SOCIETy, Journal o Hymenoptera Research Volume 3 August 1994 ISSN 1070-9428 CONTENTS ANTROPOV, A. V. A review of the agile species group of Pison (Hymenoptera: Sphecidae: Trypoxylini) 119 AZEVEDO, C. O. Descriptions of two new species and notes on the genus Bakeriella Kieffer from Brazil and Ecuador (Hymenoptera, Bethylidae) 145 BELSHAW, R. and B. BOLTON. A survey of the leaf litter ant fauna in Gahana, West Africa (Hymenoptera: Formicidae) ... 5 BITONDI, M. M. G., Z. L. P. SIMOES, A. M. do NASCIMENTO and S. L. GARCIA. Variation in the haemolymph protein composition of confined Apis mellifera and potential restoration of vitellogenin titre by juvenile hormone analogue treatment 107 BOHART, R. M. A review of North American Belomicrus (Hymenoptera, Sphecidae, Crabroninae) 207 DINIZ, I. R. and K. KITAYAMA. Colony densities and preferences for nest habitats of some social wasps in Mato Grosso State, Brazil (Hymenoptera, Vespidae) 133 FIELD, S. A. and M. A. KELLER. Localization of the female sex pheromone gland in Cotesia rubecula Marshall (Hymenoptera: Braconidae) 151 GUPTA, V. K. A review of the world species of Orthomiscus Mason (Hymenoptera: Ichneumonidae: Tryphoninae) 157 HERATY, ]. M., ]. B. WOOLLEY and D. C. DARLING. Phylogenetic implications of the mesofurca and mesopostnotum in Hymenoptera 241 KERR, W. E and E. B. SOARES REZENDE. Genetic characters of African bees that have high adaptive value in the tropics . 1 KUGLER, C. Revision of the ant genus Rogeria with descriptions of the sting apparatus (Hymenoptera: Formicidae) 17 NIEVES-ALDREY, J. L. Revision of West-European genera of the tribe Aylacini Ashmead (Hymenoptera, Cynipidae) 175 OVRUSKI, S. M. Immature stages of Aganaspis pelleranoi (Brethes) (Hymenoptera: Cynipoidea: Eucoilidae), a parasitoid of Ceratitis capitata (Wied.) and Anastreplm spp. (Diptera: Tephritidae) 233 POLASZEK, A. and K. V. KROMBEIN. The genera of Bethylinae (Hymenoptera: Bethylidae) 91 QUICKE, D. L. J. Myosomatoides gen. nov. (Hymenoptera: Braconidae), a Neotropical larval parasitoid of stem-borer pests, Diatraea (Lepidoptera: Pyralidae) 227 QUINTERO A., D. and R. A. CAMBRA T. Systematics of Pseudomethoca areta (Cameron): sex association, description of the male and a gynandromorph, and a new synonymy (Hymenoptera: Mutillidae) 303 TANG, Y. and P. M. MARSH. A taxonomic study of the genus Ascogaster in China (Hymenoptera: Braconidae: Cheloninae) . 279 Additions and corrections to Volume 2, Number 1, 1993 309 INTERNATIONAL SOCIETY OF HYMENOPTERISTS Organized 1982; Incorporated 1991 OFFICERS FOR 1994 George C. Eikwort, President Donald L. J. Quicke, President-Elect Michael E. Schauff, Secretary Gary A. P. Gibson, Treasurer Paul M. Marsh, Editor Subject Editors John Huber, Arnold Menke, David Rosen, Mark Shaw, Robert Matthews All correspondence concerning Society business should be mailed to the appropriate officer at the following addresses: President, Department of Entomology, Cornell University, Ithaca, New York 14853; President- Elect, Department of Biology, Imperial College at Silwood Park, Ascot, Berks SL5 7PY, England; Secretary, c/o Department of Entomology, NHB 168, Smithsonian Institution, Washington, DC 20560; Treasurer, Biological Resources Division, CLBRR, Agriculture Canada, K.W. Neatby Building, Ottawa, Ontario, Canada K1A 0C6; Editor, P. O. Box 384, North Newton, Kansas 67117. Membership. Members shall be persons who have demonstrated interest in the science of entomology. Annual dues for members are $25.00 (U.S. currency) per year, payable to The International Society of Hymenopterists. Requests for membership should be sent to the Treasurer (address above). Journal. The journal is published once a year by the International Sociey of Hymenopterists, c/o Department of Entomology, NHB 168, Smithsonian Institution, Washington, DC 20560, U.S.A. Members in good standing receive the Journal of Hymenoptera Research. Nonmember subscriptions are $50.00 (U.S. currency) per year. The Society does not exchange its publications for those of other societies Please see inside back cover of this issue for information regarding preparation of manuscripts. Statement of Ownership Title of Publication: Journal of Hymenoptera Research. Frequence of Issue: Once a year (currently). Location of Office of Publication, Business Office of Publisher and Owner: International Society of Hymenopterists, c/o Department of Entomology, NHB 168, Smithsonian Institution, Washington, DC 20560, U.S.A. Editor: Paul M. Marsh, P. O. Box 384, North Newton, Kansas 67117. Managing Editor and Known Bondholders or other Security Holders: none This issue was mailed 15 October 1994 George Campbell Eickwort 1940-1994 We were saddened to hear of the untimely death of George Eickwort, President, International Society of Hymenopterists, on July 11, 1994. George died from injuries suffered in an automobile accident while vacationing in Jamaica. This issue of the Journal of Hymenoptera Research is dedicated to his memory and to honor his contributions to the study of Hymenoptera and to this Society. First Announcement Third International Hymenoptera Conference August 12-17, 1995 University of California Davis, California, USA Talks, poster presentations and symposia on all aspects of Hymenoptera, including: •Biological control •Behavior •Morphology •Genetics •Systematics Also, possiPle tours of the Bohart Museum, California Academy of Sciences, San Francisco and the northern California wine country To be put on mailing list tor second announcement with registrationand call for papers, send name, address, phone, fax and/or e-mail numbers to: Dr. Lynn S. Kimsey Department of Entomology University of California Davis, California 95616 USA Phone: (916) 753-5373 FAX (916) 752-1 537 A few grants to cover travel costs may Pe availaPle to participants from countries with foreign currency or other financial proPlems. Mention your need when requesting the second accouncement. J. HYM. RES. Vol 3, 1994, pp. 1-4 Genetic Characters of African Bees That Have High Adaptive Value in the Tropics Warwick Estevam Kerr and Eduardo Badue Soares Rezende Universidade Federal de Uberlandia, Departamento de Biociencias, 38400-902, Uberlandia, MG, Brazil Abstract- The Ac gene present in Africanized Apis mellifera populations is male limited and confers bronze color to the abdomen; females are non-affected. The gene ac confers yellow abdomen equally to Italian female and male bees. Afncan-mt- DNA is found in African and most Africanized populations, and European-mt-DNA occurs in European populations, and descendents. The original frequency in Camaqua in 1957 of ac was 0.939 and for Ac 0.061. In 1991, after about 68 generations, these frequencies changed to 0.038 toacand 0.962 to Ac, what gives an adaptive value to ac in the tropics of 0.952 (considering 1.00 to Ac). The same mav happen tothemt-DNAs, what mav cause the mates European-mt-DNA queens X Afncan-mt-DNA males to be less fit than the Afncan-mt-DNA queens X European-mt-DNA males. INTRODUCTION There is a constant search in tropical research for characters that give to the bearer species a higher adaptive value when compared to Euro- pean ones. This happened with corn, wheat, cab- bage, apples, cattle, goats, etc. Many morphologi- cal, behavioral and physiological characters are being studied in European and African bees, and in their hybrids under different ecological condi- tions. Among them, two of the African characters are particularly interesting because within a few generations they became predominant in a tropi- cal environment, but not in a temperate one. These characters are the gene Ac and the African-mt- DNA (that acts as a single gene). The population of this area is being well studied as far as their enzymes are concerned and show that the Africanized honeybee is an admixture of Apis mellifera mellifera (19.5%), Apis mellifera ligustica (3.8%) and Apis mellifera scutellata (76.7%) (Lobo et al. 1989; Del Lama et al. 1990). The African-mt-DNA has been studied by Hall and Muralidharan (1989), Sheppard et al. (1991), Sheppard et al. (1991), Soares (1992) and others. It shows indications of high adaptive value in the tropics and low in southern South America (that has European-like climate), where it is domi- nated by European mt-DNA. The Ac gene has been observed by Brazilian bee biologists since 1957. In 1969 the information published that it is male sex limited: it confers bronze color to the abdomen, especially to the tergites, while workers are yellow (Kerr 1969). It was present in 100% of the 145 queens collected in Africa and brought to Brazil in 1956. There was some information (Prof.V. Portugal Araujo, p.c.) that yellow males were occasionally present in Angolan populations. However, they were never seen by W. E. Kerr in his 1956 trip to Africa. It is an allele of b(=black) (Woyke and Kerr 1989), segre- gates lAc: lac in different genetic background, and is not linked to five components of agressive behavior (Stort 1978). MATERIAL AND METHODS Between 19 July 1990 and 17 September 1991, 14 swarms of Africanized bees (Apis mellifera Linne) entered emptv stingless bee hive boxes located in the Apiary of the Universidade Federal de Uberlandia, MG, Brazil. Hive boxes varied in size, from 15 to 45 litres in volume, located 420 km from Camaqua, the point of introduction of African bees, in 1957. All swarms contained many males. A sample of males was taken from each swarm and the numbers expressing the Ac and ac alleles were determined. Journal of Hymenoptera Research Males from an additional 17 colonies of Africanized bees were sampled from commercial honeybee hives that were occupied by Africanized swarms. Frequencies of the ac and Ac genes were calculated and compared with estimates of the frequencies of the same alleles in Camaqua in 1957. RESULTS Six hundred and ninety (690) Ac drones (0.9623) and 27 ac drones (0.0377) were sampled from the 31 colonies. No data for 1992 and 1993 were collected because all swarms (22 and 9, re- spectively) had only Ac drones. Drones in a swarm come from several colonies. Some additional observations made are im- portant to note: 1. Of the 14 swarms, only one had two queens (the same proportion found in Kerr et al. 1970), 2. in three swarms many bees with wax scales were seen, which indicates that bees of different ages were in the swarms, 3. in 1992 two and in 1993 two swarms arrived without males. The lack of ac drones in 1992 and 1993 swarms is assumed to be a consequence of the population reaching fixation (100% Ac genes) or near fixation. DISCUSSION Since the gene Ac segregates independent of the xo gene, is independent of genes for defensive behavior and is an allele of black (b), the hypoth- esis that it was linked and continues to be linked with genes for high fitness after 68 generations of meiosis was discarded. Several traits of Africanized bees have been and are being selected for Brazilian conditions (Page and Kerr 1991; Kerr 1992). The bees are becoming less aggressive; they are being selected for greater resistance to the Varroa jacobsoni mite (Moretto et al. 1991); they do not any more reject Italian foundation and they use fewer armadillo holes (Kerr 1992). The high frequencies of Ac and of African-mt-DNA are also a result of natural selection, since the degree of natural crosses be- tween Africanized x Italian and Italian x Africanized are about equal when queens and drones of both races use the same mating ground (Kerr and Bueno 1970). In 1956, there were 400 Italian hives located in the same Eucalyptus forest into which 26 swarms of Africanbees escaped. All Italian colonies had ac and all African colonies had Ac genes. Therefore, assuming equal contributions of all escaped colo- nies to the breeding population, a frequency of 0.939 for the gene ac and 0.061 for Ac in 1957 was obtained for that original population. According to Winston (1992 pg. 40-42) an individual Africanized colony swarms about 16 times per year. Of course, this can only happen in the expan- sion phase of the population, before the popula- tion approaches the carrying capacity of the envi- ronment. According to Nascimento (1981 page 166), however, this figure, obtained in 1980, is 1.5 swarms per year. For our estimate a conservative "intermediate" figure of two successful swarms per colony per year will be used. The original frequency F1"57 of the gene ac (0.939) in 1957, multiplied, per generation, by its relative adaptive value (w) will give us the fre- quency F1*"1 for ac in 1991, that is 0.0377, 34 years later, or after 68 generations. Therefore: F1991 = wn. F11^7. In this formula n, the number of generations, is 68, F1957 is 0.939, F,MM1 is 0.0377 and our estimation for w, the fitness of ac, is 0.952 (considering 1.00 to Ac). This relatively low adaptive value w of ac is the reason for the almost universal presence of Ac in feral populations of Apis meUifern in tropical South America. It may be a similar reason for the the high frequency of African-mt-DNA found in Africanized populations, that is colonies of Euro- pean-mt-DNA queens x African-mt-DNA drones should be less fit than colonies of African-mt- DNA queens x European-mt-DNA drones. The bronze color of the Ac drones may not be the cause of its fitness, since the workers are yel- low and heat preservation by dark color would be better in temperate climate; the physiological rea- son are being studied. ACKNOWLEDGMENTS We thank the FAPEMIG (Stateof Minas Gerais Research Foundation) and CNPq (Brazilian National Research Coun- cil) for financial help and Dr. Robert E. Page, Jr. forcorrecting our English and for sound suggestions. Volume 3, 1994 LITERATURE CITED Del Lama, M. A., J. A. Lobo, A. E. E. Soares, S. N. Del Lama, 1990. Genetic differentiation estimated by isozymic analysis of Africanized honeybee populations from Bra- zil and from Central America. Apidologie 21: 271-280. Hall, H. G. and K Muralidharan. 1989. Evidence from the mitochondrial DNA that African honey bees spread as continuous maternal lineages. Nature 339: 211-213. Kerr, W. E. 1969. Genetica e melhoramento de abelhas. In: Melhoramento e Genetica. Organized by W.E.Kerr, in homage to Prof. F.G. Bneger. EDUSP, Melhoramentos, USP, Sao Paulo. Cap. XIV pg. 263-295. Kerr, W. E. 1992. Abejas africanas: su introduccion y expan- sion en el continente americano. Subespecies y ecotipos africanos. Industria Apicola N" 13: 12-21. Kerr, W. E. and D. Bueno. 1970. Natural crossing between Apis mellifera adansonii and Apis mellifera ligustica. Evolu- tion 24(1): 145-148. Kerr, W. E., L. S. Goncalves, L. F. Blotta and H. B. Maciel. 1970. Biologia comparada entre abelhas italianas (Apis mellifera ligustica), africana (Apis mellifera adansonii) e suas hibridas. Anais do V Congresso Brasileiro de Apicultura (Flonanopolis, SC) pg. 151-185. Lobo, J. A., M. A. Del Lama, and M. A. Mestnner. 1989. Population differentiation and racial admisture in the Africanized honeybee (Apis mellifera L.). Evolution 43(4): 794-802. Moretto, G., L. S. Goncalves and D. De Jong. 1991 . Africanized bees are more efficient at removing Varroa jacobsoni. Preliminary data. American Bee Journal. 131: 434. Nascimento, A. F., Jr. 1981. Estudo da mfluencia de fatores ambientais no comportamento enxameatono, migratono e no desenvolvimento de colmeias africanizadas. M.Sc. Thesis presented to the University of Sao Paulo at Ribeirao Preto, Brazil. Page, R. E, Jr., and W. E. Kerr. 1991 . Honey bee genetics and breeding. 8th article of The "African" Honey Bee, pp 157 - 186. Ed. Maria Spivak, David J. C. Fletcher and Michael D. Breed. Westview Studies in Insect Biology. Westview Press, Boulder, USA. Sheppard, W. S., A. E. E. Soares, D. De Jong and H. Shimanuki 1991. Hybrid status of honey bee populations near the historic origin of afncanization in Brazil. Apidologia 22: 643-652. Sheppard, W.S., T.E. Rinderer, J. A. Mazolli, J. A. Steiner and H. Shimanuki. 1991. Gene flow between African and European derived honey bee population in Argentina. Nature 349: 782-784. Soares, A.E.E. 1992. A utilizacao da Genetica molecular e da morfometna na caractenzacao de populacoes de abelhas africanizadas. Naturalia (Edic;ao Especial, 14 a 18 de setembro de 1992) pg. 117-125. Stort, A. C. 1978. Genetic study of the aggressiveness of two subspecies of Apis mellifera in Brazil. VII. Correlation of the various aggressiveness characters among each other and with the genes for abdominal color. Ciencia e Cultura 30(4): 492-496. Winston, Mark L. 1992. Killer Bee - The Africanized Honey Bee in the Americas. Harvard University Press, London, En- gland Woyke, I. and W.E. Kerr. 1989. Linkage test between a sex limited color gene and sex alleles in the honey bee. Brazilian Journal of Genetics 12(1): 9-15. Journal of Hymenoptera Research Table 1. Data on swarms and hives at Uberlandia, Minas Gerais, Brazil, with reference to frequency of Ac and ac alleles. N° Day Arrival Hour Hive Volume (Liters) Frequency Of Genes Place 01 19.07.90 15:00 27 25 Ac Kerr's House 02 05.08.90 11:00 27 16 Ac Kerr's House 03 12.08.90 08:45 (1)(3) 10 Ac Kerr's House 04 19.08.90 10:45 27 16 Ac Kerr's House 05 11.09.90 12:30 27 20 Ac Kerr's House 06 21.07.91 13:15 27 24 Ac Kerr's House 07 14.08.91 14:20 27 15 Ac Kerr's House 08 04.08.91 15:15 27 08 Ac Kerr's House 09 05.08.91 14:28 (li 05 Ac Kerr's House 10 08.08.91 15:00 27 10 Ac Kerr's House 11 09.08.91 12:00 45 23 Ac Kerr's House 12 14.08.91 14:20 27 (3) 14 Ac Kerr's House 13 27.08.91 12:29 27 47 Ac Kerr's House 14 24.11.91 (1) 15 (3) 34 Ac Kerr's House 15 12.05.91 (2) — 43 Ac Alvorada Farm 16 11.08.91 (2) — 23 Ac Rio Pedras Farm 17 11.08.91 (2) — 23 Ac Rio Pedras Farm 18 11.08.91 (2) — 23 Ac Rio Pedras Farm 19 11.08.91 (2) — 23 Ac Alvorada Farm 20 11.08.91 (2) — 23 Ac Alvorada Farm 21 11.08.91 (2) — 23 Ac Sta. Rita Farm 22 06.08.91 (2) — 24 Ac Sta. Rita Farm 23 06.08.91 (2) — 24 Ac St\ Rita Farm 24 04.09.91 (2) — 24 Ac Alvorada Farm 25 04.09.91 (2) — 23 Ac Alvorada Farm 26 04.09.91 (2) — 24 Ac Rio Pedras Farm 27 06.08.91 (2) — 24 Ac Rio Pedras Farm 28 17.09.91 (2) — 24 Ac Alvorada Farm 29 17.09.91 (2) — 23 Ac Alvorada Farm 30 17.09.91 (2) — 23 Ac Alvorada Farm 31 17.09.91 (2) 34 Ac 690 Ac (96.23%) 27 ac (3.77%) Rio Pedras Farm (l)Hour not noticed. (2) Langstroth hive. (3) Enormous swarm. J. HYM. RES. Vol. 3, 1994, pp. 5-16 A Survey of the Leaf Litter Ant Fauna in Ghana, West Africa (Hymenoptera: Formicidae) Robert Belshaw and Barry Bolton Biodiversity Division, Department of Entomology, Natural Historv Museum, Cromwell Road, London SW7 5BD, U.K. Abstract. — Leaf litter samples were taken from 34 sites scattered across the moist tropical forest zone in Ghana. They included areas of primary forest, secondary forest and cocoa. Over 40,000 individual ants were extracted using Winkler bags and identified. The species found are listed together with their abundance and a summary of their distribution. A total of 176 species was found (excluding stray workers of arboreal and surface-foraging species), almost two-thirds of which were Myrmicinae. The composition of the fauna is discussed and compared with that found in other tropical forests The species composition at the different sites showed little variation either between the different forest types or with geographic distance. INTRODUCTION In the West African forest belt there has been little quantitative sampling of the ant fauna; stud- ies have been carried out in the Tai Forest Reserve, Cote d'lvoire (see Levieux 1982 and included ref- erences) and the Reserve de Campo, Cameroun (Halle and Pacal 1992: 65-109). In Ghana there has been no quantitative sampling except in the main tree-crop, cocoa. This has been studied in detail (e.g. Majer 1976 and included references) but the extent to which it resembles the original forest fauna is not known. Nevertheless, this research has led to the taxonomy of West African ants being more advanced than that of most tropical ant faunas (e.g. Bolton 1987 and included references). In consequence we can survey elements of this fauna with the hope of accurately identifying much of it to species. Ghana has two main terrestrial biomes, sa- vannah and forest, and these have distinct ant faunas. In turn, the forest zone is readily divisible into a canopy and a ground fauna. In this paper we survey the leaf litter element of the forest ground fauna by sampling at different localities across Ghana. In addition to identifying the species present, we discuss the composition of the fauna and compare it with that from other tropical for- ests. We also examine how the species composi- tion at the sites varies geographically and between the different forest types. Voucher specimens of all taxa are deposited in the Natural History Museum, London. METHODS Sites. — The locations of the sampling sites in Ghana are shown in Figure 1, with brief descrip- tions and sampling dates given in Table 1. Sites designated by the same letter but with different numbers (e.g. jl and j2) are within 3km of each other. With one exception, the sites are within the moist semi-deciduous forest zone of Hall and Swaine (1976). We sampled in a wide range of the forest habitats found in Ghana, including 14 areas of primary forest (in the broad sense of forest with a closed high canopy), 10 areas of secondary forest (of varying age, mostly on agricultural land) and 10 cocoa farms. Sampling was carried out between December 1991 and November 1992. Sampling. — At each site an area of approxi- mately 1000m: was measured out. Within this area ten lm2 quadrats were placed at random. All the leaf litter inside a quadrat was collected, shaken through a 1cm sieve, and then left for three days in a Winkler bag. The extracted ants were combined to form a single total for each site, each site being sampled on only one occasion. All sampling was done between 9.30 a.m. and 3.00 p.m. At three sites (h, ql and cj2) an additional soil Journal of Hymenoptera Research sample was taken from each quadrat. This was done by collecting the soil from a 25cm by 25cm quadrat (= 0.0625m2) to a depth of approximately 5cm. This soil was then sieved and left for three days in Winkler bags in the same manner as the overlying leaf litter. The Winkler bag (Besuchet el al. 1987) oper- ates in a similar fashion to a Berlese Funnel except that the material is left hanging in a mesh bag to dry in air rather than exposed to a heat source. Winkler bags are much cheaper and easier to use and to transport than Berlese Funnels. Litter-sift- ing followed by extraction in Winkler bags records many species which do not turn up in pitfall traps (Olson 1991). Our extraction period of three days was chosen on the basis of a trial extraction, with daily sorting of a sample over a two week period. We found that within the first three days 86% of the individuals and 88% of the species had emerged. We ignored winged reproductives and wing- less queens found without workers, except in spe- cies where the queen is known to forage during nest foundation. Analyses. — In order to assess the complete- ness of our survey for the region sampled, i.e. the moist semi-deciduous zone plus disturbed habi- tats within, we plotted a species accumulation curve. We first arranged the sites in five random sequences. In each sequence we calculated the number of species found at the first site, the first two sites combined, the first three sites combined, and so on. Finally, the mean of the five sequences was calculated. The extent to which the resulting curve flattens out indicates the proportion of the actual fauna which has been recorded; a failure to flatten out indicates that additional species would have been found if the sampling had been contin- ued. Palmer (1990, 1991 ) compares and tests meth- ods for estimating the species richness of a region from samples taken within it. He concludes that the first-order jacknife is the most precise method, i.e. the one whose estimates are closest to the true value, and we therefore also apply this analysis to our data. We converted the body lengths of each species to biomass (= dry weight) using the following equation, taken from Gowing and Recher (1984). Logn weight(mg) = -4.0 + 2.5(logn length(mm)) Table 1. Sampling sites with dates and habitat description. Sacred groves are small pieces of forest left in agricultural areas for religious reasons. (Note that Ghanaian place name spellings are often variable.) a Sui River Forest Reserve, 1.x. 1992, primary forest, b Mabang, 18.xii.1991, secondary forest, c Tinte Bepo Forest Reserve, 31.iii.1992, primary forest, d Mankrang Forest Reserve, 11. in. 1992, primary forest, e Poano, 9.1x1992, cocoa, f near Ofinso, 2.X1.1992, cocoa. g Jachie, 20. iv. 1992, sacred grove (28 acres), primary for- est, h Effiduase, 17.xi.1992, cocoa. i Bobiri. 1 - 6.1V.1992, Forest Reserve (primary forest); 2 8. iv. 1992, Forest Reserve (primary forest but all mature trees killed with sodium arsenide in 1947); 3 - 13. iv. 1992, secondary forest (farmland left in 1982 and burnt in 1983). I Juaso. 1 - 21. ix. 1992, secondary forest (area of Dome River Forest Reserve burnt in 1983); 2 - 23.ix.1992, sec- ondary forest (farmland left for c.20 years); 3 - 8.ix.l992, cocoa k Southern Scarp Forest Reserve (North-West of Mpraeso near Osubeng), 23.x. 1992, secondary forest (burnt in 1983). 1 Kade. 1 - 6.x. 1992, primary forest (in Aiyeola Forest Reserve); 2 - 12.x. 1992, secondary forest (farmland left in 1957); 3- 12.x. 1992, cocoa, m Esukawkaw Forest Reserve, 27.x. 1992, primary forest, n Nkawanda (near Nkawkaw), 12.xii.1991, roadside sec- ondary forest, o Atewa Forest Reserve, primary forest. 1 - 2.iii.l992, near Kibi; 2 - 24.ih.1992, near Potrase; 3 - 26.h.l992, near Sagymasi (logged in 1970's); 4 - 27. m. 1992, nearSagymasi (logged in 1970's). p Asiakwa, 1. v. 1992, cocoa. q Bunso. 1 -6.xi.1992, Crops Research Institute arboretum (15 acres), primary forest; 2 - 17.iv.1992, secondary for- est (primary forest partially cleared ca 20 years previ- ously, left undisturbed for ca 12 years); 3 - secondary forest (cocoa left in 1981);4 -24.ii.1992, cocoa;5-6.iii. 1992, cocoa, r Old Tafo sacred grove (ca 3 acres), 31. i. 1992, primary forest. s New Tafo (Cocoa Research Institute of Ghana). 1 - 23.xii.1991, secondary forest (farmland left for en 40-50 years); 2 - ll.xii.1991, cocoa, t Nankasi, 17.ix.1992, cocoa. Body lengths (= the outstretched length of a point-mounted worker including mandibles) were taken from the literature or from an average of five specimens. In species with a dimorphic worker caste we did not count major and minor workers separately. Instead we used an estimate of 30:1 for the ratio of minor to major workers in all cases. To assess the effect of the distance between Volume 3. 1 994 1 1 1 IVORY r -8° y 1 "\ \n TOGO coastt \\ \ V ( »" V / DS • !''• 6'' cl> 0 f; / 1 * - , ^ Sunyani -7°/ b C / MS Pvf ' / (9 S- - - / r < \ ' / 9^ > \ ' - \ \ N - - -a, , . \ v Wiawso -6° W Kumasi ***. ^^ h ] e Lake Bosumtwi a 1 rs ^\ C /''Ho\ Koforidua ~~~X / \ \ ME Dunkwa % *' Accra WE \ \ i ,' ■' _ *Lape Coast t J^ Takoradi C*J O^ 100 km 3° 2° i° 0° 1° 1 " Fig. 1 . Map of the forest zone of Ghana showing sampling sites (see table 1 for details). Forest and forest-type boundaries are shown by dotted lines ( — ). Forest-type abbreviations: DS = Dry Semi-deciduous; MS = Moist Semi-deciduous; ME = Moist Evergreen; WE = Wet Evergreen. From Hall and Swaine (1981). sites on their species composition we first mea- sured the similarity between all possible pairs of sites using the Morisita Index (see Wolda 1981). Their values were then plotted against the loga- rithm of the distance between the sites, and a regression line fitted. As the data points are not independent, this procedure is of questionable statistical validity; it does, however, allow us to see the general pattern. RESULTS AND DISCUSSION We extracted 43,824 ants in 197 species and 47 genera. At least 17 of these species and two of the genera were undescribed (one of these genera is now described in Belshaw and Bolton (1994) and the other is currently being described by Prof. W.L. Brown under the name Loboponera). A further 57 species are in groups which lack keys or synthesising taxonomic studies, and we can there- fore neither identify them nor recognise them as being undescribed. The species are listed in Table 2, along with the total number of individuals found in each species and a summary of their distribution among the sites. Of the 197 species recorded, three were only found in the soil samples and a further 18 are known not to forage in the leaf litter and are classed here as tourists. The latter are all arboreal species except for Camponotus, which forage on the surface of the ground. Although these tourists s Journal of Hymenoptera Research comprised 9% of the species found, they com- prised only 2% of the total number of individuals. In the leaf litter the average density of indi- viduals (excluding tourists) was 117m2. Using Berlese funnels, other studies have found similar densities of ants in leaf litter: in tropical moist forest on Barro Colorado Island, Panama, in the wet season it was ca 200m: (Levings 1983); in temperate deciduous forest in Maryland, U.S.A. between May and September the mean monthly density was 194m 2 (Lynch and Johnson 1988). Composition of the fauna. — The relative impor- tance of the different subfamilies in the combined soil and leaf litter samples is shown in Fig. 2. The fauna is dominated by Myrmicinae. Where their biology is known, the species we found are all generalist predators / scavengers except for most Dacetonini, the workers of which forage singly for Collembola and other soft-bodied arthropods (us- inghighly specialised mandibles), and Decamorium decern (Forel), which has been observed attacking termites (Bolton pers. obs.). Paedalgus distinctus Bolton and Belshaw has also been found in asso- ciation with termites (Bolton and Belshaw 1993, see below). The Ponerinae is the second most important subfamily in our survey. Many ponerines are large ants, so although the subfamily comprises only 8% of the total number of individuals, it comprises 27% of the total biomass. A good example is Paltothyreus tarsatus (Fabricius): only 25 individu- als of this species were found during the survey, but its contribution to the total ant biomass was exceeded by only two other species. This subfam- ily contains a high proportion of genera which are specialised predators. In our survey we recorded Ambh/opone (preys on geophilomorph centipedes), Plectroctena macgeei Bolton (other members of the genus prey on millipedes), Paltothi/reus tarsatus (preys on termites), and Discothyrea (preys on arthropod eggs) (all sources in Holldobler and Wilson 1990: 559). Driver ants (Aenictinae and Dorylinae) have highly aggregated distributions - their colonies are nomadic and very large (with between 60,000 and 20,000,000 workers in other species (Holldobler and Wilson 1990: 581)). They are represented in our survey by single workers found at three sites and by almost 1500 workers from one soil quadrat that hit part of a Dorylus nest. This group has clearly not been adequately sampled in our sur- vey, and it is probably a much more important component of African forests than our results indicate (see Holldobler and Wilson 1990: 588). We found one worker of Apomyrma stygia Brown, Gotwald and Levieux in damp leaf litter a mile inside the Esukawkaw Forest Reserve. This is the sole described species in the Apomyrminae and was recorded previously only from one local- ity in Cote D'lvoire (Brown etal. 1970), where four nests were found in soil under gallery forest and one under adjacent unburnt savannah. The spe- cies is clearly subterranean (the workers are blind ), and the remains of a geophilomorph centipede were found in one of the Cote D'lvoire nests, raising the possiblity that the species specialises on this prey item. Habitat and geograpliic variation. — In a separate paper we have investigated in detail the differ- ences in the ant assemblages between the primary forest, secondary forest and cocoa sites in this study (Belshaw and Bolton 1993). We found that they did not differ significantly either in species composition or in species richness. Only two com- mon species show an association with a particular forest type (see Table 2): Serrastruma hijae (Forel) was with the exception of a single individual found only in the Esukawkaw and Atewa Forest Reserves, and one Oligomyrmex species (sp.indet.4) mostly occurred in cocoa. Given the large number of species present, one would predict that, even if all species were randomly distributed among the three forest types, a handful of such apparent associations would occur purely from chance. The slope of the regression line in Fig. 4 is very shallow, and the (logarithmically-transformed) distance between sites only accounts for 1.9% of the variation in similarity. The distance between sites therefore had at the most only a slight effect on species composition. Comparison with other faunas. — Several other studies of tropical forest ant faunas contain lists of leaf litter species: dry forest at Kimberley, N. Aus- tralia (Andersen and Majer 1991); moist forest on Barro Colorado Island, Panama (Levings 1983); and dry forest in Madagascar (Olson and Ward, in press). Following elimination of obvious arboreal species, these studies all show the Myrmicinae to be the largest subfamily, followed by the Ponerinae Volume 3, 1994 100 80 o c o 60 40 20 0 [1 Biomass ■ Individuals D Species ii L nl ^1 nL, 11 J I 1 In ^1 'V* Subfamily Fig. 2. Relative importance of the different ant subfamilies in the total fauna (leaf litter and soil samples combined, tourists excluded). Names are abbreviated (see table 2). (in our study comprising 63% and 22% of the total number of species respectively). They comprised 52% and 30% at Kimberley (33 species in total), 65% and 27% on Barro Colorado Island (93 species in total), and 55% and 20% in Madagascar (44 species in total). The figures in the last study would have been higher but for the unusually large number of Cerapachys species. The absence of doryline and aenictine ants from the island has possibly permitted diversification in this genus of specialised predators of other ants (Wilson 1971: 68). Unfortunately, it is not possible to compare the species richness of the four areas owing to the differences in sampling area, effort and method (Berlese funnel, Winkler bag and pitfall trapping). 10 Journal of Hymenoptera Research ■M e u 200 o o, 150 Cm O ,0 a 100 50 0 0 10 20 30 40 Number of sites Fig. 3. Species accumulation curve for the leaf Utter samples. Line fitted by DWLS smoothing option of computer program SYSTAT (Wilkinson 1990). Efficiency of survey. — The species accumula- tion curve is shown in Fig. 3. If the fitted line is extrapolated beyond the data it does not continue to rise. Estimating the true regional species rich- ness using the first-order jacknife method, we find our total of 176 species represents 81% of the actual species present. We infer from this that, within the area of Ghana sampled, we recorded a large ma- jority of the species foraging within leaf litter. In tropical forests the majority of ground ant species nest in the leaf litter, either in small pieces of rotting wood or between compressed leaves (Wilson 1959; Bolton pers. obs.). In addition, some other species nest in the top l-2cm of the soil but forage in the leaf litter. However, there are other smaller elements of the ground fauna which, be- cause of their nesting or foraging habits, are likely to have been missed by our sampling method. 1) Completely subterranean species (= ones which nest and forage only in the soil). In Ghana only a few such species are known, e.g. Plectroctena anops Bolton and P. hastifera (Santschi) (the work- ers of which are either blind {anops) or with very small eyes {hastifera)). However, this microhabitat Volume 3. 1994 11 s 75 1.0 X 0.8 > 0.4 £ 0.2 0.0 0 1 Log10 (distance (km) + 1) Fig. 4. Plot of similarity in species composition against distance for all possible pairwise comparisons between sites. Line fitted by linear regression (y = 0.392 - 0.060x), r2 = 0.019. has to date hardly been sampled at all, and its fauna is not understood. 2) Species which nest in rotting logs. Such species usually also restrict their foraging to this microhabitat (Wilson 1959). For example, we failed to find Plectroctena minor Emery, which is not uncommon in eastern Ghana where it is usually found in or under rotten logs (Bolton 1974). Many Tetramoriini also nest in rotting logs (see Bolton 1980, and below). 3) Obligatory termitolestic species (= species which nest and forage only within termitaria). This is a small group: in Ghana the habit is con- firmed only in species of Carebara and Centromyrmex. We found Paedalgus distinctus and P. saritus Bolton and Belshaw foraging in leaf litter which indicates that, despite observations on other members of the genus (see Bolton and Belshaw 1993), they do not belong in this group. 4) Species which only forage nocturnally are likely to be under- rep resented. In Ghana, among the ground fauna, this habit appears to be re- stricted to species of Leptogenys and Camponotus. As expected, we found only one of the 12 species 12 Journal of Hymenoptera Research of Leptogem/s previously recorded from Ghana (Bolton 1975), plus one undescribed species. The genus Camponotus is poorly known taxonomically, but there are more than 15 species from Ghana in the collection of the Natural History Museum which we did not find. Non-arboreal members of this genus, although nesting in the soil, are fast- moving surface foragers, which may also have contributed to their poor representation in our survey. The three largest myrmicine tribes are suffi- ciently well known taxonomically for us to com- pare the list of species recorded in our survey with that of species already known from Ghana. In the Dacetonini, 25 species were previously known (Bolton 1983), including two known to be arbo- real. We found 18 of the remaining 23 species, plus 14 previously unrecorded in Ghana. In the Solenopsidrni, 15 species were known (Bolton 1987), including three savannah species. We found 10 of the remaining 12 species, plus six previously unrecorded . These results show that we have found many small cryptic species which had previously been overlooked. The world-wide household pest Monomorium pharaonis (L.) (Pharaoh's Ant) was recorded from nine sites (including primary for- est). Its range is clearly not restricted to disturbed habitats in Ghana. In contrast, we found a smaller proportion of the Tetramoriini species known from Ghana. Pre- viously 53 species were known (Bolton 1976, 1980), including 10 savannah and three arboreal species. We found only 21 of the remaining 40 species, plus six previously unrecorded in Ghana. Members of this tribe, however, often nest in exposed sites or in rotten logs, which may explain its relatively poor representation in our survey. We cannot compare our results for the Pheidologetonini, Pheidolini or Crematogastrini as no synthesising taxonomic studies have been undertaken. The remaining myrmicine tribes con- tain a total of 40 species previously recorded from Ghana (Bolton 1981a, 1981b, 1982) including 20 arboreal (chiefly Cataidacus), four savannah, and two species which tunnel in living wood (Melissotarsus). We found 8 of the remaining 14 species: one of the three Cardiocondyla, one of the three Leptothorax, two of the four Calyptomyrmex, both Pristomyrmex, the sole Meranoplus and the sole Baracidris. In most cases nothing is known of the ecology of the species not found in our survey except for earlier collection localities and dates. However, their absence from our survey shows that they either do not forage in the leaf litter or are rare. The soil samples. — The density of individuals was eight times higher in the soil samples than in the leaf litter (mean of the three soil samples = 982 m 2 (S.E.= 62). Lynch and Johnson (1988) also found that in a temperate forest the density of ants was higher in the soil than the leaf litter. These data may, however, be misleading as in both studies the soil samples probably included species which, although nesting in the top few cm of the soil, forage primarily in the leaf litter. Although similar in overall species composi- tion to the leaf litter samples (also found by Lynch and Johnson 1988), the three soil samples con- tained a small distinctive subterranean element. We extracted 30 workers of Leptanilla boltoni Baroni Urbani from a single soil quadrat in a small patch of primary forest at Bunso. The sole Afrotropical representative of the genus, it is known previously only from 6 specimens extracted by Berlese funnel from leaf litter in a cocoa plantation at Mampong, Ghana (Baroni Urbani 1977). The genus contains a total of 33 described species, all of which appear to be very rare (Holldobler and Wilson 1990: 590). The biology of only one species (L. japonica Baroni Urbani) is known: it is strictly subterranean and appears to have true legionary behaviour, em- ploying both group predation (of geophilomorph centipedes) and colony migration (Masuko, 1990). Of the 17 individuals of Acropyga sp., 10 were found in soil samples. This species resembles oth- ers in the genus which tend coccids in subterra- nean nests for their sugary secretions (Holldobler and Wilson 1990: 527). It is the only species found in our survey which is known not to be primarily predacious, although the habit probably occurs in species of several other formicine genera (Pseudolasius, Paratrechiiia and Lepisiota). One of the species of the undescribed Ponerini genus also appeared to be soil-dwelling. The results of our few soil samples point tantalisingly to a rich and poorly known subterra- nean fauna that would repay further collecting. Volume 3, 1994 13 Table 2. Species recorded during the survey (see table 1 for details of sites, and figure 1 for their location). "Widespread" denotes the species was recorded from more than five differently-lettered sites (the number is given in parentheses). An asterisk (*) denotes species classed as tourists (see text). Species Total number () = from soil samples Sites Species Total number () = from soil samples Sites AENICTINAE OECOPHYLLINI Aenictus sp. (indet) (1) 1 si "Oecophylla longinoda 15 d,i3,j3,p A. sp. (indet) (2) 1 o3 PLAGIOLEPIDINI APOMYRMINAE Acropyga sp. (indet.) 17(10) ai.h.ql Apomyrma stygia 1 m Lepisiota sp. (indet.) (1) 4 f,sl L. sp. (indet.) (2) 2 o3 L. sp. (indet.) (3) 1 o4 CERAPACHYINAE L. sp. (indet.) (4) 10 f,q5 Cerapachvs foreh 55 widespread (1 1) L. sp. (indet.) (5) 1 qs C. nitidulus 299(134) widespread (12) L. sp. (indet.) (6) 1 q5 C. n.sp. 3 m,o2,sl Plagiolepis sp. (indet.) LEPTANILLINAE 19 d DOLICHODERINAE Leptanilla boltoni 30(30) qi Tapinoma sp. (indet.) 2 r Technomyrmex sp. (indet.) (1) 821(2) widespread (14) MYRMICINAE T. sp. (indet.) (2) 163(8) widespread (11) CATAULACINI T. sp (indet.) (3) 3 si 'Cataulacus guineensis 1 d T. sp (indet.) (4) 3 d,p T. sp. (indet.) (5) 530 widespread (10) CREMATOGASTR1NI *Crematogaster depressa 499 widespread (6) *C. bequaerti 9 o2,p,q2 DORYLINAE *C. clariventris 24 p,q4 Dorvlus sp. (indet.) 1477(1468) 12,ql C. stnatula *C. sp. (indet. ) (1) 2210 1 widespread (13) d C. sp. (indet.) (2) 1 s2 FORMICINAE C. sp. (indet.) (3) 1 o3 CAMPONOT1NI C. sp. (indet.) (4) 9 o3 "Camponotus acvapimensis 2 q4-5 *C. sp. (indet.) (5) 103 i3 *C. flavomarginatus 15 q4-5,s2 *C. maculatus 2 qs-t DACETON1NI *C. vividus 2 qs Epitntus laticeps 27 k,ll-3,m,q4 "Polyrhachis decemdentata 2 q4,r E. roomi 53(16) widespread (12) *P. militans 24 p,q2 E. tiglath 9 d,f,k,Il-2 *P. rufipalpis 2 a,o4 Glamyromyrmex crvpturus 57(5) f,il,o3,q2-3,sl *P. weissi 1 r G. sistrurus 110 widespread (6) G tetragnathus 31 widespread (6) LASIINI G tukultus 20 e.ql Paratrechina sp. (indet.) (1) 794 widespread (12) Microdaceton tibialis 53 widespread (7) P. sp. (indet.) (2) 48 d,i3,jl,m,q4-5 Quadnstruma emmae 1 n P. sp. (indet.) (3) 148(48) h,i2-3,q2,t Serrastruma concolor 238 widespread (8) P. sp. (indet.) (4) 1 P S. ludovici 20 g,12,o2,t Pseudolasius sp. (indet.) 270(4) t,il,o3,ql-3 S lujae 802 j3,m,ol-4 S. serrula 823(4) widespread (16) Smithistruma cavinasis 3 11 S. enkara 7 a,q2 14 Journal of Hymenoptera Research Species Total number () = from soil samples Sites Species Total number () = from soil samples Sites S. fulda 2 a P. sp. (indet.) (3) 180(74) d,g,i2,p,q2 S. hensekta 58 widespread (9) P. sp. (indet.) (4) 67 d,q5,s2 S. malaplax 69 sl-2 P. sp. (indet.) (5) 213(1) widespread (9) S. marginata 1 d P. sp. (indet.) (6) 10 a,o4,p,t S. minkara 55 f,12,o3,q3 P. sp. (indet.) (7) 98 o4 S. ninda 55 f,h,n,ql+5 P. sp. (indet.) (8) 8 ol,h S. sharra 25 d,f,ll,o2-3 P. sp. (indet.) (9) 496 widespread (7) S. facta 56 widespread (8) P. sp .(indet.) (10) _1 e,i3,q4 S. tignlla 9 d,ll,o2 P. sp (indet.) (11) 30 a S. n.sp. 4 h,q5 Strumigenys korahvla 1 d SOLENOPSIDINI S- nimbrata 70(4) widespread (7) Monomonum bicolor 1 d S. petiolata 1439(31) widespread (20) M. cryptobium 1475(242) widespread (16) S. rogen 177 widespread (6) M. draxocum 5 d,g S. rufobrunea 1059 widespread (15) M. exiguum 605 widespread (16) S. spathoda KD ^ M. floncola 13 13 S. vazerka 22(1) a,13,o2,ql+3,s2 M. egens 8 widespread (7) M. gabnelense 42 a,b,g,o2-3,t FORMICOXENINI M guineense 55 g.'2 Cardiocondyla neferka 1 d M. invidium 3432(62) widespread (19) Leptothorax angulatus 1 qs M. pharaonis 17 widespread (6) "Terataner piceus 1 n M. rosae 7 f,q4-5 M. tanysum 2 d MERANOPLINI M. trake 10 d,i2 Meranoplus inermis 1 si M. n.sp. (1) 22 d M. n.sp. (2) (exiguum group) 13 p,q5 MYRMECININI M n.sp. (3) 75 i2,jl+3,p Pristomyrmex afncanus 79(3) widespread (9) P. orbiceps 505(3) widespread (13) STENAMMINI Baracidns meketra 8(3) j3,12,m,n,ql-2 PHEIDOLOGETONINI Calyptomyrmex kaurus 18 k,ll-2,o3 Afroxyidris cngensis 2 e,f C. nummuliticus 1 1 12,o3,p,q3 Oligomyrmex sp. (indet.) (1) 440(7) widespread (10) O. sp. (indet.) (2) 1957(333) widespread (15) TETRAMORIINI O. sp- (indet.) (3) 1532(175) widespread (12) Decamonum decern 119 widespread (7) O. sp. (indet.) (4) 123(10) e.j3,13,q2+5,s2,t "Tetramonum aculeatum 3 ol+4,i3 O. sp. (indet.) (5) 245 widespread (6) T. amentete 106 widespread (8) O. sp. (indet.) (6) 967(81) widespread (13) T. antrema 162 widespread (10) O. sp. (indet.) (7) 98(92) ql-2+5 T ataxium 140(4) widespread (11) O. sp. (indet ) (S) 602 widespread (6) T brevispinosum 118 widespread (6) O. sp. (indet.) (9) 156(34) ql,o3 T. camerunense 2 ol O. sp. (indet.) (10) 197 g,jl+3,m T. distincrum 552(16) widespread (16) O. sp. (indet.) (11) 39(27) h T flavithorax 350 widespread (17) O. sp. (indet.) (12) 104 ll-3,q3 T. furtivum 689(33) widespread (8) Paedalgus distinctus 228(22) widespread (13) I guineense 417 widespread (10) P. saritus 42 |3,n,o2 T ictidum 1 o2 T. invictum 65 widespread (6) PHEIDOLINI T. |ugatum 4 d,12,t Pheidolesp. (indet.) (1) 778 widespread (10) T. lucayanum 3 e,13 P. sp. (indet ) (2) 5474(94) widespread (20) T. menkaura 21 d,o2-4,sl,t Volume 3, 1994 15 Species Total number () = from soil samples Sites T. minimum 368(25) widespread (15) T. muralti 189 widespread (10) T. muscorum 18 widespread (6) T. peutli 28 widespread (6) T. quadndentatum 1 q4 T. rhetidum 971(6) widespread (9) T. youngi 7 o2-3 T. zambesium 2293(13) widespread (17) T. zapyrum 33 d,r,t T. n.sp. (weitzeckeri group; 30 H,o4,q2 T. n.sp. (convexum grou P) 5 i3,m T. n.sp. (dumezi gro up) 2 q3 Species Total number () = from soil samples Sites PONERINAE AMBLYOPONINI Amblyopone mutica 3 i2,t A. santschii 6 f,g,jl+3,ql ECTATOMMINI Discothyrea oculata 1 q2 D. mixta 3 f,ll,o4 D. n.sp (1) 5(1) f,ll,o4,ql D. n.sp (2) 3 a PLATYTHYREINI •Platythyrea conradti 1 r Probolomyrmex guineensis 46 f,k,ll-3,sl PONERINI Anochetus afncanus 170 g,ll-3,ol+3-4,p,ql-3 A. bequaerti 43(2) widespread (9) A. katonae 568(70) widespread (17) A. maynei 1 si A. siphneus 6 f,h,]3,t A. n.sp. 4 o3-4 Asphinctopone silvestni 5(1) f,ql-3 Cryptopone n.sp. 2 jl Hypoponera sp. (indet.) (1) 244(29) widespread (13) H. sp. (indet.) (2) 1828(35) widespread (19) H. sp. (indet.) (3) 11 d,g,k H. sp. (indet.) (4) 98(16) widespread (6) H. sp. (indet.) (5) 223(2) widespread (9) H. sp. (indet.) (6) 6(1) i3,ol,q2 Leptogenys nuserra 15(1) 13,m,ol-2,q2,t L. n.sp. 1 o2 Odontomachus assiniensis 8 si O. troglodytes 22 widespread (7) Pachycondyla ambigua 24 a,d,f,12,ol-2+4 P. brunoi 24(9) widespread (8) P. caffraria 86 widespread (8) P. fugax 12 h,il,12,o3,t P. pachyderma 16 widespread (7) P. soror 2 ]3,q2 P. sp. (indet.) 18(7) p,ql+3,sl P. n.sp. 1 i3 Paltothyreus tarsatus 25 widespread (7) Phrynoponera bequaerti 13 11+3 P. gabonensis 24(1) widespread (6) Plectroctena macgeei 3 jl,o3 n. gen., n.sp. (1) 3(3) q2,h n. gen., n.sp. (2) 1 13 ACKNOWLEDGMENTS This project was funded by the Leverhulme Trust (Grant F.696A). The field work was based at the Cocoa Research Institute of Ghana, New Tafo, and we are very grateful for their hospitality and assistance. We would also like to thank the Ghanaian Forestry Department; Cocoa Services Division; Crops Research Institute; the Ghana Association for the Con- servation of Nature; and the Tafohene, Old Tafo. In addition, Paul Eggleton and two reviewers made valuable comments on the manuscript. LITERATURE CITED Andersen, A. N. and J. D. Ma]er. 1991. The structure and biogeography of rainforest ant communities in the Kimberley region of north-western Australia, pp. 333- 346. In McKenzie N.L..R.B. Johnston and P.G. Kendrick, eds. Kimberley Rainforests of Australia. Surrey Beatty and Sons, Chipping Norton, Australia. 490 pp. Baroni Urbani, C. 1977. Materiah per una revisione della sottofamiglia Leptanillinae Emery (Hymenoptera: Formicidae). Entomologica Brasiliensia 2: 427-488. Belshaw, R. and B. Bolton. 1993. The effect of forest distur- bance on the leaf litter ant fauna in Ghana. Biodiversity and Conservation 2: 656-666. Belshaw, R. and B. Bolton. 1994. A new myrmicine ant genus from cocoa leaf litter in Ghana (Hymenoptera: Formicidae). Journal of Natural History 28: 631-634. Besuchet, C, D. H. Burckhardt and I. Lobl. 1987. The "Winkler/Moczarski" eclector as an efficient extractor for fungus and litter Coleoptera. The Coleopterists Bulle- tin 41: 392-394. Bolton, B. 1974. A revision of the ponenne ant genus Plectroctena F. Smith (Hymenoptera: Formicidae). Bulle- tin of the British Museum (Natural History) (Entomology) 30: 309-338. Bolton, B. 1975. A revision of the ant genus Leptogenys Roger (Hymenoptera: Formicidae) in the Ethiopian region: with a review of the Malagasy species. Bulletin of the British Museum (Natural History) (Entomology) 31: 235- 305. L6 Journal of Hymenoptera Research Bolton, B. 1976. The ant tribe Tetramornni (Hymenoptera: Formicidae): constituent genera, review of smaller gen- era and revision of Triglyphothrix Forel. Bulletin of the British Museum (Natural History) (Entomology) 34: 283- 379. Bolton, B. 1980. The ant tribe Tetramornni (Hymenoptera: Formicidae): the genus Tetramorium Mayr in the Ethio- pian zoogeographical region. Bulletin of the British Mu- seum (Natural History) (Entomology) 40: 193-384 Bolton, B. 1981a. A revision of the ant genera Meranoplus F. Smith, Dieroaspts Emery and Calyptomyrmex Emery (Hy- menoptera: Formicidae) in the Ethiopian zoogeographi- cal region. Bulletin of the British Museum (Natural His- tory) (Entomology) 42: 43-81. Bolton, B. 1981b. A revision of six minor genera of My rmicmae (Hymenoptera: Formicidae) in the Ethiopian region. Bulletin of the British Museum (Natural History) (Entomol- ogy) 43: 245-307. Bolton, B. 1982. Afrotropical species of the myrmicine ant genera Cardiocondyla, Leptothorax, Melissotarsus, Messor and Cataulaeus (Formicidae). Bulletin of the British Mu- seum (Natural History) (Entomology) 45: 307-370. Bolton, B. 1983. The Afrotropical dacetine ants (Formicidae) Bulletin of the British Museum (Natural History) (Entomol- ogy) 46: 267-416. Bolton, B. 1987. A review of the Solenopsis genus-group and revision of Afrotropical Monornorium Mayr (Hy- menoptera: Formicidae). Bulletin of the British Museum (Natural History) (Entomology) 54: 263-452. Bolton, B. and R. Belshaw. 1993. Taxonomy and biology of the supposedly lestobiotic ant genus Paedalgus (Hym.: Formicidae). Systematic Entomology 18: 181-189. Brown, W. L., W. H. Gotwald Jr. and J. Levieux. 1970. A new genus of ponerine ants from West Africa (Hvmenoptera: Formicidae) with ecological notes. Psyche 77: 259-275. Gowing, G. and H. F. Recher. 1984. Length-weight relation- ships for invertebrates from forests in south-eastern New South Wales. Australian Journal of Ecology 9: 5-8 Hall, J. B. and M. D. Swaine. 1976. Classification and ecologv of closed-canopy forest in Ghana. Journal of Ecology h4 913-951. Hall, J. B. and M. D. Swaine. 1981. Distribution and ecology of vascular plants in a tropical rain forest: forest vegeta- tion in Ghana. Geobotany 1: 1-383. Halle, H and O. Pacal, eds. 1992. Biologie d'une canopee deforet equatoriale: volume 2. Unpublished report. Holldobler, B. and E.O. Wilson 1990 The ants Springer- Verlag, Berlin. 732 pp. Levieux, J. 1982. A comparison of the ground dwelling ant populations between a guinea savanna and an ever- green rain forest of the Ivory Coast, pp. 48-53. In Breed, M.D., CD. Michener, and HE. Evans, eds. The biology of social insects. Westview Press, Boulder, Colorado. 419 PP- Levings, S. C. 1983. Seasonal, annual, and among-site varia- tion in the ground ant community of a deciduous tropi- cal forest: some causes of patchy species distributions. Ecological Monographs 53: 435-455. Lynch, J. F. and A. K. Johnson. 1988. Spatial and temporal variation in the abundance and diversity of ants (Hy- menoptera: Formicidae) in the soil and litter layers of a Maryland forest. The American Midland Naturalist 119: 31-44. Majer, J. D. 1976. The influence of ant and ant manipulation on the cocoa farm fauna. Journal of Applied Ecology 13: 157-175. Masuko, K. 1990, Behavior and ecology of the enigmatic ant Leptamlla japonica Baroni Urbani. hisectes Socwux 37: 31- 57. Olson, D. M. 1991. A comparison of the efficacy of litter sifting and pitfall traps tor sampling leaf litter ants (Hvmenoptera, Formicidae) in a tropical wet forest, Costa Rica. Biotropica 23: 166-172. Olson, D. M. and P. S. Ward, in press. The ant fauna (Hy- menoptera: Formicidae) of Kinndy Forest (tropical dry forest) in Western Madagascar. In Ganzhorn, J., ed. The ecology and economy of a tropical dry forest. Springer Verlag, Berlin Palmer, M. W. 1990. The estimation of species richness by extrapolation. Ecology 71: 1195-1198. Palmer, M. W. 1991. Estimating species richness: the second order jacknife reconsidered. Ecology 72: 1512-1513. Wilkinson, L. 1990. SYSTAT: the system for statistics- Evanston, IL: SYSTAT, Inc. Wilson, E. O. 1959. Some ecological characteristics of ants in New Guinea rain forests. Ecology 40: 437-447. Wilson, E. O. 1971. The insect societies Belknap Press, Cam- bridge, Massachusetts. Wolda, H. 1981 . Similarity indices, sample size and diversity. Oecologia 50: 296-302. J. HYM. RES. Vol. 3, 1994, pp. 17-89 Revision of the Ant Genus Rogeria (Hymenoptera: Formicidae) with Descriptions of the Sting Apparatus Charles Kugler Biology Department, Radford University, Radford, VA 24142 USA Abstract. — This is the first complete revision of the century-old ant genus Rogeria. The revision recognizes 39 species, of which 19 are new species: alzatei n. sp. (Panama, South America, Dominican Republic), besucheti n. sp. (Paraguay, Colombia, Peru), carinata n. sp. (Caribbean), ciliosa n. sp. (Venezuela, Ecuador), cornuta n. sp. (Belize, southern Mexico), cuneola n. sp. (Mexico, Central America), gibba n. sp. (Colombia, Ecuador), innotabilis n. sp. (Mexico, Central America), leptonana n. sp. (Mexico, Central America, Dominican Republic) hrata n. sp. (Trinidad, South America), megastigmatica n. sp. (Solomon Islands), merenbergiana n. sp. (Colombia, Ecuador), neilyensis n sp. (Costa Rica), nevadensis n. sp. (Colombia), prommula n. sp. (Brazil), scobinata n. sp. (Trinidad, South America), terescandens n. sp. (Costa Rica), tribrocca n. sp. (Colombia), and unguispina n. sp. (Venezuela). Previously recognized species are belli Mann 1922, blanda Fr. Smith 1858, bruclu Santschi 1922, brunnea Santschi 1930, creightoni Snelling 1973, curvipubens Emery 1894, exsulans Wilson and Tavlor 1967, foreli Emery 1894, germaim Emery 1894, inermis Mann 1922, lacertosa Kempf 1963, micromma Kempf 1961, minima Kusnezov 1958, pellecta Kempf 1963, procera Emery 1896, scandens Mann 1922, sicaria Kempf 1962b, stigmatica Emery 1897, subarmata Kempf 1962a, and tonduzi Forel 1899. Nine new synonyms are proposed: caraiba Santschi 1936, cubensis Santschi 1936, habanica Santschi 1936, and scabra Weber 1934 = brunnea Santschi 1930; gaigei Forel 1914, and huachucana Snelling 1973 = foreh Emery 1894, minensis Santschi 1923 = germaini Emery 1894; sublei'inodis Emery 1914, and manni Santschi 1922 = stigmatica Emery 1897. Seven new combinations are proposed: epmotalis Mann 1919, rugosa Mann 1921, striatella Mann 1921, tortuosa Mann 1921, tortuosa levifrons Mann 1921, tortuosa pohta Mann 1921, and tortuosa stonert Mann 1925 to Lordomyrma Three species occur in Polynesia and Melanesia; the rest are New World: southwestern United States to southern Paraguay A key to the workers is provided All species are briefly described and illustrated to show variation. Available biological information is summarized. Six informal species groups are defined for 29 species; possible affiliations of the remaining 10 species are indicated. A new character system, the sting apparatus, is used to help define the species and species-groups. Pilositv also provides an important new set of characters for species identification in this genus. INTRODUCTION AND DISCUSSION This revision will redefine the genus, establish new synonymies, create informal species groups, redescribe species, provide a key to the workers, and summarize ecological information. Keys to the reproducti ves are not included, because queens are known or proposed for less than half of the Rogeria species and males for only four species. The genus Rogeria was erected by Emery ( 1 894 ) for a few neotropical ants then in Myrmica and Tetramorium. He then extended the distribution of the genus to New Guinea with the addition of stigmatica (Emery 1897). Discovery of other neotropical species led Emery (1915) to create the subgenus Irogera for those with a clavate petiole. This subdivision was attractive, and for a brief period Irogera even enjoyed generic status (Brown 1953), but with a better understanding of the varia- tion in petiole shape within species, the splitting of Rogeria along these lines became untenable ( Kempf 1965). Meanwhile, Melanesian and Polynesian species were being added to the genus with no discussion of the disjunct distribution of the genus or comparisons of New World and Old World species (Mann 1919, 1921, 1925; Santschi 1922, 194 1 ). Brown (1953) recognized that many of these were actually Lordomyrma species, but still, the Pacific species including stigmatica, epinotalis, and later, exsulans remained in Rogeria. Wilson and Taylor (1967) called this "...one of the most anoma- lous discontinuous distributions found in ants." 18 Journal of Hymenoptera Research The disjunct distribution of Rogeria was ques- tioned by Kusnezov (1958) on the basis of palpal formula differences between stigmatica and a South American species, and by Kugler (1978b) on the basis of the sting apparatus. In contrast to all other myrmicine genera, there were two distinctly dif- ferent sting apparatuses within the genus, and that difference coincided with the disjunct distri- bution of the genus. The sting of inermis from Central America was strong, acute and with an unusual low dorsal flange; the sting of stigmatica from the Pacific had a weak, narrowly spatulate sting shaft and no dorsal flange (among other differences). I believed a revision of Rogeria using sting apparatus characters would separate the Pacific and New World species into distinct gen- era. The potential of the sting apparatus for defin- ing ant genera and inferring phylogeny has been demonstrated by a number of comparative mor- phological studies (Kugler 1978b, 1980, 1986, 1992). In addition, Bolton (1973, 1982, 1987) has often used externally visible parts of the sting to help define some myrmicine taxa. But this is the first time that characters from the whole apparatus have been used as an integral part of a taxonomic revision. This work, then, also tests the usefulness of the sting apparatus in the alpha-taxonomy of ants. I examined sting apparatuses of 79 workers and queens in 27 Rogeria species and seven Lordomyrma species. In the five species in which both queens and workers were dissected, there were only the usual individual differences, such as numbers of sensilla. Contrary to expectations, the Western Samoan species exsulans is unmistakably related to Central American Rogeria species in sting apparatus, pygidium, and external anatomy. Moreover, some South American species (ciliosa, gibba, besucheti) had sting apparatuses and pygidia with the same distinctive features of stigmatica. It seems the distribution of Rogeria really is disjunct. Sting and pygidial characters provided sup- port for numerous other taxonomic decisions as well. For example, they provided strong support for creating the stigmatica-group of species, and could conceivably be used to distinguish that group as a separate genus. Sting apparatus anatomy also supported Brown's (1953) belief that levifrons, striatella, and tortuosa are Lordomyrma species. Autapomorphic shape of the sting shaft confirmed that germaini and lirata are sister species. The un- usual shape of the spiracular plate ultimately con- vinced me to propose innotabilis as a new species after long deliberation over the welter of variation within the creightoni-gioup. This character system must be used with caution however. The apparatus is prone to reduction and convergence of form. The sting of Lordomyrma epinotalis is more like those of the Rogeria stigmatica- group than those of Lordomyrma. In this case, I based my decision to transfer epinotalis to Lordomyrma on external characters, since the com- mon sting features could well be due to reduction convergence. Hair has not been used previously in Rogeria systematics, but in this revision I came to rely more and more on characteristics of pilosity for distin- guishing species. Consequently it has become nec- essary for species descriptions to identify pre- cisely the several kinds of hairs and their distribu- tions on the ants, sometimes including the num- bers of pairs of erect hairs on the head, mesosoma, and nodes. This means, of course, that care must be taken to not disturb the hair while cleaning and mounting the ants. I have attempted to take a fairly conservative approach to synonymy and description of new species, choosing in ambiguous situations to err on the side of not changing the number of species in the genus. Thus, for lack of sufficient evidence, some species may actually contain several sibling species: alzatei, belli, creightoni, foreli, leptonana, and scandens. On the other hand, I have retained some names that may fall when collections improve: brnchi,micromma. Treatment of the creightoni-group has perhaps been somewhat less conservative (See the Species Group section). All told, this study recognizes 39 species (in- cluding 19 new species), establishes nine new synonymies and transfers seven species to Lordomyrma. It fails to resurrect the subgenus Irogera. It retains three Polynesian /Melanesian species (stigmatica, megastigmatica, exsulans) with the Neotropical bulk of the genus. As presently constituted, members of the ge- nus Rogeria are distributed from Buenos Aires to southern Texas and Arizona, and in the Pacific between ION and 25°S from Tahiti to the western end of the island of New Guinea. So far it is unknown in Australia or southeastern United States. Table 1 shows the distribution of species in Volume 3, 1994 19 Table 1. Geographic distribution of Rogena species. The North American region is from southwestern United States through Panama. The northern region of South America extends from the north and west coasts through the Amazonian basin. The southern region of South America includes Paraguay, ad)acent Brazil, and Argentina south to Buenos Aires. Trinidad is the only known Caribbean locality of three species (blandajirata, scobinata). A question mark indicates that presumed queens, but no workers, were collected in that region. See text for further discussion. Speices North Caribbean America South America Pacific Northern Southern alzatai + belti + besucheti blanda + bruchi brunnea carinata ciliosa + cornuta creightoni + cuneola + cunnpubens + exsulans foreli + germaini gibba uierrms + mnotabilis + lacertosa leptonana + lirata metastigmatica merenbergiana micromma minima neilyensis + nevadensis pellecta procera prominula scandens + scobinata + sicarta stigtnatica subarmata terescandens + tonduzi + tnbrocca unguispina +? + + + 20 Journal of Hymenoptera Research more detail. In the North American region, most species occur below the Isthmus of Tehuantepec, but two species (creightoni, cuneola ) extend north- ward through the eastern lowlands of Mexico. Only foreli and creightoni have been collected in the United States. The North American region con- tains 8-10 endemic species (depending on the un- certain distributions of innotabilis and leptonana); the Caribbean, two endemic species; South America (including Trinidad), 19 endemic species and the Pacific, three endemic species. Only 5-7 species (depending on innotabilis and leptonana) are found in both Central America and northern South America. Little is known of the biology of these cryptic ants. Collection records usually range from sea level to 1000m, but five species extend higher and two (ungiiispina and merenbergiana) can be found at 2000m. Rogeria species are generally collected in moist forests (primary or secondary forests, coffee or cacao plantations), but at higher elevations can be found in pastures {leptonana, merenbergiana). Several species (creightoni, cuneola, foreli) have been found in both moist and dry climates. Rogeria foreli is the most unusual, with some members dwelling at over 1800m in the temperate mountains of south- ern Arizona. Most Rogeria species have only been collected as strays or by Berlese or Winkler sampling, usu- ally in leaf litter and rotten wood, but occasionally among epiphytes and moss (belti, creightoni, exsulans). Nests of several species (belti, blanda, merenbergiana) have been found under loose bark of rotten logs. Nests of blanda and tondnzi have been taken from the trunks of cacao trees. A nest of leptonana was found at 1750m under a rock in a pasture. Because nests are so rarely found, males are known for only four species (belti, blanda, leptonana and stigmatica), and queens associated through nest series for only nine species (See Synonymic List of Species). METHODS AND TERMINOLOGY Dissections Mouthparts, sting apparatus, pygidium, hy popy gium, and male genitalia preparations were obtained by rehydrating ants in 70% ethanol, dis- secting these structures from the ants, clearing in 55-60°C lactophenol for 5 minutes (or longer if necessary), rinsing twice in 70% ethanol, and rins- ing twice in 95% ethanol. After clearing, sting apparatuses and genitalia were usually cut in half, except for the aedeagus and sting, which were separated from the other sclerites. Stings, pygidia, and hypopvgia were mounted in glycerin jelly for ease of repositioning. Mouthparts, genitalia, and other sting apparatus sclerites were soaked in xylene, then mounted in Canada balsam. Occa- sionally, whole ants were similarly cleared and mounted in balsam. Preparations were examined with a Zeiss KF-2 phase contrast microscope. The "Materials Examined" sections of each species description identifies which specimens were dissected. Following the locality of the dis- sected ant, brackets identify the structures slide- mounted ("sting" means sting apparatus, py- gidium and hypopygium). Dissected specimens are workers unless otherwise identified. Pinned vouchers are identified with the label "Kugler 1991 Dissection Voucher." All are depos- ited in the collection of the MCZ, except for vouch- ers of lacertosa, pellecta, subarmata, the Agudos, Brazil voucher of alzatei, and the Surinam vouch- ers of curvipubens, which are all in the MZSP. Illustrations Drawings were made using a grid eyepieces and grid-backed tracing paper. Drawings of sting apparatus preparations were made at 400X with estimated accuracy of <0. 001mm. Scanning elec- tron micrographs were taken on an AMR-900 and a JEOL JSM-35C. Dorsal views of heads are in the same full-face view used to measure HL. Lateral views of mesosomas are in the same view used to measure WL, SpL, and MHI. Drawings attempt to show all hairs on the left side of the body that project well above the body outline. Care was taken to present typical pilosity, rather than matted or damaged hairs. Hair is not included in line drawings of: 1) dorsal views of heads, 2) ventral portions of head profiles, 3) most dorsal views of waists, and 3) queens and males. All line drawings of sting apparatuses are from workers. Male genitalia drawings show the lateral view of the aedeagus and medial view of the right paramere and volsella (with associated part of the Volume 3, 1994 21 phallobase). Nontype specimens are identified in figure legends by affixing their collection localities in parentheses. Measurements and Indices Measurements and drawings of whole ants, except for the very largest, were made using a Zeiss SR stereomicroscope at 125X magnification and a fiber optic ring lamp. Estimated accuracy: ± 0.01mm. All specimens were measured for Weber's Length (WL) then at least the largest and smallest from each locality were measured completely. The maximum and minimum of each measurement and index were double-checked, as were all mea- surements of holotypes. The number of specimens (N) that follows the list of metric characters at the beginning of a species description is the number of ants that were measured completely. The numbers at the ends of "Material Examined" sections are total specimens studied. In listing metric characters for new species, the ranges found in all type material are followed by the holotype measurements in parentheses. If no queens or males are described for a species, none were available for study. CL Cephalic index: HW/HL EL Eye length: Maximum diameter of compound eye viewed straight on. FLW Frontal lobe width: In full face dorsal view, maxi- mum transverse distance across frontal lobes. GL Gaster length: In lateral view, from anterior edge of first tergum to posteriormost point usually T2 or T3, but sometimes to end of pvgidium) GW Gasterwidth: Maximum transversedistanceacross gaster. HL Head length: Maximum longitudinal distance from the antenormost portion of the clypeus (usually the clypeal apron, but sometimes the more promi- nent body of the clypeus) to the midpoint of a line across the back of the head. HW Head width: Maximum width in same view as for HL, excluding eves. For males, HW includes com- pound eves. ML Mandible length: In same view as for HL, from antenormost portion of head to apex of closed mandibles. OI Ocular index: EL/HW. PetL Petiole length: In lateral view and with petiole not strongly flexed up or down, the axial distance from the dorsal corner of the posterior peduncle to the nearest edge of the metapleural lobe. PetW Petiole width: Maximum transversedistanceacross the node. PpetL Postpetiole length: In lateral view, the axial dis- tance from base of node in front to tip of posterior peduncle. PpetW Postpetiole width: Maximum transverse distance across the postpetiole. PSI Petiolar spine index: SpL/WL. PW Pronotal width: Maximum width across pronotum in dorsal view. SI Scape index: SL/HW. SL Scape length: Maximum length excluding basal condyle and neck. SpL Propodeal spine length (Fig. 1): In same view as for WL, from tip of propodeal spine to nearest edge of propodeal spiracle pentreme. MH Mesosoma height (Fig 1): In perfect lateral view, the vertical distance from a line tangent to the ventralmost points of the of meso- and metasterna to the highest point on the mesosoma profile (for a broadlv arching pronotum, estimated as the mid- point of the arc). MHI Mesosoma height index (Fig. 1): Mesosoma height (MH) divided by the horizontal distance from the base of the MH line to the posteroventral corner of the propodeum. TL Total length: Sum of ML + HL + WL + PetL + PpetL + GL to the nearest 0.1mm. WL Weber's length (Fig. 1): In perfect lateral view of the mesosoma, diagonally from posteroventral corner of mesosoma to farthest point on anterior face of pronotum, excluding the neck. Additional Terminology Terms such as vertex, occiput, malar and genal areas are of dubious accuracy when applied to ants. Consequently, there is disagreement over how to name the regions of the head (Trager 1989; Snelling 1989; Wheeler 1989). Moreover, these terms are confusing jargon that make the use of keys and descriptions difficult for the uninitiated. Instead, I prefer to rely, as much as possible, on the generally understood directional terms, anterior, lateral, etc., assuming that the mandibles are ante- rior. Five regions of the head usually contain dis- tinctive sculpture in Rogeria (Fig. 1). 1) The middorsum of the head is the median portion of the dorsal surface between the maximum retrac- tion of the scapes and extending from the frontal area to a change in sculpture near the rear of the head. 2) The posterior (region of the) head in- cludes the actual posterior surface of the head and the posterodorsum, the dorsal surface between the middorsum and the posterior outline of the head. The "posterior head" seems equivalent to 22 Journal of Hymenoptera Research the "vertex" of some authors who consider the mandibles to be ventral and the back of the head to be the top, rather than posterior. 3) The laterodorsa are the portions of the head seen in dorsal view that lie beneath the sweeps of the scapes. 4) The venter is the ventral surface of the head not seen in lateral view. 5) The sides of the head lie between the venter and the laterodorsa (between the lateral outline of the head in dorsal view and the ventral outline of the head in lateral view). In triangular mandibles the basal angle is distinct and nearly a right angle (e.g. Figs. 43, 49, 63); in subtriangular mandibles the basal angle is reduced and more obtuse (e.g. Figs. 14, 58, 61). Sometimes tiny denticles are intercalated between or replace teeth (e.g. Figs. 35, 55); these are not included in tooth counts. The palpal formula is the number of maxillary palpomeres followed by number of labial palpomeres. Palpal formula is usually determined by dissection, thus the sample size is very small and may not show the variation within the species. The body of the clypeus is the medial portion of the clypeus anterior to the fron- tal lobes and dorsal to the clypeal apron, which is the anterior rim just above the mandibles (Fig. 1). Clypeal apron shape is described as seen from dorsal view. In describing eye shape, elliptical or ellipsoid means similarly rounded at both ends, whereas oval or ovoid means the anterior end is more narrowly rounded. The shape of the poste- rior outline of the head is as seen in the full dorsal view used for measuring HL since shape some- times varies with the pitch of the head. The nuchal grooves are longitudinal impressions in the posteroventral corners of the head, into which the lower corners of the pronotum fit when the head is retracted. A rough index of the degree of inclination of the propodeal spines is expressed by imagining a bisecting line that divides the lateral aspect of the spine in half and extends forward over the mesosoma side, and noting where it intersects the edge of the mesopleuron or pronotum (Fig. 1). It is important that the mesosoma be in perfect lateral view. The diameter of a propodeal spiracle in- cludes the peritreme and is as seen in full lateral view of the mesosoma. The spiracle's position is measured from the outer edge of the peritreme to the nearest edge of the propodeum, which is the free edge of the infradental lamella. The most common shapes of postpetioles as seen from dorsal view are subrectangular, which have evenly convex sides that are widest midlength (e.g. Figs. 53, 58), and subtrapezoidal, which have convex sides that diverge caudad (e.g. Figs. 49, 50, 51). The terga and sterna of the gaster are referred to by abbreviations Tl, SI, etc. Terminal seg- ments of gaster rotated ventrad means that T2 and T3 are distinctly longer then their correspond- ing sterna, causing the pygidium and hypopygium (T4 and S4) to shift anteroventrad and making T2 or T3 the distalmost sclerite of the gaster. Figs. 28, 40, 43, and 63 show rotated terminal segments; Figs. 2, 5, and 8 show unrotated segments. Pygidial gland sculpture refers to a pair of microareolate patches on the anterior edge of the pygidium. The presence of this characteristic sculp- ture strongly suggests the presence of pygidial glands, but its absence does not necessarily mean the glands are absent (Holldobler, et al. 1976; Kugler 1978a; Holldobler and Engel 1978). In most species I was able to see the gland reservoirs attached to these patches, but failure to see the reservoirs does not mean they are absent; they could easily be lost during dissection. Sting apparatus terms are identified in Fig- ures 3 and 42. Sculpture descriptions follow illustrations and definitions of Harris (1979). It seems to me that the term "areolate" in Harris' nomenclature best de- scribes much of the sculpture in Rogeria, but I suspect many would prefer the term "reticulate." I have called the ridges on the head and mesosoma of many species "rugae," because they are irregu- lar, and wrinkled to varying degrees, but others may prefer to call them "carinae." Also following Harris, in compound descriptors, such as rugose- areolate, the predominant type of sculpture is first. The following are used to describe approxi- mate hair inclination as seen in side view (Fig. 1): erect (> 70 ), suberect (70-50), subdecumbent (50-25), decumbent (25-5), appressed (< 5 ) (in- ference from Wilson 1955). To arrive at the num- bers of erect hairs on the mesosoma dorsum, I used both dark and light backgrounds and a vari- ety of positions to count all erect hairs on one half of the dorsum from the anterior edge of the pronotal disc to the bases of the propodeal spines. Since the hairs are distributed symmetrically, that count gives the number of pairs of erect hairs. Frequently, Volume 3. 1994 23 both sides were counted in order to check for the presence of broken hairs. Erect hair on head dor- sum means on frontal lobes, middorsum, and posterior head, but not under the sweep of the scapes. Erect hair on scapes means on the dorsal or anterior surfaces. Usually no attempt is made to describe hair on ventral and lateral aspects of body, nor on other than extensor surfaces of femora and tibiae. DEPOSITORIES ANIC BMNH CKC CUIC DMOC FML JTLC LACM MCSN MCZ MHN MIZA MZSP NMB PSWC USNM Australian National Insect Collection, CSIRO, Canberra, Australia. British Museum (Natural History), London, U.K. Charles Kugler Collection, Radford University, Radford, VA, U.S.A. Cornell University Insect Collection, Ithaca, NY, U.S.A. David M. Olson Collection, University of Califor- nia, Davis, CA, U.S.A. Fundacion Miguel Lillo, Institute de Zoologia, Miguel Lillo 251, Tucuman, Argentina John T. Longino Collection Los Angeles County Museum, Los Angeles, CA, U.S.A. Museo Civico di Storia Naturale "Giacomo Doha," Genoa, Italy. Museum of Comparative Zoology, Cambridge, MA, U.S.A. Museum d'Histoire Naturelle, Geneva, Switzer- land. Museo Instituto de Zoologia Agricola, Universidad Central de Venezuela, Maracay, Venezuela. Museu de Zoologia da Universidade de Sao Paulo, Sao Paulo, Brazil. Naturhistorisches Museum, Basel, Switzerland. Philip S. Ward Collection, University of California, Davis, CA, U.S.A. United States National Museum, Washington, DC, U.S.A. Genus ROGERIA Emery Myrmica (in part); Fr. Smith 1858:131. Type: Formica rubra Linneaus, 1758, designation of Wheeler, 1911 Tetramorium (in part); Mayr 1886:359. Type: Formica caespitum Linneaus, 1758. Rogeria Emery 1894:188. Type: Rogeria curvipubens Emery, 1894, by designation of Wheeler, 1911. Rogeria subgenus Irogera Emery 1915:191. Type: Rogeria procera Emery, by original designation. Macromedia (in part); Mann 1922:30. Type: Macromischa purpurata Roger 1863, by designation of Wheeler, 1911. Irogera; Brown 1953:4. Irogera; Kempf 1962a:435. [Redescnbed]. Rogeria; Kempf 1963:195. [Key to South American species], Irogera; Kempf 1964:66. [Key to species] Rogeria; Kempf 1965:185. Rogeria; Kempf 1972:227. [Catalog of New World species]. Worker Diagnosis. — Monomorphic myrmi- cines. Antenna 12-segmented; scape not reaching posterior margin of head; distinct 3-segmented club longer than rest of funiculus; apical antennomere longer than combined lengths of other two club segments. No antennal scrobes or fossae. Clypeus projects narrowly between frontal lobes at least to posterior edge of antennal inser- tions. Body of clypeus with one or more pair of longitudinal carinulae. Lateral clypeus not raised into a ridge in front of antennal insertions. Nuchal grooves present on posteroventral corners of head . Anteroventral corners of pronotum angular to dentate and fitting into nuchal grooves. Propodeal spiracle 3 diameters or less from the edge of the propodeum below the propodeal spines. Metapleural lobes not sharply pointed. Worker Description.— Mandibles usually trian- gular. Except in some ciliosa and foreli, mandibles with 5-7 teeth (3 apical teeth decreasing in size basad, followed by 2-3 smaller basal teeth). Addi- tional denticles may occur among basal teeth or any basal tooth may be replaced by a pair of denticles. Palpal formula usually 3,2 or 2,2, but 3,3 in some stigmatica-group, and 2,1 in the very tiny minima. Scape neither elbowed nor ridged at the base, nor with an apron around the peduncle. Clypeus in profile usually with a very narrow anterior apron. Body of clypeus rises near verti- cally in most species, but occasionally projecting beyond the clypeal apron. Frontal lobes narrow, but covering antennal insertions; at most feebly notched behind. Frontal triangle small, depressed slightly. Eyes with 1-100 facets; located on sides, in the anterior half of the head (excluding mandibles). Sides of head widest just behind the eyes, forming rounded corners with the posterior head, which is weakly concave to weakly convex in full face view. Mesosoma generally compact, broad shoul- dered. Anterior face of pronotum rises nearly ver- tically from the neck and usually forms a distinct, rounded angle with the dorsal surface. Mesosoma 24 Journal of Hymenoptera Research dorsum without sutures; no mesonotal groove; metanotal groove absent to distinct. Anterior edge of propodeum often marked by a transverse ca- rina. Propodeal spines absent to long. Metapleural lobes low carinae to rounded triangular. Legs not incrassate. No tibial spurs on middle or hind legs. Tarsal claws simple. Petiolar peduncle with or without a ventral keel; inferior process dentate except in stigmatica group. Node unarmed; poorly to well differenti- ated from peduncle. Postpetiole with short pe- duncles and a low node that is broader than long. Terminal segments of gaster rotated ventrad in all but the stigmatica group. All pygidia dissected have a pair of small pygidial gland reservoirs and/or paired microareolate patches present on anterior edge, except in the stigmatica group. Common features of the sting apparatus are: 1) medial connection of spiracular plate incompletely sclerotized, 2) gonostylus single-segmented, 3) dorsoterminal chaeta present, 4) at least one companion seta (except gibba), 5) each lancet with a single moder- ate to large valve, 6) sting bulb large, with arched sting base. Mandibles usually predominantly smooth, with piligerous punctures and vestigial carinulae at insertions, but carinulae stronger and more extensive in some members of the stigmatica-group. In all but ciliosa, the body of the clypeus is smooth with a pair of carinulae extending from the frontal lobes and stopping short of the clypeal apron; these are sometimes accompanied laterally by 1-2 shorter, weaker carinulae. Lateral extremities of clypeus and adjacent cheeks with longitudinal carinulae. Frontal triangle smooth, except in ciliosa. If macrosculpture present, frontal lobes and middorsum with diverging longitudinal rugae; rest of body areolate, rugose, or occasionally cari- nate. Microsculpture when present usually microareolate, appearing granular or punctured at lower magnifications. Posterior face of propodeum smooth, except in gibba. Legs smooth and shiny. First segment of gaster smooth and shiny; less so mprominula, and minima. Pygidium and sometimes other terminal terga with micro- scopic areolate sculpture on exposed posterior surfaces; pygidium may also possess minute piligerous tubercles. Short appressed or decumbent pilosity com- mon, most dependably on legs (except blanda and procera ) and antennae. Terminal segments of gaster with erect hair. No erect hair on laterodorsa of head (under sweep of scapes). Body of clypeus just above apron with pairs of erect hairs; members of stigmatica group each with an additional median hair. Color from brownish yellow to black, with man- dibles, antennae, and legs lighter. Most species also with a lighter triangle on clypeus, cheeks, and frontal area. Malpighian tubules in the one species exam- ined (belti) 5 in the worker (N=2); 5 to 6 in the male (N=3) (Brown 1988). Wings. — Notation (Fig. 30) follows Brown and Nutting (1949) and Holldobler and Wilson (1990:9). First and second radial crossveins of primitive formicid wings are lost, creating a single large submarginal cell. The distal portion of the radial sector vein (Rsf4+5) vein is straight or curves forward, but never reaches the edge of the wing. The r-m crossvein present (Fig. 30) or absent (Fig. 37). Median and cubital veins of variable length, but not reaching wing margins. Medio-cubital crossvein always present in females, but may be lost in males. Larvae. — Similarities among the larvae of belti, scandens, procera and blanda found by Wheeler and Wheeler (1973, 1976, 1986, 1988) are summarized here. Form pheidoloid, i.e. "abdomen short, stout and straight; head ventral near anterior end, mounted on short stout neck, which is the protho- rax; ends rounded, one end more so than the other" (Wheeler and Wheeler 1976:8). Thorax and abdomen with at least bifid and anchor-tipped hairs; bifid hair on head (sometimes others). Cra- nium subtrapezoidal; clypeus bulging; antennae with 3 sensilla; labrum bilobed and narrowed dorsally. Mandibles ectatommoid (stout, gradu- ally tapered and curved, with an apical tooth and additional teeth and denticles in some species). Maxilla shape paraboloid or lobose, with cylindri- cal galea and palps of varying length. Labium narrow; palps only slight elevations. Maxillae and labial surfaces coarsely spinose. COMPARISONS WITH SIMILAR GENERA Some members of Leptotlwrax (placed with Rogeria in the Leptothoracini) have antennae like Rogeria and some have a narrow posterior lobe of the clypeus, but these have rounded anteroventral Volume 3, 1994 25 corners of the pronotum and no nuchal grooves. Of 65 species of Leptothorax examined at the MCZ, only one had an angular inferior corner of the pronotum, but in that species the scapes extend beyond the head, the posterior lobe of the clypeus is wider, and nuchal grooves are absent. Some species of Lordomymm (Myrmecinini), a genus sometimes confused with Rogeria in pacific islands, have similar antennal and clypeal fea- tures, but have a rounded anteroventral corner of the pronotum and lack nuchal grooves. The stings of Rogeria species are dramatically different from those of five Lordomyrma species I have examined (caledonica Ern Andre, levifrons Mann, punctiventris Wheeler, stnatella Mann, tortuosa Mann) in shape of quadrate plate, oblong plate, gonostylus, trian- gular plate and sting. Another species, L. epinotalis is very different from the other Lordomyrma I ex- amined and has a sting apparatus somewhat like that of Rogeria besucheti or R. ciliosa, but lacks the pronotal corner or nuchal grooves of Rogeria. Wings of three Lordomyrma species I have been able to examine (epinotalis, striatella, leae Wheeler) are dis- tinct from the known Rogeria wings in having the Rsf4+5 vein curve anteriorly to the wing margin. In L. caledonica this vein is like Rogeria in failing to reach the margin, but the m-cu vein is absent. The larvae of an unidentified Lordomyrma species dif- fers from the known larvae of Rogeria as follows: form pogonomyrmecoid, cranium subelliptical, hairs fewer, none anchor-tipped (Wheeler and Wheeler 1977). Stenamma (Pheidolini) workers are similar in form of clypeus, including narrow posterior por- tion between frontal lobes, and some have 3-seg- mented antennal clubs, but in that case the apical segment is shorter than the combined length of the other two segments. Also, Stenamma has no nuchal grooves, the anteroventral corner of the pronotum is rounded, and the metanotal groove is generally more distinct than in Rogeria species. Larvae of Stenamma differ from those of Rogeria as follows: form aphaenogastroid; cranium subhexagonal; mandibles pogonomyrmecoid (similar to ectatommoid); body hairs bifid or denticulate, not anchor-tipped (Wheeler and Wheeler 1953, 1976). Synonymic List of Species Castes that have been collected are recorded in brackets in the right column. W = worker, Q = queen, M = male. Upper case Q and M are used if at least some reproductives are associated through a nest series. Lower case q and m are used if no reproductives are linked to workers by a nest series. An asterisk (*) is added if reproductives were not even found in the same localities as workers. alzatei n. sp. [W, q] belti Mann, 1922 [W, Q, Mj besucheti n. sp. [W] blanda Fr. Smith, 1858 [W, Q, M] =foveata Kempf, 1964 (synonymy by Kempf 1965) bruchi Santschi, 1922 [W] brunnea Santschi, 1930 [W, q] =caraiba Santschi, 1936 n. syn. =cubensis Santschi, 1936 n. syn. =habamca Santschi, 1936 n. syn. =scabra Weber, 1934 n. syn. cannata n. sp. [W] cilwsa n. sp. [W, Q] cornuta n. sp. [W] creightoni Snelling, 1973 [W, q*[ cuneola n. sp. [W, q] curvipubens Emery, 1894 [W, Q] exsulans Wilson and Taylor, 1967 [W] foreli Emery, 1894 [W, Q] =gaigei Forel, 1914 n. syn. =huachucana Snelling, 1973 n. syn. germaini Emery, 1894 [W, q*] =minensis Santschi, 1923 n. syn. gibba n. sp. [W, q] inermis Mann, 1922 (W, Q] innotabilis n. sp. [W, q*] lacertosa Kempf, 1963 [W| leptonana n. sp. [W, q", M) lirata n. sp. [W, q*| megastigmatica n. sp. [W] merenbergiaim n. sp. [W, Q] micromma Kempf, 1961 [W] minima Kusnezov, 1958 (Q] neilyensis n. sp. \\\ | nevadensis n. sp. [W] pellecta Kempf, 1963 [wj procera Emery, 1896 [W] =brasiliensis Borgmeier, 1953 (Synonymy by Kempf 1962a) pronnnula n. sp. [W] scandens Mann, 1922 [VV, Q] scobinata n. sp. [VV, q] sicaria Kempf, 1962b [VV] stigmatica Emery, 1897 [W, Q, M] =sublevinodis Emerv, 1914 n. syn. =manni Santschi, 1922 n. syn. 26 Journal of Hymenoptera Research subarmata Kempf, 1962a terescandens n. sp. tonduzi Forel, 1899 tribrocca n. sp. unguispma n. sp. [W] [W] [W] [W] [W] Names Transferred to Other Genera. azumai Santschi 1941 (to Lordomyrma, Brown 1952) epinotalis Mann 1919 to Lordomyrma, n. comb. rugosa Mann 1921 to Lordomyrma, n. comb. tortuosa Mann 1921 to Lordomyrma, n. comb. tortuosa levifrons Mann 1921 to Lordomyrma, n. comb. tortuosa pohta Mann 1921 to Lordomyrma, n. comb. tortuosa stonen Mann 1925 to Lordomyrma, n. comb. striatella Mann 1921 to Lordomyrma, n. comb. SPECIES GROUPS Some species of Rogeria can be assembled into more or less distinct species groups. Others can not be placed easily in any group, or seem to link several groups. These incertae sedis species are described with the group to which thev may be most related. The stigmatica-group may be distinct enough to be treated as a separate genus, but for now seems tied to the rest of Rogeria through blanda. In the creightom species group, extensive varia- tion and often limited and scattered collections made it difficult to ascertain which specimens were simple variants and which were distinct species. At one time or another, I considered rec- ognizing between four and 20 species in this group. Ultimately, I tried to steer a middle course be- tween these extremes; neither masking the varia- tion by describing large conglomerate species, nor encumbering the genus with a lot of dubious new species names. As a result, I have been somewhat more venturesome about naming new species in this group, yet there remain a number of rather heterogeneous species (alzatei, brunnea, creightom, leptonana). I have described and illustrated some of the variants in these heterogeneous species in order to alert other taxonomists to remaining spe- cies problems and the need for extensive collect- ing, especially in the Caribbean Islands and circumcaribbean countries. In the diagnoses below, if a species is listed as undissected, the palpal formula, pygidium, and sting apparatus characters are unknown. stigmatica-group: besucheti, ciliosa, gibba, megastigmatica, prominula, stigmatica. Diagnosis. — (prominula andmegastigmaticanot dissected). WL 0.67-1. 30mm. Palpal formula 3,3 or 3,2. Eye with 4-41 facets. Metanotal groove distinct and extends down sides of mesosoma. Petiolar peduncle without a keel and inferior process a small step or absent. Nodes and gaster wide. Gaster with terminal segments not rotated ventrad, or only slightly rotated (besucheti). No pygidial gland sculpture or any evidence of reservoirs. Sting ap- paratus: weakly sclerotized, spiracle occupying more than 1/4 of spiracular plate; no anal plate; valve chamber not arising abruptly from base of sting shaft; sting shaft and lancets weak, sparulate (except perhaps for prominula); sting shaft apex without any flange; furcula a simple arch (no dorsal arm). Predominantly dense areolate macrosculpture on head (except middorsum), mesosoma (except some stigmatica), and nodes. Mandibles with piligerous punctures and stron- ger, more extensive carinulate sculpture than in other Rogeria species. No microsculpture on most of body. Body of clypeus with a median erect hair low near anterior margin. Similar incertae sedis species: blanda. This species has sculpture and pilosity like ciliosa and gibba, but has a 2,2 palpal formula, weak metanotal groove, dentate inferior petiolar process, and gaster, py- gidium and sting apparatus typical of other groups. scandens-group: scandens, subarmata, terescandens Diagnosis. — (terescandens not dissected). WL 0.78-1. 17mm. Palpal formula 3,2 Eyes large (30-80 facets). Propodeal spiracle faces strongly caudad. Metapleural lobes prominent, triangular. Petiolar node low, anterior face indistinct from peduncle (except some subarmata). Inferior petiolar process dentate. Postpetiolar sternum large. Pygidial gland sculpture and reservoirs present, at least in subarmata. Sting apparatus like that of inermis (Fig. 42). Head, mesosoma, and nodes predominantly rugose; sides of head below eye smooth. Dorsa of head, mesosoma nodes, and gaster Tl with sparse, stiff, untapered, erect hairs and sparse, incon- spicuous appressed pilosity (occasional decum- bent hairs in some subarmata). Scapes and legs with appressed hair only. Volume 3, 1 994 27 Similar incertae sedis species: procera, tonduzi. These both lack the stiff hairs and inflated postpetiole that make the group so distinctive, but otherwise are so similar that perhaps they should be included. germaini-group: germaini, lirata, lacertosa. Diagnosis. — WL 0.65-1. 05mm, usually > 0.80mm. Mandibles triangular. Palpal formula 3,2. Body of clypeus not projecting over apron. Eye small (6-20 facets; OI 0.09-0.14). Propodeal spines rather long (PSI 0.18-0.23); directed caudad. Peti- ole with prominent keel and dentate inferior pro- cess. Petiolar node rises abruptly from anterior peduncle and with more or less distinct anterior, dorsal and posterior surfaces. Posterior two- thirds of postpetiolar node with parallel or slightly con- vergent sides. Postpetiolar sternum in side view rather long, its anterior corner perpendicular. Py- gidial gland sculpture and reservoirs present. Sting apparatus similar to that of inermis, but sting shaft and lancets weak; lancet lacks barbule. Laterodorsa, sides, and posterior head rugose-areolate (some- times effaced on sides). Mesosoma predominantly rugose or vermiculate-rugose. Microsculpture weak; intervals on head and mesosoma shiny and nearly smooth. Mesosoma dorsum and gaster Tl with abundant fine, tapered, erect-suberect hairs, but very little, if any, decumbent or appressed pilosity. Similar incertae sedis species: pellecta, sicaria (sicaria not dissected) These have different petiole shapes and more decumbent hair on the mesosoma. Rogeria sicaria also differs in clypeus, propodeal spines, and erect pilosity. creightoni-group:alzatei,brunnea,carinata,creightoni, innotabilis, leptonana, merenbergiana, nevadensis, scobinata, tribrocca, unguispina Diagnosis. — (brunnea, nevadensis, and tribrocca not dissected). WL 0.51-0.93mm. Mandibles trian- gular, except as noted. Palpal formula 3,2 or 2,2. Eyes small (6-19 facets). Petiole with a distinct node and dentate inferior process. Pygidial gland sculpture and reservoirs present. Sting apparatus like that of inermis unless otherwise noted. Mesosoma predominantly rugose or carinate. Head dorsum, mesosoma, nodes, and gaster cov- ered with two distinct kinds of hairs: 1) abundant short, decumbent pilosity and 2) equally or less abundant longer, erect-suberect hair (except for lack of erect hair on head of some scobinata.). Usually more than 10 pairs on mesosoma dorsum (9 in some leptonana, 8 in carinata) and 2 or more posterodorsally projecting hairs on each node. All hairs tapered. Similar incertae sedis species: inermis, belti, cornuta, neilyensis, exsulans (cornnta and neilyensis not dissected). These do not have two distinct types of hair on the mesosoma dorsum. curvipiibens-group: cuneola, curvipubens Diagnosis. — WL 0.50-0.63mm. Mandibles tri- angular. Palpal formula 2,2. Clypeal apron usu- ally convex (medially flattened in Haitian curvipubens). Eye small (4-11 facets). PSI 0.13-0.18. Petiole with distinct node, weak keel, and dentate inferior process. Postpetiole from above usually as in Fig. 74. Pygidial gland sculpture and reservoirs present. Sting apparatus as in inermis (Fig. 42), except for sting shaft and lancets of curvipubens. Median head and cheeks with weak longitudinal rugose-areolate macrosculpture. Mesosoma with predominantly rugose macrosculpture and ar- eolate microsculpture, both often weak. Body cov- ered with appressed-decumbent pilosity. Erect hairs sparse: none on scapes or extensor surfaces of legs, 0-16 on head (if present, short and limited to posterior head), usually 2-7 pairs on mesosoma dorsum (rarely 8), 0-1 pair posterodorsally pro- jecting and 0-1 laterally projecting hairs on each node. Anterior portions of gaster Tl often lack erect hair. All hairs tapered. Similar incertae sedis species: micromma, minima (neither dissected). Rogeria micromma has a subrectangular postpetiole in dorsal view and abundant erect hair on head. Rogeria minima has stiff, cuneate hairs and a 2,1 palpal formula. /ore//-group: bruchi,foreli Diagnosis. — WL 0.50-0. 80mm, usually < 0.75mm. Mandibles triangular. Palpal formula 2,2. Eye 5-20 facets. Propodeal spines wide at base. Petiole with distinct node and dentate inferior process. Pygidial gland sculpture and reservoirs present. Sting apparatus almost identical to that of inermis (Fig. 42). Sculpture predominantly microareolate on head, mesosoma, and waist; 28 Journal of Hymenoptera Research macrosculpture feeble and limited to head. Most clypeus, ventral head, coxae, trochanters, sterna of of body with short, appressed-decumbent pilosity gaster, and terminal terga of gaster. only. Erect-suberect hairs present on mandibles, KEY TO WORKERS OF ROGERIA 1 Pilosity appressed on mesosoma dorsum and gaster Tl. A very small number of short decumbent hairs may also be present , 2 1 ' Erect, suberect, or subdecumbent hairs present on mesosoma or gaster Tl, often in addition to appressed or decumbent pilosity 4 2(1) Body of clypeus strongly projecting over apron. Basal angle of mandibles very reduced, (amazonian Brazil) (Fig. 14) prominula 2' Body of clypeus at most projecting only slightly over apron. Mandibles triangular 3 3(2') Clypeal apron emarginate. Mandible with 5 teeth. (Paraguay, Argentina) (Fig. 82) bntclii 3' Clypeal apron convex, often with faint median angle. Mandible with 6-7 teeth. (Caribbean, southwestern United States through northern South America) (Figs. 79-81, 104-105) foreli 4(1') Very dense, flexuous, erect hairs on dorsa of head, mesosoma and gaster; no decumbent or appressed pilosity. Promesonotum uniformly and densely areolate 5 4' Erect hair not so dense or flexuous. Decumbent hair present. Promesonotum may have some rugae or carinae 7 5(4). Metanotum forms an abrupt declivity between promesonotum and propodeum. (lowland Colombia, Ecuador) (Figs. 5-6) gibba 5' Shallow metanotal groove hardly interrupts mesosoma profile 6 6(5') Petiolar node relatively short and tall . Gaster larger (GW / WL 0.94-0.97). (lowland Ecuador, Venezuela) (Figs. 2-4) ciliosa 6' Petiolar node relatively long and low. Gaster smaller (GW/ WL 0.63-0.70). (Central and South America) (Figs. 83-84) blanda 7(4'). Ventral process of petiolar peduncle reduced to a small step in petiolar profile, or absent 8 Ventral process of petiolar peduncle dentate 10 8(7) Propodeal spines longer (PSI 0.17-0.20). Ventral process of petiolar peduncle present as a small step. (Paraguay, Peru, Colombia) (Figs. 15-16) besucheti 8' Propodeal spines shorter (PSI 0.07-0.13). Petiolar peduncle lacking a ventral process 9 9(8') WL 0.72-0.92mm. (Pacific Islands) (Figs. 7-12) stigmatica 9' WL 1.18-1. 30mm. (Solomon Islands) (Fig. 13) megastigmatica 10(7') WL 1.28-1. 53mm. Eye large (80-100 facets). Postpetiole with an inconspicuous sternum. (Guyana, amazonian Brazil) (Fig. 18) procera 10' WL and eyes usually much smaller. If WL and eyes nearly as large, then postpetiolar sternum conspicuous 11 11(10) Sparse, stiff, erect-suberect hairs and short appressed hairs on dorsa of head, mesosoma, waist, and gaster Tl. WL > 0.72mm. Eye with > 25 facets 12 11' Erect hairs more curved and tapered; shorter pilosity, if present, usually decumbent. If some erect hairs are stiff, then WL and eye are smaller 14 12(11) Propodeal spines very short (PSI 0.09-0.12). Pygidium with a pair of median tubercles just above the sting, (eastern Brazil) (Figs. 22-23) subarmata 12' Propodeal spines longer (PSI > 0.15). Pygidial tubercles lacking 13 13(12') Macrosculpture vestigial or absent on lateral pronotum and petiolar node. Waist slender (PpetW/PpetL 1.08-1.16). (Costa Rica) (Fig. 21) terescandens 13' Lateral pronotum with distinct rugae, petiolar node strongly rugose to areolate-rugose. Waist inflated (PpetW/PpetL 1.38-1.61). (Central America) (Figs. 19-20) scandens 14(11) Petiole clavate. Eye with 39-48 facets. (Central America) (Figs. 85-86) tonduzi 14' Petiole with a distinct node and/or smaller eyes 15 15(14) Undamaged hair on mesosoma dorsum mostly erect to suberect. Decumbent hairs, if present, much less abundant than longer, more erect pilosity 16 15' Mesosoma dorsum with abundant appressed or decumbent pilosity, usually in addition to longer erect to suberect hair 21 Volume 3. 1994 29 16(15) EL/SpL > 1.00. Gaster Tl usually lacks decumbent and appressed pilosity (occasionally 1 or 2 decumbent hairs present). (Central America) (Figs. 40-42) inermis 16' EL/SpL usually < 0.90. If greater (some belti), then gaster Tl with abundant decumbent pilosity.... 17 17(16) Clypeal apron with median tooth 18 17' Clypeus emarginate to evenly convex 19 18(17) Promesonotum with thick rounded vermiculate rugae. Metapleural lobes well developed, triangular. (Trinidad, Guyana, amazonian Peru and Colombia, Mato Grosso) (Figs. 28-30, 89-90) lirata 18' Promesonotum areolate-rugose to vermiculate rugose; rugae narrower, sharper. Metapleural lobes low, broadly rounded. (Paraguay, Mato Grosso and Minas Gerais) (Figs. 24-27, 87-88) germaini 19(17') Eyes small and propodeal spines long and horizontal (EL/SpL < 0.46). (Belize, southern Mexico) (Fig. 45) cornuta 19' EL/SpL > 0.46. Propodeal spines more inclined 20 20(19') Eye smaller, elliptical (OI 0.13-0.14). Petiolar peduncle with lamellate keel. Sides of postpetiolar node distinctly areolate. (southern Brazil) (Fig. 31) lacertosa 20' Eye larger, oval (OI 0.18-0.22). Petiolar keel not lamellate. Sides of postpetiolar node smooth, or nearly so. (southern Mexico to Colombia) (Figs. 35-39, 91-92) belti 21(15') Scapes with longer erect to suberect hair in addition to shorter decumbent to appressed pilosity (in exsidans and some brunnea, erect hairs can be sparse and little longer than decumbent hair) 33 21' No erect hair on scapes 22 22(21) WL 0.90-0.99mm. EL/SpL 0.48-0.56. (southern Brazil) (Figs. 32-33) pellecta 22' Smaller. If WL approaches 0.90mm, then EL/SpL either < 0.30 or > 0.70 23 23(22') Propodeal spines very long (PSI 0.29), strongly inclined. Eyesmall (EL/SpL0.13). (southern Brazil) (Fig. 34) sicaria 23' Propodeal spines usually much shorter. If similar in size, then not strongly inclined. EL/SpL > 0.20 .. 24 24(23') WL > 0.71mm. Pilosity on gaster not differentiated into two distinct kinds. (Costa Rica) (Fig. 43) neilyensis 24' WL usually < 0.71mm. If similar in size, hair on gaster differentiated into short-decumbent and longer- erect 25 25(24') Clypeal apron with median tooth. Gaster T3 with short median spine. (Colombia) (Fig. 63) tribrocca 25' Clypeal apron without a median tooth. Gaster T3 unarmed 26 26(25) Erect-suberect hairs absent from head dorsum (or short and limited to posterior head) and sparse on mesosoma. WL 0.51 -0.63mm 27 26' Erect hair present on middorsum and posterior head. If missing (scobinata), then > 10 pairs of erect hairs on mesosoma dorsum 28 27(26) Postpetiolar sternum wedge-shaped in side view; node less vaulted. Sides of head and pronotum strongly microareolate, making macrosculpture (if present) difficult to see. (central Mexico to Costa Rica) (Figs. 77-78, 103) cuneola 27' Postpetiolar sternum not wedge-shaped; node more vaulted. Sides of head and pronotum with more or less smooth, shiny intervals. (Caribbean Islands, northern South America; possibly Central America) (Figs. 74-76, 101-102) curvipubens 28(26') Part of gaster Tl microareolate. Most erect hairs stiff, cuneate-fimbriate. (Argentina) (Figs. 72-73) minima 28' Gaster Tl smooth. Stiff hairs absent (except on some micromma) 29 29(28') Rugae on pronotal dorsum and sides unbranching and nearly straight, with smooth and shiny intervals. Sides of head below eye nearly smooth, strongly shining, (northeastern Colombia over 1000m) (Figs. 65, 94) nevadensis 29' Rugae on promesonotum and sides of pronotum undulating and with lateral spurs or branches, sometimes forming areolate patches; intervals appear granular, not strongly shining. Sides of head rugose-areolate 30 30(29') Eye with 2-5 facets. Mesosoma sides opaque with dense areolate microsculpture, but macrosculpture absent. Mesosoma dorsum with 8-10 pairs of erect hairs. (Surinam, Para State of Brazil) (Fig. 71) .... micromma 30' Eye with 4-21 facets (rarely < 7). Mesosoma sides with rugose macrosculpture in addition to weak or distinct microsculpture. Mesosoma dorsum with > 12 pairs of erect hairs 31 30 Journal of Hymenoptera Research 31(30) Macrosculpture tuberculate on posterodorsum of head (Fig. 100). Erect hair usually absent from head dorsum, but sometimes short, sparse and mostly limited to posterior rim. (Trinidad, South America east of Andes to Paraguay) (Figs. 61-62, 100) scobinata 31' Posterodorsum of head rugose to areolate; no tubercles. Middorsum of head with long erect hair .32 32(31 ) Clypeal apron convex. Nuchal groove clearly visible in lateral view. Propodeal spines generally wider. Petiolar keel distinct. (Mexico, Nicaragua; possibly into Colombia) (Figs. 55-57, 97-98) .. innotabilis 32'. Clypeal apron truncate in Central America. Nuchal groove not clearly visible in lateral view. Propodeal spines usually slender. Petiolar peduncle with little or no keel. (Panama, South America, Dominican Republic) (Figs. 58-60, 99) alzatei 33(21) WL 0.93-1. 