to the high number of wasps and bees concentrated on the rela- tively few Pliacelia plants still flowering. Wasps visited an average of 24.5 ± 5.3 (16.4-31.2, n = 20) Pliacelia flowers per minute. Nectar visits were quite quick, averaging only 0.8 ± 0.3 (0.3-2.0, n - 40) 272 Journal of Hymenoptera Research Figs. 11-12. 11. Pseudomasaris phaceliae female collecting pollen of Phacelia integrifolia. 12. P. phaceliae female nectaring at flower of P. integrifolia. seconds per flower while pollen-collecting visits were longer at 4.1 ± 3.0 (1.1-11.2, n = 20) seconds. During a foraging bout, the proportion of flowers worked for pollen was low, averaging only 13.7 ± 14.6% (0- 41.7, n = 20). This is presumably a re- flection both of the low availability of fresh flowers with available pollen, and the fact that some foragers were constructing their nests and not foraging for pollen at the time. If we exclude foraging series where no flowers were visited for pollen, the proportion of flowers visited for pollen rises only slightly to 19.6 ± 13.7% (3.4-41.7, n = 14). A nectar foraging bout involved landing on medial portions of the exserted staminal filaments and rapidly inserting the re- markably extensible proboscis (~ 4.5 mm long when fully extended) to reach the nectary at the base of corolla tube (Fig. 12). The initial approach to a flower presum- ably involved some assessment of pollen availability, since foraging behavior chan- ged when a pollen forager encountered a flower with obvious available pollen. A pollen forager would grasp the staminal filament near the anther with her hind and mid-tarsi while hovering (Fig. 11). She would then grasp an anther with her mandibles and extract pollen by scraping pollen to her mouthparts with the tarsal brushes of her forelegs. Depending on pollen availability, a wasp might work several anthers on a flower before moving down the filaments to insert her mouth- parts to gather nectar. A female observed on 17 June 2006 took 8.40 hrs and 15 foraging trips to provision a cell. Pollen foraging appeared to com- mence immediately after oviposition. Be- cause of poor light conditions, we were not able to determine if the female deposited an initial pollen and nectar load immedi- ately after ovipositing as has been reported for Pseudomasaris edwardsii (Torchio 1970). Pollen trips averaged 31.53±7.10 min (19.32-41.72, n = 13) and deposition time in the nest between trips averaged 2.06±0.48 min (1.50-3.52, n = 14). As noted for P. edwardsii (Torchio 1970), the female rotated within the cell during the later phases of pollen deposition as she de- posited pollen pellets. Development and Voltinism. — Only frag- mentary data were obtained on develop- ment. The only intact egg recovered was smooth, slightly asymmetrical, and 3.7 mm long with a medial width of 1.0 mm. The posterior end of the egg was attached VI Ml tfl Volume 16, Number 2, 2007 273 Table 1. Collection dates for nests of Pseudomasaris phaceliae from Monahans Sandhills with emergence dates ind sex ratio of the wasps. Nest Collection date Emergence date Males Females H-l 16-vi-2005 by 12-vii-2005 1 14 M-l 18-vii-2005 19-vii-2005 0 2 \J-2 18-vii-2005 23-vii-2005 0 1 SJ-3 18-vii-2005 3-iv-2006 0 1 vJ-4 15-viii-2005 28-viii-2005 1 3 M-5 15-viii-2005 l-ix-2005 0 1 sr-6 18-vii-2005 19-iv-2006 0 2 )5-06 18-vii-2005 approx. 3-iv-2006 0 1 )5-07 18-vii-2005 approx. 3-iv-2006 1 1 15-02 18-vii-2005 18-iv-2006 2 0 )6-01 23-iv-2006 26 to 29 iv-2006 1 5 )6-02 19-V-2006 1 to 7 vi-2006 1 8 16-03 18-vi-2006 27-vi to 11 -vii-2006 1 7 16-04 18-vi-2006 1 to7 vii-2006 4 3 16-05 18-vi-2006 5 toll vii-2006 1 3 16-06 18-vi-2006 12-vii-2006 1 1 fotal 14 53 perpendicularly to the inner surface of the :ell wall, just below the hemispherical, nner end of the cell. The ventral surface of he egg, which paralleled the flat inner ;urface of the provision mass, was some- vhat flattened, but the dorsal surface of the ?gg arched slightly into the curved space of he hemispherical inner cell cap. In two nstances where late instar larval feeding vas noted, the larva fed along the side of he provision, eating its way down the provision mass toward its distal end. Sometime after completing feeding, the arva spun a thin, translucent cocoon that idhered tightly to the cell wall. Defecation >ccurred after the completion of the co- :oon. Feces were typically deposited as an rregular ring of smooth, flattened, semi- .pherical pellets (0.8-1.0 mm wide, and 1.5-0.6 mm tall), around the inner end of he cell, although some of these are ometimes pressed into a flattened cake vith unrecognizable individual pellets. \Iest dissections indicated wasps overwin- er as prepupae. Adults emerge by chew- ng through the nest plug (Fig. 10). The total number of generations per year )f Pseudomasaris phaceliae at Monahans is inknown but at a minimum it is two and perhaps as many as four. Pseudomasaris phaceliae was active at Monahans for at least 93 days (18 June to 18 September) in 2005 and 116 days (26 April to 19 August) in 2006. Judging from the extensive wing wear of females collected in June of 2005, the phenology of PJiacelia iutegrifolia, and the emergence times of P. phaceliae in the lab (Table 1.), flight at Monahans during 2005 probably began in early April sug- gesting a flight period in excess of 156 days in 2005. Emergence patterns of wasps from nests collected in 2005 and 2006 are indicated in Table 1. All nests collected before July had their inhabitants emerge that same year, but for nests collected in July or later, some emerged the same year but others went into larval diapause and emerged the following April. The instances of late 2005 (July to September) emergence were from nests taken from infructescenses on rela- tively fresh green plants, and it is clear that they had been provisioned in 2005. Phacelia iutegrifolia stems usuallv break down com- O J J pletely over the winter and the glandular epidermis, to which the nests are frequent- ly attached, falls away soon after the plant dies. It is probable that over-wintering 274 Journal of Hymenoptera Research nests fall from the disintegrating plants was not confirmed. Additionally, two and spend the winter in the sand. newly provisioned cells were encountered Mating and Sex Ratio.— Males of Pscudo- with small lateral slits, possibly the results masaris phaceliae were observed to forage of bird probes. One of these probed cells for nectar at plants of Phacelia integrifolia contained a larva and partially consumed and patrol P. integrifolia inflorescences, but pollen mass while the other was being mating was not observed. Newly emerged raided and emptied by ants, males confined with newly emerged fe- We found no indication of empty (closed male nest mates attempted to mount their but unprovisioned) cells in Pseudomasaris sisters, although it was not determined if phaceliae nests at Monahans. Empty cells mating was successful. Females were much have been invoked as an anti-parasite more common than males at flowers at all defense strategy for Pseudomasaris vespoides times during our visits, and the sex ratio of (Cresson), in which roughly 30% of the nest wasps emerging from nests was heavily cells are empty (Tepedino et al. 1979). female biased 3.8:1 (53 females/ 14 males, Table 1). In all cases where emergence DISCUSSION order was determined, males emerged Our observations of Pseudomasaris phace- from the first provisioned cell(s) of a nest. Hae at Monahans suggest this species has Nest Associates and Predators. — Nest par- a broader behavioral range than previously asitization was uncommon. One female observed in Pseudomasaris. The most obvi- Chrysurissa densa (Cresson) (Chrysididae) ous differences from previous reports are emerged on 18 July 2005 from a Pseudoma- the incidence of multivoltinism and use of saris phaceliae nest and a second female plant stems for nest placement. Multivoltin- emerged from a different nest on 13 ism has not previously been reported in September 2005. Chrysurissa densa appar- Pseudomasaris and appears to be rare in the ently is a specialist on Pseudomasaris, since Masarinae, at least among temperate zone besides P. phaceliae, its only known hosts species (Gess 1996). Presumably, this is true are four other Pseudomasaris species. In because these wasps are typically oligolectic addition, its range mirrors that of Pseudo- (or at least have a narrow range for floral masaris (Bohart and Kimsey 1982). The only hosts) and the flowering periods of their other nest parasites where 10 females and 2 floral hosts usually are quite temporally males of an unidentified Monodontomerus restricted. Multivoltinism is possible at sp. (Torymidae) which emerged from a cell Monahans because of the unusual extended of another nest of P. phaceliae. As parasite flowering of its floral host in this distinctive emergence occurred only two days after habitat, allowing a flight season that can collection of the nest, it was clearly the extend from April into September in favor- result of field, rather than laboratory, able years. However, it is possible that infestation. multivoltinism occurs in other non-dune Indications of nest predation were occa- populations of P. phaceliae associated with P. sionally encountered. Several nests ob- integrifolia. We have collected P. phaceliae on served in 2005 and 3 of 21 nests measured this species at other west Texas sites in June, in June 2006 had all their cells opened Wing wear indicated these wasps were laterally (Fig. 5). The cells in these nests relatively newly emerged even though the contained empty cocoons but had intact few flowering P. integrifolia plants at these cell caps. Birds are the most likely pre- sites were in very poor condition and most dators although we cannot rule out small plants in the populations were dead or mammals. A small, unidentified wood- fruiting. pecker was seen perching and searching As we encountered neither flowering on Phacelia stems, but actual nest predation Phacelia nor Pseudomasaris phaceliae during Volume 16, Number 2, 2007 275 July, August, or September visits to Mon- ahans prior to 2005, multivoltinism may be a facultative phenomenon for P. phaceliae. The wasps could use environmental cues such as temperature and humidity, which may predict extended Phacelia bloom, to "determine" whether to pupate and emerge immediately or proceed to dia- pause. Increasing humidity, a predictor of flowering in desert plants, has experimen- tally been shown to be an important cue in breaking diapause in Macrotera portalis Timberlake (Andrenidae), a desert bee with extended diapause (Danforth 1999), while temperature (above or below 29°C) determines whether Nomia melanderi Cock- erell (Halictidae) pupates and emerges immediately or proceeds to diapause (Ste- phen 1965). Even in years with extended Phacelia flowering, flowering declines greatly late in the year. The observation that some larvae in late provisioned nests (July or later) pupate and emerge immedi- ately while others diapause until the following spring suggest a bet-hedging strategy predicated on the decreasing chances of encountering adequate floral resources late in the year. Published reports have indicated that surfaces of rocks are the preferred nest substrates utilized by six Pseudomasaris species (Hicks 1929, Hungerford 1937, Parker 1967). In Torchio's 1970 glasshouse study, Pseudomasaris edwardsii females con- structed nests on a variety of substrates, but not plant stems. However, the use of bamboo stakes as a nest substrate in Torchio's study suggests they may occa- sionally use twigs or plant stems under natural conditions. The only report on Pseudomasaris texana (Cresson) indicates it constructs its nests on twigs (Bequaert 1940). Pseudomasaris vespoides has repeat- edly been reported constructing its nests on twigs or plant stems (Cockerell 1913, Davidson 1913, Hicks 1929, Bequaert 1940, Torchio 1970), although it also is known to use rocks (Hicks 1927). Pseudomasaris mar- ginalis (Cresson) was found to nest in beetle borings in logs at a high altitude site in Colorado (Dorr and Neff 1982) although this report has been questioned (Gess 1996). In the only previous report on the biology of P. phaceliae, rock surfaces were the only reported nest substrate (Parker 1967). Since rocks are rarely encountered in the Monahans Sandhills, it is not surprising that most of the Pseudomasaris phaceliae nests we encountered were attached to plants. The relative advantages of rock surfaces and plant stems as nest substrates are unclear. Rock surfaces are obviously more permanent and stable than herba- ceous plant stems, but the importance of this difference is not clear for wasps like these that do not reuse their nests. The issue of difference of substrate permanence could easily be erased by using the stems of perennials, although P. phaceliae does not seem to regularly do this at Monahans. If heat stress is a problem, a strong possibility in a habitat like Monahans Sandhills where soil surface temperatures regularly exceed 40° during the late spring and summer, constructing nests well above the soil surface on plant stems might be advanta- geous relative to building nests on low, exposed rock surfaces. The nest biology of most Pseudomasaris species is poorly known, often based on only a single population, and, in some cases, a single nest. When the biologies of more populations of more species of Pseudomasaris are known, it will be in- teresting to see if other Pseudomasaris species are similarly flexible in their pat- terns of voltinism and /or nest substrate usage. ACKNOWLEDGEMENTS We thank the Texas Parks and Wildlife Department (Scientific Study Permit 27-05) and the staff at Monahans Sandhills State Park for the opportunity to study the Monahans Pseudomasaris population. Beryl B. Simpson (The University of Texas) improved early drafts of the manuscript. Sarah Gess and an anonymous reviewer made many useful comments on the submitted manuscript. 276 Journal of Hymenoptera Research LITERATURE CITED Ajilvsgi, G. 2003. Wild/lowers of Texas, revised edition. Shearer Publishing, Fredericksburg, Texas, xix + 524 pp. Bequaert, J. 1940. Notes on the distribution of Pseudo- masaris and on the foodplants of the Masaridinae and Gayellinae (Hym., Vespidae). Bulletin of the Brooklyn Entomological Society 35: 37-45. Bohart, R. M. and L. S. Kimsey. 1982. Chrysididae in America North of Mexico. Memoirs of the American Entomological Institute 33: 1-266. Carpenter, J. M. 1982. The phylogenetic relationships and natural classification of the Vespoidea (Hy- menoptera). Systematic Entomology 7: 11-38. . 2001. Checklist of species of the subfamily Masarinae (Hymenoptera: Vespidae). American Museum Novitates 3325: 1-40. Cockerell, T. D. A. 1913. Pseudomasaris bred in California. Proceedings of the Entomological Society of Washington 15: 107. Correll, D. S. and M. C. Johnston. 1970. Manual of the Vascular Plants of Texas. Texas Research Founda- tion, Renner, Texas, xv + 1881 pp. Danforth, B. N. 1999. Emergence dynamics and bet hedging in a desert bee, Perdita portalis. Proceed- ings of the Royal Society of London, B 266: 1985-1994. Davidson, A. 1913. Masaria vespoides. Bulletin of the Southern California Academy of Science 12: 17-18. Dorr, L. J. and J. L. Neff. 1982. Pseudomasaris marginalis nesting in logs in Colorado (Hymenoptera: Masaridae). Pan-Pacific Entomologist 58: 124-128. Gess, S. K. 1996. The Pollen Wasps. Ecology and Natural History of the Masarinae. Harvard University Press, Cambridge, Massachusetts, x + 340 pp. Hicks, C. H. 1927. Pseudomasaris vespoides (Cresson), a pollen provisioning wasp. Canadian Entomologist 59: 75-79. . 1929. Pseudomasaris edwardsii Cresson, another pollen provisioning wasp, with further notes on P. vespoides (Cresson). Canadian Entomologist 61: 122-125. Hungerford, H. B. 1937. Pseudomasaris occidentalis (Cresson) in Kansas (Hymenoptera - Vespidae). Journal of the Kansas Entomological Society 10: 133-134. Machenberg, M. D. 1984. Geology of Monahans Sand- hills State Park, Texas. Bureau of Economic Geology, The University of Texas at Austin. 39 pp. ' Muhs, D. R. and V. T. Holliday. 2001. Origin of late Quaternary dune fields on the Southern High Plains of Texas and New Mexico. Geological Society of America Bulletin 113: 75-87. Parker, F. D. 1967. Notes on the nests of three species of Pseudomasaris Ashmead (Hymenoptera: Masar- idae). Pan-Pacific Entomologist 43: 213-14. Richards, O. W. 1966. New records of Pseudomasaris Ashmead (Hymenoptera: Vespoidea, Masaridae), with notes on P. phaceliae Rohwer and P. cazieri R. M. Bohart. Proceedings of the Royal entomological Society, London (B) 35: 47-55. Stephen, W. P. 1965. Temperature effects on the development and multiple generations in the alkali bee, Nomia melanderi Cockerell. Entomolo- gica Experimentalis et Applicata 8: 228-240. Tepedino, V. J., L. L. McDonald, and R. Rothwell. 1979. Defense against parasitization in mud- nesting Hymenoptera: Can empty cells increase net reproductive output. Behavioral Ecology and Sociobiology 6: 99-104. Torchio, P. F. 1970. The ethology of the wasp, Pseudomasaris edwardsii (Cresson) and a descrip- tion of its immature forms (Hymenoptera: Ves- poidea, Masaridae). Los Angeles County Museum Contributions in Science 202: 1-32. IH J. HYM. RES. Vol. 16(2), 2007, pp. 277-280 Mydrosoma micheneri Packer, new species, a New Diphaglossine Bee from Brazil (Hymenoptera: Colletidae) Laurence Packer Department of Biology, York University, 4700 Keele St., Toronto, Ontario, M3J 1P3, CANADA; email: bugsrus@yorku.ca Abstract. — Mydrosoma micheneri Packer, new species, is described and illustrated. The sole known specimen, a female, is from the Mato Grosso of Brasil and was collected almost 40 years ago. It is distinctive in having a longer head and malar space than other members of its tribe. The purpose of this paper is to describe a somewhat unusual species in the genus Mydrosoma. It differs from other species by the comparatively elongate malar space and clypeus. The genus Mydrosoma is one of three genera in the tribe Dissoglottini, the others being Mydrosomella, with two species (Graf and Urban 2001), and the monotypic Ptiloglossidia (Michener 2007). Mydrosoma occurs from southern Brazil to Mexico and none of its nine species have been collected frequently. The only biological data suggest that these may be late afternoon flying bees, with short activity periods; this could help explain their apparent rarity. In the description below, standard ter- minology for bee morphology is employed, following Michener (1986, 2007). Puncture density is indicated by the relative dis- tances between punctures in terms of interspace (i) to puncture diameter (d) ratios (e.g. i = 2d). Flagellomeres are numbered 1-10, and metasomal terga and sterna indicated by T and S, respectively. Hair length is indicated relative to the diameter of the median ocellus - MOD. Mydrosoma micheneri Packer new species Figs 1-2 Diagnosis. — The new species has the standard combination of characteristics of the tribe Dissoglottini: pre-episternal groove absent and notaulus weak or absent. It is clearly a member of the genus Mydrosoma as indicated by the presence of arolia, second and third submarginal cells subequal in area and basitibial plate in- complete. The new species is readily separated from other Mydrosoma by the comparatively elongate head, with clypeus only 1.5 times as wide as long and malar space almost as long as basal depth of mandible (Fig. 1). Description. — Female. Body length 14 mm, forewing length 9 mm, head width 3.05 mm, intertegular span 2.8 mm. Colouration: Black with lower face, antenna, legs and metasoma dark brown; following parts orange: anterior surface of flagellum (except F2 red-brown), entire apical flagellomere, tegula, fore tibia and fore tarsus, wing veins; wing membrane pale amber; metasomal terga with metallic reflections; apical impressed areas straw. Pubescence: Hairs plumose with nu- merous short branches. Bright fuscous on dorsal and lateral surfaces of mesosoma, pale yellowish on face, ventral surface of mesosoma, legs and metasoma. Outer surface of hind tibia with brown hairs. Prepygidial fimbria dark brown, hairs on disks of T2-T5 blackish. Hairs on face short, 1.5MOD; slightly longer on vertex, genal area and mesoscutum, 2MOD; longer on mesopleuron, scutellum and metano- tum <3MOD; longest hairs on lateral 278 Journal of Hymenoptera Research Fig. 1. Lateral habitus of Mydrosoma micheneri, Packer, n. sp. surface of propodeum and hind femoral scopa, 4MOD. Hind tibia with shorter hairs <2MOD except longer on ventral surface <3MOD. Tl with simple erect hairs 2MOD, longer and plumose laterally, 3MOD. Apical bands of appressed hairs on T2-T4 <2MOD. SI with short erect hairs, 1.5MOD; S2-S5 with posteroven- trally directed long subapical hair bands, hairs with numerous branches on anterior of rhachis only, longest on S3, 3.5MOD. Surface sculpture: Microsculpture weak, surface shiny, except somewhat dull on lower face. Punctures on face below anten- nae distinct and moderately dense i ~ d; frons with punctures finer and shallower but equally dense; narrow transverse im- punctate band between antennal bases; area between lateral ocellus and compound eye and vertex immediately behind ocelli al- most impunctate. Mesoscutum with shal- low, moderately dense punctures, i ^ d; scutellum with deeper, more distinct and irregularly spaced punctures, i = l-3d; metanotum with punctures almost crowded laterally, i > d medially; dorsal area of propodeum impunctate, lateral surface densely punctate above and along ventral margin, i < d, sparsely punctate below; mesopleuron with dense, somewhat effaced punctures, almost crowded dorsally, i < 1.5MOD below; Tl with punctures sparse on disk, i = 2-4d, denser towards sub- marginal zone i = l-2d; punctures increas- ingly dense on more posterior terga, i ^ d on T5. Apical impressed areas with tiny dense punctures apically. Structure: Head as wide as long. La- bium flat with weak transverse