02mm. Propodeal spines long (SpL > 0.20mm; PSI > 0.25), horizontal. Gaster with few or no decumbent hairs. (Belize, southern Mexico) (Fig. 45) comuta 33' WL usually < 0.90mm. If larger (some belt i, brunnea, creightoni), then decumbent hair abundant on gaster 34 34(33) Gaster Tl with two distinct kinds of pilosity: shorter, decumbent hairs and longer, erect hairs 36 34' Hair on gaster Tl not clearly sorting into two distinct kinds 35 35(34) Pilosity on gaster Tl more dense and decumbent to subdecumbent. Petiolar peduncle with very large keel. (Samoa) (Fig. 44) exsulans 35' Pilosity non gaster Tl more sparse and erect to subdecumbent. Petiolar peduncle with little or no keel. (Central America) (Figs. 40-42) biennis 36(34) Eyes larger (> 19 facets), oval. Macrosculpture on pronotal disc usually predominantly areolate and extending uninterrupted onto anterior face of pronotum (if predominantly rugose, then eye with > 25 facets), (southern Mexico to Colombia) (Figs. 35-39, 91-92) belti 36' Eyes smaller, often elliptical. Pronotal disc predominantly longitudinally rugose, vermiculate, or carinate. Anterior edge of disc transversely rugose, rugose-areolate, or carinate 37 37(36') Promesonotum with nearly straight, longitud inal cannae; no lateral branches; intervals very smooth and shiny. Eye small, elliptical (EL < 0.10. (Caribbean Islands) (Fig. 64) carinata 37' Promesonotum longitudinally rugose, vermiculate, or rugose-areolate. If intervals smooth and shiny, then eyes oval and EL > 0.10 38 38(37) Clypeus evenly convex. Postpetiolar node from above subtrapezoidal; sternum long, flat, without anterior lip. (Caribbean Islands) (Figs. 50, 93) brunnea 38' Clypeus usually emarginate or truncate. If convex, then postpetiolar node not subtrapezoidal; sternum shorter, with distinct anterior lip 39 39(38') Eye elongate-oval. A strong transverse carina runs across pronotal shoulders. Pronotal disc and sides with weakly undulating rugae and nearly smooth interrugal spaces. Propodeal spines often with downcurved tips. (Venezuela) (Fig. 49) unguispina 39' Eyes not elongate. Anterior pronotum may have 1 or more transverse rugae, but not carinae. Pronotal disc rugose or rugose-areolate. Propodeal spines straight 40 40(39') WL 0.54-0. 66mm. Mesosoma low, slender (PW < 0.37mm). Propodeal spines short, sometimes absent (SpL < 0.11mm). Postpetiole generally widest in anterior half (Fig. 66) (southern Mexico, Central America, Dominican Republic) (Figs. 66-70) leptonana 40' Generally larger, with MHI > 0.90, PW > 0.37mm, and SpL > 0.10mm. Postpetiole subtrapezoidal or subrectangular (Figs. 51, 53) 41 41(40') Metanotum strongly interrupts contour of mesosoma profile. Sides of head rugose or effaced rugose- areolate. EL/SpL > 0.65. (Colombian and Ecuadorian Andes above 1000m) (Figs. 46-48) meretibergiana 41' Metanotum with little or no effect on overall shape of mesosoma profile. EL/SpL < 0.65. Sides of head rugose-areolate; not effaced, (southern Texas to Panama) (Figs. 51-54, 95-96) creightoni Volume 3, 1 994 31 SPECIES DESCRIPTIONS Stigmatica-Group and Related Species Rogeria ciliosa new species Figs. 2-4 Additions to stigmatica-group diagnosis. WL 1 .06-1 .20mm. Mandibles triangular; coarsely punc- tured, weakly carinulate. Anterior clypeus evenly convex. Propodeal spines long (> 0.20mm), strongly inclined and diverging. Metapleural lobes well developed; angular. Inferior petiolar process reduced to a small step. Dorsal head, mesosoma and gaster densely covered with long flexuous hairs; terminal segments of gaster with dense, stiff erect hair. Holotype and Paratype Workers. — TL 4. 1-4. 8 (4.1), HL 0.90-1.02 (0.90), HW 0.78-0.89 (0.785), SL 0.69- 0.77 (0.69), EL 0.14-0.16 (0.15) (32-42 facets), PW 0.60-0.66 (0.60), WL 1.06-1.20 (1.06), SpL 0.24-0.28 (0.245), PetL 0.44-0.53 (0.44), PpetL 0.23-0.29 (0.235)mm, CI 0.84-0.87 (0.87), OI 0.17-0.19 (0.19), SI 0.85-0.90 (0.88), PSI 0.21-0.23 (0.23). N=9 Mandibles with 7-9 teeth (3 apical teeth de- creasing in size basad, followed by 4-6 small, subequal teeth). Palpal formula 3,2. No clypeal apron. Body of clypeus not projecting over ante- rior edge. Nuchal grooves weak. Pronotal shoul- ders rounded. Ventral petiole with a low median carina rather than a distinct keel. Node distinct, wider than long. Postpetiole dorsal view shape as in Fig. 66. Postpetiolar sternum short, with a dis- tinct anterior lip followed by a narrow sulcus. Gaster large (GW/WL 0.94-0.97). Quadrate plate of sting apparatus with somewhat reduced apodeme that lacks lobes on anterodorsal corner; oblong plate ventral arm very reduced (see also Fig. 3 and stigmatica-group diagnosis). Body of clypeus with fragmented longitudi- nal rugulae surrounded by effaced areolate sculp- ture. Longitudinal rugae on head dorsum mostly confined to the frontal lobes, frontal area, and midline. Frontal lobes rugose-areolate in some paratypes. Rest of dorsum, cheeks, venter, and posterior head densely areolate; intervals bearing shallow piligerous punctures. Promesonorum with the same areolate sculpture. Mesopleura, metapleura, and sides of propodeum with more irregular and confused areolate sculpture. Met- anotal groove scrobiculate. Dorsal face of propodeum transversely rugose with undulating, smooth intervals or largely areolate with a few carinulae between the spines. Most of petiole and postpetiole strongly areolate as well. Sculpture on anterior petiolar node effaced; dorsum of peduncle smooth. First tergum and sternum of gaster rather coarsely punctured in front and more finely punc- tured caudad; smooth and shiny between punc- tures. Fine, long, flexuous, erect to suberect hair covers middorsum of head, dorsum of mesosoma, dorsum and sides of waist, and first segment of gaster. On terminal segments of gaster these be- come denser, stif fer and more erect to form brush- like rings. Shorter, subdecumbent hairs occur on lateral and ventral surfaces of head, dorsal sur- faces of scapes, sides of mesosoma, and sometimes dorsal face of propodeum. Very short, decumbent to appressed pilosity on extensor surfaces of legs and ventral petiole. Median hair on clypeus fine and often obscured by surrounding paired hairs. Body uniformly rusty-brown; appendages lighter, more yellowish-brown. Paratype Queen.— TL5.3,HL 1.04, HW 0.91, SL 0.80, EL0~27, PW 0.97, WL 1.50, SpL 0.30, PetL 0.58, PpetL 0.30mm, CI 0.88, SI 0.88, PSI 0.20. N=l Differing from paratypes only in the normal queen characters (Fig. 4). Mandible8-toothed. Short parapsidal furrows present on mesoscutum. Both mesoscutum and mesoscutellum with same dense areolate sculpture as in worker. Metanotum vaguely microareolate. Dorsal face of propodeum with transverse carinulae mesad; areolate laterad. Discussion. — Rogeria gibba (Figs. 5-6) from western Ecuador and Colombia resembles ciliosa in size, sculpture and pilosity, but has different mandibles, clypeal margin, propodeal spines, and metapleural lobes. Rogeria stigmatica and megastigmatica (Figs. 7-13) from the Pacific have much shorter propodeal spines, less punctured gaster, and different pilosity. The name ciliosa refers to its dense covering of flexuous hairs. Distribution. — This species is known only from lowland rain forests in the amazonian basin of Ecuador and the Orinocan basin of Venezuela. Material Examined. — Holotype locality. EC- UADOR: Napo Province, Limoncocha, 250m, 18- VI-1976, #B-348 (S. and J. Peck) [MCZ]. Paratype locality. 9 workers, 1 queen, VEN- EZUELA: Bolivar State, Campamento Rio Grande, 8.07N61.42W, 250m, 14- VIII-1986, sifted leaf mold 32 Journal of Hymenoptera Research and rotten wood, #8572-12 (P. S. Ward) [2 workers dissected: mouthparts, 2 stings] [BMNH, LACM, MCZ, MIZA, MZSP, USNM]. Rogeria gibba new species Figs. 5-6 Additions to stigmatica-group diagnosis. WL 0.85-0.93mm. Mandibles subtriangular. Clypeal apron slightly convex medially, with sharp cor- ners on either side. Eyes with 16-20 facets. Mesosoma profile humpbacked. Propodeal spines short (< 0.15mm), not strongly inclined. Propodeal spiracle rather large, within one diameter of pos- terior edge of mesosoma. Metapleural lobes re- duced to short carinae, sometimes nearly absent. No inferior petiolar process. Dorsal head, mesosoma and gaster densely covered with long flexuous hairs. Holotypeand Paratype Workers. — TL 3.5-3.8 (3.7), HL 0.78-0.85 (0.83), HW 0.73-0.81 (0.78), SL 0.54- 0.60 (0.59), EL 0.09-0.10 (0.10) (15-22 facets), PW 0.51-0.57 (0.55), WL 0.85-0.96 (0.93), SpL 0.11-0.14 (0.11), PetL 0.40-0.49 (0.46), PpetL 0.21-0.23 (0.22)mm, CI 0.91-0.96 (0.94), OI 0.12-0.13 (0.13), SI 0.72-0.77 (0.76), PSI 0.12-0.15 (0.12). N=10 Mandibles with 5 teeth diminishing in size basad. Palpal formula 3,3. Body of clypeus does not project over apron. Head capsule about as wide as long. Nuchal grooves visible from below or behind. Promesonotal dorsum almost flat and dropping abruptly to metanotum (Fig. 5). Petiolar peduncle lacks inferior petiolar process. Petiolar node distinct, bulbous, wider than long. Postpetiole node subrectangular from above; sternum short. Gaster large (GW/ WL 0.90-0.97). Sting apparatus like that of ciliosa, except for: 1 ) a longer tulcral arm on the oblong plate (like inermis Fig. 42), 2) no companion seta on gonostylus, 3) more enlarged lancet apex, and 4) no anterolateral processes on sting base (Fig. 6). Middorsum of head longitudinally rugose becoming rugose-areolate behind level of eyes. Laterodorsa, posterior, sides, and ventral surfaces of head areolate with minutely granulate ridges; intervals smooth except for some piligerous punc- tures. Promesonotum with the same sculpture as sides and back of head. Meso- and metapleura with more confused areolate sculpture, but simi- lar ridges and intervals. Metanotal groove scrobiculate. Dorsal face of propodeum and part of posterior face transversely rugulose. Rest of posterior face shagreened. Petiolar peduncle colliculate; anterior face of node smooth, except for piligerous punctures. Rest of petiolar node and all of postpetiolar node transversely areolate-rug- ose, with granulate ridges and smooth to weakly punctured intervals, as on head and prome- sonotum. Gaster smooth and shiny, dotted by numerous small, shallow piligerous punctures. Dorsum of head, mesosoma, top and sides of waist, and all sides of gaster densely covered with long, flexuous, erect to suberect hairs and without interspersed shorter pilosity. Body rusty-brown, gaster slightly darker. An- tennae, legs and mandibles yellowish-brown to yellow. Queens.— TL 4.2-4.3, HL 0.83-0.85, HW 0.80, SL 0.58-0.59, EL 0.19-0.20, PW 0.72-0.73, WL 1.14- 1.15, SpL 0.15-0.17, PetL 0.49-0.53, PpetL 0.24- 0.25mm, CI 0.94-0.96, SI 0.73-0.74, PSI 0.13-0.15. N=4 Queens differ from the workers in the usual and the following ways. Short parapsidal furrows present. Longitudinal areolate-rugose sculpture on median head extends to posterior of head. Median pronotum transversely rugose. Mesoscutum with longitudinally rugose sculp- ture that becomes more areolate on mesoscutellum. Discussion. — See the ciliosa discussion for com- parison. Like stigmatica also in many features, but differs in pilosity and shape of promesonotum. The name gibba is from Latin meaning humpback. Distribution. — One worker is from the eastern side of the central cordillera in northern Colombia, the rest are from about 1000km away on the west- ern slope of the Andes and the coastal range of northern Ecuador. In all three areas they were collected at 300-800m elevation in natural rain forest, probably by Berlese sampling. Material Examined. — Holotype locality. EC- UADOR: Pichincha Province, 4 km E. Santo Domingo de los Colorados, 22-VI-1975, #B-304 (S. and J. Peck) [MCZ]. Paratype localities. COLOMBIA: 1 worker, Antioquia Department, near El Bagre, Providencia, Estacion Biologica, Zona Buenos Aires, 30-31-XII- 1977 (C. Kugler) [MCZ]; 1 worker, Choco Depart- ment, Rio Napipi, 1968 (P. A. Silverstone) [LACM]. ECUADOR: 5 workers, holotype locality, 22- VI- 1975 and 8-VII-1976 (S. and J. Peck) [2 mouthparts, stings] [BMNH, CKC, MCZ, MZSP]; 2 workers, Volume 3, 1994 33 Pichincha Province, 47km S. Santo Domingo, Rio Palenque Station, 23-V-1976 (S. and J. Peck [MCZ]; 1 worker, Pichincha Province, Tinalandia, 16km SE. Santo Domingo de los Colorados, 4-VI-1976 (S. and J. Peck) [MCZ]. Nontype localities. ECUADOR: 3 queens, Pichincha Province, 47km S Santo Domingo, Rio Palenque Station, 1975 (S. and J. Peck); 1 queen, Manabi Province, 73km NE Chone, 12-VI-1976 (S. and J. Peck) [MCZ]. Rogeria stigmatica Emery Figs. 7-12 Rogeria stigmatica Emery 1897:589. Syntype workers, NEW GUINEA: Friednch-Wtlhelmshafen [=Madang] (Biro) [MHN] (Both syntypes examined]. Rogeria stigmatica subsp. sublevinodis Emerv 1914:415. Svntvpe workers, LOYALTY ISLANDS: Mare, Raoua [MHN] [Both syntypes examined). N. syn. Rogeria sublevinodis; Wilson and Taylor 1967:76, Fig. 61. Rogeria m 0.10mm. Propodeal spiracle < half its diameter from edge of infradental lamella. Propodeal spines short (PSI 0.07-0.13), strongly inclined dorsad. Metapleural lobes much reduced. Inferior petiolar process absent. Abundant de- cumbent hair on head, mesosoma, waist and gaster; erect hairs rather sparse, not flexuous. Workers.— TL 3.0-3.7, HL 0.66-0.83, HW 0.58- 0.71, SL 0.46-0.61, EL 0.05-0.15 (7-34 facets), PW 0.44-0.52, WL 0.72-0.92, SpL 0.05-0.10, PetL 0.32- 0.42, PpetL 0.19-0.23mm, CI 0.84-0.92, OI 0.08- 0.22, SI 0.79-0.87, PSI 0.07-0.13. N=28 Mandibles subtnangular (usually) to triangu- lar; with 5 subequal teeth or 3 apical teeth followed by 3-4 (rarely 5) smaller teeth or denticles. Palpal formula 3,3. Little or no clypeal apron; median clypeal margin truncate, weakly convex, or weakly angular. Body of clypeus not projecting over clypeal margin. Eyes oval, large (EL 0.10-0. 15mm and 17- 34 facets), except in some of the Papua New Guinea workers (EL 0.05-0.07mm; 7-9 indistinct facets). Nuchal groove distinct from behind, but not clear in side view. Promesonotum with evenly convex profile. Metanotal suture narrow, emphasized by a sharp ridge at anterior edge of propodeum. Node large, wider than long, more or less sym- metrical in side view (Fig. 8). Postpetiolar node in side view rounded front to back; usually widest in anterior half, much as in Fig. 66, but sometimes widest midlength. Postpetiolar sternum short. Sting apparatus like that of ciliosa (Fig. 3), except for: 1) spiniform medial and and lateral projec- tions from anterodorsal corner of quadrate plates, 2) smaller valve chamber, and 3) lack of anterolateral processes on sting base (Fig. 9). Gonostylus sometimes with no clear sensillar gap; sometimes lacking a companion seta. The "Rogeria (stigmatica group) spp. 1 and 2" in Kugler (1978b) are both stigmatica. The sting shown here is more accurate than the previous one, which was not in full lateral orientation when drawn. Middorsum of head longitudinally rugose; rest of head, including venter, coarsely areolate. Dorsum of promesonotum coarsely areolate to rugose (intermediate specimens predominantly areolate, but with elongate cells or short rugae medially; rarely, rugae also occur on shoulders). Anterior and sides of mesosoma areolate, finely so on neck and ventrad on meso- and metapleura. Spaces in sculpture smooth except for piligerous punctures. Dorsal face of propodeum areolate along very anterior margin, followed by either transversely rugulose or densely punctate sculp- ture, or both in varying degrees of density and definition. Petiolar peduncle finely colliculate or smooth. Anterior and apex of node smooth or areolate; posterior face and sides areolate, some- times with a few rugae. Postpetiole transversely rugose-areolate behind, becoming more effaced anteriorly, often leaving anterior face smooth and shining. Decumbent to subdecumbent pilosity covers most of body. Sparser erect to suberect hair also on dorsa of scapes, head, mesosoma, nodes, and gaster. Erect hair moderately abundant on gaster Tl of most specimens (Fig. 8), but sparse on speci- mens from Papua New Guinea. Body of clypeus with strong median seta. Color of mandibles, frontoclypeal region, an- tennae, and legs light brownish-yellow to brown. Rest of body light brown to blackish-brown. Queens.— TL 3.6-4.5, HL 0.73-0.85, HW 0.66- 0.75, SL 0.53-0.62, EL 0.19-0.24, PW 0.60-0.71, WL 1.00-1.17, SpL 0.10, PetL 0.41-0.48, PpetL 0.21- 0.26mm, CI 0.87-0.90, SI 0.80-0.84, PSI 0.09-0.10. N=6 34 Journal of Hymenoptera Research As in worker except for the usual caste differ- ences. Mesosoma habitus as in Fig. 10. Queen from McAdam Park, Papua New Guinea with median bulge on pronotum. Pronotum areolate on sides; finer and transversely rugose-areolate mesad. Mesoscutum longitudinally rugose; mesoscu- tellum areolate-rugose. Metanotum smooth. Mesosoma sides confused areolate, except for smooth area on mesokatepisterna and costulate metapleural gland bullae. Wing venation as in Hrata (Fig. 30), except for Rs vein as in be/f; (Fig. 37). Males.— TL 2.6-3.1, HL 0.45-0.54, HW 0.56- 0.66 SL 0.27-0.35, EL 0.20-0.26, PW 0.50-0.62, WL 0.84-1.04, PetL 0.22-0.30, PpetL 0.14-0. 19mm, CI 1.22-1.27, SI 0.48-0.53. N=3 Mandibles with a large apical tooth and 4 others decreasing in size basad. Posterior lobe of clypeus projects more broadly between antennae than in worker; anterior clypeal margin weakly convex. Frontal lobes absent. No distinctly im- pressed frontal area. Funicular segment 6 curved and longer than 4 and 7; more extremely curved and elongate on one side of the head than the other. Posterior outline of head medially concave; sharp crests run from ocelli to posteroventral cor- ners of head, which project slightly and fit around prothoracic sternum when head is retracted. Mesosoma and waist as shown in Fig. 1 1 . Genitalia as shown in Fig. 12. Head integument vaguely roughened. Mesosoma and waist smooth, except along furrows and on sides of propodeum, metepimera, and petiolar peduncle. Gaster smooth and shining. Pilosity all erect to suberect, except around eyes. Propodeum nude. Color variation as in worker. Discussion. — According to Emery (1914) and Wilson and Taylor (1967), sublevinodis differs from stigmatica in having larger size, coarser sculpture on head and mesosoma, and smooth nodes with coarser punctures on other parts of the waist. When two of Emery's stigmatica syntypes and two of his sublevinodis syntypes in the MHN were compared side by side, the TL and WL of the stigmatica syntypes fell within the range of the sublevinodis syntypes. The sculptural characteris- tics were not distinct either, except on the dorsal face of the propodeum, which is transversely ru- gulose and very weakly punctate and shiny in stigmatica syntypes, but densely punctate and lack- ing rugulae in the sublevinodis syntypes. However, intergrades with rugulae and various degrees of punctation occur in the Solomon Islands, Irian Jaya, and Pohnpei. Santschi (1922) claimed that manni differed from stigmatica in a variety of ways. After examin- ing manni types, I could confirm only one clear way they differ from the types of stigmatica and sublevinodis: the presence of rugose sculpture in- stead of areolate sculpture on the promesonotum. However, in non-type material, I found all inter- mediate states, sometimes within the same local- ity. Other supposedly different characters also intergrade or are due to the manni types being at the small end of the size distribution. I have too few collections from Papua New Guinea to know if those specimens with unusually small eyes and few erect hairs on gaster Tl might be a distinct species. See sister species megastigmatica description. See also ciliosa and gibba for discussions of related species in South America and exsulans for discus- sion of a not so related species from the Pacific. Distribution and Behavior. — Rogeria stigmatica is known only from the Central and West Pacific, from as far east as Tahiti to the western tip of the Island of New Guinea and from about 22 S to 7°N. Most collections come from berlesate or sift- ing of leaf mould, rotten wood, soil, moss, or bases of fern epiphytes in rain forest. The one nest series with ecological data (Sorong, Irian Jaya) is from rotten wood. Mann (1921:451) found colonies be- neath stones and logs. Twelve specimens, were collected on imported coconuts in Honolulu. If nesting occurs in coconuts, colonization of Polynesia and Melanesia from South America may have occurred by rafting on the South Equatorial Current. Mann (1921:451) observed workers produc- ing long, worm-like stands from the anal area when the nest was disturbed. Material Examined.— SOCIETY ISLANDS: Ta- hiti, Punaauia District (J. Dixon). AMERICAN SAMOA [=E. Samoa]: Tutuila [sting; whole speci- men]; Tafuna; Alega (T. E. Woodward). WEST- ERN SAMOA: Falepuna [sting]; Matautu; Vaipoto; Poutasi (T. E. Woodward); Le Mafa; Gagaifomauga (G Ettershank); Apia (H. Swale). WALLIS IS- LANDS: Nukuione (G. Hunt) [mouthparts, sting, whole specimen]. ILES DE HORNE [=Hoorn Is- lands]: Futuna (G. Hunt). FIJI: Vanua Ava; Waiyanitu; Ovalau; Lasema; Somosomo; Nagasau; Saiaro; Munia; Nadarivatu; Labasa (W. M. Mann); Volume 3, 1 994 35 Sigatoka (W. L. and D. E. Brown) [2 stings; male genitalia]. VANUATU [=New Hebrides]: Espiritu Santo (E. O. Wilson). LOYALTY ISLANDS: Mare. SOLOMON ISLANDS: Ysabel, Fulakora [sting]; Ugi [=Uki] (W. M. Mann) [sting]. PAPUA NEW GUINEA: New Guinea, Huon Peninsula (E. O. Wilson); Wau, McAdam Park (S. Peck); Friedrich- Wilhelmshafen [=Madang] (Biro). INDONESIA: Irian Jaya, Sorong (W. L. Brown). CAROLINE ISLANDS: Ponape [=Pohnpei] (Y. Kondo). 170 workers, 9 queens, 3 males [CUIC, MCZ, MHN, USNM], Rogeria megastigmatica new species Fig. 13 Holotype and Paratype Workers. — TL 4.5-4.9 (4.5), HL ~1.01-1.09 (1.01), HW 0.90-1.00 (0.91), SL 0.77-0.83 (0.77), EL 0.16-0.19 (0.16) (41-52 facets), PW 0.62-0.70 (0.62), WL 1.18-1.30 (1.18), SpL 0.10- 0.11 (0.11), PetL 0.51-0.52 (0.52), PpetL 0.28-0.31 (0.28)mm, CI 0.88-0.92 (0.90), OI 0.18-0.19 (0.18), SI 0.83-0.87 (0.85), PSI 0.08-0.09 (0.09). N=5 Like stigmatica in most respects, but markedly larger (at least 35% larger than stigmatica on other Solomon Islands). Metanotal groove not as nar- row or sharply defined. Petiolar node strongly asymmetrical in side view (Fig. 13). Mandibles triangular, with at least 6 teeth, some basal denticles may have been abraded. Median clypeal apron convex. Mandibular carinulae effaced. Pronotum areolate with a rug- ose-areolate patch in center of disc. Dorsal face of propodeum transversely rugulose. Petiolar pe- duncle smooth; sides of node areolate; posterior face transversely rugulose-areolate or areolate- rugose. Material Examined. — Holotype locality. SOLOMON ISLANDS: Guadalcanal, Ilu Bush, 16- III-1962, #1181 (P.J. M. Greenslade) (MCZ). Paratype localities. SOLOMON ISLANDS: 1 worker, holotype locality [MCZ]; 3 workers, San Cristobal, Kira Kira, 24-IV-1962, #1579 (P. J. M. Greenslade) [MCZ]. Rogeria prominula new species Fig. 14 Additions to stigmatica-group diagnosis. WL 0.78mm. Basal angle of mandibles greatly reduced. Body of clypeus projects strongly over the trun- cate clypeal apron. Frontal region elevated and laterodorsa slightly concave. Eyes very small. Metapleural lobes and metapleural gland bulla reduced. Body almost devoid of erect hair. Holotype Worker.— TL 2.88, HL 0.66, HW 0.56, SL 0.51, EL 0.05 (4 facets), PW 0.435, WL 0.78, SpL 0.15, PetL 0.30, PpetL 0.19mm, CI 0.85, OI 0.09, SI 0.91, PSI 0.19. Mandibles with 5 teeth, basal angle greatly reduced (Fig. 14). Frontal lobes wide, closely ap- proximated. Eyes small, round, sunken; facets indistinct. Nuchal groove visible laterally as a distinct notch. Pronotal shoulders well rounded. Dorsal face of propodeum wide, weakly concave. Propodeal spiracle about 2 diameters from edge of infradental lamella. Metapleural lobes reduced to low carinae. Metapleural gland bulla small. Peti- olar node large, wider than long. Petiolar pe- duncle short, with weak keel and small, dentate inferior process. Posfpetiole subrectangular in dorsal view. Postpetiolar sternum projects shelf- like under articulation with petiole. GW/WL 0.84. Terminal segments of gaster not rotated ventrad. Shaft of undissected sting slender, with slightly enlarged apex; lancets acute and appear strong. Mandibles strongly carinulate for most of length. Frontal area impressed, smooth. Median head longitudinally rugose-areolate. Cheeks near antennal insertions microareolate. Laterodorsa, sides and posterior head densely macroareolate; cells small, usually obscuring the effaced microareola te background. Promesonotal dorsum with same areolate sculpture as on back and sides of head. Mesosoma sides strongly microareolate and macroareolate; in different places one or the other predominates. Scrobiculate sculpture in mesosoma sutures. Dorsal face of propodeum dis- tinctly microareolate, with branching rugulae crossing between propodeal spines. Peduncle and anterior face of petiole as well as sterna of petiole and postpetiole microareolate; rest of petiole and postpetiole macroareolate, somewhat effaced mesad. Gaster Tl densely covered with piligerous punctures; punctures weaker caudad. Other terga smooth except for vaguely roughened posterior margins. Most of body covered with very fine, ap- pressed pilosity. Sparse erect hairs on clypeus, frontal lobes and nearby head middorsum. A few short, decumbent hairs on mesosoma dorsum; dense erect hairs on terminal segments of gaster. 36 Journal of Hymenoptera Research No pilosity on ventral petiole. Discussion. — This is perhaps the most aber- rant Rogeria species, with its unusual head shape and pilosity. But it has the characteristic Rogeria antennae, nuchal grooves and square anteroventral corners of the pronotum. It has affinities with ciliosa, gibba, and stigmatica, as described in the stigmatica-group discussion. The name prominula, meaning little promi- nence, describes the body of the clypeus. Material Examined. — Holotype locality. BRA- ZIL: Amazonas, Ig. Mananil, Rio Branco Road, 24km NE Manaus, 22-VIII-1962, #M-2 (W. L. Brown) [MCZ]. Rogeria besucheti new species Figs. 15-16 Additions to stigmatica-group diagnosis. WL 0.67-0.75mm. Eye small (8-10 facets). Propodeal spiracles small, more than 3/4 diameter from infradental lamella. PSI 0.17-0.20. Metapleural lobes well developed. Inferior petiolar process a small step. Head, mesosoma, nodes and gaster with abundant decumbent pilosity and more sparse erect hairs. Holotyi>e and Paratype Workers. — TL 2.7-3.1 (2.85), HL 0.65-0.71 (0.66), HW 0.59-0.63 (0.60), SL 0.43-0.49 (0.44), EL 0.05-0.08 (0.06) (8-10 facets), PW 0.42-0.46 (0.43), WL 0.67-0.75 (0.69), SpL 0.12- 0.15 (0.13), PetL 0.30-0.34 (0.32), PpetL 0.16-0.18 (0.18)mm, CI 0.87-0.90 (0.90), OI 0.08-0. 13 (0.10), SI 0.72-0.78 (0.72), PSI 0.17-0.20 (0.18). N=7 Holotype mandible with 5 visible teeth de- creasing in size basad. In paratypes, mandibles always with 3 apical teeth, but basal teeth may have additional denticles or be replaced by pairs of denticles. Palpal formula 3,2. Clypeal apron truncate; body of clypeus projecting enough to block view of apron in full dorsal view. Pronotal shoulders rounded. Shallow metanotal groove shallow present on dorsum and sides of mesosoma. Propodeum lacking a distinct transverse carina at anterior border. Peduncle of petiole with weak ventral keel; inferior process reduced to a small step. Petiolar node bulbous, wider than long. Postpetiolar node widest in anterior half (as in Fig. 74). Postpetiolar sternum short, anterior lip not greatly prominent. GW/WL 0.91-0.98. Terminal segments of gaster slightly rotated ventrad, but not enough to make T3 the distalmost point of the gaster. Sting apparatus much like that of ciliosa (Fig. 3), but: 1) anterior apodeme of spiracular plate widest midlength, 2) anterodorsal corner of quadrate plate longer, narrower, 3) anterior apodeme of oblong plate longer, 4) gonostylus a little longer, with two companion setae and less distinct gap in setation and 5) sting base lower and without anterolateral processes (Fig. 16). Longitudinally rugose macrosculpture on frontal lobes becomes rugose-areolate on middorsum. Laterodorsa, sides, and posterior head areolate with rather small areolae; intervals smooth and shining, except for piligerous punctures. Pronotal disc varies from all rugose-areolate to all areolate. Rest of promesonotum slightly less coarsely areolate. Intervals smooth, except for piligerous punctures. Dorsal face of propodeum densely microareolate, with or without overlying transverse rugulae. Dorsum and anterior face of petiolar node smooth; rest of petiolar and postpetiolar nodes effaced areolate. Gaster predominantly smooth and shiny; Tl and SI with piligerous punctures that in some specimens become weaker caudad. Remaining terga and sterna very weakly rough- ened and shiny. Paraguayan specimens have suberect hair on scapes; others do not. Head dorsum with suberect hairs. Pilosity on mesosoma dorsum and nodes ranges from short and decumbent to long and erect; all curving quite strongly toward midline. Gaster Tl similar, but with no erect hairs. Termi- nal segments of gaster with rather dense erect hair and decumbent pilosity. No hair on ventral peti- ole. Extremities and mandibles light brownish yellow. Rest of body brown with more yellowish than reddish accents; frontoclypeal area and ter- minal segments of gaster lighter. Discussion. — Rogeria besucheti differs from ciliosa, gibba, prominula, and blanda in pilosity. It differs from stigmatica and megastigmatica in mesosoma shape, propodeal spine length, and generally smaller eye size. This species is named for Claude Besuchet, who as director of the MHN in Geneva was most helpful and patient in loaning material valuable for this work. Distribution. — Paraguayan specimens come from gallery forest with some bamboo. Peruvian specimens are from mixed broadleaf primary for- Volume 3, 1 994 37 est on a steep hillside at 1000m. In both localities collections resulted from Berlese and Winkler sam- pling of leaf litter and rotten wood. Material Examined. — Holotype locality. PARA- GUAY: Alto Parana Province, Puerto Santa Teresa, 3-XI-1979 (F. Baud, et al.) [MHN]. Paratype localities. COLOMBIA: 3 workers, Putumayo Department, Villa Garzon, 23-VII-1977 (D. Jackson) [BMNH], PERU: 2 workers, Pasco Department, near Pozuzo, 1000m (C. Kugler and R. R. Lambert) [mouthparts, sting] [MCZ, MZSP]. PARAGUAY: 1 worker, San Benito Province, Itapua, 29-X-1982 (F. Baud, et al.) [MHN]. Rogeria blanda Fr. Smith Figs. 17, 83-84 Myrmica blanda Fr. Smith 1858:131. Syntvpe workers, BRA- ZIL: Amazonas, Ega [=Tefe] [BMNH] [4 syntypes exam- ined]. Irogera foveaia Kempf 1964:64, Figs. 19-20. Holotype and paratype workers, BRAZIL: Amazonas, Manaus (K. Lenko) [MZSP] [Paratype examined] Rogeria blanda; Kempf 1965:185. Diagnosis. — WL0.87-l.15mm. Palpal formula 2,2. Eye rather large, oval. Propodeal spines long (PSI 0.20-0.29). Petiolar node long and low. Gaster with terminal segments rotated ventrad. Inferior petiolar process dentate. Head, mesosoma, and nodes densely areolate. Body rather densely cov- ered with long, fine, soft, erect hair; no decumbent hair. Workers.— TL 3.0-4.0, HL 0.73-0.92, HW 0.65- 0.82, SL 0.49-0.62, EL 0.12-0.16 (27-59 facets), PW 0.48-0.63, WL 0.87-1.15, SpL 0.20-0.30, PetL 0.35- 0.49, PpetL 0.18-0.25mm, CI 0.85-0.91, OI 0.18- 0.21, SI 0.73-0.80, PSI 0.20-0.29. N=20 Additions to description and figures of Kempf (1964). Mandibles triangular, with 6 large teeth decreasing in size basad, then basal tooth larger than neighbor. One or two denticles sometimes added between basal teeth. Clypeal apron with a shallow median notch. Nuchal grooves not visible in lateral view. Metanotal groove weak (Fig. 83) to absent. Propodeal spiracle small, strongly directed caudad. Propodeal spines long, straight or weakly upturned at apex; distinctly longer in the two specimens from Ecuador (PSI 0.29) than in the others (PSI 0.20-0.25). Metapleural lobes triangu- lar; apex blunt to subacute. Petiolar peduncle curved, with weak keel and dentate inferior peti- olar process. Petiolar node widest in posterior half. Postpetiolar node highest in posterior half; shape from above as in Fig. 32 or Fig. 49. Gaster not enlarged (G W/ / WL 0.63-0.70); terminal segments rotated ventrad. Pygidial gland sculpture present. Sting apparatus nearly identical to that of inermis (Fig. 42). Longitudinal rugae on frontal lobes rapidly give way to areolate sculpture usually by mid-eye level and continuing onto posterior head. Sides of head strongly sculptured: rugose-areolate in front of eye to areolate behind. Mesosoma and nodes (Figs. 83-84) also densely areolate, except for trans- verse carinulae between spines and smooth poste- rior face of propodeum. Some elongate cells on pronotal disc. Very apex of postpetiolar node some- times smooth. No microsculpture. Scapes with very long erect hairs and shorter hairs ranging from suberect to decumbent. Erect hair on terminal segments of gaster not brush-like. Legs with little decumbent or appressed pilosity. Head, mesosoma, and waist black to yellow- ish-brown; appendages lighter. Gaster darker than rest of body. Head sometimes with black patches around and/or between eyes. Queens.— TL 3.6-4.1, HL 0.79-0.87, HW 0.70- 0.78, SL 0.52-0.58, EL 0.19-0.20, PW 0.61-0.70, WL 1 .02-1 .17, SpL 0.24-0.30, PetL 0.40-0.52, PpetL 0.22- 0.25mm, CI 0.87-0.90, SI 0.73-0.81, PSI 0.22-0.26. N=8 Median pronotum sometimes transversely rugose-areolate; mesonotum longitudinally rug- ose-areolate. Otherwise, differing from the worker only in the normal queen characters. Wing vena- tion like that of belti (Fig. 37). Males.— TL 3.2-3.4, HL 0.60-0.62, HW 0.70- 0.75, SL 0.23-0.26, EL 0.32-0.35, PW 0.79-0.86, WL 1.10-1.24, PetL 0.34-0.35, PpetL 0.18-0.20mm, CI 1.16. SI 0.32-0.35. N=4 All males came from Rio Akaban, Venezuela. Mandibles with 4 subequal teeth. Anterior edge of clypeus with weak median notch. Frontal area a distinctly impressed triangle. No frontal lobes. Flagellomeres 2-11 straight, subequal in length and width. Habitus much like male of belti (Fig. 38), but junction of dorsal and posterior faces of propodeum has blunt lateral corners, and the propodeal spiracle faces more caudad. Back of head microareolate with piligerous 38 Journal of Hymenoptera Research punctures in the pits; median vertex with addi- tional longitudinal rugulae. Pronotum and mesopleura largely smooth. Mesonotum densely and finely rugulose with scattered punctures in intervals. Posterior face of propodeum, nodes, and gaster smooth. Genitalia as shown in Fig. 17. Hairs shorter, less flexuous than in worker; erect to suberect and moderately dense over much of body. Color dark brown with lighter brown ap- pendages. Discussion. — Rogeria ciliosa and gibba from low- land South America resemble blanda in having long, soft, dense pilosity and areolate sculpture, but see the stigmatica group diagnosis. Distribution. — Rogeria blanda is found in south- ern Central America and in South America east of the Andes to southern Brazil. Elevations range from 50m (Costa Rica) to 1000m (Venezuela). Nests have been found in trunks of cacao trees in Costa Rica and in a small rotten log suspended about 50cm above the ground in Peru. Material Examined.— COSTA RICA: Heredia Province, Puerto Viejo de Sarapiqui, La Selva Sta- tion (L. Garling); Santa Clara Province, Hamburg Farm (F. Nevermann); Puntarenas Province, Osa Peninsula, Corcovado (J. Longino). PANAMA: Barro Colorado Island (Brown and McCluskey; D. E. Wheeler; J. Zetek). TRINIDAD: Basin Hill Re- serve (N. A. Weber) [mouthparts, sting, whole specimen]. VENEZUELA: Falcon State, near Curimagua, Haitoncito; Monagas State, Caripe; Bolivar State, Talud, south of Amarawai Tepuii and Rio Akaban (J. Lattke) [2 male genitalia]. GUYANA [=British Guiana]: Oko River (N. A. Weber). BRAZIL: Amazonas State, Tefe; Manaus vicinity ( W. L. Brown; K. Lenko); Para State, Icoarci (W. L. Brown), Belem (N. Rosa), Jacareacanga (M. Alvarenga); Mato Grosso State, Municipal Diamantino (W. L. Brown); Espiritu Santo State, Linhares (M. Alvarenga). ECUADOR: Napo Prov- ince, Limoncocha (R. Chadab). PERU: Madre de Dios Department, Puerto Maldonado, Lake Sandoval (C. Kugler) [mouthparts, sting]. 93 work- ers, 8 queens, 4 males [BMNH, CKC, CUIC, MCZ, MIZA, MZSP, USNM]. Scandens-Group and Related Species Rogeria scandens Mann Figs. 19-20 Macrormscha scandens Mann 1922:30, Fig. 14. Syntype work- ers, HONDURAS: Lombardia (Mann) (USNM) [5 of b syntypes examined]. Irogera scandens; Kempf 1962a:436, 438. Rogeria scandens; Kempf 1965:185. Additions to scandens-group diagnosis. WL 0.93-1. 17mm. Eye very large (about 60-80 facets). Propodeal spines long (PSI 0.25-0.28). Petiole clav- ate, with rather large node (PetW/PetL 0.49-0.58). Postpetiole wide (PpetW/PpetL 1.38-1.61). Mesosoma and petiolar node strongly macro- sculptured. Erect hairs with dentate ends. Workers.— TL 3.5-4.2, HL 0.83-1.02, HW 0.70- 0.85, SL 0.59-0.73, EL 0.17-0.20 (60-80 facets), PW 0.53-0.62, WL 0.93-1.17, SpL 0.23-0.30, PetL 0.42- 0.55, PpetL 0.23-0.27mm, CI 0.82-0.86, OI 0.21- 0.24, SI 0.84-0.86, PSI 0.25-0.28. N=6 Mandible in most specimens with 5 teeth de- creasing in size until large basal tooth. A denticle may appear between basal and penultimate tooth. Nuchal grooves shallow, not visible in lateral view. Metanotal groove in lateral view broad and shal- low to absent; groove sometimes accentuated by a low transverse ridge at anterior of propodeum. Propodeal spines more strongly inclined in Hon- duran (Fig. 19) than Panamanian specimens. Metapleural lobes somewhat longer and narrower in Honduran specimens. Petiolar node slightly more distinct in Honduran specimens (Fig. 19). Sting apparatus like that of biennis (Fig. 42), except for wider anterior apodemes and a more rounded posterodorsal corners on spiracular plates and larger anterolateral processes on the sting base (pygidium and anal plate lost in preparation). Median clypeus with additional 1-2 pairs of fairly distinct carinulae lateral to the usual pair for the genus. Posterior head longitudinally rugose (continuing from middorsum), transversely arch- ing rugose, or transversely rugose-areolate. Interrugal spaces on head distinctly to weakly granular; microsculpture weaker and surface shinier on posterior. Oval area on ventral half of sides of head largely smooth and very shiny. An- terior surface and neck of pronotum smooth; rest of promesonotum with widely spaced longi- tudinal rugae, which become vermiculate on Volume 3. 1994 39 mesonotum. Rugae on Panamanian specimens not so vermiculate on mesonotum, and with lat- eral spurs. Sides of pronotum with parallel, upcurved carinae, which are more numerous and more distinct on Honduran specimens. Dorsal face of propodeum confused areolate or areolate- rugose. Intervals in mesosoma macrosculpture smooth and shiny. Petiolar node heavily areolate- rugose on sides and posterior face; sculpture weaker dorsad; no clear microsculpture. Postpetiolar node weakly rugose or rugose-ar- eolate on sides and posterior surface, becoming weaker, sometimes absent toward midline; microsculpture vaguely microareolate to nearly smooth. Scapes and extensor surfaces of legs without erect hair. Dorsa of head, mesosoma, nodes, and gaster Tl sparsely covered with fine appressed hair and longer, stiff, thick, erect-suberect hair with toothed apex (Fig. 20). Mandibles, clypeus and terminal segments of gaster with long erect hair that is tapered and less stiff. Color of Honduran specimens reddish-brown. Panamanian specimens blackish-brown. Append- ages often lighter. Queen.— TL 4.2, HL 0.95, HW 0.82, SL 0.63, EL 0.22, PW 0.69, WL 1.20, SpL 0.28, PetL 0.50, PpetL 0.30mm, CI 0.86, SI 0.76, PSI 0.23. N=l Differing from the Panamanian workers in the usual queen characteristics. Pronotum transversely rugose. Mesonotum longitudinally rugose with- out cross-connections or vermiculate appearance. Discussion. — The Honduran and Panamanian specimens may be separate species. They differ slightly in shape of mesosoma and petiole, sculp- ture, and color, and the Honduran specimens are a little larger (WL 1.13-1. 17mm) than the Panama- nian specimens (WL 0.93-1. 08mm). I prefer to call these geographic variants, however, until we have more specimens from more localities. See the terescandens description for compari- sons with its sister species. Ecology. — Some specimens from the Canal Zone were collected from Heliconia. Material Examined.— HONDURAS: Lombardia (W. M. Mann). PANAMA: Canal Zone, Barro Colo- rado Island (W. L. Brown and E. S. McCluskey; D. Wheeler; J. Zetek) [2 workers: mouthparts, whole specimen; sting]. 31 workers, 1 queen [CKC, LACM, MCZ, USNM]. Rogeria terescandens new species Fig. 21 Diagnosis. — Like that of scandens, except: 1) Propodeal spines shorter, 2) petiolar node lower, more slender (PetW/PetL0.40-0.41)/3)postpetiole narrower (PpetW/PpetL 1.08-1.16), and 4) macrosculpture very weak, especially on mesosoma and waist, and 5) eye slightly smaller. Holotype and Paratype Workers.— TL 3.6, HL (0.88)-0.89, HW (0.68)-0.70, SL (0.70)-0.71, EL 0.16 (49-54 facets), PW (0.50)-0.52, WL 1.00, SpL (0.18)- 0.21, PetL (0.45)-0.47, PpetL 0.25mm, CI (0.77)- 0.79, OI 0.23-(0.24), SI 1.01-(1.03), PSI (0.18)-0.21. N=2. Also differing from scandens in the following ways. Mandible with 6 teeth. Profile of mesosoma dorsum almost evenly convex; no metanotal groove or ridge at front of propodeum. Median carinulae on clypeus weak; no lateral carinulae. Head dorsum densely microareolate; overlain on middorsum by wisps of of longitudinal rugulae, on laterodorsa by faint reticulations, and on poste- rior of head by fragmented, transversely arching rugulae. Microsculpture on sides and posterior head effaced. Mesosoma dorsum, including dor- sal face of propodeum, densely microareolate, with superimposed patches of fine rugulose-ar- eolate macrosculpture. Pronotal sides shiny and coriarious with some effaced longitudinal rugulae. Mesopleura and metapleura also shiny near coxae, but more opaque dorsad with microareolate and confused rugulose sculpture. Sides of petiolar node shiny, with effaced microareolate background and vestigial longitudinal rugulae. Postpetiole smooth and shiny. Discussion. — The name of this species refers to sculpture like that of R. scandens, but smoother, as if rubbed (teres L., rubbed off). Distribution. — Both specimens of terescandens were taken from trees in lowland forest on the Pacific side of Costa Rica. The holotype was col- lected in a two week old treefall by general collect- ing on trunks. The paratype was on or beneath a thick epiphyte mat on the base of a fallen branch (J. Longino unpublished field notes). Material Examined. — Holotype locality. COSTA RICA: Osa Peninsula, Sirena, 8.28N 83.35W, 50m, 31-111-1982, #0950 (J. Longino) [MCZ]. Paratype locality. 1 worker, holotype locality, 28- V-1981, #1100 (J. Longino) [LACM]. 40 Journal of Hymenoptera Research Rogeria subarmata Kempf Figs. 22-23 Irogera subarmata Kempf 1962a:438, Figs. 1-4. Holotype and paratype workers, BRAZIL: Guanabara, Rio de Janeiro, Deodoro (A. Ronna) [MZSP] [12 of 38 paratypes exam- ined; holotype not examined]. Rogeria subarmata; Kempf 1965:185. Rogeria subarmata; Kempf 1975:367 [new records]. Additions to scandens-group diagnosis. WL 0.78-1. 00mm. Eye with 30-53 facets. Propodeal spines short (PSI 0.09-0.12). Pygidium with a pair of median piligerous tubercles near caudal mar- gin. Strong macrosculpture on mesosoma and petiolar node. Erect hairs not as rigid as in scandens; tips acute. Workers.— TL 2.9-3.7, HL 0.69-0.87, HW 0.60- 0.75, SL 0.46-0.57, EL 0.12-0.16 (30-53 facets), PW 0.45-0.57, WL 0.78-1.00, SpL 0.08-0.10, PetL 0.37- 0.47, PpetL 0.20-0.26mm, CI 0.86-0.89, OI 0.20- 0.23, SI 0.73-0.77, PSI 0.09-0.12. N=6 Additions to Kempf's (1962a) description. Mandibles usually with 5 teeth that decrease in size basad. Sometimes basal tooth replaced by two very small teeth, or 1-2 denticles are found be- tween the basal and penultimate tooth. Clypeal apron weakly notched medially to evenly convex. Frontal lobes narrow as in scandens (Fig. 19). Nuchal grooves shallow, forming only a weak notch in lateral view. Figs. 22-23 show the range of propodeal spine size and angle, but tips some- times more rounded. Petiole clavate to rather dis- tinctly set off from peduncles (Figs. 22-23). Postpetiole from above much as in Fig. 21. Poste- rior surface of pygidium with a caudal pair of long, columnar, piligerous tubercles that are vis- ible at 50X with a dissection microscope. Sting apparatus nearly identical to that of inerniis (Fig. 42); sting as in pellecta (Fig. 33). Median clypeus with 1-2 pair of fairly distinct extra carinulae lateral to the usual pair. Posterior head with transversely arching rugose-areolate macrosculpture. Head covered with dense, indis- tinctly microareolate roughening that appears punctate or granular at lower magnifications. Mesosoma dorsum longitudinally rugose; rugae with numerous lateral spurs that occasionally con- nect rugae on shoulders. Macrosculpture on sides of mesosoma and dorsal face of propodeum con- fusedly rugose to rugose-areolate. Mesosoma microsculpture as on head. Petiolar node ver- miculate-rugose to rugose-areolate. Postpetiolar node similar, but rugae straighter, more effaced. Microsculpture on nodes slightly weaker than on head and mesosoma. Scapes and extensor surfaces of legs lack erect hair. Rest of body with both short, appressed- decumbent and longer, erect-suberect hairs. Erect hairs are nearly as stiff as those of scandens and terescandens (Fig. 20), but seem to have acute tips. Color dark brown to yellowish-brown with a reddish tint on mesosoma, waist and middle of gaster; appendages and ends of gaster lighter. Distribution. — All available specimens are from localities along the coast of Brazil. The type series was collected from the stomach of an anteater (Tamandua tetradactyla). Material Examined. — BRAZIL: Para State, Belem (N. Rosa) [mouthparts, sting]; Bahia State, Itabuna (J. A. Winder); Espirito Santo State, Pedro Canario near Conceicao da Barra (M. Alvarenga); Guanabara State [=Rio de Janeiro State], Rio de Janeiro, Deodoro (A. Ronna). 20 workers [MCZ, MZSP]. Rogeria procera Emery Fig. 18 Rogeria procera Emery 1896:92, Fig. 19. Holotype worker, BRAZIL: Para, Belem [MCSN] [Holotype examined]. Rogeria {Irogera) procera; Emery 1915:191. Macromischa brasiliensis Borgmeier 1953(1951 ):107, Figs. 1-4. Holotype worker, BRAZIL: Para, RioCumina, Cachoeira do Breu [MZSP] [Holotype examined). [Synonymy by Kempf 1962a:437], Irogera procera; Kempf 1962a:436 [partial description]. Irogera procera; Kempf 1964:66 [partial key]. Rogeria procera; Kempf 1965:185. Diagnosis. — WL 1.28-1. 53mm. Eye very large. Propodeal spines long, not inclined. Postpetiolar node small, subconical; sternum inconspicuous and without a differentiated peduncle. Middorsum and posterior head with fine, nearly straight longi- tudinal rugae. Mesosoma and petiolar node with thick, vermiculate rugae. Erect hairs tapered; rarely any appressed or decumbent hairs on mesosoma dorsum, gaster, or legs. Scapes and extensor sur- faces of legs with erect hairs and appressed to decumbent pilosity. Palpal formula, propodeal spiracle, metapleural lobes, petiole, pygidium, and sting apparatus as in scandens-group diagnosis. Workers.— TL 4.5-5.4, HL 1.07-1.19, HW 0.91- 1.08, SL 0.66-0.78, EL 0.19-0.23 (about 80-100 fac- Volume 3, 1 994 41 ets), PW 0.67-0.80, WL 1.28-1.53, SpL 0.26-0.32, PetL 0.60-0.73, PpetL 0.28-0.33mm, CI 0.85-0.91, OI 0.20-0.22, SI 0.70-0.75, PSI 0.18-0.23. N=7 Mandibles triangular with 6 teeth; basal larger than neighbor. Clypeal apron with shallow me- dian notch. Nuchal grooves shallow, not visible in lateral view. Sting apparatus like that of inermis (Fig. 42), but spiracular plate with more rounded anteroventral corner and gonostylus with sepa- rate proximal and distal patches of sensilla. Laterodorsa of head longitudinally rugose- areolate; sides below eye smooth and shiny. Microsculpture vaguely microareolate on anterior laterodorsa; more effaced on rest of head, impart- ing a vaguely granular, shiny appearance between rugae. Promesonotum longitudinally vermiculate- rugose. Meso- and metapleura rugose. Interrugal spaces on mesosoma and petiole almost smooth. Postpetiolar node shiny with vague rugae and weak microsculpture. Scapes and head dorsum with erect hair and much shorter appressed to decumbent pilosity. Mesosoma, legs, waist, and gaster with erect hairs, but rarely any decumbent or appressed hairs. Mesosoma and waist black to dark brown; appendages and sometimes gaster lighter brown. Material Examined.— GUYANA (British Guiana): Oronoque River (N. A. Weber). BRAZIL: Para State, Rio Cumina (A. Sampaio), Ourem (A. Schulz); Amazonas State, Manaus (K. Lenko); Manaus to Itacoatiara (W. L. Brown) [mouthparts, sting]. 23 workers [CUIC, MCZ, MZSP]. Rogeria tonduzi Forel Figs. 85-86 Rogerm tonduzi Forel 1899:53. Holotype worker, COSTA RICA (Tonduz) (MHN) [Holotype examined). Irogera tonduzi, Emery 1915:191. Irogera tonduzi; Kempf 1962a:436. Rogeria tonduzi; Kempf 1965:185. Diagnosis.— WL 0.81-0.90mm. Eye large. Pal- pal formula 2,2. Propodeal spiracle faces laterally. Propodeal spines long. Postpetiolar sternum not enlarged. Posterior head with transversely arch- ing rugae. Rugae on mesosoma and petiolar node thick and rounded. Decumbent hair abundant on head dorsum and legs; little on gaster; no decum- bent or appressed hair on mesosoma or nodes. Scapes, head dorsum, mesosoma, nodes and gaster with abundant flexible, tapered, erect hair; none on extensor surfaces of legs. The following as in scandens- group diagnosis: metapleural lobes, peti- ole, pygidial gland sculpture, sting apparatus, and sculpture. Workers.— TL 3.0-3.2, HL 0.72-0.78, HW 0.61- 0.68, SL 0.48-0.53, EL 0.12-0.15 (39-48 facets), PW 0.46-0.50, WL 0.81-0.90, SpL 0.18-0.21, PetL 0.37- 0.40, PpetL 0.17-0.19mm, CI 0.85-0.88, OI 0.19- 0.24, SI 0.78-0.81, PSI 0.20-0.25. N=7 Mandibles triangular; most specimens with 6 teeth, the first 5 decreasing in size basad then a large basal tooth. In others, the penultimate basal is replaced by 2-3 denticles. Clypeal apron convex with median angle or small tooth. Body of clypeus rises perpendicularly. Eyes oval with narrow an- terior point. Nuchal groove inconspicuous. Pronotum in lateral view lacks a distinct angle between anterior and dorsal faces. Metanotal groove very weak to absent. Propodeal spines long, weakly inclined. Postpetiole subtrapezoidal from above; sternum flat, with a distinct peduncle. Sting apparatus like that of inermis (Fig. 42) except for more elongate anterolateral processes on sting bulb. Middorsum of head longitudinally rugose; laterodorsa and dorsal part of sides rugose-ar- eolate. Mesosoma (Figs. 85-86) with thicker rugae and narrower interrugal spaces than on head. Rugae transverse on anterior pronotum, trans- verse to confused on metanotum and dorsal face of propodeum, predominantly longitudinal on sides and pronotal disc. Petiolar node longitudi- nally rugose on sides; smooth along midline. Postpetiole smooth. Microsculpture weak or ab- sent throughout, integument very shiny. Scapes and head with erect to suberect hair along with the typical short decumbent pilosity. Mesosoma dorsum and waist generally with erect to suberect hair of a variety of lengths; no decum- bent-appressed pilosity. Gaster with long, erect hair and a few short, decumbent hairs. Color shiny black with dark brown mandibles, scapes and legs to reddish-brown with vellowish- brown appendages. Discussion. — Rogeria belti occurs in the same localities and could be confused with tonduzi, but belti has a distinct petiolar node and predomi- nantly areolate pronotal sculpture. Rogeria lirata (Figs. 28-30, 89-90) from northern South America is similar to tonduzi in size, clypeus, and sculpture, but lirata has smaller eyes (6-12 facets), a distinct 42 Journal of Hymenoptera Research petiolar node, and scapes with suberect hair only. Distribution. — With the exception of one speci- men from Guatemala, all tonduzi specimens come from Costa Rica at elevations of 0-100m on both sides of the cordillera. Most specimens were col- lected by Jack Longino as strays on ground and vegetation. He found one worker in a Cyphomyrmex nest and another "...on the base of a small tree, amongst some Pheidole workers" (unpublished field notes). Lyn Garling found a nest with a "tubular entrance with white 'fuzz'" in a cacao tree (field note on label). Material examined. GUATEMALA (no locality or collector). COSTA RICA: Heredia Province, Puerto Viejo de Sarapiqui, La Selva Station (L. Garling); Puntarenas Province Carara Biological Reserve (P. S. Ward), Osa Peninsula, Corcovado National Park (J. Longino) [mouthparts, sting]. 24 workers [BMNH, CKC, JTLC, LACM, MCZ, MZSP, USNM]. Gerwaim-Group and Related Species Rogeria germaini Emery Figs. 24-27, 87-88 Rogeria germaini Emery 1894:189. Syntype workers, BRAZIL: MatoGrosso (Germain) [MCSN] [1 of 2 syntypes exam- ined] Rogeria germaini minensis Santschi 1923:1262. Lectotype and paratype workers, BRAZIL: MinasGerais, PassaQuatro (Reichensperger) [NMB] [Both lectotype and paratype examined]. N. syn. Rogeria germaini; Kempf 1962b:20, Figs. 18, 19 [Redescnbed] Rogeria minensis; Kempf 1963:189, Figs. 1, 2 [Redescnbed, raised to species]. Additions to gemiaini-group diagnosis. WL 0.74-0. 90mm. Clypeal apron with median tooth. Metanotal groove very weak or absent. Metapleural lobes reduced, not angular. Petiolar keel not lamel- late. Sting shaft apex weak, lacks dorsal flange; lancets spatulate. Macrosculpture effaced on side of head below eye; sometimes nearly smooth. Promesonotal rugae sharp and narrow like those on head; rugae low on sides of pronotum do not continue onto me tanotum. Macrosculpture on both nodes and petiolar peduncle vestigial. Scapes with- out erect hair. Mesosoma with erect-suberect only. Head and gaster with both erect and decumbent pilosity. Workers.— TL 2.7-3.5, HL 0.66-0.85, HW 0.59- 0.74, SL 0.45-0.59, EL 0.06-0.10 (7-13 facets), PW 0.44-0.51, WL 0.74-0.90, SpL 0.15-0.21, PetL 0.30- 0.44, PpetL 0.19-0.23mm, CI 0.86-0.89, OI 0.09- 0.14, SI 0.75-0.80, PSI 0.20-0.23. N=9 Additions to Kempf (1962b). Dentition vari- able; simplest pattern is 6 teeth of decreasing size, however, it seems that any or all of the last 3 teeth may be replaced by a pair of denticles. FLW/HW 0.32-0.36. Posterior outline of head concave to convex. Anterior propodeum marked by a trans- verse carina. Petiolar node distinct; Figs. 24 and 26 show extremes of shape. Postpetiolar node with a distinct posterior face; shape from above as in Figs. 24 or 32. Sting apparatus differing dramati- cally from that of inermis (Fig. 42) in some features: 1) sting shaft and lancets very weak, 2) lancets spatulate as in Fig. 29, 3) sting shaft apex with eroded sides, and no dorsal flange (Fig. 27), and 4) furcula with a shorter dorsal arm that broadly merges with the lateral arms, thus appearing broadly V-shaped in anterior view. Posterior head transversely arching rugose to rugose-areolate. Promesonotal dorsum varies from predominantly areolate (with occasional elongate cells) (Fig. 88) to predominantly vermiculate-rug- ose. Pronotal sides rugose; rugae subparallel with ventral edge of pronotum to diagonal. Microsculpture vestigial, leaving irregular, but shiny spaces in macrosculpture; sides of mesosoma especially smooth. Color brownish-yellow to brown with darker gaster; appendages at times slightly lighter. Queen.— TL 4.1, HL 0.90, HW 0.79, SL 0.73 EL 0.17, PW 0.67, WL 1.12, SpL 0.15, PetL 0.48, PpetL 0.28mm, CI 0.88, SI 0.92, PSI 0.13. N=l Differing from the worker in the normal queen characteristics and the following. Notch formed by nuchal groove not so distinct in lateral view. Anteroventral corner of pronotum not as clearly dentate. Parapsidal furrows not distinguishable from furrows in sculpture. Diverging rugae on middorsum of head continue onto posterior head, with few or no lateral spurs. Laterodorsa and sides of head similarly rugose. Anterior face of prono- tum transversely areolate, mesonotum with longi- tudinal rugae diverging from an anterior point, then parallel in posterior half; some branching, but no cross-ridges; less vermiculate than in worker. Discussion. — Kempf (1963) examined the types of germaini and minensis and noted that they were very similar, but chose to retain both names on the basis of a list of differences he saw in those speci- Volume 3. 1994 43 mens. I have also examined the types as well as 1 1 specimens collected in Paraguay in 1979 and 1982. These new collections bridge the gap between the types oigermaini and minensis. They are all inter- mediate in size between the smaller germaini type and the larger minensis types, have convex poste- rior heads like ihegermain i type, have promesonotal sculpture varying from nearly as extensively rug- ose as thege rmaini type, to areolate like the minensis types; and some have petiolar nodes intermediate between the more abruptly arising tninensis-hke and the more obliquely arisingger??!fl;>n-like nodes. Rogeria lirata (Figs. 28-30, 89-90) from more northern parts of South America is germaini' s clos- est relative (See lirata description for compari- sons). Rogeria lacertosa (Fig. 31), also from southern Brazil, differs in size, clypeal shape, sculpture on side of head, and pilosity . Rogeria pellecta (Figs. 32- 33), collected further south in Brazil, differs in clypeal shape, metapleural lobes, sting, and pilos- ity. Distribution. — So far germaini is known only from southern Brazil and Paraguay. Most speci- mens have come from Winkler apparatus collect- ing by expeditions from the MHN in Geneva. Specimens were extracted from rotting leaf litter and wood in forests. Material Examined. — BRAZIL: Minas Gerais State, Passa Quatro (Reichensperger); Mato Grosso State; Sao Paulo State, Anhembi [=Piramboia, 29km E Botucatu] (Kempf et al.). PARAGUAY: Alto Parana Province, Puerto Santa Teresa; Misiones Province, Panchito Lopez; Itapua Province, San Benito Island [2 workers: whole specimen; mouth- parts, sting]; Itapua Province, Santa Maria; Cen- tral Province, Asuncion (F. Baud et al.). 15 work- ers, 1 queen [BMNH, CKC, MCSN, MCZ, MHN, NMB], Rogeria lirata new species Figs. 28-30, 89-90 Diagnosis. — As in germainibut: 1) metapleural lobes larger, triangular, and 2) rugae on promesonorum more rounded and thicker than on head; one ruga begins near anteroventral cor- ner of pronotum and continues unbroken to the metanotum. Holotype and Paratype Workers. — TL 2.6-3.4 (3.3), HL 0.63-0.80 (0.78), HW 0.53-0.70 (0.67), SL 0.45-0.56 (0.55), EL 0.05-0.08 (0.08) (6-12 facets), PW 0.39-0.51 (0.51), WL 0.65-0.87 (0.86), SpL 0.12- 0.20 (0.175), PetL 0.31-0.47 (0.44), PpetL 0.17-0.23 (0.21)mm, CI 0.83-0.88 (0.86), OI 0.09-0.12 (0.12), SI 0.79-0.85 (0.82), PSI 0.19-0.23 (0.20). N=10 Dentition as in germaini. Nuchal groove forms a weak notch in lateral view. No clear angle be- tween anterior and dorsal faces of pronotum. Spi- racle faces slightly caudad, posterior edge within one diameter of nearest edge of propodeum. Peti- olar peduncle with a sharp, but not lamellate keel. Petiolar node profile angular. Postpetiolar node in dorsal view shaped as in Figs. 24 or 49; sternum flat. Sting apparatus as described for germaini. Divergent longitudinal rugae on frontal lobes grade into areolate sculpture at level of eyes. Sculp- ture of posterior head transversely arching rugose to rugose-areolate. Laterodorsa and sides of head longitudinally rugose-areolate; sides may be ef- faced to nearly smooth. Head microsculpture ves- tigial; intervals nearly smooth, quite shiny. Ante- rior face of pronotum areolate, becoming ver- miculate-rugose on disc (Fig. 90). Thick, rounded rugae of promesonotum may merge, but are rarely joined by cross-ridges. Rugae on meso- and metapleura sharper and more separated. Anterior and dorsal faces of petiolar node smooth. Sides and posterior faces of node in holotype longitudi- nally rugose-areolate, but weaker and more rug- ose in Trinidad specimen and nearly smooth in specimens from Peru. Microsculpture weak or absent; integument shiny. Scapes with decumbent to suberect hair (holo- type) or with short, uniformly subdecumbent hair. Head dorsum with subdecumbent to erect hair. In most, mandibles, frontoclypeal region, an- tennae and legs brownish-yellow; other parts rusty- brown, becoming darker, almost black, on dorsa of mesosoma and petiole. Trinidad and Guyana specimens lighter on all parts. Queens.— TL 3.6-3.7, HL 0.76-0.80, HW 0.69- 0.70, SL 0.52-0.57, EL 0.14-0.15, PW 0.58-0.60, WL 0.95-1 .00, SpL 0.18-0.20, PetL 0.46-0.52, PpetL 0.23- 0.24mm, CI 0.88-0.90, SI 0.76-0.81, PSI 0.18-0.21. N=4. Besides the usual features of an alate female, differing from the worker in the following ways. Mandible with 7 teeth or 3 teeth and 5 denticles. Parapsidal furrows not distinct from furrows in sculpture. Propodeal spines project caudad or posteroventrad. Rugae on sides of pronotum al- most vertical to shoulder, then bend across the 44 Journal of Hymenoptera Research anterior face of the pronotum. Mesonotal rugae longitudinal with occasional fusions and cross- ridges. Wing venation (Fig. 30) most similar to that of stigmatica. Radial sector and median veins long, nearly reaching wing margin; r-m vein present. Discussion. — See the tonduzi discussion for comparisons with that species. The name lirata refers to the characteristic rugae of the mesosoma, which resemble the ridges thrown up by a plow. Material Examined. — Holotype locality. CO- LOMBIA: 7km N Leticia, forest litter, 10-25-11- 1972, #B-230 (S. and J. Peck) [MCZ]. Paratype localities. TRINIDAD: 3 workers, #191 (N. A. Weber) [MCZ, MZSP]; 1 worker, Mayaro, Trinity Hills Reserve, 5-VIII-1976 (J. Noyes) [BMHN]. GUYANA (British Guiana): 1 worker, R. Mazaruni Forest Settlement, 20-VIII-1935, #304 (N. A. Weber) [MCZ]. COLOMBIA: 2 workers, holotype locality [mouthparts, sting; one coated for electron microscopy] [CKC, MCZ]. PERU: 2 workers, Loreto Department, 15km WSW Yurimaguas, 5.59S 76.13W, 200m, 22-VII-1986, #8701-24 and #8701-25 (P. S. Ward) [mouthparts, sting] [LACM, MCZ]. Nontype localities. TRINIDAD: 1 queen, Nariva Swamp, 23-IV-1935, #140 (N. A. Weber) [MCZ]. BRAZIL: 3 queens, Mato Grosso State, Sinop, 55.37W 12.31S, X-1974 #12314 (Alvarenga), #12576 (Alvarenga and Roppa) [MZSP]; 2 queens, Goias State, Jatai, XI-XII-1972, #8857, #8939 (F. M. Oliveira) [MZSP]. Rogeria lacertosa Kempf Fig. 31 Rogeria lacertosa Kempf 1963:194, Figs 5-6. Holotype and paratype workers, BRAZIL: Rio Grande do Sul State, Sinimbu (F. Plaumann) [MZSP] [All 4 paratypes exam- ined; holotype not examined]. Additions to germaini-group diagnosis. WL 0.93-1. 05mm. Clypeal apron with median notch. Metanotal groove weak to strong. Petiolar keel with single lamellate carina. Sting shaft apex has dorsal flange; lancets acute; both weak. Macrosculprure not effaced on side of head below eye. Promesonotal rugae sharp and narrow like those on head. Sides of both nodes distinctly macroareolate. Erect hairs on scapes. No decum- bent hair on gaster; little if any on mesosoma dorsum. Workers.— TL 3.6-4.0, HL 0.83-0.91, HW 0.72- 0.81, SL 0.59-0.61, EL 0.10-0.11 (19-20 facets), PW 0.52-0.60, WL 0.93-1.05, SpL 0.17-0.20, PetL 0.43- 0.45, PpetL 0.23-0.26mm, CI 0.87-0.88, OI 0.13- 0.14, SI 0.75-0.79, PSI 0.18-0.20. N=4 The following supplements Kempf (1963). All specimens at hand with 6 mandibular teeth de- creasing in size basad, except for a large basal tooth. Eyes elliptical. Nuchal groove forms a notch in in lateral view of head. Anterior edge of propodeum not marked by a transverse carina. Metapleural lobes low and broadly rounded or triangular with more narrowly rounded apex. Postpetiolar node from above like that of pellecta (Fig. 32). Sting apparatus as biennis (Fig. 42), but sting shaft and lancets are less sclerotized (easily twisted) and the lancets lack the barbule. Laterodorsa of head predominantly rugose to rugose-areolate. Back of head areolate in a trans- versely arched pattern. Microsculpture vestigial; intervals with a shiny, effaced granular appear- ance. Anterior face of pronotum transversely ar- eolate; disc with diverging, longitudinal, vermicu- late rugae with variable number of cross-ridges imparting a rugose-areolate appearance in places. Sides of mesosoma predominantly longitudinally rugose, but with occasional cross-ridges making elongate cells. Intervals in mesosoma macrosculprure shiny, nearly smooth, especially on sides. Anterior and sometimes dorsal faces of nodes weakly sculptured. Nodes slightly dulled by vestigial microsculprure. Head dorsum with erect-suberect hairs in addition to the typical decumbent pilosity. Color yellowish-brown; gaster slightly darker. Legs and, sometimes, antennae lighter, more yel- lowish. Mandibles often slightly darker than head capsule. Discussion. — Because minensis has been syn- onymized with gertnaini, some of Kempf's (1963) list of characters that distinguish lacertosa are no longer valid, however, gertnaini and lacertosa are still distinguishable on the basis of a number of characters (see gertnaini discussion). Rogeria pellecta (Figs. 32-33), also from southern Brazil, differs in petiole keel, sting and lancets, promesonotal sculp- ture, and pilosity. Distribution. — Rogeria lacertosa is known only from the type material collected from 100-200m elevation in southern Brazil. No ecological data are available. Material Examined. — BRAZIL: Rio Grande do Volume 3. 1 994 45 Sul State, Pardinho (F. Plaumann), Sinimbu (F. Plaumann) [mouthparts, sting]. 4 workers [MZSP]. Rogeria pellecta Kempf Figs. 32-33 Rogeria pellecta Kempf 1963:191, Figs. 3-4. Holotype worker, BRAZIL: Santa Catanna State, Nova Teutonia (F. Plaumann) [MCZ, MZSP] [6 of 28 paratype workers examined, including 4 from holotype locality; holotype not examined] Diagnosis.— WL 0.90-0.99mm. Clypeus with median notch. Metanotal groove distinct. Metapleural lobes prominent, triangular. Petiolar node arises gradually from peduncle. Sting and lancets strong, acute; sting shaft with dorsal flange; lancet with barbule. Laterodorsa of head longitu- dinally rugose. Promesonotum coarsely areolate to rugose-areolate. Back of petiolar node strongly areolate; postpetiole vestigially areolate. Scapes lack erect hair. Dorsa of head, mesosoma, nodes and gaster with erect and decumbent pilosity. Mandibles, palpal formula, eye, propodeal spines, postpetiole, other aspects of petiole, sculpture, and pilosity as in germaini-group diagnosis. Workers.— TL 3.4-3.7, HL 0.81-0.89, HW 0.71- 0.78, SL 0.55-0.60, EL 0.09-0.11 (16-21 facets), PW 0.50-0.55, WL 0.90-0.99, SpL 0.17-0.21, PetL 0.39- 0.46, PpetL 0.20-0.22mm, CI 0.87-0.88, OI 0.12- 0.14, SI 0.77-0.79, PSI 0.19-0.22. N=6 The following supplements Kempf (1963). Palpal formula 3,2. Anterior edge of clypeus weakly emarginate. Eyes oval. Metanotal groove may be bordered behind by a transverse costa. Posterior face of petiolar node vertical or slightly concave in lateral view. Postpetiole widest anteriorly, as in Fig. 32, or evenly convex, as in Fig. 53. Sting apparatus like that of inermis (Fig. 42), except for the sting (Fig. 33), which has larger anterolateral processes on sting base, a stronger, relatively thicker sting shaft, and a lower dorsal flange. Posterior head sculpture is rugose-areolate, with rugae longitudinally diverging or transversely arching across back of head. Rugae of laterodorsa sometimes broken and with lateral spurs. Sides of head areolate around eye, but effaced and nearly smooth ventrad. Dorsal face of propodeum trans- versely carinulate to coarsely areolate. Petiolar node with areolate macrosculpture that becomes weaker and less defined anteriorly. Ventral peti- ole with a pair of longitudinal carinae arising from an anterior keel. Head and mesosoma microsculpture vestigial, leaving nearly smooth, shiny intervals in macrosculpture. Sides of peti- olar peduncle microareolate; nodes roughened by obscure microsculpture and not as shiny as head and mesosoma. Color yellowish-brown; gaster slightly darker. Legs, and sometimes antennae, lighter, more yel- lowish. Mandibles often slightly darker than head capsule. Gynecoid Workers. — As described by Kempf (1963). Discussion. — In the southern Brazil /Paraguay area occur related species germaini, lacertosa, and sicaria. Comparisons with these species are found in the "Species Groups" section and in the germaini and lacertosa discussions. Distribution. — Rogeria pellecta is known only from the 33 type specimens taken in southern Brazil from berlesate of leaf litter. Material Examined.— BRAZIL: Santa Catarina State, Nova Teutonia (F. Plaumann) [mouthparts, sting]. 6 workers, 1 gynecoid [MCZ, MZSP]. Rogeria sicaria Kempf Fig. 34 Rogeria sicaria Kempf 1962b:22, Figs. 20, 21 . Holotype worker, BRAZIL: Sao Paulo, Agudos [19km SSE Bauru] (W W. Kempf) [MZSP] [Holotype examined]. Diagnosis. — WL 0.75mm. Eye very small. Clypeal apron convex. Metanotal groove strong. Propodeal spines very long, strongly inclined dor- sad and divergent. Petiole long (PetL/PW 0.95), with prominent keel. Sting apex strong, with dor- sal flange. Sides of head below eye, mesosoma, and petiolar node areolate; coarsely so on mesosoma. No erect hair on scapes. Head, mesosoma dorsum, nodes, and gaster Tl with stiff, erect-suberect hairs and shorter, more abun- dant, appressed-decumbent hair. Holotype Worker.— TL 2.8, HL 0.67, HW 0.56, SL 0.48, EL 0.03 (8 facets), PW 0.37, WL 0.75, SpL 0.22, PetL 0.35, PpetL 0.17mm, CI 0.83, OI 0.05, SI 0.86, PSI 0.29. The following supplements Kempf (1962b). Mandible triangular, with 5 teeth decreasing in size to base. Clypeal apron evenly convex; body of clypeus projecting over mandibles. Frontal lobes narrow (FLW/HW 0.29). Eyes small, circular, somewhat sunken in head. Petiolar peduncle long (PetL/PW 0.95), with strong keel and dentate 4h Journal of Hymenoptera Research inferior process. Postpetiole highest in caudal half; dorsal view subtrapezoidal as in Fig. 49. Posterior head areolate and more coarse than elsewhere on head. Intervals shiny, but dulled somewhat by vestigial microsculpture. Areolate sculpture on petiolar node not as well defined as on head and mesosoma and even more effaced on postpetiolar node. Nodes roughened by vague microsculpture. Color uniformly golden-brown. Discussion. — This species is still known only from the holotype. It seems related to germaini, pellecta, and lacertosa, also from southern Brazil. See the "Species Groups" section for comparisons. Creightoni-Group and Related Species Rogeria merenbergiana new species Figs. 46-48 Additions to creightoni-group diagnosis. WL 0.69-0. 83mm. Palpal formula 3,2. Clypeal apron with median concavity. Eye small, elliptical; EL/ SpL 0.67-0.86. Anterior edge of pronotal disc with- out a strong transverse carina. MHI 0.91-1.01. Promesonotum and metanotum with distinct pro- files— promesonotum convex, metanotum usu- ally flat. Propodeal spines short, straight. Postpetiole from above subtrapezoidal. Mesosoma and sides of head rugose (sides of head sometimes effaced rugose-areolate). Erect hair on scapes and middle and hind tibiae. Holotype and Paratype Workers. — TL 2.7-3.1 (3.0), HL 0.69-0.71 (0.71), HW 0.57-0.62 (0.61), SL 0.46-0.52 (0.50), EL 0.07-0.09 (0.08) (8-14 facets), PW 0.42-0.47 (0.45), WL 0.74-0.83 (0.80), SpL 0.10- 0.14 (0.12), PetL 0.28-0.36 (0.335), PpetL 0.15-0.18 (0.16)mm, CI 0.83-0.88 (0.85), OI 0. 13-0. 15 (0.13), SI 0.80-0.84 (0.83), PSI 0.13-0.17 (0.15), MHI 0.91-1.00 (0.93). N=6 Mandibles with 6-7 teeth decreasing in size basad and basal tooth as large or larger than penultimate basal. If 6 teeth, may have 1-2 addi- tional denticles among basal teeth. Body of clypeus not strongly produced. Nuchal groove weak. Pos- terior outline of head evenly convex. Propodeal spiracle one diameter, or slightly less, from edge of infradental lamella. Metapleural lobes low; edges perpendicular. Petiolar node with a more or less distinct angle between dorsal and posterior faces. Petiolar peduncle with weak keel. Postpetiole with longer anterior face and short posterior face meet- ing at a narrow apex. Postpetiolar node from above as in utiguispina (Fig. 49). Postpetiolar ster- num short, weakly concave; anterior edge square. Laterodorsa of head longitudinally rugose- areolate. Posterior head with rugae diverging from midline or transversely arching; few if any cross ridges. Sides of head longitudinally rugose. Ante- rior pronotal disc transversely rugose to rugose- areolate; rest of promesonotal dorsum longitudi- nally rugose. Dorsal face of propodeum areolate- rugose. Lateral mesosoma rugose; less regularly so on meso- and metapleura. Microsculpture on head and mesosoma effaced; intervals quite smooth and shiny, especially on sides and back of head. Petiole distinctly microareolate; sides of node vaguely rugose to vaguely areolate; posterior face with strong ridges. Postpetiole effaced microareolate and very sparsely and vaguely rug- ose. Color very dark reddish brown to yellowish brown; appendages and often sides of mesosoma and gaster lighter. Nontype Workers.— TL 2.6-3.0, HL 0.63-0.71, HW 0.53-0.60, SL 0.43-0.49, EL 0.08-0.09 (12-15 facets), PW 0.40-0.43, WL 0.69-0.77, SpL 0.10-0.13, PetL 0.29-0.33, PpetL 0.16-0.18mm, CI 0.84-0.85, OI 0.14-0.15, SI 0.79-0.82, PSI 0.14-0.17, MHI 0.95- 1.01. N=3 Little difference from types. Some slightly smaller, but with slightly larger eye for size. Gen- erally higher mesosoma height index. Posterior head sometimes transversely arching rugose-ar- eolate. Sides of head rugose-areolate, but effaced posterior to eye. Promesonotal dorsum longitudi- nally rugose to rugose-areolate. Lateral mesosoma rugose-areolate. Paratype And Nontype Queens. — TL 3.3-3.7, HL 0.70-0.75," HW 0.61-0.64, SL 0.48-0.52, EL 0.16-0.17, PW 0.58-0.60, WL 0.98-1.15, SpL 0.15-0.17, PetL 0.34-0.39, PpetL 0. 19-O.20mm, CI 0.86-0.88, SI 0.78- 0.80, PSI 0.14-0.16, MHI 1.13-1.19. N=3 Mandible with 6-7 teeth; if 6, basal larger than penultimate tooth. Clypeus evenly convex to weakly emarginate. Posterior outline of head strongly convex. Mesosoma and waist habitus shown in Fig. 48. Sculpture mostly like worker. Anterior pronotum areolate, becoming more rug- ose on sides. Half to nearly all of ventral half of mesopleura smooth; the rest of meso- and metapleura and meso- and metanota longitudi- Volume 3, 1994 47 nally rugose-carinate. Dorsal face of propodeum transversely rugose. Intervals on mesosoma nearly smooth; shiny. Petiole and postpetiole more strongly microareolate and weakly areolate. Discussion. — Rogeria unguispina (Fig. 49) from the mountains of Venezuela (1100-2000m) is also very similar in size, sculpture, pilosity and color to merenbergiana, and the two species intergrade with respect to mesosoma shape and petiole shape. It is quite possible that these are geographic variants of the same species, but all unguispina specimens presently available have more elongate eyes and have a strong carina running across the shoulders, and specimens dissected differ in palpal formula. Most unguispina specimens also have downcurved propodeal spines, concave posterior face of peti- olar node, and a broad, shallow metanotal groove. See also the nevadensis discussion for comparisons with that species from high elevations in Colom- bia. Rogeria alzatei occurs in Colombia, but so far has not been found above 1000m. These ants are usually smaller than merenbergiana (WL 0.51- 0.68mm), lack erect hair on the scapes, and lack a distinct metanotal profile. Rogeria creightoni, so far only known from North and Central America, is related to merenbergiana , but generally has: 1) longer, less inclined propodeal spines, 2) less distinct metano- tum, and 3) stronger, rugose-areolate sculpture on sides of head (Figs. 51-54, 95-96). Rogeria belti from Central America is the same size and also generally has a more or less definite step in mesosoma profile at metanotal groove, but has: 1) larger eye, 2) longer propodeal spines, 3) more rounded postpetiolar node, 4) more areolate pronotum, and 5) mesosoma dorsum without two distinct types of hairs. This species is named in honor of Gunther Buch and his family, who have heroically main- tained part of their farm, Finca Merenberg, as a nature preserve and permitted the collection of these and other ants. Distribution. — Rogeria merenbergiana is found at high elevations (1300-2300m) in the Andes of Ecuador and southern Colombia. It has been taken in litter samples in moist broadleaf forest and bamboo-moss forest. A nest was found in a rotten log in a pasture. Material Examined. — Holotype locality. CO- LOMBIA: 14 workers, 1 queen, Huila, 12km W Belen, Santa Leticia, Finca Merenberg, 2300m, 9- 13-1-1978, rotten jog in pasture (C. Kugler and J. Hahn) [1 mouthparts; 4 stings] [MCZ]. Paratype localities. COLOMBIA: 13 workers, 1 queen, holotype locality [BMNH, LACM, MCZ, MZSP, USNM]; 2 workers, Huila, Las Cuevas de los Guacharos National Park, 1900-2300m, near Palestina, about 20km S Pitalito, 15-17-1-1978 (C. Kugler and J. Hahn) [whole specimen slide mounted] [MCZ]. Nontype localities. ECUADOR: 2 workers, Pichincha Province, 16km E Tandapi, 2000m, in litter, 20-VI-1975, B-302 (S. and J. Peck); 3 workers, 2 queens, Pichincha Province, 3km E Tandapi, 1 300m, litter in wet ravine, VI-1975 (S. and J. Peck); 4 workers, Pichincha Province, 20-30km ENE Alluriquin on Chiriboga Road, 1400-1800m, 1975, B-301 (S. and J. Peck) [CKC, MCZ]. Rogeria unguispina new species Fig. 49 Additions to creightoni-group diagnosis. WL 0.64-0.83mm. Palpal formula 2,2. Clypeal apron with median concavity. Eye moderately large, elongate-oval. Strong carina across shoulders. Metanotal groove broad and shallow. Propodeal spines usually with downcurved tips. Postpetiole from above subtrapezoidal. Sculpture on side of head vestigial, leaving a small or extensive smooth area. Pronotal disc and sides with weakly undu- lating rugae and nearly smooth, shiny interrugal spaces. Erect-suberect hair on scapes, and middle and hind tibiae. Holotype and Paratope Workers. — TL 2.4-3.1 (3.0), HL 0.60-0.72 (0.71), HW 0.50-0.61 (0.60), SL 0.40-0.49 (0.47), EL 0.10-0.11 (0.11) (15-19 facets), PW 0.36-0.46 (0.46), WL 0.64-0.83 (0.81), SpL 0.09- 0.13 (0.13), PetL 0.26-0.34 (0.34), PpetL 0.14-0.19 (0.19)mm, CI 0.82-0.86 (0.84), OI 0.12-0.16 (0.13), SI 0.79-0.82 (0.78), PSI 0.14-0.16 (0.15). N=7 Mandibles with 5-6 teeth, basal larger than penultimate tooth; some with a denticle between basal and penultimate teeth. Body of clypeus not projecting over apron. Nuchal groove makes a notch in side of head. Pronotum rounded on front and sides. Propodeal spiracle large (Fig. 49) to moderate; less than one-half diameter from near- est edge of propodeum. Free edges of metapleural lobes form an oblique angle. Posterior face of petiole concave in holotype and some paratypes. 48 Journal of Hymenoptera Research Petiolar peduncle lacks a distinct keel. Postpetiolar node highest in posterior half. Postpetiolar ster- num with a short, flat profile and square anterior corner. Laterodorsa of head rugose, sometimes ar- eolate caudad (holotype). Posterior head rugose- areolate in a more or less transversely arching pattern. Interstices on head vaguely roughened by effaced microsculpture; especially shiny on sides and back of head. Promesonotal dorsum longitu- dinally rugose in front, becoming more vermicu- late with partial or complete cross-ridges behind. Pronotal sides longitudinally rugose with few cross-ridges. Meso- and metapleura with broken, irregular macrosculpture. Dorsal face of propodeum transversely areolate. Interstices on disc and sides of pronotum shiny, nearly smooth; rest of mesosoma vaguely microareolate and quite shiny. Petiolar node weakly to vestigially microareolate; posterior face of node with longitu- dinal carinulae. Postpetiolar node nearly covered with vague microsculpture to nearly smooth; al- ways weaker than on petiole. Color blackish-brown with yellowish-brown man- dibles, flagellum of antennae, legs, and terminal segments of gaster. Discussion. — Rogena nevadensis is a similar ant from the mountains of northeastern Colombia, but nevadensis: 1) is usually smaller, 2) has smaller eyes, 3) has straight propodeal spines, 4) has more effaced sculpture, and 5) lacks erect hair on the scapes and tibiae. See also the merenbergiana dis- cussion. The name unguispina refers to the claw-like shape of the propodeal spines. Distribution. — So far, unguispina is known only from Venezuelan rain forest between 200m and 2000m elevation. Material Examined. — Holotype locality. VEN- EZUELA: Miranda state, Rio Capaya at Quebrada El Bagre, near Salmeron, E of Caracas, 200m, nest beneath log on forest floor in limestone gorge, 8-II- 1982, #238 (J- Lattke) [MCZ]. Paratype localities. VENEZUELA: 9 workers, holotype locality [mouthparts, sting] [BMNH, CKC, LACM, MCZ, MZSP, USNM]; 1 worker, Falcon State, Sierra de San Luis [mountain range just south of Paraguana Peninsula], Haitoncitos, near Curimagua, 1180m, primary forest, 3- VIII- 1982, #271 (J- Lattke) [MCZ]; 1 worker, Aragua State, Rancho Grande, 15-VI-1960, 4563a (J. qasti) [MCZ]; 1 worker, Aragua State, Rancho Grande, 14-VIII-1967, 1100m (R. W. Poole) [MCZ]; 2 work- ers, Lara State, vicinity of Caspo, 18km S Sanare, 1620m, forest rotten wood, 6-XII-1985 (J. Lattke and W. L. Brown) [MIZA]. Rogeria brunnea Santschi Figs. 50, 93 Rogena curvipubens brunnea Santschi 1930:79. Syntype work- ers, CUBA: Pinar del Rio, Sierra de los Organos, Rangel (A. Bierig) [NMB] [Both syntypes examined]. N. comb. Rogena scabra Weber 1934:27, Fig. 2. Syntype workers, queen, CUBA: Cienfuegos, Soledad, Limones Seboruco (Darlington and Weber) (MCZ) [Queen and 1 of 2 worker syntypes examined], N. syn. Rogeria caraiba Santschi 1936:198, Fig. 5. Syntype workers, CUBA: Habana, Playa de Mananao [NMB] [All 8 syntypes examined]. N. syn. Rogena cubensis Santschi 1936: 199, Figs. 3-4. Holotype worker, CUBA: Pinar del Rio, Sierra del Rosano (A. Bierig) [NMB] [Holotype examined]. N. syn. Rogena cubensis liabanica Santschi 1936:200. Holotype worker, CUBA: Playa de Marianao (Bierig) [NMB] [Holotype examined]. N. syn. Additions to creightoni-group diagnosis. WL 0.61-0. 91mm. Clypeal apron convex. Eye small, elliptical. Nuchal groove weakly visible in lateral view. MHI 0.84-0.99. Pronotal shoulders well rounded. Propodeal spines straight, inclined. Propodeal spiracle less than 1 diameter from edge of infradental lamella. Postpetiole long (PpetL/ PW 0.42-0.49), subtrapezoidal in dorsal view; ster- num with long, flat profile and receding anterior edge (except Dominican Republic). Both macro- and microsculpture weak overall (stronger in Dominican Republic). Rugae fine, low on head and mesosoma (Fig. 93); vestigial on petiole. Erect hair sometimes sparse and only slightly longer than decumbent hair. Scapes have erect hairs; extensor surfaces of legs do not. Workers.— TL 2.3-3.3, HL 0.57-0.80, HW 0.45- 0.70, SL 0.36-0.52, EL 0.06-0.09 (9-18 facets), PW 0.35-0.49, WL 0.61-0.91, SpL 0.09-0.15, PetL 0.25- 0.34, PpetL 0.15-0.22mm, CI 0.79-0.87, OI 0.12- 0.16, SI 0.73-0.89, PSI 0.15-0.17, MHI 0.84-0.94. N=16 Mandibles with 5-7 teeth; basal tooth equal to or slightly larger than penultimate basal; a den- ticle may occur between basal teeth. Body of clypeus not projecting over apron. Posterior out- line of head convex to weakly concave. Mesosoma with broadly rounded shoulders; metanotal groove Volume 3. 1 994 49 a weak impression or absent; no ridge at anterior of propodeum. Propodeal spines narrow, moder- ately long; a bisecting line passes dorsal to propodeal spiracle and through or below anteroventral corner of pronorum. Metapleural lobes low, angular. Petiolar peduncle with strong keel; node large and somewhat bulbous. Postpetiolar node with long, low profile and weakly defined posterior and anterior faces (more distinct in Dominican Republic). Laterodorsa and sides of head rugose-areolate; in most Cuban specimens ridges become very weak on sides, even absent in spots. Posterior head with transversely arching rugose-areolate macrosculp-rure. Anterior pronotal disc trans- versely rugose to rugose-areolate; rest of prono- rum longitudinally rugose, usually with incom- plete cross-ridges between rugae. Mesonotum more rugose-areolate. Pronotal sides basically rug- ose (areolate-rugose in Dominican Republic); meso- and metapleura confused rugose-areolate with some smooth patches. Dorsal face of propodeum transversely rugose. Vague rugosi- ties on sides and /or posterior petiolar node; postpetiolar node smooth. Colorbrown to yellowish-brown, with lighter, more yellowish appendages and frontoclypeal region. Queens.— TL 2.8-3.4, HL 0.59-0.74, HW 0.52- 0.65, SL 0.39-0.50, EL 0.11-0.15, PW 0.45-0.52, WL 0.81-0.95, SpL 0.10-0.16, PetL0.26 (estimated)-0.40, PpetL 0.17-0.22mm, CI 0.86-0.89, SI 0.74-0.78, PSI 0.13-0.16, MHI (Cuban only; could not measure in Bahamian) 0.97-1.06. N=3 Apparently collected as strays, so not defi- nitely associated with workers described above. The Bahamian queen distinctly smaller (all of the smallest measurements above); Cuban queens nearly identical in size. Mandible with 6 teeth; basal and penultimate tooth subequal in size. Clypeus evenly convex. Posterior outline of head medially concave. Mesosoma of Cuban queens robust, shaped like that of scobinata (Fig. 62), but larger and with more prominent, angular metapleural lobes. Mesosoma of Bahamian queen seems like that of Cuban queens, but is obscured by legs. Waist as in workers, except for lack of keel on the petiolar peduncle in Cuban specimen. PpetW/PpetL 1.05-1.14. Sculpture much as in workers. Macrosculpture on sides of head weak. Most of ventral half of mesopleura smooth. Pilos- ity and color as in workers. Discussion. — Workers of this species have two disjunct sizes. In the smaller size range (WL 0.62- 0.70mm) are three specimens from Pinar del Rio Province (including brunnea syntypes), one from Soledad (scabra syntype), one from I. Pinas, three from Dominican Republic, and one from Baha- mas. The larger workers (WL 0.80-0. 91mm) are from Sierra del Rosario (cubensis holotype), Soledad, Playa Marianao (habanica holotype), Si- erra Maestra, and Sierra Anale. Aside from size, the smaller workers show little consistent differ- ence from the larger workers, and in ways that often vary within Rogeria species: 1) slightly more macrosculpture on petiolar node, and 2) slightly more extensive microareolate sculpture on sides of mesosoma. The specimens from Dominican Republic differ from the rest in having a more compact mesosoma, more distinct postpetiolar node and sternum in side view, and slightly stron- ger sculpture. Workers of creightoni from San Jose, Costa Rica have the same size as larger brunnea and much the same pilosity, nuchal groove, and shape of mesosoma and petiole. Rogeria creightoni speci- mens from Yucatan and Chiapas have size, eye size, postpetiole, and pilosity like the larger brunnea, and the Yucatan specimen has similar mesosoma shape. But brunnea workers can be distinguished by the combination of characters in the diagnosis. Material Examined. — CUB A: Pinar del Rio Prov- ince, Sierra de los Organos, Sierra del Rosario (Bierig), Las Acostas (E. O. Wilson), San Vincente (E. O. Wilson); Habana Province, Playa Marianao (Bierig); Cienfuegos Province, Soledad (P. J. Darlington; N. A. Weber; E. O. Wilson; D. Bates and G. Fairchild); Las Villas Province, Caibanen (E. O. Wilson); Sierra Anale (Bierig); Sierra Maestra, Uvero (L. Armas). I. Pinas, S. Casas (L. B. Zayas). DOMINICAN REPUBLIC: 16km ENE Pedemales (P. S. Ward). BAHAMAS: Andros Island, Nassau Island (W. M. Wheeler). 20 workers, 3 queens [LACM, MCZ, NMB]. 50 Journal of Hymenoptera Research Rogeria creightoni Snelling Figs. 51-54, 95-96 Rogeria creightoni Snelling 1973:2, Fig. 1. Holotype and paratype worker, UNITED STATES: Texas, Cameron County, La Fena (W. S. Creighton) [LACM] (Holotype and paratype examined]. Additions to creightoni-group diagnosis. WL 0.63-0. 93mm. Clypeal apron usually with weak to distinct median notch. Eyes small, oval to circular (EL/SpL 0.29-0.64). Metanotal groove weak or absent. MHI usually 0.90-1.06. Propodeal spines long (PSI usually > 0.17), often nearly horizontal. Postpetiolar node from above usually subtrapezoidal, sometimes subrectangular; ante- rior edge of sternum not prominent. Sides of head distinctly rugose-areolate. Pronotal disc trans- versely rugose-areolate on anterior edge and lon- gitudinally rugose-areolate behind (Fig. 96). Scapes and extensor surfaces of tibiae with erect-suberect hairs. Workers.— TL 2.4-3.5, HL 0.58-0.81, HW 0.51- 0.72, SL 0.38-0.55, EL 0.05-0.10 (8-16 facets), PW 0.37-0.52, WL 0.63-0.93, SpL 0.11-0.20, PetL 0.27- 0.40, PpetL 0.14-0.22mm, CI 0.85-0.91, OI 0.08- 0.17, SI 0.68-0.80, PSI 0.16-0.28, MHI 0.87-1.08. N=21 Mandibles with 5-7 teeth, if 5, may have 1-2 additional basal denticles. Basal tooth subequal or larger than penultimate basal. Palpal formula 3,2. Body of clypeus projecting to anterior edge of apron or beyond. Nuchal groove weak to strong. Posterior margin of head weakly concave to con- vex. Mesosoma habitus variable (Figs. 51-53). Propodeal spines moderately long and angled to very long and nearly horizontal, a bisecting line usually passes well above anteroventral corner of pronotum. Propodeal spiracle not especially large or prominent; located < 1/2 diameter to almost one diameter from edge of infradental lamella. Metapleural lobes moderately to very prominent; corner broadly to narrowly rounded. Petiolar node more or less distinct (Figs. 51, 53). Petiolar keel absent to moderately well developed. Sting appa- ratus of specimens from four localities like that of inermis (Fig. 42), except for less angular anteroventral corner of spiracular plate, some- what lower valve chamber height, and in Belize and Oaxaca specimens the sting shaft is higher and slightly upturned. Laterodorsa of head longitudinally rugose to areolate; posterior head transversely arched ar- eolate-rugose. Pronotal sides longitudinally rug- ose to areolate-rugose; meso- and metapleura con- fused rugose (Fig. 95). Dorsal face of propodeum marked anteriorly by a distinct transverse carina, then areolate and /or transversely rugose. Head and mesosoma microsculpture obscure; intervals in macrosculpture moderately shiny. Top and sides of petiolar node effaced microareolate with very weak, indistinct macrosculpture; posterior face with weak longitudinal rugulae. Postpetiolar node with vague microsculpture on sides; smooth dor- sally. Color brownish-yellow to brownish-yellow; dorsa of head and gaster slightly darker, append- ages lighter. Queen. — Uncertain; see discussion. Discussion. — This species becomes quite het- erogeneous with this revision. Ants from almost every locality are different from the others in some conspicuous way. Specimens from Belize have the longest propodeal spines (PSI 0.24-0.28), more prominent metapleural lobes, weak or absent peti- olar keel and thicker ridges in macrosculpture. The Texas specimens have similar habitus, but have somewhat shorter propodeal spines (PSI 0.21- 0.23), less prominent metapleural lobes, more dis- tinct petiolar keel and less thickened macrosculpture. A Tamaulipas specimen is simi- lar to the Texas specimens, but has longer, dis- tinctly inclined propodeal spines. Specimens from Yucatan, and Chiapas (Fig. 52) are the largest ( WL 0.78-0. 93mm), have a prominent clypeal body, somewhat longer scapes, small circular eye, strong nuchal groove, relatively shorter propodeal spines (PSI 0.19-0.22), and head macrosculpture with sharp ridges and unusually large areolae. Speci- mens from La Selva, Costa Rica are similar, but have normal clypeus and sculpture. Others from Costa Rica (Fig. 53) have shorter, more inclined propodeal spines (PSI 0.16-0.20), a more distinct petiolar node and keel, and subrectangular postpetiolar node. One specimen from Oaxaca is more like those from Belize; another is more like the San Jose specimens. With little material from most localities, I did not feel confident naming a new species for each variant, but as collections improve, I will not be surprised if this species undergoes fission. At one time during the revision I also consid- ered describing cornuta, innotabilis, leptonana, and Volume 3, 1 994 51 alzatei as variants of creightoni. But now I believe they can delineated. Comparison of creightoni with leptonana is found in key couplet 40; comparisons with conuita and alzatei, in their respective discus- sions. All known specimens of innotabilis fall within the geographic range of creightoni and the two species are sympatric in at least one locality. All specimens of creightoni differ from innotabilis (Figs. 55-57, 97-98) in several ways: 1) erect hair on scapes and tibiae, 2) palpal formula 3,2, 3) clypeal apron with weak to distinct median notch, and 4) postpetiolar sternum not prominent. The regional variants differ in additional, but inconsistent ways. A single queen from Trinidad (N. A. Weber #129) is very much like the three Costa Rican and Colombian queens provisionally assigned to innotabilis (Fig. 57) in size, general habitus, trian- gular mandibles, convex clypeus, robust mesosoma (MHI 1.28), propodeal spine size and shape, and subrectangular postpetiolar node, but differs in having erect hair on scapes and tibiae, little or no keel on petiolar peduncle, undulate ventral profile of postpetiole, and longitudinally oriented sculp- ture on the posterior head. On the basis of the pilosity, I provisionally assign the Trinidad queen to creightoni, but because of its great similarity in other respects to the innotabilis-like queens, and because neither set of queens is from an area where workers of creightoni or innotabilis have been collected, I do not feel confident of these assignments. See also discussions of alzatei, belti, brunnea, and cornuta. Distribution. — Rogeria creightoni ranges from southern Texas to Costa Rica; from about 1,500m in parts of Chiapas and Costa Rica to sea level. Types were collected in a residential area formerly mesquite-acacia savannah. Other specimens come from riparian woodland, palm-thorn forest, rain forest, pine-oak forest, cecropia forest, and cacao plantation. Most were taken in leaf litter on the ground. Some Belize specimens were under a termite nest; another in orchids. Material Examined.— UNITED STATES: Texas, Cameron County (W. S. Creighton) [LACM]; Live Oak County (P. S. Ward) [mouthparts, sting] [MCZ]. MEXICO: Tamaulipas, Antiguo Morelos (S. & J. Peck); Vera Cruz, Pueblo Nuevo (E. O. Wilson); Oaxaca, 1 mi. E Reforma, nr. Tuxtepec (A. Newton) [mouthparts, sting]; 9 mi. E El Cameron (A. Newton); Chiapas, 12 mi. NW Ocozocoautla (A. Newton); Yucatan, Chichen Itza [MCZ]. BELIZE: Intercepted in Brownsville, Texas on or- chids (Heinrich); nr. Belmopan (S. & J. Peck); Caves Branch (S. & J. Peck) [mouthparts, sting] [BMNH, CKC, MCZ, MZSP, USNM]. COSTA RICA: Heredia Province, La Selva (Talbot & VanDevender) [sting] [LACM], (D. M. Olson) [DMOC]; San Jose (H. Schmidt) [MZSP]; Jurrucarres (A. Bierig) [MZSP]; no locality (F. Nevermann) [MZSP]. 32 workers. Rogeria innotabilis new species Figs. 55-57, 97-98 Additions to creightoni-group diagnosis. WL 0.66-0. 73mm. Mandibles triangular to slightly subtriangular. Palpal formula 2,2. Clypeal apron convex or slightly flattened medially. Eye small, elliptical. Nuchal groove clearly visible in lateral view, but not notching ventral outline of head. Mesosoma with nearly flat dorsal profile; a strong carina marks anterior edge of propodeum; MHI 0.90-1.00. Propodeal spines straight, wide. Peti- olar keel moderate to large. Postpetiole subrectangular in dorsal view; sternum weakly to strongly prominent. Spiracular plate of sting ap- paratus with long spine on posteroventral corner. Sides of head areolate-rugose; mesosoma predomi- nantly rugose; microareolate sculpture distinct on mesosoma sides. No erect hair on scapes or exten- sor surfaces of tibiae. Holotype and Paratope Workers. — TL 2.4-2.7 (2.7), HL b.60-0.65 (0.65), HW 0.50-0.57 (0.54), SL 0.40-0.45 (0.43), EL 0.06-0.08 (0.08) (7-13 facets), PW 0.37-0.41 (0.41), WL 0.66-0.73 (0.72), SpL 0.10- 0.15 (0.13), PetL 0.26-0.30 (0.29), PpetL 0.14-0.16 (0.16)mm, CI 0.83-0.87 (0.84), 010.11-0.15(0.15), SI 0.76-0.80 (0.78), PSI 0.15-0.20 (0.18), MHI 0.90-1.00 (0.98) N=10 Mandible with 5-7 teeth. If 5 or 6 teeth, it may have 1-3 denticles among proximal teeth. If 7 teeth, 4 proximals small. Basal tooth equal to or larger than penultimate tooth. Body of clypeus not pro- jecting to anterior edge of apron. Eye small, ellip- tical. Posterior outline of head strongly to weakly convex. Metanotal groove absent or weak. Propodeal spines wide, straight; a bisecting line passes through or below anteroventral corner of pronotum. Petiolar keel weak or absent. Petiolar node evenly rounded from front to back, or apex somewhat flattened (Fig. 55). Postpetiolar node 52 Journal of Hymenoptera Research highest in posterior half; node in dorsal view much as in Fig. 53. Sting apparatus like that of inermis (Fig. 42), except for spiracular plate shape (Fig. 56). Head macrosculpture areolate on laterodorsa; areolate to transversely arching rugose-areolate on posterior head. Areolae on posterior head are larger and ridges sharper than elsewhere on head. Head dorsum slightly dulled by vestigial microsculpture; intervals smooth and shiny on sides and back. Anterior pronotal disc with 1-2 transverse rugae; rest of promesonotal dorsum longitudinally rugose with lateral spurs (Fig. 98) to rugose-areolate. Lateral pronotum areolate; meso- and metapleura mixed rugose and rugose- areolate. Anterior edge of propodeum marked by weak to strong transverse carina; dorsal face crossed by one or more transverse, sometimes branching, rugulae. Dorsum with vestigial microsculpture, but ventral sides quite distinctly microareolate (Fig. 97). Macrosculpture on sides and back of petiolar node very weakly areolate; microsculpture vague and effaced. Postpetiolar node shiny, almost completely smooth. Color golden brown with yellowish append- ages to brown with slightly lighter frontoclypeal area, mandibles, mesosoma sides, gaster, and legs. Possible Queens.— TL 2.6-2.9, HL 0.59-0.61 , H W 0.52-0.54, SL 0.40-0.41, EL 0.11-0.13, PW 0.44-0.50, WL 0.72-0.83, SpL 0.13-0.14, PetL 0.28-0.31, PpetL 0.14-0.17mm, CI 0.88-0.89, OI 0.21-0.24, SI 0.75- 0.77, PSI 0.16-19, MHI 1.15-1.22. N=3 Collected outside known range of innotabilis workers, but have innotabilis traits: shapes of clypeus, nuchal groove, and postpetiole, and lack of erect /suberect hair on scapes and tibiae. Differing from the worker by normal caste differences and the following. Parapsidal furrows indistinguishable from furrows in sculpture. Pronotum with 2-3 transverse rugae mesad, be- coming longitudinally rugose to rugose-areolate on mesosoma sides; anterior katepisternum smooth. Mesonotum longitudinally rugose. Ve- nation nearly identical to that of belti (Fig. 37). Discussion. — For comparison with related spe- cies, see creightoni, leptonana, and alzatei discus- sions. Remarks about the queens are found in the creightoni and curvipubens discussions. The name innotabilis (L., not remarkable) re- flects my inability to identify any one salient fea- ture on which to base the name for this species. Distribution. — If the queens truly belong to innotabilis, this species ranges from Chiapas to northern Colombia, but workers have not been collected south of Nicaragua. Workers come from moist forest litter at 1000-1200m elevation. Collec- tion sites for queens range from 10-1200m. Material Examined. — Holotype locality. MEXICO: Chiapas State, 12 mi. NW Ocozocoautla, 3200 ft., 4-5-IX-1973, forest leaf litter (A. Newton) [MCZ], Paratype Localities. MEXICO: 7 workers, ho- lotype locality [mouthparts, sting, 1 coated for SEM] [BMNH, MCZ, MZSP]; 10 workers, Chiapas, 6-XII-1951, #51 11574 (Cary) [2 mouthparts, stings] [USNM], NICARAGUA: 2 workers, Km 139 near Matagalpa, Hotel Selva Negra, ca. 1200m, 18- VI- 1978, forest leaf litter (C. Kugler & J. Hahn) [mouth- parts, sting] [LACM, MCZ]. Nontype localities. COSTA RICA: Puntarenas Province, Monteverde, 1200m, 23-V-1979, #3496 (P. Ward) [MCZ], COLOMBIA: Magdalena De- partment, Parque Tayrona, 210m, l-X-76 (C. Kugler) [MCZ]; Guajira, Don Diego, 25-50m, 22- VI-76 (W. L. Brown & C. Kugler) [queen mouth- parts, sting] [MCZ]. 3 queens. Rogeria alzatei new species Figs. 58-60, 99 Diagnosis. — WL 0.51-0. 70mm, most < 0.65mm. Mandibles subtriangular (usually) to triangular. Clypeus usually truncate. Eye usually moderately large, oval (10-14 facets, OI 0.17-0.20 in types), but occasionally half as large and elliptical. Nuchal groove weak to strong. Mesosoma compact (MHI 0.90-1.07). Propodeal spines usually slender, in- clined. Petiolar peduncle with little or no keel. Postpetiolar node from above usually subrectangular; anterior edge of sternum not prominent. Macrosculpture areolate on head sides. Posterior head transversely rugose-areolate (Fig. 99). Pronotal disc rugose to rugose-areolate. No erect hair on scapes or legs. Holotype and Paratype Workers. — TL 1.9-2.2 (2.2), HL 0.51-0.56 (0.54), HW 0.45-0.48 (0.47), SL 0.32-0.36 (0.35), EL 0.07-0.09 (0.09) (11-14 facets), PW 0.32-0.37 (0.35), WL 0.51-0.61 (0.59), SpL 0.10- 0.12 (0.11), PetL 0.21-0.26 (0.23), PpetL 0.12-0.14 (0.13)mm, CI 0.87-0.88 (0.87), OI 0.17-0.20 (0.19), SI 0.72-0.74 (0.74), PSI 0.19-0.21 (0.19), MHI 1.01-1.07 (1.05). N=ll Volume 3. 1 994 53 Mandible subtriangular; 5-6 teeth, two basal teeth small, subequal. Palpal formula 2,2. Clypeal apron truncate, with distinct corners; body of clypeus does not project to edge of apron. Posterior outline of head continuously convex. Nuchal groove in- distinct in lateral view. Mesosoma's evenly con- vex dorsal profile ends abruptly fore and aft by rather sharp angles and by transverse sculpture. Metanotal groove weak or absent. Propodeal spines straight, narrow; a bisecting line passes through or just above anteroventral corner of pronotum. Cor- ner of metapleural lobes a narrowly rounded right angle. Petiolar node evenly rounded front to back. Pygidium in the five specimens dissected with no piligerous tubercles. Head laterodorsa rugose-areolate to areolate; sides more coarsely areolate. Microsculpture ves- tigial, producing irregular intervals on head dor- sum and nearly smooth intervals on sides and posterior (Fig. 99). Anterior edge of pronotal disc with two transverse rugae. Promesonotal dorsum longitudinally rugose with numerous lateral spurs that sometimes connect, forming areolae. No macrosculpture on dorsal face of propodeum. Lateral pronotum longitudinally rugose-areolate. Meso- and metapleura with scattered, irregular rugae. Vestigial rugae on sides and back of peti- olar node. Vague microsculpture makes rough- ened intervals on mesosoma, petiole and much of postpetiole; apex of postpetiolar node smooth. Body yellowish-brown to reddish-brown with slightly darker head and gaster dorsa (black to naked eye); legs and antennae lighter, more yel- lowish or yellowish-brown. Nontype Workers.— TL 1.9-2.5, HL 0.50-0.66, HW 0.42-0.56, SL 0.31-0.46, EL 0.05-0.10 (4-21 fac- ets), PW 0.29-0.39, WL 0.51-0.70, SpL 0.09-0.13, PetL 0.21-0.28, PpetL 0.11-0.15mm, CI 0.81-0.88, OI 0.08-0.20, SI 0.71-0.85, PSI O.15-0.21, MHI 0.90- 1.07. N=32 Specimens from Guyana, French Guiana, and the Caribbean coast of Colombia extend the de- scription as follows: Mandibles triangular with 3- 4 teeth plus 4 basal denticles. Clypeal apron evenly convex, sometimes with medial emargination. Some with only 7-8 facets in the eyes. Some with wider propodeal spines (Fig. 60) or slightly downturned tips. Petiolar keel sometimes distinct (Fig. 60). Macrosculpture on head or pronotum more areolate. Dominican Republic specimens have a con- vex clypeus, a distinct metanotal groove as in Fig. 46, and subtrapezoidal postpetiolar node as in Fig. 51. Specimens from Peru, Brazil, and Paraguay extend the description somewhat differently: Pal- pal formula 3,2. Clypeal apron usually truncate with rounded corners; occasionally convex. Sev- eral specimens with a somewhat prominent clypeal body (Fig. 59). Generally with smaller, elliptical eyes (OI of most 0.10-0.16), but the four workers from the Puerto Maldonado vicinity of Peru pos- sess both the smallest eyes (Fig. 59) and largest eyes of the species (4 and 21 facets, OI 0.08 and 0.20). Nuchal groove forms a distinct notch in the head of one Peruvian worker (Fig. 59). Mesosoma generally less compact (MHI 0.93-1.03). Petiolar keel weak to distinct (Fig. 59). Some with postpetiolar node slightly wider in anterior half and shape similar to Fig. 66. Posterior pygidium with small tubercles in a Paraguayan specimen dissected. Sting apparatus with reduced anterodorsal corner of quadrate plate in a Brazil- ian worker. Paratype and Nontype Queens. — TL 2.4-2.8, HL 0.56-0.60,"hW 0.51-0.53, SL 0.35-0.38, EL 0.13-0.14, PW 0.44-0.49, WL 0.70-0.79, SpL 0.14-0.15, PetL 0.26-0.30, PpetL 0.15-0. 1 6mm, CI 0.88-0.92, OI 0.25- 0.26, SI 0.70-0.72, PSIO.19-0.20, MHI 1.17-1. 18. N=2 Both collected in the same localities as alzatei workers (BCI, Panama; Quebrada Susumuco, Co- lombia), but not in nest series. They most resemble alzatei workers in characteristic shapes of clypeus, mandible, mesosoma, propodeal spines, petiole and postpetiole, as well as sculpture and pilosity. Mesoscutellum projects slightly beyond posterior edge of metanotum. Posterior head rugose-ar- eolate; no tubercles. Middorsum of head covered with erect hairs. Discussion. — The extent of variation within this species makes me question whether this is a single species, but external characters that can be used in a key are either not marked enough to discount individual variation or not concordant. Further complicating the picture is the surprising amount of variation within four specimens from the same region of Peru, and unique sets of fea- tures found in single specimens from northern Colombia, Rio de Janeiro, and Peru. Specimens from Dominican Republic strongly resemble creightoni workers from the same localities, but differ in lacking erect hair on the scapes and tibiae 54 Journal of Hymenoptera Research and in having narrower propodeal spines. Given the geographic variation within alzatei, the species as a whole is difficult to distinguish clearly from creightoni, innotabilis, and leptonana. But if one compares only those specimens oialzatei that are in sympatry or parapatry with those three species, the differences are clear. The Panamanian alzatei workers differ from Central American creightoni, innotabilis, and leptonana in having: 1) subtriangular mandibles, 2) truncate clypeal apron, 3) larger, elongate-oval eye (OI 0.17-0.20), 4) more compact mesosoma (MHI 1.01-1.07), and 5) peti- olar peduncle without a keel. They are also smaller than creightoni and innotabilis workers (WL 0.51- 0.61mm vs. 0.66-0. 93mm in creightoni and 0.66- 0.73mm in innotabilis). Additional differences from leptonana are palpal formula (2,2 vs. 3,2 in leptonana), longer propodeal spines (PSI 0.19-0. 20 vs. < 0.16 in leptonana), and lack of erect hair on scapes. See also the description of sibling species scobinata and the discussions of belli, curvipubens, micromma, tribrocca and merenbergiana. The species is named in honor of a courageous Colombian intellectual and friend, Isaac Alzate. Distribution. — The range of alzatei is from Panama to Paraguay, from western slopes of the Andes in Colombia to the Eastern coast of Brazil. All specimens were apparently taken as strays or in Berlese or Winkler sampling in moist forests from sea level to 1000m. Material Examined. — Holotype locality. PANAMA: Canal Zone, Barro Colorado Island (BCI), 2-VII-79, W. L. Brown [MCZ]. Paratype localities. PANAMA: 3 workers, 1 queen from holotype locality [mouthparts, sting, 1 worker coated for SEM] [CKC, MCZ]; 2 workers, BCI, IV-V-1942, #4953 (J- Zetek) [USNM]; 2 work- ers, BCI, VI-X-1943, #5105 (J. Zetek) [whole speci- men slide mounted] [USNM]; 1 worker, BCI, I- 1960 (W. L. Brown & E. S. McCluskey) [MCZ]; 1 worker, BCI, 7-III-1975, FP#10 (C. Toft & S. Levings) [LACM]; 1 worker, Punta de los Chivos, W side Gatun Lake, 3km SW Gahin, 100m, 3-9-VII-1979 (W. L. Brown) [MCZ]; 1 worker, Bocas del Toro, Pipeline Road, 300m, 8.53N 82.10W, 18-VII-1987, #633 (D. M. Olson) [MZSP]. Nontype localities. DOMINICAN REPUBLIC: 16km ENE Pedernales, 800m (P. S. Ward). CO- LOMBIA: Guajira, Rio Don Diego (W. L. Brown & C. Kugler); Magdalena Department, San Pedro, 550m (P. Ward) [MCZ]; Choco Department, 10km SW San Jose del Palmar, Rio Torito, Finca los Guaduales, 800m (C. Kugler) [mouthparts, sting] [CKC, MCZ, MZSP]; Antioquia Department, Providencia, Estacion Biologico, 600-800m, (C. Kugler) [MCZ]; Cundinamarca Department, Bogota to Villavicencio, Km 79 (W. L. Brown & I. del Polania) [MCZ]; Meta Department, Bogota to Villavicencio, 23km NW Villavicencio, Quebrada Susumuco, 1000m (S. & J. Peck; C. Kugler) [1 mouthparts, 2 stings, 1 whole specimen] [MCZ]. GUYANA (British Guiana): Oronque River (N. A. Weber) [MCZ]. BRAZIL: Sao Paulo State, Agudos (W. Kempf; C. Gilbert) [mouthparts, sting] [MZSP, USNM]; Rio de Janeiro (T. Borgmeier) [MZSP]; Federal District (C. A. C. Seabra) [MZSP]. PERU: Madre de Dios Department, Puerto Maldonado vicinity, 260m (C. Kugler) [CUIC, MCZ]. PARA- GUAY: Central Province (F. Baud, et al.) [mouth- parts, sting] [BMNH, MCZ (voucher), MHN]; Concepcion, and Paraguari Provinces (F. Baud, et al.) [BMNH, MHN]. 62 workers, 1 queen. Rogeria scobinata new species Figs. 61-62, 100 Diagnosis. — As inalzatei, except the following: Clypeal apron truncate (none emarginate). MHI 0.96-1.14. Posterior head with tuberculate macrosculprure (Fig. 100). Erect hair usually ab- sent from head dorsum; if present, it is short and usually confined to posterior margin. Holotype and Parati/pe Workers. — TL 1.9-2.5 (2.2), HL 0.50-0.60 (0.55), HW 0.42-0.52 (0.475), SL 0.31-0.40 (0.35), EL 0.06-0.09 (0.07) (10-14 facets), PW 0.30-0.40 (0.35), WL 0.50-0.67 (0.585), SpL 0.08- 0.12 (0.095), PetL 0.19-0.26 (0.23), PpetL 0.12-0.16 (0.14)mm, CI 0.82-0.86 (0.86), OI 0.14-0.18 (0.15), SI 0.74-0.78 (0.74), PSI 0.16-0.19 (0.16), MHI 0.96-1.09 (1.00). N=5 Nontype Workers.— TL 1.9-2.5, HL 0.50-0.61, HW 0.44-0.52, SL 0.31-0.40, EL 0.06-0.09 (7-15 fac- ets), PW 0.30-0.40, WL 0.51-0.68, SpL 0.08-0.13, PetL 0.19-0.28, PpetL 0.12-0.17mm, CI 0.84-0.88, OI 0.14-0.18, SI 0.70-0.78, PSI 0.16-0.20, MHI 1.00- 1.14. N=32 Mandibles subtriangular, 5-toothed (some- times with 1-2 additional basal denticles), decreas- ing in size basad; basal tooth small. Palpal formula 2,2. Median clypeus of some nontype workers from Colombia like that of alzatei (Fig. 58), but type specimens with less prominent corners (Fig. 61) Volume 3, 1994 55 and other nontypes (Bolivia, some Brazil) have an almost evenly convex clypeal apron. Body of clypeus not projecting over clypeal apron. Poste- rior outline of head weakly concave medially to weakly convex. Nuchal groove clearly visible in side view. Eye oval to elliptical. Anterior and dorsal faces of pronotum may join smoothly, or in a weak angle (Fig. 61). Metanotal groove broad, slightly less to slightly more impressed than shown in Fig. 61, bordered behind by a transverse carina. Propodeal spines inclined. Metapleural lobes moderately prominent; corner varies from sharply angular (Ecuador, some Peru) to rounded as in Fig. 44 (some Paraguay). Ventral petiolar peduncle usually with a weak, nonlamellate keel, but some Ecuadorian specimens with distinct keel. Postpetiolar node in dorsal view subrectangular as in Fig. 58. Pygidium with a pair of median, columnar, piligerous tubercles near caudal edge (barely visible in dissection microscope at 50X). Laterodorsa and sides of head densely ar- eolate. Posterior head with short triangular, blunt tubercles in more or less distinct rows (Fig. 100). Tuberculate sculpture usually extends across pos- terior quarter of head, but in a few specimens from Leticia, Benjamin Constant, and Paraguay, the ridges between the tubercles are not always com- pletely lost, so the posterior head appears mostly fragmented-rugose, with only a few of the trian- gular tubercles. Interstices on most of head some- what dulled by indistinct areolate microsculpture, but smoother and quite shiny between tubercles on back of head; sides sometimes rather strongly microareolate. Anterior edge of pronotal disc with 1-4 more or less transverse rugae. Rest of promesonotum longitudinally rugose with fre- quent incomplete lateral spurs. Mesosoma sides weakly and sparsely rugose to rugose-areolate, but more strongly microareolate than on pronotal disc. Dorsal face of propodeum usually lacking macrosculpture, but rather strongly microareolate. Rest of mesosoma with indistinct microareolate sculpture. Petiolar node with broken vestigial macrosculpture. Postpetiole without macrosculp- ture; nearly smooth on top. Sides of nodes with weak microareolate sculpture that imparts a granu- lar appearance; microsculpture usually weaker on postpetiole. Workers from Leticia have 8-10 erect hairs along posterior rim of head and those from Ben- jamin Constant have sparse, short erect hairs on the posterior rim and along the midline. Color brown to golden brown. Legs and an- tennae generally lighter than rest of body; gaster sometimes darker. Queens.— TL 2.4-2.6, HL 0.55-0.58, HW 0.46- 0.52, SL 0.34-0.38, EL 0.11-0.13, PW 0.41-0.45, WL 0.68-0.74,SpL0.11-0.14,PetL0.25-0.27,PpetL0.14- 0.16mm, CI 0.84-0.90, SI 0.70-0.75, PSI 0.16-0.19, MHI 0.64-0.69. N=7 Habitus shown in Fig. 62. Parapsidal furrows indistinguishable from grooves in sculpture. An- terior pronotum transversely rugose to rugose- areolate, becoming longitudinal on sides. Mesoscutum with longitudinal, often diverging rugae; mesoscutellum rugose or rugose-areolate. Discussion. — Rogeria alzatei is a sibling species of scobinata, with which it is sympatric in Peru, Brazil, and Paraguay, but can be distinguished by characters in the diagnosis and key. The pair of columnar tubercles on the pygidium may also be distinctive. The name scobinata, meaning having the na- ture of a rasp, refers to the sculpture on the poste- rior head, which has rows of teeth like a rasp. Distribution. — Rogeria scobinata ranges from the north coast of South America to Paraguay at elevations below 1000m. All specimens were taken as strays in tropical forest, mostly by Berlese or Winkler sampling of leaf litter, rotten wood, or moss. Material Examined. — Holotype locality. PERU: Madre de Dios Department, 3km N Puerto Maldonado, 260m, primary forest remnant by side of road, berlesate of leaf litter and rotten wood, 13- 16-VI-1981 (C. Kugler and R. R. Lambert) [MCZ]. Paratype localities. PERU: 2 workers, 1 queen, holotype locality [MCZ]; 22 workers, 2 queens, 5km E Puerto Maldonado on Rio Tambopata, Finca Medina, 260m, primary forest berlesate, 13-16- VI- 1981 (C. Kugler and R. R. Lambert) [mouthparts, sting, 1 worker coated for SEM] [BMNH, CKC, LACM, MCZ, MZSP, MHN, USNM]. Nontype localities. TRINIDAD: Nariva Swamp (N. A. Weber) [MCZ]. SURINAM: La Poulle, Vank, and Tambahredjo (J. van der Drift) [MZSP]. COLOMBIA: 7km N Leticia (S. and J. Peck) [MCZ]. ECUADOR: Napo Province, Limoncocha and 20km S Tena (S. and J. Peck); Pastaza Province, 25km N Puyo, (S. and J. Peck) [MCZ]. PERU: Loreto Department, Ramon Castillo (S. and J. Peck), 15km WSW Yurimaguas (P. S. Sb Journal of Hymenoptera Research Ward) [PSWC]; Pasco Department, near Pozuzo, (C. Kugler and R. R. Lambert); Madre de Dios Department, Puerto Maldonado vicinity near Lake Sandoval (C. Kugler and R. R. Lambert) [CKC, CUIC, MCZ]. BOLIVIA: La Paz Department, Caranavi (C. Kugler and R. R. Lambert) [MCZ]. BRAZIL: Para State, (collector name unreadable); Amazonas State, Benjamin Constant and vicinity (W. L. Brown); Mato Grosso State, Utiariti, Rio Saueruina [=R. Papagaio] (Lenko and Pereira); Sao Paulo State, Piracununga Municipality, Cachoeira das Emas (EEBP) (W. L. Brown); Rio de Janeiro (T. Borgmeier) [MCZ, MZSP]. PARA- GUAY: Alto Parana, Amambay, Caaguazii, Canendiyu, and Concepcion Provinces (F. Baud et al.). 122 workers, 13 queens. Rogeria tribrocca new species Fig. 63 Additions to creightoni-group diagnosis. WL 0.62mm. Eye small, elliptical. Clypeal apron with three acute processes. Metapleural lobes very low, broadly rounded. Postpetiolar node widest in an- terior half. Posterior edge of gaster T3 with me- dian spine that seems to arise from a weak concav- ity. Sides of head smooth. Promesonotal dorsum with fine rugae diverging and branching from anterior pronotum, becoming weak on meso- and metanota; interrugal spaces rather wide. No erect hair on scapes or legs. Holotype Worker.— TL 2.3, HL 0.61, HW 0.51, SL 0.40, EL 0.05 (7 facets), PVV 0.36, WL 0.62, SpL 0.10, PetL 0.26, PpetL 0.13mm, CI 0.83, OI 0.10, SI 0.78, MHI 0.93, PSI 0.16. One mandible with six teeth; the other with the penultimate basal replaced by two denticles (Fig. 63). Nuchal groove forms a weak notch in lateral view of head. Pronotal shoulders seen from above subangular. Propodeal spiracles 1 diameter from nearest edge of infradental lamella. Ventral petiole somewhat obscured by glue, but probably with distinct keel and dentate process. Petiolar node narrow; nearly symmetrically rounded in lateral view. Postpetiolar node with broadly confluent anterior and dorsal faces and short pos- terior face. Postpetiolar sternum with perpendicu- lar anterior corner and flat ventral profile (Fig. 63); node in dorsal view widest in anterior half, as in Fig. 74. Sting shaft projecting from gaster acute, with low dorsal flange. Laterodorsa of head longitudinally rugose- areolate. Posterior head with transversely arching rugae. Intervals with weak microsculpture. Dor- sal face of propodeum with 3 weak transverse rugulae. Mesosoma sides with sparse, indistinct, longitudinal rugulae. Microsculpture on mesosoma even weaker than on head; interrugal spaces shiny, nearly smooth. Petiolar and postpetiolar nodes nearly smooth, shiny. Mesosoma dorsum with eight pairs of erect hairs (some matted down); nodes each with two pairs of posterodorsally projecting hairs. Color of body and appendages brownish-yel- low; dorsum of gaster Tl slightly darker. Discussion. — Boihciiri'ijntbens (Figs. 74-76, 101- 102) and alzatei (Figs. 58-60, 99), have also been collected in Quebrada Susumuco or nearby Villavicencio environs. But they do not have a three-toothed clypeal apron or a median spine on the third tergum of the gaster. The name tribrocca (brocca L., projection of teeth) refers to the three tooth-like projections of the clypeal apron. Material Examined. — Holotype locality. CO- LOMBIA: Meta Department, Quebrada Susumuko [or Susumuco], 23km NW Villavicencio, 1000m, forest litter, 5-III-1972, #B-234 (S. and J. Peck) [MCZ]. Rogeria carinata new species Fig. 64 Rogeria curvipubens: M. Smith 1937:851. Misidentification. Additions to creightoni-group diagnosis. WL 053-0.67mm. Clypeal apron convex. Eye small. Propodeal spines small. Mesosoma low (MHI 0.86- 0.96). Postpetiole subrectangular in dorsal view; sternum with distinct, but not projecting anterior lip. Sides of head rugose. Promesonotal dorsum carinate, with broad, shiny intervals. Short erect- suberect hairs on scapes, but not on extensor sur- faces of legs. Holotype and Parati/pe Workers. — TL 2.0-2.5 (2.2), HL 0.51-0.61 (0.55), HW 0.44-0.54 (0.46), SL 0.35-0.42 (0.38), EL 0.05-0.06 (0.06) (7-10 facets), PW 0.31-0.41 (0.34), WL 0.53-0.67 (0.57), SpL 0.05- 0.10 (0.08), PetL 0.19-0.24 (0.22), PetW 0.10-0.13 (0.11), PpetL 0.11-0.14 (0.12), PpetW 0.14-0.18 (0.15)mm, CI 0.86-0.89 (0.86), OI 0.11-0.12 (0.13), SI 0.78-0.80 (0.83), PSI 0.09-0.15 (0.14), MHI 0.86-0.96 (0.92). N=15 Volume 3. 1 994 57 Mandibles with 5 teeth and 0-2 denticles be- tween the two basal teeth; basal tooth larger than penultimate basal. Eyes oval to elliptical. Nuchal groove weak. Posterior outline of head weakly convex to weakly concave. Mesosoma from above broad, angular at shoulders, but strongly con- stricted behind; metanotum half as wide as shoul- ders. Propodeal spiracle small, one diameter from posterior edge of propodeum. Metapleural lobes small, rounded. Ventral petiole with weak keel. Postpetiole highest in posterior half. Sting appara- tus with slightly lower valve chamber than that of inermis (Fig. 42). Laterodorsa of head rugose-areolate in Puerto Rican specimens. Posterior head transversely arched areolate-rugose. Interrugal spaces on head weakly and irregularly roughened on dorsum; smoother on sides and posterior. Anterior face of pronotum with transverse carinules that continue onto shoulders and curve dorsad. Meso- and metapleura with irregular longitudinal rugae; in- tervals smooth dorsally, but vestigially microareolate below level of propodeal spiracle. Dorsal face of propodeum mostly smooth with 0- 3 transverse rugulae. Petiolar node vaguely microareolate; macrosculpture absent or indis- tinct on sides. Postpetiole vaguely microareolate on venter; smooth and shiny on dorsum. Gaster smooth and shiny. Mesosoma dorsum with 10 pairs of erect hairs. Color brownish-yellow; head and mesosoma slightly darker than rest of body. Some Puerto Rican specimens slightly darker, with reddish hue to head and mesosoma. Discussion. — The carinate macrosculpture of the promesonotum with very shiny intervals serves to differentiate carinata from others of the creightoni- group. See also the discussion of nevadensis, which has similar sculpture. Material Examined. — Holotype locality. BRIT- ISH VIRGIN ISLANDS: Tortola, Long Bay, berlesate from soil and leaves, 25-VII-1965 (I. Proj. Staff) [MCZ]. Paratype localities. 1 worker, holotype local- ity [MCZ]. PUERTO RICO: 3 workers, Guanica, 7- IV-1982, #75 (J. A. Torres) [LACM]; 1 worker, Cayo Ratones, 17-VII-1982, #77 (J. A. Torres) [LACM]; 8 workers, Cayo L. Pena, Humacao, 9-X- 1982, #91 (J. A. Torres) [2 stings] [BMNH, CKC, LACM, MCZ, MZSP]; 1 worker, Ensenada, #1409, (no collector) [USNM]. Rogeria nevadensis new species Fig. 65, 94 Additions to creightoni-group diagnosis. WL 0.58-0.65mm. Clypeal apron emarginate. Eye small. Shoulders rounded; promesonotum and propo- deum meet without a strong angle. Postpetiolar node widest in anterior half. Sides of head largely smooth and shiny. Promesonotal dorsum longitu- dinally rugose, with low, rounded ridges and no lateral spurs; interrugal spaces wide and nearly smooth. No erect hair on scapes or extensor sur- faces of legs. Holotype and Paratype Workers.— TL 2.2-2.5 (2.2), HL 0.59-6.62 (0.59), HW 0.49-0.51 (0.49), SL 0.40- 0.42 (0.40), EL 0.06-0.08 (0.06) (7-10 facets), PW 0.34-0.38 (0.34), WL 0.58-0.65 (0.58), SpL 0.09-0.10 (0.09), PetL 0.22-0.25 (0.22), PpetL 0.12-0.14 (0.12)mm, CI 0.81-0.83 (0.82), 010.13-0.15(0.13), SI 0.82, PSI 0.15, MHI 0.93-0.95 (0.93). N=3 Mandibles with 5 teeth plus two denticles, or six teeth plus 1-2 denticles among basal teeth; basal tooth not larger than penultimate basal. Eye elliptical. Nuchal groove weak. Promesonotum slopes to join dorsal face of propodeum without an angle or transverse carina; metanotal groove weak or absent. Propodeal spines narrow; a bisecting line passes through anteroventral corner of prono- tum. Propodeal spiracle within one diameter of edge of infradental lamella. Petiolar peduncle with indistinct keel. Postpetiolar node with a short vertical anterior face and long, sloping dorsal face; posterior face slightly concave in two paratypes. Postpetiole widest in anterior half as in Fig. 66. Postpetiolar sternum with flat ventral profile and somewhat prominent anterior edge. Sting shaft projecting from apex is like that otinermis (Fig. 42). Laterodorsa of head rugose-areolate. Poste- rior head with transversely arching rugae. Head shiny with effaced microsculpture, especially on sides and back. Anterior pronotum crossed by several incomplete rugae. Lateral mesosoma (Fig. 94) with little distinct macrosculpture and smooth, shiny intervals. Dorsal face of propodeum with transverse rugae and effaced microareolate sculp- ture. Both nodes with weak, effaced microareolate sculpture and vague, vestigial rugose macrosculpture. Mesosoma dorsum with 10-11 pairs of erect hairs; 2-3 pairs project posterodorsally from each node. 58 Journal of Hymenoptera Research Color brown to dark reddish-brown (mesosoma and head dorsum may be darkest), with brownish-yellow appendages and frontoclypeal area. Gynecoid Worker. — A nontype specimen is worker-like in most respects, but is larger (WL 0.70mm; PSI 0.19), has distinct mesonotal and metanotal sutures and partially formed wing at- tachments. Pronotum transversely rugose; mesoscutum longitudinally rugose. Sides of postpetiolar node from above more evenly con- vex. Discussion. — Rogeria merenbergiana (Figs. 46- 48), which is also found at high elevations in Colombia and Ecuador, has similar shapes of clypeus, mesosoma and propodeal spines, and somewhat reduced sculpture, but: 1) are a little larger than nevadensis (WL 0.69-0. 83mm), 2) have a metanotal impression, 3) have erect hair on second and third tibiae and scapes, and 4) are more distinctly sculptured, with sides of head rugose. Rogeria carinata (Fig. 64) from Puerto Rica and Tortola have the same size, pilosity, shape of waist, and extensive shiny spaces on mesosoma, but: 1) the metanotal-propodeal junction is more angu- lar, 2) the sides of the head are distinctly rugose, 3) the promesonotal dorsum has straighter parallel ridges and smoother intervals, and 4) the propodeal spines are generally shorter (PSI 0.11-0.15). Rogeria nevadensis takes its name from the Sierra Nevada de Santa Marta, site of the type localities. Material Examined. — Holotype locality. CO- LOMBIA: Magdalena Department, SE Santa Marta, Cuchilla San Lorenzo, vicinity of El Campano, 1340m, rain forest, VI-1976 (W. L. Brown) [MCZ]. Paratype localities. 1 worker, same site as holotype [coated for SEM] [MCZ]. 1 worker, CO- LOMBIA: Magdalena, E Orihueca, 74.03W 10.55N, San Pedro de la Sierra, 1300m, rotten log in coffee plantation, 10-11-1977 (C. Kugler) [LACM]. Nontype locality. One gynecoid worker, berlesate at the San Pedro de Sierra site (C. Kugler) [MCZ]. Rogeria leptonana new species Figs. 66-70 Additions to creightoni-group diagnosis. WL 0.54-0. 66mm. Median clypeal apron concave to convex. Eye small, nearly circular. Propodeal spines small to absent. Mesosoma low, slender (MHI 0.78-0.90; PW/WL 0.53-0.58). Petiole with moder- ate to large keel. Postpetiolar node from above as in Fig. 66. Anterior lip of postpetiolar sternum not prominent. Sides of head distinctly macro- or microsculptured. Mesosoma rugose, sometimes with strongly microareolate sides. Erect hair on scapes and usually on extensor surfaces of tibiae. Holotype and Paratype Workers. — TL 2.1-2.3 (2.1), HL 0.50-0.55 (0.51), HW 0.43-0.46 (0.44), SL 0.32-0.34 (0.32), EL 0.05-0.07 (0.06) (7-10 facets), PW 0.32-0.36 (0.33), WL 0.57-0.61 (0.57), SpL 0.07- 0.10 (0.07), PetL 0.20-0.23 (0.21), PpetL 0.12-0.13 (0.13)mm, CI 0.83-0.86 (0.86), OI 0.11-0.15 (0.14), SI 0.73-0.75 (0.73), PSI 0.12-0.16 (0.12), MHI 0.83-0.90 (0.83). N=4 Mandibles triangular, with 5 teeth and occa- sional denticle; basal tooth larger than penultimate tooth. Palpal formula 3,2. Median clypeal apron concave. Body of clypeus projecting slightly over edge of apron. Posterior outline of head very weakly concave or flat. Nuchal groove indistinct in lateral view. Mesosoma dorsal profile nearly flat. Propodeal spines short with wide base; a bisecting line passes just above anteroventral cor- ner of pronotum. Spiracle about 1 diameter from nearest edge of infradental lamella. Petiolar node small. Postpetiole widest in anterior half (Fig. 66). Ventral profile of sternum concave; anterior lip not prominent. Head rugose-areolate to areolate on laterodorsa and sides, transversely areolate-rug- ose on posterior head. Macrosculpture well de- fined on posterior head, weaker on dorsum and sides. Areolate microsculpture gives a granular appearance to dorsum and sides of head, where it largely obscures the weak macrosculpture; inter- vals on posterior head nearly smooth. Anterior edge of pronotal disc transversely rugose-areolate. Rest of promesonotal dorsum longitudinally rug- ose with occasional laterals; microsculpture ob- scure. Mesosoma sides with confused areolate macro- and microsculpture; the latter quite strong, giving a granular appearance. Petiole appears granular, with vestigial overlying macrosculpture. Volume 3, 1 994 59 Postpetiolar node smooth; sides and venter ap- pear granular. Mesosoma dorsum with more than 12 pairs of emergent erect hairs; nodes each with at least 2 pairs of long posterodorsally projecting hairs. Tibiae with erect hairs. Body yellowish-brown to brownish-yellow; frontoclypeal area and appendages lighter, more yellowish. Nontype Workers.— TL 2.2-2.5, HL 0.54-0.60, HW 0.43-0.50, SL 0.34-0.40, EL 0.04-0.08 (5-9 fac- ets), PW 0.33-0.35, WL 0.54-0.66, SpL 0.03 (no spines)-0.11, PetL 0.21-0.26, PpetL 0.12-0.16mm, CI 0.80-0.83, OI 0.09-0.12, SI 0.77-0.83, PSI 0.05 (no spines)-0.18, MHI 0.78-0.90. N=8 Type description extended as follows. Man- dibles subtriangular in Dominican Republic speci- mens. Clypeal apron convex in the Chiriqui, Panama specimens and truncate in Chiapas and Dominican Republic. Body of clypeus not project- ing beyond clypeal apron in Chiapas specimen. Basal tooth may equal penultimate basal in size. Scapes slightly longer (SI 0.77-0.83). The Ocosingo, Mexico specimen has slightly longer spines than types (PSI 0.18); Cerro Pico Blanco, Costa Rica specimens have very short spines (PSI 0.07-0.10) those from Panama are devoid of armature (Fig. 67). Short propodeal spines of Costa Rican speci- mens are more inclined than in types; a bisecting line passes just below the axilla. Postpetiole of Chiapas and Pedernales, Dominican Republic specimens narrower than others (PpetW/ PpetL 1.23-1.28 vs. 1.41-1.56); Pedernales postpetioles are widest in posterior half. Microsculpture on head may be less distinct than in types. Panama- nian specimens lack erect hair on extensor sur- faces of tibiae and have only 9 pairs of erect hairs on the mesosoma dorsum. Possible Queens.— TL 2.6-2.8, HL 0.60-0.61, HW 0.52-0.53, SL 0.40, EL 0.12-0.13, PW 0.45, WL 0.73- 0.78, SpL 0.14-0.17, PetL 0.28-0.30, PpetL 0.14- 0.16mm, CI 0.86-0.88, OI 0.20-0.23, SI 0.75-0.77, PSI 0.19-0.22, MHI 0.92-0.97. N=2 Both queens collected on the north coast of Colombia. Though workers of leptonana are un- known from South America, these queens (Fig. 68) have strongest affinities to leptonana workers: Mandible with 5 teeth; basal as large as or larger than penultimate basal. Clypeal apron emargin- ate. Posterior outline of head flat. Propodeal spines wide. Propodeal spiracle 2 diameters from edge of infradental lamella. Mesosoma low. Petiolar pe- duncle with prominent lamellate keel; node short. Pygidium and sting apparatus as in workers. Sculpture also as in workers. Microsculpture present but indistinct on head and mesosoma. Pronotum with 1-2 transverse rugae medially; laterally rugose. Meso- and metanota longitudi- nally rugose, with nearly smooth, shiny intervals. Meso- and metapleura longitudinally rugose, ex- cept for smooth, shiny mesokatepisterna. Dorsal face of propodeum smooth. Petiole, including peduncle, and postpetiole nearly smooth. There are differences, however. Palpal for- mula 2,2 in the dissected Guajiran queen (vs. 3,2 in workers). One queen has a nearly subrectangular postpetiole. Neither has erect hair on the tibiae. Males.— TL2.9,HL0.51-0.53,HW0.55-0.56,SL 0.32-0.34, EL 0.25, PW 0.50-0.53, WL 0.91-0.95, PetL 0.32-0.33, PpetL0.15-0.16mm, CI 1.04-1. 10, SI 0.58-0.61. N=4. All four specimens from a nontype nest series, Cerro Pico Blanco, near San Jose, Costa Rica. Man- dible with 4 teeth. Clypeal margin convex. Anten- nal flagellomeres 2-11 subequal in length and width; not twisted. Lateral habitus shown in Fig. 69. May rian and parapsidal sutures present. Forew- ing venation as in Fig. 30 hind wing as in Fig. 37. Genitalia shown in Fig. 70. Mandibles smooth; clypeus nearly so. Head dorsum areolate; sides behind eyes rugose; posterior head rugose-areolate. Mesosoma longitudinally rugose on lateral prono- tum and dorsal meso- and metapleura. Anterior and ventral mesopleura smooth. Ventral metapleura diagonally rugose. Anterior portion of mesonotum with vague effaced microsculpture; longitudinally rugose macrosculpture begins at level of wing and continues onto mesoscutellum. Propodeum areolate. Petiole microareolate, with a few weak, fine longitudinal rugae. Postpetiole and gaster smooth. Abundant erect and decumbent pilosity on scapes, head, mesosoma, waist, and gaster. Head brown, except for yellowish-brown frontoclypeal area and yellow mandibles. Rest of body and appendages brownish-yellow; gaster Tl and SI darker. Discussion. — The nontype specimens from the high elevations of Costa Rica may belong to a distinct species defined by the complete absence of propodeal spines, but presently they seem linked to the types by the two specimens from Panama, which have intermediate spine length. 60 Journal of Hymenoptera Research The different palpal formula of the workers and queens and the fact that workers and queens specimens come from different countries cause doubt about whether the queens are correctly assigned. But most morphological characters fit leptonana better than any other species. See also the curvipubens discussion. The range imwtabilis (Figs. 55-57, 97-98) over- laps that of leptonana at least in southern Mexico and Nicaragua, perhaps more, if queens are cor- rectly assigned. Workers of innotabilis differ as follows: 1) no erect hair on scapes, 2) generally larger (WL 0.66-0.73mm), 3) clypeus evenly con- vex, 4) palpal formula 2,2, 6) postpetiolar node widest midlength and sternum moderately to strongly prominent. See also creightoni, curvipubens, alzatei, inermis, and neilyensis discussions. The name from leptos (G., slender) and nanus (L., dwarf) describes the habitus of this species. Distribution. — Specimens range from Chiapas, Mexico to the northern coast of Colombia; from sea level to 1750m. Type specimens come from lowland rain forest, Colombian specimens from forest receiving about 2m strongly seasonal an- nual precipitation. Montane specimens from Costa Rica were taken under large rocks in a pasture; those from Panama were collected in debris under a stump in unknown habitat. Material Examined. — Holotype and paratype locality. PANAMA: Canal Zone, Barro Colorado Island; 2 workers including holotype, H-III-1943, #5059 (J. Zetek) [USNM, lot 43 3035]; 5 workers, VI-X-1943, #5105 (J. Zetek) [1 whole specimen slide mounted] [USNM, lot 43-16534]; 4 workers, 14-11-1976 (A. Newton) [MCZ]; 1 worker, 27-11- 1976 (A. Newton) [MCZ]; 1 worker, 1-1960, #B-9 (W. L Brown & E. S. McCluskey) [mouthparts, sting] [MCZ]; 2 workers, 7-III-1975, FP#6 (C. Toft & S. Levings) [LACM]. Nontype localities. MEXICO: 1 worker, Chiapas, Ocosingo Valley, Finca El Real, 1-7- VII- 1950 (Goodnight & Stannard). COSTA RICA: 6 workers, 4 males, Cerro Pico Blanco, near San Jose, 1750m, 15-III-1987(W. Eberhard & W. L. Brown) [2 male genitalia] [MCZ], PANAMA: 2 workers, Chiriqui Province, near Nueva California, W of Finca Palo Santo, 1550m, 5-III-1959, #B-377 (H. S. Dybas) [MCZ]. COLOMBIA: 1 queen, Guajira, near Don Diego, Quebrada Guacoche, 10m, 22- VII-1976 ( W. L. Brown & C. Kugler) [queen mouth- parts, sting] [MCZ]; 1 queen, Magdalena Depart- ment, Parque Nacional Tayrona, near Pueblito, 210m, ll-XI-1976 (C. Kugler) [MCZ]. DOMINI- CAN REPUBLIC: 3 workers 16km ENE Pedernales, 800m, 9-IX-1992 (P. S. Ward); 1 worker 4km NNW Villa Altagracia, 200m, 12-IX-1992 (P. S. Ward) [MCZ]. Rogeria belti Mann Figs. 35-39, 91-92 Rogeria belti Mann 1922:31. Syntype workers, HONDURAS: Progreso (Mann) [USNM] [2 of 12 syntypes examined]. Diagnosis. — WL 0.61-0. 94mm. Clypeal apron usually with median notch, sometimes convex. Eyes large, oval. Propodeal spines moderately long, inclined (bisecting line usually passes well below anteroventral corner of pronotum). Promesonotal dorsum usually strongly areolate; postpetiolar node smooth. Scapes with erect hair. Hair on mesosoma dorsum decumbent to erect; not clearly sorting into two kinds. Head dorsum and gaster Tl with short decumbent and long erect/suberect hair. No erect hair on extensor surfaces of tibiae. Workers.— TL 2.4-3.4, HL 0.56-0.80, HW 0.48- 0.69, SL 0.39-0.55, EL 0.10-0.14 (20-36 facets), PW 0.34-0.51, WL 0.61-0.94, SpL 0.11-0.20, PetL 0.25- 0.39, PpetL 0.13-0.21mm, CI 0.82-0.89, OI 0.18- 0.25, SI 0.76-0.89, PSI 0.14-0.24. N=31 Mandibles with 5-6 teeth; typically 5 teeth evenly decreasing in size basad followed by a large basal tooth, or 4 teeth are followed by 1 or 2 denticles and a large basal tooth (Fig. 35). A few specimens in scattered localities have an addi- tional denticle between the third and fourth teeth. Palpal formula 3,2. Clypeal apron medially emar- ginate in most, but evenly convex in specimens from Miami, Costa Rica. Body of clypeus not pro- jecting over clypeal apron. Posterior outline of head weakly convex, sometimes with a slight median concavity. Eyes sometimes narrower than shown in Fig. 35; widest in Miami, Costa Rica specimens (Fig. 36). Pronotum from the side usu- ally subangular at junction of anterior and dorsal faces; from above, shoulders well rounded. Met- anotal groove either hardly visible, or a distinct narrow groove, or a small step in mesosoma pro- file; usually marked behind by a transverse ridge (Fig. 35). Propodeal spiracles small, slightly more than 1 diameter from posterior edge of propodeum, facing slightly caudad. Metapleural lobes nar- Volume 3, 1 994 61 rowly to broadly rounded. Petiolar keel not lamel- late. Petiolar node usually dome-shaped (Fig. 35), but sometimes slightly flattened on top. Postpetiolar node highest in posterior half; dorsal view subtrapezoidal as in Fig. 49; sternum long and flat. Sting apparatus as in creightoni-group diagnosis. Longitudinal rugae on frontal lobes become rugose-areolate by midlength of head; middle pair of rugae usually remain free of cross-ridges until posterior head. Cheeks and laterodorsa areolate to rugose-areolate. Sides of head strongly areolate in most, but nearly smooth in Miami, Costa Rica specimens. Posterior head areolate with no clear transverse pattern in most, but transversely arch- ing rugose to rugose-areolate in ants from Miami, Costa Rica and Colombia. Promesonotum gener- ally coarsely areolate to areolate-rugose (Figs. 91- 92), but lower in relief and more rugose in speci- mens from Miami, Costa Rica and some from Colombia. Meso- and metapleura confused ar- eolate to rugose. Dorsal face of propodeum pre- dominantly transversely rugose, except for Boquete, Panama specimens, in which dorsal face is indistinctly areolate. Petiolar node not as heavily sculptured as mesosoma, but distinctly areolate to weakly rugose-areolate; dorsal and anterior faces often free of macrosculpture. Intervals irregular and usually shiny. Microsculprure generally weak and indistinct, producing quite shiny intervals in macrosculpture, especially on sides of head, mesosoma, and on petiolar node. Head dorsum and petiolar node of Miami, Costa Rica specimens with more pronounced microsculprure, lending a granular appearance. Most dark reddish-brown with yellowish- brown mandibles, antennae and legs. Specimens from Boquete, Panama and Miami, Costa Rica lighter, with more yellowish appendages and end of gaster. Mexican specimens reddish-black with brown appendages. Queens.— TL 3.0-3.4, HL 0.68-0.74, HW 0.58- 0.64, SL 0.48-0.52, EL 0.18-0.20, PW 0.48-0.59, WL 0.88-1.00, SpL0.13-0.21,PetL0.29-0.37,PpetL0.18- 0.20mm, CI 0.84-0.85, SI 0.77-0.86, PSI 0.14-0.21. N=5 Like the worker except for normal queen struc- tures and the presence of predominantly rugose sculpture on the mesonotum. Parapsidal furrows indistinguishable from furrows in sculpture. Habitus similar to that of merenbergiana queen (Fig. 48), but mesonotum slightly more convex and propodeal spines longer and narrower. Wings as in male (Fig. 37). Males.— TL 3.4-3.5, HL 0.59-0.61, HW 0.65- 0.70, SL 0.32-0.35, EL 0.32, PW 0.71-0.72, WL 1.12- 1.19, PetL 0.40-0.41, PpetL 0.19-0.20mm, CI 1.11- 1.15, SI 0.50. N=2 All five specimens from La Ceiba, Honduras. Mandibles with 5 teeth. Clypeus weakly emargin- ate. Flagellomeres 2-11 straight, subequal in length and width. Wings, habitus and genitalia as shown in Figs. 37-39. Rs vein curves away from leading edge of wing; r-m vein lost; M vein short. Man- dibles smooth except for piligerous punctures. Face very finely areolate, opaque. Posterior head longitudinally rugose. Pronotum weakly punc- tate. Mesonotum and posterior face of propodeum finely rugulose and densely punctured. Mesopleura longitudinally striate right under wing, otherwise smooth and shiny with piligerous punctures. Metapleura and sides of propodeum longitudinally rugose. Posterior face of propodeum rugose-areolate. Petiolar node finely punctate to smooth. Postpetiole and gaster smooth and shiny. Hair as in worker, but more dense. Color brown- ish-black, with brown legs and antennae and brownish-yellow mandibles. Discussion. — The 18 specimens from Miami, Costa Rica (Fig. 36) differ from typical belti in having: 1) an evenly convex clypeus, 2) a larger eye (31-36 facets vs. 20-30 in typical belti), 3) smaller propodeal spines (PSI 0.15-0.18 vs. 0.19-0.24 in typical belti), 4) generally narrower head, 5) head and mesosoma more rugose than areolate, and 6) weaker macrosculpture in general, but microsculprure stronger on head and petiolar node. Queens from Miami, Costa Rica like their workers have narrower heads, shorter propodeal spines and weaker sculpture than the four from Guate- mala, Panama and the Osa Peninsula of Costa Rica. I considered these Miami ants a new species until the arrival of new specimens from Costa Rica and Colombia. One worker from Heredia, Costa Rica has a convex clypeus like the Miami ants, but eye and propodeal spine size just within the typi- cal belti range. Other workers from Colombia have eyes as large as any of the Miami specimens, but an emarginate clypeus like typical belti, and they have intermediate sculpture. While it is still pos- sible that the Miami population represents a dif- ferent species, I am not confident of that at present. 62 Journal of Hymenoptera Research Some alzatei (Figs. 58-60, 99) specimens strongly resemble specimens of belti from Costa Rica in size, habitus, spine length, and sculpture, but alzatei: 1) is smaller (WL 0.52-0.65mm), 2) has a truncate clypeal apron in Central America, 3) has smaller eyes (11-14 facets), and 4) lacks erect hair on the scapes. The geographic ranges of creightoni (Figs. 51- 54, 95-96) and belti overlap in much of Central America and some members look quite similar in size, sculpture, and pilosity. Workers of creightoni can be distinguished by: 1) generally smaller, el- liptical eyes with longer propodeal spines (< 19 facets; EL/SpL 0.29-0.64 vs. 0.62-1.23 in belti), 2) more horizontal propodeal spines, 3) promesonotal dorsum rugose to rugose-areolate, and 4) hair on mesosoma dorsum clearly sorting into two dis- tinct kinds. See discussions of inermis, tonduzi, merenbergiana, and exsulans for comparisons with other similar species. Distribution. — Rogeriabelti is so far known from southern Mexico to the Canal Zone and southwest Colombia. Elevations range from 200m to 1500m in Mexico and 50m to 950m in Panama. Labels with ecological data show them collected in rain forest "in Philodendron," "frond sheaths on ground," leaf litter, around an epiphyte mat on a fallen branch, rotten wood, and in dead trees or branches on the ground. Jack Longino found a colony "nesting under loose bark of a still solid log fallen across trail." On another rotten log he lured workers from under the bark to a freshly killed tabanid (Longino, unpublished field notes). Material Examined. — MEXICO: Vera Cruz State, Los Tuxtlas (P. S. Ward); Chiapas State, El Bosque (A. Newton). GUATEMALA: Bobas (W. M. Mann). BELIZE [=British Honduras]: Caves Branch (S. and J. Peck). HONDURAS: Progreso [or El Progreso] (W. M. Mann), Lancetilla and La Ceiba (W. L. Brown) [mouthparts, sting; 2 male genitalia]. EL SALVADOR: between Apopa and Nejapa (W. L. Brown). COSTA RICA: Miami (un- known collector) [mouthparts, sting] [USNM]; Heredia Province, (J. Longino); Santa Clara [ = Limon?] Province, Hamburg Farm and Parismina Br. (F. Nevermann); Puntarenas Prov- ince, Manuel Antonio National Park and Reserva Biologica Carara (P. S. Ward), Osa Peninsula, Corcovado National Park (J. Longino; P. S. Ward). PANAMA: Chiriqui Mountains, Boquete (F. M. Gaige); Cerro Campana (A. Newton); El Valle (H. S. Dybas); Canal Zone, Barro Colorado Island (J. Zetek; A. Newton). COLOMBIA: Nariho Depart- ment, La Guayacana (P. Silverstone) [LACM]. 117 workers, 15 queens, 5 males [CUIC, JTLC, LACM, MCZ, MZSP, USNM]. Rogeria inermis Mann Figs. 40-42 Rogcrm inermis Mann 1922:32, Fig. 15. Syntype workers, HONDURAS: Progreso, Lombardia (Mann) [MCZ] [3of 5 syntypes examined]. Diagtiosis. — WL 0.60-0.85mm. Clypeal apron medially emarginate. Eye oval, moderately large. Nuchal grooves weakly visible in lateral view. Propodeal spines short to absent. Propodeal spi- racles rather large, prominent, face somewhat caudad, usually < 1/2 diameter from edge of infradental lamella. Promesonotal dorsum rugose to areolate-rugose; pronotal sides rugose, shiny. Scapes with erect hair. Head dorsum with erect hairs and shorter, decumbent pilosity. Hair on mesosoma and gaster Tl subdecumbent to erect; not two distinct kinds. Pilosity on gaster rather sparse. Extensor surfaces of tibiae with decum- bent to suberect hairs. Workers.— TL 2.2-3.0, HL 0.55-0.71, HW 0.48- 0.63, SL 0.38-0.50, EL 0.09-0.13 (15-28 facets), PW 0.34-0.47, WL 0.60-0.85, SpL 0.03(no spines)-0.10, PetL 0.25-0.37, PpetL 0.14-0.20mm, CI 0.82-0.91, OI 0.18-0.22, SI 0.73-0.88, PSI 0.04(no spines)-0.14. N=15 Mandibles with 5 or 6 teeth, usually five de- creasing in size basad then a large basal tooth. Palpal formula 3,2. Median clypeal apron weakly to strongly emarginate. Posterior outline of head convex. Metanotal groove weak to absent. Ante- rior edge of propodeum sometimes marked by a transverse ridge. Propodeal spines usually short and acute (Fig. 40), but may also be quite long (Fig. 41), blunt, or absent. Petiole (Fig. 40) with a dis- tinct, rounded node; peduncle slender, venter with strong tooth, but little or no keel. Postpetiolar node highest in posterior half; shape in dorsal view subtrapezoidal as in Fig. 49. Anterior lip of postpetiolar sternum square, not protruding. Py- gidium as in creightoni-gioup diagnosis. Sting apparatus as in Fig. 42. Spiracular plates with incomplete medial connection; spiracle not enlarged. Anterodorsal corner of quadrate plate Volume 3, 1994 63 prominent, with medial and lateral lobes. Oblong plate anterior apodeme long, slender; ventral arm tall, with with long, slender fulcral arm. Gonostylus with dorsoterminal and companion setae; no gap between these and other sensilla. Lancets strong, acute, with a single minute apical barb. Sting shaft slender, straight; apex strong, acute, with low dorsal flange. Valve chamber and sting bulb large, subequal in height and length. Sting base much lower than bulb height, with weak anterolateral processes. Furcula an inverted Y-shape. See Kugler (1978b) for a more complete description. Microsculpture on head and mesosoma re- duced to vague irregularities in otherwise smooth and shiny interrugal spaces. Middorsum of head longitudinally rugose with lateral spurs, grading to areolate-rugose on laterodorsa and sides. Sculp- ture below and behind the eye sometimes vesti- gial, leaving a shiny area. Posterior head in most specimens transversely arching rugose (with lat- eral spurs) to areolate-rugose, but longitudinally rugose in two workers from the Osa Peninsula of Costa Rica. Rugae on mesosoma are sharp, with broad interrugal spaces that are even smoother and shinier than on the head. Anterior pronotum with 1-3 transverse rugae; disc with longitudinal rugae; meso- and metanorum vermiculate-rugose or areolate-rugose. Pronotal sides, mesopleura, and metapleura predominantly rugose; intervals very shiny. Dorsal face of propodeum transversely rugose to areolate. Sides of petiolar node weakly rugose to rugose-areolate, producing an uneven, shiny surface; apex smooth. Postpetiolar node largely smooth and shining with only slight un- evenness of surface on sides. Color varies from a reddish-brown (almost black) body with yellowish brown appendages to a yellowish-brown body with darker gaster and brownish-yellow appendages. Queens.— TL 2.9-3.0, HL 0.62-0.64, HW 0.55- 0.57, SL 0.41-0.45, EL 0.15-0.16, PW 0.47-0.48, WL 0.81-0.84, SpL 0.09, PetL 0.32-0.36, PpetL 0.18- 0.19mm, CI 0.89, SI 0.74-0.79, PSI 0.11. N=3 Mandibles with 6 teeth, or 5 teeth and 2 den- bdes. Parapsidal furrows not discernible. Sculp- ture on posterior head of Panamanian specimens transversely rugose-areolate; more longitudinally oriented in Costa Rican specimens. Pronotum transversely rugose-areolate. Mesonotal sculpture predominantly rugose, with some closed cells. Dorsal face of propodeum transversely rugose. Petiolar node with weak macrosculpture, postpetiole smooth. Discussion. — The two specimens from the Osa Peninsula of Costa Rica differ from the other speci- mens in having noticeably longer propodeal spines and longitudinal, rather than transverse sculpture on the back of the head. These may be members of a sibling species, but spine length is quite variable in inermis (e.g. PSI 0.04-0.09 in one Berlese sample). Thus, I hesitate to name a new species until more specimens show consistent differences. Rogena leptonana (Figs. 66-70) from Cerro Pico Blanco, Costa Rica and Nueva California, Panama have reduced or absent propodeal spines, but differ from inermis in: 1) generally smaller size (WL 0.57-0.66mm; only 1 of 49 inermis workers with WL < 0.67mm), 2) smaller eye (6-8 facets; OI 0.11-0.12), and 3) mesosoma and gaster Tl dorsa with 2 distinct types of pilosity. Rogeria belti (Figs. 35-39, 91-92), often collected with inermis, has longer propodeal spines and more areolate head and pronotal sculpture. See also the neilyensis dis- cussion. Distribution. — All specimens are from Guate- mala to Panama, at elevations from 10m to 1250m. Some were collected in fruit bat guano; others, in Berlese samples from rain forest leaf litter. Material Examined. — GUATEMALA: Alta Verapaz Department, Lanquin (S. and J. Peck). HONDURAS: Progreso [or El Progreso]; Lombardia (W. M. Mann). COSTA RICA: (A. Bierig; F. Nevermann); Heredia Province, Puerto Viejo, La Selva Field Station (C. Kugler; Talbot & VanDevender) [2 stings] ; Limon Province, Cahuita (P. S. Ward); Puntarenas Province, Osa Peninsula (S. Peck), Manuel Antonio National Park (P. S. Ward), Carara Biological Reserve (P. S. Ward); Cartago Province, Turrialba vicinity (A. Bierig) (W. L. Brown) [mouthparts, sting]. PANAMA: Bocas del Toro, 3 localities (D. M. Olson); Barro Colorado Island (W. L. Brown and E. S. McCluskey [mouthparts, whole specimen]; N. A. Weber; J. Zetek); Gamboa (P. S. Ward); Panama City (W. L. Brown). 49 workers, 3 queens [CKC, CUIC, L ACM, MCZ, MZSP,USNM]. Rogeria neilyensis new species Fig. 43 Diagnosis. — WL 0.75-0.76mm. Clypeal apron convex, with weak median angle. Eye small. 64 Journal of Hymenoptera Research Nuchal groove makes a deep notch in lateral view of head. Propodeal spines short, wide. Propodeal spiracles > 1 diameter from edge of infradental lamella. Postpetiolar sternum not enlarged. Ter- minal segments of gaster rotated ventrad. Sides of head with distinct areolate-rugose to rugose macrosculpture; mesosoma predominantly rug- ose. Scapes and legs lack erect hairs. Head dorsum with short, dense, decumbent pilosity and longer, more sparse, erect-suberect hairs. Mesosoma hairs range continuously from short, decumbent to long, erect (i.e. not sorting into two distinct kinds as on head.). Hair on gaster Tl dense, subdecumbent- suberect (decumbent hairs few or absent). Holotype and Paratope Workers. — TL 2.8-2.9 (2.8), HL 0.66-0.68 (0.67), HW 0.58-0.60 (0.59), SL 0.45, EL 0.06-0.07 (0.06) (9-13 facets), PW 0.42-0.43 (0.42), WL 0.75-0.76 (0.75), SpL 0.08-0.10 (0.08), PetL 0.30-0.34 (0.30), PpetL 0.18-0.19 (0.18)mm, CI 0.88-0.89 (0.88), OI 0.10-0.12 (0.10), SI 0.76-0.77 (0.76), PSI 0.11-0.13 (0.11), MHI 0.96-1.00 (0.96). N=2 Mandible 6-toothed; basal tooth larger than penultimate tooth. On left mandible, fifth tooth replaced by 2 denticles. Body of clypeus slightly prominent. Frontal lobes rather wide (FLW/HW 0.43-0.44). Anteroventral corner of pronotum pro- jecting tooth-like. No distinct metanotal groove. Propodeal spiracles moderately large, weakly bulging from sides. Metapleural lobes low, broadly rounded. Petiolar peduncle with large ventral keel and tooth. Postpetiolar node lacking a distinct posterior face; subtrapezoidal from above, as in Fig. 50. Longitudinal rugae on frontal lobes become areolate-rugose halfway tobackof head. Posterior head transversely areolate-rugose. Laterodorsa areolate-rugose. Intervals in macrosculpture of head shiny, but wrinkled by vague, irregular microsculpture. Anterior edge of pronotal disc areolate; rest of promesonotal dorsum longitudi- nally rugose-areolate; more rugose mesad, more areolate laterad. Dorsal face of propodeum trans- versely rugose. Intervals on mesosoma generally wider and smoother than on head. Petiolar node vaguely areolate; posterior face and peduncle with longitudinal carinae. Surface of postpetiolar node uneven, but predominantly smooth and shiny. Gaster Tl and SI smooth and shiny, but other segments progressively dulled caudad by dense microareolate sculpture. Color yellowish-brown, with slightly more yellowish appendages and anterior of head; first tergum of gaster dark brown. Discussion. — This species is similar in size and reduced petiolar spines to another Central Ameri- can species, biennis (Figs. 40-42), which differs as follows: 1) clypeal apron medially emarginate, 2) nuchal groove weak, 3) eyes larger (OI 0.18-0.22; 15-28 facets), 4) propodeal spiracle < 1 diameter from edge of propodeum, 5) scapes with erect hair, and 6) mesosoma and gaster Tl with little, if any, decumbent hair. The leptonana (Figs. 66-70) specimens from Cerro Pico Blanco, Costa Rica, have low, wide propodeal spines like neilyensis, but leptonana has: 1 ) an emarginate clypeus, 2) weak nuchal grooves, 3) lower lower mesosoma profile (MHI 0.84-0.90), and 4) scapes with erect hair. The name neilyensis refers to the type locality. Material examined. — Holotype locality. COSTA RICA: Puntarenas Province, 3km N Ciudad Neily [=Villa Neily], 8.41N 82.57W, 210m, 31-VII-1985, #7771-5 (P. S. Ward). From sifted litter (leaf mold and rotten wood) in second-growth rain forest [MCZ]. Paratype locality. 2 workers, holotype locality [LACM, MCZ]. Rogeria exsulans Wilson and Taylor Fig. 44 Rogena exsulans Wilson and Taylor 1967:74, Fig. 60. Holotype worker, SAMOA: Upolu (T. E. Woodward) [Holotype MCZ; paratypes ANIC, MCZ] [Holotype and 30 paratypes examined]. Diagnosis. — WL 0.68-0. 84mm. Eye fairly large, oval. Propodeal spiracle faces posterolaterally. Petiole with lamellate keel and dentate inferior process. Terminal segments of gaster rotated ventrad. Sides of head areolate-rugose; posterior head and promesonotum strongly areolate. Scapes with decumbent and a few short, suberect hairs. No erect hair on extensor surfaces of legs. Mostly abundant decumbent to subdecumbent hair on dorsa of head, mesosoma, nodes and gaster Tl (sparse erect hair on head and mesosoma); not sorting into two distinct kinds of pilosity. Workers.— TL 2.5-3.0, HL 0.62-0.71, HW 0.52- 0.64, EL 0.08-0.11 (20-26 facets), PW 0.37-0.45, WL 0.68-0.84,SpL0.10-0.15,PetL0.26-0.33,PpetL0.15- 0.17mm, CI 0.84-0.90, OI 0.19-0.21, SI 0.81-0.85, PSI Volume 3. 1 994 65 0.14-0.19. N=4 Some mandibles with 6 teeth (5 decreasing in size then a large basal), but usually the small penultimate basal tooth is replaced by 2 denticles, or by a gap and 1 denticle. Palpal formula 3,2. Clypeal apron medially emarginate. Back of head in dorsal view convex or flat medially; temples broadly rounded (Wilson and Taylor's Fig. 60). Promesonotal profile may be evenly rounded, but more often angular (Fig. 44). Metanotal groove usually distinct, producing a step between promesonotum and propodeum, but weak in two specimens. Propodeal spiracles more than one diameter from nearest edge of inf radental lamella. Node short, evenly convex. Postpetiolar sternum relatively long, flat; postpetiole from above subtrapezoidal, as in Fig. 51. Sting apparatus as in creightoni-group diagnosis. Longitudinal rugae of frontal lobes and me- dian head gradually change to strongly areolate on posterior head. Cheeks and laterodorsa longi- tudinally areolate-rugose. Intervals on head dor- sum vaguely undulate, but shiny; intervals on sides and posterior head smooth and shiny. Promesonotum for the most part strongly areolate, but several elongate cells may occupy the midline of the pronotal disc, and often the pronotal sides have weaker areolate sculpture. Meso- and metapleura confused areolate-rugose. Mesosoma microsculpture reduced; intervals in macrosculp- ture uneven, but very shiny. Petiolar node with weak to vestigial areolate macrosculpture and vague microareolate background. Postpetiolar node shiny, nearly smooth throughout. Head, mesosoma dark reddish-brown, waist and anterior and posterior ends of gaster some- what lighter. Mandibles, clypeus, antennae, and legs yellowish-brown. Discussion. — The only other Pacific species, stigmatica and megastigmatica are easily distin- guished from exsulans by their lack of a petiolar keel and inferior process, as well as other features listed in the diagnoses. The species most similar to exsulans is Central American. Rogeria belti has some members with similar size, habitus, eye size, clypeus shape, a distinct petiolar keel, strong areolate sculpture, and intergrading types of hairs on mesosoma. However, belti workers still differ in having gener- ally longer propodeal spines (PSI 0.19-0.24), less prominent petiolar keel, and distinct long/erect and short /decumbent pilosities on the gaster. Distribution. — Rogeria exsulans has been col- lected only at 600-700m on a single island in the middle of the South Pacific. Most collections specify a rain forest habitat, where it has been found under dead bark, in rotten logs, in moss and ferns on trees and in berlesate of moss on logs and tree trunks. Material Examined.— WESTERN SAMOA: Upolu, Afiamalu (T. E. Woodward; R. W. Taylor; E. C. Zimmerman) [2 mouthparts, stings]. 31 work- ers [MCZ], Rogeria cornuta new species Fig. 45 Diagnosis. — WL 0.93-1. 02mm. Eye relatively small. Nuchal groove makes strong notch in lat- eral view of head. Propodeal spines very long (EL/SpL < 0.50); not inclined dorsad; distal por- tions subparallel with midline. Propodeal spiracles prominent, less than 1/2 diameter from edge of infradental lamella. Metapleural lobes prominent. Petiole with little or no keel. Sides and posterior head strongly areolate; mesosoma predominantly rugose. Erect hair on scapes. Gaster Tl lacks de- cumbent hair. Holotifpe and Paratype Workers. — TL (3.5)-3.8, HL (0.80)-0.89, HW (0>3)-0.78, SL (0.55)-0.58, EL 0.10 (16-18 facets), PW (0.55)-0.60, WL (0.93)-1.02, SpL (0.25)-0.27, PetL (0.42)-0.45, PpetL (0.21)- 0.24mm, CI 0.88-(0.92), OI 0.13-(0. 14), SI 0.74-(0.75), PSI 0.26-(0.27), MHI (0.98)-1.02. Mandible with 6 teeth; basal larger than penultimate basal. Clypeal apron medially emar- ginate; body of clypeus projecting slightly over apron. Posterior outline of head broadly and weakly concave. Sides of pronotum with anterior grooves for insertion of corners of head; shoulders from above angular. No meso- or metanotal grooves. Paratype promesonotal dorsum less con- vex and more angular in front and back than shown for holotype (Fig. 45). Propodeal directed caudad so that a bisecting line would extend just below shoulder; distal half of spines curve inward, almost paralleling the midline. Postpetiolar node somewhat flattened on top; subtrapezoidal in dor- sal view. Postpetiolar node of paratype lower in front than behind. Postpetiolar sternum long, not projecting anteriorly. Head macrosculpture coarse (especially be- 66 Journal of Hymenoptera Research hind), with sharp ridges and shiny interstices. Longitudinal rugae on front break up at midlength of head and give way to a transversely arching areolate-rugose pattern on the posterior head. Laterodorsa confused rugose-areolate. Mesosoma macrosculpture also coarse with smooth inter- stices, but ridges are rounded. Anterior face of pronotum transversely rugose-areolate, pronotal disc longitudinally rugose to vermiculate-rugose (holotype). Mesonotum vermiculate-rugose with some cross-ridges. Pronotal sides areolate-rugose (holotype) or broken and confused; rest of sides confusedly longitudinally rugose with few con- necting ridges. Anterior edge of propodeum marked by a sharp transverse ridge. Petiolar node weakly areolate on sides and posterior; smooth along anterior and dorsal midline. Postpetiolar node weakly areolate on sides; smooth along mid- line. Scapes, head dorsum and tibiae with short decumbent and long erect-suberect hair. Hair on mesosoma and waist ranges from decumbent to erect and varies in length, but not clearly segre- gated into two distinct types. Hair on gaster erect- suberect. Color dark reddish-brown, with lighter frontoclypeal area and ends of gaster; legs and antennae yellowish-brown. Discussion. — Some creightoni, also from Belize, have the same habitus as cornuta, including long, horizontal propodeal spines, but these creightoni members are much smaller (WL 0.63-0. 71mm) and have abundant decumbent pilosity on the gaster Tl. Other creightoni from La Selva, Costa Rica are the same size as the cornuta holotype, but have shorter propodeal spines, slightly smaller eyes, and abundant decumbent pilosity on the gaster Tl. The name cornuta means horned, referring to the long, horn-like propodeal spines. Material Examined. — Holotype locality. BELIZE (British Honduras): 2.5 mi. S Belmopan, 4- VIII-1972, S. and J. Peck, #242 [MZSP]. Paratype locality. 1 worker, MEXICO: Chiapas State, Ocosingo, 2-VI-1967, J. M. Campbell [MCZ] . Curvipubens-Group and Related Species Rogeria curvipubens Emery Figs. 74-76, 101-102 Rogeria curvipubens Emery 1894:190. Worker and queen syntypes, U. S. VIRGIN ISLANDS: St. Thomas (Eggers) [MCSN] [Worker syntype examined]. Additions to curvipubens-group diagnosis. Postpetiolar node strongly vaulted and with small posterior peduncle. Anterior edge of postpetiolar sternum not strongly produced; junction of poste- rior and ventral edges angular (Fig. 75). Sting shaft and lancets weak, spatulate. Sides of head smooth and shiny. Promesonotum with vestigial microsculprure, making interrugal spaces nearly smooth and weakly to strongly shiny. Dorsal face of propodeum usually with 1-5 transverse rugulae. Workers.— TL 1.9-2.3, HL 0.48-0.58, HW 0.38- 0.49, SL 0.32-0.39, EL 0.04-0.07 (4-11 facets), PW 0.28-0.38, WL 0.50-0.63, SpL 0.05-0.11, PetL 0.20- 0.24, PpetL 0.10-0.13mm, CI 0.80-0.86, OI 0.10- 0.15, SI 0.77-0.81, PSI 0.13-0.18, MHI 0.84-1.05. N=22 Mandibles typically with 6 teeth, but occa- sionally with an additional denticle or tooth; basal tooth little if any larger than penultimate basal. Haitian specimens with 5 teeth. Anterior edge of clypeus evenly convex or with median concavity (Haiti). Body of clypeus, though not especially prominent, is generally the anteriormost point of the head in full dorsal view; little or no shelf-like apron. Posterior outline of head generally broadly convex, but sometimes with a weak median con- cavity (Fig. 74). Nuchal groove weak. Eye small, elliptical. Pronotum weakly angular or rounded in front and on sides. Metanotal groove absent or suggested by a broad, very shallow impression. Anterior propodeum marked by transverse carinula that often does not interrupt mesosoma profile. Propodeal spines short, wide, with curved (Fig. 74) or straight edges; a bisecting line passes below the anteroventral corner of pronotum. Metapleural lobes low, subangular to rounded. Sting apparatus like that of inermis (Fig. 42) in most respects, but lancets and sting shaft are weak and the lancets are blunt and spatulate. Though twisted in all preparations, the sting shaft does seem to have a dorsoterminal flange. Posterior head transversely arching rugose to rugose-areolate, intervals shiny (Fig. 101). Most of Volume 3. 1 994 67 head with vague microsculpture. Anterior edge of pronotal disc areolate to rugose-areolate; disc lon- gitudinally rugose (Fig. 102), sometimes effaced on meso- and metanota. Pronotal sides with one or more weak longitudinal rugae; meso- and metapleura confused rugose to rugose-areolate. Petiole and postpetiole microareolate and devoid of macrosculpture, except for small carinulae on petiolar peduncle of some specimens. Microsculpture weaker on apices of nodes, espe- cially postpetiolar node, which is shiny and nearly smooth. Head with 0-6 short suberect hairs; mesosoma dorsum with 2-7 pairs. Extent of erect pilosity on gaster Tl variable; usually covering whole sur- face, but may cover as little as the posterior third. Color uniformly yellow, to golden or light brown body with lighter appendages, frontoclypeal region and ventral gaster. Queens.— TL 2.3-2.6, HL 0.51-0.59, HW 0.44- 0.51, SL 0.34-0.39, EL 0.09-0.12, PW 0.38-0.45, WL 0.63-0.71, SpL 0. 1 1-0.14, PetL 0.22-0.27, PpetL 0.12- 0.15mm, CI 0.83-0.90, SI 0.73-0.78, PSI 0.16-0.19, MHI 0.96-1.17. N=6 Queens (Fig. 76) vary like workers in shapes of head, propodeal spines, petiole and postpetiole. Mandibles have 6 teeth in most; plus 2 denticles in the Guatemalan specimen. Parapsidal furrows weak or indistinguishable. Wing venation as in belti (Fig. 37). Sting apparatus of a queen from the north coast of Colombia like that of workers from the region, including spatulate lancets. Sculpture, pilosity, and color also vary similarly. Sides of head may be partly rugose, but some portion smooth. Pro-, meso-, and metanota rugose. Meso- and metapleura longitudinally rugose dorsad; anterior half of mesokatepisterna very smooth and shiny. Microsculpture as in workers. Num- bers of erect hairs on various parts range as in workers, except for one queen with 10 pairs on mesosoma dorsum. Queens from Mexico and Guatemala are not associated with workers. Moreover, the Guatema- lan specimen lacks a postpetiole and gaster. Both have the side of the head rather strongly rugose- areolate, but I am guessing that they are cunnpubens, rather than cuneola on the basis of the strongly shiny mesokatepisterna and shape of the postpetiolar sternum of the Mexican specimen. Discussion. — Workers come from the Carib- bean Islands and northern South America. How- ever, if the two queens from Mexico and Guate- mala are in fact cunnpubens, the range of cunnpubens broadly overlaps that of sister species cuneola. The two species are distinguished by characters in the cuneola diagnosis and description. Rogeria alzatei (Figs. 58-60, 99) and leptonana (Figs. 66-70) are very similar to curvipubens and cuneola and have been considered curvipubens in the past, but I believe they can be distinguished, usually by general habitus, but especially by more abundant erect pilosity than described in the curvipubens-group diagnosis. Moreover, side-by- side comparisons of sympatric specimens from the Villavicencio vicinity of Colombia and from Barro Colorado Island reveal more differences: the alzatei specimens have a truncate clypeal apron, narrower propodeal spines, and generally stron- ger sculpture. The leptonana specimens have an emarginate clypeal apron, lower mesosoma, and larger petiolar keel. The only question of identity arises on the north coast of Colombia, where a single specimen has wider propodeal spines and convex clypeus like curvipubens, which are com- mon in the area, but has the abundant pilosity and stronger macrosculpture of alzatei. I call that speci- men alzatei, since spine and clypeal shapes vary in that species. Assigning stray queens has also been prob- lematic. In two localities on the north coast of Colombia (Pueblito, Tayrona Park; Don Diego, Guajira) I collected three distinct kinds of queens. One set has a distinctly lower mesosoma (MHI 0.92-0.97) and larger petiolar keel than the other two and has abundant erect pilosity on scapes, whole head, mesosoma, waist, and gaster. These and other characters suggest those are leptonana queens. The other two kinds of queens have a more compact mesosoma (MHI 1 .05-1 .22), but one set has no erect hair on scapes, head, waist, or first tergum of gaster, little or no erect hair on the mesosoma dorsum, more effaced sculpture with nearly smooth sides of head and transversely arch- ing rugae on posterior head, and spatulate lancets, just like the curvipubens workers that are common in both areas. The third set of queens have distinct macrosculpture, distinctly areolate sides and pos- terior head, short erect hair on the whole head dorsum and more abundant erect /suberect hair on mesosoma, waist and gaster and acute lancets. I have tentatively assigned them to innotabilis be- cause of their evenly convex clypeus and 68 Journal of Hymenoptera Research posteroventral spine on the spiracular plate of the sting apparatus. See also micromma, and tribrocca discussions. Distribution. — Rogeria curvipubens has been collected in by Berlese sampling in tropical rain forest, secondary growth forest, and dry tropical forest. Collection sites have all been below 500m elevation. Material Examined.— VIRGIN ISLANDS: St. Thomas (Eggers; Balzan); St. Croix (I. Proj. staff). HAITI: Aquin (no collector). JAMAICA: St. Eliza- beth Parish, Black River (no collector). MEXICO: Vera Cruz, Cuatatolapan (M. Abarca). GUATE- MALA: Alta Verapaz, Trece Aguas (Schwarz and Barber). PANAMA: Barro Colorado Island (N. A. Weber; W. L. Brown). COLOMBIA: Guajira, Serrania de Macuira and vicinity Rio Don Diego (W. L. Brown and C. Kugler) [1 mouthparts, 2 stings, 1 whole specimen slide mounted]; Magdalena Department, Tigrera near Santa Marta (W. L. Brown and C. Kugler), Tayrona National Park (C. Kugler) [worker and queen mouthparts, sting]; Meta Department, Caho El Buque near Villavicencio (Kugler). VENEZUELA: Bolivar State, Campamento Rio Grande (P. S. Ward); Monagas State, Parque de Laguna Grande (P. F. Kukuk). SURINAM: Tambahredjo [queen sting], Sidoredjo [sting], Dirkshoop, and La Poulle (I. van der Drift). 57 workers, 13 queens [BMNH, CKC, CUIC, LACM, MCZ, MZSP, USNM]. Rogeria cuneola new species Figs. 77-78, 103 Additions to curi'ipubens-group diagnosis. Postpetiolar node weakly vaulted and with no posterior peduncle. Anterior edge of sternum strongly produced; posterior and ventral edges merge insensibly (Fig. 78). Sting shaft and lancets strong, acute; sting shaft with dorsal flange; lancet with barbule. Sides of head and mesosoma with strong microareolate sculpture that obscures weak macrosculpture and makes intervals opaque. Dor- sal face of propodeum without transverse rugulae. Holotype and Paratype Workers. — TL 2.0-2.1 (2.0), HL 0.50-0.54 (0.51), HW 0.43-0.46 (0.43), SL 0.32-0.35 (0.33), EL 0.05-0.06 (0.05) (6-7 facets), PW 0.31-0.33 (0.31), WL 0.51-0.56 (0.52), SpL 0.07-0.09 (0.08), PetL 0.20-0.22 (0.20), PpetL 0.10-0.12 (O.ll)mm, CI 0.84-0.87 (0.84), OI 0.10-0.13 (0.13), SI 0.73-0.78 (0.77), PSI 0.15-0.17, MHI 0.94-1.02 (1.02). N=7 Nontype Workers.— TL 2.0-2.3, HL 0.50-0.56, HW 0.43-0.48, SL 0.31-0.36, EL 0.04-0.06 (5-10 fac- ets), PW 0.30-0.37, WL 0.52-0.60, SpL 0.07-0.10, PetL 0.21-0.24, PpetL 0.11-0.13mm, CI 0.83-0.86, OI 0.10-0.13, SI 0.72-0.78, PSI 0.14-0.17, MHI 0.93- 1.03. N=9 Like curvipubens, but differing in the follow- ing ways in addition to diagnosis. Relative widths of nodes with slightly different ranges (PetW/ PetL 0.56-0.70); PpetW/PpetL 1 .38-1 .52). Sting ap- paratus of specimens from Oaxaca (paratypes) and Vera Cruz, Mexico with strong, acute sting shaft and lancets as in inennis (Fig. 42). Posterior head with transversely arching rug- ose-areolate macrosculpture. Compared to curvipubens, rugae on mesosoma dorsum with more lateral spurs that may connect rugae and create areolae on anterior pronotum and on met- anotum. Macrosculpture on mesosoma sides ab- sent or weakly rugose-areolate. Dorsal face of propodeum lacks macrosculpture. Head dorsum with 0-16 hairs suberect hairs; mesosoma dorsum with 1-8 pairs (usually 2-7). Erect hair on gaster Tl usually limited to posterior margin, but entirely absent from Jalisco specimen and entirely covering the tergum of the Yucatan specimen. The Yucatan specimen is also unique in having some stiff, spatulate hairs on head, mesosoma and gaster. Paratype and Nontype Queens. — TL 2.3-2.5, HL 0.53-0.56,HW 0.45-0.50, SL 0.35-0.38, EL 0.10-0.11, PW 0.39-0.45, WL 0.65-0.72, SpL 0.11-0.14, PetL 0.22-0.25, PpetL0.13-0.15mm, CI 0.85-0.89, SI 0.76- 0.78, PSI 0.16-0.19. N=2 Queen as in curvipubens, except for shape of postpetiole, sting, and sculpture as in workers of cuneola. Mandibles with 6 or 7 teeth. Sides of head and mesokatepisterna strongly microareolate and opaque. Paratype queen with erect-suberect hair over whole gaster Tl. Discussion. — See micromma and minima dis- cussions for comparisons with other tiny Rogeria. The name cuneola (L., small wedge) refers to the shape of the postpetiolar sternum in lateral view. Distribution. — These tiny ants are most often taken in berlesate of leaf litter and rotten wood. Some come from sittings under termite mounds and one was collected in a Cattelya orchid. Habitat of most specimens is rain forest or mesic forest, either primary or secondary growth, but one speci- men was found in Yucatan thorn forest. Volume 3, 1994 69 Material Examined. — Holotype locality. MEXICO: Oaxaca State, 1 mi. E Reforma, 15-VIII- 1973, litter, tropical evergreen forest (A. Newton) [MCZ]. Paratype localities. MEXICO: 14 workers, ho- lotype locality [ BMNH, CKC, LACM, MCZ, MZSP, USMN]; 1 worker, 1 queen, Oaxaca State, 1 mi. E Reforma, near Tuxtepec,12-15-VIII-1973, litter for- est floor (A. Newton) [3 stings, 1 worker coated for SEM] [MCZ]. Nontype localities. MEXICO: 1 worker, San Luis Potosi State, El Salto, 6-VII-1969, #B-166 (S. and J. Peck) [MCZ]; 2 workers, Vera Cruz State, Pueblo NuevonearTetzonapa,17-VIII-1953,#B.F. (E. O. Wilson) [MCZ]; 2 workers, Vera Cruz State, El Palmar near Tetzonapa, 8-VIII-1953, #173 (E. O. Wilson) [sting, whole specimen] [MCZ]; 1 worker, Jalisco State, 6km N El Tuito, 31-XII-1987, #9327-6 (P. S. Ward) [MCZ]; 7 workers, 1 queen, Chiapas State, Palenque, 4-IX-1974 (E. M. and J. L. Fisher) [LACM]; 1 worker, Yucatan State, Uxmal, 27- VII- 1963, #118 (E. O. Wilson) [MCZ]. BELIZE: 4 work- ers, Belmopan, 1-15-VIII-1972, #B-243 (S. and J. Peck). GUATEMALA: 1 worker, 14-15- V-1946[?], SF-21214, 46-6134 (no locality orcollector) [USNM] . HONDURAS: 2 workers, La Lima, 4-IX-1961, UCF 217-128 (C. Evers Q.) [MZSP]. COSTA RICA: 1 worker, Guanacaste Province, Santa Rosa National Park, 14-VII-1983 (P. S. Ward) [CKC]; 3 workers, Puntarenas Province, Manuel Antonio National Park, 27-VII-1985, #7692-9 (P. S. Ward) [MCZ]. Rogeria micromma Kempf Fig. 71 Rogeria micromma Kempf 1961:509 (Figs. 12-13) . Holotype worker, SURINAM: Dirkshoop (J. vander Drift) [MZSP] [Holotype and La Poulle paratype examined] Diagnosis. — WL 0.45-0. 52mm. Clypeal apron medially flattened. Eye tiny. Postpetiolar node subrectangular in dorsal view and not strongly vaulted in side view; anterior lip of sternum not prominent, posterior edge angular in side view. Sides of head and mesosoma and dorsal face of propodeum opaque with dense microareolate sculpture. Laterodorsa and sides of head also finely macroareolate. Propodeum free of macrosculpture. No erect hair on scapes or extensor surfaces of legs. Dorsa of head, mesosoma, nodes and gaster Tl with short, appressed hairs and longer, erect to suberect hairs. Mesosoma dorsum with 8-10 pairs of erect hair; each node with 2 pairs of posterodorsally projecting hairs. Workers.— TL 1.7-1.8, HL 0.44-0.51, HW 0.37- 0.45, SL 0.28-0.35, EL 0.02-0.04 (2-5 facets), PW 0.28-0.30, WL 0.45-0.52, SpL 0.07-0.08, PetL 0.17- 0.18, PetW 0.10-0.13, PpetL 0.10-0.11, PpetW 0.13- 0.15mm, CI 0.84-0.88, OI 0.05-0.08, SI 0.76-0.79, PSI 0.15-0.16, MHI 0.92-0.96. N=3 The following supplements diagnosis and Kempf (1961). Mandibles 5-toothed to 6-toothed; basal tooth not larger than penultimate basal. Mesosoma profile of holotype interrupted by a weak metanotal groove followed by two trans- verse carinulae, but profiles of La Poulle and Bra- zil specimens uninterrupted. Brazil specimen with narrower propodeal spines than in types, and propodeal spiracle closer to posterior edge of propodeum. Petiole short (PetL//WL 0.35-0.38), with ventral keel and tooth. In Surinam specimens, vague microareolate microsculpture densely covers head, dorsum of mesosoma, dorsal face of propodeum, petiole and postpetiole of types, producing a weakly shining, granular appearance. Brazil specimen similar, ex- cept for smoother ventral petiolar peduncle and postpetiolar dorsum. Sides of mesosoma and pos- terior surface of head more distinctly microareolate. Anterior of pronotal disc with 1 to 2 transverse rugae followed by longitudinally rugose to ar- eolate-rugose macrosculpture that disappears into microsculpture on meso- and metanota. Sides of mesosoma with sparse, faint longitudinal rugulae on meso- and metapleura. Nodes microareolate; more effaced on postpetiole. Erect hairs of Dirkshoop specimen all trichoid; but thicker and stiff on the La Poulle specimen, at least some cuneate; erect hairs on the Belem speci- men seem intermediate. Color brownish-yellow; legs, mandibles and antennae sometimes slightly lighter. Frontocly peal region not lighter than rest of head. Discussion. — It may be that the three micromma specimens are just unusual curvipubens or cuneola, but at present there are enough differences to provisionally retain this species. Eighteen curvipubens (Figs. 74-76, 101-102) workers from other van der Drift collections in Dirkshoop and La Poulle, are very similar to micromma in size and shape, but differ as follows: 1 ) clypeal apron evenly convex, 2) sides of head and mesosoma shinier with effaced microsculpture, 3) sides of head with 70 Journal of Hymenoptera Research rugose macrosculpture, and 4) reduced pilosity. Some cuneola (Figs. 77-78, 103) specimens are also very similar to micromma in shape and size, and one from Yucatan has stiff erect hairs on head, mesosoma, waist and gaster Tl , but cuneola work- ers differ in shape of the postpetiolar sternum, as well as in clypeal shape and pilosity. Some alzatei from Panama, Colombia, Guyana, and French Guiana are only slightly larger (WL 0.51-0.68mm) than micromma and have the same pilosity and similar structure and sculpture, but they generally have distinctly larger eyes with more than 10 facets, have a higher, more compact mesosoma (MHI 1.00-1.04), and generally nar- rower propodeal spines. Severalrt/za to from north- ern Colombia have reduced eyes (7-8 facets) and one has wider propodeal spines, but those Colom- bian ants are larger, have a higher mesosoma, and more abundant erect pilosity. See also the minima discussion. Material Examined. — SURINAM: Dirkshoop (J. van der Drift), La Poulle 0- van der Drift). BRAZIL: Para, Pirelli Plantation (Iriboca) nr. Belem, (W. L. Brown). 3 workers [MCZ, MZSP]. Rogeria minima Kusnezov Figs. 72-73 Rogeria minima Kusnezov 1958:44, Figs. 1-3. Holotypedealate queen, ARGENTINA: Tucuman [FML] [Holotvpe ex- amined]. Known only from a single queen mounted on a microscope slide showing dorsal head (nearly split in half length- wise), ventral maxillae and labium, lateral mesosoma and petiole, and a mostly ventral view of postpetiole and gaster. Diagnosis. — WL of worker probably < 0.60mm. Mandibles triangular. Palpal formula 2,1. Sting shaft and lancets spatulate. Postpetiole widest in anterior half; anterior lip of sternum not promi- nent. Mesosoma predominantly rugose. No microsculpture on head dorsum, little on mesosoma sides; but microareolate sculpture present on gaster Tl and SI. No erect hairs on scapes or extensor surfaces of legs. Head dorsum with abundant erect hair; mesosoma dorsum with more than 12 pairs; petiolar node with 2 pair projecting posterodorsally. Most erect hairs cu- neate-fimbriate. Queen.— TL 2.3, HL 0.55, HW 0.53, SL 0.34, EL 0.10, WL 0.65, SpL 0.12, PetL 0.22, PpetL 0.11mm, CI 0.96, SI 0.64, PSI 0.18. Mandibles with 5 teeth; basal only slightly larger than penultimate basal. Palpal formula 2,1. Clypeus torn; shape of apron unclear. Eyes mod- erately large, with about 20 facets. Mesosoma as shown in Fig. 72. Parapsidal furrows cannot be discerned. Petiole with ventral tooth and nonlamellate keel. Postpetiole wider (0. 18mm) than long; sides of postpetiole seem to be convex in front, then tapered, as in many curvipnbens (Fig. 74); sternum seems low and not prominent or wedge-shaped. Pygidial gland sculpture present, no tubercles on posterior surface. What is visible of the sting apparatus looks like that of inermis (Fig. 42), except that lancet apices lack barbules and sting shaft seems to have little, if any, terminal flange. Sinceboth sting shaft and lancets are folded, they are probably weakly sclerotized. Median head with diverging rugae continu- ing on posterior head as diverging rugose-areolate sculpture; laterodorsa areolate-rugose. No macrosculpture on median pronotum; sides weakly and incompletely areolate, especially ventrad. Mesonotum longitudinally rugose. Dor- sal half of mesopleura longitudinally rugose; most of mesokatepisterna smooth, but with some weak areolate sculpture along posterior margins. Metapleural and propodeal sides confused ar- eolate-rugose. Mesosoma lacks microsculpture, except on median pronotum, sides of propodeal spines, and metapleural lobes (metanotum and median propodeum could not be examined). All surfaces of petiole, at least venter and sides of postpetiole, and at least sterna of gaster minutely and shallowly areolate with sharp, thin partitions. Much of dorsal aspect of gaster difficult to see, but at least anterior and lateral portions of Tl, and the terminal terga are also shallowly microareolate. Body covered with appressed to decumbent setiform pilosity. In addition, dorsa of head, mesosoma nodes, and gaster Tl and SI with erect cuneate-fimbriate hairs (Fig. 73). Mesosoma dor- sum with more than 12 pairs of erect hairs; petiolar node with 2 pairs of posterodorsally projecting erect hairs (postpetiolar hairs hidden). All hair on terminal segments of gaster are setiform. Discussion. — A worker of micromma from La Poulle, Surinam and a worker of cuneola from Uxmal, Yucatan come very close in size, sculpture, and in having cuneate pilosity, but the gasters of both are smooth except for piligerous punctures. I cannot tell if any hairs are fimbriate. Volume 3. 1 994 71 Fore/i-Group Rogeria foreli Emery Figs. 79-82, 104-105 Rogeria foreli Emery 1894:191. Holotype worker, VIRGIN ISLANDS: St. Thomas (Eggers) [MHN] [Holotype ex- amined). Rogeria foreli gaigei Forel 1914:617. Holotype worker, CO- LOMBIA (Gaige) [MHN]) [Holotypeexamined], N. syn. Rogeria huaciiucana Snelling 1973:4, Fig. 1. Holotype and paratype worker, USA: Arizona, Cochise County (Snelling) [LACM] [Holotype and 1 paratype exam- ined]. N. syn. Additions to foreli-gioup diagnosis. Basal mandibular teeth abruptly smaller than apical teeth. Clypeal apron convex, often with a faint median angle. Eyes usually 10 or more facets. Workers.— TL 1.9-2.9, HL 0.50-0.71, HW 0.43- 0.62, SL 0.32-0.51, EL 0.06-0.10 (7-20 facets), PW 0.30-0.45, WL 0.50-0.80, SpL 0.07-0.15, PetL 0.20- 0.32, PpetL 0.12-0.19mm, CI 0.83-0.89, SI 0.74-0.85, OI 0.12-0.17, PSI 0.14-0.20. N=25 Mandibles with 4-7 teeth and 0-3 denticles. Generally teeth 1-4 decrease in size gradually, then teeth 5-7 (if present) abruptly smaller and possibly interspersed with one or more denticles. Sometimes basal tooth is distinctly larger than penultimate tooth. Body of clypeus often project- ing slightly over the anterior clypeal margin. Pos- terior outline of head flat to weakly convex. Eyes small, oval. Nuchal grooves inconspicuous in lat- eral view. Pronotal shoulders well rounded. Met- anotal groove generally absent (Fig. 79), but may be weakly to distinctly (Fig. 80) visible. Anterior border of propodeum not marked by a ridge. Metapleural lobes small, broadly rounded. Peti- olar node shape varies between extremes shown in Figs. 79 and 81; smaller nodes are as long as wide, larger nodes are longer than wide. Postpetiolar node peaks in posterior half; subrectangular in dorsal view. Anterior lip of postpetiolar sternum small (Fig. 79), or prominent (Figs. 81, 104). Mandibles, median clypeus, legs, posterior face of propodeum, gaster, and sometimes sides of petiolar peduncle smooth, except for minute piligerous punctures. Rest of body densely microareolate or microcolliculate, often appear- ing granular at low magnification. Microareolate sculpture on head is more distinct near antennal insertions and more effaced caudad, sometimes nearly smooth on sides of head. Microareolate pattern distinct on meso- and metapleura and generally on dorsal face of propodeum; indistinct on promesonotum and petiole and vestigial on postpetiole (Figs. 104-105). Microsculprure over- lain by very fine longitudinal rugulae on lateral clypeus, cheeks, frontal lobes, middorsum and sometimes laterodorsa; posterior head with very fine transverse or diverging rugulae. Color yellow with a slightly brownish gaster to chestnut-brown with brownish-yellow or light brown appendages. Queens.— TL 2.4-3.1, HL 0.54-0.69, HW 0.48- 0.59, SL 0.36-0.49, EL 0.10-0.16, PW 0.38-0.51, WL 0.64-0.87, SpL0.12-0.17,PetL0.24-0.32,PpetL0.13- 0.20mm, CI 0.86-0.93, SI 0.72-0.83, PSI 0.17-0.20. N=8 Differing from the workers in the normal queen attributes and in the following. One specimen with 3 teeth and 5 denticles; the others with the same variation as in the workers. Posterior outline of head with weak median concavity in some. Parapsidal furrows are barely discernable. Mesoscutum in all specimens longitudinally ru- gulose; mesoscutellum vaguely rugose to areolate- rugose. Discussion. — Side by side comparison of the types of foreli and gaigei revealed that the only difference between them is size; but the gaigei type is well within the size variation of foreli specimens. Although the types of huachucana come from Ari- zona, far from other known foreli specimens, and were collected in an unusual habitat, they differ from the foreli holotype only in having: 1 ) 7-8 facets in the eyes (vs. 12), and 2) a weak metanotal groove (vs. none). Since eye size and metanotal groove development vary continuously and not concor- dantly in foreli specimens from Central and South America, and since equally small eyes and even more distinct metanotal grooves are present in those specimens, I am unable to distinguish the huachucana specimens as a distinct species at this time. While working in Northern Colombia for two years I regularly collected two kinds of foreli, which I could distinguish at a glance by the shapes of their postpetiolar sterna (Figs. 79, 81). Specimens from Barro Colorado Island, Panama, and Trinidad also have the same two types of sterna. In Colom- bia and Panama, both morphs have been taken from the same locality, but it is not known whether 72 Journal of Hymenoptera Research the two morphs come from different colonies or not. I considered calling the specimens with a more prominent, shelf-like postpetiolar sternum a new species, but the difference is sometimes subtle and individuals difficult to assign. Without cor- roboration from another character, I decided against erecting a new species at this time. Rogeria foreli is closely related to bruchi, which at present is known only from Argentina and Paraguay, much farther south than known foreli specimens (see the bruchi description). Distribution. — Rogeria foreli has a disjunct dis- tribution: Panama and northern South America, the Caribbean, and Southwestern United States. Specimens come from Om to 610m in Panama, 0m to 240m in Colombia and Venezuela generally from berlesate of leaf mold and rotten wood. The Arizona specimens were under stones in an oak- juniper woodland at 1783-1814m. Material Examined. — USA: Arizona, Cochise County (R. R. Snelling). PUERTO RICO: Rio Piedras (J. A. Torres). ST. CROIX: Buck Island. ST. THO- MAS (Eggers). TRINIDAD: (N. A. Weber). PANAMA: Barro Colorado Island (W. L. Brown, W. L. Brown and E. S. McCluskey, A. Newton, N. A. Weber, J. Zetek); Gahin (W. L. Brown); Ancon (no collector) [sting]; Gamboa (P. S. Ward); Colon Province, Frijoles (J. Ventocilla); Cerro Azul (A. Newton). COLOMBIA: no locality (Gaige); Magdalena Department, Tayrona Park (C. Kugler, P. S. Ward) [mouthparts, sting]; Tigrera (W. L. Brown and C. Kugler) [queen mouthparts, whole specimen; worker mouthparts, sting] ; Guaj ira, Don Diego (W. L. Brown and C. Kugler). VENEZU- ELA: Barinas State, near Ciudad Bolivia (P. Ward); Sucre State, 32km W. Campano (S. & J. Peck). 106 workers, 17 queens, 4 males [CKC, CUIC, LACM, MCZ, MCZ, MZSP, USNM]. Rogeria bruchi Santschi Fig. 82 Rogeria bruchi Santschi 1922:352. Holotype worker, ARGEN- TINA, Buenos Aires, Monte Veloz (C. Bruch) [NHM] [Holotype examined). Workers.— TL 2.2-2.6, HL 0.56-0.65, HW 0.48- 0.57, SL 0.38-0.46, EL 0.04-0.06 (5-7 facets), PW 0.35-0.40, WL 0.58-0.70, SpL 0.10-0.17, PetL 0.21- 0.29, PpetL 0.13-0.17mm, CI 0.61-0.70, OI 0.10- 0.12, SI 0.76-0.81, PSI 0.17-0.24. N=5 The five known specimens of bruchi are just like the foreli with nonprojecting clypeus, weak or absent metanotal groove, and nonprojecting postpetiolar sternum, but differ in the following ways: 1) Mandible with five large teeth, second to fifth subequal (Fig. 82), 2) clypeus with a median notch, 3) eyes smaller than in most foreli. Material Examined. — PARAGUAY: Misiones Province, 30km S. of San Juan Bautista (F. Baud et al.); Concepcion Province, 50km E. of San Lazaro (F. Baud, et al.) [mouthparts, sting]. ARGENTINA: Buenos Aires Province, Monte Veloz (C. Bruch); Misiones Province, Posadas (A. Breyer). 5 workers [MCZ, MHN, MZSP, NHM], ACKNOWLEDGEMENTS I am very grateful to the following people for loans or gifts of specimens: Cesare Baroni Urbani (NMB), Claude Besuchet (MHN), Barry Bolton (BMNH), Carlos Roberto Brandao (MZSP), William L.Brown, Jr. (MCZ), John E.Lattke, Jack Longino, Alfred Newton (MCZ), David M. Olson, Roberto Poggi (MCSN), David Smith (USNM), Roy R. Snelling (LACM), Philip S Ward, and E. O. Wilson (MCZ). Jack Longino also generously sent me his unpublished field col- lections notes. Lily Fainter of the Virginia-Maryland Regional College of Veterinary Medicine and Bob Honeycutt of Virginia Poly- technic Institute and State University, Department of Forest Products assisted with the scanning electron microscopy. Reed R. Lambert helped collect ants in Peru and Bolivia. Richard L. Hoffman, W. L. Brown, Jr., Andre Francoeur, and R. W. Taylor provided valuable advice. Special thanks to Roy Snelling, who made many helpful criticisms of a draft. None of these bear any responsibility for the product. The work was supported financially by NSF grant #DEB- 8022177 and Radford University. LITERATURE CITED Bolton, B. 1973. The ant genera of West Africa: A synonymic synopsis with keys (Hymenoptera: Formicidae). Bulle- tin of the British Museum (Natural History), Entomology 27(6):319-368. Bolton, B. 1982. Afrotropical species of the myrmicine ant genera Cardiocondyla, Leptothorax, Melissotarsus, Messor and Cataulacus (Formicidae). Bulletin of the British Mu- seum (Natural History), Entomology 45(4):307-370. Bolton, B. 1987. A review of the Solenopsis genus-group and revision of afrotropical Monomorium Mayr (Hy- menoptera: Formicidae). Bulletin of the British Museum (Natural History), Entomology 54(3):263-452. Borgmeier, T. 1953 (1951). Algumas formigas do genero Macromischa Roger. Arquwos do Museu Nacwnal Rio de /amero 42:107-112. Brown, W. L. 1952. Synonymous ant names. Psyche 58(3): 124. Brown, W. L. 1953. Characters and synonymies among the Volume 3. 1 994 73 genera of ants. Part II. Brevwra no. 18:1-13. Brown, W. L. 1988. Data on malpighian tubule numbers in ants (Hymenoptera: Formicidae). Pp. 17-27 in: J. C. Trager (ed.) Advances in myrmecology . Leiden: E. J. Brill, xxvii+551 pp. Brown, W. L., and W. L. 1949. Wing venation and the phylog- eny of the Formicidae. Transactions of the American Ento- mological Society 75(3-4):113-134. Emery, C. 1894. Studi sulle formiche della fauna neotropica. V1-XV1. Bullettnio della Societa Entomologica Italiana26:\37- 242 + 4 pi. Emery, C. 1896. Studi sulle formiche della fauna neotropica. XVII-XXV. Bullettmo della Societa Entomologica ltaliana 28:33-107. Emery, C. 1897. Formicidarum species novae vel minus cognitae in collectione musaei nationalis Hungarici quas in Nova-Guinea, colonia germanica, collegit L. Biro. Termeszetrajzi Fuzetek 20:571-599 + 2 pi. Emery, C. 1914. Les fourmis de la Nouvelle-Caledonie et des lies Loyalty. P 415 In: F. Sarasin and J. Roux (eds.) Nova Caledonia recherches scientifiques en Nouvelle Caledonie et aux iles Loyalty (Zoologie). Wiesbaden: C. W. Kreidels Verlag, l(4):389-436+13 pi. Emery, C. 1915. Noms de sous-genres et de genres proposes pour la sous-famille des Myrmicinae. Modifications d la classification de ce groupe (Hymenoptera Formicidae). Bulletin de la Societe Entomologique de France 1915:189- 192. Forel, A. 1899. Formicidae. Biologia Centrali-Americana. Insecta. Hymenoptera 3:1-160+4 pi. Forel, A. 1914. Einige amerikanische ameisen. Deutsche Entomologische Zeitschrift 1914:615-620. Harris, Rick A. 1979. A glossary of surface sculpturing. Occa- sional Papers in Entomology. Department of Food and Agri- culture, State of California no. 28:1-31. Holldobler, B. and H. Engel. 1978. Tergal and sternal glands in ants. Psyche 85(4):285-330. Holldobler, B., R. Stanton, R., and Engel. 1976. A new exo- crine gland in Novomessor (Hymenoptera: Formicidae) and its possible significance as a taxonomic character. Psyche 83(1 ): 32-41. Holldobler, B. and E. O. Wilson. 1990. The Ants. Cambridge, Massachusetts: Belknap Press, xii+732 pp. Kempf, W. W. 1961. A survey of the ants of the soil fauna in Surinam. Studia Entomologica 4(l-4):481-524. Kempf, W. W. 1962a (1961). Remarks on the ant genus Irogera Emery, with the description of a new species (Hy- menoptera, Formicidae). Rei'istaBrasileiradeEntomologia 21(4):435-441. Kempf, W. W. 1962b. Miscellaneous Studies on Neotropical Ants. II. (Hymenoptera, Formicidae). Studia Entomologica 5(l-4):l-38. Kempf, W. W. 1963. Additions to the neotropical ant genus Rogeria Emery, with a key to the hitherto recorded South American species (Hym., Formicidae). Remsta Brasileira de Entomologia 23(2):189-196. Kempf, W. W. 1964. Miscellaneous studies on neotropical ants. III. Studia Entomologica 7(l-4):45-71. Kempf, W. W. 1965. Nota preliminar sobre algumas formigas neotropicas, descritas por Frederick Smith (Hy- menoptera, Formicidae). Revista Brasileira de Entomologia 25(2):181-186. Kempf, W. W. 1972. Catalogo abreviado das formigas de regiao neotropical (Hym. Formicidae). Studia Entomologica 15(l-4):3-344. Kempf, W. W. 1975. Miscellaneous studies on neotropical ants. VI. (Hym. Formicidae). Studia Entomologica 18(1- 4):341-380. Kugler, C. 1978a. Pygidial glands in the myrmicine ants (Hymenoptera, Formicidae). Insectes Sociaux 25(3):267- 274. Kugler, C. 1978b. A comparative study of the myrmicine sting apparatus (Hymenoptera, Formicidae). Studia Entomologica 20:413-548. Kugler, C. 1980. The sting apparatus in the primitive ants Nothomyrmecia and Myrmecia. Journal of the Australian Entomological Society 19:263-267. Kugler, C. 1986. Stings of ants of the tribe Pheidologetini (Myrmicinae). Insecta Mundi l(4):221-230. Kugler, C. 1992 (1991). Stings of ants of the tribe Ectatommini (Formicidae: Ponerinae). Insecta Mundi 5(3-4):153-166. Kusnezov, N. 1958. La posicion sistematica del genero Rogeria, con descripcion de una nueva especie. Acta Zoologica Lilloana 15:41-45. Linneaus, C. 1758. Systema Naturae Regnum Animate ed. 10, Tom 1. Stockholm, VI+824pp. Mann, W. M. 1919. The ants of the British Solomon Islands. Bulletin of the Museum of Comparative Zoology, Harvard 63(7):273-391 + 2 pi. Mann, W. M. 1921. The ants of the Fiji Islands. Bulletin of the Museum of Comparative Zoology, Harvard 64(5):401-499. Mann, W. M. 1922. Ants from Honduras and Guatemala. Proceedings of the United States National Museum 61(13):1- 54. Mann, W. M. 1925. Ants collected by the University of Iowa Fiji-New Zealand expedition. University of Iowa Studies m Natural History ll(4):5-6. Mayr, G. L. 1886. Notizen ueber die Formiciden-Sammlung des British Museum in London. Verhandlungen der Zoologisch-Botanischen Gesellschaft in Wein 36:353-368. Roger, J. 1863. Die neu aufgefuehrten Gattungen und Arten meines Formicidenverzeichnisses, nebst Ergaenzungen einiger frueher gegebenen Beschreibungen. Berliner Entomologische Zeitschrift 7:131-214. Santschi, F. 1922. Myrmicines, dolichoderines et autre formicides neotropiques. Bulletin de la Societe Vaudoise Sciences Naturelles 54(205):345-378. Santschi, F. 1923. Description de quelques nouvelles fourmis du Bresil. Reinsta (Museu Paulista, Sao Paulo, Brazil) 13:1253-1264, Santschi, F. 1930. Quelques fourmis de Cuba et du Bresil. Bulletin de la Societe Royale de Botamque d'Egypte Mai 1930(2-3):75-83. Santschi, F. 1936. Contribution & l'etude des fourmis de l'Amerique du Sud. Revista de Entomologia, Rio de Janeiro 6(2):196-218. Santschi, F. 1941. Quelques fourmis japonaises inedites. Mitteilungen der Schweizerischen Entomologischen Gesellschaft 18:273-279. Smith, F. 1858. Catalogue of the hymenopterous insects in the collection of the British Museum. Part VI. Formicidae. Lon- don, 216 pp + 14 pi. Smith, M. 1937. The ants of Puerto Rico. University of Puerto Rico, Journal of Agriculture 20(4):819-875. Snelling, R. R. 1973. Two ant genera new to the United States. Contributions in Science, Natural History Museum of Los Angeles County no. 236:1-8. Snelling, R. R. 1989. Untitled comment. Notes From Under- 74 Journal of Hymenoptera Research ground: A Myrmecological Newsletter. MCZ Laboratories, Harvard, MA no. 3, pp. 7-8. Trager, J. C. 1989. A pitch for standardized orientation of the ant head, and related matters. Notes from Underground: A Myrmecological Newsletter. MCZ Laboratories, Harvard, MA. no. 1, pp. 5-7. Weber, N. A. 1934. Notes on neotropical ants, including the descriptions of new forms. Revista de Entomologia, Rw de Janeiro 4(l):22-59. Wheeler, G. C. 1989. Vive la myrmecologie. Notes from Under- ground: a Myrmecological Newsletter. MCZ Laboratories, Harvard, MA. no. 3, pp. 6-7. Wheeler, G. C. and J. Wheeler. 1953. The ant larvae of the myrmicine tribe Pheidolini. Proceedings of the Entomo- logical Society of Washington 55(2):49-84. Wheeler, G. C. and J Wheeler. 1973. Ant larvae of four tribes: Second supplement (Hymenoptera: Formicidae: Myrmicinae). Psyche 80(l-2):70-82. Wheeler, G. C. and J. Wheeler. 1976. Ant larvae: Review and synthesis. (Memoires of the Entomological Society of Washington, no. 7). Washington: Entomological Society of Washington D.C. vi + 108 pp. Wheeler, G. C. and J. Wheeler. 1977. Supplementary studies on ant larvae: Myrmicinae. Transactions of the American Entomological Society 103:581-602. Wheeler, G. C and J. Wheeler. 1986. Supplementary studies on ant larvae: Myrmicinae (Hymenoptera: Formicidae). journal of the New York Entomological Society 94(4):489- 499. Wheeler, G. C. and J. Wheeler. 1988. Notes on ant larvae: Myrmicinae. Transactions of the American Entomological Society 114:319-327. Wheeler, W M. 1911. A list of the type species of the genera and subgenera of Formicidae. Annals of the Neiv York Academy of Sciences 21:157-175. Wilson, E. O 1955. A monographic revision of the ant genus Lasius. Bulletin of the Museum of Comparative Zoology, Harvard 113(1):1-201. Wilson, E. O. and R. W. Taylor. 1967. The ants of Polynesia (Hymenoptera: Formicidae). Pacific Insects Monograph 14:1-109. Frontal area Lateral clypeus Denticles Posterior head Middorsum Laterodorsum Body of clypeus Clypeal apron >,/ Nuchal groove Side Venter Erect 1 Suberect Appressed Fig. 1. Illustration of some terms and measurements used in descriptions. Dotted lines on the heads show approximate boundaries of the regions. The dashed line is an extension of a line bisecting the propodeal spines. Where it crosses the mesosoma outline gives a rough measure of spine inclination. SpL = spine length. WL = Weber's length. The mesosoma height index (MHI) is distance a (mesosoma height) divided by distance b. Volume 3, 1 994 75 gibba Figs. 2-6. Figs. 2-4. Rogerm ciliosa. 2. Holotype profile; dorsal head. 3. Paratype sting apparatus. Above: lateral views spiracular plate (SP), quadrate plate (QP), oblong plate (OP), gonostylus (Go), triangular plate (TP), and lancet (La). Below left: ventral sting, furcula. Below right: lateral sting, furcula (Fu). 4. Paratype queen lateral mesosoma. Figs. 5-6. Rogeria gibba. 5. Holotype profile; dorsal head. 6. Paratype sting with furcula in lateral and ventral views; lateral view of lancet apex. 76 Journal of Hymenoptera Research 14 prominula Figs. 7-14. Figs. 7-12. Rogeria stigmatica. 7. Syntype dorsal mandibles, clypeus. 8. Syntype profile (pilosity reconstructed by comparison with other specimens); nontype lateral gaster showing typical pilosity (Fulakora, Solomon Islands). 9. Lateral and ventral views of sting with furcula (Falepuna, Western Samoa). 10. Queen mesosoma profile (Viti Levu, Fiji). 11. Male (Sigatoka, Viti Levu). 12. Genitalia (same male). Fig. 13. Rogeria megastigmatka holotype profile. Fig. 14. Rogeria prominula holotype profile (appressed pilosity not shown); dorsolateral view mandibles and clypeal margin; dorsal head. Same scale for all external views. Volume 3, 1994 77 subarmata Figs. 15-23. Figs. 15-16. Rogeria besucheti. 15. Holotype profile; dorsal head. 16. Paratype ventral and lateral views of sting with furcula. Fig. 17. Rogerui blanda male genitalia (Rio Akabin, Venez.). See Figs. 83-84 for worker. Fig. 18. Rogeria procera profile (Rio Cuming, Braz. — " brasiliensis" holotype). Figs. 19-20. Rogeria scartdens syntype. 19. Profile; dorsal head; dorsal waist. 20. Erect hair. Fig. 21. Rogeria terescandens holotype profile; dorsal waist. Figs. 22-23. Rogeria subarmata. 22. Paratype profile. 23. Nontype lateral propodeum, petiole (Belem, Braz). 78 Journal of Hymenoptera Research germami lirata lacertosa Figs. 24-31. Figs. 24-27. Rogeria germaini. 24. Profile (mesosoma hairs had been matted down somewhat), dorsal head; dorsal postpetiole (Passa Quatro, Brazil — "minensis" lectotype). 25. Mesosoma dorsum profile showing typical pilosity (Central Prov., Para.). 26. Lateral petiole (hair omitted) (Central Prov., Para.). 27. Lateral sting, furcula; ventral sting shaft (San Benito Is., Para.). Figs. 28-30. Rogeria lirata. 28. Holotype profile; dorsal head. 29. Paratype lateral lancet (nr. Yunmaguas, Peru). 30. Nontype queen wings (Jatai, Braz.) Cubital vein (Cu), median vein (M), radio-medial crossvein (r-m), radial sector vein (Rs), submarginal cell (SM). Fig. 31. Rogeria lacertosa paratype profile. Same scale for all external views except wings. Volume 3, 1 994 79 Figs. 32-39. Figs. 32-33. Rogerm pellecta. 32. Paratype profile; dorsal postpetiole. 33. Paratype lateral sting, furcula. Fig. 34. Rogeria sicaria holotype profile. Figs. 35-39. Rogeria belti. 35. Syntype profile; dorsal head. 36. Nontype lateral head, mesosoma (hair omitted) (Miami, C. R.). 37. Male wings (La Ceiba, Hond.). 38. Male (La Ceiba, Hond.). 39. Genitalia (same male). Same scale for all external views. 80 Journal of Hymenoptera Research neilyensis Figs. 40-43. Figs. 40-42. Rogeria inermis. 40. Syntype profile showing smallest propodeal spines. 41. Nontype mesosoma dorsum profile (hairs omitted) showing longest propodeal spines (Osa Penin., C. R.). 42. Sting apparatus (La Selva Station, C. R). Right: lateral views of spiracular plate, quadrate plate (QP), oblong plate (OP), gonostylus, triangular plate and lancet; dorsal view anal plate (AP). Below: Lateral and ventral views sting, furcula. Fig. 43. Rogeria neilyensis holotype profile; dorsal head. Volume 3, 1 994 81 exsulans cornuta merenbergiana Figs. 44-48. Fig. 44. Rogeria exsulans holotype profile. Fig. 45. Rogeria cornuta holotype profile; dorsal head. Figs. 46-48. Rogeria merenbergiana. 46. Holotype profile. 47. Nontype mesosoma dorsum profile (hair omitted) showing the least distinct metanotum (Cuevas de los Guacharos, Col.). 48. Paratype queen lateral mesosoma and waist. Same scale for all figures. 82 Journal of Hymenoptera Research unguispma creightoni Figs. 49-54. Fig. 49. Rogena unguispma holotype profile; dorsal head; dorsal waist. Fig. 50. Rogena brunnea profile; dorsal postpetiole (Sierra del Rosario, Cuba — "cubensis" holotype). Figs. 51-54. Rogena creightoni. 51. Paratype profile; dorsal head, dorsal waist. 52. Nontype profile (Ocozocoautla, Mex). 53. Nontype profile; dorsal waist (Costa Rica, Nevermann leg.). 54. Lateral sting, furcula (Caves Branch, Belize). Same scale for all external views. Volume 3. 1 994 83 Figs. 55-62. Figs. 55-57. Rogeria innotabilis. 55. Holotype profile; dorsal head. 56. Paratype spiracular plate from sting apparatus (Chiapas, Mex). 57. Nontype queen profile (Parque Tayrona, Col.). Figs. 58-60. Rogeria alzatei. 58. Holotype profile, dorsal head; dorsal waist. 59. Nontype profile (Pto. Maldonado, Lake Sandoval, Peru). 60. Nontype profile (R. Don Diego, Col.). Figs. 61- 62. Rogeria scobinata. 61. Holotype profile; dorsal head. 62. Paratype queen profile (holotype loc). Same scale for all external views. K4 Journal of Hymenoptera Research tribrocca carinata nevadensis leptonana Figs. 63-70. Fig. 63. Rogeria tribrocca holotype profile; dorsal head. Fig. 64. Rogena carinata holotype profile; dorsal head. Fig. 65. Rogeria nevadensis holotype profile, dorsal head. Figs. 66-70. Rogeria leptonana. 66. Holotype profile; dorsal head; dorsal postpetiole. 67. Nontype profile (nr. Nueva California, Pan.). 68. Nontype queen profile (Parque Tayrona, Col.). 69. Male (Cerro Pico Blanco, C. R). 70. Male genitalia (same loc). Same scale for all external views. Volume 3. 1 994 85 micromma 1 ^Q curvipubens cuneola Figs. 71-78. Fig. 71. Rogeria micromma holotype profile. Figs. 72-73. Rogeria minima holotype. 72. Lateral mesosoma, petiole; ventral postpetiole, gaster. 73. Erect hair. Figs. 74-76. Rogeria curvipubens. 74. Nontype profile; dorsal head; dorsal waist (BCI, Pan). 75. Slide-mounted worker lateral petiole (Parque Tayrona, Col.). 76. Queen profile (R. Don Diego, Col). Figs. 77-78. Rogeria cuneola. 77. Holotype profile. 78. Slide-mounted worker lateral petiole (El Palmar, Mex.). Same scale for all figures except 73, 75, and 78. 86 Journal of Hymenoptera Research 79/ foreli 1.0 mm bruchi Figs. 79-82. Figs. 79-81. Rogerm foreli. 79. Holotype profile. 80. Nontype mesosoma dorsum profile showing strongest metanotal groove (Cerro Azul, Pan.) 81. Lateral waist (Don Diego vie, Col). Fig. 82. Rogeria bruchi holotype profile; dorsal mandibles and clypeus. All to same scale Volume 3. 1 994 87 ;W3fe| 86 tonduzi N« I l» :""' lirata Figs. 83-90. Figs. 83-84. Rogeria blanda (L. Sandoval, Peru). 83. Lateral mesosoma, waist. 84. Dorsal head, mesosoma, petiole. Figs. 85-86. Rogeria tonduzi (La Selva Station, C. R.). 85. Lateral mesosoma and waist. 86. Dorsal mesosoma. Figs. 87-88. Rogerw germaini (Alto Parana Prov., Para.). 87. Lateral mesosoma. 88. Dorsal mesosoma, posterior head. Figs. 89-90. Rogeria lirata paratype. 89. Lateral mesosoma. 90. Dorsal mesosoma. Scale lines = 0.10mm. 88 mm'*/® '''■*IW mm Journal of Hymenoptera Research creightoni |98 |fc inotabilis Figs. 91-98. Figs. 91-92. Rogeria belti (nr. El Bosque, Mex.). 91. Lateral head (part), mesosoma. 92. Dorsal mesosoma and waist of a. smaller worker. Fig. 93. Rogeria brunnea lateral mesosoma (Blanco's Woods, Soledad, Cuba). Fig. 94. Rogeria nevadensis paratype lateral mesosoma. Figs. 95-96. Rogeria creightoni (Belmopan, Belize). 95. Lateral mesosoma, petiole. 96. Dorsal mesosoma. Figs. 97-98. Rogeria innotabilis paratype. 97. Lateral head, mesosoma. 98. Posterior head, dorsal mesosoma. Scale lines = 0.10mm. Volume 3, 1994 89 Figs. 99-105. Fig. 99. Rogeria alzatei paratype posterodorsal head, dorsal pronotum. Fig. 100. Rogeria scobinata paratype posterodorsal head. Figs. 101-102. Rogeria curvipubens (R. Don Diego vie, Col). 101. Posterodorsal head. 102. Dorsal mesosoma. Fig. 103. Rogeria cuneola paratype dorsal mesosoma. Figs. 104-105. Rogeria foreli (Parque Tayrona, Col.). 104. Lateral waist showing prominent anterior lip of postpetiolar sternum of some specimens. 105. Posterior head, dorsal mesosoma. Scale lines = 0.10mm. J. HYM. RES. Vol. 3, 1994, pp. 91-105 The Genera of Bethylinae (Hymenoptera: Bethylidae) Andrew Polaszek and Karl V. Krombein (AP) Department of Entomology, Wageningen Agricultural University, The Netherlands International Institute of Entomology, c/o The Natural History Museum, London, UK (KVK) Department of Entomology, Smithsonian Institution, Washington, D. C. USA. Abstract- — The taxonomic status of the genera comprising the bethylid subfamily Bethylinae is reassessed using computerised phylogenetic analysis. From this analysis seven genera are considered valid, and the following are synonymised: Trissomalus Kieffer 1905 with Odontqjyns Kieffer 1904; and Anoxus Thomson 1862 with Bethylus Latreille 1802. Several species are transferred generically, and several new combinations are presented. The distribution and biology of the Bethylinae are summarised. INTRODUCTION The aculeate family Bethylidae (Chrysidoidea) is a moderately large family of about 1,900 de- scribed species with probably at least as many undescribed. Bethylids are mainly gregarious ectoparasitoids, the Bethylinae mostly develop- ing on larval Lepidoptera, the remainder mostly on Coleoptera, although the hosts are known for only a small proportion of the family. Bethylids are frequently encountered as parasitoids of crop pests, especially in tropical areas, and several spe- cies have been used in attempted biological con- trol (Gordh & Evans 1976; Hempel 1934). Success- ful identification of Bethylidae is therefore impor- tant for many economic entomologists, but has been extremely difficult because the most recent revision was published eighty years ago (Kieffer 1914) and contains many errors. Even at the ge- neric level identification can be very difficult, par- ticularly for the Palaeotropical species. For the Nearctic and Neotropical species this problem has largely been alleviated by the work of Evans (1964, 1978). During our independent studies on African and Asian Bethylinae we have each discovered related undescribed species that donotfitKieffer's (1914) concepts of the bethyline genera, but which we do not regard as meriting description in new genera. Kieffer's and subsequent authors' con- cepts of the bethyline genera needed modification to accommodate the degree of morphological varia- tion which we had discovered. Computerised phylogenetic analysis was selected as the most objective method of assessing character states within the bethyline genera, while providing a more stable classification of the subfamily. The following study was carried out primarily to set new limits to some of these genera, and to facilitate their identification. Here we address the internal phylogeny of the Bethylinae. In comparison to the other bethylid subfamilies, the Bethylinae have been little af- fected at the generic level since Kieffer's (1914) revision. Only one genus, Neoclystopsenella Kurian (1955), was assigned to the Bethylinae since Kieffer's work. Neoclystopsenella was synonymised with Tapinoma (Formicidae) by Brown (1987), but mistakenly retained by Gordh & Moczar (1990) in the Bethylidae. Gordh & Moczar (1990) also mis- takenly transferred Odontepyris Kieffer to Epyrinae, thus assigning a total of eight genera to the Bethylinae (Table 1). When attempting to identify Bethylinae genera using Kieffer's (1914) key, the primary source of confusion is his treatment of the genera Goniozus Fbrster, Parasierola Cameron and Perisierola Kieffer. The latter two genera have since correctly been synonymised with Goniozus (Evans 92 Journal of Hymenoptera Research 1978), but from our studies it is clear that Kieffer intended something different from what he inad- vertently achieved when he keyed and diagnosed these genera in his revision. Goniozus (sensu Evans 1978) contains species that either possess, or do not possess, a closed discoidal cell (areola, areolet) (Fig. 1). Kieffer assigned those species with a dis- coidal cell to Perisierola, and those without one to Goniozus. Kieffer (1907, 1911, 1914), Brues (1907a 1907b) and Muesebeck (1940) all mistakenly se- lected Parasierola Cameron (1883) to accommodate species with both a closed discoidal cell and with one or more longitudinal carinae on the propodeum, this lattercharacter being absent from Kieffer's concepts of the other two genera. Fur- thermore, Kieffer (1914) transposed his concepts of Parasierola and Perisierola when going from his key to genera (1914:238) to his generic diagnoses (1914:533, 542). Thus began 80 years of confusion surrounding these bethyline genera. METHODS Selection of taxa To clarify the status of the genera of Bethy linae we analysed 11 taxa of Bethylinae for 22 morpho- logical characters using the parsimony programme 'Hennig86' (Farris 1988). To polarise characters the genus Lytopsenella Kieffer was selected as the outgroup. Lytopsenella possesses all the characters common to all the remaining bethyline genera in their hypothetical plesiomorphic conditions (see character selection). Characters that are prone to reduction (e.g., number of antennal segments, maxillary and labial palp segments, and wing venation) are found at their maxima within Lytopsenella. Lytopsenella has previously been cho- sen as a basal group, not just for Bethylinae but for Bethylidae as a whole (Evans 1964; Sorg 1988). Representatives of each of the currently valid bethylid genera were included in the analysis. In cases of existing doubt or controversy surround- ing the limits of some genera, type species of both current genera and formerly recognised genera were examined. Particular attention was paid to previous authors' concepts of Goniozus, Odontepyris, Parasierola, Perisierola and Trissomalus. Three species of Goniozus (in the broad sense, i. e. that of Evans 1978) were selected to cover the range of known variation in propodeal and wing venation characters which are important for de- ducing the phylogeny of the subfamily. Current interpretations of the genera Anoxus and Bethylus differ from each other only in whether the eyes are setose or not. This is a character that we have observed to vary intragenerically, so only one representative species of Bethylus was included in the analysis. The taxa selected are given in the data matrix below (Table 2) in the generic combinations which have resulted from this study. Former ge- neric combinations can be found in the treatments of Bethylus and Odontepyris (see below). Selection of Characters We consider the following to represent the ground plan characters of the subfamily Bethylinae. This character list is based partly on the work of Evans (1964) and of Sorg (1988) but largely on our own independent assessments. Ground plan characters within Bethylinae Antennae 13-segmented; clypeus with a well- developed keel, and frontal streak present; maxil- lary palps 6-segmented; labial palps 3-segmented; notauli and parapsidal lines present; scutellum flat, with two small grooves at its proximal corners (Fig. 2); propodeum without carinae (but see be- low); fore femora expanded; prostigma absent; discoidal cell (areolet) present, submarginal and marginal cells present; subdiscoidal cell absent (=discoidal cell of Evans 1964); sternum of petiole with a complete keel, sexual dimorphism limited to genital characters and head shape. Characters Analysed The following characters include all of those which have been used previously for the discrimi- nation of genera within the Bethylinae (except eye setation, see above) as well as some which have not been used previously. We generally agree with Sorg (1988) concerning both the selection and polarization of characters. However, we disagree with Sorg's polarization of the scutellar foveae (character 6, below; Figs. 3, 4). Sorg considers that the occurence of scutellar foveae in the Embolemidae (Sorg 1988: p 30) suggests plesiomorphy. The probability is that, at least in the Bethylinae, they are represented in their Volume 3, 1 994 93 Table 1: Genera of Bethylinae Kieffer (1914) Gordh & Moczar (1990) This paper (Bethylini) (Bethylinae) (Bethylinae) Anoxus Anoxus Bethylus Bethylus Bethylus Bethylus Clyslopsenella not mentioned Scolebythidae (Evans, 1963) Digoniozus Goniozus Goniozus Eupsenella Eupsenella Eupsenella Gomozus Gomozus Gomozus Kathepyris Pristocerinae Pristocerinae (Evans 1964) Lytopsenella Lytopsenella Lytopsenella Neoclystopsenella (incertae sedis) Formicidae (Brown 1987) Odontepyris (Bethylini) Odontepyris (Epy rinae) Odontepyris (Bethylinae) Parasierola Goniozus Goniozus Perisierola Goniozus Gomozus Progoniozus Goniozus Goniozus Prosierola Prosierola Prosierola Swrola Sierola Sierola Trissomalus Trissomalus Odontepyris plesiomorphic condition in Lytopsenella. Assumed primitive (plesiomorphic) charac- ters are coded '0' with '1', '2' representing as- sumed derived (apomorphic) states. 1. Number of antennal segments: The plesiomorphic condition in the Bethylinae is possession of 13-segmented antennae. In Bethylus (including Anoxus) a reduction to 12-segmented antennae has occurred. 0= an- tennae 13-segmented; 1= antennae 12-seg- mented. 2. Number of labial palp segments: The plesiomorphic condition is possession of 3- segmented labial palps. In Bethylus and Sierola the labial palps are 2-segmented. 0= labial palps 3-segmented; 1= labial palps 2-seg- mented. 3. Number of maxillary palp segments: The plesiomorphic condition is possession of 6- segmented maxillary palps, found in both Eupsenella and Lytopsenella. In the remaining taxa the number is reduced to five, except Sierola which has 4-segmented maxillary palps. 