basal ridge. Mandible three times as long as basal depth, subapical tooth short. Galeal comb with approximately 28 teeth. Clypeus 1.5 times as wide as long, apical rim slightly upturned. Supraclypeal area 1.25 times as Volume 16, Number 2, 2007 279 Fig. 2. Frontal view of head of M. micheneri Packer n. sp. to show elongate mandible, clypeus and malar area. long as apical breadth. Interantennal dis- tance 0.75 X antennocular distance. Malar space long, —0.87X as long as basal depth of mandible. Genal length subequal to width of compound eye; longitudinal axis of compound eye just posterior to mid- depth of mandible; almost at right angle to axis of mandible base. Facial fovea in- distinct, marked by weak ridge at inner margin of upper paraocular area approxi- mately 1.5MOD in length. Ocellocular distance subequal to interocellar distance and less than twice MOD. Vertex behind lateral ocellus subequal in length to MOD, weakly depressed. Fl equal to combined lengths of F2 and F3; F2-F9 somewhat broader than long, F10 almost twice as long as wide. Notaulus not evident. Scutellum weakly depressed medially. Dorsal surface of propodeum convex, mostly declivous. Basal vein thickened for apical half. First recurrent vein enters second submarginal cell two vein widths from lr-m. Posterior margin of second submarginal cell 10% longer than that of third submarginal cell. Basitibial plate indicated by posterior carina 2.5MOD in length. Hind basitarsus twice as long as greatest depth, dorsal margin strongly and ventral margin weak- ly convex. Jugal lobe slightly less than half as long as vannal lobe. Male. — Unknown. Etymology. — It is a pleasure to name this species after Charles Michener in recogni- tion of his stellar achievements in melittol- °gy- Material Studied. — Holotype female: BRAZIL: Mato Grosso 12 50'S 51 47'VV, 2.iv.l968. O.W. Richards. A second label states: R.S. & R.G.S. Expedition B.M. 1968- 260. A third label states "Gallery forest". A fourth, handwritten, label states "Paracol- letini, n. gen. N. sp.!" [Though the species is clearly a member of the Diphaglossinae 280 Journal of Hymenoptera Research as indicated by the tiny stigma and strongly bifid glossa.] The specimen be- longs to the Natural History Museum (BMNH). Comments. — This species keys out to M. inusitatum (Snelling) in Michener (1986), but can be easily distinguished from that species by the malar area which is longer than in all other species of Mydrosoma , but linear (absent) in M. inusitatum. It might seem to belong to Friese's Bicornelia (sunk within Mydrosoma by Michener 1986; see also Snelling 1980), although that species group is defined based upon secondary sexual characteristics of males, which are unknown for M. micheneri. It does not key out to either species of "Bicornelia" using Michener (1986). The locality where the species was collected is now largely agricultural, al- though gallery forest remains along water- courses. There is a substantial area of forest to the west of the type locality however, the Parque Nacional do Xingu. It is possible that this species may persist in this region. ACKNOWLEDGEMENTS I am grateful to George Else (BMNH) for the opportunity to borrow the specimen described herein. Jason Gibbs prepared the images in figures 1 and 2, for which I am grateful. Funding for my research is provided by the Natural Science and Engineering Research Council of Canada. LITERATURE CITED Graf, V. and D. Urban. 2001. Mydrosomella cleia, uma especie nova do sul do Brasil (Hymenoptera, Colletidae). Acta Biologica Paranaense 30: 1715-179. Michener, C. D. 1986. A review of the tribes Diphaglossini and Dissoglottini (Hymenoptera: Colletidae). University of Kansas Science Bulletin 53: 183-214. — . 2007. The bees of the world [2"d edition]. Johns Hopkins University Press, Baltimore, Mary- land. Snelling, R. R. 1980. The genus Bicornelia (Hymenop- tera: Colletidae). Contributions in Science, Natural History Museum of Los Angeles County 327: 1-6. J. HYM. RES. Vol. 16(2), 2007, pp. 281-292 Phenology and Social Organization of Halictus (Seladonia) tripartitus (Hymenoptera: Halictidae) Laurence Packer, Anne-Isabelle D. Gravel and Gretchen Lebuhn (LP, A-IDG) Department of Biology, York University, 4700 Keele St., Toronto, Ontario M3J 1P3, CANADA (GL) Department of Biology, San Francisco State University, San Francisco, CA, 94132, USA Abstract. — We present data on the social biology of Halictus (Seladonia) tripartitus Cockerell based upon samples from California. This bee is at least partially eusocial and overwintered gynes and first brood workers differed in size by 5.14%. 35.5% of the first brood workers had developed ovaries, 11.2% had ovaries with the equivalent of at least one fully developed oocyte but less than 14% were mated. In July and August, female reproductive options seemed highly variable: approximately 55% were sterile workers, 2.6% had better developed ovaries than spring foundresses and over 20% had at least the equivalent of one fully developed oocyte whereas an additional 20-30% may have been capable of overwintering as gynes. Ovarially developed workers were larger than those that remained sterile. Sixteen to 28% of the late summer workers were mated, but mated individuals were not more likely to have developed ovaries or to be larger than unmated bees. We discuss these findings in the light of the climate of the study area and compare this species with other members of its subgenus. Detailed assessment of the social organi- sation of halictine bees requires painstak- ing field observations and nest excavations over the course of several flight seasons. Nonetheless, useful information can be obtained from less detailed analyses. Be- cause of the evolutionary lability of social- ity in these bees, even comparatively superficial data can be of utility in terms of plotting presence or absence of sociality upon a phylogeny or in obtaining estimates of sociobiologically important variables such as levels of morphological and phys- iological caste differentiation (e.g. Dunn et al. 1998). The subgenus Seladonia (of the genus Halictus) contains species that are primarily solitary, such as H. (S.) virgatellus Cockerell (Eickwort et al. 1996), some with weak eusociality (Michener's 1974 terminology for social categories is used herein) as in H. confnsus Smith (Dolphin 1971) to others with the largest colony sizes ever re- corded for a halictine - H. (S.) lutescens Friese (Sakagami and Okazawa 1985) or the largest morphological caste differenti- ation - H. (S.) land Moure (Janjic and Packer 2001). However, it should be cau- tioned that data for these, and most other species in the subgenus, are mostly frag- mentary. Halictus (Seladonia) tripartitus Cockerell is a halictine for which the only sociobiolog- ical data available are i) the surprising observation that, at a nest aggregation, smoke blown down one nest entrance came up out of the other entrances, demonstrating that the nests were con- nected underground (Eickwort personal communication 1988) and ii) a statement that the species is solitary (Amdam et al. 2006). It is a somewhat divergent member of the subgenus, forming the sister group to the remainder (Danforth et al. 1999), probably along with two rare species, H. harmonius Sandhouse and H. pinguismentus Janjic and Packer (Janjic and Packer 2001, Janjic and Packer unpublished data). 282 Journal of Hymenoptera Research For a halictine population with an species of bee (Leong and Thorp 1999), the archetypal eusocial phenology (and indeed bee fauna was surveyed using a recently other annual eusocial insects in temperate developed standard protocol which uti- climates), only mated and comparatively lizes a mix of sweep-netting and pan unworn females are expected to survive trapping (protocols can be found at the winter. In spring they establish nests, ). Pan their wings and mandibles increase in traps were set out between 09:00 and 15:00 wear and their ovaries become well de- and sweep-netting was done for one hour veloped as they produce a brood com- in the morning and one hour in the posed primarily of worker females. Indi- afternoon per sampling day. viduals in this brood are usually smaller Some samples of females from 2003 were than their mother and have reduced levels preserved in 70% alcohol and this permit- of ovarian development and a low fre- ted dissection for sociobiological data (see quency of mating (partly dependent upon below). All samples collected in 2002 were the availability of males produced in the pinned so only phenological data could be "worker" brood). Depending upon the obtained from them. length of the flight season, additional Sociobiological Data. — The preserved bees broods, primarily of workers, may be were observed under a microscope to produced. Towards the end of the colony evaluate alar and mandibular wear (based cycle, a brood composed of males and next upon the right hand side), measure head year's colony foundresses is produced, width, and dissect females for insemina- Much of the variation in these parameters tion and ovarian development status fol- can, to some extent, be estimated from field lowing slight modifications of standard collected samples in the absence of nest protocols (Ordway 1965, Abrams and excavations (Dunn et al. 1998). In this Eickwort 1980). Mandibular wear was paper we present some phenological and scored on a scale from 0 - completely social organisation data for H. tripartitus unworn, apex sharp; to 6 - mandible worn based upon field samples, in order to add away to the base of the subapical tooth, to the intensity of taxonomic sampling for Wing wear was assessed by counting the such data in these behaviourally diverse number of nicks in the wing margin, bees. completely abraded wings were scored as having 15 nicks - the number that seems to MATERIALS AND METHODS result in complete abrasion of the margin. Sampling. — Halictus tripartitus adults Total wear was calculated by adding the were collected on the Kunde Wine Estate, scores for wing and mandibular wear. It is 2.3 km southeast of Kenwood in Sonoma useful to exclude freshly emerged adults County, California, USA (38°24'15N, from some analyses, such as of ovarian 122 31'43W). This site is a low-elevation, development and mating, because they gently sloping oak woodland dominated may not have had enough time to mate by blue oak (Quercus douglasii), adjacent to or develop their ovaries. Consequently, in a large vineyard. The oak woodland con- many analyses only bees with a total wear sists of 4-10 m tall trees that are generally score greater than one were considered, widely spaced with few shrubs inter- Comparisons of ovarian development spersed among them. At the site, a 100 m2 among samples were performed with bees sampling plot was established. parasitized by conopids removed from the Bees were sampled in 2002 and 2003 on sample. In later samples, when a mixture clear, low wind days. As preliminary data of ageing workers and newly emerged and previous monitoring studies show that gynes might be expected, putative exam- different colored pan traps attract different pies of the latter were assumed to have Volume 16, Number 2, 2007 283 a total wear of at most one and to have entirely undeveloped ovaries. Individuals with a total wear of three or more and with completely undeveloped ovaries were as- sumed to be permanently sterile workers in these samples. This protocol seemed appropriate in this study as these bees appear to become worn rather slowly, their wings seem to become "nicked" particu- larly slowly. The spermatheca of each bee was ob- served for sperm, which make the other- wise glassy-transparent spermatheca opa- que. On the few occasions when there was uncertainty as to whether a bee had mated or not, the spermatheca was gradually squashed between a cover slip and a mi- croscope slide under high magnification to look for spermatozoa. Ovarian develop- ment was evaluated by estimating the proportion of a fully developed oocyte present in the bees and summing them across the 6 ovarioles. Bees with developed ovaries (defined as with at least one ovariole with at least one quarter of a fully developed oocyte) are sometimes referred to as OD+ in the account that follows, bees without ovarian development are referred to as OD — . Caste size dimorphism is estimated as (q-w)/q where q is mean putative queen head width and w is mean putative worker head width. Climatic data. — Because weather varia- tion, both within and among years, influ- ences sociobiological parameters for sweat bees (Richards and Packer 1995), we report temperature and rainfall data for the study area both in terms of data from the period when bees were sampled and for longer term average conditions. Data were taken from the Santa Rosa Airport Weather Station as indicated on the following web- pages: http://www.pressdemocrat.com/ nbwx/srweather/index.cfm and http:// www.wunderground.com/history/airport/ KSTS/1993/3/l/MonthlyHistory.html# calendar. Data presented here are for mean daily temperature and rainfall for the months of March through August 2002 and 300 -i r 18 co250" ® 200 - w \ \ \ \ - 16 - 14 - 12 CO E 150 - / / t 1 \ \ \ \ - 10 - 8 £ ioo- ^ - 6 - 4 50 - 0 - • - 2 1 1 V May Jun Jul Aug Sep Oct (f) CO * Fig. 1. Phenology of H. tripartitus. The left-hand axis indicates the number of females (•) while the right axis shows the number of males (A), all collected in 2002. 2003, along with averages and maxima and minima for the same months over the 10 year period 1992-2001. RESULTS Phenological Patterns. — During the May- September period, 2002, a total of 765 H. tripartitus were collected at the Kunde Wine Estate. Of these, 726 were females and 39 were males. Numbers of H. triparti- tus females increased gradually, nearly at a constant rate, from May to mid-July (Fig. 1) peaking at the beginning of August after a temporary decrease in sample size, and then decreased rapidly until the end of the season. Males show a similar pattern, but peak in the sampling period after the females. The discrepancy between the number of males and number of females in the samples is surprising; even at the peak of male abundance, there are approx- imately 8 times as many females as males. Based upon observations of museum hold- ings, this ratio would seem not to be an artefact as males of H. tripartitus are rare in collections (Packer unpublished observa- tions). The alcohol preserved material demon- strates that the species is active as early as March (see below), and inspection of museum holdings shows that a few fe- males can be found as late as October (Packer unpublished observations). The March 28,h sample of dissected bees suggest that nest initiation had only just begun at this time: only 39% of the bees 284 Journal of Hymenoptera Research 3.5 3 A A c 2.5 0) E Q. o 2 0) > ? 15- A A A A A A A A A A * k A A A (0 1 1 * A A ^ A ' A A 0.5 A U 1 — 1.8 i 1.9 1 * 2 2.1 2.2 2.3 Head width (mm) Fig. 2. Ovarian development and size variation in females of the May sample. had mandibular wear, 16% had wing wear and 34% of them had some ovarian de- velopment with mean summed oocyte fractions of 0.16 per individual (Table 1). There was no size difference between ovarially developed and undeveloped fe- males (mean head width OD+ females = 2.04 mm, SD = 0.11, n = 54; OD- females head width = 2.06 mm, SD = .093, n= 28; t = 0.93, ns). Almost all females mated (97.2%), both unmated females were above average size for the sample. In contrast, the May 16th sample was entirely of mated bees with worn mand- ibles, 94% had worn wings and all except a parasitized individual had developed ovaries with the summed fractions of developing oocytes averaging over 1.6 per bee. Ovarian development was indepen- dent of bee size (Fig. 2). The mean size of March and May sample bees was 2.04 (SD = 0.10) and 2.05 mm (SD = 0.09) respec- tively, not significantly different (t = 0.35, p > 0.5). Table 1. Summary sociobiological data by sample date. % of 9 with % of 9 with ".i of 9 with* Mean ovarian Mean 9 Head Sample date N mandible wear wing wear % of 9 mated developed ovaries score* width March 28 82 39 16 97.2 34 0.16 2.04 May 16 31 100 94 100 100 1.6 2.05 June 20 67 81.5 26 13.8 43 0.4 1.93 June 30 50 96 62 6.5 26 0.14 1.94 July 18 166 68.7 70.5 13.7 55 0.44 2.01 August 20 118 56 36.2 34.2 34.5 0.28 2.01 I fema riitted from ovarian development data. Volume 16, Number 2, 2007 285 The two samples from June show an increased variance in wear but decreased ovarian development in comparison to the May sample. On June 20* 81.5% of the bees had worn mandibles, 26% had worn wings, 43% had developed ovaries and 13.8% were mated. This was the only sample in which unworn bees were signif- icantly less likely to have developed ovaries (x2 = 7.18, P < 0.01; all other samples p > 0.16), suggesting that many of these individuals were young and had not had time to develop their ovaries. None- theless, worn bees with undeveloped ova- ries made up 46.5% of this sample. On June 30th comparable data are: 96% with worn mandibles, 62% with worn wings, 26% with developed ovaries (31% if only worn bees are included) and 6.5% were mated. These data suggest increased average age/ activity levels of the bees over the in- tervening ten days but a decrease in ovarian development. The average sum of fractions of oocytes shows the same de- crease over time, the values were 0.40 for [une 20th and 0.14 for June 30th. Of all June bees combined, only one individual (ie < 1%) had more enlarged ovaries than the average individual in the May sample. Ovarially developed worn bees were significantly larger than worn OD- bees in the June 30th sample (mean head width OD+ bees = 1.98 mm, SD = 0.076, n - 11; OD- = 1.92 mm, SD = .089, n = 28; t = 2.1, p< 0.05), but not in the June 20th one (OD+ = 1.95 mm, SD = 0.94, n = 23, OD- = 1.93 mm, SD = .096, n = 19; t = 0.91, ns). The proportion of bees that were mated did not differ between the two June samples (Fisher's exact test, p = 0.35). The data suggest that these early sum- mer bees are offspring of the generation sampled in March and May and that they represent a worker brood. The pattern of ovarian development and age in these bees suggests that more workers initiate ovarian development soon after eclosion than manage to maintain developed ovaries a short while later. Furthermore, the data indicate that individuals that do maintain developed ovaries, are disproportionately larger than those that do not. Bees in the two June samples did not differ in size (1.93 mm and 1.94 mm for June 20th and June 30th respectively). Three individuals collected in June were extreme- ly worn, (wing margin completely abrad- ed), suggesting that they may be ageing overwintered females, perhaps from mul- tiple foundress associations. All three were mated, one was parasitized with a conopid larva (Diptera, Conopidae), one had no ovarian development, the remaining in- dividual had Va of a developed oocyte. Inclusion of these individuals in the com- parison of ovarian development data be- tween worn and unworn bees did not alter the statistical patterns noted above. The July 18th sample had a lower pro- portion of bees with mandibular wear than in June (68.7%), a higher proportion with worn wings (70.5%) and a higher pro- portion with developed ovaries (55%, 54.4% when only worn bees are included). Similar to the June 20th sample, 13.7% of females were mated. The wear differential between mandibles and wings suggests that bees at this time are performing relatively more foraging or less digging than in the earlier samples. The average sum of fractions of oocytes was 0.44 and 6 individuals (3.8%) had more ovarian de- velopment than the average overwintered female in May. Worn OD+ bees were significantly larger than worn OD- bees (head width OD+ = 2.02 mm, SD = 0.07, n = 82; head width OD- - 1.98 mm, SD = 0.09, n = 44 respectively; t = 2.56, p < 0.05). Overall, July bees were intermediate in size between the smaller June and larger earlier samples, averaging 2.01 mm in head width; the sample of six individuals with very well developed ovaries had exactly the same average head width as the rest of the sample. On August 20th the proportion of bees with worn mandibles decreased to 56% while 36.2% had worn wings. The pro- 286 Journal of Hymenoptera Research Wear and Reproductive Data for Different Sample Dates q 0.5 in all comparisons). Similarly, the mated bees were not larger than unmated ones (worn bees only) in any of the analyses (p > 0.5 in each case) except for the July sample in which the mated bees were significantly larger than the unmated ones (head width mated = 2.05 mm, SD = 0.08, n = 14, head width unmated = 1.99 mm, SD = 0.06, n - 75; t = 7.14, p << 0.001). Parasitism. — One female was found with a large nematode in the metasoma, she was from the March sample of overwintered females. Twenty females were parasitized with one conopid larva in each metasoma and three contained two parasite larvae. It is likely that these parasitism rates are underestimates as small conopid larvae and nematodes would likely have been missed in the dissections, especially if they were primarily underneath the first meta- somal tergum, which was not removed. Of the 23 females with conopids, one was found in the May sample, four each in the two June samples, eight in July and six in August. The overwintered females were significantly less affected by conopids than were later females (Fisher's exact test p = 0.0377). The long period during which conopid larvae were found suggests either that the parasite has more than one generation per year, or that multiple species of conopid are involved. Bees with conopid larvae did not differ in size from the other individuals in their samples (p >0.2 in all cases), but they did have more wear than their contemporaries in July and August (Mann Whitney U test, U = 991, p = 0.007 and U = 536, p = 0.014 respectively) but not in either of the June samples (U = 191, p = 0.059 and U = 130, p = 0.187). Climate and Weather. — The collection pe- riod for H. tripartitus was from May to August 2002 for the pinned specimens and from March to August 2003 for the pickled ones. The weather data for the months March to August are shown in Fig. 5 along with averages for the previous ten year period. 2002 was drier, but 2003 was wetter than the average for the previous ten year period. In both years April was an extreme month, with more than double the average rainfall in 2003 and less than a quarter the normal amount in 2002. The temperature data show that 2002 varied little from the average conditions whereas in 2003 April was cooler than any year between 1992 and 2002 and July was warmer than in any of these other years. DISCUSSION Halictine bees are well known for their variable social behaviour, not only within and between populations and species but also among individuals within a colony. Taken in their entirety, our data strongly support the view that H. tripartitus is primarily eusocial at our study site. Per- haps the strongest evidence for this comes from all overwintered females being mated whereas the first brood offspring were almost entirely unmated. Indeed, as a whole, the data are mostly consistent with the view that this species is a typical annual eusocial species with overwintered females initiating nests in March and Volume 16, Number 2, 2007 289 35 30 25 20 O (0 2 15 £ fi 10 0 march april may june July august - - •- - average lowest highest - -a- - 2002 a 2003 5 - 1 4 JS 3 a D _i i march april may lune July august 2003 □ average 0 2002 Fig. 5. Weather data for 2002, 2003 and summaries for the ten previous years 1992-2001. foraging until May, workers emerging in June and overwintering gynes commenc- ing emergence in July. Less usual features are the apparent multiple nest entrances, low frequency of males and the nature of the sample in August. We return to these three aspects towards the end of the discussion after considering worker fitness options and making comparisons between H. tripartitus and other species of the subgenus Seladonia. Fitness options for bees eclosing in summer within a eusocial family structure include remaining as a sterile worker, attempting to produce some offspring di- rectly or initiating a nest either the same year as they eclose (Richards et al. 2003) or after overwintering (Yanega 1988). Given 290 Journal of Hymenoptera Research the Mediterranean climate of the region and the dryness and hardness of the soil in summer, initiating a nest in summer is probably not a very profitable option (McCorquodale 1989). Furthermore, as the most ovarially developed summer bees were unmated, it is unlikely that they are founding new nests at this stage, unlike some summer females of Halictus (Halictus) sexcinctus (Richards et al. 2003). Halictus tripartitus worker brood individ- uals seem to take advantage of at least three of these options. Of the July and August samples, approximately 55% seem to be sterile workers and perhaps 20% of them seem capable of reproducing. Over 2% of the bees collected in July and August had ovaries more fully developed than did the foundresses in spring, suggesting that either some are nesting solitarily or they have the potential to reproduce in their natal colony (as has been demonstrated using genetic markers in several other species, Packer and Owen 1994, Richards et al. 1995). As none of these most ovarially developed later workers were mated, the offspring they produce must be male. However, few males are ever seen in this species. This would be in agreement with Packer and Owen (1994) where genetic data showed that considerable levels of ovarian development resulted in few in- stances of successful oviposition by work- ers of Lnsioglossum laevissimum. The fourth option, mated worker brood females overwintering and initiating a nest the following spring, cannot be confirmed or refuted with the data at hand. It remains possible that brood divalency may occur in H. tripartitus. The easiest way to document this would be to mark large numbers of emerging "worker" brood females as they leave the nest and search for them when the foundresses begin activity the follow- ing spring. Few species of the subgenus Seladonia have received detailed sociobiological in- vestigation. The best studied is H. (S.) hesperus Smith, which has large colonies and very large morphological caste differ- entiation with concomitantly low worker ovarian development (<1% of workers seemingly capable of reproduction, Brooks and Roubik 1983, Packer 1985). Halictus (S.) lutescens has even larger colony sizes and in the one nest excavated by Sakagami and Okazawa (1985) approximately one quarter of the workers seemed capable of repro- duction, though it remains likely that this colony was studied after the death of the queen. A third species, H. (S.) lanei, has the largest caste size dimorphism of any halictine known (Janjic and Packer 2001), although its social organisation may not fit the standard eusocial model (Gravel et al. in preparation). All three of these species are tropical and their workers seem not to have the range of options that may be available to H. tripartitus. In contrast, the alpine H. (S.) virgatellus is predominantly solitary with a few nests being shared (Eickwort et al. 1996). The north temperate H. (S.) confusus and its sibling species H. (S.) tumulorum (L.) are weakly eusocial, perhaps reverting to solitary behaviour at the northern edge of their ranges (Dolphin 1971, Sakagami and Ebmer 1979). Worker brood individuals in these species likely have a similar range of options as sug- gested above for H. tripartitus, albeit usually within a shorter summer activity period, but appropriately detailed studies remain to be published. Richards and Packer (1995) found that variations in local weather patterns affect- ed the demography and behaviour of H. (H.) ligatus colonies: warmer and drier weather resulted in larger broods with better survival rates and lower nest-failure and more reproduction by workers. Thus, warm, dry weather resulted in weaker eusociality. Conversely, colder, wetter con- ditions gave rise to smaller workers with greater reproductive differentiation be- tween the castes but also lower brood- survival rates and more nest-failure. Colder and wetter weather resulted in stronger eusociality. The area where H. Volume 16, Number 2, 2007 291 tripartitus was studied was unusually cold and wet in April of 2003, during foundress provisioning. If the weather influences this species in the same manner as it does with H. ligatus, it is likely that early summer colonies were smaller, with stronger phys- iological caste differentiation than is usual for this population. The poor weather in April may also explain the apparently extended period of activity of overwin- tered foundresses. There are no detailed studies of social sweat bees from Mediterranean climates in North America. The restriction of precipi- tation to the winter and early spring is a feature that can limit the duration of colonies of social bees in the area and few individuals of H. tripartitus have been collected from late August onwards: even though temperatures are adequate for activity, the dry conditions severely re- stricts the availability of forage. It is likely that the late summer activity of this species is a comparatively recent phenom- enon as at this time of year they rely upon flowering of plants that are maintained by agricultural irrigation. This could ex- plain the unusual nature of the August sample. Two other unusual aspects of this spe- cies' biology are worthy of comment, and may be related. First, the very low pro- portion of males, either as sampled here or in museum collections, is surprising. To- wards the end of the colony cycle in most eusocial halictines, males are approximate- ly as common as females, yet in H. tripartitus they are almost an order of magnitude less common. Second, the ob- servation of nests apparently being con- nected underground is entirely unique among bees. It remains possible that males primarily search for mates underground, thereby uniting these two unusual phe- nomena. Certainly this species is worthy of more detailed field investigations, some- thing we hope this paper will stimulate. Lastly, the observation that worker ovar- ian development seemed to decline over time is of interest. This phenomenon has been observed in numerous social insects. There are two main reasons this might be the case. First, the act of work might make it less likely that a bee has the energy resources to develop oocytes - which are remarkably large in comparison to bee body size in halictines. Alternatively, it could be that workers develop their ovaries primarily to coincide with male production when reproductive broods are protan- drous. The fact that first brood workers are active at a time of year when few males are being produced in H. tripartitus, argues against the second explanation. AKNOWLEDGEMENTS The senior author's research is funded by the Natural Sciences and Engineering Research Council of Canada. We are grateful for the comments of Miriam Richards on an earlier version of this manuscript and to Luana Sciullo for processing the weather data. The junior author's research is funded by the Integrated Hardwoods Range Management Program. We also thank Cynthia Fenter, Erin Rentz and many SFSU students for their work collecting these insects. LITERATURE CITED Abrams, J. and G. C. Eickwort. 1980. Biology of the communal sweat bee, Agapostemon virescens (Hy- menoptera: Halictidae) in New York state. Search (Cornell University Agriculture Experiment Station) 1: 1-20. Amdam, G. V., A. Csondes, M. K. Fondrk, and R. E. Page, Jr. 2006. Complex social behaviour derived from maternal reproductive traits. Nature 439: 76-78. [Cover photo caption associated with article.] Brooks, R. W. and D. W. Roubik. 1983. A Halictine bee with distinct castes: Halictus hesperus (Hymenop- tera: Halictidae) and its bionomics in Central Panama. Sociobiology 7: 263-282. Danforth, B. N. 2002. Evolution of sociality in a primitively eusocial lineage of bees. Proceedings of the Natural Academy of Science of the United States of America 99: 286-290. , H. Sauquet, and L. Packer. 1999. Phytogeny of the bee genus Halictus (Hymenoptera: Halictidae) based on parsimony and likelihood analyses of nuclear EF-loc sequence data. Molecular Phyloge- netics and Evolution 13: 605-618. Dolphin, R. E. 1971. Observations of Halictus confusus Smith (Hymenoptera: Halictidae) on woodland 292 Journal of Hymenoptera Research and field flowers. Proceedings of the Indiana Academy of Science 81: 182-185. Dunn, M, P. L. Mitchell, and L. Packer. 1998. Phenology and social biology of two sibling species of Halictus in an area of sympatry. Canadian Journal of Zoology 76: 2207-2213. Eickwort, G. C, J. M. Eickwort, J. Gordon, and M. K. Eickwort. 1996. Revision to solitary behavior from eusocial ancestry in the sweat bee Halictus rubi- cundus in the Rocky Mountains, and its implica- tions for high-althtude and high-latitude adapta- tions and holarctic distributions of social insects. Behavioral Ecology and Sociobiology 38: 227-233. Janjic, J. and L. Packer, L. 2001. New descriptions of Halictus (Seladonia) from the New World (Hyme- noptera: Halictidae). journal of Hymenoptera Re- search 10: 55-75. Leong, J. M. and R. W. Thorp. 1999. Colour-coded sampling: the pan trap colour preferences of oligolectic and nonoligolectic bees associated with a vernal pool plant. Ecological Entomology 24: 329-335. McCorquodale, D. B. 1998. Soil softness, nest initiation and nest sharing in the wasps, Cerceris antipodes (Hymenoptera: Sphecidae). Ecological Entomology 14: 191-196. Michener, C. D. 1974. The Social Behavior of Bees: A Comparative Study. Harvard University Press, Cambridge, MA. Ordway, E. 1965. Caste differentiation in Augochlorella (Hymenoptera, Halictidae). Insectes Sociaux 12: 291-308. Packer, L. 1985. The social organisation of two halictine bees from southern Mexico with notes on two bee-hunting philanthine wasps. Pan- Pacific Entomologist 51: 291-298. and R. E. Owen. 1994. Relatedness and sex ratio in a primitively eusocial halictine bee. Behavioral Ecology and Sociobiology 34: 1-10. Richards, M. H. and L. Packer. 1995. Annual variation in survival and reproduction of the primitively eusocial sweat bee Halictus ligatus (Hymenoptera: Halictidae). Canadian Journal of Zoology 73: 933-341. , L. Packer, and J. Seger. 1995. Unexpected patterns of parentage and relatedness in a prim- itively eusocial bee. Nature 373: 239-241. -, E. J. von Wettberg, and A. C. Rutgers. 2003. A novel social polymorphism in a primitively eu- social bee. Proceedings of the National Academy of Sciences of the United States of America 100: 7175-7180'. Sakagami, S. F. and P. A. W. Ebmer. 1979. Halictus (Seladonia) tumulorinu higashi ssp. Nov. from the Northeastern Palaearctic (Hymenoptera: Apoi- dea; Halictidae). Kontyu 47: 543-549. and T. Okazawa. 1985. A populous nest of the Halictine bee Halictus (Seladonia) lutcsccus from Guatemala (Hymenoptera, Halictidae). Kontyu 53: 645-651. Yanega, D. 1988. Social plasticity and early-diapaus- ing females in a primitively social bee. Proceedings of the National Academy of Sciences of the United States of America 85: 4374-4377. J. HYM. RES. Vol. 16(2), 2007, pp. 293-296 The Status of Liris magnificus Kohl, 1884, and Trachogorytes costaricae R. Bohart, 2000 (Hymenoptera: Crabronidae: Crabroninae, Bembicinae) WOJCIECH J. PULAWSKI Department of Entomology, California Academy of Sciences, 875 Howard Street, San Francisco, California 94103, USA; email: wpulawski@calacademy.org Abstract. — The Australian Liris magnificus Kohl, 1884, currently treated as a subspecies of Liris haetnorrhoidalis (Fabricius, 1803) from the Palearctic and Afrotropical Regions, is an independent, full species. The taxonomic history of the species is reviewed and the differences with haemorrhoidalis are discussed. Trachogorytes Bohart, 2000, a monotypic genus described for Trachogorytes costaricae Bohart, 2000, is actually a junior synonym of Mellinus Fabricius, 1790. Mellimis costaricae Bohart, 2000, comb, nov., is redescribed. Liris magnificus Kohl by Cardale (1985) and Naumann (1993). Liris magnificus Kohl, 1884:356, E (as magnifica, Bohart and Menke (1976)' on the other incorrect original termination). Holotype or hand, regarded the subspecific status of syntypes: E, northern Australia: no specific magnificus as tentative, locality (Naturhistorisches Museum Wien). - Having recently examined five females Kohl, 1892:228 (in key to world Liris s.s.); Turner, and six males of magnificus, I conclude that 1908:473 (as new synonym of Liris haemorrhoi- it is actually a full species, and not a geo- dalis); Dollfuss, 1989:10 (type material in graphic form of haemorrhoidalis. The differ- NHMW). - As Larra magnifica: Kohl, 1885:245 u . .u ( n T ; , , ,,. r , , r x ences between them are as follows, in (new combination, m checklist of world Larra); .r. , .. ... , ,, ^ ,, T 1on„„„ ,. . , , ., magnificus, the median swelling of the Dalla Torre, 1897:669 (in catalog of world * J ' & Hymenoptera). - As Liris haemorrhoidalis Perioral collar is wider (Fig. 2b); in the magnifica: Williams, 1928:49 (new status, nest- female, the carina emerging from the ing habits); Bohart and Menke, 1976:245 (as clypeal lobe corner is about twice as long tentative subspecies of Liris haemorrhoidalis); as the midocellar width (Fig. 2a) and the Cardale, 1985:235 (in catalog of Australian dark apical coloration of fore wing does Sphecidae); Naumann, 1993:185 (Australia: not extend into the cell area (Fig. 2c); the Queensland: Heathlands area in Cape York). male hind tarsomere II is simple, as in most Lin's magnificus was described as a full of the congeners; in most males the hind species, but was synonymized with hue- coxa is concave ventrally and carinate morrhoidalis (Fabricius) by Turner (1908). along inner margin (Fig. 2d), but slightly Surprisingly, the author stated "I cannot convex ventrally and obtuse along inner detect any appreciable difference in the margin in one specimen from Wonga male", but in fact the males are strikingly Beach. In haemorrhoidalis, the median swell- different (see below). Both species, howev- ing of the pronotal collar is narrower er, are similar in having a non-emarginate (Fig- lb); in the female, the carina emerging posterior mandibular margin, red legs, and horn the clypeal lobe corner is about as conspicuously golden body setae, a combi- long as midocellar width (Fig. la) and the nation unique within the genus. Williams dark apical area of fore wing covers (1928) treated magnificus as a subspecies of marginal as well as second and third haemorrhoidalis, an interpretation followed submarginal cells (Fig. lc); in the male the 294 Journal of Hymenoptera Research ' -sF&.iiiiiiB ■...■— ■■*•" ^ i ^i\ ■■ - «&.' wFm' < ■"I 1 mm * 1 mm Fig. 1. Liris haemorrhoidalis (Fabricius): a - lateral carina of female clypeal lobe in oblique view; b - female pronotum; c - apical half of female fore wing; d - male hind tarsomere II. Fig. : Liris magnificus Turner: a - lateral carina of female clypeal lobe in oblique view; b - female pronotum; c - apical half of female fore wing; d - male hind coxa. Volume 16, Number 2, 2007 295 hind coxa is neither concave nor carinate, and hind tarsomere II is conspicuously expanded (Fig. Id), a unique such feature in the genus. The genitalia appear identical in both species. The two species do not occur sympatrically: Lin's magnificus is known only from Australia, whereas hae- morrhoidalis occurs throughout Africa, Spain, the Canary Islands, and southwest- ern Asia to western India and Sri Lanka. Records (all specimens are at the California Academy of Sciences). — AUSTRALIA: Queens- land: Armstrong Beach ca 15 km E Sarina at 21 27.3'S 149 17.5'E, 29 Oct. 2006, W.J. Pulawski (1 e?); Balgal Beach 51 km NW Townsville at 19 02.5'S 146°25.2'E, 18 May 2007, V.E. Ahrens and W.J. Pulawski (1 9); Blacks Beach ca 8 km N Mackay at 21 03.6'S 149 ll'E, 1 Dec 2006, W.J. Pulawski (1 J); Burdekin River 20 km NE Charters Towers at 20 00.1 'S 146 26.3'E, 26 Nov 2006, W.J. Pulawski, and 21-22 May 2007, V.E. Ahrens and W.J. Pulawski (1 9, 1 $); Crystal Cascades 10 km W Cairns, 9-10 July 1983, TW. Davies (1 9); 69 road km WNW Mount Carbine at 16T3.2'S 144°43.8'E, 13 May 2007, V.E. Ahrens and W.J. Pulawski (1 9); Wonga Beach 11 km NNE Mossman at 16 19.9'S 14525.3', W.J. Pulawski, 19 Nov 2006 (1 9, 1 o) and 21 Nov 2006 (1 S). Mellinus costaricae (R. Bohart, 2000), new combination Trachogon/tes costaricae R. Bohart, 2000:168, 9. Holotype: 9/ Costa Rica: Puntarenas: San Vito (University of California, Davis). - Amarante, 2002:19 (in catalog of Neotropical Crabronidae). R. Bohart (2000) published an important revision of the Neotropical Gorytini in which he described eight new genera and a number of new species. One of them was Trachogorytes costaricae, based on a single female from Costa Rica. I have examined that specimen and found that it is a member of Mel Hints based on the wing venation (second submarginal cell not receiving any of the recurrent veins), non-emarginate posterior mandibular margin, short tongue, absence of an omalus and oblique scutal carina, raised and well separated pronotal collar, scutellum, and metanotum, evident Fig. 3. Mellinus costaricae (Bohart), holorvpe: a - whole body in lateral view; b - mesopleuron; c - propodeal enclosure. notaulus, propodeal dorsum with well-de- fined enclosure, submarginal cell III long and distally acute, mid-coxa simple, pres- ence of two mid-tibial spurs, and a pedun- culate gaster (Bohart and Menke 1976). In Menke's key (1996) to Neotropical Mellinus, this species runs to henseni Menke. It differs from henseni and all other currently 296 Journal of Hymenoptera Research recognized Mellinus (Siri and Bohart 1974, Menke 1996) by its unique sculpture: the mesopleuron is longitudinally ridged in the posterior half (Fig. 3b) rather than punctate or uniformly microsculprured, the propo- deal enclosure is all coarsely rugose (except at the very apex), the propodeal side is longitudinally ridged, and the propodeal posterior surface is rugose (Fig. 3c), not punctate, as stated in the original descrip- tion. Additionally, the tentorial pit is closer to the antennal socket than to the eye margin and the propodeal side is separated from the posterior surface by a conspicuous carina, as in the Palearctic arvensis (Lin- naeus) and crabroneus (Thunberg). Most of the body is black (Fig. 3a), but the following are whitish: narrow paraorbital strip in the ventral half of the frons, clypeus (except along frontoclypeal margin), scape ventral- ly, mandible (except apically), mesally interrupted fascia on pronotal collar, pro- notal lobe apically, anterior half of tegula, tiny median spot on metanotum, tergum I laterally (except in basal half) and apically, and a pair of preapical, widely separated spots on tergum II. The femora are blackish basally, then reddish brown and yellow; the tibiae are reddish brown and yellow; and the tarsi are yellow. ACKNOWLEDGMENTS I thank Mr. Pavel G. Nemkov, Vladivostok, Russia, who first drew my attention to the correct generic position of Trachogorytes costaricae during our visit to the Bohart Museum of Entomology, University of California, Davis, on 17 April 2007. I am indebted to Steven L. Heydon, Bohart Museum of Entomology, and Brian Harris, United States National Museum of Natural History, for lending holotypes of Trachogor- ytes costaricae and Mellinus henseni, respectively. Robert L. Zuparko, California Academy of Sciences, and Arnold S. Menke, Bisbee, Arizona, kindly reviewed a draft of the manuscript and made a number of significant improvements. LITERATURE CITED Amarante, S. T. P. 2002. A synonymic catalog of the Neotropical Crabronidae and Sphecidae (Hyme- noptera: Apoidea). Arquivos de Zoologia 37: 1-139. Bohart, R. M. 2000. A review of Gorytini in the Neotropical Region (Hymenoptera: Sphecidae: Bembicinae). Contributions on Entomology, Interna- tional 4: 111-259. and A. S. Menke. 