0= maxillary palps 6-segmented; 1 = maxillary palps 5-segmented; 2= maxillary palps 4-segmented. Presence of an unsculptured streak fron- tally: In many bethyline taxa a narrow impunctate, sometimes shiny, streak extends from the proximal end of the clypeal carina to the frontal ocellus (Fig. 12, see also Sorg 1988: 121, Fig. 33d, for its presence in Epyrinae: Chilepyris). In Bethylus (including Anoxus), Goniozus and Sierola the sculpturing of the head is reduced (i. e. smooth rather than strongly rugose) and this streak is absent. 0= frontal streak present; 1= absent. Presence of notauli: The notauli, a pair of longitudinal furrows on the mesoscutum, occur throughout the Hymenoptera, and are often reduced. In the Bethylinae they are present only in the basal taxa, Eupsenella (Figs. 2, 3) and Lytopsenella. In all remaining Bethylinae the notauli are absent. The notauli should not be confused with the parapsidal lines, which are an additional pair of shallow grooves lateral to the position of the notauli (Menke 1993) (Fig. 2). 0= notauli present; 1= absent. Presence of scutellar grooves or pits: In the basal taxa Eupsenella and Lytopsenella the scutellum has a pair of grooves at its anterior corners (Figs. 2, 3, 5). In several bethyline taxa 94 Journal of Hymenoptera Research pterostigma marginal cell 10. 11. 12. position of prostigma median cell submedian cell subdiscoidal cell (absent in Bethylinae) Fig. 1. Bethylidae: terminology of wing venation these have apparently become enlarged to form pits, often connected by a broad trans- verse groove (Figs. 4, 6, 7). 0= scutellar pits or foveae present as grooves; 1= scutellar foveae enlarged. Expansion of mesopleuron: Kieffer (1905) characterised his genus Odontepyris by the possession of dentate processes on the mesopleura (Figs. 6, 14). Although an apomorphic character in one lineage, it shows varying degrees of development in related taxa. 0= mesopleuron not expanded to form a dentate process; 1= mesopleuron moder- 13. ately expanded; 2=mesopleura with dentate processes. Presence of posterior transverse propodeal carina (Figs. 2, 6, 7, 9, 10): 0= posterior trans- verse propodeal carina present; 1= posterior transverse propodeal carina absent. Presence of median longitudinal propodeal carina (Figs. 2, 5, 6, 8, 9): 0= median longitu- dinal propodeal carina present; 1= median longitudinal propodeal carina absent. Presence of discal longitudinal propodeal 14. carinae (Figs. 2, 5, 6, 7, 8, 9, 10): 0= discal longitudinal propodeal carinae present; 1= discal longitudinal propodeal carinae absent. Presence of median propodeal pits or foveae (Figs. 4, 7, 10): The presence of these struc- tures is characteristic of the genus Prosierola. 0= median propodeal pits absent; 1= median propodeal pits present. 15. Development of a smooth, triangular area on the dorsal propodeum: In several bethyline taxa, particularly Goniozus and Prosierola, the proximal dorsum of the propodeum is characterised by a smooth, unsculptured triangular area which shows modifications in several taxa (Figs. 7, 10, 11). It is absent in the basal groups. The develop- ment of this character in some bethyline lin- eages should not be confused with the "propodeal triangle" in Apidae and Sphecidae (Brothers 1976). 0= smooth, trian- gular area absent from dorsal propodeum (Figs. 2, 5, 6, 8, 9); 1= smooth, triangular area present on dorsal propodeum (Figs. 7, 10, 11). Petiole ventrally keeled: To our knowledge, this character has not been used previously in defining the genera of Bethylinae, but is important for separating Goniozus from su- perficially similar taxa in Odontepyris (see key to genera, below). In most Bethylinae, a complete longitudinal keel is present on the ventral petiole (Fig. 13). This keel is reduced in several taxa (e. g. Fig. 15). 0= petiole with a complete ventral keel; 1= petiolar keel re- duced; 2= petiolar keel absent. Expansion of the fore femora: Expansion of the fore femora for fossorial use is often en- countered among the Bethylidae, and also occurs in related chrysidoid taxa and other aculeates. Other modifications of the fore femora are frequent in the Aculeata. 0= fore femora strongly expanded; 1= fore femora less strongly expanded, half as wide as long. Development of the prostigma: The expan- sion of the junction of the subcostal and basal veins into a secondary pterostigma is charac- teristic of the genus Goniozus (Figs. 20, 25), but also occurs elsewhere within the subfam- Volume 3. 1994 Table 2. Character matrix: 95 Characters 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Lytopsenella sp. 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Eupsenella sp. 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 1 0 0 0 0 Sierola sp. 0 1 2 1 1 0 0 0 1 0 0 0 0 0 1 0 0 0 0 1 0 1 Bethylus cq>halotes 1 1 1 1 1 0 0 0 0 0 0 0 2 0 0 0 0 0 1 1 1 1 Odontepyris flavinervis 0 0 1 0 1 1 2 1 1 1 0 0 0 1 0 1 1 0 0 1 1 0 Odontepyns transvaalensis 0 0 1 0 1 1 1 1 1 1 0 0 0 0 0 1 1 0 1 1 1 0 Odontepyns xanthoneurus 0 0 1 0 1 1 1 1 1 1 0 0 0 1 1 1 1 0 0 1 1 0 Prosierola sp. 0 0 1 0 1 1 1 1 0 1 1 1 0 0 0 1 1 0 0 1 1 0 Conwzus sp. 0 0 1 1 1 0 0 1 0 0 0 1 1 0 1 0 0 0 0 1 1 0 Goniozus gallicus 0 0 1 1 1 0 0 0 0 0 0 1 1 0 1 0 0 0 0 1 1 0 Gomozus mdicus 0 0 1 1 1 0 0 0 0 0 0 1 1 0 1 0 0 0 1 1 1 0 ily. Its presence outside Goniozus has previ- ously led to the generic misplacement of certain species. 0= prostigma absent; 1= 19. prostigma present. 16. Length of rs+m: In the basal taxa, fore wing vein rs+m forms the upper distal part of the discoidal cell (Fig. 1). In all Goniozus species it is present and well-developed, despite the fact that in Goniozus (sensu stricto) the discoi- dal cell is open. The genus Odontepyris is here characterised partly by the reduction of this vein (Figs. 22-24). The absence of this vein in Bethylus is clearly convergent with its reduc- 20. tion in Odontepyris (see below). 0= rs+m long; as long as, or longer than, rs (Figs. 16-18, 20, 25); 1= rs+m shorter than rs, or absent (Figs. 19, 22-24). 17. Length of rs: The length of rs relative to m (Fig. 1) characterises the Prosierola / 21. Odontepyris clade, in which rs has become elongated. 0= rs equal to or shorter than m (Figs. 16-20, 25); 1= rs clearly longer than m (Figs. 21-24). 18. Length of marginal (radial) cell: In Eupsenella, 22. one of the basal taxa, the marginal cell is characteristically shortened (Fig. 17). In the remaining taxa which have a closed marginal cell {Lytopsenella, Sierola), the marginal cell is of normal dimensions. For those taxa in which the marginal cell is open, it is necessary to calculate the size of a hypothetical closed marginal cell by the length of the radial vein (2r - rs of Sorg 1988; r + Rs of Evans 1964). 0= marginal cell long (Fig. 16); 1= marginal cell shortened (Fig. 17). Presence of discoidal cell (areolet): Within all Bethylinae the discoidal cell appears to be homologous, being plesiomorphically present. In Goniozus it is either clearly present or absent, whereas in Odontepyris various intermediate degrees of reduction of the dis- coidal cell are apparent (c. f. Figs. 22, 23, 24). 0= discoidal cell present (Figs. 16, 17, 18, 21, 22, 23, 25); 1= discoidal cell absent (Figs. 19, 20, 24). Presence of submarginal cell In the Bethylinae, the submarginal cell is present and complete only in Eupsenella and Lytopsenella. It is absent in all remaining taxa. 0= submarginal cell present (Figs. 1, 16, 17); 1= submarginal cell absent (Figs. 18-25). Presence of closed marginal cell: In the basal taxa, and in Sierola, the marginal cell is closed, the plesiomorphic condition for Bethylinae. 0= marginal cell closed (Figs. 1, 16-18); 1 = marginal cell open (Figs. 19-25). Angle of radial vein: The radial vein (Fig. 1; = vein 2r - rs of Sorg 1988; vein r + Rs of Evans 1964) is distinctly angled in Bethylus and Sierola (Figs. 18, 19), genera respectively with an open and a closed marginal cell. The angle of the radial vein is thus independent of the closure of the marginal cell. 0= radial vein without a sharp angle; 1 = radial vein sharply angled. 96 Journal of Hymenoptera Research Fig. 2. Bethylidae: terminology of mesosoma. The character matrix was analysed using the 'implicit enumeration' option of Hennig86, the procedure which computes the maximum pos- sible number of cladograms, or 'trees'. Multistate characters (3: number of maxillary palps; 5: pres- ence of notauli; 7: expansion of the mesopleuron; 13: reduction of the petiolar keel) were treated both as unordered (non-additive) and ordered (additive). RESULTS A single cladogram (tree) resulted from the analysis, having the following characteristics: length 35 steps, consistency index 71 and retention index 80 (Fig. 26). Ordering the data had no effect on the topology of the tree, only increasing its length by one step, and reducing the consistency index by two. Characters Supporting Monophyly of Bethyline Genera / Clades 1. (node A, Fig. 26): {(Prosierola + Odontepyris) + ((Sierola + Bethylus) + Goniozus))). Monophyly of this clade is supported by the following synapomorphies: 3: reduction of number of maxillary palp segments from six to five; 5: loss of notauli; 20: loss of submar- ginal cell; 21: marginal cell open (reversed in Sierola). 2. (node B, Fig. 26): (Prosierola + Odontepyris) Monophyly of this clade is supported by the following synapomorphies: 6: expansion of the scutellar pits; 7: expansion of the mesopleura; 8: presence of a posterior trans- verse propodeal carina; 10: presence of discal carinae; 16: rs+m shorter than rs; 17: rs longer than m. 3. (node C, Fig. 26): Odontepyris Volume 3. 1994 97 Figs. 3-4. 3, Eupsenella sp.: detail of mesosoma. 4, Prosierola sp.: scutellum and propodeum. Monophyly of Odon tepyris is supported by the following synapomorphy: 9: presence of a median propodeal carina. 4. (node D, Fig. 26): Prosierola Monophyly of Prosierola is supported mostly by convergent characters, except for the autapomorphic character (11) presence of propodeal pits. 5. (node E, Fig. 26): ((Sierola + Bethylus) + Goniozns)) Monophyly of this clade is supported by a single synapomorphy: 4: loss of the frontal streak. 6. (node F, Fig. 26): Goniozus Monophyly of Goniozus is supported by a single synapomorphy: 13: reduction of the petiolar keel. Character 12, presence of a smooth triangular area on the propodeum is a synapomorphy for Goniozus which occurs as a homoplasy in Prosierola. 7. (node G, Fig. 26): (Sierola + Bethylus) Monophyly of this clade is supported by the following synapomorphies: 2: reduction of the number of labial palp segments from three to two; 22: radial vein sharply angled. 8. (node H, Fig. 26): Sierola Monophyly of Sierola is supported by 3: max- illary palps 4-segmented, as well as by the closed marginal cell (21 ) which is here a rever- sal. 9. (node I, Fig. 26): Bethylus Monophyly of Bethylus is supported by the following synapomorphies: 1: reduction of the number of antennal segments from 13 to 12; 13: petiolar keel absent. DISCUSSION The analysis supports monophyly of our modi- fied concept of Odontepyris, including Trissomalus Kieffer, and Parasierola Cameron sensu Kieffer (1914), not Cameron (1883). The currently accepted definitions of most of the remaining genera are also supported by the analysis. Trissomalus (Kieffer 1905) was characterised in a key, but not formally diagnosed until Kieffer's (1914) generic revision. Odontepyris (Kieffer 1904) was characterised mainly by the possession of dentate processes on the enlarged mesopleura 98 Journal of Hymenoptera Research Figs. 5-7. 5. Eupsenella sp.: mesosoma. 6, Odontepyris sp.: mesosoma. 7, Prcsierola sp.: mesosoma. (Fig. 14). We have examined many described and undescribed species belonging to the Odontepyris I Trissomalus group, and conclude that the degree of variation in the development of the mesopleural processes, coupled with a successive reduction in the size of the discoidal cell (Figs. 22-24) suggests the group should be treated as a single genus, for which Odontepyris Kieffer is the oldest available name. The alternative would be to treat as new genera all intermediate species or species-groups. Examination olAnoxus specimens shows that the genus differs from Bethylus only in the extent of setation of the eyes, other characters are identical in the two genera. Furthermore in some Bethylus species some setation can be found on the eyes (e. g. B. amoenus Fouts). Since we know that the de- gree of eye setation is a character that clearly varies interspecifically within other bethyline genera, we do not see how retaining Anoxns as a distinct genus can be justified, and it is therefore synonymised below. GENERIC SYNONYMY ODONTEPYRIS Kieffer Odontepyris Kieffer 1904: 378. Type species Odontepyris flavinervis Kieffer 1904 by original designation. Holo- type female: SUMATRA: Pangherang-Pisang x. 90 e iii. 91 (E. Modigliani) (MCSN) [examined]. Trissomalus Kieffer 1905: 105. Type species Goniozus transvaalensis Du Buysson, 1888: 354 by subsequent des- ignation. Holotype female: [SOUTH AFRICA:] Transvaal, Hamman's Kraal 1893 (E. Simon) (MNHN) [examined] syn. n. Diagnosis. — Antennae 13-segmented. Palpal formula 5:3. Frontal streak present. Notauli ab- sent. Scutellar foveae well developed. Mesopleuron expanded, sometimes developed into a dentate prominence (Fig. 14). Median, discal and posterior transverse carinae present on propodeum, the longitudinal carinae occasionally reduced. Petiole with a complete ventral keel. Prostigma present or absent, rs + m reduced, rs elongate. Discoidal cell present, reduced or absent. Marginal and submar- ginal cells absent. Radial vein smoothly curved. Included species. — Odontepyris argyriae Kurian; O. batrae Kurian; O. cameroni (Kieffer) (comb. n. from Trissomalus, holotype examined); O. cirphi Kurian; O. erucarum (Szelenyi) (comb. n. from Parasierola); O. flavinervis (Kieffer) (holotype ex- amined); O. fuscicrus (Kieffer) (comb. n. from Trissomalus, holotype examined); O. hypsipylae (Kurian) (comb. n. from Goniozus); O. indicus Kurian (comb. n. from Trissomalus); O. moldavicus (Nagy) (comb. n. from Prosierola); O. peringueyi (Kieffer) (comb. n. from Trissomalus); O. quadrifoveatus (Muesebeck) (comb. n. from Parasierola, holotype examined); O. ruficeps Kieffer; Volume 3, 1 994 99 KEY TO GENERA OF BETHYLEMAE 1 Closed submarginal cell present (Figs. 1, 16, 17) 2 — Closed submarginal cell absent (Figs. 18-25) 3 2 Marginal cell elongate, larger than submarginal (Figs. 1, 16) Lytopsenella — Marginal cell short, smaller than submarginal (Fig. 17) Eupsenella 3 Marginal cell closed (Figs. 1, 16-18) Sierola — Marginal cell open (Figs. 19-25) 4 4 Antennae 12-segmented. Wing venation as in Fig. 19; radial vein at apex turned abruptly upward, but not reaching wing margin. Fore wing with rs+m always absent Bethylus — Antennae 13-segmented. Wing venation different; radial vein shorter and evenly curved towards wing apex (Figs. 20-25). Fore wing with rs+m present or absent 5 5 Propodeum without well-developed lateral carinae (fig 11). Scutellum without large foveae, with small grooves (as in Fig. 3). Petiole ventrally with a reduced, forked keel (Fig. 15) Goniozus — Propodeum with well-developed lateral carinae (Figs. 9, 10). Scutellum with large foveae (Figs. 6, 7). Petiole ventrally with a complete keel (Fig. 13) 6 6 Median longitudinal propodeal carina present (Fig. 9). Propodeum without median foveae Odontepyris — Median longitudinal propodeal carina absent (Fig. 10). Propodeum with median foveae (Fig. 10) Prosierola O. transvaalensis (De Buysson) (comb. n. from Goniozus, holotype examined); O. zvaterhousei (Kieffer) (comb. n. from Parasierola, holotype ex- amined); O. xanthoneurus (Kieffer) (comb. n. from Parasierola, holotype examined). Comments. — Tryapitsin (1978) and Terayama (1990) mistakenly included the Neotropical bethyline genus Prosierola in their respective keys to Russian and Japanese bethylid genera. In each case the error was due to misidentification of an Odontepyris species. BETHYLUS Latreille Bethylus Latreille, 1802: 315. Type species Omalus fusacorms Junne 1807: 301 by subsequent designation (Interna- tional Commision on Zoological Nomenclature, opin- ion 153). Holotype 'female [not examined, ?lost). Anoxus Thomson 1862: 451. Type species Anoxus boops Thomson 1862: 452. Monotypic. Lectotype female (Naturhistonska Rijksmuseet, Stockholm) [not exam- ined] syn. n. For full synonymy see Gordh & Moczar 1990. Diagnosis. — Antennae 12-segmented. Palpal formula 5:2. Frontal streak absent. Eyes with or without long hairs. Notauli absent. Scutellar foveae narrow, groove-like. Mesopleuron not expanded. Propodeum without carinae. Petiole with the ven- tral keel absent. Prostigma and discoidal cells absent. Marginal and submarginal cells absent. Radial vein usually sharply angled. Species transferred from Anoxus B boops (Thomson) comb. n. ; B. coniceps (Kieffer) comb. n. , B pilosus (Kieffer) comb. n. . Comments. — It might appear questionable to synonymise/iHOXKS with Bethylus without having examined the type species of either genus. How- ever, there has not been any controversy sur- rounding these genera since Kieffer's (1914) ge- neric revision, and we have examined sufficient material conforming to the original descriptions and Kieffer's interpretations of Anoxus and Bethylus. DISTRIBUTION OF BETHYLINAE The most primitive bethyline, Lytopsenella, is known only from two extant species, both from Chile. Three fossil species are known from Baltic amber (Evans 1964). Eupsenella is known so far only from Australia. Odontepyris is predominantly Palaeotropical, but its distribution spans southeastern Europe to South Africa and northern Queensland. The small genus Prosierola is primarily Neotropical although one species ranges into the extreme southern United States; a fossil species from Baltic Amber is certainly incorrectly assigned to Prosierola (Brues 1933). Goniozus, with about 150 described species, 100 Journal of Hymenoptera Research ■m Figs. 8-11. 8, Eupsenella sp.: propodeum. 9, Odontepyris sp.: propodeum. 10, Prosierola sp.: prapodeum. 11, Goniozus sp.: propodeum. Volume 3. 1 994 101 Figs. 12-15. 12, CWoHf£7n/nssp.:head. 13, Odontepyns sp.: ventral showing dentate process. 15, Goniozus sp.: ventral petiole and is cosmopolitan although only one species is known from the Pacific Islands. Approximately three dozen species each were described from the Ori- ental, Nearctic and Neotropical regions, and about a dozen each from the Palaearctic, Ethiopian and Australian regions. A number of species have been introduced into other countries as biological control agents. A few fossil species have been petiole and sternite 2. 14, Odontepyris sp.: side of mesopleuron sternite 2. described from Baltic and Dominican amber. Sierola contains almost 200 described species, mostly from Hawaii, with three endemic species in Australia. The genus underwent tremendous speciation in Hawaii, and a secondary small spe- ciation in the Marquesas Islands. Evans (1978) suggested that the single Calif ornian species might be introduced, and this is possibly also true of the 102 Journal of Hymenoptera Research 17 19 25 Figs. 16-25. Fore wings of various Bethylidae: 16, Lytopsenella sp. 17, Eupsenella sp. ; 18, Sierola sp. 19, Bethylus sp. 20. Coniozus (sensu stricto) 21, Prosierola sp. 22-24, Odonttpyris spp. 25, Goniozus ("Parasierola") sp. Volume 3, 1 994 103 26 F ■ H G ■ ■ I Lytopsenella sp. Eupsenella sp. Odontepyris flavinervis Odontepyris xanthoneurus Odontepyris transvaalensis Prosierola sp. Goniozus sp. Goniozus gallicus Goniozus indicus Sierola sp. Bethylus cephalotes Fig. 26. Cladogram of the Bethylinae. single Chinese species. A fossil species has been described from Baltic amber (Sorg 1988). Bethylus, containing some 30 described spe- cies, was previously considered as Holarctic, and predominantly Palaearctic, only two species be- ing definitely known from the Nearctic. Recently, the first author examined several specimens of Bethylus from South Africa. Homalus lamplipennis Motschulsky was erroneously placed by Gordh & Moczar (1990) in Bethylus; Krombein (1987) trans- ferred it to Holepyris. BIOLOGY OF BETHYLINAE Bethylinae are characterized in part by para- sitizing lepidopterous larvae, whereas members of the other subfamilies attack primarily coleopter- ous larvae. A few exceptions to these host prefer- ences are documented in Bethylinae, Epyrinae and Pristocerinae, but the few known hosts of Mesitiinae are case-bearing coleopterous larvae. The preferred hosts within the Bethylinae are mostly microlepidopterous as listed by Evans (1978) and Gordh & Moczar (1990). However, there are a few anomalous or questionable host records. It should be noted that, under experimen- tal conditions, females of the epyrine Sclerodermus immigrans Bridwell (1920) will successfully para- sitize a variety of insect larvae other than their normal coleopterous host. These include other Hawaiian beetles, bees, wasps, ants, and even braconid and chalcidoid parasites of the normal host larva. Bridwell also reared S. immigrans from normal workers, dealate adults, and nymphs of the termite Neotermes castaneus (Burmeister), al- though not from the soldiers. Perhaps some of the questionable records below may reflect this capa- bility in other bethylids. Evans (1964) reported that a specimen of Lytopsenella herbsti (Kieffer) was labelled as attack- ing an adult cantharid beetle. This record is dubi- ous, inasmuch as hosts of all other bethylids are larvae. The only other records of bethylines attack- ing a host stage other than the larva are those of Nagy (1976). He reported Odonte^n/ris moldavicus (Nagy) (as Prosierola moldavica) as reared from "pupae of a nocruid moth" and Goniozus plugarui Nagy attacking a lepidopterous pupa. These 104 Journal of Hymenoptera Research records of bethylids reared from pupae are dubi- ous, and the most probable explanation is that "pupal parasites" were reared from larvae which had spun cocoons but had not yet pupated. Evans (1962) reported that Bethylus amoenus Fouts was reared from an olethreutid moth and a nitidulid beetle. The latter record was later omit- ted from Evans' (1978) list of host records, and to us seems doubtful. A species tentatively identified as Goniozus gestroi (Kieffer) was reported by Richards (1955) as being reared from larvae of the anobiid beetle, Lasioderma. The rearing was not questioned, only the specific identity of the Goniozus, and we therefore accept this record as authentic. Goniozus morindae Kurian (1952), de- scribed from a single male, was reared from the gall of acecidomyiid fly , Asphondyla morindae Mani in flowers ofMorinda tinctoria. The record is anoma- lous, and is the first for a dipteran as a host of a bethylid. The most recent authentic record of a hymenopterous host for a bethy line is that of Melo and Evans (1992) who reported Goniozus microstigmi Evans as being reared from a brood cell of the sphecid wasp Micros tigmus xylicola Melo, a predator of nymphal Thysanoptera. The nests were in abandoned beetle galleries in beams of an exposed roof, and a dozen G. microstigmi females were collected while walking near and entering nests of M. xylicola and M. similis Melo. So far as is known, all Bethylinae are gregari- ous parasitoids, laying a clutch of eggs on each host larva. The host is stung, sometimes repeat- edly, behind the gula. Paralysis of the host larva may be temporary or permanent, depending upon the species of wasp. The number of eggs per clutch varies according to the size of the host, as well as interspecifically (Gordh & Evans 1976), ranging from one or two to as many as 40. Placement of the eggs depends upon the species, eggs being depos- ited either intersegmentally or longitudinally, and either dorsally, ventrally or laterally. Eggs hatch about two days after oviposition, and the larvae complete feeding in 2 to 5 days. Each larva spins a cocoon on the substrate near the host remains. The pupal period varies from 8-14 days, depending upon the species and ambient temperature. Males emerge a day or two before females are ready to eclose; they have been observed chewing into a cocoon containing a female and mating with her while she is still teneral. The progeny from a clutch usually consists of a single male and a number of females, and sibling mating is common (Hardy 1992). Maternal care of the larvae has been ob- served in Bethylus and Prosierola, and the mother may subsequently mate again with one of her male offspring. Maternal care has also been ob- served in Goniozus (Hardy & Blackburn 1991). Adults of both sexes feed on honey in culture, and females have been observed feeding on haemolymph exuding from the paralysed host. ACKNOWLEDGMENTS This work was carried out while the first author was in receipt of a grant from the Directorate General for Interna- tional Cooperation of the Netherlands Government (DGIS). We thank the following for the loan of material: J. Casevitz- Weulersse and M. LaChaise (Museum National d'Histoire Naturelle, Pans MNHN), E. De Coninck (Musee Royal de l'Afrique Centrale, Tervuren, Belgium), V Rainen (Museo Civico di Stona Naturale, Genoa, Italy MCSN) and M Terayama (Biological Laboratory, Toho Institute of Educa- tion, Tokyo). J. M. Carpenter (American Museum of Natural History, New York), H. E. Evans (Dept of Entomology, Colo- rado State University), I. C. W Hardy (Dept of Population Biology, University of Leiden), A. Menke (Systematic Ento- mology Lab. , USDA) and M. Mickevich (Maryland Center for Systematic Entomology, University of Maryland) are thanked for their comments on the manuscript, and R. W. Hodges (Systematic Entomology Laboratory, USDA) for as- sistance with Lepidoptera systematics. LITERATURE CITED Brothers, D. J. 1976. Modifications of the metapostnotum and origin of the 'propodeal triangle' in Hymenoptera Aculeata. Systematic Entomology 1: 177-182. Brown, W. L. 1987. Neoclystopsenella (Bethylidae) a synonym of Tamnoma (Formicidae). Psyche 94: 337. Brues, C T. 1907. Notes and descriptions of North American parasitic Hymenoptera IV. Bulletin of the Wisconsin Natu- ral History Society 5: 96-1 1 1 Brues, C. T. 1933. The parasitic Hymenoptera of the Baltic Amber. Part 1. Bernstein-Forschungen 3 (1932): 4-172. Cameron, P. 1883. Descriptions of new genera and species of Hymenoptera. Transactions of the Entomological Society of London 1883: 187-197. Du Buysson, R 1898. Voyage de M. E. Simon dans l'Afrique australe (janvier-avril 1893). Annates de la Societe Entomologiaue de France 66: 351-363. Evans, H. E. 1962. The genus Bethylus in North America Breviora 16: 1-12. Evans, H. E. 1964 A synopsis of the American Bethylidae (Hymenoptera: Aculeata). Bulletmof the Museum of Com- parative Zoology, Harvard Unwersity 133:67-151. Evans, HE. 1978. The Bethylidae of America north of Mexico. Memoirs of the American Entomological Institute 27: 1-332. Volume 3. 1994 105 Farns, J. S. 1988. Henmg86 Reference. Version 1. 5. User's Manual 17 pp. Gordh, G. and H. E. Evans. 1976. A new species of Gomozus imported into California from Ethiopia for the biologi- cal control of pink bollworm and some notes on the taxonomic status of Parasierola and Gomozus (Hy- menoptera: Bethylidae). Proceedings of the Entomological Society of Washington 78: 479-489. Gordh, G. and L. Moczar. 1990. A catalog of the World Bethylidae (Hymenoptera: Aculeata). Memoirs of the American Entomological Institute 46: 1-364. Hardy, I. C. W. 1992. Non-binomial sex allocation and brood sex ratio variances in the parasitoid Hymenoptera. Oikos 65: 143-158. Hardy, I. C. W. and T. M. Blackburn. 1991. Brood guarding in a bethylid wasp. Ecological Entomology 16: 55-62. Hempel, A. 1934. A Prorops nasuta Waterston no Brasil. Archizvs do lnstituto Bwlogico Sao Paolo 5: 197-212. Junne, L. 1807. Nouvelle methode de classer les Hymenopteres et les Dipteres. J. J. Paschaud, Geneva. 1-319. Kieffer, J. J. 1904. Description de nouveaux Dryininae et Bethylinae du Musee Civique de Genes. Annali Museo Civico di Storm Naturale di Genova 41: 351-412. Kieffer, J. J. 1905. Description de nouveaux proctotrypides exotiques. Annales de la Societe Saentiftque de Bruxelles29: 95-142. Kieffer, J. J. 1907 (1906). Beschreibung neuer in Naturhistorischen Museum zu Hamburg aufbewahrter Proctotrypiden and Evaniiden. Berliner Entomologische Zeitschrift 51: 258-278. Kieffer,]. J. 1911. Nouveaux Bethylides et Dry mides exotiques du British Museum du Londres. Annales de la Societe Scientifique de Bruxelles 35: 200-233. Kieffer, J. J. 1914. Hymenoptera, Proctotrupoidea, Bethylidae Das Tierreich 41: 1-595. Krombein, K. V. 1987. Synonymic notes on the Bethylidae described by V. de Motschulsky (Hymenoptera: Aculeata). Proceedings of the Entomological Society of Wash- ington 89: 356-358. Kurian, C. 1952. Descriptions of four new and record of one known Bethyloidea (Parasitic Hymenoptera) from In- dia. Agra University journal of Research 1: 63-72. Kurian, C. 1955. Bethyloidea (Hymenoptera) from India Agra University Journal of Research 4: 67-155. Latreille, P. A. 1802. Histoire Naturelle des Crustaces et Insectes. Tome 3, 1-467. Melo, G. A. R. and H. E. Evans. 1993. Two new Mtcrostigmus species (Hymenoptera, Sphecidae), with the descrip- tion of their parasite, Gomozus microstigmi sp. n. Proceed- ings of the Entomological Society of Washington 95: 258-263. Menke, A. S. 1993. Notauli and parapsidal lines: just what are they? Sphecos 24: 9-12. Muesebeck, C. F. W. 1940 Two new hymenopterous para- sites of sugarcane borers in India. Proceedings of the Entomological Society of Washington 42: 120-122. Nagy, C. G. 1976. Bethylidae (Hymenoptera) parasitizing orchard caterpillars. Revue Roumame de Biologie, Serie de Bwlogie Animate 21: 103-108. Richards, O. W. 1955. On the Bethyloidea (Hymen. ) of Israel. Bulletin of the Research Council of Israel 4: 357-359. Sorg, M. 1988. Zur phylogenie und systematik der Bethylidae (Insecta: Hymenoptera: Chrysidoidea). Geologisches lnstitut der Umversitaet zu Koeln Sonderveroeffentlichungen 63: 1-146. Terayama, M. 1990. Keys to the Japanese Bethylidae (Hy- menoptera, Aculeata) I. Subfamilies and Genera. Bulle- tin of the Toho Gakuen 5: 19-43. Thomson, C. G. 1862. Svenges Proctotruper (Fortsattning). Svenska Vetenskapsakadenuen. Stockholm. Oefrersigt af K. Academinens. Forhandlmgar. 18: 451-453. Tryapitsyn, V. A. 1978. [Identification of the insects of the European part of the USSR. Volume 3. Hymenoptera. Bethyloidea, Bethylidae. ] Opredeliteh Faune SSSR 120: 3- 16 (in Russian). J. HYM. RES. Vol. 3, 1994, pp. 107-117 Variation in the Haemolymph Protein Composition of Confined Apis Mellifera and Partial Restoration of Vitellogenin Titre by Juvenile Hormone Analogue Treatment M. M. G. Bitondi, Z. L. P. Simoes, A. M. do Nascimento and S. L. Garcia Dept. Biology, Faculdade de Filosofia, Ciencias e Letras de Ribeirao Preto Umversidade de Sao Paulo, CEP. 14040-901 Ribeirao Preto, S.P., Brasil (AMN and SLG are undergraduate fellows of the Convenio CNPq/USP) Abstract . — Haemolymph proteins, especially vitellogenin ( Vg), were investigated in confined Apis mellifera workers, that were fed different diets and treated with juvenile hormone (JH) I, III or with pyriproxifen (PPN). Vg and total protein concentrations were drastically decreased in the haemolymph of workers removed from the colony and confined for different periods of time. SDS-PAGE analysis demonstrated that confinement also caused induction and repression of the synthesis of certain haemolymph proteins. All of these changes occurred even when the confined workers were fed a protein-rich diet. In workers with Vg deficiency induced by confinement PPN, but not JHI or JHIII, induced a partial increase in Vg concentration. INTRODUCTION The vitellogenin (Vg) of Apis mellifera, the precursor of vitellin — the major protein of yolk — is a glycolipoprotein produced in the fat body (Harnish and White 1982; Trenczek and Engels 1986; Shapiro et al. 1988; Wheeler and Kawooya 1990). In most insects, Vg synthesis may be con- trolled by the diet, which obviously provides the materials and energy needed for this process (Bianchi and Pereira 1987; Bownes 1989; Bownes and Reid 1990). However, other factors in addition to diet are also involved. Many studies have been conducted on the role of neurosecretory cells (Elliott and Gillott 1978) and neurohormones (Keeley and Mckercher 1985; Keeley et al. 1988; Girardie et al. 1992), and of juvenile hormones and ecdysteroids in the regulation of Vg synthesis (Adams et al. 1985; Borowsky et al. 1985; Hagedorn 1985; Schwartz et al. 1985; Wojchowsky and Kunkel 1987; Adams and Filipi 1988; Keeley et al. 1988; Ma et al. 1988; Roseler and Roseler 1988; Wyatt 1988; Bownes 1989; Bownes and Reid 1990; Davis et al. 1990; Hatakeyama and Oishi 1990; Yin et al. 1990; Agui et al. 1985, 1991; Don- Wheeler and Engelman 1991;Sochaeffl/. 1991). In Apis mellifera, a highly eusocial insect, the control exerted by the queen over the workers represents an additional factor influencing Vg synthesis. A queen pheromone inhibits oocyte development in the workers. As a result Vg is not incorporated into the oocytes, although Vg is de- tected in the workers haemolymph. However, in queenless colonies the Vg titre of workers in- creases, reaching a level similar to that observed in the queens, followed by oocyte growth and ovipo- sition (Engels 1974; Engels and Fahrenhorst 1974). How Vg synthesis is regulated in A. mellifera still remains an interesting question, and aspects of this regulation have been studied in queens, workers and drones. In queens Vg synthesis does not depend on functional corpora allata and is not mediated by JH (Engels and Ramamurty 1976; Kaatz 1985). Similarly, JH topical application does not increase Vg synthesis in drones (Trenczek et al. 1989). But in workers, Rutz etal. (1976) and Flurief al. (1977) observed that low JH doses applied on 6 day old workers stimulate Vg synthesis whereas high doses have an inhibitory effect. Furthermore, Rutz et al. (1976) observed a correlation in vivo between low JH titre and Vg synthesis. Within the first few days after worker emergence, character- ized by low JH titres in haemolymph there is an increase in Vg synthesis. After this period Vg 108 Journal of Hymenoptera Research synthesis decreases while JH titre increases. This increase in JH titre was confirmed by Huang et al. (1991) and the temporal changes in Vg titre ob- served by Rutz et al. (1976) were similar to those related by Engels et al. (1990). The investigation of factors that affect A. mellifera Vg synthesis can help to understand how this protein is regulated. In the present work, the effect of some factors such as changes in social environment (confinement of workers with or without a queen) and diet (protein-rich or not) and JH or PPN (pyriproxyfen, 2-[l-methyl-2(4- phenoxyphenoxy) ethoxyl] pyridine, a JH ana- logue, were investigated not only on worker Vg synthesis but also on other haemolymph proteins. MATERIALS AND METHODS Apis mellifera We used "wild type" Africanized Apis mellifera bees (hybrids of European A. m. ligustica, A. m. carriica, A. m. mellifera and the African bee A. m. scutellata) from the Experimental Apiary of the Department of Genetics, Faculty of Medicine of Ribeirao Preto, University of Sao Paulo. Combs containing workers ready to emerge were removed from colonies and placed in an incubator whose temperature (34°C) and R.H. (80%) were similar to those in the colony. The workers that emerged within 15-20 hours were collected. About 100 newly emerged workers were marked on the thorax and put back into a small colony, formed by a queen, approximately 3000 workers (hive and forager bees), brood (eggs, lar- vae and pupae) and, sometimes, a few drones. Presence of nectar and pollen into the combs were also checked. Confinement The remaining workers were separated into groups of 150-200 and immediately submitted to confinement in 8 x 11 x 13 cm wooden cages with a sliding glass door and a screened bottom. The workers in these cages were placed in an incubator at 34°C and 80% R.H. and confined for 6 days (short confinement) or 15-16 days (long confine- ment). The confined bees received water and food ad libitum. The diet consisted of 50% sugar in water (syrup), a mixture of pollen from the comb (bee- bread) and candy (powdered sugar and honey), or only candy. In three experiments, three naturally mated queens aged 60-90 days were removed from the respective colonies and confined with groups of 150-200 newly emerged workers. These groups formed by workers and a queen were confined for 15-16 days, and fed on the mixture of beebread and candy. Water was also supplied. Treatment of confined workers with JHIII, JHI and PPN Some groups of 150-200 workers confined for 6 days were treated with JHIII, JHI or PPN applied topically to the abdominal cuticle. Each worker received 1 ul of a hormone solution in acetone at a concentration of 1 ug/ul, administered in two equal doses, the first immediately after emergence, before the introduction into the cages, and the other on the third day after rapid anesthesia with gaseous nitrogen. Two worker groups were treated with JHIII, two groups with JHI and nine groups with PPN. Three control groups were prepared in paral- lel: the first consisted of marked workers reintro- duced into the colony (control a), the second formed by confined workers treated with two 1 ul doses of acetone on the first and third days, respectively (control b). This group was also submitted to a rapid anesthesia with gaseous nitrogen at third day, immediately before acetone treatment. The third group consisted of untreated confined work- ers (control c). All worker groups, except control a, were allowed to feed ad libitum on the mixture of beebread and candy. Water was also supplied. Haemolymph For collecting haemolymph the workers were anesthetized with gaseous nitrogen and immobi- lized on dissection plates. Haemolymph was ex- tracted through a small superficial incision in the dorsal cuticle between the 2nd and 3rd tergites. Haemolymph was withdrawn from 6 day old and 15-16 day old confined workers that were fed on different diets, treated or not with hormones or acetone and mantained with or without a queen (Table I). Haemolymph was also extracted from colony reared workers (6 or 15-16 day old) and from newly emerged workers before confining or Volume 3, 1 994 109 returning them to the colony. Haemolymph pools were prepared from groups of at least 80 workers obtained from the same confinement cage. Similar pools were pre- pared with the haemolymph of at least 20 workers of the same age maintained in the colony and of 1 2 newly emerged workers. Phenylthiourea was added to the pools. Haemolymph pools were cen- trifuged at 3080 g for 10 min at 0°C and the super- natant was stored at -20°C. Rocket Immunoelectrophoresis Immunoelectrophoresis was used for the quan- titative determination of the vitellogenin fraction in haemolymph. Monospecific vitellogenin anti- serum produced in rabbits (Simoes 1980) was added at a 1% concentration to 1% agarose gels prepared with 0.06 M Tris-HCl buffer, pH 8.6. Immunoelectrophoresis was carried out at 10°C for 16 hours, at 0.08 V/cm gel. The same buffer used in the gel was used in the electrode compart- ments at a concentration of 0.3 M. The gels were stained with Coomassie Blue R-250. The height of the peaks detected (reported as mm) was propor- tional to the amount of antigen. The values ob- tained for confined workers were compared to those for workers of the same age maintained in colonies. The Vg peaks detected in colony reared workers (control c) was considered to be 100%. Total protein Protein concentration in the haemolymph pools was determined using bovine serum albu- min as a standard (Bradford 1976). SDS-PAGE Soluble haemolymph proteins were separated by SDS-PAGE according to the method of Laemmli (1970) except that SDS was not used in the separat- ing and stacking buffers. A 5-12% acrylamide gradient was used on a 0.7-mm thick gel. Electro- phoresis was run at 12mA constant current at 10°C until bromophenol blue tracking dye reached the bottom of the slab. Haemolymph samples (5 \i\) from confined and newly emerged workers (diluted 1: 2, v/v, in sample buffer) and from workers maintained in the colonies (diluted 1: 20 or 1: 40, v/v) were applied to the acrylamide gel. RESULTS Confinement blocks the increase of haemolymph Vg The confinement of workers for a period of 6 days after emergence prevented the increase of Vg titre in haemolymph that normally occurs in work- ers in colonies. This occurred even when bee- bread, the natural source of bee protein was sup- plied to the confined bees. In some of the haemolymph pools an extremely low Vg titre was detected (Fig. 1, wells 5 and 6), but, on average, Vg titres in these pools corresponded to 6.0 ± 2.1% (Table II) of that present in workers maintained in the colonies under natural conditions (Fig. 1 , wells 1 and 2). A small increase in Vg titre was observed in 15-16 day confined workers (Fig. 1, wells 3 and 4). This only occurred when the workers were fed beebread and candy. Even with this diet, the Vg concentration was much lower than in workers of the same age maintained in the colonies (Fig. 1, wells 1 and 2). Confinement changes the protein pattern of haemolymph obtained by SDS-PAGE The pattern of soluble haemolymph proteins from six day confined workers (Fig. 2, lane 4) differed from that observed in workers main- tained in colonies during the same period of time (Fig. 2, lane 3). In the confined bees, in addition to the fact that the Vg band (as determined by Trenczek et al. 1989) was very weak, the a and c polypeptides were not observed, whereas the b polypeptide formed a strong band. This polypep- tide corresponded to a weak band in the workers living in colonies. These were the most obvious differences, however differences between low molecular weight polypeptides, were also ob- served. In Fig. 2, lanes 1, 2 correspond to the haemolymph protein pattern of newly emerged workers, collected immediately before the bees were confined or returned to the colony. This pattern changed as the bees maintained in the colony aged (Fig. 2, lane 3). However, the changes depended on the social environment as shown by the protein pattern of confined workers (Fig. 2, lane 4). 110 Journal of Hymenoptera Research vg ^*t|l'"iWWwwi«AM*' J | 2 3 4 5 6 Fig. 1. Rocket Immunoelectrophoresis of worker haemolymph. 150 |il queen egg antiserum in 15 ml agarose gel. Antigen: 2 ul of a haemolymph pool per well. Staining: Coomassie Brilliant Blue R-250. Adult workers (6 days) main- tained naturally in the colony (1, 2), confined for 15-16 days (3, 4) and fed beebread and candy and confined for 6-days (5, 6) and fed the same diet. The haemolymph pattern of workers confined for 15-16 days and fed candy and pollen (Fig. 3, lanes 5 and 6) did not differ from that obtained after a short confinement (Fig. 2, lane 4). However, a discrete increase was observed in the Vg band of workers confined for 15-16 days. The value of diet protein content for protein synthesis is clearly shown in Fig. 3, lanes 3, 4, corresponding to the protein pattern of a haemolymph pool from work- ers maintained on a carbohydrate (syrup) diet for 15-16 days. In the same figure, the protein pattern of haemolymph from confined workers can also be compared with that of newly emerged ones (lanes 7, 8) and with that of workers maintained in the colony for 15 days (lanes 1, 2). There is no Vg in workers confined for 15-16 days with a queen Workers confined for 15-16 days in the ab- sence of a queen had Vg in their haemolymph b c 4 5 6 7 8 9 0 M(kDa) Fig. 2. SDS-PAGE (5.0-12%). Coomassie Blue staining. Pat- terns of worker haemolymph proteins. (1, 2) Newly emerged workers. Haemolymph diluted 1: 2 (v/v) in sample buffer; (3, 7) 6-day old workers maintained naturally in the colony Haemolymph diluted (3) 1: 20 (v/v) and (7) 1! 40 (v/v) in sample buffer. Workers confined for 6 days after emergence and treated with PPN (10), JHI (5,9), JHIII (6), or acetone (8, 11), or untreated (4). Haemolymph diluted 1: 2 (v/v) in sample buffer. (M) - molecular weights in kDa according to marker proteins. Note: Columns should be numbered 1-11. when fed on the beebread and candy mixture, as can be seen by Immunoelectrophoresis Fig. 1, wells 3 and 4. However if a queen is confined together with a group of 150-200 workers during this same time interval, from emergence until 15-16 days, the workers will not have Vg in the haemolymph. Thus, the queen effect on Vg synthesis could be observed even in a different environment condi- tion, i.e. that established by confinement. The haemolymph protein SDS-PAGE pattern of these workers was similar to that of bees confined in the absence of a queen, except for a weak Vg band present in the latter (results not shown). PPN induces Vg titre increase in workers confined for 6 days Workers confined for six days after emer- Volume 3. 1994 111 vg Wt* b ^-Htm* 180 116 .97 •66 -45 I — 29 12 3 4 5 6 7 8 M(kDa) Fig. 3. SDS-PAGE (5.0-12%), Coomassie Blue staining. Pat- terns of worker haemolymph proteins. (1, 2) 15-day old workers maintained naturally in the colony. Haemolymph diluted 1: 20 (v/v) in sample buffer; (3, 4) workers confined for 15-16 days after emergence and fed syrup or (5, 6) bee- bread and candy; haemolymph diluted 1: 2 (v/v) in sample buffer; (7, 8) Newly emerged workers. Haemolymph diluted 1: 2 (v/v) in sample buffer. (M) - molecular weights in kDa according to marker proteins. gence, fed beebread and candy and treated with PPN consistently showed a significant increase (p<0.001) in haemolymph Vg concentration (Fig. 4, wells 3, 4, and Fig. 5, wells 3, 4) when compared to confined workers treated with acetone (Fig. 4 wells 1, 2, 7, 8; Fig. 5 well 1) or untreated (Fig. 4, wells 9, 10). This increase corresponded to ap- proximately 17% of the Vg levels in the haemolymph of workers naturally maintained in the colonies (Fig. 4 wells 11, 12; Fig. 5, well 5). To calculate Vg percentage, we used a total of 9 haemolymph pools derived from 9 experiments, each consisting of 150-200 workers treated with PPN. In this case, Vg concentration in control a (workers maintained in the colonies) was consid- ered to be 100% (Table II). Table II also shows that the percentage of Vg detected in acetone-treated workers (control b, 7.4 + 3.2%) did not differ (0.5>p>0.4) from that detected in the untreated control c (6.0 ±2.1%). The Vg titre detected in workers treated with PPN was similar to that detected in workers con- fined for 15-16 days and feeding on pollen and candy (Fig. 1, wells 3 and 4). This Vg concentra- tion, however, was never detected in controls c (workers confined for six days and feeding on pollen and candy) or b (workers confined for six days, feeding on pollen and candy and treated 3 4 7 8 9 10 11 12 13 14 15 16 Fig. 4. Rocket Immunoelectrophoresis of worker haemolymph. 150 ul queen egg antiserum in 15 ml agarose gel. Antigen: 2 ul of a haemolymph pool per well. Staining: Coomassie Brilliant Blue R-250. Adult workers confined for 6 days after emergence and treated with JHIII (13, 14), JHI (5, 6, 15. 16), PPN (3, 4), or acetone (control b, 1, 2, 7, 8), or untreated (control c, 9, 10). These confined workers were fed beebread and candy. Wells 11 and 12 correspond to the Vg of adult 6-day old workers maintained naturally in colonies. 112 Journal of Hymenoptera Research o 1 2 I Fig. 5. Rocket Immunoelectrophoresis of worker haemolymph. 150 ul queen egg antiserum in 15 ml agarose gel. Antigen: 2 (il of a haemolymph pool per well. Staining: Coomassie Brilliant Blue R-250. Adult workers confined for 6 days after emergence, fed beebread and candy and treated with acetone (control b, well 1), JHI (well 2), or PPN (wells 3 and 4). Well 5 corresponds to the Vg of adult 6-day old workers maintained naturally in colonies. with acetone). Thus, under conditions of prolonged confinement, there is an increase in Vg concentra- tion in haemolymph during the second week of confinement, but this increase never occurs dur- ing the first week of confinement unless the work- ers are treated with PPN. PPN does not change the SDS-PAGE protein pattern The protein pattern (Fig. 2) from workers con- fined for six days and treated with PPN (lane 10) was similar to that observed in the acetone-treated (lanes 8, 1 1) and untreated controls (lane 4), except by an increase in Vg band. PPN does not significantly change the total protein concentration in the haemolymph There was no significant difference in haemolymph protein content between PPN-rreated workers and their controls: acetone treated (con- trol b, p>0.35) and untreated (control c, p>0.15). (Table III). Comparison between untreated groups (con- trol c) and the groups maintained in the colonies (control a) demonstrated a drastic decrease in haemolymph protein concentration in the first group, maintained under confinement conditions, even when beebread and candy was supplied. JHI and jHIll did not induce Vg titre increase in workers confined for 6 days Treatment with JHI (Fig. 4, wells 5, 6, 15, 16 and Fig. 5, well 2) or JHIII (Fig. 4, wells 13, 14) under the same experimental conditions as those with PPN did not increase Vg concentration in haemolymph. The Vg peaks detected in the work- ers treated with JHI or JHIII did not differ statisti- cally (p>0.15 and p>0.5, respectively) from the acetone-treated control b (Fig. 4, wells 1, 2, 7, 8 and Fig. 5, well 1 as showed in Table IV). However, the Vg peaks obtained from workers treated with JHI, a hormone not synthesized by Apis mellifera (Robinson et al. 1987), were found to be slightly higher than the peaks obtained for workers treated with acetone or with JHIII, the natural hormone of these bees. This result should be considered with caution since it is based only on the data obtained for two haemolymph pools from workers treated with JHI or JHIII. But we do not exclude the possibility of this homologue (JHI) being more effective on Apis mellifera than JHIII. JHI and JHIII did not change the SDS-PAGE protein pattern The haemolymph protein pattern of workers confined for six days and treated with JHI (Fig. 2, lanes 5, 9) or JHIII (Fig. 2, lane 6) was similar to that observed in the controls treated with acetone (Fig. 2, lanes 8, 11) or untreated (Fig. 2, lane 4). Volume 3, 1 994 113 DISCUSSION Effect of confinement, queen presence and diet on haemolymph protein composition Our results show that the normal Vg titre in workers mainly depends on social environment established in the colony. When workers were removed from the colony, and maintained during 6 days in an appropriate environment (where mortality was practically zero) and on a protein- rich diet, profound physiological changes occurred, that inhibited Vg titre and probably Vg synthesis. The initiation of Vg synthesis in these workers can be observed if confinement is lenghthened for 15- 16 days. But the haemolymph Vg peak detected by immunoelectrophoresis in these workers (after 15-16 days of confinement) is smaller than that observed in colony reared workers. The onset of Vg in the haemolymph of 15-16 days confined workers is dependent on administration of a pro- tein-rich diet: Vg is not detected in these confined workers fed a diet without protein such as syrup, or a low protein content diet, such as candy made with honey. Besides if a queen is confined with the workers for 15-16 days no Vg is produced al- though a protein rich diet (beebread and pollen) had been available. The influence of factors related to the social environment, on Vg synthesis has been studied in social Hymenoptera, especially the interaction among individuals of different castes in the colo- nies. Reproductively active queens characteristi- cally inhibit Vg synthesis and egg-laying in other females capable of reproduction. This fact has been well documented in A. mellifera (Engels et al. 1990), Melipona (Engels and Imperatriz-Fonseca 1990), and Camponotus festinatus (Martinez and Wheeler 1991) among other social insects. We verify that A. mellifera queens can inhibit Vg syn- thesis in workers even under conditions of con- finement, i.e., far from the normal colony environ- ment. Engels et al. (1990) maintained groups of 25 workers for three weeks confined from emergence on a piece of comb, in the presence and absence of a queen. They detected that both worker groups had Vg in the haemolymph, but the Vg titre of workers confined in the absence of a queen was higher than that of workers confined in the pres- ence of a queen. In our experiments, we found that workers confined without a queen for 15-16 days produced Vg when fed beebread and candy. How- ever, when the workers were confined with a queen during this same period of time, no Vg was detected in the haemolymph. Perhaps, the experi- mental conditions employed by Engels et al. 1990, with a piece of comb (with brood?) within the confinement cage, provided a more favorable en- vironment (maybe more similar to that of a normal colony) that permitted Vg synthesis even in the presence of a queen. For comparison, other condi- tions employed in both experiments should also be taken into account, such as the age of the queens used, the size of the worker population and the time of confinement. Confinement affects not only Vg synthesis, which is also controlled by the queen and by the food available, but also the synthesis of other haemolymph proteins, as determined by SDS- PAGE. The polypeptide a showed marked reduc- tion, the polypeptide c was not detected, whereas another, b, showed markedly increased titer (Figs. 2 and 3). This indicates that confinement can si- multaneously provoke contrasting gene expres- sions. This was observed in confined workers fed proteic or non-proteic diets. Workers maintained on a syrup diet had a lower protein content, but similar protein pattern in the haemolymph (Fig. 3) when compared to workers maintained on a proteic diet. Thus the social interactions and not the pro- tein supply are responsible for the contrasting gene expressions mentioned above. Action ofPPN, JHIII and ]H1 on haemolymph protein composition changed by confinement A significant (p<0.001) increase in Vg titre ocurred in workers with confinement-induced Vg deficiency, after treatment with PPN. However the observed induction was partial. Apparently other factors are also involved in the regulation of normal Vg synthesis. The action of these factors may be prevented or impaired by the confinement conditions employed. PPN seems to be specific for Vg synthesis, since the haemolymph SDS-PAGE pattern of the other proteins in confined workers apparently did not change when treated with this JH analogue. However as the small increase in Vg titre induced by PPN was not reflected in total protein measure- ment we can not exclude the possibility that PPN also partially inhibits other haemolymph 114 Journal of Hymenoptera Research protein(s). Contrary to what is observed with PPN, the natural JHIII (Robinson et al. 1987) and its homo- logue JHI did not significantly increase Vg con- centration in haemolymph. JHIII and JHI also did not seem to modify the haemolymph protein pat- tern induced by confinement. We should consider that PPN may have caused an increase in Vg titre in haemolymph by being a more potent analogue. PPN is considered to be one of the most effective juvenile hormone analogues known for locusts (De Kort and Koopmanschap 1991), with a strong juvenilizing effect on these insects. PPN has a strong morphogenetic effect when topically ap- plied to A. mellifera larvae. Larvae treated with 1 fig undergo metamorphosis, but the pupae show dras- tic changes in pigmentation, especially in the eyes and thorax, and can also die before emergence (Bitondi et al., unpublished data). Similarly to what occurs in Locusta, the effect of PPN on A. mellifera larvae and pupae is more drastic when compared to the effects induced by JHIII or JHI. Kaatz (1985) proposed a model of regulation of Vg synthesis in A. mellifera queens. According to this model, Vg synthesis may be influenced by JH but also by ecdysteroids and by a haemolymph factor. These factors mentioned by Kaatz (1985) may be involved in the regulation of normal Vg synthesis in workers. However, the simple adap- tation to workers of the model proposed for queens should be considered with caution since some data obtained in studies on Vg synthesis regula- tion in queens indicate that this regulation may differ between the two castes (Engels et al. 1990). In summary the absence of one or more types of social interaction induced by confinement caused a reduction of total protein content in haemolymph, induction and repression of protein synthesis and impairment of Vg synthesis in A. mellifera workers. Only haemolymph Vg titer can be partially recovered by PPN treatment of con- fined workers. ACKNOWLEDGMENTS This work was supported by grants from the Fundacao de Amparo a Pesquisa do Estado de Sao Paulo (FAPESP) Proc. 91/2587-8. We thank Dr. H Oouchi, Sumitomo Chemical Co., Ltd, Osaka, Japan, for providing the pynproxyfen, and PR. Epifanio and L. R. Aguiar for the technical assistance We thank also Dr. David De Jong for correcting the English. LITERATURE CITED Adams, T. S. and P A. Filipi. 1988. Interaction between ]uvenile hormone, 20-hydroxyecdysone, the corpus - cardiacum - allatum complex, and the ovaries in regulat- ing vitellogenin levels in the housefly, Musca domestica. journal of Insect Physiology 34: 11-19. Adams, T. S., H. H. Hagedorn and G. D. Wheelock. 1985. Haemolymph ecdysteroid in the housefly, Musca domestica, during oogenesis and its relationship with vitellogenin levels, journal of Insect Physiology 31: 91-97. Agui, N., S. Izumi and S. Tomino. 1985. The role of ecdysteroids and juvenoids in vitellogenin levels and follicle devel- opment in the housefly, Musca domestica. Applied Ento- mology and Zoology 20: 179-188. Agui, N., T. Shimada, S. Izumi and S. Tomino. 1991. Hor- monal control of vitellogenin mRNA levels in the male and female housefly, Musca domestica. Journal of Insect Physiology 37: 383-390. Bianchi, A. G. de and S. D. Pereira. 1987. Time of synthesis of Musca domestica vitellogenin during the first gonotrophic cycle. Comparative Biochemistry, and Physiology 86B(4): 697-700. Borovsky, D., B. R. Thomas, D. A. Carlson, L. R. Whisenton and M. S. Fuchs. 1985. Juvenile hormone and 20- hydroxyecdysone as primary and secondary stimuli of v\te\\oger\esism Aedes aegypti. Archwes of Insect Biochem- istry and Physiology 2'. 75-90. Bownes, M. 1989. The roles of juvenile hormone, ecdysone and the ovary in the control of Drosophila vitellogenesis. Journal of Insect Physiology 35(5): 409-413. Bownes, M and G. Reid. 1990. The role of the ovary and nutritional signals in the regulation of fat body yolk protein gene expression in Drosophila melanogaster. jour- nal of Insect Physiology 36(7): 471-479. Bradford, M.. 1976. A rapid and sensitive method for the quantification of microgram quantities of protein utiliz- ing the principle of protein dye binding. Analytical Bio- chemistry 72: 248-254. Davis, R. E., T. J. Kelly, E. P. Masler, H. W. Fescemyer, B. S. Thyagaraja and A. B. Borkovec. 1990 Hormonal control of vitellogenesis in the gypsy moth, Lymantria dispar (L.): suppression of haemolymph vitellogenin by the juvenile hormone analogue, methoprene. Journal of In- sect Physiology 36(4): 231-238. De Kort, C. A. D. and A. B. Koopmanschap. 1991. A juvenile hormone analogue affects the protein pattern of the haemolymph in last-instar larvae of Locusta nugraloria. Journal of Insect Physiology 37(2): 87-93. Volume 3, 1 994 115 Don-Wheeler, G. and F. Engelmann. 1991. The female - and male - produced vitellogenins of Leucophaea maderae. Journal of Insect Physiology 37(12): 869-882. Elliot, R. H. and C. Gillot. 1978. The neuro-endocrine control of protein metabolism in the migratory grasshopper, Melanoplus sangumipes. Journal of Insect Physiology 24: 119-126. Engels, W. 1974. Occurrence and significance of vitellogenins in female castes of social Hymenoptera. American Zoolo- gists 14: 1229-1237. Engels, W. and H. Fahrenhorst. 1974. Alters - und kasten - spezifische Veranderungen der Haemolymph - Protein - Spektren bei Apis mellifica. Wilhelm Roux' Archiv 174: 285-296. Engels, W. and V. L. Imperatriz-Fonseca. 1990. Caste devel- opment, reproductive strategies, and control of fertility in honey bees and stingless bees. In: Engels, W. (ed.) Social Insects. An evolutionary approach to castes and repro- duction. Springer-Verlag, Berlin, ppl67-230. Engels, W., H. Kaatz, A. Zillikens, Z. L. Paulino-Simoes, A. Trube, R. Braun and F. Dittrich. 1990. Honey bee repro- duction: vitellogenin and caste-specific regulation of fertility. In: M. Hoshi and O. Yamashita (eds.) Advances in invertebrate reproduction 5, 495-502. Elsevier, Amsterdam. Engels, W and P. S. Ramamurty. 1976. Effects of carbon dioxide on vitellogenin metabolism in unmated queen honeybees. Journal of Apiculture Research 15: 3-10. Fluri, P., H. Wille; L. Gerig and M. Luscher. 1977. Juvenile hormone, vitellogenin and haemocyte composition in winter worker honey bees (Apis mellif era). Experientia 33: 1240-1241. Girardie, J., O. Richard and A. Girardie. 1992. Time-depen- dent variations in the activity of a novel ovary maturat- ing neurohormone from the nervous corpora cardiaca during oogenesis in the locust, Locusta migratoria migratorwides. Journal of Insect Physiology 38: 215-221. Hagedorn, H. H.. 1985. The role of ecdysteroids in reproduc- tion. In Comprehensive Insect Physiology, Biochemistry and Pliarmacology. (Edited by G. A. Kerkut and L. I. Gilbert), vol. 8, pp 205-262. Pergamon Press, Oxford. Harnish, D. G. and B. N. White. 1982. Insect vitellins: identi- fication, purification, and characterization from eight orders. Journal of Experimental Zoology 320: 1-10. Hatakeyama, M. e K. Oishi. 1990. Induction of vitellogenin synthesis and maturation of transplanted previtellogenic eggs by juvenile hormone III in males of the sawfly, Athalia rosae. Journal of Insect Physiology 36: 791-797. Huang, Z. Y., G. E. Robinson, S. S. Tobe, K. J. Yagi, C. Strambi, A. Strambi and B. Stay. 1991. Hormonal regulation of behavioral development in the honey bee is based on changes in the rate of juvenile hormone biosynthesis. Journal of Insect Physiology 37(10): 733-741. Kaatz, H. H.. 1985. Initiation und Regulation der Vitellogenin- Synthese bei der Bienenkbnigin (Apis mellifera L.). Inau- gural-Dissertation, Universitat Tubingen, pp. 146-159. Keeley, L. L. and S. C. McKercher. 1985. Endocrine regula- tions of ovarian maturation in the cockroach Blaberus discoidales. Comparatwe Biochemistry and Physiology 80A: 115-121. Keeley, L. L., S. M Sowa, T. K. Hayes and]. Y. Bradfield. 1988. Neuroendocrine and juvenile hormone effects on fat body protein synthesis during the reproductive cycle in female Blaberus discoidalis cockroaches. General Com- paratwe Endocrinology 72: 364-373. Laemmli, U. K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685. Ma, M, J. -Z. Zhang and H. Gong. 1988. Permissive action of juvenile hormone on vitellogenin production by the mosquito Aedes aegypti. Journal of Insect Physiology 34: 593-596. Martinez, T. and D. Wheeler. 1991. Effect of the queen, brood and worker caste on haemolymph vitellogenin titre in Camponotus festinatus workers. Journal of Insect Physiol- ogy 37(5): 347-352. Robinson, G. E., Strambi, A., Strambi, C, Paulino-Simoes, Z.L., Tozeto, SO., Barbosa, J. M. N. 1987. Juvenile hor- mone titers in Africanized and European honey bees in Brazil. Gen. Comp. Endocrinol. 66: 457-459. Roseler, P. F. and I. Roseler. 1988. Influence of juvenile hormone on fat body metabolism in ovanolectomized queens of the bumblebee, Bombus terrestris. Insect Bio- chemistry 18(6): 557-563. Rutz, W., L. Gerig;, H. Wille and M. Luscher. 1976. The function of juvenile hormone in adult worker honey- bees, Apis mellifera. Journal of Insect Physiology 22: 1485- 1491. Schwartz, M. B., T. J. Kelly, R. B. Imberski and E. C. Rubenstein. 1985. The effects of nutrition and methoprene treatment on ovarian ecdysteroid synthesis in Drosophila melanogaster. Journal of Insect Physiology 31: 947-957. Shapiro, J. P., J. H. Law and M. A. Wells. 1988. Lipid transport in insects. Annual Rei'iew of Entomology 33: 297-318. Simoes, Z. L. P. 1980. Estudo da vitelogenina e da vitelina em Apis mellifera L. (Hymenoptera : Apoidea). Tese de doutoramento, Universidade de Sao Paulo, pp.: 1-108. Socha, R., J. Sula, D. Kodrik and I. Gelbic. 1991. Hormonal control of vitellogenin synthesis in Pyrrhocoris apterus (L.) (Heteroptera). Journal of Insect Physiology 37( ): 805- 816. Trenczek, T. and W. Engels. 1986. Occurrence of vitellogenin in drone honeybees. International Journal of Invertebrate Reprodution and Development 10: 307-311. Trenczek, T., A. Zillikens and W. Engels. 1989. Developmen- tal patterns of vitellogenin haemolymph titre and rate of synthesis in adult drone honey bees (Apis mellifera). Journal of Insect Physiology 35: 475-481. Wheeler, D. E. and J. K. Kawooya. 1990. Purification and characterization of honey bee vitellogenin. Archives of Insect Biochemistry and Physiology 14: 253-267. Wojchowski, D. M. and J. G. Kunkel. 1987. Purification of two distinct oocyte vitellins and identification of their corre- sponding vitellogenins in fat body and hemolymph of Blaberus discoidalis. Insect Biochemistry 17: 189-198. Wyatt, G. R.. 1988. Vitellogenin synthesis and the analysis of juvenile hormone action in locust fat body. Canadian Journal of Zoology 66: 2600-2610. Yin, C. -M„ B. -X. Zou, S. -X. Yi and J. G. Jr. Stoffolano. 1990. Ecdysteroid activity during oogenesis in the black blow- fly, Phormia regina (Meigen). Journal of Insect Physiology 36: 375-382. 116 Journal of Hymenoptera Research Table I. Haemolymph pools prepared from workers confined for 6 or 15-16 days in the presence or absence of a queen, fed on different diets and submitted or not to hormonal treatment No. of Confinement Diet Hormonal or pools (days from acetone analysed emergence) treatment 1 6 Syrup _ 1 6 Candy - 5 6 breebread and candy - 2 6 breebread and candy JHIII 2 6 breebread and candy JHI 9 6 breebread and candy PPN 6 6 breebread and candv Acetone 2 15-16 Syrup - 1 15-16 Candy - 3 15-16 breebread and candy - 3 15-16 breebread and candy - Queen absent absent absent absent absent absent absent absent absent absent present Table II - Vitellogenin (Vg) in haemolymph pools from confined workers treated with PPN or acetone (control b), from untreated confined workers (control c) and from workers maintained in colonies (control a). PPN Acetone Untreated Maintainec treatec treated in the colonies nent (control b) (control c) (control a) Experir Peak Vg# Peak Vg# Peak Vg# Peak Vg# No. Heigh (mm) (%) Height (mm) (%) Height (mm) (%) Height (mm) (%) 1 5.0 16.13 3.5 11.29 1.5 4.39 31.0 100 : 5.5 17.74 * * * » • * 3 411 12.90 » * • * • • 4 6.5 20.97 * * * • • * 5 6.5 15.12 2.0 4.65 3.0 6.98 43.0 100 6 7.0 17.95 1 i) 2.56 3.0 7.69 39.0 100 7 8.5 21.80 3.5 8.97 3.0 7.69 + + 8 4.5 14.52 2 5 8.06 1.0 3.22 31.0 100 9 6.5 14.61 4 (I 8.99 z z 44.5 100 MEAN ±SD 16.9±3.0 7.4±3.2 6.0±2.1 100 # Percent Vg in relation to workers maintained in the colony (control a = 100% Vg); * Since experiments 1-4 were performed simultaneously, the same controls (a, b and c) were used in each; z There was no control c for experiment 9; + Experiments 6 and 7 were performed during subsequent weeks. The same control a was used. Volume 3, 1 994 117 Table III. Protein titre* (ug/|il haemolymph in BSA equivalents) of 6-day old workers confined from emergence and treated with PPN or acetone (control b), untreated (control c) and maintained in the colony (control a). Experiment PPN treated Acetone Untreated Maintained in No. treated the colonies (control b (control c) (control a) |lg/(ll M-g/^l Hg/|d \ig/\d 1 6.76 7.22 5.04 53.15 2 7.03 — — — 3 6.81 — — — 4 8.38 — — — 5 — — — 33.88 6 7.0 6.62 5.64 31.05 7 7.98 6.89 7.93 — MEANiSD 7.3+0.7 6.9±0.3 6.2+.1.5 39.4+12.0 (')Protein concentrations were only measured in 6 of the 9 experimental groups treated with PPN, in 3 of the 6 groups treated with acetone, in 3 of the 5 untreated groups, and in 3 of the 5 groups maintained in the colonies. Table IV. Vitellogenin (Vg) percent in haemolymph pools from confined workers treated with JHIII, JHI or acetone (control b) and from workers maintained in the colonies (control a). Experiment JHI JHIII Acetone Maintained in No. treated treated treated the colonies Peak Vg Peak Vg Peak Vg Peak Vg Height Height Height Height (mm) (%) (mm) (%) (mm) (%) (mm) (%) 1 4.5 10.00 4.0 8.89 * * 2 4.0 8.89 3.5 7.78 1.5 3.33 45 100 3 2.0 4.60 2.0 4.60 43.5 100 MEAN± SD 9.4+0.8 6.2±2.2 5.6±2.9 * Since experiments 1 and 2 were performed on subsequent days, the same control a was used. J. HYM. RES. Vol. 3, 1994, pp. 119-132 A Review of the Agile Species Group of Pison (Hymenoptera: Sphecidae: Trypoxylini) Alexander V. Antropov Zoological Museum of the Moscow State University, Herzen Street 6, Moscow K-9, 103009 Russia Abstract. — The agile species group of Pison is redescribed, the 12 included species are reviewed, and a key is provided for identification. The agile group is restricted to the Oriental and eastern Palearctic Regions, although one species, koreense, is adventive in North America. Five species are new: agiloides from Sri Lanka; chrysoptilum from Borneo; mngyuenfuense from southwestern China; vechti from Malaya and Indonesia; and pulawskii from India. Other species of the group are: agile (Smith) from southern India and Sri Lanka; erythropus Kohl from western India; koreense (Radoszkowski) from eastern Asia and North America; rothneyi Cameron from southeast Asia; browni (Ashmead) from the Philippines; different Turner from Assam, India; and hissancum Gussakovskij from Uzbekistan and Tajikistan. Lectotypes are designated for agile, differens and rufipes (Smith) and a neotype is designated for koreense. Pison koreense is removed from synonymy with agile. INTRODUCTION The genus Pison, which contains nearly 200 described species (Bohart and Menke 1976; Menke 1988), is well represented in all faunal regions except North America where only an adventive species, koreense, is known. In this paper I review the agile species group which contains 12 species, five of which are new. Members of this assemblage have been placed in the subgenera Pisonoides and Krombeiniellum, but Menke (1988) used species groups rather than subgenera for infrageneric groups. The agile group is restricted to the eastern Palearctic and the Orien- tal Regions, except for the east Asian species koreerxse which was introduced into North America presumably after World War II (Krombein 1958a). Morphological terminology used here follows Bohart and Menke (1976) and Menke (1988). The following abbreviations are used in the text: OOD = ocello-ocular distance; OD = ocellus diameter; POD = distance between posterior ocelli. The following museum and institutions lent specimens for this study (abbreviations are used in the text): AUZM BMNH Universiteit van Amsterdam, Zoologisch Museum, Amsterdam, The Netherlands (W. Hogenes). The Natural History Museum, London, England (C. R. Vardy, L. Ficken). CAS California Academy of Sciences, San Francisco, California (W. J. Pulawski). NMNH Nationaal Natuurhistorisch Museum, Leiden, The Netherlands (C. van Achterberg). OUM Hope Entomological Collections, University Mu- seum, Oxford, England (C. OToole). USNM U.S. National Museum, Washington, DC, USA (K. V. Krombein, A. S. Menke). ZIN Zoological Institute, Russian Academy of Sciences, St. Petersburg, Russia (V. I. Tobias). ZMK Zoological Museum, Copenhagen, Denmark (O. Lomholdt). The specimens from the collection of the Zoo- logical Museum of the Moscow State University (ZMUM) were also used in the review. THE AGILE GROUP This group is characterized as follows: com- pound eyes densely setose, antenna clavate, oc- cipital carina complete or nearly so, anterior pronotal pit small, subomaulus present, episternal sulcus straight, forewing with only two submar- ginal cells (true second submarginal cell lost through diminution), hindcoxa without dorsolat- eral carina, metapleural flange usually lamelli- form, propodeum without lateral carina or crenu- late ridge, male sternum VIII narrow, gonostyle simple, volsella small, penis valve compressed 120 Journal of Hymenoptera Research laterally and without teeth or notched ventrally. Menke (1988) regarded the setose eyes, the clavate antenna, the two submarginal cells, the presence of a subomaulus, the straight episternal sulcus, the lamelliform metapleural flange and claw shape as apomorphies of the agile group. Its species apparently represent a monophyletic as- semblage. Morphologically they are very similar and differ mainly in leg color, tergal bands, punc- tation, vestiture and proportions of tergum I. Male genitalia are also similar. One of the new species, pulaivskii, stands apart from the others in the group because its metapleural flange is narrow. I regard it as a plesiomorphic state. Two other species have unique autapomorphies: in browni the propodeal dorsum is delimited laterally by a shallow, broad sulcus and tergum I has a distinct preapical de- pression; in agiloides submarginal cell II is open distally. Description. — Inner orbits of eyes moderately emarginate, parallel (eyes equidistant at vertex and clypeus) or slightly converging below (rarely above); eyes covered densely with short, erect setae (Fig. 3); clypeus convex, in female rounded (Figs. 4a, 5a), in male angulate or prominent apically (Figs. 4b, 5b); frons convex; antennae clavate, com- paratively short, with distal flagellomeres wider than long; labrum subquadrate, truncate or slightly emarginate apically; occipital carina a complete (or almost complete) circle, narrowly separated from hypostomal carina; male mandible simple, that of female with inner tooth slightly distad of midpoint; pronotum with small round pit anteri- orly, without lamellae; scutum and mesopleuron moderately, uniformly punctate; episternal sul- cus almost straight, not curved forward ventrad; subomaulus recurved ventrad; omaulus and ac- etabular carina absent; mesopleural sulcus paral- leled anteriorly by a row of foveolae; metapleuron smooth; metapleural flange usually broadly lamel- liform posteriorly (Figs. 6-8); tegula entirely punc- tate; forewing media diverging after cu-a; forew- ing with two submarginal cells (Figs. 10-13), 2nd (really 3rd) usually not petiolate; recurrent veins received by 1st and 2nd submarginal cells or 2nd recurrent vein interstitial between 1st and 2nd submarginal cells; hind coxa dorsum with low inner carina, without outer carina; legs finely sculp- tured, without stout spines on tibiae and tarsomeres; all tarsomeres IV with small plantulae; tarsal claw thick to just before apex; propodeum rounded, without lateral carinae or lines of foveae and crenulate ridges, punctate with smooth interspaces; propodeal dorsum not delimited (browni with shallow lateral sulcus), with medium furrow containing short to complete ridge; abdo- men compact; tergum I simple or with preapical transverse depression (Figs. 15-19); apical bands of terga often translucent and with silvery or golden pubescence; male sternum VIII long, narrow, rounded or weakly notched apically; genitalia compact, compressed laterally; volsellae small, rounded and weakly setose; gonostyle triangular, simple, with long, coarse lateral setae curved be- neath; penis valve compressed laterally, consider- ably widened apically, without teeth or notches ventrally. Included Species. — agile (Smith), agiloides sp. n., browni (Ashmead), chrysoptilum sp. n., differens Turner, erythropus Kohl, hissaricum Gussakovskij, koreense (Radoszkowski), ningxjuenfuense sp. n., pulaivskii sp. n., rothneyi (Cameron), vechti sp. n. Biology. — Information is available for two spe- cies of the group, koreense and erythropus. The former was studied by Iwata (1964) in Japan and by Sheldon (1968) in North America. The Indian species, en/thropus, was studied by Home (1870). These wasps construct small, clay cells, placing them separately or in groups on variable surfaces, but not within linear cavities (borings in wood or empty stems of plants). The groups may include up to 21 cells which do not merge in a common mass but keep their independence. Prey consists of 6-31 paralyzed, small, usually immature spi- ders. KEY TO SPECIES OF THE AGILE GROUP 1 All femora completely reddish 2 — Mid- and hindfemora brown, forefemora mainly brown 6 2(1) Metapleural flange narrow (Fig. 9); India pulaivskii sp. n. — Metapleural flange broadly lamelliform posteriorly (Figs. 6-8) 3 Volume 3. 1994 121 3(2) Pronotum, scutum, scutellum, metanotum and propodeal dorsum and hindface with suberect golden setae; Borneo chrysoptilum sp. n. Thorax and propodeum with only silvery pubescence 4 4(3) All legs including tarsi and base of trochanters yellowish-red; western India erythropus Kohl — Trochanters, hintibiae apically and tarsi dark brown 5 5(4) Abdominal tergum I densely punctate (punctures separated by a diameter or less), dull due to dense microsculpture; translucent apical bands of terga I-III whitish, at middle hardly broader than hindtarsal diameter; Uzbekistan, Tajikistan hissaricum Gussakovskij — Abdominal tergum I finely, sparsely punctate (punctures more than a diameter apart), surface smooth, weakly shiny in spite of microstriae; translucent apical bands of terga golden, broad, those of 1 1— 1 1 1 at middle almost equal to diameter of hindtibia; Malaysia, Indonesia vechti sp. n. 6(1) Abdominal tergum I polished, sculpture nearly same as following terga 7 — Abdominal tergum I dull or weakly shiny in contrast to following terga 8 7(6) Tibiae reddish; scutal punctures fine, dense, less than a diameter apart; translucent apical bands of abdominal terga golden, apical bands of terga I-III at middle at least twice as broad as diameter of hindtarsomere I; submarginal cell II open distally (Fig. 14); Sri Lanka agiloides sp. n. — Mid- and hindtibiae brown, foretibia partly reddish; scutal punctures coarse, more than a diameter apart; translucent apical bands of abdominal terga mainly whitish, those on I-III equal to or hardly broader than diameter of hindtarsomere I (Fig. 15); submarginal cell II closed; southern India and Sri Lanka agile (Smith) 8(6) Lamelliform part of metapleural flange mainly dark; translucent apical bands of abdominal terga not broader than diameter of hindtarsomere I 9 Lamelliform part of metapleural flange reddish at least posteriorly; translucent apical bands of abdominal terga (at least of terga II-III) obviously broader than diameter of hindtarsomere I 10 9(8) Abdominal tergum I dull, very densely, finely sculptured, transversely depressed preapically (Fig. 17); apical bands of abdominal terga II-III dark brown, that of tergum I narrower than hindtarsal diameter; propodeal dorsum delimited by shallow depression at least laterally; Philippines browni (Ashmead) — Abdominal tergum I shiny dorsally, coarsely, sparsely punctate, faintly transversely depressed preapically; apical bands of abdominal terga II-III yellowish, band on tergum I almost as broad as hindtarsal diameter and half as wide as band on tergum II; propodeal dorsum not delimited by depressions; southeastern China ningyuenfuense sp. n. 10(8) Translucent apical bands of abdominal terga I - 1 1 bright golden, of equal width; propodeum as long as wide (seen from above), with hind surface mostly punctate apically; ocelli (especially in male) small (OOD>OD); India: Assam differens Turner Translucent apical bands of abdominal tergum I very narrow or absent; propodeum wider than long (seen from above), hind surface transversely carinate apically; ocelli larger (OODPOD>OOD. Clypeus, frons, vertex, pronotum, scutum, scutellum and mesopleuron finely, densely punc- 130 Journal of Hymenoptera Research tate (punctures 0.5-1.5 diameters apart), shiny; metapleural flange broad, spoon-shaped, widened part translucent; propodeum more coarsely, sparsely punctate laterally and dorsally (at least 2 diameters apart), hindface coarsely and densely punctate (0.5-1 .0 diameters apart), with transverse ridges ventrallv; propodeal dorsum not enclosed by sulcus or carina, with narrow median furrow containing simple carina; tergum I punctate (l-# diameters apart), shiny in spite of microstriahon; sternum I with coarser punctures (as on propodeal hindface), shiny; other abdominal sclerites micro- scopically but distinctly, densely and uniformly punctate (approximately a diameter apart), shiny; translucent apical margins of terga I-IV twice as broad as diameter of first hind tarsomere (on tergum V and sterna III-V slightly less than diam- eter). Recurrent vein I ending on submarginal cell I, recurrent vein II ending near base of submarginal cell II. Length 8.2 mm. Variation in Females (2 specimens). — Paratype female differs by its smaller size (6.2 mm) and almost parallel inner orbits of eye. Male. — Unknown. Discussion. — Unlike all other members of the agile group, chrysopdlwn has golden pubescence on the frons, vertex, thorax and propodeum. It resembles vechti because of its brightly colored legs and golden translucent bands on the abdomi- nal terga. Pison chrysoptilum and vechti may be two forms of one species, but more material will be necessary to resolve this. Etymology. — This species name is derived from the Greek words din/sos (=gold) and ptilon (=down, fluff) emphasizing the color of the pubescence. Range. — Northern Borneo (Sarawak, Brunei). Types. — Holotype female: MALAYSIA, Sarawak: 4th div., Gn. Mulu, RGS Exp., 17.IX- 23.X.1977, D. Hollis (BMNH). Paratype (1 female): BRUNEI: Ulu Temburong, Base camp hut, 300 m, 115 16'E,4 26'N, 16.II-9.III.1982, M. C. Day (BMNH). Pison pulawskii Antropov, new species Figs. 5, 9, 13, 19 Description of Holotype Female. — Black except the following reddish-orange: palpi, mandible largely, labrum, fore legs distad of coxal apex, mid and hind legs except basal parts of coxae and tarsomeres III-V; antennal articles beneath, apical margin of clypeus, tegula, mid and hind tarsomeres III-V reddish-brown. Pubescence silvery only; clypeus, frons, ver- tex, pronotal collar, scutum, scutellum, metano- tum, metapleuron and propodeum with erect se- tae, longest on propodeum posterolaterally and shortest on scutum (scarsely longer than on femora); gena, mesopleuron and abdomen with suberect or appressed short setae (longest on api- cal bands of terga and shortest on sterna). Labrum truncate apically; clypeus compara- tively broadly rounded (Fig. 5A); inner orbits of eyes almost parallel; OOD=OD=POD. Frons densely punctate, shiny; scutum very densely, finely punctate, weakly shiny; mesopleuron uniformly, sparsely punctate (1.0- 1.5 diameters apart), polished; metapleural flange not lamellate (Fig. 9); propodeal dorsum obliquely carinate basally and along median furrow, dor- sum not enclosed by sulcus or carina, sparselv punctate (as on propodeal side) (1-3 diameters apart); tergum I uniformly, densely ounctate (as on mesopleuron), shiny; tergum II more densely and finely punctate, shiny; remaining terga weakly shiny because of very dense microscopic punc- tures; sternum I dull, punctured as frons; sternum II shiny, punctate as mesopleuron; other sterna with dense micropunctures, weakly shiny; trans- lucent apical bands of terga I-V and sterna II-IV whitish, as broad as diameter of hindtarsomere I (Fig. 19). Recurrent vein I ending on submarginal cell I, recurrent vein II ending on submarginal cell II near its base (Fig. 13). Length 7.0 mm. Variation in Females (11 specimens). — Mid tarsi completely reddish-orange in three specimens from holotype locality and one from Kurumbagaram; hind tibiae and tarsi brownish posteriorly and antennal flagellum completely dark-brown in one specimen from Coimbatore. Length 6.0-7.0 mm. Males (3 specimens). — As in females except: clypeus with acute median prominence spicallv (Fig. 5b); labrum slightly emarginate; mandibles without inner teeth. Length 4.8-5.2 mm. Discussion. — The nonlamellate metapleural flange of pulawskii sets the species apart from other in the agile group. Other features of the species are the form of the clypeus, the mostly reddish-orange Volume 3. 1994 131 legs and the narrow whitish translucent apical bands of the abdominal segments. Reddish legs occur in erythropus and whitish abdominal bands occur in erythropus, agile and ningyuenfitense, but these species have a lamellate metapleural flange. Etymology. — This species is dedicated to Wojciech J. Pulawski. Range. — Western and southern India. Types. — Holotype female: INDIA, Rajasthan: Uddaipur (24 35'N), 27.V.1989, W. J. Pulawski (CAS). Paratypes (3 males, 9 females): same place, date and collector as holotype (CAS); same place and collector as holotype, 21-25. V. 1989 (CAS, ZMUM); Gujarat: Disa (=Deesa), 4-6. VI. 1989, W. J. Pulawski (CAS); Karnataka: Bangalore, 915 m, 26.V.1980, 30.V.1980, K. D. Ghorpade (ZMK); Karikal Territory: Kurumbagaram, III.1947, P. S. Nathan (USNM). ACKNOWLEDGMENTS I express my gratitude to all colleagues who lent mate- rial for study. I am grateful to Dr. A. P. Rasnitsyn for his help in searches for the type specimen of koreense in Radoszkowski's collection. I am extremely grateful to Dr. A. S. Menke who read earlier versions of the manuscript and suggested many important improvements of the text and who helped me in the English. Finally, I thank Dr. K. V. Krombein and Dr. W. J. Pulawski who critically and attentively reviewed the entire manuscript and who provided manv useful comments. LITERATURE CITED Ashmead, W. H. 1905. Additions to the recorded Hymenopter- ous fauna of the Philippine Islands, with descriptions of new species. Proceedings of the United States National Museum 28(1413): 957-971. Bingham, C. T. 1987. The fauna of British India. Hymenoptera, vol. 1. Taylor and Francis, London. 179 pp. Bohart, R. M. and A. S. Menke. 1976. Sphecid wasps of the World, a generic revision. University of California Press, Los Angeles, London, iv + 695 pp. Cameron, P. 1897a. IV. Hymenoptera Orientalia, or contribu- tions to a knowledge of the Hymenoptera of the Oriental zoological region. Part V. Memoirs and Proceedings of the Manchester Literary & Philosophical Society 41(2): 1-144. Dalla Torre, C. G. 1894. Catalogus Hymenopterorum hucusque descriptorum systematica et synonymicus. Vol. IX. G. En- gelmann, Lipsiae. 181 pp. Dalla Torre, C. G. 1897. Catalogus Hymenopterorum hucusque descriptorum systematica et synonymicus. Vol. VIII. G. Engelmann, Lipsiae. 750 pp. Gussakovskij, V. V. 1937. Especes palearctictiqques des genres Didmeis Wesm., Pison Latr. et Psen Latr. (Hymenoptera, Sphecodeal). Travaux de I'Institut Zoohgie de I ' Academie des Sciences de I'URSS 4(3-4): 599-695. Hawkins, W. A. 1974. A western record for an introduced Pison (Hymenoptera, Sphecidae) journal of the Kansas Entomological Society 47(2): 279. Home, C and F. Smith. 1870. Notes on the habits of some hymenopterous insect from the north-west provinces of India. With an appendix, containing descriptions of some new species of Apidae and Vespidae collected by Mr. Home. Transactions of the Zoological Society of London 7(3): 161-196. Iwata. K 1964. Ethological notes on four Japanese species of Pison (Hymenoptera, Sphecidae). Mushi 38: 1-6. Kazenas, V. L. 1980. Materials to the fauna of digger wasps (Hymenoptera, Sphecidae) of the Far East of the USSR. Taxonomy of insects of the Far East, Vladivostok, pp. 80-94. Kohl, F. F. 1885. Die Gattungen und Arten der Larnden Auctorum. Verliandlungen derk. k. zoologisch-botamschen Gesellschaft in Wien 34: 171-268. Krombein, K. V. 1958a. Pison tParaceramius) koreense (Rad.), a new adventive wasp in the eastern United States. Enfo- mologtcal News 69: 166-167. Krombein, K. V. 1958b. Superfamily Sphecoidea. In: Krombein, K. V., et al., Hymenoptera of America North of Mexico - Synoptic Catalog. Agriculture Monograph 2, First Supplement, U. S. Department of Agriculture, Washington, DC. pp. 186-204. Krombein, K. V. 1967. Superfamily Sphecoidea. In: Krombein, K. V., B. D. Burks et al., Hymenoptera of America North of Mexico — Synoptic Catalog. Agriculture Monograph 2, Second Supplement, U S. Department of Agriculture, Washington, DC. pp. 386-421. Krombein, K. V., P. D. Hurd, D. R. Smith and B. D. Burks. 1979. Catalog of Hymenoptera in American North of Mexico. Vol. 2. Smithsonian Institution Press, Washington, DC. pp. 1199-2209. Menke, A. S. 1968a. New genera and species of wasps of the tribe Trypoxylonini from the neotropical region (Hy- menoptera: Sphecidae: Larnnae). Los Angeles County Museum Contributions in Science 135: 1-9. Menke, A. S. 1968b. A review of the New World species of Pison. I. The subgenus Krombeiniellum (Hymenoptera: Sphecidae). The Canadian Entomologist 100(10): 1100- 1107. Menke, A. S, 1988. Pison in the New World: a revision (Hymenoptera: Sphecidae. Trypoxylonini). Contribu- tions of the American Entomological Institute 24(3): 1-171. Radoszkowski, O. 1887. Hymenopteres de Koree. Horae Scoitatis Entomologicae Rossicae 21(3-4): 428-436. Richards, O. W. 1962. A revisional study of the masarid wasps. Wm. Clowes and Sons, London. 294 pp. Sheldon, J. K. 1968. The nesting behavior and larval morphol- ogy of Pison koreense (Radoszkowski). Psyche 75: 107- 117. Shuckard, W. E. 1838. Descriptions of new exotic aculeate Hymenoptera. Transactions of the Entomological Society of London 2(1): 68-82. Smith, F. 1869. Descriptions of new species of the genus Pison, and a synonymic list of those previously described. Transactions of the Entomological Society of London. 9: 289- 300. Tsuneki, K. 1974. A contribution to the Knowledge of Sphecidae occuring in Southeast Asia (Hym.). Polskie 132 Journal of Hymenoptera Research Pismo Entomologiczne 44(3): 585-660. Tsuneki, K. 1983. Further studies on the Larrinae of the Philippine Islands, with remarks on the Indian species of the genus Lyroda (Hymenoptera, Sphecidae). Special Publications of the japan Hymenopterists Association 24: 1- 116. Turner, R. E. 1916. Notes on the wasps of the genus Pison and some allied genera. Proceedings of the Zoological Society of London 1916: 591-629. Yasumatsu, K. 1935. The genus Pison Spinola of the Japanese Empire (Hymenoptera, Trypoxylonidae). Annotationes zoologicae Japonenses 15(2): 227-238. Yasumatsu, K. 1939. Notes supplementaires sur le genre Pison Spinola du Japon (Hymenoptera, Trypoxylonidae). Festschrift zum 60 Geburtstage von Professor Dr. Embnk Strand, Riga 5: 81-84. J. HYM. RES. Vol. 3, 1994, pp. 133-143 Colony Densities and Preferences for Nest Habitats of Some Social Wasps in Mato Grosso State, Brazil (Hymenoptera, Vespidae) IVONE R. DiNIZ AND KlNITI KlTAYAMA Departamento de Zoologia, Institute de Biologia Universidade de Brasilia, 70910-900 Brasilia DF, BRAZIL Abstract — Studies of colony densities and preferences for nest habitats of some social wasps were conducted in cerrado vegetation in Southern Mato Grosso state in Central Brazil. Wasp colony densities were estimated in three habitats (campo umido, cerrado sensu-stricto and gallery forest). Regarding the wasp nest habitats preferences, we found 100 colonies of 30 species in 15 genera nesting in six different habitats in Southern Mato Grosso. In Northern Mato Grosso, based on Richards (1978), we found 199 colonies of 51 species in 14 genera nesting in six different habitats. Considering both regions together, we came out with 299 colonies of 61 species in 16 genera nesting in nine different habitats. INTRODUCTION METHODS AND STUDY AREA The effectiveness and efficacy of predation by- social wasps on many insects (Gobbi et al. 1984, Gobbi and Machado 1985, Machado et al. 1987, Raw 1988) confers on them a fundamental impor- tance as biological control agents. Nonetheless, the study of the nests of these insects has focused mainly on their architecture (Jeanne 1975, Kojima 1982) and on the numbers of individuals per colony (Richards 1978), with few reports on pairs of spe- cies nesting close together (Windsor 1972, Starr 1988). Information on preferences for nest habitats (Richards 1978, Reed and Vinson 1979, Forsyth 1980) and on colony densities (Rau 1942, Kitayama et al. 1989) of neotropical social wasps is still very scarce. In central Brazil, 130 species of wasps have been collected (Richards 1978, Raw pers. comm. and our own collections), but studies on their ecology are very fragmentary. The aim of this report is to add new informa- tion on colony densities and preference for nest habitats of social wasps in the cerrado sensu-lato of Central Brazil. The study was conducted in cerrado vegeta- tion during a five week survey (August, October through December 1988 and June 1989) at the Rio Manso Hydroelectric Power Station (HPS-Rio Manso) (14°52' S and 55°50' W), in the Chapada dos Guimaraes county, southern Mato Grosso State in central Brazil. Cerrado sensu-lato is a semideciduous xeromorphic vegetation dominant in Central Bra- zil, occupying about 20% of the whole country. It occurs in various structures from closed forest- like forms to pure grasslands (Eiten 1972). "Habi- tat" is used here to refer to a vegetation subtype within the cerrado sensu-lato. The censuses of social wasp colonies (nests with resident adults) were conducted in campo umido, vereda, campo sujo, cerrado sensu-stricto, gallery forest and surrounding dirt roads. Charac- terizing these habitats very briefly it could be said that campo umido is a wet grassland without visible woody plants (Fig. 1); vereda is a broad marshy valley bottom grassland with buriti palm (Mauritia vinifera) galleries (Fig. 2); campo sujo is a cerrado grassland with a few, very scattered, low conspicuous shrubs or acaulescent palms (Fig. 3); 134 Journal of Hymenoptera Research AiiP*:' =*#■;*> a-- S^^Ss M 1 K? Fig. 1. Area of 100 nv of campo umido where wasp nests were surveyed. Fig. 2. Area of Vereda (Photo by Dr. Roberto Cavalcante) 135 1 v^g»HK Fig. 3. Area of campo sujo. Fig. 4. Area of cerrado sensu-stricto (Photo by Dr. Roberto Cavalcante). 136 Journal of Hymenoptera Research '4MB£^4< Fig. 5. Area of gallery forest (Photo by Dr. Roberto Cavalcante). cerrado sensu-stricto includes forms with the total woody plant cover of about 30-40% and with a canopy generally less than seven meters (Fig. 4) and gallery forest is usually a narrow evergreen mesophytic forest following the streams (Fig. 5) (Eiten 1972). Searches for nests were conducted in four different months, covering two seasons: dry sea- son (June and August) and wet season (October and December) in six habitats. Each habitat re- ceived the same intensity of searching during each of these seasons. From each nest, adult wasps were collected to guarantee identification of the species. All the nests and insects sampled are deposited in the Laboratorio de Zoologia, Departamento de Zoologia in the Universidade de Brasilia. The data on Xavantina and Serra do Cachimbo (10°50'S and 51°47'W), in northern Mato Grosso state were compiled from Professor Richards' (1978) species descriptions of nest habitats using only the data where there was no doubt on the location of the nests). Richards' censuses were conducted in campo sujo, dirt roads and clearings, cerrado sensu-stricto, Cerradao (the medium tall arboreal form of cerrado with a closed or semi- open canopy with 30-40% tree crown cover, Eiten 1972), gallery forest and dry forest (deciduous and semi deciduous mesophytic forest, Eiten 1972). For the study on colony density, known areas were sampled, in gallery forest (7500 m2) and in cerrado sensu-stricto (10000 m2). Nests were lo- cated among leaves, on branches, on the trunks of trees and in holes in tree-trunks and in the ground. Binoculars were used to locate nests in the canopy. In addition, five squares of 100 m2 each were surveyed in campo timido, where nests of wasps were sought among tufts of grasses and sedges (Fig.l). The survey for estimates of the density of colonies, within the study area, was done during the wet season. Three people spent approximately 360 hours searching for nests. In each habitat the time spent per square meter was about one per- son-minute. For the estimates of density of adult wasps/ ha, the number of adults within some of the nests were counted and also, for some species, data from Richards (1978) were compiled. In each table the habitats, when it was pos- sible, were arranged in order of increasing com- plexity of structure because this makes it easier to see general patterns in species preferences for habitats. Volume 3. 1 994 137 RESULTS AND DISCUSSION 1. Nesting Habitat Preference of Wasps A total of 100 colonies, comprising 30 species in 15 genera of wasps were encountered in six different habitats in Chapada dos Guimaraes (Table 1). Among them, 18 species nested in only one habitat, nine in two, and three (Polybia (Myrapetra) ruficeps Schrottky — a very common species of the region, Chartergus chartarius (Olivier) and Clwrtegelhis communis Richards) in three habitats (Table 1). Of the species that nested in only one habitat, most were found in cerrado sensu-stricto (seven species), in gallery forest (five species) and in campo umido (four species) (Table 1). Of the 30 species collected in this study, only Polistes (Epicnemius) pacificus liliaciosus de Saussure was not collected by Richards (Table 1 and 2). This subspecies has been collected only rarely in cen- tral Brazil and, until Richards' 1978 record, its distribution was thought to be restricted in Brazil to the North (Amazonas, Amapa and Para states). Polybia ruficeps was the commonest species of wasp at Rio-Manso with 23 colonies, of which 70% were found in cerrado sensu-stricto. Three other species, Epipona tatua (Cuvier), Synoeca surinama (L) and Oiartergus chartarius, with eight colonies each, were also common (Table 1). Richards (1978) recorded 199 colonies com- prising 51 species in 14 genera nesting in six habi- tats (Table 2). Most of the species he collected (25 of the 51), were found nesting only in one habitat (eight of them in dry forest), 12 species nested in two habitats, eight nested in three habitats, three nested in four, and one species Polybia (Myrapetra) platycephala Richards nested in five habitats. In Xavantina and Serra do Cachimbo, Polybia (Myrapetra) occidentalis (Olivier) with 24 colonies (12%), was the most abundant, followed by Polybia ruficeps (10%), Polybia (Apopolybia) jurinei de Saussure (8%), Polybia (Trichothorax) ignobilis (Haliday) (7%) and Polybia (Formicicola) rejecta (F) (5%) (Table 2). The data collected by us and by Richards were combined and for the same type of habitat the data were lumped (Table 3 representing the sum of Tables 1 and 2). At the nine habitats 299 colonies, comprising 61 species and 16 genera, were col- lected. Twenty eight species (46%) nested in only one habitat, 13 species nested in two, 10 in three, seven in four and three in five habitats (Table 3). The area surveyed was different in each habitat so the number of colonies found do not represent one absolute abundance but the relative habitat pref- erence for nesting. Of the total number of colonies recorded by Richards and us from Mato Grosso, the species with the largest number of colonies was Polybia ruficeps (14%), followed by Polybia occidentalis (9%), Polybia jurinei (6%), Polybia (Myrapetra) erythrothorax Richards (5%), Polybia ignobilis (5%), Paraclmrtergus fratemus (Gribodo) (4%) and Epipona tatua (4%). The genus Polybia represents 60% of the colonies reported from the two regions of Mato Grosso State (Table 3), ranging from 67% in Xavantina and Serra do Cachimbo (Table 2) to 45% in Rio Manso (Table 1). Among all the colonies of the 19 species of Polybia collected, 58% were found in cerrado sensu-stricto (Table 3) ranging from 31% in Rio Manso and 55% in Xavantina and Serra do Cachimbo (Tables 1 and 2). Richards (1978) suggested that the choice of habitat for nesting is very characteristic and less diverse than that used for foraging. His opinion is confirmed in our survey at Rio Manso. For ex- ample, both Synoeca suririama (L), which nests in gallery forest, and Polybia (Trichothorax) sericea (Olivier) in cerrado sensu-stricto, were collected when they hunted in gallery forest, cerrado sensu- stricto, campo sujo and campo umido. Of the 30 species of wasps at Rio Manso, 21 (70%) were encountered foraging in more than one habitat (in preparation). A water source is important for wasps to nest successfully, (Rau 1942, Forsyth 1980). The latter author concluded that the greater colony density of wasps in gallery forest was due mainly to the presence of water. Wet habitats at Rio Manso (gallery forest, eight/ha and campo umido 60/ha) also had higher colony densities than did dry habitats (cerrado sensu-stricto five/ha). Polistes (Epicnemius) subsericeus de Saussure and Mischocyttarus (Mischocyttarus) drewseni de Saussure which nest in campo umido, had higher colony density compared to other species in drier habitats, such as cerrado sensu-stricto. Availability of water all year round and nest protection by tufts of grasses and sedges, could be the cause of that high density. Most animals have preferences for particular habitats (Partridge, 1978). In Mato Grosso ap- 138 Journal of Hymenoptera Research Table 1. Wasps nesting habitats in Rio Manso Chapada dos Guimaraes: Mato Grosso, Brazil. Species CU VE cs OF CE GF TT Apoica c.f. pollens 0 0 0 0 1 1 2 Brachygastra bilineolata 0 0 0 0 1 0 1 Cliartegellus communis 0 1 1 0 1 0 3 Clwrtergus chartarius 0 0 4 0 3 1 8 Epipona tatua 0 0 0 0 5 3 8 Metapolybia cingulala 0 0 0 0 1 0 1 Mischocyttarus cerberus 0 0 0 0 2 0 2 Mischocyttarus drewseni 2 0 0 0 0 0 2 Mischocyttarus labiatus 0 0 0 0 1 0 1 Mischocyttarus matogrossoensis 1 0 0 0 0 0 1 Mischocyttarus mcthathoracicus 0 0 0 0 1 0 1 Parachartergus fratemus 0 1 0 0 4 0 5 Polistes canadensis 1 0 0 0 0 0 1 Polistes pacificus 0 0 0 0 0 1 1 Polistes subsericeus 4 0 0 0 0 0 4 Polybia emaciata 0 0 0 0 0 1 1 Polybia erythrothorax 0 0 2 0 4 0 6 Polybia jurinei 0 0 0 0 2 0 2 Polybia occiden talis 0 0 0 1 2 0 3 Polybia paulista 0 0 0 0 0 2 2 Polybia quadricincta 0 0 0 2 0 0 2 Polybia ruficeps (1 0 0 5 16 2 2< Polybia sericea 0 0 0 1 2 0 3 Polybia singularis 0 11 11 0 1 2 3 Protopolybia exigua 0 0 0 1 1 0 2 Pseudochartergus chartergoides 0 0 0 0 0 1 1 Pseudopolybia compressa 0 0 0 0 0 1 1 Pseudopolybia vespiceps 0 (1 0 0 1 0 1 Stelopolybia lobipleura 0 0 1 0 0 0 1 Synoeca surinama 0 0 0 2 0 6 8 Total of nests 08 02 08 12 49 21 100 Total of species 04 02 04 06 18 11 30 CU - Campo umido VE - Vereda CS - Campo sujo OF - Old field CE - Cerrado sensu-stricto GF - Gallery forest TT- Total Volume 3, 1994 139 Table 2. Wasp nesting habitats of Xavantina and Serra do Cachimbo: Mato Grosso, Brazil (Data from Richards, 1978) Species CS DC CE CD GF DF TT Apoica gellida 0 0 0 2 0 1 3 Apoica pollens 0 0 0 0 1 2 3 Brachygastra augustii 0 3 1 3 0 0 7 Brachygastra moebiana 0 0 1 1 0 0 2 Brachygastra scutellaris 0 0 0 0 0 2 2 Chartegellus communis 0 0 1 1 0 0 2 Cluirtergus chartarius 0 0 1 1 0 0 2 Chartergus metanotalis 0 0 1 0 0 0 1 Clypearia humeralis 0 1 0 0 0 0 1 Epipona tatna 0 1 2 0 0 0 3 Mischocyttanis flavicornis 0 2 0 0 0 0 2 Mischocyttarus latior 0 0 1 0 0 0 1 Mischocyttanis matogrossoensis 1 0 0 0 0 0 1 Mischocyttanis melanoxanthus 0 1 0 0 0 0 1 Mischocyttanis metluithoracicus 0 0 0 0 0 1 1 Mischocyttanis omicron 0 1 0 0 0 0 1 Mischocyttarus surinamensis 0 0 0 0 1 0 1 Mischocyttarus undulatus 0 0 0 0 2 0 2 Parachartergus fraternus 0 4 4 0 0 0 8 Polistes billardieri 1 0 0 0 0 0 1 Polistes cinerascens 0 0 1 0 0 0 1 Polistes goeldii 1 0 0 0 0 0 1 Polybia chrysothorax 0 0 2 0 0 0 2 Polybia dimidiata 0 0 0 2 2 2 6 Polybia emaciata 0 0 2 0 0 2 4 Polybia en/throthorax 0 1 2 6 0 0 9 Polybia gorytoides 0 1 0 0 3 0 4 Polybia ignobilis 2 0 7 0 0 5 14 Polybia jurinei 0 0 2 5 6 2 15 Polybia liliacea 0 0 0 1 0 1 2 Polybia micans 0 0 0 0 0 1 1 Polybia occidentalis 11 6 1 6 0 0 24 Polybia platycepluila 0 1 1 1 1 2 6 Polybia quadricincta 0 0 0 0 0 2 2 Polybia rejecta 0 0 1 1 2 6 10 Polybia ruficeps 0 6 7 0 7 0 20 Polybia scrobalis 0 0 0 0 2 0 2 Polybia sericea 0 3 1 2 0 0 6 Polybia singularis 0 1 0 3 0 1 5 Polybia striata 0 0 0 0 0 2 2 Protopolybia acutiscutis 0 0 0 0 0 1 1 Protopolybia exigua 1 0 0 0 0 0 1 Protopolybia sedula 0 0 1 0 0 1 2 Pseudopolybia compressa 0 0 0 1 0 0 1 Pseudopolybia vespiceps 0 0 0 1 0 0 1 140 Journal of Hymenoptera Research Species CS DC CE CD GF DF TT Stelopolybia angulata 0 0 0 1 2 1 4 Stelopolybia fulvofasciata 0 0 0 0 0 1 1 Stelopolybia lobipleura 1 0 0 0 0 0 1 Stelopolybia testacea 0 0 0 0 2 1 3 Synoeca chalybea 0 0 0 2 0 0 2 Synoeca surinama 0 0 0 0 0 1 1 Total of nests 18 32 40 40 31 38 199 Total of species 07 14 20 L8 12 21 51 CS - Campo sujo DC - Dirt roads and clearings CE - Cerrado sensu-stricto CD - Cerradao GF - Gallery forest DF - Dry forest TT- Total Table 3. Wasp nesting habitats in Chapada dos Guimaraes and Xavantina and Serra do Cachimbo: Mato Grosso, Brazil. Species CU VE CS OF DC CE CD GF DF TT Apoica gellida 0 0 0 0 0 0 2 0 1 3 Apoica pallens 0 0 0 0 0 1 0 2 2 5 Brachygastra augustii 0 0 0 0 3 1 3 0 0 7 Brachygastra bilineolata 0 0 0 0 I) 1 0 0 0 1 Brachygastra moebiana 0 0 0 0 0 1 1 0 0 2 Brachygastra scutellaris 0 0 0 0 0 0 0 0 2 2 Cluirtegellus communis 0 1 1 0 I) 2 1 0 0 5 Clwrtergus cliartarius 0 0 4 0 0 4 1 1 0 10 Owrtergus metanotalis 0 0 0 0 0 1 0 0 0 1 Clypearia humeralis 0 0 0 0 1 0 0 0 0 1 Epipona tatua 0 0 0 0 1 7 0 3 0 11 Metapolybia cingulata 0 0 0 0 0 1 0 0 0 1 Mischocyttarus cerberus 0 0 0 0 0 2 0 0 0 2 Mischocyttarus dreioseni 2 0 0 0 o 0 0 0 0 2 Mischocyttarus flavicornis 0 0 0 0 2 0 0 0 0 2 Mischocyttarus labiatus 0 0 0 0 0 1 0 0 0 1 Mischocyttarus latior 0 0 0 0 0 1 0 0 0 1 Mischocyttarus matogrossoensis 1 0 1 0 0 0 0 0 0 2 Mischocyttarus melanoxanthus 0 0 0 0 1 0 0 0 0 1 Mischocyttarus metliathoracicus 0 0 0 0 0 1 0 0 1 2 Mischocyttarus omicron 0 0 0 0 1 0 0 0 0 1 Mischocyttarus surinamensis 0 0 0 0 0 0 0 1 0 1 Mischocyttarus undulatus 0 0 0 0 0 0 0 2 0 2 Parachartergus frateruus 0 1 0 0 4 8 0 0 0 13 Polistes billardieri 0 0 1 0 0 0 0 0 0 1 Polistes canadensis 1 0 0 0 0 0 0 0 0 1 Volume 3. 1 994 141 Species CU VE CS OF DC CE CD GF DF TT Polistes cinerascens 0 0 0 0 0 1 0 0 0 1 Polistes goeldii 0 0 1 0 0 0 0 0 0 1 Polistes pacificus 0 0 0 0 0 0 0 1 0 1 Polistes subsericeus 4 0 0 0 0 0 0 0 0 4 Poly bia chrysothorax 0 0 0 0 0 2 0 0 0 2 Polybia dimidiata 0 0 0 0 0 0 2 2 2 6 Polybia emaciata 0 0 0 0 0 2 0 1 2 5 Polybia erythrothorax 0 0 2 0 1 6 6 0 0 15 Polybia gorytoides 0 0 0 0 1 0 0 3 0 4 Polybia ignobilis 0 0 2 0 0 7 0 0 5 14 Polybia jurinei 0 0 5 0 0 4 5 6 2 17 Polybia liliacea 0 0 0 0 0 0 1 0 1 2 Polybia micans 0 0 0 0 0 0 0 0 1 1 Polybia occidentalis 0 0 11 1 6 3 6 0 0 27 Polybia paulista 0 0 0 0 0 0 0 0 2 2 Polybia platycephala 0 0 0 0 1 1 1 1 2 6 Polybia qimdricincta 0 0 0 2 0 0 0 0 2 4 Polybia rejecta 0 0 0 0 0 1 1 2 6 10 Polybia ruficeps 0 0 0 5 6 23 0 9 0 43 Polybia scrobalis 0 0 0 0 0 0 0 2 0 2 Polybia sericea 0 0 0 1 3 3 2 0 0 9 Polybia singnlaris 0 0 0 0 1 1 3 2 1 8 Polybia striata 0 0 0 0 0 0 0 0 2 2 Protopolybia acutiscutis 0 0 0 0 0 0 0 0 1 1 Protopolybia exigiia 0 0 1 1 0 1 0 0 0 3 Protopolybia sedula 0 0 0 0 0 1 0 0 1 2 Pseudoclmrtergus clwrtergoides 0 0 0 0 0 0 0 1 0 1 Pseudopolybia compressa 0 0 0 0 0 0 1 0 1 2 Pseudopolybia z>espiceps 0 0 0 0 0 1 1 0 0 2 Stelopolybia angulata 0 0 0 0 0 0 1 2 1 4 Stelopolybia fitlvofasciata 0 0 0 0 0 0 0 0 1 1 Stelopolybia lobipleura 0 0 1 0 0 1 0 0 0 2 Stelopolybia testacea 0 0 0 0 0 0 0 2 1 3 Synoeca chalybea 0 0 0 0 0 0 2 0 0 2 Synoeca surinama 0 0 0 2 0 0 0 6 1 9 Total of nests 08 02 24 12 32 88 39 48 41 299 Total of species 04 02 10 06 14 29 17 19 23 61 CU — Campo umido VE— Vereda CS — Campo sujo OF— Old field DC — Dirt roads and clearings CE — Cerrado sensu-stricto CD — Cerradao GF — Gallery forest DF — Dry forest TT— Total 142 Journal of Hymenoptera Research Table 4. Wasp colonies density in three different habitats in Rio Manso - Chapada dos Guimaraes: Mato Grosso, Brazil. Campo umido Cerrado sensu stricto Gallery forest Area (m2) 500 Number of nests 3 Nest density/ha 60 Density of adult wasps/ha 480 Number of species 2 10000 5 5 1300 5 7500 6 8 50000 5 proximately 50% of wasps were restricted to a single habitat for nesting (Table 3). Only a few ubiquitous species like Polybia occidentalis (5), Polybia platycephala (5), Polybia (Pedothoeca) si)igularis Ducke (5), Chartergellus communis (4), Chartergus chartarius (4), Polybia erythrothorax (4), Polybia rejecta (4), Polybia ruficeps (4) and Polybia sericea (4) use more than three habitats for nesting (Table 3). Substrates for nesting in the cerrado is almost unlimited (Henriques el al. 1992), but availability of food, water and protection could be limiting factors. 