1976. Sphecid Wasps of the World. A generic revision. University of California Press, Berkeley, Los Angeles, London. 1 color plate, IX + 695 pp. Cardale, J. 1985. Sphecidae. Pp. 218-303 in D. W. Walton, ed. Zoological Catalogue of Australia, 2. Hymenoptera. Formicoidea, Vespoidea and Sphecoidea. Australian Government Publishing Service, Can- berra, i-vi, 381 pp. Dalla Torre, C. G. 1897. Catalogus Hymenopterorum hucusque descriptorum systematicus et synonymicus. Volumen VIII: Fossores (Sphegidae). Guilelmi Engelmann, Lipsiae. 749 pp. Dollfuss, H. 1989. Verzeichnis der Grabwespentypen am Naturhistorischen Museum in Wien (Hyme- noptera, Sphecidae). Kataloge der wissenschaftlichen Sammlungen des Naturhistorischen Museums in Wien. Entomologie 7 (4): 1-26. Kohl, F. F. 1884 (1883). Neue Hymenopteren in den Sammlungen des k. k. zool. Hof-Cabinetes zu Wien. II. Verhandlungen der kaiserlich-koniglichen Zoologisch-Botanischen Gescllschaft in Wien 33: 331-386, pis. XVIIa-XVIII. — . 1885 (1884). Die Gattungen und Arten der Larriden Autorum [sic]. Verhandlungen der kaiser- lich-koniglichen Zoologisch-Botanischen Gesellschaft in Wien 34: 171-268, pis. VIII-IX, 327-454, pis. XI- XII. . 1892. Neue Hymenopterenformen. Annalen des k.k. Naturhistorischen Hofmuseums 7: 197-234, pis. XIII-XV. Menke, A. S. 1996. Neotropical Mellinus: a review (Hymenoptera: Sphecidae). Memoirs of the Entomological Society of Washington 17: 125- 141. Naumann, I. D. 1993. Results for aculeate wasps. Pp. 175-187 in I. D. Naumann, E. D. Edwards, T. A. Weir, and D. C. F. Rentz. Insects of the Heath- lands area, Cape York Peninsula, Queensland. Cape York Peninsula scientific expedition. Wet Season 1992. Report. The Royal Geographical Society of Queensland Inc. Vol. 2: 173-203. Turner, R. E. 1908. Notes on the Australian fos- sorial wasps of the family Sphegidae, with descriptions of new species. Proceedings of the General Meetings for Scientific Business of the Zoological Society of London 1908: 457-535, pi. XXVI. Williams, F. X. 1928. Studies in tropical wasps - their hosts and associates (with descriptions of new species). Bulletin. Reports of Work of the Experiment Station of the Hawaiian Sugar Planters' Association. Entomological Series 19: 1-179. J. HYM. RES. Vol. 16(2), 2007, pp. 297-310 Interspecific Variation in Hunting Behavior of Pepsis grossa (Fabricius) and Pepsis thisbe Lucas (Hymenoptera: Pompilidae): A field study Fred Punzo Department of Biology, Box 5F, University of Tampa, 401 W. Kennedy Blvd., Tampa, Florida 33606, USA; email: fpunzo@ut.edu Abstract. — Field studies were conducted on encounters between the spider wasps Pepsis grossa (Fabricius) and P. thisbe Lucas, and females of their host spider, Aphonopelma steindachneri (Ausserer) (Theraphosidae), in Big Bend National Park, Texas. Females of P. grossa were significantly larger than those of P. thisbe. Number of eggs found in ovarioles of P. grossa and P. thisbe ranged from 6-14 and 3-12, with a mean of 11.3 and 8.4, respectively. Behavioral acts comprising hunting behavior of both species included antennation of a spider's burrow (BA), evicting spiders from their burrow (EVB), initial approach and antennation of spider (AA), moving away and grooming (MG1), attack and paralysis (AP), moving away/grooming (MG2), drinking behavior (DB), burial of spider and egg deposition (BO), and closure of the burrow entrance (BC). Antennae of most wasps made initial contact with the forelegs or palps of a spider. During AP, wasps typically grasped leg 3 or 4 of the host before inserting their stings. Most wasps of P. grossa (78%) inserted their sting into the intersegmental membrane between the sternum and coxa 2 of the spider; 88% of P. thisbe females chose a site between the sternum and coxa 1. Only 33 and 26% of P. grossa and P. thisbe, respectively, drank fluids from a spider's mouth or from sting insertion site (LB). Pepsis thisbe required significantly more time (mean: 129.1 min) to complete all behavioral acts of hunting than P. grossa (mean: 101.4 min). Wasps were successful in paralyzing spiders in all observed encounters, and no spider attempted to attack a wasp. Aculeate spider wasps of the genus Pepsis (Hymenoptera: Pompilidae) include at least 133 species varying in size from >60 mm to <12 mm in length (Vardy 2000). Most members of this genus have been referred to as tarantula hawk wasps because females selectively hunt mygalo- morph spiders of the family Theraphosidae (Cazier and Mortenson 1964, Punzo and Garman 1989, Vardy 2002). They paralyze and store these spiders (hosts) in under- ground nests (Williams 1956, Punzo and Ludwig 2005) as a food source for their carnivorous larvae (Punzo 1994a). They occur throughout the New World, from the United States and West Indies, south to Patagonia (Hurd 1952, Vardy 2000). Pepsis grossa (Fabricius 1798) and P. thisbe Lucas (1895) are large, long-legged wasps and are conspicuous components of the arthropod fauna of desert regions in the southwestern United States and northern Mexico (Hurd 1952, Vardy 2000, Punzo, 1994b, 2006a). In Big Bend National Park (BBNP; Brewster County, Texas, USA), females of these two species, as well as P. mildei Stal (1844) selectively hunt and paralyze the large theraphosid spider, Aphonopelma steindachneri (Ausserer 1929) which they use as a host for their de- veloping larvae (Punzo 2005a). Adult wasps feed on nectar which is obtained from flowers of a variety of plants (Evans and West-Eberhard 1970, Punzo 2000, 2006a). In BBNP, where ambient temperatures in late spring and summer may exceed 43 C, it is not uncommon for females to fly over considerable distances during daylight hours in search for flowers and host spiders (Punzo 1994b, Schmidt 298 Journal of Hymenoptera Research 2004). It is important for females to obtain adequate amounts of required nutrients because insect flight places high metabolic demands on insects (Nation 2002). Females of Aphonopelma steindachneri (Ausserer), like other spiders of this genus, excavate burrows (or occupy abandoned rodent burrows) where they remain for most of their lives (Baerg 1958, Gabel 1972, Punzo 2007a). Females seize prey that approaches close to the burrow entrance. Adult males actively wander over the ground surface during the mating season when they search for conspecific females (Punzo 2000, 2007b). Female Pepsis wasps initiate their search for suitable spider hosts after mating (Punzo 1994b, 2006a,b). They typically fly over the ground and are thought to detect occupied spider burrows at a distance using visual and/or odor cues. They in- termittently interrupt flight by landing on the ground and walk rapidly over the surface, tapping the ground surface fre- quently with their antennae as they search for spider burrows (Kurczewski and Kurc- zewski 1968). When a suitable spider burrow is located, the female wasp typi- cally stops at the entrance and taps the edge of the opening with her antennae (burrow antenna tion, BA). After a variable period of time, she cuts through the silk covering over the burrow entrance with her mandibles, enters the burrow, and usually forces the spider out of its burrow and onto the surface (EVB, eviction behav- ior). Male theraphosids are usually en- countered as they move about searching for food and mates (Minch 1979, Punzo 2005b, 2007b). Unlike burrow-dwelling females, males of the genus Aphonopelma from dessert regions usually seek shelter within or under rock crevices, or under surface debris (Smith 1994, Punzo and Henderson 1999). A specific sequence of behavioral acts are exhibited once a spider has been forced to the ground surface. These acts comprise the overall attack behavior of Pepsis wasps found in the tropics and desert regions of the southwestern United States (Petrunke- vitch 1926, 1952, Cazier and Mortenson 1964, Punzo and Garman 1989, Punzo 1991, 1994b, 2005a,c). A female wasp typically approaches the spider and touches its body surface with her antennae (approach and antennation, AA) (Punzo and Garman 1989). In some cases, the spider does not move away, although it may twitch one of its forelegs or raise its palps off the ground (Punzo 2007b). In other instances, tactile stimulation by a wasp elicits a threat posture from the spider which elevates the anterior legs and exposes its fangs (Petrunkevitch 1952, Punzo 1994b). How- ever, the spider rarely strikes at the wasp, although if presented with another insect (cricket, etc.) it typically strikes quickly, seizing the insect and eating it. It has been suggested that chemosensory cues associ- ated with the wasp's cuticle inhibit the spider's strike response (Punzo 2000). Once a spider has been identified as a suitable host, the wasp moves a short distance away and exhibits grooming behavior by passing its antennae through the mandibles. This behavioral component is known as 'moving away and grooming' (MG1, Punzo 1991). After a few minutes, the wasp turns to face its host and then walks under the ventral body region of the spider. In response to this, spiders usually extend their legs, elevating their body off the ground. The wasp then grasps one of the spider's legs and quickly inserts its sting through the ventral body region into the prosomal nerve mass resulting in a rapid paralysis of the spider. This component of hunting is referred to as attack and paralysis (AP). Once a spider is immobilized, the wasp moves away and repeats the grooming sequence described above (MG2). It then returns to the spider and in some instances may either drink fluids from the spider's mouth cavity or drink spider hemolymph that leaks out of the puncture wound made by the insertion of the sting. This is known as drinking or Volume 16, Number 2, 2007 299 lapping behavior, DB (Punzo 2000). How- air temperatures ranging from 5.4 C in ever, because DB does not occur in most January to 33.8 C in August (US Dept. of encounters, the question arises as to what Interior 2005). Annual rainfall is between factor(s) may be responsible for its occur- 13.8-30.1 cm, depending on location and rence. altitude, with 65-70% occurring from May The paralyzed host is then dragged into through October (Parent 1996). Topogra- the spider's burrow (or one excavated by phy of the Park is diverse and includes the wasp) and a single egg is deposited on gypsum formations, igneous rocks, and the ventral surface of the spider's abdomen limestone deposits that provide different (burial and oviposition, BO). The wasp substrates including alluvial fans, gypsum then closes the burrow entrance (BC, flats, saline playas, siliceous and gypsum burrow closure) using soil particles and dunes, fine-textured basins, canyons, small pebbles and flies off to search for mountain ridges, and freshwater springs another host. and seeps, all supporting a diverse plant Although behavioral acts of the hunting fauna categorized within distinctive vege- sequence have been studied, most observa- tative zones (Powell 1988). tions are based on laboratory encounters between wasps and hosts (Punzo 1991, MATERIALS AND METHODS 1994b, 2007). Few detailed observations of I conducted field studies over a 4-year encounters in the field have been de- period (2002-2005) within BBNP from May scribed. In addition, there are a number through September, when male wasps of questions that remain. For example, were establishing perch sites (territories) when attacking a spider does a wasp and females were searching for spiders, exhibit any preference for grasping a par- Adults of P. grossa and P.thisbe were ticular leg? Are there specific sites on the observed within a 5-km radius of Tornillo spider's body where a wasp inserts its Flat (TF; 29 01'N, 102 59W), a site where sting? How commonly does lapping be- both of these species are abundant, as well havior occur? Is there a preference for the as their host spider, A. steindachneri, are site at which it occurs? Finally, are there abundant (Punzo 2000, 2007b). I had interspecific differences associated with located and marked numerous occupied these behaviors? The present study was tarantula burrows during previous field conducted in order to analyze hunting studies in this area over the last 12 years (;/ behavior of P. grossa and P. thisbe under = 946). Females of A. steindachneri typically natural conditions and to address these remain within a single burrow for most of questions in these two sympatric species of their adult lives (Punzo, unpubl. data). As wasps that are found in similar microhab- a result, I knew the locations of host itats in Big Bend National Park (BBNP), spiders and concentrated my field observa- where both utilize A. steindachneri as a host, tions at these burrow sites. Voucher speci- mens of wasps, wasp eggs, and spiders DESCRIPTION OF GENERAL have been deposited in the invertebrate STUDY AREA collection at BBNP. Pepsis grossa and P. thisbe occur through- Based on my knowledge of locations for out Big Bend National Park (BBNP) is burrows occupied by a female tarantula, I located in Brewster County, Texas, and lies examined 148 burrows whose entrances within the northern region of the Chihua- had been closed from previously unob- huan Desert. Its western, southern and served encounters with wasps to deter- eastern boundaries are bordered by the Rio mine whether it contained a paralyzed Grande River. Climatic conditions range host. During the course of this study I also from arid to semiarid, with mean monthly monitored 96 burrows containing an adult 300 Journal of Hymenoptera Research female spider, and observed 54 and 42 ses. Another Pepsis wasp, P. mildei also encounters between a wasp and spider for occurs at TF, but is far less abundant P. grossa and P. thisbe, respectively. Because (Punzo, unpubl. data). After measurements encounters between a wasp and a male were recorded, each wasp was preserved spider usually occur while the male is in 70% ethanol for subsequent determina- wandering over the ground surface, op- tion of number of eggs in ovarioles. Re- portunities to witness such encounters moval of wasps ensured that the same occur far less frequent. Therefore, only wasp was not involved in more than one encounters involving a female spider were encounter for the data set. used for analyses. For each encounter I observed the general I observed all encounters at close range behavior of the wasp as it approached the (1-2 m from combatants), and used a 35- burrow entrance and interacted with a spi- mm Nikon FE2 camera to photograph der, as well as the concomitant behaviors some of the encounters. Encounters were exhibited by the host. I recorded: (1) the observed during daylight and evening amount of time required to complete the hours (0830-0200 h, Central Standard hunting sequence (from initial contact to Time). At the end of each encounter closure of the burrow), as well as all (following burial, oviposition, and closure behavioral components of hunting, using of the burrow) I collected the wasp with a stopwatch; (2) which spider leg was a sweep net and anesthetized it using initially seized by the wasp as it attempted a portable C02 cartridge. Once inactivated, to insert its sting; (3) site at which wasp I verified species identification and used sting had been inserted into the body of the a Unitron dissecting microscope fitted with spider; and (4) whether or not lapping an ocular micrometer to record body behavior occurred, and if so, where, length and head capsule width. Although All statistical procedures followed those Aphonopelma steindachneri is the only ther- described by Sokal and Rohlf (1995). All aphosid known to occur at the TF site data conformed to conditions of normality (Smith 1994, Punzo 2007b), I opened each as assessed using a Bartlett's test for burrow to verify species identification of homogeneity of variances and a G-test for paralyzed spiders. I examined the body normality. Comparisons on means for mor- surface of paralyzed spiders with a dissect- phometric data between the two species of ing microscope in order to locate the site wasps as well as for male and female where the wasp's sting had been inserted spiders were tested using a t test. Differ- into the host during the paralyzation ences between the proportion of burrows sequence. I also recorded the width of the containing paralyzed males versus females, carapace and total body length to the and frequency at which specific spider legs nearest 0.1mm using a digital caliper, were grasped by a wasp during initiation of After removing the wasp's egg, I recorded attack behavior, were tested using a Chi weight of wasp eggs (to the nearest Square test (X2). Comparisons between time 0.01 mg) and spiders (to the nearest required by each wasp species to complete 0.01 g) using a portable electronic balance, overall hunting sequence was tested using Egg length and width were recorded to the an analysis of variance (ANOVA), and nearest 0.01 mm using a dissecting micro- a Scheffe F test was used for ad hoc scope. comparisons between individual behavioral The site of the puncture wound could be components of hunting, readily identified by hemolymph that oozed out of the wound. Only data RESULTS obtained for P. grossa, P. thisbe and Of the 148 burrows whose entrances had host A. steindachneri were used for anal} been closed with soil (following previously Volume 16, Number 2, 2007 301 Table 1. Morphometric data on females of Pepsis grossa (n = 54) and P. thisbe (n = 42) and female host spiders, Apkonopelma steindachneri (n = 96 females, 46 males) at Tornillo Flat, Big Bend National Park, Texas. Data from wasps and spiders examined from 2002 to 2005. Data expressed as means; values in parentheses represent (±SE). BL (body length); HCW (head capsule width); CW (cephalothorax width); BW (body weight); NEO (number of eggs found in ovarioles). Values in rows followed by a different letter are statistically significant (t tests; P < 0.05). Pepsis grossa P. thisbe A. steindachneri BL (mm) HCW (mm) NEO CW (mm) BW(g) 42.7a (2.4) 4.9a (0.2) 11.3a (2.4) 35.9b (1.7) 3.9b (0.3) 8.4b (0.7) Males 13.4a (0.7) 6.8a (0.5) Females 15.2b (0.5) 10.8b (1.1) unobserved encounters), 128 (87%) con- tained a paralyzed spider. Although the species of wasp responsible for the paral- ysis cannot be known unless the offspring is reared, 87 of these 128 burrows (68%) contained a female spider, and 32% held a male (Chi Square test: X2 = 10.89, P < 0.03). Paralyzed male spiders ranged in weight from 5.2-6.4 g (mean: 5.77 ± 0.42), while the range was 6.9-10.4 g (mean: 7.94 ± 0.37) for females. For data obtained from observed encounters, female spiders para- lyzed by P. grossa and P. thisbe had a mean weight of 8.7 ± 0.83 g (range: 6.2-10.6) and 7.7 ± 0.64 g (range: 5.7-10.7), respectively. Mean values for length, width, and weight for eggs of P. grossa were 4.34 ± 0.03 mm, 1.34 ± 0.01 mm, and 7.14 ± 0.31 mg, as compared to 4.29 ± 0.05 mm, 1.24 ± 0.02 mm, and 6.88 ± 0.41 mg for P. thisbe. Morphometric data for wasps observed encountering spiders, as well as for hosts, are shown in Table 1. Concerning host spiders, based on width of cephalothorax (t = 3.04, P < 0.05) and body weight (t = 2.46, P < 0.05), females were significantly larger than males. For the two species of wasps, females of P. grossa were signifi- cantly larger than those of P. thisbe, based on body length (t = 3.35, P < 0.05) and head capsule width (t = 2.88, P < 0.05). Data on wasps collected from the field indicated that the mean number of eggs found in ovarioles for P. grossa and P. thisbe was 11.3 and 8.4, respectively (Table 1), with a range of 6-14 and 3-12. Because wasps require a host for each egg, and number of previous encounters with a spi- der was unknown, the number of eggs produced by each species of wasp prior to any hunting experience could not be de- termined from field data. However, I have reared both species of wasps from larvae feeding on A. steindachneri females (mean weight: 9.89 ± 0.61 g) in the laboratory and found that non-mated P. grossa (10-12 days of age) produced 9-21 eggs /female (mean: 16.2 + 3.2 g SE, n = 78) as compared to 4-14 (mean: 10.2 + 2.2, n = 38) for P. thisbe (t = 7.09, P < 0.05) (Punzo, unpubl. data). For all observed encounters between P. grossa or P. thisbe and a host spider, wasps entered an occupied burrow by cutting through the silk covering over the burrow entrance and then forced the host to the surface. Attack and paralysis never oc- curred within the burrow. A wasp would typically approach the entrance of a spider burrow and tap its antennae along the edge of the opening. The time allocated by these wasps for each behavioral component of hunting behavior is shown in Table 2. Pepsis thisbe females required signifi- cantly more time to complete the overall hunting sequence as compared to P. grossa (F = 19.27, P < 0.05) (Table 2). No significant interspecific differences were found for eviction behavior (EVB; Scheffe F, P > 0.50) or moving away and grooming (MG1, MG2, P > 0.60). Interspecific differ- 302 Journal of Hymenoptera Research Table 2. Time (in min) allocated by females of Pepsis grossa (n = 54) and P. thisbe (« = 42) for various behavioral components of the overall hunting sequence during encounters with a host spider, Aphonopelma steindachneri. Data are expressed as means; numbers in parentheses represent tSE. Values in rows followed by a different letter are statistically significant (P < 0.05). BA (burrow antennation); EVB (eviction behavior); AA (approach and antennation); MG1, MG2 (moving away and grooming); AP (antennation and paralysis); LB (lapping behavior); BO (burial and oviposition); BC (burrow closure). Behavioral component Pepsis grossa Pepsi's thisbe BA 4.8a (0.8) 7.6b (1.1) EVB 3.4a (0.3) 3.9a (0.4) AA 8.3a (1.7) 5.1b (0.8) MG1 4.6a (0.4) 5.2a (0.3) AP 1.8a (0.2) 2.3a (0.5) MG2 3.2a (0.6) 2.9a (0.3) LB1 3.6a (1.1) 5.1b (1.8) BO2 14.4a (3.5) 23.2b (4.6) BC 57.3a (7.1) 73.8b (6.9) Total: 101.4a 129.1b 1 Lapping behavior occurred in 18 of 54 encounters for P. grossa (33.3%), and in 11 of 42 encounters for P. thisbe (26.2%). 2 Represents data for situations in which a spider was buried in its own burrow. ences for all other behavioral components were significant. As compared to P. grossa, P. thisbe females allocated significantly- more time for burrow antennation (BA; Scheffe F = 6.2, P < 0.05), drinking behavior (DB; F = 4.9, P < 0.05), burial and oviposition (BO; F = 7.1, P < 0.05), and burrow closure (BC; F = 6.9, P < 0.05), and significantly less time for approach and antennation (AA; F = 5.5, P < 0.05). Wasps of both species approached spi- ders that they had forced out of their burrows and then tapped the spider's body surface with their antennae. Antennae of P. grossa and P. thisbe initially made contact with the tarsus of one of the spider's forelegs in 50 of 54 (92.5%) and 39 of 42 (92.8%) encounters, respectively. In other cases, the antennae initially made contact with one of the spider'spalps. Subsequent- ly, wasps of both species used their antennae to explore the lateral region of a spider's cephalothorax and abdomen. During antennation by P. grossa, 34 of 54 spiders (63%) exhibited no bodily move- ments as compared to similar values observed for P. thisbe (28 of 42, 67%, P > 0.60). For encounters with P. grossa, other spiders either remained stationary but exhibited slight movements of their foreleg (n = 2, 4%) or an elevation of the palps (n = 6, 11%), while the remainder (n = 12, 22%) exhibited a threat posture. Compara- ble values for similar responses of spiders to P. thisbe were 5% (n = 2), 9% (n = 4), and 19% (n = 8), respectively. In no case did a spider attempt to flee back into its burrow or attack the wasp, and wasps 'won' all observed encounters. When initiating attack, wasps of both species showed a preference for grasping leg 3 or 4 of the spider (Table 3). Eight-one and 57% of P. grossa and P. thisbe, re- spectively, exhibited a rapid dash under the ventral region of the spider and used their mandibles to grasp leg 3 or 4 before attempting to insert their sting. Leg 1 was never grasped, and leg 2 in only 2.4-7.4% of encounters. Sting insertion sites for P.grossa and P. thisbe are shown in Fig. 1. Examination of spiders post-paralysis showed that 78% of P. grossa females inserted their sting into the intersegmental membrane between the sternum and coxa 2 of the spider, and 22% between the sternum and pedipalp. In comparison, 88 and 12% of P. thisbe females, respectively, inserted their sting into the membrane between the sternum and coxa 1 or at the junction between the abdomen and cephalothorax. Mean time that elapsed between insertion of sting and immobilization (paralysis) of spider (in- dicated by curling of the legs under the spider's body) was 6.2 s ± 0.4 SE (range: 4- 8 s) for P. grossa and 12.8 s ± 1.1 SE (range: 8-16 s) for P. thisbe. Regardless of insertion site, there was no significant difference in time required for paralysis for either species of wasp. Volume 16, Number 2, 2007 303 Table 3. Leg of spider (Aphonopelma steindachneri) grasped by females of Pepsis grossa and Pepsis thisbe when initiating attack. Data derived from a single observation of each wasp/spider encounter for 54 and 42 encounters, respectively, between P. grossa and P. thisbe, and a host spider. Spider legs on right and left side of the body (based on position of spider when a wasp was facing it) are designated as R and L, respectively, and legs are numbered 1 (forelegs) through 4 (hindlegs). Values in parentheses represent frequency of occurrence (%). Spider leg grasped Pepsis grossa (n = 54) Pepsis thisbe (n = 42) Rl 0 0 LI 0 0 R2 4 (7.4) 1 (2.4) L2 0 1 (2.4) R3 20 (37) 14 (33.3) L3 24 (44.4) 10 (23.8) R4 4 (7.4) 7 (16.6) L4 2 (3.7) 9 (21.4) Only 33.3 and 26.2% of P. grossa and P. thisbe females, respectively, exhibited drinking behavior (DB, Table 2). Eleven of 18 (61%) females of P. grossa were observed to drink fluid oozing from the wound site (sting insertion site), while 7 (39%) wasps drank fluids from the spider's mouth cavity. For P. thisbe, the percentage of females that engaged in DB was 8 of 11 (73%) and 3 of 11 (27%) for the wound site and mouth cavity, respectively. DISCUSSION Physical dimensions and number of eggs for P. grossa and P. thisbe were similar to values reported for other species of Pepsis wasps of similar size. For example, eggs of P. cerberus Lucas and P. mexicana Lucas from another area of BBNP ranged from 4.19-4.29 and 4.25 vs. 4.31 mm (length), 1.26-1.38 and 1.18-1.29 mm (width), and 7.18-7.32 and 7.06-7.17 g (weight), respec- tively (Punzo 2005c). In like manner, number of eggs produced per female for P. grossa and P. thisbe were similar to values reported for other Pepsis wasps which range from 2-44/female (Haupt 1952, Evans and West-Eberhard 1970, Punzo 2000, 2005c). It has been shown that the number of eggs produced by Pepsis females (Evans 1953, Punzo 2005c), as well as in many other insects (Price 1975, Ito 1980, Nation 2002) is positively correlated with body size. Although a majority of paralyzed spi- ders found with an attached Pepsis egg (where encounters with a wasp had not been observed) contained female spiders (68%), almost 1/3 contained a male. Thus, it appears that Pepsis wasps are opportu- nistic hunters and will readily attack a male tarantula even though males are usually smaller than females. This is in general agreement with previous laboratory or field observations showing that Pepsis wasps will attack, paralyze, and deposit an egg on male and female theraphosid hosts (Kurczewski and Kurczewski 1968, Punzo 2000). An experimental protocol that might allow us to determine whether female Pepsis wasps have any 'preference' for spiders of different sexes would be to observe the response of mated female wasps when given a choice between a male and female spider. If naive females are used (no previous encounter with a spider), one can also assess whether such a prefer- ence, if exhibited, had an innate compo- nent. Little information is available on the number of eggs produced per female for Pepsis wasps. A previous study yielded some data for pepsine wasps collected from Persimmon Gap, a site 48 km to the northeast of TF that also lies within BBNP (Punzo 2005c). Number of eggs in ovarioles from wasps collected immediately after mating (before they started to hunt for hosts) ranged from 5-26 (mean: 12.7 ± 2.8 SE) and 4-20 (mean: 11.8 ± 1.7) eggs per female, for P. cerberus and P. mexicana, respectively. It is difficult to extrapolate and compare these data with values reported for P. grossa (11.3) and P. thisbe (8.4) in the present study because there was no way of knowing how many prior encounters these wasps had with a host before they were collected. These compare 304 Journal of Hymenoptera Research Pedipalp Chelicera Pedipalp Spinnerets Fig. 1. Ventral body region of Aphonopelma steindachneri showing insertion sites (wound sites) for the sting of Pepsi's grossa and P. thisbe. Solid circle and square (intersegmental membrane between sternum and pedipalp, and sternum and coxa 2, respectively) are insertion sites for P. grossa: solid triangle and diamond (membrane between sternum and coxa 1, and at junction of cephalothorax and abdomen, respectively) are insertion sites for P. thisbe. Legs: L1-L4. to 9-21 eggs for P. grossa and 4-14 for P. thisbe reared in the laboratory (Punzo, unpubl. data). Spiders paralyzed by P. grossa and P. thisbe varied in body weight. Previous research has demonstrated that adult size (as assessed by head capsule width, length of legs or wings) attained by other species of spider wasps is positively correlated with the mass attained by their last-instar larvae (Vinson 1984, Punzo 2005c), which in turn is positively correlated with the mass of the spider that the larvae fed on (Price 1997, Punzo 2005c). This most likely accounts for the ranges in adult size observed in male and female Pepisis wasps in the field. Although empirical evidence is lacking, it would be interesting to de- termine to what extent (if any) a decision made by a female Pepsis wasp concerning size of host suitable for attack may be influenced by the wasp's size. There are relatively few detailed obser- vations and analyses on foraging (Cazier Volume 16, Number 2, 2007 305 and Mortenson 1964, Punzo and Ludwig 2005, Punzo 2006a,b), territoriality (Rau and Rau 1918, Punzo 2000), dispersal (Evans and West-Eberhard 1970), diel periodicity (Punzo 2005c), and hunting behavior (Petrunkevitch 1952, Williams 1956, Punzo 2005a) in Pepsis wasps under natural conditions. Although many aspects of overall hunting behavior for wasps of this genus are similar across species, the results of this field study indicate that there is some interspecific variation associated with certain behavioral components of hunting between P. grossa and P. thisbe which include frequency of drinking be- havior and sting insertion sites. At the Tornillo Flat (TF) study site, females of both species approached a burrow occu- pied by adult females of A. steindachneri and tapped their antennae along the edges of the opening (burrow antennation, BA). Wasps then entered the burrow after cutting through the silk covering and forced the spider out onto the ground surface (eviction behavior, EVB). These behaviors have been reported for encoun- ters between theraphosid spiders and other species of Pepsis wasps (Buckley 1862, Petrunkevitch 1926, 1952, Passmore 1936, Punzo 2005a), as well as P. grossa (as formosa) (Punzo and Garman 1989, Punzo 1991) and P. thisbe (Punzo 1994b), that were staged in the laboratory. Because A. steindachneri is the only species of theraphosid spider known to occur at TF, all females of P. grossa and P. thisbe utilized this species as a host for their larvae. At sites 25-80 km to the north, these two species of wasps, along with P. mildei Stal, are known to utilize both sexes of two other theraphosids, Aphonopelma (as Rhe- chostica) hentzi and Dugesiella (as Aphono- pelma) echina Hentzi as a host (Punzo and Garman 1989, Punzo 1991). In southern Texas (Hidalgo County), P. grossa (as formosa) utilizes both sexes of the thera- phosids A. harlingeninn (Chamberlin) and A. heterops (Chamberlin) as hosts, although it shows a strong preference for A. harlin- geninn (Punzo 2006b). Farther to the west, in Arizona and California, P. grossa is known to hunt another theraphosid, A. chalcodes (Chamberlin) (Cazier and Mor- tenson 1964). In southern California, P. thisbe hunts two theraphosid species, A. reversum Simon and Eurypelma (as Aphono- pelma) eutylenum (Ausserer) (Williams 1956). From the same region, P. mildei has been reported to utilize females of the trapdoor spider Bothriocyrtum californicum (Chamberlin and Ivie) (Passmore 1933), a mygalomorph spider from an entirely different family (Ctenizidae). However, I have never found a nest containing any species of paralyzed trapdoor spider with a Pepsis larvae, or egg at any of a number of locations in BBNP or Big Bend Ranch State Park (Presidio Co., Texas) (Punzo unpubl. data). These various host records suggest that Pepsis wasps have the ability to utilize immatures and adult males and females of a variety of theraphosid spiders as hosts, depending on the theraphosids available at any particular site. Presumably, during the course of evolution in pompilid wasps, selection favored a preference for hunting a single, larger host for each wasp larva. Many species of spider wasps in this genus are among the largest wasps in the New World (Hurd 1952, Vardy 2000, 2002) and adults develop from larvae that attain lengths in excess of 27 mm and can weigh over 5 g (Punzo 2000). In order for a female to provide an adequate amount of food for such large larvae, each larvae would have to be provided with a high number of smaller hosts either together or over a continuous period of time (progressive provisioning) as opposed to providing all required food at one time (a single large host or mutliple small ones, e.g. by Trypoxylon wasps) at one time {mass pro- visioning) (O'Neill 2001). There are obvious trade-offs that are involved. Progressive provisioning involving several smaller prey would require more energy and increase exposure of wasps to potential 306 Journal of Hymenoptera Research predators, while requiring more handling FAPs, which are characteristic of instinc- time but less risk from the prey. In contrast tive (innate) behavior (Tinbergen 1951). a single, larger, more formidable host, may More recent studies, based on sequential require less handling time but pose a higher laboratory-staged contests between Pepsis risk from the prey, whilst requiring less wasps and spiders, have shown that the overall energy expenditure and a decreased amount of time required to perform some probability of encountering a predator. behavioral acts decreases as a function of The data collected to do not point to increasing number of encounters (experi- specific niche divisions between P. grossa ence) (Punzo and Garman 1989, Punzo, and P. ihisbe at the TF site. I did not observe 1991, 2005a). any significant interspecific differences in The term 'modal action pattern' (MAP) the sex or size of spiders selected, nor in has been used to describe components of temporal patterns of hunting activity, innate behavior that exhibit some degree of Adult wasps of both species begin to plasticity (Barlow 1977). The acts that have emerge from their nests during late March been shown to improve with experience and continue to do so until August, include initial approach and antennation Females of both species were observed (AA) and attack and paralysis (AP) (Ta- hunting during daylight and evening ble 2), while the time required for other hours. At sites further south (Zapata Co., behavioral components does not. This Texas), P. thisbe begins to emerge in mid- suggests that some degree of learning is March, whereas P. grossa adults are not associated with hunting behavior in these seen until mid-April (F. Punzo unpubl. wasps (Punzo 1996). It is interesting to note data). Perhaps host spiders occurred at that those acts which a wasp can perform sufficiently high densities during the more quickly with experience are precisely course of the present study at the TF site those which present the most risk for so that any need for resource paritioning a wasp. AA requires that a wasp approach was reduced. within a close distance of the spider and Future studies should further analyze actually touch the spider's body with its host preference in Pepsis wasps. For exam- antennae, placing it well within the strike pie, at locations where a Pepsis wasp is distance of the spider (Punzo 2007b). AP known to utilize more than one therapho- requires the wasp to move directly under sid species, it would be instructive to a spider's body, often passing directly assess any differences that may occur with below its fangs, grasp a leg, and insert its respect to duration of embryonic develop- sting. In contrast, performance of behav- ment, larval growth rate and number of ioral acts that pose no risk because the larval instars, size of emerging adults, spider has already been immobilized adult longevity, fecundity, flight endur- (MG2), or are most likely subject to bio- ance, and host preference, for wasps de- mechanical constraints (BO, and BC), do veloping on different hosts. not 'improve' with experience (Punzo 1991, Overall hunting behavior of Pepsis wasps 1994b). Similarly, MG1, which increases (P. grossa, P. tliisbe, P. mildei, P. marginata, the distance between protagonists and and Pompilus spp). can be categorized into thereby decreases risk for a wasp, is not several distinct behavioral components performed more rapidly with increasing (Table 2), one of which may or may not number of encounters (Punzo and Garman occur (DB) (Petrunkevitch 1926, Passmore 1989, Punzo 2000). 1936, Evans 1953). These behavioral com- During attack, these wasps showed ponents were traditionally interpreted as a marked preference for grasping legs 3 examples of genetically-determined, inflex- or 4 of a spider before stinging it. Perhaps ible acts referred to a fixed action patterns, the positions of legs 3 and 4, relative to the m Volume 16, Number 2, 2007 307 center of gravity for a spider, make it easier for a wasp to obtain the leverage required to insert its sting through an appropriate site on the spider's ventral surface in the least amount of time. The faster a wasp insert its sting the more rapidly it can immobilize a formidable host and reduce the probability of retaliation by the spider. Experiments where a wasp is presented with a spider whose legs 3 or 4 have been removed, would force a wasp to grasp leg 1 or 2, or refuse to attack at all, and would provide a way to assess any possible biomechanical advantage associated with grasping various legs. All observations of encounters between Pepsis wasps and theraphosids have shown that a wasp's sting is directed into the ventral body region of a spider (Punzo 2007b, and references cited therein). Fe- males of P. grossa and P. thisbe showed a marked preference when choosing a site on the spider's body in which to insert their sting. Little information is available for sting insertion sites for Pepsis wasps. Petrunkevitch (1926) observed a female of P. marginata insert her sting between the third and fourth right coxae when attack- ing the theraphosid, Cyrtopholis portoricae Simon. Another wasp inserted its sting into the intersegmental membrane between the sternum, maxilla and coxa 1. In an encoun- ter with the theraphosid Dugesielln hentzi, a Ptysis wasp of undetermined species inserted its sting through the membrane between coxa 3 and coxa 4 (Baerg 1958). The site preferred by females of P. grossa and P. thisbe was the intersegmental mem- brane between coxa 2 and sternum, and between coxa 1 and sternum, respectively. These, as well as the other sting insertion sites observed in this study (membrane between pedipalp and sternum, P. grossa; junction between abdomen and cephalo- thorax, P. thisbe), all allow a wasp to deliver its venom into the prosomal nerve mass which supplies motoneurons to muscles involved in movments of all legs, chelicer- ae, and fangs (Foelix 1996, Punzo 2007b). Some investigators have observed in- stances in which a wasp failed to locate an insertion site during its first attack on a spider, moved a short distance away, and then attacked again, successfully paralyz- ing its host (Petrunkevitch 1952, Williams 1956, Baerg 1958). Nonetheless, once the sting delivers venom into the prosomal nerve mass, paralysis of the spider occurs very rapidly as indicated by a curling of the spider's legs under its body, slight twitching movements of some appendages, and then complete immobilization (Petrun- kevitch 1926, Cazier and Mortenson 1964, Punzo 2000). These results on sting insertion sites also indicate that there is behavioral variation exhibited by these wasps. This suggests several interesting questions that future studies should address: do individual wasps choose the same insertion site for all encounters, or do they vary? If sting insertion sites are 'fixed' for individual females this would suggest that the behav- ioral program has a genetic basis (innate). If so, breeding experiments involving males with females showing different behavioral phenotypes might shed some light on the patterns of inheritance in- volved in this behavior. Secondly, is there a relationship between a particular inser- tion site and time required to immobilize a spider? Nutritional state (body condition) may afford a possible explanation for why drinking behavior (DB) occurs in only some encounters. Drinking hemolymph oozing from a host's wound site may provide necessary nutrients to meet the energetic demands of flight which wasps engage in when searching for hosts, as well as those of venom production, handling time, and burrow closure. Drinking fluids from a spider's mouth cavity may help wasps to maintain proper water balance of body fluids. Similar behavior has been reported for species in other wasp families. For example, Tinbergen (1972) observed that females of the digger wasp, Philanthus 308 Journal of Hymenoptera Research triangulum Fabricius, which selectively hunt honeybees, Apis mellifera L., press the abdomen of a paralyzed bee through their mandibles and lick up the fluid (nectar) extruded from the bee's mouth. Species of spider wasps from other genera are also known to drink fluids from a host's mouth cavity or wound site (Petrunkevitch 1952, Evans 1953, Williams 1956, Evans and West-Eberhard 1970, Punzo 2000). It may be that wasps engage in DB only after a certain number of foraging bouts have occurred resulting in a need to replenish nutrients and /or body fluids. Experimental protocols using a tethered flight apparatus should be used in future studies to test this hypothesis. Different female wasps could be subjected to forced flight tests for varying periods of time and then allowed to encounter a host. If the hypothesis is true, wasps that are subjected to longer bouts of flight (and thus expend more energy, and lose more water by evaporation) should be more likely to engage in DB than wasps subjected to flying for shorter periods of time. Such an experimental design would also allow one to determine the amount of flight time required to initiate LB in a particular wasp species. The confines of a spider's burrow might not provide enough room for a wasp to maneuver in such a way as to effectively administer a sting to its host. This may account for the fact that all wasps observed at the TF site forced a spider out of its burrow and onto the ground surface (eviction behavior, EVB) before attacking their host. Similar EVB has been observed for Pepsis wasps presented with tarantulas that have been allowed to excavate bur- rows within their cages under laboratory conditions (Petrunkevitch 1926, Punzo and Garman 1989, Punzo 