2. Density In an area of 2500 m2 of the gallery forest (Fig. 5) three colonies of two species of wasps were found, two of Polybia (Myrapetra) paulista H.von Ihering and one of Polybia ruficeps. In a contiguous area of approximately 5000 m2 three colonies of three species (Polybia (Pedothoeca) emaciata Lucas, Polybia (Myrapetra) sp. and Polybia (Pedothoeca) singularis (Ducke) were found, which correspond to an average of eight colonies and five species per ha (Table 4). In an area of 10000 m2 of cerrado sensu stricto (Fig. 4), near a stream, three colonies of the following species were found: Mischocyttarus ( Kappa) metathoracicus(deSaussure),Mischoa/ttarus (Mischocyttarus) labiatus (F) and Polybia ruficeps. Colonies of Brachygastra bilineolata Spinola and Stelopolybia lobipleura Richards were also found on trees close to each other, but away from the stream with an average of colonies of five species per ha (Table 4). A similar result was found in the cerrado sensu-stricto of Brasilia DF ( Ecological Reserve of IBGE), where six colonies were found in an area of 1.5 ha corresponding to an average of five colo- nies per ha (Henriques et al, 1992). In an area of 500 m2 of campo umido (Fig. 1), two colonies of Polistes subsericeus and one of Mischocyttarus drewseni were found. Based on these numbers estimates were 60 colonies per ha for both species, or 40 per ha for Polistes subsericeus and 20 per ha for Mischocyttarus drewseni (Table 4). Although the colony density in campo umido was larger than that found in gallery forest and cerrado sensu-stricto, the number of individuals per nest was smaller, due to low numbers of adults per nest (eight per nest). In gallery forest and cerrado sensu stricto, the colonies of Synoeca surinama, Epipona tatua and Polybia spp, were larger, with hundreds to thousands of individuals per nest (Table 4). In a relatively simply-structured habitat, such as campo umido (Fig. 1) with a small area (500 m2), the density estimate was more accurate than in more complex habitats. A ranking of habitats in order of reliability of estimate is: campo umido, cerrado sensu-stricto and gallery forest. Volume 3. 1 994 143 ACKNOWLEDGMENTS We acknowledge the contribution made by the late Professor O.W. Richards to studies of social wasps in general. It is not unusual for subsequent workers to discover interesting in- formation in the data he published. Financial support was provided by the North Brazilian Electricity Company (Eletronorte) and the University of Brasilia. We are espe- cially grateful to Dr. Anthony Raw and Dr. Linda Caldas of the Universidade de Brasilia and the anonymous reviewer for useful comments on the manuscript. Dr. Raw also iden- tified several of the species of wasps. Dr. Paul M. Marsh made some editorial changes on the manuscript. Sra. Beatnz Mossri kindly assisted in the collection of field data. We also thank Dr. Antonio Jose E. Brussi, for his help and encourage- ment LITERATURE CITED Eiten, G. 1972. The Cerrado vegetation of Brazil. The Botanical Review 38:201-341. Forsyth, A. 1980. Nest site and Habitat selection by the social wasp, Metapolybia azteca Araujo (Hymenoptera: Vespidae) Brenesia 17: 157-162. Gobbi, N. & Machado, V. L. L. 1985. Material capturado e utihzado na alimentacao de Polybia (Myrapetra) paulista Ihering 1896 (Hymenoptera-VespidaeJ./lnais da Sociedade Entomologica do Brasil 14(2): 189-195. Gobbi, N. Machado, V. L. L. & Tavares Filho, J. A. 1984. Sazonalidade das presas utilizadas na alimentacao de Polybia occidentalis (Olivier 1791) (Hymenoptera: Vespidae). Anais da Sociedade Entomologica do Brasil 13(1): 63-69. Henriques, R. P. B. Rocha, I. R. D. & Kitayama, K. 1992. Nest density of some social wasps species in Cerrado vegeta- tion of Central Brazil (Hymenoptera: Vespidae) Entomologia Generalis 17(4): 265-268. Jeanne, R. L. 1975. The adaptiveness of social wasp architec- ture. Quarterly Review of Biology 50: 267-287. Kitayama, K. Rocha, I. R. D. & Bulhoes B. M. 1989. Densidade de ninhos de vespideos em diversos habitats de Mato Grosso. I . Shnpdsio Latino-americano sobre Inseto Socials Neotropicais. Rio Claro SP. R-67 P 44. Kojima, J. 1. 1982. Nest architecture of three Ropalidia species (Hymenoptera: Vespidade) on Leyte Island, The Philip- pines. Biotropica 14(4): 272-280. Machado, V. L. L. Gobbi, N. & Simoes, D. 1987. Material capturado naalimentai;aodeSrp/opo/i/l)i(!fi«//if)fs (Olivier 1791) (Hymenoptera: Vespidae) Anais da Soctedade Entomologica do Brasil 16 (1): 73-79. Partridge, L. 1978. Habitat selection In Behavioural Ecology (Eds. JR. Krebs & N.B. Davies) Blackwell Scientific Publications, Oxford, 351-376. Rau, P. 1942. Habitat preferences of Polistes wasps (Hy- menoptera: Vespidae) Entomological Nezus 53: 293-295. Raw, A. 1988. Social wasps (Hymenoptera: Vespidae) and insect pests of crops of the Surui and Cinta Larga Indians in Rondonia, Brazil. The Entomologist 107: 104- 109. Reed, H. C. & Vinson, S. B. 1979. Nesting ecology of paper wasps (Polistes) in a Texas urban area (Hymenoptera: Vespidae). Journal of the Kansas Entomological Society 52 (4): 673-689. Richards, O. W. 1978. The social wasps of the Americas. British Museum of Natural History. London 580 pp. Starr, C. K. 1988. The nesting association of the social wasps Mischocyttarus immarginatus and Polybia spp. in Costa Rica. Biotropica 20(2): 171-173. Windsor, D. M. 1972. Nesting association between twoNeotropical polybiine wasps (Hymenoptera: Vespidade)Biofropic«. 4: 1-3. J. HYM. RES. Vol. 3, 1994, pp. 145-150 Descriptions of two New Species and Notes on the Genus Bakeriella Kieffer from Brazil and Ecuador (Hymenoptera: Bethylidae) Celso Oliveira Azevedo Departamento de Biologia, Universidade Federal do Espinto Santo, Av. Marechal Campos 1468, Maruipe, 29040-090 Vitoria, ES Brasil Abstract. — Bakeriella mcompleta sp.n., from rubber-tree crops of Sao Paulo, Brazil and Bgrossensissp. n. fromMatoGrosso, Brazil are described and illustrated. Taxonomic data on B. reclusa (Evans) and B. montivaga (Kieffer) are included, and the latter recorded from Brazil for the first time. A key to females with a transverse carina on the pronotal disc is presented. INTRODUCTION Bakeriella Kieffer is a neotropical genus, rang- ing from Southern United States to Northern Ar- gentina (Azevedo 1991). Kieffer (1910a) described this genus based on one species from Para, Brazil, B. flavicornis, which has the pronotum carinate laterally and medially as well as anteriorly. This genus is related to Epyris Westwood by having the scutellar groove divided (Evans 1964). Kieffer (1910b) described B. depressa from Peru which has a carinate pronotum also. Evans (1964) expanded the definition of Bakeriella when he described ten species, in some of which the median or lateral carinae of the pronotum are weak or absent. Evans (1969), in his revision of Epyris, pointed out the relationship of the monthwgiis-group of Epyris to Bakeriella by the fact that both groups have similar scutellar pits. Evans (1979) reviewed Bakeriella, providing identification keys for both males and females and arranging the genus into four species- groups. In this revision, Evans described one spe- cies, B. erythrogaster, which has the scutellar groove similar to the genus Rliabdepyris Kieffer. Evans also transfered members of the montivagus-group from Epyris to Bakeriella simply by the fact that they closely resemble each other. Curiously, he did not include B. subcarinata Evans 1965 in his revision. Gordh & Moczar (1990) cited in their catalog 19 species for the world and Azevedo (1991) described B. dentata from Sao Paulo State, Brazil. The present study focuses on four species of this genus, B. montivaga (Kieffer), B. reclusa (Evans) and two new species. Ideally what is needed is a complete revsion of the genus. Unfortunately, specimens are rare in collections and until more material is available for study, I have chosen to present these notes and new species descriptions. Specimens for this study belong to the follow- ing collections (abbreviations according to Arnett & Samuelson (1986)): IBSP, Instituto Biologico de Sao Paulo, SP, Brasil; MZSP, Museu de Zoologia da Universidade de Sao Paulo, SP, Brasil; UFES, Colecao de Entomologia da Universidade Federal do Espirito Santo, Vitoria, ES, Brasil. The last col- lection is not included in Arnett & Samuelson (1986) because it is new. Abbreviations used follow Evans (1964) and are as follows: LFW, length of forewing; WH, maxim un width of head, including eyes; LH, length of head, measured from median apical margin of clypeus to median point of vertex; WF, minimun width of frons; HE, maximun height (or length) of eye; WOT, maximun width of ocellar triangle, including posterior ocelli; OOL, shortest ocello- ocullar line. The nomenclature of the integument follows Eady (1968). 146 Journal of Hymenoptera Research Bakeriella incompleta, sp. n. Figs. 1-2 Holotype female.— Brazil: SP,Ibitinga;05.X.1988; yellow pan trap, rubber-tree crop, no collector given. (UFES). Description of holotype female. — Length of body 6.0 mm; LFW 2.85 mm; head and thorax black, abdomen black except tip dark brownish; palpi light brownish; mandible brownish, black at base and with a light brownish spot outside, teeth reddish; antennal scape dark brownish, apex brownish, pedicel and flagellum somewhat infuscated; tegula brownish; fore and hindcoxae black, midcoxa darkbrownish; trochanters, femora, tibiae and tarsi dark brownish, the anterior being lighter; wings subhyaline somewhat brownish, veins, prostigma and stigma dark brownish. Head (Fig. 1). Mandible with 5 teeth, apical tooth largest (Fig. 1). Clypeus with subangulate median lobe, with median carina straight in lat- eral profile, clypeus forming a small excavated reentrance in the region of contact with frons. WH 1.03 X LH; WF 0.5 XWH;WF 1.34 X HE; WOT 1.33 X OOL. Antennal scrobe not carinate. Posterior ocelli about 3 X their diameter from the vertex. Vertex nearly straight and with lateral angles rounded off. Distance from the top of eyes to crest of vertex about 0.7 X HE. Length of first four antennal segments in a ratio of about 11:5:4:4. Segments 5 to 12 about as wide as long. Temple not carinate. Frons somewhat alutaceous and bril- liant. Mesosoma(Fig. 1). Somewhat more alutaceous than frons, mesoscutum polished anteriorly. Pronotal disc slightly convex anteriorly, tranverse anterior carina without teeth, median carina weak, situated in a slight depression and absent anteri- orly, not intersecting transverse carina. Posterior margin of pronotal disc paralelled by a series of small foveae; lateral margin of disc somewhat sharpened, not carinate, and slightly diverging posteriorly; median width of the disc slightly greater than its length; disc slightly shorter than mesoscutum and scutellum together. Notauli not reaching the anterior margin of mesoscutum, con- verging posteriorly, widened posteriorly, and slightly curved anteriorly. Parapsidal furrows straight, somewhat more enlarged posteriorly, occupying only the posterior half of mesocurum (fig. 1). Scutellar pits tranverse, much wider than long, separated by a thin septum which is lower than margin of foveae. Propodeal disc as wide as long, with three discal carinae, median one com- plete and the others incomplete, occupying about two thirds of disc and converging posteriorly but not reaching median carina; space among discal carinae with tranverse striae; disc without sublateral carinae; inner side of lateral carinae paralelled by a series of small irregular depres- sions; anterior part of space between lateral and discal carinae occupied by a rather large depres- sion; posterior angle of the disc foveolate; decliv- ity of propodeum with median carina and perpen- dicular striae. Metapleuron with longitudinally sinuous striae. Mesopleuron granulate, with large fovea anteriorly and small, deep pit at anterior margin, with small pentagonal fovea and elon- gate upper fovea (fig. 2). Mesosternum with small median pit, anterior margin of acetabular carina paralelled by striae. Midtibia spinose. Forefemur 2.0 X as long as thick, midfemur 1.65 X as long as thick. Claw trifid, first tooth rounded off, the oth- ers sharp. Metasoma. Sterna setose on posterior of each sternum. Terga with few setae on the posterior half of each tergum. Terga I and II with lateral setae. Tranverse section of metasoma circular. Length of mesosoma 1.35 X length of metasoma. Valvae III dark brownish with only its end setose, setae range between one third and a half length of the valvae. Paratype female. — BRAZIL: SP, Balsamo; 21. IX. 1988, white pan trap, no collector given. (IBSP). The paratype differs from holotype as fol- lows: WH 1.00 XLH;WF0.57XWH; WF 0.76 X HE; OOL 1.62 X WOT; propodeal disc 1.08 as large as long. Remarks. — This species is known only from the female. Bakeriella incompleta differs from other species in the genus in having the median longitu- dinal pronotal carina incomplete anteriorly, not reaching the anterior transerve carina, but reach- ing the posterior margin. In B. cristata Evans this carina is incomplete posteriorly. Etymology. — The specific name refers to the fact that the median longitudinal carina of the pronotal disc is incomplete. Volume 3. 1 994 147 Figs 1-2. Bakeriella mcompleta, sp n. 1, head and mesosoma, dorsal view; 2, mesopleurum, lateral view. Scale = 0.5 mm. Bakeriella grossensis, sp. n. Holotype female. — Brazil: Barra do Tapirape, MT; XII.1960; unknown trap; B. Malkin, col. (MZSP). Description of holotype female. — Length of body 3.75 mm; LFW 2.4 mm; head and mesosoma black, metasoma black with apex somewhat brownish; palpi brownish; mandible yellowish with teeth darker; antenna yellowish with scape darker; tegula yellowish; coxae and femora brown, tro- chanters, tibiae and tarsi yellowish; wings subhyaline, veins brown. Head. Mandibles with 5 teeth, the basal teeth small and rounded off, the 2 apical teeth sharp- ened, anterior tooth largest, the second of median size. Clypeus with median lobe angulate, without median carina, clypeus forming a small excavated reentrance in the region of contact with frons. Eyes with sparse long hairs. WH 1.08 X LH; WF 0.58 X WH; WF 1.16 X HE; WOT 1.62 X OOL. Antennal scrobe not carinate. Distance posterior ocelli from vertex about 2.8 X diameter of ocelli, angle of ocellar triangule less than a right angle. Vertex rounded, very slightly angled laterally. Distance from the top of eyes to crest of vertex about 0.57 X HE. Length of first four antennal segments in a ratio about 20:8:7:7. Temple not carinate. Frons coriaceous with punctures which are separated by 1-3 X their own diameters. 148 Journal of Hymenoptera Research Mesosoma. Pronotal disc coriaceous, punc- tures similar to frons, with straight anterior trans- verse carina, without median longitudinal carina, posterior margin paralelled by series of small foveae; sides of disc subparalell; median width of disc 1.17 X its length. Pronotal disc 0.77 longer than mesoscutum and scutellum together. Ante- rior half of mesoscutum alutaceous, posterior half coriaceous. Notauli straight and very thin, vesti- gial anteriorly, widened posteriorly, reaching both anterior and posterior margin of mesoscutum. Parapsidal furrows straight, occupying only pos- terior half of mesoscutum and reaching its poste- rior margin. Scutellar pits transverse, 2.0 X as wide as long, separated by thin septum. Propodeal disc 1.05 X as wide as long, with three discal carinae, median one complete and lateral ones converging behind, not reaching median carina; space be- tween discal carinae with small transverse striae, space between discal carinae and lateral carina polished; disc without sublateral carinae; lateral carinae boardered by a depression with small transverse striae; posterior angle of disc foveolate; declivity of propodeum with median carina and perpendicular striae. Metapleuron alutaceous, without striae. Mesopleuron coriaceous, with large ventral fovea occupying the lower half of mesopleuron, with deep pit in upper margin of lower fovea, with a fovea in the anterior part of mesopleuron and with long fovea in upper part of mesopleuron opened anteriorly. Mesosternum coriaceous, with pit in the middle of posterior region. Midtibia spinose. Forefemur 2.5 X as long as thick, midfemur 1.7 X. Claws trifid, first tooth rounded off, the others sharpened. Metasoma. Wholly polished, terga and sterna with few setae. Setae of terga on posterior and lateral margins. Transverse section of metasoma subcircular. Length of mesosoma slightly longer than the metasoma. Valvae III brown with setae on apex. Remarks. — This species is known only from the holotype. It runs to B. inconspicua and B. polita in the key written by Evans (1979). B. grossensis differs from B. inconspicua in having the scutellar pits tran verse, the claws trifid, the eyes not strongly setose, the metasoma black with reddish apex and the mandible 5-toothed. B. grossensis differs from B. polita in lacking a median carina in the clypeus, in lacking a lateral carina on the pronotal disc, the head being not well developed above the eyes and the occipital carina being obscured in full frontal view of head. In addition, the mandible of B. grossensis has two big apical teeth, while B. polita has one and B. inconspicua has three. Etymology. — The specific name is in reference to the Mato Grosso State, region of origin of the holotype. Bakeriella montivaga (Kieffer) 1910b Epyris montivagus Kieffer 1910b:31; Evans 1969:324 Bakeriella montivaga: Evans 1979:261 This species is recorded from Peru, Colombia (Kieffer 1910b), Bolivia, Ecuador, Venezuela, Panama, Costa Rica, Honduras, Mexico (Evans 1969), and now from Brazil. Four females have been studied: Balsamo, SP, Brazil, one female, 28. IV. 1988, yellow pan trap, rubber-tree crop, no collector given (UFES). The others were from Pichincha, Centro Cientifico R. Palangue, Ecuador, 21.XII.1980, 1.1.1981, 4.1.1981, forest, S. Sandoval, col. (MZSP). These specimens are similar to the type; pronotal disc with the sides diverging posteriorly; WH 0.99 to 1.17 X LH; WF 0.53 to 0.59 X WH; WF 1.18 to 1.46 X HE; OOL 1.63 to 1.94 X WOT; propodeal disc slightly wider than long. The line of small foveae near the posterior margin of the pronotal disc is weakly angled anteriorly in the middle. This character is common only in males. Evans (1969) considered the possibility that B. quinquepartita (Kieffer) and B. montivaga represent variations of a single species. Indeed, both species are very similar and the only differences between them were pointed out in Evans'(1979) key to females. Bakeriella reclusa (Evans) 1969 Epyris reclusus Evans 1969:329 Bakeriella reclusa: Evans 1979:261 This species was described from Costa Rica and El Salvador. Evans (1979) studied a female from Ecuador. I have examined a male from Ecua- dor: Morona Santiago, Los Tayos, 3. VIII. 1976, DeVries, col. (MZSP). Measurements for this spe- cies are: LFW = 2.13 mm, WH 1.2 X LH; WF 0.55 X WH; WF 1 .06 X HE; WOT 1 .44 X WOT. The pronotal disc is slightly wider than long, and the propodeal disc is as long as wide. Volume 3. 1994 149 The following key to females with an anterior transverse carina on the pronotal disc is modified from Evans (1979:258) and includes B. dentata Azevedo 1991, B. subcarinata Evans 1965, B. incompleta, sp. n., and B. grossensis, sp. n. 1 Pronotum without median longitudinal carina; head black 2 1' Pronotum usually with complete median carina; head black or dark green 7 2(1) Legs wholly yellowish; propodeal disc covered with weak transverse striations ... floridana Evans 1964 2' Coxae and femora largely black; metapleuron shining, at most with very weak sculpture 3 3(2') Transerve carina of pronotum weakly angled forward medially; median area of propodeum rather weakly sculptured 4 3' Transerve carina of pronotum evenly arched; median area of propodeum with strong transverse rigdes 5 4(3) Mandible with 5 teeth, basal 3 small; frons with small punctures which are separated by 1.5-3.0 X their own diameters olmeca Evans 1964 4' Mandible with only two large apical teeth; frons with somewhat stronger and more widely spaced punctures brasiliana Evans 1964 5(3) Sides of pronotal disc carinate anteriorly; scutellar pits ovoid; mandible with 3 strong apical teeth inconspicua Evans 1964 5' Sides of pronotal disc not carinate; scutellar pits transverse; mandible with 1 or 2 strong apical teeth 6 6(5') Sides of pronotal disc sharpened; mandible with one large apical tooth; scutellar pits slightly wider than long polita Evans 1964 6' Sides of pronotal disc angled, but not truly sharpened; mandible with two large apical teeth; scutellar pits 2.0 X wider than long grossensis, sp. n. 7(1') Median carina of pronotal disc incomplete 8 7' Median carina of pronotal disc complete 9 8(7) Median longitudinal carina of pronotum lacking posteriorly; body black with green metallic reflexions cristata Evans 1964 8' Median longitudinal carina of pronotum lacking anteriorly; body black without metallic reflexions incompleta, sp. n. 9(7') Internal margin of tranverse carina of pronotum with tooth-like emargination .... dentata Azevedo 1991 9' Internal margin not like above 10 10(9) Median longitudinal carina of pronotum somewhat weak and not paralleled by a depression subcarinata Evans 1965 10' Median longitudinal carina of pronotum well evident and situated in a depression inca Evans 1964 ACKNOWLEDGMENTS I express thanks to Ehana Bergmann for the loan of material from IBSP and Dr. Carlos Brandao for the loan of material from MZSP. I also thank Oscar Shibatta for prepar- ing the illustrations. 150 Journal of Hymenoptera Research LITERATURE CITED Arnett, R. H., Jr & G. A. Samuelson (eds). 1986. The insect and spider collections of the world. Gainesville, E.J. Brill /Fauna & Flora Publications, 220p. Azevedo, C. O. 1991. Especie nova e notas sobre Bakeriella Kieffer (Hymenoptera, Bethylidae) da regiao de Sao Carlos, SP. Revista Brasileira de Entomologia 35(3): 535-8 Eady, R. D. 1968. Some illustrations of microculpture in the Hymenoptera. Procedings of the Royal Entomological Soci- ety of London, series A General Entomology 43(3-6): 66-72. Evans, H. E. 1964. A Synopsis of the American Bethvlidae (Hymenoptera : Aculeata). Bulletin of the Museum of Com- parative Zoology, Harvard University 132(1): 1-222. Evans, H. E. 1965. Further studies on Neotropical Epynni (Hymenoptera: Bethylidae). Psychef Cambngde) 72(4): 265-78. Evans, H. E. 1969. A revision of the genus Epyris in the Americas (Hymenoptera: Bethylidae). Transactions of the American Entomological Society 95(2): 181-352. Evans, H. E. 1979. A reconsideration of the genus Bakeriella (Hymenoptera: Bethylidae). Journal of the Nexv York En- tomological Society 87(3): 256-66. Gordh, G. and L. Moczar. 1990 A catalog of the world Bethylidae (Hymenoptera). Memoirs of the American En- tomological Institute 46: 1-364. Kieffer, J. J 1910a. Description de Nouveaux Micro- hymenopteres du Bresil. Annates de la Societe de Entomologique de France 78: 287-348. Kieffer, J. J. 1910b. Description de Nouveaux Microhy- menopteres du Bresil. Annates de la Societe de Entomologique de France 79: 31-56. J. HYM. RES. Vol. 3, 1994, pp. 151-156 Localization of the Female Sex Pheromone Gland in Cotesia rubecula Marshall (Hymenoptera: Braconidae) Scott A. Field And Michael A. Keller Department of Crop Protection, University Of Adelaide, Waite Campus, Glen Osmond, South Australia 5064, Australia Abstract - A combination of behavioural, histological and electron microscopic techniques was used to localize the female sex pheromone gland in the parasitic wasp Cotesia rubecula (Marshall) (Hymenoptera: Braconidae). The genital capsule was identified as the source of the sex pheromone by presenting males with a choice of body parts in a flight tunnel. Histological examination and electron microscopy further reduced the possibilities for pheromone production to two structures. One of these was a gland in a position similar to that reported for other braconid wasps, and which was associated with surface pores but lacked any obvious secretory products. The other was an active secretory gland which is associated with the moving parts of the ovipositor. Further studies aimed at isolating and identifying the chemicals involved are required to resolve this issue. The presence of sex pheromone glands has been demonstrated in a number of species of para- sitic wasps (Weseloh 1976, 1980; Tagawa 1977, 1983). Although unable to localize the sex phero- mone source more precisely than the abdominal region, Vinson (1978) suggested that Dufour's gland produced a sex identification pheromone in Cardiochiles nigriceps. Obara and Kitano (1974) first demonstrated that the source of the female sex pheromone in Cotesia glomerata is the tip of the abdomen, near the ovipositor. Tagawa (1977) later described a pair of secretory glands located at the base of the second valvifer on the ninth abdominal tergite of this species, which he suggested was responsible for sex pheromone production. A sub- sequent study (Tagawa 1983) revealed the exist- ence of similar glands in a corresponding position in all seven braconid species examined. Upon revising his initial findings (Weseloh 1976), Weseloh (1980) likewise concluded that paired epidermal glands on the last abdominal tergite were the sex pheromone source in Cotesia melanoscelus. Previous field observations (Keller unpub- lished) established that courtship in male Cotesia rubecula (Marshall) (Hymenoptera: Braconidae) is elicited by a female sex pheromone. The present study combined behavioural tests with micro- scopic techniques in order to identify the position of the sex pheromone gland in C. rubecula. MATERIALS AND METHODS Behavioural Experiments. — C. rubecula was reared in the laboratory on larvae of Pieris rapae L. using the methods of Keller (1990). Females be- tween one and four days of age were frozen at -15° C for approximately 30 minutes before being dis- sected. Using the same experimental protocol, we performed two experiments to test the respon- siveness of males to different female body parts. In the first, the head and mesosoma (thorax plus first abdominal segment) were tested against the metasoma (remaining abdominal segments), and in the second the genital capsule (terminal ab- dominal segment plus ovipositor) was tested against the remaining anterior portion of the metasoma. A glass microscope slide was divided into two halves and the two body parts being tested were placed randomly in the middle of either half, presenting a choice to the males. The slide was clamped at its midpoint and held horizontally at a 152 Journal of Hymenoptera Research height of 30 cm by a small metal stand. Tests were conducted in a flight tunnel (Keller 1990) at a wind speed of 30 cm/s. Males were kept in a separate cage and removed for testing individually with a minimum of disturbance. They were released 30 cm downwind of the slide and observed to deter- mine (1) which half of the slide they landed on, (2) whether subsequently they oriented to either body segment and performed typical elements of court- ship (Field and Keller 1993), and (3) whether they attempted to copulate with either body segment. The same males were then tested again, but this time the female body segments were washed in diethyl ether prior to testing. This solvent re- moved all contaminating pheromone from the surface of the body segment (Golub and Weatherston 1984). Therefore any subsequent emission of pheromone could be due only to con- tinued leaking of the pheromone onto the surface from an internal glandular reservoir. The prefer- ence of courting males for a particular segment after washing would therefore localize the phero- mone gland to somewhere within that segment. Most males were successful in navigating to the slide; those that did not were discarded after a few attempts. To lessen the effects of pheromone contamination and depletion of pheromone re- serves, the stand was washed with diethyl ether and body segments and slides were replaced at regular intervals throughout the course of the experiments. Variation in attractiveness of females was assumed to be negligible. Fifty males were tested in each experiment and data were analysed using two-tailed binomial tests (Zar 1984). Scanning Electron Microscopy (SEM). — Two methods were used to prepare specimens for SEM observations. In the first, newly-dissected speci- mens were immersed in Peterson's KAA (kero- sene (10%), 95% alcohol (75%), glacial acetic acid (15%)) prior to washing in 0.1 M phosphate buffer and dehydration through an alcohol series. This technique is commonly used to preserve larval insects because it distends structures that may otherwise lose shape (Smithers 1981). The second method was to "clear" the speci- mens by gently warming in 10% KOH for 30-60 minutes. This removed most tissue and allowed dissection of the specimens so that both internal and external cuticular structures could be exam- ined under SEM. Following dehydration, the KAA- treated and cleared specimens were critical-point dried in an EMSCOPE CPD 750 and mounted on TAAB aluminium stubs using Acheson Electrodag 915. They were coated by vacuum evaporation with carbon and gold palladium and examined in an ETEC Autoscan scanning electron microscope at an accelerating voltage of 20 kV. Light Microscopy (LM). — Genital capsules were removed and immersed for 4 hours in fixative (3% glutaraldehyde/3% formaldehyde made up in 0.1 M phosphate buffer, pH 7.4, to which had been added 2.5% polyvinyl pyrrolidone). They were washed in 0.1 M phosphate buffer overnight and then dehydrated by passing through a series of alcohols. After washing in propylene oxide, they were infiltrated with increasing concentrations of TAAB epoxy embedding resin over 48 hours and then embedded in resin by curing at 60° C for a further 48 hours. Using glass knives in a Sorvall MT2-B "Porter- Blum" ultramicrotome, serial transverse sections of 0.5 urn thickness were cut, starting from near the base of the ovipositor and proceeding anteriorly. These were stained using 0.025% Toluidine Blue in 0.5% Borate buffer and photographed with a Wild MPS 45 photoautomat on Ilford Pan-f 50 ASA film. Agfa x 3 light green and Wratten 58 filters were used to achieve suitable contrast. Males were also subject to the same processing proce- dures and serial sections were cut from the poste- rior tip of the abdomen. Transmission Electron Microscopy (TEM). -Tis- sue samples were prepared in the same mariner as for LM. Gold to silver sections (0.08 mm) were cut with a Diatome diamond knife using a Reichert- Jung Ultracut and stained with saturated uranyl acetate in 70% alcohol for 20 minutes, followed by lead citrate for 12 minutes (Reynolds, 1963). They were examined in a JEOL 100S transmission elec- tron microscope at an accelerating voltage of 80 kV. RESULTS Behavioural Experiments. — In the first experi- ment, neither the head/mesosoma nor the metasoma was found to be more attractive when the segments had not been washed in ether. How- ever, when surface contamination had been re- moved with ether, males responded preferentially toward the metasoma, indicating that it was the source of the sex pheromone (Table 1). Volume 3. 1994 153 Fig. 1. SEM of lateral view of genital capsule in Cotesia rubecula. Arrow indicates position of hair plate and pores (see Fig. 2); n - ninth tergite; o - ovipositor. Scale bar = 100 urn Fig. 2. High magnification SEM of hair plate and pores. Scale bar = 10 urn Fig. 3. Transverse section of genital capsule, showing under LM two possible sex pheromone-producing glands: gp - glands associated with surface pores; gb - gland at base of genital capsule; hp - hair plate. Scale bar = 30 urn Fig. 4. TEM of basal section of genital capsule, p - pore canals; s - secretory droplets. Scale bar = 5 urn 154 Journal of Hymenoptera Research Table 1 . Response of male Cotesia rubecula to a choice between the metasoma and the head+mesosoma of females, washed in ether or unwashed. * indicates significant differences (P<0.001). Treatment Number Responding Response Metasoma Head+Mesosoma Unwashed Washed Land Court Copulate Land Court Copulate 23 23 10 36 36 19 27 27 22 l" r r Table 2. Response of male Cotesia rubecula to a choice between the genital capsule and the anterior portion of the metasoma of females, washed in ether or unwashed. * indicates significant differences (P<0.001). Treatment Number Responding Response Anterior Metasoma Genital Capsule Unwashed Washed Land 17 Court 18 Copulate Land Court Copulate 19 3 4 3 31 27 1* 28* 21* I In the second experiment, there was no prefer- ence shown by males for either the unwashed genital capsule or the unwashed anterior portion of the metasoma in terms of frequencies of landing and courting (Table 2). Several males moved be- tween the body segments after landing in close proximity to one of them. More males attempted to copulate after courting an unwashed metasoma (18/18, plus one that courted to the genital capsule initially) than an unwashed genital capsule (1/ 27). After these body parts were washed with ether, males showed a clear preference for the genital capsule in terms of landing and courting, but not copulating. Microscopy. — Having identified the genital capsule as the source of the sex pheromone, we examined KAA-treated genital capsules under the SEM to determine whether there were any surface pores that could serve as outlets for pheromone secretion. The only pores found were positioned slightly below and posterior to a mechanosensory hair plate (Figs. 1,2). Dissection of a cleared speci- men revealed the internal attachment site of these mechanosensory hairs to be at the base of the second valvifer, which is an internal cuticular plate comprising part of the ovipositor system (see Snodgrass, 1935). The LM sections revealed only one region of tissue that was a possible candidate for the sex pheromone source. A pair of glands was located in a position corresponding closely to that of the pores, i.e., just below the hairs and appearing immediately before the hairs in serial transverse sections (Fig 3). No pores were apparent, but this Volume 3, 1994 155 may have been a result of the small proportion of sections cut (approx. 1/10) that were actually re- trieved, stained and mounted for examination. The results of ultrastructural examination of this gland were inconclusive and failed to offer any indication of the organelles present or whether it was secretory in nature. TEM examination revealed the presence of a second pair of secretory glands (Fig. 4), located at thebase of the genital capsule and adjacent to folds of membrane which are presumably stretched and compressed during movements of the ovipositor. The ultrastructure of these glands was also indis- tinct, but pore canals leading to the surface and an abundance of secretory bodies were visible on or just under the surface of the cuticle. DISCUSSION The behavioural experiments clearly demon- strate that the genital capsule is the source of the sex pheromone in C. rubeculn. With all surface pheromone removed, males were attracted to the metasoma in preference to the head/mesosoma and, when given a further choice, made their initial approach more often to the genital capsule rather than to the anterior portion of the metasoma. This is indicative of the continuous emission of pheromone from an internal reservoir located in the genital capsule, which provides an olfactory cue to the searching wasp. There are, however, two anomalous results which require explanation. Firstly, a single male chose the washed head/mesosoma over the metasoma in the first experiment. This could have been due to incomplete washing of the body parts. Secondly, although significantly more males landed near and courted the genital capsule as opposed to the anterior portion of the metasoma, they attempted copulation with the genital cap- sule less frequently. This is not considered to be evidence that the sex pheromone gland is in the anterior portion of the metasoma. Rather, it sug- gests that a visual cue is involved in triggering attempts at copulation by the male. Upwind flight, landing, orienting and courting all appear to be stimulated entirely by chemical cues. The pres- ence of a particular threshold amount of chemical may "prime" males so that they are ready to mount and copulate after a certain amount of courtship, but an object of appropriate size, shape and colour in the vicinity of a pheromone source may be required to release copulation behavior. One male did attempt copulation with the tiny genital capsule (0.5 mm in length), so the phero- mone alone may be sufficient to elicit attempts at copulation in some cases. The histology and electron microscopy failed to resolve the exact location of the sex pheromone gland within the genital capsule, but narrowed the possibilities to two structures. One of these is a group of cells located directly underneath surface pores which are adjacent to the mechanosensory hair plate on the genital capsule (Figs. 1-3). The hair plate was in turn juxtaposed to the second valvifer, which places this gland in approximately the same position as the putative pheromone glands indicated by Tagawa (1977, 1983) for a number of other closely-related braconids. Al- though this constitutes circumstantial evidence that this is the sex pheromone gland in C. rubecula, the absence of any clearly defined secretory prod- ucts in the ultrastructural examinations leaves some uncertainty. The pores associated with these glands required thorough washing before they became visible by SEM. Another possible source of sex pheromone is a gland located further toward the base of the genital capsule. The most striking feature of this gland is an abundance of associated secretory bodies amassed on and under the surface of the cuticle which are visible under TEM (Fig. 4). One possibility is that these secretions act as a lubricant for the ovipositor as it is extended and retracted from the metasoma . However, if it is the sex phero- mone gland, then such movements would un- doubtedly smear the secretions over the cuticle, which would be ideal for pheromone dispersal. An additional possibility is that the one secretion has the dual function of lubrication and sex attrac- tant. In the absence of isolation and identification of the pheromone, the question of which of the two glands is the source of sex pheromone in C. rubecula will remain unresolved. 156 Journal of Hymenoptera Research ACKNOWLEDGEMENTS We thank Chris Leigh for the generous use of his facili- ties for histological processing, preparation of TEM sections and help with the SEM, Dr Adam Lockett for the use of his photographic equipment, Frances FitzGibbon for printing the micrographs and Paul Dangerfield for proofreading the manuscript. LITERATURE CITED Field, S.A. and MA. Keller. 1993. Courtship and intersexual signalling in the parasitic wasp Cotesia rubecula. journal of Insect Behaviour 6: 737-750 Golub, MA. and Weatherston, I. 1984. Techniques for ex- tracting and collecting sex pheromones from live insects and from artificial sources. In Hummel, H. and Miller, T., eds. .Techniques in pheromone research. Springer- Verlag, New York. 464 pp. Keller, MA. 1991 . Responses of the parasitoid Cotesia rubecula to its host Pieris rapae in a flight tunnel. Entomologia Experiementalis et Applicata, 57: 243-249. Obara, M. and Kitano, H. 1974. Studies on the courtship behaviour of Apanteles glomeratus L. -1. Experimental studies on releaser of wing-vibrating behaviour in the male. Kontyu.il: 208-214. Reynolds, E.S. 1963. The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. Journal of Cell Biology, 17: 208-212. Smithers, C. 1981 Handbook of Insect Collecting. A.H. and A.W. Reed, Sydney. 120 pp. Snodgrass, RE. 1935. Principles of Insect Morphology. McGraw- Hill Book Co., New York. 667 pp. Tagawa, J. 1977. Localization and histology of the sex phero- mone-producing gland in the parasitic wasp, Apanteles glomeratus. journal of Insect Physiology, 23: 49-56 Tagawa, J. 1983 Female sex pheromone glands in the para- sitic wasps, genus Apanteles. Applied Entomology and Zoology, 18: 416-427. Vinson, SB. 1978. Courtship behavior and source of a sexual pheromone from Cardiochiles nigriceps. Annals of the Entomological Society of America, 71: 832-837. Weseloh, R.M. 1976. Dufour's gland: Source of sex phero- mone in a hymenopterous parasitoid. Science, 193: 695- 697. Weseloh, R.M. 1980. Sex pheromone gland of the gypsy moth parasitoid, Apanteles melanoscelus : revaluation and ul- trastructural survey. Annals of the Entomological Society Of America, 73: 576-580. Zar, J.H. 1984 Biostatistical Analysis. 2nd edition. Prentice- Hall, Englewood Cliffs. 718 pp. J. HYM. RES. Vol. 3, 1994, pp. 157-173 A Review of the World Species of Orthomiscus Mason (Hymenoptera: Ichneumonidae: Tryphoninae)* Virendra K. Gupta Entomology and Nematology Department, University of Florida, Gainesville, Florida 32611-0620, U.S.A. Abstract. — The world species of the exentenne genus Orthomiscus Mason are reviewed. O. amurensis Kasparyan is transferred to Kristotomus , leaving six species under Orthomiscus in the Holarctic Region: O. eridolius Kasparyan, medusae Kasparyan, pectoralis (Hellen), -platyura Mason, simplex (Mason) and unicinctus (Holmgren). Diagnostic characters, species relationships and a key to the species are given. Diagnostic characters are illustrated by diagrams and photographs. INTRODUCTION The genus Orthomiscus was described by Ma- son (1955) to accommodate two new Nearctic species, O. platyura and O. leptura. It belongs to the Kristotomus-Complex of genera in the tribe Exenterini (Gupta 1990), including Kristotomus Mason, Kerrichia Mason and Orthomiscus Mason. This group is characterized by having the apical rim of the hind tibia with a fringe of long, close bristles on the inner side and with a flat polished area on the lower and inner sides between the apical fringe of bristles and the tarsal socket; by having the hind tibia widest at apex and truncate; and by its short ovipositor which hardly surpasses the tip of abdomen. These genera also exhibit the usual exenterine characters such as the presence of only one spur on the middle tibia and the absence of spurs on the hind tibia. Mason (1962) commented on the relationships of the genus and included two European species: Exenterus unicinctus Holmgren, 1858 (with which he synonymized O. leptura) and Cteniscus pectoralis Hellen, 1951. Kasparyan (1976, 1986, 1990) de- scribed three additional new species from Russia: O. medusae, O. amurensis, and O. eridolius. Gupta (1990) transferred the Japanese Kristotomus sim- plex Mason, 1962 to this genus. In the present study, O. amurensis Kasparyan is transferred to the genus Kristotomus as it exhibits most of the charac- ters of that genus, like the tapered and slender ovipositor and sheaths, lower valve of ovipositor without teeth and egg with a simple stalk and anchor. Orthomiscus thus is recognized as a Hol- arctic genus with six species. Biology. — Not much is known about the biol- ogy of Orthomiscus except that O. pectoralis has been reared from late larval cimbicids (Trichiosoma lucorum L. and T. tibiale Stephens) that feed on birch and ash (Betula and Crataegus) in Europe (Jussila 1975). Material. — This paper is based on the speci- mens present in the collection of the American Entomological Institute, Gainesville (GAINES- VILLE). The types of the Russian species were borrowed from the collections of the Zoological Institute, Russian Academy of Science, St. Peters- burg (ST. PETERSBURG) through the kindness of Dr. D. R. Kasparyan. Under synonymy of various species, the location of the type is indicated by the 'Florida Agricultural Experiment Station Journal Series R-03128. 158 Journal of Hymenoptera Research city where the type is housed (see Townes, Momoi & Townes 1965 for details). The key to the species is based on the females which are more diagnostic. Genus ORTHOMISCUS Mason Orthomiscus Mason, 1955. Canadian Journal of Zoology, 33: 63. Type: Orthomiscus platyura Mason; original designation. Taxonomy: Mason 1962: 1273, Gupta 1990: 7; Kasparyan 1990: 36. decurved, not tapered apically and not extending beyond apex of abdomen. Upper valve of oviposi- tor flattened dorsally in basal half. Lower valve of ovipositor with an obliquely serrated edge (Figs. 10, 12, 18, 29). Ovipositor sheaths short, broadly triangular or sometimes a little slender. Female subgenital plate thin, creased medially and pinched apically. Egg stalk short, anchor simple, button- like or flattened and enclosing part of the egg (Figs. 4-14). Diagnostic characters. — Head subcuboidal, swollen behind eyes. Mandibular teeth equal or subequal. Occipital and hypostomal carinae com- plete. Occipital carina bent inwards, meeting hypostomal carina at a right angle above base of mandible by a distance equal to the basal width of mandible. Epomia distinct. Epicnemial carina ex- tending at least to the middle of hind margin of pronotum. Basal area of propodeum confluent with areola or only partly separated from it. Legs slender, elongate. Tarsal claws with 2-6 basal teeth. Areolet present, oblique, sometimes second intercubitus partly weak. Nervellus inclivous, in- tercepted in its lower 0.1 5 to 0.3. Tergitel elongate, usually about 2.0x as long as its apical width, its dorsomedian and dorsolateral carinae sharp and distinct; dorsolateral carina usually extending to the entire length of the tergite and passing just dorsad of spiracle and not broken there. Tergite 2 usually with a pair of strong to weak oblique basolateral grooves. Ovipositor short, stout and SPECIES RELATIONSHIPS On the basis of egg structure, the six species of Orthomiscus form two distinct species groups. Group 1 comprises platyura Mason (type-species), medusae Kasparyan and eridolius Kasparyan (Figs. 4, 7, 9), where the egg stalk is short and wide and the anchor is four-pronged and encloses part of the egg. Group 2 comprises pectoralis (Hellen), unicinctus (Holmgren) and simplex (Mason) (Figs. 8, 12, 30) where the egg has a simple short stalk and a short button-like anchor. In the latter group, the stalk arises apicoventrally in pectoralis and unicinctus while in simplex it is central and the anchor is flat and oval. Gupta (1990) stated that the plesiomorphic state of the egg in Cteniscini was probably an egg with a short stalk and anchor. This type of egg is seen in some generalized members of the genus Kristotomus (like claviventris, santoshae and ctenonyx). Kristotomus also exhibits more general- eridolius platyura T2 with deep grooves modified egg anchor/stalk pectoralis Kristotomus stalk medusae wide mandible (reversal) body punctate unicinctus second segment -of maxillary palp swollen simplex simple egg anchor/stalk mandible widened apically Diagram depicting the relationships of Orthomiscus species Volume 3, 1994 159 ized characters from which Orthomiscus and Under group 1, Orthomiscus platura has deep Kerrichia may be derived. The Orthomiscus species oblique grooves on tergite 2 separating it from O. closest to Kristotomus appears to be O. simplex, medusae and eridolius. In medusae the mandibles are which was originally described in Kristotomus and widened apically and the lower tooth is longer considered a primitive member of the genus by than the upper (reversal). It also has the basal vein Mason (1962), primarily because the mandible is more strongly curved medially (Fig. 38). Under slightly widened apically in this species and in group 2, O. simplex has the mandible widened several species of Kristotomus. O. medusae also apically with the lower tooth longer than the up- exhibits a widened mandible but the egg structure per. This is not the condition in O. pectoralis and O. is very different. I consider that the widened man- unicinctus. O. pectoralis has the body, including dible of these two species arose by convergence. face, punctate and the propodeum rugose and The relationships of all Orthomiscus species may punctate in places. O. unicinctus has the second be depicted as follows: segment of maxillary palp swollen (Fig. 39). KEY TO THE SPECIES la Basal vein of fore wing slightly curved close to its junction with the medius vein. Petiolar area of propodeum almost pentagonal, about 0.75x the combined length of areola and basal area (Fig. 3). Apical transverse carina of propodeum angulate at its junction with lateral and median longitudinal carinae (Fig. 23). Mandibular teeth equal in length (Figs. 2, 15). Hypostomal carina not raised at its junction with occipital carina. Egg stalk situated apicoventrally or in the middle 2 lb Basal vein of fore wing more strongly curved, curvature in the middle of the vein. Petiolar area of propodeum small, circular, about half the combined length of areola and basal area. Apical transverse carina of propodeum almost evenly curved between the lateral longitudinal carinae. Mandible slightly widened apically and the lower tooth longer than the upper (Figs. 27, 31). Hypostomal carina slightly raised at its junction with occipital carina. Egg stalk situated in the middle of the egg (Figs. 7, 30) ... 5 2a Second segment of maxillary palp swollen, about 2. Ox as wide as the rest. Abdomen black dorsally, tergites 2-3 often with triangular yellow patches or their apical margins yellow. Pronotum yellow. Mesopleurum partly yellow. Mesosternum yellow. Ovipositor slender (Figs. 20, 21). Egg stalk short and apicoventral (Figs. 6, 11). Egg anchor button shaped. Holarctic 2. unicinctus (Holmgren) 2b Second segment of maxillary palp similar to the others, not swollen. Abdomen beyond first tergite black or brown with middle segments of ten with large yellowish brown patches. Egg stalk and anchor various (Figs. 4, 8, 9) 3 3a Face with definite punctures. Propodeum rugose and punctate in places. Tergites 4-5 as long as wide. Ovipositor sheaths slender. Propleurum yellow. Mesopleurum largely black. Abdominal tergites black with narrow apical yellow stripes. Egg stalk short and apicoventral (Figs. 8, 12). Egg anchor button- shaped. Eurasia, Japan 1. pectoralis (Hellen) 3b Face polished, only with sparse punctures. Propodeum polished to subpolished, without distinct punc- tures or rugosities. Tergites 4-5 transverse. Ovipositor thick, its sheaths triangular (Figs. 18,25). Egg oval, its stalk very short, central. Egg anchor partly enclosing the egg (Figs. 4, 5, 9, 14) 4 4a Tergite 2 with deep oblique basolateral grooves. Abdominal tergites beyond first largely reddish brown. Pronotum, mesopleurum ventrally and mesosternum yellow. Nervellus intercepted at its lower 0.25-0.3. Egg stalk short (Fig. 4), anchor 4-pronged (Fig. 5). U.S.A., Canada 5. platyura Mason 4b Tergite 2 without oblique grooves. Abdominal tergites black, with margins of tergites 2-3 (sometimes 4-5 also) yellow. Mesosternum of female black. Nervellus intercepted at its lower 0.15-0.2. Egg anchor (Fig. 9) broad and partly enclosing the egg. Russia: Far East 6. eridolius Kasparyan 5a Mesopleurum of female yellow ventrally, in male with yellow spots. Tarsal claws with 4-5 teeth. Female subgenital plate sharply creased medioventrally. Ovipositor slender (Fig. 29). Egg with a simple, oval anchor (Fig. 30). japan 3. simplex (Mason) 5b Mesopleurum black. Tarsal claws with a pair of weak teeth. Female subgenital plate convex ventrally, not strongly creased. Ovipositor thick (Fig. 10). Egg with a 4-lobed anchor (Fig. 7). Russia: Far East, Japan 4. medusae Kasparyan 160 Journal of Hymenoptera Research 1. Orthomiscus pectoralis (Hellen) Figs. 8, 12 Ctemscus pectoralis Hellen 1951. Notulae Entomologicae Helsingfors, 31: 31. F. des. Type: Female, Finland: Esbo (HELSINKI). — Kernch 1952. Bulletin of the British Museum (Natural History) Entomology Series, 2 (6): 432. F. des., distr. Orthomiscus pectoralis: Mason, 1962. Canadian Entomologist, 94: 1 274. n. comb. Japan: Nagano Prefecture: Kamikochi. — Townes, Momoi & Townes, 1965. Memoirs of the American Entomological Institute, 5: 109. cat. — Jussila, 1975. Acta Entomologica Fennica,41 (2): 53. des. of male, host records. — Kasparyan, 1976. Entomologicheskoe Obozrenie, 55: 137. — Kasparyan, 1977. Nasekomye Mongoliaca, 5: 463. Mongolia. — Kasparyan, 1990. Fauna of USSR Insecta, Hymenoptera, 3 (2): 43. key, des., fig. Russia. No specimens of this species were available for study. Kerrich (1952) and Kasparyan (1990) have provided detailed descriptions of this spe- cies. The description that follows is adapted from Kerrich (1952). Diagnosis. — This species appears different from the others by having the face and thorax extensively punctate and the propodeum rugoso- punctate (cf. Kerrich, 1952 description). All other species have only weak to indistinct body puncta- tion. Tergites 4-5 as long as wide. The egg has a short apicoventral stalk and a small button-like anchor (Fig. 8). Male and female. — Face moderately punctate. Clypeus with fine punctures. Epomia short and weak. Mandibular teeth equal in length. Notauli sharply impressed. Scutellum strongly convex, finely punctate. Epicnemial carina incomplete medially. Sternaulus superficially impressed. Propodeum rugose to punctate in places. Areola about as long as wide, confluent with basal area. Abdomen elongate and slender. Abdomen elon- gate and slender. Tergite 1 about 1.6x as long as its apical width, its dorsomedian and dorsolateral carinae strong. Postpetiole with a pitlike depres- sion bounded by carinae. Tergite 2 with small but distinct thyridia. Tergites 4-5 as long as their basal width. Apical abdominal segments compressed in female. Female subgenital plate less sharply folded than in O. unicinctus (with which Kerrich com- pared this species). Ovipositor sheaths slender. Ovipositor comparatively stout as is usual in the genus (Fig. 12). Color. — Ground color black. Mouthparts, clypeus, face, and temples broadly, yellow. Scape and pedicel pale brownish. Thorax and propodeum black, with pronotum, scutellum, metascutellum, a mark above epicnemial carina, mark above middle coxa, tegula and subtegular ridge, yellowish brown. Propleurum yellow. Fore and middle coxae and trochanters yellow. Fore and middle femora yellowish brown, their tibiae and tarsi paler. Hind coxa reddish brown. Hind femur, tibia and tarsus more brownish to reddish brown. Abdomen black above, yellow below. Tergites 1-6 with narrow, pale apical margins and tergite 7 with an obscure, triangular yellow mark. Ovipositor sheaths yellow, infuscate above and rufous towards apex. Egg. — (Fig. 8) Egg with a short apicoventral stalk and a button-like anchor. Length. — 5.0-10.0 mm. Fore wing 3.8-8.5 mm.; Ovipositor about 0.5 mm. Distribution. — Europe (Finland), Russia, Mongolia, Japan. Mason (1962) reported this species from Japan and included two females from Nagano Prefec- ture: Kamikochi, collected by Townes in 1954. These specimens could not be located in the AEI collections. 2. Orthomiscus unicinctus (Holmgren) Figs. 6, 11, 19-23,36,39 Exen terusumanct us Holmgren, 1858. SvenskaVetensk.-Akad. Handlirsch (n. f), 1: 234. M, F. key, des. Lectotype (selected by Roman, 1914), Female, Sweden: Lapland: Tarna (STOCKHOLM). Exenterus macrocephalus Holmgren, 1858. Svenska Vetensk.- Akad. Handlirsch (n. f), 1: 243. M. des. Type: Male, Sweden (STOCKHOLM). Syn. by Roman, 1914. Ctemscus macrocephalus:]acob &Tosquinet, 1896. Annalesde la Societe Entomologique Beige, 34: 108. new comb. Cteniscus unicinctus: Roman, 1914. Arkiv for Zoologie, 9 (2): 17. syn., des. — Kerrich, 1942. Transactions of the Soci- ety for British Entomology, 8 (2): 63-64. Britain. — Kerrich, 1952. Bulletin of the British Museum (Natural History) Entomology Series, 2 (6): 434. key, des., fig. Localities in Britain, France, Belgium, Germany, Swe- den. Orthomiscus leptura Mason, 1955. Canadian Journal of Zool- ogy, 33: 64. M, F. key, des., fig. Type: Female, U.S.A.: New Hampshire: Pinkham Notch (AEI, Gainesville) Examined. Syn. by Mason, 1962. Orthomiscus unicinctus: Mason, 1962. Canadian Entomolo- gist, 94: 1274. syn., distr. Japan: Sapporo. — Townes, Momoi & Townes, 1965. Memoirs of the American En- tomological Institute, 5: 109. cat. — Mason, 1966. Cana- dian Entomologist, 98: 48. key. Japan. — Carlson, 1979. Catalog of Hymenoptera in America North of Mexico, 1 : 377. cat. U.S.A., Canada. — Kasparyan, 1990. Fauna of Volume 3. 1994 161 USSR Insecta, Hymenoptera, 3 (2): 45. key, des., fig. Russia. Diagnostic features.— Second segment of max- illary palp widened and flat, about 2.0x as wide as the rest (Fig. 39). Oblique impressions on tergite 2 faint. Ovipositor comparatively more slender than in the previous species (Figs. 20, 21). Egg with a short apicoventral stalk and a button-like anchor (Figs. 6, 11). Temple yellow only on its basal half. Scutellum black except at apex. Hind coxa with blackish brown patches. Abdomen largely black and with yellow patches on tergites 2 and 3. Tho- rax of male largely black, that of female generally yellow in lower half, but color variable. Male and female— Flagellum with 23-24 seg- ments in male and 23-26 segments in female. First segment about 1.4x as long as the second. Second segment of maxillary palp widened and flat, about 2.0x as wide as the rest of the segments (Fig. 39). Mandibular teeth equal in length but lower tooth more prominent than the upper. Hypostomal ca- rina of normal shape, not conspicuously raised at its junction with occipital carina. Vertex widened posteriorly. Interocellar distance 0.25x the ocellocular distance. Thorax subpolished. Mesoscutum rather strongly convex. Notauli distinctly impressed up to middle of mesoscutum. Lateral carinae of scutel- lum distinct only at base of scutellum. Epomia sharp across pronotal collar. Sternaulus and epicnemial carina moderately impressed. Propodeum convex, subpolished. Areola (Fig. 23) widened at its junction with costulae, confluent with basal area. Basal vein of fore wing uniformly but weakly arched (Fig. 36). Nervellus intercepted in its lower 0.25-0.33. Abdomen subpolished. Tergite 1 slender (Fig. 19), its dorsomedian carinae strong and extending to 0.7-0.8 the length of the tergite, its dorsolateral carina sharp and extending to the entire length of the tergite. Tergite 2 with distinct thyridia and faint oblique grooves. Female subgenital plate folded medially and pointed apically. Ovipositor slender, of uniform diameter (Fig. 21). Ovipositor sheaths (Figs. 20, 21) comparatively narrow. Egg anchor small and situated apicoventrally (Fig. 6). Color. — Ground color black. Thorax almost wholly black in male and some females. Abdomi- nal tergites largely black. Scape, pedicel below, mouthparts, clypeus, face, inner orbits up to top of eyes, malar space, and temples on lower half, yellow. Propleurum, lower half of pronotum, its upper margin, tegula, subtegular ridge, tip of scutellum, epicnemium, mesosternum, lower half of mesopleurum, abdominal venter, and sides of abdomen, yellow . Legs yellow to yellowish brown. Fore and middle coxae and trochanters yellow. Hind coxa largely brownish or yellow with large black or brown patches. Hind femur brown. Hind coxa largely yellow in specimens from Ontario and blackish in specimens from Alaska. Tergites 2 and 3 often with triangular yellow patches or with their apical margins yellow. In males the yellow along the inner orbits extending only to the level of antennal sockets, temple yellow on its lower 0.3, and thorax almost wholly black. The color is quite variable. The European specimens were not examined to ascertain the range of variation. Kerrich (1952: 434) and Kasparyan (1990: 45) have described this species in detail. Egg — (Fig. 6, 11, 21). Egg small, oval, with a short apicoventral stalk and a button-like anchor. Length. — 6.0-8.5 mm.; fore wing 5.0-7.0 mm.; ovipositor 0.4-0.6 mm. Kerrich (1952) stated that a female measuring 4.0 mm. was seen from Lapland. Specimens examined. — CANADA: Quebec: Stoneham, 1 male, 1 female, 21.VI.1938, H. & M. Townes. U.S.A.: New Hampshire: Pinkham Notch, 1 female, 25.VI.1938, H. & M. Townes; Mt. Madi- son, 3 males, 24.VI.1938, H. & M. Townes. Ver- mont: L. Willoughby, 2 male, 17, 25.VI.1945, C.P. Alexander. (Types and paratypes in AEI, Gainesville). Non-type material: U.S.A.: Oregon: Corvallis, 2 males, 1 female, V-VII. 1978, H. & M. Townes. Alaska: Anchorage, 1 female, 6- 16.VII.1976, Petre Rush. CANADA: Ontario: Cumberland, 1 male, 3 females, VI.1975, L. Ling. Distribution.— -U.S.A., Canada, Europe (Brit- ain, France, Belgium, Germany, Sweden), Russia, Japan. 3. Orthomiscus simplex (Mason) Figs. 26-30, 42 Kristotomus simplex Mason, 1966. Canadian Entomologist, 98: 46. M, F. key, des., fig. Type: Female, Japan: Nagano Prefecture: Kamikochi (GAINESVILLE. Examined. — Kasparyan, 1976. Entomologicheskoe Obozrenie, 55: 150. (Entomological Review, 55: 108). key. Orthomiscus simplex: Gupta, 1990. Contrib. Amer. Ent. Inst., 25 (6): 8. n. comb. — Kasparyan, 1990. Fauna of USSR Insecta, Hymenoptera, 3 (2): 44. key, des., fig. Japan. 162 Journal of Hymenoptera Research Diagnostic features. — Mandible widened apically, its lower tooth wider and longer than the upper (Fig. 27). Hypostomal carina slightly raised at its junction with occipital carina (Fig. 42). Basal vein of fore wing curved. Egg reniform-ovate, with a short central stalk and a simple anchor. Egg surface coarsely granular. Head and thorax in female yellow on lower half and black on upper half, in male largely black. Abdominal tergites black, their apical margins yellow. Male and female. — Scape about 1.5x as long as wide. Flagellum with 24 segments. First flagellar segment 1.4x as long as the second. Face (and rest of head also) smooth and shiny. Mandible wid- ened apically, the lower tooth wider and slightly longer than the upper (Fig. 27). Malar space 0.4x the basal width of mandible. Maxillary palp seg- ments normal, not swollen (Fig. 42). Hypostomal carina slightly raised at its junction with occipital carina (Fig. 42). Vertex widened behind eyes (Fig. 26). Interocellar distance 0.4x the ocellocular dis- tance. Thorax largely smooth and shiny. Notauli distinct anteriorly. Lateral carinae of scutellum distinct. Epomia normal, not very sharp. Epicnemial carina extending to half the height of mesopleurum. Sternaulus indistinct. Propodeum normally areola ted. Petiolar area small and round. Apical transverse carina uniformly curved be- tween its junction with lateral longitudinal cari- nae. Basal vein in fore wing medially curved, more strongly so than in the preceding species. Second intercubitus of areolet sometimes absent. Abdomen smooth and shiny. Tergite 1 about 2.25x as long as its apical width, its dorsomedian carinae extending up to about 0.75 its length, its dorsolateral carina complete, sinuate and passing just above the spiracle (Fig. 28). Tergite 2 with weak basolateral oblique grooves. Female subgenital plate strongly creased and pointed apically. Ovipositor (Fig. 29) somewhat slender and cylindrical, its teeth weak but present at the tip. Ovipositor sheaths small and slender. Male claspers broader than in other species. Color. — Ground color black. Male compara- tively darker than the female. Face, clypeus, man- dibles, lower half of temples, pronotal collar, hind corner of pronotum, whole of fore and middle legs, hind coxae and trochanters, and abdominal venter, yellow. Ventral side of scape and pedicel and apex of scutellum yellowish-brown. Hind femur, tibia and tarsus blackish brown (male) to yellowish brown (female). Pronotum of female more extensively yellow and mesosternum and lower part of mesopleurum may also be yellow. Abdominal tergites black with yellow apical stripes, particularly on tergites 2-3. Egg. — Reniform-ovate (Fig. 30). Stalk shorter than its own diameter, arising near the middle of the egg. Anchor about 0.33 as long as the egg, flat and oval. Egg surface coarsely granular. Length. — 4.5-6.5 mm.; fore wing 4.5-6.0 mm.; ovipositor about 0.5 mm. long. Specimens examined. — JAPAN: Nagano Prefec- ture: Kamikochi, 6 males and 2 females (holotype, allotype and paratypes), 23-31. VII. 1954, Townes family (AEI, Gainesville). Distribution. — Japan. Discussion. — Mason (1966) placed this species under Kristotomus, stating that it was the most primitive member of the genus and closely related to Ortlnmuscus, particularly in the structure of the egg and development of the dorsolateral carina of the petiole. Apart from the widened mandible, this species shares most characters of Orthomiscus, particularly the structure of the ovipositor, in which the lower valve has teeth-like indentations. The dorsolateral carina of the petiole and the egg struc- ture also relate it to the present genus. 4. Orthomiscus medusae Kasparyan Figs. 7, 10, 31,38, 41 Orthomiscus medusae Kasparyan, 1976. Entomologicheskoe Obozrenie, 55: 140 (Entomological Review, 55 (1): 99). M, F. des., fig. Type: Female, Russia: Sakhalin: Kunashir Island (ST. PETERSBURG). Examined — Kasparyan, 1990. Fauna of USSR Insecta, Hymenoptera, 3 (2): 42. key, des., fig. Russia. Diagnostic features. — Mandible widened apically, the lower tooth longer than the upper (Fig. 31). Hypostomal carina raised at its junction with occipital carina. Scutellum subconvex. Basal vein of fore wing strongly curved medially (Fig. 38). Tergite 2 with very faint to indistinct oblique grooves. Female subgenital plate convex ventrally, not sharply creased. Egg (Figs. 7, 10) with a short medioventral stalk and a four-pronged anchor enclosing the egg. Thorax and abdomen black; sometimes pronotal collar and presternum yellow and tergites 2-3 with yellow marks, particularly in females. Volume 3. 1 994 163 Male and female. — Flagellum with 23-24 seg- ments in male and 25-27 segments in female. First flagellar segment about 1 .3x as long as the second. Face and clypeus subpolished, with sparse punc- tures. Malar space about 0.33-0.4.X (male) to 0.4- 0.5x (female) the basal width of mandible. Man- dible slightly widened apically (Fig. 31), its lower tooth a little longer than the upper. Maxillary palp slender, normal (Fig. 41). Vertex widened posteri- orly, smooth and shiny. In terocellar distance 0.25x the ocellocular distance. Epomia present but not very strong. Notauli distinct to middle of mesoscutum. Scutellum subconvex, its lateral carinae confined to base. Epicnemial carina strong, ending in middle of hind margin of pronotum. Sternaulus weakly in- dicated anteriorly. Propodeum areolated but its longitudinal dorsal and sublateral carinae quite often erased beyond the costulae or absent; com- bined length of basal area and areola approxi- mately 1 .5x the length of apical area (according to the original description). In one paratype and in the Japanese specimens examined in the AEI col- lection, the areola is narrow, 2. Ox as long as wide, confluent with the basal area, and a little widened at the costulae. Basal vein curved medially (Fig. 38). Tarsal claws with a pair of weak teeth at base. Tergite 1 2.0x as long as its apical width; its dorsomedian carinae extending to 0.75 its length and its dorsolateral carinae strong and complete. Postpetiole with a median oval depression. Terg- ite 2 with rather weak oblique basolateral impres- sions. Female subgenital plate convex, not strongly creased ventrally, apically narrow and pinched. Ovipositor stout (Fig. 10). Ovipositor sheaths broad. Color. — Ground colorblack. Scape and pedicel reddish brown dorsally and lighter ventrally. Fla- gellum yellow. Face, clypeus, mouthparts, and lower half of temple, yellow. Thorax black with pronotal collar and prosternum yellow in female. Apex of scutellum brown. Coxae and trochanters yellow. Fore and middle legs otherwise yellowish brown. Hind femur, tibia and tarsus blackish brown. Abdomen black with tergites 2 and 3 with triangular yellow marks. Tergite 4 with a faint yellow mark in the paratype examined. In one male studied only yellow apical lines are present on the abdominal tergites. Egg-— Oval in shape (Fig. 7, 10), with a medioventral short stalk, its anchor four-pronged, enclosing part of the egg, somewhat similar to that seen in O. platyura. Length. — 7.0-9.0 mm.; fore wing 5.5-7.0 mm.; Ovipositor about 0.5 mm. Specimens examined. — RUSSIA: SAKHALIN: Kunashir Island: Sernovodsk, 1 female (paratype), 26.VIII.1973, D. R. Kasparyan. JAPAN: Nagano Prefecture: Kamikochi, 1 male, 22. VII.1954, 1 male, 24. VII. 1954 (labeled O. unicinctus Tow.), Townes family (AEI, Gainesville). Distribution. — Russia, Japan. Relationships. — Structurally this species is close to O. simplex (Mason) from which it differs mainly in the complex structure of the egg anchor, black mesopleurum, male tarsal claws with a pair of weak teeth at base, female subgenital plate convex but not strongly creased ventrally and by possess- ing a comparatively thicker ovipositor. The egg structure relates it to O. plati/ura. 5. Orthomiscus platyura Mason Figs. 1-5, 15-18,37,40 Orthomiscus platyura Mason, 1955. Canadian Journal of Zool- ogy, 33: 66. M, F. key, des., fig. Type: Female, U.S.A.: Rhode Island: Westerly (GAINESVILLE). Examined. — Carlson, 1979. Catalog of Hymenoptera in America North of Mexico, 1:377. cat. U.S.A. New Hampshire, Vermont, California. Canada: Quebec, Ontario. Diagnostic features. — Areola narrow, parallel- sided and confluent or partly separated from the basal area. Tergite 2 with rather deep oblique basolateral grooves. Ovipositor stout, its sheaths broadly triangular. Egg with a four-lobed anchor, as large as the egg. Scutellum mostly yellow. Hind coxa yellow. Abdominal tergites beyond first red- dish brown to reddish yellow. Male and female. — Scape about 1.5x as long as wide. Flagellum with 24 segments in male and 26 segments in female. First segment 1.32 to 1.35x as long as the second segment. Face (Fig. 15) pol- ished, depressed laterally. Clypeus convex and polished. Mandibular teeth subequal, upper tooth slightly longer than the lower. Malar space 0.45x the basal width of mandible. Second segment of maxillary palp slender and similar to other seg- ments (Fig 40). Vertex polished, quadrate, wid- ened behind eyes, 1.75x as wide as eye in dorsal view. Interocellar distance 0.22x the ocellocular distance. Temple in lateral view 1.7x to 1.75x as wide as eye. 164 Journal of Hymenoptera Research Thorax mostly smooth and shiny. Mesoscutum convex, pubescent. Notauli deep and distinct to middle of mesoscutum. Scutellum flat, with lateral carina confined to base. Prono- tum somewhat dull, epomia strong and extend- ing close to upper edge of pronotum. Epicnemium dull, epicnemial carina strong and extending to 0.75 the height of mesopleurum. Sternaulus indi- cated on anterior 0.4 of mesopleurum. Propodeum convex, shiny. Areola narrow, parallel-sided, about 2.0x as long as wide (Fig. 3), confluent with basal area or only partly separated from it. Abdomen shiny. Tergite 1 2. Ox as wide at apex as at base, about 2. Ox as long as its apical width, its dorsomedian carinae extending up to about 0.8 its length (carinae weaker apicaUy, wid- ened, and enclosing a depressed area between them), its dorsolateral carina complete and strong (Fig. 17). Tergite 2 with deep, oblique, basolateral grooves. Ovipositor thick, decurved (Fig. 18). Ovipositor sheaths short, broadly triangular. Color. — Black and yellow. LowerO. 66 of head, scape, flagellum beyond 4th segment, pronotum, scutellum, metascutellum, presternum, mesoster- num, lower part of mesopleurum, fore and middle coxae and trochanters, sternites, ovipositor sheaths, and subgenital plate, yellow. Legs otherwise and abdomen (except tergite 1 ), yellowish brown. Terg- ite 1 black. Brown to blackish patches present on abdominal tergites and pronotum. Wings hyaline. Egg— (Figs. 4, 5, 18). Short, oval, ventrally flat and bearing on nearly half its length a short but wide stalk; egg anchor large and enclosing most of the egg. Anchor bearing two long posterior and two long anterolateral arms that extend toward the dorsal surface of the egg. Length. — 6.0-8.0 mm.; fore wing 5.0-6.5 mm.; ovipositor about 0.4 mm. Specimens examined. — CANADA: Ontario: Cumberland, 1 female, ll.VI.1975,L. Ling. U.S.A.: Rhode Island: Westerly, 1 male, 2 females, 9- 1 1 .VI. 1936, M. Chapman. New York: McLean Bogs, Tompkins Co., 1 female, 30. VI. 1939, J. G. Franclemont. New Jersey: Tabernacle, 1 male, 11.VI.1939, H. K. Townes; Millwood, 1 female, 21.VI.1936, H. K. Townes. Maryland: Takoma Park, 1 male, 2.VII.1944, H. & M. Townes. Pennsylvania: Trout Run, 1 female, 29.V.1938, H. K. Townes; Spring Brook, 1 female, 27.VI.1945, H. K. Townes; Michigan: Huron Mrs., 2 males, 2 females, VI. 1961, H. & M. Townes; Iron River, 1 female, 23. VI. 1969, H. & M. Townes. South Carolina: Cleveland, 3 females, V-VI.1971, G. Townes. Alabama: Gulf Shores, 1 male, 22.IV.1968, H. & M. Townes. All these specimens are in the AEI collection at Gainesville, and those collected prior to 1955 are paratypes (see Mason, 1955: 67). Additional paratypes were described by Mason from Maine, Vermont, New Hampshire, Massachusetts, Con- necticut, and Quebec. Distribution. — U.S. A., Canada. 6 Orthomiscus eridolius Kasparyan Figs. 9, 14, 24-25 Orthomiscus eridolius Kasparyan, 1990. Fauna of USSR Insecta, Hymenoptera, 3(2):44. M, F. key, des., fig. Type: Female, Russia: Far East: Khabozavodsk Region (ST. PETERSBURG). Paratypes examined. This species has been described in detail by Kasparyan (1990). Structurally it resembles O. sim- plex (Mason) but shows differences in the curva- ture of the basal vein, shape of the areola, shape of tergite 2, and in the egg structure. The diagnostic features are given below. Diagnosis. — Temple more buccate than in O. simplex. Mandibular teeth equal. Epomia present, moderately strong. Areola slightly widened at its junction with costulae and confluent with basal area. Apical transverse carina of propodeum angulate at its junction with lateral and median longitudinal carinae. Petiolar area appearing pen- tagonal. Basal vein of fore wing only slightly curved near its junction with the medius vein. Nervellus intercepted at its lower 0.15-0.2. Tergite 2 without oblique grooves. Ovipositor short and thick (Fig. 25). Ovipositor sheaths short, tapered. Egg (Figs. 9, 14, 25) oval, with a ventral flattened anchor that encloses almost half of the egg. Coloration essentially similar to that of O. simplex. Head and thorax in female yellow in lower half and black in upper half. Mesosternum of female black or brownish-black. Face of male partly yellow. Abdomen beyond tergite 1 brownish black, with yellow spots. Mesosternum black in female. Face of male largely black. Length. — 4.5-6.0 mm.; fore wing 3.8-4.8 mm.; ovipositor about 0.4 mm. Distribution. — Russia: Far Eastern Region. Orthomiscus amurensis Kasparyan, 1986 is hereby transferred to Kristotomus. Volume 3, 1994 165 Kristotomus amurensis (Kasparyan), new combination Figs. 13, 32-35 Orthomiscus amurensis Kasparyan, 1986: 55. Kasparyan, 1990. Fauna of USSR Insecta, Hymenoptera, 3 (2): 40. key. des., fig. Type: Female, Russia: Far East: Khabozavodsk Region (ST. PETERSBURG). Examined. The holotype of this species was examined in 1992. The ovipositor and its sheaths are tapered and slender and both extend beyond the apex of abdomen. The lower valve of the ovipositor is without teeth. The egg has a simple stalk and anchor. The interocellar distance is about 0.8x the ocellocular distance. The malar space is 0.5x the basal width of mandible. The temple is moder- ately swollen, and is about as wide as the eye width. The dorsolateral carina of tergite 1 is inter- rupted just above the spiracle. The above mentioned characters and the gen- eral body form would place this species in Kristotomus, to which genus this species is hereby transferred. It appears related to Kristotomus bitccatus Kasparyan. ACKNOWLEDGMENTS My thanks are due to Dr. D. R. Kasparyan (St. Peters- burg) for the loan of the type specimens and for his com- ments. The collections of the American Entomological Insti- tute (AEI) were readily available for my studies, for which I am thankful to the President of the Institute. The draft of the paper was kindly read by and commented upon by Drs. John Barron and Charles Porter and their comments have been incorporated. I am also thankful to Mr. Andrei Sourakov for translating Kasparyan's key to the Russian species of the genus for my use. Subsequently I received a similar transla- tion from Dr. Kasparyan. The use of published figures from the works of Henry Townes, D. R. Kasparyan and W. R. M. Mason is duly acknowledged and the source is cited under figure legends. LITERATURE CITED Carlson, R. W. 1979. Family Ichneumonidae. In: Krombein, Smith & others, Catalog of Hymenoptera in America North of Mexico 1: 315-741 (Ichneumonidae pages). Washing- ton, DC. Gupta, Virendra K. 1990. The taxonomy of the Kristotomus- Complex of genera and a revision of Kristotomus (Hy- menoptera: Ichneumonidae: Tryphoninae). Contributions of the American Entomological Institute 25 (6): 1-88. Hellen, W. 1951. Neue Ichneumoniden aus Ostfennoskandien. I. Notulae Entomologicae 31: 25-31. Holmgren, A. E. 1858. Forsok till upstallning och beskrifning av de i Sverige fauna Tryphonider [Monographia Tryphonidum Sueciae]. Svenska Vetensk.-Akad. Handlirsch (N.F.) 1: 93-246, 305-394. Jacob, J. C. and J. Tosquinet. 1890. Catalogue des Ichneumonides de la Belgique appartenant au groupe des Tryphonides. Annales de la Societe Entomologique Beige 34: 44-135. Jussila, R. 1975. Ichneumonological (Hym.) reports from Finland III. Annales Entomologici Fennici 41: 49-55. Kasparyan, D. R. 1976. [New species of ichneumonids of the tribe Cteniscini (Hymenoptera, Ichneumonidae) from eastern Asia. The genera Cycasis Townes, Orthomiscus Mason and Kristotomus Mason.] Entomologicheskoe Obozrenie 55(1): 137-150. [English translation in Entomo- logical Review 55 (1): 99-108.] Kasparyan, D. R. 1977. [Ichneumonids of the subfamilies Pimphnae and Tryphoninae (Hymenoptera, Ichneu- monidae) new for Mongolia and Transbaikalia.] Nasekomye Mongolica [Insects of Mongolia] 5: 456-470. Kasparyan, D. R. 1986. [Two new species of the tribe Exentenni (Hymenoptera, Ichneumonidae) of the Far East] In: Ler, P. A. (Editor) [Systematics and ecology of insects from the Far East.] Academiya Nauk SSSR, Vladivostock, 155 pages. (Pp. 54-57). Kasparyan, D. R. 1990. Fauna of USSR Insecta, Hymenoptera Vol. Ill, No. 2. Ichneumonidae, Subfamily Tryphoninae: Tribe Exentenni, Subfamily Adelognathinae. Leningrad: Nauka Publishing House. 340 pages. Kerrich, G. J. 1942. Second review of literature concerning British Ichneumonidae (Hym.), with notes on palaearctic species. Transactions of the Society for British Entomology 8 (2): 43-77. Kerrich, G. J. 1952. A review, and a revision in greater part, of the Cteniscini of the Old World. Bulletin of the British Museum (Natural History) Entomology Series 2 (6): 305- 460. Mason, W. R. M. 1955. A revision of the Nearctic Cteniscini (Hymenoptera: Ichneumonidae) I. Eudiaborus Kerrich and a new genus. Canadian Journal of Zoology 18-73. Mason, W. R. M. 1962. Some new Asiatic species of Exenterini (Hymenoptera: Ichneumonidae) with remarks on ge- neric limits. Canadian Entomologist 94: 1273-1296. Mason, W.R.M. 1966. A primitive new species of Kristotomus Mason (Hymenoptera: Ichneumonidae). Canadian Ento- mologist 98: 46-49. Roman, A. 1914. Beitrage zur schwedischen Ichneu- monidenfauna. Arkivfor Zoologie 9 (2): 1-40. Townes, H. 1969. The genera of the Ichneumonidae Part I. Memoirs of the American Entomological Institute 11: 1-300. Townes, H., S. Momoi and M. Townes. 1965. A Catalog and reclassification of the Eastern Palearctic Ichneumonidae. Memoirs of the American Entomological Institute 5:1-671. 166 Journal of Hymenoptera Research Figs. 1-14. 1-5. Orthomiscus platyura Mason: 1, habitus, 2, head; 3, propodeum and tergite 1; 4, lateral view of egg; 5, ventral view of egg. Figs. 6-9. Lateral view of egg of: 6, O unicinctus (Holmgren); 7, O medusae Kasparyan; 8, O pectoralis Hellen; 9, O. eridolius Kasparyan. Figs. 10-14. Apex of abdomen showing subgemtal plate, ovipositor and ovipositor sheath of: 10, O. medusae; 11, O. unicinctus; 12, 0. pectoralis; 13, 0. [=Kristotomus] amurensis (Kasparyan); 14, O. eridolius. [Figs. 1-3 after Townes, 1969; 4-6 after Mason, 1955; 7-14 after Kasparyan, 1990.] Volume 3, 1 994 167 17 Figs. 15-18. O. platyura: 15, face and mandible; 16, tip of abdomen showing ovipositor, subgenital plate and ovipositor sheath; 17, tergite 1; 18, enlarged view of ovipositor + egg. 168 Journal of Hymenoptera Research 20 t> (■P 21 Figs. 19-21. Orthomiscus unianctus: 19, tergite 1; 20, apical half of abdomen; 21, ovipositor with egg, ovipositor sheath and subgenital plate. [Photos taken from specimens of O leptura Mason.] Volume 3. 1 994 169 25 Figs. 22-25. 22, vertex of Orthomiscus unicinctus; 23, propodeum of O. unicinctus; 24, vertex of O. eridolius; 25, ovipositor + egg of O. eridolius. [22-23 of O. leptura Mason; 24-25 from holotype of eridolius.] 170 Journal of Hymenoptera Research 28 Figs. 26:31. 26-30. Orthomiscus simplex (Mason): 26, vertex; 27, face and mandible; 28, tergite 1; 29, tip of abdomen showing ovipositor and subgemtal plate; 30, egg. 31, Face and mandible of O medusae. [Figs. 27, 29, 30 after Mason, 1966; 26 and 28 of paratype and 30 after Kasparyan, 1990.] Volume 3. 1994 171 32 35 Figs. 32-35. Kristotomus amurensis (Kasparyan): 32, habitus; 33, face; 34, ovipositor and sheaths; 35, abdomen. [Photos of holotype, Orthonuscus amurensis Kasparyan] 172 Journal of Hymenoptera Research Figs. 36-38. Fore wing showing the curvature of the basal vein in: 36, Orthomiscus unicinctus; 37, O. platyura; 38, O. medusae. Volume 3. 1 994 173 Figs. 39-42. Maxillary palp: 39, Orthomiscus unicinctus; 40, O. platyura; 41, O. medusae; 41, O. simplex. J. HYM. RES. Vol. 3, 1994, pp. 175-206 Revision of West-European Genera of the Tribe Aylacini Ashmead (Hymenoptera, Cynipidae) J. L. Nieves-Aldrey Museo Nacional de Ciencias Naturales (Biodiversidad), c/Jose Gutierrez Abascal 2, 28006 Madrid, Spain Abstract. — The west European genera of Aylacini Ashmead (Hym., Cynipidae) and the genus Antistrophus Walsh from North America are reviewed. Diagnoses of all the genera are presented and an illustrated key for the identification of the genera is provided. Four genera are described as new: Barbotuua gen. n., Hedickwna gen. n., Iraelk gen. n. and Neaylax gen. n. The genus Timaspis Mayr is resurrected and the following synonymies are proposed: Aylacopsis Hedicke 1923 = Timaspis Mayr 1881; Vantina (Hedicke), described within Aulacidea Ashmead. In common with Aulacidea, the new genus exhibits a closed radial cell but it can be distinguished from Aulacidea as follows: antenna with 1 2 segments; mesopleuron not clearly longitudinally striated, in part irregularly reticu- late; incomplete notauli and mesoscutum sculp- tured with piliferous punctures; base of third ab- dominal tergum lacking lateral pubescent patch. Distribution and Biology. — Eastern Mediterra- nean; recorded from Crete, Syria and Israel (Hedicke 1928). The only known species, Hedickiana levantina (Hedicke), induces conspicuous juicy galls in stems of Salvia triloba L. (Lamiaceae). IRAELLA gen. n. Figs. 23-27 Type species: Aulax luteipes Thomson 1877:807. (= Timaspis papavens Kieffer in Goury et Gignon). By present desig- nation (examined) Derivation: Feminine gender. Dedicated to my wife Ira Description. — Head viewed dorsally scarcely less than two times broader than long; temples not expanded behind eyes; viewed frontally (Fig. 26) slightly broader than high; transfacial line slightly more than height of eye ; radiating striae on face absent; clypeus trapezoid-shaped, moderately projecting anteriorly, supraclypeal area slightly prominent, reticulate; frons and vertex with re- ticulate-coriaceous sculpture; female antenna, 13- 14-segmented, male antenna 14-15-segmented; A3 almost of equal length to A4 (Fig. 27). Pronotum (Fig. 23) dorsally relatively short; submedial pits present; mesoscutum reticulate-coriaceous; notauli faint in anterior one third of mesoscutum; median mesoscutal impression visible in posterior 1/3 of mesoscutum; scutellar foveae transverse, confluent; scutellum reticulate; mesopleuron (Fig. 24) reticulate. Wings hyaline; radial cell of forew- ing (Fig. 25) open; three times longer than wide; distal wing margins with hair fringe long. Third abdominal tergum basally with lateral pubescent patch; following segments without punctures. Comments. — The following morphological fea- tures define this new genus: face lacking visible radiating striae; clypeus projecting anteriorly; pronotum relatively short dorsally but with vis- ible submedian pits; mesopleuron entirely reticu- late; radial cell open on front margin. Some char- acters, like the general body shape, face and mesopleuron sculpture and shape of the scutellar foveae, approach the new genus to the Timaspis- Phanacis complex, but the presence of submedian pits in the pronotum is a differentiating trait; other features that distinguish them are: the entirely reticulate face with total absence of radiating striae; clypeus projecting anteriorly; third antennal seg- ment being of almost the same length as fourth and radial cell of the forewing open on front margin. The genus is erected for a single species: Iraella luteipes (Thomson) comb. n. (= Timaspis papaveris Kieffer). Redescription of Aylax luteipes Thomson including statement of the synonymy indicated above was recently published (Nieves-Aldrey, in press). Distribution and Biology. — Genus distributed from Sweden to Spain and Iran. The only known species of the genus is associated with Papaver somniferum L. (Papaveraceae), on the stems of which it produces inconspicuous galls. I have had the chance to examine in the Natural History Museum of London some specimens from Iran obtained from stems of Papaver pseudorientale and P. bracteatum Lindley, which might correspond to a different species within the same genus. ISOCOLUS Forster Isocolus Forster 1869:330, 334 ( = Eubothrus Forster). Type spe- cies: Diastrophus scabiosae Giraud 1859. Monobasic and original designation (examined) Diagnosis. — Head, viewed dorsally, about two times broader than long; temples at most only slightly expanded behind eyes; viewed frontally 1.3 times broader than high; transfacial line around 1.5 times the height of eye ; malar space shorter than height of eye; radiating striae visible later- ally, usually incomplete, sometimes present even in the supraclypeal area; antenna of female with 1 3 antennomeres, A3 shorter than A4; antenna of male with 14 segments. Pronotum dorsally long and submedial pits conspicuous, broadly sepa- rated; mesoscutum with coriaceous-rugulose sculpture, sometimes with transverse rugae; notauli usually complete; median mesoscutal im- pression usually visible at least in posterior one third of mesoscutum; scutellar foveae distinct, well defined; scutellum rugulose; mesopleuron with conspicuous longitudinal striae; wings hya- 186 Journal of Hymenoptera Research line; radial cell of forewing open; Rl and Rs not quite reaching anterior margin of wing; distal margin of wing with hair fringe very short. Third abdominal tergum usually without lateral pubes- cent patch, rarely with pubescence; abdominal tergites usually punctate, sometimes weakly, from fourth tergite; ventral spine of hypopygium short. Tarsal claws simple. Comments. — The genus is defined mainly on the basis of wing characters as follows: Rl and Rs not quite reaching to anterior margin of wing, radial cell clearly open and wing fringe absent or very short. Distribution and Biologi/. — The distribution is Palaearctic: five species are known in Western Europe. From the biological point of view the genus is associated with Asteraceae; the species inducing galls in stems, achenes or Lnvolucral bracteae of species of Centaurea and Serrahda (Asteraceae). Redescription Of Some Species Isocolus scabiosae (Giraud) Diastrophus scabiosae Giraud 1859:368 Diastrophus areolatus Giraud 1859:369. Syn. n. Isocolus scabiosae (Giraud): Forster 1869:334 Isocolus rogenhoferi Wachtl 1880:542. Syn. n. Aulax scabiosae (Giraud): Mayr 1882: 10 Aylax scabiosae (Giraud): Dalla Torre and Kieffer 1910:671 Isocolus scabiosae (Giraud): Rohwer and Fagan 1917:369 Isocolus scabiosae (Giraud): Weld 1952:271 Material examined. — The type series of D. scabiosae Giraud, in the MNHP, comprises 63 speci- mens: 30<5, 339, mounted on micro-pins, as well as some galls on achenes of Centaurea scabiosa. 1 9 , set next to a male, with handwritten label "3 mai" and "Museum Paris, coll. GIRAUD", is hereby designated lectotype. Paralectotypes, 18 6 and 12 9 Redescription.— FEMALE. (Figs. 71-79). Length 2.8-3.3 mm. Black; gaster red-brown to black; an- tenna and legs red brown; coxa, base of femora and first two segments of antenna darker. Head viewed dorsally (Fig. 71) 2.2 times broader than long; viewed frontally (Fig. 72) 1.3 times broader than high; temples only slightly expanded behind eyes; frons and vertex with sculpture coriaceous- alutaceous; POL slightly shorter than OOL , OOL about four times greatest diameter of lateral ocel- lus; transfacial line around 1.6 times height of eye; malar space slightly shorter than height of eye; face laterally with radiating striae; supraclypeal area slightly raised, with some striae; antenna (Fig. 73) about 2/3 the length of the body, with 13 antennomeres, A3 slightly shorter than A4. Prono- tum (Fig. 75) coriaceous, laterally pubescent; dor- sally long, with distinct submedial pits ; mesoscutum (Fig. 76) with coriaceous-rugulose sculpture posteriorly with some conspicuous trans- verse rugae; notauli complete, well impressed posteriorly; median mesoscutal impression dis- tinct in posterior half of mesoscutum; scutellar foveae large, rounded; shining and almost smooth; separated by broad posteriorly septum; scutellum ovate, rugulose with the posterior half more or less flat and sloping backwards; mesopleuron (Fig. 77) longitudinally striated. Wings hyaline or very lightly infuscated; Rl and Rs of forewing not quite extending to wing margin and radial cell open (Fig. 79); Rs slightly curved apically; length of the radial cell about 3 times the width; areolet indis- tinct; fore wing without distinct hair fringe on wing margins. Third abdominal tergum (Fig. 78) basally smooth without sculpture or pubescence; posterior third of third tergite and following seg- ments conspicuously punctate; ventral spine of hypopygium very short; tarsal claws simple. MALE, differs from female in antenna (Fig. 74) with 14 antennomeres; A3 slightly curved in the middle. Comments. — The complete series from the Giraud Collection, sent by the MNHP, comprises 63 individuals. In contrast with the specimens predominantely black described above, other in- dividuals of a lighter, chestnut-reddish shade and some completely reddish individuals are included in the series. The latter also exhibit other morpho- logical differences such as a stronger mesoscutal sculpture and less impressed notauli and medial mesoscutal impression. Also included in the series are some galls consisting of transformed achenes from flower heads of Centaurea scabiosa L. and in compliance with the handwritten labels from Giraud: "caput centaureae scabiosae" and the date, at least part of the specimens of the series should have been obtained from these galls. Giraud origi- nally described the species Diastrophus scabiosae as obtained from conspicuous galls in the stems of Centaurea scabiosa L. (Giraud 1859), but these galls are not included in the series. The specimens ob- tained from galls in achenes could belong to Isocolus rogenhoferi Wachtl. However, the status of the Volume 3, 1994 187 latter has not been clear, since only differences in the coloration of the antennal flagellum together with the different position of the galls have been pointed out in the literature (Kieffer 1901, Eady and Quinlan 1963) to separate it from /. scabiosae. For these reasons, I consider that there is only one species with a wide morphological variability be- ing capable of inducing galls on both stems and flower heads of species of Centaurea, mainly C. scabiosa L. It is possible that the two types of galls correspond to two different generations of the same species. In the same work, Giraud described another species, Diastrophus areolatus, from 2 fe- males caught by netting near Vienna. The differ- ences between the latter and I. scabiosae were stated to be reddish coloration of the antennal flagellum and presence of areolet in the fore-wings. In the examined series, I found 2 specimens which could correspond to D. areolatus labeled "captur 13 juill, Turksch" . The presence or absence of the areola, as well the general coloration, is quite variable in the whole of the series, so I consider D. areolatus as conspecific with J. scabiosae and, therefore, I estab- lish the corresponding new synonymy. Isocolus serratulae Mayr comb. n. Aulax serratulae Mayr 1882:9 Aylax serratulae (Mayr): Kieffer 1901:313 Type material. — Series composed of 31 speci- mens: 126", 19 9, most of them set on micro-pins (NHMW). 1 9 , with handwritten labels "serratulae ydetRogenhofer" and "Luxenburg,Rogenh 1881", hereby designated lectotype. Other specimens carry handwritten labels "Aid. serratulae det. G. Mayr" and "Serratula heterophylla"; designated paralectotypes. Redescription.— (Figs. 80-87). FEMALE. Length 2.5 mm. Black; gaster red-brown to dark-brown; antenna and legs light brown; coxa, base of femora and two first segments of antenna darker. Head viewed dorsally (Fig. 80) about 2 times broader than long; viewed frontally (Fig. 81) 1.2 times broader than high; temples not expanded behind the eyes; frons and vertex with sculpture minutely coriaceous-reticulate; POL:OOLas 11:9; OOL about three times the greatest diameter of lateral ocellus transfacial line around 1 .5 times the height of eye malar space around 0.6 times the height of eye, face laterally with radiating striae; supraclypeal area slightly raised, without striae; antenna (Fig. 82) with 13 antennomeres, A3 clearly shorter than A4. Pronotum (Fig. 83) reticulate-coriaceous, lat- erally pubescent; dorsally long, with distinct submedial pits ; mesoscutum (Fig. 84) with reticu- late-coriaceous sculpture; notauli complete, well impressed posteriorly; median mesoscutal impres- sion distinct in posterior 1/5 of mesoscutum; scutel- lar foveae large, subtriangle-shaped; not smooth; separated by a posteriorly broader septum; scutel- lum reticulate-coriaceous, with longitudinal fur- row variably impressed; mesopleuron (Fig. 85) longitudinally striated. Wings hyaline, slightly longer than the body; Rl and Rs of forewing not quite reaching to wing margin and radial cell open (Fig. 86); length of the radial cell about 2.6 times the width; areolet indistinct; fore wing without distinct hair fringe on wing margins. Third ab- dominal tergum (Fig. 85) basally with a lateral pubescent patch; posterior third of third tergite and following segments conspicuously punctate; ventral spine of hypopygium very short; tarsal claws simple. MALE. Differs from female in an- tenna with 14 antennomeres (Fig. 87), A3 not modified only slightly curved. Comments. — This species was described by Mayr from material collected by Rogenhofer on Serratula heterophylla auct ross non (L.) Desf. The gall is probably produced on this plant, but it has not been described. Since the description, the spe- cies has been recorded once from Ukrania, includ- ing galls Zerova et al. (1988), so it is only known from Austria and that country. The features of this species fit well in the genus Isocolus, except for the third abdominal tergum with basal pubescent patch, so I transfer it to this genus. Isocolus lichtensteini (Mayr) comb. n. Aulax lichtensteini Mayr 1882:7 Aylax lichtensteini (Mayr): Kieffer 1901:297 Isocolus lavaresi Nieves-Aldrey 1984:243. Syn. n. Type material. — Type series composed of 2 9 ; 19, mounted on micro-pin, with handwritten la- bels " Aulax lichtensteini, det. G. Mayr" and "Cen- taur, gall lich " designated lectotype. Paralectotype, one female with label "Centaur, salmant. gall Lich". Comments. — This species was described from material collected in Montpellier (France), pro- ducing galls in the stems of Centaurea salmantica L. 188 Journal of Hymenoptera Research (=Microlonchus salmanticus) (Mayr, 1882). Isocolus tavaresi Nieves-Aldrey, described from material reared from galls on Centaurea aspera L. in Spain (Nieves-Aldrey, 1984), later also found on Centau- rea melitensis L., (Nieves-Aldrey, unpublished), is conspecific with the two type specimens studied and, therefore, I establish the corresponding syn- onymy. The type specimens differ from individuals collected in Spain (Nieves-Aldrey 1984) only in their lighter chestnut-reddish coloration, also ex- tended for antenna and legs, including the coxae. Therefore, the types exhibit only a slight colour contrast between the two first antennal segments and the flagellum, and the longitudinal furrow of scutellum is more pronounced compared to the Spanish specimens collected. LIPOSTHENES Forster Liposthenus (!) Forster 1869:332. Liposthenes Forster 1869:336. Type species: Aulax glechomae Hartig. Monotvpic and orig. desig. (examined). The types of Aulax glechomae Hartig, in the ZSM, consist of 4 9 , mounted on pins. One of them carry label of lectotype ? designated by Weld in 1931 (Weld 1952). Diagnosis. — Head, viewed dorsally, slightly less than two times broader than long; in frontal view slightly broader than high; transfacial line 1.5 times height of eye; malar space shorter than height of eye; frons and vertex finely reticulate; radiating striae incomplete, weakly impressed; antenna of female with 13 segments, A3 slightly longer than A4; male antenna 14-segmented. Pronotum dorsally long, laterally strongly pubes- cent and with some longitudinal striae or rugae; submedial pits conspicuous, asociated with a dis- tinct pronotal plate; mesoscurum regularly coria- ceous and shining; notauli well impressed and complete; median mesoscutal impression very short ; scutellar foveae deep, rounded, smooth and shining; mesopleuron finely and clearly striated longitudinally . Wings hyaline; radial cell of forew- ing open on front margin; length of radial cell slightly less than three times width. Third ab- dominal tergum with conspicuous lateral pubes- cent patch; following segments not punctate. Tar- sal claws simple. Comments. — The genus is morphologically similar to Isocolus Forster, but can be separated by the pronotum being strongly pubescent laterally; presence of wing fringe on distal margin; third antennal segment being longer than fourth; third abdominal tergum having a lateral pubescent patch and in general sculpture and biology. Distribution and Biology. — The genus com- prises a single known species from Western Eu- rope which causes the formation of galls in the runners, stems and leaves of Nepeta (=Glechoma) hederacea L. (Lamiaceae). NEAYLAX gen. n. Type species: Aylax salviae (Giraud 1859):369. By present designation, (examined) Derivation: from Neos and Aylax (new Aylax) Description. — Head, viewed dorsally, two times broader than long; temples not expanded behind the eyes; radiating striae on face well im- pressed laterally; supraclypeal area raised, with- out striae; frons and vertex with coriaceous or alutaceous sculpture; antenna 12-13- segmented in females; 14-15-segmented in males; A3 as long as A4 or shorter than A4. Pronotum dorsally long; submedial pits clearly visible, not broadly sepa- rated; mesoscutum coriaceous or alutaceous, some- times punctate; notauli complete or faint anteri- orly; scutellar foveae large, rounded; mesopleuron longitudinaly striated. Wings hyaline; radial cell of forewing open on front margin but Rj reaching margin; distal wing margins with hair fringe mod- erately long. Third abdominal tergum basally smooth or with a lateral pubescent patch; follow- ing segments finely punctate but punctures not conspicuous; ventral spine of hypopygium short. Legs with tarsal claws simple. Comments. — Neai/lax is erected to comprise some species formerly included in Aylax Htg, but which do not fit well in that genus. The new proposed genus is closely related to Isocolus Forster, from which it differs as follows: radial cell of forewing not completely open along the margin and at the base; first abscissa of radius reaching the wing margin and presence of hair fringe on the distal margin of the wing (absent in Isocolus). Biologically Neai/lax differs from Isocolus in its association with Lamiaceae rather than with Asteraceae. The dorsally long pronotum, the rounded scutellar foveae, and the association with Lamiaceae and not with Papaveraceae separate it from Aylax Htg. Volume 3, 1 994 189 Distribution and Biology. — The genus com- prises two species distributed in centre and south of Europe which produce galls on species of Salvia (Lamiaceae) and a third species, only recorded from the Iberian Peninsula whose host plant and gall are unknown. REVISED SPECIES Neaylax salviae (Giraud), comb. n. Aulax salviae Giraud 1859:369 hocolus salviae (Giraud): Nieves-Aldrey 1988:224. Syn. n. Type material. — The type series of A. salviae, in the NHMW, sent to me by M. Fischer, consists of 12 specimens. 1 6 , mounted on a rectangular card with labels: red "Type" and white handwritten "Aulax salviae Giraud", is designated lectotype. 1 8 , 1 9 , mounted on micro-pin, with labels "Col- lect. Mayr" and "Aulax salviae Gir, det. G. Mayr" clearly do not belong to the type series, although they are conspecific with A. salviae Giraud. In the same series 16, 29, collected by Mayr, deter- mined as Aulax salviae Giraud, are cynipids but do not correspond to that species. Finally 2 speci- mens, identified as A. salviaeby Fahringer, belong to the family Eucoilidae. Redescription.— MALE. (Figs. 96-103). Length 1 .8 mm. Red-brown; antenna red-brown; legs with coxae red-brown, tibiae and tarsi paler. Head viewed dorsally (Fig. 96) about 2 times broader than long; temples not expanded behind the eyes; frons and vertex with sculpture coriaceous- alutaceous; POL as long as OOL; OOL about 4 times the greatest diameter of lateral ocellus; face laterally with radiating striae; supraclypeal area almost raised, without striae; antenna (Fig. 98) with 14 antennomeres, A3 clearly shorter than A4. Pronotum (Fig. 100) coriaceous; dorsally long, with distinct submedial pits; mesoscutum (Fig. 101) finely alutaceous and with some obsolete punctures medially; notauli relatively expanded posteriorly, complete but slightly faint anteriorly; median mesoscutal impression indistinct in poste- rior 1/4 of mesoscutum; scutellar foveae large, rounded, smooth and shining, separated by a narrow septum; scutellum weakly rugulose, with a weak longitudinal furrow; mesopleuron (Fig. 102) longitudinally striated. Wings hyaline; radial cell of forewing (Fig. 103) open on front margin but Rj reaching margin; length of radial cell about three times the width; hair fringe on distal margin of wing long. FEMALE, differs from male as fol- lows: antenna with 13 antennomeres (Fig. 98); third abdominal tergum without lateral pubes- cent patch; following segments with fine and in- conspicuous punctures; ventral spine of hypopygium short. Distribution and Biology. — The species was described from specimens collected in Dalmatia, reared from galls in the fruits of Salvia officinalis L. Species of apparently Eastern Mediterranean dis- tribution. SPECIES TRANSFERRED TO NEAYLAX GEN. N. Neaylax verbenacus (Nieves-Aldrey), comb. n. Isocolus verbe>\acus Nieves-Aldrey 1988:221. Syn. n. This species is closely related to N. salviae (Giraud). It was described within Isocolus from Spain. Its galls are produced in fruits of Salvia verbenaca L. Neaylax versicolor (Nieves-Aldrey), comb. n. Aylax versicolor Nieves-Aldrey 1985:122. Syn. n. This species was described under Aylax (Nieves-Aldrey 1985). It is known only from Spain. Gall unknown. PANTELIELLA Kieffer Pantelia Kieffer 1901:248 (non Bolivar 1887, Orthopf.) Panteliella Kieffer 1901:324. Type species: Aulax fedtschenkoi (Rubsaamen 1896). Monobasic Endocaulonia Ionescu and Roman 1960:222. Type species: Endocaulonia bicolor Ionescu and Roman 1960. By origi- nal designation. Weld (1952), stated that the types, consisting of slide-mounts of two adults in balsam, were in the Berlin Museum. Later on, Quinlan (1968) stated that the type material consisted of two specimens deposited in the Vienna Museum but Dr. Fischer informed me that the only material of A. fedtschenkoi housed in that museum consisted in galls. I have had the chance to examine one of the preparations of the Berlin Museum, consisting of a head in bad condition, with one of the two antennae being 190 Journal of Hymenoptera Research complete, a leg and the gaster. The preparation has a red type label and is printed "Aulaxfedtschenkoi; Bijuk Onlar, juni 1893; Blattgallen on Phlomis tuberosa". The type species was redescribed by Weld (1930) and by Quinlan (1968). Endocanlonia was described by Ionescu and Roman in 1960 based on E. bicolor, a species discov- ered by these authors in Rumania producing galls in the stems of Phlomis tuberosa L. Although I have not been able to examine the types, it seems clear from the detailed description that this genus is very similar to Panteliella Kieffer only differing slightly in the relative lengths of pedicel and third antennal segment. The descriptions of the two genera, except for the mentioned feature, are fully coincident, as is the gall and the host plant, which makes me consider that Endocaulonia Ionescu and Roman is synonymous with Panteliella Kieffer. The type species of Endocaulonia appears to be synonymous with Panteliella fedtschenkoi, although in this case I do not venture to formally establish the synonymy without a previous examination of the type. I present the following diagnosis of the genus based on Quinlan's redescription (1968). Diagnosis. — Female antenna 14-segmented; male 15-segmented; A3 and A4 of same length in the two sexes. Mesoscurum reticulate; notauli and median mesoscutal impression very faintly indi- cated; scutellar foveae distinct separated by a broad and striate septum; mesopleuron longitudinally striated. Radial cell of forewing open on the mar- gin; wing fringe on distal margin distinct. Third abdominal tergum without lateral pubescent patch; following tergites weakly punctate on lateral mar- gins. Tarsal claws simple. The genus is close to Neaylax Nieves-Aldrey and to Rhodus Quinlan. It can be separated from the former by the obsolete notauli, scutellar foveae broadly separated and female antenna 14-seg- mented. From Rhodus it is distinguished by the distinct fringe on distal margin of forewing, the obsolete notauli and A3 not being shorter than A4. Distribution and Biology. — Only one known species: Panteliella fedtschenkoi (Rubsaamen, 1896) distributed in central and eastern Europe and eastern Mediterranean. It produces galls in the leaves and stems of Phlomis tuberosa L. (Lamiaceae). Panteliella bicolor (Ionescu and Roman) is posibly a second species of the genus, only known from Rumania. PHANACIS Forster Plmnacis Forster 1860:145. Type species: Plianacis centaureae Forster 1860. Monobasic. Gillettea Ashmead, 1897:69. Type species: Gillettea taraxaci Ashmead 1897. Original designation. Weld (1952) pointed out that Forster's types in the Berlin Museum were destroyed by dermestids but he could observe good specimens of P. centau- rea in Vienna. However, I have abundant Iberian material of the type species at my disposal from which I have produced the diagnosis of the genus. Gillettea Ashmead was synonymized with Phanacis Forster by Eady and Quinlan (1963). I have examined the type species of Gillettea and I confirm the synonymy. Diagnosis. — Head, viewed dorsally two times broader than long; viewed frontally slightly broader than high; frons and vertex reticulate- coriaceous; radiating striae of face weakly im- pressed, incomplete; female antenna 13-seg- mented; male antenna with 14 antennomeres; A3 clearly longer than A4. Pronotum dorsally moder- ately long, without distinct submedial pits; mesoscurum reticulate-coriaceous; notauli usu- ally faint anteriorly; scutellar foveae transverse, confluent, not well defined or indistinctly closed posteriorly; mesopleuron reticulate. Wings fully developed in females, sometimes reduced in males; radial cell of forewing usually closed, sometimes incompletely on the margin; hair fringe on distal wing margins moderately long. Third abdominal tergum without lateral pubescent patch. Tarsal claws simple. Comments. — Phanacis is easily distinguished from other genera of Aylacini excepting Timaspis by the absence of pronotal submedian pits on the pronotum and the reticulate sculpture of the mesopleuron. The genus was considered by Eady and Quinlan (1963) to comprise also the species included in Timaspis Mayr. However, when some species of the latter genus not examined by Eady and Quinlan, are included in Phanacis, the hetero- geneity of the grouping becomes evident. This is the case with, e.g., Timaspis cichorii Kieffer and T. phoenixopodos Mayr. The boundary between the two genera does not, however, seem clearly de- fined and some species, particularly P. caulicola (Hedicke), P. hypochoeridis (Kieffer) and Timaspis lampsanae Perris, are not easily assigned to one or Volume 3, 1 994 191 the other genus. All the known species of Phanacis (s. str.) produce galls on plants of the family Asteraceae. Distribution and Biology. — The genus has a palaearctic distribution with two species intro- duced in North America. All the known species are associated with Asteraceae. Galls are produced mostly on stems. Phanacis taraxaci (Ashmead) Gillettea taraxaa Ashmead 1897:69. Phanacis taraxaci (Ashmead): Eady and Quinlan, 1963:18. Examined material. — The type material con- sists of two female specimens; one in the USNM and the other in the NHMW (both examined); the specimen from Wien, in better condition, is desig- nated lectotype. The species is being studied in detail by Shorthouse (personal communication). Appar- ently the species has been introduced in North America, together with the host plant Taraxacum, from Europe. But the existence of the species in Europe is documented only by a few gall records, so the european origin of this species does not seem clear. The species is widely distributed and relatively abundant in North America (Shorthouse, pers. comm.). RHODUS Quinlan Rhodus Quinlan 1968:282. Type species: Rhodus oriundus Quinlan 1968. Monotypic and original designation (ex- amined) Diagnosis. — Figs. 104-110. Head viewed dor- sally (Fig. 104) two times broader than long; viewed frontally (Fig. 105) clearly broader than high; f rons and vertex with reticulate sculpture; radiating striae of face present laterally; supraclypeal area raised; female antenna (Fig. 106) 13-segmented; male with 14 antennomeres; A3 clearly shorter than A4. Pronotum (Fig. 107) dorsally long, with distinct submedial pits; mesoscutum (Fig. 108) with dull reticulate sculpture and some scattered piliferous punctures; notauli distinct in posterior 3/4 of mesoscutum; faint or obsolete anteriorly; scutellar foveae large, rounded, sculptured; sep- tum relatively broad; mesopleuron (Fig. 109) lon- gitudinally striated. Rl and Rs not quite reaching anterior margin of wing; radial cell of forewing (Fig. 110) open; hair fringe on distal wing margins absent. Third abdominal tergum without lateral pubescent patch. Tarsal claws simple. Comments. — The genus was based on one spe- cies, Rhodus oriundus Quinlan, found in Rhodes Island, galling apical buds of Phlomis cretica Presl (Quinlan 1968). Its morphological features relate it to the genera Isocohis Forster and Neaylax Nieves- Aldrey; it differs from the former in the conforma- tion of the notauli, the sculpture of the mesoscutum and the distal margin of the wing completely lacking hairs. The wing venation with Rl and Rs not quite reaching the anterior margin of wing and the absent hair fringe on distal wing margin sepa- rate it from Neaylax. Distribution and Biologi/. — The genus comprises one described species, only recorded from Rhodes (Greece) associated with Phlomis (Lamiaceae). TIMASPIS Mayr Timaspis Mayr 1881:18. Type species: Diastrophus lampsanae Karsh. Monotypic. (Type. — Timaspis phoenixoptodos Mayr 1882. Designated by Ashmead 1903 p.214). Phanacis Forster (part): Eady and Quinlan 1963:18. Aylacopsis Hedicke 1923:81. Syn. n. Type species: Aylacopsis heraclei Hedicke 1923. Monotypic and original designa- tion. The type material of Timaspis lampsanae, in the Museum of Vienna, consists of a 2 mounted on micro-pin on polyporous strip labeled "Tim. lampsanae det Karsch Type"; "lampsanae Karsch Type"; "collect G. Mayr" and red label "Holo- type". Diagnosis. — Radiating striae present laterally on the face; incomplete. Female antenna of 13-14 segments; 14-15 in males; A3 relatively long; at least 1.5 times longer than A4. Pronotum dorsally moderately long, without distinct submedial pits; notauli usually complete but often not well im- pressed anteriorly; median mesoscutal impres- sion usually present; scutellar foveae small, confluent, usually indistinctly closed posteriorly; mesopleuron reticulate-rugulose, rugulose- alutaceous or striate-alutaceous. Radial cell of forewing closed on the margin, sometimes incom- pletely so anteriorly; hair fringe on distal margins of wing long. Third abdominal tergum without lateral pubescent patch. Tarsal claws simple. Comments. — Timaspis can be distinguished from Phanacis on the basis of its rugulose-reticu- 192 Journal of Hymenoptera Research late sculpture of the mesopleuron and complete notauli. Timaspis was synonymyzed with Phanacis Forster by Eady and Quintan (1963) but these authors' concept of Timaspis was incomplete be- cause it was limited to the type species of the genus, T. lampsanae. But if other species described within Timaspis are considered, especially T. phoenixopodos Mayr, I consider it advisable to keep the genera separate. It must be emphasized, how- ever, that the boundary between the two genera is not clearly defined and some species are difficult to assign to one genus or the other. Aylacopsis Hedicke, was described from a spe- cies that produces galls in the stems of Heracleum sphondylum L. (Apiaceae) (Hedicke 1923). After having examined the type species, I consider that it fits well within Timaspis Mayr. Consequently, I synonymize Aylacopsis with Timaspis Mayr. Distribution and Biology. — As understood herein, the genus comprises 9 species in West- Europe. All the species, excepting one, are associ- ated with Asteraceae Timaspis heraclei (Hedicke), comb. n. Aylacopsis heraclei Hedicke 1923:81. Syn. n. The type of Aylacopsis heraclei Hedicke in the Berlin Museum consists of a 9 specimen glued on a card triangle. The specimen carries the following labels: white handwrit- ten "Doicul, Leipzig 26-2-15"; "Aylacopsis heraclei 9 Hedicke" and red printed "Type". We can contribute the following complemen- tary data to the original description: Figs. 111-118. Head, viewed dorsally (Fig. Ill), about two times broader than long; viewed frontally (Fig. 112) slightly broader than high; frons and vertex finely coriaceous; POL slightly less than two times OOL; transfacial line about 1.6 times height of eye; face laterally with radiating striae; supraclypeal area without striae; with weak median protuberance- malar space about 0.7 times height of eye; antenna (Fig. 113) slender, filiform, with 14 antennomeres; A3 slightly curved in the middle, about 1.7 times longer than A2; 1.3 times longer than A4. Prono- tum (Fig. 114) dorsally only moderately long; scarcely pubescent, without submedial pits; mesoscutum (Fig. 1 15) finely reticulate-coriaceous; notauli complete; median mesoscutal impression impressed in posterior 1 / 3 of mesoscutum; scutel- lar foveae confluent, indistinctly closed posteri- orly; scutellum ovate, weakly rugulose; mesopleuron (Fig. 116) rericulate-rugulose. Forew- ing with radial cell (Fig. 117) open anteriorly on margin but R, clearly reaching wing margin; length of radial cell 2.6 times width; hair fringe on wing margins long. Third abdominal tergum (Fig. 118) without lateral pubescent patch; following seg- ments not punctate; ventral spine of hypopygium quite short, slightly visible. Tarsal claws simple. Comments. — Within Timaspis Mayr this spe- cies is morphologically close to Timaspis cichorii from which it is differentiated by the coloration, sculpture of the mesoscutum and conformation of the scutellar foveae. According to the original description, this species produces galls in the stems of Heracleum sphondylum (Apiaceae) at that time being the first cynipid species demostrated to be associated with this family of plants. Afterwards this species has not been recorded again and, consequently, the host has not been confirmed. It may, however, be pointed out that Phanacis eryngii Diakonchuk, another Aylacini species, was re- cently found in Eastern Europe associated with Apiaceae (En/ngium) (Diakonchuk 1984). Timaspis lusitanica Tavares Timaspis lusitanicus Tavares 1904:301 Timaspis lusitanica Tavares 1904: D.T. and Kieffer 1910 Emend. Phanacis crepidos Weidner 1965:1. Syn. n. Timaspis lusitanica Tavares was described from Portugal (Tavares 1904) reared from galls on Barkausia taraxacifolia DC. (Asteraceae). More re- cently, it has been recorded from Spain, on the same host plant (Nieves-Aldrey 1992). Phanacis crepidos was described from Germany (Weidner 1965) from material reared from galls on Crepis biennis L. (Asteraceae). Having examined the type material from the Museum of Hamburg, I have concluded that it is the same species described by Tavares, so I establish the corresponding new synonymy. Volume 3, 1 994 193 XESTOPHANES Forster LITERATURE CITED Xestophanes Forster 1869:332, 337. Type species: Cynips potentillae De Villers. Monotypic and original designa- tion. Diagnosis. — Face laterally with radiating striae; supraclypeal area without striae; with weak me- dian protuberance; frons and vertex almost smooth and shining; antenna of female 13-segmented, A3 as long as or slightly longer than A4; antenna of male 14 segmented. Pronorum dorsally long, lat- erally pubescent; submedial pits conspicuous; mesoscutum smooth and shining; notauli com- plete or faint anteriorly; median mesoscutal im- pression slightly indicated; scutellar foveae smooth; mesopleuron smooth and shining. Forew- ing with radial cell open on the margin but R, reaching to wing margin; hair fringe moderately long. Third abdominal tergum laterally smooth and glabrous. Tarsal claws with weakly devel- oped basal lobe or tooth . Comments. — The genus, together with the al- lied genus Diastrophus, form a distinct unit within the Aylacini: morphologically, by the tarsal claws having an acute basal lobe or tooth and frons, vertex and mesopleuron without sculpture; bio- logically, by their association exclusively with Rosaceae (the genera Potentilla and Rubus). The genus is very likely monophyletic, based on the fusion of third and fourth abdominal terga in the females Distribution and Biology. — Two species of Xestophanes are known in Western Europe, both gall makers on stems and runners of species of Potentilla. . ACKNOWLEDGMENTS I am very grateful to the curators of the collections listed under materials for the loan of type material and other specimens. Special thanks to A. Menke (USNM) and R. Danielsson (MZLU) also for sending some interesting bibli- ography. I am also most indebted to one anonymous re- viewer and specially to F. Ronquist for a critical revision of the manuscript including linguistic suggestions. M. A. Alonso- Zarazaga helped with nomenclatonal questions. I would like to thank Inaqui Diez for the rotulation of the drawings made by myself. Financial support was provided by the Research project "Fauna Ibenca II", DGICY PB89 0081. Ashmead, W. H. 1885. A Bibliographical and Synonymical Catalogue of the North American Cynipidae, with de- scription of new species. Transactions of the American Entomological Society 12: 291-304. Ashmead, W. H. 1887. On the Cynipidous galls of Florida, with descriptions of new species and synopses of the described species of North America. Transactions of the American Entomological Society 14: 125-158. Ashmead, W. H. 1896. Description of new cynipidous galls and gall wasps in the United States National Museum. Proceedings of the United States National Museum 19: 113- 136. Ashmead, W. H. 1897a. Description of some new genera in the family Cynipidae. Psyche 8: 67-69. Ashmead, W. H. 1897b. Description of five new genera in the family Cynipidae. Canadian Entomologist 29: 260-263. Ashmead, W. H. 1903. Classification of the gall-wasps and the parasitic cynipoids, or the super-family Cynipoidea. I-IV. Psyche 10: 140-155 Askew, R. R. 1984. The biology of gall wasps, pp. 223-271 . In T. N. Ananthakrishnan (ed.), Biology of Gall Insects. Ed- ward Arnold, London. Barbotin, F. 1964. Sur une nouvelle galle et deux cynipides en provenance d'Algerie. Marcellia 31: 151-157. Belizin, V. I. 1959. Gall wasps of the tribe Aylaxim (Hy- menoptera, Cynipoidea) new for the fauna of the USSR. Entomologkheskoe Obozreme 38: 662-674. Burks, B. D. 1979. Cynipoidea, pp. 1045-1107. In K. V. Krombein, P. D. Hurd Jr., D. R. Smith, and B. D. Burks (eds), Catalog of Hymenoptera in North America North of Mexico. Volume 1, Smithsonian Institution Press, Washington, DC. Dalla Torre, C. W. and J. Kieffer. 1910. Cynipidae. Das Tierreich 24: 1-891. Diakonchuk, L. A. 1984. New species of Cynipidae (Hy- menoptera) from Georgian SSR. Vestnik Zoology 3: 74-77 (In Russian). Eady, R. D. and J. Quinlan. 1963 Hymenoptera: Cynipoidea. Key to families and subfamilies and Cynipinae (includ- ing galls). Handbooks for the Identification ofBritish,lnsects 8(la): 1-81. Folhot, R. 1964. Contribution a l'etude de la biologie des cynipides gallicoles (Hymenopteres, Cynipoidea). Annates des Sciences Naturelles, Zoologie, Paris. 12 ser. 6: 407-564. Forster, A. 1860. Die zweite Centune neuer Hymenoptera (Eucoela et Phanacis). V erhandlungen derNaturhistorischen Vereins der Preussischen Rhemlande und Westphalens 17: 143-146. Forster, A. 1869. Ueber die Gallwespen. V erhandlungen der kaiserlich-koemglichen zoologisch-botamschen Gesellschafi in Wien 19: 327-370. Giraud, J. 1859. Signalements de quelques especes nouvelles de cynipides et de leurs galles. Verhandlungen der kaiserlich-koemglichen zoologisch-botanischen Gesellschaft in Wien 9: 337-374. Hartig, T. 1840. Uber die familie der Gallwespen. Germar Enlomologische Zeitung 3: 322-358. Hartig, T. 1843. Zweiter nachtrag zur naturgeschichte der 194 Journal of Hymenoptera Research Gallwespen. Zeitschrift fur Entomologie 4: 395-422. Hedicke, H. 1923. Beitrage zur kenntnis dercvnipiden (Hym.) XII. Ein neues cecidozoon an Heracleum spondylium. Zeitschrift fur Pflanzenkrankheiten 33(1-2): 81-83. Hedicke, H. 1928. Beitrage zur kenntnis der cynipiden (Hym.) XIV. Der Erzeuger der levantimschen salviagalle Deutsche Entomologische Zeitschrift 14: 81-85. Ionescu, M. A. and N. Roman. 1960. Un genre nouveau de Cynipides gallicoles Endocauloma bicolor n g., n. sp. (Hymenoptera, Cynipoidea). Revue de Bwlogie.Acadenuci Republicu Populare Ronune Bucaresti 5(3): 221-225. Ionescu, M. A. and N. Roman. 1962. New genera of gall maker Cynipidae in the fauna of the rumanian people's repub- lic: Weldiella aequalis n. gen. and Cecconia valerianeliae Thomson. Revue de Biologie Academia Republicii Populare Romine Bucaresti 7: 551-559. KiefferJ.J. 1901 . Monographic desCynipides d' Europe et d' Algene Hermann edit. Paris, t.I, 687 pp. Kieffer, J. J. 1902. Revision du genre Aulax et des genres limitrophes d'Aulax, avec quelques notes sur divers autres Cynipides. Bulletin de la Societe d Histoire N aturelle deMetz. 10:91-97. Kinsey, A.C. 1920. Phylogeny of cynipid genera and biologi- cal characteristics. Bulletin of the American Museum of Natural History 62: 357-402. Kovalev, O. B. 1982. Cynipid gall-makers of the subfamily Aylacinae (stat. n.) (Hym., Cynipidae) and their species described within the family Figitidae. Trudy Zoologicheskogo Instituta Akademiya Nauk SSR 110:85-93. (in russian) Mayr, G. 1881. Die Genera der gallenbewohnenden Cynipiden. 20, fahresbenchte der Communal-Oberrealschule im l. Bezirke pp. 1-38. Mayr, G. 1882. Die europaischen arten der gallen- bewohnenden cynipiden. 21, fahresberichte der Commu- nal-Oberrealschule im I. Bezirke pp. 1-44. Nieves-Aldrey, J. L. 1984. Notas sobre los Aylaxini (Hvm, Cynipidae, Cynipinae) de la Peninsula Iberica con descripcion de una nueva especie de Isocolus. Eos 60: 235-250. Nieves-Aldrey, J. L. 1985. Nuevos Aylaxini (Hym., Cynipidae) para la Peninsula Iberica con descripcion de una nueva especie de Aylax Htg. Boletim da Sociedade Porttuguesa de entomologia, suppl. 1, Adas do II Congresso Iberico de Entomologia 117-128. Nieves-Aldrey, J. L. 1987. Estado actual de conocimiento de la subfamiha Cynipinae (Hym.. Parasitica, Cynipidae) en la Peninsula Iberica. Eos 63: 179-195. Nieves-Aldrey, J. L. 1988. Descripcion de una nueva especie de Isocolus Forster con notas de otras especies de Aylaxini nuevas para la Peninsula Iberica (Hym., Cynipidae). Eos 64: 221-227. Nieves-Aldrey, J. L. 1992. Adiciones a la lista de cinipidos ibericos, especialmente Aylaxini, incluyendo la descripcion de una nueva especie de/lH/aniifa Ashmead (Hymenoptera, Cynipidae). Boletim da Sociedade portuguesa de entomologia, suppl n°3, Adas do II Congresso Iberico de Entomologia v. 1: 493-500. Nieves-Aldrey, J. L. "In press". The types of Aylax Hartig described by C.G.Thomson (Hymenoptera, Cynipidae). Entotnologica Scandmavica. Nordlander, G. 1984. Vad vet viom parasitiska Cynipoidea?. Entomologisk Tidsknft 105: 36-40 Quinlan, J. 1968. Cynipinae (Hymenoptera) occurring on Phlomis L. Transactions of the Royal Entomological Society of London 120:275-286. Ritchie, A. J. 1984. A revision of the higher classification of the mquiline gall-wasps (Hymenoptera: Cynipidae) and a revi- sion of the Neartic species of Periclistus Forster. Ph. D. Thesis, Carleton University, Ottawa, ON. 368 pp. Rohwer, S. A. and M M. Fagan. 1917. The type-species of the genera of the Cynipoidea, or the gall wasps and para- sitic cynipoids. Proceedings of the United States National Museum, 53: 357-378. Ronquist, F. In press. Evolution of parasitism among closely related species: phylogenetic relationships and the ori- gin of inquilinism in gall wasps (Hymenoptera, Cynipidae). Evolution. Ronquist, F. and G Nordlander 1989. Skeletal morphologv of an archaic cympoid, Ibalia rufipes (Hymenoptera: Ibaliidae). Entomological scandmavica Suppl. 33: 1-60. Tavares, J. da S. 1904. Descnpcao de un Cynipide novo. Broteria 3: 301-302. Thomson, C. G. 1877. Oefversigt af sveriges Cynips Arter. Opuscula Entomologica_8: 732-841. Wachtl, F. 1880. Beitrage zur Kenntnis der gallenerz. Insecten Europas. Verhandlungen der kaiserlich-koeniglichen zoologisch-botamschen Cesellscliaft in Wien 30: 531-545. Wachtl, F. 1891. Eine neue Gallwespe. Wiener Entomologische Zeitung pp. 277-280. Walsh, B. D. 1869. Galls and their architects. The American Entomologist 2: 70-74. Weidner, H. 1965. Phanacis crepuios sp. nov., eine neue gallwespe an Crepis biennis L. Entomologische Mttteilungen 52: 1-7. Weidner, H. 1968. Zur Kenntnis der Gallwespentnbus Aulacini (Hymenoptera, Cynipidae). Entomologische Zeitschrift 10: 105-120. Weld.L. H. 1930. Notes on Types (Hymenoptera: Cynipidae). Proceedings of the Entomological Society of Washington 32: 8. Weld, L. H. 1952. Cynipoidea (Hym.) 1905-50. Privately printed, Ann Arbor, MI, 351 pp. Wiebes-Rijks, A. A. 1979. A character analysis of the species of Synergus Hartig, Section II (Mayr, 1872) (Hy- menoptera, Cynipidae). Zoologische mededelingen 53(28): 297-321. Zerova, M. D., Diakonchuck, L. A. and V. M. Ermolenko. 1988. Gall-formers of the european part of the USSR Part 2 . Hymenoptera. Naukova Dumka, Kiev. 180 pp. Volume 3, 1 994 195 APPENDIX Check list of the west-European genera and species of Aylacini including their host plant genera. AYLACINI Ashmead 1903 Host plant genus AULACIDEA Ashmead, 1897 abdominalis (Thomson, 1877) ? arnicae Hoffmeyer, 1930 Arnica follioti Barbotin, 1972 Sonchus hieracii (Bouche 1834) Hieracium artenusiae (Thomson, 1877) crassinervis (Thomson, 1877) foveiger (Thomson, 1877) kerneri (Wachtl, 1891) comb, n Nepeta kieffen ' Cotte, 1915 laurae Nieves-Aldrey, 1992 ? nibletti Quinlan, 1969 Hieracium pilosellae (Kieffer, 1901) Hieracium scorzonerae (Giraud, 1859) Scorzonera subterminalis Niblett, 1946 Hieracium tragopogonis (Thomson, 1877) Tragopogon tavaresi Nieves-Aldrey, 1984 syn. n. scabiosae (Giraud, 1859) Centaurea rogenhoferi Wachtl, 1880 syn. n. serratulae (Mayr, 1882) comb, n Serratula . Nepeta LIPOSTHENES Forster glechomae (Linnaeus, 1758)... latreilki (Kieffer, 1898) NEAYMXgen.n. salviae (Giraud, 1859) comb, n Salvia verbenacus (Nieves-Aldrey, 1988) comb, n Salvia versicolor (Nieves-Aldrey, 1985) comb, n ? PANTELIELLA Kieffer, 1902 ENDOCAULONIA Ionescu and Roman, 1960 syn. n. fedtschenkoi (Rubsaamen, 1896) Phlomis bicolor (Ionescu and Roman, 1960) Phlomis AYLAX Hartig, 1840 hypecoi Trotter, 1912 Hypecoum minor Hartig, 1840 Papavcr papaveris (Perris, 1839) Papaver artenusiae Weidner, 1968 misident. BARBOT1N1A gen. n. oramensis (Barbotin, 1964) comb, n Papaver CECCON1A Kieffer, 1902 WELD/ELM Ionescu and Roman, 1962 syn. n. valerianellae (Thomson, 1877) Valenanella aequalis (Ionescu and Roman, 1962) Valenanella DIASTROPHUS Hartig, 1840 rubi (Bouche, 1834) Rubus mayri Remhart, 1876 Potentilla HED1CK1ANA gen. n. levantina (Hedicke, 1928) comb, n Salvia IRAELLA gen. n. luteipes (Thomson, 1877) Papaver papaveris Kieffer in Goury and Gignon, 1905 1SOCOLUS Forster, 1869 fitchi (Kieffer, 1898) Centaurea jaceae (Schenck, 1863) Centaurea lichtensteini (Mayr, 1882) comb, n Centaurea PHANACIS Forster, 1860 GILLETTEA Ashmead, 1897 caulicola (Hedicke, 1939) Picris centaureae Forster, 1860 Centaurea punctipleuris (Thomson, 1877) hypochoendts (Kieffer, 1887) Hypochoens taraxaci (Ashmead, 1897) Taraxacum RHODUS Quinlan, 1968 onundus Quinlan, 1968 ... . Phlomis T1MASP1S Mayr, 1881 AYLACOPSIS Hedicke, 1923 syn. n. cichorn Kieffer, 1909 Cichonum heraclei (Hedicke, 1923) comb, n Heracleum lampsanae Perris, 1873 Lampsana lusitanica Tavares, 1904 Crepis crepidos (Weidner,1968) syn. n. phoenixopodos Mayr, 1882 Lactuca pilicornis (Thomson, 1877) ? rufipes Ionescu and Roman, 1959 Crepis sonchi (Steffarti, 1900) Sonchus urospermi Kieffer, 1901 Urospermum XESTOPHANES Forster, 1869 brevitarsis (Thomson, 1877) Potentilla potentillae (Retzius in Degeer, 1873) Potentilla foveicollis (Thomson, 1877) abreviatus (Thomson, 1877) 196 Journal of Hymenoptera Research Figs. 1-15. Cynipidae. 1-3, Pronotum, dorsal view. 1, Aylax papaveris, 2, Aulacidea hieracii. 3, Cynips divisa 9. 4, Scutellum, Pediaspis acens. 5, Gula, Penchstus brandtii. 6-7, Ceroptres 6, Gaster, lateral (9)7, Head, frontal view. 8, Gula, Isocolus. 9, Gaster, lateral, Synergus ilicinus (2). 10, Gaster, lateral, Cecconia valerianellae. (9). 11-13, Eschatocerus. 11, Mesoscutum and scutellum, dorsal view. 12, Forewing (hairs omitted). 13, Head, dorsal view. 14-15, Diplolepis rosae. 14, Mesopleuron. 15, Gaster, lateral (2). Volume 3, 1994 197 Figs. 16-34. Aylacini. 16-20, Tarsal claws. 16, Diastrophus rubi (hind). 17, Xestophanes potentillae (front). 18, X. potentillae (hind). 19. Barbotmia oraniensis (hind). 20, Isocolus Hchtensteini (hind). 21, Forewing, X. potentillae. 22, Forewing, D. rubi. 23-27, lraclla luteipes (9). 23, Pronotum. 24, Head and thorax, lateral. 25, Forewing. 26, Head, frontal view. 27, Antenna, first segments. 28, Mesopleuron, Phanacis caulicola. 29-30, Cecconia valerianellae 8. 29, Head, frontal view. 30, Forewing. 31-33, Liposthenus glechomae 9. 31, Head and thorax, lateral. 32, Mesoscutum and scutellum, dorsal view. 33, Antenna, first segments. 34, Forewing At/lax papaveris (9). 198 Journal of Hymenoptera Research Figs. 35-42. Antistrophuspisum. 35, Head, dorsal view. 36, Head, frontal view. 37, Antenna ( 9 ). 38, Antenna ( 6 ). 39, Pronotum. 40, Mesoscutum and scutellum. 41, Body, lateral. 42, Forewing (hairs omitted). Volume 3, 1994 199 Figs. 43-55. Aulacidea. 43-50, Aulacidea liarringtom. 43, Head, dorsal view. 44, Head, frontal view. 45, Antenna ( ? ). 46, Antenna (§).47, Pronotum.48, Mesoscutum and scutellum. 49, Body, lateral. 50, Forewing (hairs omitted). 51-55, Aulacidea kerneri. 51, Antenna (?). 52, Pronotum, 53, Mesoscutum and scutellum. 54, Body, lateral. 55, Radial cell of forewing (hairs omitted). 200 Journal of Hymenoptera Research Figs. 56-62. Barbotinia oraniensis. 56, Head, dorsal view. 57, Head, frontal view. 58, Antenna ( 9 ). 59, Antenna ( 6 ). 60, Pronotum. 61, Mesoscutum and scuteilum. 62, Body and wings. Volume 3, 1 994 201 %vj& Figs. 63-70. Hedickiana levantina. 63, Head, dorsal view. 64, Head, frontal view. 65, Antenna ( 9 ) and Scutellum. 68, Mesopleuron. 69, Forewing (hairs omitted). 70, Gaster, lateral. 66, Pronotum. 67, Mesoscutum 202 Journal of Hymenoptera Research Figs 71-79. hocolus scabwsae. 71, Head, dorsal view. 72, Head, frontal view. 73, Antenna (9). 74, Antenna (6). 75, Pronotum. 76, Mesoscutum and scutellum. 77, Mesopleuron. 78, Gaster, lateral. 79, Forewing (hairs omitted). Volume 3, 1 994 203 Figs. 80-95. Isocolus and Aulacidea. 80-87, Isocolusserratulae. 80, Head, dorsal view. 81, Head, frontal view. 82, Antenna (9). 83, Pronorum. 84, Mesoscutum and scutellum. 85, Body, lateral. 86, Forewing (hairs omitted). 87, Antenna {6). 88-95, Aulacidea scorzonerae. 88, Head, dorsal view. 89, Head, frontal view. 90, Antenna ( 9 ). 91, Pronorum. 92, Mesoscutum and scutellum. 93, Body, lateral. 94, Forewing (hairs omitted). 95, Antenna (6"). 204 Journal of Hymenoptera Research Figs. 96-103. Neaylax salviae. 96, Head, dorsal view. 97, Head, frontal view. 98, Antenna ( 9 ). 99, Antenna ( 9 ). 100, Pronotum. 101, Mesoscutum and scutellum. 102, Body, lateral. 103, Forewing (hairs omitted) Volume 3, 1994 205 105 104 j^Vi^? Figs. 104-110. Rhodus oriundus. 104, Head, dorsal view. 105, Head, frontal view. 106, Antenna (9). 107, Pronotum. 108, Mesoscurum and scutellum. 109, Body, lateral. 110, Forewing (hairs omitted). 20b Journal of Hymenoptera Research Figs. 111-118. Timaspis heraclei. Ill, Head, dorsal view. 112, Head, frontal view. 113, Antenna (9). 114, Pronotum, 115, Mesoscutum and scutellum. 116, Mesopleuron. 117, Forewing (hairs omitted). 118, Gaster, lateral. J. HYM. RES. Vol. 3, 1994, pp. 207-226 A Review of North American Belomicrus (Hymenoptera, Sphecidae, Crabroninae) Richard M. Bohart Department of Entomology, University of California, Davis, CA 95616 Abstract. — The North American Belomicrus are reorganized into five groups containing a total of 40 species. Of these, 14 are described as new: californicus (central Sierra of California), costalis (central California), darwini (sw. U.S.); desertus (s. California); myo (Inyo Co., California), longiceps (sw. U.S.), melanus, (central Sierra of California), montanus (central Sierra of California), oraibi (Arizona and s. Utah), palhdus (s. California), powelli (coast range mts. of California), siccatus (s. California), sierrae (central Sierra of California), texensis (Texas and Baja California Sur). Keys, illustrations, and distributions are given. New synonymy: quemaya Pate is raised to full species; and jurumpa Pate is synonymized under quemaya. Belomicrus Costa (1871) is one of several crabronine genera in the tribe Oxybelini. Bohart and Menke (1976:359-370) gave an overview of present day knowledge of the tribe. They pointed out that the presence of metanotal projections (squamae) and a propodeal projection (mucro) in addition to a lateral carina on terga I-III, and sometimes IV-V, characterize Belomicrus. As far as known, distribution of the genus is holarctic and Ethiopian. Bohart and Menke listed 63 species of which 25 were North American. I now recognize 40 species in the latter region, of which 14 are new. Among earlier writers the work of Pate (1940a,b) was outstanding. His overall assessment of Belomicrus and its relatives was brilliant, and his descriptions were meticulous. The only criticisms that can be made are that some of his phylogenetic discussions suffered from lack of material, and his descriptions contain much detail irrelevant at the species level. Belomicrus are ground-nesting wasps, mostly in sandy areas. Prey of the forbesii group are nymphal Miridae (Bohart 1956, Evans 1969). The franciscus group provisions with adult dasytid beetles (Williams 1936, Bohart and Menke 1976). Prey of the other groups are unknown. Members of the genus are few in most collec- tions. Through museum visits and borrowings I have been able to examine about 600 specimens. The majority of these are in the University of California at Riverside museum and were col- lected by P. H. Timberlake. Diligent work by Bohart Museum collectors at Davis has made some 3,000 more specimens available for study. Principal cooperating museums and individu- als are listed below. Museum designations are identified by their city locations in capital letters. The manuscript was read by Dr. Arnold Menke and Dr. Eric Grissell. Their comments and correc- tions were quite helpful, and their assistance is gratefully acknowledged. BERKELEY— J. Powell, P. D. Hurd (deceased), Essig Mu- seum, University of California at Berkeley. DAVIS — L. S. Kimsey, S. Heydon, Bohart Museum, Univer- sity of California at Davis. LOGAN — G. E. Bohart, Entomology Museum, Utah State University. NEW YORK — ]. Rozen, Entomology Museum, American Museum of Natural History. OTTAWA — L. Masner, Biological Resources Division, Agri- culture Canada. PHILADELPHIA— S. Roback, Daniel Otte, D. Azuma, Acad- emy of Natural Sciences. RIVERSIDE— P. H. Timberlake (deceased), S. Frommer, En- tomology, Museum, University of California at River- side. SAN FRANCISCO— W. J. Pulawski, P. Arnaud, Entomology Museum, California Academy of Sciences. VIENNA — M. Fischer, Naturhistorische Museum, Vienna, Austria. WASHINGTON— A. S. Menke, K. V. Krombein, U.S. Na- tional Museum. 208 Journal of Hymenoptera Research Terms used in the species key and in descrip- tions which may need explanation are: epipleural "button", small raised area just behind pronotal lobe; T-I, T-II, etc. terga after propodeum; S-I, S-II, etc., sterna after propodeum; LID, least interocular distance; PD, puncture diameter; post-tegula, basal wing sclerite. The generotype, Belomicrus italicus A. Costa (1871), has a palearctic distribution and is similar to species in the forbesii group. However, it should be placed in a different group because the clypeus has a considerable flattened area on the median lobe, and the more posteriorly blunt metanotal projection is divided by a median carina. The ten groups of North American Belomicrus which Pate (1940b) proposed are now reduced to five. Bohart and Menke (1976) considered only 2 groups, forbesii and cladothricis. With the discovery of additional species it now seems desirable to divide the forbesii group into four according to the following key. The purpose of this and other keys in the paper is to give a quick means of making identifi- cations. It is understood that the key user should have access to at least a representative named collection of Belomicrus for comparative purposes. KEY TO THE SPECIES GROUPS OF NORTH AMERICAN BELOMICRUS Metanotal projections (squamae) divided into two distinct lobes (Figs. 28-43), basally joined or separated but without an outer membranous area; body rarely more than 4mm long; postocular tubercles well defined; (Figs. 18-24) cladothricis group Metanotal projections joined to form a roughly triangular plate with a posterior notch and an outer partly membranous edge (Figs. 11-14), postocular tubercles various 2 Mandible of females and some males with an obtuse angle subapically on lower edge (Fig. 15),mesopleuron with an omaulus vanyume group Mandible with at most a slight inferior angle subapically, mesopleuron without omaulus 3 Clypeus with an anchorlike apical projection in female (Fig. 16), body about 2 mm long apache group Clypeus simply pointed or with a deflected apical bevel, body 3.5 mm long or more 4 Clypeus medially hoodlike, often somewhat wrinkled, not deflected apically but ending in a small point (Fig. 17) forbesii group Clypeus with an apicomedial polished triangle or deflected bevel (Figs. 3-7) franciscus group Belomicrus forbesii group The only previous key to the forbesii group is that given as part of a more extensive one by Pate (1940b:210). The forbesii part was based on about 65 specimens, of which some 43 were forbesii and columbianus, 10 were cookii, seven were penuti (as a subspecies of forbesii), 3 were coloratus, and two were querecho. I have studied about 1,500 speci- mens of the group. Instead of the six now valid species recog- nized by Pate, this figure has been raised to 10, which includes four new species. Members of the forbesii group have relatively few "structural" differences. However, the colora- tion and color patterns seem to be quite constant in long series. Therefore, I am treating columbianus as distinct from forbesii, sierrae as distinct frompenuti, and desertus as distinct from cookii. Details on differences are given in the key, which follows: Volume 3. 1994 209 KEY TO THE BELOMICRUS FORBESII SPECIES GROUP 1 T-I with basal slope all or mostly red 2 — T-I with basal slope dark reddish brown, dark brown, or black 5 2 Postocular tubercles weak and obtuse, tergal pale yellow bands usually well developed on T-I and often on T-II, weak or faint on T-III and following montanus R. Bohart — Postocular tubercles well developed, tergal markings various 3 3 S-II and following with distinct, close, moderately fine punctures; S-II with a broad median depression querecho Pate — S-II and following with indistinct, close, fine punctures or other microsculpture; S-II with at most a median flattened area 4 4 Terga without distinct pale yellow bands, these sometimes faintly visible; band across pronotum nearly always broken a little inside pronotal lobe, wing membrane stained powelli R. Bohart Terga with distinct pale yellow bands on most terga, band across pronotum complete to pronotal lobes, wing membrane nearly clear desertus R. Bohart 5 Epipleural "button" cup shaped, cup usually with a membranous outer edge, or pronotal collar all black 6 — Epipleural "button" not cup shaped, only slightly indented, pronotal collar nearly always maculate 7 6 Pronotal collar all black, T-I to IV all dark or with indistinct bands, epipleural "button" pocket without a lighter colored edge, postocular tubercles weak sierrae R. Bohart — Pronotal collar maculate, terga with pale bands or lateral spots usually on T-I to V, epipleural "button" pocket deeper and with a lighter colored edge, postocular tubercles often well developed penuti Pate 7 Tergal markings weak, pale bands indistinct at least on T-III to V; pronotal collar black laterally and sometimes medially columbianus (Kohl) Tergal markings prominent; pronotal collar nearly always completely banded 8 8 T-I, II and sometimes III in female with dark ground color, following terga with red ground color; both sexes with small but sharp postocular tubercles cookii Baker T-I and most following terga in both sexes with reddish brown to black ground color; both sexes with at most weakly to moderately developed, but not sharp, postocular tubercles 9 9 Tergal coloration in front of yellow bands brownish red or reddish brown, postocular area without definite tubercles coloratus Baker Tergal coloration in front of yellow bands dark brown to nearly black, postocular area nearly always with low but definite tubercles forbesii Robertson Belomicrus coloratus Baker Belomicrus columbianus (Kohl) Belomicrus coloraia Baker 1909:29. Holotype female, Ormsby Oxybelus columbianus Kohl 1892:208. Lectotype female (ex- Co., Nevada (DAVIS), examined. - amined and here designated), Revelstoke, British Co- lumbia (VIENNA). Treated as a subspecies of forbesii by I have studied 422 males and 140 females from Bohart and Menke (1976). New status. Nevada and California. They were all taken in the Oxybelus larimerensis Rohwer 1908:417. Holotype female, months of June and July at elevations of 4,000 to Lanmer Co " Colorado (WASHINGTON), examined, oonnr lm j j • i i ^ , Treated as a synonym of forbesii bv Bohart and Menke 8,000feet. Nevada county records include Ormsby, (1976) Revised syn;nymy. Douglas, and Eureka. California counties are Al- pine, El Dorado, Inyo, Lassen, Mono, Nevada, In all/ 54 males ^j 69 females have been Placer, Sierra, and Tuolumne. A single Oregon srudied. These were taken during the months of record is 5 mi e. Bly, Klamath Co. May (rarely), June, July, and August, mostly at The principle diagnostic features, especially elevations above 4,000 feet. Western state records the tergal markings, are given in the key. include British Columbia, Alberta, Colorado, Idaho, Montana, Nevada, Oregon, Utah, and Wash- ington. California records are: Bell Echo Camp, 210 Journal of Hymenoptera Research Dorris, and Rattlesnake Meadow, Siskiyou Co.; Morrison Meadow and Snowslide Park, Trinity Co. The new status of columbianus and its syn- onym, larimerensis is indicated by the difference in markings from forbesii as outlined in the key. Belomicrus cookii Baker Belomicrus cookii Baker 1909:29. Lectotype male, Claremont, Los Angeles Co., California (WASHINGTON). Lecto- type designated by Pate (1940a). This species is known from a few southern California localities. I have seen 8 males and 11 females from Los Angeles Co. (Claremont, Newhall) and Riverside Co. (Riverside, Gavilan), all taken during the months of April and May at low elevations in foothill locations. Belomicrus desertus R. Bohart, new species Female holotype. — Length 5 mm. Black, red are: clypeal apex, foretarsus partly, tegula dully, terga mostly, sterna except for dark median blotches on T-I to T-III; whitish yellow are: mandible basally, scape partly, flagellum beneath, apical spots on femora, tibiae outwardly, squamal triangle, apical bands on T-I to III, weak on IV-V; wings nearly clear. Pubescence moderate, silvery and appressed on clypeus, lateral frontal band enlarged above, short supraclypeal spot, postocular, mesopleural, and lateral tergal areas moderately silvered. Punc- tation fine and close, a little shiny across frons below ocelli; pygidial plate with coarse, separated punctures. Postocular tubercles well developed, epipleural "button" weakly indented. Male. — Length 4-5 mm. Whitish yellow are: mandible mostly, clypeus across apex, basal tarsal segment, apical bands on T-I to VII (usually). Holotype female (DAVIS), 3 mi s. Kramer Junction, San Bernardino Co., California, IV-6-66 (R.M. Bohart). Paratypes, 123 males, 31 females topotypical, collected in April of various years by R.M. Bohart, E.I. Schlinger, M.E. Irwin, D.S. Horning, J.C. Hall, F.D. Parker. Other paratypes from San Bernardino Co., California: 3 males, 5 females, Joshua Tree National Monument, (H.K. Court); 13 males, 4 females, Adelanto (E.I. Schlinger, et al.): 6 males, 2 females, Apple Valley (W. R. Mason, P. D. Hurd). Other specimens (not paratypes) were from California counties: San Ber- nardino (Victorville, Morongo Valley, Red Moun- tain), Imperial (Palo Verde), Los Angeles (Llano, near Lovejoy Buttes), Ventura (Chuchupate Ranger Sta., Frazier Mt.), Kern (Dove Well), inyo (Panamint Mts., 8 mi w. Brown). Paratypes in cooperating museums. This species is similar to powelli which also has well developed postocular tubercles, the basal slope of T-I red, and the epipleural "button" scarcely indented. However, desertus usually has extensive whitish yellow tergal markings. Some- times, these may be on T-I only. See also poioelli. Etymology. — specific name derived from Latin adjective, desertus = abandoned. Belomicrus forbesii (Robertson) Figs. 12, 17 Oxybelus forbesii Robertson 1889:85. Holotype male, Colo- rado (PHILADELPHIA), examined. I have studied 118 males and 122 females, collected from May to August at elevations of 4,000 to 10,000 feet in the following states and counties: California (Alpine, Lassen, Modoc, Mono, Nevada, Sierra, Siskiyou, Trinity), Colorado (Larimer, Denver), Montana (Missoula), Nevada (Elko, Pine, Washoe), Utah (Box Elder, Kane, Sum- mit), and Wyoming (Sublette, Teton, Uinta). Fe- male face (Fig. 17). Diagnostic characters, particularly the abun- dant tergal markings, are given in the key. Belomicrus montanus R. Bohart, new species Female holotype. — Length 5 mm. Black, red are: clypeal apex, terga and sterna mostly except for dark blotches on S-I-II; whitish are: mandible ba- sally, scape apically, flagellum beneath, pronotal collar and lobe, apical spot on forefemur, tibiae outwardly, squamal triangle, T-I apex faintly; wings lightly stained. Pubescence light, silvery and appressed on clypeus, lateral frontal area, short supraclypeal spot, postocular and mesopleural areas moderately. Punctation fine and close, not shiny across frons below ocelli; pygidial plate with coarse, separated punctures. Postocular tubercles quite weak, epipleural "but- ton" slightly indented. Male. — Length 4-4.5 mm. Foretarsus partly Volume 3, 1 994 211 off-white, T-I whitish apically. Holotype female (DAVIS), White Mts., Mono Co., California, 10,000 ft., VII-10-68 (R.M. Bohart). Paratypes, 14 males, 13 females, topotypical but collected from June 22 to July 23 by J. W.MacSwain, J. Powell, and G.I. Stage. Other California paratypes: 1 pair, Benton, Mono Co., V-23-86 (R.M. Bohart); 3 males, Westgard Pass, Inyo Co., V-VI-37 (G.A. Hamsher, CD. Michener); 2 females, Deep Springs, Inyo Co., V-13, V-14 (L.D. French, N.J. Smith). Paratypes in cooperating museums as far as possible. There is some similarity to querecho, powelli, and desertus, since the basal slope of T-I is mainly red, and the epipleural "button" is scarcely in- dented. From all three of these the weakly indi- cated postocular tubercles of montanus are differ- entiating. Also, it may be separated from querecho by the distinct sternal punctation of the latter, as well as its broad depression on S-II. Etymology. — specific name derived from Latin adjective, montanus = dwelling on mountains. Belomicrus penuti Pate Belomicrns forbesii penuti Pate 1940:27. Holotype male, Yosemite Valley, Mariposa Co., California (PHILADEL- PHIA). Raised to species status by Bohart and Menke (1976). The 1,115 males and 51 females I have studied were collected during April to September in Cali- fornia and neighboring Nevada at elevations mostly above 5,000 feet. California counties are: Alpine, El Dorado, Fresno, Lassen, Mariposa, Mono, Nevada, Placer, Calaveras, Sierra, Tulare, Amador, and Tuolumne. Nevada records are In- cline Village and Little Valley, Washoe Co. Also, Pate (1940a:28) recorded penuti from southern Oregon: Crater Lake and Lake of the Woods, Kla- math Co. I have not studied these specimens. The relationships to columbianus and forbesii are outlined in the key. Belomicrus powelli R. Bohart, new species Female holotype. — Length 5 mm. Black, red are: clypeal apex, terga and sterna mostly except for dark blotches on S-II-III; whitish are: mandible basally, flagellum beneath, pronotal collar medi- ally, pronotal lobe, apical spot on femora, tibiae outwardly, foretarsus dully, squamal triangle; wings moderately stained. Pubescence silvery appressed on clypeus, lateral frontal area, stout supraclypeal triangle, postocular and mesopleural areas (weakly). Punctation fine and close, not shiny across f rons below ocelli, somewhat shiny on terga; pygidial plate with coarse, separated punctures. Postocular tubercles well developed, epipleural "button" slightly indented. Male. — Length 4-4.5 mm. Characters about as in female. Scape whitish at apex. Holotype female (SAN FRANCISCO), La Panza, 12 mi ne. Pozo, San Luis Obispo Co., Cali- fornia, IV-29-62 (J. Powell). Paratypes, 5 males, 16 females (all from California collected in April and May): Monterey Co.: 1 male, 5 females, Hastings Reserve (D.L. Linsdale); 1 pair, Arroyo Seco (D. Burdick, P. Torchio); female, 12 mi n. Cholame (D.H. Janzen); San Luis Obispo Co.: 2 pair, topotypical (J. Powell, R.W. Thorp); female, 5 mi ne. Santa Margarita (R.W. Thorp); Fresno Co.: 12 mi w. Coalinga (J.W. MacSwain); Ventura Co.: 3 males, 8 females, Chuchupate Ranger Sta., Frazier Mt. (J. Powell, P.D. Hurd). Paratypes in SAN FRANCISCO, DAVIS. The well developed postocular tubercles are also found in desertus, which see. The all red terga and broken yellow pronotal band of powelli are distinguishing. Although primarily a coastal form, poioelli occurs with desertus on Frazier Mt., Ventura Co., at Chuchupate Ranger Station. In fairly long series the differences given in the key seem to hold. The species is named for my friend, Jerry Powell, a lepidopterist who has also worked with wasps. Belomicrus querecho Pate Belomicrus querecho Pate 1940:36. Holotype male, Alamogordo, Otero Co., New Mexico (PHILADEL- PHIA), examined. This southwestern species is largely restricted to the Chihuahuan Life Zone. I have studied 3 males and 20 females, collected in April, May, and June in the following states: Arizona (near Apache and Willcox, Cochise Co.; Tubac, Santa Cruz Co.; Oak Creek Valley Road, Yavapai Co.; 10 mi w. Jacob Lake, Coconino Co.). Utah (25 mi s. Moab, Grand Co.). New Mexico (Skeleton Canyon, Peloncillo Mts., and Rodeo, Hidalgo Co.; 212 Journal of Hymenoptera Research Alamogordo, Otero Co.; 5 mi e. Las Cruces and Leasburg Dam, Dona Ana Co.). Nevada (4 mi s. Warm Springs, Nye Co. ) and Texas (Rankin, Upton Co.). Belomicrus sierrae R. Bohart, new species Female holotype. — Length 5 mm. Black or dark brown, red are: flagellum beneath dully, foretarsus dully, apical half of pygidial plate; whitish yellow are: mandible basally, femorotibial joints, tibiae outwardly, apicomedial spot on squamal triangle (all yellow in some paratypes); wings moderately stained. Pubescence silvery appressed on clypeus laterally, narrowly on lateral frontal area, short supraclypeal spot, weakly on postocular and mesopleural areas. Punctation fine, close, dull; clypeus wrinkled, with a pair of coarse punctures above polished apex, pygidial plate with coarse, separated punctures. Postocular tubercles present but weak, epipleural "button" deeply indented but not covered with a thin membrane; lateral propodeal carina flattened, somewhat bent in- ward. Male. — Length 4.5-5 mm. About as in female. Pronotal lobe sometimes dully pale. Holotype female (DAVIS), Sagehen Creek, Nevada Co., California, 6,500 feet, VI-25-68 (R.M. Bohart). Paratypes, 118 males, 97 females, topotypical, V-29 to VII-25 (R.M. Bohart, J. A. Skin- ner, L.S. Kimsey, M.E. Irwin, D.S. Horning, J.E. Slansky, J. Powell, B. Villegas). Also paratypes, 10 males, 19 females, Grass Lake, El Dorado Co., California, 8,000 ft., VII-5-62 and VII-16-62 (R.M. Bohart et ai). Other specimens (not paratypes) have been studied from other California counties at 6,000 ft. or above: Alpine, Calaveras, Plumas, Lassen, Placer, and Trinity. Paratypes in all coop- erating museums. The dark basal slope of T-I and the deeply indented epipleural "button" ally sierrae with penuti. However, the "button" is not as thinly lidded above as in penuti. Futhermore, the latter seems always to have at least several of the terga with whitish bands. The dark abdomen ot sierrae is similar to that of colwnbianus, and conceivably the two species might occur together in California near the Oregon border. In addition to the epipleural "button" difference, columbianns has the lateral propodeal carina simple, whereas in sierrae it is not sharp but flattened and bent in- ward. Etymology. — the specific name refers to the California Sierra. Belomicrus apache group The apache group contains only the single known species, apache Pate. It was described in detail from a single female by Pate (1940a) and placed in a key by Pate (1940b). This is one of the smallest known Belomicrus, approximately 2 mm in length. The female has a peculiar anchoriform terminal process figured by Pate (1940a) and herein (Fig. 16). The male clypeus is more ordinary. Fea- tures of the group are characterized in the key. A brief description of the species, including the pre- viously unknown male, and additional locality records follow: Belomicrus apache Pate Figs. 13, 16 Belomicrus apache Pate 1940a: 15. Holotype female, Las Cruces, Dona Ana Co., New Mexico (PHILADELPHIA), exam- ined. Female. — Length 2 mm. Black with red unhanded abdomen; legs partly whitish yellow, frons at narrowest point about equal to eye breadth, clypeus with an apical anchoriform projection (Fig. 16), no postocular tubercles or omaulus, squamal triangle translucent laterally, deeply in- cised posteriorly (Fig. 13). Male. — Length 2 mm. Much as in female but LID slightly greater and clypeal bevel triangular rather than anchoriform. I have seen three males and seven females from the following western states: New Mexico (Las Cruces, Doha Ana Co.; Rodeo, Hidalgo Co.), Arizona (Aztec, and 7 mi s. Quartzite, Yuma Co.; Quijotoa and Sells, Pima Co.; Willcox, Cochise Co. on Euphorbia mat), California (Cathedral City, Riv- erside Co. on Euphorbia poly carpa). Collections were made from August to October. Belomicrus vanyume group The vanyume group consists of three species of which one is described below as new. Pate (1940a, b) placed vanyume in a monotypic group of its own. The vanyume group, as I consider it, is unified by the distinct omaulus running vertically along the widest point of the mesopleuron, and ending Volume 3, 1994 213 in a tooth or angle. This is followed halfway to the midcoxa by another tooth. Females have the forefemur right-angled subbasally. Except for vanyume, males also have this feature. Further, females have the mandible broadened before the apex (Fig. 15). KEY TO SPECIES OF THE BELOM1CRUS VANYUME GROUP Postocular area a little swollen but not tuberculate, clypeus broadly rounded, scape light yellow, squamal unit with windowlike clear areas vanyume Pate Postocular area with prominent tubercles, clypeus with a longitudinal raised area medially, scape dark, squamal unit various 2 Pronotum all dark, mesopleuron completely punctate and dull, clypeus (as seen in side view) not dentate above bevel, T-II to V or III to V black texensis R. Bohart Pronotum usually partly whitish, pronotal lobe white, clypeus ending above bevel in a forward-directed tooth (Fig. 15), T-I to V usually red, especially in females maricopa Pate Belomicrus maricopa Pate Figs. 14, 15 Belomicrus maricopa Pate 1947:54. Holotype female, Higley, Maricopa Co., Arizona (WASHINGTON), examined. Belomicrus mariposa Pate 1947:55. In error. I have seen 44 males and 58 females from Arizona (Higley, Willcox, 28 mi s. Quartzite, Picacho Canyon), New Mexico (5 mi e. Las Cruces, 9 mi n. Cotton City), California (18 mi w. Blythe, Palo Verde), and Sonora, Mexico (Guaymas). This species is similar to texensis, described below and differences enumerated. The unde- scribed male is much like the female but is a little smaller and the mandible is somewhat less ex- panded. Also, the last few terga are darker. The principal differences between the two species are the shinier mesopleuron and apically toothed clypeus of maricopa (Fig. 15). Belomicrus texensis R. Bohart, new species Female holotype. — Length 5 mm. Black, light yellow are basal two-thirds of mandible, foretibia outwardly; brownish are: legs partly; reddish are: forefemur at base, T-I to III; wings clear, veins mostly orange. Silvery appressed pubescence on clypeus mostly, broad patch along eye margin reaching up as far as an imaginary line drawn across beneath midocellus, short but large supraclypeal patch. Punctation of mesonotum and interocellar area coarse and separated by 1-2 PD of microsculpture, mesopleuron similar but rugose below, tergal punctation moderate and a little shiny, coarse on pygidial plate. LID about 1.7x scape length, clypeus with median raised area bearing longitudinal depression, clypeal bevel transverse and nearly linear, frontal groove well impressed, postocular tubercles well developed, gena in side view bent outward at middle below, mandible strongly expanded at distal one-third, front dorsal margin of pronotum obtusely cor- nered, squamal unit as in Fig. 14, forefemur rightangled near base, pygidial plate narrowly triangular. Male. — Length 4 mm. Legs mostly brownish red, forewing veins dark, clypeal bevel broadly triangular, T-I-II red, III a little red laterally and posteriorly, VII dully red. Female holotype (DAVIS), Kingsville, South Pasture, Kleberg Co., Texas, VI-16-68 (J. E. Gillaspy). Paratypes, 4 females (DAVIS, NEW YORK), same data as holotype but V-6-67 and V- 6-68; on Ratibida columnaris; paratype male (SAN FRANCISCO), Mexico: Baja California Sur, e. edge of Sierra Placeres, 111-24-84 (W. J. Pulawski). The large size (for Belomicms\), large postocular tubercles, expanded mandible subapically and all dark antenna are also found in maricopa. However, texensis has many differences: pronotum all dark, 214 Journal of Hymenoptera Research punctation of mesopleuron and mesonotum much closer and not shiny, female clypeus with a longi- tudinal flattened and shallowly grooved median ridge, clypeus also without a forward-pointing apical tooth, wing veins of female mostly orange instead of black, marginal cell of forewing some- what shorter, and T-III to V mostly or all black. Etymology- — the specific name indicates "of or from" Texas. Belomicrus vanyume Pate Belomicrus vanyume Pate 1940a:17. Holotype female, Victorville, San Bernardino Co., California (PHILADEPHIA), examined. In addition to the type I have seen five males and three females, all from California counties: Riverside (Box Canyon, Thousand Palms), San Bernardino (Apple Valley, Kramer Hills, Adelanto, 1 1 mi w. Ludlow), Imperial (Fish Creek Mts.), Inyo (Darwin Falls). These were all taken in April, May, and June. The previously undescribed male is much like the female but the last two terga may be dark. Belomicrus franciscus group The absence of an omaulus, triangular median squamal complex (Fig. 11), mandible not angled beneath toward apex, and clypeus with an apical deflected bevel or polished triangle, when taken together distinguish the group. I have identified eight species, of which three are herein described as new. Species characters are the extent of tergal yel- low markings, flagellarcoloration, pronotal macu- lation, punctation of the postmandibular area of males, development of the mandibular midtooth, and form of the clypeal apex. Since females may be difficult to distinguish, the following key is based on males. KEY TO MALES OF THE BELOMICRUS FRANCISCUS GROUP Ventral area of head just posterior to mandibular insertion nearly all polished, flagellum relatively short, flagellomeres hardly longer than broad (Fig. 2) 2 Ventral area of head just posterior to mandibular insertion distinctly but not closely punctate, flagellum various 4 Clypeal apex arched (Fig. 5), T-I usually with a subapical yellow band or spot (central California below 3,000 feet) costalis R. Bohart Clypeal apex not much, if any, arched; terga usually all red or with only a narrow streak of yellow on T- I 3 Flagellum pale yellow beneath on I-IX or X; mandible tooth on inner margin well developed (Fig. 9); punctures of interocellar area close but not contiguous, area a little shiny (central Sierra of California at 4,000 to 5,000 feet) californicus R. Bohart Flagellum orange beneath before apex; mandible tooth on inner margin small (Fig. 8), punctation of interocellar area contiguous and dull (low elevations in California from San Luis Obispo Co. to San Diego Co.) serrano Pate Ground color on T-I to T-VI black, tergal markings whitish (central Sierra of California above 5,000 feet) melanus R. Bohart Ground color on T-I to T-III, at least, red; tergal markings pale yellow : 5 Clypeal bevel produced downward laterally (Fig. 4), terga maculate 6 Clypeal bevel narrowed laterally, maculation various 7 Flagellum unusually long, mostly orange, many flagellomeres longer than broad as viewed laterally (Fig. 1); clypeal bevel often with a small median point (Fig. 4); terga often with yellow spots but rarely with complete yellow bands (desert areas from Inyo Co., California and Clark Co., Nevada south to Sonora, Mexico) quemaya Pate Flagellum rather stout, dark above (Fig. 2), flagellomeres about as broad as long in lateral view; clypeal bevel without a median denticle; terga with complete yellow banding (San Francisco) franciscus Pate Clypeal apex convex overall, bevel simple below, femora often extensively red (southern California) cahuilla Pate Clypeal apex more nearly flat, bevel margined below by a slight inflection, femora black with a small amount of pale yellow (central Sierra of California at 5,000-7,000 feet) mono Pate Volume 3, 1 994 215 Belomicrus cahuilla Pate Belonucrus cahuilla Pate 1940:39. Holotype male, Andreas Canyon, Riverside Co., California (PHILADELPHIA), examined. The male differs from that of mono by the simple and convex clypeal bevel. In the female the bevel is more indented than that in mono or quemaya. Both sexes have dark median areas on the terga, which may extend all across on T-III and follow- ing. Only a single female has been identified, so variation cannot be assessed. The type series of four males came from Andreas Canyon (Palm Springs). I have also seen five males from close by (Taquitz Canyon, F. D. Parker and L. A. Stange). Other southern California records are: male, fe- male, near Cajon Junction, San Bernardino Co. (J. C. Hall, E. I. Schlinger); males, Walker Pass and West Wofford Heights, Kern Co. (J. Powell); male, 2 mi e. Banner, Riverside Co. (H. C. Dickson). Dates of capture ranged from April 16 to June 7 (males), and July 4 (female). Belomicrus californicus R. Bohart, new species Figs. 7, 9 Belomicrus franciscus Pate of Bohart, in Bohart and Menke 1976:363. Misidentification. Male holotype. — length 4.5 mm. Black, whitish are: mandible mostly, scape in front, flagellum beneath except at apex, pronotal collar medially, lobe, squamal unit, tip of mucro, femora distally, tibiae outwardly, foremetatarsus dully; red are: clypeal bevel mostly, abdomen except for median dark blotches on T-III to VI; wings weakly stained. Silvery appressed pubescence along inner eye margin, stout supraclypeal spot, weak pubescence on postocular area, mesopleuron, and terga. Punc- tation sparse on mostly polished postmandibular area of head venter, fine and close on dorsum of body, a little shiny on interocellar area. Clypeal bevel quite thin (Fig. 7), distinctly darkened later- ally; mandibular inner tooth well developed. Female. — Length 5 mm. Clypeal bevel slightly margined below, mandible (Fig. 9). Holotype male (DAVIS), Baxter, Placer Co., California, elev. 5,000 ft., VI-16-56 (R. M. Bohart). Paratypes: 18 males, 21 associated females, all from the California Sierra at 4,000 to 5,000 ft. elevation during May to July, in the following counties: Placer (Dutch Flat, Baxter, Colfax), Tuolumne (Strawberry), El Dorado (Pyramid Ranger Station and near Icehouse Road). Collec- tors were J. G. Rozen, J. W. MacSwain, R. M. Bohart, W. J. Pulawski, and H. M. Kimball Court. Paratypes are deposited in museums listed in acknowledgements. This species is close to Serrano. Both have the male postmandibular area practically impunctate, but in californicus the mandible tooth is well devel- oped, and punctation of the interocellar area is a little less dense. J. MacSwain and I observed a large nesting area in 1956 near Pyramid Ranger Station north of Placerville, California. Females were provisioning ground nests with dasytine beetles, Amecocerus cervicalis Blaisdell (Melyridae). Etymology. — the specific name indicates "of or from" California. Belomicrus costalis R. Bohart, new species Fig. 5 Male holotype. — Length 4.5 mm. Black marked with whitish yellow: scape and mandible mostly, flagellum beneath, pronotum all across, post- tegula, squamal unit, large forefemoral spot, small distal midfemoral spot, tibiae and metatarsi out- wardly, subapical band on T-I, attenuate laterad; red are: clypeal bevel, T-I to III mostly, T-IV to VI except for median dark blotches, T-VII, wings weakly stained. Silvery appressed pubescence in strip along inner eye margin, a little expanded above, supraclypeal spot, postocular area and mesopleuron moderately, terga lightly. Puncta- tion sparse on mostly polished postmandibular area of head venter, fine and close on dorsum of body, somewhat reflective on interocellar area. Clypeal bevel somewhat arched (Fig. 5). Female. — Length 5 mm. Terga all red, scape pale in front only, flagellum reddish yellow be- neath. Clypeal bevel flat, not margined below; mandibular inner tooth well developed. Holotype male (DAVIS), Cache Creek Can- yon, Yolo Co., California, IV-30-54 (R. M. Bohart). Paratypes (all from California at low to moderate elevations): male (DAVIS), 4 mi nw. Lake Berryessa, Napa Co., V-12-61 (F. D. Parker); 2 males (DAVIS, WASHINGTON), Sacramento, Sacramento Co., V-19-61 (R. M. Bohart); male (BERKELEY), Alpine Lake, Marin Co., VI-6-57 (J. Powell); 2 males 216 Journal of Hymenoptera Research (DAVIS), Midlake, Lake Co., V-10-56 (E. I. Schlinger); male (BERKELEY), near Quincy, Plumas Co., Vl-22-49 (J. W. MacSwain); 2 males (NEW YORK), Tuolumne City, Tuolumne Co., V- 30-53 (J. G. Rozen); male (DAVIS), Boca, Nevada Co., VI-19-62 (R. M. Bohart); male (DAVIS), Kyburz Flat, Sierra Co., VII-15-76 (R. M. Bohart); male (DAVIS), Daffodill Hill, Amador Co., VI-5-63 (R. M. Bohart). Also, 6 presumed and associated fe- males from localities listed above: Cache Creek Canyon, Boca, Daffodill Hill, Sacramento. The band on the pronotal collar is sometimes broken in both sexes. The arched clypeal apex (Fig. 5) differentiates males from serrano and californicus which agree with costalis in having the postmandibular area polished. Females are best distinguished by association with males. Etymology . — specific name derived from Latin noun, costa = side. The species occurs especially on the western side of California. Belomicrus franciscus Pate Figs. 2, 3, 11 Belomicrus franciscus Pate 1 331 :77.Holotype male, Lone Moun- tain, San Francisco, San Francisco Co., California (PHILA- DELPHIA), examined. I have studied several of the type series. The rather stout male flagellum (Fig. 2), extensively yellow-banded terga, and partly punctate male postmandibular area distinguish this species. Fe- male face (Fig. 3). The type series was collectecd by F. X. Williams (1936) in sandhills of San Francisco. Williams found a colony of the species in late April and early May of 1930. Females were provisioning with Trichochrous antennatus Mots. (Melyridae). In all probability the type locality has been taken over by a housing development. Belomicrus melanus R. Bohart, new species Fig. 6 Male holotype. — Length 4.5 mm. Black or brown, whitish are: scape in front, mandible mostly, flagellum beneath, pronotum medially, pronotal lobe, post-tegula, squamal unit, tip of mucro, outer distal spots on fore and midfemora, tibiae and foremetatarsus outwardly, subapical band on T-I, weak subapical bands on T-II to VI, apex of VII; dark red: clypeal bevel; wings weakly stained. Silvery appressed pubescence in strip along inner eye margin nearly to level of midocellus, stout supraclypeal spot, postocular area, mesopleuron moderately, terga lightly. Punctation moderate on postmandibular area of head venter, fine and close on dorsum of body, slightly reflective on interocellar area. Clypeal bevel thin all across (Fig. 6). Female. — Length 5 mm. Whitish bands on T-I- V sometimes better developed in para types, prono- tum sometimes white all across, pygidial plate partly red. Postmandibular area and rest of head venter polished. Mandibular inner tooth well de- veloped, clypeal bevel thin but impressed over its entire breadth. Holotype male (DAVIS), Sonora Pass, Tuolumne Co., California, elev. 9,624 ft., VII-6-61 (R. M. Bohart). Paratypes (all from California Si- erra at elevations above 5,000 ft.): female, same data as holotype; 3 females, Hope Valley, Alpine Co., VII, 1948, 1978 (R. M. Bohart); male, Fred's Place, El Dorado Co., VII-10-67 (R. M. Bohart); male, Gold Lake, Sierra Co., VII-8-54 (R. M. Bohart). Paratypes all in DAVIS museum. This species belongs to the subgroup in which males have the postmandibular area somewhat punctate. It differs from other members of the subgroup by the extensively black (rather than red) ground color of the terga. In cahmlla the terga may be partly black but the male clypeal bevel is not rimmed above, and the terga are not spotted with whitish yellow. Etymology. — specific name derived from the Greek melanos = black. Belomicrus mono Pate Belomicrus serrano mono Pate 1940:46. Holotype female, 5 mi n. Round Valley, Mono Co. California (PHILADEL- PHIA), examined. Belomicrus mono Pate, of Bohart and Menke 1976:363. The male belongs to the subgroup with punc- tate postmandibular area on the head venter. The characters given in the key separate it. Females are best identified by association with males. I have seen 83 males and 65 associated females from 5,000 to 7,000 feet, in the California Sierra of the following counties: Mono (Round Valley holo- type), Placer (Lake Tahoe), Sierra (Yuba Pass, Si- erra Valley, Independence Lake, Sattley), Nevada Volume 3, 1 994 217 (Sagehen Creek), Glenn (Plaskett Meadows). The Sagehen Creek locality is represented by a long series. Collection dates are mostly in June and July. Belomicrus quemaya Pate Figs. 1, 4, 10 Belomicrus quemaya Pate 1940:47. Holotype male, PalmSprings, Riverside Co., California (PHILADELPHIA), examined. Belomicrus jurumpa Pate 1940:53. Holotype male, Gavilan, Riverside Co., California (PHILADELPHIA), examined. New synonymy. Belomicrus franciscus quemaya Pate, of Bohart and Menke 1976:363. The punctate postmandibular area of the head venter in the male places quemaya in its subgroup. The unusually long and lightly colored flagellum of both sexes and the distinctive clypeal bevel of the male (Figs. 1, 4) are distinguishing. Wings are nearly clear in both sexes. Females are best identi- fied by association with males. B. jurumpa appears to be a simple synonym based on an examination of the holotypes. I have studied 60 males and 30 associated females collected from March to May. California county records are: Riverside (Whitewater Can- yon, Gavilan, Thousand Palms, 18 mi w. Blythe, Palm Springs), San Bernardino (Kramer Jet., Adelanto), Kern (Short Canyon, Iron Canyon), Imperial (Kane Springs, Fish Creek Mts., Palo Verde), Inyo (Mazourka Canyon, Lone Pine), San Diego (Borrego). Other records are Clark Co., Nevada (Searchlight, Jean), and Sonora, Mexico (9 mi e. Kino). Belomicrus serrano Pate Fig. 8 Belomicrus serrano serrano Pate 1940:42. Holotype male, Tujunga Wash, San Gabriel Mts., Los Angeles Co., Cali- fornia (PHILADELPHIA), examined. The polished postmandibular area of the male places the species in the same subgroup as costalis and califomicus. The dull interocellar area and weak mandible tooth (Fig. 8) separate serrano. Also, the relatively simple male clypeal bevel, and the indented one of the female are additional characters. I have seen 46 males and 27 females in addi- tion to the type series. The species is widespread in southern California. The range includes low el- evation localities in the following counties: Los Angeles, Riverside, San Bernardino, San Diego, Kern, Ventura, and San Luis Obispo. Collection dates are from March 27 to May 7. Belomicrus cladothricis group The deeply divided squamal lobes of the met- anotum (Figs. 28-43), and short body length, char- acterize the group. With the exception of cladothricis Cockerell (1895), all previously known species were described by Pate (1940a, 1947). The group appears to be largely confined to western United States, only a few species occurring east of the 100th meridian or in Sonora and the Baja Califor- nian peninsula of Mexico. Considering a small amount of synonymy and the 6 new species de- scribed herein, the total number of known species in the group is now 18. In most of these the squamal lobes are separated at the base by dark integu- ment. In some species, however, the pale color and setal texture may be continuous basally, at least in females. These are cladothricis, eriogoni, and sechi. Species characters of the tiny forms in this group are squamal structure, scutal and interocellar punctation, frons breadth, clypeal bevel details, and the somewhat variable coloration. In one spe- cies, sechi, the unusual amount of silvery pubes- cence, and shape of the male scape are distinguish- ing. Although little is known about the life history of these species, they are all presumably ground nesting (personal observation of eriogoni at Antioch, California), and they are most frequently collected on the flowers of Eriogonum. Other flower hosts reported are Chilopsidis, Lepidospartum, Euphorbia, Cladothrix, Cleomella, Rhamnus, Chrysothamnus, Solidago, Cercidium, Prosopis, Sphaeralcea, Asclepias, Baeria, and Salsola. Because of the small size of species in the cladothricis group (2-4 mm long), it can be surmised that they, like many other di- minutive sphecids, provision with thrips. A great deal of descriptive matter on the group was given by Pate (1940a, b, 1947). However, he had a limited amount of material to work with and many of his new species were based on one or two specimens, often of a single sex. His remarks on geographical range must not be taken too literally. For instance, on timberlakei he said (1940a:93), "Known at present only from the two specimens 218 Journal of Hymenoptera Research recorded above from Riverside, timberlakei will in Transverse Ranges district of southern Califor- all probability eventually be found rather wide- nia". As I have shown later in this paper, Pate's spread throughout the southern piedmont of the conclusion was far off the mark. KEY TO SPECIES OF THE CLAD0THR1CIS GROUP 1 Scutum partly polished, punctures widely spaced 2 — Scutum with rather close punctation 3 2 Forewing veins normally dark, squamae separated by less than 3x mucro width (Fig. 34) mescalero Pate Forewing veins practically all white, squamae separated by about 4x mucro width (Fig. 33) pallidus R. Bohart 3 Pronotum all dark including lobe 4 Pronotum partly whitish, at least on pronotal lobe 6 4 Distance between squamal apices about half length of scutellum (Fig. 35), area between squamae not pointing posteriorly timberlakei Pate Distance between squamal apices at least two-thirds length of scutellum (Figs. 41, 42), area between squamae various 5 5 Area between squamae pointed posteriorly (Fig. 43), scutal punctures moderate and close, terga brown to black bridwelli Pate — Area between squame truncate (Fig. 41), scutal punctation quite coarse, T-I or T-I-I1I with some red potawatomi Pate 6 Squamae relatively long and angled inward so that apices are not or hardly farther apart than width of mucro (Figs. 37, 39,40), LID 1.3x to 1.6x length of scape 7 — Squamal apices not unusually long or angled inward, apices farther apart than width of mucro, LID various 9 7 Scutal punctation fine and close, not reflective, interocellar area likewise, T-I not yellow spotted or banded, LID 1.4-1.5x length of scape pachappa Pate — Scutal punctation a little uneven, some punctures medium fine; surface reflective, interocellar area likewise, tergal markings and LID various 8 8 Scutal and interocellar punctation a little more coarse and distinct, terga without yellow markings tuktum Pate — Scutal and interocellar punctation a little finer and less distinct, T-I or T-I-II often with yellow markings istam Pate 9 Abdomen black; squamae slender, crescent shaped (Fig. 22), scutal punctation medium coarse with polished interspaces of 1 PD or more; scape whitish, flagellum partly pale beneath; pronotal collar whitish all across; tibiae and tarsi whitish vierecki Pate Abdomen with some pale markings or at least with T-I usually somewhat reddish, squamae not so crescent shaped, scutal punctation with microsculptured interspaces, antenna and pronotal collar various, tibiae and tarsi various 10 10 Head unusually long and narrow (Figs. 21, 24), inner eye margins nearly parallel below ocelli, postocular tubercles prominent and sharp, female genal area depressed and densely silvery below (genal carina undeveloped), frontal groove extending below broadly rounded brow, hindtibia and hindmeta tarsus all whitish outwardly 11 Head not unusually long and narrow (Figs. 18, 20), inner eye margins more strongly curved, female genal area not depressed, frontal groove and leg markings various 12 11 Flagellum extensively pale beneath, forewing veins mostly pale, T-I to III (male) or T-I to VI (female) with prominent whitish bands, squamae somewhat curved but shorter than their interapical distance (Fig. 36) oraibi R. Bohart Volume 3, 1 994 219 — Flagellum dark beneath, forewing veins mostly brown, T-I-VI unbanded, squamae strongly curved within and longer than their interapical distance (Fig. 38) longiceps R. Bohart 12. Body with exceptional amounts of silvery pubescence (Fig. 30), interocellar area and vertex silvery setose, propodeum usually extensively red, male scape strongly swollen distally (Fig. 27) sechi Pate — Body with moderate amounts of silvery pubescence, interocellar area and vertex not silvery, propodeum black, male scape not swollen 13 13 Flagellum at least partly lighter colored toward base 14 — Flagellum practically all dark 16 14 Flagellum mostly light orange with some dark spots laterally, squamae moderately stout (Fig. 32) darivini R. Bohart — Flagellum somewhat lighter toward base beneath; squamae short and stout 15 15 Interocellar area, vertex, and scutum dull; terga red, at least toward base; female frons length below midocellus 1.6x LID (Fig. 22) eriogoni Pate — Interocellar area, vertex, and scutum somewhat shiny; T-I or II often pale marked, female frons length below midocellus 1.9x LID (Fig. 18) cladothricis Cockerell 16. Frons length below midocellus 1.2-1.3x LID (Fig. 19), wing veins often mostly orange (female) or light brown (male) rather than dark brown to black siccatus R. Bohart — Frons length below midocellus 1.7-1.9x LID (Fig. 20), wing veins mostly dark brown to black 17 17. Mid and hindmetatarsi brown or brownish red (Fig. 26) postocular tubercles small as seen laterally or dorsally, LID in female usually nearly equal to scape length, female scape whitish in front inyo R. Bohart — Mid and hindmetatarsi white (Fig. 25), female only known, postocular tubercles large, LID a little broader, scape all whitish cucamonga Pate Belomicrus bridwelli Pate Fig. 43 Belomicrus bridwelli Pate 1940a:93. Male holotype, Clifton, Fairfax Co., Virginia (PHILADELPHIA), examined. In addition to the holotype and four topotype paratypes, I have also studied the following: male, Branford, Suwannee Co., Florida, VII-31-30; 2 males, female, Fort Bragg, Cumberland Co., North Carolina; female, Dayton, Rhea Co., Tennessee, VI-20-54; male, 15 mi nw. Big Spring, Howard Co., Texas, VI-1 3-63. According to these meager records, bridwelli occurs east of 102° latitude in May, June, and July. Female squamae and mucro (Fig. 43). Belomicrus cladothricis (Cockerell) Fig. 18 Oxybelus cladothricis Cockerell 1895:309. Female lectotype, Las Cruces, New Mexico (PHILADELPHIA), examined. Lectotype designated by Cresson (1928). Belomicrus cladothricis prosopidis Pate 1940a:72. Male holo- type, Palm Springs, Riverside Co., California (PHILA- DELPHIA), examined. Belomicrus minidoka Pate 1940a:79. Male holotype, Hagerman, Gooding Co. Idaho (WASHINGTON), examined. This rather wide-ranging and relatively abun- dant species is represented in the DAVIS collec- tion by 450 males and 198 females taken at 51 localities during March to October. All of the captures were made south of 41.5° longitude and west of 100° latitude. Outlying states in the range are Nebraska, Texas, Utah, Idaho, and California. Elevations were moderate to low and many were in strictly desert areas. As might be expected of such a common and widespread species, there is a considerable amount of variation. The short and rather broad squamae are a consistent feature. Also, the narrow LID of females (Fig. 18) helps identification. Many western examples have yel- low spots on basal terga. Punctation of the scutum and interocellar area is close but allows some shininess not found in several related species such as eriogoni. The legs are usually extensively red- dish but this occurs in some other species. T-I-II are often reddish in males but II may be all dark brown as in the type of prosopidis, or I-II brown as in the type of minidoka. 220 Journal of Hymenoptera Research Belomicrus cucamonga Pate Fig. 25, 31 Belomicrus cucamonga Pate 1940a:76. Female holotype, Camp Baldy, San Gabriel Mts., San Bernardino Co., California (PHILADELPHIA), examined. In addition to the holotype I have seen three females from Big Pine, Inyo Co., California, VII- 10-61 (R. M. Bohart); and a female from Batchelder Springs, Inyo Co., VI-10 (H. K. Court). The more abundant whitish of the scape and tarsi (Fig. 25), as well as the larger postocular tubercles and broader frons, distinguish cucamonga from inyo. Female squamae and mucro (Fig. 31). Belomicrus darwini R. Bohart, new species Fig. 32 Female holotype. — Length 3 mm. Black or dark brown, white are: mandible mostly, scape in front, pronotal lobe, squama, mucro distally; reddish are: clypeal bevel, flagellum (pale orange, spotted laterally), tibia and tarsi mostly (brownish red), hindtibia lighter in basal ring, dull apical bands on T-I to III; wings clear, veins black. Silvery ap- pressed pubescence extending upward along two- thirds of inner eye margin, short supraclypeal patch, weak on genal and mesopleural areas. Punc- tures of scutum and interocellar area medium fine, slightly separated by microsculpture. LID about 1.4x scape length, frontal suture relatively deep halfway from midocellus to clypeus, clypeal bevel margined and triangular, postocular tubercle small, squamae short and slightly curved (Fig. 32). Male. — Length 2.5 mm. Scape mostly brown- ish, flagellum a little spotted laterally as viewed in front. Holotype female (DAVIS), Darwin Falls, Inyo Co., California, May 17, 1970 (R. M. Bohart). Paratypes, 2 males, 2 females, same data as holo- type but collected by E. E. Grissell and R. M. Bohart. Other paratypes, female (DAVIS), Wikiup, Mohave Co., Arizona, IV-11-57 (T. R. Haig); fe- male (RIVERSIDE), Kyle Canyon, Clark Co., Ne- vada, V-4-41 (P. H. Timberlake); female (DAVIS), near Boron, Clark Co., Nevada, VI-7-41 (P. H. Timberlake). As indicated in the key, the rather close punc- tation, moderate LID, brownish legs and terga, short squamae, light orange flagellum, and long frontal suture characterize the species. The flagel- lum is a little longer and more slender than in other species. Etymology. — named for Dr. Darwin French, who discovered Darwin Falls in 1860. Belomicrus eriogoni Pate Fig. 22 Belomicrus cladothrias eriogoni Pate 1940a:70. Male holotype, San Lucas, Monterey Co , California (PHILADELPHIA), examined. Belomicrus eriogoni Pate of Bohart and Menke (1976). I have studied 240 males and 165 females in the DAVIS collection. These were taken during May to September in California, Arizona, New Mexico, Nevada, and Utah. California records are from Mono Co., Lassen Co., Sacramento Co., and Contra Costa Co., south to the Mexican border. Mexican records are from Baja California. At times the species is abundant, especially on the flowers of Eriogonum fasciculatum. Female facial propor- tions (Fig. 22). Belomicrus inyo R. Bohart, new species Fig. 20, 26, 28 Female holotype. — Length 2.5 mm. Black, white are: mandible mostly, scape in front, pronotal lobe, squama, foretibia outwardly, mid and hindtibiae on basal one- third; reddish are: clypeal bevel, abdomen (more brownish on T-II to VI); wings clear, veins black. Silvery appressed pubes- cence is narrow along lower half of inner eye margin, supraclypeal patch conical, moderate on gena and mesopleuron. Punctures of scutum and interocellar area medium fine, slightly separated by microsculpture. LID about equal to scape length (Fig. 20); frontal suture not deep; faintly visible halfway from midocellus to clypeus; clypeal bevel triangular, postocular tubercles small; squamae curved, well separated distally (Fig. 28). Male. — Length 2.5 mm. Hindtibia mostly or- ange with dirty white basal area, mid and hindmetatarsi orange or reddish brown. LID about 1.7x scape length. Holotype female (DAVIS), Antelope Springs, Inyo Co. California, V-10-61 (R. M. Bohart). Paratypes, 2 males, 1 female (DAVIS, BERKE- LEY), topotypes, VI-29-61, VII-1-61, VIII-24-60, on Volume 3, 1 994 221 Solidago (P. D. Hurd, G. I. Stage). The orange to brown metatarsi (Fig. 26), dark flagellum, narrow female LID (Fig. 20), basally banded mid and hindtibiae (Fig. 26), close and dull punctation, and well separated squamae (Fig. 28) characterize the species. Etymology. — specific name is an acronymic noun inspired by Inyo Co., where all of the type series was collected. Belomicrus istam Pate Fig- 37 Belomicrus istam Pate 1940a:81. Male holotype, 4.5 mi nw. Edom, Riverside Co., California (PHILADELPHIA), examined. I have seen a total of 37 males and 62 females collected from March to June in sandy areas of 11 localities. These range in California from Sacra- mento to Borrego Valley, and in Nevada from Nixon to Stillwater. I collected a long series of both sexes lighting on a sandy slope in Sand Canyon, Kern Co., California, on April 14, 1962. Female squamae and mucro (Fig. 37). Belomicrus longiceps R. Bohart, new species Figs. 24, 38 Female holotype. — Length 2 mm. Black, white are: mandible mostly, pedicel partly, pronotal lobe, post-tegula, squama, mucro, tibiae outwardly, tarsi; reddish are: clypeal bevel, abdomen; wings clear, veins brown. Silvery appressed pubescence present along lower half of eye margin, short supraclypeal patch, dense on genal area and mesopleuron. Punctation on scutum and interocellar area fine and close but a little irregu- lar, somewhat shiny. Head unusually long and narrow (Fig. 24), inner eye margins nearly straight (Fig. 24), frontal carina distinct one-third of dis- tance from midocellus to clypeus, LID 1.7x length of scape, clypeal bevel broader than high, postocular tubercle prominent and sharp, genal area flattened to a little concave inside row of erect hairs, squamae short but curved and well sepa- rated (Fig. 38). Male. — length 1.8 mm. T-I-II partly black, T- III-VII black. Holotype female (DAVIS), 6 mi nw. Adelanto, San Bernardino Co. California, V-ll-79, on Spliaeralcea (R. M. Bohart). Para types collected April to August, 4 males, female (DAVIS, WASHING- TON), 18 mi w. Blythe, Riverside Co., California (M. E. Irwin, F. D. Parker, R. M. Bohart, R. W. Brooks); male (DAVIS), 15 mi w. Baker, San Ber- nardino Co., California (N. J. Smith); female (BER- KELEY), Surprise Canyon, Inyo Co., California (P. D. Hurd); male (DAVIS), Patrick, Washoe Co., Nevada (E. E. Grissell); female (NEW YORK), 2 mi n. Rodeo, Hidalgo Co., Nevada (J. G. Ehrenfeld). Characteristic are the long head (Fig. 24), dark antenna, sharp postocular tubercles, extensively silvery female gena and mesopleuron, brown wing veins, and unhanded abdomen. The only other species with a rather long head (compare Figs. 21, 24) is the much more abundantly marked oraibi, which see. Belomicrus mescalero Pate Fig. 34 Belomicrus mescalero Pate 1940a:87. Female holotype, Alamogordo, Otero Co., New Mexico (PHILADEL- PHIA), examined This species is known to occur in April and May. In addition to the three female types, one of which came from Needles, California, I have stud- ied the following: male, female, 3 mi sw. Rodeo, Hidalgo Co., New Mexico, IV-30-65 (J- G. Rozen). The partly polished scutum, together with the dark wing veins, mostly red terga, and whitish- yellow spotted T-I are distinguishing. The male is in general agreement with the female description given by Pate, but T- VI- VII are reddish. Female squamae and mucro are shown in Fig. 34. Belomicrus oraibi R. Bohart, new species Figs. 21, 36 Female holotype. — Length 3 mm. Black, white are: mandible mostly, scape in front, pronotal lobe, tegula, post-tegula, metanotum, mucro dis- tally, tibiae and tarsi outwardly (a little reddish), T-I to VI mostly; reddish are: scape partly, flagel- lum beneath, clypeal bevel, femora distally, tibiae partly, terga basally; wings clear, most veins light orange. Silvery appressed pubescence along lower three-fourths of eye margin, short supraclypeal patch, dense on genal area and mesopleuron. Punc- tation fine, close and dull on scutum, a little reflec- 222 Journal of Hymenoptera Research tive on interocellar area. Head longer than usual (Fig. 21 ), inner eye margins nearly straight, frontal carina distinct, LID 1.7x length of scape, clypeal bevel broader than high, postocular tubercles prominent and sharp, genal area slightly concave inside row of erect hairs; squamae short, curved, well separated (Fig. 36). Male . — Length 2.5 mm. About as in female but T-IV-VI black. Genal area not concave. Holotype female (DAVIS), Oraibi, Navajo Co., Arizona, VIII-30-76, on Chrysothamnus (R. M. Bohart). Paratype female (WASHINGTON), Dinnebito Wash, 5 mi nw. Hotevilla, Navajo Co., Arizona, VIII-31-86 (A. S. Menke); paratype male (DAVIS), 23 mi nw. Page, Kane Co., Utah, VI- 22-70 (N. J. Smith); paratype male (SAN FRAN- CISCO), Green River, 5 mi ne. Jensen, Uinta Co., Utah, VI-26-82 (W. J. Pulawski). Although generally similar to lotigiceps, which also has an unusually long head, oraibi differs by its pale wing veins and extensively white terga. The concavity of the lower gena of the female is even more pronounced. The presumed males from Utah agree with the female in punctation and wing vein color. However, they have T-IV to VI dark. The presently known range includes north- ern Arizona and southern Utah. Etymology. — the name is an acronymic noun inspired by the Oraibi indians of Navajo Co., Arizona. Belomicrus pachappa Pate Fig. 40 Belomicrus pacliappa Pate 1940a:73. Female holotype, River- side, Riverside Co., California (PHILADELPHIA), ex- amined I have studied 11 males and 30 females col- lected from April to August at 4 localities in Cali- fornia: Weed, Siskiyou Co. (J. Powell); Sand Can- yon, Kern Co. (R. M. Bohart); Chuchupate Ranger Station, Ventura Co. (J. Powell, P. D. Hurd); River- side, Riverside Co. (P. H. Timberlake). Nevada records axe from Washoe Co.: Wadsworth (R. M. Bohart, G. I. Stage); Patrick (R. M. Bohart). The 3 species, pachappa, tuktum, and istam, have the squa- mae elongate and nearly touching distally. The close and dull punctation of pachappa separates it. The female has T-I-VI red, but the male has the abdomen dark following T-I or T-II. Female squa- mae and mucro (Fig. 40). Belomicrus pallidus R. Bohart, new species Fig. 23, 33 Male holotype. — Length 2 mm. Black, white are: mandible mostly, antenna mostly but dark distally in front, pronotum all across, tegula and post-tegula, wing veins, squamae and mucro, legs beyond basal two-thirds of femora; reddish are: legs basad of white markings, abdomen; wings clear. Silvery appressed pubescence on face below midocellus except for a median bare spot (Fig. 23), genal and mesopleural areas densely silver, scutum with scattered silvery setae. Punctation fine and sparse on extensively polished mesonorum, a little less so on interocellar area. LID (just above anten- nal insertion) 1.4x scape length, inner eye margin weakly curved below midocellus (Fig. 23); frontal groove and clypeal bevel (if any) obscured by silvery pubescence; postocular tubercle promi- nent, sharp; squamae short, curved, pubescent, well separated (Fig. 33). Female.— Unknown. Male holotype (SAN FRANCISCO), Borrego, San Diego Co., California IV-25-54 (P. D. Hurd). Paratype male (DAVIS), Blythe, Riverside Co., California, on Asclepias, VII-8-56 (M. Wasbauer). The rather extensive silvery pubescence is reminiscent of sechi, but there are many points of difference, one of which is the simple scape in pallidus male. General features characterizing pallidus are the considerably polished mesonorum, whitish wing veins, and red and white legs. Etymology, — specific name based on the Latin adjective pallidus = pale. Belomicrus potawatomi Pate Fig. 41 Belomicrus potawatomi Pate 1947:57. Female holotype, Sioux City, Woodbury Co., Iowa (WASHINGTON), exam- ined. The holotype was collected July 12, 1935 by C. N. Ainslee. The only other record of which I am aware is: female, Regnier, Cimarron Co., Okla- homa, June 9 (NEW YORK). This species is similar to bridwelli which also has the squamae far apart. However, the metanorum is posteriorly pointed (compare Figs. 41, 43) in bridwelli and its scutal punctation is less coarse. Volume 3, 1994 223 Belomicrus sechi Pate Figs. 27, 30 Belomicrus sechi Pate 1940a:60. Male holotype, 6 mi n. Palm Springs, Riverside Co. California (PHILADELPHIA), examined. I have seen 17 males and five females of this extensively silvery pubescent species. They were taken from May to August in the following desert localities of southern California: Palm Springs, Magnesia Canyon, and Rancho Mirage, all River- side Co.; 11 mi w. Ludlow, San Bernardino Co.; and Borrego Valley, San Diego Co.; a male (SAN FRANCISCO) comes from Arizona: Roper Lake State Park, Graham Co., V-26-83 (W. J. Pulawski). The male has a unique, distally swollen scape (Fig. 27). Otherwise, the female agrees with the descrip- tion given by Pate. The female squamae and mu- cro are shown in Fig. 30. Belomicrus siccatus R. Bohart, new species Figs. 19, 29 Female holotype. — Length 2.5 mm. Black, white are: mandible basally, pronotal tubercle, squa- mae, mucro distally; light dull orange are: wing veins, tibiae and tarsi; dark red are: T-I-II, py- gidium; wing membrane clear. Silvery appressed pubescence in a broad patch reaching up along two- thirds of inner eye margin, supraclypeal patch undeveloped, postocular and mesopleural patches moderate. Punctation of scutum and interocellar area fine and a little separated, somewhat shiny. LID 2x scape length, inner eye margin gently curved (Fig. 19), postocular tubercles present but not sharp, squamae nearly straight (Fig. 29). Holotype female (DAVIS), Thousand Palms Canyon, Riverside Co., California, IV-9-64 (R. M. Bohart). Paratypes, from Riverside Co.: 2 males (DAVIS), Whitewater, IV-17-87 (N. J. Smith); 7 males (DAVIS), SAN FRANCISCO, WASHING- TON), same data as holotype; 2 male topotypes (DAVIS), 111-29-77 (R. M. Bohart, N. J. Smith); 2 male topotypes (DAVIS), 111-31-63 (F. D. Parker); female topotype (DAVIS), IV-11-70 (R. M. Bohart). Also a paratype female (RIVERSIDE), Apple Val- ley, San Bernardino Co., California, V-20-41 (P. H. Timberlake). The combination of unusually broad frons (Fig. 19) black flagellum, nearly straight squamae (Fig. 29) orange (female) to brownish (male) wing veins, and orange-tinted leg markings character- ize the species. Etymology. — specific name derived from the Latin adjective siccus = dry. Belomicrus timberlakei Pate Fig. 35 Belomicrus timberlakei Pate 1940a:91. Male holotype, River- side, Riverside Co., California (PHILADELPHIA), ex- amined. This species has been collected from April to September. I have studied 41 males and 1 1 females from California, Nevada, Utah, Arizona, and New Mexico. In California 13 localities range from Siskiyou Co. to Riverside Co., mostly in the Tran- sition Life Zone. One record is at 8,000 ft. on Carson Pass, Alpine Co. Non-California records are: Cochise Co., Arizona (Willcox, Apache, Santa Rita Mts.); Eureka Co., Nevada (Eureka); Churchill Co. Nevada (Frenchman); Grant Co., Utah (Moab); Emery Co. Utah (Greenriver); Hidalgo Co., New Mexico (Rodeo); Baja California, Mexico (10 mi e. San Quintin). Of the 3 species with all dark pronotum (timberlakei, potazoatomi and bridzoelli) only tnnberlakei has the squamae only moderately sepa- rated (Fig. 35). The female resembles the male in nearly all respects but its LID equals 2x the scape length instead of 1.7. Belomicrus tuktum Pate Fig. 39 Belomicrus tuktum Pate 1940a:84. Male holotype 5 mi e. Edom, RiversideCo., California (PHILADELPHIA), examined. In addition to the holotype I have studied 24 males and 25 females collected from March to May. The 13 localities represented include the following California counties: Inyo (Darwin Falls, Little Lake, Surprise Canyon), Kern (Sand Can- yon), Ventura (Quatal Canyon), Riverside (Thou- sand Palms, Deep Canyon, 18 mi w. Blythe, Whitewater Canyon), San Bernardino (5 mi n. Barstow), Imperial (Glamis), and San Diego (Borrego Valley). This species is quite similar to istam. Females have the abdomen essentially all dark red, but males usually have T-I dark red, the rest brown. Female squamae and mucro are shown in Fig. 39. 224 Journal of Hymenoptera Research Belomicrus vierecki Pate Fig. 42 Belomicrus vierecki Pate 1940a:56. Male holotype, Alamogordo, Otero Co., New Mexico (PHILADELPHIA), examined. This is a relatively abundant and widespread species. I have seen 272 males and 47 females collected from March to July. I swept a long series from flowers of Chilopsis linearis at 15 mi w. Baker, San Bernardino Co., California in May. Other pre- ferred hosts are Cercidium and Prosopis. The 15 localities for California are all sandy situations at low elevations south of 38.6 longitude. Arizona records are from Wickenburg, Tucson, and Santa Rita Mts. New Mexico records are Alamogordo, Las Cruces, and near Laguna, Valencia Co. Mexi- can material has been seen from Sonora (Alamos), Puebla (Petlalcingo), Durango (Nombre de Dios), and Baja California Sur (La Ribera). The most eastern record in the United States is Presidio, Texas at about 104.5° latitude. The dark brown abdomen, white tibiae and tarsi, extensively polished interocellar area, nar- row LID, and well separated crescentic squamae (Fig. 42), make identification relatively simple. REFERENCES Baker, C. F. 1909. Studies in Oxybelidae I. Pomona journal of Entomology 1:27-30. Bohart, R. M. 1956. Prey captures of Belomicrus penuti and B coloratus. In Bohart and Menke 1976: 363. Ibid. Bohart, R. M. and A. S. Menke. 1976. Sphecid Wasps of the World. A Generic Revision, ix + 695 pp. Univ. California Press, Berkeley Cockerell, T. D. A. 1895. The second Anacrabro and the small- est American Oxybelus. Canadian Entomologist 27: 308- 309. Costa, A. 1871 . Prospetto sistematico degh Immenotterologia Itahana. Ann Mus. Zool. Univ. Napoli 6: 28-83. Cresson, E. T. 1928. The types of Hvmenoptera in the Acad- emy of Natural Sciences of Philadelphia other than those of Ezra T. Cresson. Memoirs of American Entomo- logical Society 5: 1-90. Evans, H. E. 1969. Notes on the nesting behaviour of Pisonopsis clypeata and Belomicrus forbesu. Journal of the Kansas Entomological Society 42: 117-125. Kohl, F. F. 1892. Neue Hymenopterenformen. Ann. Nat Hist Hofmus Vienna 7: 197-234. Pate, V. S. L. 1931 . A new Belomicrus from the West. Ent. News 42: 77-78. Pate, V. S. L. 1940a. The taxonomy of the Oxybeline wasps I. A review of the genera Belomicroides, Bnmocelus and Belomicrus with particular reference to the nearctic spe- cies. Transactions of the American Entomological Society 66: 1-99. Pate, V. S. L. 1940b. The taxonomy of the Oxybeline wasps II. The classification of the genera Belomicrus and Enchenncrum . Transactions of the American Entomological Society 66: 209-264. Pate, V. S. L. 1947. New North American Belomicrus. Proceed- ings of the Entomological Society of Washington 49: 54-57. Robertson, C. 1889. Synopsis of North American species of Oxybelus. Transactions of the American Entomological Soci- ety 5&. 77-85. Rohwer, S. A. 1908. Four new Hvmenoptera. Entomological News 19:417-420. Williams, F. X. 1936. Notes on two oxybehd wasps in San Francisco. Pan-Pacific Entomologist 12: 1-8. Volume 3. 1 994 225 1. quemaya 2. franciscus 3. franciscus 4. quemaya 5. costalis 6. melanus 7. californicus 8. serrano 15. maricopa 16. apache 17. forbesii Figs. 1-17. 1, 2, male antenna xlOO; 3, female face; 4-7, male clypeus, anterior view xlOO; 8-10, female mandible x75; 11-14, female scutellum, squamae, and mucro, not to scale; 15-17, female face, not to scale. 226 Journal of Hymenoptera Research 18. cladothricis 19. siccatus 20. inyo 21. oraibi 33. pallidus 34. mescalero 35. timberlakei 36. oraibi 37. istam 38. longiceps 39. tuktum 41. potawatomi 42. vierecki 43. bridwelli Figs. 18-24, female face x50; 25-26, female hindtibia and metatarsus x50; 27, male antenna xlOO; 28-43, female squamae and mucro, dorsal xlOO. Silvery scales on face and squamae shown as dark setal areas. J. HYM. RES. Vol. 3, 1994, pp. 227-231 Myosomatoides Gen. Nov., (Hymenoptera: Braconidae), A Neotropical Larval Parasitoid of Stem-borer Pests, Diatraea (Lepidoptera: Pyralidae) Donald L. J. Quicke Department of Biology, Imperial College at Silwood Park, Ascot, Berks SL5 7PY, England Abstract.-A new genus, Myosomatoides Quicke (type species: Myosoma pennipes Westwood 1882), from South America is described and illustrated. Ipobracon pennipes Myers 1931 (non Westwood), originally recorded as a parasitoid of an unspecified Diatraea species from British Guiana and subsequently reared from D. angustella Dyar, is congeneric with Myosomatoides pennipes (Westwood) comb. nov. and is therefore a junior subjective homonym of the former. A new name, Myosomatoides myersi is proposed for Myers's species and a key is provided to enable its separation from M. pennipes (Westwood) comb. nov. INTRODUCTION Diatraea, a large genus of New World pyralid moths, whose larvae bore in the stems of graminaceous plants, includes several species of considerable economic importance in both the temperate and particularly the tropical regions where they attack sugarcane, corn and sorghum (Elias 1970, Chippendale 1979, Agnew et al. 1988, Pashley et al. 1990). Several genera of parasitic wasps of the braconid subfamily Braconinae have been recorded as attacking species of Diatraea, viz. Bracon Fabricius (Muesebeck 1925, Shenefelt 1978), Digonogastra Viereck (Wharton et al. 1989), Myosoma Brulle (Quicke 1989) and Palabracon Quicke (Quicke 1988a). All of these are believed to be ectoparasitoids attacking relatively late instar host larvae. In addition to these, Myers (1931) described and illustrated three species of Braconinae reared from Diatraea spp. in British Guiana, all of which he placed in Ipobracon Thomson, though he recognized that one of his new species, I. pennipes Myers (non Westwood), was an "isolated species". Ipobracon was subse- quently shown to be a junior synonym of Cyanopterus Haliday (Quicke 1985, 1987) and the limits of that genus were restricted such that the great majority of New World species described under both Ipobracon and Iphiaulax Foerster, and including the other two species described by Myers, actually belong to a distinct, though related genus, Digonogastra Viereck (Quicke 1988b, Wharton et al. 1989). As Myers implied; however, I. pennipes does not belong to Digonogastra, and in fact, it is a member of a new genus near Myosoma Brulle which is described below. By coincidence, Westwood (1882) had earlier described another species belonging to the same new genus under the name of Myosoma pennipes, which therefore becomes a senior, subjective hom- onym of pennipes Myers. In addition to the type specimens of Myosomatoides pennipes (Westwood) comb. nov. (Hope Entomological Collections, Ox- ford) and M. myersi comb, et nom. nov. (Natural History Museum, London: type no. 3C 418) the author has seen seven other specimens belonging to the new genus from Argentina, Brazil, Colom- bia, British Guiana, Ecuador, Paraguay, and Peru, though none of these has any associated host data. Both M. myersi and M. pennipes appear to be wide- spread in South America. Available material suggests that there are only two species of Myosomatoides in South America, viz. M. pennipes and M. myersi nom. nov. A key is provided to enable their separation. Some varia- tion is apparent for both species. In particular, specimens of M. pennipes vary in the extent of the black coloration on the body, the wing coloration, the extent and length of setosity, the degree of compression of the hind leg, the shape of the head, the relative lengths of hindwing veins lr-m and SC+RI, and the relative length of the ovipositor. 228 Journal of Hymenoptera Research Because no two specimens examined were found to be entirely similar, I consider it most likely that they represent a single rather variable species rather than a diverse polyspecific complex though the latter possibility cannot be completely ex- cluded. Myers (1931) provided a fairly complete de- scription of M. myersi (= pennipes Myers non Westwood), in which he illustrated the rather distinctive metasoma. Following Myers' original record, the species was recorded as a parasitoid of Diatraea, on several further occasions (Myers 1932, Thomson 1953) and, in particular, of D. angustella Dyar in Argentina (Parker et al. 1953, De Santis & Esquival 1966). Members of the related genus Myosoma are similarly idiobiontectoparasitoids of stem-boring pyralid moth larvae, including Diatraea and Chilo species (Quicke 1989, Quicke & Wharton 1989). At least one species of Myosoma is gregarious. Myosomatoides is being described here to make the name available to biocontrol workers in South and Central America and for a forthcoming manual to the New World genera of Braconidae (Marsh & Wharton, in preparation). Terminology follows that of Achterberg (1979, 1988). MYOSOMATOIDES Quicke gen. nov. Type species: Myosoma pennipes Westwood, 1882 Diagnosis. — Superficially similar to Myosoma except that the 2nd metasomal tergite is strongly pinched-up in the middle to form a mid-longitudi- nal ridge (Fig. 4). In addition, the hind femur and tibia are extremely broad and laterally compressed and have very long, dense setosity (Fig. 3). In Quicke & Sharkey's (1989) key to the North Ameri- can genera of Braconinae, Myosomatoides will key to Bracon Fabricius, from which it can be distin- guished by its extremely compressed hind femur and tibia. Description. — Head. Antenna with approxi- mately 50 flagellomeres. Terminal flagellomere acuminate. Medial flagellomeres wider than long. Scapus small, shorter ventrally than dorsally in lateral aspect, emarginate apico-laterally, not emar- ginate apicomedially . Clypeus very shallow; with- out a transverse carina separating reflexed ventral part from dorsal part; dorsal part punctate. Clypeus separated from face by weak groove. Malar area weakly impressed, punctate. Face densely setose except for smooth, shiny, glabrous supraclypeal triangular area. Antennal sockets approximately level with middle of eye. Frons very weakly im- pressed with indistinct midlongitudinal sulcus; with moderately dense short setosity. Mesosoma. Mesosoma smooth, shiny. Notauli indicated only by weak depressions at anterior of mesoscutum. Scutellar sulcus narrow, smooth. Propodeum simple, without carinae. Forewing. Veins 1-M and 1-SR+M straight, vein cu-a interstitial (or virtually so), the 2nd sub- marginal cell rather long (i.e. vein 3-SR 0.67-0.95 times SRI), vein m-cu less than or equal to 0.5 times length of 2-SR. Hindwing. Vein lr-m more or less straight, slightly shorter than, to slightly longer than, vein SC+R1. Apex of vein C+SC+R with one especially thickened bristle. Base of wing evenly and densely setose. Vein 2-1A absent. Legs. Claws with large, acutely pointed, basal lobes. Telotarsi swollen in dorsal aspect compared to basal tarsal articles. Hind leg and often mid-leg (though less so) extremely laterally compressed, tibia markedly expanded in lateral aspect (Figs. 1,2,4). Metasoma. Smooth, shiny, sparsely setose. First metasomal tergite rather long, approximately 1.5-2.0 times longer than wide, and without cari- nae. Second tergite transverse, with midlongitudinal carina. Third to seventh tergites without transverse grooves subbasally or sub- posteriorly, without anterolateral areas, and mem- branous posteriorly. Hypopygium not extending beyond apex of metasoma, pointed in lateral as- pect. Ovipositor (part exserted beyond apex of metasoma) approximately two-thirds length of metasoma, with a pre-apical dorsal nodus and apicoventral serrations. Male genitalia. Digitus with two rather small, well separated apical teeth. Parameres not extend- ing beyond base of digitus; setation restricted to a fairly narrow band at apex. Volsellae virtually glabrous. Basal ring moderately produced medioanteriorly, approximately as long as wide. Internal anatomy. Rectum small with four circular rectal pads. Male with a pair of weak, pouch-shaped, intertergal gland reservoirs be- tween abdominal tergites 7 and 8, and with similar but weaker intertergal glands between tergites 6 Volume 3. 1 994 229 Figs 1-4. Myosomaloides penmpes (Westwood) gen. et comb, nov., Light photomicrographs of male from Brazil. 1. Habitus, lateral aspect. 2. Detail of hind leg. 3. Head and mesosoma lateral aspect. 4. Metasoma and hind legs, dorsal aspect. Scale lines: 1=2 mm; 2=0.6 mm; 3=0.7 mm, 4=0.8 mm. 230 Journal of Hymenoptera Research flattened hind tibiae and general habitus simply represent convergence cannot be totally excluded. As with Myosomatoides, members of the genus Myosoma are parasitoids of lepidopterous grass- stem borers including Pyralidae and Sesiidae (Maeto 1992). ACKNOWLEDGMENTS and 7; abdominal tergum 8 without obvious glands. Etymology. — Name based on the existing ge- neric name Myoscma and the Greek suffix -oides indicating close affinity; gender masculine. Distribution. — M. myersi: Argentina, Brazil, British Guiana, Colombia, Ecuador, and Paraguay; M. pennipes: Brazil (Amazonia) and Peru. Biology. — M. myersi nom. nov. is a larval para- sitoid of members of the pyralid borer genus Diatraea. Nothing is known of the biology of M. I would like to thank Tom Huddleston (The Natural pennipes (Westwood). History Museum, London), Paul Marsh (USNM), Chris ' O'Toole (Hope Entomological Collections, Oxford), L.Stange (Florida State Collection of Arthropods) and David Wahl (American Entomological Institute, Gainesville) for allowing KEY TO THE SPECIES OF MYOSOMATOIDES me access to specimens in their care. This work was partly supported by the American Entomological Institute, a scien- 1 Wings uniformly brown tific exchange visit to Budapest funded by the Royal Society, M. muer si nom. nov. dr,d by a research grant jointly to DLJQ and Mike Fitton (BMNH) from the NERC. — Wings largely hyaline with the apical third of the forewing brown (Fig. 1) LITERATURE CITED M. pennipes (Westwood) Achterberg, C. van, 1979. A revision of the subfamily Zelinae auct. (Hymenoptera, Braconidae). Tijdschrift voor Entomologie 122: 241-479. Achterberg, C. van, 1988. Revision of the subfamily Blacinae Foerster (Hymenoptera, Braconidae). Zoologische Verliandelmgen, Leiden 249: 1-324. Agnew, C.W., Rodriguez-del-Bosque, L. A., Smith, J. W. Jr. 1988. Misidentification of Mexican stalkborers in the subfamily Crambinae (Lepidoptera: Pyralidae). Folia Entomologica Mexicana 75: 63-75. Chippendale, G.M. 1979. The southwestern corn borer, Diatraea grandiosella: case history of an invading insect. Bulletin of the Missouri Agricultural Experiment Station, no. 1031, 1-52. De Santis, L., Esquivel, L. 1966. Tercera lista de himenopteros parasiticos y predatores de los insectos de la Republica Argentina. RevistadelMuseo de La P/ _■• ,, .• i .