1991, 1994b). Interspecific differences in the amount of time required to complete the overall hunting sequence among P.grossa and P. thisbe in the field may be associated with some wasps of either species having had more encounters with hosts than other wasps. It may also reflect genetically-based differences in synaptic events associated with afferent (sensory) neural pathways involved with detection and identification of hosts and /or efferent (motor pathways) involved in the control of bodily move- ments required for various behavioral acts. On a final note, theraphosid spiders are typically aggressive and innately strike at arthropods that wander within their prey awareness area (Punzo 2007b). The fact that no spider attempted to seize a wasp suggests that the spider's attack response is somehow inhibited. It has been suggested that these wasps may release some chemi- cal compound (s) or possess chemosensory cues associated with their epicuticle that inhibit spider's from attacking them (Pet- runkevitch 1952, Punzo 2000). Petrunke- vitch (1926) observed that P. marginata from Puerto Rico produced a "pungent odor" when initially making contact with a theraphosid host and argued that the substance responsible for this odor might somehow diminish the aggressiveness of the spider. Others have pointed out that the smooth surface of a wasp's cuticle, combined with its high degree of hardness, makes it difficult for a spider's fangs to penetrate a wasp's integument (Petrunke- vitch 1926, Passmore 1936). It should be pointed out that Petrunke- vitch (1926) observed a theraphosid (Cyrto- pholis portoricae Simon) that unsuccessfully attempted to grasp a female of P. marginata Lucas with her fangs as the wasp passed under the spider. Cazier and Mortenson (1964) observed a Pepsis grossa (as formosa) female entering a burrow occupied by Aphonopelma sp., and after a few minutes the spider emerged from its burrow with its anterior two legs wrapped around the wasp and its chelicerae inserted into the wasp's abdomen. Nonetheless, the wasp was able to sting the spider. After being stung, the spider released the wasp which exhibited erratic movements and was un- Volume 16, Number 2, 2007 309 able to fly. The spider's right leg was rigidly extended forward, making locomo- tion awkward. After several minutes, the wasp and spider were placed in a screened plastic container and initially both animals avoided one another. When observed 45 min later, the spider was engaged in eating the wasp. These observations sug- gest that: (1) cues potentially responsible for inhibiting a spider's strike response may not always be effective (2) varying degrees of effectiveness may be species- specific; or (3) mutations may account for differences in the chemical profile of the wasp's cuticle and certain profiles may be less effective at deterring a spider's strike than others. ACKNOWLEDGMENTS I thank C. Bradford, L. Ludwig, G. Broad, and anonymous reviewers for commenting on an earlier draft of the manuscript, and L. Ludwig, J. Bottrell, K. Crawford, and S. Madragon for assistance in field observations. B. Garman generously provided consul- tation on statistical analyses. This research was supported by a Faculty Development Grant from the University of Tampa. Field work was conducted with permission of the National Park Service, with logistical support provided by R. Skiles, Big Bend National Park. LITERATURE CITED Baerg, W. J. 1958. The Tarantula. University of Kansas Press, Lawrence, Kansas. 88 pp. Barlow, G. W. 1977. Modal action patterns. Pp. 98-134 in: T. A. Seboek, ed. How Animals Communicate. Indiana Univ Press, Bloomington, Indiana. Buckley, S. B. 1862. The tarantula (Mygale hentzii Girard) and its destroyer {Pompilus formosus Say). Proceedings of the Entomological Society of Philadel- phia 1: 138-139. Cazier, M. A. and M. Mortenson. 1964. Bionomical observations on tarantula hawks and their prey (Hymenoptera: Pompilidae): Pepsis. Annals of the Entomological Society of America 57: 533-541. Evans, H. E. 1953. Comparative ethology of spider wasps. Systematic Zoology 2: 155-172. and M. J. West-Eberhard. 1970. The Wasps. University of Michigan Press, Ann Arbor, Michi- gan, vii + 265 pp. Foelix, R. F. 1996. The Biology of Spiders. Oxford University Press, London, v + 321 pp. Gabel, J. 1972. Further observations of theraphosid tarantula burrows. Pan-Pacific Entomologist 48: 72-73. Haupt, H. 1952. Alte und neue Pepsis Arten (Hyme- noptera-Sphecoidea). Nova Acta Eeopoldina N.F. 15: 309-414. Hurd, P. D. 1952. Revision of the Nearctic species of the pompilid genus Pepsis (Hymenoptera: Pom- pilidae). Bulletin of the American Museum of Natural History 98: 257-334. Ito, Y. 1980. Comparative Ethology. Second Ed. Cam- bridge University Press, Cambridge. 478 pp. Kurczewski, F. E. and E. J. Kurczewski. 1968. Host records for some North American Pompilidae (Hymenoptera) with a discussion of factors in prey selection. Journal of the Kansas Entomological Society 41: 1-33. Minch, E. 1979. Annual activity patterns in the tarantula Aphonopelma chalcodes Chamberlin. No- vitates Arthropodae 1: 1-34. Nation, J. L. 2002. Insect Physiology and Biochemistry. CRC Press, Boca Raton, Florida. 485 pp. Parent, L. 1996. Big Bend National Park. Falcon Publishing, Helena, Montana, viii + 171 pp. Passmore, L. 1933. California trapdoor spider per- forms engineering marvels. National Geographic Magazine 64: 195-211. . 1936. Tarantula and tarantula hawk. Nature Magazine 27: 155-159. Petrunkevitch, A. 1926. Tarantula vs. tarantula hawk: a study in instinct, journal of Experimental Zoology 43: 367-397. . 1952. The spider and the wasp. Scientific American 187: 20-33. Powell, A. M. 1988. Trees and Shrubs of Trans Pecos Texas. Big Bend Natural History Association, Big Bend, Texas. 536 pp. Price, P. W. 1975. Strategies for egg production. Evolution 28: 76-84. . 1997. Insect Ecology. Third Ed. J. Wiley, New York, vi + 874 pp. Punzo, F. 1991. Neurochemical events associated with learning and hunting behavior in the spider wasp, Pepsis formosa (Hymenoptera: Pompilidae). Florida Scientist 54: 51-61. . 1994a. The biology of the spider wasp, Pepsis thishe (Hymenoptera: Pompilidae) from Trans Pecos Texas. I. Adult morphometries, larval development and the ontogeny of feeding behav- ior. Psyche 101: 229-241. . 1994b. The biology of the spider wasp, Pepsis thishe (Hymenoptera: Pompilidae) from Trans Pecos Texas. II. Temporal patterns of activity and hunting behavior with special reference to effects of experience. Psyche 101: 243-256. . 1996. Localization of brain function and neurochemical events associated with learning in insects. Recent Trends in Comparative Biochemis- try and Physiology 2: 9-16. . 2000. Desert Arthropods: I ife History Variations. Springer, Heidelberg, Germany, xi + 310 pp. 310 Journal of Hymenoptera Research — . 2005a. Experience affects hunting behavior of the wasp, Pepsis mildei Stal (Hymenoptera: Pom- pilidae). journal of the New York Entomological Society 113: 222-229. — . 2005b. The ability of Aphonopelma steindachneri (Ausserer) (Arachnida: Theraphosidae) to detect and respond to chemosensory cues associated with a predator, Hogna carolinensis (Arachnida: Lycosidae). Bulletin of the British Arachnological Society 13: 169-172. — . 2005c. Studies on the natural history, ecology, and behavior of Pepsis cerberus and P. mexicana (Hymenoptera: Pompilidae) from Big Bend Na- tional Park, Texas, journal of the New York Entomological Society 113: 84-95. — . 2006a. Plants whose flowers are utilized by adults of Pepsis grossa Fabricius (Hymenoptera: Pompilidae) as a source of nectar. Journal of Hvmenoptera Research 15: 171-176. — . 2006b. Effect of host odor cues on behavioral responses of the spider wasp, Pepsis formosa (Hymenoptera: Pompilidae). Texas journal of Science 58: 45-55. — . 2007a. Microhabitat utilization, diet composi- tion, and diel periodicity in five sympatric species of desert arachnids: a wolf spider (Hogna caroli- nensis), tarantula spider (Aphonopelma steindach- neri), solifuge (Eremobates palpisetulosus), giant whipscorpion (Mastigoproctus giganteus), and scorpion (Diplocentrus bigbendensis). Bulletin of the British Arachnological Society 14: 1-8. — . 2007b. Spiders: Biology, Ecology, Natural History and Behavior. Brill Academic Publishers, Leiden, The Netherlands, iii + 431 pp. — and B. Garman. 1989. Effects of encounter experience on the hunting behavior of the spider wasp, Pepsis forniosa (Say) (Hymenoptera: Pom- pilidae). South-western Naturalist 34: 513-518. — and L. Henderson. 1999. Aspects of the natural history and behavioral ecology of the tarantula spider Aphonopelma hentzi (Chamberlin) (Orthog- natha, Theraphosidae). Bulletin of the British Arachnological Society 11: 121-128. and L. Ludwig. 2005. Behavioral responses of Pepsis thisbe (Hymenoptera: Pompilidae) to che- mosensory cues associated with host spiders. Journal of Insect Behavior 18: 757-766. Rau, P. and N. Rau. 1918. Wasp Studies Afield. Princeton University Press, Princeton, New Jer- sey. 372 pp. Schmidt, J. O. 2004. Venom and the good life in tarantula hawks (Hymenoptera: Pompilidae): how to eat, not be eaten, and live long, journal of the Kansas Entomological Society 77: 402-413. Smith, A. 1994. Tarantula Spiders: Tarantulas of the U.S.A. and Mexico. Fitzgerald Press, London. 312 Pp. Sokal, R. R. and F. J. Rohlf. 1995. Biometry. Third Ed. W. H. Freeman, New York. 881 pp. Tinbergen, N. 1951. The Study of Instinct. Oxford University Press, Oxford. 228 pp. — . 1972. The Animal in its World: Field Studies. Harvard University Press, Cambridge, Massachu- setts. 343 pp. U. S. Department of Interior. 2005, Division of Publications, National Park Service, National Park Handbook, No. 122, Washington, D.C., 216 PP- Vardy, C. R. 2000. The New World tarantula-hawk wasp genus Pepsis Fabricius (Hymenoptera: Pompilidae). Part 1. Introduction and the P. rubra species-group. Zoologische Verhandelingen 332: 1-86. . 2002. The New World tarantula-hawk wasp genus Pepsis Fabricius (Hymenoptera: Pompili- dae). Part 2. The P. grossa to P. deaurata groups. Zoologische Verhandelingen 338: 1-135. Vinson, S. B. 1984. Parasitoid-host relationship. Pp. in: W. J. Bell, and R. T. Carde eds. Chemical Ecology of Insects. Sinauer Associates, Sunderland, Massa- chusetts Williams, F. X. 1956. Life history studies of Pepsis and Hemipepsis wasps in California (Hymenoptera: Pompilidae). Annals of the Entomological Society of America 49: 447-466. J. HYM. RES. Vol. 16(2), 2007, pp. 311-325 Systematic Studies on the Pompilidae Occurring in Japan: Genus Irenangelus Schulz (Hymenoptera: Pompilidae: Ceropalinae) Akira Shimizu and Raymond Wahis (AS) Department of Biological Sciences, Graduate School of Science and Engineering, Tokyo Metropolitan University, Minami-Ohsawa 1-1, Hachioji, Tokyo, 192-0397 Japan; email: shimizu-akira@cmetro-u.ac.jp (RW) Entomologie fonctionnelle et evolutive, Faculte universitaire des Sciences agronomiques, B. 5030 Gembloux, Belgique; email: entomologie@fsagx.ac.be and raymond.wahis@skynet.be Abstract. — The Japanese species of the genus Irenangelus Schulz (Pompilidae: Ceropalinae) are revised. Three new species are described: I. hikosamis Wahis, /. nantbui Shimizu, and I. punctipleuris Wahis. Irenangelus hikosamis occurs in Japan (Honshu, Shikoku, Kyushu), Korea, Taiwan, and China; J. nambui occurs in Japan (Honshu) and Korea; J. punctipleuris is broadly distributed from Japan through the Philippines and Malaysia to India and Sri Lanka. Irenangelus Schulz 1906, like Ceropales Latreille 1796, is one of the most aberrant genera of the family Pompilidae. All members of both genera have an uncoiled antenna, reniform eyes with the inner orbits strongly emarginate and diverging above, a fully exserted labium in both sexes, and a strongly compressed metaso- mal sternum VI produced beyond tergum VI in the female. Evans (1969, 1987) reviewed the Neo- tropical species of Irenangelus. In these papers he treated ten species, seven of which were new, and regarded Xantham- pulex Schulz 1906 as a synonym of Irenan- gelus. Kimsey and Wasbauer (2004) revised the New World species of Irenangelus, adding two new species. As regards the Old World species of Irenangelus, Cameron (1891, 1896), Bingham (1896), Schulz (1906), Turner (1910), Rohwer (1919), Banks (1934), and Wahis (1988) described either single species or, at most, a few new species, but no comprehensive revisionary studies have been published. Species of Irenangelus have long been recognized as cleptoparasites of other pompilids, as are species of Ceropales. Thus, Williams (1919) reared an Irenangelus wasp from a nest of Auplopus nyemitawa (Rohwer 1919) and several wasps in this genus from cocoons of nests of Tachypompilus analis (Fabricius 1781) in the Philippines. At least one wasp reared belonged to I. luzonensis (Rohwer 1919). In Costa Rica, Wcislo et al. (1988) observed females of /. eberhardi Evans 1987 fly or perch near nests of Auplopus seinialatus Dreisbach 1963, enter an open cell containing a spider, and extend her gaster deep into the cell. They reared several wasps of this parasite from one of the nests of A. semialatus. Shimizu (see below) found females of a Japanese species of Irenangelus attempting to ovipos- it eggs into a slit of the booklung of heteropodid spiders that the host pompi- lids had captured and then transported. Irenangelus is mainly distributed in the Oriental and Neotropical regions. In Japan only one species of the genus has been known since Yasumatsu (1933) recorded the species from Honshu, Kyushu, and Taiwan as Xanthampulex pernix (Bingham 1896). This species is distributed also in Korea and China, and differs from X. pernix, which was originally recorded from 312 Journal of Hymenoptera Research "Tenasserim" (Burma). Recently speci- vis, California, USA; ELKU, Collection of mens of a further two species of this genus Entomological Laboratory, Kyushu Uni- were collected from Honshu, Japan. One of versify, Fukuoka, Japan; AEIC, American them is found in Korea in addition to Entomological Institute, Gainesville, Flor- Japan, and the other occurs from Japan ida, USA; FSAG, Entomologie fonction- through Southeast Asia to South Asia. It nelle et evolutive, Faculte universitaire was found that these three species are des Sciences agronomiques, Gembloux, undescribed. Belgique; RMNH, Nationaal Naturhistor- In this paper, we review the generic ische Museum, Leiden, Netherlands; characters and phylogenetic relationships BMNH, Natural History Museum, London, of Irenangelus, describe three new species UK; OMNH, Osaka Museum of Natural and provide a key to their identification. History, Osaka, Japan; CNC, Canadian National Collection of Insects, Ottawa, MATERIALS AND METHODS Ontario, Canada; NSMT, Department of The terminology of the wing veins and Zoology, National Science Museum, To- cells follows Day (1988). The following kyo> JaPan; TMUB, Laboratory of Zoolog- morphological terms and abbreviations are ical Systematica Department of Biological used: antennocular line, the anterior mar- Sciences, Tokyo Metropolitan University, gin of the frons in dorsal view; scutal Tokyo, Japan; USNM, National Museum of groove, a pair of longitudinal grooves Natural History, Washington, D. C, USA. between the notaulus and parapsidal sul- cus on the mesoscutum (Evans (1969) and Kimsey and Wasbauer (2004) called this ^ F , c , , y , „. _ T_ t \ . , Genus Irenangelus Schulz the notaulus ); LID, the lower interocular distance; MID, the middle interocular Irenangelus Schulz 1906: 175. Type of genus: distance; OOL, the ocello-ocular line; irenangelus hornus Schulz 1906: 160, by mono- POL, the postocellar line; SMC, the sub- v \ , c , , 1fin^ 1DO T , . r .. . . . . TTTT^ , Xanthampulex Schulz 1906: 1 83. Type of genus: marginal cell of the tore wing; UID, the v ,, , , c c , , /ani- 1QQ u & &/ ' Xmitluimpulex tnfur Schulz 1906: 183, by upper interocular distance. monotvpv Measurements were made in the follow- ing ways: clypeus length versus breadth, Description. — Further to the descriptions being measured comparing the length of of this genus by Evans (1969) and Kimsey the clypeus from the uppermost point of and Wasbauer (2004) we note the following the front-clypeal sulcus to the apical characteristics: gena flattened or concave margin to the maximum breadth of the just posterior to outer orbit at least below; clypeus; labrum length versus breadth, posterolateral margin of pronotum almost being measured comparing the longest straight (Figs 2F, 3D); scutal groove deeply part of the labrum to the breadth across impressed in many species (Figs 1A, 2C); the base of the labrum; breadth of flagello- metapostnotum well developed at least mere I, being measured across the maxi- medially (Figs 1A, D, 2 A, F, 3D); fore wing mum breadth of flagellomere I in dorsal vein M reaching outer wing margin view. In the description of each species, the (Figs 2A, 5A, B); metatibia with longitudi- measurements of the holotype are given in nal sharp groove along upper margin of parentheses. brush on inner side (Fig. 2A); apicoventral Specimen depositories are abbreviated seta on metatarsomere V long and setiform as follows: ZMUC, Zoological Museum, (see Shimizu et al. 1998: fig. 3); female University of Copenhagen, Copenhagen, laterosterna of metasomal sternum VI Denmark; UCDC, R. M. Bohart Museum of extending dorsad, scarcely overlapping to Entomology, University of California, Da- envelope sting apparatus; sting almost Volume 16, Number 2, 2007 313 Fig. 1. Irenangelus hikosanus n. sp. (A-F, holotype female; G-I, para type male from Japan) and female /. pernix Bingham from Bali, Indonesia (J). A, Head and mesosoma, dorsal view; B, head, lateral view; C, head, frontal view; D, mesosoma, lateral view; E, mesosternum and mesocoxa, ventrolateral view; F, right metatarsal claw, outer view; G, genitalia (left half, ventral view; right half, dorsal view); H, subgenital plate, ventral view; I, sternum VI, ventral view. Scale lines: 0.5 mm. straight; male sternum VI without a pair of sublateral hook-like projections posteriorly (Fig. II); digitus volsellaris with large semicircular emargination on inner margin (Figs 1G, 21, 3L). ' Diagnosis. — Irenangelus is closely related to Ceropales, forming a monophyletic group, the subfamily Ceropalinae (Shimizu 1994, Pitts et al. 2006). These two genera are distinguishable on the characteristics shown in Table 1. Phylogenetic relationships. — On the basis of morphological characteristics, both Shi- mizu (1994) and Pitts et al. (2006) treated the Ceropalinae (Ceropales + Irenangelus) as the most basal clade in the Pompilidae. This subfamily is thus considered to be the sister group to the rest of the family. Based on this hypothesis, there is a biological inconsisten- cy: although the Ceropalinae are placed as the earliest offshoot of the pompilid stock, all species of the subfamily whose behav- iour is known are cleptoparasitic. To avoid this contradiction, Shimizu (1994) consid- ered that the behavioural type of the Ceropalinae has evolved directly from that of the parasitoids, which is the most likely life history of a common ancestor of the 314 Journal of Hymenoptera Research Table 1. Comparison of differential characters between Irenangelus and Ceropales in the broad sense, including Priesnerius and Hemiceropales. Irenangelus Ceropales 1. Metacoxa normal-sized, less than 1.5X as long as Metacoxa much larger than mesocoxa, more than 1.5X mesocoxa. as long as mesocoxa. 2. Ventral angle of pronotum short and blunt, not Ventral angle of pronotum long and acute, partly attaining dorsal margin of procoxa (Fig. 3D). 3. Posterolateral margin of pronotum almost straight (Figs 2F, 3D). 4. Inner margin of male protarsomere V not produced. 