• i ■ lack of reliable studies on the structure of the thorax" for understanding phylogenetic relationships, a Snodgrass 1910 p 37 toPlc recently addressed by various authors (Rasnitsyn 1969, 1980, 1988, Brothers 1975, Since Snodgrass (1910) first attempted to ex- Konigsmann 1977, 1978a 1978b, Carpenter 1986, pand our knowledge of the structure of the hy- Brothers ^d Carpenter 1993). menopteran thorax, additional studies have de- The mesofurca is an invagination of the ster- scribed the skeleto-musculature of single species num mto the thorax that forms a central Point of or single families (Weber 1925, 1927, Tulloch 1935, attachment f°r the ventral longitudinal muscles, Maki 1938, Duncan 1939, Michener 1944, Bucher the sterno-pleural muscles, the coxal and trochant- 1948, Alam 1951, Saini et al. 1982, Daly 1964, eral muscles' and the mesopostnotal muscles Gibson 1986, 1993). Fewer studies have compared (KelseY 1957< Matsuda 1970). The mesofurca is thoracic structures among families of Hy- comPnsed of a basal plate (discrimenal lamella) menoptera (Snodgrass 1942, Daly 1963, Rasnitsyn that rises vertically f™ the discrimen, slopes posteriorly to the furcal base, and divides dorsally 242 Journal of Hymenoptera Research into two lateral arms, termed the sternal apophy- ses (Snodgrass 1927, Chapman 1992, Lawrence et al. 1992). The mesopostnotum is one of the pri- mary dorsal sclerites involved in flight through the posterior inflection of the antecosta (second phragma) which forms the posterior attachment of the longitudinal flight muscles. Recent studies involving skeleto-musculature of the hy- menopteran thorax have focused on the pleural attachments (Shcherbakov 1980, 1981, Gibson 1985, 1993), the extrinsic musculature of the mesocoxa (Johnson 1988), and the development of the metapostnotum (Whitfield et al. 1992). Rasnitsyn (1969, figs. 187-194) was the first person to com- pare the different skeletal structures for the mesofurca of 8 families of Symphyta. His illustra- tions show the transformation series for Symphyta that are discussed in this paper. Rasnitsyn (1988) refers to the furca for features supporting Tenthredinoidea and for Cephoidea + Siricoidea + Apocrita. Similarly, Snodgrass (1942) presented a pictorial evolutionary history for development of the axillary lever of Apoidea. This work expands upon these initial studies and extends the com- parative aspects of these works to include most families of Apocrita. This study of the mesofurca and mesopostnotum began as an attempt to under- stand the polarity and homology of mesofurcal structures and muscles of Aphelinidae (Chalcidoidea) and the phylogenetic implications of these attributes within the Chalcidoidea. Even- tually the entire Hymenoptera needed to be sur- veyed to resolve what we initially thought were relatively simple questions. In this study, all muscles attaching to the mesofurca and mesopostnotum are identified and compared to homologous muscle groups in Neuropterida and Mecopterida (sensu Kristensen 1992), as they are considered to be phylogenetically close to Hy- menoptera (Kristensen 1992), and have a mesotho- rax which is structured similar to the Symphyta. Within Hymenoptera, we have concentrated our analysis on the skeletal structure of the mesofurca and mesopostnotum, and on the muscles attach- ing between the thoracic furcae and the laterophragma of the mesopostnotum. The evi- dence provided by the mesofurca and mesopostnotum for relationships within the Chalcidoidea will be discussed in a subsequent paper. MATERIA! S AND METHODS Terms for structures and muscles generally follow Snodgrass (1910, 1942), Daly (1963), Gibson (1985, 1986, 1993) and Ronquist and Nordlander (1989). Muscles were identified using the systems proposed by Kelsey (1957) and Daly (1963) (Table 1). Figures 1 and 2 are used to place the skeleto- musculature within the context of the mesosoma. Muscles and stuctures are extensively labeled in Figs. 3 and 4. The Kelsey system uses a fixed set of numbers and is useful for comparisons across the Endopterygota. Daly's system is preferred for clar- ity because the insertion-origin of attachment sites are readily identified and new muscles can be added to the system; for example, the new muscle fu,-ba3 was given the abbreviation fbl for the Kelsey system as it could not be assigned a nu- meric value that would signify its relative position to other muscles in the mesothorax. Terms pro- posed by Matsuda (1970) are comprehensive and may be referenced across orders of insects; how- ever his abbreviated system is difficult to use and is not followed here. Several new classifications of families within Hymenoptera have been proposed recently that differ largely in placement of certain families as separate superfamilies, families, or subfamilies. We follow the classification of Huber and Goulet (1993), as it represents the most current synthesis of information across the order. Dissections were based on specimens pre- served in 70%ethanol or initially fixed in Dietrich's or Kahle's solution and then transferred to etha- nol. All specimens were critical point-dried prior to dissection. The mesosoma of Monomachns (Monomachidae) was rehydrated using Barber's solution, transferred through increasing concen- trations of ethanol to 98% and then critical point- dried. For each dissection, the mesosoma was anchored onto a standard SEM stub using chloro- form-based silver paint. Dissections were made using hooked minuten pins or fragments of razor blades. Dried haemolymph and extraneous tis- sues were removed from dissections using small amounts of glue obtained by dragging a hooked minuten pin across clear sticky tape (Gibson 1985). Exemplar taxa were chosen to represent the maximum variation within taxa. In some groups (e.g. Apoidea), there was virtually no variation; whereas within some taxa (e.g. Diapriidae) both Volume 3. 1994 243 structure and presence of muscles varied and more genera were dissected to characterize this varia- tion. Our primary concern was for establishing groundplan states for higher taxa, although autapomorphies are discussed. The taxa exam- ined for internal characters are listed in Table 2. Numerous Chalcidoidea were also dissected as part of a comprehensive study of the mesofurca in that superfamily. Representative dissections are housed at the Canadian National Collection (CNCI), Royal Ontario Museum (ROM) and Texas A&M University (TAMU). When possible, con- specific adults of the dissected specimens are de- posited as voucher specimens in the above collec- tions. The majority of specimens were obtained from the CNCI alcohol collection. The mesofurcal-mesopostnotal complex was broken up into 12 characters with a total of 36 character states. Seventy hymenopteran taxa were scored (Appendix 2) based on the examination of internal characters for 119 species (Table 2). A single outgroup taxon was scored based on dissec- tions of 5 families of Neuropterida and 3 families of Mecopterida. Characters 1, 3 and 9 are postu- lated as unique characters for Hymenoptera; in each of these cases, the outgroup is coded as a unique character state (state 0). The state value "?" was used to denote uncertain homology, not miss- ing data. Illustrations were made with a camera lucida. Outlines of muscles present but not illustrated are represented by dashed lines. Some muscles not central to this study were not consistently figured (e.g. muscle 180 for Symphyta) and caution should be exercised in deriving additional interpretation from the illustrations. Abbreviations referring to muscles are circled in all figures, skeletal charac- ters are not. The mesofurcal-mesopostnotal com- plex is abbreviated as MF-MPN complex. RESULTS AND DISCUSSION Character Analysis Character 1. Mesofurcal bridge The most significant modification of the mesofurca in Hymenoptera is the fusion of the elongate anterior arms (af) into a mesofurcal bridge (fb) that characterizes virtually all Apocrita. In basal groups of Symphyta (excluding Cephoidea + remaining Hymenoptera), the anterior furcal arms are either absent, short, or elongate and well separated along their entire length. The anterior arms, or the equivalent region on the lateral arms of the mesofurca, form the posterior attachment sites for the ventral intersegmental muscles (muscle 124, fiij-fUj, and muscle 127, fu-rsps,). With fusion of the anterior arms, the furca separates the gastric and nervous systems and the ventral nerve cord passes through the foramen bounded by the mesofurcal bridge and the lateral arms of the mesofurca. In most Neuropterida and Mecopterida, the interfurcal muscles attach directly to the anterior face of the lateral arms of the mesofurca (state 0, no arms) (Snodgrass 1927, Kelsey 1957, Matsuda 1970). In Xyelidae (Fig. 3) and Pamphiliidae (Fig. 6) (and probably Megalodontidae, cf. fig. 187, Rasnitsyn 1969), the interfurcal muscles attach to anterior projections (af, anterior furcal arms) that are long, robust, and separated along their entire length (state 1). Of the outgroup taxa examined, only Brachynemurus (Myrmeleontidae) have anterior arms (state 1) similar to Xyelidae, and thus their presence could be plesiomorphic or apomorphic for Hymenoptera. Because anterior arms are present only in a derived member of the outgroup, we consider the presence of long anterior arms in Xyeloidea and Megalodontoidea to be apomorphic for Hymenoptera. In Tenthredinoidea, the anterior arms (af) can be reduced in size (Cimbicidae Fig. 8, Blasticotomidae Fig. 20, and Nematus (Tenthredinidae) Fig. 21), modified into support- ing cup-like structures (Diprionidae, Fig. 7), or lost entirely so that the interfurcal muscles attach to the anterior face of the lateral furcal arms ( Argidae, Pergidae Fig. 5, and most Tenthredinidae). Rasnitsyn (1988) treated "fore arm short" as a character state (his 2-3d) supporting Tenthredinoidea and "fore arm reduced" (his 6- 8a) for Argidae + Pergidae. Short arms, as in Blasticotomidae (Fig. 20), are probably plesiomorphic for Tenthredinoidea, and further modifications of the arms or complete loss are derived within Tenthredinidae. It is possible to code for several different character states within Tenthredinoidea (small, absent, cup-shaped, etc.), but this would only introduce unnecessary ho- moplasy into the analysis (i.e. "absence" derived 2 or more times) or a series of autapomorphic char- 244 Journal of Hymenoptera Research Table 1. Homology and terms for muscles of the mesopostnotum and mesofurca examined in Neuroptera and Hymenoptera Letter m added to distinguish metathoracic muscles. Muscle fbl described in text. Kelsey 1957 — Con/dalus No Name Hymenoptera Matsuda Daly Johnson Present 1970 1964 1988 designation Notes Mesothoracic dorsal indirect muscles 112 internal (ventral] longitudinal tl4 lph-2ph 112m metathoracic longitudinal tl4 2ph-3ph 114 scutellar-metanotal tl3 t,-t, lph-2ph longitudinal flight muscle Kelsey (1957) recognized both ventral (internal, 112) and dorsal (external, 113) muscles Recognition of the two muscles in Hymenoptera is unnecessary 2ph-3ph reduced and attached laterally in most Hymenoptera t,-t, in Hymenoptera, paired medial muscles passing from meta- notum (t,) to scutellum over mesopostnotum (PN,), external in some sawflies From PN; in Neuroptera 116 second dorsal diagonal Mesothoracic ventral muscles 124 mesothoracic interfurcal 127 mesofurcal-spina t,-2ph sl3 fu,-fu, sl4 fu,-lsps t,-pn, dorsomedial attachment on mesoscutum to anterior face of laterophragma (pn,), attachment to dorsal axillar surface where transscutal arhculahon present fu,-fu, from lateral face of fu, to anterior face or arms of fu, Kelsey (1957) treats as three muscles (124-126) and Matsuda (1970) as bundles of the same muscle Hymenoptera have maximum of two bundles and homology of each is uncertain fu,-sps, spina of profurca to apex of lateral arm of fu,. 181 metathoracic interfurcal sl3 fuj-fu, • fu,-fu, posterior face of lateral arms of fu; to fu, Kelsey (1957) treats as three muscles (181-183) Hymenoptera have maximum of two 180 metafurcal-spina sl4 fUj-spSj spina of mesofurca to apex of lateral arm of fu, Tergopleural muscles 137 laterophragmal-basalare t-p87 Furcal muscles 150a posterior furcal-laterophragmal t-sl fu,-2ph 150b anterior furcal-laterophragmal t-sl 151 furcal-pleural arm 170 coxal arhculation-furcal fbl furcal-basalare p-sl pl,-fu, s-cx2 p-s3? fu2-pl37 t -ba, posterior face of laterophragma ( = t,) to apodeme of basalare (ba,) Synonomy with t-p8 is questionable fu,-pn, posterior attachment on lateral arm of fu, to anterior process of laterophragma (ap) fu,-pn, attachment anterior to 150a on anterior or lateral arm of fu, to posterior lobe of laterophragma (pn,) pl,-fu,_ lateral surface of fu, to pleural ridge pl,-fu,t lateral surface of fu2 to coxal process on pleuron fu,-ba, apex of lateral arm of fu2 to basalare (ba,), may be homologous with pupal muscle of Apis (Daly 1964), well developed with dorsal cap and apodeme in Dipnon (one preparation where ba, dissected with fu;-ba, and pl,-ba, both attached). Possibly homologous with Matsuda's p-s3 from furca to anterior margin of succeeding episternum Walking and indirect flight muscles 169 anterior furcal-coxal s-cx6 fu,-cx fu-cx fu,-cx 173 posterior furcal-coxal s-cx3 fu2-cx fu-cx fu2-cx. anterior base of fu; to median nm of coxa. attachment on fu, posterior to muscle 169 to posteromesal rim of 174 furcal depressor of trochanter s-trl hy h", fu,-tr, anterior base of fu; (or arms) to trochanteral apodeme Volume 3, 1 994 Table 2. Taxa dissected for study of the mesofurcal-mesopostnotal complex. 245 Section: NEUROPTERIDA Mantispidae Chrysopidae Corydalidae Myrmeleontidae Rhaphidiidae Mantispa sp. Chrysopa sp. Corydalus sp. Brachynemurus sp. Rhaphidia sp. Section: MECOPTERIDA Bittacidae Bittacus sp. Meropeidae Merope tuber Newman Panorpidae Panorpa sp. Order: HYMENOPTERA SYMPHYTA Xyeloidea Xyelidae Macroxyelinae: Macroxyela ferruginea (Say); Xyelinae: Pleuroneura sp., Xyela minor Norton Megalodontoidea Pamphilndae Cephalciinae: Acantholyda sp., Pamphiliinae: Pamphilius sp. Tenthredinoidea Blasticotomidae Tenthredinidae Dipnonidae Cimbicidae Pergidae Argidae Cephoidea Cephidae Siricoidea Anaxyehdae Siricidae Xiphydrioidea Xiphydriidae Orussoidea Orussidae APOCRITA Stephanoidea Stephanidae Blasticotoma sp. Heteranthinae: Profenusa canadensis (Marlatt); Nematinae: Nematus sp.; Selandriinae: Aneugmenus flavipes (Norton), Strongylogaster fflcirn(Norton); Tenthredininae: Filacus sp., Macrophya sp. Dipnoninae: Diprion similis (Hartig) Cimbicinae: Cimbex americana Leach; Ambnnae: Zaraea americana Cresson Acordulecerinae: Acordulecera sp., Syzygoninae: Lagideus Imexicana Smith Arginae: Arge sp., Durgoa matogrossensis Mai. Cqjhus cinctus Norton, Hartigia trimaculata (Say) Syntexis libocedrii Rohwer Siricinae: Urocerus albicornis (Fabricius); Tremicinae: Tremex columba (Linnaeus) Xipiudria abdommalis Say Orussus terminalis Newman Megischus bicolor (Westwood) Trigonalyoidea Trigonalyidae Ceraphronoidea Megaspilidae Ceraphronidae Evanoidea Aulacidae Evaniidae Gasteruptiidae Ichneumonoidea Braconidae Ichneumonidae Chrysidoidea Plumarndae Sclerogibbidae Embolemidae Dryinidae Bethylidae Chrysididae Orthogonalys pulchella (Cresson) Megaspilus fuscipennis Trichosteresis sp. Ceraphron sp. (Ashmead), Prislaulacus strangahae (Rohwer) Hyptw thoracica (Blanchard), Evania sp. Gasteruptwn sp. Alysiinae: Coelwius sp.; Hybrizontinae: Hybrizon sp.; Macrocentnnae: Macrocentrus sp.; Miscogastennae: Apanteles sp.; Rogadinae: genus ?. Ephialtinae: genus ?, Scambus sp., Megarhyssa sp.; Ophioninae: Enicospilus sp. Plumana sp. Probethylus sp. Etnbolemus nearcticus (Brues) Anteoninae (female), Gonatopodinae (male) Anisepyris sp., Epyris sp. Amiseginae: Adelphe anisomorphae Krombein; Chrysidinae: Chrysis sp., Parnope sp.; Cleptinae: Cleptes sp. Vespoidea Tiphiidae Sapygidae Mutillidae Sierolomorphidae Pompilidae Rhopalosomatidae Bradynobaenidae Formicidae Scoliidae Vespidae Apoidea Crabronidae Heterogynaeidae Myzinum sp. Sapyga sp. Sphaeropthalminae (males, 3 genera) Sierolomorpha canadensis Provancher Pepsinae: Calicurgus hyalinatus Fabr.; Pompilinae: Aporinella galapagensis Rohwer, Aporus sp. Rhopalosoma sp. Bradynobaenus sp. Myrmicinae: Solenopsis invicta Buren (queen & worker); Formicinae: Camponotus planus F. Smith (queen), Paratrechina sp. (queen) Scolia sp. Eumeninae: Odynerus sp., Parancistrocerus sp.; Vespinae: Dolichovespula sp., Vespula sp. Ectetnnius sp., Larra sp. Heterogyna sp. continued on next page 246 Journal of Hymenoptera Research Table 2 continued Andrenidae Anthophondae Apidae Halictidae Megachilidae Proctotrupoidca Diaprndae Monomachidae Vanhorniidae Helondae Ropronhdae Proctotrupidae Pelecimdae Cynipoidea Ihalndae Eucoihdae AUoxvstidae Platygastroidea Scelionidae Platygastndae sP. Andrena sp. Nomada sp., Ceratma sp. Apis mellifera L., Bombus sp., Trigona sp. Agnpostemon sp., Laswglossum sp. Megachile sp. Ambositrinae: Dissoxylabis sp.; Belytinae: Aclista sp., Acropiesta sp., Aneurrynchus sp., Oxylabis sp., Diaprunae: Coptera sp., Paramenia sp. Spilomicrus sp., Trichopna sp. Monomachus sp. Vanhornia eucnemidarum Crawford Helorus sp. Roprorua sp. Exallonyx sp., Mwta vera (Fouts) Pelecinus polyturator (Drury) Jbfl/ifl sp. genus 7 Alloxysta sp. Scehonmae: y4n/cris sp., Calotelea sp., Oemasfobaeus sp., Gryon sp. (wingless), Macrcteleia absona Muesebeck, Sparasion sp., Teleasinae: Tnmorus sp. Inostemmatinae: hocybus sp., Inoslemma Mymarommatoidea Mymarommatidae Palaeomymar sp Chalcidoidea (additional dissections for work in progress not listed) Mymandae Gonafoccrus sp. Pteromalidae Cleonvminae: Cleonymus sp., Oodera sp. acter states. Coding of character states must in part reflect the level of analysis. At a different level, for example in an analysis of the relationships of Tenthredinoidea that is associated with a more extensive survey of taxa, itmightbe appropriate to further partition the various shapes. These modi- fications are difficult to characterize and here we have combined them into one apomorphic state, the reduction of well-separated anterior arms (state 2). A clear transformation series leading to the development of the mesofurcal bridge is found within the Symphy ta beginning with the Cephidae. In Cephidae (Fig. 9), the anterior arms are narrow and elongate and only the extreme apices of each arm are joined (state 3). The anterior arms are considered to be homologous with those of Xyeloidea and Megalodontoidea; therefore, state 3 is probably derived from state 1. Rasnitsyn (1988; his character 19b) includes Cephoidea with Siricoidea + Vespina based on having the fore- arms of the mesofurca long and fused for some distance. Although true for Siricoidea + Vespina, the arms are only apically fused in Cephidae. In Anaxyelidae (Fig. 10) and Siricidae (Fig. 11), the anterior arms are elongate and fused along most of their length (state 4). The anterior arms of Siricidae are laterally flattened (Fig. lib) and in dorsal view each arm can be distinguished along its entire length (Fig. 11a). Anaxyelidae have a similar structure, including anterior placement of muscle 150b, but fusion of the arms is more com- plete (Fig. 10a). Posteriorly, the anterior furcal arms of Anaxyelidae are connected by a thin hori- zontal plate of cuticle. InXiphydrndae(Fig. 12), theanteriorarmsare completely fused and form a transverse mesofurcal bridge (state 5), and the entire dorsal and lateral surfaces form the posterior attachment for muscle 124. Displacement of muscles 150a and 150b to the lateral arms suggests that fusion of the arms in Xiphydriidae may be independent of the fusion in Siricidae and Anaxyelidae. Orussidae (Fig. 13) exhibit complete fusion of the anterior arms into a smooth and bowed mesofurcal bridge (state 6) with a strong median anterior process that is the attachment site for muscles 150 and 124, as occurs in Anaxyelidae and Siricidae. The groundplan states for the mesofurca of Apocrita consists of 1) a complete mesofurcal bridge (as in Xiphydriidae and Orussidae), 2) an anterior medial projection supporting muscle 124 (fu^-fUj) (as in Xiphydriidae and Orussidae), and 3) lateral displacement of muscle 150a (fu,-pn, ) (as in Xiphydriidae). No Apocrita have muscle 150 originating on the anterior projection of the mesofurca as in Orussidae. Because of the lateral displacement of muscle 150a, which is similar to Xiphydriidae (see character 2), Apocrita with a mesofurcal bridge are coded as character state 4. The mesofurcal bridge is absent (state 7) in Ceraphronoidea (Fig. 25), Pelecinidae (Fig. 31), some Chalcidoidea (including all Mymaridae), Mymarommatidae, and Platygastridae (Fig. 28). We consider that absence of the bridge is an Volume 3, 1 994 247 apomorphic loss of the type of bridge found in Xiphydriidae (state5), which is similar to the type found in most Apocrita. Assuming that presence of a bridge is a groundplan state for Apocrita (whether states 4, 5 or 6), losses within each of these taxa are considered irreversible (unlikely that a bridge can be regained) and probably inde- pendent. In Ceraphronidae, Pelecinidae, Mymarommatidae, Platygastridae, and some Mymaridae, the mesofurca is lyre-shaped with the lateral arm terminating in a cup-shaped process that supports muscle 150a. When the furcal bridge is lost in Chalcidoidea (Aphelinidae, Encyrtidae, Rotoitidae, Signiphoridae and Tricho- grammatidae), the shape of the mesofurca and attachment of muscle 150a are considerably differ- ent. Synonymy for mesofurcal bridge: Siricidae & Vespidae: mesofurcal ring (Tenthredinidae, Vespa, Weber 1925). Ichneu- monoidea: mesofurcal bridge (Stenobracon, Alaml951). Aculeata:mesofurcalbridge(ArF2) (Vespula, Duncan 1939); supraneural bridge (Apis, Snodgrass 1942). Chalcidoidea: arch of the furca (Monodontomerus, Bucher 1948); ten- dinous arch of the mesofurca (Tetramesa, James 1926). Character 2. Laterophragma of mesopostnotum (pn2) In Neuropterida, Mecopterida, Xyelidae and Pamphiliidae, the laterophragma of the mesopostnotum forms a lobe (pn,, Figs. 4, 6) (state 0) that extends obliquely into the mesothorax from the lateral corners of the mesopostnotum, mediad of the lateral attachment to the upper mesepimeron and lateral to the second phragma. The laterophragma in all of these taxa forms the attach- ment site for muscle 150 (fu2-pn,, see Character 3), muscle 1 16 (t2-pn„ see Character 4) and muscle 137 (pnj-ba,). The posterior face of the laterophragma forms the attachment site for muscle 137 (pn,-bav Fig. 3). Muscle 137 is usually small and difficult to trace, but it is apparently lost in Pergidae and all Apocrita. In Xyelidae (Figs. 3, 4) and most Symphyta (Figs. 6, 8, 9), an apodeme (ap) is present on the anterolateral margin of the lobe that serves as the attachment site for muscle 150a (fu,-pn2 ). The laterophragma is functionally coupled with muscles 150 and 116, and the loss of either of these muscles is associated with a corresponding change in structure of the laterophragma. To avoid dupli- cation of character coding, the presence or absence of certain structures of the laterophragma are treated under other characters. For example, the anterior apodeme is lacking in Neuropterida and Mecopterida but this was not coded as a different state for this character because it is reflected in the differentiation of muscle 150a in Hymenoptera, which is dealt with as Character 3. A broad laterophragma with a small anterior apodeme (state 0) occurs in Xyelidae, Pamphiliidae (and probably Megalodontidae), Blasticotomidae, Tenthredinidae, Argidae, Diprionidae, and Cephidae. The laterophragma of Cimbicidae (Fig. 8) pos- sesses an exaggerated apodeme (ap) and an en- larged posterior lobe (pn,), which is fused with the second phragma (2ph). The laterophragma is unique in form and apparently autapomorphic. In Xiphydriidae (Fig. 12), Monomachidae (Fig. 14), Vanhorniidae (Fig. 15), Cynipoidea (Fig. 22), Ceraphronoidea (Fig. 25) and most Diapriidae (Fig. 29), the laterophragma is excised between the elongate apodeme (ap) and the posterior lobe (pn,) (statel). In most of these taxa the apodeme of the laterophragma extends medially and horizon- tally into the thoracic cavity, but in Cynipoidea (cf . Fig. 22), the apodeme is vertical. The posterior lobe is lost in most Apocrita (incuding some Diapriidae), but this was not coded as an additional state change for this characterbecause it reflects the loss of the muscle 116 (Character 4). The anterior apodeme (=axillary lever) or associated attach- ment for muscle 150a is present in all Apocrita. Additional character state changes for the laterophragma in Apocrita are based only on changes in the shape of the anterior apodeme. In Pergidae (Fig. 5) and Orussidae (Fig. 13), the laterophragma is reduced to a narrow triangu- lar process that forms an attachment for the ten- don of muscle 150 (state 2). This reduction is also associated with the loss of muscle 116. State 2 could be derived from either state 0 or state 1. In Apocrita, the axillary lever (ap) occurs in a variety of shapes that probably have different effects on leverage of the laterophragma with re- spect to the fourth axillary sclerite and the second phragma. In all Apocrita the apex of the axillary lever maintains a connection with muscle 150a 24S Journal of Hymenoptera Research (fu^-piv )• The plesiomorphic lever for Apocrita is narrow and inflected medially into the thorax as found in Xiphidriidae (statel) and occurs either in association with a posterior lobe (most Diapriidae, Monomachidae (Fig. 14), Vanhorniidae (Fig. 15), Ceraphronoidea (Fig. 25) and Cynipoidea (Fig. 22) or without the posterior lobe (Tngonalyidae (Fig. 23), Stephanidae (Fig. 24), Diapriidae (Fig. 29), Proctotrupidae (Fig. 32), some Ichneumonidae, Mymarommatidae, and Chrysidoidea (Figs. 34- 39) (except some Chrysididae). The axillary lever exists in a variety of forms in Apocrita but in shapes that are difficult to separate into discrete states, especially without a more thorough survey of the apocritan taxa. In all cases where the lever was narrow and inflected medially the character was coded as state 1. In Platygastridae (Fig. 28), some Scelionidae and most Chalcidoidea, the axillary lever is de- flected ventrally (state 3). In Mymarommatidae, the lever is robust and inflected medially (statel). In Sparasion (Scelionidae; Fig. 16), the lever is reduced to a small cup-shaped lobe on the laterophragma (autapomorphic and not coded). In other Scelionidae and Mymaridae the lever is short, narrow and medially inflected (statel). The axillary lever in Stephanidae (Fig. 24) and most Chrysidoidea (Figs. 35-38, also in Fig. 34 but this view is slightly different) is elongate and strongly inflected medially (state 1). This confor- mation of the lever is likely the precursor to the lever found in Apoidea and Vespoidea, discussed below. However, in Cleptinae and Chrysidinae (Chrysididae) the lever is reduced (state 4) form- ing a short, broad process attaching to a broad tendon of the shortened muscle 150a (Fig. 40). In Apoidea (Figs. 45, 46, 54) and Vespoidea (Figs. 41, 42, 44) the axillary lever is robust and strongly appressed to the inner surface of the second phragma (state 5). This type of lever was found in all Apoidea examined and is associated with a robust muscle 150a. Except in Bradynobaenidae and Formicidae, muscle 150a is conical and attached to the lever by a narrow tendon or robust and nearly tubular (Mutillidae and Sapygidae). In Sierolomorphidae, the lever is not closely appressed to the second phragma (in- flected medially about 30°), but this was not treated as a different state (intermediate between states 1 and 5). In Bradynobaenidae, the lever is reduced to a rounded knob (attaching to a narrow tendon of muscle 150a) (state 4?). In Formicidae (Fig. 43), the lever forms a narrow, twisted apodeme (state 1?). Neither Bradynobaenidae and Formicidae were broadly surveyed and other forms may exist. In Ichneumonidae the axillary lever is either closely appressed to the second phragma and similar to the lever found in Vespoidea and Apoidea (state 5), or rod-like and extending medi- ally into the mesolhorax (state 1), or inflected medially and dorsally (Fig. 17a). A robust lever is not found in Braconidae, including Hybrizontinae (Fig. 33), in which it was knob-like or short and slightly deflected ventrally (state 1), similar to Trigonalyidae and some Proctotrupoidea. The le- ver of Gasteruptiidae (Fig. 26) is similar to some Ichneumonidae but is oriented vertically with re- spect to the second phragma. In Oodera (Pteromalidae), the lever is robust, appressed to the second phragma and horizontal (state 5). This form of the lever appears to be convergent with Ichneumonidae and Aculeata, since in Oodera muscle 150a attaches to the entire ventral aspect of the apodeme as in other Chalcidoidea. All Apiformes have the axillary lever sepa- rated (state 6) as an independent sclerite (Snodgrass 1942). This feature was verified in all of the Apiformes examined here. The apical connection of the mesopostnotum is weak and encased by the cup-shaped basal process of the axillary lever (Figs. 45,54). Synonomy for laterophragma: Symphyta: lateral lobe of second phragma (2ph) (Daly 1963). Icnneumonoidea: muscle bearing process of mesopostphragma (Stenobracon, Alam 1951); axillary lever (Snodgrass 1942). Cynipoidea: mesopostnotal apodeme (Ibalia, Ronquist and Nordlander 1989); 2ph (Daly 1963). Aculeata: axillary le- ver (Apis, Snodgrass 1957; Bombus, Pringle, 1957, 1960, 1961 ); accessory sclerite of the fourth axillary sclerite (Apis, Snodgrass 1910); inner process of mesopostphragma (Vespa, Weber 1925 [states that lever turns apex of mesophragma and the connected axillary 4]); anterior process of mesopostphragma (MPPh J (Vespula, Duncan 1939). Hymenoptera: axil- lary lever (Matsuda 1970). Character 3. Furcal-laterophragmal muscle In Neuropterida and Mecopterida, only a Volume 3, 1994 249 single muscle (150, fu2-pn2) acts between the mesofurca and laterophragma of the mesopostnotum (state 0) (Kelsey 1957, Matsuda 1970). In the majority of Symphyta, including Xyelidae and Pamphiliidae, two distinct muscles (150a and 150b) operate antagonistically on the enlarged laterophragma (Figs. 2, 3, 4, 6). The pre- sumed division of muscle 150 into two muscles is considered as an apomorphy for Hymenoptera (state 1). Muscle 150a (fu,-pn^) originates on the lateral arms of the mesofurca posterior or lateral to the origin of 150b and inserts onto an anterior apodeme or process of the laterophragmal lobe (Figs. 4, 6, 7, 9). Muscle 150b (fu,-pn2 ) originates anteriorly or medially on the lateral or anterior arms of the mesofurca and broadly attaches to the margin of the laterophragmal lobe (Figs. 4, 6, 7, 9). The fibres of muscle 150b are often arranged so that the anteriormost fibres attach to the posteriormost or innermost margin of the laterophragmal lobe, although this is dependent on the size of the muscle and lobe. If both muscles are present, muscle 116 (t,-pn,) is also present. Loss of muscle 150b (state 2) occurs in Cimbicidae, Monomachidae, Vanhorniidae, Diapriidae, Cynipoidea, and Ceraphronoidea, all of which have retained both the laterophragmal lobe (presumably the former point of insertion of muscle 150b) and muscle 116. All Apocrita have lost muscle 150b. In Anaxyelidae (Fig.10), muscle 150b is attached to the posterior face of the poste- rior lobe (versus the margin) and muscle 150a is retained. Because of a similar placement and at- tachment of muscles in Siricidae (Fig. 1 1), it can be postulated that muscle 150a is lost and 150b re- mains (state 3). When both the laterophragmal lobe and muscle 116 are missing, as in Pergidae (Fig. 5) and Orussidae (Fig. 13), it is difficult to assess which muscle, 150a or 150b, has been lost. We could assume that muscle 150b is lost in both families. However, the forward attachment of muscle 150 to the mesofurca in Orussidae suggests that it is homologous with muscle 150b of Anaxyelidae and Siricidae. Therefore, the con- figuration in Pergidae and Orussidae could also be treated as an parallel loss of muscle 150a. Be- cause of this uncertainty, we code the reduction to a single muscle in Pergidae and Orussidae as of questionable homology (state ?). Based on the presence of muscle 150a in the apocritan families mentioned above, we assume that muscle 150b is lost in all Apocrita without a posterior laterophragmal lobe. Autapomorphic losses of muscle 150a occurs in workers of Formica and Camponotus (Markl 1966, Saini et al. 1982). Synonymy for muscle 150: Tenthredinidae: Ilisml (Dolerus, Schizocerus, Tenthredo, Weber 1927), 25 & 26 (Euthomostethus, Maki 1938). Ichneumonidae: 25 (Psilopsyche, Maki 1938); 63 (Stenobracon, Alam 1951). Aculeata: 25 (Vespa, Maki 1938); 78 (Formica, Markl 1 966; Apis, Snodgrass 1942); IIdv2 (Vespula, Duncan 1939); Ilism (Vespa, Weber 1925); fu2-2ph (Apis, Daly 1964). Hy- menoptera: t-sl (Matsuda 1970). Character 4. Second dorsal diagonal (=phragmal flexor) muscle Muscle 116 (t2-pn2) is found attaching between the mesonotum and the anterior face of the laterophragmal lobe in Neuropterida, Mecopterida and most Symphyta (Figs. 2, 3, 12, 14, 15, 20), and is therefore plesiomorphic for Hymenoptera (state 0). Muscle 116, and the associated posterior lobe of the laterophragma, is lost (statel) in Pergidae, Orussidae and most Apocrita. It seems unlikely that a functional complex composed of both muscle 116 and the corresponding posterior lobe of the laterophragma could be regained, and we con- sider the apomorphic state (loss of muscle and lobe) to be irreversible. An assumption of irrevers- ibility has obvious consequences for models of character state change which are discussed in the later section on parsimony analyses. Muscle 116 and the corresponding posterior lobe are present in Monomachidae, Vanhorniidae, Cynipoidea, Ceraphronoidea and most Diapriidae, and it is likely that this complex is the groundplan for Apocrita (Daly 1963; Gibson 1985). The loss of this complex in Pergidae and Orussidae is therefore convergent with the loss in most Apocrita. Synonomy for muscle 116: Symphyta: 21 (Euthomostethus, Maki 1938). Symphyta, Monomachidae, Diapriidae, Vanhorniidae, Ceraphronoidea, Cynipoidea: t,-2ph (Daly 1963; Gibson 1985). Hymenoptera: tl2 (Matsuda 1970). 250 Journal of Hymenoptera Research Character 5. Mesopostnotum and scutellar- metanotal muscle In Neuropterida and Mecopterida, the mesopostnotum is broadly exposed medially; al- though in Corydalus (Megaloptera) the mesopostnotum is weakly sclerotized medially and appears to be split. The scutellar-postnotal muscle (muscle 114, t2-t3; absent in Bittacus) passes internally from its origin on the scutellar area of the mesonorum to the anterior medial margin of the mesopostnotum (Kelsey 1957; Matsuda 1970). Within Hymenoptera, changes in the structure of the mesopostnotum and attachment of muscle 114 are correlated; however, the attachment of muscle 114 to the mesopostnotum is unique with respect to the outgroup, and the outgroup is scored as a separate state (state 0). In Hymenoptera, muscle 114 (t,-t3) arises dorsomedially from a fan-shaped attachment on the mesonorum and passes medially over the mesopostnotum (rarely under), to a narrow me- dial attachment on the anterior margin of the metanotum. In all Symphyta, our dissections indi- cate a posterior attachment of muscle 114 to the metanotum. Our observations are supported by illustrations of attachments to the metanotum in Weber (1925) for Vespa (his IldlmJ and Markl (1966) for Apis (his 70). In Eutomostethus (Tenthredinidae), muscle 114 was described as inserting on the median membrane that divides the mesopostnotum (Maki 1938, his 19 & 20), but in Tenthredinidae muscle 114 passes through the membrane to the metanotum. The attachment of muscle 114 to the metanotum is apomorphic for Hymenoptera. Determination of the groundplan condition for Hymenoptera is complicated by the presence of different character states in the two extant sub- families of Xyelidae. In Xyelinae (Xyelidae) (based on Pleuroneura, Fig. 47, and Xyela), the mesopostnotum (PNJ is broadly exposed dor- sally and the scutellar-metanotal muscle (1 14, t;-t,) passes under the mesopostnotum through a small emargination of its posterior margin (state 1, broadly exposed and 114 internal). In Macroxyelinae (Xyelidae) (based on Macroxycla ferrunginea (Say), Megaxyela tricolor Norton, Megaxyela aviingrata (Dyar) and Xyelicia neurotica Ross), Megalodontoidea (Fig. 48),Tenthredinoidea and Cephoidea (Fig. 49), the mesopostnotum is exposed dorsally (depending on the contortion of the mesosoma) and muscle 114 passes externally over the mesopostnotum as a tendon enclosed by a sheath of connective tissue (state 2, broadly exposed and 114 external). In groups with an external muscle and and an exposed mesopostnotum (state 2), the anterior medial mar- gin may be shallowly or deeply emarginate (ap- pearing split) underneath muscle 114. In some genera of Tenthredinidae and Dipnonidae, the mesopostnotum is obscured by posterior devel- opment of the scutellum, but otherwise conforms to state2 (see Character 6). In Anaxyelidae, the mesopostnotum and muscle 114 are the same as in Xyelinae (Fig. 47). Initially, we coded this as the same character state (state 1). However, this caused problems in devel- oping an additive coding scheme that would force the anaxyelid state to be aucapomorphic and not transitional between states 2 and 4. To resolve this problem, Anaxyelidae were assigned a new char- acter state (state 3), which is treated as an autapomorphy. In all of the analyses, coding for Anaxyelidae and Xyelinae as state 1 or state 3 had no effect on tree topology. In Siricidae, a median vertical process on the anterior margin of the metanotum forms the pos- terior attachment of muscle 114, which is internal and lacks any connective sheath. Although the process is unique for Hymenoptera, this confor- mation is considered to be a modification of state 2 even though the mesopostnotum is generally not exposed. The scutellum often extends over the metanotal process and, as in some Tenthre- dinoidea, the internalization of muscle 114 and mesopostnotum may be secondary. Snodgrass (1910) reported that the mesopostnotum of Tremex cohtmba was exposed medially; however, we ob- served an exposed mesopostnotum only in some specimens of Urocerus. Some Cephidae (Fig. 49) also have a similar attachment to a peg-like pro- cess (mp) on the metanotum. In Xiphydriidae (Fig. 50), Orussidae and Apocrita, muscle 114 (t2-t3) and the mesopostnotum (medially) are completely internal and muscle 1 14 passes medially over the second phragma (state 4). An autapomorphic modification of state 4 is found in Xiphydriidae, in which the cuticle of the second phragma (between the lobes of the pseudophragma) encircles the muscle tendon as it passes through to the metanotum. Further devel- opment of the mesopostnotum and its associated Volume 3. 1 994 251 phragma in Apocrita is toward an even greater degree of internalization and a lateral shift of the points of articulation with the scutellum and met- anotum (Snodgrass 1910) (see character 10). In Aculeata (Fig. 53), the second phragma is almost entirely disassociated mesally and the notal at- tachments are reduced to lateral lamellae along the extreme lateral margins (Snodgrass 1942). Synonomy for muscle 114: Tenthredinidae: 19 & 20 (Euthomostethus, Maki 1938). Ichneumonidae: 20 & 21 (Psilopsyche, Maki 1938). Aculeata: IIdlm2 {Vespa, Weber 1925); 23 (Vespa, Maki 1938)'; Ilisl (Vespula, Duncan 1939); 70 (Apis, Markl 1966); t,-t3 {Apis, Daly 1963). Hymenoptera: tl3 (Matsuda 1970, Brodskiy 1992). Character 6. Formation of pseudophragma The postphragma (2ph) originates at the junc- tion of the mesopostnotum and metanotum and forms the posterior attachment site for the longitu- dinal flight muscles (112, lph-2ph) (Snodgrass 1910). In Mecopterida, Neuropterida, Xyelidae (Fig. 47), Megalodontoidea (Fig. 48) and Tenthre- dinoidea, the anterior margin of the mesopostnotum is broadly attached to the poste- rior margin of the mesonotum (scutellum) by a narrow conjunctiva (60, term from Ronquist and Nordlander 1989) (state 0). In Cephidae (Fig. 9b), Anaxyelidae, Siricidae (Fig. lib), Xiphydriidae and Apocrita, the anterior margin of the mesopostnotum is developed into a broad bilobed internal plate (pseudophragma) that extends an- teriorly and dorsally beyond the dorsal attach- ment of the mesopostnotum to the mesonotum (state 1). The pseudophragma is an extention of the second phragma and both form the posterior attachment for the longitudinal flight muscles (lph- 2ph). A pseudophragma is not apparent in Orussidae (Fig. 13) and some Ichneumonidae and has probably been lost. Characters 7-8. Furcal-spina muscles In Xyeloidea (Fig. 3) and Pamphiliidae (Fig. 2, indicated by dashed line), muscles 127 (fu,-sps,) and 180 (fu?-sps,) connect the spina with the furca of the following segment (stateO). Muscle 127 is present in both Neuropterida and Mecopterida and muscle 180 is found only in Neuropterida (Maki 1938, Kelsey 1957, Matsuda 1970); we veri- fied their presence only in Corydalus. Muscle 127 is lost (state 1, character 7) in all Hymenoptera ex- cept the Xyeloidea and Megalodontoidea. Rasnitsyn (1969, fig. 204) illustrates a muscle con- nection (127?) between sps, (his sst) and fu2 in Tremex (Siricidae), but we did not observe this in either Tremex or Urocerus. The loss of muscle 180 (state 1, character8) is a synapomorphy of Apocrita. Synonomy for muscle 127: Apidae: fu,-lsps [pupal muscle only] (Apis, Daly 1964). Hymenoptera: sl2 (Matsuda 1970; not listed by Matsuda for Hymenoptera). Synonomy for muscle 180: Symphyta: living (Dolerus, Schizocerus, Tenthredo, Weber 1927). Character 9. Furcal-basalare muscle The furcal-basalare muscle (fbl, fu,-ba3) ap- pears to be found only in Hymenoptera (state 1). The muscle is apparently absent in the outgroup (state 0). Muscle fbl could be homologous with Matsuda's p-s3 which connects the furca to the anterior margin of the succeeding episternum. Muscle p-s3 is found in Neuropterida (IIpml7, Korn 1916 for Myrmeleon), but apparently not in other Neuropterida or Mecopterida (Matsuda 1970). In Symphyta (Figs. 2, 3, 5-13) muscle fbl extends from the extreme lateral apex of the mesofurcal arm to the apodeme of the basalare of the hindwing (ba?). This muscle was first illus- trated, but not discussed, by Rasnitsyn (1969, fig. 204) for Tremex. In Diprionidae (Fig. 7), muscle fbl ends in a sclerotized cap attached to the apodeme of the basalare, along with muscle 137 (t,-ba ) which originates from the posterior face of the laterophragma (t3). Muscle fbl is absent (state 0) in most Apocrita. However, what appears to be the same muscle occurs in Mutillidae (Fig. 18), Rhopalosomatidae (Fig. 19) and Bradynobaenidae (state 1). The muscle in these taxa originates laterally on the mesofurcal arms (dorsal to the furcal-trochanteral muscle, musclel74) and narrows to a fine tendon that enters the pleural area near the base of the hind wing. The dorsal point of insertion was not ob- served but the muscle enters the pleural area pos- terior to the subalar muscle of the mesothorax and may insert onto the basalare or axillary sclerites of the hind wing. No such muscle was observed in 252 Journal of Hymenoptera Research other Apocrita. The loss of muscle fbl in Apocrita could be the result of a shift of the muscle origin from the furca to the pleural wall. Gibson (1986) reported a unique muscle in Eupelmidae and Pteromalidae (pl2-3ax3a) that attached between the lower mesepimeron and possibly the metatho- racic third axillary sclerite (of uncertain insertion; Gibson pers. comm. 1993). The homology of fbl with the pleural muscles is beyond the scope of this study. Character 10. Lateral articulation of mesopostnotum In Xyeloidea, Megalodontoidea (Fig. 51) and Tenthredinoidea and the outgroup taxa, the mesopostnotum is broadly and evenly joined to the upper mesepimeron (state 0). In Cephidae (Fig. 52), Anaxyelidae and Siricidae, the juncture is invaginated and connected by weak cuticle within the invagination, but maintains a strong cuticular connection at the anteriormost point of attachment (state 1). In Xiphydnidae, Orussidae and Apocrita, the mesopostnotum is completely internal and the lateral connection of the mesopostnotum to the upper mesepimeron is re- duced to the anteriormost point of attachment (state 2). Character 11. Fusion of lateral arms of mesofurca and metafurca In most Aculeata the lateral arms of the mesofurca and metafurca are closely appressed, but as in most other Hymenoptera and the outgroup taxa, they are broadly separated and joined by the interfurcal muscle 181 (state 0) (Fig. 3). In Apoidea (Figs. 46, 53), the lateral arms of both furcae are fused at the junction with the mesofurcal bridge and the interfurcal muscle is lost (state 1). The lower furcal arms form a four- cornered brace for the furcal complex, and the metafurcal arms are divided lateral to the furcal bridge. A similar fusion of the mesofurca and metafurca occurs in Amiseginae (Chrysididae; Fig. 39) and Gasteruptiidae (Fig. 26), but in both cases muscle 181 is retained and these are considered to be non-homologous and autapomorphic changes. Additional coding as autapomorphic states is un- necessary for this analysis but may be warranted in more extensive studies of related taxa. Character 12. Furcal process for trochanteral muscle 174 In most Hymenoptera and the outgroup taxa, the furcal depressor of the trochanter (muscle 174, fu^-tr^) attaches to the lateral arms of the mesofurca (Fig. 3) (state 0). In Chrysididae (Cleptinae and Chrysidinae), muscle 174 attaches to anterior ex- tensions of the lateral furcal arms (Fig. 40, vp) (state 1). In Pompilidae (Fig. 42), Sapygidae, Scoliidae, Sierolomorphidae and Vespidae (Fig. 41), muscle 174 attaches to similar extensions (vp) that arise instead from the furcal bridge (state 2). In Vespoidea, muscle 174 has two origins - the lateral furcal arms and the ventral process (= scoop shaped processes, schT, Weber 1925; anterior pro- cess of mesofurcal bridge, PF„ Duncan 1939). PHYLOGENETIC HYPOTHESIS FOR HYMENOPTERA Figs. 55, 56 The higher classification of Hymenoptera has come under increasing scrutiny following reviews published by Konigsmann (1977, 1978a) and Rasnitsyn (1969, 1980). Konigsmann (1978a) pre- sented evidence that Hymenoptera was comprised of two monophyletic sister taxa, the Symphyta (excluding Cephoidea) and Cephoidea + Apocrita, with most apocritan families presented as a series of unresolved lineages. Although this more tradi- tional concept of the Symphyta is still used (Zessin 1985), it has not been accepted in more recent papers except to present an alternate hypothesis of relationships (Gauld and Bolton 1988). Rasnitsyn (1980) presented evidence that Symphyta is paraphyletic relative to Apocrita with Cephoidea as the sister group of Apocrita. Studies subsequent to Rasnitsyn (1980) (Gibson 1985, 1986; Naumann and Masner 1985; Carpenter 1986) led Rasnitsyn to modify his original hypotheses in a later paper (Rasnitsyn 1988), in which Cephoidea were placed as a more basal lineage of Hymenoptera and Orussidae as the sister group of Apocrita. A far more resolved set of relationships among the para- sitic Apocrita (Fig. 56) were presented by Rasnitsyn (1988). Some of Rasnitsyn's hypotheses have been corroborated in recent morphological or phyloge- netic treatments (Johnson 1988; Whitfield et al. 1989), although Gibson (1993) presented evidence against the biphyletic origin of Hymenoptera from Xyelidae and Brothers and Carpenter (1993) sup- Volume 3, 1 994 253 ported a different hypothesis for Chrysidoidea and Aculeata. For our data, these modifications affect only the distribution of character 12 in the Aculeata, and have no effect on the parsimony arguments presented in a later section. Cladograms comparing the hypotheses of Rasnitsyn and Konigsmann are presented in Whitfield (1992). Rasnitsyn (1980, 1988), Shcherbakov (1980, 1981) and Gibson (1993) have proposed Xyelidae as the most basal extant lineage of Hymenoptera. As well as possessing a branched Rs vein, Xyelidae retain the anepisternum as an integral part of the pleuron and retain the plesiomorphic condition of a relatively equally proportioned meso- and meta- thorax (Gibson 1993). Other than the structure of the antenna, which may be plesiomorphic, there are no features that support the monophyly of the Xyelidae, and Rasnitsyn (1988) proposed that the two subfamilies of Xyelidae belong to two lin- eages of Hymenoptera (Hymenoptera biphyletic with respect to the two subfamilies). Gibson (1993 and personal communication, 1993) argues that Hymenoptera excluding Xyelidae are monophyl- etic but that no characters support the monophyly of the Xyelidae. The mesofurcal-mesopostnotal complex of Pamphiliidae and both subfamilies of Xyelidae are basically the same and we consider them to represent the plesiomorphic condition for Hymenoptera. Rasnitsyn (1988) classified the suborder Siricina into three infraorders that include the following extant families: Xyelomorpha (Xyelidae), Tenthredinomorpha (Argidae, Blasticotomidae, Cimbicidae, Pergidae and Tenthredinidae [includ- ing Diprionidae]), and Siricomorpha (including Megalodontoidea [=Pamphilioidea]: Pamphiliidae and Megalodontidae; Cephoidea: Cephidae; Siricoidea: Anaxyelidae, Siricidae and Xiphydriidae). Rasnitsyn (1988) treated Xyelidae as paraphyletic with respect to Tenthredinoidea and remaining Hymenoptera, and the three fami- lies of Siricoidea are paraphyletic with respect to the Orussidae + Apocrita (Suborder Vespina). Rasnitsyn (1988) placed Cephoidea as basal to the Siricoidea + Apocrita, rather than sister group to the Apocrita as in his earlier treatments (Rasnitsyn 1969, 1980). As a framework for evaluating our results, we have compiled the results of Rasnitsyn (1988) and Gibson (1985, 1993) to derive the cladogram pre- sented in Figure 55. Contrary to Rasnitsyn (1988), Hymenoptera excluding Xyeloidea are treated as monophyletic based on the presence of a postspiracular sclerite (=detached anepisternum) (Gibson 1985, 1993) and an apically simple Rs vein that is not bifurcate (Rasnitsyn 1988). Xyelidae are treated as paraphyletic to allow for discussion of Character 5. Tenthredinoidea are monophyletic and placed here as sister group to the remaining Hymenoptera, excluding Xyeloidea and Megalodontoidea, based on a reduction in the number of preapical tibial spurs (from 2-4 spurs on the hind legs in Xyeloidea and Megalodontoidea tolessthanl in Tenthredinoidea, l-2inCephoidea and 0 in remaining Symphyta and Apocrita; H. Goulet and G. Gibson, pers. comm., 1993), and the loss of the ventral mesofurcal-spina muscle, fu2- spSj, which is present only in Xyeloidea and Megalodontoidea (see Character Analysis). Remaining infraorders of extant Vespina pro- posed by Rasnitsyn (1988) include the Orussomorpha, Evaniomorpha (Stephanoidea, Evanioidea), Proctotrupomorpha (Proctotru- poidea, Cynipoidea, Chalcidoidea), Ichneu- monomorpha (Ichneumonoidea), and Vespomorpha (Chrysidoidea, Sphecoidea, Pompiloidea, Scolioidea, Formicoidea and Vespoidea). Recent classifications (Naumann 1992; Huber and Goulet 1993) treat the five families of Rasnitsyn's Stephanoidea as higher taxa (Ceraphronoidea [Ceraphronidae and Megaspilidae], Megalyroidea, Stephanoidea and Trigonalyoidea). Following Brothers (1975) and Huber and Goulet (1993), the Aculeata are re- ferred to only three superfamilies (Chrysidoidea, Apoidea and Vespoidea) and Mymarommatidae are placed as a superfamily. GROUP ANALYSIS Character states are referred to in brackets as [charactenstate]. Relationships are discussed ac- cording to the phy logenetic hypotheses illustrated in Figures 55 and 56. The distribution of character states for each taxon are listed in Appendix 2. 254 Journal of Hymenoptera Research Xyeloidea + Megalodontoidea + remaining Hymenoptera Figs. 1, 2, 3, 4, 6, 47, 48, 51 Four characters of the mesofurcal- mesopostnotal (MF-MPN) complex are not found in the outgroups and therefore support the mono- phyly of Hymenoptera: long anterior arms of the mesofurca [1:1]; two furcal laterophragmal muscles [3:1]; muscle 114 attaching to the metanotum (part of Character 5, see below); and furcal-basalare muscle (fbl, fu,-ba.,) present [9:1]. Xyeloidea and Megalodontoidea are plesiomorphic for all characters postulated for Hymenoptera in the mesofurcal-mesopostnotal complex. Megalondontidae were verified for ex- ternal characters only and the presence of anterior furcal arms [1:1] was based on fig. 189 in Rasnitsyn (1969). Rasnitsyn (1969) also refers to the mesofurco-metapre-episternal muscle as found only in Xyela. This muscle was not identified in our dissections, although it may be synonymous with muscle 127 (fu^sps,). The mesopostnotum is external and broad medially in both Xyeloidea and Megalodontoidea. In extant Xyelinae (Xyelidae), the mesopostnotum is entire medially and muscle 114 is internal [5:1, Fig. 47]. A similar conformation is found in the outgroup, but the attachment of muscle 114 is different [5:0]. In Macroxyelinae (Xyelidae), Megalodontoidea and Tenthredinoidea, the mesopostnotum is almost completely separated medially and muscle 114 passes externally [5:2, Fig. 48]. The deep emargination of the mesopostnotum associated with state 5:2 is visible externally and can be interpreted from fossil Hy- menoptera. Based on illustrations in Rasnitsyn (1969), external passage occurs in the fossil xyelids: Archexyelinae: Asioxyela (fig. 34), Leioxyela (fig. 41), Triassoxyela (fig. 32), Xiphoxyela (figs. 45, 46); Macroxyelinae: Agaridyela Rasn. (fig. 41), Chaetoxyela Rasn. (fig. 76), Ceroxyela Rasn. (fig. 77); and Xyelinae: Eoxyela Rasn. (fig. 59), Lydoxyela Rasn. (fig. 68), Spathoxyela Rasn. (fig. 42), Xyelisca Rasn. (fig. 61), and Xyela mesozoica Rasn. (fig. 62). If correctly interpreted in fossil Xyelinae, the inter- nal passage of muscle 114 and an entire mesopostnotum are derived within extant Xyelinae, and the external passage of muscle 114 and an emarginate mesopostnotum are probably apomorphic for Hymenoptera Tenthredinoidea + remaining Hymenoptera Monophyly of Tenthredinoidea and the re- maining Hymenoptera is supported by the loss of muscle 127 (fu2-spSj) [7:1]. Tenthredinoidea Figs. 5, 7-8, 20, 21 The anterior mesofurcal arms are reduced or lost in all Tenthredinoidea [1:2]. Blashcotomidae (Fig. 20) The plesiomorphic combination of muscles is present. Anterior furcal arms are present, but they are short and thin. As discussed earlier (analysis of characterl ), because of uncertainties in the homol- ogy of reduced arms, we have scored both re- duced and absent anterior furcal arms as the same character state (1:2). The anterior apodeme of the laterophragma is absent and both 150a and 150b attach to the apex of the posterior lobe (Fig. 20a). This is considered to be an autapomorphic devel- opment, but there is also a general similarity with the laterophragma of Nematus (Fig. 19), in which short anterior furcal arms are also present. Tenthredinidae (Fig. 21) The plesiomorphic combination of muscles is present in all taxa examined, and the structure of the laterophragma is the same as in Xyelidae and Pamphiliidae. The anterior arms of the mesofurca vary from narrow, elongate apodemes extending from the anterior face of the lateral arms (Nematus), to short cup-like structures similar to those found in Diprionidae, to completely absent (most Tenthredinidae). Muscle 180 (fu^sps,) was not found in Eutomostethus (Blennocampinae) (Maki 1938). Diprionidae (Fig. 7) The plesiomorphic combination of muscles and structures for Tenthredinoidea are present but with the following apomorphic features: ante- rior arms reduced to cup-like receptacles for muscle 124; anterior process of laterophragma expanded and forming a broad attachment for 150a; and muscle 150b originates medially to 150a on the lateral furcal arms. Volume 3. 1 994 255 Cimbicidae (Fig. 8) The anterior arms are reduced to small spoon- shaped processes, the posterior lobe of mesopostnotum is fused to second phragma form- ing attachment for a large muscle 116, and muscle 150b is lost [3:2]. The anterior process of the laterophragma is enlarged and extends ventrally into the cup-shaped lateral arm of the mesofurca (autapomorphy). Argidae + Pergidae The lateral arms of mesofurca are tubular and hollow (apomorphic), the anterior arms of the mesofurca are absent, and the attachment site for muscle 124 (f Uj-fiij) is to the rounded surface of the lateral arms. Although they would suggest a sister group relationship for these two families, these characters were not surveyed throughout Hy- menoptera and were not coded for the parsimony analysis. Argidae The laterophragma and its associated muscle attachments are as in Tenthredinidae. Pergidae (Fig. 5) This family displays a number of autapomorphies: loss of the tergal depressor muscle (116) [4:1] and the associated posterior lobe of the laterophragma [2:2], and loss of one of the furcal laterophragmal muscles (possibly 150b) [3:?]. The laterophragma is reduced to a small triangular lobe that ends in a narrow attachment to muscle 150. The lateral arms form a cup-shaped attachment for muscles 150 and fbl. A sclerotized apodeme, probably derived from the lateral arm and muscle 151 (pl,-fu,rt), extends laterally to the pleural ridge. Ventrolateral extensions of the lat- eral arms are developed as insertions for the en- larged muscle 170 (pl2-fu2b). Cephidae + remaining Hymenoptera Figs. 9, 49, 52 Monophyly of the Cephidae and the remain- ing Hymenoptera is supported by a suite of char- acters: apical connection of the anterior mesofurcal arms [1:3], formation of a pseudophragma [6:1], and invagination of the mesopostnotum laterally [10:1]. The pseudophragma is not present in all Hymenoptera, but its complete absence in the lower Hymenoptera and presence in Cephidae, Anaxyelidae, Siricidae, Xiphydriidae and most Apocrita suggest that it is a synapomorphy at this level. There are no published accounts which make reference to a pseudophragma occuring elsewhere within Endopterygota. Cephidae are plesiomorphic for other charac- ters of the mesofurcal-mesopostnotal complex: muscles 116, 150a, 150b, and fbl are present, and the laterophragma is divided into a large posterior lobe and a small anterior process. The mesopostnotum is usually exposed medially but is sometimes concealed by the scutellum. Muscle 114 passes externally over the mesopostnotum in all genera (plesiomorphic) and may arise dorsomedially from a vertical process on the metatergum {Calameuta Konow, Ceplms Latrielle [mp, Fig. 49], Monoloplopus Konow, Syrista Konow and Tracheitis Jurine) as in Siricidae, or from an emargination of the metatergum (Caenocephalus Konow and Janus Stephens) as in Anaxyelidae. Anaxyelidae + (Siricidae + (Xiphydriidae + (Orussidae + Apocrita))) Fig. 10 Monophyly of the Anaxyelidae and remain- ing Hymenoptera is based on a more complete fusion of the anterior arms of the mesofurca [1:4]. Fusion of the anterior mesofurcal arms in Anaxyelidae (Fig. 10a) appears to be more com- plete than in Siricidae (Fig. 11a) in which the arms are fused in about the anterior half. The anterior arms of Anaxyelidae (Fig. 10a) are still traceable in dorsal view and they are connected posteriorly by a thin plate of cuticle. The greater degree of fusion of the arms is probably autapomophic in Anaxyelidae; however, in Anaxyelidae and Siricidae the fusion is intermediate between the apical connection found in Cephidae and the com- plete fusion including a strong furcal bridge found in Xiphydriidae + Orussidae + Apocrita. In Anaxyelidae, Siricidae and Xiphydriidae, muscle 150b inserts onto the posterior face of the laterophragmal lobe rather than to the ventral margin as in Xyeloidea, Megalodontoidea, Tenthredinoidea and Cephoidea. As the muscle is absent in Orussidae and Apocrita, the importance of this unique attachment as a character for sup- porting monophyly of Siricoidea (including Anaxyelidae and Xiphydriidae) or Siricoidea + 256 Journal of Hymenoptera Research Orussoidea + Apocrita cannot be assessed. In Anaxyelidae, the presence of both furcal- laterophragmal muscles (150a & 150b) and ante- rior origin of muscle 150b on the f ureal arms is plesiomorphic. The mesopostnotum is broadly exposed medially and muscle 114 (t,-t3) passes under the mesopostnotum medially (autapo- morphy; 5:3). The dorsal enclosure of muscle 114 in Anaxyelidae may result from the extension and medial fusion of the anterior margin of the mesopostnotum. Rasnitsyn (1969, fig. 99) illus- trates a split mesopostnotum [5:2] for the fossil Anaxyela gracilis Martynov. This supports the con- tention that an internal muscle 114 is derived within Anaxyelidae. Siricidae + (Xiphydriidae + (Orussidae + Apocrita)) Fig. 11 No characters of the mesofurcal- mesopostnotal complex demonstrate the mono- phyly of Siricidae and the remaining Hymenoptera. The anterior mesofurcal arms of Siricidae (Fig. 1 1 ) are vertically flattened and in dorsal view are distinct along their entire length. The degree of fusion of the anterior arms is intermediate be- tween Cephidae and Xiphydriidae + Orussidae + Apocrita, but, as discussed above, is comparable to that found in Anaxyelidae. Siricidae are autapomorphic for the loss of muscle 150a (along with the anterior process of the laterophragma) [3:3]. The posterior attachment of muscle 114 (t,-t,) to a vertical process of the metanotum is shared with some Cephidae but not with Xiphydriidae, Orussidae or Apocrita, which lack a process. Inter- nal passage of muscle 114 in Siricidae may result from the posterior extension of the scuteOum rather than internalization of the mesopostnotum as in Cephidae and Tenthredinoidea. Rasnitsyn (1988) proposed that Siricidae + remaining Hymenoptera were monophyletic based on four synapomorphies: 1) compound third an- tennal segment reduced and subequal to the fol- lowing segments, 2) head capsule with postgenae subcontiguous, divided with narrow hypostomal bridge, 3) transscutal articulation present, and 4) prepectus concealed. In Anaxyelidae, the primi- tive state of a compound third antennal segment is found only in the fossil anaxyelid genus Sphenosyntexis Rasnitsyn; the derived state is found in the remaining genera of Anaxyelidae including the only extant species, Syntexis libocedrii Rohwer. However, the antennal flagellum is apomorphic (reduced and divided) in Xyeloidea, Megalo- dontoidea, Tenthredinoidea, Cephoidea and within Anaxyelidae, and thus its reduction in Siricidae and remaining Apocrita is not a reliable character. The postgenae meet broadly along the medial line to form a postgenal bridge (=genaponta sensu Ross 1937) in Anaxyelidae, Siricidae and Xiphydriidae; we agree with Ross (1937) that the postgenae of Anaxyelidae and Xiphydriidae are similar and thus Anaxyelidae cannot be excluded from Siricoidea based on this character. Gibson (1985) demonstrated that a transscutal articula- tion is absent in Cephidae, Anaxyelidae and Siricidae. The presence of a concealed prepectus is useful for supporting the monophyly of Siricidae + Xiphydriidae + Orussidae + Apocrita (Rasnitsyn 1988; Gibson 1985). Xiphydriidae + (Orussidae + Apocrita) Fig. 12 There is strong support in the MF-MPN com- plex for the monophyly of Xiphydriidae, Orussidae and Apocrita. The anterior mesofurcal arms are completely fused into an anterior furcal process (fp) and posteriorly strengthened to form a trans- verse mesofurcal bridge (fb) [1:4], with the excep- tion of some Apocritan taxa in which the mesofurcal bridge is lost. In addition, the mesopostnotum is completely internal medially [5:4] and laterally [10:2], and the laterophragma is strongly incised with the anterior process (ap) developed into a strong apodeme [2:1]. The plesiomorphic combi- nation of muscles in the MF-MPN complex is retained. Among the Cephoidea, Siricoidea and Orussoidea, Xiphydriidae are unique in having muscle 150b attached to the lateral arms of the mesofurca (versus the anterior arms or furcal pro- cess). Muscle 124 (fu^-fUj) is broadly attached to the furcal process and bridge. In Xiphydriidae, the tendon of muscle 114 (t2-t3) is enclosed medially, and internally, by the cuticle of the second phragma (autapomorphic). Orussidae + Apocrita Fig. 13 No synapomorphies were found in the Volume 3, 1994 257 mesofurcal-mesopostnotal complex that support the monophyly of Orussidae + Apocrita without Xiphydriidae. It is possible that the loss of muscle 150a supports the monophyly of Orussidae + Apocrita [3:? = 3:2] . However, the homology of the remaining muscle 150 in Orussidae is uncertain (see discussion of Character 3). Therefore, it is also possible that muscle 150b is lost in Orussidae and 150a is lost in Apocrita. Orussidae have at least three apomorphic fea- tures in the MF-MPN complex, two of which are also found in Pergidae and Cimbicidae. The furcal bridge (fb) is rounded and bowed anteriorly with a distinct median process (fp) [1:6]. The attach- ment of muscle 150 to the furcal process is found only in Orussidae. In Apocrita, muscle 150a is usually attached to the furcal bridge or lateral arm. Otherwise the furcal bridge and process in Orussidae are similar to Xiphydriidae and other Apocrita. Muscle 116 is lost [4:1] in Orussidae, and there is an associated reduction of the laterophragma to a small triangular lobe that forms the attachment for muscle 150 [2:2]. Additional autapomorphies in Orussidae include loss of the pseudophragma [6:0] and the configuration of muscle 180 (fu3-sps2) which arises from the spina as a single narrow tendon that bifurcates about half-way along its length to the metafurcal arms (the plesiomorphic configuration is completely separated along its entire length). Apocrita Figs. 14-19, 22-46, 53, 54 The loss of two muscles in the MF-MPN com- plex supports the monophyly of the Apocrita: the metafurcal-spina muscle (180, fu3-sps2) [8:1] and the furcal-basalare muscle (fbl, fu2-ba3) [9:0]. Muscle 150b (fu2-pn2a) [3:2] is lost in all Apocrita including those that retain the posterior lobe of the laterophragma. However, as discussed above, the homology of the remaining muscle 150 in Orussidae is not clear, and therefore loss of 150a is either a synapomorphy of Orussidae + Apocrita or an autapomorphy of Apocrita. Loss of musclel37 may also support the monophyly of Apocrita (see discussion for Character 2). Shcherbakov (1981) stated that muscle 170 (pl2-fu2b, s-cx2) is connected to the coxal process of the mesopleuron in all Hymenoptera. Based on our observations, it is well developed in all Symphy ta and attaches laterally to the mesofurcal arms (Figs. 3, 5, 9b, 12b, 13). In Apocrita, the position of muscle 170 shifts from the apex of the furcal arms to the base of the furca and is more nearly horizontal. This is most apparent in Vespoidea (cf. Duncan 1939, muscle IIfpl2 (71)). The laterophragma of Cynipoidea, some Diapriidae, Monomachidae (Fig. 14) and Vanhorniidae (Fig. 15a) are similar in structure with an anterior apodeme (ap) narrowly sepa- rated by a deep incision from the posterior lobe (pn,) [1:5], which supports muscle 116 [4:0]. In Megaspilidae (Fig. 25), the apodeme and lobe may be broadly or narrowly separated. Muscle 116 [4: 1 ] is present only in the families listed above and probably has been lost numerous times within Apocrita (Gibson 1985). It is unlikely that this muscle and associated posterior lobe could be regained once lost. Under the present hypothesis of relationships in Apocrita, muscle 116 and the lobe are lost at least 9 times (10 if also lost in Austroniidae) (Fig. 56). Of these, only in Diapriidae is the muscle lost within an entire family. Groundplan states for the Apocrita are exem- plified by Monomachidae (Fig. 14), Vanhorniidae (Fig. 15), most Diapriidae (Fig. 29) and Cynipoidea (Fig. 22), in which muscle 116 and the posterior lobe of the laterophragma are retained (Daly 1963; Gibson 1985). Plesiomorphic attributes of Apocrita include l)presence of a median furcal process (fp) forming the attachment for muscle 124, 2) furcal bridge complete (fb) [1:4], 3) muscle 150a originat- ing laterally on the bridge or the lateral arm of the mesofurca, 4)laterophragma divided into an ante- rior apodeme and posterior lobe [2:1], 5)presence of the tergal depressor muscle (116) [4:0], 6) mesopostnotum internal [10:2] and 7) prephragma present [6:1]. Ceraphronoidea (Fig. 25) have lost the mesofurcal bridge [1:7] but retain a divided laterophragma, and in some Diapriidae (Fig. 22) the posterior lobe of the laterophragma is lost [4:1] Apart from changes in the laterophragma, apomorphic changes of the MF-MPN complex in Apocrita include loss of the mesofurcal bridge [1:7] and changes in the structure of the anterior apodeme (=axillary lever) of the laterophragma [2:3-6]. Changes in the shape of muscle 150a, lat- eral arms, furcal bridge and second phragma are homoplastic and probably related to changes in thoracic shape. Notable exceptions in Aculeata include development of an axillary lever [2:5] that 258 Journal of Hymenoptera Research is separated in Apiformes [2:6], fusion of the lat- eral arms of the mesofurca and metafurca [11:1], development of processes on the mesofurca for the furcal-trochanteral muscle [12:1,2], and a me- dial reduction of the attachment points (60 & 61) between the second phragma and terga. The cladogram presented in Figure 56 is con- sistent with the relationships proposed by Rasnitsyn (1988) and Brothers and Carpenter ( 1 993) for extant Apocrita, and is used as the model for examination of character state change within Apocrita. Apart from the modif ications in Aculeata, the MF-MPN complex offers little support for phylogenetic relationships proposed for Apocrita. Within Apocrita a total of 22 steps are required to explain the distribution of character states as mod- elled on Figure 56 (excluding changes in the Chalcidoidea + Platygastroidea + Mymarom- matoidea trichotomy, Austroniidae and Peradeniidae). Given that the tergal depressor muscle and posterior lobe of the laterophragma are present [4:0] in what are assumed to be rela- tively derived groups (e.g., Cynipoidea, Diapriidae and Monomachidae), this complex must be lost a minimum of 9 times (or 10 times if it is absent in Austroniidae). Only 8 losses are required if Pelecinidae and Vanhorniidae are monophyletic (Fig. 56). An additional 3 losses of muscle 116 [4:1] are required using Konigsmann's (1978a) hypoth- esis for Apocrita. Obviously, loss of this muscle does not offer strong evidence for the monophyly of Megalyroidea, Stephanoidea and Trigo- nalyoidea (Fig. 56). Among the parasitic families of Apocrita, the mesofurcal complex of Gasteruptiidae (Fig. 26) and Evaniidae (Fig. 27) were very different from each other. However, the mesofurcal bridge of Aulacidae and Evaniidae are similar in shape (broad and flattened with muscle 124 attaching under the anterior margin of the bridge). In all three families the axillary lever is strongly in- flected and perpendicular to the lateral wall of the mesosoma (Figs. 26, 27), and the lever is connected by a long thin tendon to muscle 150a. Scelionidae and Chalcidoidea both have a similar mesofurcal bridge (bowed with the median process virtually absent) which may indicate a sister-group rela- tionship between Chalcidoidea and Platy- gastroidea. However, given that the bridge is lost in Platygastridae, Mymarommatidae, Mymaridae and some Chalcidoidea [1:7], it is not clear what the groundplan state of this character is in either superfamily. Also, the axillary lever is deflected [2:3] in Platygastridae, some Scelionidae and most Chalcidoidea. Almost all Chalcidoidea including Mymaridae are unique in that muscle 150a at- taches to the entire length of the deflected axillary lever (versus only apical attachment). Mymarommatidae have a medially inflected lever [2:1] and muscle 150a is conical and attaches only to the apex of the lever (plesiomorphic). Proctotrupidae (Fig. 32) have a similiar axillary lever to that of Platygastridae (Fig. 28), although it is shorter and not as strongly deflected (Figs. 29- 32). In Heloridae (Fig. 30) and Pelecinidae (Fig. 31), the lever is followed by a short spine (pn,?) that has no muscle connection and may be the remnant of the posterior lobe of the laterophragma. Some Proctotrupidae have a blunt triangular lobe in a homologous position. Heloridae, Pelecinidae and Proctotrupidae are the only Hymenoptera with a vestigial posterior lobe of the mesopostnotum. Ropronndae were identical to Heloridae inter- nally except for the lack of a vestigial posterior lobe. Most Ichneumonoidea (Fig. 17) have a mesofurca consisting of a broad mesofurcal bridge and a strong medial projection. The furcal bridge of Hybrizontinae (Braconidae) (Fig. 33) lacks a median process but this could be autapomorphic. The axillary lever in Ichneumonidae is similar to Vespoidea and Apoidea [2:5]. This may be evi- dence for the close relationships to Aculeata that have been proposed (Rasnitsyn 1988, Sharkey and Wahl 1992), but a similar lever has not been found in Braconidae or Chrysidoidea. Chrysidoidea do not have any defining apomorphies of the mesofurca or axillary lever. The arched bridge of Plumariidae (Fig. 34), Sclerogibbidae (Fig. 35) and Embolemidae (Fig. 36) are similar but this feature may be plesiomorphic, as this shape is similar to Stephanidae. The furcal bridge of the single fe- male dryinid examined (Anteoninae) was similar to Bethylidae (Fig. 38) (relatively straight with a strong apical projection), and the bridge of the male dryinid (Fig. 35) was reduced to a straight bar. Obviously more taxa need to be surveyed to understand the importance of furcal shape in Chrysidoidea. Chrysididae have apomorphies that distinguish taxa at the subfamily level: the fusion of the lateral arms of the mesofurca and metafurca Volume 3, 1 994 259 in Amiseginae (Fig. 39), and extensions of the furca that support the furcal depressor of the trochanter [12:1] in Chrysidinae (Fig. 40) and Cleptinae. These latter two subfamilies also share a reduction of the axillary lever [2:4]. Reduction of the axillary lever and presence of ventral pro- cesses in Cleptinae and Chrysidinae do not sup- port the hypothesis of relationships among Chrysididae proposed by Kimsey and Bohart (1990) of Cleptinae + (Amiseginae + Chrysidinae). Based on relationships for Chrysidoidea proposed by Brothers and Carpenter (1993), Carpenter (1986) and Rasnitsyn (1988), the plumariid type of bridge is plesiomorphic, Dry inidae have an independently derived bridge, and Chrysididae + Bethylidae have a relatively broad, straight furcal bridge with a strong medial projection. Monophyly of the Vespoidea and Apoidea is supported by the development of an axillary lever [2:6], which is subsequently reduced in Bradynobaenidae [2:4] and modified in Formicidae [2:1]. Most Aculeata also have a broad, flattened furcal bridge with muscle 124 attaching to the dorsal and lateral surfaces (Fig. 18), although the bridge is not as well developed in Rhopalo- somatidae (Fig. 19), Sierolomorphidae and some Sphecidae (Fig. 46). In Tiphiidae (Fig. 44), the furca is convoluted and broadly excavated dorsally. The furcal bridge has several shapes in Vespoidea but their importance for assessing relationships will require a much broader survey of taxa. A strong ventral process (vp) for the furcal depres- sor of the trochanter (174, fu,-tr,) confluent with the furcal bridge [12:2] is present in Sapygidae, Sierolomorphidae, Pompilidae (Fig. 42), Scoliidae and Vespidae (Fig. 41). Under the most recent hypothesis of relationships for Vespoidea (Broth- ers and Carpenter 1993), the ventral process is derived independently in Sapygidae, Pompilidae and Scoliidae + Vespidae, although several mod- els of state change are possible. Mutillidae (Fig. 18), Rhopalosomatidae (Fig. 19) and Bradyno- baenidae have apparently retained muscle fbl (fu2-ba3) [9:1] and have a characteristically shaped mesofurca, however these families are not regarded as sister taxa (Brothers 1975, Carpenter and Broth- ers 1993). Apoidea have been separated into two groups, the Apiformes and the Spheciformes (Goulet and Huber 1993). Heterogynaidae have been included either within the Spheciformes (see Goulet and Huber 1993) or as the sister group of Spheciformes + Apiformes (Carpenter and Brothers 1993). Apoidea (Figs. 45, 53), including Heterogynaidae, have the arms of the mesofurca and metafurca fused at the junction with the mesofurcal bridge and muscle 181 (fu3-fu2) is lost [11:1]. This feature is unique in Hymenoptera. In Vespoidea (Figs. 41, 42, 44), Spheciformes (Fig. 46) and Heterogynaidae, the base of the axillary lever is broadly fused to the lateral arm of the mesopostnotum. All Apiformes have the axillary lever separated from the mesopostnotum as an independent sclerite [2:6] (Fig. 45) (Snodgrass 1942). PARSIMONY ANALYSIS The mesofurcal-mesopostnotal complex rep- resents a single system of inter-related characters, and it should not be used alone to form new hypotheses of relationships. However, it is of in- terest to determine if our interpretations of charac- ter state evolution are in fact the most parsimoni- ous for these data. We tried two approaches to explore this question. First, we superimposed the data in Appendix 2 on existing hypotheses of relationships. PAUP Version 3.0s (Swofford 1985) was used to optimize the character state data in Appendix 2 on Figure 55 for Symphyta and Fig- ures 55 and 56 for the entire Hymenoptera (using one fully resolved tree for Apocrita based on Rasnitsyn (1988) and Brothers and Carpenter (1993)). Ibaliidae was used to represent the groundplan of Apocrita in the former case, and each matrix included the outgroup (Appendix 2). All most parsimonious models of character state evolution were examined on these tree topologies. Characters 1 and 4 are sensitive to assump- tions about transformation series. As discussed above (Character Analysis), it seems most reason- able to assume that character 1 can evolve from state 0 to state 1 (unique for Hymenoptera), from state 1 to state 2 or state 3, from state 3 to state 4 and from state 4 to state 5; states 6 and 7 are indepen- dent derivations from state 5. Character 1 was ordered using a character state tree (or "stepmatrix" in PAUP) that specified the above transformation series. As discussed earlier, character 4 is treated as irreversible. The assumptions of additive cod- ing for characters 1 and 4 are referred to as the Ordered Characters. Other characters were al- ways treated as nonadditive (unordered). 260 Journal of Hymenoptera Research For Symphyta, the models of character state change shown in Figure 55 are in fact the most parsimonious for that tree topology. 27 steps are required to explain the data (indices of homoplasy are not possible for characters ordered using step matrices). Different models of state change are possible for character 3 (because it is scored as missing in Pergidae and Orussidae) and character 5 (unknown polarity for ingroup). For character 5, the presence of state 2 in fossil Xyelinae argues that this condition is groundplan for Hymenoptera, but it is equally parsimonious (at least for extant taxa) to assume that state 1 is groundplan and state 2 arises in the interval below Macroxyelinae and remaining Hymenoptera (dashed line, Fig. 55). The latter solution would make Xyelidae paraphyletic. Treating all characters as nonaddi- tive (unordered) does not change the tree length (27 steps, consistency index 0.81, retention index 0.77). In fact, the only effect is to allow for a number of alternate models of state change for character 1 . In this case, it is equally parsimonious to assume any of states 0, 1, 2 or 3 in the nodes below Tenthredinoidea + remaining Hymenoptera and Cephidae + remaining Hymenoptera, a result that essentially discards the information in the trans- formation series. Optimizing the ordered data onto alternate hypotheses of Symphyta resulted in longer trees. Rasnitsyn's (1988) tree, with Macroxyelinae as sister group to Tenthredinoidea, is one step longer, because of the parallel loss of muscle 127 [7:1]. Konigsmann's (1977) tree, with Symphyta monophyletic and Cephidae as sister group to Apocrita, requires 37 steps (character 3 optimized). For Hymenoptera, similar results were ob- tained for optimizing the Ordered Character data onto one tree, which is summarized for Apocrita in Figure 56. 58 steps were required to explain the data. Different optimizations were possible for Character 1 forChalcidoidea + Mymarommatoidea + Platygastroidea (if parallel loss of mesofurcal bridge [1:7] is preferred), and Character 9 in Vespoidea. In the former case, although it is equally parsimonious to assume loss of the mesofurcal bridge (1:7) followed by a gain [1:6], parallel loss of the bridge seems more likely. Likewise, it seems more likely to postulate a parallel loss of the f urcal- basalare muscle [9:0] in Vespoidea. Treating all characters as nonadditive decreases the length of the hymenopteran tree by 5steps because of a different optimization for Character 4 (treated as irreversible in the ordered analysis) and reversals (4:1 to 4:0) are favoured. For this result, the loss of the posterior lobe of the laterophragma and asso- ciated muscle 1 16 (4:1) would be a synapomorphy of Orussidae and Apocrita and then the presence of both structures (4:0) would have to be regained at least 4 times within Apocrita. Although this may be a more parsimonious solution in terms of the number of steps, we feel it is an unacceptable model of character evolution and the assumption of irreversibility should be invoked. Parsimony analysis using the branch and bound algorithm (Symphyta) or branch-swapping (Hymenoptera) was then performed on these data to determine if another tree topology would result in a more parsimonious solution. When characters are ordered as above, for Symphyta, 24 trees of 27 steps result (after derooting, condensing and rerooting the resulting trees in PAUP, which is necessary when using directed characters). Cod- ing Anaxyelidae as state 1 (homoplastic) or state 3 (autapomorphic) had no effect on the number of steps or resulting trees. As should be apparent from Figure 55, the basal taxa (Xyelinae, Macroxyelinae and Pamphiliidae) are unresolved in each of the 24 trees. Tenthredinoidea is always resolved as monophyletic but with little or no further resolution and Cephidae is consistently placed as sister group to remaining Hymenoptera, as in Figure 55. Anaxyelidae and Siricidae form a trifurcation with Xiphydriidae + Orussidae + Apocrita while the latter three taxa are essentially unresolved. In other words, there is no more par- simonious interpretation of these data and the most parsimonious solutions are congruent with the data shown on Figure 55. When characters are unordered for Symphyta, 138 trees of 26 steps result. The strict consensus solution for these trees is almost completely unresolved. For Hy- menoptera, the lack of character support in Apocrita resulted in a considerably shorter tree (40 steps for the Ordered Characters) but also provided no resolution of Apocritan relationships; the structure of the Symphyta portion remained unchanged. CONCLUSIONS The mesofurcal-mesopostnotal complex is a conservative character system that generally ex- Volume 3. 1 994 261 hibits few changes within families of Hymenoptera but undergoes considerable change at higher taxo- nomic levels. The most significant changes occur in the Symphyta, but additional characteristics of the mesofurcal-mesopostnotal complex provide characters that are useful in inferring relation- ships within Apocrita. Of the competing hypotheses for relationships of Symphyta, those in Figure 55 are the most parsimonious using the hypotheses of character state transformation discussed in this paper. Exact search methods using additive (ordered) charac- ters reproduce these relationships with the same number of steps and identical indices of character homoplasy and fit. Siricoidea are recognized as a paraphyletic group with respect to Apocrita. If Xyelidae are paraphyletic with the rest of Hy- menoptera (Rasnitsyn 1980, 1988) and Macroxyelinae + Tenthredinoidea are monophyl- etic, then a minimum of one extra step is required for character 7 (loss of fUj-spSj). In sharp contrast, a monophyletic grouping for Symphyta with Cephidae as sister group to the Apocrita (Konigsmann 1977) requires 11 additional steps for the same data. The phy logenetic hypothesis for Apocrita proposed by Rasnitsyn (1988; Fig. 56) was shorter than Konigsmann's by only three steps (for character 4). Hypotheses based on parsi- mony are much shorter for Apocrita, even when characters are ordered, but these hypotheses should not be accepted until they can be included in a larger data set with other character systems. Monophyly of Orussidae + Apocrita (Vespina se nsu Rasnitsyn) is supported by 10 external char- acters of adults (Rasnitsyn 1988) and a shift of the furcal-coxal muscles from a posterior to an ante- rior attachment on the discriminal lamella of the mesosoma (Johnson 1988). No features of the mesofurcal-mesopostnotal complex discussed here provide support for the monophyly of Orussidae + Apocrita, without Xiphydriidae. The mesofurca and laterophragma of Orussidae is highly derived and the only potential synapomorphy for Vespina (Character 3: loss of muscle 150b, fu2-pn, ) is prob- ably not homologous. The lack of 2ph-3ph in Siricidae and Xiphydriidae has been proposed as a potential synapomorphy for Xiphydriidae + Siricidae (Whitfield et al. 1992), but this muscle has also been lost independently in Goniozus (Bethy lidae), Cleptes (Chrysididae), Formicidae and Ampulex (Sphecidae) (Daly 1963). Monophyly of Xiphydriidae, Orussidae and Apocrita is supported by four synapomorphies. No evidence in the mesofurcal-mesopostnotal complex supports the inclusion of Xiphydriidae or Orussidae within Siricoidea or the monophyly of Anaxyelidae + Siricidae. The monophyly of Apocrita is supported based on losses of the muscles fu2-pn, , fu3-sps2, fu2-ba3, pl2-fu2b, and probably t -ba . Using the evolution- ary scheme proposed by Rasnitsyn (1988), charac- ters of the mesofurca and mesopostnorum were less useful for demonstrating relationships within Apocrita. Taxa which have retained the posterior lobe of the laterophragma and the associated muscle 116 are treated by Rasnitsyn (1988) as relatively derived taxa (especially Diapriidae and Cynipoidea). Rasnitsyn's hypothesis requires at least nine independent losses of character 4, if loss of both lobe and muscle is irreversible. This ap- pears to be a poor character for postulating rela- tionships. Loss of the mesofurcal bridge [1:7] could indicate support for the monophyly of Ceraphronoidea, but it is homoplastic within both Chalcidoidea and Platygastroidea. Some changes in shape of the axillary lever or mesofurcal bridge within Apocrita are difficult to categorize or to place into distinct transition series. Distinctive features such as the vestigial posterior laterophragmal lobe of Pelecinidae, Heloridae and Proctotrupidae are reductions and possibly con- vergent. The development and separation of the axillary lever, fusion of the mesofurca and metafurca, and the retention (or redevelopment) of muscle fbl provide strong support for relation- ships within the Aculeata. Changes in the mesofurcal-mesopostnotal complex probably result from an increased em- phasis on the fore wings for flight, reduction of the metathorax and fusion of the first abdominal seg- ment. Although the indirect flight muscles are the main wing depressors, the direct muscles of the mesothorax control flight through modifications of the pronation and rotation of the wings during the downstroke and by controlling the tension of the longitudinal flight muscles, thereby affecting the amplitude of the wing beat (Pringle 1957, 1960, 1961 ). In Apocrita, the axillary lever acts to turn the apex of the mesophragma and the associated fourth axillary sclerite (Weber 1925). This is a shift in function from the lower Hymenoptera where muscles 150a and 150b twist the posterior lobe of 262 Journal of Hymenoptera Research the laterophragma, initiating a greater force upon the indirect phragmal flexor (muscle 116, t;-pn2). The excision of the laterophragma in Xiphydriidae would reduce the effect of the anterior apodeme (= axillary lever) on the posterior lobe. Pringle (1960, 1961 ) proposed that the pleurosternal and axillary lever muscles control the power generated by the indirect flight muscles and thus the amplitude of the wing beat. The flexibility of the axillary lever reaches a maximum in the Apiformes in which complete separation from the laterophragma re- sults in more powerful leverage upon the associ- ated axillary sclerites of the forewing (Snodgrass 1942). Homologies may not be clear in taxa from opposite ends of a phylogenetic tree. For example, Weber (1925) discusses the mesofurcal ring (=mesofurcal bridge) of Siricidae and Vespidae as non-homologous. Superficially they do look dif- ferent, but by examining the intermediate stages the homologies can be verified. Changes in the mesofurca and mesopostnotum provide a num- ber of characters that show informative transitions from the plesiomorphic states in Xyelidae, but only by comparing a large number of taxa do these transformation series become apparent. Nonaddi- tive multistate characters are not as useful in build- ing a classification because unlinked character states support only their inclusive members. How- ever, a priori ordering of transformation series are dependent on previous classifications or on the presumed homology of sometimes very divergent character states. For example, placement of Cephidae as sister group to the Apocrita led to a misinterpretation of the hypostomal bridge by Ross (1937) and the mesofurcal bridge was not discussed as an important character by Rasnitsyn (1969, 1980, 1988). Present interpretations of changes in the mesofurca and mesopostnotum would have made little sense without the direc- tion provided from new classifications for Symphyta proposed by Gibson (1985) and Rasnitsyn (1988). The current distribution of char- acter states of the mesofurca and mesopostnotum within Apocrita seems incongruent with other character information. Perhaps as more characters are used to construct improved classifications of Apocrita, this will also change. ACKNOWLEDGEMENTS Gary Gibson and two anonymous reviewers provided valuable comments and criticisms. We thank Henri Goulet, John Huber, Lubomir Masner and especially Gary Gibson (all CNC) for providing specimens and guidance. Jeff Cumming (CNC) and Dave Smith (USNM) provided speci- mens of miscellaneous families that were important addi- tions to the material examined. This study was supported by Hatch Project TEXO-6705 of J.B.W. for initial work carried out at Texas A&M University and an NSERC grant to D.C.D. for subsequent work done in Ottawa. LITERATURE CITED Alam, S.M. 1951. The skeleto-muscular mechanism of Stenobracon deesae Cameron (Bracomdae, Hymenoptera) - An ectoparasite of sugarcane and juar borers of India. Part I. Head and thorax. Aligarh Muslim University Pub- lications (Zoological Series) 3: 76 pp. + 9 pis. Brodskiy, A.K. 1992. Structure, function, and evolution of the terga of wing-bearing segments of insects. II. Organiza- tional features of the terga in different orders of insects. Entomological Review 71: 8-28. Brothers, D. J. 1975. Phvlogeny and classification of the aculeate Hymenoptera, with special reference to Mutillidae. Kansas University Science Bulletin 50: 483- 648. Brothers, D. J. and J. M. Carpenter. 1993. Phylogeny of Aculeata: Chrysidoidea and Vespoidea. Journal of Hy- menoptera Research 2: 227-304. Bucher, G. E. 1948. The anatomy of Monodontomerus dentipes Boh., an entomophagous chalcid. Canadian journal of Research (D) 26: 230-281. Carpenter, J. M 1986. Cladistics of the Chrysidoidea (Hy- menoptera). journal of the New York Entomological Society 94: 303-330. Chapman, R. F. 1992. General anatomy and function, pp. 33- 67. In The Insects of Australia, 2nd edition. Volume 1. Melbourne University Press, Carlton. 1137 pp. Daly, H. V. 1963. Close-packed and fibrillar muscles of the Hvmenoptera. Annals of the Entomological Society of America 56: 295-306. Daly, H. V. 1964. Skeleto-muscular morphogenesis of the thorax and wings of the honey bee Apis mellifera (Hy- menoptera: Apidae). University of California Publications in Entomology 39: 77 pp. Duncan, C. D 1939. A contribution to the biology of North American vespine wasps. Stanford University Publica- tions (Biological Sciences) 8 (1): 272 pp. Gauld, I. and B. Bolton. 1988. The Hymenoptera Oxford Uni- versity Press, New York. 332 pp. Gibson, G. A. P. 1985. Some pro- and mesothoracic structures important for phylogenetic analysis of Hvmenoptera, with a review of terms used for the structures. Canadian Entomologist 117: 1395-1443. Gibson, G. A. P. 1986. Mesothoracic skeletomusculature and mechanics of flight and jumping in Eupelminae (Hy- menoptera, Chalcidoidea: Eupelmidae) Canadian Ento- mologist 118:691-728 Volume 3, 1994 263 Gibson, G. A. P. 1993. Groundplan structure and homology of the pleuron in Hymenoptera based on a comparison of the skeletomusculature of Xyelidae (Hymenoptera) and Rhaphidndae (Neuroptera). Canadian Entomologist 165: 165-187. Huber, J. T. and H. Goulet. 1993. Families of Hymenoptera of the World. Research Branch, Agriculture Canada. Publication 1894/E. 668 pp. James, H. C. 1926. The anatomy of a British phytophagous chalcidoid of the genus Harmolita {Isosoma). Proceedings of the Zoological Society of London [1926] 75-182. Johnson, N.F. 1988. Midcoxal articulations and the phylog- eny of the order Hymenoptera. Annals of the Entomologi- cal Society of America 81: 870-881. Kelsey, L.P. 1957. The skeleto-motor mechanism of the Dob- son Fly, Corydalus cornutus. Part II. Pterothorax. Mem- oirs of the Cornell University Agricultural Experimental Station 346: 31 pp. + 55 figs. Kimsey, L. and R.M. Bohart 1990. The Chrysidid Wasps of the World. Oxford University Press: Oxford. 652 pp. Korn, W. 1916. Die muskulatur des Kopfes und des thorax von Myrmeleon europaes und ihr metamorphose. Zoologische Jahrbucher 68: 273-330. Konigsmann, E. 1977. Das phylogenetische System der Hy- menoptera. Teil 2: "Symphyta." Deutche Entomologische Zeitschrift (N.S ,) 24: 1-40. Konigsmann, E. 1978a. Das phylogenetische System der Hy- menoptera. Teil 3 "Terebrantes" (Unterordung Apocrita). Deutche Entomologische Zeitschrift (N.S. )25: 1- 55. Konigsmann, E. 1978b. Das phylogenetische System der Hy- menoptera. Teil 4 Aculeata (Unterordung Apocrita). Deutche Entomologische Zeitschrift (N.S.) 25: 365-435. Knstensen, N.P. 1992. Phylogeny of extant Hexapods. pp. 125-140. In The Insects of Australia, 2nd edition. Volume 1. Melbourne University Press, Carlton. 1137 pp. Lawrence, J.F., E.S. Nielson and I.M. Mackerras. 1992. Skel- etal anatomy and key to orders, pp. 3-32. In The Insects of Australia, 2nd edition. Volume 1. Melbourne University Press, Carlton. 1137 pp. Maki, T. 1938. Studies on the thoracic musculature of insects. Memoirs of the Faculty of Science and Agriculture, Taihoku Imperial University 24: 343 pp. + 17 pis. Markl, H. 1966. Peripheres Nervensystem und Muskulatur lm Thorax der Arbeiterin von Apis mellifera L., Formica polyctena Foerster und Vespa vulgaris L., und der Grundplan der Innervierung des Insektenthorax Zoologische Jarbucher 83: 107-184. Matsuda, R. 1970. Morphology and evolution of the insect thorax. Memoirs of the Entomological Society of Canada 76: 431 pp. Michener, CD., 1944. Comparative external morphology, phylogeny, and a classification of the bees (Hy- menoptera). Bulletin of the American Museum of Natural History 82: 157-326. Naumann, I.D. 1992. Hymenoptera. pp. 916-1000. In The Insects of Australia, 2nd edition. Volume 1. Melbourne University Press, Carlton. 1137 pp. Naumann, I.D. and L. Masner. 1985. Parasitic wasps of the proctotrupoid complex: a new family from Australia and a key to world families (Hymenoptera: Proctotrupoideasensu lata). Australian journal of Zoology 33: 761-783. Pnngle, J. W. S. 1957. Insect Flight. University Press, Cam- bridge. 132 pp. Pnngle, J. W. S. 1960. The function of the direct flight muscles in the bee. Proceedings of the 1 2th International Congress of Entomology p. 660. Pnngle, J. W. S. 1961. The flight of the Bumblebee Natural History 70 (7): 20-29. Rasnitsyn, A. P. 1969. [Origin and evolution of the lower Hymenoptera.) Trudy Paleontologicheskogo Instituta, Akademiya Nauk SSSR 123: 196 pp. Rasnitsyn, A. P. 1980. [Origin and evolution of Hymenoptera. ] Trudy Paleontologicheskogo Instituta, Akademiya Nauk SSSR 174: 190 pp. Rasnitsyn, A. P. 1988 An outline of evolution of the hy- menopterous Insects (Order Vespida). Oriental Insects 22: 115-145. Ronquist, F. and G. Nordlander. 1989. Skeletal morphology of an archaic cynipoid, Ibalia rufipes (Hymenoptera: Ibaliidae). Entomologica Scandinavica, Supplement 33: 60 pp. Ross, H. H. 1937. A generic classification of the Nearctic sawflies (Hymenoptera, Symphyta). Illinois Biological Monographs 34: 173 pp. Saini, M.S., S.S. Dhillon, and R. Agarwal. 1982. Skeletomuscular differences in the thorax of winged and non-winged forms of Camponotus camelinus (Smith) (Hym., Formicidae). Deutche Entomologische Zeitschrifte (N.S.) 29: 447-458. Sharkey, M. J. and D. B. Wahl. 1992. Cladistics of the Ichneumonoidea (Hymenoptera). Journal of Hymenoptera Research 1: 15-24. Shcherbakov, D. E. 1980. [Morphology of the pterothoracic pleura in Hymenoptera. 1. Groundplan]. Zoologicheskn Zhurnal 59: 1644-1653. Shcherbakov, D. E. 1981. [Modifications of the pterothoracic pleura in Hymenoptera. 2. Modtficat ions of the groundplan]. Zoologicheskii Zhumal 60: 205-213. Snodgrass, R. E. 1910. The thorax of Hymenoptera. Proceed- ings of the United States National Museum 39: 37-91 + 16 pis. Snodgrass, R. E. 1927. Morphology and mechanism of the insect thorax. Smithsonian Miscellaneous Collections 80 (1): 108 pp. Snodgrass, R. E. 1942. The skeleto-muscular mechanisms of the honey bee. Smithsonian Miscellaneous Collections 103 (2) 120 pp. Tulloch, G. S. 1935. Morphological studies of the thorax of the ant. Entomologica Americana 15: 93-130. Weber, H. 1925. Derthorax derHornisse. Zoologische jarbucher 67: 1-100 + 4 pis. Weber, H. 1927. Die Gliedrung der Sternalregion des Tenthredinidenthorax. Ein Beitrag zur vergleichenden Morphologie des Insecktenthorax. Zeitschrift fur ivissenschaftliche Insektenbiologie 22: 161-198 + 6 pis. Whitfield, J. B. 1992. Phylogeny of the non-aculeate Apocrita and the evolution of parasitism in the Hymenoptera. journal of Hymenoptera Research 1: 3-14. Whitfield, J. B„ N. F. Johnson and M. R. Hamerski. 1989. Identity and phylogenetic significance of the 264 Journal of Hymenoptera Research metaposmotum in nonaculeate Hymenoptera. /birni/so/ Zessin, W. 1985. Neue oberliassiche Apocnta und die the Entomological Society of America 82: 663-673. Phylogenie der Hymenoptera (Insecta, Hymenoptera). Deutsche Entomologische Zeitschrift, N.S. 32:129-142. Appendix 1 List of abbreviations used for figures. Abbreviations used by Ronquist and Nordlander (1989) in brackets. af ap ex dl fb fp fu, FWB hP HWB If pn2 anterior furcal arm [91] PN, anterior process/axillary pph lever of laterophragma (pn;) sps cenchrus t, coxa t3 discrimenal lamella [dl] ume mesofurcal bridge vp mesofurcal process [91] lph metafurca 2ph fore wing base 60 horizontal plate of mesofurca [90] hind wing base hi lateral furcal arm [fu;] laterophragma of mesopostnotum (posterior face of lobe =t,) [pnap] mesopostnotum pseudophragma |pph] spina mesonorum metanotum upper mesepimeron ventral process of lateral furcal arm first phragma second phragma attachment line of mesopostnotum (PN2) to scutellum (t;) [60) attachment line of mesopostnotum (PN,) to metanotum (t,) [61] Appendix 2. Character states of the mesofurcal-mesopostnotal complex attributed to Hymenoptera. Characters and states are discussed in text. Superscript values refer to other observed states within a taxon. Characters Character? 1 2 3 4 5 6 7 8 9 111 11 12 1 2 3 A 5 6 / 8 9 10 11 12 Outgroup i) 0 ii 1) 0 0 0 0 0 II " 0 Proctotrupidae 5 2 3 1 1 1 0 2 0 0 Xyehnae 1 i) 1 •> 1 0 0 0 1 0 0 n Mvmandae 7 ~i 3 1 1 1 0 2 0 0 Macroxyelinae 1 ii 1 u 1 0 0 0 1 I) 11 il Pteromalidae 5 2 3 1 1 1 0 2 0 0 Pamphilndae 1 0 1 0 ■' 0 0 0 1 (1 11 ii Mymarommatidae7 2 3 1 1 1 0 2 0 0 Blasticotomidae 2 0 1 Ii 2 0 1 0 1 0 1.1 ii Platygastndae " T 3 1 1 0 2 0 0 Tenthredinidae 2 0 1 Ii 2 0 1 0 1 0 11 ii Scehonidae 5 3' 2 3 1 1 1 0 i 0 11 Dipnomdae 2 0 1 0 p 0 1 0 1 0 II (i Bracomdae 5 ■> 3 1 1 0 i 1) II Cimbicidae 2 0 2 0 2 0 1 0 1 0 0 ii Ichneumonidae 5 2 3 r 1 0 i 0 II Pergidae i 1 ? 1 ) 0 1 0 1 0 1) 0 Plumariidae 5 -) 3 1 1 0 2 0 0 Argidae 2 0 1 1) 2 0 1 0 1 11 II 1.1 Sclerogibbidae 5 -> 3 1 1 11 2 0 0 Cephidae 3 1) 1 1) : 0 1 1 11 II Embolerrudae 5 T 3 1 1 0 2 0 II Anaxvehdae -1 (1 1 1) 3 0 1 1 11 11 Dryinidae 5 2 3 1 1 0 ^ II (1 Sincidae 4 0 3 II 3 0 1 1 II 11 Bethylidae 5 T 3 1 1 0 2 0 0 Xiphvdrndae 5 1 1 1) 1 1 2 11 II Chrysidid.it' Orussidae 6 2 ? 1 3 1 -1 0 0 Amiseginae 5 1 2 3 1 1 0 -i 0 0 Stephanidae 5 2 1 3 0 2 II 0 Cleptinae 5 4 ~> 3 1 1 0 t 11 1 Trigonalvidae 5 2 1 3 1.1 -i II 0 Chrysidinae 5 4 T 3 1 1 0 -> 1 Megaspilidae 7 2 11 i 0 2 11 II Bembicidae 5 5 T 3 1 1 0 2 0 Ceraphronidae 7 2 11 i II 2 II II Larndae 5 5 2 3 1 1 0 2 II Aulacidae 5 ) 1 3 0 2 0 0 Heteri ^vnaeidae 5 5 2 I 3 1 1 0 T 0 Evanndae 5 2 1 3 1) -i II 0 Andremdae 5 6 2 I 3 1 1 0 2 II Gasteruptudae 5 -i 1 1 11 2 II 0 Anthophoridae 5 6 2 1 3 1 1 0 2 II [baliidae 5 2 0 1 II i 0 0 Apidae 5 6 -i 3 1 1 0 2 II Eucoilidae 5 2 11 3 II 2 II II Colletidae 5 b 2 1 3 1 1 1 0 T 0 Alloxystidae 5 j II 3 0 2 11 II Halictidae 5 b 2 1 1 1 1 1 0 2 0 Diaprndae Megachihdae 5 b 2 1 3 1 1 1 0 i 0 Ambositnnae 5 2 0 1 (1 -t 1.1 II Tiphndae 5 5 T 1 3 1 1 1 0 2 11 Diaprnnae 5 2 11 3 II 2 II 11 Sapvgidae 5 5 2 1 3 1 I 1 1) 2 2 Belytinae Mutilhdae 5 5 2 1 3 1 1 1 1 2 11 Acropiesta 5 i II 3 II 2 11 II Sierolomorphida* 5 5 2 1 3 1 1 1 11 2 II 2 Aneurrynchus r. 2 II 3 0 2 II 0 IVmpihdae 5 5 T 1 3 1 1 1 0 i 11 2 Oxylabis 5 2 1 1 1.1 2 II 11 Rhopalosomatidae5 5 2 1 3 1 1 1 1 1 0 0 Monomachidae 5 i II 3 11 2 II II Bradvnobaenid.u 5 4 2 1 3 1 1 1 1 -> 11 0 Ropronndae 5 2 1 3 11 2 0 0 Formicidae 5 1 2 1 3 1 1 1 0 2 II II Helondae 5 2 1 1 (1 2 II 0 Scoliidae 5 5 2 1 3 1 1 1 11 "> 1 2 Pelecinidae " 2 1 3 II 2 II 0 Vespidae s 5 2 1 3 1 1 1 11 2 II 2 Vanhornudae 5 -> (1 } II 2 II 0 Volume 3, 1 994 265 dl Figs. 1-2. Pamphiliidae. l,Pamphilius sp., lateral habitus of mesosoma; 2, Acantholyda sp., sagittal section of mesosoma. Muscle 127 removed and location indicated by dashed line. Abbreviations as listed in Table 1 and Appendix 1. 266 Journal of Hymenoptera Research 3 XYELIDAE sps2 4 XYELIDAE 5 PERGIDAE Figs. 3-5. 3, Xyela minor, oblique subsagittal section of mesothorax. 4, Xyela minor, skeletal components of Fig. 3, musculature removed. 5, Acordulecera sp., frontolateral view of MF-MPN complex. Abbreviations as listed in Table 1 and Appendix 1. Volume 3, 1 994 267 6 PAMPHILIIDAE 7 DIPRIONIDAE 8 CIMBICIDAE Figs. 6-8. 6, Pamphilius sp.: a, dorsal view of MF-MPN complex; b, lateral view of MF-MPN complex. 7, Diprion sp., supralateral view of MF-MPN complex. 8, Zaraea sp., dorsal view of MF-MPN complex. Abbreviations as listed in Table 1 and Appendix 268 Journal of Hymenoptera Research 9CEPHIDAE 10 ANAXYELIDAE 1 1 SIRICIDAE Figs. 9-11. MF-MPN complex. 9, Cephus cinctus: a, dorsal view; b, supralateral view. 10, Syntexis libocedrii: a, dorsal view; b, supralateral view. 11, Tremex sp.: a, dorsal view; b, lateral view. Abbreviations as listed in Table 1 and Appendix 1 Volume 3, 1 994 269 12 XIPHYDRIIDAE 1 3 ORUSSIDAE 14M0N0MACHIDAE 15 VANHORNIIDAE Figs. 12-15. MF-MPN complex. 12, Xiphydria abdommalis: a, dorsal view; b, supralateral view. 13, Orussus sayii, supralateral view. 14, Monomachus sp., supralateral view. 15, Vanhornia eucnemidarum: a, laterophragma; b, supralateral view. Abbrevia- tions as listed in Table 1 and Appendix 1. 270 Journal of Hymenoptera Research 16 SCELIONIDAE 17 ICHNEUMONIDAE 18 MUTILLIDAE 19 RHOPALOSOMATIDAE Figs. 16-19. MF-MPN complex. 16, Sparasion sp., supralateral view. 17, Megarhyssa sp. (Ichneumorudae): a, laterophragma (inner view); b, supralateral view. 18, Mutilhdae (male), supralateral view. 19, Rhcpalosoma sp., supralateral view. Abbrevia- tions as listed in Table 1 and Appendix 1. Volume 3, 1 994 271 20 BLASTICOTOMIDAE 23 TRIGONALYIDAE 21 TENTHREDINIDAE 24STEPHANIDAE 22 IBALIIDAE 25 MEGASPILIDAE 26GASTERUPTIIDAE 27 EVANIIDAE 28 PLATYGASTRIDAE Figs. 20-28. MF-MPN complex of Apocrita: right laterophragma (upper figure); mesofurca in dorsal view (lower figure). 20, Blasticotoma sp.; 21, Nematus sp.; 22, Ibalm sp.; 23, Orthogonalys pukhella; 24, Megischus sp.; 25, Megasptlus sp.; 26, Gasteruption sp.; 27, Hyptia sp.; 28, Isocybus sp. 272 J JNAI C)( HvMENOPTERA RfSLARi H JLji 29DIAPRIIDAE 30 HELORIDAE 31 PELECINIDAE 32 PROCTOTRUPIDAE 33BRACONIDAE, 34PLUMARIIDAE HYBRIZONTINAE 35 SCLEROGIBBIDAE 36EMBOLEMIDAE 37 DRYINIDAE 38BETHYLIDAE Figs. 29-38. MF-MPN complex of Apocnta: right laterophragma (upper figure); mesofurca in dorsal view (lower figure). 29, Oxylabis sp.; 30, Helorus sp.; 31, Pelecinus polyturator; 32, Proctotrupidae; 33, Hybnzon sp., 34, Plumana sp.; 35, Sclerogibbidae, 36, Embolemus sp.; 37, Gonatopodinae (male); 38, Goniozus sp. Volume 3, 1994 273 ap — 39 CHRYSIDIDAE, AMISEGINAE 40CHRYSIDIDAE, CHRYSIDINAE 41 VESPIDAE 42 POMPILIDAE 43F0RMICIDAE 45 APIDAE 44 TIPHIIDAE 46 LARRIDAE Figs. 39-46. MF-MPN complex of Apocrita: right laterophragma (upper figure); mesofurca in dorsal view (lower figure). 39, Amiseginae; 40, Chn/sis sp.; 41, Vespula sp.; 42, Aporinella galapagoensis; 43, Camponotus planus, 44, Myzmuni sp.; 45, Tngona sp.; 46, Larra sp. 274 Journal of Hymenoptera Research PN2 47 XYELIDAE 48 PAMPHILIIDAEl ■*-,. ^ . ** 50 XIPHYDRIIDAE Figs. 47-50. Mesopostnotum, dorsal view. 47, Pleuroneura sp.; 48, Pamphilius sp, 49, Cephus cinctus; 50, X,PMn« abfommoto Abbreviat.ons as listed in Appendix 1. Split f.gures are magnifications of highlighted area in upper figure. Volume 3, 1 994 275 Figs. 51-54. 51-52, Mesopostnotum in lateral view with magnification of connection to upper mesepimeron. 51 , Pamphilius sp.; 52, Cqyhus ductus. Arrow indicates articulation of laterophragma with upper mesepimeron. 53-54, MF-MPN complex of Lasioglossum, dorsolateral view. 53, fused meso- and metafurca; 54, magnification of axillary lever (ap). Other abbreviations as listed in Table 1 andAppendix 1. Split figures are magnifications of highlighted area in upper figure 276 Journal of Hymenoptera Research Fig. 55. Phylogenetic hypothesis for extant Hymenoptera based on Rasnitsyn (1988) and Gibson (1985, 1993), with characters of the MF-MPN complex superimposed. Characters and states discussed in text. Solid circles are unique apomorphies; shaded circles indicate convergence; open circle indicates reversal. Xyelidae are not demonstrably paraphyletic, but under some optimizations character 5:2 is treated as a synapomorphv of Macroxyehnae and the remaining Hymenoptera (dotted line to alternate placement, see text for discussion). Character 3:? was not optimized, and this tree is two steps longer than found in the parsimony analyses (see text). Volume 3. 1994 277 Fig. 56. Phylogenetic hypothesis for Apocrita based on Rasnitsyn (1988), with characters of the MF-MPN complex superim- posed. Characters and states discussed in text. Placement of Peradenndae and Vanhorniidae by Masner (pers. comm., 1993). Superfamily names follow Huber and Goulet (1993), those names in parentheses were not examined for internal characters and question marks indicate unknown character states. Solid circles are unique apomorphies; shaded circles indicate convergence; open circle indicates reversal. Multiple states indicate groundplan first followed by proposed derived state changes; changes listed in Appendix 2. J. HYM. RES. Vol. 3, 1994, pp. 279-302 A Taxonomic Study of the genus Ascogaster in China" (Hymenoptera: Braconidae: Cheloninae) Yuquing Tang and Paul M. Marsh (YT) Biological Control Research Institute, Fujian Agricultural College, Fuzhou, Fujian, People's Republic of China; (PMM) Cooperating Scientist, Systematic Entomology Laboratory, U. S. Department of Agriculture (present address: P. O. Box 384, North Newton, Kansas 67117) Abstract. — The chelonine genus Ascogaster is studied for the first time from the largely unknown area of mainland China and Taiwan. Twenty-three species are described and keyed, of which 1 3 are new to science and six are recorded for the first time from this area. INTRODUCTION MATERIALS AND METHODS Ascogaster Wesmael is a cosmopolitan bra- conid genus whose species are parasitic upon microlepidoptera, predominantly Tortricidae. Some species have been recorded frequently as parasitoids of several economically important in- sect pests and have been considered for use in biological control programs. About 110 species of Ascogaster have been recorded in the world. Huddleston (1984) revised the 30 Palaearctic spe- cies; Tobias (1986a, 1986b, 1988) and Papp (1989) added 12 species to this region. Shaw (1983) re- viewed 11 species for the Nearctic Region and described a related new genus, Leptodrepana (4 species). Walker and Huddleston (1987) recorded 12 species of Ascogaster in New Zealand. The fauna of the Indo-Australian Region is largely unknown, although Baker (1926) and Szepligeti (1905, 1908) described a few species. The Ascogaster fauna of China is especially poorly studied with only five species recorded, three from mainland China (Fahringer 1934; He, et al 1989) and two from Taiwan (Sonan 1932). In this work 23 species of Ascogaster in China are treated, of which 13 species are new to science and six are recorded for the first time from China. "Throughout this paper, China includes mainland China and Taiwan. Approximately 1,200 specimens of Chinese Ascogaster were examined in this study, mostly from the collections of the Fujian Agricultural College, the Zhejiang Agricultural University, the Zoology Research Institute of Academia Sinica, the Taiwan Agricultural Research Institute and the American Entomological Institute. The fol- lowing acronyms are used to identify collections that provided specimens for this study and several as the depositories for type material: AEIG American Entomological Institute, Gainesville, Florida, USA BM Bishop Museum, Honolulu, Haiwaii, USA CNC Candian National Collection, Agriculture Canada, Ottawa, Canada ELKU Entomological Laboratory, Kyushu University, Fukuoka, Japan FAC Biological Control Research Institute, Fujian Agricul- tural College, Fuzhou, China HNHM Hungarian Natural History Museum, Budapest, Hungary IRSNB Institut Royal des Sciences Naturelles de Belgique, Brussels, Belgium TARI Taiwan Agricultural Research Institute, Wufeng, Tai- wan UEI Entomological Institute, Hokkaido University, Sapporo, Japan USNM National Museum of Natural History, Washington, D. C, USA ZAU Department of Plant Protection, Zhejiang Agricul- tural University, Hungzhou, China 2S0 Journal of Hymenoptera Research ZRJ Zoology Research Institute, Academia Simca, Beijing, China We also examined available Palaearctic and Indo-Australian material in the U. S. National Museum of Natural History and some type mate- rial from other institutions as follows: USNM - abdominatorDahlbom,albitarsusRemhard,annulans Nees von Esenbeck, atamiensis Ashmead (holo- type, synonymized with bidentida Wesmael by Huddleston 1984), bicarinata Herrich-Shaffer, bidentida Wesmael, brevicornis Wesmael, canifrons Wesmael, cinctus Baker (holotype), coelioxoides (Baker) (holotype), consobrnw Curtis, detectus Baker (holotype), distinctus Baker (holotype), fullawayi (Baker) (type series), inconspicuus Baker (holo- type), intensus Baker ( holotype), laeviventris Baker (holotype), longulus Baker (holotype), luzonensis Baker (holotype), maculaticeps Baker (holotype), malayanus (Baker) (holotype), modestus Baker (ho- lotype), philippinensis Baker (type series), quadradentata Wesmael, reticulata Watanabe (type series), rufidens Wesmael, rufipes (Latreille), vividus Baker (holotype); BM - argentea Fullaway (syntypes); rugosa Fullaway (holotype); IRSNB - armata Wesmael (lectotype), bidentida Wesmael (lectotype), brevicorinis Wesmael (lectotype), canifrons Wesmael (lectotype); limitatus Wesmael (syntypes, synonymized with scabricula (Dahlbom) by Huddleston 1984), quadridentata Wesmael (type- series), rufidens Wesmael (type-series), varipes Wesmael (type series); UEI - epinotiae Watanabe (type series, synonymized with quadridentata Wesmael by Huddleston 1984); reticulata Watanabe (type series); ELKU -longicornis Huddleston (holo- type, synonymized with formosensis Sonan, new synonymy); HNHM - nilena Papp (holotype). The general braconid morphological termi- nology and measurements used in this study are mostly after van Achterberg (1988). The terminol- ogy used for wing veins is illustrated and explaned in Sharkey ( 1988). Microsculprure terms are based on Harris (1979). Two character states of the ocel- lar triangle are defined after Huddleston (1984) as follows: where a straight line drawn between the anterior borders of posterior ocelli also touches the anterior ocellus, the ocelli are referred to as being 'on line'; where a line between the posterior ocelli passes behind the anterior ocellus without touching it, the ocelli are referred to as being 'not on line'. The measurements CO and OD refer, respectively, to the ocellar-ocular distance and the diameter of the posterior ocellus. The ratio CL/ CW is carapace length (CL) to carapace width (CW). Shaw (1983) and Huddleston (1984) dis- cussed the important taxonomic characters of Ascogaster in detail and we have used most of the specific characters used by Huddleston (1984) for the description of new species in this study (see discussion under generic diagnosis). All dates from labels on specimens examined have been stan- dardized thus, 1-1-93, to avoid confusion. BIOLOGY Members of Ascogaster, like all the chelonines, are internal egg-larval parasitoids of microlepi- doptera, principally Tortricidae. They lay their eggs into the host egg, but their larval develop- ment is delayed at the first instar until the host larva is mature. Little detailed work has been done on the biology of Ascogaster. Huddleston (1984) and Shaw (1983) reviewed the hosts of Ascogaster for the Palaearctic and Nearctic Regions respec- tively. Cox (1932), Rosenberg (1934) and Boyce (1936) gave fairly detailed accounts of the biology of A. quadridentata Wesmael. A research group in Japan has recently studied in some detail host- searching, host-preference, mating and oviposi- tion behavior of A. reticulata W atanabe, an impor- tant parasitoid of the smaller tea tortix, Adoxophyes sp., in Japan (Kainoh 1986, 1988; Karnoh, el al. 1982; Kainoh, et a/.1991; Kainoh and Tamaki 1982, 1988; Kainoh, et fl/,1990; Kainoh, et al. 1989; Kamano, et al. 1989; Kawakami 1985; Kawakami and Kainoh 1985, 1986). Little is known at present about the biology of the Chinese Ascogaster species, how- ever He et al. (1987) gave a full host list of A. reticulata in mainland China. SYSTEMATICS The Cheloninae are distinguished from all other subfamlies of Braconidae by the combina- tion of having the first three metasomal terga fused to form a carapace (Figs. 21-26), having a complete postpectal carina and three submarginal cells in the forewing (Fig. 1). Two tribes of the Cheloninae, Chelonini and Phanerotomini, are represented in the Palaearctic and Indo- Austrialian regions. In the Phanerotomini the carapace is di- vided into three tergites by two transverse su- Volume 3. 