5*. Male subgenital plate flat or gently convex (Figs 1H, 2H, 3L). covering dorsal margin of procoxa. Posterolateral margin of pronotum strongly curved inward. Inner margin of male protarsomere V produced ventrally. Male subgenital plate tectate with median carina. * Character treated by Kimsey and Wasbauer (2004). Pompilidae. Similarly, Day (1988: 16) stated: "it seems more probable that Ceropales has evolved from an ancestral group already specialized as ectoparasitioids." This pre- sumption seems likely given that the two important characteristics of Ceropalinae females are shared with other pompilids known to behave as parasitoids, viz- (i) the strongly exposed clypeus, and (ii) the laterally compressed sternum VI. Biology. — One or more Philippine species of Irenangelus are known to be cleptopar- asites of Auplopus nyemitawa (Pepsinae) and Tachypompilus analis (Pompilinae), and I. eberlmrdi a cleptoparasite of Auplopus semialatus (Pepsinae) (Williams 1919). Shi- mizu observed the cleptoparasitic behav- iour of /. lukosanus Wahis n. sp., which pursues its host pompilid Platydialepis ryoheii (Ishikawa 1956) (Pepsinae) as the latter transports her prey, Heteropoda for- cipata (Karsch 1881) (Heteropodidae) to her nest. Eventually the parasite pounces on the spider and extends her gaster, attempt- ing to insert its tip into a slit of the spider's booklung. The details of this behaviour will be treated in a separate paper. Distribution. — This genus is known from Oriental, Neotropical, Australian, East Asian, and Madagascan Regions (see Wa- his 1988), but is best represented in the first two regions. KEY TO FEMALES AND MALES OF IRENANGELUS OCCURRING IN JAPAN Flagellum crenulate in profile, i.e., flagellomeres II— X each with angular swelling below (Fig. 3A); all tarsal claws bifid, inner ray truncate (Fig. 3G); propodeum transversely striate; metasomal tergum I gradually narrowed and petiolate (tergum I narrower immediately behind articulation with propodeum than width at articulation itself) (Fig. 3E) or parallel-sided basally; head and mesosoma with irregularly-distributed punctures (Figs 4C-F). (Head and mesosoma black, varie- gated with bright yellow markings; metasoma and legs predominantly reddish brown; body length 8-12 mm) I. punctipleuris Wahis, n. sp. Flagellum not crenulate in profile; all tarsal claws dentate (Fig. IF) or sub-bifid (Fig. 2G), inner ray acute; propodeum smooth, never striate; metasomal tergum I abruptly narrowed, not petiolate or parallel-sided basally (Fig. 2A); head and mesosoma impunctate (Fig. 4B) 2 Propodeum with lateral tubercle between spiracle and posterior rim (Fig. 1A); interantennal area distinctly raised (Figs IB, 4B); fore wing crossvein cu- a originating at or slightly distal to separation of vein M+Cu (Fig. 5A); fore wing Volume 16, Number 2, 2007 315 SMC3 removed by approximately its own length from outer wing margin; apical margin of subgenital plate strongly convex (Fig. 1H); body predominantly yellowish brown; body length 8-15 mm J. hikosanus Wahis, n. sp. Propodeum without lateral tubercle between spiracle and posterior rim (Fig. 2A); interantennal area not raised, continuous to upper frons (Fig. 2D); fore wing crossvein cu-a originating slightly basad of separation of vein M+Cu (Fig. 2A); fore wing SMC3 removed by much more than its own length from outer wing margin; apical margin of subgenital plate slightly emarginate or almost truncate (Fig. 2H); body predominantly blackish brown to black; body length 5-9 mm I. ttambui Shimizu, n. sp. Irenangelus hikosanus Wahis, new species (Figs 1A-I, 4B, 5A) Xanthampulex pernix: Yasumatsu 1933: 143, figure 1, Jo, misidentification; Kim 1970: 807. Irenangelus pernix: Lelej et al. 1994: 145; Lelej et al. 1995: 46; Shimizu 1994: 45; Shimizu 1996: 507; Shimizu et al. 1998: 429, figure 3. Female. — Length: Body 9.1-14.5 (9.8) mm; fore wing 8.7-11.3 (8.7) mm. Coloration: Body and appendages predominantly yel- lowish brown and polished. Following light yellow: mandible (apical portion dark brown), clypeus (lateral side sometimes yellowish brown), frons along inner orbit, gena along outer orbit, ventral margin of scape, maxillary and labial palpi, prono- tum posteriorly and laterally, discs of scutellum and metanotum, posterior rim of propodeum, procoxa, and sometimes interantennal tubercle ventrally, labrum, and episternum posteriorly. Apical 4 or 5 flagellomeres black dorsally. Basilateral and posterior portions of metasomal ter- gum I and posterior portions of following terga more or less darkened. Metatarsus becoming darker towards apex. Wings hyaline with yellowish brown tint, irides- cent in certain lights, narrowly and weakly infuscate along outer margins. Pterostigma dark brown. Punctation: Body devoid of punctures. Pubescence and setae: Pubescence on body usually very short and decum- bent; metanotum, lateral portion of meta- postnotum, and propodeum with long, sub-erect, brown pubescence. Vertex, la- brum, mandible, propleuron, and sterna V- VI with short yellowish brown to brown setae; remainder of body and legs almost devoid of setae. Head: 1.1-1.2 (1.1 )x as broad as long. Vertex moderately to strongly convex between eye tops (Fig. 1C). Frons distinctly tuberculate be- tween antennal sockets (Figs IB, 4B); upper frons broadly depressed along median line, the latter being sharply impressed on antennal tubercle but becoming obscure near anterior ocellus. Antennocular line depressed beside antennal tubercle (Fig. 1A). Inner orbits distinctly emarginate at upper 1/3, gently convergent below (Fig. 1C). UID:MID:LID = 9.1-9.4:10:6.3-6.7 (9.4:10:6.3). MID 0.57-0.60 (0.58) X head width. Ocelli forming acute triangle, this area being distinctly raised. POL:OOL = 1:2.5-3.1 (3.1). Clypeus slightly convex, 1.8-2.0 (2.0) X as broad as long; anterior margin truncate, weakly and arcuately emarginate (Fig. 1C); lateral sides strongly convergent towards apex. Labrum 1.8-2.1 (2.1) X as broad as long; anterior margin feebly and triangularly emarginate. Man- dible narrowly rounded without sharp carina laterally. Malar space short (Fig. IB). Genae 0.4-0.5 (0.5) X eye width in profile, roundly receding in dorsal view. Scape with carina long but not sharp beneath; face slightly concave laterally in dorsal view. Flagellomere I 2.1-2.9 (2.5) X as long as wide and 0.34-0.48 (0.44) X as long as UID; flagellomeres I and II in ratio of 10:9.3-11 (10:10). Mesosoma: Pronotum 316 Journal of Hymenoptera Research short (Fig. 1A); anterior margin of disc Male. — Very similar to female. Length: arcuately convex in dorsal view, its ante- Body 7.1-12.4 mm; fore wing 7.1-11.4 mm. rolateral corner rounded; lower anterolat- Head: 1.1-1.2X as broad as long. UID:MID: eral tubercle not much swollen, being LID = 9.0-9.3:10:6.5-6.8. MID 0.57-0.62X almost concealed by disc in dorsal view head width. POL:OOL= 1:2.3-2.6. Clypeus (compare Fig. 1A with Fig. 1J: /. pernix); 1.8-2.0X as broad as long. Labrum 1.8- posterior margin arcuate with small medi- 1.9 X as broad as long. Gena 0.4-0.5 X eye an notch. Mesoscutum with scutal groove width in profile. Flagellomere I 2.0-2.5 X as sharply impressed anteriorly, becoming long as wide, 0.34-0.43 X as long as UID; shallower and broader posteriorly, but flagellomeres I and II in ratio of 10:9.2-11. almost attaining scuto-scutellar sulcus; Mesosoma: Metapostnotum 0.8-1 X length parapsidal sulcus appearing as a fine, of metanotum at midline. Legs: Longer raised line; posterolateral margin broadly spur of metatibia 0.69-0.76 X length of reflexed. Discs of scutellum and metano- metatarsomere I. Wings: Fore wing SMC2 turn remarkably projecting, the latter steep- receiving crossvein lm-cu at basal 0.56- ly falling posteriorly (Fig. ID). Posterome- 0.70. SMC3 narrowed on vein Rs by 0.67- dian lobes of mesosternum well developed 0.79X its length on vein M, 1.1-1.4X length and digitate, apices close to each other of SMC2 on vein M, 1.1-1.4X as long as (Fig. IE). Metapostnotum 0.7-1 (0.7) X as SMC2 on vein Rs, receiving crossvein 2m- long as metanotum at midline, deeply cu at basal 0.55-0.67. Subgenital plate sunken between metanotum and propo- (Fig. 1H): Lateral sides gently convergent deum (Fig. ID), with fine, transverse striae, towards apex; apical margin sub-triangu- Propodeum strongly depressed along ante- larly convex; ventral surface covered with rior margin, almost linearly sloping in minute setae except for subapical portion, profile, with one or two lateral tubercles Genitalia (Fig. 1G): Paramere with strong between spiracle and posterior rim setae apicomedially; parapenial lobe slight- (Fig. 1A); infrastigmal tubercle roundly ly extending beyond apex of aedeagus. raised; median groove impressed only Distribution. — Japan (Honshu, Shikoku, anteriorly; surface smooth, not striate. Me- and Kyushu), Korea, Taiwan (Yasumatsu tasoma: Slender and almost parallel-sided 1933), and China (Fig. 6). medially. Tergum I abruptly narrowed, not petiolate or parallel-sided basally. Legs: T^e ^erial.-Holotype 9 (ELKU), Japan, Longer spur of metatibia 0.69-0.77 (0.73)X K^lshup M*" Hik°san' ^r1** £.uYasUr , T _ , , , matsu. Paratypes: Japan: Kyushu: Mt. Hikosan, as ong as metatarsomere I. Tarsal claws Fukuoka Pref 5 viiU940/ K yasumatsu, 1«J with vertical tooth near middle (Fig. IF). (ELKU). Lake Yamashita, Kokonoe-machi, Oita Wings: Fore and hind wing venation as Pref., 9.ix.l997, R. Matsumoto, lo (OMNH). Mt. shown in Fig. 5A. Fore wing crossvein 2r-rs Ariake-yama, Izuhara-machi, Tsushima Is., originating beyond middle of pterostigma. 24.vii.2001, R. Oomuta, 1 J (TMUB). Japan: Crossvein cu-a originating at or slightly Shikoku: Mt. Ishizuchi-san, Omogo-mura, distal to point of separation of vein M+CuA. Ehime Pref., 17.viii.2002, M. Shiraishi, lo SMC2 trapezoid, receiving crossvein lm-cu (TMUB). Omogo, Omogo-mura, Ehime Pref., at basal 0.54-0.70 (0.58). SMC3 narrowed on 16.viii.1951, T. Esaki, 1 J (ELKU); 23.viii.1953, T. vein Rs by 0.70-0.82 (0.77) X its length on EdashiSe' W (TMUB); 23.ix.1999, A. Shimizu, 29 ,,«;,-. \/f 1 i 1 a n qw i Cw^n (TMUB). Japan: Honshu: Jomine Shrine, Yano, vein M, 1.1-1.4 (1.3) X as long as SMC2 on \, .. J . „ . J _, , _ ... ' . ' \/f 1 t 1 r /i a\., i ™ ,^~ Kamnzumi-mura, Saitama Pref., 3.vm.l994, T. vein M, 1.2-1.5 (1.4) X as long as SMC2 on M , 1im;Trm ~ ,.' ' . . . . ' b Nambu, 1J (TMUB). Onouchi, Ogano-machi, vein Rs, receiving crossvein 2m-cu at basal Saitama Pref ># 10.x.1992/ T. Nambll/ 1o (TMUB). 0.55-0.61 (0.55). Hind wing crossvein rs-m Onagata, Yoshida, Saitama Pref., 18.viii.1988, T. straight, oblique to vein M. Crossvein cu-a at Nambu, 69IJ (TMUB), 19 (FSAG); 26.viii.1988 ngle of approximately 150° to vein A. (29: TMUB, FSAG), 1, 3.viii.l994 (I9: TMUB), 24, Volume 16, Number 2, 2007 317 26.viii.1995 (2Q: TMUB), 28.vii.2001 (26*: TMUB), A. Shimizu. Riv. Ochi-gawa, Otaki-mura, Chi- chibu, Saitama Pref., 6.ix.l970, T. Nambu, 29 (TMUB); 6.ix.l999, A. Shimizu, 19 (TMUB). Kawamata, Otaki, Chichibu, 18.viii.2005, A. Shimizu, I9 (TMUB). Mt. Komaga-take, 1050 m, Hakone, Kanagawa Pref., 8.viii.2005, A. Shimizu, I3 (TMUB). East of Fujikawagu- chiko-machi, Minami-tsuru-gun, Yamanashi Pref., 5.viii.2006, H. Takahashi, 26* (TMUB). Mt. Sanage, Evergreen forest, Aichi Pref., deciduous forest, 8.ix.2002, P. Tripotin, I9. Jeollanamdo, Gurye-gun, Toji-myeon, Nae- dong-li, Piakol Valley, on wild wine flowers, 3.viii.2001, P. Tripotin, 19. Etymology. — This species is named after the type locality. Remarks. — This new species is similar to I. pernix, but the following characters distinguish them: /. hikosanus 1. Lower anterolateral tubercle of pronotum slightly and roundly produced, almost concealed by disc in dorsal view (Fig. 1A). 2. Interantennal tubercle merging into upper frons, with median line finely impressed (Fig. 4B). /. pernix Lower anterolateral tubercle of pronotum angulate, markedly projecting beyond disc in dorsal view (Fig. 1J). Interantennal tubercle abruptly raised from slightly depressed upper frons, with median line deeply and broadly impressed (Fig. 4A). 28.viii-3.ix. 1992, T. Kanbe, Malaise trap, I9 (TMUB). Hio, Kanazawa-shi, Ishikawa Pref., 27.viii, 1998, Y.Tazaki, 1 6* (NSMT). Misaka-dani, Izumi-mura, Ono-gun, Fukui Pref., 13.ix.2002, H. Takahashi, I9 (TMUB). Kaizuka-shi, Izumi- katsuragisan, Osaka, l.x.2000, R. Matsumoto, I9 (OMNH); 13-23.vii.2002 (16*), 23.vii-2.viii.2002 (26*), 2-10.viii.2002 (23), 20.viii-2.ix.2002 (I9), 2- 14.ix.2002 (19), 23.ix-2.x.2002 (I9), 2-11.X.2002 (16*), Malaise trap, R. Matsumoto, (OMNH). Kishiwada-shi, Izumi-katsuragisan, Osaka, 20- 30.vi.2002 (19), 30.vi-13.vii.2002 (I9I6*), 13- 23.vii.2002 (29), 23.vii-2.viii.2002 (16*), 2- 10.viii.2002 (26*), 10-20.viii.2002 (l;), 23.ix-2.x. 2002 (49), Malaise trap, R. Matsumoto, (OMNH). Six stage of Mt. Atago-yama, Ukyo-ku, Kyoto-shi, 27.viii.1987, A. Ichikawa, 19 (OMNH). Hanase Pass, Kyoto-shi, 10.ix.1999, R. Matsumoto, 19 (OMNH). Mimuro, Shingo-cho, Okayama Pref., 6.ix.l992, R. Matsumoto, 19 (OMNH). Kozagawa, Wakayama Pref., 20. ix. 1957, S. Momoi, 39 (TMUB). Daisen, Tottori Pref. (Hoki), 19.viii.1932, S. Yasimoto, 1J (ELKU). Korea: Chungcheong- namdo, Keumsan, Poseoksa, 10.viii.1998 (I9), 22.viii.1998 (I9), 24.ix.2000 (I9), ix.2001 (29), P. Tripotin, (FSAG). Kyeongsangnamdo, Jirisan, Hamyang-gun, Macheon-myon, Samjeong-li Jir- isan, 700 m, 23-25.viii.2002 (49), 10-20.ix.2003 (16*), Malaise trap, P. Tripotin, (FSAG). CHINA: Szechwan, Suifu, 1000-1500 m, l-21.vi.1928, D. Graham, 19 (USNM). Non-type material. — Korea: Chungcheong- namdo, Keumsan, Poseoksa, along trail in Irenangelns nambui Shimizu, new species (Hg. 2) Female. — Length: Body 4.5-7.6 (5.5) mm; fore wing 4.7-6.9 (5.2) mm. Coloration: Body predominantly blackish brown to black and polished. Following ivory-white: clypeus and labrum laterally, mandible (apical portion brown), ventral margin of scape, maxillary and labial palpi, procoxa (basal portion more or less dark brown), protrochanter, and sometimes profemur, mid and hind coxae, trochanters, femora and tibiae partly. Remainder of fore leg light brown. Mid and hind legs predomi- nantly brown, darker than fore leg, but somewhat lighter ventrally than dorsally. All tibial spurs ivory-white to yellowish light brown. Posterolateral margin of pro- notum, lateral and posterior portions of metasomal terga, and posterior portions of metasomal sterna light brown to ferrugi- nous. Wings hyaline, iridescent in certain lights, weakly infuscate along outer mar- gins. Pterostigma dark brown. Punctation: Body devoid of punctures. Pubescence ami setae: Body and legs with short, appressed white pubescence, longer and denser on lower frons, clypeus, lower pronotum, propleuron, mesopleuron, lower meta- pleuron, propodeum, and coxae. Upper 318 Journal of Hymenoptera Research Fig. 2. Irenangelus nambui n. sp. (A-G, holotype female; H-I, paratype male from Japan). A, Whole body, dorsal view; B, head, frontal view; C, head, pronotum, and mesoscutum, dorsal view; D, head, lateral view. E, mesosternum and mesocoxa, ventral view; F, mesosoma, lateral view; G, right metatarsal claw, outer view; H, subgenital plate, ventral view. I, genitalia (left half, ventral view; right half, dorsal view). Scale lines: 0.5 mm. irons, vertex, clypeus, labrum, mandible, apices of terga VI and sterna IV-V, and sternum VI with short pale setae. Head: 1.2X as broad as long. Vertex strongly convex in frontal view (Fig. 2B). Frons with interantennal area not tuberculate but slightly overhanging antennal sockets (Fig. 2D); median line impressed only on lower half. Antennocular line nearly trans- verse (Fig. 2C). Inner orbits slightly emar- ginate a little above middle, strongly di- vergent above. UID:MID:LID=9.8-10.1:10: 6.8-7.1 (10:10:6.9). MID 0.57-0.60 (0.60) X head width. Ocelli forming acute triangle, this area being scarcely raised. POL: OOL= 1:2.5-3.6 (1:3.6). Clypeus slightly convex, 2.1-2.4 (2.2) X as broad as long; anterior margin truncate, weakly and arcuately emarginate (Fig. 2B); lateral sides arcuately convergent towards apex. La- brum 2.5-3.0 (3.0) X as broad as long; anterior margin arcuately emarginate. Mandible carinate laterally. Malar space short. Genae 0.3-0.4 (0.4) X eye width in profile, roundly receding in dorsal view. Scape sharply carinate beneath; lateral face slightly concave in dorsal view. Flagello- mere I 2.9-3.5 (3.0) X as long as wide and 0.45-0.49 (0.47) X length of UID; flagello- meres I and II in ratio of 1:0.94-1.0 (1:0.98). Mesosoma: Pronotum short; anterior margin of disc arcuately convex in dorsal view, its anterolateral corner gently rounded (Fig. 2C); lower anterolateral tubercle not much swollen, being completely concealed by disc in dorsal view; posterior margin arcuate with small median notch. Mesos- cutum with scutal groove sharply im- pressed anteriorly, becoming shallower and broader posteriorly, obsolete just an- Volume 16, Number 2, 2007 319 terior to scuto-scutellar sulcus; parapsidal sulcus finely impressed; posterolateral margin narrowly reflexed. Discs of scutel- lum and metanotum distinctly projecting (Fig. 2F), the latter being pyramidal. Pos- teromedian lobes of mesosternum short but bilobed, their inner lobes close to each other (Fig. 2E). Metapostnotum 0.7-0.9 (0.7) X length of metanotum at midline, with few very fine striae anteriorly and distinct longitudinal median groove. Pro- podeum weakly convex in profile (Fig. 2F), scarcely depressed along anterior margin, without lateral tubercle or infrastigmal tubercle (Fig. 2A); median groove obsolete; surface smooth, not striate. Metasoma: Slenderly fusiform. Tergum I abruptly narrowed, not petiolate or parallel-sided basally. Legs: Longer spur of metatibia 0.65-0.74 (0.69) X metatarsomere I. Tarsal claws sub-bifid: inner ray sub-parallel to outer ray, acute. Wings: Fore and hind wing venation as shown in Fig. 2A. Fore wing crossvein 2r-rs originating before middle of pterostigma. Crossvein cu-a orig- inating slightly basad of point of separation of vein M+CuA. SMC2 rhomboid, receiving crossvein lm-cu at basal 0.43-0.56 (0.52). SMC3 narrowed on vein Rs by 0.42-0.53 (0.51) X its length on vein M, 1.2-1.5 (1.3) X as long as SMC2 on vein M, 0.69-1.1 (0.87) X as long as SMC2 on vein Rs, receiving crossvein 2m-cu at apical 0.54-0.66 (0.58). Hind wing crossvein rs-m almost straight, oblique to vein M. Crossvein cu-a forming angle of 135-140 to vein A. Male. — Very similar to female. Length: Body 3.9-8.3 mm; fore wing 3.7-6.5 mm. Head: 1.2X as broad as long. UID:MID: LID=9.9-10.1:10:7.0-7.7. MID 0.57-0.60 X head width. POL:OOL=l:2.1-3.1. Clypeus 2.3-2.5 X as broad as long. Labrum 2.6- 3.2 X as broad as long. Gena 0.3-0.4 X eye width in profile. Flagellomere I 2.2-2.7 X as long as wide and 0.38-0.43 X as long as UID; flagellomeres I and II in ratio of 1:0.96-1.1. Mesosoma: Metapostnotum 0.8- lx length of metanotum at midline. Legs: Longer spur of metatibia 0.66-0.73 X meta- tarsomere I. Wings: SMC2 receiving cross- vein lm-cu at basal 0.44-0.60. SMC3 narrowed on vein Rs by 0.44-0.73 X its length on vein M, 1.1-1.6X as long as SMC2 on vein M, 0.64-1 .4 X as long as SMC2 on vein Rs, receiving crossvein 2m- cu at apical 0.50-0.66. Subgenital plate (Fig. 2H): Broadened medially; apical mar- gin slightly emarginate or truncate; ventral surface covered with minute setae. Genita- lia (Fig. 21): Paramere without strong setae apicomedially; parapenial lobe short, not attaining apex of aedeagus. Distribution. — Japan and Korea (Fig. 6). Type material. — Holotype 9 (TMUB), Nageishi Pass, Higashi-Mikabo, Gunma Prefecture, 28.viii.1986, T. Nambu. Paratypes: Japan: Hon- shu: Showa, Mt. Hakase, 1000 m, Beech forest, Fukushima Pref., 29.vi-26.vii.1998 (1 ;), 27.vii- 23.viii.1998 (697^), 24.viii-19.ix.1998 (192^), Ma- laise trap, T. Muroi, (TMUB). Imperial Palace, Chiyoda-ku, Tokyo, 28.V.1999, T. Nambu, 1<$ (TMUB). Mt. Komaga-take, 1000-1300 m, Ha- kone, Kanagawa Pref., ll.vii.2000 (1^), 30.viii.2000 (I9), H. Nagase, (TMUB); 18.vii.2001 (1976*: TMUB; 1 £ FSAG), l.viii.2001 (26*: TMUB), A. Shimizu. Takekurabe-yama, Maruoka-cho, Fukui Pref., 5.ix.l994, Y. Haneda, I9 (TMUB). Akausagi-yama, Ohno-shi, 23. ix. 1974, Y. Ha- neda, 19 (FSAG). Shitara, Beech forest, 900 m, Uradani, Aichi Pref., 25-31.vii.1994 (1 J), 29.viii- 4.ix.l994 (1 ;), Malaise trap, K. Yamagishi, (TMUB); 29.viii-4.ix.1994, Emergence trap, K. Yamagishi, I9I 5 (TMUB); 19-25.ix.1994, Pan trap, K. Yamagishi, I9 (TMUB); l-7.viii.1994 (1J: TMUB), 22-28.viii.1994 (I9I6*: TMUB; 19, FSAG), Malaise trap, T. Kanbe. Asahi, Yawata, 650 m, Deciduous forest, 17-26.vi.1998 (2;), 12- 21.viii.1998 (I9I ?), 15-25.ix.1998 (19), Malaise trap, M. Ozawa, (TMUB). Mt. Sanage, Evergreen forest, Aichi Pref., 28.viii-3.ix.1992, Emergence trap, K. Shima, I9 (TMUB); 4-10.ix.1992, Malaise trap, T. Kanbe, 19 (TMUB); 16-22.ix.2002, Ma- laise trap, M. Kiyota, 1 J (TMUB). Korea: Kyeongsangnamdo, Jirisan, Hamyang-gun, Ma- cheon-myon, Samjeong-li, 700 m, 35 20'55N 127 38'21E, Malaise trap, 10-20.ix.2003, P. Tripo- tin, 292; (FSAG). Etymology. — This species is named in honor of the provider of the holotype specimen. 320 Journal of Hymenoptera Research Fig. 3. Irenangelus punctipleuris n. sp. (A-G, I, holotype female; H, J-K, paratype females: H, from Sulawesi; J, from Brunei; K, from Japan); L-M, paratype male from Japan). A, Head and antenna, lateral view; B, head, frontal view; C, head, dorsal view; D, mesosoma, lateral view; E, metasomal tergum I, dorsal view; F, mesosternum and mesocoxa, ventral view; G, right metatarsal claw, outer view; H-K, pronotum, dorsal view; L, genitalia (left half, dorsal view; right half, ventral view); M, sternum VII and subgenital plate, ventral view. Scale lines: 0.5 mm. Remarks. — In Irenangelus this species is unique in its predominantly dark brown to black body and wholly transparent wings. Irenangelus punctipleuris Wahis, new species (Figs 3, 4C-F, 5B-C) Female. —Length: Body 8.1-12.4 (10.0) mm; fore wing 6.2-9.3 (7.9) mm. Coloration: Head, mesosoma and coxae black with following bright yellow: clypeus and la- brum (lateral portions black), frons be- tween and below antennal sockets, upper frons along inner orbit and gena along outer orbit broadly, scape and pedicel (dorsal faces dark brown to black), prono- tal disc (lateral margin black), ventral and posterolateral margins of pronotum broad- ly, posteromedian elliptic spot and lateral streak on mesoscutum, median spot on scutellum, metanotal disc, oblong spot on upper mesopleuron, this spot being some- times obsolete, two large spots on lower mesopleuron, these often being continuous (Figs 4E-F), median triangular and lateral longitudinal marks on propodeum, these being continuous posteriorly, oblique spot on upper metapleuron, this spot being * . a ■ - Volume 16, Number 2, 2007 321 Fig. 4. Female head (A-D) and mesopleuron (E-F) of Irenangelus (A-B, dorsolateral view; C-D, frontal view; E-F, lateral view). A, /. pernix Bingham from Bali, Indonesia; B, /. hikosanus n. sp., holotype; C-F, /. punctipleuris, n. sp. (C, E, holotype; D, F, paratype from Japan). sometimes obsolete, posterodorsal mark on lower metapleuron, this being continuous with lateral propodeal mark, apical greater parts of coxae, and sometimes side of metanotum and anterior portion of meta- postnotum. Following reddish brown: fla- gellomeres I-IV or -X ventrally (remainder of flagellum dark brown to black), maxil- lary and labial palpi, trochanters (basally dark brown to black), femora (dorsolateral portion of profemur and sometimes ventral portions of meso- and metafemora bright yellow; sometimes all femora dark brown ventrally and/or laterally), tibiae (protibia bright yellow dorsolaterally; sometimes dorsal portion of mesotibia and basidorsal Fig. 5. Female wings (A, fore and hind wings; B, fore wing; C, hind wing). A, /. hikosanus n. sp., holotype; B-C, /. punctipleuris n. sp., paratype from the Philippines. 