1 994 281 tures; in the Chelonini the carapace has no trace of sutures. The Chelonini are represented in the Palaearctic and Indo-Australian regions by three genera, Ascogaster Wesmael, Megascogaster Baker and Chelonus Panzer. The former two genera can be separated from Chelonus by having the first submarginal cell always separated from the discal cell by vein 1-Rs+M (Rs+M of some authors) (Fig. 1). Many authors (e.g. Baker 1926; Shaw 1983; Huddleston 1984; Walker and Huddleston 1987; van Achterberg 1990b; Zettel 1990) state that Ascogaster is also characterized by having hairless eyes. However, we have described one new spe- cies, A. setula from Taiwan, with eyes densely and distinctly setose (Fig .8), but with most of the other characters of typical,4scogasfer. We include it within Ascogaster until more species with this unusal character have been found. Megascogaster Baker is distinguished from Ascogaster only by having a very elongate carapace (distinctly longer than head and thorax) and a slender stigma in the forewing. The taxonomic position of Megascogaster Baker must wait until more material from the Indo- Australian region has been studied. Shaw (1983) cited thirteen characters for distinguishing Ascogaster and Leptodrepana, however intermedi- ate forms of most characters occur in many unde- scribed Indo-Australian Ascogaster which we have examined and we have chosen to follow van Achterberg's synonymy. This concept makes Ascogaster an apparent polyphyletic group and we feel that a thorough phylogenetic study of the entire genus is needed before the correct position of Leptodrepana and the many other unusual forms from the Indo-Australian Region can be settled. That is beyond the scope of this limited study which aims only to identify species from the pre- viously unknown area of China. Ascogaster Wesmael Ascogaster Wesmael 1835:226. Type-species: Ascogaster mstabilis Wesmael \-abdonunator (Dahlbom)], subse- quently desigated by Foerster (1862). Cascogaster Baker 1926:482. Type-species: Cascogaster fullawayi Baker, original desigation, syn. by Watanabe (1937). Leptodrepana Shaw 1983:37. Type-species: Leptodrepana opuntiae Shaw, original designation; syn. by van Achterberg (1990b). Diagnosis. — Distinguished from all other bra- conid genera by having: (a) the first three metasomal terga fused into a rigid carapace with- out transverse sutures; and (b) the forewing with three submarginal cells and vein 1-Rs+M present, separating the first submarginal and discal cells. KEY TO THE ASCOGASTER SPECIES OF CHINA 1 Face punctate or rugose-punctate, occasionally nearly smooth; mesonotum usually predominantly punctate, notauli usually distinct 2 Face strongly areolate or areolate-rugose, at least finely areolate-rugose; mesonotum usually coarsely areolate-rugose so that notauli often are indistinct 13 2(1) Setae on upper part of face pointing upwards; ocellar triangle acute, ocelli usually not on line or almost on line; propodeum without mediodorsal and apicolateral tubercles 3 Setae on upper part of face pointing downwards; ocellar triangle always obtuse, ocelli usually on line; propodeum with distinct mediodorsal and apicolateral tubercles 6 3(2) Clypeus with conspicuous dentate flanges laterally; legs completely black or sometimes fore tibia reddish-brown; ovipositor sheaths broad, knife-like semenovi Telenga Clypeus without dentate flanges laterally; legs mostly yellow or yellow-brown; ovipositor sheaths not conspicuously broad and knife-like 4 4(3) Clypeus with a large median incision apically; female carapace with a conspicuous dorsal prominence at base ditnorpha Tang and Marsh, new species Clypeus without any medial apical tubercle or incision; carapace without a dorsal prominence at base 5 5(4) Vein r of forewing about 1.0-1.2 times as long as 3-RS; hind coxa of male completely yellow townesi Tang and Marsh, new species Vein r of forewing about twice as long as 3-RS; hind coxa of male mostly black yunnanica Tang and Marsh, new species 282 Journal of Hymenoptera Research 6(2) Eyes distinctly setose; antennae short, with 26-29 flagellomeres; body 2.9-3.1 mm setula Tang and Marsh, new species Eyes without setae or at most with only a few minute setae; antennae longer, generally with more than 30 flagellomeres; body generally more than 3.2 mm 7 7(6) Mesopleuron completely coarsely areolate-rugose so that precoxal suture not distinct; antennae very long, 45-49 flagellomeres; apical border of clypeus weakly marginate medially ..fortnosensis Sonan Mesopleuron at least in part smooth or punctate, precoxal suture usually distinct (except vanpes); antennae shorter, at most with 40 flagellomeres; apical border of clypeus either produced or with one or two small tubercles 8 8(7) Apical border of clypeus transversely impressed and produced medially but without medial excision or tubercles 9 Apical border of clypeus with one or two distinct teeth medially 11 9(8) Hind coxa strongly strigate; carapace short, oval in dorsal view; mandible at base with a deep, semicircular depression varipes Wesmael Hind coxa largely smooth or finely punctate; carapace elongate, clavate in dorsal view; mandible at base with vertical, parallel-sided depression 10 10(9) Legs yellow except hind coxa at base black, apex of hind femur and tarsi infuscate consobrina Curtis Legs black except tibiae at base and basal tarsus of hind leg ivory or pale yellow albitarsus Reinhard 11(8) Apical border of clypeus with two medial tubercles, generally with a small excision between them; if excision absent, then medial clypeal border with a narrow, spatulate projection, but never a medial tooth arisanica Sonan Apical border of clypeus with a single medial pointed toot 12 12(11 ) Carapace short, oval, broadest at about middle point in dorsal view; ventral opening of carapace shorter in front of apex; hind coxa completely yellow perkinsi Huddleston Carapace longer, clavate, broadest in posterior third in dorsal view; ventral opening of carapace long at apex of carapace; hind coxa predominantly black lint Tang and Marsh, new species 13(1) Carapace elongate, CL/CW more than 2.0 14 Carapace shorter, usually oval, CL/CW generally 1.7 or less, never more than 2.0 15 14(13) Hind coxa strigate and black; ventral opening of carapace short, reaching only slightly beyond middle of carapace grandis Tang and Marsh, new species Hind coxa smooth or finely punctate, yellow; ventral opening of carapace longer, reaching apical 1 /4 of carapace macrogaster Tang and Marsh, new species 15(13) Gena long and straight in face view; hind coxa strigate and yellowc/iooi Tang and Marsh, new species Gena more or less rounded in face view; hind coxa often punctate or if strigate, then always mostly black 16 16(15) Interantennal carina strongly raised into an erect triangular flange between scapes 17 Interantennal carina present but never strongly raised into a triangular flange between scapes 19 17(16) Carapace rounded, without a tubercle apically; clypeus matte, rugose-punctate, its apical border straight with no medial tooth annatoides Tang and Marsh, new species — Carapace apically pointed, with a tubercle at apex; clypeus smooth, sparsely punctate, its apical border pointed with a medial tooth 18 18(17) Carapace oval; face and temple areolate-rugose acutus Tang and Marsh new species — Carapace acutely pointed; face and temple strongly and coarsely rugose fullawayi (Baker) 19(16) Hind coxa smooth, or finely punctate, yellow or infuscate at base; hind tibia with a medial pale colored band; clypeus transversly impressed apically, without a medial tooth or if with a medial tooth, then lower part of face smoother than upper part 20 Hind coxa strigate, at least at part, always black; hind tibia generally completely black or pale at base, but never with a medial pale-colored band medially; clypeus always with a distinct medial tooth; face evenly areolate or strongly areolate-rugose 21 20(19) Clypeus matte, rugose-punctate, straight apically, with no medial tooth; face strongly areolate reticulata Watanabe Clypeus smooth, finely and sparsely punctate, with a distinct medial apical tooth; face finely areolate- rugose rugulosa Tang and Marsh, new species 21(19) Anterior corners of middle mesonotal lobe swollen, notauli distinct; palpi black gibbosa Tang and Marsh, new species Volume 3. 1994 283 — Anterior corners of middle mesonotal lobe not swollen, notauli indsitinct; palpi yellow or reddish-brown 22 22(21). Temple short, about equal in width to eye in dorsal view; carapace short, oval in dorsal view, deep in lateral view quadridentata Wesmael — Temple rounded, distinctly longer than eye in dorsal view; carapace longer, clavate in dorsal view, not so deep in lateral view hei Tang and Marsh, new species Ascogaster acutus Tang and Marsh, new species Fig. 23 Female. — Length of forewing 2.5-2.7 mm, of body 3.0-3.2 mm. Head. — Antenna with 28-32 flagellomeres, slightly dilated medially, medial flagellomeres slightly broader than long, apical ones longer than broad. Temple slightly contracted behind eyes, distinctly longer than length of eye in dorsal view. Occiput strongly concave. Ocelli small, on line, 00= 4.0 OD. Frons strongly depressed, finely rugose or smooth, medial carina distinct, extend- ing from uper part of face betwen antennae to anterior ocellus and strongly expanded between antennae into an erect triangular lamella. Eyes protuberant, glabrous. Face slightly protuberant, about twice as broad as high, strongly areolate- rugose. Clypeus punctate, strongly divided from face, apical border produced medially into a small pointed tooth. Mesosoma. — Pronotum slightly projecting in front of mesonotum, dorsolaterally rugose-punc- tate. Notauli not distinct; mesonotum completly coarsely rugose. Precoxal suture indistinguish- able; mesopleuron strongly rugose. Propodeum coarsely areolate-rugose, divided by a transverse carina which is rasied into a medial pair and a lateral pair of dentate flanges. Hind coxa strigate. Vein r of forewing about 1.0-1.5 times as long as 3- RS. Metasoma. — Carapace (Fig. 23) short, with pointed tubercle at apex, areolate-rugose. Ventral opening of carapace longer, not distinctly in front of apex. Ovipositor concealed. Color. — Black; carapace yellow at base; all legs yellow except hind coxa at base, tibia at apex and tarsi black. Male. — Same as females except antenna not dilated medially, smaller yellow spot at basal carapace. Holotype Female. — BEIJING CITY: Mentougou, VIII-25-81, M. S. Shi. Deposited in ZAU. Paratypes.— BEIJING CITY: 1 female, Lugouqiao, IX-57, T. L. Chen; 1 male, IX-78, J. H. He. GUIZHOU PROVINCE: 1 female, Guiyang, V-13-85, Q. H. Luo. FUJIAN PROVINCE: 1 female, Shaowu, VII-20-45, H. F. Chao; 1 female, Shanhan, VII-19-21-88, Y. Ma. TAIWAN: 1 female Duruanpi, VIII-5-82, K. C. Chou and C. N. Lin; 1 female, Wushe, IV-13-83, H. Townes; 1 female, Wushe, rV- 19-83, H. Townes; 1 female, Wushe, V-3-83, H. Townes; 2 females, Wushe, V-10-83, H. Townes; 1 male, Wushe, V-15-83, H. Townes. Deposited in AEIG, FAC, TARI, ZAU and ZRI. Host. — Unknown. Distribution. — China (Beijing City, Guizhou Province, Fujian Province),Taiwan. Remarks. — This species is very close tofidlawayi (Baker) from which it can be distinguished by the characters mentioned in the key. Etymology. — The specific name is from the Latin acutus meaning pointed in reference to the pointed tubercle at the apex of the carapace. Ascogaster albitarsus Reinhard Ascogaster albitarsus Reinhard 1867: 364. Lectotype male, POLAND: Gdansk (designated by Huddleston 1984). Shenefelt 1973: 815; Huddleston 1984: 364; Papp 1989: 297. Ascogaster leptopus Thomson 1874: 584. Lectotype female, SWEDEN (designated by Huddleston 1984 and syn. by Hellen 1953) Diagnosis. — Length of foewing 3.3-3.5 mm, of body 3.9-4.0 mm. Antenna with 36-38 flagellomeres, slightly dilated medially (female); ocelli almost on line; face about 1 .5 times as broad as high, rugose-punctate; clypeus with its apical border produced medially and slightly reflexed forwards; notauli distinct, fovealate, coalescing posteriorly in a large areolate-rugose area, rest of mesonotum densely areolate-punctate; precoxal suture strongly areolate-rugose anteriorly, rather weak posteriorly, mesopleuron anterodorally strongly rugose, rest of mesopleuron finely punc- tate; hind leg predominately dark with base of 284 Journal of Hymenoptera Research tibia and tarsus pale-yellow; hind coxa finely punc- tate; carapace long, clavate and widest in posterior third. Specimens Examined.— HEBEI PROVINCE: 2 females. SICHUAN PROVINCE: 2 females. Additional Specimens Examined. — WESTERN EUROPE: 1 male, det. by Huddleston, 1983. Host. — A tortricid moth on pine tree. Distribution. — China (Heibei Province, Sichuan Province), Korea and some European countries. Remarks. — Our specimens from China have the mesopleuron with a polished impuncate area posteroventrally, but otherwise agreeing with the redescription of Huddleston (1984) Ascogaster arisanica Sonan Fig. 3 Ascogaster arisanicus Sonan 1932: 79. Holotype male, TAI- WAN: Alishan (=Arisan). Watanabe 1937: 77 (as syn. of rufipes (Latreille), Chou 1981 : 72 (notes on the locality of types). Ascogaster arisanica Sonan: emended by Huddleston 1984: 365. Diagnosis. — Length of forewing 2.9-3.2 mm, of body 3.3-3.7 mm. Antenna with 34-38 flagellomeres, moderately dilated medially (fe- male); ocellar triangle obtuse, ocelli almost on line; face (Fig. 3) about 1.5 times as broad as high, densely punctate or rugose-punctate; apical bor- der of clypeus produced, with a distinct medial excision flanked by two small tubercles; notauli shallow, coarsely rugose, coalescing posteriorly in an areolate-rugose area, rest of mesontum punc- tate or rugose-punctate; precoxal suture broad, shallow, foveolate; rest of mesopleuron sparsely punctate except dorsally rugose; hind coxa finely punctate; carapace elongate, broadest in distal half; ventral opening of carapace conspicuously in front of apex. Specimens Examined.— ZHEJIANG PROV- INCE: 16 females, 9 males. SICHUAN PROV- INCE: 2 females. YUNNAN PROVINCE: 1 male. GUANGXI PROVINCE: 1 female. HAINAN PROVINCE: 1 female. TAIWAN: 115 females, 35 males. Host. — Unknown. Distribution. — China (Zhejiang Province, Sichuan Province, Yunnan Province, Guangxi Province, Hainan Province), Taiwan, Japan. Remarks.— Huddleston (1984) stated that "This species is structurally very close to bidentula, but it can be distinguished by the more massive head and the more elongate carapace." However, our specimens generally have the heads not so mas- sive as described by Huddleston. In Chinese speci- mens examined here, the temple is only 1.0-1.3 times as long as eye in dorsal view. Sonan (1932) named this species as arisanicus, however, the holotype is labelled 'Ascogaster artsensis Sonan'. Ascogaster armatoides Tang and Marsh, new species Fig. 19 Female. — Length of forewing 3.5-3.6 mm, of body 4.2-4.4 mm. Head. — Antenna with 42 flagellomeres, mod- erately dilated medially, medial segments about as broad as long. Temple constricted behind eyes, slightly shorter than length of eye in dorsal view. Occiput strongly concave. Ocelli very small, on line, OO = 4.0 OD. Frons strongly excavate behind antennae, smooth, with a distinct medial carina which is expanded between antennae into an erect triangular flange (Fig. 19). Eyes glabrous, strongly protuberant. Face slightly protuberant, about 1.5 times as broad as high, strongly areolate-rugose. Clypeus dull, rugose-punctate, not very distinctly divided from face; apical border convex, not pro- duced medially and without distinct tooth or tu- bercle. Mesosma. — Pronotum slightly projecting in front of mesonotum, areolate-rugose laterally Notauli indistinct; mesonotum strongly rugose. Precoxal suture indistinct; mesopleuron com- pletely areolate-rugose. Propodeum strongly ar- eolate-rugose, divided into dorsal and posterior surfaces by a transverse carina which is raised into a pair of medial blunt teeth and laterally a pair of less stout teeth. Hind coxa strongly strigate. Vein r of forewing about as long as 3-RS. Metasotna. — Carapace short, CL/CW = 1.6, oval in dorsal view, somewhat pointed in lateralview, areolate-rugose. Ventral opening of carapace short, in front of apex. Ovipositor convealed. Male. — Unknown. Holotype Female.— GUANGXI PROVINCE: Longzhou, V-18-82, J. H. He. Deposited in ZAU. Paratypes. — INDIA: 4 females, Ammatti, S. Coorg, 11-52, P. S. Nathan. Deposited in CAN, USNM. Volume 3, 1994 285 Host. — Unkown. Distribution. — China (Guangxi Province), In- dia. Remarks. — This species is very close to arniata Wesmael from which it can be distinguished by its longer antenna and less massive head. Etymology. — The specific name is in reference to the similarity of this species to arniata. Ascogaster chaoi Tang and Marsh, new species Female. — Length of forewing 3.0-3.4 mmm, of body 3.5-4.0 mm. Head. — Antenna with 40-42 flagellomeres, strongly dilated medially, tapered at apex, medial segments slightly broader than long, apical seg- ments about as long as broad. Temple constricted behind eyes, slightly longer than length of eyes in dorsal view. Occiput deeply concave. Ocelli small, OO =3.5 OD, on line. Frons moderately depressed, finely punctate, with distinct medial carina. Eyes slightly protuberant, glabrous without distinct setae. Gena long and straight in face view. Face slightly protuberant, about 1.5 times as broad as high, areolate-rugose. Clypeus protuberant, punc- tate, distinctly divided from face; apical border slightly impressed laterally, produced medially into stout dentate flange. Mesosoma. — Pronotum projecting in front of mesonotum, dorsolaterally areolate-rugose. Notauli indistinguishable; mesonotum strongly areolate-rugose. Mesopleuron completely rugose so that precoxal suture not distinct. Propodeum strongly areolate-rugose, with a medial transverse carina raised into a pair of medial stout dentate and a pair of lateral less stout dentate. Hind coxa strigate. Vein r of forewing 1.0-1.5 times as long as 3-RS. Metasoma. — Carapace shorter, CL/CW = 1.5- 1.7, somewhat pointed in dorsal view, rounded and deep in lateral view, areolate-rugose. Ventral opening of carapace short, distinct in front of apex. Hypopygium short. Ovipositor sheath clavate. Color. — Black; carapace yellow at base; all legs yellow except hind femur and tibia black at apex. Male. — Unknown. Holotype Female.— FUJIAN PROVINCE: Huanggangshan, VII-14-85, Y. Q. Tang. Deposited inFAC. Paratypes.— FUJIAN PROVINCE: 1 female, same as holotype; 6 females, Tongmuguan (Wuyishan), VII-11-82, J. C. Huang; 1 female, Wuyishan, VI-80, H. F. Chao; 2 females, Shanguan (Wuyishan), VII-5-85, D. H. Huang and G. Zheng; 1 female, Wuyishan, VII-15-86, J. S. Wang. GUANGXI PROVINCE: 1 female, Tainlin, V-30- 82, J. H. He ; 1 female, Longsheng, VI-24-84, J. H. He. ZHEJIANG PROVINCE: 5 females, Songyang, VII-15-89, J. H. He and H. L. Chen; 1 female, Tianmushan, VI-17-83, Y. Ma. ANHUI PROV- INCE: 1 female, Huangshan, VII-30-86, S. C. Zhang. JILING PROVINCE: 1 female, Dongliao, VII-20- 31-88, X. M. Luo. Deposited in FAC, USNM, ZAU. Host. — Unknown. Distribution. — China (Jiling Province, Anhui Province, Zhejiang Province, Guangxi Province, Fujian Province). Remarks. — This species is close toquadridentata from which it can be easily distinguished by its longer antenna, straight and long gena, and yel- low hind coxa. Etymology. — This species is named for Prof. H. F. Chao, Fujian Agricultural College, in apprecia- tion for his outstanding contribution to the tax- onomy of parasitic Hymenoptera in China. Ascogaster consobrina Curtis Figs. 20, 27 Ascogaster consobrmus Curtis 1837: 672. Holotype male, GREAT BRITAIN: England. Shenefelt 1973: 819; Huddleston 1984: 367; Papp 1989: 297. Diagnosis. — Antenna with 32-34 flagellomeres, slightly dilated medially (female); ocellar triangle obtuse, ocelli almost on line; face (Fig. 20) about 1 .5 times as broad as high, areolate-punctate, hairs downwards; apical border of clypeus not retracted, medially produced into a broad, blunt point, with- out impression or tubercle; notauli distinct, foveolate, rest of mesonotum punctate except with a broad areolate-rugose area posteromedially; precoxal suture distinct foveolate, mesopleuron (Fig. 27) dorsally finely rugose-punctate except for a polished and impunctate area just dorsal precoxal suture, ventrally always sparsely punctate, shin- ing; carapace rather long, clavate, widest in poste- rior third, with a downwardly directed anterior flange; hind coxa punctate, sometimes weakly rugose dorsaly; hind leg yellow except coxa at base black, apex of femur and sometimes of tarsus infuscate. Length of forewing 3.3-3.6 mm, of body 3.8-4.3 mm. 286 Journal of Hymenoptera Research Specimens Examined.— ZHEJIANG PROV- INCE: 1 female. TAIWAN: 89 females, 12 males. Additional Specimens Examined. — Western Eu- ropean countries, 6 females, 10 males, det. by Huddleston, 1983. Hosts. — No reared material was examined. Shenefelt (1973) recorded Chelonia caja and Gelechia vulgella as its hosts, but these records need to be confirmed. Distribution. — China (Zhejiang Province), Tai- wan., Japan, Korea, and several European coun- tries (see Huddleston 1984: 367) Remarks. — All specimens examined agree well with the redescription given by Huddleston (1984) except for the slightly shorter medial flagellar segments of the female, which are generally about as long as broad. Ascogaster dimorpha Tang and Marsh, new species Fig. 4 Female. Length of forewing 3.8 mm, of body 4.4 mm. Head. Antenna, with 26 flagellomeres, flagel- lum slightly dilated medially, all segments longer than broad. Temple more or less rounded behind eyes,, slightly longer than eye in dorsal view. Oc- ciput strongly concave. Ocelli large, OO = 4.5 OD; ocellar triangle acute, ocelli not on line. Frons behind antenna moderately excavated, smooth anteriorly, rugose-punctate posteriorly. Eyes mod- erately protuberant, nearly glabrous with only few scattered minute setae. Gena contracted be- low in face view. Face (Fig. 4) slightly protuberant, about twice as broad as high, densely and regu- larly rugose-punctate, moderately hairy, the hairs on the upper part of the face pointing upwards. Clypeus moderately protuberant, rugose-punc- tate, but less densely than that of face, its apical border strongly incised medially. Mesosoma. — Pronotum projecting in front of mesonotum, rugose-foveolate dorsolaterally. Notauli not very distinct, mesonotum finely punc- tate except with an areolate-rugose area posteromedially. Precoxal suture shallow foveolate, rest of mesopleuron punctate. Propodeum slightly impresed and punctate medi- ally, strongly areolate-rugose laterally, but with- out distinct tubercles. Hind coxa smooth and shining. Vein r of forewing about 2.5 times as long as 3-RS. Metasoma. — Carapace long, CL/CW = 2.2, ar- eolate-rugose, but finely and sparsely punctate apically, with a distinct dorsal prominence at base. Hypopygium large and broad. Ovipositor long slender, upcurved. Color. — Black; legs yellow-brown except hind coxa at base and tarsis black, and hind tibia infuscate with a pale-yelow band medially; cara- pace with yellow-pale spots at anterolateral sides; papli yellow-pale, Males. — Same as female except antenna longer, not dilated medially, with 30-31 flagellomeres and carapace more slender, CL/CW = 2.3-2.5, more rounded and deeper apically, without a dorsal prominence at base. Length of forewing 3.6-3.9 mm, of body 4.3-4.5 mm. Holotype female.— ZHEJIANG PROVINCE: Songyang, VII-18-31-89, J. H. He. Deposited in ZALL Paratypes.— ZHEJIANG PROVINCE: 3 males, Songyang, VII-15-17-89, J. H. He. TAIWAN: 5 males, Meifeng 2150 m, -24-26-81, K. S. Lin and W. S. Tang (TARI); 1 male, Meifeng, VI-22-26-83, K. S. Lin and S. C. Lin. Deposited in TARI, ZAU. Hosts. — Unknown. Distribution. — China (Zhejiang Province), Tai- wan. Remarks. — This species belongs to the caucasica- group (sensu Huddleston 1984) by virtue of its punctate face, hair on the upper part of the face pointing upwards and the bidentate clypeus. It appears to be related to caucasica Kokujev and bicarinata Herrich-Schaffer, but is easily distin- guished from the later two species by the charac- teristics of the carapace. Etymology. — The specific name refers to the sexual dimorphism of the metasomal carapace. Ascogaster formosensis Sonan Fig. 5 Ascogaster formosensis Sonan 1932:78. Holotype female, TAI- WAN: Arisan, Kunkiko. Watanabe 1937:76; Shenefelt 1973: 822; Chou 1981: 72 (notes on the locality of types); Papp 1989: 297. Ascogaster formosanus (!): Watanabe 1934: 198. Ascogaster longicomis Huddleston 1984: 368. Holotype fe- male, JAPAN: Mt.Tachibana (ELKU). New synonymy. Diagnosis. — Antenna with 45-49 flagellomeres, medial segment weakly dilated; ocelli on line or Volume 3, 1 994 287 almost on line; face (Fig. 5) rugose-punctate, 1.2- 1.5 times as broad as high; clypeus with its apical border rounded except its medial 1/4 weakly emarginate; notauli foveolate, rest of mesonotum densely punctate except posteromedially areolate- rugose; mesopleuron coarsely rugose so that precoxal suture not distinct; propodeum areolate- rugose with 4 prominent tubercles; hind coxa finely punctate; carapace always yellow at base, oval in dorsal view, clavate in lateral view, ventral open- ing of carapace distinctly in front of apex. Larger species, length of forewing 5.0-5.4 mm, of body 5.8-6.4 mm. Specimens Examined.— YUNNAN PROVINCE: 1 female. TAIWAN: 15 females, 7 males. Additional Specimens Examined. — NEPAL: 1 female, 1 male. INDIA: 6 females, 6 males. JAPAN: Holotype of longicorinis Huddleston. Hosts. — Unknown. Distribution. — China (Yunnan Province), Tai- wan, India, Japan, Nepal. Remarks. — This species is easily distinguished from other species of Chinese Ascogaster by its conspicuously long antennae and by the charac- teristics of its face and clypeus. It is very similar to philippinensis Baker from which it differs only in the sclupture of the mesonotum. We treat it here as a valid species given no intermediate forms are known. Ascogaster fullawayi (Baker) Figs. 7, 24 Cascogaster fullawayi Baker 1926: 483. Holotype female (cited as male), PHILIPPINES: Baguio, Benguet. Ascogaster fullawayi: Shenefelt 1973: 822. Diagnosis. — Antenna with 34-35 flagellomeres, slightly dilated medially; scrobes outwardly mar- gined by high, thin and complete carina middle of which is raised into sharply angulate teeth; ocelli almost on line; face (Fig. 7) narrow, less 1.5 times as broad as long, strongly and coasely rugose; clypeus smooth and shining, sparsely punctate, apical border acute; mesonotum between notauli swollen, strongly areolate-rugose; mesopleuron coarsely areolate-rugose, precoxal suture indis- tinguishable; hind coxa rugose-punctate; carapace (Fig. 24) acute with a tubercle at apex; ventral opening of carapace short, distinctly in front of apex. Speimens Examined. — TAIWAN: 2 females. Additional Specimens Examined. — Holotype of fullawayi (Baker), female, PHILIPPINES: Baguio, Benguet (USNM). Holotype of fullawayi var. maquilingensis (Baker), female, PHILIPPINES: Mt. Makiling, Luzon (USNM). Paratype of fullawayi (Baker), 1 female, PHILIPPINES: Imugin, N, Viscaya (USNM). Hosts. — Unknown. Distribution. — Taiwan, Philippines. Remarks. — Two Taiwanese speicmens exam- ined agree well with the type of fullawayi (Baker) except that the antennae are shorter. This species is the only Philippine Ascogaster described by Baker (1926) found in China. It is distinguished from the other Chinese Ascogaster by the characteristics of the carapace and sculpture of the head. Ascogaster gibbosa Tang and Marsh, new species Fig. 28 Females.- — Length of forewing 3.0-3.2 mm, of body 3.5-3.8 mm. Head. — Antenna with 37-40 flagellomeres, slightly dilated medially, medial segments about as long as broad. Temple rounded behind eye, slightly shorter than eye in dorsal view. Occiput concave. Ocelli small, OO = 3.5-4.0 OD, on line. Frons moderately depressed, smooth, with a weak medial carina. Eyes protuberant, without distinct setae. Gena in face view contracted. Face about 1.5 times as broad as high, coarsely areolate-rugose. Clypeus distinctly divided from face, punctate, apical border produced medially into a strongly pointed large tooth. Mesosoma. — Pronotum slightly projecting in front of mesonotum, dorsolaterally areolate-rug- ose. Notauli present, areolate-rugose; mesonotum (Fig. 28) between notauli strongly protuberant, rugose-punctate, rest of mesonotum areolate-rug- ose. Precoxal suture indistinguishable; mesopleuron completely areolate-rugose. Propodeum strongly areolate-rugose, divided by transverse carina which is raised medially and laterally into prominent dentate flanges. Hind coxa dorsally strigate. Vein r of forewing about as long as 3-RS. Metasoma. — Carapace moderately long, CL/ CW = 1.6-1.7, oval in dorsal view, somewhat pointed in lateral view. Hypopygium short. 288 Journal of Hymenoptera Research Ovipostor short, straight. Color. — Almost completely black except ex- treme apex of hind coxa and of femur yellow. Male. — Unknown. Holotype Female. — TAIWAN: Tsuifeng, VI-3- 80, L. Y. Chou and C. C. Chen. Deposited in TARI. Paratypes. — TAIWAN: 2 females, same as ho- lotype; 2 females, Tsuifeng, VII-16-82. S. C. Lin and C. N. Lin; 1 female, Tsuifeng, VI-21-79, K. S. Lin and B. H. Chen; 1 female, Meifeng, VII-31-IX-2-82, L. Y. Chou and K. C. Chou; 1 female, Tayuling, IX- 12-15-80, K. S. Lin and C. H. Wang. Deposited in TARI, USNM. Host. — Unknown. Distribution. — Taiwan. Remarks. — This species can be easily distin- guished from other species of the quadridentata- group (see Huddleston 1984:371 ) by the character- istics of the mesonotum and by the strongly pointed clypeus tooth and completely black body. Etymology.— The specific name is from the Latin gibbosus meaning hunched or humped in reference to the strongly humped mesonotum. Ascogaster grandis Tang and Marsh, new species Fig. 6, 21 Females. — Length of forewing 4.0-4.1 mm, of body 4.8-5.0 mm. Head. — Antenna with 37 flagellomeres; flagel- lum dilated medially, strongly tapered at apex, medial segments about as long as broad. Temple slightly constricted behind eyes, at least 1.5 times as long as eye in dorsal view. Occiput deep, con- cave. Ocelli very small, on line, OO = 4.0 OD. Frons strongly depressed, rugose-punctate. Eyes protu- berant, glabrous, without distinct setae. Gena swol- len in lateral view. Face (Fig. 6) slightly protuber- ant, about 1.5 times as broad as high, areolate- rugose, with medial carina dorsally. Clypeus slightly protuberant, punctate, distinctly divided from face, apical border impressed laterally, pro- duced medially into a very small pointed tooth. Mesosoma. — Pronotum strongly projecting in front of mesonotum, dorsolaterally rugose-punc- tate, smooth ventrally. Notauli not distinct, mesonotum strongly areolate-rugose. Precoxal suture indistinguishable, mesopleuron strongly areolate-rugose. Propodeum completely coarsely areolate-rugose, divided by a transverse carina produced into medial and postero-lateral pairs of stout teeth. Hind coxa strongly strigate or ar- eolate-rugose. Vein r of forewing about as long as 3-RS. Metasoma. — Carapace (Fig. 21) very long, CL/ CW = 2.3-2.5, areolate-rugose. Ventral opening of carapace very short, ending almost in middle of carapace. Ovipositor concealed. Color. — Black; fore and middle legs yellow except coxae and femora black basally, hind leg black except trochanters reddish-brown; papli pale yellow. Male. — Same as female except antenna not dilated medially. Holotype Female.— ZHEJIANG PROVINCE: Xitianmushan, VII-29-84, X. J. Wu. Deposited in ZAU. Paratypes.— ZHEJIANG PROVINCE: 1 male, same as holotype; 1 female, Xitianmushan, VII-27- 84, X. J. Wu; 1 female, Xitianmushan, VII-22-87, X. M. Lou. Deposited in USNM, ZAU. Host. — Unknown. Distribution. — China (Zhejiang Province). Remarks. — Both this species and macrogaster, new species, run into the quadridentata-group in the key of Huddleston (1984), but both of them are distinguished from species of this group by the characteristics of head shape and their elongate carapaces. Etymologi/. — The specific name is from the Latin grandis meaning large in reference to the large expanded temple. Ascogaster hei Tang and Marsh, new species Figs. 15, 16 Females. — Length of forewing 4.0-4.2 mm, of body 4.6-5.0 mm. Head. — Antenna with 33-37 flagellomeres, moderately dilated medially, medial flagellomeres slightly broader than long. Temple rounded be- hind eyes, distinctly longer than length of eye in dorsal view. Occiput deeply concave. Ocelli small, OO = 3.5-4.0 OD, almost on line. Frons (Fig. 15) moderately depressed behind antennae, rugose- punctate, with a weak medial carina. Eyes protu- berant, nearly glabrous but with a few scattered setae. Gena in face view contracted. Face (Fig. 16) about twice as broad as high, strongly areolate- rugose. Clypeus distinctly divided from face, punc- tate; apical border produced medially into a small Volume 3. 1 994 289 pointed tooth. Mesosoma. — Pronotum projecting in front of mesonotum, dorsolaterally areolate-rugose. Notauli not distinct; mesonotum strongly areolate- rugose; precoxal suture indistinguishable; mesopleuron coarsely areolate-rugose. Propodeum strongly areolate-rugose, divided by a transverse carina which is raised medially and laterally into stout dentate flanges. Hind coxa strongly strigate. Vein r of forewing about as long as 3-RS. Metasoma. — Carapace longer, CL/CW =1.6- 1.7, clavate in dorsal view. Ventral opening of carapace longer, not very distinctly in front of apex. Hypopygium short. Ovipositor sheaths clav- ate. Color. — Almost completely black except fore tibia, and sometimes middle and hind coxae testa- ceous apically. Male. — Same as females except antennae not dilated medially, with 34-35 flagellomeres. Holotype Female.— ZHEJIANG PROVINCE: Fengyangshan, VIII-12-84, L. R. Sheng. Deposited inZAU. Paratypes.— JILING PROVINCE : 2 females, 3 males, Dongjiang, VII-20-31-88, X. M. Luo. HEILONHJIANG PROVINCE: 1 female, Qingdinzhi, V-77, Y. Y. Hun. ZHEJIANG PROV- INCE: 1 male, same as holotype; 1 female, Xitainmushan, VIII-3-84, L. K. Sheng; 1 female, Songyang, VII-15-17-89, J. H. He; 1 male, Zhuji, VI- 10-85, X. X. Chen. FUJAIN PROVINCE: 1 female, 1 male, Guadun (Wuyishan), VI-11-82, J H. Xiu; 1 female, Wuyishan, VI-20-80, N. Q. Lin; 1 female, Wuyishan, VI-30-80, J. H. Xiu; 1 female, Wuyishan, VIII-3-83, Y. Ma; 1 male, Wuyishan, VII-30-81, J. S. Weng. Deposited in FAC, USNM, ZAU. Host. — Unknown. Distribution. — China (Heilongjiang Province, Tiling Province, Zhejiang Province, Fujian Prov- ince). Remarks. — This species can be distinguished from other species of the quadridentata-group by the more massive head, strongly strigate-rugose hind coxa, and larger and stout body. Etymology. — This species is named for Prof. J. H. He, Zheijiang Agricultural University, in ap- preciation for his helpfulness in this work. Ascogaster lini Tang and Marsh, new species Fig. 9 Females. — Length of forewing 3.9-4.3 mm, of body 4.4-4.8 mm. Head. — Antenna with 33-36 flagellomeres, weakly dilated medially, slightly tapered at apex, medial segments about as broad as long, basal segments distinctly longer than broad, apical seg- ments slightly longer than broad. Temple strongly constricted behind eyes, slightly shorter than eye in dorsal view. Occiput deeply concave. Ocelli small, on line, OO = 4.0 OD. Frons with two depressed, polished impunctate areas behind an- tenna, median carina distinct. Eyes moderately protuberant, glabrous. Face (Fig. 9) slightly protu- berant, about 1.5 times as broad as high, finely punctate, with a median carina in upper part; Clypeus weakly divided from face, protuberant, sparsely punctate, smoother than face; apical bor- der of clypeus strongly impressed laterally, me- dial area raised and produced forwards into a distinct pointed tooth. Mesosoma. — Pronotum projecting in front of mesonotum, laterally shining and smooth, dor- sally rugose-foveolate. Notauli strong, foveolate, coalescing posteriorly in an areolate-rugose area; rest of mesonotum rugose-punctate. Precoxal su- ture shallow, foveolate; mesopleuron above precoxal suture rugose-foveolate anteriorly, smooth and punctate posteriorly, rest of mesopleuron ventrally densely punctate. Propodeum completely areolate-rugose divided medially by a transverse carina with medial and lateral tubercles, the medial pair broad and stout, lateral tubercles more prominent. Hind coxa finely punctate. Vein r of forewing about 1.5-2.0 times as long as 3-RS. Metasoma. — Carapace long, CL/CW =1.8-2.0, clavate in dorsal view, finely areolate-rugose, with two weak medial carinae at base. Ventral opening of carapace longer, almost at apex of carapace. Hypopygium short. Ovipositor sheaths clavate, short. Color. — Black except carapace yellow at base; antenna infuscate except scape yellow; legs yel- low except hind coxa at base, mid and hind femur at apex, and all tarsi infuscate. Males. — Same as females except medial flagellomeres not dilated, carapace completely black and scape infuscate as in flagellum. 290 Journal of Hymenoptera Research Holotype Female.— TAIWAN: Tsuifeng 2300 m, malaise trap, VI-84, K. S. Lin and K. C. Chou. Deposited in TARI. Paratypes. TAIWAN: 1 female, 3 males , Tsuifeng," VI-21-79, K. S. Lin and B. H. Chen; 3 males, Tsuifeng, VI-3-80, L. Y. Chou and C. C. Chen; 1 male, Tsuifeng, V-8-81, K. S. Lin and S. C. Lin; 1 female, Tsuifeng, VII-1-3-81, T. Lin and W. S. Tang; 12 females, 4 males, VI-25-27-81, K. S. Lin and W. S. Tang; 2 females, 1 male, Tsuifeng, V-23- 82, L. Y. Chou; 9 females, 1 male, VII-16-82, S. C. Lin and C. N. Lin; 4 females, 5 males, Tsuifeng, VI- 23-25-83, K. S. Lin and S.C. Lin; 13 females, 14 male, same as holotype; 20 females, 8 males, Tsuifeng, VII-XI-84, Malaise trap, K. S. Lin and K. C. Chou; 1 male, Tsuifeng, IX-85, Malaise trap, K. S. Lin; 1 female, 1 male, Meifeng, V-10-79, K. C. Chou; 2 females, Meifeng, VI-20-22-79, K. S. Lin and B. H. Chen; 2 females, 1 male, Meifeng, VI-2- 4-80, L. Y. Chou and C. C. Chen; 1 female, Meifeng, VI-5-8-80, C. C. Chen; 1 female, Meifeng, VI-8-80, K.S. Lin and B. H. Chen; 1 female, Meifeng, X-5-9- 80, C. C. Chen and C. C. Chien; 1 female, 10 males, Meifeng, V-7-9-81, K. S. Lin and S. C. Lin; 2 fe- males, Meifeng, VI-24-26-81, K. S. Lin and W. S. Tang ; 2 males, Meifeng, V-22-82, L. Y. Chou; 1 male, Meifeng, V-3-83, H. Townes; 5 males, Meifeng, V-10-83, H. Townes; 2 males, Meifeng, V- 15-83, H. Townes; 1 female, 4 males, V-22-83, H. Townes; 4 females, Meifeng, VI-22-26-83, K. S. Lin and S. C. Lin; 1 male, Meifeng, V-8-11-84, K. C. Chou and C. C. Pan; 1 female, 3 males, Meifeng, VI-84, Malaise trap, K. S. Lin and K. C. Chou; 1 female, Meifeng, VIII-84, Malaise trap, K. S. Lin and K. C. Chou; 1 female, Meifeng, X-84, Malaise trap, K. S. Lin and K. C. Chou; 1 female, Meifeng, X-85, Malaise trap, K. S. Lin; 1 female, Sungkang 2100 m, X-84, Malaise trap, K. S. Lin and K. C. Chou; 1 female, 2 males, Sungkang, X-XI-85, Mal- aise trap, K. S. Lin; 8 females, 3 males, Tayuling 2560 m, VI-9-16-80, Malaise trap, K. S. Lin and B. H. Chen; 1 female, 2 males, Alishan 2400 m, VI-12- 16-65, T. Maa and K. S. Lin; 4 females, Anmashan 2275m, VII-6-9-79, L. Y. Chou. Deposited in AEIG, FAC, TARI, USNM. Host. — Unknown. Distribution. — Taiwan. Remarks. — This species is very similar to perkinsi in the characteristics of the face and clypeus, but it is rather different in the shape of the carapace and color patterns of both legs and antennae. Etymology. — It is a pleasure to name this spe- cies in honor of its collector, Mr. K. S. Lin, who also collected a large amount of Braconidae in Taiwan. Ascogaster macrogaster Tang and Marsh, new species Females. — Length of forewing 3.5-4.2 mm, of body 4.3-4.9 mm. Head. — Antenna with 33 flagellomeres, slightly dilated medially, medial segments about as long as broad. Temple slightly contracted behind eyes, slightly longer than eyes in dorsal view. Occiput concave. Ocelli larger, OO =3.0 OD, on line. Frons slightly excavated, rugose-punctate; medial ca- rina distinct. Eyes not protuberant, glabrous with- out distinct setae. Gena expanded in lateral view. Face slightly protuberant, about 1 .5 times as broad as high, areolate-rugose. Clypeus punctate; apical border impressed laterally, produced medially into a small pointed tooth. Mesosma. — Pronotum strongly projecting in front of mesonotum, dorsolaterally areolate-rug- ose. Notauli not distinct; mesonotum rugose-punc- tate anteriorly, rest of mesonotum foveate. Mesopleuron deep punctate, but precoxal suture not distinguishable. Propodeum strongly areolate- rugose, not distinctly divided into dorsal and pos- terolateral surfaces. Hind coxa smooth, shining. Vein r of forewing about 1.5 times as long as 3-RS. Metasoma. — Carapace very long, CL/CW = 2.5, areolate-rugose, rounded apically in lateral view. Ventral opening of carapace at 1/4 apex of carapace. Hypopygium broad. Ovipositor short, pointed apically; ovipostor sheath clavate. Color. — Black; carapace sometimes yellow at base; all legs yellow except hind femur and tibia brownish apically; palpi pale yellow. Male. — Same as female except antenna not dilated medially, all flagellomeres longer than broad, and smaller yellow spot at basal carapace. Holotype Female.— FUJIAN PROVINCE: Tongmugun (Wuyishan), VI-23-82, J. C. Huang. Deposited in FAC. Paratypes.— FUJIAN PROVINCE: 1 female, same as holotype; 1 female, Wuyishan, VI-80, H. F. Chao; 1 female, Wuyishan, VI-80. J. C. Huang; 1 male, Huanxi, V-30-86, N. Q. Lin. Deposited in FAC. Host. — Unknown. Distribution. — China (Fujian Province). Volume 3, 1 994 291 Remarks. — This species is very close to grandis from which it can be distinguished not only by the characters mentioned in the key but also by the lack of distinct tubercles on the propodeum. Etymology. — The specific name is from the Greek makros meaning long and gaster meaning stomach in reference to the long metasomal cara- pace. Ascogaster perkinsi Huddleston Figs. 10, 22 Ascogaster perkmsi Huddleston 1984: 368. Holotype female, JAPAN: Mt.Tachibana, Fukuoka. Diagnosis. — Length of forewing 2.6-2.9 mmm, of body 3.0-3.5 mm. Antenna with 34-36 flagellomeres, slightly dilated medially (female); ocellar triangle obtuse, ocelli almost on line; face (Fig. 10) about 1.5 times as broad as high, rugose- punctate; apical border of clypeus strongly im- pressed laterally, medial area projecting and pro- duced forwards into a distinct pointed tooth; notauli weak, foveolate, rest mesonotum rugose- punctate; precoxal suture foveolate, mesopleuron above precoxal suture areolate-foveolate anteri- orly, sparsely punctate posteriorly, rest of mesopleuron densely punctate; carapace short, oval in dorsal view, clavate in lateral view, gener- ally yellow at base; ventral opening of carapace (Fig. 22) short, distinctly in front of apex; hind coxa finely punctate; all legs yellow except middle tibia dark at base, hind femur and tibia dark at base, mid and hind tarsi infuscate. Specimens Examined. — ZHEJIANG PROV- INCE: 41 females, 24 males. HUNAN PROVINCE: 1 male. FUJIAN PROVINCE: 4 females, 3 males. TAIWAN: 11 females, 6 males. Host. — Unknown. Distribution. — China (Zhejiang Province, Hunan Province, Fujian Province), Taiwan, Japan. Remarks. — All specimens examined agree well with the description of Huddleston (1984). This is a new species record for the fauna of China. Ascogaster quadridentata Wesmael Figs. 17, 18 Ascogaster quadridentatus Wesmael 1835: 237. Lectotype fe- male, BELGIUM: Brussels (designated by Shaw 1983). Chelonus impressus Herrich-Schaffer 1838: 153. Syntypes, GERMANY ( syn. by Reinhard 1867) Ascogaster nigricornis Thomson 1892: 1719. Lectotype female, SWEDEN (designated and syn. by Huddleston 1984). Shenefelt 1973: 826. Ascogaster cynipum Thomson 1892: 1720. Holotype male, SWEDEN (syn. by Huddleston 1984). Shenefelt,1973: 820. Ascogaster egregius Kokujev 1895: 83. Holotype male, USSR (syn. by Huddleston 1984). Fahnnger 1934: 525; Shenefelt 1973: 821. Chelonus nigrator Szepligeti 1896: 303. Holotype female, YU- GOSLAVIA: Buccari (syn. by Huddleston 1984). Shenefelt 1973: 859. Chelonus carpocapsae Viereck 1909: 43. Holotype female, USA: Michigan (syn. by Rosenberg 1934). Ascogaster epinotiae Watanabe, 1937: 76. Holotype female, JAPAN: Hokkaido, Sapporo (syn. by Huddleston 1984). Shenefelt 1973: 821. Ascogaster quadridentata; Shenefelt 1973: 828; Shaw 1983: 32; Huddleston 1984: 376 Walker and Huddleston 1987: 343; Papp 1989: 258. Diag)iosis. — Antenna with 27-32 flagellomeres, slightly dialted medially; ocelli on line or almost on line (Fig. 17); face (Fig. 18) about twice as broad as high, genearlly finely rugose; clypeus finely punctate, apical border produced medially into a pointed tooth; notauli and precoxal suture indis- tinguishable, mesonotum and mesopleuron com- pletely areolate-rugose; hind coxa always mostly black and strongly strigate-rugose; carapace oval, short, genearlly deep in lateral view, sometimes yellow at base; ventral opening of carapace short, usually distinctly in front of apex. Specimens Examined. — BEIJING CITY: 1, fe- male, 1 male. JILFNG PROVINCE: 1 male. HUNAN PROVINCE: 2 females. JIANGXI PROVINCE: 1 male. GUANGXI PROVINCE: 1 female, 1 male. YUNNAN PROVINCE: 2 females, 2 males. GUIZHOU PROVINCE: 3 females, 2 males. ZHEJIANG PROVINCE: 29 females, 30 males. FUJIAN PROVINCE: 18 females, 11, males. TAI- WAN: 26 females, 7 males. Additional Specimens Examined. — Lectotype of quadridentatus Wesmael, female, BELGIUM: Brus- sels (IRSNB). Lectoparatypes of quadridentatus Wesmael, 1 female, 5 males, same as lectotype (IRSNB). USA: 1 female, 1 male, WA, Pullman, August 30, 1989, ex. Cydia pomonella, det. by Shaw, 1990 (FAC). Hosts.— Huddleston (1984) and Shaw (1983) recorded the following hosts in the Palaearctic and Nearctic Regions: Cydiafunebrana (Treitsche), Cydia nigricana Steph., Cydia pallifrontana (Lienig & Zeller), Cydia pomonella (L.), Epiblema uddmanniana 292 Journal of Hymenoptera Research (L.), Grapholitlui molesta (Busck), Grapholitapnmivora (Walsh), Spilonota ocellana (Denis & Schif fermuller) (Lepidoptera: Tortricidae); Endopiza viteana (Clem.); Yponomeuta padella (L.) (Lepidoptera: Yponomeutidae). Distribution. — China (Beijing City, Jiling Prov- ince, Jiangsu Province, Zhejiang Province, Yunnan Province, Guizhou Province, Guangxi Province, Fujian Province), Taiwan, Japan, Korea, New Zealand, West Palaearctic Region, Nearctic Re- gion. Remarks. — This species has been studied widely as a parasitoid of several economically important insect pests in Europe and North America. Its hosts and biology were reviewed by Huddleston (1984) and Shaw (1983). There is very little published work on this species in China. Fahringer (1934) recorded it in Jiangsu Province (=Kiangfu). This species is characterized by its coarse sculpture, short oval and deep carapace, short temple and antenna, highly strigate and black hind coxa, and the single tooth on its cly peus. Ascogaster reticulata Watanabe Asc ogaster reticulatus Watanabe 1967: 41 . Holotype male (cited as female), JAPAN: Hokkaido, Ashigawa. He, Chen and Ma 1989: 438. Ascogaster reticulata: Huddleston 1984: 377; Papp 1989: 298. Diagnosis. — Length of forewing 2.8-3.2 mm, of body 3.4-3.8 m. Antenna with 34-36 flagellomeres, weakly dilated medially; ocelli large, on line; face about 1.5 times as broad as high, coarsely irregu- larly rugose; clypeus strongly punctate, matt and its apical border almost flat without any trace of tooth or tubercle; mesonotum areolate-rugose, notauli indistinct; precoxal suture indistinguish- able, mesopleuron strongly areolate-rugose; cara- pace short, deep, sometimes yellow at base; ven- tral opening of carapace short, distinctly in front of apex; hind coxa yellow, smooth or sometimes finely punctate; hind tibia black with a pale yellow medial band. Specimens Examined.— BEIJING CITY: 2 fe- males. HENAN PROVINCE: 13 females. SHANDONG PROVINCE: 8 females, 5 males. SHANXI PROVINCE: 8 females. SHAANXI PROVINCE: 3 females, 1 male. ANHUI PROV- INCE: 1 male. JIANGSHU PROVINCE: 9 females, 2 males. JIANGXI PROVINCE: 2 females. ZHEJIANG PROVINCE: 7 females, 4 males. YUNNAN PROVINCE: 3 females. GUANHXI PROVINCE: 1 female. FUJIAN PROVTNCE: 1 fe- male, 1 male. TAIWAN: 31 females, 36 males. Additional Specimens Examined. — Holotype of reticulatus Watanabe, male, JAPAN: Hokkaido, Asahigawa, June 15, 1966 (UEI). Paratypes of reticulatus Watanabe: 3 females, JAPAN, May 16- July 20, 1966 (UEI); 1 female, 1 male, JAPAN, July 20-23, 1966 (USNM). Hosts. — Acleris fimbriana (Thunberg), Adoxophyes orana (Fischer von Roslerstamm), Archips issikii Kodama, Archips oporana (L.), Archips pulchra (Butler), Carposiatw nipponensis Walsingham (Lepidoptera: Tortricidae). Distribution. — China (Beijing City, Henan Province, Shandong Province, Shanxi Province, Shannxi Province, Anhui Province, Jiangshu Prov- ince, Jiangxi Province, Zhejiang Province, Yunnan Province, Guangxi Province, Fujian Province), Taiwan, Japan, Korea, Czechoslovakia. Remarks. — This species is easily distinguished from other species of the quadridentata-gToup by the matt and no tooth clypeus and by the yellow and smooth hind coxa and the hind tibia with a yellow-pale band medialy. Ascogaster rugulosa Tang and Marsh, new species Fig. 13 Females. — Length of forewing 3.0-3.2 mm, of body 3.4-3.7 mm. Head. — Antenna with 31-35 flagellomeres, slightly dilated medially, flagellomeres 1-11 longer than broad, rest of flagellum about as broad as long. Temple rounded behind eyes, about equal to length of eye in dorsal view. Occiput concave. Ocelli small, OO =3.5-4.0 OD, on line. Eyes protu- berant, glabrous. Frons moderately concave, smooth, with a weak medial carina. Face (Fig. 13) about 1.5 times as broad as high, finely areolate- rugose, generally more finely scluptured ventrally than dorsally. Clypeus punctate, apical border produced medially into a pointed tooth. Mesosoma. — Pronotum little projecting in front of mesonotum, rugose-punctate dorsolaterally. Notauli present but not very distinct; mesonotum rugose-punctate except an areolate-rugose area posteromedially. Precoxal suture very weak, shal- low foveolate; rest of mesopleuron sparsely but coarsely punctate. Propodeum coarsely areolate- Volume 3. 1 994 293 rugose, divided by a tranverse carina which is raised into a pair of medial dentate flanges and a pair of lateral dentate flanges. Hind coxa finely punctate. Vein r of forewing about as long as or slightly longer than 3-RS. Metasoma. — Carapace short, oval in dorsal view, clavate in lateral view, areolate-rugose. Ven- tral opening of carapace less distinctly in front of apex. Ovipositor short, tapered apically. Color. — Black; fore and middle legs yellow except middle tibia at apex and all tarsi infuscate, hind leg black except apex of coxa and base of femur yellow and with a pale yellow medial band on tibia; carapace always yellow laterally at base, but black medially. Males. — Same as females except antenna not dilated medially. Holotype Female.— TAIWAN: Meifeng, V-7-9- 81, K. S. Lin and S. C. Lin. Deposited in TARI. Paratypes.— ZHEJIANG PROVINCE: 1 male, Xitainmushan, VI-2-4-90, L. G. Weng. HAINAN PROVINCE: 1 male, Shuiman, V-26-60, X. F. Li. TAIWAN: 7 females, 1 male, same as holotype; 1 female, Meifeng, VII-26-78, K. C. Chou; 1 female, Meifeng, IV-8-9-78, K. S. Lin; 1 male, Meifeng, VII- 18-79, K. C. Chou; 1 female, Meifeng, V-15-22-79; 1 female, Meifeng, VI-20-21-79, K. S. Lin and B. H. Chen; 1 female, Meifeng, X-5-9-80, C. C. Chen and C. C. Chien; 1 male, Meifeng, VI-24-26-81, K. S. Lin and W. S. Tang; 3 males, Meifeng, VIII-28-29-81, L. Y. Chou and S. C. Lin; 1 female, 1 male, VII-31-IX- 2-82, L. Y. Chou and S. C. Lin; 2 females, 2 males, Meifeng, V-8-11-82, K. C. Chou and C. C. Pan; 1 male, Meifeng, VII-30-83, L. Y. Chou; 1 female, Meifeng, VI-22-26-83, K. S. Lin and S. C. Lin; 1 female, Meifeng, X-4-7-82, K. C Chou; 1 male, V- 8-11-84, K. C. Chou and C. C Pan; 1 female, Meifeng, VII-84, Malaise trap, K. S. Lin and K. C. Chou; 1 female, Tungpu, X-18-21-82, K. C. Chou and S. C. Lin; 1 female, Tungpu, VI-20-24-83, K. C. Chou and C. Y. Wong; 1 female, Tungpu, VII-85, Malaise trap, K. S. Lin; 1 female, Tsuifeng, VI-21- 79, K. S. Lin and B. H. Chen; 1 female, Tsuifeng, VI- 3-80, L. Y. Chou and C. C. Chen; 1 female, Tsuifeng, V-8-81, K. S. Lin and S. C. Lin; 4 females, 1 male, Tsuifeng, VI-25-27-81, K. S. Lin and W. S. Tang; 1 male, Tsuifeng, VIII-27-81, L. Y. Chou and S. C. Lin; 1 male, Tsuifeng, IX-1-3-82, L. Y. Chou and K. C. Chou; 3 females, Tsuifeng, VI-23-25-83, K. S. Lin and S. C. Lin; 2 females, Tsuifeng, VI- VII-84, Mal- aise trap, K. S. Lin and K. C. Chou; 3 males, Tsuifeng, IX-85, Malaise trap, K. S. Lin; 4 males, Tsuiofeng, IX-12-14-84, K. S. Lin and S. C. Lin; 3 males, Wushe, IV-26-83, H. Townes;l male, Wusha, VII-25-78, K. C. Chou; 1 female, Sungkang, VIII-6- 84, K. S. Lin; 1 female, 1 male, Sungkang, X-84, Malaise trap, K. S. Lin and K. C. Chou; 2 males, Sungkang, Malaise trap, IX-X-85, K. S. Lin; 1 fe- male, 2 males, Shengkuang, IX-20-68, K. S. Lin; 1 female, Wuling, VI-27-29-79, K. S. Lin and L. Y. Chou; 1 female, Tayuling, VI-9-16-80, K. S. Lin and B. H. Chen; 1 male, Lishan, IX-12-68, K. C. Chou. Deposited in AEIG, FAC, TARI, USNM, ZAU, ZRI. Host. — Unknown. Distribution. — China (Zhejiang Province, Hainan Province), Taiwan. Remarks. — This species is close to reticulata from which it can be distinguished not only by the characters mentioned in the key but also by its smooth mesopleuron, less deep carapace and mostly black hind coxa. Etymologic — The specific name is in reference to the rugulose sculpturing on the clypeus. Ascogaster semenovi Telenga Figs. 11, 25 Ascogaster semenovi telenga 1941: 310, 453. Holotype female, MONGOLIA: Alashan, Dyn-juan-in. Shenefelt 1973: 825; Huddleston 1984: 352. Ascogaster kyushuensis Yoneda 1978: 291. Holotype female, JAPAN: Kyushu, Fukuoka Pref., Fukuoka City, Hakozaki ( syn. by Huddleston 1984 ). Diagnosis. — Length of forewing 4.3-4.7 mm, of body 5.2-5.6 mm. Antenna of female with 20 flagellomeres and medial segments dilated, genearlly broader than long, male with 25-26 flagellomeres, medially not dilated, all segments longer than broad; ocellar triangle acute, ocelli not on line; face (Fig. 11) about twice as broad as high, rugose-punctate, the hairs on the upper part of face pointing upwards; clypeus with its apical border rounded medially, produced laterally into broad dentate flanges, without tooth or incision; notauli distinct, f oveolate, rest of mesonotum punc- tate with an areolate-rugose area posteromedially; precoxal suture deep foveolate, rest of mesopleuron punctate with a deep foveolate groove anterodorsally; hind coxa finely and sparsely punc- tate; carapace (Fig. 25) long, CL/CW = 2.2-2.4, pointed in dorsal and lateral views; ovipositor 294 Journal of Hymenoptera Research sheaths broad, knife-like. Almost completely black except sometimes fore tibia reddish-brown. Specimens Examined.— JIANGSU PROVINCE: 2 females, 6 males. SHANGHAI CITY: 8 females, 3 males. ZHEJIANG PROVINCE: 2 females, 8 males. Hosts. — Unkown. Distribution. — China (Jiangsu Province, Shang- hai City, Zhejiang Province), Mongolia, Japan. Remarks. This remarkable speices is easily dis- tinguished from other species of Ascogaster by the lateral dentate flanges on its clypeus and by the broad and knife-like ovipositor sheaths. It is new to the fauna of China. Ascogaster setula Tang and Marsh, new species Fig. 8 Females. — Length of forewing 2.4-2.7 mm, of body 2.9-3.1 mm. Head. — Antenna with 26-29 flagellomeres; fla- gellum moderately dilated medially, medial seg- ments slightly broader than long, apical segments about as broad as long. Temple rounded behind eyes, about equal to or slightly shorter than eye in dorsal view. Occiput concave. Ocelli small, OO = 4.0 OD, almost on line. Frons moderately depressed behind antenna. Eyes slight protuberant, with dense and distinct hairs. Malar space shorter, about 0.5 time as eye high. Gena strongly constricted in face view. Face (Fig. 8) protuberant, coarsely rug- ose-punctate, about 1.5 times as broad as high. Clypeus not distinctly divided from face, slightly convex, its apical border more or less straight, with a very small tooth medially. Mesosoma. — Pronotum protecting in front of mesonotum, dorsolaterally foveate. Notauli shal- low foveolate, rest of mesonotum punctate except an areolate-rugose area posteromedially . Precoxal suture not very distinct, mesopleuron generally coarsely rugose-punctate, sometimes areolate- rugose. Propodeum coarsely areolate-rugose, di- vided by a transverse carina which is raised into a medial pair and a lateral pair of stout dentate flanges. Hind coxa smooth. Vein r of forewing 1.0- 1.3 times as long as 3-RS. Metasoma. — Carapace shorter, CL/CW = 1.5- 1.8, areolate-rugose, oval, deeper in lateral view. Ventral opening of carapace distinctly in front of apex. Hypopygium short. Ovipositor short, its sheath clavate. Color. — Black; carapace always yellow at base; all legs yellow except hind femur and tibia apically in fu scare. Males. — Same as female except antennae not dilated medially and carapace black at base. Holotype Female.— TAIWAN: Tungpu, XI-18- 21-82, K. C. Chou and S. C. Lin. Deposited in TARI. Paratopes. — TAIWAN: 4 females, 1 male, same as holotype; 3 females, 2 males, Tungpu, IX-25-29- 80, L. Y. Chou and T. Lin; 1 female, Tungpu, IV-28- V-2-81, T Lin and C. J. Lee; 2 females, 1 male, Tungpu, X-5-8-81, T. Lin and W. S. Tang; 1 female, Tungpu, XI-18-23-81, T. Lin and W. S. Tang; 6 females, Tungpu, X-XI-85, Malaise trap, K. S. Lin; 4 males, Tungpu, IV-16-VII-27-84, K. C. Chou and C. H. Yung; 2 females, 2 males, Lienhuachi, III-VII- 84, Malaise trap, K. S. Lin and K. C. Chou. Depos- ited in FAC, TARI, USNM. Hosts. — Unknown. Distribution. — Taiwan. Remarks. — Previous authors (Baker 1926, Shaw 1983, Zettel 1990) treated glabrous eyes as a char- acter to distinguish Ascogaster. Strikingly this spe- cies has distinct and dense setae on its eyes, but except for this character it agrees well with typical Ascogaster. Its very small body and the character- istics of the clypeus are also useful to distinguish this species from other Chinese Ascogaster. Etymology. — The specific name is the diminu- tive form of the Latin seta meaning bristle in reference to the unusually hairy eyes. Ascogaster townesi Tang and Marsh, new species Figs. 12, 26 Females. — Length of forewing 4.3-4.5 mm, of body 4.6-4.9 mm. Head. — Antenna with 23 flagellomeres, me- dial segments not dilated, all segments longer than broad. Temple rounded behind eyes, slightly longer than eye in dorsal view. Occiput deeply concave. Ocelli small, 00= 4.0-4.5 OD, ocellar triangle acute, ocelli not on line. Frons behind antenna slightly depressed, sparsely punctate. Eyes moderately protuberant, glabrous with few scat- tered minute setae. Malar space short. Gena in face view strongly constricted. Face (Fig. 12) slightly convex, about twice as broad as high, less hairy, the hairs on the upper part of face pointing up- wards, finely punctate, with a medial carina on the Volume 3, 1994 295 upper part. Clypeus slightly protuberant, more scatteredly punctate than face; apical border straight, without medial teeth or flanges. Mesosoma. — Pronotum projecting little in front of mesonotum, dorsolaterally smooth, shallow fovea te. Notauli deep, foveolate, rest of mesonotum punctate except with a depressed and areolate- rugose area posteromedially . Precoxal suture shal- low foveate; rest of mesopleuron sparsely punc- tate, smooth except anterodorsally foveolate. Propodeum not distinctly divided into dorsal and posterlateral surfaces; strongly rugose with no tubercles. Hind coxa smooth. Vein r of forewing 1.0-1.2 times as long as 3-RS. Metasoma. — Carapace (Fig. 26) very long, CL/ CW = 2.3-2.5, clavate in dorsal view, flat in lateral view, areolate-rugose, apically polished. Ventral opening of carapace at apex of carapace. Hypopyium large and broad. Ovipositor long, slender and upcurved. Color. — Black; legs yellow-brown except hind coxa basally, hind and middle femora and tibia apically and tarsi brownish; antenna infuscate except scape reddish brown; palpi yellowish brown. Males. — Same as females except antennae longer with 32 flagellomeres, flater carapace and yellow hind coxa. Holotype Female. — TAIWAN: Meifeng 2150 m, V-10-83, H. Townes. Deposited in AEIG. Paratypes.— TAIWAN: 1 male, Meifeng 2150 m, IV-19-21-83, K. C. Chou and P. Huang; 4 males, Meifeng.2150 m, IV-26-83, H. Townes; 1 female, 30 males, Meifeng 2150 m, V-3-83, H. Townes; 2 females, 17 males, same as holotype. Deposited in AEIG, FAC, TARI, USNM. Host. — Unknown. Distribution. — Taiwan. Remarks. — This species is easily distinguished from other species of Ascogaster by the hairs on the upper part of the face, very flat carapace, no tu- bercle on its clypeus, and long, upcurved oviposi- tor. Etymology. — It is a pleasure to name this spe- cies in honor the late Henry Townes who collected many of the type series and who contributed greatly to our knowledge of the World Ichneumonoidea during his lifetime. Ascogaster varipes Wesmael Fig. 14 Ascogaster imnpes Wesmael 1835: 234. Lectotype female, BEL- GIUM: Brussels (designated by Huddleston 1984). Shenefelt 1973: 837; Huddleston 1984: 370; Tobias 1986: 305; Papp 1989:298. Ascogaster cavifrons Thomson 1874: 585. Lectotype female , SWEDEN: Skane, Torekov (desiganted and syn. by Huddleston 1984). Shenefelt 1973: 818. Ascogaster sternalis Thomson 1874: 587. Lectotype female, SWEDEN: Smaland (designated by Huddleston 1984 and syn. by Telenga 1941). Ascogaster jaroslawensis Kokujev 1895: 86. Holotype female, USSR: 'Jaroslaw' (syn. by Huddleston 1984 ). Shenefelt 1973: 824; Tobias 1986: 304 (as a valid species). Ascogaster variipes; Telenga 1941: 322. Diagnosis. — Antenna with 32-36 flagellomeres, slightly dilated medially (females); ocellar triangle obtuse, ocelli almost on line; face (Fig. 14) gener- ally rugose-punctate, all hairs downwards; clypeus with its apical border transversely impressed, without medial tooth or excision; mandibles with a deep semicircular depression at base; notauli distinct, foveolate, rest of mesonotum punctate except posteriorly areolate-rugose; mesopleuron coarsely rugose so that precoxal suture not easily distinguished; propodeum completely areolate- rugose; hind coxa strongly transversely strigate; carapace short, oval, deep in lateral view, ventral opening of carapace short, distinctly in front of apex. Specimens Examined.— SHANDONG PROV- INCE: 2 females, Lao-shan, 800 m. Additional Specimens Examined. — Lectotype of varipes Wesmael, female, BELGIUM: Brussels (IRSNB). Lectotype and paralectotypes of varipes Wesmael, BELGIUM: 2 females, 3 males (IRSNB). Hosts.— Shenefelt (1973) listed many host records. However, much of this information is of little value because of doubt about the accuracy of identification of the parasite species involved (see Huddleston 1984). Distribution. — China (Shandong Province), Korea, "USSR" and many other European coun- tries (see Huddleston 1984). Remarks. — This species is characterized on account of the semicircular depression at the base of the mandibles, the strongly strigate hind coxa, the characteristics of the clypeus and the shape of the carapace. The two female Chinese specimens examined here differ from the type series of imripes 296 Journal of Hymenoptera Research Wesmael in having a rugose-punctate face instead of a finely areolate-rugose face as in the type series. Ascogaster yunnanica Tang and Marsh, new species Male. — Length of forewing 2.7 mm, of body 3.3 mm. Head. — Antenna incomplete, 1st flagellomere about 3.0 times as long as broad, 2-6 th flagellomeres 2.0-2.5 times as long as broad. Temple constricted behind eyes, slightly shorter than eye in dorsal view. Occiput deeply concave. Ocelli small OO = 4.0 OD; ocellar triangle acute, ocelli not on line. Frons behind antenna moderately depressed, smooth. Eyes slightly protuberant, glabrous with- out distinct setae. Malar space short. Gena in face view strongly constricted. Face slightly convex, finely and sparsely punctate, about twice as broad as high, the hairs on the upper part of face pointing upwards. Clypeus slightly protuberant, more sparsely punctate than face, apical border straight without flange or tooth. Mesosoma. — Pronotum projecting little in front of mesonotum, deep foveate dorsolateraly . Notauli foveolate, coalescing posteriorly in a fine areolate- rugose area; rest of mesonotum punctate. Precoxal suture shallow foveate anterodorsally, indistinct posteroventrally; rest of mesopleuron smooth posteroventraly, punctate anterodorsaly. Propodeum not distinctly divided into dorsal and posterolateral surfaces, strongly rugose but with no dentates. Hind coxa smooth. Vein r of forewing about twice as long as 3-RS. Metasoma. — Carapace very long, CL/CW = 2.5, oval in dorsal view, deeper and not so strongly flat in lateral view, areolate-rugose, but sparsely punctate apically. Ventral opening of carapace long, at apex of carapace. Color. — Black; all legs yellow except hind coxa mostly black, hind femur and tibia apically and tarsi infuscate; palpi yellow-brown. Female. — Unknown. Holotype Male.— YUNNAN PROVINCE: Kunming, 111-30-81, J. H. He. Deposited in ZAU. Host. — Unknown. Distribution. — China (Yunnan Province). Remarks. — Morphologically this species is simi- lar to townesi from which it differs not only in the characters mentioned in the key but aslo in the smaller body and the less flat carapace. Etymology. — The species name is in reference to the type locality. ACKNOWLEDGEMENTS We thank the following people for loan of types and much of unidentified specimens or for providing valuable information: C. van Achterberg (The Netherlands), A. Austin (Australia), H. F. Chao (FAC), X. X. Chen (ZAU), L. Y. Chou (TARI), P. Dessart (IRSNB), J. H. He (ZAU), T. Huddleston (BMNH), K. Maeto (UEI), I. Nauman (Australia), J. Papp (HNHM), D. J. Preston (BM), M. J. Sharkey (CNC), S. R. Shaw (Wyoming), G. N. Shida (BM), M. Suwa (UEI), D. Wahl (AEIG), J. Y. Wang (ZRI). We also thank C. M. Liu (FAC) for help with translations of some Russian papers. The senior author offers special thanks to Prof. H. F. Chao for his continued helpfulness and encouragement and to Chunying for her patience and understanding. This work was partly supported by a grant to Y. Q. Tang by the National Educational Commission of the People's Republic of China. LITERATURE CITED Achterberg, C. van. 1976. A preliminary key to the subfami- lies of the Braconidae (Hymenoptera). Tijdschrift voor Entomologie 119: 33-78. Achterberg, C van. 1988. Revision of the subfamilv Blacmae Foerster (Hymenoptera: Braconidae). Zoologische Verhandelingen 249: 1-324. Achterberg, C. van. 1990a. Illustrated kev to the subfamilies of the Holarctic Braconidae (Hvmenoptera: Ichneumonoidea). Zoologische Mededelmgen 64: 1-20 Achterberg, C. van. 1990b. Revision of the western Palaearctic Phanerotomini (Hymenoptera: Braconidae). Zoologiche Verliandelingen 255: 1-106. Baker, C. F. 1926. Braconidae-Cheloninae of the Philippines, Malaya, and Australia. The Philippine Journal of Science 31: 451-489. Bovce, H R. 1936. Laboratory breeding of Ascogaster carpocapsae Viereck with notes on biology and larval morphology. Canadian Entomologist 68: 241-246. Cox, J. A. 1932. Ascogaster carpocapsae Viereck, an important parasite of the codling moth and the oriental fruit moth. Technical Bulletin of the New York State Agricultural Ex- periment Station 188: 1-26. Fahringer, J. 1934. Opuscula Braconologica, Palaearktische Re- gion 2: 321-594. Fullaway, D. T. 1919. New genera and species of Braconidae mostly Malayan, journal of Straits Branch, Royal Asiatic Society 80: 39-61. Harris, R. A. 1979. A glossary of surface sculpturing. Occa- sional Papers in Entomology, California Department of Food and Agriculture 28: 1-31. Volume 3. 1 994 297 He, J. H., X. X Chen and Y. Ma. 1989. The Braconid (Hy- menoptera) parasites of Adoxophyes orana Fischer von Rosterstamm (Lepidoptera: Tortncidae) from China Acta Agricultae Universitatis Zhejiangensis 15 (4): 437-439. Huddleston, T. 1984. The Palaearctic species of Ascogaster (Hymenoptera: Braconidae). Bulletin of the British Mu- seum (Natural Htston/) (Entomology) 49: 341-392. Kainoh, Y. 1986. Mating behavior of Ascogaster reticulatus Watanbe (Hymenoptera: Braconidae, an egg-larval para- sitoid of the smaller tea tortrix moth, Adoxophyes sp. (Lepidoptera: Tortncidae) I. Diel patterns of emergence and mating, and some condintions for mating. Applied Entomology and Zoology 21: 1-7. Kainoh, Y. 1988. Some factors influencing sex ratio in Ascogaster reticulatus Watanabe (Hymenoptera: Braconidae). Applied Entomology and Zoology 23:35-40. Kainoh, Y., T. Hiyori and Y. Tamaki. 1982. Kairomone of the egg-larval parasitoid, Ascogaster reticulatus Watanabe (Hymenoptera: Braconidae). Applied Entomology and Zo- ology \7: 102-110. Kainoh, Y., T. Nemoto, K. Shimizu, S. Tatsuki, T. Kusano and Y. Kuwahara. 1991. Mating behavior of Ascogaster reticulatus Watanabe (Hymenoptera: Braconidae), an egg-larval parasitoid of the smaller tea tortrix, Adoxophyes sp. (Lepidoptera: Tortricidae) II. Identification of a sex pheromone. Applied Entomology and Zoology 26: 543-549. Kainoh, Y. and Y. Tamaki. 1982. Searching behavior and oviposition of the egg- larval parasitoid, Ascogaster reticulatus Watanabe (Hymenoptera: Braconidae). Ap- plied Entomology and Zoology 17: 194-206. Kainoh, Y. and S. Tatsuki. 1988. Host egg kairomones essen- tial for egg-larval parasitoid, Ascogaster reticulatus Watanabe (Hymenoptera: Braconidea I, Internal and external kairomones. journal of Chemical Ecology 14: 1475- 1484. Kainoh, Y., S. Tatsuki. and T. Kusano. 1990. Host moth scales; a cue for host location for Ascogaster reticulatus Watanabe (Hymenoptera: Braconidae). Applied Entomology and Zoology 25: 17-25. Kainoh, Y., S. Tatsuki, H. Sugie and Y. Tamaki. 1989. Host egg kairomones essential for egg-larval parasitoid, Ascogaster reticulatus Watanabe (Hymenoptera: Braconidae) II. Identification of internal kairomone. Journal of Chemical Ecology 15: 1219-1229. Kamano, Y., K. Shimizu, Y. Kainoh and S. Tatsuki. 1989. Mating behavior of Ascogaster reticulatus Watanabe (Hy- menoptera: Braconidae), an egg-larval parasitoid of the smaller tea tortix, Adoxophyes sp. (Lepidoptera: Tortricidae). Applied Entomology and Zoology 24: 372-378. Kawakami, T. 1985. Development of the immature stages of Ascogaster reticulatus Watanabe (Hymenoptera : Braconidae), an egg-larval parasitois of the smaller tea tortrix moth, Adoxophyes sp. (Lepidopatera: Tortricidae). Applied Entomology and Zoology 20: 380-386. Kawakami, T. and Y. Kainoh. 1985. Host discrimination and competition in the egg larval parasitoids, Ascogaster reticulatus Watanabe (Hymenoptera: Braconidae). Ap- plied Entomology and Zoology 20: 362-364. Kawakami, T. and Y. Kainoh. 1986. Host specificity of the egg-larval parasitoid, Ascogaster reticulatus Watanabe (Hymenoptera: Braconidae) among five tortricid and one noctuid species. Applied Entomology and Zoology 21 : 8-14. Papp, ]. 1989. Braconidae (Hymenoptera) from Korea XI. Acta Zoologica Hunganca 35(3-4): 295-326. Quicke, D. L. J. and C. van Achterberg. 1990. Phy logeny of the subfamilies of the family Braconidae (Hymenoptera: Ichneumonoidea). Zoologische Verhandehngen 258: 1-95. Rosenberg, H. F. 1934. The biology and distribution in France of the larval parasites of Cydia pomonella L. Bulletin of Entomological Research, 25: 201-256. Shaw, M. R. and T. Huddleston. 1991. Classification and biology of braconid wasps (Hymenoptera: Braconidae). Handbooks for the Identification of British Insects 7 (11): 1- 126. Shaw, S.R. 1983. A taxonomic study of Nearctic Ascogaster and a description of a new genus Leptodrepana (Hy- menoptera: Braconidae). Entomography 2: 1-54. Shenefelt, R. D. 1973. Catalogus Hymenopterorum (Noi<. ed. part 10 Braconidae, 6:813-936. s'Gravenhage. Sonan, J. 1932. Notes on some Braconidae and Ichneumomdae from Formosa with descriptions of 18 new species. Transactions of the Natural History Society of Formosa 22: 66-86. Szepligeti, G. V. 1905. Exotische Braconiden aus den Aethiopischen, Dnentalischen und Austtralischen Regionen. Annates Musei Nationalis Hungarici 3: 25-55. Szepligeti, G. V. 1908. Braconidae und Ichneumomdae. Die Fauna Sudtoest- Australiens 1(9): 317-324. Telenga, N. A. 1941. Insects Hymenoptera, Family Braconidae, Subfamily Braconinae (continued) and Sigalphinae Fauna SSSR 5 (3): 1-466. Tobias, V. I. 1986a. New species of Subfamily Cheloninae (Hymenoptera: Braconidae) from the Far East of the USSR. Trudy Zoologicheskogo Instituta Akademiya Nauk SSSR 159: 3-17 (in Russian). Tobias, V. I. 1986b. Guide to the insect of European part of the USSR Hymenoptera. Opredeliteli Faune SSSR, 3 (4): 1-500 (in Russsian). Tobias, V. I. 1988. Two new species of Braconidae of the subfamlily Cheloninae (Hymenoptera) from the pro- tected territoties of the Liruanian SSR. Acta Entomology Lituaman 9: 89-94. Walker, A. and T. Huddleston. 1987. New Zealand chelonine braconid wasps (Hymenoptera: Braconidae). journal of Natural History 21: 339-361. Watanabe, C. 1937. A contribution to the knowledge of the Braconid fauna of the Empire of Japan, journal of the Faculty of Agriculture Hokkaido Imperial University 42(1): 1-188." Watanabe, C. 1967. Description of a new species of the genus Ascogaster Wesmael and notes on synonym of Apanleles species (Hymenoptera: Braconidae). Insecta Matsumurana 29(2): 41-44. Wesmael, C. 1835. Monographic des Braconides de Belgique. Noui'eaux Memoires de V Academic Royale des sciences et belles-lettres de Bruxelles 9: 1-252. Yoneda, Y. 1978. A new species of the genus Ascogaster Wesmael (Hymenoptera, Braconidae) from Japan. Kontyu 46: 291-296. Zettel, V.H.I 990. Eine Revision der Gattungen der Cheloninae (Hymenoptera, Braconidae) mit Beschreibungen neuer 298 Journal of Hymenoptera Research Gattungen und Arten. Annates Naturlustorisches Mu- seum Wien 91(B): 147-196. NOTE ADDED IN PROOF The following paper came to our attention after our manuscript was sent to the printer: Chen Jiahua, Huang Juchang and Wu Zhishan. 1994. Notes on two new species and six new records of the genus Ascogaster Wesmael from China (Hy- menoptera: Braconidae: Cheloninae). Journal of the Fujian Agricultural College (Natural Sciences Edi- tion) 23(1)51-57. Although we were not able to study the specimens used by Chen et al., we have made comments below on each species mentioned by them based on a translated version of their paper. Ascogaster albitarsis Sonan. The characters mentioned in their key are not significant to separate this species from reticulatus based on our study (see discussion under reticulatus). Ascogaster arisanica Sonan and Ascogaster bidentula Wesmael. During our study we examined the type ma- terial of arisanica, bidentula and atamiensis (synonym of bidentula). All of the material we examined under the name arasanica agreed with the type but none agreed with types of bidentula or atamiensis. Furthermore, we found that the characters used by Chen et al. and Huddelston (1984) are not realiable in distinguishing bidentula and arisanica (see our discussion of arasanica). Most of the material examined by Chen et al. of bidentula and arasanica was collected at the same locality and same date; thus we suspect that their material is prob- ably arasanica. This problem is further complicated by the fact that atamiensis was recorded from China by Fahringer (1938, Ark. Zool. 30:3) and Shenefelt (1973). Obviously this species complex needs to be studied further Ascogaster consobrina Curtis and Ascogaster inf acetus Chen and Huang. The new species described in Chen et al., mfacetus, will run to consobrina in our key. Fig. IB in Chen et al. which is an illustration of the mesopleuron is not clear. Thus, we must reserve comment on this spe- cies until the type can be observed Ascogaster longicornis Huddleston. We treated this species as a synonym olformosensis based on type examination. Ascogaster perkinsi Huddleston. See notes under wuyiensis below. Ascogaster reticulatus Watanabe. Although this species was mentioned in their key, it was not discussed in the text. Thus, we must see their specimens before deciding upon the relationship of this species and albitarsis. Ascogaster ruf idens Wesmael. We did not find this species in any of the material that we studied. Again, their mate- rial should be studied before we confirm this species is in China. Ascogaster wuyiensis Chen and Huang. This species will run to perkinsi in our key based on the description in Chen etal, and the color of the clypeus and theshapeof the yellow mark at the base of the carapace may be only variation in perkinsi. We described a new species, lini, which is related to wuyiensis and perkinsi. The validity of these three species must wait until specimens of wuyiensis are examined. Figs. 1-2. Fore wings, somewhat diagrammatic. 1, Ascogaster. 2, Chelonus. SM=submarginal cells; D=discal cell. Volume 3. 1994 299 Figs. 3-8. Faces of Ascogaster species. 3, arisanica Sonan. 4, dimorpha n. sp. 5 , formosensis Sonan. 6, grandis n. sp. 7,fullaivayi (Baker). 8, setula n. sp. 300 Journal of Hymenoptera Research Figs. 9-14. Faces of Ascogaster species. 9, lini n. sp. 10, perkinst Huddleston. 11, stmenoi'i Telenga. 12, townesi n. sp. 13, rugulosa n. sp. 14, varipes Wesmael. Volume 3. 1 994 301 Figs. 15-20. Heads of Ascogaster species. 15-16, hein.sp.: 15, dorsal view; 16, face. 17-18, quadridentataVJesmael. 17, dorsal view; 18, face. 19, armatoides n. sp. 20, consobrina Curtis. 302 Journal of Hymenoptera Research ■. ■ .■;.■■■ .: ' "■" -^^^Sk ,.'-' ' ,--■-, . ■ -" . - . ■"- i^v^B *>.! 'T/ >. ■* '' ■ '• ' "•'-' • ^~ • ' - r -""^IS^sJHf ^■'^-:>L* v.r:;;.;:- .■•.,•;..•. ^titm K © Mm mm Figs. 21-28. Morphology of Ascogaster species. 21-26, metasomal gaster: 21, grandis n. sp.; 22, perkmsi Huddleston; 23, acutus J. HYM. RES. Vol. 3, 1994, pp. 303-308 Systematics of Pseudomethoca areta (Cameron): Sex association, description of the male and a gynandromorph, and a new synonymy (Hymenoptera: Mutillidae). DlOMEDES QUINTERO A. AND ROBERTO A. CaMBRA T. Museo de Invertebrados G.B. Fairchild, Universidad de Panama, Estafeta Universitaria, Panama, Repiiblica de Panama; (DQA) Smithsonian Tropical Research Institute, Unit 0948, APO AA 34002-0948. Abstract. — A gynandromorph of Pseudomethoca areta (Cameron, 1895) is described, and previously published cases of gynandromorphism in Mutillidae are reviewed. Sex association permits recognition of the undescribed male of P. areta. We place Pseudomethoca panamensis (Cameron, 1895) in New Synonymy with P. areta. Resumen. — Se describe un individuo ginandromorfo de Pseudomethoca areta (Cameron, 1 895) y se presenta un resumen de casos previamente descritos de ginandromorfia en Mutillidae. Se lleva a cabo la asociacion sexual y se describe al macho de P. areta, hasta ahora desconocido. Pseudomethoca panamensis (Cameron, 1895) se coloca como Nueva Sinonimia de P. areta. INTRODUCTION Mutillids are solitary parasitoid wasps that exhibit great sexual dimorphism, making sex as- sociations difficult. The New World genus Pseudomethoca Ashmead illustrates this problem; only one-fifth of its 103 described species are known from both sexes. Distinctly fewer sex associations have been obtained for Neotropical than for Ne- arctic species of Pseudomethoca. Seventeen out of 45 Nearctic species known (37.8%) have both sexes recognized (Krombein, 1992). Of the remaining 28 species, 20 are known only from females, and eight only from males. In contrast, only four of some 58 Neotropical species of Pseudomethoca are known from both sexes (6.9%) (Nonveiller, 1990; Cambra & Quintero, 1992). Of the remaining 54, 46 are known from females only, and eight from males only. Success in associating the sexes will facilitate future biological work on the group and will solve some of the annoying taxonomic prob- lems. We suspect that only about one-third of the species of Pseudomethoca in the Neotropics have been described. Previous taxonomic work on Pseudomethoca was done by Mickel (1924, 1935, revision of North American species; 1952, key to females of Guyanan species), Schuster (1945, key to Caribbean species), and Krombein (1992). Gynandromorphy is a developmental phe- nomenon useful for associating the sexes in some extremely dimorphic animals, including mutillids (Mickel 1928, 1936, 1952; Bischoff 1931). Unfortu- nately, gynandromorphs are rare in Mutillidae. After examining more than 15,000 mutillid speci- mens, we have discovered only two gynandro- morphs. A review of the literature revealed only six previously published cases (Table 1). We re- port here the second known Neotropical mutillid gynandromorph. We recently discovered a gy- nandromorph of Timulla labdace Mickel, from Panama, that will be described in a separate pub- lication. 304 Journal of Hymenoptera Research Pseudomethoca areta (Cameron) Figs. 1 - 4 Sphaerophthalma areta Cameron, 1895: 332, pi. 14, fig. 12, female. Bugaba, Chinqui Province, Panama, Champion col., BM(NH), London, Type 15.822, examined; Pseudomethoca areta: Cambra & Quintero, 1992: 474. Sphaerophthalma panamensis Cameron, 1895: 334-35, female. Bugaba, Chinqui Province, Panama, Champion col., BM(NH), London, Type 1 5.833, examined; Pseudomethoca panamensis: Cambra & Quintero, 1992: 475. NEW SYN- ONYMY. Diagnosis. — In Mickel's revised key (1935) it runs to P. vanduzei Bradley in couplet five. The male of P. areta differs from P. vanduzei as follows: posterior half of tegula bent downward so as to form a posterior face at a sharp angle with the dorsal surface (in P. vanduzei the tegula is uni- formly convex, without a posterior face); anterior margin of clypeus in P. areta is bidentate (it lacks teeth in P. vanduzei); mandibles tridentate at the tip in P. areta (bidentate in P. vanduzei); integument of abdomen mostly ferruginous in P. areta (totally black in P. vanduzei). Pseudomethoca areta is en- demic to Panama, and P. vanduzei is present in the southeastern United States (Krombein, 1979). Description. — Integument black, except apex of tergum one and abdominal segments two to six, orange. Head large, subrectangular in dorsal view, as wide as thorax, clothed with long, erect and recumbent white pubescence; row of six to eight long, erect, dark hairs near inner margin of eyes; posterolateral angles of head not dentate. Man- dibles tridentate distally; clypeus strongly bidentate medially on the cephalic margin; disk of clypeus densely punctate. Scape with a strong longitudinal carina beneath; first flagellomere equal in length to second. Front, vertex and genae coarsely and confluently punctate. Antennal scrobes and genae not carinate. Ocelli small, dis- tance between eye margin and lateral ocelli equal to approximately five times the greatest diameter of the latter. Pronotum, mesonotum and scutellum with close, more or less confluent punctures, punctures about the size of those on head. Propodeum strongly and coarsely reticulate dorsally and pos- teriorly. Tegula punctate throughout; posterior part of tegula bent downward so as to form a posterior face at a sharp angle with the dorsal surface. Humeral angles of pronotum without any evidence of a carina. Propleura and mesopleura with close punctures; metapleura smooth, with- out punctures; sides of propodeum with only a few scattered punctures. Pronotum, scutellum, metanotum and dorsal face of propodeum, clothed with sparse, long erect white pubescence; mesonotum clothed with sparse black pubescence; tegula clothed with intermixed, black and white pubescence; metapleura and sides of propodeum almost bare, with only sparse white micropubescence. Anterior and intermediate coxae without teeth or keels; posterior coxa with a keel on inner mar- gin. Legs clothed with sparse white pubescence. Calcaria pale. Abdomen with segment one completely sessile with second. Terga one and two with small, sepa- rated punctures, except the apical margins with close punctures; terga three to six with small, close punctures. Pygidium rugose. Felt line 0.6 x as long as lateral margin of tergum two. Sternum one almost smooth, with only a few, sparse punctures, and with a low, median longitudinal carina on anterior two-fifths. Sternum two with sparse, mod- erate punctures. Sterna three to six with close, moderate punctures. Posterior margin of hypopygium evenly convex. Tergum one clothed with sparse, long erect, white pubescence. Terga two to six with sparse long erect, orange pubes- cence, the apical margins with a band of dense, recumbent, orange pubescence. Last tergum clothed with black pubescence. Sterna one to six clothed with white pubescence, except lateral margins of sterna two to four, with orange pubes- cence. Last sternum clothed with white pubes- cence and a few intermixed fuscous hairs. Wings infuscated, especially apically; forewing with two well defined submarginal cells and traces of a third. Body length; 10.6 mm. ALLOTYPE Male Information.— PANAMA: Darien Province, Cruce de Mono, Estacion INRENARE, 8 Feb 1993 (yellow trap). R. Cambra T., J. Coronado, Museo de Invertebrados "G. B. Fairchild", Universidad de Panama (MIUP). Additional Material Examined. — PANAMA: Darien Province, Cruce de Mono, Estacion INRENARE, R. Cambra, J. Coronado, 144 females and 27 males, 8 Feb - 4 Mar 1993, deposited in MIUP, NMNH-Smithsonian Institution, Univer- sity of Minnesota Insect Collection, and BM(NH). Body length varies in males from 8.3 to 1 1 .0 mm, in Volume 3, 1 994 305 Figs. 1-4. Pseudomethoca areta. 1. Male allotype, dorsal habitus. 2. Gynandromorph, dorsal habitus. 3. Male genitalia, split drawing, dorsal = right; ventral = left. 4. Penis valve, side view. 306 Journal of Hymenoptera Research females from 7.8 to 11.1 mm. Comments on Sex Association and Nezo Syn- onymy. — Sex association is based on observations of males courting and mating with females in the field, and has been corroborated in the laboratory with mating experiments. Courtship is very brief, lasting 15-40 seconds, and consists of bursts of rapid vibrations of the wings and antennae, inter- spersed by short hopping flights. The male climbs onto the female and grasps her neck with his mandibles and attempts to mate with her. We have not seen heterospecific courtship or mating in Psendomethoca. Erroneous heterospecific sex as- sociations may be made if containers or outdoor sites become contaminated by pheromones re- leased from a female of a different species that recently occupied that site (personal observations). Males of Psendomethoca areta, like those of the genotype and males of numerous other species of Pseudomethoca we have examined, have a strong longitudinal carina beneath the scape. Therefore, we consider erroneous Casal's (1965) observation that males of the ill-defined genus Darditilla Casal (genotype is only male known, and 35 other spe- cies, known only from females) differ from those of Psendomethoca in having a carina beneath the scape, said to be absent in the latter. The male genitalia of P. areta are symmetrical (Fig. 3), as are those of all other Neotropical sphaeropthalmine males we have examined, and phoretic mating is absent in this group. In contrast, the Neotropical mutilline males of Timulla (Timnlla) present strongly asymmetrical genitalia and phoretic mat- ings (Cambra & Quintero, 1992). The asymmetry in the male genitalia possibly functions to provide a better hold, or grasp, of the female while air- borne. Cameron's types of areta and panamensis are both from Bugaba, and we found them to be identical; the name areta has page precedence over that of panamensis. Psendomethoca areta is closely related to P. hecate (Gerstaecker, 1874), from Costa Rica, differing only in the integumental coloration of the vertex and dorsolateral areas of the thorax. We suspect that they are the same species. To confirm the synonymy we need to examine Gerstaecker's type specimen and to compare the genitalia of males sexually associated with hecate females with those of the males described here. Gynandromorph Individual of Psendotnethoca areta Fig. 2 Description. — Head identical to that of a nor- mal female, without a trace of male characters. Thorax and legs identical to those of a normal male, without a recognizable trace of female char- acters. The anterior wings have abnormally thin, translucent venation; they are torn along their posterior half, and we suspect were never func- tional. Abdomen with six segments, as in the female. First tergum completely male. Second ter- gum a mosaic: right half with coloration and pu- bescence of male and female; left half is com- pletely male-like. Third tergum, right half with female characteristics only; left half is a mosaic with integument coloration and pubescence both of male and female. Abdominal segments four to six are female only. Second sternum with right half female and left half male, same as other sterna, except for sternum three, identical to that of a female. Body length: 11.1 mm. Data and comments on the gynandromorph. — The gynandromorph individual was collected on 26 February 1993, at 10:00 AM, in the general locality of the allotype. When we first noticed it, the individual was walking over dry leaves on the ground. Shortly afterward, we watched a male arrive, flying upwind, attracted by what we thought was a normal female. The male quickly attempted to mate but encountered indifference on the part of the female-like individual. Female mutillids are known to attract winged males by means of wind-dispersed pheromones (see Cambra & Quintero, 1993). The upwind arrival of the male suggests that the gynandromorph individual was secreting female pheromones. The abnormally thin and quite battered forewings, suggest that the animal was unable to fly, although it had perhaps attempted to. The specimen exhibits anterior/ posterior division of male and female compo- nents, as well as mosaic segments, a type of gynan- dromorphy not previously described for mutillids (see Table 1). Volume 3. 1994 307 ACKNOWLEDGMENTS We thank the Smithsonian Tropical Research Institute (STRI), particularly Ira Rubinoff, for providing research fa- cilities. We are grateful to the British Embassy in Panama, in particular Ambassador Thomas H. Malcomson, for securing travel funds to England for R.A.C.; to the Entomology De- partment personnel of the Natural History Museum, Lon- don, BM(NH) for providing R.A.C. with working facilities and assistance during his visit. We also thank Indra Candanedo and Roberto Arango, Instituto Nacional de Recursos Naturales Renovables, who aided us with permits and logistics in the Darien National Park. This project was financed in part by Vicerectoria de Investigacion y Postgrado, Universidad de Panama, Fund No. 1-4500-91-12. Our thanks to Karl V. Krombein, Smithsonian Institution, Annette Aiello and David W. Roubik, STRI, and an anonymous reviewer, for reading and making comments on the manuscript, James Coronado, now at STRI, for providing very valuable help in the field, and to Angel Aguirre, STRI, for bibliographic assis- tance and location of references on gynandromorphs. LITERATURE CITED Bischoff, H. 1931. Der typus der Mutilla dubia ¥.— ein gynander. Mitteilungen deutschen entomologischen gesellschaft 2(4): 54-56." Cambra, R. A. and D. Quintero A. 1992. Velvet ants of Panama: distribution and systematics (Hymenoptera: Mutillidae), pp. 459-478. In: Quintero A., D. & A. Aiello (eds.) Insects of Panama and Mesoamerica: Selected studies. Oxford University Press, Oxford. Cambra, R. A. and D. Quintero A. 1993. Studies on Timulla Ashmead (Hymenoptera: Mutillidae): New distribu- tion records and synonymies, and descriptions of previ- ously unknown allotypes. Pan-Pacific Entomologist 69(4): 296-310. Cameron, P. 1895-1896. Hymenoptera, vol. 2. Biologia Centrali Americana, pp. 262-360, plates 13-14. Casal, O. H. 1965. Darditilla nuevo genero Neotropical de Sphaeropthalminae (Hym. Mutillidae). Eos, Madrid, 41: 9-18. Dalla Torre, C. G. and C. Friese. 1899 (1898). Die hermaphroditen und gynandromorphen Hymenop- teren. Berichte des naturxuissenschaftlich - Medizmischen Vereins in Innsbruck 24: 1-96. Gerstaecker, C. E. 1874. Mutillarum Americae mendionalis indigenarun synopsis systematica et synonymica. Archil' fur Naturgeschiechte 40: 41-77. Krombein, K. V. 1979. Family Mutillidae, pp. 1276-1314. In: Krombein, K. V. et al., eds. Catalog of Hymenoptera m America North of Mexico, Vol. 2. Smithsonian Institution Press, Washington, D.C Krombein, K. V. 1992. Host relationships, ethology and sys- tematics of Pseudomethoca Ashmead (Hymenoptera: Mutillidae, Andrenidae, Hahctidae and Anthophondae). Proceedings of the Entomological Society of Washington 94(1): 91-106. Maeklin, F. W. 1856. Om hermafroditism bland insekterna, samt beskrifning ofver en i Helsingfors funnen hermafrodit af Mutilla obscura Nyl. ofver. af Finska Vetenskaps-Societens Forliandlingar 3: 106-112. Mann, W. M. 1915. A gynandromorphous mutillid from Montana. Psyche 22: 178-180. Mickel, C. E. 1924. A revision of the mutillid wasps of the genera Myrmilloides and Pseudomethoca ocurring in America North of Mexico. Proceedings of the United States National Museum 64(15): 1-51. Mickel, C. E. 1928. Biological and taxonomic investigations on the mutillid wasps. United States National Museum Bulletin 143: 1-351. Mickel, C. E. 1935. Descriptions and records of mutillid wasps of the genera Myrmilloides and Pseudomethoca (Hymenoptera: Mutillidae). Transactions of the American Entomological Society 61: 383-398. Mickel, C. E. 1936. New species and records of Nearctic mutillid wasps of the genus Dasymutilla. Annals of the Entomological Society of America 29: 29-60. Mickel, C. E. 1952. The Mutillidae (wasps) of British Guiana. Zoologica 37(3): 105-150. Nonveiller, G. 1990. Catalog of the Mutillidae, Myrmosidae and Bradynobaenidae of the Neotropical region including Mexico (Insecta: Hymenoptera). SPB Academic Publishing bv, The Netherlands, 150 pp. Schuster, R. M. 1945. A new species of Pseudomethoca (Mutillidae) from the West Indies. Bulletin of the Brooklyn Entomological Society 40: 7-8. Wheeler, W. M. 1910. A gynandromorphous mutillid. Psyclie 17: 186-190. 308 Journal of Hymenoptera Research Table 1. Previously published cases of gynandromorphy in Mutillidae Species Type Wing Country Ref. & tegula Mutilla europea obscura bilateral, right f, left m Finland Maeklin 1856 DallaTorre & Friese 1899 Dasi/nmtilla cypris [=hora] head, thorax, abdom. segm. 5-7 m; 1-4 f USA Mickel 1928 Dasymutilla gloriosa [=reperticia] Dasymutilla vestita [=euchroa =fulvohirta] Pseudomethoca frigida [=canadensis] TraumatomutiUa dubia not described decussated, head half: right m, left f; thorax & abdomen: right f; left m bilateral, right m, left f head, thorax, 1st abdom. segment m, rest abdomen mosaic USA Mickel 1936 USA Mann 1915 USA Wheeler 1910 Am. Mer. Bischoff 1931 [Guyana] Mickel 1952 Abbreviations: f, female; m, male; +, well developed; 0, absent. Volume 3, 1994 309 Additions and Corrections to Volume 2, Number 1, 1993 Publication date: the correct date of publication for Volume 2, Number 1, is November 17, 1993. All ">", "<" and "=" signs were inadvertently left out of the text in the following two articles resulting in a significant loss of information. The affected text is as follows: Davidson, D. W. and D. McKey, "The evolutionary ecology of symbiotic ant-plant relationships," pp. 13-83. P. 14, column 2, line 42 "(N = 8)" P. 14, column 2, line 43 "(N = 6)" P. 14, column 2, line 44 "(N = 4)" P. 14, column 2, line 46 "(N = 5) and "(N = 6)" P. 15, column 1, lines 11, 12 "(N = 8), (N = 7) and (N > 5)" P. 15, column 1, line 35 "(N = 50-60 ant-plant species)" P. 15, column 1, line 36 "(N = 20)" P. 15, column 1, line 37 "(N = 15)" P. 15, column 1, lines 38, 39 "(N = 12 species) and (N > 6)" P. 15, column 1, line 40 "(N = 15)" P. 15, column 1, line 42 "(N = 3)" P. 15, column 2, line 6 "(N = 23)" Additional corretions are as follows: P. 69, Appendix 1 : reference to Manriquez & Dirzo 1990 should read "Iborra-Manriquez & Dirzo 1990." P. 70, Appendix 1: entry for Barteria should read ". . . HFOWEb Y Z . . . " P. 74, Appendix 1: entry for Maieta should read "... B;D; Vasconcelos 1990, 1991, IP; Herre et al. 1986" P. 78, Appendix 1: entry for Cordia should read ". . . ufRIy Y,i' GH . . ." Ward, P. S., "Systematic studies on Pseudomyrmex acacia-ants," pp. 117-168. P. 121, column 2, line 30 "PLI2 > 0.77" P. 121, column 2, line 45 "worker PLI > 0.71, queen PLI > 0.64" P. 122, column 1, line 36 "HW < 0.85" P. 122, column 1, line 37 "HW > 0.85" P. 122, column 1, line 41 "PL/HW > 0.71" P. 122, column 2, lines 8-9 "CI > 0.94 and/or HW > 0.96" P. 122, column 2, line 15 "SL/HL > 0.22" P. 122, column 2, line 16 "SL/HL < 0.21" P. 130, line 33 "CI = 1.12" P. 130, line 35 "CI = 0.80" P. 132, line 40 "SL/HL < 0.21" P. 132, line 42 "SL/HL > 0.22" P. 133, lines 4-5 "CI > 0.94 and/or HW > 0.96" P. 133, column 1, line 8 "MD8/MD9 = 0.70" P. 133, column 1, line 11 "MD4/MD5 = 0.74" P. 135, column 2, line 4 "worker REL < 0.50, queen REL < 0.48" P. 135, column 2, lines 5-6 "worker PLI < 0.71, queen PLI < 0.63" P. 135, column 2, lines 29-30 "queen PLI = 0.65, queen PL/HL = 0.49" P. 145, column 1, lines 26-27 "REL < 0.45, REL2 < 0.56, EL/LHT < 0.61' 310 Journal of Hymenoptera Research P. 145, column 1, line 30 "FCI = 0.055" P. 145, column 2, lines 7-8 "worker PWI3 > 0.50, worker PPW1 < 1.30" P. 146, column 2, line 10 "CI = 0.61" P. 146, column 2, line 12 "LHT/HW = 1.12" P. 152, column 2, lines 14-15 "(< 0.10 mm)" P. 155, column 1, line 10 "REL = 0.66" P. 157, column 1, lines 20-21 "PLI = 0.55" P. 157, column 2, line 12 "MD8/MD9 = 0.70" P. 157, column 2, line 17 "MD4/MD5 = 0.65" P. 158, column 1, lines 7-8 "PWI3 = 0.60" P. 158, column 1, line 41 "REL = 0.65" P. 158, column 1, lines 42-43 "MFC = 0.02" P. 158, column 1, line 44 "CI = 0.85" P. 159, column 1, line 3 "CI = 1.12" Additional corrections are as follows: P. 163, Table 2, column 31: Pseudomyrmex nigrocinctus and P. particqis should be "1", not "0" for character 31. P. 121, column 1, line 41: for "public domain software" read "Shareware software". P. 131, line 34: for "Figs. 10, 34" read "Figs. 11, 34". Kazenas, V. L. and B. A. Alexander, "The nest, prey, and larva of Etttomosericus kaufmani", pp. 221- 226. Nearly all references to illustrations should be corrected as follows: P. 221, column 2, line 19: for "Fig. 2" read "Fig. 4". P. 221, column 2, line 23: for "Fig. 3" read "Fig. 6". P. 223, column 1, lines 7-8: for "Fig. 4" read "Fig. 7". P. 223, column 1, line 13: for "Figs. 5, 8" read "Figs. 8, 11". P. 223, column 1, line 22: for "Fig. 8" read "Fig. 9". P. 223, column 1, line 28: for "Figs. 5, 8": read "Figs. 8, 11". P. 223, column 1, lines 29-30: for "Figs. 10, 11" read "Figs. 2, 3, 8". P. 223, column 1, line 31: for "Figs. 5, 6" read "Figs. 5, 8". P. 223, column 1, line 37: for "Fig. 6" read "Fig. 5". P. 223, column 2, line 3: for "Fig. 7" read "Fig 10". INSTRUCTIONS FOR AUTHORS General Policy The journal of Hymenoptera Research invites papers of high scientific quality reporting comprehensive research on all aspects of Hymenoptera, including biology, behavior, ecology, systematics, taxonomy, genetics, and morphology. Taxonomic papers describing single species are unlikely to be accepted unless a strong case is evident, such as importance in economic entomology or with concurrent biology or ecology. Manuscript length generally should not exceed 50 typed pages; however, no upper limit on length has been set for papers of exceptional quality and importance, including taxonomic monographs at generic or higher level. All papers will be reviewed by at least two referees. The referees will be chosen by the appropriate subject editor. However, it would be helpful if authors would submit the names of two persons who are competent to review the manuscript. The language of publication is English. Summaries in other languages are acceptable. Format and Preparation Three copies of each manuscript, including copies of illustrations, should be submitted on letter size or A4 paper, double spaced, with at least 25 mm margins on all sides. On the upper left of the title page give name, address and telephone and fax numbers of the author to whom all correspondence is to be sent. The paper should have a concise and informative title, followed by the names and addresses of all authors. The sequence of material should be: title, author(s), abstract, text, acknowledgments, literature cited, appendix, figure legends, figure copies (each numbered and identified), tables (each numbered and with heading). Each of the following should start a new page: (1) title page, (2) abstract, (3) text, (4) literature cited, (5) figure legends, (6) footnotes. Following acceptance of the manuscript, the author should provide the editor with two copies of the manuscript accompanied by a copy on diskette using DD, double sided computer diskettes - IBM compatible MS DOS 5.25 inch or IBM and Macintosh 3.5 inch diskettes. (Authors who do not have access to a computer should submit three copies of the manuscript.) The paper may be submitted in most PC and Mac word processor programs such as Microsoft Word, FullWrite Professional, WordPerfect, WriteNow, Nisus, MacWrite, or MacWrite II; it is important that the manuscript also be formatted as an ASCII or TEXT file. If possible, all words that must be italicized should be done so, not underscored. Use of the words male and female is preferred to sex symbols. Tables may be formatted in a spread sheet program such as MS Works or MS Excel. Text should be double spaced typing, with 25 mm left and right margins. Tables should be put in a separate file. Diskettes should be accompanied by the name of the software program used (e.g., WordPerfect, Microsoft Word). Authors should keep backup copies of all material sent to the Editor. The Society cannot be responsible for diskettes or text mislaid or destroyed in transit or during editing. Illustrations should be planned for reduction to the dimension of the printed page (15.3 X 22 cm, column width 7.5 cm) and allow room for legends at the top and bottom. Do not make plates larger than 14" X 18" (35.5 X 46 cm). Individual figures should be mounted on a suitable drawing board or similar heavy stock. Photographs should be trimmed, grouped together and abutted when mounted. Figure numbers should be on the plate, but it is strongly recommended that names be included after the numbers (e.g., Fig. 2, texanus). Include title, author(s) and address(es), and illustration numbers on back of each plate. Original figures need not be sent until requested by the editor, usually after the manuscript has been accepted. Reference to figures /tables in the text should be in the style "(Fig. ')" "(Table 1)". Measurements should be in the metric system. All papers must conform to the International Code of Zoological Nomenclature. The first mention of a plant or animal should include the full scientific name including the authority. Genus names should not be abbreviated at the beginning of a sentence. In taxonomic papers type specimens must be clearly designated, type depositories must be clearly indicated, and new taxa must be clearly differentiated from existing taxa by means of keys or differential diagnoses. Authors are required to deposit all type material in internationally recognized institutions (not private collections). Voucher specimens should be designated for specimens used in behavioral or autecological studies, and they should be deposited similarly. Acceptance of taxonomic papers will not require use of cladistic methods; however, authors using them will be expected to specify the phylogenetic program used (if any), including discussion of program options used. A data matrix should be provided. The number of parsimonious cladograms generated should be stated and the reasons for the one adopted. Lengths and consistency indices should be provided. References in the text should be (Smith 1999), without a comma, or Smith (1999). Two articles by a single author should be (Smith 1999a, 1999b) or Smith (1999a, 1999b). For papers in press, use "in press," not the expected publication date. The Literature Cited section should include all papers referred to in the paper. Journal names should be spelled out completely. Charges Publication is free to members of the International Society of Hymenopterists. At least one author of the paper must be a member. Reprints are charged to the author and must be ordered when returning the proofs; there are no free reprints. Author's corrections and changes in proof are also charged to the author. Color plates will be billed at full cost to the author. All manuscripts and correspondence should be sent to: Paul M. Marsh, Editor P. O. Box 384 North Newton, Kansas 67117 (316) 284-0990 Bound .Tb-Wcase FEB 00 uccTFR INDIAN^"6962!