322 Journal of Hymenoptera Research 5 "5 "5- o E c ^ TO 3 Volume 16, Number 2, 2007 323 portion of metatibia dark brown to black), tarsi, and metasoma (tergum I dark brown to black anterodorsally and sublaterally; terga II— IV sometimes becoming darker dorsally). Posterolateral margin of mesos- cutum, tegula, and wing bases yellowish brown. Mandible black; apical 1/3 dark rufous. Wings hyaline, iridescent in certain lights, weakly infuscate along outer mar- gins. Pterostigma light to dark brown. Punctation: Upper frons (Figs 4C-D), vertex between eye and ocellus, pronotum, me- soscutum, discs of scutellum and metano- tum, mesopleuron (Figs 4E-F), and lower metapleuron with irregularly-spaced, shal- low punctures, these being larger and denser along median line of frons and scutal groove, and sometimes on meso- pleuron. Pubescence and setae: Body and legs with short, appressed, white to pale brown pubescence but devoid of long bristly setae; vertex, mandible, propleuron, lateral side of pronotal disc, mesopleuron, meta- notum, and posterolateral portion of pro- podeum with short, dense, white to pale brown setae. Head: 1.1-1.2 (1.2)X as broad as long. Vertex strongly convex in frontal view (Fig. 3B). Frons without interantennal tubercle (Fig. 3A); median line finely im- pressed from interantennal area close to anterior ocellus. Antennocular line slightly depressed nearby antennal base (Fig. 3C). Inner orbits distinctly emarginate at upper 1/3, strongly convergent below. UID: MID:LID = 8. 5-9.4:10:5.6-6.5 (9.3:10:5.9). MID 0.59-0.63 (0.61) x head width. Ocelli forming slightly acute triangle, this area being slightly raised. POL:OOL = 1:1. 6-2.5 (1.9). Clypeus feebly convex, 1.6-1.9 (1.8) X as broad as long; anterior margin truncate, weakly and arcuately emarginate (Fig. 3B); lateral sides strongly convergent towards apex. Labrum 2.0-2.3 (2.2) X as broad as long; anterior margin with small median notch. Malar space very short (Fig. 3A). Mandible sharply carinate laterally. Gena 0.4-0.5 (0.4) X eye width in profile, feebly rounded in dorsal view. Scape carinate on apical half ventrally; lateral face flattened but scarcely concave in dorsal view. Fla- gellomere I 2.1-2.4 (2.4) X as long as wide and 0.38-0.43 (0.38) X length of UID; flagellomeres I and II in ratio of 1:0.84-1.0 (0.94). Mesosoma: Pronotum short; anterior margin of disc almost straight in dorsal view, its lateral corner sub-angulate, but degree of angulation variable (Figs 3H-K); lower anterolateral tubercle not much swollen, being concealed by disc in dorsal view; posterior portion narrowly but dis- tinctly depressed along posterior margin, the latter being arcuate. Mesoscutum with scutal groove shallowly impressed on anterior 1/4-3/4; parapsidal sulcus ap- pearing as a fine, raised line; posterolateral margin narrowly reflexed. Discs of scutel- lum and metanotum strongly raised (Fig. 3D). Metapostnotum 0.59-1.0 (1.0) X length of metanotum at midline, deeply sunken between metanotum and propo- deum, with fine, transverse striae anterior- ly and short oblique striae posteriorly. Posteromedian lobes of mesosternum tri- angularly produced, their apices removed from each other (Fig. 3F). Upper meta- pleuron finely and obliquely striate. Pro- podeum short, barely convex in profile (Fig. 3D), deeply depressed along anterior margin, without lateral tubercle; infrastig- mal tubercle weak; surface finely and transversely striate, with weak to rudimen- tary median groove. Metasoma: Much slen- derer than mesosoma. Tergum I gradually narrowed and petiolate or parallel-sided basally (Fig. 3E). Legs: Longer spur of metatibia 0.63-0.72 (0.68) X length of meta- tarsomere I. Tarsal claws bifid; inner ray of claw truncate (Fig. 3G). Wings: Fore and Fig. 6. Map showing the known distribution of Irenangelus hikosanus, I. nambui, and /. punctipleuris. In fapan only certain localities of specimens, including type localities, have been selected for /. hikosanus and /. nambui. 324 Journal of Hymenoptera Research hind wing venation as shown in Figs 5B and C, respectively. Fore wing crossvein 2r-rs originating slightly before middle of pter- ostigma. Crossvein cu-a usually originating at or slightly basad of point of separation of vein M+CuA. SMC2 almost rectangular, receiving crossvein lm-cu at basal 0.43- 0.57 (0.54). SMC3 narrowed on vein Rs by 0.75-0.85 (0.84) X its length on vein M, 1.3- 1.8 (1.8)x as long as SMC2 on vein M, 1.2- 1.8 (1.8) X as long as SMC2 on vein Rs, receiving crossvein 2m-cu at basal 0.39-0.61 (0.50). Hind wing crossvein rs-m almost vertical to vein M. Crossvein cu-a forming angle of approximately 150 to vein A. Male. — Very similar to female. Length: Body 6.7-8.4 mm; fore wing 5.7-7.0 mm. Head: 1.2X as broad as long. UID:MID: LID=9.0:10:6.7-7.3. MID 0.62-0.65X head width. POL:OOL=l:1.9-2.1. Clypeus 2.0x as broad as long. Labrum 2. 1-2.2 X as broad as long. Gena 0.4-0.5 X eye width in profile. Flagellomere I 2.3-2.4 X as long as wide and 0.32-0.35 X UID; flagellomeres I and II in ratio of 1:0.87-0.90. Mesosoma: Metapostnotum 0.75 X length of metano- turrt at midline. Legs: Longer spur of metatibia 0.72 X metatarsomere I. Wings: SMC2 receiving crossvein lm-cu at basal 0.42-0.55. SMC3 narrowed on vein Rs by 0.77-0.78 X its length on vein M, 1.4-1.7X as long as SMC2 on vein M, 1.3-1.5 X as long as SMC2 on vein Rs, receiving cross- vein 2m-cu at basal 0.45-0.67. Subgenital plate (Fig. 3M): Lateral sides gradually convergent towards apex; apical margin sub-triangularly produced; ventral surface with minute setae apically. Genitalia (Fig. 3L): Paramere with strong setae api- comedially; parapenial lobe long and slen- der, decurved apically, extending beyond apex of aedeagus. Distribution. — From Japan through the Philippines and Malaysia to India and Sri Lanka (Fig. 6). Type material— Holotype 9 (ZMUC), Philip- pines, Balabac Dalawan Bay, 7.x. 1961, Noona Dan Exp. 61-62. Paratypes: Philippines: Tawi Tawi, Tarakawan, north of Batu Batu, 4.xi.l961 (19: ZMUC), 10.xi.1961 (19: FSAG), 12.xi.1961 (19: ZMUC), Noona Dan Exp. 61-62. Brunei: Ulu Temburong, Base camp hut, 300 m, 115 16'E 4 26'N, 16.ii-9.iii.1982, M. C. Day, 29 (BMNH, FSAG). Sulawesi: Central Sulawasi, Napu-valley, 100 km S/O, Palu, near Lore- Lindu National Park, 9.U.2001, A. -M. Klein, I9 (FSAG). Utara, Dumoga-Bone Nat. Park, ii.1985, 19 (BMNH). Bali: W. Bali, near Negara, rain- forest above Batuagung, 550 m, 4-6.xii.1911, C. v. Achterberg, I9 (RMNH). Java: W. Java, Djampang-Tengah, Mrs. Walsh, 19 (RMNH). Borneo: Sarawak, S.W. Gunung Buda, 64 km S. Limbang 4 13'N 114 56'E, 8-15.xi.1996, Malaise trap, S. L. Heydon & S. Fung, 19 (UCDC). Malaysia: S. E. Sabah, Danum Valley Field C, 117 48'E 4 58'N, x-xii.1986, P. Eggleton, 49 (BMNH), 19 (FSAG). S. E. Sabah, near Danum Valley Field, ca.150 m, 26.v-20.vi. 1987 (49: RMNH; 39: FSAG), 20.vi-12.vii.1987 (I9: RMNH), 13.ix-4.x.l987 (I9: RMNH), Malaise trap, C. v. Achterberg & D. Kennedy. S. W. Sabah, near Long Pa Sia (West), 1020 m, 25.xi- 9.xii.l987, Malaise trap, C. v. Achterberg, 19 (RMNH). S. W. Sabah, near Long Pa Sia (East), 1000 m, l-13.iv.1987 (I9), 25.xi-9.xii.1987 (19), Malaise trap, C. v. Achterberg, (RMNH). Pasoh Forest Reserve, Negeri S., 22.vii.1978 (I9: AEIC), 7.ix.l978 (19: FSAG), 6.xi.l978 (I9: AEIC), 8.i.l979 (19: AEIC), P. & M. Becker. Bukit Kutu, 304.1930, H. T. Padgen, 19 (BMNH). India: U. P. Garjia, 610 m, 26-29.iv.1969, Gupta, No. 335, 1J (FSAG). S. India, Madras ST., Anamalai Hills, 3500 f, v.1964, P. S. Nathan, 19 (CNC). Kerala, Periyar A. Sanctuary, 5-15.X.1979, 19 (BMNH). Sri Lanka: Kandy District, Udawattakele Sanc- tuary, 1800 f, l-3.ix.1980, Malaise trap, K. V. Krombein et al, 19 (FSAG). Japan: Kawamata, Otaki-mura, Chichibu, Saitama Pref., 30. vi- l.vii.2004, A. Shimizu, 19 (TMUB). Maruno- machi, Nirasaki-shi, Yamanashi Pref., 27.vi- 5.vii.2005, Malaise trap, K. Hosoda, 13 (TMUB). Kanegasaki-cho, Tsuruga-shi, Fukui Pref., 3.vii.2001, H. Takahashi, 19 (TMUB). Etymology. — The species name is derived from the punctate mesopleuron; puncti- (punctate) + pleuris (pleuron). Remarks. — The present species is similar to "Ceropales" tennatus Turner 1910 occur- ring in Australia in that 1) the flagellum is crenulate; 2) all tarsal claws are bifid; 3) the frons is devoid of an interantennal tubercle; Volume 16, Number 2, 2007 325 and 4) the metasomal tergum I is gradually narrowed and petiolate or parallel-sided basally. However, this new species is easily distinguished from the latter by the almost entirely rufous metasoma and the distinctly punctate mesopleuron. ACKNOWLEDGMENTS We thank Emeritus Professor R. Ishikawa (Tokyo Metropolitan University) and Dr N. Springate (Natu- ral History Museum, London) for reviewing our manuscript and providing critical comments. For the gift or loan of specimens, our thanks are also due to the following: M. C. Day, T. Edashige, T. Esaki, Y. Haneda, K. Hosoda, L. S. Kimsey (UCDC), K. V. Krombein (USNM), R. Matsumoto (OMNH), S. Mo- moi, H. Nagase, T. Nambu, R. Oomuta, A. Shinohara (NSMT), M. Shiraishi, O. Tadauchi (ELKU), H. Takahashi, Y. Tazaki, P. Tripotin, L. B. Vilhelmsen (ZMUC), K. Yamagishi, Y. Yasimoto, and K. Yasu- matsu. LITERATURE CITED Banks, N. 1934. The Psammocharidae of the Philip- pines. Proceedings of the American Academy of Arts and Sciences 69: 1-117. Bingham, C. T. 1896. On some exotic fossorial Hymenoptera in the collection of the British Museum, with descriptions of new species and of a new genus of the Pompilidae. journal of the Linnean Society of London (Zoology) 25: 422-A45. Cameron, P. 1891. Hymenoptera Orientalis; or Con- tributions to a knowledge of the Hymenoptera of the Oriental Zoological Region. Part 3. Memoirs and Proceedings of the Manchester Literary and Philosophical Society (4) 4: 431-481, pi. 3. . 1896. Hymenoptera Orientalis; or Contribu- tions to a knowledge of the Hymenoptera of the Oriental Zoological Region. Part 5. Memoirs and Proceedings of the Manchester Literary and Philo- sophical Society 41 (4): 1-144, pis 3, 4. Day, M. C. 1988. Spider wasps, Hymenoptera: Pompilidae. Handbooks for the Identification of British Insects 6 (4): 1-60. Evans, H. E. 1969. Studies on Neotropical Pompilidae (Hymenoptera) VII. Irenangeliis Schulz. Studio Entomologica 12: 417-431. . 1987. A new species of Irenangeliis from Costa Rica (Hymenoptera: Pompilidae: Ceropalinae). Proceedings of the Entomological Society of Washing- ton 89: 559-561. Kim, C.-W. 1970. Illustrated Encyclopedia of Fauna and Flora of Korea. Vol. 11, part 3. Samwha-Chulpansa, Seoul. 891 pp. [In Korean.] Kimsey, L. S. and M. S. Wasbauer. 2004. Revision of New World species of the cleptoparasitic pompilid genus Irenangeliis Schulz (Hymenop- tera: Pompilidae). journal of the Kansas Entomolog- ical Society 77: 650-668. Lelej, A. S., T. Saigusa, and C. E. Lee. 1994. Spider wasps (Hymenoptera, Pompilidae) of Korea. Russian Entomological journal 3: 135-148. , T. Tano, and H. Kurokavva. 1995. Spider wasps (Hymenoptera, Pompilidae) from Cheju- Do Island, Southern Korea. Transactions of Essa Entomological Society 75: 44-47. Pitts, J. P., M. S. Wasbauer, and C. D. Dohlen. 2006. Preliminary morphological analysis of relation- ships between the spider wasp subfamilies (Hymenoptera: Pompilidae): revising an old problem. Zoologica Scripta 35: 63-84. Rohwer, S. A. 1919. Philippine wasp studies. Part. 1. Descriptions of new species. Bulletin. Hawaiian Sugar Planters' Association Experiment Station, Entomological Series 14: 5-18. Schulz, W. A. 1906. Spolia Hymenopterologica. A. Pape, Paderborn. 355 pp. Shimizu, A. 1994. Phylogeny and classification of the family Pompilidae (Hymenoptera). Tokyo Metro- politan University Bulletin of Natural History 2: 1-142. . 1996. Key to the genera of the Pompilidae occurring in Japan north of the Ryukyus (Hyme- noptera) (Part 2). Japanese Journal of Entomology 64: 496-513. -, M. S. Wasbauer, and M. Ujiie. 1998. Taxo- nomic importance of the female apicoventral setae on metatarsomere V of the Pompilidae (Hymenoptera), with special reference to phvlog- eny of the family. Entomological Science 1: 427-439. Turner, R. E. 1910. Additions to our knowledge of the fossorial wasps of Australia. Proceedings of the Zoological Society of London 1910: 253-356, pis 31, 32. Wahis, R. 1988. Hymenopteres Pompilides de Mada- gascar. Genus Ceropales Latreille et Irenangeliis Schulz. (Hymenoptera: Pompilidae). Revue de Zoologie Africaine 102: 213-221. Wcislo, W. T., M. J. West-Eberhard, and W. G. Eberhard. 1988. Natural history and behavior of a primitively social wasp, Auplopus semialatus, and its parasite, Irenangeliis eberhardi (Hymenoptera: Pompilidae). Journal of Insect Behavior 1: 247-260. Williams, F. X. 1919. Philippine wasp studies. Part 2. Descriptions of new species and life history studies. Bulletin. Hawaiian Sugar Planters' Associ at ion Experiment Station, Entomological Series 14: 19-186. Yasumatsu, K. 1933. [Two unrecorded Psammochar- idae from Japan and Formosa.] Fukuoka Hakubu- tsugaku Zasshi, 1: 143-149. [In Japanese.] J. HYM. RES. Vol. 16(2), 2007, pp. 326-335 Natural History and Larval Behavior of the parasitoid Zatypota petronae (Hymenoptera: Ichneumonidae) Ju-Lin Weng* and Gilbert Barrantes1 Escuela de Biologia, Universidad de Costa Rica, Ciudad Universitaria Rodrigo Facio, San Jose, Costa Rica Abstract. — The koinobiont ectoparasitoid Zatypota petronae Gauld (Ichneumonidae) parasitizes medium-sized immatures of the cobweb spider Theridion evexum Keyserling (Theridiidae). Zatypota petronae apparently attacks the spider inside its retreat. An egg is glued on the antero-lateral dorsal section of the spider's abdomen. First-instar larvae remain partially inside the egg chorion which is attached to the spider's abdomen. In later instars, a layer of a brownish material (saddle), to which the 7th and 8th abdominal segments of the larva adhere ventrally, anchors the larva to the spider. In the last instar the saddle includes the egg chorion and the shed exoskeletons of previous instars. A row of retractile, dorsal protuberances, crowned with hooklets, is present on abdominal segments 1 to 8 of the final-instar larva. The larva uses the hooklets to grab silk lines of the retreat of the spider's web. Hanging on the spider's web the larva kills the spider and sucks out its body tissues. Then the larva pushes vigorously laterally with its head against the spider's corpse, and alternately presses the corpse against the saddle. These movements, in combination with peristaltic movements, free the larva from the saddle that falls to the ground with the dead spider. The larva then constructs its pupal cocoon. Prior to cocoon construction, the larva induces the spider to reinforce the retreat by adding more threads. Parasitism rate and host behavior are also described. The polysphinctine pimplines are koino- biont ectoparasitoids of spiders in several families (Nielsen 1923, 1932, Fincke et al. 1990, Hanson and Gauld 1995, Gauld et al. 1998). Nielsen (1923, 1932) described in detail the behavior of the larvae and hosts of several European polysphinctine spe- cies. The parasitism rates and life cycle of Hymenoepiiuecis robertsae Gauld on the neotropical tetragnathid Nephila clavipes (L.) was described by Fincke et al. (1990). However, larval behavior of neotropical polysphinctine wasps has been described in detail for only one species, H. argyr- aphaga Gauld on the tetragnatid Plesiometa argyra (Walker) (Eberhard 2000a, 2000b, 2001). Current address: Department of Entomology, Kansas State University, Manhattan, Kansas, USA ' Author for correspondence Parasitoid wasps of the cosmopolitan speciose polysphinctine genus Zatypota Forster parasitize spiders in at least five families (Dictynidae, Agelenidae, Tetra- gnathidae, Araneidae and Theridiidae) (Shaw 1994, Gauld et al. 1998). In the neotropics the only two host records were Theridion species: T. contreras Levi for an unidentified Zatypota species (Jimenez 1987) and T. evexum Keyserling for Z. petronae Gauld (Barrantes and Weng in press). The larval behavior of Zatypota sp. (Jimenez 1987) differs in some aspects from that of European polysphinctine species (Nielsen 1923, 1932) and H. argyrapmaga (Eberhard 2000a). The larva of Zatypota sp. was said to hold on to the spider by biting the dorsum or sides of the anterior section of the spider's abdomen. This description is likely wrong as detailed descriptions of the behavior of the larva of Z. albicoxa Volume 16, Number 2, 2007 327 (Nielsen 1923) and the larva of H. argyr- aphaga (Eberhard 2000a) show that some posterior segments of the larvae lodge ventrally in a "saddle", probably coagulat- ed spider's hemolymph, that adheres tight- ly to the spider's abdomen (Nielsen 1923, Eberhard 2000a). There is no further in- formation on the biology of larvae of this Zatypota species. Here we describe the intensity of parasitism and behavior of the larva of Z. petronae and its host T. evexum. We describe for the first time how a polysphinctine larva frees itself from the spider's corpse. MATERIALS AND METHODS Field observations were made from October 2005 to October 2006 in a 250 m2 plot in the understory of a middle-eleva- tion wet forest patch (9 54'N, 84 03' W; elevation 1200 m), the Reserva Biologica Leonel Oviedo on the Universidad de Costa Rica campus, San Jose Province, Costa Rica. All spiders (or nearly so) from third-forth instar outside the egg sac to adults were checked for parasites every two weeks; most spiderlings disperse from the mother's web at fourth instar. The small eggs and early instar larvae probably went undetected. Theridion evexum constructs most webs between 0.20 to 1.5 m above the ground (Barrantes and Weng in press), making it possible to find practically all webs. Addi- tionally, we collected seven parasitized spiders and kept them on their webs indoors to observe the behavior of larvae and spiders. In two cases we transplanted the plant on which the parasitized spider had constructed its web indoors, allowing us to observe the larva and host behavior with little disturbance. The complete larval development was not observed in all cases, so sample sizes are not always the same. Behavior and morphological features of more than 10 larvae were observed under a dissecting microscope. Video recordings of behavior were made using a Sony DCR - VX 1000 camcorder with +5 close-up lenses. Drawings of larval behavior were traced from video recordings. Voucher specimens of wasps and spiders were deposited in the Museo de Zoologia of the Universidad de Costa Rica. Wasp species names follow Gauld et al. (1998). RESULTS Percentage of parasitism. — Only juveniles of T. evexum were found to be parasitized by Z. petronae. A second instar larva was feeding on a juvenile spider, possiblv a third instar. However, final instar larvae were found feeding on large immature spiders, possibly juveniles of fourth to fifth instars. The parasitism in T. evexum was very low (mean percentage of parasitism/ census = 1.39%, SD = 1.80, n = 53 bi- weekly censuses). The reproduction in T. evexum is extremely seasonal, and the abundance of immature spiders (4th instar or larger) susceptible to attack by Z. petronae increased in March and declined drastically through August (Fig. 1). Be- tween September and February the popu- lation consists, first, of mature females, and then of very small spiderlings (Fig. 1). Parasitized spiders occurred primarily from March through August. Spider web and wasp attack. — The web of T. evexum includes a folded leaf that forms a conical retreat, with a tangle in front of the retreat opening, and long viscid threads extending from the tangle to other leaves (Barrantes and Weng in press). An additional tangle is constructed by the spider inside on the upper side of the retreat. We witnessed one attack bv a female Z. petronae wasp. The wasp approached the web and hovered in front of the spider's retreat opening. The wasp then flew inside the retreat. A few seconds later, the spider, with the wasp perched on its dorsum, dropped about 10 cm below the retreat, and hung on its dragline. They struggled for a few seconds and then the wasp flew out of sight. The spider began to climb towards the retreat but after advancing 328 Journal of Hymenoptera Research 350 n Dec Feb Apr Jun Aug Sep Nov Date Fig. 1. Temporal changes in the number of immature spiders (black rhombus), males (open circles), and reproductive females (black circles) of T. evexutn. about four centimeters, it became para- lyzed and fell back, motionless for about 10 min, dangling from its dragline. The spider recovered its motion slowly, and with clumsy movements cleaned some of its legs before ascending to the retreat. When we returned, 30 min later, the spider had fully recovered its mobility. We did not ascertain whether the spider had an egg on its abdomen. Larvae. — We observed one egg of Z. petronae glued on the antero-lateral dorsal section of the abdomen of a spider collect- ed in the field, a first instar larva emerged about four hours later. All eleven larvae of different instars checked under the dissect- ing microscope were attached by their rear end to the cuticle of the antero-lateral surface of the spider's abdomen (Fig. 2A). The first instar larva (n = 3) had its posterior end lodged inside the egg chori- on, with its head, thorax, and some abdominal segments protruding; the cho- rion remained attached to the spider. In the "second" instar, larvae (n = 5) were completely outside the collapsed, flattened egg chorion that was embedded in an apparently rigid, semitransparent layer of brownish material (Fig. 2B) (the "saddle" of Nielsen 1923). The ventral surface of two or three posterior abdominal segments rested on the saddle. In subsequent instars, the cuticles of the previous molts became incorporated into the saddle as they ad- hered to its upper surface, against the ventral surface of the larva. The egg chorion was near the spider's surface, but not in contact with it. The saddle was attached by a short pedicel to the spider's abdomen (Fig. 2B), and the larva's abdom- inal segments 7 and 8 secured it to the saddle. Feeding scars were observed on the nearby dorsal and lateral surface of the spider's abdomen (Fig. 2B). In the final instar, larvae had dorsal, two-lobed, retractable tubercles on eight abdominal segments (1st to 8th); these structures were absent in previous stages. Volume 16, Number 2, 2007 329 A * larva B y A i v v a •-- ■